diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,39731 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 5604,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 116.68827514412096,
+      "learning_rate": 5.91715976331361e-08,
+      "loss": 1.459,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 137.05559220326518,
+      "learning_rate": 1.183431952662722e-07,
+      "loss": 1.5293,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 96.59057528089231,
+      "learning_rate": 1.775147928994083e-07,
+      "loss": 1.4766,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 66.60254143032729,
+      "learning_rate": 2.366863905325444e-07,
+      "loss": 1.2988,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 103.34499410914529,
+      "learning_rate": 2.958579881656805e-07,
+      "loss": 1.3887,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.76380439002071,
+      "learning_rate": 3.550295857988166e-07,
+      "loss": 1.1865,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.37017508979888,
+      "learning_rate": 4.1420118343195276e-07,
+      "loss": 1.3906,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 94.36297438079606,
+      "learning_rate": 4.733727810650888e-07,
+      "loss": 1.3477,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 45.10300787040884,
+      "learning_rate": 5.32544378698225e-07,
+      "loss": 1.2451,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.24939665836075,
+      "learning_rate": 5.91715976331361e-07,
+      "loss": 1.3477,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 102.10130791881714,
+      "learning_rate": 6.50887573964497e-07,
+      "loss": 1.2109,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.836605877501068,
+      "learning_rate": 7.100591715976332e-07,
+      "loss": 1.1523,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.91418906814854,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.2031,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 55.18879990724547,
+      "learning_rate": 8.284023668639055e-07,
+      "loss": 1.1562,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 111.28485143610422,
+      "learning_rate": 8.875739644970415e-07,
+      "loss": 1.2539,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 64.53655153369745,
+      "learning_rate": 9.467455621301776e-07,
+      "loss": 1.1113,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 49.81258432944396,
+      "learning_rate": 1.0059171597633138e-06,
+      "loss": 1.0557,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 81.3821602984751,
+      "learning_rate": 1.06508875739645e-06,
+      "loss": 1.1436,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 77.89797760634401,
+      "learning_rate": 1.1242603550295859e-06,
+      "loss": 1.0498,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.51657421815158,
+      "learning_rate": 1.183431952662722e-06,
+      "loss": 1.1523,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.98384966020216,
+      "learning_rate": 1.242603550295858e-06,
+      "loss": 0.8809,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.909290919081947,
+      "learning_rate": 1.301775147928994e-06,
+      "loss": 0.8525,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.7361005439023,
+      "learning_rate": 1.3609467455621303e-06,
+      "loss": 0.8418,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 26.264367353380752,
+      "learning_rate": 1.4201183431952664e-06,
+      "loss": 0.876,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 35.89694409386303,
+      "learning_rate": 1.4792899408284026e-06,
+      "loss": 0.8018,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 52.88511927832449,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.8203,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.430991723939343,
+      "learning_rate": 1.5976331360946749e-06,
+      "loss": 0.7646,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 40.02223588407742,
+      "learning_rate": 1.656804733727811e-06,
+      "loss": 0.7539,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.523542799757628,
+      "learning_rate": 1.7159763313609468e-06,
+      "loss": 0.8047,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 58.78591457424737,
+      "learning_rate": 1.775147928994083e-06,
+      "loss": 0.8398,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.018173530627244,
+      "learning_rate": 1.834319526627219e-06,
+      "loss": 0.7764,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.224774251543503,
+      "learning_rate": 1.8934911242603552e-06,
+      "loss": 0.7041,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.267158783210224,
+      "learning_rate": 1.952662721893491e-06,
+      "loss": 0.6865,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.71449038710854,
+      "learning_rate": 2.0118343195266275e-06,
+      "loss": 0.7051,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.808645253515795,
+      "learning_rate": 2.0710059171597635e-06,
+      "loss": 0.7783,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 36.10291312881198,
+      "learning_rate": 2.1301775147929e-06,
+      "loss": 0.6826,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.047754674174154,
+      "learning_rate": 2.1893491124260358e-06,
+      "loss": 0.6885,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.948882559824707,
+      "learning_rate": 2.2485207100591717e-06,
+      "loss": 0.6152,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 26.159147372349974,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.7139,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.488442848627141,
+      "learning_rate": 2.366863905325444e-06,
+      "loss": 0.6602,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.239851284814467,
+      "learning_rate": 2.42603550295858e-06,
+      "loss": 0.6924,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.408441779673309,
+      "learning_rate": 2.485207100591716e-06,
+      "loss": 0.6289,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.76752777666436,
+      "learning_rate": 2.5443786982248527e-06,
+      "loss": 0.6436,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.405290702619705,
+      "learning_rate": 2.603550295857988e-06,
+      "loss": 0.6064,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 22.56239327853637,
+      "learning_rate": 2.6627218934911246e-06,
+      "loss": 0.6279,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.96378836265653,
+      "learning_rate": 2.7218934911242605e-06,
+      "loss": 0.6484,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.297707654533205,
+      "learning_rate": 2.7810650887573965e-06,
+      "loss": 0.5977,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 21.368931517016037,
+      "learning_rate": 2.840236686390533e-06,
+      "loss": 0.6211,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.055034780572917,
+      "learning_rate": 2.8994082840236688e-06,
+      "loss": 0.6504,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.170299121299626,
+      "learning_rate": 2.958579881656805e-06,
+      "loss": 0.5391,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.65980083079512,
+      "learning_rate": 3.017751479289941e-06,
+      "loss": 0.6562,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.141571770193977,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.5576,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 38.18938821430388,
+      "learning_rate": 3.1360946745562134e-06,
+      "loss": 0.6465,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 36.485584647459994,
+      "learning_rate": 3.1952662721893497e-06,
+      "loss": 0.6182,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.1172270983848,
+      "learning_rate": 3.2544378698224853e-06,
+      "loss": 0.6304,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.35522270180552,
+      "learning_rate": 3.313609467455622e-06,
+      "loss": 0.5791,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.919799862854324,
+      "learning_rate": 3.3727810650887576e-06,
+      "loss": 0.6064,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 37.86337212965342,
+      "learning_rate": 3.4319526627218935e-06,
+      "loss": 0.6719,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 32.4241795097629,
+      "learning_rate": 3.49112426035503e-06,
+      "loss": 0.6816,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.843367212342677,
+      "learning_rate": 3.550295857988166e-06,
+      "loss": 0.5415,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.033086617607612,
+      "learning_rate": 3.609467455621302e-06,
+      "loss": 0.5898,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.78012764506264,
+      "learning_rate": 3.668639053254438e-06,
+      "loss": 0.5967,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.240783033272136,
+      "learning_rate": 3.7278106508875745e-06,
+      "loss": 0.5645,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.123512981508057,
+      "learning_rate": 3.7869822485207104e-06,
+      "loss": 0.5703,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.660633641874856,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.6182,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.27816226916257,
+      "learning_rate": 3.905325443786982e-06,
+      "loss": 0.623,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.991620078648673,
+      "learning_rate": 3.964497041420119e-06,
+      "loss": 0.5547,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 29.73306262944105,
+      "learning_rate": 4.023668639053255e-06,
+      "loss": 0.624,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.983350773990825,
+      "learning_rate": 4.0828402366863906e-06,
+      "loss": 0.5938,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.203249085331949,
+      "learning_rate": 4.142011834319527e-06,
+      "loss": 0.6152,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.903248382781342,
+      "learning_rate": 4.201183431952663e-06,
+      "loss": 0.5425,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.743327563970325,
+      "learning_rate": 4.2603550295858e-06,
+      "loss": 0.5518,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.530873380507195,
+      "learning_rate": 4.319526627218935e-06,
+      "loss": 0.5859,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.306285659968713,
+      "learning_rate": 4.3786982248520715e-06,
+      "loss": 0.585,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.30837385866365,
+      "learning_rate": 4.437869822485207e-06,
+      "loss": 0.7109,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.972681142302918,
+      "learning_rate": 4.497041420118343e-06,
+      "loss": 0.5781,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.69935251821666,
+      "learning_rate": 4.55621301775148e-06,
+      "loss": 0.5889,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 30.74542232405432,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.5996,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.053844021932429,
+      "learning_rate": 4.674556213017752e-06,
+      "loss": 0.6523,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.660654525701556,
+      "learning_rate": 4.733727810650888e-06,
+      "loss": 0.5903,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 33.06855879264672,
+      "learning_rate": 4.792899408284024e-06,
+      "loss": 0.5947,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.027708084710582,
+      "learning_rate": 4.85207100591716e-06,
+      "loss": 0.5249,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.927455268454475,
+      "learning_rate": 4.911242603550296e-06,
+      "loss": 0.5024,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.12037453816752,
+      "learning_rate": 4.970414201183432e-06,
+      "loss": 0.5884,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.758818975299697,
+      "learning_rate": 5.029585798816569e-06,
+      "loss": 0.5601,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.52708954325852,
+      "learning_rate": 5.088757396449705e-06,
+      "loss": 0.4854,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.86444076854706,
+      "learning_rate": 5.14792899408284e-06,
+      "loss": 0.4858,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.60677062559464,
+      "learning_rate": 5.207100591715976e-06,
+      "loss": 0.5625,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.10508123587462,
+      "learning_rate": 5.266272189349113e-06,
+      "loss": 0.5068,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.276312006219614,
+      "learning_rate": 5.325443786982249e-06,
+      "loss": 0.6006,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.278189599056375,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.5049,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 24.04098331248457,
+      "learning_rate": 5.443786982248521e-06,
+      "loss": 0.4922,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 29.519688923776314,
+      "learning_rate": 5.502958579881657e-06,
+      "loss": 0.5532,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.073075096267969,
+      "learning_rate": 5.562130177514793e-06,
+      "loss": 0.5015,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.04049428861296,
+      "learning_rate": 5.621301775147929e-06,
+      "loss": 0.4731,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.43611431386209,
+      "learning_rate": 5.680473372781066e-06,
+      "loss": 0.5938,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.37496355488797,
+      "learning_rate": 5.739644970414202e-06,
+      "loss": 0.6074,
+      "step": 97
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 23.4534463120769,
+      "learning_rate": 5.7988165680473375e-06,
+      "loss": 0.5552,
+      "step": 98
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.224551348565953,
+      "learning_rate": 5.857988165680474e-06,
+      "loss": 0.5259,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.043175008582129,
+      "learning_rate": 5.91715976331361e-06,
+      "loss": 0.5615,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.886760864897116,
+      "learning_rate": 5.976331360946747e-06,
+      "loss": 0.564,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.62984639534972,
+      "learning_rate": 6.035502958579882e-06,
+      "loss": 0.5703,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.366791130831217,
+      "learning_rate": 6.0946745562130185e-06,
+      "loss": 0.4766,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.11640373494133,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.5801,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.344028697162539,
+      "learning_rate": 6.21301775147929e-06,
+      "loss": 0.6074,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.171423052067137,
+      "learning_rate": 6.272189349112427e-06,
+      "loss": 0.5859,
+      "step": 106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.52701822263934,
+      "learning_rate": 6.331360946745563e-06,
+      "loss": 0.5063,
+      "step": 107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.13390826805979,
+      "learning_rate": 6.3905325443786995e-06,
+      "loss": 0.5391,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.893616746652306,
+      "learning_rate": 6.449704142011834e-06,
+      "loss": 0.6562,
+      "step": 109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.267811806709746,
+      "learning_rate": 6.5088757396449705e-06,
+      "loss": 0.5327,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.573660726227825,
+      "learning_rate": 6.568047337278107e-06,
+      "loss": 0.501,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.424349082496683,
+      "learning_rate": 6.627218934911244e-06,
+      "loss": 0.479,
+      "step": 112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.516504502014815,
+      "learning_rate": 6.686390532544379e-06,
+      "loss": 0.7539,
+      "step": 113
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.53470669412648,
+      "learning_rate": 6.745562130177515e-06,
+      "loss": 0.4805,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.690074282712688,
+      "learning_rate": 6.8047337278106515e-06,
+      "loss": 0.5947,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.491351335831942,
+      "learning_rate": 6.863905325443787e-06,
+      "loss": 0.5752,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.709408023616227,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.5063,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.782856613289166,
+      "learning_rate": 6.98224852071006e-06,
+      "loss": 0.5957,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.968430706562753,
+      "learning_rate": 7.041420118343196e-06,
+      "loss": 0.6045,
+      "step": 119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.252540842800897,
+      "learning_rate": 7.100591715976332e-06,
+      "loss": 0.5908,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.269334605120518,
+      "learning_rate": 7.159763313609468e-06,
+      "loss": 0.5405,
+      "step": 121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.810048379769524,
+      "learning_rate": 7.218934911242604e-06,
+      "loss": 0.5518,
+      "step": 122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.265067489943311,
+      "learning_rate": 7.278106508875741e-06,
+      "loss": 0.5259,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.5951034364997,
+      "learning_rate": 7.337278106508876e-06,
+      "loss": 0.5137,
+      "step": 124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.46844506700403,
+      "learning_rate": 7.396449704142013e-06,
+      "loss": 0.4946,
+      "step": 125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.353931220439996,
+      "learning_rate": 7.455621301775149e-06,
+      "loss": 0.5303,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.291482953488973,
+      "learning_rate": 7.5147928994082845e-06,
+      "loss": 0.5142,
+      "step": 127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.676486048527362,
+      "learning_rate": 7.573964497041421e-06,
+      "loss": 0.519,
+      "step": 128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.100948739888956,
+      "learning_rate": 7.633136094674556e-06,
+      "loss": 0.5303,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.616985829007,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.4478,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.999044255166673,
+      "learning_rate": 7.751479289940829e-06,
+      "loss": 0.6455,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 50.80837956748365,
+      "learning_rate": 7.810650887573965e-06,
+      "loss": 0.6377,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.987804243664623,
+      "learning_rate": 7.869822485207102e-06,
+      "loss": 0.6377,
+      "step": 133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.857441145744092,
+      "learning_rate": 7.928994082840237e-06,
+      "loss": 0.6055,
+      "step": 134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 31.853850755139643,
+      "learning_rate": 7.988165680473373e-06,
+      "loss": 0.5396,
+      "step": 135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 26.922060020459558,
+      "learning_rate": 8.04733727810651e-06,
+      "loss": 0.6792,
+      "step": 136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.884182601970089,
+      "learning_rate": 8.106508875739646e-06,
+      "loss": 0.5957,
+      "step": 137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.414653577595885,
+      "learning_rate": 8.165680473372781e-06,
+      "loss": 0.5635,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.83898607971713,
+      "learning_rate": 8.224852071005918e-06,
+      "loss": 0.6191,
+      "step": 139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.238831882725698,
+      "learning_rate": 8.284023668639054e-06,
+      "loss": 0.5503,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.273476773493702,
+      "learning_rate": 8.343195266272191e-06,
+      "loss": 0.5918,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.25326134871138,
+      "learning_rate": 8.402366863905327e-06,
+      "loss": 0.6064,
+      "step": 142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 26.52884271993335,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.6025,
+      "step": 143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.41160283896373,
+      "learning_rate": 8.5207100591716e-06,
+      "loss": 0.4663,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 35.434570167809206,
+      "learning_rate": 8.579881656804735e-06,
+      "loss": 0.6113,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.227787194473562,
+      "learning_rate": 8.63905325443787e-06,
+      "loss": 0.5303,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.767727771237986,
+      "learning_rate": 8.698224852071006e-06,
+      "loss": 0.6484,
+      "step": 147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.707533062635017,
+      "learning_rate": 8.757396449704143e-06,
+      "loss": 0.5186,
+      "step": 148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.412761457820203,
+      "learning_rate": 8.816568047337279e-06,
+      "loss": 0.5659,
+      "step": 149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.467634594644565,
+      "learning_rate": 8.875739644970414e-06,
+      "loss": 0.5151,
+      "step": 150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.925519538332352,
+      "learning_rate": 8.934911242603551e-06,
+      "loss": 0.5879,
+      "step": 151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.71700779047759,
+      "learning_rate": 8.994082840236687e-06,
+      "loss": 0.5835,
+      "step": 152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.84943528345089,
+      "learning_rate": 9.053254437869822e-06,
+      "loss": 0.5264,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.688901386150174,
+      "learning_rate": 9.11242603550296e-06,
+      "loss": 0.6367,
+      "step": 154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.867220435355714,
+      "learning_rate": 9.171597633136095e-06,
+      "loss": 0.6777,
+      "step": 155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 26.02685451565874,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.5967,
+      "step": 156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.482134608562367,
+      "learning_rate": 9.289940828402368e-06,
+      "loss": 0.5615,
+      "step": 157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.88805449508632,
+      "learning_rate": 9.349112426035503e-06,
+      "loss": 0.5684,
+      "step": 158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.661463282963666,
+      "learning_rate": 9.40828402366864e-06,
+      "loss": 0.6445,
+      "step": 159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.35562044235245,
+      "learning_rate": 9.467455621301776e-06,
+      "loss": 0.5557,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.980779661002774,
+      "learning_rate": 9.526627218934912e-06,
+      "loss": 0.5474,
+      "step": 161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.268903766575969,
+      "learning_rate": 9.585798816568049e-06,
+      "loss": 0.5225,
+      "step": 162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.12465799994179,
+      "learning_rate": 9.644970414201184e-06,
+      "loss": 0.5547,
+      "step": 163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.6737374134887215,
+      "learning_rate": 9.70414201183432e-06,
+      "loss": 0.5859,
+      "step": 164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.06983956812275,
+      "learning_rate": 9.763313609467457e-06,
+      "loss": 0.5649,
+      "step": 165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.848185642792462,
+      "learning_rate": 9.822485207100593e-06,
+      "loss": 0.5591,
+      "step": 166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.283247757964315,
+      "learning_rate": 9.88165680473373e-06,
+      "loss": 0.5698,
+      "step": 167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.36449736428342,
+      "learning_rate": 9.940828402366864e-06,
+      "loss": 0.5366,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.504342411667235,
+      "learning_rate": 1e-05,
+      "loss": 0.5571,
+      "step": 169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 28.526999718293393,
+      "learning_rate": 9.999999164703534e-06,
+      "loss": 0.6133,
+      "step": 170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.676287862832698,
+      "learning_rate": 9.999996658814406e-06,
+      "loss": 0.4307,
+      "step": 171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.27838137396804,
+      "learning_rate": 9.999992482333461e-06,
+      "loss": 0.4771,
+      "step": 172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.098596330705657,
+      "learning_rate": 9.99998663526209e-06,
+      "loss": 0.4785,
+      "step": 173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.364350114790247,
+      "learning_rate": 9.99997911760225e-06,
+      "loss": 0.5928,
+      "step": 174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 32.439748171009725,
+      "learning_rate": 9.99996992935645e-06,
+      "loss": 0.6206,
+      "step": 175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.869970364498535,
+      "learning_rate": 9.99995907052776e-06,
+      "loss": 0.5176,
+      "step": 176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 28.853561378084798,
+      "learning_rate": 9.99994654111981e-06,
+      "loss": 0.5718,
+      "step": 177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 30.030018985404727,
+      "learning_rate": 9.999932341136785e-06,
+      "loss": 0.6338,
+      "step": 178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.898208289891516,
+      "learning_rate": 9.999916470583429e-06,
+      "loss": 0.5049,
+      "step": 179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.836549153691703,
+      "learning_rate": 9.999898929465047e-06,
+      "loss": 0.5605,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 31.456036267942125,
+      "learning_rate": 9.999879717787495e-06,
+      "loss": 0.6118,
+      "step": 181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.965440018657493,
+      "learning_rate": 9.999858835557197e-06,
+      "loss": 0.5356,
+      "step": 182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.241499025434392,
+      "learning_rate": 9.999836282781128e-06,
+      "loss": 0.5215,
+      "step": 183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.343720644680257,
+      "learning_rate": 9.999812059466825e-06,
+      "loss": 0.6118,
+      "step": 184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 28.654780626684733,
+      "learning_rate": 9.999786165622379e-06,
+      "loss": 0.6016,
+      "step": 185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.17157492958299,
+      "learning_rate": 9.999758601256441e-06,
+      "loss": 0.624,
+      "step": 186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.176557440775614,
+      "learning_rate": 9.999729366378224e-06,
+      "loss": 0.5527,
+      "step": 187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.253723481448025,
+      "learning_rate": 9.999698460997493e-06,
+      "loss": 0.5601,
+      "step": 188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.43944022216274,
+      "learning_rate": 9.999665885124577e-06,
+      "loss": 0.5273,
+      "step": 189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.913171126499616,
+      "learning_rate": 9.99963163877036e-06,
+      "loss": 0.6587,
+      "step": 190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 22.224898845103755,
+      "learning_rate": 9.99959572194628e-06,
+      "loss": 0.666,
+      "step": 191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.83510331844381,
+      "learning_rate": 9.999558134664342e-06,
+      "loss": 0.6099,
+      "step": 192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.477618783981828,
+      "learning_rate": 9.999518876937102e-06,
+      "loss": 0.4771,
+      "step": 193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.778400243235488,
+      "learning_rate": 9.999477948777678e-06,
+      "loss": 0.5562,
+      "step": 194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.459065251806433,
+      "learning_rate": 9.999435350199745e-06,
+      "loss": 0.5342,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.538105722452126,
+      "learning_rate": 9.999391081217536e-06,
+      "loss": 0.627,
+      "step": 196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.7688376682692,
+      "learning_rate": 9.999345141845842e-06,
+      "loss": 0.5293,
+      "step": 197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.99862969600929,
+      "learning_rate": 9.99929753210001e-06,
+      "loss": 0.5005,
+      "step": 198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.055362447650138,
+      "learning_rate": 9.999248251995951e-06,
+      "loss": 0.5659,
+      "step": 199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.011136685984399,
+      "learning_rate": 9.999197301550127e-06,
+      "loss": 0.5586,
+      "step": 200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.565468854396421,
+      "learning_rate": 9.999144680779564e-06,
+      "loss": 0.5127,
+      "step": 201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.322706038393164,
+      "learning_rate": 9.999090389701844e-06,
+      "loss": 0.5396,
+      "step": 202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.689469240544318,
+      "learning_rate": 9.999034428335103e-06,
+      "loss": 0.5366,
+      "step": 203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.174468406049195,
+      "learning_rate": 9.998976796698043e-06,
+      "loss": 0.6064,
+      "step": 204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.26408461209251,
+      "learning_rate": 9.998917494809917e-06,
+      "loss": 0.5181,
+      "step": 205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.482605171650068,
+      "learning_rate": 9.998856522690538e-06,
+      "loss": 0.6626,
+      "step": 206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.98758204165801,
+      "learning_rate": 9.998793880360283e-06,
+      "loss": 0.48,
+      "step": 207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 22.66544425395466,
+      "learning_rate": 9.998729567840077e-06,
+      "loss": 0.6836,
+      "step": 208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.309350432593693,
+      "learning_rate": 9.998663585151409e-06,
+      "loss": 0.5674,
+      "step": 209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 27.286620280815722,
+      "learning_rate": 9.998595932316327e-06,
+      "loss": 0.6514,
+      "step": 210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 14.739890462945858,
+      "learning_rate": 9.998526609357432e-06,
+      "loss": 0.5947,
+      "step": 211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.551376294742977,
+      "learning_rate": 9.998455616297889e-06,
+      "loss": 0.5879,
+      "step": 212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.631161156381427,
+      "learning_rate": 9.998382953161417e-06,
+      "loss": 0.6865,
+      "step": 213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 18.33142486833869,
+      "learning_rate": 9.998308619972292e-06,
+      "loss": 0.6357,
+      "step": 214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.60392425953813,
+      "learning_rate": 9.998232616755354e-06,
+      "loss": 0.5732,
+      "step": 215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.976381604704594,
+      "learning_rate": 9.998154943535996e-06,
+      "loss": 0.5645,
+      "step": 216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.10651009987485,
+      "learning_rate": 9.998075600340166e-06,
+      "loss": 0.582,
+      "step": 217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.196857537217384,
+      "learning_rate": 9.997994587194381e-06,
+      "loss": 0.564,
+      "step": 218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.280398372848552,
+      "learning_rate": 9.997911904125704e-06,
+      "loss": 0.6353,
+      "step": 219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.816198078589931,
+      "learning_rate": 9.997827551161762e-06,
+      "loss": 0.5684,
+      "step": 220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.183296652188321,
+      "learning_rate": 9.997741528330739e-06,
+      "loss": 0.5449,
+      "step": 221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.185998935163878,
+      "learning_rate": 9.997653835661376e-06,
+      "loss": 0.5967,
+      "step": 222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.288601461686234,
+      "learning_rate": 9.997564473182976e-06,
+      "loss": 0.5698,
+      "step": 223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.321856011834956,
+      "learning_rate": 9.997473440925394e-06,
+      "loss": 0.5771,
+      "step": 224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.311288966424027,
+      "learning_rate": 9.997380738919045e-06,
+      "loss": 0.5259,
+      "step": 225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.481033073870574,
+      "learning_rate": 9.997286367194903e-06,
+      "loss": 0.6689,
+      "step": 226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.550467772007261,
+      "learning_rate": 9.9971903257845e-06,
+      "loss": 0.5928,
+      "step": 227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.702124607152298,
+      "learning_rate": 9.997092614719926e-06,
+      "loss": 0.5181,
+      "step": 228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 17.241256236619744,
+      "learning_rate": 9.996993234033826e-06,
+      "loss": 0.5918,
+      "step": 229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.044871821417644,
+      "learning_rate": 9.996892183759407e-06,
+      "loss": 0.5811,
+      "step": 230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.21185115842485,
+      "learning_rate": 9.99678946393043e-06,
+      "loss": 0.5791,
+      "step": 231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.301100574961,
+      "learning_rate": 9.996685074581216e-06,
+      "loss": 0.5332,
+      "step": 232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.335320291463777,
+      "learning_rate": 9.996579015746645e-06,
+      "loss": 0.5742,
+      "step": 233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.8646367315459935,
+      "learning_rate": 9.996471287462151e-06,
+      "loss": 0.5376,
+      "step": 234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.966657193641781,
+      "learning_rate": 9.99636188976373e-06,
+      "loss": 0.5762,
+      "step": 235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.07991289086175,
+      "learning_rate": 9.996250822687932e-06,
+      "loss": 0.5405,
+      "step": 236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.807407534670729,
+      "learning_rate": 9.996138086271869e-06,
+      "loss": 0.585,
+      "step": 237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.991884688831597,
+      "learning_rate": 9.996023680553204e-06,
+      "loss": 0.5181,
+      "step": 238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.4817825314537645,
+      "learning_rate": 9.995907605570167e-06,
+      "loss": 0.5957,
+      "step": 239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.175799986215807,
+      "learning_rate": 9.995789861361538e-06,
+      "loss": 0.6895,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.737883477841022,
+      "learning_rate": 9.995670447966658e-06,
+      "loss": 0.4727,
+      "step": 241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.747945457840865,
+      "learning_rate": 9.995549365425426e-06,
+      "loss": 0.5635,
+      "step": 242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.188385247219632,
+      "learning_rate": 9.995426613778297e-06,
+      "loss": 0.6445,
+      "step": 243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.735363655951899,
+      "learning_rate": 9.995302193066286e-06,
+      "loss": 0.5112,
+      "step": 244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.187157770296569,
+      "learning_rate": 9.995176103330962e-06,
+      "loss": 0.5776,
+      "step": 245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.366215466061506,
+      "learning_rate": 9.995048344614455e-06,
+      "loss": 0.6016,
+      "step": 246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.101901236927242,
+      "learning_rate": 9.994918916959453e-06,
+      "loss": 0.5952,
+      "step": 247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.494534788687338,
+      "learning_rate": 9.994787820409198e-06,
+      "loss": 0.5625,
+      "step": 248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.350350461703404,
+      "learning_rate": 9.994655055007491e-06,
+      "loss": 0.5278,
+      "step": 249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.815623310077835,
+      "learning_rate": 9.994520620798696e-06,
+      "loss": 0.5273,
+      "step": 250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.04445898352151,
+      "learning_rate": 9.994384517827726e-06,
+      "loss": 0.6157,
+      "step": 251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.217464632831893,
+      "learning_rate": 9.994246746140057e-06,
+      "loss": 0.5576,
+      "step": 252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.624589818254435,
+      "learning_rate": 9.99410730578172e-06,
+      "loss": 0.5503,
+      "step": 253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.459911000864924,
+      "learning_rate": 9.993966196799304e-06,
+      "loss": 0.5166,
+      "step": 254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.972917737075516,
+      "learning_rate": 9.993823419239959e-06,
+      "loss": 0.6016,
+      "step": 255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.26491066381988,
+      "learning_rate": 9.993678973151388e-06,
+      "loss": 0.4448,
+      "step": 256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.959824992641648,
+      "learning_rate": 9.993532858581853e-06,
+      "loss": 0.6025,
+      "step": 257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.17714303025146,
+      "learning_rate": 9.993385075580173e-06,
+      "loss": 0.605,
+      "step": 258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.583037935905157,
+      "learning_rate": 9.993235624195728e-06,
+      "loss": 0.5659,
+      "step": 259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.251932963533232,
+      "learning_rate": 9.993084504478448e-06,
+      "loss": 0.5811,
+      "step": 260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.8640641134105,
+      "learning_rate": 9.99293171647883e-06,
+      "loss": 0.4863,
+      "step": 261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.539968969637762,
+      "learning_rate": 9.992777260247916e-06,
+      "loss": 0.5469,
+      "step": 262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.333336259322376,
+      "learning_rate": 9.99262113583732e-06,
+      "loss": 0.5479,
+      "step": 263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 31.227206000118187,
+      "learning_rate": 9.992463343299203e-06,
+      "loss": 0.6367,
+      "step": 264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.373876295931044,
+      "learning_rate": 9.992303882686288e-06,
+      "loss": 0.5479,
+      "step": 265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.169578673800072,
+      "learning_rate": 9.99214275405185e-06,
+      "loss": 0.5327,
+      "step": 266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.948032240297028,
+      "learning_rate": 9.991979957449729e-06,
+      "loss": 0.6313,
+      "step": 267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 28.97323801726882,
+      "learning_rate": 9.991815492934318e-06,
+      "loss": 0.6411,
+      "step": 268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.141849461327332,
+      "learning_rate": 9.991649360560565e-06,
+      "loss": 0.6318,
+      "step": 269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.920257209617739,
+      "learning_rate": 9.99148156038398e-06,
+      "loss": 0.5425,
+      "step": 270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.115581788358424,
+      "learning_rate": 9.991312092460626e-06,
+      "loss": 0.4868,
+      "step": 271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.252432772054797,
+      "learning_rate": 9.991140956847128e-06,
+      "loss": 0.6235,
+      "step": 272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 22.131525985269647,
+      "learning_rate": 9.990968153600664e-06,
+      "loss": 0.6006,
+      "step": 273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.41445550986435,
+      "learning_rate": 9.990793682778973e-06,
+      "loss": 0.52,
+      "step": 274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.317553621436222,
+      "learning_rate": 9.990617544440346e-06,
+      "loss": 0.5083,
+      "step": 275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.54955596823522,
+      "learning_rate": 9.990439738643635e-06,
+      "loss": 0.5161,
+      "step": 276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.964194414552132,
+      "learning_rate": 9.99026026544825e-06,
+      "loss": 0.5083,
+      "step": 277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.720177088068684,
+      "learning_rate": 9.990079124914156e-06,
+      "loss": 0.522,
+      "step": 278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.459906145881508,
+      "learning_rate": 9.989896317101873e-06,
+      "loss": 0.4951,
+      "step": 279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.765981248006945,
+      "learning_rate": 9.989711842072482e-06,
+      "loss": 0.6133,
+      "step": 280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.835633957228747,
+      "learning_rate": 9.989525699887619e-06,
+      "loss": 0.5205,
+      "step": 281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.649027140963295,
+      "learning_rate": 9.989337890609478e-06,
+      "loss": 0.5625,
+      "step": 282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.40438475972807,
+      "learning_rate": 9.98914841430081e-06,
+      "loss": 0.4858,
+      "step": 283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.573834463528204,
+      "learning_rate": 9.988957271024922e-06,
+      "loss": 0.542,
+      "step": 284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.091426545143989,
+      "learning_rate": 9.988764460845676e-06,
+      "loss": 0.5542,
+      "step": 285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.36487501871403,
+      "learning_rate": 9.9885699838275e-06,
+      "loss": 0.4185,
+      "step": 286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.691682917775056,
+      "learning_rate": 9.988373840035366e-06,
+      "loss": 0.541,
+      "step": 287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.560245915679797,
+      "learning_rate": 9.988176029534814e-06,
+      "loss": 0.543,
+      "step": 288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.393486733090167,
+      "learning_rate": 9.987976552391933e-06,
+      "loss": 0.4878,
+      "step": 289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.696902281023876,
+      "learning_rate": 9.987775408673373e-06,
+      "loss": 0.603,
+      "step": 290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.542550692108352,
+      "learning_rate": 9.987572598446337e-06,
+      "loss": 0.5083,
+      "step": 291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.93517301011984,
+      "learning_rate": 9.987368121778594e-06,
+      "loss": 0.4785,
+      "step": 292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.117485021835481,
+      "learning_rate": 9.98716197873846e-06,
+      "loss": 0.4951,
+      "step": 293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 29.587346547838976,
+      "learning_rate": 9.98695416939481e-06,
+      "loss": 0.7002,
+      "step": 294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.306933072284718,
+      "learning_rate": 9.986744693817077e-06,
+      "loss": 0.561,
+      "step": 295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.606210464051424,
+      "learning_rate": 9.986533552075252e-06,
+      "loss": 0.5801,
+      "step": 296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.005787274091094,
+      "learning_rate": 9.986320744239883e-06,
+      "loss": 0.5742,
+      "step": 297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.512421082684124,
+      "learning_rate": 9.98610627038207e-06,
+      "loss": 0.5532,
+      "step": 298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.006107344505082,
+      "learning_rate": 9.985890130573474e-06,
+      "loss": 0.5298,
+      "step": 299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.111526173571763,
+      "learning_rate": 9.98567232488631e-06,
+      "loss": 0.5459,
+      "step": 300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.911631119108614,
+      "learning_rate": 9.985452853393353e-06,
+      "loss": 0.5425,
+      "step": 301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.375016686561212,
+      "learning_rate": 9.985231716167933e-06,
+      "loss": 0.5298,
+      "step": 302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.179532269489528,
+      "learning_rate": 9.985008913283933e-06,
+      "loss": 0.5459,
+      "step": 303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.104521872821598,
+      "learning_rate": 9.984784444815799e-06,
+      "loss": 0.6201,
+      "step": 304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.983836084666084,
+      "learning_rate": 9.984558310838528e-06,
+      "loss": 0.5645,
+      "step": 305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.16177018185428,
+      "learning_rate": 9.984330511427676e-06,
+      "loss": 0.5693,
+      "step": 306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.784575497017391,
+      "learning_rate": 9.984101046659353e-06,
+      "loss": 0.4595,
+      "step": 307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.33312354424948,
+      "learning_rate": 9.983869916610232e-06,
+      "loss": 0.4668,
+      "step": 308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.786060507815215,
+      "learning_rate": 9.983637121357534e-06,
+      "loss": 0.5557,
+      "step": 309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.611215764264122,
+      "learning_rate": 9.983402660979042e-06,
+      "loss": 0.6064,
+      "step": 310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.263169097847555,
+      "learning_rate": 9.983166535553093e-06,
+      "loss": 0.5977,
+      "step": 311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.022215974074236,
+      "learning_rate": 9.98292874515858e-06,
+      "loss": 0.5137,
+      "step": 312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.936834299449774,
+      "learning_rate": 9.982689289874956e-06,
+      "loss": 0.5898,
+      "step": 313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.511642609064424,
+      "learning_rate": 9.982448169782226e-06,
+      "loss": 0.5967,
+      "step": 314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.598567525889859,
+      "learning_rate": 9.98220538496095e-06,
+      "loss": 0.5234,
+      "step": 315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.979819387251359,
+      "learning_rate": 9.98196093549225e-06,
+      "loss": 0.5054,
+      "step": 316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.574667202148518,
+      "learning_rate": 9.9817148214578e-06,
+      "loss": 0.5801,
+      "step": 317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.56598373183688,
+      "learning_rate": 9.981467042939833e-06,
+      "loss": 0.5732,
+      "step": 318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.776851149154552,
+      "learning_rate": 9.981217600021133e-06,
+      "loss": 0.5469,
+      "step": 319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.737203552004415,
+      "learning_rate": 9.980966492785048e-06,
+      "loss": 0.5742,
+      "step": 320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.915557318934983,
+      "learning_rate": 9.980713721315473e-06,
+      "loss": 0.4888,
+      "step": 321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.65537704300492,
+      "learning_rate": 9.98045928569687e-06,
+      "loss": 0.5425,
+      "step": 322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.172596910986449,
+      "learning_rate": 9.98020318601424e-06,
+      "loss": 0.4824,
+      "step": 323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.440835376005527,
+      "learning_rate": 9.97994542235316e-06,
+      "loss": 0.5522,
+      "step": 324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 19.118269130078797,
+      "learning_rate": 9.979685994799753e-06,
+      "loss": 0.6069,
+      "step": 325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.200575202995402,
+      "learning_rate": 9.979424903440695e-06,
+      "loss": 0.5405,
+      "step": 326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.163978783652977,
+      "learning_rate": 9.979162148363222e-06,
+      "loss": 0.5332,
+      "step": 327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.535886529238546,
+      "learning_rate": 9.978897729655127e-06,
+      "loss": 0.5405,
+      "step": 328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 38.988384289584985,
+      "learning_rate": 9.978631647404755e-06,
+      "loss": 0.6826,
+      "step": 329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.249997402471777,
+      "learning_rate": 9.97836390170101e-06,
+      "loss": 0.5259,
+      "step": 330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.01736287591341,
+      "learning_rate": 9.978094492633353e-06,
+      "loss": 0.5601,
+      "step": 331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.939324146161377,
+      "learning_rate": 9.977823420291796e-06,
+      "loss": 0.5688,
+      "step": 332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.361987880781351,
+      "learning_rate": 9.97755068476691e-06,
+      "loss": 0.5605,
+      "step": 333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 22.600799019220247,
+      "learning_rate": 9.977276286149821e-06,
+      "loss": 0.6226,
+      "step": 334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.773235886574865,
+      "learning_rate": 9.977000224532211e-06,
+      "loss": 0.5332,
+      "step": 335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.06416768362281,
+      "learning_rate": 9.976722500006318e-06,
+      "loss": 0.6416,
+      "step": 336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.827713447093007,
+      "learning_rate": 9.976443112664932e-06,
+      "loss": 0.5957,
+      "step": 337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.629610673442508,
+      "learning_rate": 9.976162062601407e-06,
+      "loss": 0.5527,
+      "step": 338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.528967178159732,
+      "learning_rate": 9.97587934990964e-06,
+      "loss": 0.5762,
+      "step": 339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.608082935381697,
+      "learning_rate": 9.975594974684096e-06,
+      "loss": 0.5659,
+      "step": 340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.717599293321035,
+      "learning_rate": 9.975308937019787e-06,
+      "loss": 0.5278,
+      "step": 341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.812143947109021,
+      "learning_rate": 9.975021237012286e-06,
+      "loss": 0.5552,
+      "step": 342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.925371312954011,
+      "learning_rate": 9.974731874757717e-06,
+      "loss": 0.5596,
+      "step": 343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.339309116508288,
+      "learning_rate": 9.974440850352762e-06,
+      "loss": 0.582,
+      "step": 344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.614761768586927,
+      "learning_rate": 9.974148163894658e-06,
+      "loss": 0.5303,
+      "step": 345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.71369707758011,
+      "learning_rate": 9.973853815481196e-06,
+      "loss": 0.5601,
+      "step": 346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.093172429150728,
+      "learning_rate": 9.973557805210724e-06,
+      "loss": 0.5283,
+      "step": 347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.19951189726805,
+      "learning_rate": 9.973260133182145e-06,
+      "loss": 0.5615,
+      "step": 348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.218505188713452,
+      "learning_rate": 9.972960799494915e-06,
+      "loss": 0.5591,
+      "step": 349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.343689711468452,
+      "learning_rate": 9.972659804249047e-06,
+      "loss": 0.4824,
+      "step": 350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.095683927102183,
+      "learning_rate": 9.972357147545113e-06,
+      "loss": 0.5591,
+      "step": 351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.28619957306297,
+      "learning_rate": 9.972052829484231e-06,
+      "loss": 0.5586,
+      "step": 352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.927219563114457,
+      "learning_rate": 9.971746850168084e-06,
+      "loss": 0.543,
+      "step": 353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.070239504065867,
+      "learning_rate": 9.971439209698902e-06,
+      "loss": 0.6523,
+      "step": 354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.674351013034489,
+      "learning_rate": 9.971129908179474e-06,
+      "loss": 0.541,
+      "step": 355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.105592852551204,
+      "learning_rate": 9.970818945713145e-06,
+      "loss": 0.5659,
+      "step": 356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.0981327588150025,
+      "learning_rate": 9.970506322403813e-06,
+      "loss": 0.4458,
+      "step": 357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 31.65511002299402,
+      "learning_rate": 9.970192038355928e-06,
+      "loss": 0.6401,
+      "step": 358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.640887314827702,
+      "learning_rate": 9.969876093674502e-06,
+      "loss": 0.5005,
+      "step": 359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.615326153238845,
+      "learning_rate": 9.969558488465097e-06,
+      "loss": 0.4995,
+      "step": 360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.0121028825566,
+      "learning_rate": 9.969239222833829e-06,
+      "loss": 0.5254,
+      "step": 361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.942638756744063,
+      "learning_rate": 9.968918296887374e-06,
+      "loss": 0.48,
+      "step": 362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.8607942912242965,
+      "learning_rate": 9.968595710732955e-06,
+      "loss": 0.5239,
+      "step": 363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.584219880853102,
+      "learning_rate": 9.968271464478357e-06,
+      "loss": 0.6064,
+      "step": 364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.74816464727213,
+      "learning_rate": 9.967945558231917e-06,
+      "loss": 0.624,
+      "step": 365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.18424589657994,
+      "learning_rate": 9.967617992102526e-06,
+      "loss": 0.5688,
+      "step": 366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.963994900078255,
+      "learning_rate": 9.967288766199628e-06,
+      "loss": 0.5151,
+      "step": 367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.107368415735328,
+      "learning_rate": 9.966957880633225e-06,
+      "loss": 0.5117,
+      "step": 368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.726462628184319,
+      "learning_rate": 9.966625335513873e-06,
+      "loss": 0.5464,
+      "step": 369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.415412889353508,
+      "learning_rate": 9.96629113095268e-06,
+      "loss": 0.6455,
+      "step": 370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.29743204783179,
+      "learning_rate": 9.965955267061309e-06,
+      "loss": 0.564,
+      "step": 371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.177772883312987,
+      "learning_rate": 9.965617743951982e-06,
+      "loss": 0.4883,
+      "step": 372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.407931454753083,
+      "learning_rate": 9.965278561737466e-06,
+      "loss": 0.4746,
+      "step": 373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.137420154130437,
+      "learning_rate": 9.964937720531094e-06,
+      "loss": 0.5532,
+      "step": 374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.056581957504337,
+      "learning_rate": 9.964595220446744e-06,
+      "loss": 0.5771,
+      "step": 375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.993928512302421,
+      "learning_rate": 9.964251061598853e-06,
+      "loss": 0.583,
+      "step": 376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.30189244870148,
+      "learning_rate": 9.96390524410241e-06,
+      "loss": 0.5552,
+      "step": 377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.305556797808645,
+      "learning_rate": 9.96355776807296e-06,
+      "loss": 0.5098,
+      "step": 378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.335808377185364,
+      "learning_rate": 9.9632086336266e-06,
+      "loss": 0.5747,
+      "step": 379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.797481794532704,
+      "learning_rate": 9.962857840879983e-06,
+      "loss": 0.5664,
+      "step": 380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.172910648343924,
+      "learning_rate": 9.962505389950317e-06,
+      "loss": 0.6455,
+      "step": 381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.971801104641452,
+      "learning_rate": 9.962151280955359e-06,
+      "loss": 0.5317,
+      "step": 382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.019462783189976,
+      "learning_rate": 9.961795514013424e-06,
+      "loss": 0.6611,
+      "step": 383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.015095631990192,
+      "learning_rate": 9.961438089243384e-06,
+      "loss": 0.54,
+      "step": 384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.693443657141236,
+      "learning_rate": 9.961079006764659e-06,
+      "loss": 0.6846,
+      "step": 385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.388009134623115,
+      "learning_rate": 9.960718266697223e-06,
+      "loss": 0.4805,
+      "step": 386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.305245922345355,
+      "learning_rate": 9.960355869161609e-06,
+      "loss": 0.5625,
+      "step": 387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.98682854904853,
+      "learning_rate": 9.959991814278898e-06,
+      "loss": 0.5361,
+      "step": 388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 30.855897554320187,
+      "learning_rate": 9.95962610217073e-06,
+      "loss": 0.6348,
+      "step": 389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.186818656200053,
+      "learning_rate": 9.959258732959296e-06,
+      "loss": 0.6367,
+      "step": 390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.1692607171307,
+      "learning_rate": 9.958889706767341e-06,
+      "loss": 0.6035,
+      "step": 391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.89752794089645,
+      "learning_rate": 9.95851902371816e-06,
+      "loss": 0.6279,
+      "step": 392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.94232313274701,
+      "learning_rate": 9.95814668393561e-06,
+      "loss": 0.625,
+      "step": 393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.263112378321651,
+      "learning_rate": 9.957772687544094e-06,
+      "loss": 0.6211,
+      "step": 394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 18.694064854132726,
+      "learning_rate": 9.95739703466857e-06,
+      "loss": 0.6113,
+      "step": 395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.054888000787919,
+      "learning_rate": 9.957019725434554e-06,
+      "loss": 0.6055,
+      "step": 396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.852195125531498,
+      "learning_rate": 9.956640759968111e-06,
+      "loss": 0.4897,
+      "step": 397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.66681348604124,
+      "learning_rate": 9.956260138395857e-06,
+      "loss": 0.5479,
+      "step": 398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 20.557192888856104,
+      "learning_rate": 9.955877860844969e-06,
+      "loss": 0.6069,
+      "step": 399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.942620031874748,
+      "learning_rate": 9.955493927443171e-06,
+      "loss": 0.4883,
+      "step": 400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.005501795532552,
+      "learning_rate": 9.955108338318743e-06,
+      "loss": 0.6094,
+      "step": 401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.342699104317559,
+      "learning_rate": 9.954721093600517e-06,
+      "loss": 0.541,
+      "step": 402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.090322647328717,
+      "learning_rate": 9.95433219341788e-06,
+      "loss": 0.5225,
+      "step": 403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.150636904880326,
+      "learning_rate": 9.953941637900769e-06,
+      "loss": 0.666,
+      "step": 404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.523930157702077,
+      "learning_rate": 9.953549427179676e-06,
+      "loss": 0.5566,
+      "step": 405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.643994089560804,
+      "learning_rate": 9.953155561385646e-06,
+      "loss": 0.5015,
+      "step": 406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.03294877321269,
+      "learning_rate": 9.952760040650278e-06,
+      "loss": 0.562,
+      "step": 407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.801467808368454,
+      "learning_rate": 9.95236286510572e-06,
+      "loss": 0.519,
+      "step": 408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.80918459345022,
+      "learning_rate": 9.95196403488468e-06,
+      "loss": 0.5747,
+      "step": 409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.774285807444773,
+      "learning_rate": 9.951563550120412e-06,
+      "loss": 0.5752,
+      "step": 410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.463568068842072,
+      "learning_rate": 9.951161410946725e-06,
+      "loss": 0.5527,
+      "step": 411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.73368659445093,
+      "learning_rate": 9.950757617497983e-06,
+      "loss": 0.4585,
+      "step": 412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.942045154978281,
+      "learning_rate": 9.950352169909101e-06,
+      "loss": 0.4893,
+      "step": 413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.393249457904501,
+      "learning_rate": 9.949945068315544e-06,
+      "loss": 0.5684,
+      "step": 414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.911557122462217,
+      "learning_rate": 9.949536312853334e-06,
+      "loss": 0.5786,
+      "step": 415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.653566902215857,
+      "learning_rate": 9.949125903659042e-06,
+      "loss": 0.6289,
+      "step": 416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.471009617251026,
+      "learning_rate": 9.948713840869797e-06,
+      "loss": 0.5283,
+      "step": 417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.083527199424468,
+      "learning_rate": 9.948300124623274e-06,
+      "loss": 0.4492,
+      "step": 418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.197608537282099,
+      "learning_rate": 9.947884755057703e-06,
+      "loss": 0.5186,
+      "step": 419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.630148938077044,
+      "learning_rate": 9.947467732311868e-06,
+      "loss": 0.5669,
+      "step": 420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.24383405339626,
+      "learning_rate": 9.947049056525104e-06,
+      "loss": 0.5068,
+      "step": 421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.288937926215693,
+      "learning_rate": 9.9466287278373e-06,
+      "loss": 0.5103,
+      "step": 422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.534922729005983,
+      "learning_rate": 9.946206746388892e-06,
+      "loss": 0.5278,
+      "step": 423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.99533118513947,
+      "learning_rate": 9.94578311232087e-06,
+      "loss": 0.585,
+      "step": 424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.150688092265762,
+      "learning_rate": 9.945357825774786e-06,
+      "loss": 0.5933,
+      "step": 425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.850098763453946,
+      "learning_rate": 9.944930886892731e-06,
+      "loss": 0.5488,
+      "step": 426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.389796398632541,
+      "learning_rate": 9.944502295817353e-06,
+      "loss": 0.5278,
+      "step": 427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.254074622509817,
+      "learning_rate": 9.944072052691853e-06,
+      "loss": 0.5723,
+      "step": 428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.682353391544783,
+      "learning_rate": 9.943640157659984e-06,
+      "loss": 0.4854,
+      "step": 429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.532300495721698,
+      "learning_rate": 9.94320661086605e-06,
+      "loss": 0.4614,
+      "step": 430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 23.899338345320164,
+      "learning_rate": 9.942771412454906e-06,
+      "loss": 0.749,
+      "step": 431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.323403099794728,
+      "learning_rate": 9.942334562571961e-06,
+      "loss": 0.5317,
+      "step": 432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.057160250544284,
+      "learning_rate": 9.941896061363173e-06,
+      "loss": 0.5894,
+      "step": 433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 19.87921251443284,
+      "learning_rate": 9.941455908975054e-06,
+      "loss": 0.5293,
+      "step": 434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.374070788814263,
+      "learning_rate": 9.941014105554668e-06,
+      "loss": 0.6357,
+      "step": 435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.07075036921681,
+      "learning_rate": 9.94057065124963e-06,
+      "loss": 0.5859,
+      "step": 436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.935502918801745,
+      "learning_rate": 9.940125546208107e-06,
+      "loss": 0.4937,
+      "step": 437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.849912769259404,
+      "learning_rate": 9.939678790578813e-06,
+      "loss": 0.5679,
+      "step": 438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.909255099305636,
+      "learning_rate": 9.93923038451102e-06,
+      "loss": 0.5547,
+      "step": 439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.00963416001379,
+      "learning_rate": 9.938780328154549e-06,
+      "loss": 0.6104,
+      "step": 440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.4109051578140575,
+      "learning_rate": 9.938328621659775e-06,
+      "loss": 0.5923,
+      "step": 441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.359334459897687,
+      "learning_rate": 9.937875265177615e-06,
+      "loss": 0.5879,
+      "step": 442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.83227778183207,
+      "learning_rate": 9.937420258859547e-06,
+      "loss": 0.563,
+      "step": 443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.415148747386903,
+      "learning_rate": 9.9369636028576e-06,
+      "loss": 0.5352,
+      "step": 444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.775852096977613,
+      "learning_rate": 9.936505297324346e-06,
+      "loss": 0.5283,
+      "step": 445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.923034553787746,
+      "learning_rate": 9.936045342412917e-06,
+      "loss": 0.5732,
+      "step": 446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.630684830040368,
+      "learning_rate": 9.93558373827699e-06,
+      "loss": 0.5078,
+      "step": 447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.677385634372398,
+      "learning_rate": 9.935120485070799e-06,
+      "loss": 0.5557,
+      "step": 448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.267582350250366,
+      "learning_rate": 9.934655582949123e-06,
+      "loss": 0.4868,
+      "step": 449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.591276642156359,
+      "learning_rate": 9.934189032067296e-06,
+      "loss": 0.5791,
+      "step": 450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.200236779301449,
+      "learning_rate": 9.933720832581197e-06,
+      "loss": 0.5679,
+      "step": 451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.316615652129823,
+      "learning_rate": 9.933250984647266e-06,
+      "loss": 0.5435,
+      "step": 452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.91748370377269,
+      "learning_rate": 9.932779488422484e-06,
+      "loss": 0.5562,
+      "step": 453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.369736880224455,
+      "learning_rate": 9.93230634406439e-06,
+      "loss": 0.5498,
+      "step": 454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.79281460550446,
+      "learning_rate": 9.931831551731067e-06,
+      "loss": 0.585,
+      "step": 455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.320334157458905,
+      "learning_rate": 9.931355111581154e-06,
+      "loss": 0.6392,
+      "step": 456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.65106240403786,
+      "learning_rate": 9.930877023773837e-06,
+      "loss": 0.5015,
+      "step": 457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.87208988766515,
+      "learning_rate": 9.930397288468853e-06,
+      "loss": 0.5522,
+      "step": 458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.81270031499954,
+      "learning_rate": 9.929915905826494e-06,
+      "loss": 0.4966,
+      "step": 459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.623328763405382,
+      "learning_rate": 9.9294328760076e-06,
+      "loss": 0.5083,
+      "step": 460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.176704828042569,
+      "learning_rate": 9.928948199173552e-06,
+      "loss": 0.5142,
+      "step": 461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.452044947551833,
+      "learning_rate": 9.928461875486297e-06,
+      "loss": 0.4746,
+      "step": 462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.515116021434167,
+      "learning_rate": 9.927973905108323e-06,
+      "loss": 0.5566,
+      "step": 463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.279657241609938,
+      "learning_rate": 9.927484288202671e-06,
+      "loss": 0.5566,
+      "step": 464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 23.899242007190043,
+      "learning_rate": 9.926993024932929e-06,
+      "loss": 0.5767,
+      "step": 465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.739763048191918,
+      "learning_rate": 9.926500115463238e-06,
+      "loss": 0.5396,
+      "step": 466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.322105279702368,
+      "learning_rate": 9.926005559958287e-06,
+      "loss": 0.5317,
+      "step": 467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.685933746852832,
+      "learning_rate": 9.925509358583319e-06,
+      "loss": 0.5044,
+      "step": 468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.819644131934337,
+      "learning_rate": 9.92501151150412e-06,
+      "loss": 0.5361,
+      "step": 469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.096851486127841,
+      "learning_rate": 9.924512018887036e-06,
+      "loss": 0.54,
+      "step": 470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.4073975042839,
+      "learning_rate": 9.924010880898952e-06,
+      "loss": 0.5059,
+      "step": 471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.34808552011236,
+      "learning_rate": 9.923508097707306e-06,
+      "loss": 0.6025,
+      "step": 472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.169077335799752,
+      "learning_rate": 9.923003669480094e-06,
+      "loss": 0.562,
+      "step": 473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.814349393155531,
+      "learning_rate": 9.922497596385848e-06,
+      "loss": 0.5376,
+      "step": 474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.91476877486239,
+      "learning_rate": 9.92198987859366e-06,
+      "loss": 0.458,
+      "step": 475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.796586778125109,
+      "learning_rate": 9.921480516273168e-06,
+      "loss": 0.5645,
+      "step": 476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.08880736913912,
+      "learning_rate": 9.920969509594558e-06,
+      "loss": 0.5273,
+      "step": 477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.015029790773587,
+      "learning_rate": 9.920456858728567e-06,
+      "loss": 0.4678,
+      "step": 478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.165530523776276,
+      "learning_rate": 9.919942563846482e-06,
+      "loss": 0.4663,
+      "step": 479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.515499096006561,
+      "learning_rate": 9.919426625120137e-06,
+      "loss": 0.5649,
+      "step": 480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.393807256918857,
+      "learning_rate": 9.918909042721918e-06,
+      "loss": 0.5576,
+      "step": 481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.408074858410433,
+      "learning_rate": 9.918389816824759e-06,
+      "loss": 0.6514,
+      "step": 482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.273121385573273,
+      "learning_rate": 9.917868947602144e-06,
+      "loss": 0.6157,
+      "step": 483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.733233581807566,
+      "learning_rate": 9.917346435228102e-06,
+      "loss": 0.6221,
+      "step": 484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.1906994483687,
+      "learning_rate": 9.916822279877217e-06,
+      "loss": 0.4849,
+      "step": 485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.21592645313153,
+      "learning_rate": 9.91629648172462e-06,
+      "loss": 0.5352,
+      "step": 486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.57311751283234,
+      "learning_rate": 9.915769040945984e-06,
+      "loss": 0.54,
+      "step": 487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.283261788372716,
+      "learning_rate": 9.915239957717542e-06,
+      "loss": 0.5034,
+      "step": 488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.597381341995393,
+      "learning_rate": 9.91470923221607e-06,
+      "loss": 0.5898,
+      "step": 489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.86577970600292,
+      "learning_rate": 9.914176864618891e-06,
+      "loss": 0.5303,
+      "step": 490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.003835054098607,
+      "learning_rate": 9.913642855103881e-06,
+      "loss": 0.4561,
+      "step": 491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.025050964061947,
+      "learning_rate": 9.913107203849464e-06,
+      "loss": 0.5947,
+      "step": 492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.663144493285422,
+      "learning_rate": 9.912569911034607e-06,
+      "loss": 0.6509,
+      "step": 493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.22745250393881,
+      "learning_rate": 9.912030976838832e-06,
+      "loss": 0.5649,
+      "step": 494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.50368431179958,
+      "learning_rate": 9.911490401442205e-06,
+      "loss": 0.5723,
+      "step": 495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.650046429480334,
+      "learning_rate": 9.910948185025345e-06,
+      "loss": 0.582,
+      "step": 496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.16731458869406,
+      "learning_rate": 9.910404327769414e-06,
+      "loss": 0.5161,
+      "step": 497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.48953220742883,
+      "learning_rate": 9.909858829856127e-06,
+      "loss": 0.5527,
+      "step": 498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.6203128320521,
+      "learning_rate": 9.909311691467744e-06,
+      "loss": 0.479,
+      "step": 499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.61392255485728,
+      "learning_rate": 9.908762912787073e-06,
+      "loss": 0.5513,
+      "step": 500
+    },
+    {
+      "epoch": 0.36,
+      "eval_avg_AUC": 0.7617813721540047,
+      "eval_avg_Accuracy": 0.688204575596817,
+      "eval_avg_Accuracy-right": 0.8745924090256946,
+      "eval_avg_Accuracy-wrong": 0.3632021833068001,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6387441987762204,
+      "eval_last_AUC": 0.7801619916516469,
+      "eval_last_Accuracy": 0.7156415782493368,
+      "eval_last_Accuracy-right": 0.7742924220686057,
+      "eval_last_Accuracy-wrong": 0.6133727541505573,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6543256471238833,
+      "eval_max_AUC": 0.6908342239463399,
+      "eval_max_Accuracy": 0.6420755968169761,
+      "eval_max_Accuracy-right": 0.9715012390765619,
+      "eval_max_Accuracy-wrong": 0.06765976802365249,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5464614776782707,
+      "eval_min_AUC": 0.7800757977941195,
+      "eval_min_Accuracy": 0.712823275862069,
+      "eval_min_Accuracy-right": 0.7380331289943916,
+      "eval_min_Accuracy-wrong": 0.6688651353195361,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6532406424802264,
+      "eval_prod_AUC": 0.7709256779799856,
+      "eval_prod_Accuracy": 0.6040285145888594,
+      "eval_prod_Accuracy-right": 0.41932959436546235,
+      "eval_prod_Accuracy-wrong": 0.9260859677052535,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6340698606601375,
+      "eval_runtime": 247.7022,
+      "eval_samples_per_second": 97.407,
+      "eval_steps_per_second": 3.044,
+      "eval_sum_AUC": 0.6279793611726269,
+      "eval_sum_Accuracy": 0.6357758620689655,
+      "eval_sum_Accuracy-right": 0.9995434981087779,
+      "eval_sum_Accuracy-wrong": 0.0014782806458949283,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6170558343639955,
+      "step": 500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.540498504399512,
+      "learning_rate": 9.908212493997473e-06,
+      "loss": 0.5596,
+      "step": 501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.128203532313956,
+      "learning_rate": 9.90766043528285e-06,
+      "loss": 0.5674,
+      "step": 502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.257665582917742,
+      "learning_rate": 9.907106736827654e-06,
+      "loss": 0.5996,
+      "step": 503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.247378361884045,
+      "learning_rate": 9.906551398816886e-06,
+      "loss": 0.4883,
+      "step": 504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.880394950980597,
+      "learning_rate": 9.9059944214361e-06,
+      "loss": 0.543,
+      "step": 505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.949181718064018,
+      "learning_rate": 9.905435804871387e-06,
+      "loss": 0.5933,
+      "step": 506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.010979155843657,
+      "learning_rate": 9.904875549309391e-06,
+      "loss": 0.6396,
+      "step": 507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.747061457011934,
+      "learning_rate": 9.904313654937308e-06,
+      "loss": 0.6533,
+      "step": 508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.93160977054565,
+      "learning_rate": 9.903750121942873e-06,
+      "loss": 0.5938,
+      "step": 509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.696205755954405,
+      "learning_rate": 9.903184950514378e-06,
+      "loss": 0.5122,
+      "step": 510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.53643132677427,
+      "learning_rate": 9.90261814084065e-06,
+      "loss": 0.5781,
+      "step": 511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.83447427750056,
+      "learning_rate": 9.902049693111077e-06,
+      "loss": 0.6719,
+      "step": 512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.7557972910997,
+      "learning_rate": 9.901479607515587e-06,
+      "loss": 0.5342,
+      "step": 513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.403858483082648,
+      "learning_rate": 9.900907884244654e-06,
+      "loss": 0.5684,
+      "step": 514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.25591218016626,
+      "learning_rate": 9.900334523489303e-06,
+      "loss": 0.5972,
+      "step": 515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.831414048462934,
+      "learning_rate": 9.899759525441101e-06,
+      "loss": 0.5625,
+      "step": 516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.09169003682686,
+      "learning_rate": 9.899182890292171e-06,
+      "loss": 0.5913,
+      "step": 517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.221486795019896,
+      "learning_rate": 9.898604618235175e-06,
+      "loss": 0.6055,
+      "step": 518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 19.080050148413697,
+      "learning_rate": 9.898024709463322e-06,
+      "loss": 0.6074,
+      "step": 519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.676674384714692,
+      "learning_rate": 9.897443164170375e-06,
+      "loss": 0.5405,
+      "step": 520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.304810818457433,
+      "learning_rate": 9.896859982550636e-06,
+      "loss": 0.4937,
+      "step": 521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.777058423733208,
+      "learning_rate": 9.89627516479896e-06,
+      "loss": 0.5742,
+      "step": 522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.391882166187722,
+      "learning_rate": 9.895688711110739e-06,
+      "loss": 0.5264,
+      "step": 523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.925774976971661,
+      "learning_rate": 9.895100621681923e-06,
+      "loss": 0.666,
+      "step": 524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.998754301379017,
+      "learning_rate": 9.894510896709003e-06,
+      "loss": 0.5176,
+      "step": 525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.840708276690297,
+      "learning_rate": 9.893919536389017e-06,
+      "loss": 0.5288,
+      "step": 526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.687147803937831,
+      "learning_rate": 9.89332654091955e-06,
+      "loss": 0.5239,
+      "step": 527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.768270858341022,
+      "learning_rate": 9.892731910498731e-06,
+      "loss": 0.5415,
+      "step": 528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.905157062520896,
+      "learning_rate": 9.892135645325238e-06,
+      "loss": 0.5664,
+      "step": 529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.120351808706744,
+      "learning_rate": 9.891537745598293e-06,
+      "loss": 0.5811,
+      "step": 530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.407988555074818,
+      "learning_rate": 9.89093821151767e-06,
+      "loss": 0.5762,
+      "step": 531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.3711923602355105,
+      "learning_rate": 9.89033704328368e-06,
+      "loss": 0.5137,
+      "step": 532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.0687402397265675,
+      "learning_rate": 9.889734241097186e-06,
+      "loss": 0.5273,
+      "step": 533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.00724274100732,
+      "learning_rate": 9.889129805159595e-06,
+      "loss": 0.5386,
+      "step": 534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.481688049622676,
+      "learning_rate": 9.888523735672861e-06,
+      "loss": 0.519,
+      "step": 535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.6367957136746245,
+      "learning_rate": 9.887916032839482e-06,
+      "loss": 0.498,
+      "step": 536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.466046619158893,
+      "learning_rate": 9.887306696862504e-06,
+      "loss": 0.4902,
+      "step": 537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.116707420357166,
+      "learning_rate": 9.886695727945515e-06,
+      "loss": 0.5771,
+      "step": 538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.725720682821406,
+      "learning_rate": 9.886083126292655e-06,
+      "loss": 0.5659,
+      "step": 539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.590126217196405,
+      "learning_rate": 9.885468892108603e-06,
+      "loss": 0.5752,
+      "step": 540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.37036153224896,
+      "learning_rate": 9.884853025598587e-06,
+      "loss": 0.5225,
+      "step": 541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.911884760434983,
+      "learning_rate": 9.884235526968377e-06,
+      "loss": 0.5542,
+      "step": 542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.973277734720734,
+      "learning_rate": 9.883616396424294e-06,
+      "loss": 0.5732,
+      "step": 543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.445859091078967,
+      "learning_rate": 9.8829956341732e-06,
+      "loss": 0.5679,
+      "step": 544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.059475523503908,
+      "learning_rate": 9.882373240422503e-06,
+      "loss": 0.5146,
+      "step": 545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 24.837290624700998,
+      "learning_rate": 9.881749215380156e-06,
+      "loss": 0.5596,
+      "step": 546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.213469746306272,
+      "learning_rate": 9.881123559254658e-06,
+      "loss": 0.5791,
+      "step": 547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.571221832522987,
+      "learning_rate": 9.880496272255053e-06,
+      "loss": 0.5654,
+      "step": 548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.884368751248441,
+      "learning_rate": 9.879867354590926e-06,
+      "loss": 0.5015,
+      "step": 549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.867690100404052,
+      "learning_rate": 9.879236806472414e-06,
+      "loss": 0.5454,
+      "step": 550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.247957723804134,
+      "learning_rate": 9.878604628110194e-06,
+      "loss": 0.5864,
+      "step": 551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.43931135081461,
+      "learning_rate": 9.877970819715485e-06,
+      "loss": 0.561,
+      "step": 552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.78533565974865,
+      "learning_rate": 9.87733538150006e-06,
+      "loss": 0.5269,
+      "step": 553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.031541245747464,
+      "learning_rate": 9.876698313676225e-06,
+      "loss": 0.5073,
+      "step": 554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.535835835682214,
+      "learning_rate": 9.876059616456842e-06,
+      "loss": 0.5723,
+      "step": 555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.31605031845537,
+      "learning_rate": 9.875419290055305e-06,
+      "loss": 0.4766,
+      "step": 556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.670021692503195,
+      "learning_rate": 9.874777334685565e-06,
+      "loss": 0.6074,
+      "step": 557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.219201299200287,
+      "learning_rate": 9.874133750562108e-06,
+      "loss": 0.5947,
+      "step": 558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.904120927324652,
+      "learning_rate": 9.873488537899967e-06,
+      "loss": 0.626,
+      "step": 559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.496631368652434,
+      "learning_rate": 9.872841696914721e-06,
+      "loss": 0.5391,
+      "step": 560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.536188576252165,
+      "learning_rate": 9.872193227822492e-06,
+      "loss": 0.5508,
+      "step": 561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.528143767413258,
+      "learning_rate": 9.871543130839944e-06,
+      "loss": 0.5732,
+      "step": 562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.58970570036274,
+      "learning_rate": 9.870891406184288e-06,
+      "loss": 0.5674,
+      "step": 563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.800186998521902,
+      "learning_rate": 9.870238054073275e-06,
+      "loss": 0.5566,
+      "step": 564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.911555208685854,
+      "learning_rate": 9.869583074725206e-06,
+      "loss": 0.6074,
+      "step": 565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.231300969398722,
+      "learning_rate": 9.868926468358919e-06,
+      "loss": 0.4829,
+      "step": 566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.709909484735207,
+      "learning_rate": 9.868268235193796e-06,
+      "loss": 0.502,
+      "step": 567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.178709235999069,
+      "learning_rate": 9.867608375449772e-06,
+      "loss": 0.5166,
+      "step": 568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.354410293206907,
+      "learning_rate": 9.866946889347311e-06,
+      "loss": 0.5703,
+      "step": 569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.915312310012048,
+      "learning_rate": 9.866283777107432e-06,
+      "loss": 0.5254,
+      "step": 570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.60559354889048,
+      "learning_rate": 9.865619038951692e-06,
+      "loss": 0.5996,
+      "step": 571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.26569614452492,
+      "learning_rate": 9.864952675102193e-06,
+      "loss": 0.5723,
+      "step": 572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.002089457228422,
+      "learning_rate": 9.864284685781578e-06,
+      "loss": 0.5864,
+      "step": 573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.044339027647595,
+      "learning_rate": 9.863615071213036e-06,
+      "loss": 0.5591,
+      "step": 574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.59838106899753,
+      "learning_rate": 9.862943831620298e-06,
+      "loss": 0.4917,
+      "step": 575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.247430329121268,
+      "learning_rate": 9.862270967227636e-06,
+      "loss": 0.5332,
+      "step": 576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.838445091054146,
+      "learning_rate": 9.861596478259869e-06,
+      "loss": 0.5576,
+      "step": 577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.4489106721379,
+      "learning_rate": 9.860920364942353e-06,
+      "loss": 0.5679,
+      "step": 578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.347633105018176,
+      "learning_rate": 9.860242627500994e-06,
+      "loss": 0.5747,
+      "step": 579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.918155502814725,
+      "learning_rate": 9.859563266162231e-06,
+      "loss": 0.5552,
+      "step": 580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.352511123265138,
+      "learning_rate": 9.858882281153058e-06,
+      "loss": 0.4307,
+      "step": 581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.102252433561581,
+      "learning_rate": 9.858199672701e-06,
+      "loss": 0.499,
+      "step": 582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.073596010036917,
+      "learning_rate": 9.85751544103413e-06,
+      "loss": 0.4551,
+      "step": 583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.478615776414374,
+      "learning_rate": 9.856829586381065e-06,
+      "loss": 0.542,
+      "step": 584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 31.02955657240596,
+      "learning_rate": 9.856142108970958e-06,
+      "loss": 0.6479,
+      "step": 585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.277538328288841,
+      "learning_rate": 9.855453009033512e-06,
+      "loss": 0.5327,
+      "step": 586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.683057931417592,
+      "learning_rate": 9.854762286798965e-06,
+      "loss": 0.5645,
+      "step": 587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.215793586089497,
+      "learning_rate": 9.854069942498102e-06,
+      "loss": 0.8555,
+      "step": 588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.562234064634753,
+      "learning_rate": 9.853375976362245e-06,
+      "loss": 0.5874,
+      "step": 589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.721200107181694,
+      "learning_rate": 9.852680388623266e-06,
+      "loss": 0.5298,
+      "step": 590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.093284898328076,
+      "learning_rate": 9.85198317951357e-06,
+      "loss": 0.6309,
+      "step": 591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.346628659885774,
+      "learning_rate": 9.851284349266107e-06,
+      "loss": 0.6416,
+      "step": 592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.619550827391471,
+      "learning_rate": 9.850583898114372e-06,
+      "loss": 0.6084,
+      "step": 593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.504830357240213,
+      "learning_rate": 9.849881826292399e-06,
+      "loss": 0.5742,
+      "step": 594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.655717476878046,
+      "learning_rate": 9.84917813403476e-06,
+      "loss": 0.6416,
+      "step": 595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.761291695955354,
+      "learning_rate": 9.848472821576572e-06,
+      "loss": 0.5938,
+      "step": 596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.531377595805906,
+      "learning_rate": 9.847765889153497e-06,
+      "loss": 0.5601,
+      "step": 597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.367401037379296,
+      "learning_rate": 9.847057337001731e-06,
+      "loss": 0.6104,
+      "step": 598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.851863355122642,
+      "learning_rate": 9.846347165358014e-06,
+      "loss": 0.5552,
+      "step": 599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.152768776233245,
+      "learning_rate": 9.84563537445963e-06,
+      "loss": 0.5654,
+      "step": 600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.288895704846823,
+      "learning_rate": 9.844921964544398e-06,
+      "loss": 0.542,
+      "step": 601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.57924848248388,
+      "learning_rate": 9.844206935850687e-06,
+      "loss": 0.585,
+      "step": 602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.139809939859182,
+      "learning_rate": 9.843490288617397e-06,
+      "loss": 0.5186,
+      "step": 603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.68694326416687,
+      "learning_rate": 9.842772023083972e-06,
+      "loss": 0.4717,
+      "step": 604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.74113624317417,
+      "learning_rate": 9.842052139490403e-06,
+      "loss": 0.5591,
+      "step": 605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.814314683170982,
+      "learning_rate": 9.841330638077213e-06,
+      "loss": 0.5303,
+      "step": 606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.263159508477482,
+      "learning_rate": 9.840607519085467e-06,
+      "loss": 0.5293,
+      "step": 607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.917568543252393,
+      "learning_rate": 9.839882782756778e-06,
+      "loss": 0.5425,
+      "step": 608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.506441410922555,
+      "learning_rate": 9.839156429333291e-06,
+      "loss": 0.5986,
+      "step": 609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.668039530691972,
+      "learning_rate": 9.838428459057694e-06,
+      "loss": 0.5171,
+      "step": 610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.4133141642419424,
+      "learning_rate": 9.837698872173214e-06,
+      "loss": 0.4961,
+      "step": 611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.506554715753216,
+      "learning_rate": 9.836967668923623e-06,
+      "loss": 0.564,
+      "step": 612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.580418265887094,
+      "learning_rate": 9.836234849553228e-06,
+      "loss": 0.4878,
+      "step": 613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.792067921572484,
+      "learning_rate": 9.835500414306875e-06,
+      "loss": 0.5903,
+      "step": 614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.889230874692596,
+      "learning_rate": 9.834764363429956e-06,
+      "loss": 0.501,
+      "step": 615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.876917354950514,
+      "learning_rate": 9.8340266971684e-06,
+      "loss": 0.5684,
+      "step": 616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.194977979042537,
+      "learning_rate": 9.83328741576867e-06,
+      "loss": 0.4902,
+      "step": 617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.105001380677454,
+      "learning_rate": 9.832546519477778e-06,
+      "loss": 0.5854,
+      "step": 618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.476338145150045,
+      "learning_rate": 9.831804008543271e-06,
+      "loss": 0.6064,
+      "step": 619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.759650991581609,
+      "learning_rate": 9.831059883213234e-06,
+      "loss": 0.6025,
+      "step": 620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.320154908129068,
+      "learning_rate": 9.830314143736292e-06,
+      "loss": 0.4985,
+      "step": 621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.063611986524947,
+      "learning_rate": 9.829566790361615e-06,
+      "loss": 0.5317,
+      "step": 622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.8381031832744865,
+      "learning_rate": 9.828817823338903e-06,
+      "loss": 0.5234,
+      "step": 623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.028721530123617,
+      "learning_rate": 9.828067242918402e-06,
+      "loss": 0.5938,
+      "step": 624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.738098966810448,
+      "learning_rate": 9.827315049350895e-06,
+      "loss": 0.5249,
+      "step": 625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.549683197853444,
+      "learning_rate": 9.826561242887704e-06,
+      "loss": 0.5049,
+      "step": 626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.847026094764958,
+      "learning_rate": 9.825805823780687e-06,
+      "loss": 0.562,
+      "step": 627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.1475623479236985,
+      "learning_rate": 9.825048792282247e-06,
+      "loss": 0.5259,
+      "step": 628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.57915719785624,
+      "learning_rate": 9.824290148645322e-06,
+      "loss": 0.5098,
+      "step": 629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.748425573675469,
+      "learning_rate": 9.823529893123384e-06,
+      "loss": 0.5869,
+      "step": 630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.059327133724575,
+      "learning_rate": 9.822768025970456e-06,
+      "loss": 0.4497,
+      "step": 631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.293876053511848,
+      "learning_rate": 9.822004547441088e-06,
+      "loss": 0.6064,
+      "step": 632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.38956537939008,
+      "learning_rate": 9.821239457790373e-06,
+      "loss": 0.624,
+      "step": 633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.857454287932656,
+      "learning_rate": 9.82047275727394e-06,
+      "loss": 0.5557,
+      "step": 634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.129814430870212,
+      "learning_rate": 9.81970444614796e-06,
+      "loss": 0.5376,
+      "step": 635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.239828802765227,
+      "learning_rate": 9.81893452466914e-06,
+      "loss": 0.5459,
+      "step": 636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.759535502268271,
+      "learning_rate": 9.818162993094724e-06,
+      "loss": 0.5229,
+      "step": 637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.269358730497334,
+      "learning_rate": 9.817389851682494e-06,
+      "loss": 0.5811,
+      "step": 638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.532561868563791,
+      "learning_rate": 9.816615100690773e-06,
+      "loss": 0.5068,
+      "step": 639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.542345953735255,
+      "learning_rate": 9.81583874037842e-06,
+      "loss": 0.5195,
+      "step": 640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.295525138598315,
+      "learning_rate": 9.815060771004831e-06,
+      "loss": 0.5811,
+      "step": 641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.216808140877509,
+      "learning_rate": 9.81428119282994e-06,
+      "loss": 0.5483,
+      "step": 642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.054796365751393,
+      "learning_rate": 9.813500006114216e-06,
+      "loss": 0.5225,
+      "step": 643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.532068190094899,
+      "learning_rate": 9.812717211118673e-06,
+      "loss": 0.5259,
+      "step": 644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.617232952388049,
+      "learning_rate": 9.811932808104852e-06,
+      "loss": 0.583,
+      "step": 645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.53476537868423,
+      "learning_rate": 9.811146797334838e-06,
+      "loss": 0.5986,
+      "step": 646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.459867876200462,
+      "learning_rate": 9.810359179071255e-06,
+      "loss": 0.4854,
+      "step": 647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.442128811771815,
+      "learning_rate": 9.809569953577258e-06,
+      "loss": 0.5283,
+      "step": 648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.05485902012583,
+      "learning_rate": 9.808779121116542e-06,
+      "loss": 0.5869,
+      "step": 649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.614481323202112,
+      "learning_rate": 9.807986681953341e-06,
+      "loss": 0.5005,
+      "step": 650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.302158068476116,
+      "learning_rate": 9.807192636352422e-06,
+      "loss": 0.5518,
+      "step": 651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.246031731872442,
+      "learning_rate": 9.80639698457909e-06,
+      "loss": 0.5,
+      "step": 652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.741614003843505,
+      "learning_rate": 9.805599726899188e-06,
+      "loss": 0.5195,
+      "step": 653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.914045939553276,
+      "learning_rate": 9.804800863579094e-06,
+      "loss": 0.5059,
+      "step": 654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.602058773881305,
+      "learning_rate": 9.804000394885723e-06,
+      "loss": 0.5132,
+      "step": 655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.395330548720693,
+      "learning_rate": 9.803198321086527e-06,
+      "loss": 0.6504,
+      "step": 656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.082779853562936,
+      "learning_rate": 9.802394642449494e-06,
+      "loss": 0.5503,
+      "step": 657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.05035071434153,
+      "learning_rate": 9.801589359243147e-06,
+      "loss": 0.5625,
+      "step": 658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.74752221065336,
+      "learning_rate": 9.800782471736547e-06,
+      "loss": 0.5024,
+      "step": 659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.93020384218078,
+      "learning_rate": 9.799973980199288e-06,
+      "loss": 0.5679,
+      "step": 660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.58689659447437,
+      "learning_rate": 9.799163884901506e-06,
+      "loss": 0.5435,
+      "step": 661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.583439466485594,
+      "learning_rate": 9.798352186113867e-06,
+      "loss": 0.5093,
+      "step": 662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.423178186172265,
+      "learning_rate": 9.797538884107574e-06,
+      "loss": 0.6123,
+      "step": 663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.533199007539238,
+      "learning_rate": 9.796723979154366e-06,
+      "loss": 0.5229,
+      "step": 664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.971226162816325,
+      "learning_rate": 9.795907471526518e-06,
+      "loss": 0.5547,
+      "step": 665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.783730829396234,
+      "learning_rate": 9.79508936149684e-06,
+      "loss": 0.4644,
+      "step": 666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.03312528136628,
+      "learning_rate": 9.79426964933868e-06,
+      "loss": 0.543,
+      "step": 667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.283132104994282,
+      "learning_rate": 9.793448335325919e-06,
+      "loss": 0.4917,
+      "step": 668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.20800855604922,
+      "learning_rate": 9.792625419732969e-06,
+      "loss": 0.5264,
+      "step": 669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.415376972801536,
+      "learning_rate": 9.791800902834787e-06,
+      "loss": 0.4824,
+      "step": 670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.273820442326011,
+      "learning_rate": 9.790974784906855e-06,
+      "loss": 0.501,
+      "step": 671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.671344611499754,
+      "learning_rate": 9.790147066225198e-06,
+      "loss": 0.605,
+      "step": 672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.088843769717563,
+      "learning_rate": 9.789317747066369e-06,
+      "loss": 0.5586,
+      "step": 673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.751038170132446,
+      "learning_rate": 9.788486827707462e-06,
+      "loss": 0.6519,
+      "step": 674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.207892997115179,
+      "learning_rate": 9.7876543084261e-06,
+      "loss": 0.5347,
+      "step": 675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.857264061196519,
+      "learning_rate": 9.786820189500443e-06,
+      "loss": 0.5464,
+      "step": 676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.830984057445567,
+      "learning_rate": 9.785984471209186e-06,
+      "loss": 0.5273,
+      "step": 677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.60468250187857,
+      "learning_rate": 9.785147153831562e-06,
+      "loss": 0.5459,
+      "step": 678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.593019502309981,
+      "learning_rate": 9.784308237647329e-06,
+      "loss": 0.5757,
+      "step": 679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.955889286248375,
+      "learning_rate": 9.783467722936786e-06,
+      "loss": 0.583,
+      "step": 680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.939295729656601,
+      "learning_rate": 9.782625609980767e-06,
+      "loss": 0.5991,
+      "step": 681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.514876614342388,
+      "learning_rate": 9.781781899060635e-06,
+      "loss": 0.5703,
+      "step": 682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.553266494196397,
+      "learning_rate": 9.78093659045829e-06,
+      "loss": 0.5435,
+      "step": 683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.263926473907152,
+      "learning_rate": 9.780089684456164e-06,
+      "loss": 0.5229,
+      "step": 684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.635255900891267,
+      "learning_rate": 9.779241181337228e-06,
+      "loss": 0.5176,
+      "step": 685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.608059375423723,
+      "learning_rate": 9.778391081384979e-06,
+      "loss": 0.5645,
+      "step": 686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.41849821192734,
+      "learning_rate": 9.777539384883453e-06,
+      "loss": 0.4922,
+      "step": 687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.159527005843714,
+      "learning_rate": 9.776686092117216e-06,
+      "loss": 0.5977,
+      "step": 688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.36894102930875,
+      "learning_rate": 9.775831203371371e-06,
+      "loss": 0.5693,
+      "step": 689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.091636163606054,
+      "learning_rate": 9.774974718931551e-06,
+      "loss": 0.6221,
+      "step": 690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.969604571791074,
+      "learning_rate": 9.774116639083923e-06,
+      "loss": 0.5854,
+      "step": 691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.88504893165957,
+      "learning_rate": 9.773256964115189e-06,
+      "loss": 0.5049,
+      "step": 692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.758445464297264,
+      "learning_rate": 9.772395694312583e-06,
+      "loss": 0.603,
+      "step": 693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.955201215215652,
+      "learning_rate": 9.771532829963865e-06,
+      "loss": 0.5571,
+      "step": 694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.013451757180628,
+      "learning_rate": 9.770668371357344e-06,
+      "loss": 0.4849,
+      "step": 695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.6991625858998365,
+      "learning_rate": 9.769802318781842e-06,
+      "loss": 0.5337,
+      "step": 696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.575556491861107,
+      "learning_rate": 9.76893467252673e-06,
+      "loss": 0.5771,
+      "step": 697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.83427689960832,
+      "learning_rate": 9.768065432881903e-06,
+      "loss": 0.5859,
+      "step": 698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.23151806170552,
+      "learning_rate": 9.767194600137789e-06,
+      "loss": 0.4951,
+      "step": 699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.185321738716956,
+      "learning_rate": 9.766322174585347e-06,
+      "loss": 0.5342,
+      "step": 700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.443299162915134,
+      "learning_rate": 9.765448156516077e-06,
+      "loss": 0.4678,
+      "step": 701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.19817487137709,
+      "learning_rate": 9.764572546222e-06,
+      "loss": 0.5684,
+      "step": 702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.21262946620503,
+      "learning_rate": 9.763695343995674e-06,
+      "loss": 0.5308,
+      "step": 703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.192980873552312,
+      "learning_rate": 9.762816550130192e-06,
+      "loss": 0.5112,
+      "step": 704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.960657984729142,
+      "learning_rate": 9.76193616491917e-06,
+      "loss": 0.6094,
+      "step": 705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.679940313177855,
+      "learning_rate": 9.761054188656766e-06,
+      "loss": 0.5415,
+      "step": 706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.09877092756067,
+      "learning_rate": 9.760170621637661e-06,
+      "loss": 0.5601,
+      "step": 707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.479043385831767,
+      "learning_rate": 9.759285464157073e-06,
+      "loss": 0.5474,
+      "step": 708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.234613065252855,
+      "learning_rate": 9.758398716510751e-06,
+      "loss": 0.501,
+      "step": 709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.64323110344669,
+      "learning_rate": 9.75751037899497e-06,
+      "loss": 0.5635,
+      "step": 710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.041571595123274,
+      "learning_rate": 9.756620451906543e-06,
+      "loss": 0.5952,
+      "step": 711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.629287618121229,
+      "learning_rate": 9.75572893554281e-06,
+      "loss": 0.5493,
+      "step": 712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.5855291941636755,
+      "learning_rate": 9.754835830201645e-06,
+      "loss": 0.5557,
+      "step": 713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.8782598567148705,
+      "learning_rate": 9.753941136181448e-06,
+      "loss": 0.5273,
+      "step": 714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.85114790600747,
+      "learning_rate": 9.753044853781155e-06,
+      "loss": 0.5078,
+      "step": 715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.657250772447208,
+      "learning_rate": 9.75214698330023e-06,
+      "loss": 0.6157,
+      "step": 716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 17.576065249418974,
+      "learning_rate": 9.751247525038669e-06,
+      "loss": 0.4863,
+      "step": 717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 33.781222697734435,
+      "learning_rate": 9.750346479296998e-06,
+      "loss": 0.6094,
+      "step": 718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 185.99349497355314,
+      "learning_rate": 9.74944384637627e-06,
+      "loss": 0.625,
+      "step": 719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.364273036024208,
+      "learning_rate": 9.748539626578076e-06,
+      "loss": 0.4727,
+      "step": 720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.47582915048441,
+      "learning_rate": 9.747633820204527e-06,
+      "loss": 0.4775,
+      "step": 721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 160.92451037044142,
+      "learning_rate": 9.746726427558276e-06,
+      "loss": 0.583,
+      "step": 722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.29556717471361,
+      "learning_rate": 9.745817448942496e-06,
+      "loss": 0.6426,
+      "step": 723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.845209139079835,
+      "learning_rate": 9.744906884660894e-06,
+      "loss": 0.584,
+      "step": 724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.487953481735268,
+      "learning_rate": 9.743994735017708e-06,
+      "loss": 0.4824,
+      "step": 725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.209235710267066,
+      "learning_rate": 9.743081000317703e-06,
+      "loss": 0.6045,
+      "step": 726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.112078657361787,
+      "learning_rate": 9.742165680866173e-06,
+      "loss": 0.5244,
+      "step": 727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.970186411419759,
+      "learning_rate": 9.741248776968947e-06,
+      "loss": 0.5825,
+      "step": 728
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.315186088300615,
+      "learning_rate": 9.740330288932379e-06,
+      "loss": 0.5918,
+      "step": 729
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.457437824827904,
+      "learning_rate": 9.73941021706335e-06,
+      "loss": 0.5781,
+      "step": 730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.793356765325614,
+      "learning_rate": 9.738488561669272e-06,
+      "loss": 0.5474,
+      "step": 731
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.112475758078757,
+      "learning_rate": 9.737565323058094e-06,
+      "loss": 0.5337,
+      "step": 732
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.954732042734927,
+      "learning_rate": 9.736640501538281e-06,
+      "loss": 0.5552,
+      "step": 733
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.682917509984245,
+      "learning_rate": 9.735714097418835e-06,
+      "loss": 0.5811,
+      "step": 734
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.954785296883502,
+      "learning_rate": 9.734786111009287e-06,
+      "loss": 0.5283,
+      "step": 735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.187461253271188,
+      "learning_rate": 9.73385654261969e-06,
+      "loss": 0.5161,
+      "step": 736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.105803753991632,
+      "learning_rate": 9.732925392560634e-06,
+      "loss": 0.5781,
+      "step": 737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.780060124597431,
+      "learning_rate": 9.731992661143233e-06,
+      "loss": 0.5137,
+      "step": 738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.149918598961957,
+      "learning_rate": 9.731058348679128e-06,
+      "loss": 0.5459,
+      "step": 739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.600033304087317,
+      "learning_rate": 9.73012245548049e-06,
+      "loss": 0.6167,
+      "step": 740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.959608925181033,
+      "learning_rate": 9.729184981860023e-06,
+      "loss": 0.5547,
+      "step": 741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.421233340277922,
+      "learning_rate": 9.728245928130949e-06,
+      "loss": 0.4907,
+      "step": 742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.131878096040454,
+      "learning_rate": 9.727305294607024e-06,
+      "loss": 0.5337,
+      "step": 743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.933583907781678,
+      "learning_rate": 9.726363081602532e-06,
+      "loss": 0.4868,
+      "step": 744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.773000311800002,
+      "learning_rate": 9.725419289432287e-06,
+      "loss": 0.5586,
+      "step": 745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.30603969977933,
+      "learning_rate": 9.724473918411624e-06,
+      "loss": 0.5532,
+      "step": 746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.45632748446621,
+      "learning_rate": 9.723526968856408e-06,
+      "loss": 0.4321,
+      "step": 747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.567004859337702,
+      "learning_rate": 9.722578441083035e-06,
+      "loss": 0.4902,
+      "step": 748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.086491004356668,
+      "learning_rate": 9.721628335408423e-06,
+      "loss": 0.5205,
+      "step": 749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.63422639061617,
+      "learning_rate": 9.720676652150025e-06,
+      "loss": 0.5288,
+      "step": 750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.428381354480067,
+      "learning_rate": 9.719723391625813e-06,
+      "loss": 0.6455,
+      "step": 751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.967731907649567,
+      "learning_rate": 9.718768554154287e-06,
+      "loss": 0.6621,
+      "step": 752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.977453743909852,
+      "learning_rate": 9.717812140054479e-06,
+      "loss": 0.5083,
+      "step": 753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.976246002857728,
+      "learning_rate": 9.716854149645945e-06,
+      "loss": 0.5532,
+      "step": 754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.694789061118618,
+      "learning_rate": 9.715894583248764e-06,
+      "loss": 0.5239,
+      "step": 755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.615602095338727,
+      "learning_rate": 9.714933441183549e-06,
+      "loss": 0.4473,
+      "step": 756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.216340319991541,
+      "learning_rate": 9.713970723771432e-06,
+      "loss": 0.479,
+      "step": 757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.167611058044052,
+      "learning_rate": 9.713006431334076e-06,
+      "loss": 0.5361,
+      "step": 758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.666468658089926,
+      "learning_rate": 9.71204056419367e-06,
+      "loss": 0.5303,
+      "step": 759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.66660693531791,
+      "learning_rate": 9.711073122672928e-06,
+      "loss": 0.5464,
+      "step": 760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 29.673553271376097,
+      "learning_rate": 9.71010410709509e-06,
+      "loss": 0.6133,
+      "step": 761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.889123197476716,
+      "learning_rate": 9.70913351778392e-06,
+      "loss": 0.5479,
+      "step": 762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.884074672070827,
+      "learning_rate": 9.708161355063714e-06,
+      "loss": 0.623,
+      "step": 763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.200359924204133,
+      "learning_rate": 9.707187619259286e-06,
+      "loss": 0.46,
+      "step": 764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.9354480427811,
+      "learning_rate": 9.706212310695981e-06,
+      "loss": 0.5781,
+      "step": 765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 29.381896018391878,
+      "learning_rate": 9.705235429699666e-06,
+      "loss": 0.6304,
+      "step": 766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.16877662873349,
+      "learning_rate": 9.704256976596737e-06,
+      "loss": 0.5361,
+      "step": 767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.227617088641859,
+      "learning_rate": 9.703276951714114e-06,
+      "loss": 0.4468,
+      "step": 768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.762967619147144,
+      "learning_rate": 9.70229535537924e-06,
+      "loss": 0.4521,
+      "step": 769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.62549106013863,
+      "learning_rate": 9.701312187920084e-06,
+      "loss": 0.7197,
+      "step": 770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.332401107525445,
+      "learning_rate": 9.700327449665143e-06,
+      "loss": 0.6289,
+      "step": 771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.306780662436637,
+      "learning_rate": 9.699341140943434e-06,
+      "loss": 0.5376,
+      "step": 772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.721571034550369,
+      "learning_rate": 9.698353262084501e-06,
+      "loss": 0.4839,
+      "step": 773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.264526568674992,
+      "learning_rate": 9.697363813418414e-06,
+      "loss": 0.5029,
+      "step": 774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 28.02052415350964,
+      "learning_rate": 9.696372795275766e-06,
+      "loss": 0.6436,
+      "step": 775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.973871491555215,
+      "learning_rate": 9.695380207987675e-06,
+      "loss": 0.5459,
+      "step": 776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.86541152325245,
+      "learning_rate": 9.69438605188578e-06,
+      "loss": 0.6797,
+      "step": 777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.430529034939772,
+      "learning_rate": 9.69339032730225e-06,
+      "loss": 0.5947,
+      "step": 778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.136839582171906,
+      "learning_rate": 9.692393034569776e-06,
+      "loss": 0.5117,
+      "step": 779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.567653375487124,
+      "learning_rate": 9.69139417402157e-06,
+      "loss": 0.4937,
+      "step": 780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 38.10236907533684,
+      "learning_rate": 9.690393745991368e-06,
+      "loss": 0.6172,
+      "step": 781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 40.793834029889545,
+      "learning_rate": 9.689391750813436e-06,
+      "loss": 0.8066,
+      "step": 782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.16046506654703,
+      "learning_rate": 9.688388188822556e-06,
+      "loss": 0.7109,
+      "step": 783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.209686684667055,
+      "learning_rate": 9.687383060354038e-06,
+      "loss": 0.623,
+      "step": 784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.029930794357556,
+      "learning_rate": 9.686376365743714e-06,
+      "loss": 0.5557,
+      "step": 785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.702750453811597,
+      "learning_rate": 9.685368105327938e-06,
+      "loss": 0.5537,
+      "step": 786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 23.2055662093521,
+      "learning_rate": 9.684358279443593e-06,
+      "loss": 0.5771,
+      "step": 787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.310900740988059,
+      "learning_rate": 9.683346888428074e-06,
+      "loss": 0.4946,
+      "step": 788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.032483835960253,
+      "learning_rate": 9.68233393261931e-06,
+      "loss": 0.5244,
+      "step": 789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.215648584728573,
+      "learning_rate": 9.681319412355748e-06,
+      "loss": 0.5078,
+      "step": 790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.062322777960844,
+      "learning_rate": 9.680303327976356e-06,
+      "loss": 0.5327,
+      "step": 791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.167802022767416,
+      "learning_rate": 9.679285679820628e-06,
+      "loss": 0.501,
+      "step": 792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.46849010922445,
+      "learning_rate": 9.67826646822858e-06,
+      "loss": 0.5239,
+      "step": 793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.653596919202503,
+      "learning_rate": 9.677245693540749e-06,
+      "loss": 0.5176,
+      "step": 794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.021405696985937,
+      "learning_rate": 9.676223356098194e-06,
+      "loss": 0.4546,
+      "step": 795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.864820200236489,
+      "learning_rate": 9.675199456242499e-06,
+      "loss": 0.5151,
+      "step": 796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.137355785760509,
+      "learning_rate": 9.674173994315764e-06,
+      "loss": 0.6768,
+      "step": 797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.117836656138127,
+      "learning_rate": 9.67314697066062e-06,
+      "loss": 0.542,
+      "step": 798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.298051983963956,
+      "learning_rate": 9.672118385620209e-06,
+      "loss": 0.4927,
+      "step": 799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.163165582397156,
+      "learning_rate": 9.671088239538204e-06,
+      "loss": 0.7588,
+      "step": 800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.215200611680789,
+      "learning_rate": 9.670056532758798e-06,
+      "loss": 0.5474,
+      "step": 801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.42334654362743,
+      "learning_rate": 9.669023265626698e-06,
+      "loss": 0.6289,
+      "step": 802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.141287911310176,
+      "learning_rate": 9.66798843848714e-06,
+      "loss": 0.4873,
+      "step": 803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.552342787875665,
+      "learning_rate": 9.666952051685882e-06,
+      "loss": 0.4731,
+      "step": 804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.755001186494166,
+      "learning_rate": 9.665914105569196e-06,
+      "loss": 0.5591,
+      "step": 805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.425217131212628,
+      "learning_rate": 9.664874600483883e-06,
+      "loss": 0.5239,
+      "step": 806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.638565826764706,
+      "learning_rate": 9.663833536777256e-06,
+      "loss": 0.5005,
+      "step": 807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.967345411360103,
+      "learning_rate": 9.662790914797158e-06,
+      "loss": 0.6133,
+      "step": 808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.774853818643642,
+      "learning_rate": 9.661746734891947e-06,
+      "loss": 0.5449,
+      "step": 809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.358167390535703,
+      "learning_rate": 9.6607009974105e-06,
+      "loss": 0.5386,
+      "step": 810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.443902186847719,
+      "learning_rate": 9.659653702702223e-06,
+      "loss": 0.519,
+      "step": 811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.581607014220506,
+      "learning_rate": 9.658604851117032e-06,
+      "loss": 0.6064,
+      "step": 812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.10227881796404,
+      "learning_rate": 9.65755444300537e-06,
+      "loss": 0.5649,
+      "step": 813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.936458119810688,
+      "learning_rate": 9.656502478718197e-06,
+      "loss": 0.5459,
+      "step": 814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.720426589157082,
+      "learning_rate": 9.655448958606994e-06,
+      "loss": 0.6309,
+      "step": 815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.062061943723108,
+      "learning_rate": 9.654393883023763e-06,
+      "loss": 0.5693,
+      "step": 816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.539927314427118,
+      "learning_rate": 9.653337252321023e-06,
+      "loss": 0.584,
+      "step": 817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.01962400655284,
+      "learning_rate": 9.652279066851811e-06,
+      "loss": 0.5146,
+      "step": 818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.278364471511755,
+      "learning_rate": 9.651219326969694e-06,
+      "loss": 0.4888,
+      "step": 819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.862188944826272,
+      "learning_rate": 9.650158033028743e-06,
+      "loss": 0.5386,
+      "step": 820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.55019661530497,
+      "learning_rate": 9.64909518538356e-06,
+      "loss": 0.4985,
+      "step": 821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.736899314710898,
+      "learning_rate": 9.648030784389264e-06,
+      "loss": 0.5303,
+      "step": 822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.366946707310364,
+      "learning_rate": 9.646964830401487e-06,
+      "loss": 0.6001,
+      "step": 823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.13088503009922,
+      "learning_rate": 9.645897323776386e-06,
+      "loss": 0.5889,
+      "step": 824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.30858943174471,
+      "learning_rate": 9.644828264870634e-06,
+      "loss": 0.5371,
+      "step": 825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.315173886504947,
+      "learning_rate": 9.643757654041423e-06,
+      "loss": 0.5508,
+      "step": 826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.679272046818679,
+      "learning_rate": 9.642685491646467e-06,
+      "loss": 0.481,
+      "step": 827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.93076938329635,
+      "learning_rate": 9.641611778043992e-06,
+      "loss": 0.5361,
+      "step": 828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.131118051844421,
+      "learning_rate": 9.64053651359275e-06,
+      "loss": 0.4561,
+      "step": 829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 16.845626382905177,
+      "learning_rate": 9.639459698652e-06,
+      "loss": 0.5811,
+      "step": 830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.452226118644797,
+      "learning_rate": 9.63838133358153e-06,
+      "loss": 0.5049,
+      "step": 831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.715195201767449,
+      "learning_rate": 9.637301418741643e-06,
+      "loss": 0.5732,
+      "step": 832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.948105431930284,
+      "learning_rate": 9.636219954493157e-06,
+      "loss": 0.5264,
+      "step": 833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.493455666198669,
+      "learning_rate": 9.635136941197409e-06,
+      "loss": 0.5498,
+      "step": 834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.154843832131087,
+      "learning_rate": 9.634052379216256e-06,
+      "loss": 0.5889,
+      "step": 835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9733407722182035,
+      "learning_rate": 9.632966268912067e-06,
+      "loss": 0.5796,
+      "step": 836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.393897880638866,
+      "learning_rate": 9.631878610647734e-06,
+      "loss": 0.5762,
+      "step": 837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.346535492939477,
+      "learning_rate": 9.630789404786664e-06,
+      "loss": 0.5845,
+      "step": 838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.414814582454439,
+      "learning_rate": 9.629698651692779e-06,
+      "loss": 0.519,
+      "step": 839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.027989103111617,
+      "learning_rate": 9.62860635173052e-06,
+      "loss": 0.5229,
+      "step": 840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.691852570057074,
+      "learning_rate": 9.627512505264847e-06,
+      "loss": 0.54,
+      "step": 841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 13.672266244948228,
+      "learning_rate": 9.626417112661233e-06,
+      "loss": 0.501,
+      "step": 842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.046157418465752,
+      "learning_rate": 9.62532017428567e-06,
+      "loss": 0.5464,
+      "step": 843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.894528680541082,
+      "learning_rate": 9.624221690504663e-06,
+      "loss": 0.5454,
+      "step": 844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 19.359046304477772,
+      "learning_rate": 9.623121661685239e-06,
+      "loss": 0.5347,
+      "step": 845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.931630197988305,
+      "learning_rate": 9.622020088194934e-06,
+      "loss": 0.5518,
+      "step": 846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.126175546600982,
+      "learning_rate": 9.62091697040181e-06,
+      "loss": 0.5073,
+      "step": 847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.424497735342813,
+      "learning_rate": 9.619812308674434e-06,
+      "loss": 0.5981,
+      "step": 848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.429709124263976,
+      "learning_rate": 9.618706103381896e-06,
+      "loss": 0.5327,
+      "step": 849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.370766522473178,
+      "learning_rate": 9.6175983548938e-06,
+      "loss": 0.5356,
+      "step": 850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.936351360877747,
+      "learning_rate": 9.616489063580265e-06,
+      "loss": 0.6279,
+      "step": 851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.565406309781352,
+      "learning_rate": 9.615378229811927e-06,
+      "loss": 0.498,
+      "step": 852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.85975243185665,
+      "learning_rate": 9.614265853959935e-06,
+      "loss": 0.5449,
+      "step": 853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.376637510131717,
+      "learning_rate": 9.613151936395952e-06,
+      "loss": 0.6953,
+      "step": 854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.639153075645197,
+      "learning_rate": 9.612036477492163e-06,
+      "loss": 0.5469,
+      "step": 855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 23.55427631393824,
+      "learning_rate": 9.610919477621262e-06,
+      "loss": 0.5225,
+      "step": 856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.272759938860226,
+      "learning_rate": 9.609800937156459e-06,
+      "loss": 0.6147,
+      "step": 857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.7353396751586585,
+      "learning_rate": 9.60868085647148e-06,
+      "loss": 0.6284,
+      "step": 858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.677161643210795,
+      "learning_rate": 9.607559235940562e-06,
+      "loss": 0.4409,
+      "step": 859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.674513654587285,
+      "learning_rate": 9.60643607593846e-06,
+      "loss": 0.5229,
+      "step": 860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.208902038096937,
+      "learning_rate": 9.605311376840446e-06,
+      "loss": 0.4419,
+      "step": 861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.200913020665042,
+      "learning_rate": 9.604185139022302e-06,
+      "loss": 0.5039,
+      "step": 862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.88962045733014,
+      "learning_rate": 9.603057362860323e-06,
+      "loss": 0.6357,
+      "step": 863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.58000300790741,
+      "learning_rate": 9.60192804873132e-06,
+      "loss": 0.4673,
+      "step": 864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.975171350200098,
+      "learning_rate": 9.60079719701262e-06,
+      "loss": 0.5605,
+      "step": 865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.2559604366098025,
+      "learning_rate": 9.599664808082058e-06,
+      "loss": 0.5229,
+      "step": 866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 30.607460221946777,
+      "learning_rate": 9.598530882317992e-06,
+      "loss": 0.7324,
+      "step": 867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.12508471498375,
+      "learning_rate": 9.59739542009928e-06,
+      "loss": 0.6279,
+      "step": 868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.213276965176858,
+      "learning_rate": 9.596258421805306e-06,
+      "loss": 0.54,
+      "step": 869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.939321698210003,
+      "learning_rate": 9.595119887815962e-06,
+      "loss": 0.5654,
+      "step": 870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.54597463715274,
+      "learning_rate": 9.593979818511655e-06,
+      "loss": 0.5391,
+      "step": 871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.796494601908567,
+      "learning_rate": 9.592838214273298e-06,
+      "loss": 0.6221,
+      "step": 872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.350816902914385,
+      "learning_rate": 9.591695075482326e-06,
+      "loss": 0.5195,
+      "step": 873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.356102942266837,
+      "learning_rate": 9.590550402520683e-06,
+      "loss": 0.5361,
+      "step": 874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.949881570510586,
+      "learning_rate": 9.589404195770821e-06,
+      "loss": 0.5151,
+      "step": 875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.926146321546089,
+      "learning_rate": 9.588256455615716e-06,
+      "loss": 0.5801,
+      "step": 876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.859921527227033,
+      "learning_rate": 9.587107182438846e-06,
+      "loss": 0.5,
+      "step": 877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.050613799522687,
+      "learning_rate": 9.585956376624204e-06,
+      "loss": 0.5527,
+      "step": 878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.002907445909361,
+      "learning_rate": 9.584804038556297e-06,
+      "loss": 0.6289,
+      "step": 879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.0542998877576375,
+      "learning_rate": 9.58365016862014e-06,
+      "loss": 0.4893,
+      "step": 880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.56578174324824,
+      "learning_rate": 9.582494767201265e-06,
+      "loss": 0.5493,
+      "step": 881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.301104366990973,
+      "learning_rate": 9.581337834685713e-06,
+      "loss": 0.4365,
+      "step": 882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 21.230580902888516,
+      "learning_rate": 9.580179371460034e-06,
+      "loss": 0.5376,
+      "step": 883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.115363175931497,
+      "learning_rate": 9.579019377911296e-06,
+      "loss": 0.5518,
+      "step": 884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.063043768232241,
+      "learning_rate": 9.57785785442707e-06,
+      "loss": 0.5142,
+      "step": 885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.200092331472336,
+      "learning_rate": 9.576694801395447e-06,
+      "loss": 0.4912,
+      "step": 886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 16.048503793582384,
+      "learning_rate": 9.57553021920502e-06,
+      "loss": 0.708,
+      "step": 887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.183349614718361,
+      "learning_rate": 9.574364108244903e-06,
+      "loss": 0.459,
+      "step": 888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.786011428361001,
+      "learning_rate": 9.573196468904711e-06,
+      "loss": 0.5586,
+      "step": 889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.666551982517106,
+      "learning_rate": 9.572027301574576e-06,
+      "loss": 0.6201,
+      "step": 890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.989372858349427,
+      "learning_rate": 9.570856606645139e-06,
+      "loss": 0.543,
+      "step": 891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.978420031263653,
+      "learning_rate": 9.569684384507547e-06,
+      "loss": 0.4585,
+      "step": 892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.494441221850212,
+      "learning_rate": 9.568510635553466e-06,
+      "loss": 0.5176,
+      "step": 893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.496868914068109,
+      "learning_rate": 9.567335360175065e-06,
+      "loss": 0.5283,
+      "step": 894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.82256439758004,
+      "learning_rate": 9.566158558765026e-06,
+      "loss": 0.6777,
+      "step": 895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.172352317808183,
+      "learning_rate": 9.564980231716541e-06,
+      "loss": 0.5371,
+      "step": 896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.641915716101109,
+      "learning_rate": 9.56380037942331e-06,
+      "loss": 0.5103,
+      "step": 897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.279754257232414,
+      "learning_rate": 9.562619002279541e-06,
+      "loss": 0.5737,
+      "step": 898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.850853767539375,
+      "learning_rate": 9.561436100679959e-06,
+      "loss": 0.5244,
+      "step": 899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.349447369826582,
+      "learning_rate": 9.56025167501979e-06,
+      "loss": 0.4907,
+      "step": 900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 15.819552722595972,
+      "learning_rate": 9.559065725694775e-06,
+      "loss": 0.6021,
+      "step": 901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.09721976041659,
+      "learning_rate": 9.55787825310116e-06,
+      "loss": 0.5762,
+      "step": 902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.611541858548501,
+      "learning_rate": 9.5566892576357e-06,
+      "loss": 0.4741,
+      "step": 903
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.17797250965005,
+      "learning_rate": 9.555498739695665e-06,
+      "loss": 0.6016,
+      "step": 904
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.107430166339677,
+      "learning_rate": 9.554306699678827e-06,
+      "loss": 0.5166,
+      "step": 905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.571816321971257,
+      "learning_rate": 9.553113137983467e-06,
+      "loss": 0.5151,
+      "step": 906
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.8343904045849095,
+      "learning_rate": 9.551918055008378e-06,
+      "loss": 0.5376,
+      "step": 907
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.669597503208232,
+      "learning_rate": 9.55072145115286e-06,
+      "loss": 0.5171,
+      "step": 908
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.787777933655681,
+      "learning_rate": 9.54952332681672e-06,
+      "loss": 0.5527,
+      "step": 909
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.9497521719036,
+      "learning_rate": 9.54832368240027e-06,
+      "loss": 0.4326,
+      "step": 910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.51362496368161,
+      "learning_rate": 9.54712251830434e-06,
+      "loss": 0.4834,
+      "step": 911
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.840346790600776,
+      "learning_rate": 9.545919834930257e-06,
+      "loss": 0.4854,
+      "step": 912
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.48508364031533,
+      "learning_rate": 9.54471563267986e-06,
+      "loss": 0.6523,
+      "step": 913
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.107481431790621,
+      "learning_rate": 9.543509911955497e-06,
+      "loss": 0.5186,
+      "step": 914
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.554933322462405,
+      "learning_rate": 9.542302673160021e-06,
+      "loss": 0.5552,
+      "step": 915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.359795062769347,
+      "learning_rate": 9.541093916696793e-06,
+      "loss": 0.6514,
+      "step": 916
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.933210596647848,
+      "learning_rate": 9.539883642969681e-06,
+      "loss": 0.4824,
+      "step": 917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.030646602797862,
+      "learning_rate": 9.53867185238306e-06,
+      "loss": 0.5791,
+      "step": 918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.531996572790305,
+      "learning_rate": 9.53745854534181e-06,
+      "loss": 0.542,
+      "step": 919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.737722994826893,
+      "learning_rate": 9.536243722251321e-06,
+      "loss": 0.6357,
+      "step": 920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.176392594106526,
+      "learning_rate": 9.53502738351749e-06,
+      "loss": 0.5479,
+      "step": 921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.791505085435352,
+      "learning_rate": 9.533809529546716e-06,
+      "loss": 0.5146,
+      "step": 922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.194713334314685,
+      "learning_rate": 9.532590160745906e-06,
+      "loss": 0.5542,
+      "step": 923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.304141717636043,
+      "learning_rate": 9.531369277522475e-06,
+      "loss": 0.4966,
+      "step": 924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.781951810922592,
+      "learning_rate": 9.530146880284343e-06,
+      "loss": 0.5664,
+      "step": 925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.045579974558065,
+      "learning_rate": 9.528922969439935e-06,
+      "loss": 0.5073,
+      "step": 926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.68357233374754,
+      "learning_rate": 9.527697545398183e-06,
+      "loss": 0.5542,
+      "step": 927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.122021239823766,
+      "learning_rate": 9.526470608568521e-06,
+      "loss": 0.5415,
+      "step": 928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.454918266617198,
+      "learning_rate": 9.525242159360897e-06,
+      "loss": 0.5361,
+      "step": 929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.493825899265097,
+      "learning_rate": 9.524012198185755e-06,
+      "loss": 0.7109,
+      "step": 930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.636191715860241,
+      "learning_rate": 9.522780725454048e-06,
+      "loss": 0.5635,
+      "step": 931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.967666658359168,
+      "learning_rate": 9.521547741577232e-06,
+      "loss": 0.5757,
+      "step": 932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.728302127565175,
+      "learning_rate": 9.520313246967277e-06,
+      "loss": 0.6177,
+      "step": 933
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.518191584124642,
+      "learning_rate": 9.519077242036643e-06,
+      "loss": 0.6094,
+      "step": 934
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.997331318486447,
+      "learning_rate": 9.517839727198306e-06,
+      "loss": 0.5352,
+      "step": 935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.08940793882954,
+      "learning_rate": 9.516600702865742e-06,
+      "loss": 0.5396,
+      "step": 936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.3323126968670005,
+      "learning_rate": 9.51536016945293e-06,
+      "loss": 0.5122,
+      "step": 937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.266058320003383,
+      "learning_rate": 9.514118127374358e-06,
+      "loss": 0.5044,
+      "step": 938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.08730151307726,
+      "learning_rate": 9.512874577045016e-06,
+      "loss": 0.5518,
+      "step": 939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.676561438718075,
+      "learning_rate": 9.511629518880394e-06,
+      "loss": 0.5298,
+      "step": 940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.411348317275475,
+      "learning_rate": 9.510382953296492e-06,
+      "loss": 0.4536,
+      "step": 941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 44.62122780743628,
+      "learning_rate": 9.50913488070981e-06,
+      "loss": 0.6172,
+      "step": 942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 228.46722688523354,
+      "learning_rate": 9.50788530153735e-06,
+      "loss": 0.8828,
+      "step": 943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 109.64473438367058,
+      "learning_rate": 9.506634216196621e-06,
+      "loss": 0.9453,
+      "step": 944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 78.65828191559729,
+      "learning_rate": 9.505381625105636e-06,
+      "loss": 0.9453,
+      "step": 945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 51.832204433461676,
+      "learning_rate": 9.504127528682907e-06,
+      "loss": 0.6997,
+      "step": 946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 31.13063318378493,
+      "learning_rate": 9.502871927347452e-06,
+      "loss": 0.6172,
+      "step": 947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.291137372507908,
+      "learning_rate": 9.501614821518789e-06,
+      "loss": 0.5537,
+      "step": 948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.46365835435606,
+      "learning_rate": 9.500356211616941e-06,
+      "loss": 0.6475,
+      "step": 949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 32.29740862656474,
+      "learning_rate": 9.499096098062435e-06,
+      "loss": 0.6709,
+      "step": 950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 63.097784255629676,
+      "learning_rate": 9.497834481276293e-06,
+      "loss": 0.6182,
+      "step": 951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.902799208094248,
+      "learning_rate": 9.496571361680052e-06,
+      "loss": 0.5908,
+      "step": 952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.627462349415925,
+      "learning_rate": 9.495306739695738e-06,
+      "loss": 0.6055,
+      "step": 953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.838239735724807,
+      "learning_rate": 9.494040615745887e-06,
+      "loss": 0.5063,
+      "step": 954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.579178855565374,
+      "learning_rate": 9.492772990253535e-06,
+      "loss": 0.6553,
+      "step": 955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.534070316127274,
+      "learning_rate": 9.49150386364222e-06,
+      "loss": 0.5283,
+      "step": 956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 18.81207673614865,
+      "learning_rate": 9.490233236335977e-06,
+      "loss": 0.5747,
+      "step": 957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 35.2501349954626,
+      "learning_rate": 9.488961108759349e-06,
+      "loss": 0.5806,
+      "step": 958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.779299324475687,
+      "learning_rate": 9.487687481337377e-06,
+      "loss": 0.6221,
+      "step": 959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.734399861098689,
+      "learning_rate": 9.486412354495605e-06,
+      "loss": 0.563,
+      "step": 960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.688681103139775,
+      "learning_rate": 9.485135728660073e-06,
+      "loss": 0.5747,
+      "step": 961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.39443389421036,
+      "learning_rate": 9.48385760425733e-06,
+      "loss": 0.5088,
+      "step": 962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.663199374468476,
+      "learning_rate": 9.482577981714417e-06,
+      "loss": 0.6211,
+      "step": 963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.818151185370693,
+      "learning_rate": 9.481296861458881e-06,
+      "loss": 0.5361,
+      "step": 964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 21.38553722751277,
+      "learning_rate": 9.480014243918769e-06,
+      "loss": 0.6396,
+      "step": 965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.402379274239996,
+      "learning_rate": 9.478730129522627e-06,
+      "loss": 0.5635,
+      "step": 966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.858863875511425,
+      "learning_rate": 9.477444518699501e-06,
+      "loss": 0.6465,
+      "step": 967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.572804809224767,
+      "learning_rate": 9.476157411878937e-06,
+      "loss": 0.5537,
+      "step": 968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.627062162450095,
+      "learning_rate": 9.474868809490984e-06,
+      "loss": 0.6309,
+      "step": 969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.284249577261523,
+      "learning_rate": 9.473578711966185e-06,
+      "loss": 0.6641,
+      "step": 970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.635613200387112,
+      "learning_rate": 9.472287119735588e-06,
+      "loss": 0.5928,
+      "step": 971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.198639315955063,
+      "learning_rate": 9.470994033230735e-06,
+      "loss": 0.521,
+      "step": 972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.11429585253634,
+      "learning_rate": 9.469699452883672e-06,
+      "loss": 0.623,
+      "step": 973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 58.34165469496508,
+      "learning_rate": 9.468403379126943e-06,
+      "loss": 0.6406,
+      "step": 974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 25.195532099505844,
+      "learning_rate": 9.46710581239359e-06,
+      "loss": 0.6553,
+      "step": 975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.180951301910945,
+      "learning_rate": 9.465806753117153e-06,
+      "loss": 0.6494,
+      "step": 976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.047197342311845,
+      "learning_rate": 9.464506201731674e-06,
+      "loss": 0.6182,
+      "step": 977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.207255751064135,
+      "learning_rate": 9.463204158671687e-06,
+      "loss": 0.561,
+      "step": 978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.684448206775564,
+      "learning_rate": 9.461900624372233e-06,
+      "loss": 0.6177,
+      "step": 979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 20.50000538321963,
+      "learning_rate": 9.460595599268848e-06,
+      "loss": 0.6416,
+      "step": 980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.660593164014767,
+      "learning_rate": 9.45928908379756e-06,
+      "loss": 0.5166,
+      "step": 981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.206357340067083,
+      "learning_rate": 9.457981078394905e-06,
+      "loss": 0.6357,
+      "step": 982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.40715091937378,
+      "learning_rate": 9.45667158349791e-06,
+      "loss": 0.5601,
+      "step": 983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 36.68503954323254,
+      "learning_rate": 9.4553605995441e-06,
+      "loss": 0.6162,
+      "step": 984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.636828128912919,
+      "learning_rate": 9.4540481269715e-06,
+      "loss": 0.6436,
+      "step": 985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.970530099676225,
+      "learning_rate": 9.452734166218635e-06,
+      "loss": 0.6279,
+      "step": 986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.10424986097931,
+      "learning_rate": 9.451418717724518e-06,
+      "loss": 0.5771,
+      "step": 987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 19.04133879032038,
+      "learning_rate": 9.45010178192867e-06,
+      "loss": 0.5747,
+      "step": 988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.337405957274404,
+      "learning_rate": 9.448783359271102e-06,
+      "loss": 0.6104,
+      "step": 989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.704100149151028,
+      "learning_rate": 9.44746345019232e-06,
+      "loss": 0.5547,
+      "step": 990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.243738749349486,
+      "learning_rate": 9.446142055133333e-06,
+      "loss": 0.7139,
+      "step": 991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.327598910169833,
+      "learning_rate": 9.444819174535647e-06,
+      "loss": 0.7104,
+      "step": 992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.328880835372676,
+      "learning_rate": 9.443494808841255e-06,
+      "loss": 0.5815,
+      "step": 993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 20.53709813505537,
+      "learning_rate": 9.442168958492657e-06,
+      "loss": 0.6484,
+      "step": 994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.46766757628474,
+      "learning_rate": 9.44084162393284e-06,
+      "loss": 0.5952,
+      "step": 995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.838152911509205,
+      "learning_rate": 9.439512805605294e-06,
+      "loss": 0.5791,
+      "step": 996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.976704798054344,
+      "learning_rate": 9.438182503954002e-06,
+      "loss": 0.6211,
+      "step": 997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.574758990726995,
+      "learning_rate": 9.43685071942344e-06,
+      "loss": 0.5728,
+      "step": 998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.175433896531516,
+      "learning_rate": 9.435517452458584e-06,
+      "loss": 0.5,
+      "step": 999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.327701848585804,
+      "learning_rate": 9.434182703504904e-06,
+      "loss": 0.6045,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "eval_avg_AUC": 0.6339471196716615,
+      "eval_avg_Accuracy": 0.6359830901856764,
+      "eval_avg_Accuracy-right": 0.9985000652145559,
+      "eval_avg_Accuracy-wrong": 0.003866272458494428,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.5146999788919214,
+      "eval_last_AUC": 0.6325870765603462,
+      "eval_last_Accuracy": 0.6397546419098143,
+      "eval_last_Accuracy-right": 0.9924351115168906,
+      "eval_last_Accuracy-wrong": 0.024789629292699566,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.5155307886197232,
+      "eval_max_AUC": 0.6025879470270927,
+      "eval_max_Accuracy": 0.6355271883289124,
+      "eval_max_Accuracy-right": 0.9996087126646668,
+      "eval_max_Accuracy-wrong": 0.0006822833750284285,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5068662445848651,
+      "eval_min_AUC": 0.6212461473800316,
+      "eval_min_Accuracy": 0.640376326259947,
+      "eval_min_Accuracy-right": 0.9918481805138907,
+      "eval_min_Accuracy-wrong": 0.027518762792813282,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.5093913468479994,
+      "eval_prod_AUC": 0.5978007803736609,
+      "eval_prod_Accuracy": 0.38884283819628646,
+      "eval_prod_Accuracy-right": 0.04995434981087779,
+      "eval_prod_Accuracy-wrong": 0.9797589265408233,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.5021156954591185,
+      "eval_runtime": 249.2562,
+      "eval_samples_per_second": 96.8,
+      "eval_steps_per_second": 3.025,
+      "eval_sum_AUC": 0.4675143189480929,
+      "eval_sum_Accuracy": 0.635651525198939,
+      "eval_sum_Accuracy-right": 1.0,
+      "eval_sum_Accuracy-wrong": 0.0003411416875142142,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.5161108860078178,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.82364706060358,
+      "learning_rate": 9.432846473008363e-06,
+      "loss": 0.5806,
+      "step": 1001
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 27.37121267396446,
+      "learning_rate": 9.431508761415422e-06,
+      "loss": 0.6816,
+      "step": 1002
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.853525294894828,
+      "learning_rate": 9.430169569173034e-06,
+      "loss": 0.583,
+      "step": 1003
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.775963932541249,
+      "learning_rate": 9.428828896728645e-06,
+      "loss": 0.5537,
+      "step": 1004
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.72952999413491,
+      "learning_rate": 9.427486744530205e-06,
+      "loss": 0.5576,
+      "step": 1005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.721816684636078,
+      "learning_rate": 9.426143113026147e-06,
+      "loss": 0.6729,
+      "step": 1006
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.544564691174461,
+      "learning_rate": 9.424798002665405e-06,
+      "loss": 0.666,
+      "step": 1007
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.719641221253966,
+      "learning_rate": 9.423451413897406e-06,
+      "loss": 0.6758,
+      "step": 1008
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.0766374680692,
+      "learning_rate": 9.42210334717207e-06,
+      "loss": 0.6533,
+      "step": 1009
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.964130971436024,
+      "learning_rate": 9.42075380293981e-06,
+      "loss": 0.6338,
+      "step": 1010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.093738265546637,
+      "learning_rate": 9.419402781651537e-06,
+      "loss": 0.5977,
+      "step": 1011
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.29527691970946,
+      "learning_rate": 9.418050283758647e-06,
+      "loss": 0.6133,
+      "step": 1012
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.094752942831983,
+      "learning_rate": 9.416696309713038e-06,
+      "loss": 0.6084,
+      "step": 1013
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.37532956112354,
+      "learning_rate": 9.415340859967099e-06,
+      "loss": 0.6182,
+      "step": 1014
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.71031567566946,
+      "learning_rate": 9.413983934973709e-06,
+      "loss": 0.5996,
+      "step": 1015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.269730904625384,
+      "learning_rate": 9.412625535186242e-06,
+      "loss": 0.5771,
+      "step": 1016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 28.656034482685637,
+      "learning_rate": 9.411265661058565e-06,
+      "loss": 0.6465,
+      "step": 1017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 42.68689422622165,
+      "learning_rate": 9.409904313045038e-06,
+      "loss": 0.6816,
+      "step": 1018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.650685008374277,
+      "learning_rate": 9.408541491600511e-06,
+      "loss": 0.5723,
+      "step": 1019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 18.730637303407523,
+      "learning_rate": 9.407177197180328e-06,
+      "loss": 0.5962,
+      "step": 1020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.648761689749307,
+      "learning_rate": 9.405811430240329e-06,
+      "loss": 0.6924,
+      "step": 1021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.798406394080896,
+      "learning_rate": 9.404444191236837e-06,
+      "loss": 0.5791,
+      "step": 1022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.329130838909766,
+      "learning_rate": 9.403075480626674e-06,
+      "loss": 0.5547,
+      "step": 1023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.655850010883231,
+      "learning_rate": 9.401705298867151e-06,
+      "loss": 0.5957,
+      "step": 1024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.47268243968311,
+      "learning_rate": 9.400333646416073e-06,
+      "loss": 0.6396,
+      "step": 1025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 26.170896114421865,
+      "learning_rate": 9.398960523731735e-06,
+      "loss": 0.6133,
+      "step": 1026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.953028885048408,
+      "learning_rate": 9.397585931272919e-06,
+      "loss": 0.5537,
+      "step": 1027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.88531583467331,
+      "learning_rate": 9.396209869498905e-06,
+      "loss": 0.5498,
+      "step": 1028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.586608224218812,
+      "learning_rate": 9.39483233886946e-06,
+      "loss": 0.4663,
+      "step": 1029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.459922262332075,
+      "learning_rate": 9.393453339844842e-06,
+      "loss": 0.5684,
+      "step": 1030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.9938727234055,
+      "learning_rate": 9.392072872885802e-06,
+      "loss": 0.5859,
+      "step": 1031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.95231020980077,
+      "learning_rate": 9.39069093845358e-06,
+      "loss": 0.5669,
+      "step": 1032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.451010224992702,
+      "learning_rate": 9.389307537009902e-06,
+      "loss": 0.6187,
+      "step": 1033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.835554267305948,
+      "learning_rate": 9.387922669016992e-06,
+      "loss": 0.6094,
+      "step": 1034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.225942002568258,
+      "learning_rate": 9.386536334937557e-06,
+      "loss": 0.5293,
+      "step": 1035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.28571736427067,
+      "learning_rate": 9.385148535234799e-06,
+      "loss": 0.5771,
+      "step": 1036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.11918969171961,
+      "learning_rate": 9.383759270372408e-06,
+      "loss": 0.6396,
+      "step": 1037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.987795594520582,
+      "learning_rate": 9.382368540814563e-06,
+      "loss": 0.5493,
+      "step": 1038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.332762133791691,
+      "learning_rate": 9.380976347025932e-06,
+      "loss": 0.5781,
+      "step": 1039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.833649187371895,
+      "learning_rate": 9.379582689471671e-06,
+      "loss": 0.5078,
+      "step": 1040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 29.85600810006891,
+      "learning_rate": 9.378187568617431e-06,
+      "loss": 0.668,
+      "step": 1041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.86118062820929,
+      "learning_rate": 9.376790984929348e-06,
+      "loss": 0.5869,
+      "step": 1042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 28.051207820211953,
+      "learning_rate": 9.37539293887404e-06,
+      "loss": 0.6113,
+      "step": 1043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 32.11996073913673,
+      "learning_rate": 9.373993430918626e-06,
+      "loss": 0.5068,
+      "step": 1044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 17.79452829191096,
+      "learning_rate": 9.372592461530708e-06,
+      "loss": 0.5225,
+      "step": 1045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.681808797361166,
+      "learning_rate": 9.371190031178372e-06,
+      "loss": 0.582,
+      "step": 1046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 25.230590380011453,
+      "learning_rate": 9.369786140330198e-06,
+      "loss": 0.5239,
+      "step": 1047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 42.558740499704996,
+      "learning_rate": 9.368380789455251e-06,
+      "loss": 0.519,
+      "step": 1048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.297934374384019,
+      "learning_rate": 9.36697397902309e-06,
+      "loss": 0.5977,
+      "step": 1049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.086643595178904,
+      "learning_rate": 9.365565709503748e-06,
+      "loss": 0.5469,
+      "step": 1050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.056732327995272,
+      "learning_rate": 9.364155981367761e-06,
+      "loss": 0.4849,
+      "step": 1051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.381781397718909,
+      "learning_rate": 9.36274479508614e-06,
+      "loss": 0.519,
+      "step": 1052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.39451781815997,
+      "learning_rate": 9.361332151130396e-06,
+      "loss": 0.5249,
+      "step": 1053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.967285956708139,
+      "learning_rate": 9.359918049972512e-06,
+      "loss": 0.5493,
+      "step": 1054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 24.743475345880128,
+      "learning_rate": 9.358502492084969e-06,
+      "loss": 0.5654,
+      "step": 1055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.646634416378777,
+      "learning_rate": 9.35708547794073e-06,
+      "loss": 0.5703,
+      "step": 1056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.719559112795636,
+      "learning_rate": 9.355667008013249e-06,
+      "loss": 0.5825,
+      "step": 1057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.823118073564967,
+      "learning_rate": 9.354247082776459e-06,
+      "loss": 0.5981,
+      "step": 1058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.487103900742174,
+      "learning_rate": 9.352825702704784e-06,
+      "loss": 0.5532,
+      "step": 1059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.077491438600466,
+      "learning_rate": 9.351402868273136e-06,
+      "loss": 0.5513,
+      "step": 1060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.897696554119129,
+      "learning_rate": 9.349978579956908e-06,
+      "loss": 0.5938,
+      "step": 1061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.206740424557369,
+      "learning_rate": 9.348552838231983e-06,
+      "loss": 0.5547,
+      "step": 1062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.745893827810162,
+      "learning_rate": 9.347125643574726e-06,
+      "loss": 0.5117,
+      "step": 1063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.609051578212978,
+      "learning_rate": 9.345696996461992e-06,
+      "loss": 0.5918,
+      "step": 1064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.368231522386932,
+      "learning_rate": 9.344266897371114e-06,
+      "loss": 0.519,
+      "step": 1065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.399140300952528,
+      "learning_rate": 9.34283534677992e-06,
+      "loss": 0.5483,
+      "step": 1066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.568335263277644,
+      "learning_rate": 9.341402345166714e-06,
+      "loss": 0.6382,
+      "step": 1067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.28361345060809,
+      "learning_rate": 9.33996789301029e-06,
+      "loss": 0.626,
+      "step": 1068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.034588447871568,
+      "learning_rate": 9.338531990789926e-06,
+      "loss": 0.4995,
+      "step": 1069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.260306318615566,
+      "learning_rate": 9.33709463898538e-06,
+      "loss": 0.5732,
+      "step": 1070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.271709992114081,
+      "learning_rate": 9.335655838076902e-06,
+      "loss": 0.4492,
+      "step": 1071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.56028285693857,
+      "learning_rate": 9.33421558854522e-06,
+      "loss": 0.5234,
+      "step": 1072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.782154263459084,
+      "learning_rate": 9.332773890871548e-06,
+      "loss": 0.582,
+      "step": 1073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.094570074636316,
+      "learning_rate": 9.331330745537586e-06,
+      "loss": 0.5386,
+      "step": 1074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 20.10166588549232,
+      "learning_rate": 9.329886153025513e-06,
+      "loss": 0.5449,
+      "step": 1075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.192408971954997,
+      "learning_rate": 9.328440113817995e-06,
+      "loss": 0.5259,
+      "step": 1076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.208683818696056,
+      "learning_rate": 9.326992628398182e-06,
+      "loss": 0.5278,
+      "step": 1077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.34674029850454,
+      "learning_rate": 9.325543697249706e-06,
+      "loss": 0.6328,
+      "step": 1078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.912080486024484,
+      "learning_rate": 9.324093320856679e-06,
+      "loss": 0.5576,
+      "step": 1079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.692872107879284,
+      "learning_rate": 9.3226414997037e-06,
+      "loss": 0.5107,
+      "step": 1080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.965283753527137,
+      "learning_rate": 9.32118823427585e-06,
+      "loss": 0.6133,
+      "step": 1081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.088812339526163,
+      "learning_rate": 9.319733525058694e-06,
+      "loss": 0.5312,
+      "step": 1082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.69717168681488,
+      "learning_rate": 9.318277372538274e-06,
+      "loss": 0.5225,
+      "step": 1083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.757166076166586,
+      "learning_rate": 9.316819777201119e-06,
+      "loss": 0.5303,
+      "step": 1084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.268054362533457,
+      "learning_rate": 9.315360739534235e-06,
+      "loss": 0.6626,
+      "step": 1085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.75302943618852,
+      "learning_rate": 9.313900260025121e-06,
+      "loss": 0.4951,
+      "step": 1086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.058324474577528,
+      "learning_rate": 9.312438339161746e-06,
+      "loss": 0.5337,
+      "step": 1087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.186156422831758,
+      "learning_rate": 9.310974977432565e-06,
+      "loss": 0.5122,
+      "step": 1088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.810759557849348,
+      "learning_rate": 9.309510175326515e-06,
+      "loss": 0.5986,
+      "step": 1089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.352639000106736,
+      "learning_rate": 9.308043933333012e-06,
+      "loss": 0.54,
+      "step": 1090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.18764295150179,
+      "learning_rate": 9.306576251941957e-06,
+      "loss": 0.5889,
+      "step": 1091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.066681900530504,
+      "learning_rate": 9.305107131643729e-06,
+      "loss": 0.6221,
+      "step": 1092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.405377306247158,
+      "learning_rate": 9.303636572929188e-06,
+      "loss": 0.5723,
+      "step": 1093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 18.5355497270562,
+      "learning_rate": 9.302164576289674e-06,
+      "loss": 0.5957,
+      "step": 1094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.64764367778218,
+      "learning_rate": 9.30069114221701e-06,
+      "loss": 0.6143,
+      "step": 1095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.446851167336094,
+      "learning_rate": 9.299216271203498e-06,
+      "loss": 0.6006,
+      "step": 1096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.594638602419203,
+      "learning_rate": 9.297739963741918e-06,
+      "loss": 0.5596,
+      "step": 1097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.822830235951766,
+      "learning_rate": 9.296262220325535e-06,
+      "loss": 0.5557,
+      "step": 1098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.680693513430095,
+      "learning_rate": 9.294783041448088e-06,
+      "loss": 0.5522,
+      "step": 1099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.487159518778437,
+      "learning_rate": 9.293302427603796e-06,
+      "loss": 0.5518,
+      "step": 1100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.5217580901104935,
+      "learning_rate": 9.291820379287364e-06,
+      "loss": 0.5269,
+      "step": 1101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.594198953311631,
+      "learning_rate": 9.29033689699397e-06,
+      "loss": 0.5562,
+      "step": 1102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.991686017035859,
+      "learning_rate": 9.288851981219273e-06,
+      "loss": 0.5503,
+      "step": 1103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.804601642168482,
+      "learning_rate": 9.28736563245941e-06,
+      "loss": 0.5806,
+      "step": 1104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.916446670933475,
+      "learning_rate": 9.285877851210999e-06,
+      "loss": 0.5259,
+      "step": 1105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.312528006718804,
+      "learning_rate": 9.284388637971136e-06,
+      "loss": 0.563,
+      "step": 1106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.16927949883247,
+      "learning_rate": 9.282897993237392e-06,
+      "loss": 0.6025,
+      "step": 1107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.748674092022494,
+      "learning_rate": 9.281405917507824e-06,
+      "loss": 0.5723,
+      "step": 1108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.416556061970823,
+      "learning_rate": 9.279912411280958e-06,
+      "loss": 0.5889,
+      "step": 1109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.2599868177948,
+      "learning_rate": 9.278417475055803e-06,
+      "loss": 0.5654,
+      "step": 1110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.138906667061427,
+      "learning_rate": 9.276921109331845e-06,
+      "loss": 0.5127,
+      "step": 1111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.142722431903195,
+      "learning_rate": 9.275423314609049e-06,
+      "loss": 0.5771,
+      "step": 1112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.367984402900843,
+      "learning_rate": 9.273924091387855e-06,
+      "loss": 0.5132,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.460138534290378,
+      "learning_rate": 9.272423440169181e-06,
+      "loss": 0.5537,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.885508556319811,
+      "learning_rate": 9.270921361454424e-06,
+      "loss": 0.5698,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.642226342988765,
+      "learning_rate": 9.269417855745453e-06,
+      "loss": 0.5547,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.93459268058794,
+      "learning_rate": 9.267912923544621e-06,
+      "loss": 0.6357,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.61687681956424,
+      "learning_rate": 9.266406565354753e-06,
+      "loss": 0.5869,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.948478273276576,
+      "learning_rate": 9.26489878167915e-06,
+      "loss": 0.5884,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.224943637865685,
+      "learning_rate": 9.263389573021592e-06,
+      "loss": 0.5391,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.53281858669421,
+      "learning_rate": 9.261878939886332e-06,
+      "loss": 0.5391,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.571223439001684,
+      "learning_rate": 9.2603668827781e-06,
+      "loss": 0.5869,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.996971954221829,
+      "learning_rate": 9.258853402202106e-06,
+      "loss": 0.6143,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.562877307679958,
+      "learning_rate": 9.25733849866403e-06,
+      "loss": 0.54,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 25.681379456692,
+      "learning_rate": 9.255822172670028e-06,
+      "loss": 0.5957,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.240913002501355,
+      "learning_rate": 9.254304424726734e-06,
+      "loss": 0.48,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.171552813788881,
+      "learning_rate": 9.252785255341256e-06,
+      "loss": 0.5161,
+      "step": 1127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.921769965960761,
+      "learning_rate": 9.251264665021178e-06,
+      "loss": 0.5864,
+      "step": 1128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.1627582262793785,
+      "learning_rate": 9.249742654274554e-06,
+      "loss": 0.5283,
+      "step": 1129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.70067037372327,
+      "learning_rate": 9.24821922360992e-06,
+      "loss": 0.5396,
+      "step": 1130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 25.93057220073389,
+      "learning_rate": 9.246694373536277e-06,
+      "loss": 0.6123,
+      "step": 1131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.705923665031527,
+      "learning_rate": 9.245168104563112e-06,
+      "loss": 0.6152,
+      "step": 1132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.636045888986658,
+      "learning_rate": 9.243640417200376e-06,
+      "loss": 0.5352,
+      "step": 1133
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.505774334822952,
+      "learning_rate": 9.242111311958502e-06,
+      "loss": 0.521,
+      "step": 1134
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.567202760806081,
+      "learning_rate": 9.240580789348385e-06,
+      "loss": 0.6143,
+      "step": 1135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.503439518116549,
+      "learning_rate": 9.23904884988141e-06,
+      "loss": 0.5908,
+      "step": 1136
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.221857770793607,
+      "learning_rate": 9.237515494069417e-06,
+      "loss": 0.5098,
+      "step": 1137
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.552085086378824,
+      "learning_rate": 9.235980722424737e-06,
+      "loss": 0.5034,
+      "step": 1138
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.725639961202933,
+      "learning_rate": 9.234444535460161e-06,
+      "loss": 0.4692,
+      "step": 1139
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.41674940985237,
+      "learning_rate": 9.232906933688959e-06,
+      "loss": 0.5,
+      "step": 1140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.7960411684762825,
+      "learning_rate": 9.231367917624872e-06,
+      "loss": 0.5225,
+      "step": 1141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.017568946472704,
+      "learning_rate": 9.229827487782115e-06,
+      "loss": 0.5518,
+      "step": 1142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 31.05573818747122,
+      "learning_rate": 9.228285644675372e-06,
+      "loss": 0.7061,
+      "step": 1143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.083495042416361,
+      "learning_rate": 9.226742388819804e-06,
+      "loss": 0.5938,
+      "step": 1144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.823638652711353,
+      "learning_rate": 9.225197720731039e-06,
+      "loss": 0.606,
+      "step": 1145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.519289143758317,
+      "learning_rate": 9.223651640925181e-06,
+      "loss": 0.5732,
+      "step": 1146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.367894727037841,
+      "learning_rate": 9.222104149918804e-06,
+      "loss": 0.585,
+      "step": 1147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.655844816080055,
+      "learning_rate": 9.220555248228954e-06,
+      "loss": 0.5459,
+      "step": 1148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.43974033304336,
+      "learning_rate": 9.219004936373146e-06,
+      "loss": 0.6084,
+      "step": 1149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.847797651330511,
+      "learning_rate": 9.217453214869368e-06,
+      "loss": 0.5312,
+      "step": 1150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.437081911431463,
+      "learning_rate": 9.21590008423608e-06,
+      "loss": 0.5049,
+      "step": 1151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.470652434232925,
+      "learning_rate": 9.214345544992214e-06,
+      "loss": 0.5342,
+      "step": 1152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.034516507112599,
+      "learning_rate": 9.212789597657167e-06,
+      "loss": 0.5249,
+      "step": 1153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.99367571691055,
+      "learning_rate": 9.21123224275081e-06,
+      "loss": 0.6074,
+      "step": 1154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.864377559517447,
+      "learning_rate": 9.209673480793486e-06,
+      "loss": 0.5376,
+      "step": 1155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.346477449728334,
+      "learning_rate": 9.208113312306006e-06,
+      "loss": 0.478,
+      "step": 1156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.006532830891896,
+      "learning_rate": 9.206551737809653e-06,
+      "loss": 0.6025,
+      "step": 1157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.611614451718726,
+      "learning_rate": 9.204988757826173e-06,
+      "loss": 0.5278,
+      "step": 1158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.714149559119864,
+      "learning_rate": 9.203424372877791e-06,
+      "loss": 0.519,
+      "step": 1159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.767330554783726,
+      "learning_rate": 9.201858583487195e-06,
+      "loss": 0.5977,
+      "step": 1160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.727023867341504,
+      "learning_rate": 9.200291390177546e-06,
+      "loss": 0.4941,
+      "step": 1161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.08139287351435,
+      "learning_rate": 9.198722793472471e-06,
+      "loss": 0.4712,
+      "step": 1162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.361690233383793,
+      "learning_rate": 9.197152793896068e-06,
+      "loss": 0.5293,
+      "step": 1163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.659319120735523,
+      "learning_rate": 9.195581391972903e-06,
+      "loss": 0.6079,
+      "step": 1164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.574233768699525,
+      "learning_rate": 9.194008588228011e-06,
+      "loss": 0.5083,
+      "step": 1165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.875647206303185,
+      "learning_rate": 9.192434383186894e-06,
+      "loss": 0.5625,
+      "step": 1166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.59800269890822,
+      "learning_rate": 9.190858777375523e-06,
+      "loss": 0.5361,
+      "step": 1167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.150351861232854,
+      "learning_rate": 9.18928177132034e-06,
+      "loss": 0.5273,
+      "step": 1168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.189230279097433,
+      "learning_rate": 9.187703365548248e-06,
+      "loss": 0.5054,
+      "step": 1169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.844111824576274,
+      "learning_rate": 9.186123560586623e-06,
+      "loss": 0.7607,
+      "step": 1170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.745982102578823,
+      "learning_rate": 9.18454235696331e-06,
+      "loss": 0.5322,
+      "step": 1171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.782839753489112,
+      "learning_rate": 9.182959755206613e-06,
+      "loss": 0.5449,
+      "step": 1172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.418717825018716,
+      "learning_rate": 9.181375755845314e-06,
+      "loss": 0.5337,
+      "step": 1173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.352862983177861,
+      "learning_rate": 9.179790359408655e-06,
+      "loss": 0.5342,
+      "step": 1174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.998049401916107,
+      "learning_rate": 9.178203566426344e-06,
+      "loss": 0.5479,
+      "step": 1175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.420625471547234,
+      "learning_rate": 9.176615377428563e-06,
+      "loss": 0.5527,
+      "step": 1176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.578151600737712,
+      "learning_rate": 9.175025792945951e-06,
+      "loss": 0.5308,
+      "step": 1177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.490365222019324,
+      "learning_rate": 9.173434813509618e-06,
+      "loss": 0.4888,
+      "step": 1178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.223972990178867,
+      "learning_rate": 9.171842439651143e-06,
+      "loss": 0.5366,
+      "step": 1179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.921385835164377,
+      "learning_rate": 9.170248671902565e-06,
+      "loss": 0.4946,
+      "step": 1180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.07766786230304,
+      "learning_rate": 9.168653510796392e-06,
+      "loss": 0.7295,
+      "step": 1181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.593698266134565,
+      "learning_rate": 9.167056956865596e-06,
+      "loss": 0.5977,
+      "step": 1182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.775199062594883,
+      "learning_rate": 9.165459010643618e-06,
+      "loss": 0.5488,
+      "step": 1183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.663762375167531,
+      "learning_rate": 9.16385967266436e-06,
+      "loss": 0.5054,
+      "step": 1184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.215806260743865,
+      "learning_rate": 9.16225894346219e-06,
+      "loss": 0.5186,
+      "step": 1185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.282697942390582,
+      "learning_rate": 9.160656823571942e-06,
+      "loss": 0.5898,
+      "step": 1186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.51121958562781,
+      "learning_rate": 9.159053313528913e-06,
+      "loss": 0.5605,
+      "step": 1187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.268837485274915,
+      "learning_rate": 9.15744841386887e-06,
+      "loss": 0.5498,
+      "step": 1188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.072700613697913,
+      "learning_rate": 9.155842125128033e-06,
+      "loss": 0.5215,
+      "step": 1189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.958687423245578,
+      "learning_rate": 9.154234447843098e-06,
+      "loss": 0.5063,
+      "step": 1190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.237034674269332,
+      "learning_rate": 9.152625382551217e-06,
+      "loss": 0.4717,
+      "step": 1191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 20.153622241744905,
+      "learning_rate": 9.15101492979001e-06,
+      "loss": 0.5029,
+      "step": 1192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.47771817765446,
+      "learning_rate": 9.149403090097557e-06,
+      "loss": 0.5664,
+      "step": 1193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.451169764840035,
+      "learning_rate": 9.147789864012408e-06,
+      "loss": 0.6226,
+      "step": 1194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.500389478462552,
+      "learning_rate": 9.146175252073568e-06,
+      "loss": 0.5518,
+      "step": 1195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.352519920650707,
+      "learning_rate": 9.144559254820511e-06,
+      "loss": 0.4795,
+      "step": 1196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.579403203990102,
+      "learning_rate": 9.14294187279317e-06,
+      "loss": 0.5386,
+      "step": 1197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.402304104093776,
+      "learning_rate": 9.141323106531943e-06,
+      "loss": 0.5537,
+      "step": 1198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.906202099251876,
+      "learning_rate": 9.139702956577693e-06,
+      "loss": 0.5342,
+      "step": 1199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.994868615336212,
+      "learning_rate": 9.138081423471736e-06,
+      "loss": 0.5054,
+      "step": 1200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.708925086407962,
+      "learning_rate": 9.136458507755862e-06,
+      "loss": 0.5317,
+      "step": 1201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 22.74881444619516,
+      "learning_rate": 9.134834209972314e-06,
+      "loss": 0.5034,
+      "step": 1202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.191911807224606,
+      "learning_rate": 9.133208530663801e-06,
+      "loss": 0.4849,
+      "step": 1203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.95443208118561,
+      "learning_rate": 9.131581470373495e-06,
+      "loss": 0.5337,
+      "step": 1204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.525395622875937,
+      "learning_rate": 9.129953029645022e-06,
+      "loss": 0.4604,
+      "step": 1205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.79694996087205,
+      "learning_rate": 9.128323209022478e-06,
+      "loss": 0.5122,
+      "step": 1206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.755997307705414,
+      "learning_rate": 9.126692009050415e-06,
+      "loss": 0.4463,
+      "step": 1207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.80591783623869,
+      "learning_rate": 9.125059430273848e-06,
+      "loss": 0.5796,
+      "step": 1208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.852441160925146,
+      "learning_rate": 9.123425473238253e-06,
+      "loss": 0.5015,
+      "step": 1209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.625338450562605,
+      "learning_rate": 9.121790138489564e-06,
+      "loss": 0.5117,
+      "step": 1210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.080326610727653,
+      "learning_rate": 9.120153426574177e-06,
+      "loss": 0.5244,
+      "step": 1211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.88918740584613,
+      "learning_rate": 9.118515338038947e-06,
+      "loss": 0.48,
+      "step": 1212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.68959162364965,
+      "learning_rate": 9.11687587343119e-06,
+      "loss": 0.5259,
+      "step": 1213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.726892232656102,
+      "learning_rate": 9.115235033298682e-06,
+      "loss": 0.5601,
+      "step": 1214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.708377102841364,
+      "learning_rate": 9.113592818189661e-06,
+      "loss": 0.5708,
+      "step": 1215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.856805732469365,
+      "learning_rate": 9.111949228652816e-06,
+      "loss": 0.5498,
+      "step": 1216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.19921968394708,
+      "learning_rate": 9.110304265237304e-06,
+      "loss": 0.5098,
+      "step": 1217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.873996190520849,
+      "learning_rate": 9.10865792849274e-06,
+      "loss": 0.4893,
+      "step": 1218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.856928395460525,
+      "learning_rate": 9.107010218969191e-06,
+      "loss": 0.6113,
+      "step": 1219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.008354271851358,
+      "learning_rate": 9.10536113721719e-06,
+      "loss": 0.5874,
+      "step": 1220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.27467515244621,
+      "learning_rate": 9.103710683787728e-06,
+      "loss": 0.4824,
+      "step": 1221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.838954506974407,
+      "learning_rate": 9.102058859232247e-06,
+      "loss": 0.5596,
+      "step": 1222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.3653312710855285,
+      "learning_rate": 9.100405664102656e-06,
+      "loss": 0.5063,
+      "step": 1223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.92988348312749,
+      "learning_rate": 9.098751098951317e-06,
+      "loss": 0.6328,
+      "step": 1224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.868600191790884,
+      "learning_rate": 9.09709516433105e-06,
+      "loss": 0.5679,
+      "step": 1225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.654596470733907,
+      "learning_rate": 9.095437860795138e-06,
+      "loss": 0.5215,
+      "step": 1226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.054939394702243,
+      "learning_rate": 9.09377918889731e-06,
+      "loss": 0.5317,
+      "step": 1227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.809295897500469,
+      "learning_rate": 9.092119149191765e-06,
+      "loss": 0.5142,
+      "step": 1228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.752863951527627,
+      "learning_rate": 9.090457742233152e-06,
+      "loss": 0.4692,
+      "step": 1229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.080450373075989,
+      "learning_rate": 9.088794968576575e-06,
+      "loss": 0.5771,
+      "step": 1230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.273931732814395,
+      "learning_rate": 9.087130828777598e-06,
+      "loss": 0.5811,
+      "step": 1231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.801947879251854,
+      "learning_rate": 9.085465323392243e-06,
+      "loss": 0.5923,
+      "step": 1232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.904080105524493,
+      "learning_rate": 9.083798452976988e-06,
+      "loss": 0.584,
+      "step": 1233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.329611523622987,
+      "learning_rate": 9.082130218088762e-06,
+      "loss": 0.5698,
+      "step": 1234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.2037432427127,
+      "learning_rate": 9.080460619284954e-06,
+      "loss": 0.5283,
+      "step": 1235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.783777108155,
+      "learning_rate": 9.07878965712341e-06,
+      "loss": 0.6377,
+      "step": 1236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.5497103145234625,
+      "learning_rate": 9.077117332162427e-06,
+      "loss": 0.5791,
+      "step": 1237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.356999836172582,
+      "learning_rate": 9.075443644960761e-06,
+      "loss": 0.5063,
+      "step": 1238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.158396146128032,
+      "learning_rate": 9.07376859607762e-06,
+      "loss": 0.5376,
+      "step": 1239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.510823350991721,
+      "learning_rate": 9.072092186072675e-06,
+      "loss": 0.6494,
+      "step": 1240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.757079666813102,
+      "learning_rate": 9.070414415506038e-06,
+      "loss": 0.6143,
+      "step": 1241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.23137401597286,
+      "learning_rate": 9.068735284938288e-06,
+      "loss": 0.4785,
+      "step": 1242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.909470075068411,
+      "learning_rate": 9.067054794930452e-06,
+      "loss": 0.4731,
+      "step": 1243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.924869803084112,
+      "learning_rate": 9.065372946044014e-06,
+      "loss": 0.606,
+      "step": 1244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.833436716234493,
+      "learning_rate": 9.063689738840911e-06,
+      "loss": 0.5996,
+      "step": 1245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.76560707751374,
+      "learning_rate": 9.06200517388353e-06,
+      "loss": 0.6333,
+      "step": 1246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.385881684068905,
+      "learning_rate": 9.060319251734723e-06,
+      "loss": 0.5283,
+      "step": 1247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.674391394289866,
+      "learning_rate": 9.058631972957783e-06,
+      "loss": 0.5977,
+      "step": 1248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.351792465763545,
+      "learning_rate": 9.056943338116461e-06,
+      "loss": 0.5024,
+      "step": 1249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.53319179062876,
+      "learning_rate": 9.055253347774961e-06,
+      "loss": 0.5386,
+      "step": 1250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.528600979838268,
+      "learning_rate": 9.053562002497943e-06,
+      "loss": 0.5737,
+      "step": 1251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.959305596996799,
+      "learning_rate": 9.051869302850515e-06,
+      "loss": 0.4683,
+      "step": 1252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.58161660458802,
+      "learning_rate": 9.05017524939824e-06,
+      "loss": 0.5527,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.243428609193199,
+      "learning_rate": 9.048479842707132e-06,
+      "loss": 0.5894,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.236285055544797,
+      "learning_rate": 9.046783083343657e-06,
+      "loss": 0.4614,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.591178916038821,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.5181,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.84688810132272,
+      "learning_rate": 9.043385508867741e-06,
+      "loss": 0.4771,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.615688659064489,
+      "learning_rate": 9.041684694890492e-06,
+      "loss": 0.4736,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.846075910748137,
+      "learning_rate": 9.03998253051126e-06,
+      "loss": 0.6909,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.612951792166978,
+      "learning_rate": 9.038279016298773e-06,
+      "loss": 0.4756,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.350725640206349,
+      "learning_rate": 9.036574152822206e-06,
+      "loss": 0.5273,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.878946181687478,
+      "learning_rate": 9.034867940651186e-06,
+      "loss": 0.5635,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.333503190009848,
+      "learning_rate": 9.033160380355789e-06,
+      "loss": 0.4512,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.830536532606617,
+      "learning_rate": 9.031451472506544e-06,
+      "loss": 0.5039,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.00597678487424,
+      "learning_rate": 9.029741217674428e-06,
+      "loss": 0.5776,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.365161480301378,
+      "learning_rate": 9.02802961643087e-06,
+      "loss": 0.5635,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.387833721542348,
+      "learning_rate": 9.026316669347747e-06,
+      "loss": 0.5137,
+      "step": 1267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.41425867798381,
+      "learning_rate": 9.024602376997387e-06,
+      "loss": 0.5454,
+      "step": 1268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.311986546252852,
+      "learning_rate": 9.022886739952565e-06,
+      "loss": 0.4316,
+      "step": 1269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.496984433352765,
+      "learning_rate": 9.02116975878651e-06,
+      "loss": 0.5542,
+      "step": 1270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.783629489842369,
+      "learning_rate": 9.019451434072894e-06,
+      "loss": 0.5654,
+      "step": 1271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.656169059630007,
+      "learning_rate": 9.017731766385844e-06,
+      "loss": 0.5093,
+      "step": 1272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.338346102975163,
+      "learning_rate": 9.016010756299934e-06,
+      "loss": 0.4619,
+      "step": 1273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.6546009929447,
+      "learning_rate": 9.014288404390182e-06,
+      "loss": 0.5293,
+      "step": 1274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.49900328718962,
+      "learning_rate": 9.012564711232059e-06,
+      "loss": 0.4985,
+      "step": 1275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.799568957162913,
+      "learning_rate": 9.010839677401484e-06,
+      "loss": 0.5776,
+      "step": 1276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.442440495388319,
+      "learning_rate": 9.009113303474822e-06,
+      "loss": 0.4966,
+      "step": 1277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.780144748137946,
+      "learning_rate": 9.007385590028887e-06,
+      "loss": 0.4404,
+      "step": 1278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.298026166100321,
+      "learning_rate": 9.005656537640942e-06,
+      "loss": 0.5498,
+      "step": 1279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.495108785802219,
+      "learning_rate": 9.003926146888691e-06,
+      "loss": 0.4897,
+      "step": 1280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.611746701232708,
+      "learning_rate": 9.002194418350291e-06,
+      "loss": 0.6392,
+      "step": 1281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.718656093085405,
+      "learning_rate": 9.000461352604349e-06,
+      "loss": 0.5972,
+      "step": 1282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.384919907374725,
+      "learning_rate": 8.99872695022991e-06,
+      "loss": 0.5142,
+      "step": 1283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.111422961254767,
+      "learning_rate": 8.996991211806471e-06,
+      "loss": 0.5176,
+      "step": 1284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.61897079404457,
+      "learning_rate": 8.995254137913977e-06,
+      "loss": 0.5859,
+      "step": 1285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.626370903659802,
+      "learning_rate": 8.99351572913281e-06,
+      "loss": 0.5083,
+      "step": 1286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.1278822351560756,
+      "learning_rate": 8.991775986043814e-06,
+      "loss": 0.4365,
+      "step": 1287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.807804903096628,
+      "learning_rate": 8.990034909228262e-06,
+      "loss": 0.5439,
+      "step": 1288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.912336119687683,
+      "learning_rate": 8.988292499267885e-06,
+      "loss": 0.4189,
+      "step": 1289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.556716686548828,
+      "learning_rate": 8.986548756744852e-06,
+      "loss": 0.4966,
+      "step": 1290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.882441819426605,
+      "learning_rate": 8.98480368224178e-06,
+      "loss": 0.5439,
+      "step": 1291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.43392907102242,
+      "learning_rate": 8.98305727634173e-06,
+      "loss": 0.4883,
+      "step": 1292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.656944181948202,
+      "learning_rate": 8.981309539628212e-06,
+      "loss": 0.5811,
+      "step": 1293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.359227028201466,
+      "learning_rate": 8.979560472685174e-06,
+      "loss": 0.4385,
+      "step": 1294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.356588723580822,
+      "learning_rate": 8.977810076097013e-06,
+      "loss": 0.4492,
+      "step": 1295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.82540714825025,
+      "learning_rate": 8.97605835044857e-06,
+      "loss": 0.647,
+      "step": 1296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.504595450793861,
+      "learning_rate": 8.974305296325125e-06,
+      "loss": 0.4238,
+      "step": 1297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.877558211686209,
+      "learning_rate": 8.97255091431241e-06,
+      "loss": 0.4917,
+      "step": 1298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.50916493277105,
+      "learning_rate": 8.970795204996597e-06,
+      "loss": 0.4795,
+      "step": 1299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.204769262934178,
+      "learning_rate": 8.969038168964298e-06,
+      "loss": 0.5645,
+      "step": 1300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.851121613107946,
+      "learning_rate": 8.967279806802576e-06,
+      "loss": 0.5483,
+      "step": 1301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.587881444493236,
+      "learning_rate": 8.965520119098926e-06,
+      "loss": 0.478,
+      "step": 1302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.027251162811073,
+      "learning_rate": 8.9637591064413e-06,
+      "loss": 0.624,
+      "step": 1303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.099260672728388,
+      "learning_rate": 8.961996769418077e-06,
+      "loss": 0.5215,
+      "step": 1304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.415671378119475,
+      "learning_rate": 8.960233108618092e-06,
+      "loss": 0.5791,
+      "step": 1305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.662223641152644,
+      "learning_rate": 8.958468124630617e-06,
+      "loss": 0.5718,
+      "step": 1306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.366844981796646,
+      "learning_rate": 8.956701818045363e-06,
+      "loss": 0.4946,
+      "step": 1307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.096606101285385,
+      "learning_rate": 8.954934189452489e-06,
+      "loss": 0.4512,
+      "step": 1308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.714222025011097,
+      "learning_rate": 8.953165239442589e-06,
+      "loss": 0.5986,
+      "step": 1309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.336782119481711,
+      "learning_rate": 8.951394968606704e-06,
+      "loss": 0.5625,
+      "step": 1310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.32913837235048,
+      "learning_rate": 8.949623377536314e-06,
+      "loss": 0.5757,
+      "step": 1311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.84859426270501,
+      "learning_rate": 8.947850466823343e-06,
+      "loss": 0.4834,
+      "step": 1312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.789293590180087,
+      "learning_rate": 8.946076237060148e-06,
+      "loss": 0.5361,
+      "step": 1313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.977729367014027,
+      "learning_rate": 8.944300688839538e-06,
+      "loss": 0.5249,
+      "step": 1314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.763893603900193,
+      "learning_rate": 8.942523822754751e-06,
+      "loss": 0.5415,
+      "step": 1315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.7524094218657575,
+      "learning_rate": 8.940745639399477e-06,
+      "loss": 0.5156,
+      "step": 1316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.7945367399217975,
+      "learning_rate": 8.938966139367837e-06,
+      "loss": 0.5059,
+      "step": 1317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.0240203752149535,
+      "learning_rate": 8.937185323254395e-06,
+      "loss": 0.5151,
+      "step": 1318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.481981866447117,
+      "learning_rate": 8.935403191654155e-06,
+      "loss": 0.4126,
+      "step": 1319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.449159435304624,
+      "learning_rate": 8.933619745162559e-06,
+      "loss": 0.5938,
+      "step": 1320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.242575178472086,
+      "learning_rate": 8.931834984375492e-06,
+      "loss": 0.4771,
+      "step": 1321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.949440636644377,
+      "learning_rate": 8.930048909889272e-06,
+      "loss": 0.5474,
+      "step": 1322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.20984477728289,
+      "learning_rate": 8.928261522300665e-06,
+      "loss": 0.5073,
+      "step": 1323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.653104358487502,
+      "learning_rate": 8.926472822206869e-06,
+      "loss": 0.4878,
+      "step": 1324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.750277767714117,
+      "learning_rate": 8.924682810205519e-06,
+      "loss": 0.5728,
+      "step": 1325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.545984784195176,
+      "learning_rate": 8.922891486894692e-06,
+      "loss": 0.4961,
+      "step": 1326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.327138896740525,
+      "learning_rate": 8.921098852872904e-06,
+      "loss": 0.4985,
+      "step": 1327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.5493735701358435,
+      "learning_rate": 8.919304908739106e-06,
+      "loss": 0.5244,
+      "step": 1328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.695313836532721,
+      "learning_rate": 8.917509655092691e-06,
+      "loss": 0.5732,
+      "step": 1329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.7706710902436695,
+      "learning_rate": 8.915713092533483e-06,
+      "loss": 0.4644,
+      "step": 1330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.650752650381953,
+      "learning_rate": 8.913915221661748e-06,
+      "loss": 0.5396,
+      "step": 1331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.714879719273181,
+      "learning_rate": 8.912116043078188e-06,
+      "loss": 0.5454,
+      "step": 1332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.739222984938651,
+      "learning_rate": 8.910315557383944e-06,
+      "loss": 0.6167,
+      "step": 1333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.128982922295334,
+      "learning_rate": 8.90851376518059e-06,
+      "loss": 0.5786,
+      "step": 1334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.459299957556542,
+      "learning_rate": 8.906710667070136e-06,
+      "loss": 0.624,
+      "step": 1335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.598499948785349,
+      "learning_rate": 8.904906263655036e-06,
+      "loss": 0.6475,
+      "step": 1336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.194095670682739,
+      "learning_rate": 8.903100555538169e-06,
+      "loss": 0.5264,
+      "step": 1337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.260818438840108,
+      "learning_rate": 8.90129354332286e-06,
+      "loss": 0.5459,
+      "step": 1338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.831921547703958,
+      "learning_rate": 8.899485227612865e-06,
+      "loss": 0.5386,
+      "step": 1339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.993655493393854,
+      "learning_rate": 8.897675609012372e-06,
+      "loss": 0.5488,
+      "step": 1340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.371509471270881,
+      "learning_rate": 8.895864688126013e-06,
+      "loss": 0.5415,
+      "step": 1341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.289776737421548,
+      "learning_rate": 8.894052465558846e-06,
+      "loss": 0.5205,
+      "step": 1342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.368680291839942,
+      "learning_rate": 8.892238941916372e-06,
+      "loss": 0.5693,
+      "step": 1343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.862113997301074,
+      "learning_rate": 8.890424117804522e-06,
+      "loss": 0.5518,
+      "step": 1344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.249598220460352,
+      "learning_rate": 8.88860799382966e-06,
+      "loss": 0.6641,
+      "step": 1345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.675919356373685,
+      "learning_rate": 8.88679057059859e-06,
+      "loss": 0.5454,
+      "step": 1346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.730236623835078,
+      "learning_rate": 8.884971848718544e-06,
+      "loss": 0.5562,
+      "step": 1347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.74713854214673,
+      "learning_rate": 8.883151828797194e-06,
+      "loss": 0.5557,
+      "step": 1348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.820048698580294,
+      "learning_rate": 8.88133051144264e-06,
+      "loss": 0.5674,
+      "step": 1349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.657133326905244,
+      "learning_rate": 8.87950789726342e-06,
+      "loss": 0.5859,
+      "step": 1350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.081312677292772,
+      "learning_rate": 8.8776839868685e-06,
+      "loss": 0.5039,
+      "step": 1351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.681121735697019,
+      "learning_rate": 8.875858780867286e-06,
+      "loss": 0.5093,
+      "step": 1352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.68941777386891,
+      "learning_rate": 8.87403227986961e-06,
+      "loss": 0.6191,
+      "step": 1353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.743290633026398,
+      "learning_rate": 8.872204484485743e-06,
+      "loss": 0.5903,
+      "step": 1354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.595883157216245,
+      "learning_rate": 8.870375395326384e-06,
+      "loss": 0.4712,
+      "step": 1355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.431960114700615,
+      "learning_rate": 8.868545013002665e-06,
+      "loss": 0.4814,
+      "step": 1356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.738428919658247,
+      "learning_rate": 8.866713338126152e-06,
+      "loss": 0.6064,
+      "step": 1357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.483334098623065,
+      "learning_rate": 8.86488037130884e-06,
+      "loss": 0.542,
+      "step": 1358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 28.654635831705043,
+      "learning_rate": 8.863046113163158e-06,
+      "loss": 0.5166,
+      "step": 1359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.446905888828201,
+      "learning_rate": 8.861210564301967e-06,
+      "loss": 0.5576,
+      "step": 1360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.786577684525464,
+      "learning_rate": 8.859373725338558e-06,
+      "loss": 0.5571,
+      "step": 1361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.77386586403384,
+      "learning_rate": 8.857535596886652e-06,
+      "loss": 0.5259,
+      "step": 1362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 27.49805900372077,
+      "learning_rate": 8.855696179560402e-06,
+      "loss": 0.6602,
+      "step": 1363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.273355043162183,
+      "learning_rate": 8.85385547397439e-06,
+      "loss": 0.4653,
+      "step": 1364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 20.217716781867352,
+      "learning_rate": 8.852013480743632e-06,
+      "loss": 0.6235,
+      "step": 1365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.843685680691951,
+      "learning_rate": 8.850170200483573e-06,
+      "loss": 0.4951,
+      "step": 1366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.271803428940668,
+      "learning_rate": 8.848325633810083e-06,
+      "loss": 0.498,
+      "step": 1367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.37198649024497,
+      "learning_rate": 8.84647978133947e-06,
+      "loss": 0.5361,
+      "step": 1368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.854819527941277,
+      "learning_rate": 8.844632643688467e-06,
+      "loss": 0.6172,
+      "step": 1369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.139410040675058,
+      "learning_rate": 8.842784221474237e-06,
+      "loss": 0.5098,
+      "step": 1370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.702123104062924,
+      "learning_rate": 8.840934515314372e-06,
+      "loss": 0.6406,
+      "step": 1371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.327060121802434,
+      "learning_rate": 8.839083525826893e-06,
+      "loss": 0.562,
+      "step": 1372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.967346146693572,
+      "learning_rate": 8.837231253630247e-06,
+      "loss": 0.5122,
+      "step": 1373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.222045501981935,
+      "learning_rate": 8.835377699343318e-06,
+      "loss": 0.5391,
+      "step": 1374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.7150367345669855,
+      "learning_rate": 8.83352286358541e-06,
+      "loss": 0.501,
+      "step": 1375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.34388020346308,
+      "learning_rate": 8.83166674697626e-06,
+      "loss": 0.5498,
+      "step": 1376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.994275061577608,
+      "learning_rate": 8.829809350136027e-06,
+      "loss": 0.5469,
+      "step": 1377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.286509681654963,
+      "learning_rate": 8.827950673685306e-06,
+      "loss": 0.5586,
+      "step": 1378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.615667447671237,
+      "learning_rate": 8.826090718245112e-06,
+      "loss": 0.5747,
+      "step": 1379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.045180625435941,
+      "learning_rate": 8.824229484436894e-06,
+      "loss": 0.5361,
+      "step": 1380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.914235609616878,
+      "learning_rate": 8.822366972882523e-06,
+      "loss": 0.6753,
+      "step": 1381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.65556839435706,
+      "learning_rate": 8.820503184204299e-06,
+      "loss": 0.5171,
+      "step": 1382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.818279504647876,
+      "learning_rate": 8.818638119024949e-06,
+      "loss": 0.5488,
+      "step": 1383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 29.01882884415183,
+      "learning_rate": 8.816771777967623e-06,
+      "loss": 0.6357,
+      "step": 1384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.102254652825774,
+      "learning_rate": 8.814904161655904e-06,
+      "loss": 0.5728,
+      "step": 1385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.028759055391328,
+      "learning_rate": 8.813035270713796e-06,
+      "loss": 0.4946,
+      "step": 1386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.481350770883132,
+      "learning_rate": 8.811165105765732e-06,
+      "loss": 0.5146,
+      "step": 1387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.584523783116981,
+      "learning_rate": 8.809293667436565e-06,
+      "loss": 0.5498,
+      "step": 1388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.670757323199173,
+      "learning_rate": 8.80742095635158e-06,
+      "loss": 0.5869,
+      "step": 1389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 13.789379257935748,
+      "learning_rate": 8.805546973136481e-06,
+      "loss": 0.5391,
+      "step": 1390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.92942354858489,
+      "learning_rate": 8.803671718417407e-06,
+      "loss": 0.5396,
+      "step": 1391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.025556704510787,
+      "learning_rate": 8.80179519282091e-06,
+      "loss": 0.4453,
+      "step": 1392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.863285419079764,
+      "learning_rate": 8.799917396973976e-06,
+      "loss": 0.5576,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.510372608172871,
+      "learning_rate": 8.798038331504008e-06,
+      "loss": 0.499,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.339190789992044,
+      "learning_rate": 8.79615799703884e-06,
+      "loss": 0.5347,
+      "step": 1395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.228994238966333,
+      "learning_rate": 8.794276394206722e-06,
+      "loss": 0.4858,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.720067062308178,
+      "learning_rate": 8.792393523636337e-06,
+      "loss": 0.5122,
+      "step": 1397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.959990759829996,
+      "learning_rate": 8.790509385956784e-06,
+      "loss": 0.6104,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.7350338881120315,
+      "learning_rate": 8.788623981797592e-06,
+      "loss": 0.5569,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.122908029947811,
+      "learning_rate": 8.786737311788708e-06,
+      "loss": 0.5083,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.249960081112658,
+      "learning_rate": 8.784849376560503e-06,
+      "loss": 0.4473,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.9378761564742,
+      "learning_rate": 8.78296017674377e-06,
+      "loss": 0.3892,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.947182755307541,
+      "learning_rate": 8.781069712969726e-06,
+      "loss": 0.4663,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.735615644483297,
+      "learning_rate": 8.779177985870012e-06,
+      "loss": 0.4985,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.221979827272879,
+      "learning_rate": 8.77728499607669e-06,
+      "loss": 0.3682,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.51360598274658,
+      "learning_rate": 8.775390744222238e-06,
+      "loss": 0.4927,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.824576130882553,
+      "learning_rate": 8.773495230939567e-06,
+      "loss": 0.439,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.542083031305357,
+      "learning_rate": 8.771598456861998e-06,
+      "loss": 0.407,
+      "step": 1408
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.74463891725652,
+      "learning_rate": 8.769700422623283e-06,
+      "loss": 0.3843,
+      "step": 1409
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.052978396859257,
+      "learning_rate": 8.767801128857588e-06,
+      "loss": 0.4321,
+      "step": 1410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.845362886453344,
+      "learning_rate": 8.765900576199502e-06,
+      "loss": 0.4565,
+      "step": 1411
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.17409389016001,
+      "learning_rate": 8.763998765284036e-06,
+      "loss": 0.4888,
+      "step": 1412
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.04399663972258,
+      "learning_rate": 8.76209569674662e-06,
+      "loss": 0.4199,
+      "step": 1413
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.796687567190697,
+      "learning_rate": 8.760191371223104e-06,
+      "loss": 0.4346,
+      "step": 1414
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 13.742485700753212,
+      "learning_rate": 8.758285789349759e-06,
+      "loss": 0.479,
+      "step": 1415
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.549981447404763,
+      "learning_rate": 8.756378951763277e-06,
+      "loss": 0.4429,
+      "step": 1416
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.523418901631306,
+      "learning_rate": 8.754470859100765e-06,
+      "loss": 0.3989,
+      "step": 1417
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.12913542529696,
+      "learning_rate": 8.752561511999754e-06,
+      "loss": 0.5083,
+      "step": 1418
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 15.040083536035315,
+      "learning_rate": 8.750650911098193e-06,
+      "loss": 0.4619,
+      "step": 1419
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 14.958198799461423,
+      "learning_rate": 8.748739057034447e-06,
+      "loss": 0.457,
+      "step": 1420
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 34.28258733668816,
+      "learning_rate": 8.746825950447302e-06,
+      "loss": 0.3999,
+      "step": 1421
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.46394179525724,
+      "learning_rate": 8.744911591975967e-06,
+      "loss": 0.4434,
+      "step": 1422
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.570102832172687,
+      "learning_rate": 8.742995982260059e-06,
+      "loss": 0.4307,
+      "step": 1423
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.157526748035593,
+      "learning_rate": 8.741079121939621e-06,
+      "loss": 0.4961,
+      "step": 1424
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.562647415324578,
+      "learning_rate": 8.739161011655113e-06,
+      "loss": 0.458,
+      "step": 1425
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.069738473837281,
+      "learning_rate": 8.737241652047408e-06,
+      "loss": 0.603,
+      "step": 1426
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.918982204133165,
+      "learning_rate": 8.735321043757805e-06,
+      "loss": 0.4688,
+      "step": 1427
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.46229012344293,
+      "learning_rate": 8.73339918742801e-06,
+      "loss": 0.4341,
+      "step": 1428
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.84418719877701,
+      "learning_rate": 8.731476083700154e-06,
+      "loss": 0.4683,
+      "step": 1429
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.59064803571736,
+      "learning_rate": 8.729551733216779e-06,
+      "loss": 0.4229,
+      "step": 1430
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.516324309495134,
+      "learning_rate": 8.727626136620848e-06,
+      "loss": 0.4502,
+      "step": 1431
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 26.410113834424802,
+      "learning_rate": 8.725699294555739e-06,
+      "loss": 0.5132,
+      "step": 1432
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 20.34794795255431,
+      "learning_rate": 8.723771207665245e-06,
+      "loss": 0.5312,
+      "step": 1433
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.382266025888864,
+      "learning_rate": 8.721841876593576e-06,
+      "loss": 0.4482,
+      "step": 1434
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.996872293754613,
+      "learning_rate": 8.719911301985355e-06,
+      "loss": 0.4189,
+      "step": 1435
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.616870720209013,
+      "learning_rate": 8.717979484485628e-06,
+      "loss": 0.3623,
+      "step": 1436
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 18.78460022966816,
+      "learning_rate": 8.716046424739845e-06,
+      "loss": 0.5029,
+      "step": 1437
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 27.875048810799026,
+      "learning_rate": 8.714112123393882e-06,
+      "loss": 0.5117,
+      "step": 1438
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.085152704221128,
+      "learning_rate": 8.712176581094025e-06,
+      "loss": 0.438,
+      "step": 1439
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.939014408876105,
+      "learning_rate": 8.710239798486972e-06,
+      "loss": 0.5273,
+      "step": 1440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 36.2593949826461,
+      "learning_rate": 8.708301776219838e-06,
+      "loss": 0.4185,
+      "step": 1441
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 112.11982714132783,
+      "learning_rate": 8.706362514940153e-06,
+      "loss": 0.4478,
+      "step": 1442
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 277.4382397294177,
+      "learning_rate": 8.704422015295861e-06,
+      "loss": 0.5786,
+      "step": 1443
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 181.53470646624706,
+      "learning_rate": 8.702480277935319e-06,
+      "loss": 0.6011,
+      "step": 1444
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.931282079030481,
+      "learning_rate": 8.700537303507298e-06,
+      "loss": 0.3828,
+      "step": 1445
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 22.345280510200695,
+      "learning_rate": 8.69859309266098e-06,
+      "loss": 0.54,
+      "step": 1446
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.92976281404644,
+      "learning_rate": 8.696647646045962e-06,
+      "loss": 0.4468,
+      "step": 1447
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 15.584050288102251,
+      "learning_rate": 8.694700964312257e-06,
+      "loss": 0.499,
+      "step": 1448
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.782706492066882,
+      "learning_rate": 8.692753048110283e-06,
+      "loss": 0.5635,
+      "step": 1449
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 17.982712514408526,
+      "learning_rate": 8.690803898090878e-06,
+      "loss": 0.4897,
+      "step": 1450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.009874311456212,
+      "learning_rate": 8.68885351490529e-06,
+      "loss": 0.3774,
+      "step": 1451
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.187289386392372,
+      "learning_rate": 8.686901899205177e-06,
+      "loss": 0.4077,
+      "step": 1452
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.74330679007414,
+      "learning_rate": 8.684949051642609e-06,
+      "loss": 0.4907,
+      "step": 1453
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.533483611225794,
+      "learning_rate": 8.68299497287007e-06,
+      "loss": 0.4238,
+      "step": 1454
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.93205428371546,
+      "learning_rate": 8.681039663540454e-06,
+      "loss": 0.4351,
+      "step": 1455
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 11.711598620907111,
+      "learning_rate": 8.679083124307064e-06,
+      "loss": 0.321,
+      "step": 1456
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 20.3063051905801,
+      "learning_rate": 8.67712535582362e-06,
+      "loss": 0.4761,
+      "step": 1457
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.81069964942567,
+      "learning_rate": 8.675166358744247e-06,
+      "loss": 0.4497,
+      "step": 1458
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 25.172668334190924,
+      "learning_rate": 8.67320613372348e-06,
+      "loss": 0.6162,
+      "step": 1459
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.868299776664779,
+      "learning_rate": 8.67124468141627e-06,
+      "loss": 0.4697,
+      "step": 1460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 17.65557462571263,
+      "learning_rate": 8.669282002477975e-06,
+      "loss": 0.5439,
+      "step": 1461
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.986941627654915,
+      "learning_rate": 8.66731809756436e-06,
+      "loss": 0.4077,
+      "step": 1462
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 13.481751253152993,
+      "learning_rate": 8.665352967331604e-06,
+      "loss": 0.4507,
+      "step": 1463
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.69597798278669,
+      "learning_rate": 8.66338661243629e-06,
+      "loss": 0.4185,
+      "step": 1464
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 21.12105825698134,
+      "learning_rate": 8.661419033535419e-06,
+      "loss": 0.4966,
+      "step": 1465
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.364831925795514,
+      "learning_rate": 8.659450231286392e-06,
+      "loss": 0.4619,
+      "step": 1466
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.36575882497529,
+      "learning_rate": 8.657480206347024e-06,
+      "loss": 0.478,
+      "step": 1467
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 19.20617981129387,
+      "learning_rate": 8.655508959375536e-06,
+      "loss": 0.458,
+      "step": 1468
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 12.332507150136136,
+      "learning_rate": 8.653536491030559e-06,
+      "loss": 0.4453,
+      "step": 1469
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.375737589014147,
+      "learning_rate": 8.651562801971131e-06,
+      "loss": 0.4199,
+      "step": 1470
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.622191769632364,
+      "learning_rate": 8.649587892856698e-06,
+      "loss": 0.4438,
+      "step": 1471
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.028816389671348,
+      "learning_rate": 8.647611764347114e-06,
+      "loss": 0.4634,
+      "step": 1472
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 51.19574362423425,
+      "learning_rate": 8.64563441710264e-06,
+      "loss": 0.7012,
+      "step": 1473
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.972205154177201,
+      "learning_rate": 8.643655851783947e-06,
+      "loss": 0.3843,
+      "step": 1474
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.72003218429071,
+      "learning_rate": 8.641676069052104e-06,
+      "loss": 0.4072,
+      "step": 1475
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.009750309222879,
+      "learning_rate": 8.639695069568602e-06,
+      "loss": 0.4717,
+      "step": 1476
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.704206269175184,
+      "learning_rate": 8.637712853995324e-06,
+      "loss": 0.4814,
+      "step": 1477
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 9.870396068272045,
+      "learning_rate": 8.635729422994566e-06,
+      "loss": 0.4634,
+      "step": 1478
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.55132356224423,
+      "learning_rate": 8.633744777229029e-06,
+      "loss": 0.4365,
+      "step": 1479
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.263513038793375,
+      "learning_rate": 8.63175891736182e-06,
+      "loss": 0.3862,
+      "step": 1480
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.735243962302292,
+      "learning_rate": 8.629771844056452e-06,
+      "loss": 0.3691,
+      "step": 1481
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.39389573007216,
+      "learning_rate": 8.627783557976846e-06,
+      "loss": 0.4902,
+      "step": 1482
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.633505897881955,
+      "learning_rate": 8.62579405978732e-06,
+      "loss": 0.4678,
+      "step": 1483
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.95154363916375,
+      "learning_rate": 8.623803350152606e-06,
+      "loss": 0.4326,
+      "step": 1484
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.629591334063445,
+      "learning_rate": 8.621811429737837e-06,
+      "loss": 0.4819,
+      "step": 1485
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.321230859689711,
+      "learning_rate": 8.619818299208548e-06,
+      "loss": 0.3994,
+      "step": 1486
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.680762867215169,
+      "learning_rate": 8.617823959230683e-06,
+      "loss": 0.5298,
+      "step": 1487
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.312614949546656,
+      "learning_rate": 8.615828410470589e-06,
+      "loss": 0.5034,
+      "step": 1488
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 15.386409700658552,
+      "learning_rate": 8.613831653595013e-06,
+      "loss": 0.5503,
+      "step": 1489
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 16.553167562663948,
+      "learning_rate": 8.61183368927111e-06,
+      "loss": 0.3765,
+      "step": 1490
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 11.97547090065287,
+      "learning_rate": 8.609834518166439e-06,
+      "loss": 0.4897,
+      "step": 1491
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.51487310954413,
+      "learning_rate": 8.607834140948958e-06,
+      "loss": 0.4663,
+      "step": 1492
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.020553039764543,
+      "learning_rate": 8.60583255828703e-06,
+      "loss": 0.416,
+      "step": 1493
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.540414928748616,
+      "learning_rate": 8.603829770849421e-06,
+      "loss": 0.5366,
+      "step": 1494
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.386118227859209,
+      "learning_rate": 8.601825779305302e-06,
+      "loss": 0.4199,
+      "step": 1495
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.19089679469395,
+      "learning_rate": 8.59982058432424e-06,
+      "loss": 0.4824,
+      "step": 1496
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.744882941104166,
+      "learning_rate": 8.597814186576212e-06,
+      "loss": 0.4531,
+      "step": 1497
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.998468653111926,
+      "learning_rate": 8.595806586731589e-06,
+      "loss": 0.4771,
+      "step": 1498
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.331660088580012,
+      "learning_rate": 8.59379778546115e-06,
+      "loss": 0.5522,
+      "step": 1499
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.388963075279165,
+      "learning_rate": 8.591787783436073e-06,
+      "loss": 0.4834,
+      "step": 1500
+    },
+    {
+      "epoch": 1.07,
+      "eval_avg_AUC": 0.7334799976115187,
+      "eval_avg_Accuracy": 0.6713776525198939,
+      "eval_avg_Accuracy-right": 0.87544019825225,
+      "eval_avg_Accuracy-wrong": 0.31555606095064814,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6231060873343826,
+      "eval_last_AUC": 0.7557780548881416,
+      "eval_last_Accuracy": 0.6933438328912467,
+      "eval_last_Accuracy-right": 0.8278987870092605,
+      "eval_last_Accuracy-wrong": 0.45872185581078007,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6509237397272707,
+      "eval_max_AUC": 0.7050180895122209,
+      "eval_max_Accuracy": 0.6380968169761273,
+      "eval_max_Accuracy-right": 0.947045780618234,
+      "eval_max_Accuracy-wrong": 0.09938594496247441,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5989895777078,
+      "eval_min_AUC": 0.7259579465041305,
+      "eval_min_Accuracy": 0.663834549071618,
+      "eval_min_Accuracy-right": 0.7148819616538411,
+      "eval_min_Accuracy-wrong": 0.574823743461451,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6227692066316574,
+      "eval_prod_AUC": 0.7346491876627528,
+      "eval_prod_Accuracy": 0.6056863395225465,
+      "eval_prod_Accuracy-right": 0.46087126646667537,
+      "eval_prod_Accuracy-wrong": 0.858198771889925,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6170978703493736,
+      "eval_runtime": 251.2444,
+      "eval_samples_per_second": 96.034,
+      "eval_steps_per_second": 3.001,
+      "eval_sum_AUC": 0.6155480756158993,
+      "eval_sum_Accuracy": 0.6353199602122016,
+      "eval_sum_Accuracy-right": 0.9979783487674448,
+      "eval_sum_Accuracy-wrong": 0.0029565612917898565,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6239868444598957,
+      "step": 1500
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.466477606369253,
+      "learning_rate": 8.589776581327936e-06,
+      "loss": 0.5674,
+      "step": 1501
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 13.612158852318972,
+      "learning_rate": 8.587764179808716e-06,
+      "loss": 0.4512,
+      "step": 1502
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.617095094049752,
+      "learning_rate": 8.5857505795508e-06,
+      "loss": 0.4932,
+      "step": 1503
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 20.816769693684805,
+      "learning_rate": 8.583735781226964e-06,
+      "loss": 0.4648,
+      "step": 1504
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 12.235810035295055,
+      "learning_rate": 8.581719785510391e-06,
+      "loss": 0.4233,
+      "step": 1505
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.94140620816876,
+      "learning_rate": 8.579702593074666e-06,
+      "loss": 0.5186,
+      "step": 1506
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.10601483533534,
+      "learning_rate": 8.577684204593767e-06,
+      "loss": 0.5,
+      "step": 1507
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.588955613794905,
+      "learning_rate": 8.575664620742073e-06,
+      "loss": 0.4282,
+      "step": 1508
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 16.445563006577682,
+      "learning_rate": 8.57364384219437e-06,
+      "loss": 0.4819,
+      "step": 1509
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.081966627737616,
+      "learning_rate": 8.571621869625835e-06,
+      "loss": 0.4707,
+      "step": 1510
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.676101669447284,
+      "learning_rate": 8.569598703712045e-06,
+      "loss": 0.4351,
+      "step": 1511
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.660950506422358,
+      "learning_rate": 8.56757434512898e-06,
+      "loss": 0.5244,
+      "step": 1512
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 11.673297608958263,
+      "learning_rate": 8.565548794553016e-06,
+      "loss": 0.4507,
+      "step": 1513
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.604090359301853,
+      "learning_rate": 8.563522052660925e-06,
+      "loss": 0.5532,
+      "step": 1514
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.711609470198233,
+      "learning_rate": 8.561494120129878e-06,
+      "loss": 0.5186,
+      "step": 1515
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.518970050965722,
+      "learning_rate": 8.55946499763745e-06,
+      "loss": 0.4624,
+      "step": 1516
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 24.622164675000597,
+      "learning_rate": 8.557434685861604e-06,
+      "loss": 0.5537,
+      "step": 1517
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.341729976026523,
+      "learning_rate": 8.555403185480706e-06,
+      "loss": 0.5444,
+      "step": 1518
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 14.951861865208672,
+      "learning_rate": 8.553370497173518e-06,
+      "loss": 0.499,
+      "step": 1519
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.798380638701046,
+      "learning_rate": 8.551336621619202e-06,
+      "loss": 0.4722,
+      "step": 1520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.1744411262223,
+      "learning_rate": 8.549301559497309e-06,
+      "loss": 0.4937,
+      "step": 1521
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.356907776973655,
+      "learning_rate": 8.547265311487794e-06,
+      "loss": 0.4932,
+      "step": 1522
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.570800954915105,
+      "learning_rate": 8.545227878271004e-06,
+      "loss": 0.5479,
+      "step": 1523
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.945217800371761,
+      "learning_rate": 8.543189260527685e-06,
+      "loss": 0.4419,
+      "step": 1524
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.810431743115078,
+      "learning_rate": 8.541149458938972e-06,
+      "loss": 0.5332,
+      "step": 1525
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.488127989967904,
+      "learning_rate": 8.539108474186408e-06,
+      "loss": 0.4951,
+      "step": 1526
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.5250746108148165,
+      "learning_rate": 8.53706630695192e-06,
+      "loss": 0.4365,
+      "step": 1527
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.090581047201338,
+      "learning_rate": 8.535022957917833e-06,
+      "loss": 0.4536,
+      "step": 1528
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.713255870212436,
+      "learning_rate": 8.53297842776687e-06,
+      "loss": 0.5649,
+      "step": 1529
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 14.611298003245253,
+      "learning_rate": 8.530932717182148e-06,
+      "loss": 0.5117,
+      "step": 1530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.796700782725793,
+      "learning_rate": 8.528885826847173e-06,
+      "loss": 0.4463,
+      "step": 1531
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 10.108308604467041,
+      "learning_rate": 8.52683775744585e-06,
+      "loss": 0.416,
+      "step": 1532
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.550989790783058,
+      "learning_rate": 8.524788509662478e-06,
+      "loss": 0.4971,
+      "step": 1533
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.558016423382215,
+      "learning_rate": 8.522738084181749e-06,
+      "loss": 0.5479,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.741126259708553,
+      "learning_rate": 8.52068648168875e-06,
+      "loss": 0.4404,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.33677197442007,
+      "learning_rate": 8.518633702868955e-06,
+      "loss": 0.4131,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.431400081236161,
+      "learning_rate": 8.516579748408237e-06,
+      "loss": 0.4629,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.038672006100304,
+      "learning_rate": 8.514524618992864e-06,
+      "loss": 0.4155,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.732435386054672,
+      "learning_rate": 8.51246831530949e-06,
+      "loss": 0.4648,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 18.788487706437273,
+      "learning_rate": 8.510410838045165e-06,
+      "loss": 0.4658,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.614847193707238,
+      "learning_rate": 8.508352187887329e-06,
+      "loss": 0.4868,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.621493961264449,
+      "learning_rate": 8.506292365523816e-06,
+      "loss": 0.4014,
+      "step": 1542
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.908663513591653,
+      "learning_rate": 8.504231371642852e-06,
+      "loss": 0.5405,
+      "step": 1543
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.103135984089345,
+      "learning_rate": 8.502169206933053e-06,
+      "loss": 0.4414,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.697544626949995,
+      "learning_rate": 8.500105872083424e-06,
+      "loss": 0.4463,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 19.462066133716373,
+      "learning_rate": 8.498041367783367e-06,
+      "loss": 0.4731,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.778716367020898,
+      "learning_rate": 8.49597569472267e-06,
+      "loss": 0.5264,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 9.104157085710243,
+      "learning_rate": 8.493908853591515e-06,
+      "loss": 0.4722,
+      "step": 1548
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.793898810134827,
+      "learning_rate": 8.491840845080467e-06,
+      "loss": 0.6045,
+      "step": 1549
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.705523280095832,
+      "learning_rate": 8.489771669880489e-06,
+      "loss": 0.5137,
+      "step": 1550
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.332670535302874,
+      "learning_rate": 8.487701328682932e-06,
+      "loss": 0.4795,
+      "step": 1551
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.589426733756543,
+      "learning_rate": 8.485629822179533e-06,
+      "loss": 0.4575,
+      "step": 1552
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 13.353226183033948,
+      "learning_rate": 8.483557151062423e-06,
+      "loss": 0.4497,
+      "step": 1553
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.18956399500516,
+      "learning_rate": 8.481483316024117e-06,
+      "loss": 0.4678,
+      "step": 1554
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.368944548115262,
+      "learning_rate": 8.479408317757525e-06,
+      "loss": 0.5283,
+      "step": 1555
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.054711315944372,
+      "learning_rate": 8.477332156955942e-06,
+      "loss": 0.5078,
+      "step": 1556
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.511879702200282,
+      "learning_rate": 8.475254834313051e-06,
+      "loss": 0.5015,
+      "step": 1557
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.3864182168499,
+      "learning_rate": 8.473176350522925e-06,
+      "loss": 0.4126,
+      "step": 1558
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 10.192652705648785,
+      "learning_rate": 8.471096706280022e-06,
+      "loss": 0.5127,
+      "step": 1559
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.215018804441314,
+      "learning_rate": 8.469015902279191e-06,
+      "loss": 0.4111,
+      "step": 1560
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.060201902069851,
+      "learning_rate": 8.466933939215669e-06,
+      "loss": 0.4883,
+      "step": 1561
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 11.696202616513581,
+      "learning_rate": 8.464850817785075e-06,
+      "loss": 0.4199,
+      "step": 1562
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.514532333226486,
+      "learning_rate": 8.462766538683422e-06,
+      "loss": 0.4536,
+      "step": 1563
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 9.329536103759395,
+      "learning_rate": 8.460681102607106e-06,
+      "loss": 0.4072,
+      "step": 1564
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 18.684227028359807,
+      "learning_rate": 8.45859451025291e-06,
+      "loss": 0.6074,
+      "step": 1565
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 25.38421632073626,
+      "learning_rate": 8.456506762317998e-06,
+      "loss": 0.7139,
+      "step": 1566
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.286401468541102,
+      "learning_rate": 8.454417859499932e-06,
+      "loss": 0.5562,
+      "step": 1567
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.597284316583334,
+      "learning_rate": 8.45232780249665e-06,
+      "loss": 0.4604,
+      "step": 1568
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.314962157168296,
+      "learning_rate": 8.450236592006481e-06,
+      "loss": 0.4844,
+      "step": 1569
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.11769377664577,
+      "learning_rate": 8.448144228728135e-06,
+      "loss": 0.4971,
+      "step": 1570
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.673405541056715,
+      "learning_rate": 8.446050713360711e-06,
+      "loss": 0.4473,
+      "step": 1571
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.223436085969732,
+      "learning_rate": 8.443956046603692e-06,
+      "loss": 0.54,
+      "step": 1572
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.801140278815467,
+      "learning_rate": 8.441860229156944e-06,
+      "loss": 0.4429,
+      "step": 1573
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.623164408153567,
+      "learning_rate": 8.439763261720716e-06,
+      "loss": 0.6367,
+      "step": 1574
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.285922227520546,
+      "learning_rate": 8.43766514499565e-06,
+      "loss": 0.54,
+      "step": 1575
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.746100363903965,
+      "learning_rate": 8.435565879682759e-06,
+      "loss": 0.5107,
+      "step": 1576
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.979137821257687,
+      "learning_rate": 8.433465466483452e-06,
+      "loss": 0.5464,
+      "step": 1577
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.598799151468013,
+      "learning_rate": 8.431363906099513e-06,
+      "loss": 0.4863,
+      "step": 1578
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.105699816627178,
+      "learning_rate": 8.429261199233114e-06,
+      "loss": 0.4521,
+      "step": 1579
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.9180976781052,
+      "learning_rate": 8.427157346586807e-06,
+      "loss": 0.4756,
+      "step": 1580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.86495838313817,
+      "learning_rate": 8.42505234886353e-06,
+      "loss": 0.4922,
+      "step": 1581
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.545267152082388,
+      "learning_rate": 8.422946206766598e-06,
+      "loss": 0.4888,
+      "step": 1582
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.713434930793355,
+      "learning_rate": 8.420838920999718e-06,
+      "loss": 0.416,
+      "step": 1583
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.76577412131441,
+      "learning_rate": 8.418730492266968e-06,
+      "loss": 0.5044,
+      "step": 1584
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.502799153950787,
+      "learning_rate": 8.416620921272818e-06,
+      "loss": 0.4326,
+      "step": 1585
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.461253439409639,
+      "learning_rate": 8.414510208722111e-06,
+      "loss": 0.5684,
+      "step": 1586
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 19.10348435047004,
+      "learning_rate": 8.412398355320078e-06,
+      "loss": 0.5229,
+      "step": 1587
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.506878119981076,
+      "learning_rate": 8.410285361772328e-06,
+      "loss": 0.4419,
+      "step": 1588
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.761222151604644,
+      "learning_rate": 8.408171228784847e-06,
+      "loss": 0.4199,
+      "step": 1589
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.759892100205057,
+      "learning_rate": 8.406055957064014e-06,
+      "loss": 0.4717,
+      "step": 1590
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 15.051203228423255,
+      "learning_rate": 8.403939547316576e-06,
+      "loss": 0.4541,
+      "step": 1591
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.735328118613461,
+      "learning_rate": 8.401822000249661e-06,
+      "loss": 0.4087,
+      "step": 1592
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.260653736059055,
+      "learning_rate": 8.399703316570788e-06,
+      "loss": 0.4463,
+      "step": 1593
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.10976108191569,
+      "learning_rate": 8.397583496987846e-06,
+      "loss": 0.519,
+      "step": 1594
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.768171024664671,
+      "learning_rate": 8.395462542209106e-06,
+      "loss": 0.4648,
+      "step": 1595
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.026676798546331,
+      "learning_rate": 8.393340452943219e-06,
+      "loss": 0.501,
+      "step": 1596
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.16743911627819,
+      "learning_rate": 8.391217229899211e-06,
+      "loss": 0.5093,
+      "step": 1597
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.830636260029298,
+      "learning_rate": 8.389092873786495e-06,
+      "loss": 0.375,
+      "step": 1598
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.20546214650608,
+      "learning_rate": 8.386967385314857e-06,
+      "loss": 0.4756,
+      "step": 1599
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.439896508899231,
+      "learning_rate": 8.384840765194458e-06,
+      "loss": 0.4346,
+      "step": 1600
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 19.327934158649416,
+      "learning_rate": 8.382713014135846e-06,
+      "loss": 0.6797,
+      "step": 1601
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 9.885276182041746,
+      "learning_rate": 8.38058413284994e-06,
+      "loss": 0.5352,
+      "step": 1602
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.596961361196003,
+      "learning_rate": 8.37845412204804e-06,
+      "loss": 0.4497,
+      "step": 1603
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 11.019827177632346,
+      "learning_rate": 8.376322982441821e-06,
+      "loss": 0.4795,
+      "step": 1604
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.122204287350154,
+      "learning_rate": 8.374190714743338e-06,
+      "loss": 0.3926,
+      "step": 1605
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.991030964363688,
+      "learning_rate": 8.37205731966502e-06,
+      "loss": 0.4463,
+      "step": 1606
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.075319370068696,
+      "learning_rate": 8.369922797919672e-06,
+      "loss": 0.395,
+      "step": 1607
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.044817706502199,
+      "learning_rate": 8.367787150220481e-06,
+      "loss": 0.4814,
+      "step": 1608
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.41670683342002,
+      "learning_rate": 8.365650377281004e-06,
+      "loss": 0.4272,
+      "step": 1609
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 15.173314309019913,
+      "learning_rate": 8.36351247981518e-06,
+      "loss": 0.4678,
+      "step": 1610
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.550970965767231,
+      "learning_rate": 8.361373458537316e-06,
+      "loss": 0.374,
+      "step": 1611
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.034925091044531,
+      "learning_rate": 8.359233314162102e-06,
+      "loss": 0.439,
+      "step": 1612
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.425295339263013,
+      "learning_rate": 8.357092047404598e-06,
+      "loss": 0.3662,
+      "step": 1613
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.31408262861963,
+      "learning_rate": 8.354949658980243e-06,
+      "loss": 0.4409,
+      "step": 1614
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.260734469839917,
+      "learning_rate": 8.352806149604847e-06,
+      "loss": 0.3794,
+      "step": 1615
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.468850196741588,
+      "learning_rate": 8.350661519994596e-06,
+      "loss": 0.6748,
+      "step": 1616
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.792508635270057,
+      "learning_rate": 8.348515770866051e-06,
+      "loss": 0.564,
+      "step": 1617
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.705560827563696,
+      "learning_rate": 8.346368902936149e-06,
+      "loss": 0.5049,
+      "step": 1618
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.253537791784684,
+      "learning_rate": 8.344220916922195e-06,
+      "loss": 0.4131,
+      "step": 1619
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.961687115125363,
+      "learning_rate": 8.342071813541873e-06,
+      "loss": 0.3879,
+      "step": 1620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.258364878364446,
+      "learning_rate": 8.339921593513239e-06,
+      "loss": 0.5815,
+      "step": 1621
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.105088701540607,
+      "learning_rate": 8.337770257554721e-06,
+      "loss": 0.4336,
+      "step": 1622
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.40871548874701,
+      "learning_rate": 8.335617806385119e-06,
+      "loss": 0.3545,
+      "step": 1623
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.767330324784009,
+      "learning_rate": 8.333464240723608e-06,
+      "loss": 0.4673,
+      "step": 1624
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.648199165290603,
+      "learning_rate": 8.331309561289734e-06,
+      "loss": 0.4395,
+      "step": 1625
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.32741559592417,
+      "learning_rate": 8.329153768803415e-06,
+      "loss": 0.4731,
+      "step": 1626
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.778161659116726,
+      "learning_rate": 8.326996863984942e-06,
+      "loss": 0.5933,
+      "step": 1627
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.348733791355947,
+      "learning_rate": 8.324838847554976e-06,
+      "loss": 0.4395,
+      "step": 1628
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.740098579497312,
+      "learning_rate": 8.322679720234553e-06,
+      "loss": 0.3857,
+      "step": 1629
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.715818168299126,
+      "learning_rate": 8.320519482745076e-06,
+      "loss": 0.4917,
+      "step": 1630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.55961555884494,
+      "learning_rate": 8.31835813580832e-06,
+      "loss": 0.416,
+      "step": 1631
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.584012997946722,
+      "learning_rate": 8.316195680146431e-06,
+      "loss": 0.5322,
+      "step": 1632
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.513206719088476,
+      "learning_rate": 8.314032116481927e-06,
+      "loss": 0.5117,
+      "step": 1633
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.048893376656899,
+      "learning_rate": 8.311867445537694e-06,
+      "loss": 0.4272,
+      "step": 1634
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.270871090848454,
+      "learning_rate": 8.30970166803699e-06,
+      "loss": 0.4766,
+      "step": 1635
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.569630501193533,
+      "learning_rate": 8.307534784703438e-06,
+      "loss": 0.4082,
+      "step": 1636
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.423256634220007,
+      "learning_rate": 8.305366796261036e-06,
+      "loss": 0.4019,
+      "step": 1637
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.689591987413117,
+      "learning_rate": 8.303197703434151e-06,
+      "loss": 0.5371,
+      "step": 1638
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.2646267025812,
+      "learning_rate": 8.301027506947516e-06,
+      "loss": 0.5225,
+      "step": 1639
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.004304702541242,
+      "learning_rate": 8.298856207526234e-06,
+      "loss": 0.5005,
+      "step": 1640
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.275278524089247,
+      "learning_rate": 8.296683805895777e-06,
+      "loss": 0.4683,
+      "step": 1641
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.349760402537493,
+      "learning_rate": 8.294510302781984e-06,
+      "loss": 0.4644,
+      "step": 1642
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.34559489538677,
+      "learning_rate": 8.29233569891106e-06,
+      "loss": 0.4131,
+      "step": 1643
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.09744536399142,
+      "learning_rate": 8.290159995009586e-06,
+      "loss": 0.4858,
+      "step": 1644
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.573866063978919,
+      "learning_rate": 8.2879831918045e-06,
+      "loss": 0.4844,
+      "step": 1645
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.655351070066807,
+      "learning_rate": 8.285805290023119e-06,
+      "loss": 0.4937,
+      "step": 1646
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.922336465989796,
+      "learning_rate": 8.283626290393112e-06,
+      "loss": 0.5044,
+      "step": 1647
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.600737753837139,
+      "learning_rate": 8.28144619364253e-06,
+      "loss": 0.4478,
+      "step": 1648
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.9710210367027,
+      "learning_rate": 8.279265000499783e-06,
+      "loss": 0.5781,
+      "step": 1649
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.960701863701635,
+      "learning_rate": 8.277082711693645e-06,
+      "loss": 0.5278,
+      "step": 1650
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.16288939323412,
+      "learning_rate": 8.274899327953261e-06,
+      "loss": 0.4927,
+      "step": 1651
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 28.608511125234113,
+      "learning_rate": 8.272714850008142e-06,
+      "loss": 0.6494,
+      "step": 1652
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.7112136057185765,
+      "learning_rate": 8.270529278588158e-06,
+      "loss": 0.4153,
+      "step": 1653
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.677455868257633,
+      "learning_rate": 8.268342614423553e-06,
+      "loss": 0.4663,
+      "step": 1654
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.614598532387653,
+      "learning_rate": 8.26615485824493e-06,
+      "loss": 0.4663,
+      "step": 1655
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.086291528498144,
+      "learning_rate": 8.263966010783259e-06,
+      "loss": 0.4448,
+      "step": 1656
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.52793141998646,
+      "learning_rate": 8.261776072769878e-06,
+      "loss": 0.4453,
+      "step": 1657
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.71379532658703,
+      "learning_rate": 8.259585044936484e-06,
+      "loss": 0.4429,
+      "step": 1658
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.139946386227027,
+      "learning_rate": 8.257392928015138e-06,
+      "loss": 0.4644,
+      "step": 1659
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.658585729760876,
+      "learning_rate": 8.25519972273827e-06,
+      "loss": 0.4727,
+      "step": 1660
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.794428336250439,
+      "learning_rate": 8.253005429838667e-06,
+      "loss": 0.4209,
+      "step": 1661
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.775021748055948,
+      "learning_rate": 8.250810050049488e-06,
+      "loss": 0.4678,
+      "step": 1662
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.921459024373188,
+      "learning_rate": 8.248613584104245e-06,
+      "loss": 0.4731,
+      "step": 1663
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.454721792025683,
+      "learning_rate": 8.246416032736824e-06,
+      "loss": 0.4658,
+      "step": 1664
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.732194232724195,
+      "learning_rate": 8.244217396681461e-06,
+      "loss": 0.3638,
+      "step": 1665
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.820566534030801,
+      "learning_rate": 8.242017676672766e-06,
+      "loss": 0.4893,
+      "step": 1666
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.00873769637777,
+      "learning_rate": 8.239816873445705e-06,
+      "loss": 0.4873,
+      "step": 1667
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.979873575388726,
+      "learning_rate": 8.237614987735607e-06,
+      "loss": 0.3708,
+      "step": 1668
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.872748317763673,
+      "learning_rate": 8.235412020278164e-06,
+      "loss": 0.397,
+      "step": 1669
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.271402719404538,
+      "learning_rate": 8.233207971809427e-06,
+      "loss": 0.3921,
+      "step": 1670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.329697166946847,
+      "learning_rate": 8.23100284306581e-06,
+      "loss": 0.3176,
+      "step": 1671
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.026001535157672,
+      "learning_rate": 8.228796634784086e-06,
+      "loss": 0.5127,
+      "step": 1672
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.380805016305548,
+      "learning_rate": 8.226589347701396e-06,
+      "loss": 0.3657,
+      "step": 1673
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 38.080808210850535,
+      "learning_rate": 8.224380982555226e-06,
+      "loss": 0.4443,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.3361844592776,
+      "learning_rate": 8.222171540083442e-06,
+      "loss": 0.5322,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.113199868710224,
+      "learning_rate": 8.219961021024251e-06,
+      "loss": 0.4336,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.666400780810209,
+      "learning_rate": 8.217749426116238e-06,
+      "loss": 0.5059,
+      "step": 1677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.936998665809853,
+      "learning_rate": 8.215536756098327e-06,
+      "loss": 0.4058,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.828959837607,
+      "learning_rate": 8.21332301170982e-06,
+      "loss": 0.4595,
+      "step": 1679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.325958282764173,
+      "learning_rate": 8.211108193690369e-06,
+      "loss": 0.4141,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.793987371835103,
+      "learning_rate": 8.208892302779982e-06,
+      "loss": 0.5151,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.843795834098652,
+      "learning_rate": 8.206675339719034e-06,
+      "loss": 0.4849,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.74955973168439,
+      "learning_rate": 8.204457305248253e-06,
+      "loss": 0.499,
+      "step": 1683
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.444677514125921,
+      "learning_rate": 8.202238200108721e-06,
+      "loss": 0.5122,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.718027155024345,
+      "learning_rate": 8.200018025041887e-06,
+      "loss": 0.501,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.98702758856593,
+      "learning_rate": 8.19779678078955e-06,
+      "loss": 0.4941,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.583545079481684,
+      "learning_rate": 8.195574468093872e-06,
+      "loss": 0.4937,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.339228146259074,
+      "learning_rate": 8.193351087697366e-06,
+      "loss": 0.4468,
+      "step": 1688
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.124375880053021,
+      "learning_rate": 8.191126640342906e-06,
+      "loss": 0.4336,
+      "step": 1689
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.822649187564727,
+      "learning_rate": 8.18890112677372e-06,
+      "loss": 0.4316,
+      "step": 1690
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.84834184460974,
+      "learning_rate": 8.186674547733398e-06,
+      "loss": 0.5522,
+      "step": 1691
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.928740644270853,
+      "learning_rate": 8.184446903965875e-06,
+      "loss": 0.4897,
+      "step": 1692
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.767092152016358,
+      "learning_rate": 8.182218196215452e-06,
+      "loss": 0.5,
+      "step": 1693
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.47958841486727,
+      "learning_rate": 8.17998842522678e-06,
+      "loss": 0.417,
+      "step": 1694
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.010687951365263,
+      "learning_rate": 8.17775759174487e-06,
+      "loss": 0.5488,
+      "step": 1695
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.742980213365875,
+      "learning_rate": 8.17552569651508e-06,
+      "loss": 0.4692,
+      "step": 1696
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.073207754471264,
+      "learning_rate": 8.173292740283135e-06,
+      "loss": 0.48,
+      "step": 1697
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.048458414659237,
+      "learning_rate": 8.171058723795097e-06,
+      "loss": 0.4868,
+      "step": 1698
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.883107102405994,
+      "learning_rate": 8.168823647797401e-06,
+      "loss": 0.5278,
+      "step": 1699
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.307579695529716,
+      "learning_rate": 8.166587513036826e-06,
+      "loss": 0.5342,
+      "step": 1700
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.099245148537669,
+      "learning_rate": 8.164350320260502e-06,
+      "loss": 0.3953,
+      "step": 1701
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 20.941909833565685,
+      "learning_rate": 8.16211207021592e-06,
+      "loss": 0.5112,
+      "step": 1702
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.820604772453076,
+      "learning_rate": 8.15987276365092e-06,
+      "loss": 0.4077,
+      "step": 1703
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.473817682607748,
+      "learning_rate": 8.157632401313696e-06,
+      "loss": 0.479,
+      "step": 1704
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.74411226039115,
+      "learning_rate": 8.155390983952795e-06,
+      "loss": 0.5112,
+      "step": 1705
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.227050478751384,
+      "learning_rate": 8.153148512317117e-06,
+      "loss": 0.3857,
+      "step": 1706
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.200446672036492,
+      "learning_rate": 8.150904987155911e-06,
+      "loss": 0.5029,
+      "step": 1707
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.4110528802431,
+      "learning_rate": 8.148660409218786e-06,
+      "loss": 0.5918,
+      "step": 1708
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.448382094255253,
+      "learning_rate": 8.146414779255689e-06,
+      "loss": 0.4707,
+      "step": 1709
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.909316780039679,
+      "learning_rate": 8.144168098016933e-06,
+      "loss": 0.4331,
+      "step": 1710
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.433238461814103,
+      "learning_rate": 8.141920366253173e-06,
+      "loss": 0.5918,
+      "step": 1711
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.771308440459347,
+      "learning_rate": 8.139671584715419e-06,
+      "loss": 0.5146,
+      "step": 1712
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.892728317342609,
+      "learning_rate": 8.137421754155031e-06,
+      "loss": 0.5664,
+      "step": 1713
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.318633500646037,
+      "learning_rate": 8.13517087532372e-06,
+      "loss": 0.4043,
+      "step": 1714
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.17077151093477,
+      "learning_rate": 8.132918948973543e-06,
+      "loss": 0.4736,
+      "step": 1715
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.996362868045267,
+      "learning_rate": 8.130665975856913e-06,
+      "loss": 0.5215,
+      "step": 1716
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.616605183075755,
+      "learning_rate": 8.128411956726592e-06,
+      "loss": 0.5415,
+      "step": 1717
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.643006691319318,
+      "learning_rate": 8.126156892335686e-06,
+      "loss": 0.4834,
+      "step": 1718
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.195645679300869,
+      "learning_rate": 8.123900783437655e-06,
+      "loss": 0.5327,
+      "step": 1719
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.905175950266107,
+      "learning_rate": 8.121643630786308e-06,
+      "loss": 0.5674,
+      "step": 1720
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.886407787375669,
+      "learning_rate": 8.1193854351358e-06,
+      "loss": 0.4756,
+      "step": 1721
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.66213829319637,
+      "learning_rate": 8.11712619724064e-06,
+      "loss": 0.4854,
+      "step": 1722
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.919219445746402,
+      "learning_rate": 8.114865917855676e-06,
+      "loss": 0.4976,
+      "step": 1723
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.518986982659731,
+      "learning_rate": 8.112604597736113e-06,
+      "loss": 0.4453,
+      "step": 1724
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.744337753421034,
+      "learning_rate": 8.110342237637501e-06,
+      "loss": 0.4009,
+      "step": 1725
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.907567080982792,
+      "learning_rate": 8.108078838315732e-06,
+      "loss": 0.4307,
+      "step": 1726
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.419935864909833,
+      "learning_rate": 8.105814400527052e-06,
+      "loss": 0.415,
+      "step": 1727
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.496398651006333,
+      "learning_rate": 8.103548925028054e-06,
+      "loss": 0.5171,
+      "step": 1728
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.775365355328299,
+      "learning_rate": 8.101282412575673e-06,
+      "loss": 0.4819,
+      "step": 1729
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.67720477465964,
+      "learning_rate": 8.099014863927192e-06,
+      "loss": 0.3726,
+      "step": 1730
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.790063693984372,
+      "learning_rate": 8.096746279840245e-06,
+      "loss": 0.4429,
+      "step": 1731
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.660047686678814,
+      "learning_rate": 8.094476661072806e-06,
+      "loss": 0.4814,
+      "step": 1732
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.596433199232035,
+      "learning_rate": 8.092206008383195e-06,
+      "loss": 0.5518,
+      "step": 1733
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.474511824070156,
+      "learning_rate": 8.089934322530082e-06,
+      "loss": 0.5264,
+      "step": 1734
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.071167596382951,
+      "learning_rate": 8.087661604272477e-06,
+      "loss": 0.5366,
+      "step": 1735
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.404014720227183,
+      "learning_rate": 8.08538785436974e-06,
+      "loss": 0.5059,
+      "step": 1736
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.235368526664493,
+      "learning_rate": 8.08311307358157e-06,
+      "loss": 0.5933,
+      "step": 1737
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.232707612320379,
+      "learning_rate": 8.080837262668017e-06,
+      "loss": 0.4341,
+      "step": 1738
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.513187841921981,
+      "learning_rate": 8.078560422389472e-06,
+      "loss": 0.4668,
+      "step": 1739
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.808341804249373,
+      "learning_rate": 8.076282553506664e-06,
+      "loss": 0.4595,
+      "step": 1740
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.722292004822773,
+      "learning_rate": 8.074003656780678e-06,
+      "loss": 0.5083,
+      "step": 1741
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.773723618455808,
+      "learning_rate": 8.071723732972933e-06,
+      "loss": 0.499,
+      "step": 1742
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.009490592834382,
+      "learning_rate": 8.069442782845191e-06,
+      "loss": 0.6309,
+      "step": 1743
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.228749331770816,
+      "learning_rate": 8.067160807159566e-06,
+      "loss": 0.5264,
+      "step": 1744
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.760608376818505,
+      "learning_rate": 8.064877806678504e-06,
+      "loss": 0.4751,
+      "step": 1745
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.86883350705853,
+      "learning_rate": 8.062593782164798e-06,
+      "loss": 0.5444,
+      "step": 1746
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.526705575487213,
+      "learning_rate": 8.060308734381585e-06,
+      "loss": 0.52,
+      "step": 1747
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.766194944671241,
+      "learning_rate": 8.05802266409234e-06,
+      "loss": 0.4355,
+      "step": 1748
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.32642846265708,
+      "learning_rate": 8.055735572060883e-06,
+      "loss": 0.4399,
+      "step": 1749
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.768520958766539,
+      "learning_rate": 8.053447459051374e-06,
+      "loss": 0.4658,
+      "step": 1750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.417625812427548,
+      "learning_rate": 8.051158325828315e-06,
+      "loss": 0.4585,
+      "step": 1751
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.241606084022609,
+      "learning_rate": 8.048868173156546e-06,
+      "loss": 0.4858,
+      "step": 1752
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.691430849772492,
+      "learning_rate": 8.046577001801248e-06,
+      "loss": 0.4819,
+      "step": 1753
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.879141711500264,
+      "learning_rate": 8.044284812527949e-06,
+      "loss": 0.5547,
+      "step": 1754
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.480942295074424,
+      "learning_rate": 8.041991606102507e-06,
+      "loss": 0.6348,
+      "step": 1755
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.43827233604463,
+      "learning_rate": 8.039697383291127e-06,
+      "loss": 0.4663,
+      "step": 1756
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.948768279992116,
+      "learning_rate": 8.037402144860353e-06,
+      "loss": 0.4976,
+      "step": 1757
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.369467383801608,
+      "learning_rate": 8.035105891577064e-06,
+      "loss": 0.4478,
+      "step": 1758
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.310600570034742,
+      "learning_rate": 8.032808624208485e-06,
+      "loss": 0.4702,
+      "step": 1759
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.0022124954022,
+      "learning_rate": 8.030510343522172e-06,
+      "loss": 0.4883,
+      "step": 1760
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.2828868419992,
+      "learning_rate": 8.02821105028602e-06,
+      "loss": 0.4268,
+      "step": 1761
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.03094557514591,
+      "learning_rate": 8.025910745268276e-06,
+      "loss": 0.5122,
+      "step": 1762
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.600050046217289,
+      "learning_rate": 8.023609429237504e-06,
+      "loss": 0.3862,
+      "step": 1763
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.4892743731139815,
+      "learning_rate": 8.021307102962623e-06,
+      "loss": 0.4751,
+      "step": 1764
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.988405551016548,
+      "learning_rate": 8.019003767212881e-06,
+      "loss": 0.4453,
+      "step": 1765
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.122334044245886,
+      "learning_rate": 8.016699422757865e-06,
+      "loss": 0.5967,
+      "step": 1766
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.211595252663454,
+      "learning_rate": 8.014394070367499e-06,
+      "loss": 0.5542,
+      "step": 1767
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.735010608127107,
+      "learning_rate": 8.012087710812047e-06,
+      "loss": 0.4873,
+      "step": 1768
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.903117170564997,
+      "learning_rate": 8.009780344862101e-06,
+      "loss": 0.5122,
+      "step": 1769
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.873539053988267,
+      "learning_rate": 8.0074719732886e-06,
+      "loss": 0.4502,
+      "step": 1770
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.28213540501501,
+      "learning_rate": 8.005162596862812e-06,
+      "loss": 0.4248,
+      "step": 1771
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 22.486287226040226,
+      "learning_rate": 8.002852216356343e-06,
+      "loss": 0.5264,
+      "step": 1772
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.996115895199278,
+      "learning_rate": 8.000540832541132e-06,
+      "loss": 0.4854,
+      "step": 1773
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.038573416312534,
+      "learning_rate": 7.99822844618946e-06,
+      "loss": 0.4604,
+      "step": 1774
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.159547013032094,
+      "learning_rate": 7.995915058073933e-06,
+      "loss": 0.5234,
+      "step": 1775
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.659676859984405,
+      "learning_rate": 7.9936006689675e-06,
+      "loss": 0.4878,
+      "step": 1776
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.634718738069445,
+      "learning_rate": 7.99128527964344e-06,
+      "loss": 0.5059,
+      "step": 1777
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.586703189973363,
+      "learning_rate": 7.988968890875368e-06,
+      "loss": 0.479,
+      "step": 1778
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.900105375633403,
+      "learning_rate": 7.986651503437233e-06,
+      "loss": 0.4824,
+      "step": 1779
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.814400149182877,
+      "learning_rate": 7.984333118103318e-06,
+      "loss": 0.4585,
+      "step": 1780
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.089228994356707,
+      "learning_rate": 7.982013735648235e-06,
+      "loss": 0.3984,
+      "step": 1781
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.909003750271049,
+      "learning_rate": 7.979693356846937e-06,
+      "loss": 0.6475,
+      "step": 1782
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.987506759661084,
+      "learning_rate": 7.977371982474705e-06,
+      "loss": 0.5854,
+      "step": 1783
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.953186291279971,
+      "learning_rate": 7.975049613307151e-06,
+      "loss": 0.5356,
+      "step": 1784
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.353919440549886,
+      "learning_rate": 7.972726250120225e-06,
+      "loss": 0.4849,
+      "step": 1785
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.610513184195263,
+      "learning_rate": 7.970401893690202e-06,
+      "loss": 0.5479,
+      "step": 1786
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.063317949232149,
+      "learning_rate": 7.968076544793696e-06,
+      "loss": 0.4551,
+      "step": 1787
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.971498295810411,
+      "learning_rate": 7.965750204207647e-06,
+      "loss": 0.418,
+      "step": 1788
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.315924369534915,
+      "learning_rate": 7.96342287270933e-06,
+      "loss": 0.4507,
+      "step": 1789
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.953940518033675,
+      "learning_rate": 7.96109455107635e-06,
+      "loss": 0.3809,
+      "step": 1790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.05948822429463,
+      "learning_rate": 7.958765240086639e-06,
+      "loss": 0.5605,
+      "step": 1791
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.700046157464467,
+      "learning_rate": 7.956434940518468e-06,
+      "loss": 0.4351,
+      "step": 1792
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.454966939327795,
+      "learning_rate": 7.954103653150432e-06,
+      "loss": 0.5576,
+      "step": 1793
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.619801280870984,
+      "learning_rate": 7.951771378761455e-06,
+      "loss": 0.418,
+      "step": 1794
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.520532675801153,
+      "learning_rate": 7.949438118130797e-06,
+      "loss": 0.6309,
+      "step": 1795
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.82962380647751,
+      "learning_rate": 7.94710387203804e-06,
+      "loss": 0.5132,
+      "step": 1796
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.229460325358946,
+      "learning_rate": 7.944768641263101e-06,
+      "loss": 0.438,
+      "step": 1797
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.18358063129885,
+      "learning_rate": 7.942432426586224e-06,
+      "loss": 0.479,
+      "step": 1798
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.038226449592317,
+      "learning_rate": 7.94009522878798e-06,
+      "loss": 0.458,
+      "step": 1799
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.23538850601863,
+      "learning_rate": 7.937757048649274e-06,
+      "loss": 0.5488,
+      "step": 1800
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.148183869442587,
+      "learning_rate": 7.935417886951332e-06,
+      "loss": 0.5342,
+      "step": 1801
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.068638115217762,
+      "learning_rate": 7.933077744475713e-06,
+      "loss": 0.4331,
+      "step": 1802
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.673804499921573,
+      "learning_rate": 7.930736622004301e-06,
+      "loss": 0.4766,
+      "step": 1803
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.854117836669452,
+      "learning_rate": 7.928394520319311e-06,
+      "loss": 0.5518,
+      "step": 1804
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.307530919938724,
+      "learning_rate": 7.926051440203278e-06,
+      "loss": 0.4248,
+      "step": 1805
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.980958338000624,
+      "learning_rate": 7.923707382439073e-06,
+      "loss": 0.436,
+      "step": 1806
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.178450156206857,
+      "learning_rate": 7.921362347809888e-06,
+      "loss": 0.4692,
+      "step": 1807
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.204168068608656,
+      "learning_rate": 7.919016337099242e-06,
+      "loss": 0.4541,
+      "step": 1808
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.081522614205248,
+      "learning_rate": 7.916669351090981e-06,
+      "loss": 0.4487,
+      "step": 1809
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.916600263727217,
+      "learning_rate": 7.914321390569278e-06,
+      "loss": 0.4282,
+      "step": 1810
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.60473063045003,
+      "learning_rate": 7.911972456318629e-06,
+      "loss": 0.437,
+      "step": 1811
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.650718167681877,
+      "learning_rate": 7.909622549123855e-06,
+      "loss": 0.4644,
+      "step": 1812
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.491205118287587,
+      "learning_rate": 7.907271669770107e-06,
+      "loss": 0.4053,
+      "step": 1813
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.07315325683419,
+      "learning_rate": 7.904919819042855e-06,
+      "loss": 0.4756,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.101956566078766,
+      "learning_rate": 7.902566997727896e-06,
+      "loss": 0.4263,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.528361439531825,
+      "learning_rate": 7.900213206611353e-06,
+      "loss": 0.4673,
+      "step": 1816
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.82005333902504,
+      "learning_rate": 7.897858446479672e-06,
+      "loss": 0.4355,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 24.150479129083248,
+      "learning_rate": 7.895502718119618e-06,
+      "loss": 0.5068,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.074857103355468,
+      "learning_rate": 7.89314602231829e-06,
+      "loss": 0.4043,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.774025110056453,
+      "learning_rate": 7.8907883598631e-06,
+      "loss": 0.4058,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.169418137670785,
+      "learning_rate": 7.888429731541784e-06,
+      "loss": 0.5068,
+      "step": 1821
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.638373472077571,
+      "learning_rate": 7.886070138142407e-06,
+      "loss": 0.4453,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.724664699868582,
+      "learning_rate": 7.883709580453354e-06,
+      "loss": 0.5068,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.422813863600243,
+      "learning_rate": 7.88134805926333e-06,
+      "loss": 0.4424,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.591815386989563,
+      "learning_rate": 7.878985575361362e-06,
+      "loss": 0.4644,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.579472910793811,
+      "learning_rate": 7.876622129536801e-06,
+      "loss": 0.4619,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.59809088506851,
+      "learning_rate": 7.874257722579319e-06,
+      "loss": 0.5039,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.124095864166325,
+      "learning_rate": 7.871892355278906e-06,
+      "loss": 0.4302,
+      "step": 1828
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.067465357679238,
+      "learning_rate": 7.869526028425878e-06,
+      "loss": 0.5254,
+      "step": 1829
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.94843976061978,
+      "learning_rate": 7.867158742810866e-06,
+      "loss": 0.4839,
+      "step": 1830
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.151059746712043,
+      "learning_rate": 7.864790499224825e-06,
+      "loss": 0.437,
+      "step": 1831
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.408199801575625,
+      "learning_rate": 7.86242129845903e-06,
+      "loss": 0.458,
+      "step": 1832
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.645230899000442,
+      "learning_rate": 7.860051141305074e-06,
+      "loss": 0.5107,
+      "step": 1833
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.561675138435755,
+      "learning_rate": 7.857680028554873e-06,
+      "loss": 0.4189,
+      "step": 1834
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.925281089125331,
+      "learning_rate": 7.855307961000656e-06,
+      "loss": 0.4893,
+      "step": 1835
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.26652117905419,
+      "learning_rate": 7.852934939434977e-06,
+      "loss": 0.4399,
+      "step": 1836
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.505458999460473,
+      "learning_rate": 7.850560964650707e-06,
+      "loss": 0.6421,
+      "step": 1837
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.372565753801249,
+      "learning_rate": 7.848186037441035e-06,
+      "loss": 0.6289,
+      "step": 1838
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.881114480690239,
+      "learning_rate": 7.845810158599467e-06,
+      "loss": 0.4624,
+      "step": 1839
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.626691743280656,
+      "learning_rate": 7.84343332891983e-06,
+      "loss": 0.4785,
+      "step": 1840
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.687233987923879,
+      "learning_rate": 7.841055549196267e-06,
+      "loss": 0.521,
+      "step": 1841
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.195815391563308,
+      "learning_rate": 7.838676820223234e-06,
+      "loss": 0.3867,
+      "step": 1842
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.299770030769283,
+      "learning_rate": 7.836297142795515e-06,
+      "loss": 0.4546,
+      "step": 1843
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.8699709812237515,
+      "learning_rate": 7.833916517708203e-06,
+      "loss": 0.3662,
+      "step": 1844
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.559884344433186,
+      "learning_rate": 7.831534945756703e-06,
+      "loss": 0.54,
+      "step": 1845
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.525388260025515,
+      "learning_rate": 7.82915242773675e-06,
+      "loss": 0.4473,
+      "step": 1846
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.9216506801838,
+      "learning_rate": 7.826768964444384e-06,
+      "loss": 0.4189,
+      "step": 1847
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.147593629165257,
+      "learning_rate": 7.824384556675966e-06,
+      "loss": 0.5132,
+      "step": 1848
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.171992879631954,
+      "learning_rate": 7.821999205228168e-06,
+      "loss": 0.4893,
+      "step": 1849
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.79619518512073,
+      "learning_rate": 7.819612910897985e-06,
+      "loss": 0.541,
+      "step": 1850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.026117401917226,
+      "learning_rate": 7.817225674482717e-06,
+      "loss": 0.5181,
+      "step": 1851
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.809802504536831,
+      "learning_rate": 7.814837496779988e-06,
+      "loss": 0.3872,
+      "step": 1852
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.661493775692145,
+      "learning_rate": 7.812448378587731e-06,
+      "loss": 0.4951,
+      "step": 1853
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.532780424040105,
+      "learning_rate": 7.810058320704194e-06,
+      "loss": 0.5244,
+      "step": 1854
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.622050906657064,
+      "learning_rate": 7.807667323927941e-06,
+      "loss": 0.3711,
+      "step": 1855
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.137416584677046,
+      "learning_rate": 7.80527538905785e-06,
+      "loss": 0.4214,
+      "step": 1856
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.531652456122275,
+      "learning_rate": 7.802882516893106e-06,
+      "loss": 0.3865,
+      "step": 1857
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.761181122860886,
+      "learning_rate": 7.800488708233219e-06,
+      "loss": 0.458,
+      "step": 1858
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.986448218756069,
+      "learning_rate": 7.798093963877998e-06,
+      "loss": 0.3965,
+      "step": 1859
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.525624462916241,
+      "learning_rate": 7.795698284627575e-06,
+      "loss": 0.5312,
+      "step": 1860
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.56296095061409,
+      "learning_rate": 7.793301671282391e-06,
+      "loss": 0.4253,
+      "step": 1861
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.813705657660236,
+      "learning_rate": 7.7909041246432e-06,
+      "loss": 0.4346,
+      "step": 1862
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.044009444296448,
+      "learning_rate": 7.788505645511065e-06,
+      "loss": 0.4697,
+      "step": 1863
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.756948527676359,
+      "learning_rate": 7.786106234687362e-06,
+      "loss": 0.4351,
+      "step": 1864
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.838558191606422,
+      "learning_rate": 7.783705892973782e-06,
+      "loss": 0.5356,
+      "step": 1865
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.97549632585959,
+      "learning_rate": 7.78130462117232e-06,
+      "loss": 0.4238,
+      "step": 1866
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.230138714195853,
+      "learning_rate": 7.778902420085289e-06,
+      "loss": 0.4741,
+      "step": 1867
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.33800229572185,
+      "learning_rate": 7.776499290515304e-06,
+      "loss": 0.3765,
+      "step": 1868
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.435836369297808,
+      "learning_rate": 7.7740952332653e-06,
+      "loss": 0.4629,
+      "step": 1869
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.644876077277099,
+      "learning_rate": 7.771690249138517e-06,
+      "loss": 0.4565,
+      "step": 1870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.755377787104848,
+      "learning_rate": 7.769284338938502e-06,
+      "loss": 0.4541,
+      "step": 1871
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.27972086793472,
+      "learning_rate": 7.766877503469117e-06,
+      "loss": 0.5132,
+      "step": 1872
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.760972849630273,
+      "learning_rate": 7.764469743534529e-06,
+      "loss": 0.4404,
+      "step": 1873
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.940065164124238,
+      "learning_rate": 7.762061059939214e-06,
+      "loss": 0.4536,
+      "step": 1874
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.751090211455104,
+      "learning_rate": 7.759651453487963e-06,
+      "loss": 0.3506,
+      "step": 1875
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.25806719313458,
+      "learning_rate": 7.757240924985866e-06,
+      "loss": 0.6182,
+      "step": 1876
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.896852124606287,
+      "learning_rate": 7.754829475238323e-06,
+      "loss": 0.4888,
+      "step": 1877
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.942142350442522,
+      "learning_rate": 7.752417105051051e-06,
+      "loss": 0.5947,
+      "step": 1878
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.63242801319795,
+      "learning_rate": 7.750003815230062e-06,
+      "loss": 0.4194,
+      "step": 1879
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.915779656542515,
+      "learning_rate": 7.747589606581686e-06,
+      "loss": 0.3735,
+      "step": 1880
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.6414314110252075,
+      "learning_rate": 7.745174479912551e-06,
+      "loss": 0.4844,
+      "step": 1881
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.279768255216817,
+      "learning_rate": 7.742758436029596e-06,
+      "loss": 0.4561,
+      "step": 1882
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.118475626162798,
+      "learning_rate": 7.740341475740068e-06,
+      "loss": 0.4819,
+      "step": 1883
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.986938122701325,
+      "learning_rate": 7.737923599851519e-06,
+      "loss": 0.4072,
+      "step": 1884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.3213451137330825,
+      "learning_rate": 7.735504809171801e-06,
+      "loss": 0.4595,
+      "step": 1885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.386651252542793,
+      "learning_rate": 7.733085104509084e-06,
+      "loss": 0.4014,
+      "step": 1886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.548276619740427,
+      "learning_rate": 7.730664486671831e-06,
+      "loss": 0.4585,
+      "step": 1887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.158689999749116,
+      "learning_rate": 7.72824295646882e-06,
+      "loss": 0.3936,
+      "step": 1888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.499012467824802,
+      "learning_rate": 7.725820514709124e-06,
+      "loss": 0.4233,
+      "step": 1889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.070340859207697,
+      "learning_rate": 7.723397162202128e-06,
+      "loss": 0.5679,
+      "step": 1890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.183380030500457,
+      "learning_rate": 7.720972899757522e-06,
+      "loss": 0.4019,
+      "step": 1891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.63336525431284,
+      "learning_rate": 7.718547728185293e-06,
+      "loss": 0.5845,
+      "step": 1892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.930649239400196,
+      "learning_rate": 7.716121648295738e-06,
+      "loss": 0.4297,
+      "step": 1893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.936977513469789,
+      "learning_rate": 7.713694660899455e-06,
+      "loss": 0.3965,
+      "step": 1894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.880197223026865,
+      "learning_rate": 7.711266766807345e-06,
+      "loss": 0.478,
+      "step": 1895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.38985775819577,
+      "learning_rate": 7.708837966830615e-06,
+      "loss": 0.5176,
+      "step": 1896
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.073264625514588,
+      "learning_rate": 7.706408261780769e-06,
+      "loss": 0.4155,
+      "step": 1897
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.734620428534315,
+      "learning_rate": 7.703977652469618e-06,
+      "loss": 0.4585,
+      "step": 1898
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.71268865147328,
+      "learning_rate": 7.701546139709272e-06,
+      "loss": 0.4351,
+      "step": 1899
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.690569659507187,
+      "learning_rate": 7.69911372431215e-06,
+      "loss": 0.4819,
+      "step": 1900
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.038069885853988,
+      "learning_rate": 7.696680407090962e-06,
+      "loss": 0.4741,
+      "step": 1901
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.166802662393746,
+      "learning_rate": 7.694246188858726e-06,
+      "loss": 0.4458,
+      "step": 1902
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.652531412656241,
+      "learning_rate": 7.691811070428758e-06,
+      "loss": 0.4409,
+      "step": 1903
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.775786518184024,
+      "learning_rate": 7.689375052614681e-06,
+      "loss": 0.5034,
+      "step": 1904
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.730043609598852,
+      "learning_rate": 7.686938136230408e-06,
+      "loss": 0.4878,
+      "step": 1905
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.070741977627472,
+      "learning_rate": 7.684500322090162e-06,
+      "loss": 0.4946,
+      "step": 1906
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.66702912722839,
+      "learning_rate": 7.68206161100846e-06,
+      "loss": 0.4512,
+      "step": 1907
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.18598427399989,
+      "learning_rate": 7.679622003800122e-06,
+      "loss": 0.5322,
+      "step": 1908
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 18.592185583152887,
+      "learning_rate": 7.677181501280266e-06,
+      "loss": 0.384,
+      "step": 1909
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.483647129362556,
+      "learning_rate": 7.674740104264308e-06,
+      "loss": 0.4438,
+      "step": 1910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.56330359346813,
+      "learning_rate": 7.672297813567968e-06,
+      "loss": 0.6631,
+      "step": 1911
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.750687177326649,
+      "learning_rate": 7.669854630007257e-06,
+      "loss": 0.4619,
+      "step": 1912
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.801460674280925,
+      "learning_rate": 7.667410554398486e-06,
+      "loss": 0.5146,
+      "step": 1913
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.654003604266663,
+      "learning_rate": 7.664965587558271e-06,
+      "loss": 0.5215,
+      "step": 1914
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.252336343356576,
+      "learning_rate": 7.662519730303517e-06,
+      "loss": 0.4146,
+      "step": 1915
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.96405723828609,
+      "learning_rate": 7.660072983451433e-06,
+      "loss": 0.481,
+      "step": 1916
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.053400748038461,
+      "learning_rate": 7.657625347819522e-06,
+      "loss": 0.4648,
+      "step": 1917
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.665072797465664,
+      "learning_rate": 7.655176824225582e-06,
+      "loss": 0.4395,
+      "step": 1918
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.43877602912696,
+      "learning_rate": 7.652727413487716e-06,
+      "loss": 0.5225,
+      "step": 1919
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.53955384603415,
+      "learning_rate": 7.650277116424313e-06,
+      "loss": 0.4473,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.42532980284874,
+      "learning_rate": 7.647825933854063e-06,
+      "loss": 0.5376,
+      "step": 1921
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.36636988005007,
+      "learning_rate": 7.645373866595953e-06,
+      "loss": 0.4521,
+      "step": 1922
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.525600543992109,
+      "learning_rate": 7.642920915469265e-06,
+      "loss": 0.4688,
+      "step": 1923
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.106931536030515,
+      "learning_rate": 7.640467081293573e-06,
+      "loss": 0.5918,
+      "step": 1924
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.348679176976812,
+      "learning_rate": 7.638012364888751e-06,
+      "loss": 0.4907,
+      "step": 1925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.55252362508795,
+      "learning_rate": 7.635556767074965e-06,
+      "loss": 0.4731,
+      "step": 1926
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.419504634208726,
+      "learning_rate": 7.633100288672674e-06,
+      "loss": 0.481,
+      "step": 1927
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.121962467884394,
+      "learning_rate": 7.630642930502634e-06,
+      "loss": 0.418,
+      "step": 1928
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.289195654821805,
+      "learning_rate": 7.628184693385896e-06,
+      "loss": 0.4648,
+      "step": 1929
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.9227069619546375,
+      "learning_rate": 7.625725578143801e-06,
+      "loss": 0.4712,
+      "step": 1930
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.229125735331845,
+      "learning_rate": 7.6232655855979844e-06,
+      "loss": 0.5029,
+      "step": 1931
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.049117828200913,
+      "learning_rate": 7.620804716570376e-06,
+      "loss": 0.4653,
+      "step": 1932
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.576958024182636,
+      "learning_rate": 7.618342971883199e-06,
+      "loss": 0.5605,
+      "step": 1933
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.779629505768929,
+      "learning_rate": 7.615880352358967e-06,
+      "loss": 0.4014,
+      "step": 1934
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.50645646388632,
+      "learning_rate": 7.613416858820486e-06,
+      "loss": 0.4214,
+      "step": 1935
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.52806005097817,
+      "learning_rate": 7.6109524920908575e-06,
+      "loss": 0.5288,
+      "step": 1936
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.933316977057585,
+      "learning_rate": 7.608487252993471e-06,
+      "loss": 0.5151,
+      "step": 1937
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.6592436046183,
+      "learning_rate": 7.6060211423520095e-06,
+      "loss": 0.4072,
+      "step": 1938
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.814161825521396,
+      "learning_rate": 7.6035541609904425e-06,
+      "loss": 0.606,
+      "step": 1939
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.66882526072226,
+      "learning_rate": 7.60108630973304e-06,
+      "loss": 0.458,
+      "step": 1940
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.85672505413872,
+      "learning_rate": 7.598617589404354e-06,
+      "loss": 0.4565,
+      "step": 1941
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.858479644628172,
+      "learning_rate": 7.596148000829229e-06,
+      "loss": 0.6187,
+      "step": 1942
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.16162358033872,
+      "learning_rate": 7.593677544832802e-06,
+      "loss": 0.5293,
+      "step": 1943
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.524117704758424,
+      "learning_rate": 7.5912062222404965e-06,
+      "loss": 0.3533,
+      "step": 1944
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.818889018640792,
+      "learning_rate": 7.588734033878031e-06,
+      "loss": 0.5693,
+      "step": 1945
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.522766326382442,
+      "learning_rate": 7.586260980571407e-06,
+      "loss": 0.5547,
+      "step": 1946
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.564225336554951,
+      "learning_rate": 7.5837870631469165e-06,
+      "loss": 0.6562,
+      "step": 1947
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.668952257315175,
+      "learning_rate": 7.581312282431143e-06,
+      "loss": 0.3936,
+      "step": 1948
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.336521625375687,
+      "learning_rate": 7.578836639250958e-06,
+      "loss": 0.5151,
+      "step": 1949
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.918797922477602,
+      "learning_rate": 7.576360134433517e-06,
+      "loss": 0.4668,
+      "step": 1950
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.6676020443715975,
+      "learning_rate": 7.5738827688062676e-06,
+      "loss": 0.4961,
+      "step": 1951
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.713872938147478,
+      "learning_rate": 7.571404543196943e-06,
+      "loss": 0.4824,
+      "step": 1952
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.834704519149373,
+      "learning_rate": 7.568925458433567e-06,
+      "loss": 0.4819,
+      "step": 1953
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.91835327063667,
+      "learning_rate": 7.566445515344445e-06,
+      "loss": 0.4233,
+      "step": 1954
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.599974116531055,
+      "learning_rate": 7.563964714758172e-06,
+      "loss": 0.4458,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.201329272152451,
+      "learning_rate": 7.561483057503632e-06,
+      "loss": 0.5386,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.350442516482657,
+      "learning_rate": 7.559000544409991e-06,
+      "loss": 0.4849,
+      "step": 1957
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.802291496242244,
+      "learning_rate": 7.556517176306704e-06,
+      "loss": 0.5127,
+      "step": 1958
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.032857918894152,
+      "learning_rate": 7.554032954023508e-06,
+      "loss": 0.479,
+      "step": 1959
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.170707738509645,
+      "learning_rate": 7.55154787839043e-06,
+      "loss": 0.5752,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.33356640527272,
+      "learning_rate": 7.5490619502377805e-06,
+      "loss": 0.562,
+      "step": 1961
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.3400026057796,
+      "learning_rate": 7.546575170396153e-06,
+      "loss": 0.4961,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.41020505136803,
+      "learning_rate": 7.544087539696427e-06,
+      "loss": 0.5356,
+      "step": 1963
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.384398968489197,
+      "learning_rate": 7.541599058969766e-06,
+      "loss": 0.5757,
+      "step": 1964
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.85010786013174,
+      "learning_rate": 7.539109729047619e-06,
+      "loss": 0.5869,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.678164293750017,
+      "learning_rate": 7.5366195507617155e-06,
+      "loss": 0.5024,
+      "step": 1966
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.760310761103758,
+      "learning_rate": 7.534128524944071e-06,
+      "loss": 0.4077,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.324608405481133,
+      "learning_rate": 7.531636652426985e-06,
+      "loss": 0.4927,
+      "step": 1968
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.427863874423563,
+      "learning_rate": 7.529143934043036e-06,
+      "loss": 0.562,
+      "step": 1969
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.721667941829319,
+      "learning_rate": 7.526650370625088e-06,
+      "loss": 0.6045,
+      "step": 1970
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.373487140290482,
+      "learning_rate": 7.5241559630062896e-06,
+      "loss": 0.5576,
+      "step": 1971
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.699156015321359,
+      "learning_rate": 7.5216607120200655e-06,
+      "loss": 0.4937,
+      "step": 1972
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.49661605079921,
+      "learning_rate": 7.519164618500127e-06,
+      "loss": 0.4209,
+      "step": 1973
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.772198775491923,
+      "learning_rate": 7.5166676832804655e-06,
+      "loss": 0.4731,
+      "step": 1974
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.746747420450404,
+      "learning_rate": 7.514169907195352e-06,
+      "loss": 0.457,
+      "step": 1975
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.152389759194886,
+      "learning_rate": 7.511671291079342e-06,
+      "loss": 0.3955,
+      "step": 1976
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.057173209748903,
+      "learning_rate": 7.509171835767268e-06,
+      "loss": 0.4814,
+      "step": 1977
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.473228863127407,
+      "learning_rate": 7.506671542094246e-06,
+      "loss": 0.4976,
+      "step": 1978
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.727530037186542,
+      "learning_rate": 7.504170410895668e-06,
+      "loss": 0.4204,
+      "step": 1979
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.985145065845938,
+      "learning_rate": 7.501668443007212e-06,
+      "loss": 0.4468,
+      "step": 1980
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.866585730814796,
+      "learning_rate": 7.499165639264828e-06,
+      "loss": 0.5054,
+      "step": 1981
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.54865900787404,
+      "learning_rate": 7.496662000504752e-06,
+      "loss": 0.5615,
+      "step": 1982
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.596067173111546,
+      "learning_rate": 7.4941575275634945e-06,
+      "loss": 0.4961,
+      "step": 1983
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.522213226674722,
+      "learning_rate": 7.49165222127785e-06,
+      "loss": 0.4829,
+      "step": 1984
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.827635155968103,
+      "learning_rate": 7.489146082484882e-06,
+      "loss": 0.3564,
+      "step": 1985
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.292648628242718,
+      "learning_rate": 7.486639112021944e-06,
+      "loss": 0.4268,
+      "step": 1986
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.52207584728679,
+      "learning_rate": 7.484131310726658e-06,
+      "loss": 0.3796,
+      "step": 1987
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.540889479395993,
+      "learning_rate": 7.481622679436929e-06,
+      "loss": 0.4463,
+      "step": 1988
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.775521996498332,
+      "learning_rate": 7.479113218990934e-06,
+      "loss": 0.3867,
+      "step": 1989
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.41497161064518,
+      "learning_rate": 7.4766029302271335e-06,
+      "loss": 0.5225,
+      "step": 1990
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.663185183957201,
+      "learning_rate": 7.474091813984261e-06,
+      "loss": 0.3784,
+      "step": 1991
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.438508094185057,
+      "learning_rate": 7.471579871101326e-06,
+      "loss": 0.4409,
+      "step": 1992
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.101689467991465,
+      "learning_rate": 7.4690671024176165e-06,
+      "loss": 0.5542,
+      "step": 1993
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.147337868681644,
+      "learning_rate": 7.466553508772695e-06,
+      "loss": 0.4072,
+      "step": 1994
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.211891837722657,
+      "learning_rate": 7.4640390910064e-06,
+      "loss": 0.4326,
+      "step": 1995
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.383061627960748,
+      "learning_rate": 7.461523849958845e-06,
+      "loss": 0.5654,
+      "step": 1996
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.337379991199462,
+      "learning_rate": 7.459007786470418e-06,
+      "loss": 0.3994,
+      "step": 1997
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.488871378983164,
+      "learning_rate": 7.4564909013817845e-06,
+      "loss": 0.501,
+      "step": 1998
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.83370126958352,
+      "learning_rate": 7.45397319553388e-06,
+      "loss": 0.354,
+      "step": 1999
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.89565539397935,
+      "learning_rate": 7.451454669767919e-06,
+      "loss": 0.4561,
+      "step": 2000
+    },
+    {
+      "epoch": 1.43,
+      "eval_avg_AUC": 0.7710454060906093,
+      "eval_avg_Accuracy": 0.6791694297082228,
+      "eval_avg_Accuracy-right": 0.9080474761966871,
+      "eval_avg_Accuracy-wrong": 0.2800773254491699,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.626217598019666,
+      "eval_last_AUC": 0.7805472536484926,
+      "eval_last_Accuracy": 0.705238726790451,
+      "eval_last_Accuracy-right": 0.851506456241033,
+      "eval_last_Accuracy-wrong": 0.4501933136229247,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6337350556153214,
+      "eval_max_AUC": 0.7253894400592504,
+      "eval_max_Accuracy": 0.6378481432360743,
+      "eval_max_Accuracy-right": 0.9711099517412286,
+      "eval_max_Accuracy-wrong": 0.056743234023197635,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6102256988583293,
+      "eval_min_AUC": 0.7744442531826913,
+      "eval_min_Accuracy": 0.707145225464191,
+      "eval_min_Accuracy-right": 0.7947697926177123,
+      "eval_min_Accuracy-wrong": 0.5543552422105982,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6233897492201612,
+      "eval_prod_AUC": 0.7756218453367024,
+      "eval_prod_Accuracy": 0.6703000663129973,
+      "eval_prod_Accuracy-right": 0.5567366636233207,
+      "eval_prod_Accuracy-wrong": 0.8683193086195133,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6229021173701285,
+      "eval_runtime": 246.9142,
+      "eval_samples_per_second": 97.718,
+      "eval_steps_per_second": 3.054,
+      "eval_sum_AUC": 0.6216167742931785,
+      "eval_sum_Accuracy": 0.6359001989389921,
+      "eval_sum_Accuracy-right": 0.9986304943263337,
+      "eval_sum_Accuracy-wrong": 0.0034114168751421424,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6193609469118248,
+      "step": 2000
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.915881897626125,
+      "learning_rate": 7.448935324925386e-06,
+      "loss": 0.3965,
+      "step": 2001
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.292176761992444,
+      "learning_rate": 7.446415161848043e-06,
+      "loss": 0.4697,
+      "step": 2002
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.41157340253059,
+      "learning_rate": 7.443894181377921e-06,
+      "loss": 0.4683,
+      "step": 2003
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.703353976942797,
+      "learning_rate": 7.441372384357328e-06,
+      "loss": 0.4531,
+      "step": 2004
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.728696147279525,
+      "learning_rate": 7.438849771628844e-06,
+      "loss": 0.4814,
+      "step": 2005
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.964762875434154,
+      "learning_rate": 7.43632634403532e-06,
+      "loss": 0.4409,
+      "step": 2006
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.767032870416173,
+      "learning_rate": 7.433802102419878e-06,
+      "loss": 0.4531,
+      "step": 2007
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.973473512937122,
+      "learning_rate": 7.431277047625918e-06,
+      "loss": 0.4146,
+      "step": 2008
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.382184093999804,
+      "learning_rate": 7.428751180497104e-06,
+      "loss": 0.5166,
+      "step": 2009
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.177950650233484,
+      "learning_rate": 7.426224501877376e-06,
+      "loss": 0.5278,
+      "step": 2010
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.821415821919816,
+      "learning_rate": 7.423697012610947e-06,
+      "loss": 0.4058,
+      "step": 2011
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.12631349817775,
+      "learning_rate": 7.421168713542294e-06,
+      "loss": 0.4873,
+      "step": 2012
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.189855739113973,
+      "learning_rate": 7.418639605516172e-06,
+      "loss": 0.5205,
+      "step": 2013
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.654050014113183,
+      "learning_rate": 7.416109689377603e-06,
+      "loss": 0.5635,
+      "step": 2014
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.246721169792306,
+      "learning_rate": 7.413578965971876e-06,
+      "loss": 0.4805,
+      "step": 2015
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.730269042581263,
+      "learning_rate": 7.411047436144556e-06,
+      "loss": 0.4165,
+      "step": 2016
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.059699400483384,
+      "learning_rate": 7.408515100741471e-06,
+      "loss": 0.4656,
+      "step": 2017
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.604554128594822,
+      "learning_rate": 7.405981960608725e-06,
+      "loss": 0.4453,
+      "step": 2018
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.398182516011172,
+      "learning_rate": 7.403448016592685e-06,
+      "loss": 0.4951,
+      "step": 2019
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.429871973957303,
+      "learning_rate": 7.400913269539988e-06,
+      "loss": 0.5908,
+      "step": 2020
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.90665542864224,
+      "learning_rate": 7.398377720297541e-06,
+      "loss": 0.4917,
+      "step": 2021
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.435262236737888,
+      "learning_rate": 7.39584136971252e-06,
+      "loss": 0.4473,
+      "step": 2022
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.241789540294814,
+      "learning_rate": 7.393304218632364e-06,
+      "loss": 0.4507,
+      "step": 2023
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.58272231581781,
+      "learning_rate": 7.390766267904783e-06,
+      "loss": 0.3755,
+      "step": 2024
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.95465858537594,
+      "learning_rate": 7.3882275183777554e-06,
+      "loss": 0.5312,
+      "step": 2025
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.639047246847788,
+      "learning_rate": 7.385687970899523e-06,
+      "loss": 0.5078,
+      "step": 2026
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.953213267741265,
+      "learning_rate": 7.3831476263185965e-06,
+      "loss": 0.4751,
+      "step": 2027
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.322106542062425,
+      "learning_rate": 7.380606485483751e-06,
+      "loss": 0.4009,
+      "step": 2028
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 43.45347427250669,
+      "learning_rate": 7.378064549244031e-06,
+      "loss": 0.5298,
+      "step": 2029
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.42523425392354,
+      "learning_rate": 7.375521818448741e-06,
+      "loss": 0.5371,
+      "step": 2030
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 55.77740354141135,
+      "learning_rate": 7.372978293947459e-06,
+      "loss": 0.4761,
+      "step": 2031
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 49.09132676927039,
+      "learning_rate": 7.3704339765900205e-06,
+      "loss": 0.5615,
+      "step": 2032
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 34.9256636465853,
+      "learning_rate": 7.367888867226531e-06,
+      "loss": 0.4976,
+      "step": 2033
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.312677064646966,
+      "learning_rate": 7.365342966707359e-06,
+      "loss": 0.4487,
+      "step": 2034
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.813978743326054,
+      "learning_rate": 7.362796275883135e-06,
+      "loss": 0.4038,
+      "step": 2035
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.643306134854747,
+      "learning_rate": 7.360248795604758e-06,
+      "loss": 0.4575,
+      "step": 2036
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.868389092088139,
+      "learning_rate": 7.3577005267233885e-06,
+      "loss": 0.429,
+      "step": 2037
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.982700342237198,
+      "learning_rate": 7.355151470090449e-06,
+      "loss": 0.4712,
+      "step": 2038
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.034999940641313,
+      "learning_rate": 7.352601626557628e-06,
+      "loss": 0.4663,
+      "step": 2039
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.88549792075512,
+      "learning_rate": 7.350050996976875e-06,
+      "loss": 0.5669,
+      "step": 2040
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.208945274018559,
+      "learning_rate": 7.347499582200404e-06,
+      "loss": 0.4097,
+      "step": 2041
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.404862265993508,
+      "learning_rate": 7.344947383080687e-06,
+      "loss": 0.4634,
+      "step": 2042
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.366856522102745,
+      "learning_rate": 7.342394400470463e-06,
+      "loss": 0.3926,
+      "step": 2043
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.362697922089307,
+      "learning_rate": 7.339840635222732e-06,
+      "loss": 0.4116,
+      "step": 2044
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.008373884805586,
+      "learning_rate": 7.337286088190754e-06,
+      "loss": 0.4399,
+      "step": 2045
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.955581395387284,
+      "learning_rate": 7.334730760228049e-06,
+      "loss": 0.5054,
+      "step": 2046
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.736021901140624,
+      "learning_rate": 7.332174652188401e-06,
+      "loss": 0.5195,
+      "step": 2047
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.418786675988692,
+      "learning_rate": 7.329617764925853e-06,
+      "loss": 0.5605,
+      "step": 2048
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.156432856945289,
+      "learning_rate": 7.32706009929471e-06,
+      "loss": 0.5474,
+      "step": 2049
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.360135291080843,
+      "learning_rate": 7.324501656149532e-06,
+      "loss": 0.4775,
+      "step": 2050
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.446127423689205,
+      "learning_rate": 7.321942436345146e-06,
+      "loss": 0.324,
+      "step": 2051
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.489691101046363,
+      "learning_rate": 7.319382440736632e-06,
+      "loss": 0.5054,
+      "step": 2052
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.514088999847083,
+      "learning_rate": 7.316821670179335e-06,
+      "loss": 0.4668,
+      "step": 2053
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.159539585109048,
+      "learning_rate": 7.314260125528854e-06,
+      "loss": 0.6167,
+      "step": 2054
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 25.094579873479073,
+      "learning_rate": 7.311697807641048e-06,
+      "loss": 0.3867,
+      "step": 2055
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.846673523885213,
+      "learning_rate": 7.3091347173720386e-06,
+      "loss": 0.4932,
+      "step": 2056
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.2914105450018125,
+      "learning_rate": 7.3065708555781986e-06,
+      "loss": 0.4727,
+      "step": 2057
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.595181627149866,
+      "learning_rate": 7.304006223116162e-06,
+      "loss": 0.5098,
+      "step": 2058
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.44544649111758,
+      "learning_rate": 7.301440820842822e-06,
+      "loss": 0.4707,
+      "step": 2059
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.018485991705333,
+      "learning_rate": 7.298874649615327e-06,
+      "loss": 0.6582,
+      "step": 2060
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.661937937304063,
+      "learning_rate": 7.29630771029108e-06,
+      "loss": 0.3862,
+      "step": 2061
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.548593437781802,
+      "learning_rate": 7.293740003727745e-06,
+      "loss": 0.4663,
+      "step": 2062
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.2212974848326,
+      "learning_rate": 7.291171530783241e-06,
+      "loss": 0.4976,
+      "step": 2063
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.235414814450403,
+      "learning_rate": 7.288602292315742e-06,
+      "loss": 0.5596,
+      "step": 2064
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.313344290622826,
+      "learning_rate": 7.286032289183679e-06,
+      "loss": 0.5317,
+      "step": 2065
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.94697386715504,
+      "learning_rate": 7.283461522245736e-06,
+      "loss": 0.5137,
+      "step": 2066
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.081682104162239,
+      "learning_rate": 7.280889992360856e-06,
+      "loss": 0.5156,
+      "step": 2067
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.742609851819826,
+      "learning_rate": 7.278317700388232e-06,
+      "loss": 0.4551,
+      "step": 2068
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.514579824634136,
+      "learning_rate": 7.275744647187318e-06,
+      "loss": 0.5801,
+      "step": 2069
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.69700237796884,
+      "learning_rate": 7.273170833617818e-06,
+      "loss": 0.4678,
+      "step": 2070
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.058671415569465,
+      "learning_rate": 7.2705962605396895e-06,
+      "loss": 0.521,
+      "step": 2071
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.321934672500923,
+      "learning_rate": 7.268020928813147e-06,
+      "loss": 0.4917,
+      "step": 2072
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.94345204999045,
+      "learning_rate": 7.265444839298656e-06,
+      "loss": 0.4526,
+      "step": 2073
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.541712902362422,
+      "learning_rate": 7.262867992856934e-06,
+      "loss": 0.4409,
+      "step": 2074
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.3187862663539,
+      "learning_rate": 7.260290390348956e-06,
+      "loss": 0.5542,
+      "step": 2075
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.647404924361794,
+      "learning_rate": 7.257712032635946e-06,
+      "loss": 0.5205,
+      "step": 2076
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.207651531426404,
+      "learning_rate": 7.255132920579382e-06,
+      "loss": 0.4556,
+      "step": 2077
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.087482473274372,
+      "learning_rate": 7.252553055040991e-06,
+      "loss": 0.4121,
+      "step": 2078
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 25.25159930562991,
+      "learning_rate": 7.249972436882756e-06,
+      "loss": 0.5176,
+      "step": 2079
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.702457354457117,
+      "learning_rate": 7.247391066966909e-06,
+      "loss": 0.4609,
+      "step": 2080
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.603868415722688,
+      "learning_rate": 7.244808946155933e-06,
+      "loss": 0.438,
+      "step": 2081
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.349525316850295,
+      "learning_rate": 7.242226075312564e-06,
+      "loss": 0.437,
+      "step": 2082
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.177336362256062,
+      "learning_rate": 7.239642455299787e-06,
+      "loss": 0.4385,
+      "step": 2083
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.495636394435223,
+      "learning_rate": 7.237058086980835e-06,
+      "loss": 0.6406,
+      "step": 2084
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.827798850107106,
+      "learning_rate": 7.234472971219197e-06,
+      "loss": 0.4165,
+      "step": 2085
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.754753037992685,
+      "learning_rate": 7.231887108878606e-06,
+      "loss": 0.5586,
+      "step": 2086
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.180399436654513,
+      "learning_rate": 7.229300500823047e-06,
+      "loss": 0.4966,
+      "step": 2087
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.98964058960827,
+      "learning_rate": 7.226713147916754e-06,
+      "loss": 0.4678,
+      "step": 2088
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.21132602807958,
+      "learning_rate": 7.22412505102421e-06,
+      "loss": 0.4282,
+      "step": 2089
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.175190157330336,
+      "learning_rate": 7.221536211010147e-06,
+      "loss": 0.3574,
+      "step": 2090
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.902417216507908,
+      "learning_rate": 7.2189466287395425e-06,
+      "loss": 0.4727,
+      "step": 2091
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.995076716425732,
+      "learning_rate": 7.216356305077625e-06,
+      "loss": 0.4937,
+      "step": 2092
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.69763566486368,
+      "learning_rate": 7.21376524088987e-06,
+      "loss": 0.5181,
+      "step": 2093
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.925797313775647,
+      "learning_rate": 7.211173437042001e-06,
+      "loss": 0.3823,
+      "step": 2094
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.692483691370395,
+      "learning_rate": 7.208580894399986e-06,
+      "loss": 0.4917,
+      "step": 2095
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.59837794732827,
+      "learning_rate": 7.205987613830043e-06,
+      "loss": 0.5283,
+      "step": 2096
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 14.027224117019413,
+      "learning_rate": 7.203393596198635e-06,
+      "loss": 0.4348,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.277539316575139,
+      "learning_rate": 7.200798842372472e-06,
+      "loss": 0.5938,
+      "step": 2098
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.158219347531539,
+      "learning_rate": 7.198203353218508e-06,
+      "loss": 0.4829,
+      "step": 2099
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.788815769871793,
+      "learning_rate": 7.195607129603946e-06,
+      "loss": 0.4785,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.303724332755069,
+      "learning_rate": 7.19301017239623e-06,
+      "loss": 0.4536,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.88556880734833,
+      "learning_rate": 7.190412482463054e-06,
+      "loss": 0.4185,
+      "step": 2102
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.39510497345665,
+      "learning_rate": 7.187814060672354e-06,
+      "loss": 0.499,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.517777067379724,
+      "learning_rate": 7.1852149078923105e-06,
+      "loss": 0.4316,
+      "step": 2104
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.90115946127883,
+      "learning_rate": 7.1826150249913495e-06,
+      "loss": 0.5176,
+      "step": 2105
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.267334886302498,
+      "learning_rate": 7.18001441283814e-06,
+      "loss": 0.3643,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.066236031166735,
+      "learning_rate": 7.1774130723015955e-06,
+      "loss": 0.6748,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.466559423365932,
+      "learning_rate": 7.17481100425087e-06,
+      "loss": 0.3584,
+      "step": 2108
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.920295137461906,
+      "learning_rate": 7.172208209555365e-06,
+      "loss": 0.4668,
+      "step": 2109
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.95986098611621,
+      "learning_rate": 7.1696046890847206e-06,
+      "loss": 0.4946,
+      "step": 2110
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.786202212897535,
+      "learning_rate": 7.167000443708823e-06,
+      "loss": 0.5708,
+      "step": 2111
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.838902041674624,
+      "learning_rate": 7.164395474297798e-06,
+      "loss": 0.3911,
+      "step": 2112
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.176921095103253,
+      "learning_rate": 7.161789781722016e-06,
+      "loss": 0.4888,
+      "step": 2113
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.90810755723991,
+      "learning_rate": 7.159183366852085e-06,
+      "loss": 0.522,
+      "step": 2114
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 25.715900600535367,
+      "learning_rate": 7.156576230558859e-06,
+      "loss": 0.5347,
+      "step": 2115
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 28.28368775228431,
+      "learning_rate": 7.153968373713429e-06,
+      "loss": 0.5601,
+      "step": 2116
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.980625504053775,
+      "learning_rate": 7.1513597971871295e-06,
+      "loss": 0.4326,
+      "step": 2117
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.501120633814393,
+      "learning_rate": 7.148750501851532e-06,
+      "loss": 0.4717,
+      "step": 2118
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.712009993731347,
+      "learning_rate": 7.1461404885784545e-06,
+      "loss": 0.4873,
+      "step": 2119
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.12024516319098,
+      "learning_rate": 7.1435297582399475e-06,
+      "loss": 0.4927,
+      "step": 2120
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.745722717197969,
+      "learning_rate": 7.140918311708306e-06,
+      "loss": 0.459,
+      "step": 2121
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.688287138210573,
+      "learning_rate": 7.138306149856062e-06,
+      "loss": 0.3975,
+      "step": 2122
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.090721711186,
+      "learning_rate": 7.1356932735559905e-06,
+      "loss": 0.4312,
+      "step": 2123
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.510222651737925,
+      "learning_rate": 7.133079683681099e-06,
+      "loss": 0.5864,
+      "step": 2124
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.147602862513805,
+      "learning_rate": 7.130465381104635e-06,
+      "loss": 0.3774,
+      "step": 2125
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.962162623356548,
+      "learning_rate": 7.1278503667000885e-06,
+      "loss": 0.4297,
+      "step": 2126
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.176462569931615,
+      "learning_rate": 7.125234641341185e-06,
+      "loss": 0.4458,
+      "step": 2127
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.786769472257886,
+      "learning_rate": 7.1226182059018835e-06,
+      "loss": 0.397,
+      "step": 2128
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.022347014680557,
+      "learning_rate": 7.120001061256387e-06,
+      "loss": 0.3726,
+      "step": 2129
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.58498208142081,
+      "learning_rate": 7.1173832082791294e-06,
+      "loss": 0.5288,
+      "step": 2130
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.626054474790065,
+      "learning_rate": 7.114764647844788e-06,
+      "loss": 0.4883,
+      "step": 2131
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.442528730947034,
+      "learning_rate": 7.112145380828267e-06,
+      "loss": 0.5249,
+      "step": 2132
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.78478336790942,
+      "learning_rate": 7.109525408104717e-06,
+      "loss": 0.5713,
+      "step": 2133
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.13171845697349,
+      "learning_rate": 7.106904730549517e-06,
+      "loss": 0.4658,
+      "step": 2134
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.812889203251429,
+      "learning_rate": 7.104283349038285e-06,
+      "loss": 0.4365,
+      "step": 2135
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.244931909540133,
+      "learning_rate": 7.101661264446875e-06,
+      "loss": 0.4551,
+      "step": 2136
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.42786049799438,
+      "learning_rate": 7.099038477651371e-06,
+      "loss": 0.3369,
+      "step": 2137
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.446564970154658,
+      "learning_rate": 7.096414989528095e-06,
+      "loss": 0.4336,
+      "step": 2138
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.95393484819928,
+      "learning_rate": 7.093790800953606e-06,
+      "loss": 0.5459,
+      "step": 2139
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.975724248095739,
+      "learning_rate": 7.091165912804693e-06,
+      "loss": 0.4619,
+      "step": 2140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.463205000165559,
+      "learning_rate": 7.088540325958379e-06,
+      "loss": 0.4702,
+      "step": 2141
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.668064646053395,
+      "learning_rate": 7.085914041291921e-06,
+      "loss": 0.4897,
+      "step": 2142
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.501799641654888,
+      "learning_rate": 7.08328705968281e-06,
+      "loss": 0.5454,
+      "step": 2143
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.3244754880169625,
+      "learning_rate": 7.080659382008772e-06,
+      "loss": 0.4458,
+      "step": 2144
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.920592534439697,
+      "learning_rate": 7.078031009147759e-06,
+      "loss": 0.6465,
+      "step": 2145
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.120473562340443,
+      "learning_rate": 7.075401941977961e-06,
+      "loss": 0.4741,
+      "step": 2146
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.506397076951163,
+      "learning_rate": 7.072772181377798e-06,
+      "loss": 0.4634,
+      "step": 2147
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.898055505594296,
+      "learning_rate": 7.070141728225922e-06,
+      "loss": 0.5186,
+      "step": 2148
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.770324812358789,
+      "learning_rate": 7.067510583401217e-06,
+      "loss": 0.4585,
+      "step": 2149
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.618834750606382,
+      "learning_rate": 7.0648787477827965e-06,
+      "loss": 0.4229,
+      "step": 2150
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.901946072429947,
+      "learning_rate": 7.062246222250005e-06,
+      "loss": 0.4673,
+      "step": 2151
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.322546513477945,
+      "learning_rate": 7.05961300768242e-06,
+      "loss": 0.4419,
+      "step": 2152
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 21.827793100279667,
+      "learning_rate": 7.056979104959847e-06,
+      "loss": 0.6724,
+      "step": 2153
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.850403835999005,
+      "learning_rate": 7.054344514962319e-06,
+      "loss": 0.4351,
+      "step": 2154
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.28905672300358,
+      "learning_rate": 7.051709238570106e-06,
+      "loss": 0.5864,
+      "step": 2155
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.027979265939116,
+      "learning_rate": 7.0490732766637e-06,
+      "loss": 0.4814,
+      "step": 2156
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.594956263785003,
+      "learning_rate": 7.046436630123826e-06,
+      "loss": 0.5908,
+      "step": 2157
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.349980848273088,
+      "learning_rate": 7.043799299831438e-06,
+      "loss": 0.4062,
+      "step": 2158
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.422633603478822,
+      "learning_rate": 7.041161286667713e-06,
+      "loss": 0.4761,
+      "step": 2159
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.42059519880841,
+      "learning_rate": 7.038522591514061e-06,
+      "loss": 0.4302,
+      "step": 2160
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.084142584061851,
+      "learning_rate": 7.035883215252123e-06,
+      "loss": 0.4736,
+      "step": 2161
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.355044115107113,
+      "learning_rate": 7.03324315876376e-06,
+      "loss": 0.5054,
+      "step": 2162
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.3052309454934035,
+      "learning_rate": 7.030602422931065e-06,
+      "loss": 0.4629,
+      "step": 2163
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.739648409543541,
+      "learning_rate": 7.027961008636359e-06,
+      "loss": 0.4648,
+      "step": 2164
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.804729154765832,
+      "learning_rate": 7.025318916762185e-06,
+      "loss": 0.3853,
+      "step": 2165
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.13762113342716,
+      "learning_rate": 7.022676148191315e-06,
+      "loss": 0.604,
+      "step": 2166
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.74981032671273,
+      "learning_rate": 7.020032703806748e-06,
+      "loss": 0.4409,
+      "step": 2167
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.4004788730862,
+      "learning_rate": 7.017388584491709e-06,
+      "loss": 0.4077,
+      "step": 2168
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.52904134228692,
+      "learning_rate": 7.014743791129644e-06,
+      "loss": 0.5288,
+      "step": 2169
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.05836662525096,
+      "learning_rate": 7.012098324604231e-06,
+      "loss": 0.396,
+      "step": 2170
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.211986374727667,
+      "learning_rate": 7.009452185799368e-06,
+      "loss": 0.5439,
+      "step": 2171
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.01994379164931,
+      "learning_rate": 7.00680537559918e-06,
+      "loss": 0.5039,
+      "step": 2172
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.969090071115868,
+      "learning_rate": 7.0041578948880155e-06,
+      "loss": 0.4111,
+      "step": 2173
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.659257826403904,
+      "learning_rate": 7.001509744550446e-06,
+      "loss": 0.543,
+      "step": 2174
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.720135778440106,
+      "learning_rate": 6.998860925471267e-06,
+      "loss": 0.5352,
+      "step": 2175
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.39302176847214,
+      "learning_rate": 6.9962114385355e-06,
+      "loss": 0.4092,
+      "step": 2176
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.659365134233374,
+      "learning_rate": 6.993561284628388e-06,
+      "loss": 0.5156,
+      "step": 2177
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.458849658112828,
+      "learning_rate": 6.990910464635395e-06,
+      "loss": 0.4067,
+      "step": 2178
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.086418567695002,
+      "learning_rate": 6.9882589794422105e-06,
+      "loss": 0.418,
+      "step": 2179
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.32462739243237,
+      "learning_rate": 6.9856068299347455e-06,
+      "loss": 0.4932,
+      "step": 2180
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.227065489328242,
+      "learning_rate": 6.98295401699913e-06,
+      "loss": 0.4438,
+      "step": 2181
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.53687999777279,
+      "learning_rate": 6.980300541521721e-06,
+      "loss": 0.4766,
+      "step": 2182
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.091176799000983,
+      "learning_rate": 6.977646404389092e-06,
+      "loss": 0.3911,
+      "step": 2183
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.60903069175726,
+      "learning_rate": 6.9749916064880404e-06,
+      "loss": 0.6152,
+      "step": 2184
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.532470312845014,
+      "learning_rate": 6.972336148705583e-06,
+      "loss": 0.4365,
+      "step": 2185
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.726765751198604,
+      "learning_rate": 6.969680031928959e-06,
+      "loss": 0.4414,
+      "step": 2186
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.729404027235125,
+      "learning_rate": 6.967023257045624e-06,
+      "loss": 0.4883,
+      "step": 2187
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.006117166133711,
+      "learning_rate": 6.96436582494326e-06,
+      "loss": 0.375,
+      "step": 2188
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.304932091242616,
+      "learning_rate": 6.961707736509759e-06,
+      "loss": 0.5664,
+      "step": 2189
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.158180976223932,
+      "learning_rate": 6.959048992633241e-06,
+      "loss": 0.4287,
+      "step": 2190
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.951351287243968,
+      "learning_rate": 6.956389594202041e-06,
+      "loss": 0.4077,
+      "step": 2191
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.814399269265236,
+      "learning_rate": 6.953729542104713e-06,
+      "loss": 0.4473,
+      "step": 2192
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.228592101191628,
+      "learning_rate": 6.951068837230032e-06,
+      "loss": 0.6001,
+      "step": 2193
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.684761092197428,
+      "learning_rate": 6.9484074804669865e-06,
+      "loss": 0.4868,
+      "step": 2194
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.022339608408158,
+      "learning_rate": 6.945745472704786e-06,
+      "loss": 0.4446,
+      "step": 2195
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.806399945358521,
+      "learning_rate": 6.943082814832858e-06,
+      "loss": 0.439,
+      "step": 2196
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.733376328962677,
+      "learning_rate": 6.940419507740843e-06,
+      "loss": 0.5063,
+      "step": 2197
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.045131247868405,
+      "learning_rate": 6.937755552318606e-06,
+      "loss": 0.4028,
+      "step": 2198
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.45991812912647,
+      "learning_rate": 6.935090949456219e-06,
+      "loss": 0.4683,
+      "step": 2199
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.754774822400977,
+      "learning_rate": 6.93242570004398e-06,
+      "loss": 0.4248,
+      "step": 2200
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.026980227662335,
+      "learning_rate": 6.929759804972394e-06,
+      "loss": 0.4004,
+      "step": 2201
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.231964685107052,
+      "learning_rate": 6.92709326513219e-06,
+      "loss": 0.3926,
+      "step": 2202
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.292860309744759,
+      "learning_rate": 6.924426081414305e-06,
+      "loss": 0.4873,
+      "step": 2203
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.757286190878336,
+      "learning_rate": 6.921758254709897e-06,
+      "loss": 0.3643,
+      "step": 2204
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.606419829333902,
+      "learning_rate": 6.919089785910336e-06,
+      "loss": 0.4116,
+      "step": 2205
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.234839110113752,
+      "learning_rate": 6.916420675907207e-06,
+      "loss": 0.4482,
+      "step": 2206
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.023740197910566,
+      "learning_rate": 6.9137509255923085e-06,
+      "loss": 0.4917,
+      "step": 2207
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.280209803610372,
+      "learning_rate": 6.911080535857655e-06,
+      "loss": 0.4019,
+      "step": 2208
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.746299427986333,
+      "learning_rate": 6.908409507595472e-06,
+      "loss": 0.5361,
+      "step": 2209
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.906230626644021,
+      "learning_rate": 6.905737841698201e-06,
+      "loss": 0.4429,
+      "step": 2210
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.155722381902946,
+      "learning_rate": 6.903065539058496e-06,
+      "loss": 0.4624,
+      "step": 2211
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.196138695564693,
+      "learning_rate": 6.900392600569219e-06,
+      "loss": 0.3521,
+      "step": 2212
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.280936832961778,
+      "learning_rate": 6.897719027123451e-06,
+      "loss": 0.4634,
+      "step": 2213
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.371583647428105,
+      "learning_rate": 6.895044819614484e-06,
+      "loss": 0.3848,
+      "step": 2214
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.266982874224437,
+      "learning_rate": 6.8923699789358185e-06,
+      "loss": 0.3877,
+      "step": 2215
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.175344791502628,
+      "learning_rate": 6.88969450598117e-06,
+      "loss": 0.4609,
+      "step": 2216
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.899917847127078,
+      "learning_rate": 6.887018401644463e-06,
+      "loss": 0.5835,
+      "step": 2217
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.085073006810147,
+      "learning_rate": 6.884341666819832e-06,
+      "loss": 0.4443,
+      "step": 2218
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.194755959365828,
+      "learning_rate": 6.881664302401626e-06,
+      "loss": 0.5088,
+      "step": 2219
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.041590639227383,
+      "learning_rate": 6.878986309284401e-06,
+      "loss": 0.4272,
+      "step": 2220
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.858404710323796,
+      "learning_rate": 6.876307688362925e-06,
+      "loss": 0.5181,
+      "step": 2221
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.769923044069692,
+      "learning_rate": 6.873628440532175e-06,
+      "loss": 0.4094,
+      "step": 2222
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.760643999959065,
+      "learning_rate": 6.8709485666873375e-06,
+      "loss": 0.4131,
+      "step": 2223
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.059077802767067,
+      "learning_rate": 6.868268067723808e-06,
+      "loss": 0.498,
+      "step": 2224
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.067818695297724,
+      "learning_rate": 6.86558694453719e-06,
+      "loss": 0.4736,
+      "step": 2225
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.23251193640447,
+      "learning_rate": 6.8629051980233e-06,
+      "loss": 0.415,
+      "step": 2226
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.114819340426512,
+      "learning_rate": 6.860222829078156e-06,
+      "loss": 0.3423,
+      "step": 2227
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.762589126480844,
+      "learning_rate": 6.857539838597987e-06,
+      "loss": 0.3765,
+      "step": 2228
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.926047987180356,
+      "learning_rate": 6.8548562274792325e-06,
+      "loss": 0.4263,
+      "step": 2229
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.231555849492707,
+      "learning_rate": 6.8521719966185355e-06,
+      "loss": 0.498,
+      "step": 2230
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.73486658686708,
+      "learning_rate": 6.8494871469127474e-06,
+      "loss": 0.5493,
+      "step": 2231
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.05707560464292,
+      "learning_rate": 6.846801679258926e-06,
+      "loss": 0.5015,
+      "step": 2232
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.28340838759031,
+      "learning_rate": 6.844115594554338e-06,
+      "loss": 0.3784,
+      "step": 2233
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.26492682141842,
+      "learning_rate": 6.841428893696453e-06,
+      "loss": 0.521,
+      "step": 2234
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.269713147364001,
+      "learning_rate": 6.838741577582946e-06,
+      "loss": 0.4575,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.193115329733613,
+      "learning_rate": 6.836053647111701e-06,
+      "loss": 0.4976,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.42556116246502,
+      "learning_rate": 6.833365103180806e-06,
+      "loss": 0.4404,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.312625877201313,
+      "learning_rate": 6.830675946688552e-06,
+      "loss": 0.4473,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.467050902572582,
+      "learning_rate": 6.827986178533437e-06,
+      "loss": 0.5322,
+      "step": 2239
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.750353718427244,
+      "learning_rate": 6.825295799614163e-06,
+      "loss": 0.4478,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.352234008790848,
+      "learning_rate": 6.822604810829634e-06,
+      "loss": 0.4609,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.559244440826264,
+      "learning_rate": 6.819913213078961e-06,
+      "loss": 0.3923,
+      "step": 2242
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.88840732071109,
+      "learning_rate": 6.817221007261456e-06,
+      "loss": 0.4204,
+      "step": 2243
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.030745063332024,
+      "learning_rate": 6.814528194276636e-06,
+      "loss": 0.4292,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.309812809426399,
+      "learning_rate": 6.811834775024219e-06,
+      "loss": 0.5493,
+      "step": 2245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.66642562016404,
+      "learning_rate": 6.809140750404127e-06,
+      "loss": 0.4292,
+      "step": 2246
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.012313894198071,
+      "learning_rate": 6.8064461213164825e-06,
+      "loss": 0.4946,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.857244202004253,
+      "learning_rate": 6.803750888661611e-06,
+      "loss": 0.4478,
+      "step": 2248
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.033407840690604,
+      "learning_rate": 6.8010550533400425e-06,
+      "loss": 0.438,
+      "step": 2249
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.267146635510022,
+      "learning_rate": 6.798358616252503e-06,
+      "loss": 0.4214,
+      "step": 2250
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.114380153302204,
+      "learning_rate": 6.795661578299924e-06,
+      "loss": 0.4097,
+      "step": 2251
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.6843245806793,
+      "learning_rate": 6.792963940383436e-06,
+      "loss": 0.5952,
+      "step": 2252
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.510405242918026,
+      "learning_rate": 6.790265703404368e-06,
+      "loss": 0.4707,
+      "step": 2253
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.494076326644194,
+      "learning_rate": 6.787566868264253e-06,
+      "loss": 0.4829,
+      "step": 2254
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.25444248319594,
+      "learning_rate": 6.7848674358648195e-06,
+      "loss": 0.438,
+      "step": 2255
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.54816942321224,
+      "learning_rate": 6.782167407108001e-06,
+      "loss": 0.5273,
+      "step": 2256
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.11514410545727,
+      "learning_rate": 6.779466782895926e-06,
+      "loss": 0.4658,
+      "step": 2257
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.18230982436298,
+      "learning_rate": 6.7767655641309234e-06,
+      "loss": 0.5889,
+      "step": 2258
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.47295733121618,
+      "learning_rate": 6.7740637517155205e-06,
+      "loss": 0.5142,
+      "step": 2259
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.585772419117465,
+      "learning_rate": 6.771361346552445e-06,
+      "loss": 0.4607,
+      "step": 2260
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.554032105944135,
+      "learning_rate": 6.7686583495446164e-06,
+      "loss": 0.4375,
+      "step": 2261
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.428829908613086,
+      "learning_rate": 6.765954761595161e-06,
+      "loss": 0.5117,
+      "step": 2262
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.54475905891446,
+      "learning_rate": 6.763250583607392e-06,
+      "loss": 0.3823,
+      "step": 2263
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.52379110881373,
+      "learning_rate": 6.7605458164848316e-06,
+      "loss": 0.4619,
+      "step": 2264
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.02231162018557,
+      "learning_rate": 6.75784046113119e-06,
+      "loss": 0.5483,
+      "step": 2265
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.998871840978227,
+      "learning_rate": 6.755134518450377e-06,
+      "loss": 0.502,
+      "step": 2266
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.813653765158957,
+      "learning_rate": 6.752427989346497e-06,
+      "loss": 0.4629,
+      "step": 2267
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.768658187986116,
+      "learning_rate": 6.749720874723854e-06,
+      "loss": 0.4678,
+      "step": 2268
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.821712753766564,
+      "learning_rate": 6.747013175486944e-06,
+      "loss": 0.4683,
+      "step": 2269
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.14702978647226,
+      "learning_rate": 6.74430489254046e-06,
+      "loss": 0.542,
+      "step": 2270
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.591931087130397,
+      "learning_rate": 6.741596026789288e-06,
+      "loss": 0.5176,
+      "step": 2271
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.99774676180277,
+      "learning_rate": 6.7388865791385124e-06,
+      "loss": 0.4536,
+      "step": 2272
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.404000893575382,
+      "learning_rate": 6.736176550493411e-06,
+      "loss": 0.5005,
+      "step": 2273
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.67200743832191,
+      "learning_rate": 6.7334659417594514e-06,
+      "loss": 0.5234,
+      "step": 2274
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.823394969175009,
+      "learning_rate": 6.730754753842303e-06,
+      "loss": 0.4229,
+      "step": 2275
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.011366696228007,
+      "learning_rate": 6.728042987647818e-06,
+      "loss": 0.3921,
+      "step": 2276
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.774675579908692,
+      "learning_rate": 6.725330644082054e-06,
+      "loss": 0.5049,
+      "step": 2277
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.935065734139199,
+      "learning_rate": 6.7226177240512516e-06,
+      "loss": 0.4927,
+      "step": 2278
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.554489129584331,
+      "learning_rate": 6.7199042284618484e-06,
+      "loss": 0.4419,
+      "step": 2279
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.926780112484742,
+      "learning_rate": 6.717190158220475e-06,
+      "loss": 0.5508,
+      "step": 2280
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.840068980238367,
+      "learning_rate": 6.714475514233951e-06,
+      "loss": 0.4165,
+      "step": 2281
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.501441251965698,
+      "learning_rate": 6.71176029740929e-06,
+      "loss": 0.4741,
+      "step": 2282
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.55482426352725,
+      "learning_rate": 6.709044508653697e-06,
+      "loss": 0.4609,
+      "step": 2283
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.114294673812632,
+      "learning_rate": 6.706328148874568e-06,
+      "loss": 0.4961,
+      "step": 2284
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.349083200716997,
+      "learning_rate": 6.703611218979488e-06,
+      "loss": 0.5098,
+      "step": 2285
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.011989880992953,
+      "learning_rate": 6.700893719876234e-06,
+      "loss": 0.4868,
+      "step": 2286
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.689757524975285,
+      "learning_rate": 6.698175652472774e-06,
+      "loss": 0.4512,
+      "step": 2287
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.898403684293179,
+      "learning_rate": 6.695457017677263e-06,
+      "loss": 0.4131,
+      "step": 2288
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.240922912512271,
+      "learning_rate": 6.692737816398048e-06,
+      "loss": 0.5112,
+      "step": 2289
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.147572604976922,
+      "learning_rate": 6.6900180495436664e-06,
+      "loss": 0.3979,
+      "step": 2290
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 19.700085377690556,
+      "learning_rate": 6.68729771802284e-06,
+      "loss": 0.4995,
+      "step": 2291
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.978659517104596,
+      "learning_rate": 6.6845768227444855e-06,
+      "loss": 0.3945,
+      "step": 2292
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.810366816874936,
+      "learning_rate": 6.681855364617702e-06,
+      "loss": 0.4082,
+      "step": 2293
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.532079335607683,
+      "learning_rate": 6.67913334455178e-06,
+      "loss": 0.4424,
+      "step": 2294
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.749828735552693,
+      "learning_rate": 6.676410763456197e-06,
+      "loss": 0.4722,
+      "step": 2295
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.918601367919049,
+      "learning_rate": 6.673687622240619e-06,
+      "loss": 0.4126,
+      "step": 2296
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.349256190704107,
+      "learning_rate": 6.670963921814896e-06,
+      "loss": 0.5859,
+      "step": 2297
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.856607295110678,
+      "learning_rate": 6.668239663089069e-06,
+      "loss": 0.542,
+      "step": 2298
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.00766965784754,
+      "learning_rate": 6.665514846973361e-06,
+      "loss": 0.4756,
+      "step": 2299
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.099268884344017,
+      "learning_rate": 6.662789474378186e-06,
+      "loss": 0.5103,
+      "step": 2300
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.330890563879269,
+      "learning_rate": 6.6600635462141415e-06,
+      "loss": 0.355,
+      "step": 2301
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.992856781553082,
+      "learning_rate": 6.657337063392011e-06,
+      "loss": 0.4316,
+      "step": 2302
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.711043536387954,
+      "learning_rate": 6.654610026822761e-06,
+      "loss": 0.3696,
+      "step": 2303
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.942728892170798,
+      "learning_rate": 6.651882437417546e-06,
+      "loss": 0.4727,
+      "step": 2304
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.25898618932495,
+      "learning_rate": 6.649154296087705e-06,
+      "loss": 0.5059,
+      "step": 2305
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.960042275216363,
+      "learning_rate": 6.646425603744759e-06,
+      "loss": 0.4067,
+      "step": 2306
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.117852544611928,
+      "learning_rate": 6.643696361300418e-06,
+      "loss": 0.5503,
+      "step": 2307
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.347126680847555,
+      "learning_rate": 6.6409665696665715e-06,
+      "loss": 0.4541,
+      "step": 2308
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.520385793005548,
+      "learning_rate": 6.638236229755292e-06,
+      "loss": 0.5381,
+      "step": 2309
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.949529629176509,
+      "learning_rate": 6.635505342478838e-06,
+      "loss": 0.4204,
+      "step": 2310
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.147732369032411,
+      "learning_rate": 6.632773908749649e-06,
+      "loss": 0.4448,
+      "step": 2311
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.58683877473919,
+      "learning_rate": 6.630041929480349e-06,
+      "loss": 0.436,
+      "step": 2312
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.967781744974912,
+      "learning_rate": 6.627309405583741e-06,
+      "loss": 0.4839,
+      "step": 2313
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.185548630716545,
+      "learning_rate": 6.624576337972815e-06,
+      "loss": 0.4331,
+      "step": 2314
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.782057188604156,
+      "learning_rate": 6.621842727560737e-06,
+      "loss": 0.415,
+      "step": 2315
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.673091390131038,
+      "learning_rate": 6.6191085752608575e-06,
+      "loss": 0.4946,
+      "step": 2316
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.697402257685303,
+      "learning_rate": 6.616373881986708e-06,
+      "loss": 0.5723,
+      "step": 2317
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.647812348945917,
+      "learning_rate": 6.613638648652002e-06,
+      "loss": 0.4097,
+      "step": 2318
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.549329240009999,
+      "learning_rate": 6.610902876170631e-06,
+      "loss": 0.4482,
+      "step": 2319
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.337967711639493,
+      "learning_rate": 6.608166565456666e-06,
+      "loss": 0.4434,
+      "step": 2320
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 17.036210608178823,
+      "learning_rate": 6.605429717424359e-06,
+      "loss": 0.4116,
+      "step": 2321
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.4673537904469,
+      "learning_rate": 6.602692332988143e-06,
+      "loss": 0.4302,
+      "step": 2322
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.325826026833447,
+      "learning_rate": 6.5999544130626305e-06,
+      "loss": 0.4438,
+      "step": 2323
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.040149249921235,
+      "learning_rate": 6.597215958562608e-06,
+      "loss": 0.46,
+      "step": 2324
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.656646325351767,
+      "learning_rate": 6.5944769704030465e-06,
+      "loss": 0.4453,
+      "step": 2325
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.170279267084911,
+      "learning_rate": 6.591737449499092e-06,
+      "loss": 0.4639,
+      "step": 2326
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.82839224846332,
+      "learning_rate": 6.58899739676607e-06,
+      "loss": 0.4546,
+      "step": 2327
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.496186973092405,
+      "learning_rate": 6.586256813119482e-06,
+      "loss": 0.4648,
+      "step": 2328
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.873342997647923,
+      "learning_rate": 6.583515699475009e-06,
+      "loss": 0.4561,
+      "step": 2329
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.344690040103952,
+      "learning_rate": 6.580774056748508e-06,
+      "loss": 0.4336,
+      "step": 2330
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.872733176778468,
+      "learning_rate": 6.578031885856011e-06,
+      "loss": 0.4167,
+      "step": 2331
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.83642693495639,
+      "learning_rate": 6.575289187713731e-06,
+      "loss": 0.3911,
+      "step": 2332
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.856781054644795,
+      "learning_rate": 6.572545963238053e-06,
+      "loss": 0.4536,
+      "step": 2333
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.347855910699995,
+      "learning_rate": 6.569802213345537e-06,
+      "loss": 0.5303,
+      "step": 2334
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.646592360652438,
+      "learning_rate": 6.5670579389529255e-06,
+      "loss": 0.4077,
+      "step": 2335
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.438467277076052,
+      "learning_rate": 6.56431314097713e-06,
+      "loss": 0.3088,
+      "step": 2336
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.023040745678406,
+      "learning_rate": 6.561567820335236e-06,
+      "loss": 0.3955,
+      "step": 2337
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.033813975390757,
+      "learning_rate": 6.558821977944508e-06,
+      "loss": 0.4468,
+      "step": 2338
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.110205229735936,
+      "learning_rate": 6.556075614722383e-06,
+      "loss": 0.5103,
+      "step": 2339
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.807452959562594,
+      "learning_rate": 6.553328731586473e-06,
+      "loss": 0.5112,
+      "step": 2340
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.161963178049959,
+      "learning_rate": 6.550581329454561e-06,
+      "loss": 0.437,
+      "step": 2341
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.447457606510927,
+      "learning_rate": 6.547833409244606e-06,
+      "loss": 0.3843,
+      "step": 2342
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.377025255192994,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.4229,
+      "step": 2343
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.855103606469367,
+      "learning_rate": 6.542336018263262e-06,
+      "loss": 0.3804,
+      "step": 2344
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 27.46634631545949,
+      "learning_rate": 6.539586549328656e-06,
+      "loss": 0.7871,
+      "step": 2345
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.359851062684688,
+      "learning_rate": 6.536836565989565e-06,
+      "loss": 0.3911,
+      "step": 2346
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.053055914900025,
+      "learning_rate": 6.534086069164813e-06,
+      "loss": 0.4321,
+      "step": 2347
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.018450662610983,
+      "learning_rate": 6.531335059773392e-06,
+      "loss": 0.4824,
+      "step": 2348
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.550512997935256,
+      "learning_rate": 6.528583538734463e-06,
+      "loss": 0.5078,
+      "step": 2349
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.133602547357903,
+      "learning_rate": 6.525831506967361e-06,
+      "loss": 0.4575,
+      "step": 2350
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.318668444691664,
+      "learning_rate": 6.523078965391592e-06,
+      "loss": 0.4258,
+      "step": 2351
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.446059276582055,
+      "learning_rate": 6.520325914926831e-06,
+      "loss": 0.4482,
+      "step": 2352
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.682194650341806,
+      "learning_rate": 6.517572356492922e-06,
+      "loss": 0.4624,
+      "step": 2353
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 20.23938157311803,
+      "learning_rate": 6.514818291009881e-06,
+      "loss": 0.498,
+      "step": 2354
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.820982474486428,
+      "learning_rate": 6.512063719397894e-06,
+      "loss": 0.5381,
+      "step": 2355
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.993613065974447,
+      "learning_rate": 6.5093086425773126e-06,
+      "loss": 0.5732,
+      "step": 2356
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.37289512006142,
+      "learning_rate": 6.506553061468659e-06,
+      "loss": 0.4253,
+      "step": 2357
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.600077704313506,
+      "learning_rate": 6.5037969769926256e-06,
+      "loss": 0.4316,
+      "step": 2358
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.425829865085978,
+      "learning_rate": 6.501040390070071e-06,
+      "loss": 0.4639,
+      "step": 2359
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.9917751899752485,
+      "learning_rate": 6.498283301622022e-06,
+      "loss": 0.3745,
+      "step": 2360
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.202380861543828,
+      "learning_rate": 6.495525712569673e-06,
+      "loss": 0.4907,
+      "step": 2361
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.609299484513734,
+      "learning_rate": 6.492767623834385e-06,
+      "loss": 0.4478,
+      "step": 2362
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.853753914008019,
+      "learning_rate": 6.490009036337687e-06,
+      "loss": 0.4463,
+      "step": 2363
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.183051932609295,
+      "learning_rate": 6.487249951001276e-06,
+      "loss": 0.501,
+      "step": 2364
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.744349364358902,
+      "learning_rate": 6.484490368747012e-06,
+      "loss": 0.4519,
+      "step": 2365
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.03256193597822,
+      "learning_rate": 6.4817302904969226e-06,
+      "loss": 0.5122,
+      "step": 2366
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.807691344764418,
+      "learning_rate": 6.4789697171732024e-06,
+      "loss": 0.5269,
+      "step": 2367
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.611847017024495,
+      "learning_rate": 6.476208649698209e-06,
+      "loss": 0.4209,
+      "step": 2368
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.295897556272315,
+      "learning_rate": 6.473447088994467e-06,
+      "loss": 0.3936,
+      "step": 2369
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.443824455010702,
+      "learning_rate": 6.470685035984667e-06,
+      "loss": 0.4585,
+      "step": 2370
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.843529569578008,
+      "learning_rate": 6.467922491591658e-06,
+      "loss": 0.3989,
+      "step": 2371
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.301042404680986,
+      "learning_rate": 6.465159456738461e-06,
+      "loss": 0.4258,
+      "step": 2372
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.457468991739518,
+      "learning_rate": 6.462395932348257e-06,
+      "loss": 0.437,
+      "step": 2373
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.706783737244205,
+      "learning_rate": 6.459631919344389e-06,
+      "loss": 0.4785,
+      "step": 2374
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.469484126644588,
+      "learning_rate": 6.456867418650366e-06,
+      "loss": 0.4321,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.561227461724823,
+      "learning_rate": 6.454102431189859e-06,
+      "loss": 0.4399,
+      "step": 2376
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.39246937193766,
+      "learning_rate": 6.4513369578867026e-06,
+      "loss": 0.521,
+      "step": 2377
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.29432435609855,
+      "learning_rate": 6.448570999664894e-06,
+      "loss": 0.4775,
+      "step": 2378
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.171427132716822,
+      "learning_rate": 6.4458045574485875e-06,
+      "loss": 0.5679,
+      "step": 2379
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.83379330955623,
+      "learning_rate": 6.443037632162104e-06,
+      "loss": 0.5278,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.84513216881372,
+      "learning_rate": 6.440270224729927e-06,
+      "loss": 0.5034,
+      "step": 2381
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.928504744206203,
+      "learning_rate": 6.437502336076695e-06,
+      "loss": 0.5376,
+      "step": 2382
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.857986910329018,
+      "learning_rate": 6.4347339671272155e-06,
+      "loss": 0.3999,
+      "step": 2383
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.81608671529195,
+      "learning_rate": 6.431965118806449e-06,
+      "loss": 0.4619,
+      "step": 2384
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 27.17987199179481,
+      "learning_rate": 6.42919579203952e-06,
+      "loss": 0.5947,
+      "step": 2385
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.90191014571275,
+      "learning_rate": 6.4264259877517124e-06,
+      "loss": 0.5737,
+      "step": 2386
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.486777872297138,
+      "learning_rate": 6.423655706868468e-06,
+      "loss": 0.4072,
+      "step": 2387
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.62100655773416,
+      "learning_rate": 6.4208849503153915e-06,
+      "loss": 0.4175,
+      "step": 2388
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.205713284193536,
+      "learning_rate": 6.418113719018242e-06,
+      "loss": 0.541,
+      "step": 2389
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.27010808597464,
+      "learning_rate": 6.415342013902939e-06,
+      "loss": 0.4458,
+      "step": 2390
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.96766320470673,
+      "learning_rate": 6.412569835895562e-06,
+      "loss": 0.4741,
+      "step": 2391
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.6282420344549635,
+      "learning_rate": 6.409797185922349e-06,
+      "loss": 0.4624,
+      "step": 2392
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.407715638452274,
+      "learning_rate": 6.40702406490969e-06,
+      "loss": 0.438,
+      "step": 2393
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.853759038300608,
+      "learning_rate": 6.404250473784138e-06,
+      "loss": 0.4116,
+      "step": 2394
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.122502455763344,
+      "learning_rate": 6.401476413472404e-06,
+      "loss": 0.4565,
+      "step": 2395
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.738275465556727,
+      "learning_rate": 6.398701884901348e-06,
+      "loss": 0.4673,
+      "step": 2396
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.403912357064693,
+      "learning_rate": 6.3959268889979956e-06,
+      "loss": 0.4712,
+      "step": 2397
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.246073860215684,
+      "learning_rate": 6.393151426689522e-06,
+      "loss": 0.4727,
+      "step": 2398
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.726839635078996,
+      "learning_rate": 6.390375498903263e-06,
+      "loss": 0.5,
+      "step": 2399
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.077518387969189,
+      "learning_rate": 6.387599106566705e-06,
+      "loss": 0.3665,
+      "step": 2400
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.115234070495553,
+      "learning_rate": 6.384822250607495e-06,
+      "loss": 0.5576,
+      "step": 2401
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.158381190711447,
+      "learning_rate": 6.382044931953431e-06,
+      "loss": 0.4087,
+      "step": 2402
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.67290912672018,
+      "learning_rate": 6.379267151532467e-06,
+      "loss": 0.543,
+      "step": 2403
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.817381759789311,
+      "learning_rate": 6.376488910272709e-06,
+      "loss": 0.4165,
+      "step": 2404
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.705521476960307,
+      "learning_rate": 6.373710209102423e-06,
+      "loss": 0.4487,
+      "step": 2405
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.270499479688624,
+      "learning_rate": 6.370931048950022e-06,
+      "loss": 0.4756,
+      "step": 2406
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.738719945653308,
+      "learning_rate": 6.368151430744075e-06,
+      "loss": 0.4893,
+      "step": 2407
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.66405350297458,
+      "learning_rate": 6.365371355413306e-06,
+      "loss": 0.4688,
+      "step": 2408
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.542128682731759,
+      "learning_rate": 6.362590823886588e-06,
+      "loss": 0.4131,
+      "step": 2409
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.231576688229438,
+      "learning_rate": 6.359809837092947e-06,
+      "loss": 0.4248,
+      "step": 2410
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.471150104035912,
+      "learning_rate": 6.357028395961566e-06,
+      "loss": 0.4961,
+      "step": 2411
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.518938039693966,
+      "learning_rate": 6.354246501421777e-06,
+      "loss": 0.5054,
+      "step": 2412
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.325117897931076,
+      "learning_rate": 6.3514641544030575e-06,
+      "loss": 0.5117,
+      "step": 2413
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.695654010917847,
+      "learning_rate": 6.348681355835043e-06,
+      "loss": 0.4731,
+      "step": 2414
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.353958500414972,
+      "learning_rate": 6.345898106647521e-06,
+      "loss": 0.4497,
+      "step": 2415
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.125134539120891,
+      "learning_rate": 6.3431144077704245e-06,
+      "loss": 0.5361,
+      "step": 2416
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.29982044993289,
+      "learning_rate": 6.340330260133839e-06,
+      "loss": 0.5303,
+      "step": 2417
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.635927861507614,
+      "learning_rate": 6.337545664668001e-06,
+      "loss": 0.4492,
+      "step": 2418
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.450720986263002,
+      "learning_rate": 6.334760622303294e-06,
+      "loss": 0.4526,
+      "step": 2419
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.96898407515457,
+      "learning_rate": 6.331975133970255e-06,
+      "loss": 0.374,
+      "step": 2420
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 22.99713545008352,
+      "learning_rate": 6.329189200599566e-06,
+      "loss": 0.5244,
+      "step": 2421
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.164058078222877,
+      "learning_rate": 6.326402823122059e-06,
+      "loss": 0.3335,
+      "step": 2422
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.821666072907615,
+      "learning_rate": 6.3236160024687134e-06,
+      "loss": 0.4614,
+      "step": 2423
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.86789061992421,
+      "learning_rate": 6.3208287395706595e-06,
+      "loss": 0.4541,
+      "step": 2424
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.767069143692144,
+      "learning_rate": 6.3180410353591735e-06,
+      "loss": 0.4414,
+      "step": 2425
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.27317813903384,
+      "learning_rate": 6.315252890765678e-06,
+      "loss": 0.502,
+      "step": 2426
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.0533891840444864,
+      "learning_rate": 6.312464306721745e-06,
+      "loss": 0.4478,
+      "step": 2427
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.284405823285676,
+      "learning_rate": 6.309675284159093e-06,
+      "loss": 0.4644,
+      "step": 2428
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.273614279657444,
+      "learning_rate": 6.306885824009585e-06,
+      "loss": 0.3833,
+      "step": 2429
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.094247190767463,
+      "learning_rate": 6.3040959272052315e-06,
+      "loss": 0.3955,
+      "step": 2430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.05210332824741,
+      "learning_rate": 6.301305594678189e-06,
+      "loss": 0.4409,
+      "step": 2431
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.070593419557918,
+      "learning_rate": 6.2985148273607586e-06,
+      "loss": 0.394,
+      "step": 2432
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.162689408282903,
+      "learning_rate": 6.29572362618539e-06,
+      "loss": 0.3916,
+      "step": 2433
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.341208257731296,
+      "learning_rate": 6.292931992084672e-06,
+      "loss": 0.6147,
+      "step": 2434
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.475007541185304,
+      "learning_rate": 6.290139925991345e-06,
+      "loss": 0.5015,
+      "step": 2435
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.472537023920362,
+      "learning_rate": 6.287347428838289e-06,
+      "loss": 0.3416,
+      "step": 2436
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.81968592719629,
+      "learning_rate": 6.2845545015585275e-06,
+      "loss": 0.5249,
+      "step": 2437
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.127563834698211,
+      "learning_rate": 6.281761145085232e-06,
+      "loss": 0.4546,
+      "step": 2438
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.2131478626402,
+      "learning_rate": 6.278967360351712e-06,
+      "loss": 0.291,
+      "step": 2439
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.014830861323171,
+      "learning_rate": 6.276173148291425e-06,
+      "loss": 0.4507,
+      "step": 2440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 8.370157635005995,
+      "learning_rate": 6.273378509837969e-06,
+      "loss": 0.3735,
+      "step": 2441
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.4882237053066,
+      "learning_rate": 6.2705834459250825e-06,
+      "loss": 0.4019,
+      "step": 2442
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.073520612528323,
+      "learning_rate": 6.2677879574866515e-06,
+      "loss": 0.4048,
+      "step": 2443
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.132342613326061,
+      "learning_rate": 6.264992045456699e-06,
+      "loss": 0.4619,
+      "step": 2444
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.160719007717073,
+      "learning_rate": 6.262195710769391e-06,
+      "loss": 0.3857,
+      "step": 2445
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.472439966811617,
+      "learning_rate": 6.259398954359037e-06,
+      "loss": 0.4429,
+      "step": 2446
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.427769712154046,
+      "learning_rate": 6.256601777160082e-06,
+      "loss": 0.6123,
+      "step": 2447
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.639713264184406,
+      "learning_rate": 6.253804180107116e-06,
+      "loss": 0.4785,
+      "step": 2448
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.158135673228788,
+      "learning_rate": 6.2510061641348695e-06,
+      "loss": 0.5244,
+      "step": 2449
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.730810111574574,
+      "learning_rate": 6.248207730178211e-06,
+      "loss": 0.4771,
+      "step": 2450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.479475483349935,
+      "learning_rate": 6.245408879172148e-06,
+      "loss": 0.4209,
+      "step": 2451
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.545307163598485,
+      "learning_rate": 6.24260961205183e-06,
+      "loss": 0.4536,
+      "step": 2452
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.433203777453386,
+      "learning_rate": 6.239809929752544e-06,
+      "loss": 0.3677,
+      "step": 2453
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.724851412359518,
+      "learning_rate": 6.237009833209715e-06,
+      "loss": 0.4668,
+      "step": 2454
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 24.535444127877504,
+      "learning_rate": 6.2342093233589095e-06,
+      "loss": 0.6094,
+      "step": 2455
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.131949690353993,
+      "learning_rate": 6.231408401135828e-06,
+      "loss": 0.4727,
+      "step": 2456
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.95385057418741,
+      "learning_rate": 6.228607067476311e-06,
+      "loss": 0.425,
+      "step": 2457
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.652779199575274,
+      "learning_rate": 6.225805323316336e-06,
+      "loss": 0.5317,
+      "step": 2458
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.756602418041059,
+      "learning_rate": 6.223003169592018e-06,
+      "loss": 0.4741,
+      "step": 2459
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.349749974546135,
+      "learning_rate": 6.220200607239609e-06,
+      "loss": 0.4736,
+      "step": 2460
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.203291816830776,
+      "learning_rate": 6.217397637195497e-06,
+      "loss": 0.4951,
+      "step": 2461
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.149381480669541,
+      "learning_rate": 6.214594260396206e-06,
+      "loss": 0.4688,
+      "step": 2462
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.733718301343428,
+      "learning_rate": 6.211790477778399e-06,
+      "loss": 0.4492,
+      "step": 2463
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.776778166448288,
+      "learning_rate": 6.208986290278866e-06,
+      "loss": 0.3716,
+      "step": 2464
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.981056473195446,
+      "learning_rate": 6.206181698834544e-06,
+      "loss": 0.5264,
+      "step": 2465
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.177267724597133,
+      "learning_rate": 6.2033767043824955e-06,
+      "loss": 0.4902,
+      "step": 2466
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.497702259870394,
+      "learning_rate": 6.200571307859923e-06,
+      "loss": 0.4185,
+      "step": 2467
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.49048950885061,
+      "learning_rate": 6.197765510204161e-06,
+      "loss": 0.5205,
+      "step": 2468
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.508146338639264,
+      "learning_rate": 6.19495931235268e-06,
+      "loss": 0.4531,
+      "step": 2469
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.260708359404394,
+      "learning_rate": 6.19215271524308e-06,
+      "loss": 0.4541,
+      "step": 2470
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.1175023636757,
+      "learning_rate": 6.189345719813099e-06,
+      "loss": 0.4873,
+      "step": 2471
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.651326204773177,
+      "learning_rate": 6.186538327000609e-06,
+      "loss": 0.4849,
+      "step": 2472
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.573726463990967,
+      "learning_rate": 6.183730537743607e-06,
+      "loss": 0.4097,
+      "step": 2473
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.017710357353614,
+      "learning_rate": 6.18092235298023e-06,
+      "loss": 0.4702,
+      "step": 2474
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.84022010023255,
+      "learning_rate": 6.178113773648745e-06,
+      "loss": 0.3994,
+      "step": 2475
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.022407957306385,
+      "learning_rate": 6.175304800687551e-06,
+      "loss": 0.4067,
+      "step": 2476
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.250357096874497,
+      "learning_rate": 6.172495435035176e-06,
+      "loss": 0.4756,
+      "step": 2477
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.921552289632256,
+      "learning_rate": 6.169685677630284e-06,
+      "loss": 0.4336,
+      "step": 2478
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.198045788857288,
+      "learning_rate": 6.1668755294116655e-06,
+      "loss": 0.3325,
+      "step": 2479
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.228855338016636,
+      "learning_rate": 6.1640649913182436e-06,
+      "loss": 0.4209,
+      "step": 2480
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.828383315352083,
+      "learning_rate": 6.161254064289072e-06,
+      "loss": 0.4023,
+      "step": 2481
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.423877993770597,
+      "learning_rate": 6.158442749263332e-06,
+      "loss": 0.4683,
+      "step": 2482
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.780261625052676,
+      "learning_rate": 6.155631047180337e-06,
+      "loss": 0.4595,
+      "step": 2483
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.04469853406095,
+      "learning_rate": 6.152818958979529e-06,
+      "loss": 0.5542,
+      "step": 2484
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.285094723692712,
+      "learning_rate": 6.1500064856004796e-06,
+      "loss": 0.4995,
+      "step": 2485
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.325731810799802,
+      "learning_rate": 6.147193627982887e-06,
+      "loss": 0.3689,
+      "step": 2486
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.60771444277801,
+      "learning_rate": 6.144380387066581e-06,
+      "loss": 0.4771,
+      "step": 2487
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.951542508375255,
+      "learning_rate": 6.141566763791518e-06,
+      "loss": 0.4243,
+      "step": 2488
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.53628405098184,
+      "learning_rate": 6.138752759097778e-06,
+      "loss": 0.4272,
+      "step": 2489
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.97888607653731,
+      "learning_rate": 6.135938373925576e-06,
+      "loss": 0.4653,
+      "step": 2490
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.48936310841184,
+      "learning_rate": 6.133123609215249e-06,
+      "loss": 0.4019,
+      "step": 2491
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.096587824152824,
+      "learning_rate": 6.130308465907263e-06,
+      "loss": 0.396,
+      "step": 2492
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.227096202950863,
+      "learning_rate": 6.127492944942209e-06,
+      "loss": 0.4258,
+      "step": 2493
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.942411840921341,
+      "learning_rate": 6.124677047260805e-06,
+      "loss": 0.3965,
+      "step": 2494
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.00521904473558,
+      "learning_rate": 6.121860773803895e-06,
+      "loss": 0.4102,
+      "step": 2495
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.245096943439378,
+      "learning_rate": 6.119044125512447e-06,
+      "loss": 0.3967,
+      "step": 2496
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.70014640452883,
+      "learning_rate": 6.116227103327559e-06,
+      "loss": 0.4951,
+      "step": 2497
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.83092193293532,
+      "learning_rate": 6.113409708190447e-06,
+      "loss": 0.5278,
+      "step": 2498
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.447966368255546,
+      "learning_rate": 6.1105919410424566e-06,
+      "loss": 0.5225,
+      "step": 2499
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.135937978938692,
+      "learning_rate": 6.107773802825055e-06,
+      "loss": 0.52,
+      "step": 2500
+    },
+    {
+      "epoch": 1.78,
+      "eval_avg_AUC": 0.7652931989776042,
+      "eval_avg_Accuracy": 0.6897380636604774,
+      "eval_avg_Accuracy-right": 0.8790922133820269,
+      "eval_avg_Accuracy-wrong": 0.3595633386399818,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6470409856535183,
+      "eval_last_AUC": 0.7821256327791941,
+      "eval_last_Accuracy": 0.7235162466843501,
+      "eval_last_Accuracy-right": 0.8033128994391548,
+      "eval_last_Accuracy-wrong": 0.584375710711849,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6545993525396961,
+      "eval_max_AUC": 0.7132634593306635,
+      "eval_max_Accuracy": 0.6386770557029178,
+      "eval_max_Accuracy-right": 0.9663492891613408,
+      "eval_max_Accuracy-wrong": 0.06731862633613828,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.5960403684660525,
+      "eval_min_AUC": 0.7740940679255949,
+      "eval_min_Accuracy": 0.7144396551724138,
+      "eval_min_Accuracy-right": 0.7332724664145037,
+      "eval_min_Accuracy-wrong": 0.6816010916534,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6455547810166508,
+      "eval_prod_AUC": 0.774109644816048,
+      "eval_prod_Accuracy": 0.6302635941644562,
+      "eval_prod_Accuracy-right": 0.4789356984478936,
+      "eval_prod_Accuracy-wrong": 0.8941323629747555,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6440620351725057,
+      "eval_runtime": 247.2806,
+      "eval_samples_per_second": 97.573,
+      "eval_steps_per_second": 3.049,
+      "eval_sum_AUC": 0.6276356610336933,
+      "eval_sum_Accuracy": 0.6374336870026526,
+      "eval_sum_Accuracy-right": 0.9956306247554454,
+      "eval_sum_Accuracy-wrong": 0.01284967022970207,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6313597803483167,
+      "step": 2500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.414355809239556,
+      "learning_rate": 6.1049552944798355e-06,
+      "loss": 0.4023,
+      "step": 2501
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.539305998342051,
+      "learning_rate": 6.102136416948513e-06,
+      "loss": 0.4678,
+      "step": 2502
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.768527333824066,
+      "learning_rate": 6.099317171172929e-06,
+      "loss": 0.4272,
+      "step": 2503
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.846773419479307,
+      "learning_rate": 6.0964975580950445e-06,
+      "loss": 0.4639,
+      "step": 2504
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.851344453902673,
+      "learning_rate": 6.093677578656946e-06,
+      "loss": 0.5596,
+      "step": 2505
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.964802832185107,
+      "learning_rate": 6.090857233800839e-06,
+      "loss": 0.4653,
+      "step": 2506
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.580109191519524,
+      "learning_rate": 6.0880365244690546e-06,
+      "loss": 0.4565,
+      "step": 2507
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.766966924313547,
+      "learning_rate": 6.085215451604044e-06,
+      "loss": 0.5005,
+      "step": 2508
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.585939267007543,
+      "learning_rate": 6.082394016148379e-06,
+      "loss": 0.4629,
+      "step": 2509
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.256079889521589,
+      "learning_rate": 6.079572219044755e-06,
+      "loss": 0.4443,
+      "step": 2510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.962206455993298,
+      "learning_rate": 6.076750061235985e-06,
+      "loss": 0.4058,
+      "step": 2511
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.047853996411353,
+      "learning_rate": 6.073927543665008e-06,
+      "loss": 0.519,
+      "step": 2512
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.508769620004047,
+      "learning_rate": 6.071104667274875e-06,
+      "loss": 0.5142,
+      "step": 2513
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.47411144644649,
+      "learning_rate": 6.068281433008765e-06,
+      "loss": 0.5996,
+      "step": 2514
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.586741408151706,
+      "learning_rate": 6.0654578418099715e-06,
+      "loss": 0.5146,
+      "step": 2515
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.649138506555311,
+      "learning_rate": 6.062633894621909e-06,
+      "loss": 0.4038,
+      "step": 2516
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.188289282459092,
+      "learning_rate": 6.0598095923881105e-06,
+      "loss": 0.5435,
+      "step": 2517
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.818692335136738,
+      "learning_rate": 6.056984936052229e-06,
+      "loss": 0.4629,
+      "step": 2518
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.234001156083075,
+      "learning_rate": 6.054159926558033e-06,
+      "loss": 0.5342,
+      "step": 2519
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.075130769946417,
+      "learning_rate": 6.051334564849413e-06,
+      "loss": 0.4712,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.956699388387689,
+      "learning_rate": 6.048508851870372e-06,
+      "loss": 0.4111,
+      "step": 2521
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.649995024246259,
+      "learning_rate": 6.045682788565036e-06,
+      "loss": 0.3521,
+      "step": 2522
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.193089965908905,
+      "learning_rate": 6.042856375877644e-06,
+      "loss": 0.5518,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.405344544727377,
+      "learning_rate": 6.040029614752551e-06,
+      "loss": 0.4873,
+      "step": 2524
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.872507537818283,
+      "learning_rate": 6.037202506134234e-06,
+      "loss": 0.502,
+      "step": 2525
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.056217550868038,
+      "learning_rate": 6.03437505096728e-06,
+      "loss": 0.457,
+      "step": 2526
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.745479176249967,
+      "learning_rate": 6.0315472501963955e-06,
+      "loss": 0.4609,
+      "step": 2527
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.223610098312774,
+      "learning_rate": 6.028719104766402e-06,
+      "loss": 0.4082,
+      "step": 2528
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.50042655333063,
+      "learning_rate": 6.025890615622233e-06,
+      "loss": 0.5039,
+      "step": 2529
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.476170718540832,
+      "learning_rate": 6.023061783708941e-06,
+      "loss": 0.4048,
+      "step": 2530
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.520249627385631,
+      "learning_rate": 6.020232609971694e-06,
+      "loss": 0.439,
+      "step": 2531
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.589519698746084,
+      "learning_rate": 6.017403095355766e-06,
+      "loss": 0.5166,
+      "step": 2532
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.950886743696067,
+      "learning_rate": 6.014573240806553e-06,
+      "loss": 0.4604,
+      "step": 2533
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.551260453439275,
+      "learning_rate": 6.011743047269563e-06,
+      "loss": 0.4204,
+      "step": 2534
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.017128679094064,
+      "learning_rate": 6.008912515690415e-06,
+      "loss": 0.4873,
+      "step": 2535
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.121011247353195,
+      "learning_rate": 6.006081647014842e-06,
+      "loss": 0.4297,
+      "step": 2536
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 28.47281369820584,
+      "learning_rate": 6.00325044218869e-06,
+      "loss": 0.5586,
+      "step": 2537
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.714417153244161,
+      "learning_rate": 6.000418902157919e-06,
+      "loss": 0.5317,
+      "step": 2538
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.720034449703754,
+      "learning_rate": 5.997587027868598e-06,
+      "loss": 0.4829,
+      "step": 2539
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.914782644486152,
+      "learning_rate": 5.994754820266908e-06,
+      "loss": 0.3906,
+      "step": 2540
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.814365270699001,
+      "learning_rate": 5.991922280299143e-06,
+      "loss": 0.3979,
+      "step": 2541
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.20707607734891,
+      "learning_rate": 5.989089408911706e-06,
+      "loss": 0.4653,
+      "step": 2542
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.049185460575597,
+      "learning_rate": 5.986256207051113e-06,
+      "loss": 0.48,
+      "step": 2543
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.672521854401475,
+      "learning_rate": 5.98342267566399e-06,
+      "loss": 0.5356,
+      "step": 2544
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.541254903832398,
+      "learning_rate": 5.9805888156970714e-06,
+      "loss": 0.4609,
+      "step": 2545
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.284454916773962,
+      "learning_rate": 5.977754628097203e-06,
+      "loss": 0.4688,
+      "step": 2546
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.23610205729337,
+      "learning_rate": 5.97492011381134e-06,
+      "loss": 0.3687,
+      "step": 2547
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.11706014179341,
+      "learning_rate": 5.972085273786547e-06,
+      "loss": 0.4453,
+      "step": 2548
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.756513019937023,
+      "learning_rate": 5.969250108969995e-06,
+      "loss": 0.4448,
+      "step": 2549
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.098290335839529,
+      "learning_rate": 5.966414620308965e-06,
+      "loss": 0.4639,
+      "step": 2550
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.29559974589973,
+      "learning_rate": 5.9635788087508474e-06,
+      "loss": 0.438,
+      "step": 2551
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.829092992423524,
+      "learning_rate": 5.960742675243139e-06,
+      "loss": 0.3999,
+      "step": 2552
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.468196516461774,
+      "learning_rate": 5.957906220733447e-06,
+      "loss": 0.416,
+      "step": 2553
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.859111236122798,
+      "learning_rate": 5.9550694461694806e-06,
+      "loss": 0.4062,
+      "step": 2554
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.371396705917446,
+      "learning_rate": 5.95223235249906e-06,
+      "loss": 0.4697,
+      "step": 2555
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.74581839772436,
+      "learning_rate": 5.949394940670112e-06,
+      "loss": 0.4634,
+      "step": 2556
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.11191421574571,
+      "learning_rate": 5.946557211630667e-06,
+      "loss": 0.5122,
+      "step": 2557
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.41215503415198,
+      "learning_rate": 5.943719166328864e-06,
+      "loss": 0.4316,
+      "step": 2558
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.922238946122983,
+      "learning_rate": 5.940880805712945e-06,
+      "loss": 0.3711,
+      "step": 2559
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.556285820468634,
+      "learning_rate": 5.938042130731262e-06,
+      "loss": 0.4712,
+      "step": 2560
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.602935307175088,
+      "learning_rate": 5.935203142332267e-06,
+      "loss": 0.5796,
+      "step": 2561
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.172490980473096,
+      "learning_rate": 5.932363841464519e-06,
+      "loss": 0.3892,
+      "step": 2562
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.979141458728968,
+      "learning_rate": 5.9295242290766805e-06,
+      "loss": 0.4556,
+      "step": 2563
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.830863705403445,
+      "learning_rate": 5.9266843061175216e-06,
+      "loss": 0.4551,
+      "step": 2564
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 23.815492103215934,
+      "learning_rate": 5.92384407353591e-06,
+      "loss": 0.5635,
+      "step": 2565
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.015347774380794,
+      "learning_rate": 5.921003532280822e-06,
+      "loss": 0.4316,
+      "step": 2566
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.625830475732585,
+      "learning_rate": 5.918162683301336e-06,
+      "loss": 0.5039,
+      "step": 2567
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.919505793902855,
+      "learning_rate": 5.91532152754663e-06,
+      "loss": 0.5859,
+      "step": 2568
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.891844235140119,
+      "learning_rate": 5.91248006596599e-06,
+      "loss": 0.4126,
+      "step": 2569
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.943605729070335,
+      "learning_rate": 5.909638299508798e-06,
+      "loss": 0.4131,
+      "step": 2570
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.256434625761717,
+      "learning_rate": 5.906796229124543e-06,
+      "loss": 0.4639,
+      "step": 2571
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.457753839967175,
+      "learning_rate": 5.903953855762812e-06,
+      "loss": 0.4829,
+      "step": 2572
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.1694335296129,
+      "learning_rate": 5.901111180373298e-06,
+      "loss": 0.5698,
+      "step": 2573
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.654685001614778,
+      "learning_rate": 5.898268203905788e-06,
+      "loss": 0.4927,
+      "step": 2574
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.94449663637914,
+      "learning_rate": 5.895424927310174e-06,
+      "loss": 0.478,
+      "step": 2575
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.653556967960311,
+      "learning_rate": 5.89258135153645e-06,
+      "loss": 0.4351,
+      "step": 2576
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.724275472225052,
+      "learning_rate": 5.889737477534704e-06,
+      "loss": 0.3901,
+      "step": 2577
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.683542432644515,
+      "learning_rate": 5.886893306255129e-06,
+      "loss": 0.4609,
+      "step": 2578
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.145435964386623,
+      "learning_rate": 5.884048838648017e-06,
+      "loss": 0.5005,
+      "step": 2579
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.871255725801072,
+      "learning_rate": 5.881204075663755e-06,
+      "loss": 0.4761,
+      "step": 2580
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 24.214634872213015,
+      "learning_rate": 5.878359018252831e-06,
+      "loss": 0.6128,
+      "step": 2581
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.406317683172762,
+      "learning_rate": 5.8755136673658365e-06,
+      "loss": 0.4609,
+      "step": 2582
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.554115156101938,
+      "learning_rate": 5.872668023953449e-06,
+      "loss": 0.5054,
+      "step": 2583
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.952532962810603,
+      "learning_rate": 5.869822088966455e-06,
+      "loss": 0.4531,
+      "step": 2584
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.834444713741934,
+      "learning_rate": 5.866975863355734e-06,
+      "loss": 0.4854,
+      "step": 2585
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.54691819321447,
+      "learning_rate": 5.864129348072261e-06,
+      "loss": 0.5293,
+      "step": 2586
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.582713115393668,
+      "learning_rate": 5.861282544067112e-06,
+      "loss": 0.4829,
+      "step": 2587
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.239317398524113,
+      "learning_rate": 5.8584354522914555e-06,
+      "loss": 0.5029,
+      "step": 2588
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.723468978865414,
+      "learning_rate": 5.855588073696559e-06,
+      "loss": 0.4668,
+      "step": 2589
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.930051826297984,
+      "learning_rate": 5.852740409233785e-06,
+      "loss": 0.4092,
+      "step": 2590
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.019117275735518,
+      "learning_rate": 5.849892459854588e-06,
+      "loss": 0.3613,
+      "step": 2591
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.473807935463805,
+      "learning_rate": 5.847044226510524e-06,
+      "loss": 0.4814,
+      "step": 2592
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.961602975113415,
+      "learning_rate": 5.84419571015324e-06,
+      "loss": 0.4736,
+      "step": 2593
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.145353957021218,
+      "learning_rate": 5.8413469117344766e-06,
+      "loss": 0.4971,
+      "step": 2594
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.251775681976962,
+      "learning_rate": 5.838497832206074e-06,
+      "loss": 0.4351,
+      "step": 2595
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.772584386631225,
+      "learning_rate": 5.835648472519958e-06,
+      "loss": 0.4829,
+      "step": 2596
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.526504254371492,
+      "learning_rate": 5.832798833628156e-06,
+      "loss": 0.4814,
+      "step": 2597
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.52349073531977,
+      "learning_rate": 5.829948916482784e-06,
+      "loss": 0.4419,
+      "step": 2598
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.240904448794703,
+      "learning_rate": 5.827098722036053e-06,
+      "loss": 0.4404,
+      "step": 2599
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.162623628027008,
+      "learning_rate": 5.824248251240265e-06,
+      "loss": 0.437,
+      "step": 2600
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.063102295836982,
+      "learning_rate": 5.8213975050478155e-06,
+      "loss": 0.4668,
+      "step": 2601
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.197499697597129,
+      "learning_rate": 5.818546484411191e-06,
+      "loss": 0.4873,
+      "step": 2602
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.957695255120187,
+      "learning_rate": 5.815695190282974e-06,
+      "loss": 0.5273,
+      "step": 2603
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.149765327635293,
+      "learning_rate": 5.81284362361583e-06,
+      "loss": 0.5435,
+      "step": 2604
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.897344437580472,
+      "learning_rate": 5.809991785362525e-06,
+      "loss": 0.4624,
+      "step": 2605
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.41339345815525,
+      "learning_rate": 5.8071396764759065e-06,
+      "loss": 0.4155,
+      "step": 2606
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.588859240549306,
+      "learning_rate": 5.804287297908923e-06,
+      "loss": 0.4224,
+      "step": 2607
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.051689248011026,
+      "learning_rate": 5.801434650614601e-06,
+      "loss": 0.4731,
+      "step": 2608
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.164254844968275,
+      "learning_rate": 5.798581735546066e-06,
+      "loss": 0.4878,
+      "step": 2609
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.618691567900477,
+      "learning_rate": 5.79572855365653e-06,
+      "loss": 0.3784,
+      "step": 2610
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.41055691629168,
+      "learning_rate": 5.792875105899294e-06,
+      "loss": 0.5732,
+      "step": 2611
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.670181143911773,
+      "learning_rate": 5.790021393227747e-06,
+      "loss": 0.4133,
+      "step": 2612
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.575788308226384,
+      "learning_rate": 5.787167416595369e-06,
+      "loss": 0.4673,
+      "step": 2613
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.142364245518445,
+      "learning_rate": 5.784313176955726e-06,
+      "loss": 0.4351,
+      "step": 2614
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.236307058725137,
+      "learning_rate": 5.781458675262472e-06,
+      "loss": 0.3555,
+      "step": 2615
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.169804367113855,
+      "learning_rate": 5.778603912469349e-06,
+      "loss": 0.4067,
+      "step": 2616
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.194556650327089,
+      "learning_rate": 5.775748889530187e-06,
+      "loss": 0.5103,
+      "step": 2617
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.8031354287890995,
+      "learning_rate": 5.772893607398901e-06,
+      "loss": 0.3564,
+      "step": 2618
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.611694361969873,
+      "learning_rate": 5.770038067029496e-06,
+      "loss": 0.4175,
+      "step": 2619
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.88527859419395,
+      "learning_rate": 5.76718226937606e-06,
+      "loss": 0.3125,
+      "step": 2620
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.739686500701513,
+      "learning_rate": 5.764326215392768e-06,
+      "loss": 0.418,
+      "step": 2621
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.698752773439546,
+      "learning_rate": 5.761469906033879e-06,
+      "loss": 0.3662,
+      "step": 2622
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.707736166224288,
+      "learning_rate": 5.758613342253743e-06,
+      "loss": 0.374,
+      "step": 2623
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.383000470021422,
+      "learning_rate": 5.7557565250067896e-06,
+      "loss": 0.4565,
+      "step": 2624
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.55652728383383,
+      "learning_rate": 5.752899455247532e-06,
+      "loss": 0.3955,
+      "step": 2625
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.66660044282398,
+      "learning_rate": 5.750042133930571e-06,
+      "loss": 0.4761,
+      "step": 2626
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.605013174142563,
+      "learning_rate": 5.7471845620105925e-06,
+      "loss": 0.4524,
+      "step": 2627
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.512664823617058,
+      "learning_rate": 5.744326740442364e-06,
+      "loss": 0.4385,
+      "step": 2628
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 21.55939086501065,
+      "learning_rate": 5.741468670180737e-06,
+      "loss": 0.5186,
+      "step": 2629
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 20.560690542811283,
+      "learning_rate": 5.738610352180645e-06,
+      "loss": 0.5356,
+      "step": 2630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.948898188809688,
+      "learning_rate": 5.735751787397106e-06,
+      "loss": 0.3574,
+      "step": 2631
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 24.160391432425694,
+      "learning_rate": 5.732892976785218e-06,
+      "loss": 0.4609,
+      "step": 2632
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.426468726837316,
+      "learning_rate": 5.730033921300166e-06,
+      "loss": 0.3936,
+      "step": 2633
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.517440332362233,
+      "learning_rate": 5.7271746218972105e-06,
+      "loss": 0.4478,
+      "step": 2634
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.545964981138855,
+      "learning_rate": 5.724315079531697e-06,
+      "loss": 0.4224,
+      "step": 2635
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.167790368619306,
+      "learning_rate": 5.721455295159053e-06,
+      "loss": 0.4131,
+      "step": 2636
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.35330099624201,
+      "learning_rate": 5.7185952697347844e-06,
+      "loss": 0.5435,
+      "step": 2637
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 17.478710612288964,
+      "learning_rate": 5.71573500421448e-06,
+      "loss": 0.4561,
+      "step": 2638
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 24.345386420729913,
+      "learning_rate": 5.712874499553807e-06,
+      "loss": 0.6011,
+      "step": 2639
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.103831958980157,
+      "learning_rate": 5.710013756708513e-06,
+      "loss": 0.5371,
+      "step": 2640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.723792228428197,
+      "learning_rate": 5.707152776634427e-06,
+      "loss": 0.4746,
+      "step": 2641
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.135521772087765,
+      "learning_rate": 5.704291560287454e-06,
+      "loss": 0.5806,
+      "step": 2642
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.61923834918151,
+      "learning_rate": 5.701430108623578e-06,
+      "loss": 0.5034,
+      "step": 2643
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.595901720364187,
+      "learning_rate": 5.698568422598867e-06,
+      "loss": 0.4658,
+      "step": 2644
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.033670237094597,
+      "learning_rate": 5.69570650316946e-06,
+      "loss": 0.3877,
+      "step": 2645
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.311323300972322,
+      "learning_rate": 5.69284435129158e-06,
+      "loss": 0.5474,
+      "step": 2646
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.16920875171198,
+      "learning_rate": 5.689981967921523e-06,
+      "loss": 0.4785,
+      "step": 2647
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.505309016080973,
+      "learning_rate": 5.6871193540156666e-06,
+      "loss": 0.5347,
+      "step": 2648
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.375726119567252,
+      "learning_rate": 5.684256510530461e-06,
+      "loss": 0.5317,
+      "step": 2649
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.956873091796398,
+      "learning_rate": 5.68139343842244e-06,
+      "loss": 0.4897,
+      "step": 2650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.598497045655397,
+      "learning_rate": 5.678530138648204e-06,
+      "loss": 0.3809,
+      "step": 2651
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.743508614000168,
+      "learning_rate": 5.675666612164436e-06,
+      "loss": 0.4536,
+      "step": 2652
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.871449050564767,
+      "learning_rate": 5.672802859927895e-06,
+      "loss": 0.4248,
+      "step": 2653
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 14.838442238490044,
+      "learning_rate": 5.669938882895412e-06,
+      "loss": 0.4878,
+      "step": 2654
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.731771693748918,
+      "learning_rate": 5.667074682023896e-06,
+      "loss": 0.4346,
+      "step": 2655
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.31273234375937,
+      "learning_rate": 5.664210258270331e-06,
+      "loss": 0.5474,
+      "step": 2656
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.388124322189932,
+      "learning_rate": 5.661345612591771e-06,
+      "loss": 0.3623,
+      "step": 2657
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.206485885563511,
+      "learning_rate": 5.6584807459453515e-06,
+      "loss": 0.4312,
+      "step": 2658
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.641694644438422,
+      "learning_rate": 5.655615659288274e-06,
+      "loss": 0.4653,
+      "step": 2659
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.129600359571274,
+      "learning_rate": 5.652750353577818e-06,
+      "loss": 0.4902,
+      "step": 2660
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 25.692674590295074,
+      "learning_rate": 5.649884829771337e-06,
+      "loss": 0.5063,
+      "step": 2661
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.54868839498731,
+      "learning_rate": 5.6470190888262545e-06,
+      "loss": 0.457,
+      "step": 2662
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.289852844856199,
+      "learning_rate": 5.644153131700067e-06,
+      "loss": 0.4634,
+      "step": 2663
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 15.249877303141151,
+      "learning_rate": 5.6412869593503476e-06,
+      "loss": 0.4956,
+      "step": 2664
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.960042402355725,
+      "learning_rate": 5.638420572734733e-06,
+      "loss": 0.457,
+      "step": 2665
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.138114527336878,
+      "learning_rate": 5.63555397281094e-06,
+      "loss": 0.4009,
+      "step": 2666
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.009079366365885,
+      "learning_rate": 5.632687160536751e-06,
+      "loss": 0.4043,
+      "step": 2667
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.217920799149297,
+      "learning_rate": 5.629820136870022e-06,
+      "loss": 0.4946,
+      "step": 2668
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.941849965542726,
+      "learning_rate": 5.626952902768678e-06,
+      "loss": 0.5039,
+      "step": 2669
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.464934606221962,
+      "learning_rate": 5.624085459190717e-06,
+      "loss": 0.3403,
+      "step": 2670
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.790559734697411,
+      "learning_rate": 5.621217807094202e-06,
+      "loss": 0.353,
+      "step": 2671
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.647088939509905,
+      "learning_rate": 5.618349947437272e-06,
+      "loss": 0.4565,
+      "step": 2672
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.789802694379299,
+      "learning_rate": 5.615481881178132e-06,
+      "loss": 0.4419,
+      "step": 2673
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.442548561932774,
+      "learning_rate": 5.612613609275054e-06,
+      "loss": 0.4175,
+      "step": 2674
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.641515924971305,
+      "learning_rate": 5.609745132686383e-06,
+      "loss": 0.5254,
+      "step": 2675
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.432218412288657,
+      "learning_rate": 5.60687645237053e-06,
+      "loss": 0.4419,
+      "step": 2676
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.15077877053977,
+      "learning_rate": 5.604007569285973e-06,
+      "loss": 0.5625,
+      "step": 2677
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.468525979946765,
+      "learning_rate": 5.6011384843912605e-06,
+      "loss": 0.4912,
+      "step": 2678
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.363832033779826,
+      "learning_rate": 5.598269198645008e-06,
+      "loss": 0.4634,
+      "step": 2679
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.383705824219753,
+      "learning_rate": 5.5953997130058945e-06,
+      "loss": 0.5581,
+      "step": 2680
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.042518750173171,
+      "learning_rate": 5.5925300284326715e-06,
+      "loss": 0.5088,
+      "step": 2681
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.76598841043737,
+      "learning_rate": 5.5896601458841505e-06,
+      "loss": 0.4141,
+      "step": 2682
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.862820427563568,
+      "learning_rate": 5.586790066319217e-06,
+      "loss": 0.4126,
+      "step": 2683
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.59617718975847,
+      "learning_rate": 5.583919790696814e-06,
+      "loss": 0.4648,
+      "step": 2684
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.54969966751114,
+      "learning_rate": 5.581049319975957e-06,
+      "loss": 0.437,
+      "step": 2685
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.884177236890935,
+      "learning_rate": 5.57817865511572e-06,
+      "loss": 0.5962,
+      "step": 2686
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.066486791899976,
+      "learning_rate": 5.575307797075249e-06,
+      "loss": 0.3447,
+      "step": 2687
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.359577468848478,
+      "learning_rate": 5.572436746813748e-06,
+      "loss": 0.48,
+      "step": 2688
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.3029594415371,
+      "learning_rate": 5.5695655052904905e-06,
+      "loss": 0.4507,
+      "step": 2689
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.98486194284849,
+      "learning_rate": 5.566694073464812e-06,
+      "loss": 0.4419,
+      "step": 2690
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.087720324088506,
+      "learning_rate": 5.56382245229611e-06,
+      "loss": 0.4771,
+      "step": 2691
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.174292843227011,
+      "learning_rate": 5.560950642743847e-06,
+      "loss": 0.4883,
+      "step": 2692
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.926555916096607,
+      "learning_rate": 5.558078645767547e-06,
+      "loss": 0.4019,
+      "step": 2693
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.536627620327389,
+      "learning_rate": 5.5552064623267986e-06,
+      "loss": 0.5322,
+      "step": 2694
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.76012508956878,
+      "learning_rate": 5.5523340933812505e-06,
+      "loss": 0.4253,
+      "step": 2695
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.90679754193909,
+      "learning_rate": 5.549461539890616e-06,
+      "loss": 0.4507,
+      "step": 2696
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.420670976992822,
+      "learning_rate": 5.546588802814669e-06,
+      "loss": 0.4375,
+      "step": 2697
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.727787659346541,
+      "learning_rate": 5.543715883113241e-06,
+      "loss": 0.3931,
+      "step": 2698
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.086024809133871,
+      "learning_rate": 5.540842781746231e-06,
+      "loss": 0.3979,
+      "step": 2699
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.011343075606366,
+      "learning_rate": 5.537969499673598e-06,
+      "loss": 0.522,
+      "step": 2700
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.15708782937107,
+      "learning_rate": 5.535096037855353e-06,
+      "loss": 0.459,
+      "step": 2701
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.044008252312308,
+      "learning_rate": 5.532222397251576e-06,
+      "loss": 0.4487,
+      "step": 2702
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.232084819714416,
+      "learning_rate": 5.529348578822403e-06,
+      "loss": 0.5,
+      "step": 2703
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.659043205460982,
+      "learning_rate": 5.526474583528032e-06,
+      "loss": 0.5312,
+      "step": 2704
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.00752567623559,
+      "learning_rate": 5.523600412328716e-06,
+      "loss": 0.5352,
+      "step": 2705
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.033115280436297,
+      "learning_rate": 5.520726066184769e-06,
+      "loss": 0.5396,
+      "step": 2706
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.80140343593761,
+      "learning_rate": 5.517851546056566e-06,
+      "loss": 0.3618,
+      "step": 2707
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.135541690096266,
+      "learning_rate": 5.5149768529045355e-06,
+      "loss": 0.3765,
+      "step": 2708
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.033113241944184,
+      "learning_rate": 5.512101987689168e-06,
+      "loss": 0.3994,
+      "step": 2709
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.091847973750784,
+      "learning_rate": 5.509226951371006e-06,
+      "loss": 0.4431,
+      "step": 2710
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.550320506061626,
+      "learning_rate": 5.506351744910654e-06,
+      "loss": 0.4248,
+      "step": 2711
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.907476430192875,
+      "learning_rate": 5.503476369268773e-06,
+      "loss": 0.4434,
+      "step": 2712
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.662632331178548,
+      "learning_rate": 5.50060082540608e-06,
+      "loss": 0.3901,
+      "step": 2713
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.946005732308972,
+      "learning_rate": 5.4977251142833445e-06,
+      "loss": 0.5063,
+      "step": 2714
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.764860363419759,
+      "learning_rate": 5.494849236861397e-06,
+      "loss": 0.3701,
+      "step": 2715
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.57151587816436,
+      "learning_rate": 5.491973194101122e-06,
+      "loss": 0.4678,
+      "step": 2716
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.496999220354681,
+      "learning_rate": 5.4890969869634606e-06,
+      "loss": 0.4072,
+      "step": 2717
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.955739348741126,
+      "learning_rate": 5.486220616409403e-06,
+      "loss": 0.417,
+      "step": 2718
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.480253591668859,
+      "learning_rate": 5.4833440834e-06,
+      "loss": 0.4419,
+      "step": 2719
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.417238840494207,
+      "learning_rate": 5.480467388896353e-06,
+      "loss": 0.4951,
+      "step": 2720
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.778607214982058,
+      "learning_rate": 5.477590533859623e-06,
+      "loss": 0.4058,
+      "step": 2721
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.095377405354045,
+      "learning_rate": 5.474713519251018e-06,
+      "loss": 0.501,
+      "step": 2722
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.109288662149696,
+      "learning_rate": 5.471836346031802e-06,
+      "loss": 0.4067,
+      "step": 2723
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.606747867678797,
+      "learning_rate": 5.468959015163293e-06,
+      "loss": 0.4321,
+      "step": 2724
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.521320145257938,
+      "learning_rate": 5.46608152760686e-06,
+      "loss": 0.3298,
+      "step": 2725
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.610539194603652,
+      "learning_rate": 5.463203884323926e-06,
+      "loss": 0.541,
+      "step": 2726
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.273515699596409,
+      "learning_rate": 5.460326086275964e-06,
+      "loss": 0.5078,
+      "step": 2727
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.111661385131107,
+      "learning_rate": 5.4574481344245015e-06,
+      "loss": 0.4756,
+      "step": 2728
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.977805736536428,
+      "learning_rate": 5.454570029731115e-06,
+      "loss": 0.4663,
+      "step": 2729
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.975752561343054,
+      "learning_rate": 5.451691773157431e-06,
+      "loss": 0.4971,
+      "step": 2730
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.32339975667838,
+      "learning_rate": 5.448813365665129e-06,
+      "loss": 0.5049,
+      "step": 2731
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.315123889810822,
+      "learning_rate": 5.44593480821594e-06,
+      "loss": 0.5132,
+      "step": 2732
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.29119410355911,
+      "learning_rate": 5.443056101771643e-06,
+      "loss": 0.4316,
+      "step": 2733
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.289099728879002,
+      "learning_rate": 5.44017724729407e-06,
+      "loss": 0.3882,
+      "step": 2734
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.888644836261584,
+      "learning_rate": 5.437298245745093e-06,
+      "loss": 0.4331,
+      "step": 2735
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 48.21181573416906,
+      "learning_rate": 5.434419098086645e-06,
+      "loss": 0.5977,
+      "step": 2736
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.641134447756468,
+      "learning_rate": 5.431539805280702e-06,
+      "loss": 0.3945,
+      "step": 2737
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.965847623472408,
+      "learning_rate": 5.428660368289289e-06,
+      "loss": 0.4741,
+      "step": 2738
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.10664209523203,
+      "learning_rate": 5.42578078807448e-06,
+      "loss": 0.4512,
+      "step": 2739
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.681304900537418,
+      "learning_rate": 5.422901065598395e-06,
+      "loss": 0.4297,
+      "step": 2740
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.739932626389907,
+      "learning_rate": 5.4200212018232024e-06,
+      "loss": 0.3513,
+      "step": 2741
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.846041535480271,
+      "learning_rate": 5.41714119771112e-06,
+      "loss": 0.4302,
+      "step": 2742
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.984598238079338,
+      "learning_rate": 5.414261054224412e-06,
+      "loss": 0.4033,
+      "step": 2743
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.933328673227695,
+      "learning_rate": 5.411380772325383e-06,
+      "loss": 0.5703,
+      "step": 2744
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.854711059795555,
+      "learning_rate": 5.408500352976392e-06,
+      "loss": 0.5151,
+      "step": 2745
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.111285712834931,
+      "learning_rate": 5.40561979713984e-06,
+      "loss": 0.4084,
+      "step": 2746
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.242923858855155,
+      "learning_rate": 5.402739105778175e-06,
+      "loss": 0.4956,
+      "step": 2747
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.954715434651323,
+      "learning_rate": 5.399858279853889e-06,
+      "loss": 0.4229,
+      "step": 2748
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.3157730974535315,
+      "learning_rate": 5.39697732032952e-06,
+      "loss": 0.3398,
+      "step": 2749
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.265906166274753,
+      "learning_rate": 5.394096228167648e-06,
+      "loss": 0.4565,
+      "step": 2750
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.198284629775742,
+      "learning_rate": 5.391215004330903e-06,
+      "loss": 0.3813,
+      "step": 2751
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.31796639425881,
+      "learning_rate": 5.388333649781951e-06,
+      "loss": 0.4683,
+      "step": 2752
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.01789404392901,
+      "learning_rate": 5.3854521654835105e-06,
+      "loss": 0.4502,
+      "step": 2753
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.306465904258715,
+      "learning_rate": 5.3825705523983366e-06,
+      "loss": 0.4351,
+      "step": 2754
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.519624094511274,
+      "learning_rate": 5.37968881148923e-06,
+      "loss": 0.4707,
+      "step": 2755
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.226114092265847,
+      "learning_rate": 5.376806943719033e-06,
+      "loss": 0.4814,
+      "step": 2756
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.840072523952024,
+      "learning_rate": 5.373924950050633e-06,
+      "loss": 0.4194,
+      "step": 2757
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.039602192178293,
+      "learning_rate": 5.371042831446957e-06,
+      "loss": 0.5571,
+      "step": 2758
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.416861116474312,
+      "learning_rate": 5.3681605888709755e-06,
+      "loss": 0.4858,
+      "step": 2759
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.51902810805772,
+      "learning_rate": 5.365278223285698e-06,
+      "loss": 0.7158,
+      "step": 2760
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.508514981825446,
+      "learning_rate": 5.362395735654175e-06,
+      "loss": 0.5264,
+      "step": 2761
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.292618872818796,
+      "learning_rate": 5.3595131269395015e-06,
+      "loss": 0.4668,
+      "step": 2762
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.600159324345311,
+      "learning_rate": 5.356630398104811e-06,
+      "loss": 0.4302,
+      "step": 2763
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.912551709797079,
+      "learning_rate": 5.353747550113274e-06,
+      "loss": 0.3501,
+      "step": 2764
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.525523170385298,
+      "learning_rate": 5.350864583928106e-06,
+      "loss": 0.4648,
+      "step": 2765
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.191292892938462,
+      "learning_rate": 5.347981500512558e-06,
+      "loss": 0.3613,
+      "step": 2766
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.57847648591052,
+      "learning_rate": 5.345098300829924e-06,
+      "loss": 0.5903,
+      "step": 2767
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.335634989474457,
+      "learning_rate": 5.342214985843534e-06,
+      "loss": 0.4829,
+      "step": 2768
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.001861539425636,
+      "learning_rate": 5.339331556516755e-06,
+      "loss": 0.4902,
+      "step": 2769
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.013265462647908,
+      "learning_rate": 5.336448013812996e-06,
+      "loss": 0.4912,
+      "step": 2770
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.809090114106525,
+      "learning_rate": 5.333564358695701e-06,
+      "loss": 0.4336,
+      "step": 2771
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.281886038205364,
+      "learning_rate": 5.330680592128355e-06,
+      "loss": 0.4937,
+      "step": 2772
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.150545229018581,
+      "learning_rate": 5.3277967150744755e-06,
+      "loss": 0.4189,
+      "step": 2773
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.345814588398108,
+      "learning_rate": 5.324912728497621e-06,
+      "loss": 0.4009,
+      "step": 2774
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.091474666204597,
+      "learning_rate": 5.322028633361386e-06,
+      "loss": 0.562,
+      "step": 2775
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.754059247137123,
+      "learning_rate": 5.319144430629397e-06,
+      "loss": 0.4702,
+      "step": 2776
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.586444942071105,
+      "learning_rate": 5.316260121265323e-06,
+      "loss": 0.4595,
+      "step": 2777
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.6680666422755275,
+      "learning_rate": 5.313375706232864e-06,
+      "loss": 0.4072,
+      "step": 2778
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.091604704175747,
+      "learning_rate": 5.310491186495757e-06,
+      "loss": 0.4297,
+      "step": 2779
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.434600968392432,
+      "learning_rate": 5.307606563017772e-06,
+      "loss": 0.3491,
+      "step": 2780
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.651832210814672,
+      "learning_rate": 5.304721836762717e-06,
+      "loss": 0.4087,
+      "step": 2781
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.123238252935426,
+      "learning_rate": 5.301837008694433e-06,
+      "loss": 0.418,
+      "step": 2782
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.623468647704522,
+      "learning_rate": 5.298952079776794e-06,
+      "loss": 0.4336,
+      "step": 2783
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.961097932950663,
+      "learning_rate": 5.296067050973709e-06,
+      "loss": 0.4009,
+      "step": 2784
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.004467440225236,
+      "learning_rate": 5.29318192324912e-06,
+      "loss": 0.4385,
+      "step": 2785
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.937849862725015,
+      "learning_rate": 5.290296697566999e-06,
+      "loss": 0.4731,
+      "step": 2786
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.851718115774863,
+      "learning_rate": 5.287411374891356e-06,
+      "loss": 0.4741,
+      "step": 2787
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.247981728977928,
+      "learning_rate": 5.284525956186231e-06,
+      "loss": 0.4355,
+      "step": 2788
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.992715571855904,
+      "learning_rate": 5.281640442415695e-06,
+      "loss": 0.5229,
+      "step": 2789
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.344871517460058,
+      "learning_rate": 5.278754834543852e-06,
+      "loss": 0.4722,
+      "step": 2790
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.839020938381688,
+      "learning_rate": 5.275869133534838e-06,
+      "loss": 0.4785,
+      "step": 2791
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.33203373375532,
+      "learning_rate": 5.272983340352818e-06,
+      "loss": 0.5005,
+      "step": 2792
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.957846186020241,
+      "learning_rate": 5.270097455961991e-06,
+      "loss": 0.4048,
+      "step": 2793
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.793496740544484,
+      "learning_rate": 5.267211481326584e-06,
+      "loss": 0.3716,
+      "step": 2794
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.547713987824872,
+      "learning_rate": 5.264325417410854e-06,
+      "loss": 0.3901,
+      "step": 2795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.683395761109471,
+      "learning_rate": 5.261439265179089e-06,
+      "loss": 0.4375,
+      "step": 2796
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.88713456122952,
+      "learning_rate": 5.258553025595605e-06,
+      "loss": 0.4854,
+      "step": 2797
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.15203401175118,
+      "learning_rate": 5.255666699624749e-06,
+      "loss": 0.4604,
+      "step": 2798
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.439586255360073,
+      "learning_rate": 5.252780288230899e-06,
+      "loss": 0.4316,
+      "step": 2799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.690633931887922,
+      "learning_rate": 5.249893792378454e-06,
+      "loss": 0.4546,
+      "step": 2800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.5972252881605,
+      "learning_rate": 5.24700721303185e-06,
+      "loss": 0.3784,
+      "step": 2801
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.09788081335953,
+      "learning_rate": 5.244120551155544e-06,
+      "loss": 0.4463,
+      "step": 2802
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.183682722164911,
+      "learning_rate": 5.241233807714024e-06,
+      "loss": 0.2798,
+      "step": 2803
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.492711477273929,
+      "learning_rate": 5.238346983671805e-06,
+      "loss": 0.321,
+      "step": 2804
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.78545087675637,
+      "learning_rate": 5.235460079993429e-06,
+      "loss": 0.3223,
+      "step": 2805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.184465239974276,
+      "learning_rate": 5.232573097643462e-06,
+      "loss": 0.2476,
+      "step": 2806
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.38970510024269,
+      "learning_rate": 5.229686037586502e-06,
+      "loss": 0.3081,
+      "step": 2807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.337615749238382,
+      "learning_rate": 5.226798900787167e-06,
+      "loss": 0.3176,
+      "step": 2808
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.065520523779682,
+      "learning_rate": 5.223911688210104e-06,
+      "loss": 0.2876,
+      "step": 2809
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.050626203183986,
+      "learning_rate": 5.221024400819983e-06,
+      "loss": 0.3157,
+      "step": 2810
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.158021790081161,
+      "learning_rate": 5.218137039581504e-06,
+      "loss": 0.2568,
+      "step": 2811
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.581251837695142,
+      "learning_rate": 5.215249605459382e-06,
+      "loss": 0.3867,
+      "step": 2812
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.194547291558267,
+      "learning_rate": 5.212362099418369e-06,
+      "loss": 0.26,
+      "step": 2813
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.028079795632454,
+      "learning_rate": 5.2094745224232306e-06,
+      "loss": 0.2729,
+      "step": 2814
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.016565575074754,
+      "learning_rate": 5.206586875438759e-06,
+      "loss": 0.4033,
+      "step": 2815
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.122323867559514,
+      "learning_rate": 5.203699159429773e-06,
+      "loss": 0.2603,
+      "step": 2816
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.258210540115515,
+      "learning_rate": 5.200811375361112e-06,
+      "loss": 0.3042,
+      "step": 2817
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.530978379942863,
+      "learning_rate": 5.197923524197639e-06,
+      "loss": 0.2639,
+      "step": 2818
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.244056210394618,
+      "learning_rate": 5.195035606904237e-06,
+      "loss": 0.3047,
+      "step": 2819
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.976152241105012,
+      "learning_rate": 5.1921476244458135e-06,
+      "loss": 0.2971,
+      "step": 2820
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.62736935340582,
+      "learning_rate": 5.189259577787297e-06,
+      "loss": 0.2849,
+      "step": 2821
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.841839115792192,
+      "learning_rate": 5.186371467893638e-06,
+      "loss": 0.2444,
+      "step": 2822
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.733355341593247,
+      "learning_rate": 5.1834832957298075e-06,
+      "loss": 0.2522,
+      "step": 2823
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.20366363366365,
+      "learning_rate": 5.180595062260797e-06,
+      "loss": 0.1924,
+      "step": 2824
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.665148545436182,
+      "learning_rate": 5.177706768451619e-06,
+      "loss": 0.2285,
+      "step": 2825
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.249918360609486,
+      "learning_rate": 5.174818415267308e-06,
+      "loss": 0.1785,
+      "step": 2826
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.068358421491178,
+      "learning_rate": 5.1719300036729135e-06,
+      "loss": 0.2644,
+      "step": 2827
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.250338443568086,
+      "learning_rate": 5.169041534633511e-06,
+      "loss": 0.2855,
+      "step": 2828
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.981055766027776,
+      "learning_rate": 5.166153009114188e-06,
+      "loss": 0.2559,
+      "step": 2829
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.58373133194685,
+      "learning_rate": 5.163264428080057e-06,
+      "loss": 0.229,
+      "step": 2830
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.426613456253506,
+      "learning_rate": 5.160375792496246e-06,
+      "loss": 0.2656,
+      "step": 2831
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 28.624808178448106,
+      "learning_rate": 5.157487103327901e-06,
+      "loss": 0.4053,
+      "step": 2832
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.440233476330917,
+      "learning_rate": 5.1545983615401885e-06,
+      "loss": 0.2058,
+      "step": 2833
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.486881898609434,
+      "learning_rate": 5.151709568098289e-06,
+      "loss": 0.3364,
+      "step": 2834
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.482819056787882,
+      "learning_rate": 5.1488207239674036e-06,
+      "loss": 0.2424,
+      "step": 2835
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.817322942492323,
+      "learning_rate": 5.145931830112748e-06,
+      "loss": 0.2766,
+      "step": 2836
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.581341500866515,
+      "learning_rate": 5.1430428874995554e-06,
+      "loss": 0.2036,
+      "step": 2837
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.851916462924994,
+      "learning_rate": 5.140153897093076e-06,
+      "loss": 0.2317,
+      "step": 2838
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.85859461783392,
+      "learning_rate": 5.1372648598585725e-06,
+      "loss": 0.3672,
+      "step": 2839
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.51481537393519,
+      "learning_rate": 5.134375776761329e-06,
+      "loss": 0.2026,
+      "step": 2840
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.011970306520274,
+      "learning_rate": 5.131486648766642e-06,
+      "loss": 0.2827,
+      "step": 2841
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.623869309435559,
+      "learning_rate": 5.1285974768398205e-06,
+      "loss": 0.2432,
+      "step": 2842
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.971151752756175,
+      "learning_rate": 5.125708261946192e-06,
+      "loss": 0.2737,
+      "step": 2843
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.16313895403444,
+      "learning_rate": 5.122819005051096e-06,
+      "loss": 0.2595,
+      "step": 2844
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.834344804916485,
+      "learning_rate": 5.119929707119889e-06,
+      "loss": 0.3093,
+      "step": 2845
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.59230099518939,
+      "learning_rate": 5.117040369117937e-06,
+      "loss": 0.269,
+      "step": 2846
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.261755241964481,
+      "learning_rate": 5.114150992010621e-06,
+      "loss": 0.2363,
+      "step": 2847
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.845223713454207,
+      "learning_rate": 5.1112615767633385e-06,
+      "loss": 0.3608,
+      "step": 2848
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.952115168213247,
+      "learning_rate": 5.108372124341494e-06,
+      "loss": 0.2449,
+      "step": 2849
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.962593059947745,
+      "learning_rate": 5.105482635710509e-06,
+      "loss": 0.281,
+      "step": 2850
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.47356055701605,
+      "learning_rate": 5.102593111835815e-06,
+      "loss": 0.303,
+      "step": 2851
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.090099328504023,
+      "learning_rate": 5.099703553682854e-06,
+      "loss": 0.2979,
+      "step": 2852
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.384435155829522,
+      "learning_rate": 5.096813962217086e-06,
+      "loss": 0.2117,
+      "step": 2853
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.707780782866086,
+      "learning_rate": 5.093924338403971e-06,
+      "loss": 0.3604,
+      "step": 2854
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.41635335383991,
+      "learning_rate": 5.091034683208988e-06,
+      "loss": 0.3076,
+      "step": 2855
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.542784265367086,
+      "learning_rate": 5.088144997597627e-06,
+      "loss": 0.2373,
+      "step": 2856
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.300403445154583,
+      "learning_rate": 5.085255282535383e-06,
+      "loss": 0.3362,
+      "step": 2857
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.820500247221656,
+      "learning_rate": 5.082365538987765e-06,
+      "loss": 0.3486,
+      "step": 2858
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.963174706002892,
+      "learning_rate": 5.079475767920289e-06,
+      "loss": 0.2498,
+      "step": 2859
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.924691223131479,
+      "learning_rate": 5.076585970298481e-06,
+      "loss": 0.3442,
+      "step": 2860
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.618086517328626,
+      "learning_rate": 5.073696147087878e-06,
+      "loss": 0.2406,
+      "step": 2861
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.971342596280728,
+      "learning_rate": 5.070806299254023e-06,
+      "loss": 0.207,
+      "step": 2862
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.74822806936663,
+      "learning_rate": 5.067916427762466e-06,
+      "loss": 0.2971,
+      "step": 2863
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.841852154556609,
+      "learning_rate": 5.0650265335787685e-06,
+      "loss": 0.3123,
+      "step": 2864
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.028441837620969,
+      "learning_rate": 5.062136617668497e-06,
+      "loss": 0.2405,
+      "step": 2865
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.77779210531338,
+      "learning_rate": 5.059246680997228e-06,
+      "loss": 0.2537,
+      "step": 2866
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.536986749042004,
+      "learning_rate": 5.05635672453054e-06,
+      "loss": 0.3887,
+      "step": 2867
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.088559869194436,
+      "learning_rate": 5.053466749234023e-06,
+      "loss": 0.2421,
+      "step": 2868
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.184815205436934,
+      "learning_rate": 5.050576756073272e-06,
+      "loss": 0.3806,
+      "step": 2869
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.23293466912918,
+      "learning_rate": 5.047686746013888e-06,
+      "loss": 0.2493,
+      "step": 2870
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.369014628236714,
+      "learning_rate": 5.044796720021474e-06,
+      "loss": 0.2585,
+      "step": 2871
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.817280948904253,
+      "learning_rate": 5.041906679061643e-06,
+      "loss": 0.2686,
+      "step": 2872
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.400824784637889,
+      "learning_rate": 5.039016624100013e-06,
+      "loss": 0.3796,
+      "step": 2873
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.411506271031973,
+      "learning_rate": 5.036126556102202e-06,
+      "loss": 0.2939,
+      "step": 2874
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.73804711630787,
+      "learning_rate": 5.033236476033838e-06,
+      "loss": 0.2456,
+      "step": 2875
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.075734103259556,
+      "learning_rate": 5.0303463848605495e-06,
+      "loss": 0.2654,
+      "step": 2876
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.79768975539953,
+      "learning_rate": 5.027456283547969e-06,
+      "loss": 0.2686,
+      "step": 2877
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.816952929731688,
+      "learning_rate": 5.0245661730617344e-06,
+      "loss": 0.2668,
+      "step": 2878
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.601781932049883,
+      "learning_rate": 5.0216760543674855e-06,
+      "loss": 0.28,
+      "step": 2879
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.279934779358646,
+      "learning_rate": 5.0187859284308635e-06,
+      "loss": 0.3567,
+      "step": 2880
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.09264143614305,
+      "learning_rate": 5.015895796217514e-06,
+      "loss": 0.2632,
+      "step": 2881
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.72906261801957,
+      "learning_rate": 5.013005658693083e-06,
+      "loss": 0.2432,
+      "step": 2882
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.904243447527767,
+      "learning_rate": 5.01011551682322e-06,
+      "loss": 0.3123,
+      "step": 2883
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.56158421894508,
+      "learning_rate": 5.007225371573573e-06,
+      "loss": 0.2267,
+      "step": 2884
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.556632021145733,
+      "learning_rate": 5.004335223909797e-06,
+      "loss": 0.3115,
+      "step": 2885
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.697069853449166,
+      "learning_rate": 5.0014450747975416e-06,
+      "loss": 0.2358,
+      "step": 2886
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.361640909885745,
+      "learning_rate": 4.998554925202459e-06,
+      "loss": 0.2517,
+      "step": 2887
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 18.22972936773402,
+      "learning_rate": 4.995664776090204e-06,
+      "loss": 0.4097,
+      "step": 2888
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.143973237259363,
+      "learning_rate": 4.9927746284264275e-06,
+      "loss": 0.2427,
+      "step": 2889
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.457619103227681,
+      "learning_rate": 4.9898844831767826e-06,
+      "loss": 0.2324,
+      "step": 2890
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.882434995590293,
+      "learning_rate": 4.98699434130692e-06,
+      "loss": 0.2408,
+      "step": 2891
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.593705192329265,
+      "learning_rate": 4.984104203782488e-06,
+      "loss": 0.3152,
+      "step": 2892
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.4272420996640856,
+      "learning_rate": 4.981214071569139e-06,
+      "loss": 0.2415,
+      "step": 2893
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.71270569735412,
+      "learning_rate": 4.978323945632515e-06,
+      "loss": 0.2395,
+      "step": 2894
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.869809390534899,
+      "learning_rate": 4.975433826938267e-06,
+      "loss": 0.2932,
+      "step": 2895
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.64672924292059,
+      "learning_rate": 4.972543716452031e-06,
+      "loss": 0.2837,
+      "step": 2896
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.846800699280617,
+      "learning_rate": 4.969653615139452e-06,
+      "loss": 0.2664,
+      "step": 2897
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.340053708726927,
+      "learning_rate": 4.966763523966163e-06,
+      "loss": 0.4248,
+      "step": 2898
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.986583690207297,
+      "learning_rate": 4.963873443897799e-06,
+      "loss": 0.2932,
+      "step": 2899
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.877813445906238,
+      "learning_rate": 4.96098337589999e-06,
+      "loss": 0.2527,
+      "step": 2900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.751716315394958,
+      "learning_rate": 4.958093320938358e-06,
+      "loss": 0.2856,
+      "step": 2901
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.259213554000896,
+      "learning_rate": 4.955203279978529e-06,
+      "loss": 0.312,
+      "step": 2902
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.869350826007125,
+      "learning_rate": 4.952313253986114e-06,
+      "loss": 0.302,
+      "step": 2903
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.818949638241829,
+      "learning_rate": 4.9494232439267296e-06,
+      "loss": 0.2224,
+      "step": 2904
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.567323569857239,
+      "learning_rate": 4.946533250765977e-06,
+      "loss": 0.2588,
+      "step": 2905
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.916886883895785,
+      "learning_rate": 4.943643275469461e-06,
+      "loss": 0.2678,
+      "step": 2906
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.444466859746376,
+      "learning_rate": 4.940753319002773e-06,
+      "loss": 0.2598,
+      "step": 2907
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.782074101523976,
+      "learning_rate": 4.937863382331504e-06,
+      "loss": 0.3218,
+      "step": 2908
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.596863218172778,
+      "learning_rate": 4.934973466421234e-06,
+      "loss": 0.3345,
+      "step": 2909
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.650320442975373,
+      "learning_rate": 4.932083572237535e-06,
+      "loss": 0.2859,
+      "step": 2910
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.743783621991485,
+      "learning_rate": 4.92919370074598e-06,
+      "loss": 0.2761,
+      "step": 2911
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.118954150902352,
+      "learning_rate": 4.926303852912123e-06,
+      "loss": 0.3367,
+      "step": 2912
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.571991535980812,
+      "learning_rate": 4.9234140297015204e-06,
+      "loss": 0.2288,
+      "step": 2913
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.628440666155255,
+      "learning_rate": 4.920524232079712e-06,
+      "loss": 0.3997,
+      "step": 2914
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.12851494373273,
+      "learning_rate": 4.917634461012238e-06,
+      "loss": 0.3601,
+      "step": 2915
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.147043201032695,
+      "learning_rate": 4.914744717464617e-06,
+      "loss": 0.3708,
+      "step": 2916
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.657303874351186,
+      "learning_rate": 4.911855002402375e-06,
+      "loss": 0.3149,
+      "step": 2917
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.918834616338623,
+      "learning_rate": 4.908965316791014e-06,
+      "loss": 0.1868,
+      "step": 2918
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.338228828504006,
+      "learning_rate": 4.906075661596031e-06,
+      "loss": 0.2456,
+      "step": 2919
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.717111899061246,
+      "learning_rate": 4.903186037782917e-06,
+      "loss": 0.2471,
+      "step": 2920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.048131071947918,
+      "learning_rate": 4.900296446317146e-06,
+      "loss": 0.2698,
+      "step": 2921
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.548570881430447,
+      "learning_rate": 4.897406888164187e-06,
+      "loss": 0.3094,
+      "step": 2922
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.795066171226647,
+      "learning_rate": 4.8945173642894915e-06,
+      "loss": 0.2437,
+      "step": 2923
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.874111871990417,
+      "learning_rate": 4.8916278756585074e-06,
+      "loss": 0.2549,
+      "step": 2924
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.369183327295028,
+      "learning_rate": 4.888738423236664e-06,
+      "loss": 0.2812,
+      "step": 2925
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 15.705046693449054,
+      "learning_rate": 4.88584900798938e-06,
+      "loss": 0.2434,
+      "step": 2926
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.69346857544719,
+      "learning_rate": 4.882959630882066e-06,
+      "loss": 0.2947,
+      "step": 2927
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.467959400291866,
+      "learning_rate": 4.8800702928801124e-06,
+      "loss": 0.2712,
+      "step": 2928
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.203239862503867,
+      "learning_rate": 4.8771809949489056e-06,
+      "loss": 0.3223,
+      "step": 2929
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.718444322534985,
+      "learning_rate": 4.874291738053809e-06,
+      "loss": 0.479,
+      "step": 2930
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.328069951779979,
+      "learning_rate": 4.871402523160181e-06,
+      "loss": 0.2515,
+      "step": 2931
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.698570030098262,
+      "learning_rate": 4.868513351233359e-06,
+      "loss": 0.3232,
+      "step": 2932
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.399731299784895,
+      "learning_rate": 4.865624223238672e-06,
+      "loss": 0.3057,
+      "step": 2933
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.15982457143095,
+      "learning_rate": 4.862735140141428e-06,
+      "loss": 0.3362,
+      "step": 2934
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.120995256901526,
+      "learning_rate": 4.859846102906927e-06,
+      "loss": 0.2119,
+      "step": 2935
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.790485443872855,
+      "learning_rate": 4.856957112500446e-06,
+      "loss": 0.26,
+      "step": 2936
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.804955003772951,
+      "learning_rate": 4.854068169887254e-06,
+      "loss": 0.2522,
+      "step": 2937
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.67276784524615,
+      "learning_rate": 4.851179276032598e-06,
+      "loss": 0.2461,
+      "step": 2938
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.029967703322102,
+      "learning_rate": 4.848290431901712e-06,
+      "loss": 0.2825,
+      "step": 2939
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.701298043559392,
+      "learning_rate": 4.845401638459813e-06,
+      "loss": 0.3257,
+      "step": 2940
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 24.527517306096676,
+      "learning_rate": 4.8425128966721e-06,
+      "loss": 0.3032,
+      "step": 2941
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.060243434857231,
+      "learning_rate": 4.8396242075037555e-06,
+      "loss": 0.2424,
+      "step": 2942
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.110048308538667,
+      "learning_rate": 4.836735571919946e-06,
+      "loss": 0.2317,
+      "step": 2943
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.270855038860887,
+      "learning_rate": 4.833846990885813e-06,
+      "loss": 0.2925,
+      "step": 2944
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.546059086949068,
+      "learning_rate": 4.830958465366492e-06,
+      "loss": 0.3115,
+      "step": 2945
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.147060376074828,
+      "learning_rate": 4.828069996327088e-06,
+      "loss": 0.3071,
+      "step": 2946
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.89733878398577,
+      "learning_rate": 4.825181584732695e-06,
+      "loss": 0.354,
+      "step": 2947
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.098924759848195,
+      "learning_rate": 4.822293231548382e-06,
+      "loss": 0.3223,
+      "step": 2948
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.59459222498052,
+      "learning_rate": 4.819404937739205e-06,
+      "loss": 0.2988,
+      "step": 2949
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.9228442130931,
+      "learning_rate": 4.816516704270194e-06,
+      "loss": 0.1917,
+      "step": 2950
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.304262104161308,
+      "learning_rate": 4.813628532106363e-06,
+      "loss": 0.2844,
+      "step": 2951
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.91478464087612,
+      "learning_rate": 4.810740422212705e-06,
+      "loss": 0.2534,
+      "step": 2952
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.369008662112533,
+      "learning_rate": 4.807852375554188e-06,
+      "loss": 0.248,
+      "step": 2953
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.194937078896238,
+      "learning_rate": 4.804964393095765e-06,
+      "loss": 0.2935,
+      "step": 2954
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.543922616402899,
+      "learning_rate": 4.802076475802362e-06,
+      "loss": 0.2605,
+      "step": 2955
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.955492140429202,
+      "learning_rate": 4.799188624638889e-06,
+      "loss": 0.207,
+      "step": 2956
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.777530819834624,
+      "learning_rate": 4.796300840570227e-06,
+      "loss": 0.2734,
+      "step": 2957
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.600327023654208,
+      "learning_rate": 4.793413124561243e-06,
+      "loss": 0.2236,
+      "step": 2958
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.515292072213855,
+      "learning_rate": 4.790525477576773e-06,
+      "loss": 0.2729,
+      "step": 2959
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.024728170748826,
+      "learning_rate": 4.7876379005816325e-06,
+      "loss": 0.192,
+      "step": 2960
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.964190783912358,
+      "learning_rate": 4.784750394540619e-06,
+      "loss": 0.2013,
+      "step": 2961
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.784790275659192,
+      "learning_rate": 4.781862960418498e-06,
+      "loss": 0.2595,
+      "step": 2962
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.133720980277968,
+      "learning_rate": 4.778975599180019e-06,
+      "loss": 0.3374,
+      "step": 2963
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.406518179067787,
+      "learning_rate": 4.776088311789897e-06,
+      "loss": 0.2964,
+      "step": 2964
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.188477238179528,
+      "learning_rate": 4.773201099212835e-06,
+      "loss": 0.1699,
+      "step": 2965
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.483178808062954,
+      "learning_rate": 4.770313962413499e-06,
+      "loss": 0.2429,
+      "step": 2966
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.54657457090472,
+      "learning_rate": 4.767426902356539e-06,
+      "loss": 0.2224,
+      "step": 2967
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.491438595730546,
+      "learning_rate": 4.7645399200065745e-06,
+      "loss": 0.3599,
+      "step": 2968
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.84370673100882,
+      "learning_rate": 4.761653016328197e-06,
+      "loss": 0.2827,
+      "step": 2969
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.737199358136072,
+      "learning_rate": 4.758766192285979e-06,
+      "loss": 0.1892,
+      "step": 2970
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.368842569814777,
+      "learning_rate": 4.755879448844458e-06,
+      "loss": 0.3108,
+      "step": 2971
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.722321883118072,
+      "learning_rate": 4.752992786968153e-06,
+      "loss": 0.2773,
+      "step": 2972
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.416553515826045,
+      "learning_rate": 4.750106207621546e-06,
+      "loss": 0.2751,
+      "step": 2973
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.678335643158764,
+      "learning_rate": 4.747219711769103e-06,
+      "loss": 0.2622,
+      "step": 2974
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.178387254968198,
+      "learning_rate": 4.74433330037525e-06,
+      "loss": 0.2866,
+      "step": 2975
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.722021862525153,
+      "learning_rate": 4.741446974404396e-06,
+      "loss": 0.2549,
+      "step": 2976
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.939812320748437,
+      "learning_rate": 4.738560734820914e-06,
+      "loss": 0.2354,
+      "step": 2977
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.457539144367189,
+      "learning_rate": 4.735674582589147e-06,
+      "loss": 0.2371,
+      "step": 2978
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.031971470356236,
+      "learning_rate": 4.732788518673418e-06,
+      "loss": 0.283,
+      "step": 2979
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.290069870349917,
+      "learning_rate": 4.729902544038009e-06,
+      "loss": 0.262,
+      "step": 2980
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.646079051872375,
+      "learning_rate": 4.7270166596471825e-06,
+      "loss": 0.2522,
+      "step": 2981
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.643876500840022,
+      "learning_rate": 4.724130866465163e-06,
+      "loss": 0.26,
+      "step": 2982
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.46389033776257,
+      "learning_rate": 4.721245165456149e-06,
+      "loss": 0.2505,
+      "step": 2983
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.041824511182282,
+      "learning_rate": 4.7183595575843055e-06,
+      "loss": 0.2754,
+      "step": 2984
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.441670864552433,
+      "learning_rate": 4.715474043813771e-06,
+      "loss": 0.2422,
+      "step": 2985
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.272432898291353,
+      "learning_rate": 4.712588625108645e-06,
+      "loss": 0.2637,
+      "step": 2986
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.77655861565552,
+      "learning_rate": 4.709703302433003e-06,
+      "loss": 0.2734,
+      "step": 2987
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.616444934585852,
+      "learning_rate": 4.706818076750883e-06,
+      "loss": 0.3889,
+      "step": 2988
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.01033899307128,
+      "learning_rate": 4.703932949026291e-06,
+      "loss": 0.2466,
+      "step": 2989
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.858289061563696,
+      "learning_rate": 4.701047920223207e-06,
+      "loss": 0.293,
+      "step": 2990
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.866643961827666,
+      "learning_rate": 4.6981629913055674e-06,
+      "loss": 0.3655,
+      "step": 2991
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.132085368005846,
+      "learning_rate": 4.695278163237284e-06,
+      "loss": 0.1934,
+      "step": 2992
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.039198152601585,
+      "learning_rate": 4.692393436982229e-06,
+      "loss": 0.23,
+      "step": 2993
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.3107040441794,
+      "learning_rate": 4.689508813504246e-06,
+      "loss": 0.2344,
+      "step": 2994
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.629823411245509,
+      "learning_rate": 4.686624293767138e-06,
+      "loss": 0.2749,
+      "step": 2995
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.154888487363252,
+      "learning_rate": 4.683739878734678e-06,
+      "loss": 0.2816,
+      "step": 2996
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.03652875316943,
+      "learning_rate": 4.6808555693706045e-06,
+      "loss": 0.2156,
+      "step": 2997
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.487451590380616,
+      "learning_rate": 4.677971366638616e-06,
+      "loss": 0.2493,
+      "step": 2998
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.49534880905934,
+      "learning_rate": 4.67508727150238e-06,
+      "loss": 0.25,
+      "step": 2999
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.968859660999014,
+      "learning_rate": 4.672203284925525e-06,
+      "loss": 0.2004,
+      "step": 3000
+    },
+    {
+      "epoch": 2.14,
+      "eval_avg_AUC": 0.7804411891516083,
+      "eval_avg_Accuracy": 0.6809101458885941,
+      "eval_avg_Accuracy-right": 0.8848963088561367,
+      "eval_avg_Accuracy-wrong": 0.32522174209688426,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6521728732768947,
+      "eval_last_AUC": 0.798815075101747,
+      "eval_last_Accuracy": 0.7263759946949602,
+      "eval_last_Accuracy-right": 0.8329855223685927,
+      "eval_last_Accuracy-wrong": 0.5404821469183534,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6680679021503793,
+      "eval_max_AUC": 0.7486281620568513,
+      "eval_max_Accuracy": 0.6368534482758621,
+      "eval_max_Accuracy-right": 0.9711751662971175,
+      "eval_max_Accuracy-wrong": 0.05390038662724585,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.60539630409024,
+      "eval_min_AUC": 0.7853322845511744,
+      "eval_min_Accuracy": 0.7209051724137931,
+      "eval_min_Accuracy-right": 0.7549889135254989,
+      "eval_min_Accuracy-wrong": 0.6614737320900614,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6588892732875934,
+      "eval_prod_AUC": 0.7847266249422049,
+      "eval_prod_Accuracy": 0.706440649867374,
+      "eval_prod_Accuracy-right": 0.6447763140733012,
+      "eval_prod_Accuracy-wrong": 0.8139640664089152,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6510440752440944,
+      "eval_runtime": 252.474,
+      "eval_samples_per_second": 95.566,
+      "eval_steps_per_second": 2.986,
+      "eval_sum_AUC": 0.6452340173243202,
+      "eval_sum_Accuracy": 0.6382211538461539,
+      "eval_sum_Accuracy-right": 0.9940654754141124,
+      "eval_sum_Accuracy-wrong": 0.01773936775073914,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6368181458868507,
+      "step": 3000
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.719355683834616,
+      "learning_rate": 4.669319407871647e-06,
+      "loss": 0.2515,
+      "step": 3001
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.170964090776609,
+      "learning_rate": 4.666435641304301e-06,
+      "loss": 0.2522,
+      "step": 3002
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.923468867083184,
+      "learning_rate": 4.663551986187006e-06,
+      "loss": 0.2493,
+      "step": 3003
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.30112917716831,
+      "learning_rate": 4.660668443483248e-06,
+      "loss": 0.3486,
+      "step": 3004
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.14288841228628,
+      "learning_rate": 4.657785014156468e-06,
+      "loss": 0.2489,
+      "step": 3005
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.713009972335938,
+      "learning_rate": 4.654901699170077e-06,
+      "loss": 0.3506,
+      "step": 3006
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.161879396772502,
+      "learning_rate": 4.652018499487442e-06,
+      "loss": 0.2977,
+      "step": 3007
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.191379054117588,
+      "learning_rate": 4.649135416071896e-06,
+      "loss": 0.2581,
+      "step": 3008
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 20.471445742521716,
+      "learning_rate": 4.646252449886727e-06,
+      "loss": 0.3132,
+      "step": 3009
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.975686497998877,
+      "learning_rate": 4.6433696018951915e-06,
+      "loss": 0.3794,
+      "step": 3010
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.4897021990898,
+      "learning_rate": 4.640486873060501e-06,
+      "loss": 0.3147,
+      "step": 3011
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.363266401118915,
+      "learning_rate": 4.6376042643458254e-06,
+      "loss": 0.2346,
+      "step": 3012
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.753203788882743,
+      "learning_rate": 4.634721776714305e-06,
+      "loss": 0.2937,
+      "step": 3013
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.363526616262392,
+      "learning_rate": 4.631839411129025e-06,
+      "loss": 0.3018,
+      "step": 3014
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.110221054974863,
+      "learning_rate": 4.628957168553044e-06,
+      "loss": 0.2727,
+      "step": 3015
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.632277373728602,
+      "learning_rate": 4.6260750499493665e-06,
+      "loss": 0.4019,
+      "step": 3016
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.087285690858906,
+      "learning_rate": 4.623193056280968e-06,
+      "loss": 0.3335,
+      "step": 3017
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.025810609472378,
+      "learning_rate": 4.6203111885107735e-06,
+      "loss": 0.2375,
+      "step": 3018
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.067682340287535,
+      "learning_rate": 4.617429447601665e-06,
+      "loss": 0.2476,
+      "step": 3019
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.772321631454632,
+      "learning_rate": 4.614547834516492e-06,
+      "loss": 0.2551,
+      "step": 3020
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.09704435421464,
+      "learning_rate": 4.6116663502180495e-06,
+      "loss": 0.3059,
+      "step": 3021
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.204814449759008,
+      "learning_rate": 4.6087849956691e-06,
+      "loss": 0.2561,
+      "step": 3022
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.234181879339191,
+      "learning_rate": 4.605903771832353e-06,
+      "loss": 0.2791,
+      "step": 3023
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.496146964582882,
+      "learning_rate": 4.603022679670482e-06,
+      "loss": 0.2939,
+      "step": 3024
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 20.489599009018622,
+      "learning_rate": 4.6001417201461114e-06,
+      "loss": 0.2979,
+      "step": 3025
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.091464787515235,
+      "learning_rate": 4.597260894221826e-06,
+      "loss": 0.3162,
+      "step": 3026
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.839579410632815,
+      "learning_rate": 4.594380202860162e-06,
+      "loss": 0.3357,
+      "step": 3027
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.013418973388266,
+      "learning_rate": 4.5914996470236094e-06,
+      "loss": 0.3015,
+      "step": 3028
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.220647672670827,
+      "learning_rate": 4.588619227674619e-06,
+      "loss": 0.2327,
+      "step": 3029
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.062502514019542,
+      "learning_rate": 4.58573894577559e-06,
+      "loss": 0.322,
+      "step": 3030
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.599524075966587,
+      "learning_rate": 4.5828588022888815e-06,
+      "loss": 0.2974,
+      "step": 3031
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.933286676593458,
+      "learning_rate": 4.5799787981767975e-06,
+      "loss": 0.2861,
+      "step": 3032
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.79407778155579,
+      "learning_rate": 4.577098934401607e-06,
+      "loss": 0.2773,
+      "step": 3033
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.25668831567693,
+      "learning_rate": 4.57421921192552e-06,
+      "loss": 0.2288,
+      "step": 3034
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.332016300630848,
+      "learning_rate": 4.5713396317107115e-06,
+      "loss": 0.2136,
+      "step": 3035
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.56505719334519,
+      "learning_rate": 4.568460194719299e-06,
+      "loss": 0.1902,
+      "step": 3036
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.736750372255026,
+      "learning_rate": 4.565580901913356e-06,
+      "loss": 0.2759,
+      "step": 3037
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.30006734127021,
+      "learning_rate": 4.562701754254909e-06,
+      "loss": 0.29,
+      "step": 3038
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.077457240517598,
+      "learning_rate": 4.559822752705933e-06,
+      "loss": 0.3496,
+      "step": 3039
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.698125787141345,
+      "learning_rate": 4.556943898228358e-06,
+      "loss": 0.2375,
+      "step": 3040
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.038157371173256,
+      "learning_rate": 4.55406519178406e-06,
+      "loss": 0.2932,
+      "step": 3041
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.500307715120192,
+      "learning_rate": 4.551186634334873e-06,
+      "loss": 0.2507,
+      "step": 3042
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.239595721895437,
+      "learning_rate": 4.54830822684257e-06,
+      "loss": 0.2397,
+      "step": 3043
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 26.518571980335143,
+      "learning_rate": 4.545429970268888e-06,
+      "loss": 0.5195,
+      "step": 3044
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 21.63226226043342,
+      "learning_rate": 4.542551865575499e-06,
+      "loss": 0.314,
+      "step": 3045
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.077102386147187,
+      "learning_rate": 4.539673913724037e-06,
+      "loss": 0.2119,
+      "step": 3046
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.394200369375191,
+      "learning_rate": 4.5367961156760745e-06,
+      "loss": 0.3079,
+      "step": 3047
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 29.417594716851383,
+      "learning_rate": 4.533918472393141e-06,
+      "loss": 0.5259,
+      "step": 3048
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.457438246959063,
+      "learning_rate": 4.531040984836708e-06,
+      "loss": 0.2346,
+      "step": 3049
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.204113553765206,
+      "learning_rate": 4.5281636539682e-06,
+      "loss": 0.3069,
+      "step": 3050
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.049228198201712,
+      "learning_rate": 4.5252864807489836e-06,
+      "loss": 0.2656,
+      "step": 3051
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.606648655895928,
+      "learning_rate": 4.522409466140379e-06,
+      "loss": 0.2593,
+      "step": 3052
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.556356018045541,
+      "learning_rate": 4.5195326111036475e-06,
+      "loss": 0.3882,
+      "step": 3053
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.99828379110033,
+      "learning_rate": 4.5166559166000035e-06,
+      "loss": 0.3203,
+      "step": 3054
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 22.263085046077336,
+      "learning_rate": 4.513779383590599e-06,
+      "loss": 0.3269,
+      "step": 3055
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.933203567646347,
+      "learning_rate": 4.510903013036542e-06,
+      "loss": 0.3042,
+      "step": 3056
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.491586252287485,
+      "learning_rate": 4.508026805898878e-06,
+      "loss": 0.2708,
+      "step": 3057
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.372992871527948,
+      "learning_rate": 4.505150763138604e-06,
+      "loss": 0.2822,
+      "step": 3058
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.457990662995527,
+      "learning_rate": 4.502274885716656e-06,
+      "loss": 0.2791,
+      "step": 3059
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.701114569175404,
+      "learning_rate": 4.499399174593923e-06,
+      "loss": 0.239,
+      "step": 3060
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.167254709611056,
+      "learning_rate": 4.496523630731229e-06,
+      "loss": 0.3047,
+      "step": 3061
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.111502148347535,
+      "learning_rate": 4.493648255089347e-06,
+      "loss": 0.249,
+      "step": 3062
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.384302535844311,
+      "learning_rate": 4.490773048628997e-06,
+      "loss": 0.2898,
+      "step": 3063
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.800690735712097,
+      "learning_rate": 4.487898012310834e-06,
+      "loss": 0.2688,
+      "step": 3064
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.264879768298673,
+      "learning_rate": 4.485023147095466e-06,
+      "loss": 0.5088,
+      "step": 3065
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.264233050955779,
+      "learning_rate": 4.482148453943434e-06,
+      "loss": 0.3213,
+      "step": 3066
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.516757746284293,
+      "learning_rate": 4.479273933815232e-06,
+      "loss": 0.3726,
+      "step": 3067
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.358627634880012,
+      "learning_rate": 4.476399587671285e-06,
+      "loss": 0.4409,
+      "step": 3068
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.196044087714956,
+      "learning_rate": 4.47352541647197e-06,
+      "loss": 0.2632,
+      "step": 3069
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.996167525812037,
+      "learning_rate": 4.470651421177599e-06,
+      "loss": 0.2305,
+      "step": 3070
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.349338044623135,
+      "learning_rate": 4.467777602748425e-06,
+      "loss": 0.1885,
+      "step": 3071
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.05700812773356,
+      "learning_rate": 4.4649039621446495e-06,
+      "loss": 0.3647,
+      "step": 3072
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.042404398136627,
+      "learning_rate": 4.462030500326403e-06,
+      "loss": 0.2471,
+      "step": 3073
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.157726115560735,
+      "learning_rate": 4.459157218253769e-06,
+      "loss": 0.3259,
+      "step": 3074
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.746593922157988,
+      "learning_rate": 4.456284116886758e-06,
+      "loss": 0.2834,
+      "step": 3075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.324680514897194,
+      "learning_rate": 4.453411197185334e-06,
+      "loss": 0.2542,
+      "step": 3076
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.104263762334465,
+      "learning_rate": 4.450538460109384e-06,
+      "loss": 0.314,
+      "step": 3077
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.666298882817523,
+      "learning_rate": 4.447665906618751e-06,
+      "loss": 0.2043,
+      "step": 3078
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.855428365451152,
+      "learning_rate": 4.444793537673204e-06,
+      "loss": 0.2446,
+      "step": 3079
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.617290032096342,
+      "learning_rate": 4.441921354232455e-06,
+      "loss": 0.3394,
+      "step": 3080
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.767718908396866,
+      "learning_rate": 4.439049357256156e-06,
+      "loss": 0.2822,
+      "step": 3081
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.04967813853941,
+      "learning_rate": 4.436177547703891e-06,
+      "loss": 0.2839,
+      "step": 3082
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.124779669352307,
+      "learning_rate": 4.433305926535189e-06,
+      "loss": 0.3984,
+      "step": 3083
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.063769370375915,
+      "learning_rate": 4.430434494709509e-06,
+      "loss": 0.2358,
+      "step": 3084
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.918550212828764,
+      "learning_rate": 4.427563253186253e-06,
+      "loss": 0.1887,
+      "step": 3085
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.958914847278019,
+      "learning_rate": 4.424692202924754e-06,
+      "loss": 0.1517,
+      "step": 3086
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.898577580549401,
+      "learning_rate": 4.421821344884281e-06,
+      "loss": 0.2023,
+      "step": 3087
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.715953044475873,
+      "learning_rate": 4.418950680024046e-06,
+      "loss": 0.2214,
+      "step": 3088
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.016811447032085,
+      "learning_rate": 4.416080209303187e-06,
+      "loss": 0.2229,
+      "step": 3089
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.39587616677386,
+      "learning_rate": 4.413209933680786e-06,
+      "loss": 0.2915,
+      "step": 3090
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.239323594842487,
+      "learning_rate": 4.410339854115849e-06,
+      "loss": 0.3066,
+      "step": 3091
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.047328873557003,
+      "learning_rate": 4.407469971567331e-06,
+      "loss": 0.3145,
+      "step": 3092
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.89869317299782,
+      "learning_rate": 4.4046002869941055e-06,
+      "loss": 0.2461,
+      "step": 3093
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.751841757753965,
+      "learning_rate": 4.401730801354994e-06,
+      "loss": 0.3765,
+      "step": 3094
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.149602274951638,
+      "learning_rate": 4.39886151560874e-06,
+      "loss": 0.2488,
+      "step": 3095
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.324113331094987,
+      "learning_rate": 4.395992430714028e-06,
+      "loss": 0.2659,
+      "step": 3096
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.07732938634878,
+      "learning_rate": 4.393123547629472e-06,
+      "loss": 0.2417,
+      "step": 3097
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.730825721640787,
+      "learning_rate": 4.390254867313619e-06,
+      "loss": 0.2175,
+      "step": 3098
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.498964036214774,
+      "learning_rate": 4.387386390724947e-06,
+      "loss": 0.2483,
+      "step": 3099
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.819080003077305,
+      "learning_rate": 4.38451811882187e-06,
+      "loss": 0.2152,
+      "step": 3100
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.016314180753497,
+      "learning_rate": 4.3816500525627284e-06,
+      "loss": 0.2668,
+      "step": 3101
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.737209247832773,
+      "learning_rate": 4.3787821929057985e-06,
+      "loss": 0.2485,
+      "step": 3102
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.21172272431426,
+      "learning_rate": 4.3759145408092855e-06,
+      "loss": 0.2683,
+      "step": 3103
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.720873434696868,
+      "learning_rate": 4.373047097231324e-06,
+      "loss": 0.281,
+      "step": 3104
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.82444071744457,
+      "learning_rate": 4.370179863129979e-06,
+      "loss": 0.3105,
+      "step": 3105
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.67477919882557,
+      "learning_rate": 4.367312839463251e-06,
+      "loss": 0.3049,
+      "step": 3106
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.979444272208877,
+      "learning_rate": 4.3644460271890614e-06,
+      "loss": 0.2878,
+      "step": 3107
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.93302818612876,
+      "learning_rate": 4.361579427265268e-06,
+      "loss": 0.2407,
+      "step": 3108
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.835216618647086,
+      "learning_rate": 4.358713040649654e-06,
+      "loss": 0.2849,
+      "step": 3109
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.05126265344838,
+      "learning_rate": 4.3558468682999336e-06,
+      "loss": 0.2156,
+      "step": 3110
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.93526267989827,
+      "learning_rate": 4.352980911173747e-06,
+      "loss": 0.3201,
+      "step": 3111
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.299964275506714,
+      "learning_rate": 4.350115170228664e-06,
+      "loss": 0.2556,
+      "step": 3112
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.497945099673855,
+      "learning_rate": 4.3472496464221845e-06,
+      "loss": 0.3054,
+      "step": 3113
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.59759493393717,
+      "learning_rate": 4.344384340711728e-06,
+      "loss": 0.1531,
+      "step": 3114
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.031330972624689,
+      "learning_rate": 4.341519254054651e-06,
+      "loss": 0.1885,
+      "step": 3115
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.543383964941674,
+      "learning_rate": 4.338654387408229e-06,
+      "loss": 0.229,
+      "step": 3116
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.767553783865127,
+      "learning_rate": 4.335789741729671e-06,
+      "loss": 0.2268,
+      "step": 3117
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.926314350733149,
+      "learning_rate": 4.332925317976104e-06,
+      "loss": 0.2434,
+      "step": 3118
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.66200156018986,
+      "learning_rate": 4.330061117104589e-06,
+      "loss": 0.2202,
+      "step": 3119
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.672090923776402,
+      "learning_rate": 4.327197140072108e-06,
+      "loss": 0.2769,
+      "step": 3120
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.31880580033002,
+      "learning_rate": 4.324333387835565e-06,
+      "loss": 0.2791,
+      "step": 3121
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.507999522819803,
+      "learning_rate": 4.321469861351799e-06,
+      "loss": 0.2307,
+      "step": 3122
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.389475444013744,
+      "learning_rate": 4.318606561577562e-06,
+      "loss": 0.3716,
+      "step": 3123
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.701663568377832,
+      "learning_rate": 4.31574348946954e-06,
+      "loss": 0.2278,
+      "step": 3124
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.68598335546021,
+      "learning_rate": 4.312880645984334e-06,
+      "loss": 0.228,
+      "step": 3125
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.815726426115686,
+      "learning_rate": 4.310018032078479e-06,
+      "loss": 0.2471,
+      "step": 3126
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.09950822431688,
+      "learning_rate": 4.307155648708421e-06,
+      "loss": 0.3633,
+      "step": 3127
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.393836294451905,
+      "learning_rate": 4.304293496830542e-06,
+      "loss": 0.4065,
+      "step": 3128
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 14.434362345596266,
+      "learning_rate": 4.301431577401136e-06,
+      "loss": 0.377,
+      "step": 3129
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.846364671226091,
+      "learning_rate": 4.298569891376423e-06,
+      "loss": 0.2998,
+      "step": 3130
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.059306524932216,
+      "learning_rate": 4.2957084397125496e-06,
+      "loss": 0.3047,
+      "step": 3131
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.741521017522524,
+      "learning_rate": 4.292847223365574e-06,
+      "loss": 0.2766,
+      "step": 3132
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.152323226909704,
+      "learning_rate": 4.289986243291488e-06,
+      "loss": 0.2642,
+      "step": 3133
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.623268724924818,
+      "learning_rate": 4.287125500446193e-06,
+      "loss": 0.248,
+      "step": 3134
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.09056413640535,
+      "learning_rate": 4.284264995785521e-06,
+      "loss": 0.2935,
+      "step": 3135
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.216703895780089,
+      "learning_rate": 4.2814047302652155e-06,
+      "loss": 0.24,
+      "step": 3136
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 22.248993121674086,
+      "learning_rate": 4.278544704840948e-06,
+      "loss": 0.373,
+      "step": 3137
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.845301439748043,
+      "learning_rate": 4.275684920468306e-06,
+      "loss": 0.272,
+      "step": 3138
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.246046989550374,
+      "learning_rate": 4.272825378102791e-06,
+      "loss": 0.2622,
+      "step": 3139
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.234508413168031,
+      "learning_rate": 4.269966078699836e-06,
+      "loss": 0.2419,
+      "step": 3140
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.587106775142328,
+      "learning_rate": 4.267107023214782e-06,
+      "loss": 0.2439,
+      "step": 3141
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.882611698235124,
+      "learning_rate": 4.264248212602896e-06,
+      "loss": 0.2478,
+      "step": 3142
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 24.969669318687767,
+      "learning_rate": 4.261389647819355e-06,
+      "loss": 0.3423,
+      "step": 3143
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.55812135577993,
+      "learning_rate": 4.258531329819264e-06,
+      "loss": 0.3176,
+      "step": 3144
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.263892512075122,
+      "learning_rate": 4.255673259557636e-06,
+      "loss": 0.239,
+      "step": 3145
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.911486938832663,
+      "learning_rate": 4.252815437989408e-06,
+      "loss": 0.3022,
+      "step": 3146
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.005026801764096,
+      "learning_rate": 4.24995786606943e-06,
+      "loss": 0.377,
+      "step": 3147
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.225242396822619,
+      "learning_rate": 4.24710054475247e-06,
+      "loss": 0.2725,
+      "step": 3148
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.15552040543596,
+      "learning_rate": 4.244243474993214e-06,
+      "loss": 0.2903,
+      "step": 3149
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.337215646854586,
+      "learning_rate": 4.241386657746257e-06,
+      "loss": 0.2544,
+      "step": 3150
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.69049114869754,
+      "learning_rate": 4.2385300939661215e-06,
+      "loss": 0.3018,
+      "step": 3151
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 12.674673511882043,
+      "learning_rate": 4.2356737846072326e-06,
+      "loss": 0.2776,
+      "step": 3152
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.409534549944462,
+      "learning_rate": 4.232817730623941e-06,
+      "loss": 0.272,
+      "step": 3153
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.195418655205769,
+      "learning_rate": 4.229961932970505e-06,
+      "loss": 0.1794,
+      "step": 3154
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.817548675323104,
+      "learning_rate": 4.2271063926010995e-06,
+      "loss": 0.2317,
+      "step": 3155
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.93728757969069,
+      "learning_rate": 4.224251110469814e-06,
+      "loss": 0.313,
+      "step": 3156
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.095598064837857,
+      "learning_rate": 4.221396087530652e-06,
+      "loss": 0.251,
+      "step": 3157
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 25.89434606145267,
+      "learning_rate": 4.218541324737529e-06,
+      "loss": 0.3334,
+      "step": 3158
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.844389906885638,
+      "learning_rate": 4.2156868230442756e-06,
+      "loss": 0.2351,
+      "step": 3159
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.609452101643022,
+      "learning_rate": 4.212832583404632e-06,
+      "loss": 0.3489,
+      "step": 3160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.13110023399353,
+      "learning_rate": 4.2099786067722535e-06,
+      "loss": 0.3137,
+      "step": 3161
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.109251771071031,
+      "learning_rate": 4.207124894100707e-06,
+      "loss": 0.2671,
+      "step": 3162
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.889889573613662,
+      "learning_rate": 4.2042714463434715e-06,
+      "loss": 0.311,
+      "step": 3163
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.05359030193196,
+      "learning_rate": 4.201418264453935e-06,
+      "loss": 0.2292,
+      "step": 3164
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 29.070179041968114,
+      "learning_rate": 4.198565349385402e-06,
+      "loss": 0.3848,
+      "step": 3165
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.001684670124861,
+      "learning_rate": 4.195712702091079e-06,
+      "loss": 0.3359,
+      "step": 3166
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.578718859825187,
+      "learning_rate": 4.192860323524094e-06,
+      "loss": 0.28,
+      "step": 3167
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.171497366139107,
+      "learning_rate": 4.190008214637476e-06,
+      "loss": 0.2961,
+      "step": 3168
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.876458793899186,
+      "learning_rate": 4.187156376384171e-06,
+      "loss": 0.2766,
+      "step": 3169
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.134458157859976,
+      "learning_rate": 4.184304809717027e-06,
+      "loss": 0.311,
+      "step": 3170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 25.636667120923505,
+      "learning_rate": 4.18145351558881e-06,
+      "loss": 0.3188,
+      "step": 3171
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.453629348290125,
+      "learning_rate": 4.178602494952187e-06,
+      "loss": 0.3232,
+      "step": 3172
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.292622999034865,
+      "learning_rate": 4.175751748759737e-06,
+      "loss": 0.2307,
+      "step": 3173
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.461810881057556,
+      "learning_rate": 4.1729012779639495e-06,
+      "loss": 0.1851,
+      "step": 3174
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.995403715034513,
+      "learning_rate": 4.170051083517217e-06,
+      "loss": 0.3142,
+      "step": 3175
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 28.442476871673804,
+      "learning_rate": 4.167201166371846e-06,
+      "loss": 0.3682,
+      "step": 3176
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.11822897463826,
+      "learning_rate": 4.164351527480042e-06,
+      "loss": 0.2732,
+      "step": 3177
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.21883279375824,
+      "learning_rate": 4.161502167793928e-06,
+      "loss": 0.3286,
+      "step": 3178
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.22564232638533,
+      "learning_rate": 4.1586530882655226e-06,
+      "loss": 0.3633,
+      "step": 3179
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.812336180461106,
+      "learning_rate": 4.155804289846762e-06,
+      "loss": 0.2236,
+      "step": 3180
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.885754005735777,
+      "learning_rate": 4.152955773489479e-06,
+      "loss": 0.3079,
+      "step": 3181
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.559628066175136,
+      "learning_rate": 4.150107540145413e-06,
+      "loss": 0.3069,
+      "step": 3182
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.868335548365271,
+      "learning_rate": 4.147259590766219e-06,
+      "loss": 0.3408,
+      "step": 3183
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.395868967470347,
+      "learning_rate": 4.144411926303442e-06,
+      "loss": 0.241,
+      "step": 3184
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.321970943416959,
+      "learning_rate": 4.141564547708546e-06,
+      "loss": 0.2585,
+      "step": 3185
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.89739295436778,
+      "learning_rate": 4.138717455932888e-06,
+      "loss": 0.3022,
+      "step": 3186
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.404981314367001,
+      "learning_rate": 4.13587065192774e-06,
+      "loss": 0.2917,
+      "step": 3187
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.078981032390395,
+      "learning_rate": 4.133024136644269e-06,
+      "loss": 0.2913,
+      "step": 3188
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.308577953291362,
+      "learning_rate": 4.130177911033546e-06,
+      "loss": 0.2468,
+      "step": 3189
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.135277981344027,
+      "learning_rate": 4.127331976046553e-06,
+      "loss": 0.2318,
+      "step": 3190
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.987586352845991,
+      "learning_rate": 4.124486332634165e-06,
+      "loss": 0.3101,
+      "step": 3191
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.693395458674322,
+      "learning_rate": 4.121640981747169e-06,
+      "loss": 0.2869,
+      "step": 3192
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.221225064372197,
+      "learning_rate": 4.118795924336245e-06,
+      "loss": 0.2749,
+      "step": 3193
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.464738717879746,
+      "learning_rate": 4.115951161351985e-06,
+      "loss": 0.2207,
+      "step": 3194
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.485375562985087,
+      "learning_rate": 4.113106693744871e-06,
+      "loss": 0.3633,
+      "step": 3195
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.928871090786055,
+      "learning_rate": 4.110262522465298e-06,
+      "loss": 0.3142,
+      "step": 3196
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.569062534229731,
+      "learning_rate": 4.107418648463553e-06,
+      "loss": 0.2415,
+      "step": 3197
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.556704268124058,
+      "learning_rate": 4.104575072689827e-06,
+      "loss": 0.4214,
+      "step": 3198
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.309652598968563,
+      "learning_rate": 4.101731796094215e-06,
+      "loss": 0.2314,
+      "step": 3199
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.249740156900172,
+      "learning_rate": 4.098888819626704e-06,
+      "loss": 0.3022,
+      "step": 3200
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.021724075151338,
+      "learning_rate": 4.096046144237189e-06,
+      "loss": 0.2642,
+      "step": 3201
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.964776567576948,
+      "learning_rate": 4.093203770875458e-06,
+      "loss": 0.2451,
+      "step": 3202
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.948673464764161,
+      "learning_rate": 4.090361700491203e-06,
+      "loss": 0.2285,
+      "step": 3203
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.28404893480785,
+      "learning_rate": 4.087519934034011e-06,
+      "loss": 0.3101,
+      "step": 3204
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.473785693501132,
+      "learning_rate": 4.084678472453371e-06,
+      "loss": 0.2549,
+      "step": 3205
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.12057528554086,
+      "learning_rate": 4.081837316698665e-06,
+      "loss": 0.3501,
+      "step": 3206
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.997967895220985,
+      "learning_rate": 4.078996467719179e-06,
+      "loss": 0.3188,
+      "step": 3207
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.614239673097449,
+      "learning_rate": 4.076155926464091e-06,
+      "loss": 0.2056,
+      "step": 3208
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.47178593134632,
+      "learning_rate": 4.07331569388248e-06,
+      "loss": 0.2585,
+      "step": 3209
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.19903305701167,
+      "learning_rate": 4.07047577092332e-06,
+      "loss": 0.3623,
+      "step": 3210
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.950945054310028,
+      "learning_rate": 4.067636158535483e-06,
+      "loss": 0.2402,
+      "step": 3211
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.688010166084021,
+      "learning_rate": 4.064796857667734e-06,
+      "loss": 0.3113,
+      "step": 3212
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.84894387925838,
+      "learning_rate": 4.0619578692687405e-06,
+      "loss": 0.3286,
+      "step": 3213
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 22.02913425595009,
+      "learning_rate": 4.059119194287056e-06,
+      "loss": 0.3047,
+      "step": 3214
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.256721550325487,
+      "learning_rate": 4.056280833671139e-06,
+      "loss": 0.2534,
+      "step": 3215
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.369809286741337,
+      "learning_rate": 4.053442788369334e-06,
+      "loss": 0.2544,
+      "step": 3216
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.298685517077775,
+      "learning_rate": 4.05060505932989e-06,
+      "loss": 0.3677,
+      "step": 3217
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.898769900347371,
+      "learning_rate": 4.04776764750094e-06,
+      "loss": 0.2266,
+      "step": 3218
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.947460381359758,
+      "learning_rate": 4.04493055383052e-06,
+      "loss": 0.301,
+      "step": 3219
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.74632555217383,
+      "learning_rate": 4.042093779266553e-06,
+      "loss": 0.2559,
+      "step": 3220
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.599283467869368,
+      "learning_rate": 4.0392573247568614e-06,
+      "loss": 0.27,
+      "step": 3221
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.943002421727876,
+      "learning_rate": 4.036421191249155e-06,
+      "loss": 0.2822,
+      "step": 3222
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 29.579454388849882,
+      "learning_rate": 4.033585379691036e-06,
+      "loss": 0.4604,
+      "step": 3223
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.899580216055918,
+      "learning_rate": 4.030749891030008e-06,
+      "loss": 0.28,
+      "step": 3224
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.093984979965922,
+      "learning_rate": 4.0279147262134534e-06,
+      "loss": 0.281,
+      "step": 3225
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.406878672576628,
+      "learning_rate": 4.025079886188661e-06,
+      "loss": 0.3867,
+      "step": 3226
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.447300544742673,
+      "learning_rate": 4.022245371902796e-06,
+      "loss": 0.3687,
+      "step": 3227
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.587222605286408,
+      "learning_rate": 4.01941118430293e-06,
+      "loss": 0.3228,
+      "step": 3228
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.088881097873506,
+      "learning_rate": 4.0165773243360105e-06,
+      "loss": 0.1971,
+      "step": 3229
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.273170092461974,
+      "learning_rate": 4.0137437929488885e-06,
+      "loss": 0.2725,
+      "step": 3230
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 13.35424018521676,
+      "learning_rate": 4.010910591088296e-06,
+      "loss": 0.2815,
+      "step": 3231
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.00753162673241,
+      "learning_rate": 4.008077719700859e-06,
+      "loss": 0.3716,
+      "step": 3232
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.541628475729205,
+      "learning_rate": 4.005245179733095e-06,
+      "loss": 0.2659,
+      "step": 3233
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.232635981574267,
+      "learning_rate": 4.002412972131403e-06,
+      "loss": 0.208,
+      "step": 3234
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.224460677900014,
+      "learning_rate": 3.999581097842082e-06,
+      "loss": 0.1646,
+      "step": 3235
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.8840908089983,
+      "learning_rate": 3.99674955781131e-06,
+      "loss": 0.2446,
+      "step": 3236
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.304529312568652,
+      "learning_rate": 3.99391835298516e-06,
+      "loss": 0.2317,
+      "step": 3237
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.502911845939426,
+      "learning_rate": 3.991087484309586e-06,
+      "loss": 0.231,
+      "step": 3238
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.99359587043134,
+      "learning_rate": 3.988256952730439e-06,
+      "loss": 0.3862,
+      "step": 3239
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.178986410442256,
+      "learning_rate": 3.985426759193449e-06,
+      "loss": 0.2451,
+      "step": 3240
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.303906956030472,
+      "learning_rate": 3.982596904644236e-06,
+      "loss": 0.2224,
+      "step": 3241
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.532535397233683,
+      "learning_rate": 3.979767390028309e-06,
+      "loss": 0.2461,
+      "step": 3242
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.192913755233537,
+      "learning_rate": 3.976938216291059e-06,
+      "loss": 0.2808,
+      "step": 3243
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 11.248098877642407,
+      "learning_rate": 3.974109384377768e-06,
+      "loss": 0.2585,
+      "step": 3244
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.235266597149444,
+      "learning_rate": 3.971280895233599e-06,
+      "loss": 0.2322,
+      "step": 3245
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.562031261403643,
+      "learning_rate": 3.968452749803605e-06,
+      "loss": 0.3599,
+      "step": 3246
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.3566097178524,
+      "learning_rate": 3.965624949032723e-06,
+      "loss": 0.3271,
+      "step": 3247
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.20409787900473,
+      "learning_rate": 3.962797493865767e-06,
+      "loss": 0.2771,
+      "step": 3248
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.070021402214895,
+      "learning_rate": 3.959970385247451e-06,
+      "loss": 0.3074,
+      "step": 3249
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 25.348359088033686,
+      "learning_rate": 3.957143624122359e-06,
+      "loss": 0.3887,
+      "step": 3250
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.65763701778697,
+      "learning_rate": 3.954317211434966e-06,
+      "loss": 0.2871,
+      "step": 3251
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.067469739234614,
+      "learning_rate": 3.951491148129628e-06,
+      "loss": 0.3311,
+      "step": 3252
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.506402978506092,
+      "learning_rate": 3.948665435150589e-06,
+      "loss": 0.2527,
+      "step": 3253
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.367653284704408,
+      "learning_rate": 3.945840073441967e-06,
+      "loss": 0.2432,
+      "step": 3254
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.529703746228293,
+      "learning_rate": 3.943015063947773e-06,
+      "loss": 0.2793,
+      "step": 3255
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 28.26985063062385,
+      "learning_rate": 3.940190407611891e-06,
+      "loss": 0.333,
+      "step": 3256
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.659362639715853,
+      "learning_rate": 3.937366105378093e-06,
+      "loss": 0.1985,
+      "step": 3257
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.796592659095309,
+      "learning_rate": 3.93454215819003e-06,
+      "loss": 0.2639,
+      "step": 3258
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.950373980503407,
+      "learning_rate": 3.931718566991236e-06,
+      "loss": 0.4453,
+      "step": 3259
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.59777182080348,
+      "learning_rate": 3.9288953327251265e-06,
+      "loss": 0.3032,
+      "step": 3260
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.217410108136798,
+      "learning_rate": 3.9260724563349935e-06,
+      "loss": 0.2666,
+      "step": 3261
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.026004296899076,
+      "learning_rate": 3.923249938764016e-06,
+      "loss": 0.312,
+      "step": 3262
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.644089760277204,
+      "learning_rate": 3.920427780955247e-06,
+      "loss": 0.3784,
+      "step": 3263
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.92215295788311,
+      "learning_rate": 3.917605983851622e-06,
+      "loss": 0.2981,
+      "step": 3264
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.252312169191091,
+      "learning_rate": 3.914784548395959e-06,
+      "loss": 0.2703,
+      "step": 3265
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.325498594913125,
+      "learning_rate": 3.911963475530948e-06,
+      "loss": 0.3665,
+      "step": 3266
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.00125047262502,
+      "learning_rate": 3.909142766199163e-06,
+      "loss": 0.3264,
+      "step": 3267
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.43569373211537,
+      "learning_rate": 3.906322421343055e-06,
+      "loss": 0.2512,
+      "step": 3268
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.382819352662063,
+      "learning_rate": 3.903502441904956e-06,
+      "loss": 0.2378,
+      "step": 3269
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.338844125815825,
+      "learning_rate": 3.900682828827072e-06,
+      "loss": 0.2664,
+      "step": 3270
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.983592958009579,
+      "learning_rate": 3.897863583051488e-06,
+      "loss": 0.2817,
+      "step": 3271
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.878791410236738,
+      "learning_rate": 3.895044705520167e-06,
+      "loss": 0.2729,
+      "step": 3272
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.09056810345942,
+      "learning_rate": 3.892226197174947e-06,
+      "loss": 0.3022,
+      "step": 3273
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.416882596827744,
+      "learning_rate": 3.889408058957547e-06,
+      "loss": 0.251,
+      "step": 3274
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.418938219681527,
+      "learning_rate": 3.886590291809554e-06,
+      "loss": 0.3281,
+      "step": 3275
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.54374780153606,
+      "learning_rate": 3.883772896672443e-06,
+      "loss": 0.25,
+      "step": 3276
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.475161782689312,
+      "learning_rate": 3.8809558744875534e-06,
+      "loss": 0.3037,
+      "step": 3277
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.217374801851173,
+      "learning_rate": 3.878139226196107e-06,
+      "loss": 0.2986,
+      "step": 3278
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.47972978144969,
+      "learning_rate": 3.875322952739196e-06,
+      "loss": 0.3706,
+      "step": 3279
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.281463354122508,
+      "learning_rate": 3.872507055057793e-06,
+      "loss": 0.2288,
+      "step": 3280
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.610178959550831,
+      "learning_rate": 3.8696915340927395e-06,
+      "loss": 0.2668,
+      "step": 3281
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.2691544838789,
+      "learning_rate": 3.866876390784752e-06,
+      "loss": 0.231,
+      "step": 3282
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 36.402448271805106,
+      "learning_rate": 3.8640616260744266e-06,
+      "loss": 0.3735,
+      "step": 3283
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.171805634334017,
+      "learning_rate": 3.861247240902223e-06,
+      "loss": 0.2512,
+      "step": 3284
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 51.24345152155576,
+      "learning_rate": 3.858433236208485e-06,
+      "loss": 0.2886,
+      "step": 3285
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.317004480140044,
+      "learning_rate": 3.85561961293342e-06,
+      "loss": 0.2485,
+      "step": 3286
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.386888418538078,
+      "learning_rate": 3.852806372017115e-06,
+      "loss": 0.2256,
+      "step": 3287
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.24971292042215,
+      "learning_rate": 3.849993514399521e-06,
+      "loss": 0.2556,
+      "step": 3288
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.008978089204213,
+      "learning_rate": 3.847181041020472e-06,
+      "loss": 0.2876,
+      "step": 3289
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.822270914959086,
+      "learning_rate": 3.844368952819666e-06,
+      "loss": 0.2314,
+      "step": 3290
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.175452659496289,
+      "learning_rate": 3.84155725073667e-06,
+      "loss": 0.3372,
+      "step": 3291
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.094646202827287,
+      "learning_rate": 3.838745935710931e-06,
+      "loss": 0.2441,
+      "step": 3292
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.954423265082738,
+      "learning_rate": 3.835935008681757e-06,
+      "loss": 0.3174,
+      "step": 3293
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.370092110407349,
+      "learning_rate": 3.833124470588336e-06,
+      "loss": 0.3279,
+      "step": 3294
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.808483717392717,
+      "learning_rate": 3.830314322369717e-06,
+      "loss": 0.2334,
+      "step": 3295
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.576910490466176,
+      "learning_rate": 3.827504564964825e-06,
+      "loss": 0.2522,
+      "step": 3296
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.48583322025054,
+      "learning_rate": 3.82469519931245e-06,
+      "loss": 0.323,
+      "step": 3297
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.518800925347575,
+      "learning_rate": 3.8218862263512565e-06,
+      "loss": 0.2456,
+      "step": 3298
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.559272871616013,
+      "learning_rate": 3.819077647019772e-06,
+      "loss": 0.3108,
+      "step": 3299
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.96699943610106,
+      "learning_rate": 3.816269462256394e-06,
+      "loss": 0.3784,
+      "step": 3300
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.209507947737528,
+      "learning_rate": 3.813461672999394e-06,
+      "loss": 0.2932,
+      "step": 3301
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.387762465843936,
+      "learning_rate": 3.8106542801869007e-06,
+      "loss": 0.2808,
+      "step": 3302
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.201757791001164,
+      "learning_rate": 3.8078472847569215e-06,
+      "loss": 0.3765,
+      "step": 3303
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.253111845630949,
+      "learning_rate": 3.805040687647321e-06,
+      "loss": 0.3374,
+      "step": 3304
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 19.82147693696702,
+      "learning_rate": 3.8022344897958402e-06,
+      "loss": 0.332,
+      "step": 3305
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.343136940610577,
+      "learning_rate": 3.799428692140077e-06,
+      "loss": 0.2681,
+      "step": 3306
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.444088175662092,
+      "learning_rate": 3.7966232956175053e-06,
+      "loss": 0.2773,
+      "step": 3307
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.004902990239433,
+      "learning_rate": 3.793818301165457e-06,
+      "loss": 0.3726,
+      "step": 3308
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.09082633279314,
+      "learning_rate": 3.7910137097211345e-06,
+      "loss": 0.3049,
+      "step": 3309
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.587555974439523,
+      "learning_rate": 3.788209522221604e-06,
+      "loss": 0.2961,
+      "step": 3310
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.99106911580416,
+      "learning_rate": 3.7854057396037934e-06,
+      "loss": 0.2881,
+      "step": 3311
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.139850321676342,
+      "learning_rate": 3.7826023628045037e-06,
+      "loss": 0.2412,
+      "step": 3312
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.388827648592175,
+      "learning_rate": 3.779799392760391e-06,
+      "loss": 0.3181,
+      "step": 3313
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.958626380275174,
+      "learning_rate": 3.7769968304079833e-06,
+      "loss": 0.2668,
+      "step": 3314
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.3999162309862,
+      "learning_rate": 3.7741946766836657e-06,
+      "loss": 0.283,
+      "step": 3315
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.78486008620017,
+      "learning_rate": 3.771392932523691e-06,
+      "loss": 0.2568,
+      "step": 3316
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.359568318591949,
+      "learning_rate": 3.768591598864174e-06,
+      "loss": 0.2939,
+      "step": 3317
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 31.670155201809273,
+      "learning_rate": 3.765790676641092e-06,
+      "loss": 0.3071,
+      "step": 3318
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.970236514238282,
+      "learning_rate": 3.762990166790286e-06,
+      "loss": 0.2551,
+      "step": 3319
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.047255891257112,
+      "learning_rate": 3.760190070247458e-06,
+      "loss": 0.3247,
+      "step": 3320
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.64156184651028,
+      "learning_rate": 3.7573903879481714e-06,
+      "loss": 0.2834,
+      "step": 3321
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.664506406435355,
+      "learning_rate": 3.754591120827854e-06,
+      "loss": 0.2263,
+      "step": 3322
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.453693666344321,
+      "learning_rate": 3.7517922698217914e-06,
+      "loss": 0.2427,
+      "step": 3323
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.250753271812135,
+      "learning_rate": 3.7489938358651334e-06,
+      "loss": 0.2156,
+      "step": 3324
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.8413429198328,
+      "learning_rate": 3.746195819892885e-06,
+      "loss": 0.3213,
+      "step": 3325
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.338062073228624,
+      "learning_rate": 3.7433982228399205e-06,
+      "loss": 0.1901,
+      "step": 3326
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.241923335507073,
+      "learning_rate": 3.7406010456409648e-06,
+      "loss": 0.3037,
+      "step": 3327
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.63418710168968,
+      "learning_rate": 3.73780428923061e-06,
+      "loss": 0.2549,
+      "step": 3328
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.930254467732022,
+      "learning_rate": 3.7350079545433014e-06,
+      "loss": 0.2166,
+      "step": 3329
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.752872103252343,
+      "learning_rate": 3.7322120425133497e-06,
+      "loss": 0.1606,
+      "step": 3330
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.592309135473164,
+      "learning_rate": 3.729416554074917e-06,
+      "loss": 0.2258,
+      "step": 3331
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.525689690264553,
+      "learning_rate": 3.726621490162033e-06,
+      "loss": 0.1725,
+      "step": 3332
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.915876448357972,
+      "learning_rate": 3.7238268517085773e-06,
+      "loss": 0.353,
+      "step": 3333
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.281531039394093,
+      "learning_rate": 3.7210326396482893e-06,
+      "loss": 0.2805,
+      "step": 3334
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.802973255554388,
+      "learning_rate": 3.718238854914771e-06,
+      "loss": 0.3052,
+      "step": 3335
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.4298425578648,
+      "learning_rate": 3.7154454984414733e-06,
+      "loss": 0.3263,
+      "step": 3336
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.220340865499665,
+      "learning_rate": 3.7126525711617135e-06,
+      "loss": 0.3015,
+      "step": 3337
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.281971521389087,
+      "learning_rate": 3.7098600740086555e-06,
+      "loss": 0.2118,
+      "step": 3338
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.39960104281075,
+      "learning_rate": 3.707068007915329e-06,
+      "loss": 0.2546,
+      "step": 3339
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.371386947285565,
+      "learning_rate": 3.704276373814611e-06,
+      "loss": 0.2737,
+      "step": 3340
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.294716255095457,
+      "learning_rate": 3.7014851726392427e-06,
+      "loss": 0.2411,
+      "step": 3341
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.309078229290204,
+      "learning_rate": 3.6986944053218143e-06,
+      "loss": 0.2798,
+      "step": 3342
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.365685779974543,
+      "learning_rate": 3.69590407279477e-06,
+      "loss": 0.3003,
+      "step": 3343
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.837019788680069,
+      "learning_rate": 3.6931141759904175e-06,
+      "loss": 0.293,
+      "step": 3344
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.925370194635887,
+      "learning_rate": 3.6903247158409077e-06,
+      "loss": 0.2639,
+      "step": 3345
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.855847089664,
+      "learning_rate": 3.687535693278256e-06,
+      "loss": 0.3687,
+      "step": 3346
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.239321422376957,
+      "learning_rate": 3.6847471092343225e-06,
+      "loss": 0.2676,
+      "step": 3347
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.481727189698633,
+      "learning_rate": 3.681958964640828e-06,
+      "loss": 0.2578,
+      "step": 3348
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.534587315413534,
+      "learning_rate": 3.679171260429343e-06,
+      "loss": 0.2925,
+      "step": 3349
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.42058638916999,
+      "learning_rate": 3.676383997531288e-06,
+      "loss": 0.3088,
+      "step": 3350
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.855041948096473,
+      "learning_rate": 3.673597176877944e-06,
+      "loss": 0.2554,
+      "step": 3351
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.898301518405677,
+      "learning_rate": 3.670810799400435e-06,
+      "loss": 0.2297,
+      "step": 3352
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.44162249590785,
+      "learning_rate": 3.668024866029747e-06,
+      "loss": 0.2598,
+      "step": 3353
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.992432735538818,
+      "learning_rate": 3.665239377696706e-06,
+      "loss": 0.2859,
+      "step": 3354
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.802186033426322,
+      "learning_rate": 3.6624543353320006e-06,
+      "loss": 0.254,
+      "step": 3355
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.7282741034783,
+      "learning_rate": 3.659669739866162e-06,
+      "loss": 0.2305,
+      "step": 3356
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.747783422915768,
+      "learning_rate": 3.6568855922295776e-06,
+      "loss": 0.3083,
+      "step": 3357
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.203973860796053,
+      "learning_rate": 3.654101893352482e-06,
+      "loss": 0.2449,
+      "step": 3358
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.381045935118614,
+      "learning_rate": 3.651318644164958e-06,
+      "loss": 0.2786,
+      "step": 3359
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.200099711534078,
+      "learning_rate": 3.6485358455969454e-06,
+      "loss": 0.2385,
+      "step": 3360
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.050190434196447,
+      "learning_rate": 3.645753498578225e-06,
+      "loss": 0.1902,
+      "step": 3361
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.029174110851411,
+      "learning_rate": 3.6429716040384346e-06,
+      "loss": 0.2703,
+      "step": 3362
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.962055596198097,
+      "learning_rate": 3.6401901629070524e-06,
+      "loss": 0.3083,
+      "step": 3363
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.313655103966168,
+      "learning_rate": 3.6374091761134147e-06,
+      "loss": 0.301,
+      "step": 3364
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.61602666368444,
+      "learning_rate": 3.6346286445866953e-06,
+      "loss": 0.1937,
+      "step": 3365
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.310994068495816,
+      "learning_rate": 3.6318485692559263e-06,
+      "loss": 0.2715,
+      "step": 3366
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.739251234096173,
+      "learning_rate": 3.62906895104998e-06,
+      "loss": 0.3269,
+      "step": 3367
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.251420055737725,
+      "learning_rate": 3.6262897908975787e-06,
+      "loss": 0.3164,
+      "step": 3368
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.647741659961826,
+      "learning_rate": 3.6235110897272917e-06,
+      "loss": 0.2031,
+      "step": 3369
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.353959481918734,
+      "learning_rate": 3.620732848467535e-06,
+      "loss": 0.2383,
+      "step": 3370
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.086259437715091,
+      "learning_rate": 3.6179550680465703e-06,
+      "loss": 0.2429,
+      "step": 3371
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.140836475782853,
+      "learning_rate": 3.615177749392506e-06,
+      "loss": 0.25,
+      "step": 3372
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.462306422712425,
+      "learning_rate": 3.6124008934332956e-06,
+      "loss": 0.2981,
+      "step": 3373
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.595663064497298,
+      "learning_rate": 3.609624501096739e-06,
+      "loss": 0.2786,
+      "step": 3374
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.6865300704338,
+      "learning_rate": 3.606848573310479e-06,
+      "loss": 0.2834,
+      "step": 3375
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.459999547360587,
+      "learning_rate": 3.6040731110020065e-06,
+      "loss": 0.252,
+      "step": 3376
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.050438349312042,
+      "learning_rate": 3.6012981150986524e-06,
+      "loss": 0.3784,
+      "step": 3377
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.520037630380769,
+      "learning_rate": 3.598523586527599e-06,
+      "loss": 0.207,
+      "step": 3378
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.356693880055909,
+      "learning_rate": 3.595749526215862e-06,
+      "loss": 0.2615,
+      "step": 3379
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.210506673880975,
+      "learning_rate": 3.5929759350903117e-06,
+      "loss": 0.243,
+      "step": 3380
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.890635842665827,
+      "learning_rate": 3.5902028140776524e-06,
+      "loss": 0.3169,
+      "step": 3381
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.508691081951469,
+      "learning_rate": 3.5874301641044386e-06,
+      "loss": 0.2642,
+      "step": 3382
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.55890525274352,
+      "learning_rate": 3.5846579860970632e-06,
+      "loss": 0.2678,
+      "step": 3383
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.731659464137019,
+      "learning_rate": 3.58188628098176e-06,
+      "loss": 0.3152,
+      "step": 3384
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.571055770883325,
+      "learning_rate": 3.579115049684612e-06,
+      "loss": 0.2434,
+      "step": 3385
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.26780059174239,
+      "learning_rate": 3.576344293131533e-06,
+      "loss": 0.2771,
+      "step": 3386
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.9806207910959,
+      "learning_rate": 3.5735740122482896e-06,
+      "loss": 0.2788,
+      "step": 3387
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.608528508998647,
+      "learning_rate": 3.570804207960481e-06,
+      "loss": 0.3105,
+      "step": 3388
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.976436128368848,
+      "learning_rate": 3.5680348811935527e-06,
+      "loss": 0.2446,
+      "step": 3389
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.725885115817704,
+      "learning_rate": 3.565266032872785e-06,
+      "loss": 0.2861,
+      "step": 3390
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.014407863564019,
+      "learning_rate": 3.5624976639233056e-06,
+      "loss": 0.2568,
+      "step": 3391
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.169543696415937,
+      "learning_rate": 3.559729775270076e-06,
+      "loss": 0.2629,
+      "step": 3392
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.368641787713843,
+      "learning_rate": 3.5569623678378972e-06,
+      "loss": 0.3442,
+      "step": 3393
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.144314587647392,
+      "learning_rate": 3.554195442551416e-06,
+      "loss": 0.2119,
+      "step": 3394
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.282894560665648,
+      "learning_rate": 3.551429000335108e-06,
+      "loss": 0.3357,
+      "step": 3395
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.94605752362725,
+      "learning_rate": 3.5486630421132983e-06,
+      "loss": 0.2141,
+      "step": 3396
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.588223525170012,
+      "learning_rate": 3.5458975688101403e-06,
+      "loss": 0.2935,
+      "step": 3397
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.249548737786958,
+      "learning_rate": 3.5431325813496352e-06,
+      "loss": 0.2644,
+      "step": 3398
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.03872538160447,
+      "learning_rate": 3.540368080655612e-06,
+      "loss": 0.3416,
+      "step": 3399
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.263906287807048,
+      "learning_rate": 3.5376040676517443e-06,
+      "loss": 0.3013,
+      "step": 3400
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.90507660616009,
+      "learning_rate": 3.5348405432615407e-06,
+      "loss": 0.2251,
+      "step": 3401
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.652509111149236,
+      "learning_rate": 3.5320775084083425e-06,
+      "loss": 0.1938,
+      "step": 3402
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.948684175375187,
+      "learning_rate": 3.529314964015336e-06,
+      "loss": 0.2017,
+      "step": 3403
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.81378214693452,
+      "learning_rate": 3.526552911005533e-06,
+      "loss": 0.2417,
+      "step": 3404
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.127745326816424,
+      "learning_rate": 3.523791350301793e-06,
+      "loss": 0.2727,
+      "step": 3405
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.96529610492523,
+      "learning_rate": 3.5210302828267984e-06,
+      "loss": 0.2617,
+      "step": 3406
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.61248036670611,
+      "learning_rate": 3.5182697095030795e-06,
+      "loss": 0.3103,
+      "step": 3407
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.296153280775098,
+      "learning_rate": 3.5155096312529913e-06,
+      "loss": 0.3633,
+      "step": 3408
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.295250617043212,
+      "learning_rate": 3.5127500489987252e-06,
+      "loss": 0.2856,
+      "step": 3409
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.24988189484059,
+      "learning_rate": 3.5099909636623148e-06,
+      "loss": 0.3184,
+      "step": 3410
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.376778291796514,
+      "learning_rate": 3.5072323761656163e-06,
+      "loss": 0.3359,
+      "step": 3411
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.142529765067701,
+      "learning_rate": 3.5044742874303297e-06,
+      "loss": 0.3108,
+      "step": 3412
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.349635726849158,
+      "learning_rate": 3.501716698377979e-06,
+      "loss": 0.2485,
+      "step": 3413
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.768764085615702,
+      "learning_rate": 3.4989596099299306e-06,
+      "loss": 0.2454,
+      "step": 3414
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.424964292649417,
+      "learning_rate": 3.496203023007374e-06,
+      "loss": 0.3284,
+      "step": 3415
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.246943837578039,
+      "learning_rate": 3.4934469385313418e-06,
+      "loss": 0.3223,
+      "step": 3416
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.516172225486876,
+      "learning_rate": 3.490691357422689e-06,
+      "loss": 0.25,
+      "step": 3417
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.56752172697326,
+      "learning_rate": 3.487936280602108e-06,
+      "loss": 0.2329,
+      "step": 3418
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.265980094835774,
+      "learning_rate": 3.4851817089901203e-06,
+      "loss": 0.2244,
+      "step": 3419
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.204349860262266,
+      "learning_rate": 3.4824276435070804e-06,
+      "loss": 0.2239,
+      "step": 3420
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.692570947593712,
+      "learning_rate": 3.4796740850731716e-06,
+      "loss": 0.2324,
+      "step": 3421
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.387294099444116,
+      "learning_rate": 3.47692103460841e-06,
+      "loss": 0.2925,
+      "step": 3422
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.421806378326073,
+      "learning_rate": 3.474168493032641e-06,
+      "loss": 0.3445,
+      "step": 3423
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 16.231543949083306,
+      "learning_rate": 3.4714164612655387e-06,
+      "loss": 0.3259,
+      "step": 3424
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.791636744953154,
+      "learning_rate": 3.468664940226609e-06,
+      "loss": 0.3198,
+      "step": 3425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.274282803160158,
+      "learning_rate": 3.4659139308351885e-06,
+      "loss": 0.2417,
+      "step": 3426
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.905001077666203,
+      "learning_rate": 3.4631634340104357e-06,
+      "loss": 0.2465,
+      "step": 3427
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.736134980260223,
+      "learning_rate": 3.460413450671346e-06,
+      "loss": 0.2791,
+      "step": 3428
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.233692284060737,
+      "learning_rate": 3.457663981736739e-06,
+      "loss": 0.4175,
+      "step": 3429
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.855372305832327,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.2551,
+      "step": 3430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.398405654424193,
+      "learning_rate": 3.4521665907553957e-06,
+      "loss": 0.1948,
+      "step": 3431
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.49001839122159,
+      "learning_rate": 3.4494186705454402e-06,
+      "loss": 0.2893,
+      "step": 3432
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 13.738087130527903,
+      "learning_rate": 3.446671268413528e-06,
+      "loss": 0.2937,
+      "step": 3433
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.80364763645785,
+      "learning_rate": 3.443924385277617e-06,
+      "loss": 0.2493,
+      "step": 3434
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.451947488762436,
+      "learning_rate": 3.4411780220554937e-06,
+      "loss": 0.2285,
+      "step": 3435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.469520240369008,
+      "learning_rate": 3.4384321796647645e-06,
+      "loss": 0.3096,
+      "step": 3436
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.22828595409656,
+      "learning_rate": 3.4356868590228727e-06,
+      "loss": 0.2534,
+      "step": 3437
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 13.115450507117941,
+      "learning_rate": 3.4329420610470745e-06,
+      "loss": 0.2698,
+      "step": 3438
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.542372576236675,
+      "learning_rate": 3.4301977866544634e-06,
+      "loss": 0.2993,
+      "step": 3439
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.09705606272018,
+      "learning_rate": 3.427454036761948e-06,
+      "loss": 0.2233,
+      "step": 3440
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.02634813937635,
+      "learning_rate": 3.4247108122862703e-06,
+      "loss": 0.2429,
+      "step": 3441
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.935522500990654,
+      "learning_rate": 3.4219681141439907e-06,
+      "loss": 0.2246,
+      "step": 3442
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.667359751130242,
+      "learning_rate": 3.4192259432514934e-06,
+      "loss": 0.2803,
+      "step": 3443
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.039200499360646,
+      "learning_rate": 3.4164843005249928e-06,
+      "loss": 0.2092,
+      "step": 3444
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.756691213045567,
+      "learning_rate": 3.413743186880519e-06,
+      "loss": 0.2317,
+      "step": 3445
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.776664463230171,
+      "learning_rate": 3.4110026032339317e-06,
+      "loss": 0.2922,
+      "step": 3446
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.417900806390653,
+      "learning_rate": 3.408262550500908e-06,
+      "loss": 0.3977,
+      "step": 3447
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.178425128085712,
+      "learning_rate": 3.4055230295969556e-06,
+      "loss": 0.2422,
+      "step": 3448
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.742971833222871,
+      "learning_rate": 3.4027840414373924e-06,
+      "loss": 0.344,
+      "step": 3449
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.435224415388793,
+      "learning_rate": 3.4000455869373716e-06,
+      "loss": 0.2715,
+      "step": 3450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.634275587173237,
+      "learning_rate": 3.397307667011859e-06,
+      "loss": 0.3154,
+      "step": 3451
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.921317196391337,
+      "learning_rate": 3.394570282575642e-06,
+      "loss": 0.2876,
+      "step": 3452
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.62653176049099,
+      "learning_rate": 3.3918334345433367e-06,
+      "loss": 0.3252,
+      "step": 3453
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.047215418020556,
+      "learning_rate": 3.3890971238293703e-06,
+      "loss": 0.3218,
+      "step": 3454
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.006570126629176,
+      "learning_rate": 3.386361351347999e-06,
+      "loss": 0.2898,
+      "step": 3455
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.7530458486607,
+      "learning_rate": 3.3836261180132914e-06,
+      "loss": 0.2742,
+      "step": 3456
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.313128539846193,
+      "learning_rate": 3.3808914247391437e-06,
+      "loss": 0.2656,
+      "step": 3457
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.2334413075271,
+      "learning_rate": 3.3781572724392642e-06,
+      "loss": 0.2427,
+      "step": 3458
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.268830421432604,
+      "learning_rate": 3.3754236620271876e-06,
+      "loss": 0.2834,
+      "step": 3459
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.645365717660876,
+      "learning_rate": 3.3726905944162615e-06,
+      "loss": 0.2603,
+      "step": 3460
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.475867556089085,
+      "learning_rate": 3.3699580705196527e-06,
+      "loss": 0.271,
+      "step": 3461
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.891164807765543,
+      "learning_rate": 3.367226091250353e-06,
+      "loss": 0.2837,
+      "step": 3462
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.147315000935372,
+      "learning_rate": 3.3644946575211634e-06,
+      "loss": 0.2432,
+      "step": 3463
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.722101869142618,
+      "learning_rate": 3.36176377024471e-06,
+      "loss": 0.29,
+      "step": 3464
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.016997359020882,
+      "learning_rate": 3.3590334303334293e-06,
+      "loss": 0.3162,
+      "step": 3465
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.714073886046096,
+      "learning_rate": 3.356303638699583e-06,
+      "loss": 0.302,
+      "step": 3466
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.31122867211997,
+      "learning_rate": 3.35357439625524e-06,
+      "loss": 0.2225,
+      "step": 3467
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.443164765186587,
+      "learning_rate": 3.3508457039122965e-06,
+      "loss": 0.3494,
+      "step": 3468
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.264537881834457,
+      "learning_rate": 3.348117562582457e-06,
+      "loss": 0.3677,
+      "step": 3469
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.05702734122788,
+      "learning_rate": 3.345389973177241e-06,
+      "loss": 0.2539,
+      "step": 3470
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.800434645157752,
+      "learning_rate": 3.342662936607992e-06,
+      "loss": 0.261,
+      "step": 3471
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.088140859929942,
+      "learning_rate": 3.3399364537858594e-06,
+      "loss": 0.2424,
+      "step": 3472
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.233296365055466,
+      "learning_rate": 3.3372105256218153e-06,
+      "loss": 0.3066,
+      "step": 3473
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.936594581867434,
+      "learning_rate": 3.334485153026639e-06,
+      "loss": 0.2,
+      "step": 3474
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.685217490083236,
+      "learning_rate": 3.3317603369109332e-06,
+      "loss": 0.2756,
+      "step": 3475
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.901083967072752,
+      "learning_rate": 3.3290360781851055e-06,
+      "loss": 0.2666,
+      "step": 3476
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 9.973876232975487,
+      "learning_rate": 3.326312377759383e-06,
+      "loss": 0.2457,
+      "step": 3477
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.582000889219513,
+      "learning_rate": 3.3235892365438038e-06,
+      "loss": 0.2554,
+      "step": 3478
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.7223189804276,
+      "learning_rate": 3.3208666554482216e-06,
+      "loss": 0.2821,
+      "step": 3479
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.830264112686557,
+      "learning_rate": 3.3181446353822997e-06,
+      "loss": 0.2622,
+      "step": 3480
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.7891982910683,
+      "learning_rate": 3.315423177255516e-06,
+      "loss": 0.3813,
+      "step": 3481
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.65036520970705,
+      "learning_rate": 3.312702281977161e-06,
+      "loss": 0.2422,
+      "step": 3482
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.934632301351446,
+      "learning_rate": 3.3099819504563356e-06,
+      "loss": 0.2981,
+      "step": 3483
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.371036590772636,
+      "learning_rate": 3.3072621836019535e-06,
+      "loss": 0.2908,
+      "step": 3484
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.64869466776249,
+      "learning_rate": 3.3045429823227405e-06,
+      "loss": 0.201,
+      "step": 3485
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.730727432132026,
+      "learning_rate": 3.3018243475272282e-06,
+      "loss": 0.2419,
+      "step": 3486
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.465970571860945,
+      "learning_rate": 3.2991062801237683e-06,
+      "loss": 0.2417,
+      "step": 3487
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.95581933840102,
+      "learning_rate": 3.296388781020513e-06,
+      "loss": 0.2732,
+      "step": 3488
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.72636391101251,
+      "learning_rate": 3.293671851125434e-06,
+      "loss": 0.3291,
+      "step": 3489
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.831655869700418,
+      "learning_rate": 3.2909554913463034e-06,
+      "loss": 0.2332,
+      "step": 3490
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.681447470044946,
+      "learning_rate": 3.2882397025907114e-06,
+      "loss": 0.3584,
+      "step": 3491
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.474346090221069,
+      "learning_rate": 3.2855244857660497e-06,
+      "loss": 0.2732,
+      "step": 3492
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.735075386972751,
+      "learning_rate": 3.2828098417795267e-06,
+      "loss": 0.2156,
+      "step": 3493
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.010717824928776,
+      "learning_rate": 3.2800957715381537e-06,
+      "loss": 0.3191,
+      "step": 3494
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.239704900786217,
+      "learning_rate": 3.2773822759487497e-06,
+      "loss": 0.2515,
+      "step": 3495
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.594755050704686,
+      "learning_rate": 3.2746693559179483e-06,
+      "loss": 0.2563,
+      "step": 3496
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.933295020461518,
+      "learning_rate": 3.2719570123521816e-06,
+      "loss": 0.3687,
+      "step": 3497
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.270240218891797,
+      "learning_rate": 3.2692452461576997e-06,
+      "loss": 0.2876,
+      "step": 3498
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.502834749242435,
+      "learning_rate": 3.266534058240548e-06,
+      "loss": 0.2979,
+      "step": 3499
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.296849614190464,
+      "learning_rate": 3.2638234495065903e-06,
+      "loss": 0.2004,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5,
+      "eval_avg_AUC": 0.7877963736079466,
+      "eval_avg_Accuracy": 0.6891992705570292,
+      "eval_avg_Accuracy-right": 0.9025694535020217,
+      "eval_avg_Accuracy-wrong": 0.3171480554923812,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6606969481112277,
+      "eval_last_AUC": 0.8078098190488144,
+      "eval_last_Accuracy": 0.7378149867374005,
+      "eval_last_Accuracy-right": 0.852028172688144,
+      "eval_last_Accuracy-wrong": 0.5386627245849442,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6786984038893772,
+      "eval_max_AUC": 0.7524562987575951,
+      "eval_max_Accuracy": 0.6421999336870027,
+      "eval_max_Accuracy-right": 0.981805138907004,
+      "eval_max_Accuracy-wrong": 0.05003411416875142,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6149865113803138,
+      "eval_min_AUC": 0.7926767012641479,
+      "eval_min_Accuracy": 0.7277022546419099,
+      "eval_min_Accuracy-right": 0.7812051649928264,
+      "eval_min_Accuracy-wrong": 0.6344098248806004,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6649279504885879,
+      "eval_prod_AUC": 0.794738394115366,
+      "eval_prod_Accuracy": 0.7094247347480106,
+      "eval_prod_Accuracy-right": 0.6642754662840746,
+      "eval_prod_Accuracy-wrong": 0.788151012053673,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.660230571971425,
+      "eval_runtime": 246.6392,
+      "eval_samples_per_second": 97.827,
+      "eval_steps_per_second": 3.057,
+      "eval_sum_AUC": 0.6402790236735809,
+      "eval_sum_Accuracy": 0.6413710212201591,
+      "eval_sum_Accuracy-right": 0.996869701317334,
+      "eval_sum_Accuracy-wrong": 0.0214919263133955,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6389305505739054,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.068769827644681,
+      "learning_rate": 3.261113420861487e-06,
+      "loss": 0.3,
+      "step": 3501
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.556338925829841,
+      "learning_rate": 3.258403973210713e-06,
+      "loss": 0.2725,
+      "step": 3502
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.173921559638918,
+      "learning_rate": 3.2556951074595435e-06,
+      "loss": 0.2256,
+      "step": 3503
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.308116193516451,
+      "learning_rate": 3.2529868245130577e-06,
+      "loss": 0.3523,
+      "step": 3504
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.483362795187777,
+      "learning_rate": 3.250279125276148e-06,
+      "loss": 0.1975,
+      "step": 3505
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.818400765853749,
+      "learning_rate": 3.2475720106535036e-06,
+      "loss": 0.2515,
+      "step": 3506
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.579502198721391,
+      "learning_rate": 3.244865481549625e-06,
+      "loss": 0.2383,
+      "step": 3507
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.595647574214624,
+      "learning_rate": 3.24215953886881e-06,
+      "loss": 0.3413,
+      "step": 3508
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.566942539380886,
+      "learning_rate": 3.2394541835151692e-06,
+      "loss": 0.2266,
+      "step": 3509
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.679457066390924,
+      "learning_rate": 3.2367494163926095e-06,
+      "loss": 0.3003,
+      "step": 3510
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.640679927317136,
+      "learning_rate": 3.234045238404841e-06,
+      "loss": 0.3547,
+      "step": 3511
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.18961731562165,
+      "learning_rate": 3.2313416504553852e-06,
+      "loss": 0.3152,
+      "step": 3512
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.22642226732791,
+      "learning_rate": 3.2286386534475568e-06,
+      "loss": 0.2302,
+      "step": 3513
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.438253560959739,
+      "learning_rate": 3.2259362482844803e-06,
+      "loss": 0.2563,
+      "step": 3514
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.27856156365522,
+      "learning_rate": 3.2232344358690765e-06,
+      "loss": 0.2539,
+      "step": 3515
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.93630832767274,
+      "learning_rate": 3.220533217104075e-06,
+      "loss": 0.301,
+      "step": 3516
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.027334258786563,
+      "learning_rate": 3.217832592891999e-06,
+      "loss": 0.2263,
+      "step": 3517
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.754441023455815,
+      "learning_rate": 3.2151325641351817e-06,
+      "loss": 0.2654,
+      "step": 3518
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.74013272575728,
+      "learning_rate": 3.2124331317357506e-06,
+      "loss": 0.2236,
+      "step": 3519
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.383946788535154,
+      "learning_rate": 3.2097342965956334e-06,
+      "loss": 0.2791,
+      "step": 3520
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.32091809219386,
+      "learning_rate": 3.2070360596165667e-06,
+      "loss": 0.2312,
+      "step": 3521
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.730483344555825,
+      "learning_rate": 3.204338421700076e-06,
+      "loss": 0.3027,
+      "step": 3522
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.68177293630166,
+      "learning_rate": 3.201641383747498e-06,
+      "loss": 0.3059,
+      "step": 3523
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.746578325895099,
+      "learning_rate": 3.1989449466599574e-06,
+      "loss": 0.3159,
+      "step": 3524
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 14.046874777971698,
+      "learning_rate": 3.1962491113383896e-06,
+      "loss": 0.3032,
+      "step": 3525
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.701317007452005,
+      "learning_rate": 3.1935538786835183e-06,
+      "loss": 0.2761,
+      "step": 3526
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.573694027775247,
+      "learning_rate": 3.1908592495958747e-06,
+      "loss": 0.2598,
+      "step": 3527
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.454753496917233,
+      "learning_rate": 3.1881652249757823e-06,
+      "loss": 0.2227,
+      "step": 3528
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.312287431795568,
+      "learning_rate": 3.185471805723365e-06,
+      "loss": 0.2664,
+      "step": 3529
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.584289964239222,
+      "learning_rate": 3.1827789927385444e-06,
+      "loss": 0.2786,
+      "step": 3530
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.770661082389356,
+      "learning_rate": 3.18008678692104e-06,
+      "loss": 0.2793,
+      "step": 3531
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.506383377691748,
+      "learning_rate": 3.1773951891703668e-06,
+      "loss": 0.2068,
+      "step": 3532
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.113353081336689,
+      "learning_rate": 3.1747042003858386e-06,
+      "loss": 0.2349,
+      "step": 3533
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.615187115741376,
+      "learning_rate": 3.1720138214665643e-06,
+      "loss": 0.2405,
+      "step": 3534
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.977315544926444,
+      "learning_rate": 3.1693240533114496e-06,
+      "loss": 0.4512,
+      "step": 3535
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 14.063179683264558,
+      "learning_rate": 3.1666348968191955e-06,
+      "loss": 0.2495,
+      "step": 3536
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.892504016191015,
+      "learning_rate": 3.1639463528883007e-06,
+      "loss": 0.262,
+      "step": 3537
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.55789315186977,
+      "learning_rate": 3.161258422417055e-06,
+      "loss": 0.2847,
+      "step": 3538
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.550961610557891,
+      "learning_rate": 3.1585711063035496e-06,
+      "loss": 0.2666,
+      "step": 3539
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.899767453488833,
+      "learning_rate": 3.155884405445663e-06,
+      "loss": 0.2334,
+      "step": 3540
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.23350483466534,
+      "learning_rate": 3.153198320741074e-06,
+      "loss": 0.2043,
+      "step": 3541
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.410011852557782,
+      "learning_rate": 3.150512853087253e-06,
+      "loss": 0.325,
+      "step": 3542
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.11365917752667,
+      "learning_rate": 3.1478280033814657e-06,
+      "loss": 0.2322,
+      "step": 3543
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.164175984205736,
+      "learning_rate": 3.14514377252077e-06,
+      "loss": 0.23,
+      "step": 3544
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.426890476211831,
+      "learning_rate": 3.142460161402014e-06,
+      "loss": 0.2512,
+      "step": 3545
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.904224322008433,
+      "learning_rate": 3.139777170921847e-06,
+      "loss": 0.3203,
+      "step": 3546
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.782016167639494,
+      "learning_rate": 3.137094801976701e-06,
+      "loss": 0.2834,
+      "step": 3547
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.154378436965436,
+      "learning_rate": 3.1344130554628104e-06,
+      "loss": 0.2375,
+      "step": 3548
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.41557779364289,
+      "learning_rate": 3.131731932276193e-06,
+      "loss": 0.2896,
+      "step": 3549
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.215143879601154,
+      "learning_rate": 3.129051433312664e-06,
+      "loss": 0.3809,
+      "step": 3550
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.634755228372835,
+      "learning_rate": 3.1263715594678257e-06,
+      "loss": 0.3027,
+      "step": 3551
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.49266302111991,
+      "learning_rate": 3.1236923116370764e-06,
+      "loss": 0.2471,
+      "step": 3552
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.359540843720524,
+      "learning_rate": 3.121013690715601e-06,
+      "loss": 0.1829,
+      "step": 3553
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.828670124176265,
+      "learning_rate": 3.118335697598376e-06,
+      "loss": 0.2185,
+      "step": 3554
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.78806333330499,
+      "learning_rate": 3.1156583331801703e-06,
+      "loss": 0.2986,
+      "step": 3555
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.261812795713473,
+      "learning_rate": 3.1129815983555387e-06,
+      "loss": 0.2212,
+      "step": 3556
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 27.50178705776172,
+      "learning_rate": 3.1103054940188316e-06,
+      "loss": 0.3633,
+      "step": 3557
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.276709743000696,
+      "learning_rate": 3.1076300210641814e-06,
+      "loss": 0.2769,
+      "step": 3558
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.091825282135263,
+      "learning_rate": 3.1049551803855173e-06,
+      "loss": 0.261,
+      "step": 3559
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.497532460353776,
+      "learning_rate": 3.1022809728765486e-06,
+      "loss": 0.2812,
+      "step": 3560
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.352342983807455,
+      "learning_rate": 3.0996073994307825e-06,
+      "loss": 0.2544,
+      "step": 3561
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.47067050690066,
+      "learning_rate": 3.0969344609415076e-06,
+      "loss": 0.2544,
+      "step": 3562
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.14407357274232,
+      "learning_rate": 3.0942621583017994e-06,
+      "loss": 0.2639,
+      "step": 3563
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.376995841990846,
+      "learning_rate": 3.0915904924045294e-06,
+      "loss": 0.2681,
+      "step": 3564
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.615971289854569,
+      "learning_rate": 3.088919464142346e-06,
+      "loss": 0.2559,
+      "step": 3565
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.374764637160622,
+      "learning_rate": 3.0862490744076928e-06,
+      "loss": 0.3003,
+      "step": 3566
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.754057552940706,
+      "learning_rate": 3.0835793240927937e-06,
+      "loss": 0.3311,
+      "step": 3567
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.835687713457723,
+      "learning_rate": 3.0809102140896652e-06,
+      "loss": 0.2524,
+      "step": 3568
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.100356007256956,
+      "learning_rate": 3.078241745290103e-06,
+      "loss": 0.3794,
+      "step": 3569
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.290591029206796,
+      "learning_rate": 3.075573918585696e-06,
+      "loss": 0.2791,
+      "step": 3570
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.028395777757924,
+      "learning_rate": 3.0729067348678127e-06,
+      "loss": 0.1995,
+      "step": 3571
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.641816702608815,
+      "learning_rate": 3.0702401950276066e-06,
+      "loss": 0.1987,
+      "step": 3572
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.246342226496322,
+      "learning_rate": 3.067574299956022e-06,
+      "loss": 0.2441,
+      "step": 3573
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.844832056141563,
+      "learning_rate": 3.0649090505437804e-06,
+      "loss": 0.2976,
+      "step": 3574
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.680252229465067,
+      "learning_rate": 3.062244447681396e-06,
+      "loss": 0.3022,
+      "step": 3575
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.665615703923262,
+      "learning_rate": 3.0595804922591564e-06,
+      "loss": 0.2463,
+      "step": 3576
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.440754388319283,
+      "learning_rate": 3.0569171851671436e-06,
+      "loss": 0.2668,
+      "step": 3577
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.905524709518772,
+      "learning_rate": 3.054254527295215e-06,
+      "loss": 0.3701,
+      "step": 3578
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.807086250686762,
+      "learning_rate": 3.0515925195330148e-06,
+      "loss": 0.2522,
+      "step": 3579
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 10.476465301590263,
+      "learning_rate": 3.048931162769969e-06,
+      "loss": 0.2666,
+      "step": 3580
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.107929057524471,
+      "learning_rate": 3.0462704578952874e-06,
+      "loss": 0.2861,
+      "step": 3581
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 21.986391423429573,
+      "learning_rate": 3.0436104057979604e-06,
+      "loss": 0.2964,
+      "step": 3582
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.90626976949449,
+      "learning_rate": 3.0409510073667602e-06,
+      "loss": 0.2129,
+      "step": 3583
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.164803117431044,
+      "learning_rate": 3.038292263490242e-06,
+      "loss": 0.3071,
+      "step": 3584
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.092854006583256,
+      "learning_rate": 3.035634175056742e-06,
+      "loss": 0.3521,
+      "step": 3585
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.609423145742273,
+      "learning_rate": 3.0329767429543767e-06,
+      "loss": 0.2844,
+      "step": 3586
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.691645964790691,
+      "learning_rate": 3.030319968071043e-06,
+      "loss": 0.2329,
+      "step": 3587
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.81284676046446,
+      "learning_rate": 3.0276638512944177e-06,
+      "loss": 0.2225,
+      "step": 3588
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.769037847281707,
+      "learning_rate": 3.025008393511961e-06,
+      "loss": 0.3105,
+      "step": 3589
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.295061680310575,
+      "learning_rate": 3.022353595610909e-06,
+      "loss": 0.3152,
+      "step": 3590
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.698927211936732,
+      "learning_rate": 3.01969945847828e-06,
+      "loss": 0.293,
+      "step": 3591
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.609293397061075,
+      "learning_rate": 3.017045983000871e-06,
+      "loss": 0.2698,
+      "step": 3592
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 10.157130344506225,
+      "learning_rate": 3.014393170065256e-06,
+      "loss": 0.312,
+      "step": 3593
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.898830020128793,
+      "learning_rate": 3.0117410205577903e-06,
+      "loss": 0.2737,
+      "step": 3594
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.558372901547542,
+      "learning_rate": 3.0090895353646053e-06,
+      "loss": 0.2512,
+      "step": 3595
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.583363620015973,
+      "learning_rate": 3.006438715371614e-06,
+      "loss": 0.2542,
+      "step": 3596
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.806692519344198,
+      "learning_rate": 3.0037885614645e-06,
+      "loss": 0.2227,
+      "step": 3597
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.779965721273376,
+      "learning_rate": 3.001139074528735e-06,
+      "loss": 0.205,
+      "step": 3598
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.531815564612236,
+      "learning_rate": 2.9984902554495556e-06,
+      "loss": 0.2676,
+      "step": 3599
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.506453906728485,
+      "learning_rate": 2.995842105111987e-06,
+      "loss": 0.2793,
+      "step": 3600
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.15394731501427,
+      "learning_rate": 2.99319462440082e-06,
+      "loss": 0.2083,
+      "step": 3601
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.932997861337693,
+      "learning_rate": 2.990547814200633e-06,
+      "loss": 0.2432,
+      "step": 3602
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.849576244908592,
+      "learning_rate": 2.987901675395771e-06,
+      "loss": 0.2454,
+      "step": 3603
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.12584654833872,
+      "learning_rate": 2.985256208870357e-06,
+      "loss": 0.3494,
+      "step": 3604
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.767528797117537,
+      "learning_rate": 2.982611415508294e-06,
+      "loss": 0.2358,
+      "step": 3605
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.280128754931008,
+      "learning_rate": 2.9799672961932525e-06,
+      "loss": 0.2463,
+      "step": 3606
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.9328456698097,
+      "learning_rate": 2.9773238518086866e-06,
+      "loss": 0.176,
+      "step": 3607
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.21994488542098,
+      "learning_rate": 2.974681083237816e-06,
+      "loss": 0.262,
+      "step": 3608
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.967970336760006,
+      "learning_rate": 2.972038991363643e-06,
+      "loss": 0.2529,
+      "step": 3609
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.528992859617613,
+      "learning_rate": 2.9693975770689344e-06,
+      "loss": 0.2549,
+      "step": 3610
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.244301431577744,
+      "learning_rate": 2.9667568412362415e-06,
+      "loss": 0.2222,
+      "step": 3611
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.112330292266513,
+      "learning_rate": 2.9641167847478797e-06,
+      "loss": 0.2761,
+      "step": 3612
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.774287510401793,
+      "learning_rate": 2.96147740848594e-06,
+      "loss": 0.3308,
+      "step": 3613
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.048586990552954,
+      "learning_rate": 2.9588387133322903e-06,
+      "loss": 0.2192,
+      "step": 3614
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.648117792562253,
+      "learning_rate": 2.9562007001685644e-06,
+      "loss": 0.2556,
+      "step": 3615
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.586600872006583,
+      "learning_rate": 2.9535633698761755e-06,
+      "loss": 0.2891,
+      "step": 3616
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.723776113704208,
+      "learning_rate": 2.9509267233363005e-06,
+      "loss": 0.3936,
+      "step": 3617
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.347490053483744,
+      "learning_rate": 2.948290761429895e-06,
+      "loss": 0.2351,
+      "step": 3618
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.14243876444085,
+      "learning_rate": 2.9456554850376805e-06,
+      "loss": 0.2601,
+      "step": 3619
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.440692510978845,
+      "learning_rate": 2.943020895040155e-06,
+      "loss": 0.25,
+      "step": 3620
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 29.257448998416535,
+      "learning_rate": 2.940386992317582e-06,
+      "loss": 0.4346,
+      "step": 3621
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.111642278946286,
+      "learning_rate": 2.937753777749996e-06,
+      "loss": 0.2034,
+      "step": 3622
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.479860863761866,
+      "learning_rate": 2.9351212522172056e-06,
+      "loss": 0.3098,
+      "step": 3623
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.382932164941954,
+      "learning_rate": 2.9324894165987837e-06,
+      "loss": 0.2429,
+      "step": 3624
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.11158565572809,
+      "learning_rate": 2.9298582717740797e-06,
+      "loss": 0.2952,
+      "step": 3625
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.281584890559799,
+      "learning_rate": 2.9272278186222025e-06,
+      "loss": 0.2167,
+      "step": 3626
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.801083924218155,
+      "learning_rate": 2.9245980580220405e-06,
+      "loss": 0.2754,
+      "step": 3627
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.920589259305089,
+      "learning_rate": 2.921968990852242e-06,
+      "loss": 0.2957,
+      "step": 3628
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.695498611228306,
+      "learning_rate": 2.9193406179912297e-06,
+      "loss": 0.252,
+      "step": 3629
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.833770314366229,
+      "learning_rate": 2.91671294031719e-06,
+      "loss": 0.2573,
+      "step": 3630
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.9105743148522,
+      "learning_rate": 2.91408595870808e-06,
+      "loss": 0.2749,
+      "step": 3631
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.022981742578796,
+      "learning_rate": 2.9114596740416224e-06,
+      "loss": 0.2517,
+      "step": 3632
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.157774956816326,
+      "learning_rate": 2.908834087195308e-06,
+      "loss": 0.3579,
+      "step": 3633
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.441924468999797,
+      "learning_rate": 2.9062091990463935e-06,
+      "loss": 0.3257,
+      "step": 3634
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.911170610783156,
+      "learning_rate": 2.903585010471904e-06,
+      "loss": 0.3979,
+      "step": 3635
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.658127870248308,
+      "learning_rate": 2.9009615223486297e-06,
+      "loss": 0.3418,
+      "step": 3636
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.556064566768066,
+      "learning_rate": 2.898338735553128e-06,
+      "loss": 0.2759,
+      "step": 3637
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.750443370130997,
+      "learning_rate": 2.895716650961714e-06,
+      "loss": 0.3328,
+      "step": 3638
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 13.404219192176061,
+      "learning_rate": 2.8930952694504843e-06,
+      "loss": 0.3159,
+      "step": 3639
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.801417766512998,
+      "learning_rate": 2.8904745918952833e-06,
+      "loss": 0.24,
+      "step": 3640
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.893767277763509,
+      "learning_rate": 2.887854619171735e-06,
+      "loss": 0.2925,
+      "step": 3641
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.238072093150642,
+      "learning_rate": 2.8852353521552135e-06,
+      "loss": 0.2283,
+      "step": 3642
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.888970588309743,
+      "learning_rate": 2.8826167917208727e-06,
+      "loss": 0.2354,
+      "step": 3643
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.876703842177745,
+      "learning_rate": 2.8799989387436137e-06,
+      "loss": 0.2683,
+      "step": 3644
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.14852428141835,
+      "learning_rate": 2.8773817940981186e-06,
+      "loss": 0.2678,
+      "step": 3645
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.672569274403406,
+      "learning_rate": 2.8747653586588183e-06,
+      "loss": 0.3386,
+      "step": 3646
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.871670166521756,
+      "learning_rate": 2.872149633299913e-06,
+      "loss": 0.2263,
+      "step": 3647
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.647933339436156,
+      "learning_rate": 2.8695346188953666e-06,
+      "loss": 0.2524,
+      "step": 3648
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.424868547184808,
+      "learning_rate": 2.866920316318904e-06,
+      "loss": 0.3276,
+      "step": 3649
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.588799181559985,
+      "learning_rate": 2.8643067264440116e-06,
+      "loss": 0.3127,
+      "step": 3650
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.677503962337894,
+      "learning_rate": 2.8616938501439384e-06,
+      "loss": 0.2363,
+      "step": 3651
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.00332101941201,
+      "learning_rate": 2.8590816882916948e-06,
+      "loss": 0.2627,
+      "step": 3652
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.14346918330707,
+      "learning_rate": 2.856470241760054e-06,
+      "loss": 0.3857,
+      "step": 3653
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.979054378804332,
+      "learning_rate": 2.8538595114215472e-06,
+      "loss": 0.2969,
+      "step": 3654
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.350457278452481,
+      "learning_rate": 2.8512494981484706e-06,
+      "loss": 0.2629,
+      "step": 3655
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.518490136692607,
+      "learning_rate": 2.848640202812872e-06,
+      "loss": 0.2688,
+      "step": 3656
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.77830597016325,
+      "learning_rate": 2.846031626286574e-06,
+      "loss": 0.2463,
+      "step": 3657
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.232483324704306,
+      "learning_rate": 2.8434237694411414e-06,
+      "loss": 0.2715,
+      "step": 3658
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.486679322201393,
+      "learning_rate": 2.840816633147917e-06,
+      "loss": 0.335,
+      "step": 3659
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.147692850153896,
+      "learning_rate": 2.8382102182779846e-06,
+      "loss": 0.1785,
+      "step": 3660
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.423807909050868,
+      "learning_rate": 2.8356045257022037e-06,
+      "loss": 0.2021,
+      "step": 3661
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.72309990205576,
+      "learning_rate": 2.832999556291177e-06,
+      "loss": 0.2351,
+      "step": 3662
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.42597066218875,
+      "learning_rate": 2.8303953109152815e-06,
+      "loss": 0.3379,
+      "step": 3663
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.89552475555432,
+      "learning_rate": 2.827791790444638e-06,
+      "loss": 0.2471,
+      "step": 3664
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.378000080961417,
+      "learning_rate": 2.8251889957491317e-06,
+      "loss": 0.4006,
+      "step": 3665
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.967603756260177,
+      "learning_rate": 2.822586927698407e-06,
+      "loss": 0.2324,
+      "step": 3666
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.808368358224756,
+      "learning_rate": 2.819985587161861e-06,
+      "loss": 0.2229,
+      "step": 3667
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.91800691344456,
+      "learning_rate": 2.8173849750086513e-06,
+      "loss": 0.3875,
+      "step": 3668
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.77994963136096,
+      "learning_rate": 2.8147850921076903e-06,
+      "loss": 0.2908,
+      "step": 3669
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.899888131646092,
+      "learning_rate": 2.8121859393276475e-06,
+      "loss": 0.2932,
+      "step": 3670
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.500226797134557,
+      "learning_rate": 2.809587517536947e-06,
+      "loss": 0.2898,
+      "step": 3671
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.24302333969292,
+      "learning_rate": 2.806989827603771e-06,
+      "loss": 0.2646,
+      "step": 3672
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.78450213872708,
+      "learning_rate": 2.8043928703960565e-06,
+      "loss": 0.2385,
+      "step": 3673
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.502438761869197,
+      "learning_rate": 2.8017966467814933e-06,
+      "loss": 0.22,
+      "step": 3674
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.757377161825325,
+      "learning_rate": 2.7992011576275295e-06,
+      "loss": 0.2163,
+      "step": 3675
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.72219922258268,
+      "learning_rate": 2.7966064038013657e-06,
+      "loss": 0.1946,
+      "step": 3676
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.007379297576533,
+      "learning_rate": 2.7940123861699577e-06,
+      "loss": 0.2786,
+      "step": 3677
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.593026440283587,
+      "learning_rate": 2.7914191056000147e-06,
+      "loss": 0.2473,
+      "step": 3678
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.03444982249345,
+      "learning_rate": 2.788826562958e-06,
+      "loss": 0.2756,
+      "step": 3679
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.91137019627999,
+      "learning_rate": 2.7862347591101326e-06,
+      "loss": 0.2871,
+      "step": 3680
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.723301260051828,
+      "learning_rate": 2.7836436949223755e-06,
+      "loss": 0.2795,
+      "step": 3681
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.113635889283595,
+      "learning_rate": 2.78105337126046e-06,
+      "loss": 0.2056,
+      "step": 3682
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.904639312214186,
+      "learning_rate": 2.7784637889898534e-06,
+      "loss": 0.3232,
+      "step": 3683
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.888350346451826,
+      "learning_rate": 2.7758749489757914e-06,
+      "loss": 0.3789,
+      "step": 3684
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.763047444352006,
+      "learning_rate": 2.7732868520832455e-06,
+      "loss": 0.2673,
+      "step": 3685
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.351563872031413,
+      "learning_rate": 2.770699499176954e-06,
+      "loss": 0.2411,
+      "step": 3686
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.930545068368955,
+      "learning_rate": 2.768112891121394e-06,
+      "loss": 0.2139,
+      "step": 3687
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.380131399436635,
+      "learning_rate": 2.7655270287808045e-06,
+      "loss": 0.2854,
+      "step": 3688
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.686797655460788,
+      "learning_rate": 2.762941913019166e-06,
+      "loss": 0.2605,
+      "step": 3689
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.096895929111895,
+      "learning_rate": 2.760357544700215e-06,
+      "loss": 0.3394,
+      "step": 3690
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.428775038502973,
+      "learning_rate": 2.757773924687437e-06,
+      "loss": 0.3103,
+      "step": 3691
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.336479750202862,
+      "learning_rate": 2.755191053844068e-06,
+      "loss": 0.3137,
+      "step": 3692
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.401781691538487,
+      "learning_rate": 2.7526089330330925e-06,
+      "loss": 0.301,
+      "step": 3693
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.653098093782502,
+      "learning_rate": 2.7500275631172455e-06,
+      "loss": 0.3079,
+      "step": 3694
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.841260346341729,
+      "learning_rate": 2.74744694495901e-06,
+      "loss": 0.2393,
+      "step": 3695
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.365598543542117,
+      "learning_rate": 2.74486707942062e-06,
+      "loss": 0.2288,
+      "step": 3696
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.161027149177121,
+      "learning_rate": 2.7422879673640552e-06,
+      "loss": 0.2568,
+      "step": 3697
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.575772535414341,
+      "learning_rate": 2.7397096096510467e-06,
+      "loss": 0.3198,
+      "step": 3698
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.191135053850635,
+      "learning_rate": 2.7371320071430674e-06,
+      "loss": 0.183,
+      "step": 3699
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.824506901967847,
+      "learning_rate": 2.7345551607013475e-06,
+      "loss": 0.2175,
+      "step": 3700
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.666492088642991,
+      "learning_rate": 2.7319790711868545e-06,
+      "loss": 0.2837,
+      "step": 3701
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.142683081101087,
+      "learning_rate": 2.7294037394603135e-06,
+      "loss": 0.3069,
+      "step": 3702
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.323056173012265,
+      "learning_rate": 2.7268291663821825e-06,
+      "loss": 0.2463,
+      "step": 3703
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.448469652485555,
+      "learning_rate": 2.7242553528126842e-06,
+      "loss": 0.261,
+      "step": 3704
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.413334200894413,
+      "learning_rate": 2.72168229961177e-06,
+      "loss": 0.2285,
+      "step": 3705
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.372411343210498,
+      "learning_rate": 2.7191100076391473e-06,
+      "loss": 0.28,
+      "step": 3706
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.386689789949346,
+      "learning_rate": 2.716538477754266e-06,
+      "loss": 0.2786,
+      "step": 3707
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.380636299671224,
+      "learning_rate": 2.713967710816323e-06,
+      "loss": 0.2209,
+      "step": 3708
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.743026725229255,
+      "learning_rate": 2.7113977076842597e-06,
+      "loss": 0.262,
+      "step": 3709
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.806331974095787,
+      "learning_rate": 2.7088284692167604e-06,
+      "loss": 0.2461,
+      "step": 3710
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.668378133256628,
+      "learning_rate": 2.7062599962722563e-06,
+      "loss": 0.2358,
+      "step": 3711
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.267929385459727,
+      "learning_rate": 2.703692289708922e-06,
+      "loss": 0.1868,
+      "step": 3712
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.835870231063108,
+      "learning_rate": 2.701125350384676e-06,
+      "loss": 0.2524,
+      "step": 3713
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.563910610196777,
+      "learning_rate": 2.69855917915718e-06,
+      "loss": 0.2437,
+      "step": 3714
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.04255826697796,
+      "learning_rate": 2.695993776883839e-06,
+      "loss": 0.2261,
+      "step": 3715
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.574915744949621,
+      "learning_rate": 2.693429144421803e-06,
+      "loss": 0.2065,
+      "step": 3716
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.575572137934644,
+      "learning_rate": 2.6908652826279623e-06,
+      "loss": 0.3191,
+      "step": 3717
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.921722163471122,
+      "learning_rate": 2.688302192358952e-06,
+      "loss": 0.2988,
+      "step": 3718
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.103930678311931,
+      "learning_rate": 2.6857398744711472e-06,
+      "loss": 0.2549,
+      "step": 3719
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.661523899831044,
+      "learning_rate": 2.683178329820666e-06,
+      "loss": 0.272,
+      "step": 3720
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.55045890604639,
+      "learning_rate": 2.680617559263368e-06,
+      "loss": 0.4014,
+      "step": 3721
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.731468401982053,
+      "learning_rate": 2.6780575636548544e-06,
+      "loss": 0.2571,
+      "step": 3722
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.002086520928234,
+      "learning_rate": 2.67549834385047e-06,
+      "loss": 0.23,
+      "step": 3723
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.74586382467539,
+      "learning_rate": 2.67293990070529e-06,
+      "loss": 0.2725,
+      "step": 3724
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.370350233340924,
+      "learning_rate": 2.6703822350741483e-06,
+      "loss": 0.2493,
+      "step": 3725
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.138445754547114,
+      "learning_rate": 2.6678253478116e-06,
+      "loss": 0.2695,
+      "step": 3726
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.832773166121346,
+      "learning_rate": 2.665269239771953e-06,
+      "loss": 0.3164,
+      "step": 3727
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.550195192135488,
+      "learning_rate": 2.662713911809248e-06,
+      "loss": 0.2651,
+      "step": 3728
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.218101826282929,
+      "learning_rate": 2.6601593647772696e-06,
+      "loss": 0.2422,
+      "step": 3729
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.894184013419036,
+      "learning_rate": 2.657605599529538e-06,
+      "loss": 0.2026,
+      "step": 3730
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.05854846221324,
+      "learning_rate": 2.6550526169193148e-06,
+      "loss": 0.2878,
+      "step": 3731
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.9197967302707,
+      "learning_rate": 2.6525004177995984e-06,
+      "loss": 0.2617,
+      "step": 3732
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.975852184158727,
+      "learning_rate": 2.6499490030231255e-06,
+      "loss": 0.2622,
+      "step": 3733
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.155629799506274,
+      "learning_rate": 2.6473983734423725e-06,
+      "loss": 0.3186,
+      "step": 3734
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.758339565258634,
+      "learning_rate": 2.644848529909552e-06,
+      "loss": 0.2964,
+      "step": 3735
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.913581401033196,
+      "learning_rate": 2.6422994732766124e-06,
+      "loss": 0.2395,
+      "step": 3736
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.148566465671331,
+      "learning_rate": 2.6397512043952422e-06,
+      "loss": 0.2524,
+      "step": 3737
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.657857821556732,
+      "learning_rate": 2.637203724116865e-06,
+      "loss": 0.3242,
+      "step": 3738
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.957623757375671,
+      "learning_rate": 2.634657033292644e-06,
+      "loss": 0.2217,
+      "step": 3739
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.508785111154575,
+      "learning_rate": 2.6321111327734693e-06,
+      "loss": 0.2539,
+      "step": 3740
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.622247025051292,
+      "learning_rate": 2.6295660234099816e-06,
+      "loss": 0.291,
+      "step": 3741
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.422136750776735,
+      "learning_rate": 2.6270217060525416e-06,
+      "loss": 0.2888,
+      "step": 3742
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.53914884344272,
+      "learning_rate": 2.624478181551261e-06,
+      "loss": 0.2026,
+      "step": 3743
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.498572267274076,
+      "learning_rate": 2.62193545075597e-06,
+      "loss": 0.2798,
+      "step": 3744
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.79889320226917,
+      "learning_rate": 2.6193935145162507e-06,
+      "loss": 0.2163,
+      "step": 3745
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.306685142228023,
+      "learning_rate": 2.6168523736814035e-06,
+      "loss": 0.2278,
+      "step": 3746
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.577440943022701,
+      "learning_rate": 2.6143120291004785e-06,
+      "loss": 0.2603,
+      "step": 3747
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.96685761518886,
+      "learning_rate": 2.611772481622246e-06,
+      "loss": 0.2502,
+      "step": 3748
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.149397489215964,
+      "learning_rate": 2.609233732095218e-06,
+      "loss": 0.2128,
+      "step": 3749
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.844697752164103,
+      "learning_rate": 2.6066957813676375e-06,
+      "loss": 0.2615,
+      "step": 3750
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.15670169536182,
+      "learning_rate": 2.604158630287482e-06,
+      "loss": 0.3196,
+      "step": 3751
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.279178479626223,
+      "learning_rate": 2.60162227970246e-06,
+      "loss": 0.2178,
+      "step": 3752
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.751400189966994,
+      "learning_rate": 2.5990867304600136e-06,
+      "loss": 0.2583,
+      "step": 3753
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 15.652516885894475,
+      "learning_rate": 2.5965519834073172e-06,
+      "loss": 0.3057,
+      "step": 3754
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 19.546997677748575,
+      "learning_rate": 2.5940180393912767e-06,
+      "loss": 0.2573,
+      "step": 3755
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.260684953345004,
+      "learning_rate": 2.5914848992585293e-06,
+      "loss": 0.2771,
+      "step": 3756
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 13.383273870807141,
+      "learning_rate": 2.588952563855448e-06,
+      "loss": 0.2675,
+      "step": 3757
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 25.01046644893153,
+      "learning_rate": 2.5864210340281247e-06,
+      "loss": 0.2979,
+      "step": 3758
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.939204656676262,
+      "learning_rate": 2.5838903106224004e-06,
+      "loss": 0.2478,
+      "step": 3759
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.320199686371675,
+      "learning_rate": 2.5813603944838283e-06,
+      "loss": 0.3015,
+      "step": 3760
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.595923087132736,
+      "learning_rate": 2.578831286457708e-06,
+      "loss": 0.3175,
+      "step": 3761
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.246086611925866,
+      "learning_rate": 2.5763029873890542e-06,
+      "loss": 0.2749,
+      "step": 3762
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.986807659021308,
+      "learning_rate": 2.573775498122626e-06,
+      "loss": 0.2788,
+      "step": 3763
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.310292571942231,
+      "learning_rate": 2.5712488195028972e-06,
+      "loss": 0.3462,
+      "step": 3764
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.136729199970594,
+      "learning_rate": 2.5687229523740852e-06,
+      "loss": 0.282,
+      "step": 3765
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.782342303010028,
+      "learning_rate": 2.566197897580124e-06,
+      "loss": 0.2458,
+      "step": 3766
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.133256291376362,
+      "learning_rate": 2.5636736559646824e-06,
+      "loss": 0.2234,
+      "step": 3767
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.590584644856651,
+      "learning_rate": 2.5611502283711576e-06,
+      "loss": 0.3142,
+      "step": 3768
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.440635231131513,
+      "learning_rate": 2.5586276156426726e-06,
+      "loss": 0.2224,
+      "step": 3769
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.245808625760535,
+      "learning_rate": 2.55610581862208e-06,
+      "loss": 0.4028,
+      "step": 3770
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.723496311621378,
+      "learning_rate": 2.553584838151959e-06,
+      "loss": 0.2771,
+      "step": 3771
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.5810352576066,
+      "learning_rate": 2.5510646750746154e-06,
+      "loss": 0.2427,
+      "step": 3772
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.01714154859388,
+      "learning_rate": 2.548545330232083e-06,
+      "loss": 0.2751,
+      "step": 3773
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.33044917355502,
+      "learning_rate": 2.5460268044661215e-06,
+      "loss": 0.2717,
+      "step": 3774
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.953209080406706,
+      "learning_rate": 2.5435090986182176e-06,
+      "loss": 0.2373,
+      "step": 3775
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.763927527803714,
+      "learning_rate": 2.5409922135295827e-06,
+      "loss": 0.2861,
+      "step": 3776
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.902909129104998,
+      "learning_rate": 2.538476150041156e-06,
+      "loss": 0.2371,
+      "step": 3777
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.972557490659907,
+      "learning_rate": 2.5359609089936006e-06,
+      "loss": 0.3052,
+      "step": 3778
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.45909372522568,
+      "learning_rate": 2.533446491227305e-06,
+      "loss": 0.2371,
+      "step": 3779
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.827756309288517,
+      "learning_rate": 2.5309328975823834e-06,
+      "loss": 0.2866,
+      "step": 3780
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.749836878444942,
+      "learning_rate": 2.5284201288986744e-06,
+      "loss": 0.2866,
+      "step": 3781
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.965182716060708,
+      "learning_rate": 2.5259081860157418e-06,
+      "loss": 0.2061,
+      "step": 3782
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.07848339998907,
+      "learning_rate": 2.5233970697728673e-06,
+      "loss": 0.3752,
+      "step": 3783
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.527044251899435,
+      "learning_rate": 2.520886781009068e-06,
+      "loss": 0.2346,
+      "step": 3784
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.012392512661368,
+      "learning_rate": 2.5183773205630726e-06,
+      "loss": 0.1793,
+      "step": 3785
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 25.177862422351314,
+      "learning_rate": 2.515868689273344e-06,
+      "loss": 0.3994,
+      "step": 3786
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.33196589779338,
+      "learning_rate": 2.513360887978056e-06,
+      "loss": 0.3093,
+      "step": 3787
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.253538907290611,
+      "learning_rate": 2.510853917515119e-06,
+      "loss": 0.2842,
+      "step": 3788
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.249265295320773,
+      "learning_rate": 2.50834777872215e-06,
+      "loss": 0.2053,
+      "step": 3789
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.582172247591428,
+      "learning_rate": 2.505842472436506e-06,
+      "loss": 0.2583,
+      "step": 3790
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.66633784718915,
+      "learning_rate": 2.5033379994952493e-06,
+      "loss": 0.2407,
+      "step": 3791
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.944642252674152,
+      "learning_rate": 2.5008343607351733e-06,
+      "loss": 0.2534,
+      "step": 3792
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.936903818038365,
+      "learning_rate": 2.4983315569927895e-06,
+      "loss": 0.2915,
+      "step": 3793
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.489379129954239,
+      "learning_rate": 2.495829589104333e-06,
+      "loss": 0.3008,
+      "step": 3794
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.125082832585976,
+      "learning_rate": 2.493328457905755e-06,
+      "loss": 0.2649,
+      "step": 3795
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.20842735541152,
+      "learning_rate": 2.490828164232732e-06,
+      "loss": 0.3149,
+      "step": 3796
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.675386676649888,
+      "learning_rate": 2.4883287089206582e-06,
+      "loss": 0.1863,
+      "step": 3797
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.742830149961485,
+      "learning_rate": 2.48583009280465e-06,
+      "loss": 0.3511,
+      "step": 3798
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.230891927189097,
+      "learning_rate": 2.483332316719535e-06,
+      "loss": 0.2849,
+      "step": 3799
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.848726525249605,
+      "learning_rate": 2.4808353814998747e-06,
+      "loss": 0.2275,
+      "step": 3800
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.124014387257533,
+      "learning_rate": 2.4783392879799345e-06,
+      "loss": 0.2949,
+      "step": 3801
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.7237259687238,
+      "learning_rate": 2.4758440369937125e-06,
+      "loss": 0.2478,
+      "step": 3802
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.256139492796564,
+      "learning_rate": 2.4733496293749116e-06,
+      "loss": 0.2549,
+      "step": 3803
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.982938440215118,
+      "learning_rate": 2.4708560659569665e-06,
+      "loss": 0.2588,
+      "step": 3804
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.720889972293365,
+      "learning_rate": 2.4683633475730158e-06,
+      "loss": 0.2373,
+      "step": 3805
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.082888203055472,
+      "learning_rate": 2.465871475055931e-06,
+      "loss": 0.2601,
+      "step": 3806
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.744412266081433,
+      "learning_rate": 2.4633804492382866e-06,
+      "loss": 0.2532,
+      "step": 3807
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.257941981624063,
+      "learning_rate": 2.460890270952383e-06,
+      "loss": 0.2229,
+      "step": 3808
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.784405215041396,
+      "learning_rate": 2.4584009410302357e-06,
+      "loss": 0.2222,
+      "step": 3809
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.388475216063645,
+      "learning_rate": 2.4559124603035744e-06,
+      "loss": 0.2717,
+      "step": 3810
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.219406174751503,
+      "learning_rate": 2.4534248296038488e-06,
+      "loss": 0.2698,
+      "step": 3811
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.6815938869164615,
+      "learning_rate": 2.4509380497622208e-06,
+      "loss": 0.22,
+      "step": 3812
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.677189639592727,
+      "learning_rate": 2.448452121609571e-06,
+      "loss": 0.2183,
+      "step": 3813
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.373482728475846,
+      "learning_rate": 2.445967045976493e-06,
+      "loss": 0.3013,
+      "step": 3814
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.139596133312253,
+      "learning_rate": 2.443482823693298e-06,
+      "loss": 0.2468,
+      "step": 3815
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.189247114208078,
+      "learning_rate": 2.4409994555900125e-06,
+      "loss": 0.2351,
+      "step": 3816
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.972346774800624,
+      "learning_rate": 2.4385169424963696e-06,
+      "loss": 0.3157,
+      "step": 3817
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.086435159510867,
+      "learning_rate": 2.4360352852418305e-06,
+      "loss": 0.2595,
+      "step": 3818
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.164889104160295,
+      "learning_rate": 2.4335544846555564e-06,
+      "loss": 0.2026,
+      "step": 3819
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.857166435097096,
+      "learning_rate": 2.431074541566436e-06,
+      "loss": 0.2142,
+      "step": 3820
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.925741012441494,
+      "learning_rate": 2.4285954568030566e-06,
+      "loss": 0.3203,
+      "step": 3821
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.356894972943223,
+      "learning_rate": 2.426117231193735e-06,
+      "loss": 0.2651,
+      "step": 3822
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.0580568696073,
+      "learning_rate": 2.4236398655664834e-06,
+      "loss": 0.2673,
+      "step": 3823
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.718078840336174,
+      "learning_rate": 2.4211633607490442e-06,
+      "loss": 0.3003,
+      "step": 3824
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.376049611246835,
+      "learning_rate": 2.4186877175688576e-06,
+      "loss": 0.2839,
+      "step": 3825
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.126606193237182,
+      "learning_rate": 2.4162129368530848e-06,
+      "loss": 0.2422,
+      "step": 3826
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.206258398092292,
+      "learning_rate": 2.413739019428595e-06,
+      "loss": 0.2449,
+      "step": 3827
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.679245970025766,
+      "learning_rate": 2.41126596612197e-06,
+      "loss": 0.2478,
+      "step": 3828
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.765818992891482,
+      "learning_rate": 2.408793777759504e-06,
+      "loss": 0.2175,
+      "step": 3829
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.619739347290504,
+      "learning_rate": 2.4063224551672e-06,
+      "loss": 0.1636,
+      "step": 3830
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.46872304315626,
+      "learning_rate": 2.4038519991707725e-06,
+      "loss": 0.217,
+      "step": 3831
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.19458341714486,
+      "learning_rate": 2.4013824105956483e-06,
+      "loss": 0.2549,
+      "step": 3832
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.625624020708948,
+      "learning_rate": 2.3989136902669614e-06,
+      "loss": 0.2244,
+      "step": 3833
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.101152390439431,
+      "learning_rate": 2.396445839009558e-06,
+      "loss": 0.2581,
+      "step": 3834
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.164102827836336,
+      "learning_rate": 2.3939788576479926e-06,
+      "loss": 0.2493,
+      "step": 3835
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.073936277817232,
+      "learning_rate": 2.39151274700653e-06,
+      "loss": 0.2141,
+      "step": 3836
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.663235322211742,
+      "learning_rate": 2.389047507909143e-06,
+      "loss": 0.2766,
+      "step": 3837
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.427826685566103,
+      "learning_rate": 2.3865831411795137e-06,
+      "loss": 0.2869,
+      "step": 3838
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.972113366082914,
+      "learning_rate": 2.3841196476410337e-06,
+      "loss": 0.2744,
+      "step": 3839
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.7061464713193,
+      "learning_rate": 2.3816570281168016e-06,
+      "loss": 0.2939,
+      "step": 3840
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.571672824051713,
+      "learning_rate": 2.379195283429626e-06,
+      "loss": 0.2527,
+      "step": 3841
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.099134977591973,
+      "learning_rate": 2.3767344144020164e-06,
+      "loss": 0.2732,
+      "step": 3842
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.620004046032994,
+      "learning_rate": 2.374274421856202e-06,
+      "loss": 0.2632,
+      "step": 3843
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 8.39076395500453,
+      "learning_rate": 2.371815306614104e-06,
+      "loss": 0.2305,
+      "step": 3844
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.1166915100912,
+      "learning_rate": 2.3693570694973673e-06,
+      "loss": 0.2808,
+      "step": 3845
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.41468021588428,
+      "learning_rate": 2.366899711327326e-06,
+      "loss": 0.1987,
+      "step": 3846
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.880151286249157,
+      "learning_rate": 2.3644432329250374e-06,
+      "loss": 0.3016,
+      "step": 3847
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.01327928861911,
+      "learning_rate": 2.3619876351112486e-06,
+      "loss": 0.2637,
+      "step": 3848
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 12.736270873846344,
+      "learning_rate": 2.3595329187064282e-06,
+      "loss": 0.25,
+      "step": 3849
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.835553229983626,
+      "learning_rate": 2.3570790845307367e-06,
+      "loss": 0.2292,
+      "step": 3850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.202387368493234,
+      "learning_rate": 2.3546261334040475e-06,
+      "loss": 0.2852,
+      "step": 3851
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.411782936488706,
+      "learning_rate": 2.352174066145938e-06,
+      "loss": 0.3911,
+      "step": 3852
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.67620624008918,
+      "learning_rate": 2.3497228835756887e-06,
+      "loss": 0.2145,
+      "step": 3853
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.462214082378056,
+      "learning_rate": 2.3472725865122854e-06,
+      "loss": 0.2725,
+      "step": 3854
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.87215540966945,
+      "learning_rate": 2.344823175774418e-06,
+      "loss": 0.2715,
+      "step": 3855
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.18779316525026,
+      "learning_rate": 2.3423746521804796e-06,
+      "loss": 0.2561,
+      "step": 3856
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 12.833742282866638,
+      "learning_rate": 2.339927016548568e-06,
+      "loss": 0.3223,
+      "step": 3857
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 12.93936635239193,
+      "learning_rate": 2.3374802696964842e-06,
+      "loss": 0.3257,
+      "step": 3858
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.703243223218935,
+      "learning_rate": 2.3350344124417336e-06,
+      "loss": 0.2383,
+      "step": 3859
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.33796827383027,
+      "learning_rate": 2.3325894456015154e-06,
+      "loss": 0.2477,
+      "step": 3860
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.236913282982604,
+      "learning_rate": 2.3301453699927477e-06,
+      "loss": 0.2988,
+      "step": 3861
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.224723971064577,
+      "learning_rate": 2.3277021864320332e-06,
+      "loss": 0.2644,
+      "step": 3862
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.688501246139568,
+      "learning_rate": 2.325259895735693e-06,
+      "loss": 0.335,
+      "step": 3863
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.090819890854913,
+      "learning_rate": 2.322818498719734e-06,
+      "loss": 0.239,
+      "step": 3864
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.057125126226383,
+      "learning_rate": 2.3203779961998795e-06,
+      "loss": 0.2053,
+      "step": 3865
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.958169951613703,
+      "learning_rate": 2.317938388991541e-06,
+      "loss": 0.1948,
+      "step": 3866
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.03811936060471,
+      "learning_rate": 2.3154996779098405e-06,
+      "loss": 0.3455,
+      "step": 3867
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.677363035725163,
+      "learning_rate": 2.313061863769594e-06,
+      "loss": 0.271,
+      "step": 3868
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.65190708066888,
+      "learning_rate": 2.310624947385322e-06,
+      "loss": 0.2233,
+      "step": 3869
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.957316079617797,
+      "learning_rate": 2.3081889295712434e-06,
+      "loss": 0.2454,
+      "step": 3870
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.538888305745177,
+      "learning_rate": 2.3057538111412765e-06,
+      "loss": 0.2864,
+      "step": 3871
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.502555486177886,
+      "learning_rate": 2.3033195929090404e-06,
+      "loss": 0.2903,
+      "step": 3872
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.599925868669361,
+      "learning_rate": 2.300886275687852e-06,
+      "loss": 0.2695,
+      "step": 3873
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.740186546474655,
+      "learning_rate": 2.298453860290728e-06,
+      "loss": 0.2144,
+      "step": 3874
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 11.863302566675813,
+      "learning_rate": 2.296022347530384e-06,
+      "loss": 0.229,
+      "step": 3875
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 11.423946125684537,
+      "learning_rate": 2.293591738219233e-06,
+      "loss": 0.2622,
+      "step": 3876
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.70179065557097,
+      "learning_rate": 2.2911620331693867e-06,
+      "loss": 0.2466,
+      "step": 3877
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.731816493491,
+      "learning_rate": 2.2887332331926555e-06,
+      "loss": 0.2634,
+      "step": 3878
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.239045588894083,
+      "learning_rate": 2.2863053391005462e-06,
+      "loss": 0.2148,
+      "step": 3879
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.005638078267589,
+      "learning_rate": 2.2838783517042628e-06,
+      "loss": 0.2544,
+      "step": 3880
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 13.100939806282982,
+      "learning_rate": 2.281452271814708e-06,
+      "loss": 0.3584,
+      "step": 3881
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.093234947012963,
+      "learning_rate": 2.2790271002424794e-06,
+      "loss": 0.293,
+      "step": 3882
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.795211245967684,
+      "learning_rate": 2.276602837797872e-06,
+      "loss": 0.2092,
+      "step": 3883
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 13.552937714867474,
+      "learning_rate": 2.274179485290879e-06,
+      "loss": 0.3291,
+      "step": 3884
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.991506238164844,
+      "learning_rate": 2.271757043531184e-06,
+      "loss": 0.2747,
+      "step": 3885
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.652415067342714,
+      "learning_rate": 2.2693355133281706e-06,
+      "loss": 0.2778,
+      "step": 3886
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.559659679329272,
+      "learning_rate": 2.266914895490918e-06,
+      "loss": 0.249,
+      "step": 3887
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.410463329166966,
+      "learning_rate": 2.2644951908282e-06,
+      "loss": 0.2316,
+      "step": 3888
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.89091889726885,
+      "learning_rate": 2.262076400148484e-06,
+      "loss": 0.4292,
+      "step": 3889
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.884730752793354,
+      "learning_rate": 2.2596585242599333e-06,
+      "loss": 0.3232,
+      "step": 3890
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.28933616549401,
+      "learning_rate": 2.257241563970405e-06,
+      "loss": 0.2344,
+      "step": 3891
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.385456564823834,
+      "learning_rate": 2.254825520087451e-06,
+      "loss": 0.2449,
+      "step": 3892
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.305077098961187,
+      "learning_rate": 2.2524103934183154e-06,
+      "loss": 0.3003,
+      "step": 3893
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.773934697174711,
+      "learning_rate": 2.249996184769938e-06,
+      "loss": 0.2681,
+      "step": 3894
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.217611301648915,
+      "learning_rate": 2.2475828949489504e-06,
+      "loss": 0.353,
+      "step": 3895
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.849058612686495,
+      "learning_rate": 2.2451705247616774e-06,
+      "loss": 0.2488,
+      "step": 3896
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.548086863537335,
+      "learning_rate": 2.2427590750141364e-06,
+      "loss": 0.3018,
+      "step": 3897
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.09548543829281,
+      "learning_rate": 2.240348546512039e-06,
+      "loss": 0.2832,
+      "step": 3898
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.482779249189054,
+      "learning_rate": 2.237938940060786e-06,
+      "loss": 0.2422,
+      "step": 3899
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.515407607343274,
+      "learning_rate": 2.235530256465474e-06,
+      "loss": 0.2533,
+      "step": 3900
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.136175867369973,
+      "learning_rate": 2.233122496530884e-06,
+      "loss": 0.2546,
+      "step": 3901
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.726217970795506,
+      "learning_rate": 2.2307156610615e-06,
+      "loss": 0.2715,
+      "step": 3902
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.768133881503765,
+      "learning_rate": 2.2283097508614837e-06,
+      "loss": 0.2378,
+      "step": 3903
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.681822646089104,
+      "learning_rate": 2.225904766734702e-06,
+      "loss": 0.2346,
+      "step": 3904
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.275082284621488,
+      "learning_rate": 2.2235007094846963e-06,
+      "loss": 0.2439,
+      "step": 3905
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.454333835161174,
+      "learning_rate": 2.2210975799147143e-06,
+      "loss": 0.2776,
+      "step": 3906
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.102927596290863,
+      "learning_rate": 2.21869537882768e-06,
+      "loss": 0.2295,
+      "step": 3907
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.406469054773485,
+      "learning_rate": 2.21629410702622e-06,
+      "loss": 0.2034,
+      "step": 3908
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.923690320426484,
+      "learning_rate": 2.2138937653126393e-06,
+      "loss": 0.2527,
+      "step": 3909
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.679522304935842,
+      "learning_rate": 2.2114943544889366e-06,
+      "loss": 0.2437,
+      "step": 3910
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.2130763130063285,
+      "learning_rate": 2.2090958753568013e-06,
+      "loss": 0.2241,
+      "step": 3911
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.126425833535844,
+      "learning_rate": 2.206698328717609e-06,
+      "loss": 0.321,
+      "step": 3912
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.088147801041128,
+      "learning_rate": 2.2043017153724253e-06,
+      "loss": 0.228,
+      "step": 3913
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.08748121795016,
+      "learning_rate": 2.2019060361220036e-06,
+      "loss": 0.249,
+      "step": 3914
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 24.755148260939105,
+      "learning_rate": 2.199511291766783e-06,
+      "loss": 0.2686,
+      "step": 3915
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.238612933720066,
+      "learning_rate": 2.1971174831068944e-06,
+      "loss": 0.2566,
+      "step": 3916
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.062718282153504,
+      "learning_rate": 2.1947246109421514e-06,
+      "loss": 0.322,
+      "step": 3917
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 13.06127849954014,
+      "learning_rate": 2.192332676072061e-06,
+      "loss": 0.3306,
+      "step": 3918
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.283758870531178,
+      "learning_rate": 2.189941679295807e-06,
+      "loss": 0.2327,
+      "step": 3919
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 14.062313929604555,
+      "learning_rate": 2.1875516214122723e-06,
+      "loss": 0.2507,
+      "step": 3920
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.090141149972927,
+      "learning_rate": 2.185162503220013e-06,
+      "loss": 0.2285,
+      "step": 3921
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.824756292791596,
+      "learning_rate": 2.182774325517285e-06,
+      "loss": 0.2534,
+      "step": 3922
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.320492406815534,
+      "learning_rate": 2.180387089102016e-06,
+      "loss": 0.2246,
+      "step": 3923
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.596778425586805,
+      "learning_rate": 2.1780007947718336e-06,
+      "loss": 0.1755,
+      "step": 3924
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.455925151353494,
+      "learning_rate": 2.175615443324035e-06,
+      "loss": 0.3096,
+      "step": 3925
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 13.420610683613315,
+      "learning_rate": 2.173231035555618e-06,
+      "loss": 0.2738,
+      "step": 3926
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.592246153137603,
+      "learning_rate": 2.170847572263252e-06,
+      "loss": 0.2576,
+      "step": 3927
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.101309193662187,
+      "learning_rate": 2.1684650542432985e-06,
+      "loss": 0.2688,
+      "step": 3928
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.79780471508127,
+      "learning_rate": 2.166083482291801e-06,
+      "loss": 0.2893,
+      "step": 3929
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.65657764794894,
+      "learning_rate": 2.1637028572044867e-06,
+      "loss": 0.2734,
+      "step": 3930
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.391962577247151,
+      "learning_rate": 2.1613231797767668e-06,
+      "loss": 0.2554,
+      "step": 3931
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.977331007596224,
+      "learning_rate": 2.158944450803736e-06,
+      "loss": 0.2844,
+      "step": 3932
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.01757401617662,
+      "learning_rate": 2.1565666710801714e-06,
+      "loss": 0.271,
+      "step": 3933
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.615886862288278,
+      "learning_rate": 2.1541898414005343e-06,
+      "loss": 0.2778,
+      "step": 3934
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.59283742513515,
+      "learning_rate": 2.1518139625589663e-06,
+      "loss": 0.2664,
+      "step": 3935
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.502287160025258,
+      "learning_rate": 2.1494390353492935e-06,
+      "loss": 0.3418,
+      "step": 3936
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.93143738046724,
+      "learning_rate": 2.1470650605650235e-06,
+      "loss": 0.2114,
+      "step": 3937
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.164621540582115,
+      "learning_rate": 2.144692038999345e-06,
+      "loss": 0.2739,
+      "step": 3938
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 12.845701204985016,
+      "learning_rate": 2.142319971445129e-06,
+      "loss": 0.3232,
+      "step": 3939
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.959776190394157,
+      "learning_rate": 2.139948858694926e-06,
+      "loss": 0.2891,
+      "step": 3940
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.79971040797762,
+      "learning_rate": 2.137578701540971e-06,
+      "loss": 0.2053,
+      "step": 3941
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.638963180765387,
+      "learning_rate": 2.1352095007751754e-06,
+      "loss": 0.3005,
+      "step": 3942
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 43.00522394014673,
+      "learning_rate": 2.132841257189137e-06,
+      "loss": 0.4937,
+      "step": 3943
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.985232967151337,
+      "learning_rate": 2.1304739715741235e-06,
+      "loss": 0.3257,
+      "step": 3944
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.283080793744645,
+      "learning_rate": 2.128107644721096e-06,
+      "loss": 0.1897,
+      "step": 3945
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.552268713806683,
+      "learning_rate": 2.1257422774206816e-06,
+      "loss": 0.2751,
+      "step": 3946
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.25244065240587,
+      "learning_rate": 2.1233778704632002e-06,
+      "loss": 0.2583,
+      "step": 3947
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.46181807821369,
+      "learning_rate": 2.1210144246386378e-06,
+      "loss": 0.4048,
+      "step": 3948
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 13.599125276057082,
+      "learning_rate": 2.1186519407366725e-06,
+      "loss": 0.2998,
+      "step": 3949
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.69833834877681,
+      "learning_rate": 2.1162904195466455e-06,
+      "loss": 0.2974,
+      "step": 3950
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.782064797336112,
+      "learning_rate": 2.113929861857594e-06,
+      "loss": 0.2415,
+      "step": 3951
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.05525476633752,
+      "learning_rate": 2.1115702684582177e-06,
+      "loss": 0.2354,
+      "step": 3952
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.488998620156735,
+      "learning_rate": 2.1092116401369033e-06,
+      "loss": 0.2205,
+      "step": 3953
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.6604425215527465,
+      "learning_rate": 2.1068539776817115e-06,
+      "loss": 0.1997,
+      "step": 3954
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.936366716585368,
+      "learning_rate": 2.1044972818803816e-06,
+      "loss": 0.2666,
+      "step": 3955
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 13.919791218319379,
+      "learning_rate": 2.1021415535203294e-06,
+      "loss": 0.2935,
+      "step": 3956
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 12.83542796098615,
+      "learning_rate": 2.0997867933886467e-06,
+      "loss": 0.2839,
+      "step": 3957
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.346017326459489,
+      "learning_rate": 2.0974330022721044e-06,
+      "loss": 0.3179,
+      "step": 3958
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.944822762505321,
+      "learning_rate": 2.0950801809571466e-06,
+      "loss": 0.2749,
+      "step": 3959
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.391359360322923,
+      "learning_rate": 2.0927283302298944e-06,
+      "loss": 0.303,
+      "step": 3960
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.032353529389161,
+      "learning_rate": 2.0903774508761477e-06,
+      "loss": 0.2107,
+      "step": 3961
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.485677725607823,
+      "learning_rate": 2.0880275436813726e-06,
+      "loss": 0.1442,
+      "step": 3962
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.077591332146431,
+      "learning_rate": 2.0856786094307247e-06,
+      "loss": 0.2629,
+      "step": 3963
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.819852695103354,
+      "learning_rate": 2.0833306489090186e-06,
+      "loss": 0.356,
+      "step": 3964
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 9.563988233532719,
+      "learning_rate": 2.08098366290076e-06,
+      "loss": 0.2456,
+      "step": 3965
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.809406976811916,
+      "learning_rate": 2.078637652190112e-06,
+      "loss": 0.2141,
+      "step": 3966
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.610564298049212,
+      "learning_rate": 2.0762926175609287e-06,
+      "loss": 0.2444,
+      "step": 3967
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.367480219472963,
+      "learning_rate": 2.0739485597967237e-06,
+      "loss": 0.3579,
+      "step": 3968
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 9.40280542559146,
+      "learning_rate": 2.0716054796806916e-06,
+      "loss": 0.24,
+      "step": 3969
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.096437879516504,
+      "learning_rate": 2.0692633779956998e-06,
+      "loss": 0.2996,
+      "step": 3970
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.01349355918468,
+      "learning_rate": 2.0669222555242884e-06,
+      "loss": 0.3135,
+      "step": 3971
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.179708500485882,
+      "learning_rate": 2.064582113048669e-06,
+      "loss": 0.2285,
+      "step": 3972
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.404710632894906,
+      "learning_rate": 2.0622429513507275e-06,
+      "loss": 0.2913,
+      "step": 3973
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.974891410309862,
+      "learning_rate": 2.05990477121202e-06,
+      "loss": 0.3313,
+      "step": 3974
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.45063399330934,
+      "learning_rate": 2.0575675734137773e-06,
+      "loss": 0.2144,
+      "step": 3975
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.698967473996044,
+      "learning_rate": 2.0552313587369003e-06,
+      "loss": 0.2664,
+      "step": 3976
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 12.174398856640751,
+      "learning_rate": 2.052896127961963e-06,
+      "loss": 0.2932,
+      "step": 3977
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.769798231540381,
+      "learning_rate": 2.050561881869205e-06,
+      "loss": 0.2498,
+      "step": 3978
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.068055612015428,
+      "learning_rate": 2.048228621238547e-06,
+      "loss": 0.2527,
+      "step": 3979
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.886413081007158,
+      "learning_rate": 2.0458963468495692e-06,
+      "loss": 0.2693,
+      "step": 3980
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.972122538335825,
+      "learning_rate": 2.0435650594815338e-06,
+      "loss": 0.179,
+      "step": 3981
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.441213776975122,
+      "learning_rate": 2.0412347599133607e-06,
+      "loss": 0.2732,
+      "step": 3982
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 12.836617592039206,
+      "learning_rate": 2.0389054489236534e-06,
+      "loss": 0.2695,
+      "step": 3983
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.202241722151822,
+      "learning_rate": 2.03657712729067e-06,
+      "loss": 0.2607,
+      "step": 3984
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.254407875075486,
+      "learning_rate": 2.034249795792355e-06,
+      "loss": 0.177,
+      "step": 3985
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.31910673688725,
+      "learning_rate": 2.031923455206306e-06,
+      "loss": 0.2949,
+      "step": 3986
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.84033377167254,
+      "learning_rate": 2.0295981063098e-06,
+      "loss": 0.3264,
+      "step": 3987
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.599633875942123,
+      "learning_rate": 2.027273749879777e-06,
+      "loss": 0.2413,
+      "step": 3988
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.28616607550584,
+      "learning_rate": 2.02495038669285e-06,
+      "loss": 0.2134,
+      "step": 3989
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.91463101296468,
+      "learning_rate": 2.0226280175252966e-06,
+      "loss": 0.2627,
+      "step": 3990
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.012332249144231,
+      "learning_rate": 2.020306643153063e-06,
+      "loss": 0.2666,
+      "step": 3991
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.681187563001785,
+      "learning_rate": 2.0179862643517657e-06,
+      "loss": 0.3252,
+      "step": 3992
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.817386411445645,
+      "learning_rate": 2.015666881896684e-06,
+      "loss": 0.2441,
+      "step": 3993
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.064995665760662,
+      "learning_rate": 2.0133484965627683e-06,
+      "loss": 0.2886,
+      "step": 3994
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.865028740128867,
+      "learning_rate": 2.0110311091246333e-06,
+      "loss": 0.3228,
+      "step": 3995
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.764784698091306,
+      "learning_rate": 2.0087147203565614e-06,
+      "loss": 0.2949,
+      "step": 3996
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.094254980125088,
+      "learning_rate": 2.0063993310325013e-06,
+      "loss": 0.217,
+      "step": 3997
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.135626121609093,
+      "learning_rate": 2.0040849419260682e-06,
+      "loss": 0.251,
+      "step": 3998
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.33339179679279,
+      "learning_rate": 2.0017715538105416e-06,
+      "loss": 0.2485,
+      "step": 3999
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.79996387979217,
+      "learning_rate": 1.9994591674588677e-06,
+      "loss": 0.2537,
+      "step": 4000
+    },
+    {
+      "epoch": 2.86,
+      "eval_avg_AUC": 0.7879958586606428,
+      "eval_avg_Accuracy": 0.6990218832891246,
+      "eval_avg_Accuracy-right": 0.8709403938959176,
+      "eval_avg_Accuracy-wrong": 0.3992494882874687,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6623938351746894,
+      "eval_last_AUC": 0.8070917433092194,
+      "eval_last_Accuracy": 0.7334631962864722,
+      "eval_last_Accuracy-right": 0.8252902047737055,
+      "eval_last_Accuracy-wrong": 0.5733454628155561,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6848945312092272,
+      "eval_max_AUC": 0.7616910032003927,
+      "eval_max_Accuracy": 0.6416611405835544,
+      "eval_max_Accuracy-right": 0.9620451284726751,
+      "eval_max_Accuracy-wrong": 0.08301114396179213,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6150202437158859,
+      "eval_min_AUC": 0.7931598852081432,
+      "eval_min_Accuracy": 0.720407824933687,
+      "eval_min_Accuracy-right": 0.7332072518586148,
+      "eval_min_Accuracy-wrong": 0.6980896065499204,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6685353032201301,
+      "eval_prod_AUC": 0.7945893958373447,
+      "eval_prod_Accuracy": 0.702420424403183,
+      "eval_prod_Accuracy-right": 0.6255380200860832,
+      "eval_prod_Accuracy-wrong": 0.8364794177848534,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6701653120639736,
+      "eval_runtime": 247.3512,
+      "eval_samples_per_second": 97.545,
+      "eval_steps_per_second": 3.048,
+      "eval_sum_AUC": 0.6614690230562896,
+      "eval_sum_Accuracy": 0.6374336870026526,
+      "eval_sum_Accuracy-right": 0.9853919394808921,
+      "eval_sum_Accuracy-wrong": 0.030702751876279282,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6497230981517113,
+      "step": 4000
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.294863167005364,
+      "learning_rate": 1.9971477836436575e-06,
+      "loss": 0.3467,
+      "step": 4001
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.327232865174512,
+      "learning_rate": 1.99483740313719e-06,
+      "loss": 0.2378,
+      "step": 4002
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.22041031697138,
+      "learning_rate": 1.9925280267114e-06,
+      "loss": 0.248,
+      "step": 4003
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.21213166626409,
+      "learning_rate": 1.9902196551379006e-06,
+      "loss": 0.2129,
+      "step": 4004
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.610468088544783,
+      "learning_rate": 1.987912289187954e-06,
+      "loss": 0.2856,
+      "step": 4005
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.763270070414201,
+      "learning_rate": 1.9856059296325027e-06,
+      "loss": 0.3086,
+      "step": 4006
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.34977464133706,
+      "learning_rate": 1.9833005772421354e-06,
+      "loss": 0.1912,
+      "step": 4007
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.99525600692564,
+      "learning_rate": 1.980996232787121e-06,
+      "loss": 0.2854,
+      "step": 4008
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 8.482697059629118,
+      "learning_rate": 1.978692897037377e-06,
+      "loss": 0.2537,
+      "step": 4009
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.88855064417953,
+      "learning_rate": 1.9763905707624975e-06,
+      "loss": 0.2056,
+      "step": 4010
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.515119687094433,
+      "learning_rate": 1.974089254731727e-06,
+      "loss": 0.3047,
+      "step": 4011
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.460042983380315,
+      "learning_rate": 1.97178894971398e-06,
+      "loss": 0.1714,
+      "step": 4012
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.000891418019652,
+      "learning_rate": 1.9694896564778317e-06,
+      "loss": 0.2881,
+      "step": 4013
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.457581961485896,
+      "learning_rate": 1.9671913757915173e-06,
+      "loss": 0.1953,
+      "step": 4014
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.329358437695408,
+      "learning_rate": 1.964894108422936e-06,
+      "loss": 0.2489,
+      "step": 4015
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.507027263953846,
+      "learning_rate": 1.962597855139648e-06,
+      "loss": 0.2153,
+      "step": 4016
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.10099282908348,
+      "learning_rate": 1.960302616708873e-06,
+      "loss": 0.2883,
+      "step": 4017
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.480233105548061,
+      "learning_rate": 1.9580083938974937e-06,
+      "loss": 0.223,
+      "step": 4018
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.559695185275773,
+      "learning_rate": 1.9557151874720526e-06,
+      "loss": 0.2325,
+      "step": 4019
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.141810003892173,
+      "learning_rate": 1.953422998198754e-06,
+      "loss": 0.1979,
+      "step": 4020
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.982944575808363,
+      "learning_rate": 1.9511318268434554e-06,
+      "loss": 0.3599,
+      "step": 4021
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.211224106691908,
+      "learning_rate": 1.9488416741716877e-06,
+      "loss": 0.2395,
+      "step": 4022
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.569449118407048,
+      "learning_rate": 1.946552540948625e-06,
+      "loss": 0.2422,
+      "step": 4023
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.650414366436912,
+      "learning_rate": 1.944264427939118e-06,
+      "loss": 0.3,
+      "step": 4024
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 7.739566513501393,
+      "learning_rate": 1.941977335907659e-06,
+      "loss": 0.1785,
+      "step": 4025
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.6977439093024,
+      "learning_rate": 1.939691265618417e-06,
+      "loss": 0.3027,
+      "step": 4026
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.915637048581093,
+      "learning_rate": 1.9374062178352036e-06,
+      "loss": 0.2341,
+      "step": 4027
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.897353050548803,
+      "learning_rate": 1.935122193321499e-06,
+      "loss": 0.2625,
+      "step": 4028
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.08317279527024,
+      "learning_rate": 1.932839192840436e-06,
+      "loss": 0.2683,
+      "step": 4029
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 12.704675205411846,
+      "learning_rate": 1.930557217154809e-06,
+      "loss": 0.2295,
+      "step": 4030
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.351026891672202,
+      "learning_rate": 1.9282762670270693e-06,
+      "loss": 0.2058,
+      "step": 4031
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.34895226658225,
+      "learning_rate": 1.925996343219323e-06,
+      "loss": 0.3765,
+      "step": 4032
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.597504826581677,
+      "learning_rate": 1.923717446493336e-06,
+      "loss": 0.3281,
+      "step": 4033
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.653477339408518,
+      "learning_rate": 1.9214395776105297e-06,
+      "loss": 0.2778,
+      "step": 4034
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.944620053801449,
+      "learning_rate": 1.919162737331983e-06,
+      "loss": 0.2278,
+      "step": 4035
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.550617411583413,
+      "learning_rate": 1.9168869264184296e-06,
+      "loss": 0.2656,
+      "step": 4036
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.945812481491696,
+      "learning_rate": 1.9146121456302613e-06,
+      "loss": 0.2358,
+      "step": 4037
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.422504829133128,
+      "learning_rate": 1.9123383957275237e-06,
+      "loss": 0.2812,
+      "step": 4038
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 17.311605530516562,
+      "learning_rate": 1.91006567746992e-06,
+      "loss": 0.2559,
+      "step": 4039
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.786074439406534,
+      "learning_rate": 1.907793991616806e-06,
+      "loss": 0.2878,
+      "step": 4040
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.6922862971148,
+      "learning_rate": 1.9055233389271955e-06,
+      "loss": 0.2671,
+      "step": 4041
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.628962310466495,
+      "learning_rate": 1.9032537201597556e-06,
+      "loss": 0.218,
+      "step": 4042
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.911645743068291,
+      "learning_rate": 1.9009851360728077e-06,
+      "loss": 0.2557,
+      "step": 4043
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.595771054903054,
+      "learning_rate": 1.898717587424328e-06,
+      "loss": 0.3633,
+      "step": 4044
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.518102120271452,
+      "learning_rate": 1.8964510749719484e-06,
+      "loss": 0.2349,
+      "step": 4045
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.22905494698704,
+      "learning_rate": 1.8941855994729497e-06,
+      "loss": 0.29,
+      "step": 4046
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.207905032025637,
+      "learning_rate": 1.8919211616842703e-06,
+      "loss": 0.209,
+      "step": 4047
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 12.410396427093156,
+      "learning_rate": 1.8896577623625017e-06,
+      "loss": 0.2229,
+      "step": 4048
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.527011086500547,
+      "learning_rate": 1.887395402263888e-06,
+      "loss": 0.3206,
+      "step": 4049
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.672016334197911,
+      "learning_rate": 1.8851340821443248e-06,
+      "loss": 0.2427,
+      "step": 4050
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.125832400500446,
+      "learning_rate": 1.882873802759362e-06,
+      "loss": 0.312,
+      "step": 4051
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.046251858097985,
+      "learning_rate": 1.8806145648642005e-06,
+      "loss": 0.2979,
+      "step": 4052
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.838718306043596,
+      "learning_rate": 1.8783563692136936e-06,
+      "loss": 0.2334,
+      "step": 4053
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.943639533290575,
+      "learning_rate": 1.8760992165623465e-06,
+      "loss": 0.3025,
+      "step": 4054
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.666070064383465,
+      "learning_rate": 1.873843107664316e-06,
+      "loss": 0.2688,
+      "step": 4055
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.583556219560066,
+      "learning_rate": 1.87158804327341e-06,
+      "loss": 0.2915,
+      "step": 4056
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.0617200379471425,
+      "learning_rate": 1.8693340241430874e-06,
+      "loss": 0.1807,
+      "step": 4057
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.72852127215646,
+      "learning_rate": 1.867081051026458e-06,
+      "loss": 0.3193,
+      "step": 4058
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.749413424323619,
+      "learning_rate": 1.8648291246762818e-06,
+      "loss": 0.2511,
+      "step": 4059
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.654719307217244,
+      "learning_rate": 1.8625782458449693e-06,
+      "loss": 0.2336,
+      "step": 4060
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.407631474167662,
+      "learning_rate": 1.860328415284583e-06,
+      "loss": 0.2976,
+      "step": 4061
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.48686271000459,
+      "learning_rate": 1.8580796337468276e-06,
+      "loss": 0.1995,
+      "step": 4062
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.274457610353167,
+      "learning_rate": 1.8558319019830695e-06,
+      "loss": 0.2031,
+      "step": 4063
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.802148591200808,
+      "learning_rate": 1.853585220744311e-06,
+      "loss": 0.2812,
+      "step": 4064
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.092229309343045,
+      "learning_rate": 1.851339590781217e-06,
+      "loss": 0.2485,
+      "step": 4065
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 21.463031361010216,
+      "learning_rate": 1.8490950128440877e-06,
+      "loss": 0.355,
+      "step": 4066
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.998522968766485,
+      "learning_rate": 1.8468514876828847e-06,
+      "loss": 0.2788,
+      "step": 4067
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.552298244913189,
+      "learning_rate": 1.844609016047204e-06,
+      "loss": 0.3298,
+      "step": 4068
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.40982638450866,
+      "learning_rate": 1.8423675986863054e-06,
+      "loss": 0.3369,
+      "step": 4069
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.645489950141965,
+      "learning_rate": 1.8401272363490818e-06,
+      "loss": 0.21,
+      "step": 4070
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.723345576862725,
+      "learning_rate": 1.8378879297840818e-06,
+      "loss": 0.3018,
+      "step": 4071
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.662657815513155,
+      "learning_rate": 1.8356496797395002e-06,
+      "loss": 0.2375,
+      "step": 4072
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.360949595986845,
+      "learning_rate": 1.8334124869631765e-06,
+      "loss": 0.2791,
+      "step": 4073
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 9.685132876470785,
+      "learning_rate": 1.8311763522025994e-06,
+      "loss": 0.2039,
+      "step": 4074
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 9.895276388339477,
+      "learning_rate": 1.828941276204903e-06,
+      "loss": 0.2727,
+      "step": 4075
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.432764560809574,
+      "learning_rate": 1.8267072597168673e-06,
+      "loss": 0.2058,
+      "step": 4076
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.308441456073734,
+      "learning_rate": 1.8244743034849193e-06,
+      "loss": 0.2285,
+      "step": 4077
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 12.095081699709256,
+      "learning_rate": 1.8222424082551303e-06,
+      "loss": 0.2605,
+      "step": 4078
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.298604259683547,
+      "learning_rate": 1.820011574773221e-06,
+      "loss": 0.2684,
+      "step": 4079
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.019721690601129,
+      "learning_rate": 1.8177818037845485e-06,
+      "loss": 0.2024,
+      "step": 4080
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 16.13667342258012,
+      "learning_rate": 1.8155530960341273e-06,
+      "loss": 0.2921,
+      "step": 4081
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.865382533021625,
+      "learning_rate": 1.8133254522666033e-06,
+      "loss": 0.293,
+      "step": 4082
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.783714143176804,
+      "learning_rate": 1.8110988732262808e-06,
+      "loss": 0.2706,
+      "step": 4083
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.22742606505741,
+      "learning_rate": 1.8088733596570945e-06,
+      "loss": 0.2429,
+      "step": 4084
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 11.119724412060055,
+      "learning_rate": 1.806648912302636e-06,
+      "loss": 0.2507,
+      "step": 4085
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 8.320477068822154,
+      "learning_rate": 1.8044255319061287e-06,
+      "loss": 0.28,
+      "step": 4086
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 8.759806331612182,
+      "learning_rate": 1.8022032192104517e-06,
+      "loss": 0.2698,
+      "step": 4087
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 14.35985526948073,
+      "learning_rate": 1.7999819749581154e-06,
+      "loss": 0.3721,
+      "step": 4088
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.828623960795465,
+      "learning_rate": 1.797761799891281e-06,
+      "loss": 0.3364,
+      "step": 4089
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 11.744365228222355,
+      "learning_rate": 1.7955426947517507e-06,
+      "loss": 0.24,
+      "step": 4090
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.279599538506409,
+      "learning_rate": 1.793324660280968e-06,
+      "loss": 0.2603,
+      "step": 4091
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 12.284245558123292,
+      "learning_rate": 1.7911076972200193e-06,
+      "loss": 0.2681,
+      "step": 4092
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 10.759102641777156,
+      "learning_rate": 1.7888918063096334e-06,
+      "loss": 0.2205,
+      "step": 4093
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 10.414253236529335,
+      "learning_rate": 1.7866769882901814e-06,
+      "loss": 0.2959,
+      "step": 4094
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 12.100896779951352,
+      "learning_rate": 1.784463243901674e-06,
+      "loss": 0.25,
+      "step": 4095
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 15.39298134834118,
+      "learning_rate": 1.7822505738837648e-06,
+      "loss": 0.397,
+      "step": 4096
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 7.320159627540454,
+      "learning_rate": 1.7800389789757483e-06,
+      "loss": 0.2217,
+      "step": 4097
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.705040302239233,
+      "learning_rate": 1.7778284599165597e-06,
+      "loss": 0.2642,
+      "step": 4098
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 12.866603928279114,
+      "learning_rate": 1.7756190174447734e-06,
+      "loss": 0.2966,
+      "step": 4099
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 12.300852071369984,
+      "learning_rate": 1.7734106522986061e-06,
+      "loss": 0.2854,
+      "step": 4100
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.0406718245773545,
+      "learning_rate": 1.7712033652159133e-06,
+      "loss": 0.2153,
+      "step": 4101
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 16.366009476816807,
+      "learning_rate": 1.7689971569341907e-06,
+      "loss": 0.3169,
+      "step": 4102
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.099939081723868,
+      "learning_rate": 1.7667920281905738e-06,
+      "loss": 0.1508,
+      "step": 4103
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 17.54395094001839,
+      "learning_rate": 1.764587979721838e-06,
+      "loss": 0.3247,
+      "step": 4104
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 9.355552433889878,
+      "learning_rate": 1.7623850122643926e-06,
+      "loss": 0.1946,
+      "step": 4105
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 11.451247973760308,
+      "learning_rate": 1.7601831265542968e-06,
+      "loss": 0.2495,
+      "step": 4106
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.38373080535724,
+      "learning_rate": 1.7579823233272337e-06,
+      "loss": 0.2188,
+      "step": 4107
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 32.26992819582037,
+      "learning_rate": 1.7557826033185404e-06,
+      "loss": 0.3677,
+      "step": 4108
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 9.31935554317174,
+      "learning_rate": 1.7535839672631772e-06,
+      "loss": 0.3042,
+      "step": 4109
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.333138141169738,
+      "learning_rate": 1.7513864158957556e-06,
+      "loss": 0.1741,
+      "step": 4110
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 12.064201456722603,
+      "learning_rate": 1.7491899499505122e-06,
+      "loss": 0.2629,
+      "step": 4111
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 12.246914808812818,
+      "learning_rate": 1.746994570161334e-06,
+      "loss": 0.2664,
+      "step": 4112
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 16.025286460767063,
+      "learning_rate": 1.7448002772617324e-06,
+      "loss": 0.2764,
+      "step": 4113
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.595517375370573,
+      "learning_rate": 1.7426070719848632e-06,
+      "loss": 0.1829,
+      "step": 4114
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.033101510979936,
+      "learning_rate": 1.7404149550635173e-06,
+      "loss": 0.2468,
+      "step": 4115
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 15.395760488217663,
+      "learning_rate": 1.7382239272301221e-06,
+      "loss": 0.345,
+      "step": 4116
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 8.718081017809373,
+      "learning_rate": 1.7360339892167404e-06,
+      "loss": 0.2185,
+      "step": 4117
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 11.570151019962179,
+      "learning_rate": 1.7338451417550712e-06,
+      "loss": 0.2932,
+      "step": 4118
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 15.489633738751614,
+      "learning_rate": 1.7316573855764485e-06,
+      "loss": 0.3535,
+      "step": 4119
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 11.204623438557102,
+      "learning_rate": 1.7294707214118434e-06,
+      "loss": 0.2524,
+      "step": 4120
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 8.024037924894355,
+      "learning_rate": 1.7272851499918603e-06,
+      "loss": 0.2061,
+      "step": 4121
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 10.025616633736863,
+      "learning_rate": 1.725100672046741e-06,
+      "loss": 0.2207,
+      "step": 4122
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.237786503227326,
+      "learning_rate": 1.7229172883063556e-06,
+      "loss": 0.2344,
+      "step": 4123
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 11.400008128488517,
+      "learning_rate": 1.7207349995002192e-06,
+      "loss": 0.2041,
+      "step": 4124
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 13.04513731466515,
+      "learning_rate": 1.7185538063574692e-06,
+      "loss": 0.3313,
+      "step": 4125
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.065512097775526,
+      "learning_rate": 1.7163737096068883e-06,
+      "loss": 0.198,
+      "step": 4126
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 11.819709518012287,
+      "learning_rate": 1.7141947099768818e-06,
+      "loss": 0.2305,
+      "step": 4127
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 6.6082456220174,
+      "learning_rate": 1.7120168081955001e-06,
+      "loss": 0.168,
+      "step": 4128
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.349276591227465,
+      "learning_rate": 1.7098400049904163e-06,
+      "loss": 0.2913,
+      "step": 4129
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 8.586751031921361,
+      "learning_rate": 1.707664301088941e-06,
+      "loss": 0.2065,
+      "step": 4130
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 14.717204391015239,
+      "learning_rate": 1.705489697218019e-06,
+      "loss": 0.3105,
+      "step": 4131
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.460015525511492,
+      "learning_rate": 1.7033161941042248e-06,
+      "loss": 0.208,
+      "step": 4132
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 11.400780736034415,
+      "learning_rate": 1.7011437924737666e-06,
+      "loss": 0.3025,
+      "step": 4133
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.28553264691427,
+      "learning_rate": 1.6989724930524843e-06,
+      "loss": 0.2966,
+      "step": 4134
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.477581081372428,
+      "learning_rate": 1.6968022965658492e-06,
+      "loss": 0.183,
+      "step": 4135
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.331783836914225,
+      "learning_rate": 1.694633203738964e-06,
+      "loss": 0.281,
+      "step": 4136
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.89865181755367,
+      "learning_rate": 1.6924652152965632e-06,
+      "loss": 0.2869,
+      "step": 4137
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.774512612157912,
+      "learning_rate": 1.690298331963014e-06,
+      "loss": 0.2534,
+      "step": 4138
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.468974723185733,
+      "learning_rate": 1.6881325544623067e-06,
+      "loss": 0.2205,
+      "step": 4139
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.15613380148459,
+      "learning_rate": 1.6859678835180749e-06,
+      "loss": 0.2581,
+      "step": 4140
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.938617854657009,
+      "learning_rate": 1.6838043198535693e-06,
+      "loss": 0.2761,
+      "step": 4141
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.526184909674672,
+      "learning_rate": 1.681641864191682e-06,
+      "loss": 0.4287,
+      "step": 4142
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 10.610954712676994,
+      "learning_rate": 1.6794805172549244e-06,
+      "loss": 0.2856,
+      "step": 4143
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.130277611997542,
+      "learning_rate": 1.6773202797654486e-06,
+      "loss": 0.2515,
+      "step": 4144
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.42493725544239,
+      "learning_rate": 1.6751611524450235e-06,
+      "loss": 0.2637,
+      "step": 4145
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 10.082805920327193,
+      "learning_rate": 1.6730031360150605e-06,
+      "loss": 0.2311,
+      "step": 4146
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 8.655752640833025,
+      "learning_rate": 1.670846231196588e-06,
+      "loss": 0.2336,
+      "step": 4147
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.13420076254242,
+      "learning_rate": 1.6686904387102692e-06,
+      "loss": 0.2661,
+      "step": 4148
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 10.637960833953823,
+      "learning_rate": 1.6665357592763948e-06,
+      "loss": 0.2262,
+      "step": 4149
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.514907798772807,
+      "learning_rate": 1.6643821936148834e-06,
+      "loss": 0.2236,
+      "step": 4150
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.881896754186977,
+      "learning_rate": 1.6622297424452817e-06,
+      "loss": 0.2382,
+      "step": 4151
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 6.96329529621935,
+      "learning_rate": 1.6600784064867625e-06,
+      "loss": 0.2207,
+      "step": 4152
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.235052105560225,
+      "learning_rate": 1.6579281864581275e-06,
+      "loss": 0.2438,
+      "step": 4153
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 8.127333202419477,
+      "learning_rate": 1.6557790830778058e-06,
+      "loss": 0.2133,
+      "step": 4154
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 11.008855273874723,
+      "learning_rate": 1.6536310970638525e-06,
+      "loss": 0.2527,
+      "step": 4155
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.218850473628057,
+      "learning_rate": 1.6514842291339494e-06,
+      "loss": 0.2563,
+      "step": 4156
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 9.405684231978427,
+      "learning_rate": 1.6493384800054052e-06,
+      "loss": 0.2542,
+      "step": 4157
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 12.309209754540694,
+      "learning_rate": 1.6471938503951546e-06,
+      "loss": 0.2742,
+      "step": 4158
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 11.300836718904058,
+      "learning_rate": 1.6450503410197582e-06,
+      "loss": 0.2483,
+      "step": 4159
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 13.129412391262694,
+      "learning_rate": 1.6429079525954023e-06,
+      "loss": 0.4229,
+      "step": 4160
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 9.519718325999454,
+      "learning_rate": 1.6407666858378985e-06,
+      "loss": 0.2643,
+      "step": 4161
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 12.957849553694674,
+      "learning_rate": 1.6386265414626834e-06,
+      "loss": 0.2549,
+      "step": 4162
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 8.915315627608456,
+      "learning_rate": 1.636487520184822e-06,
+      "loss": 0.2188,
+      "step": 4163
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.583275524212816,
+      "learning_rate": 1.6343496227189948e-06,
+      "loss": 0.2233,
+      "step": 4164
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 11.15496354912215,
+      "learning_rate": 1.632212849779521e-06,
+      "loss": 0.2312,
+      "step": 4165
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 9.269917208785433,
+      "learning_rate": 1.630077202080328e-06,
+      "loss": 0.1956,
+      "step": 4166
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 12.657935965779476,
+      "learning_rate": 1.6279426803349828e-06,
+      "loss": 0.2642,
+      "step": 4167
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.290607785408874,
+      "learning_rate": 1.6258092852566625e-06,
+      "loss": 0.1294,
+      "step": 4168
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 10.711605990480061,
+      "learning_rate": 1.6236770175581807e-06,
+      "loss": 0.2771,
+      "step": 4169
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 9.135466780344332,
+      "learning_rate": 1.62154587795196e-06,
+      "loss": 0.2688,
+      "step": 4170
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 16.83078442721911,
+      "learning_rate": 1.6194158671500616e-06,
+      "loss": 0.214,
+      "step": 4171
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 10.353277203263982,
+      "learning_rate": 1.6172869858641554e-06,
+      "loss": 0.2915,
+      "step": 4172
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 9.629289861141057,
+      "learning_rate": 1.6151592348055433e-06,
+      "loss": 0.2556,
+      "step": 4173
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 8.912948810612065,
+      "learning_rate": 1.6130326146851455e-06,
+      "loss": 0.2493,
+      "step": 4174
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 12.026915751127602,
+      "learning_rate": 1.6109071262135056e-06,
+      "loss": 0.252,
+      "step": 4175
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 13.580442426663522,
+      "learning_rate": 1.608782770100789e-06,
+      "loss": 0.2695,
+      "step": 4176
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 11.492422800213438,
+      "learning_rate": 1.6066595470567825e-06,
+      "loss": 0.2375,
+      "step": 4177
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 14.85339863796703,
+      "learning_rate": 1.6045374577908944e-06,
+      "loss": 0.3145,
+      "step": 4178
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 8.415153390378142,
+      "learning_rate": 1.6024165030121542e-06,
+      "loss": 0.1987,
+      "step": 4179
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 16.294077072163837,
+      "learning_rate": 1.6002966834292116e-06,
+      "loss": 0.3169,
+      "step": 4180
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 9.594024560968199,
+      "learning_rate": 1.5981779997503405e-06,
+      "loss": 0.2385,
+      "step": 4181
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.105734484271524,
+      "learning_rate": 1.5960604526834266e-06,
+      "loss": 0.2812,
+      "step": 4182
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.49417910592248,
+      "learning_rate": 1.5939440429359888e-06,
+      "loss": 0.2888,
+      "step": 4183
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.633425145280981,
+      "learning_rate": 1.591828771215152e-06,
+      "loss": 0.2197,
+      "step": 4184
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.785768156691029,
+      "learning_rate": 1.5897146382276752e-06,
+      "loss": 0.2354,
+      "step": 4185
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.318466928652242,
+      "learning_rate": 1.587601644679922e-06,
+      "loss": 0.27,
+      "step": 4186
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 10.41034009289898,
+      "learning_rate": 1.58548979127789e-06,
+      "loss": 0.2205,
+      "step": 4187
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.937568460758499,
+      "learning_rate": 1.5833790787271819e-06,
+      "loss": 0.1968,
+      "step": 4188
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.46241310438527,
+      "learning_rate": 1.5812695077330325e-06,
+      "loss": 0.2495,
+      "step": 4189
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 16.29389775463895,
+      "learning_rate": 1.5791610790002838e-06,
+      "loss": 0.3013,
+      "step": 4190
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 13.717352648476366,
+      "learning_rate": 1.577053793233403e-06,
+      "loss": 0.2676,
+      "step": 4191
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.711414459758286,
+      "learning_rate": 1.5749476511364726e-06,
+      "loss": 0.332,
+      "step": 4192
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.705499968975047,
+      "learning_rate": 1.5728426534131946e-06,
+      "loss": 0.1589,
+      "step": 4193
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 12.787386008766704,
+      "learning_rate": 1.5707388007668877e-06,
+      "loss": 0.3003,
+      "step": 4194
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.667900077969025,
+      "learning_rate": 1.568636093900488e-06,
+      "loss": 0.2053,
+      "step": 4195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 15.07311298463385,
+      "learning_rate": 1.5665345335165488e-06,
+      "loss": 0.2358,
+      "step": 4196
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 11.966148032918056,
+      "learning_rate": 1.5644341203172415e-06,
+      "loss": 0.3,
+      "step": 4197
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 6.900175323732201,
+      "learning_rate": 1.5623348550043516e-06,
+      "loss": 0.1863,
+      "step": 4198
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 13.34318877693945,
+      "learning_rate": 1.5602367382792839e-06,
+      "loss": 0.2773,
+      "step": 4199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 14.531517790747417,
+      "learning_rate": 1.5581397708430578e-06,
+      "loss": 0.3015,
+      "step": 4200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 12.05529215822636,
+      "learning_rate": 1.556043953396309e-06,
+      "loss": 0.23,
+      "step": 4201
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 10.926738561041796,
+      "learning_rate": 1.5539492866392891e-06,
+      "loss": 0.201,
+      "step": 4202
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 11.655803933246485,
+      "learning_rate": 1.551855771271865e-06,
+      "loss": 0.2241,
+      "step": 4203
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 9.49786951614942,
+      "learning_rate": 1.5497634079935198e-06,
+      "loss": 0.2363,
+      "step": 4204
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.562319258107449,
+      "learning_rate": 1.5476721975033498e-06,
+      "loss": 0.1233,
+      "step": 4205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.707661563901169,
+      "learning_rate": 1.5455821405000703e-06,
+      "loss": 0.1448,
+      "step": 4206
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.98664541474231,
+      "learning_rate": 1.5434932376820039e-06,
+      "loss": 0.1188,
+      "step": 4207
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 7.560483987455881,
+      "learning_rate": 1.5414054897470942e-06,
+      "loss": 0.1791,
+      "step": 4208
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.315253135542903,
+      "learning_rate": 1.5393188973928957e-06,
+      "loss": 0.1033,
+      "step": 4209
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.611585149366166,
+      "learning_rate": 1.5372334613165784e-06,
+      "loss": 0.1366,
+      "step": 4210
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.021499687105682,
+      "learning_rate": 1.5351491822149255e-06,
+      "loss": 0.1254,
+      "step": 4211
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 7.3862962459769514,
+      "learning_rate": 1.533066060784333e-06,
+      "loss": 0.1672,
+      "step": 4212
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.011539859613148,
+      "learning_rate": 1.5309840977208096e-06,
+      "loss": 0.1176,
+      "step": 4213
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.460881953756751,
+      "learning_rate": 1.5289032937199793e-06,
+      "loss": 0.151,
+      "step": 4214
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.886640344653612,
+      "learning_rate": 1.5268236494770772e-06,
+      "loss": 0.1545,
+      "step": 4215
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 8.44915060491121,
+      "learning_rate": 1.5247451656869499e-06,
+      "loss": 0.163,
+      "step": 4216
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.682171952885785,
+      "learning_rate": 1.5226678430440588e-06,
+      "loss": 0.1411,
+      "step": 4217
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.312109950691412,
+      "learning_rate": 1.5205916822424755e-06,
+      "loss": 0.1438,
+      "step": 4218
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.1623724941925175,
+      "learning_rate": 1.5185166839758836e-06,
+      "loss": 0.1252,
+      "step": 4219
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.471874299848998,
+      "learning_rate": 1.5164428489375789e-06,
+      "loss": 0.167,
+      "step": 4220
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.334160602466491,
+      "learning_rate": 1.5143701778204683e-06,
+      "loss": 0.1064,
+      "step": 4221
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.790958055528099,
+      "learning_rate": 1.5122986713170712e-06,
+      "loss": 0.1418,
+      "step": 4222
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.360409064903208,
+      "learning_rate": 1.510228330119512e-06,
+      "loss": 0.1225,
+      "step": 4223
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.105021948656687,
+      "learning_rate": 1.5081591549195357e-06,
+      "loss": 0.1021,
+      "step": 4224
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 7.345073889890468,
+      "learning_rate": 1.5060911464084864e-06,
+      "loss": 0.1582,
+      "step": 4225
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 8.936176216847798,
+      "learning_rate": 1.5040243052773312e-06,
+      "loss": 0.1592,
+      "step": 4226
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.858093703881658,
+      "learning_rate": 1.5019586322166323e-06,
+      "loss": 0.1035,
+      "step": 4227
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 7.983548488354241,
+      "learning_rate": 1.4998941279165773e-06,
+      "loss": 0.1723,
+      "step": 4228
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 10.277666504790933,
+      "learning_rate": 1.4978307930669483e-06,
+      "loss": 0.1005,
+      "step": 4229
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 8.547800791086695,
+      "learning_rate": 1.4957686283571498e-06,
+      "loss": 0.1653,
+      "step": 4230
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 7.992527436033497,
+      "learning_rate": 1.4937076344761858e-06,
+      "loss": 0.1716,
+      "step": 4231
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 6.936012071215718,
+      "learning_rate": 1.4916478121126732e-06,
+      "loss": 0.1029,
+      "step": 4232
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.3255729830641885,
+      "learning_rate": 1.4895891619548374e-06,
+      "loss": 0.1046,
+      "step": 4233
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 9.902272124427949,
+      "learning_rate": 1.4875316846905113e-06,
+      "loss": 0.1895,
+      "step": 4234
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.363267661890996,
+      "learning_rate": 1.4854753810071364e-06,
+      "loss": 0.1187,
+      "step": 4235
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 10.19363846323527,
+      "learning_rate": 1.4834202515917628e-06,
+      "loss": 0.2157,
+      "step": 4236
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 8.752546304397477,
+      "learning_rate": 1.4813662971310465e-06,
+      "loss": 0.1414,
+      "step": 4237
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.430637602687925,
+      "learning_rate": 1.4793135183112523e-06,
+      "loss": 0.1292,
+      "step": 4238
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.280105835812476,
+      "learning_rate": 1.477261915818251e-06,
+      "loss": 0.2173,
+      "step": 4239
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.619705907816034,
+      "learning_rate": 1.4752114903375243e-06,
+      "loss": 0.0962,
+      "step": 4240
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.991828112437791,
+      "learning_rate": 1.473162242554151e-06,
+      "loss": 0.1031,
+      "step": 4241
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.770685175159459,
+      "learning_rate": 1.47111417315283e-06,
+      "loss": 0.1162,
+      "step": 4242
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 17.414337641647094,
+      "learning_rate": 1.4690672828178532e-06,
+      "loss": 0.1656,
+      "step": 4243
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.439371331134558,
+      "learning_rate": 1.467021572233131e-06,
+      "loss": 0.1172,
+      "step": 4244
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.740136894772873,
+      "learning_rate": 1.4649770420821663e-06,
+      "loss": 0.1099,
+      "step": 4245
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.614518214652328,
+      "learning_rate": 1.4629336930480813e-06,
+      "loss": 0.1323,
+      "step": 4246
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 7.219648721235388,
+      "learning_rate": 1.4608915258135914e-06,
+      "loss": 0.1555,
+      "step": 4247
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.825768758767164,
+      "learning_rate": 1.4588505410610283e-06,
+      "loss": 0.141,
+      "step": 4248
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 11.787094208035665,
+      "learning_rate": 1.4568107394723175e-06,
+      "loss": 0.1892,
+      "step": 4249
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 7.676246808791587,
+      "learning_rate": 1.4547721217289972e-06,
+      "loss": 0.1515,
+      "step": 4250
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.044534219415103,
+      "learning_rate": 1.4527346885122073e-06,
+      "loss": 0.1312,
+      "step": 4251
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 9.284106118194046,
+      "learning_rate": 1.450698440502692e-06,
+      "loss": 0.1499,
+      "step": 4252
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.7252324984477916,
+      "learning_rate": 1.4486633783807997e-06,
+      "loss": 0.0759,
+      "step": 4253
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 7.906025467327889,
+      "learning_rate": 1.4466295028264822e-06,
+      "loss": 0.1543,
+      "step": 4254
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.324242408244137,
+      "learning_rate": 1.4445968145192951e-06,
+      "loss": 0.1219,
+      "step": 4255
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 5.877864296276188,
+      "learning_rate": 1.4425653141383977e-06,
+      "loss": 0.14,
+      "step": 4256
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 5.312508541147513,
+      "learning_rate": 1.4405350023625514e-06,
+      "loss": 0.0874,
+      "step": 4257
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 5.865148996474216,
+      "learning_rate": 1.4385058798701223e-06,
+      "loss": 0.1302,
+      "step": 4258
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 9.287278930298898,
+      "learning_rate": 1.4364779473390767e-06,
+      "loss": 0.1621,
+      "step": 4259
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 8.177031611561803,
+      "learning_rate": 1.4344512054469855e-06,
+      "loss": 0.1541,
+      "step": 4260
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.308894066948919,
+      "learning_rate": 1.4324256548710202e-06,
+      "loss": 0.1062,
+      "step": 4261
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 8.610067458675067,
+      "learning_rate": 1.430401296287955e-06,
+      "loss": 0.1289,
+      "step": 4262
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.510987310154899,
+      "learning_rate": 1.4283781303741662e-06,
+      "loss": 0.1726,
+      "step": 4263
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 5.778662504339967,
+      "learning_rate": 1.4263561578056307e-06,
+      "loss": 0.1449,
+      "step": 4264
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.4303052868247565,
+      "learning_rate": 1.4243353792579285e-06,
+      "loss": 0.1311,
+      "step": 4265
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.0518291313422985,
+      "learning_rate": 1.4223157954062344e-06,
+      "loss": 0.1038,
+      "step": 4266
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 8.014311988987807,
+      "learning_rate": 1.4202974069253362e-06,
+      "loss": 0.1282,
+      "step": 4267
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.347104216461915,
+      "learning_rate": 1.418280214489608e-06,
+      "loss": 0.1176,
+      "step": 4268
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.026484655477969,
+      "learning_rate": 1.416264218773038e-06,
+      "loss": 0.1364,
+      "step": 4269
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.206163622677682,
+      "learning_rate": 1.4142494204492007e-06,
+      "loss": 0.1118,
+      "step": 4270
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.449486750709002,
+      "learning_rate": 1.412235820191285e-06,
+      "loss": 0.1207,
+      "step": 4271
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 7.169597104216024,
+      "learning_rate": 1.4102234186720653e-06,
+      "loss": 0.1562,
+      "step": 4272
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 7.3213306438478165,
+      "learning_rate": 1.4082122165639285e-06,
+      "loss": 0.1102,
+      "step": 4273
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.644150306624669,
+      "learning_rate": 1.4062022145388503e-06,
+      "loss": 0.0866,
+      "step": 4274
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.509551286371005,
+      "learning_rate": 1.4041934132684116e-06,
+      "loss": 0.1194,
+      "step": 4275
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 7.0171083888829395,
+      "learning_rate": 1.4021858134237892e-06,
+      "loss": 0.1792,
+      "step": 4276
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.766478418536765,
+      "learning_rate": 1.4001794156757598e-06,
+      "loss": 0.1609,
+      "step": 4277
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.7061716612934505,
+      "learning_rate": 1.398174220694699e-06,
+      "loss": 0.1375,
+      "step": 4278
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.22806612837643,
+      "learning_rate": 1.3961702291505791e-06,
+      "loss": 0.1154,
+      "step": 4279
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.054628039524962,
+      "learning_rate": 1.3941674417129714e-06,
+      "loss": 0.1053,
+      "step": 4280
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 9.438939415322599,
+      "learning_rate": 1.3921658590510434e-06,
+      "loss": 0.1853,
+      "step": 4281
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.909877879321542,
+      "learning_rate": 1.3901654818335618e-06,
+      "loss": 0.1347,
+      "step": 4282
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 9.17926909272783,
+      "learning_rate": 1.3881663107288918e-06,
+      "loss": 0.1378,
+      "step": 4283
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.0576033881925015,
+      "learning_rate": 1.386168346404988e-06,
+      "loss": 0.1311,
+      "step": 4284
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 6.618556358799638,
+      "learning_rate": 1.3841715895294138e-06,
+      "loss": 0.0911,
+      "step": 4285
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.8253753457194595,
+      "learning_rate": 1.3821760407693175e-06,
+      "loss": 0.0886,
+      "step": 4286
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 6.792261943559342,
+      "learning_rate": 1.3801817007914543e-06,
+      "loss": 0.1329,
+      "step": 4287
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 10.076624429072998,
+      "learning_rate": 1.3781885702621644e-06,
+      "loss": 0.2051,
+      "step": 4288
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.086198267318938,
+      "learning_rate": 1.3761966498473956e-06,
+      "loss": 0.0788,
+      "step": 4289
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 8.435317943548803,
+      "learning_rate": 1.3742059402126818e-06,
+      "loss": 0.1439,
+      "step": 4290
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.6714105862417075,
+      "learning_rate": 1.3722164420231565e-06,
+      "loss": 0.1266,
+      "step": 4291
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 6.760288551086754,
+      "learning_rate": 1.370228155943548e-06,
+      "loss": 0.132,
+      "step": 4292
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.607704087759127,
+      "learning_rate": 1.3682410826381816e-06,
+      "loss": 0.0842,
+      "step": 4293
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.449194178955735,
+      "learning_rate": 1.366255222770973e-06,
+      "loss": 0.1436,
+      "step": 4294
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 10.9077308370694,
+      "learning_rate": 1.364270577005436e-06,
+      "loss": 0.132,
+      "step": 4295
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 5.9434830126409635,
+      "learning_rate": 1.3622871460046778e-06,
+      "loss": 0.1117,
+      "step": 4296
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 5.335529396281524,
+      "learning_rate": 1.3603049304313992e-06,
+      "loss": 0.1206,
+      "step": 4297
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 6.400241514226316,
+      "learning_rate": 1.3583239309478953e-06,
+      "loss": 0.1349,
+      "step": 4298
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 7.442955154627488,
+      "learning_rate": 1.3563441482160562e-06,
+      "loss": 0.1261,
+      "step": 4299
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 11.286798997787256,
+      "learning_rate": 1.35436558289736e-06,
+      "loss": 0.1791,
+      "step": 4300
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.602769205972933,
+      "learning_rate": 1.3523882356528883e-06,
+      "loss": 0.1063,
+      "step": 4301
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 8.182228061518014,
+      "learning_rate": 1.350412107143303e-06,
+      "loss": 0.1431,
+      "step": 4302
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 5.4663437903465395,
+      "learning_rate": 1.3484371980288712e-06,
+      "loss": 0.1296,
+      "step": 4303
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.621578796410083,
+      "learning_rate": 1.3464635089694416e-06,
+      "loss": 0.0828,
+      "step": 4304
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 5.558582314176478,
+      "learning_rate": 1.344491040624466e-06,
+      "loss": 0.1587,
+      "step": 4305
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 6.181336673297239,
+      "learning_rate": 1.3425197936529766e-06,
+      "loss": 0.1385,
+      "step": 4306
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 10.412134988285041,
+      "learning_rate": 1.3405497687136098e-06,
+      "loss": 0.1656,
+      "step": 4307
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 7.582811581572259,
+      "learning_rate": 1.3385809664645827e-06,
+      "loss": 0.1295,
+      "step": 4308
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 8.150651347785129,
+      "learning_rate": 1.336613387563711e-06,
+      "loss": 0.1396,
+      "step": 4309
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.360245382647963,
+      "learning_rate": 1.3346470326683986e-06,
+      "loss": 0.1394,
+      "step": 4310
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.92768096906367,
+      "learning_rate": 1.3326819024356413e-06,
+      "loss": 0.1395,
+      "step": 4311
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 8.126264782030967,
+      "learning_rate": 1.3307179975220264e-06,
+      "loss": 0.1683,
+      "step": 4312
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 7.097276549928228,
+      "learning_rate": 1.3287553185837298e-06,
+      "loss": 0.1547,
+      "step": 4313
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 6.968139851957762,
+      "learning_rate": 1.3267938662765206e-06,
+      "loss": 0.1147,
+      "step": 4314
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.5357337366094725,
+      "learning_rate": 1.324833641255755e-06,
+      "loss": 0.1165,
+      "step": 4315
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 6.212179238446433,
+      "learning_rate": 1.3228746441763813e-06,
+      "loss": 0.0844,
+      "step": 4316
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 5.800820871659498,
+      "learning_rate": 1.3209168756929363e-06,
+      "loss": 0.1323,
+      "step": 4317
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 8.876523252846415,
+      "learning_rate": 1.3189603364595483e-06,
+      "loss": 0.1721,
+      "step": 4318
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 6.1759123731714825,
+      "learning_rate": 1.3170050271299316e-06,
+      "loss": 0.1298,
+      "step": 4319
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 8.861593791331382,
+      "learning_rate": 1.315050948357392e-06,
+      "loss": 0.1289,
+      "step": 4320
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 8.188259150222102,
+      "learning_rate": 1.3130981007948247e-06,
+      "loss": 0.1427,
+      "step": 4321
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 6.725476274791131,
+      "learning_rate": 1.3111464850947103e-06,
+      "loss": 0.1479,
+      "step": 4322
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.735183629373757,
+      "learning_rate": 1.3091961019091216e-06,
+      "loss": 0.1288,
+      "step": 4323
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.745017577097906,
+      "learning_rate": 1.3072469518897184e-06,
+      "loss": 0.1412,
+      "step": 4324
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 8.405808376306096,
+      "learning_rate": 1.3052990356877444e-06,
+      "loss": 0.1046,
+      "step": 4325
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 8.517500797990728,
+      "learning_rate": 1.3033523539540394e-06,
+      "loss": 0.1934,
+      "step": 4326
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.815632664792931,
+      "learning_rate": 1.3014069073390206e-06,
+      "loss": 0.1445,
+      "step": 4327
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 6.250647236307092,
+      "learning_rate": 1.2994626964927042e-06,
+      "loss": 0.1066,
+      "step": 4328
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.161414714395807,
+      "learning_rate": 1.2975197220646807e-06,
+      "loss": 0.1146,
+      "step": 4329
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 8.547923821989846,
+      "learning_rate": 1.29557798470414e-06,
+      "loss": 0.1351,
+      "step": 4330
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.445406438833862,
+      "learning_rate": 1.293637485059847e-06,
+      "loss": 0.1071,
+      "step": 4331
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 12.910301333668382,
+      "learning_rate": 1.291698223780164e-06,
+      "loss": 0.1795,
+      "step": 4332
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.415868085569591,
+      "learning_rate": 1.2897602015130306e-06,
+      "loss": 0.1145,
+      "step": 4333
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.580899357603486,
+      "learning_rate": 1.287823418905977e-06,
+      "loss": 0.1422,
+      "step": 4334
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.710105285949303,
+      "learning_rate": 1.2858878766061178e-06,
+      "loss": 0.1465,
+      "step": 4335
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.719436922598486,
+      "learning_rate": 1.2839535752601551e-06,
+      "loss": 0.1892,
+      "step": 4336
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 8.01755051568949,
+      "learning_rate": 1.2820205155143738e-06,
+      "loss": 0.1098,
+      "step": 4337
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 5.586621697986031,
+      "learning_rate": 1.2800886980146453e-06,
+      "loss": 0.1371,
+      "step": 4338
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 10.331064240612887,
+      "learning_rate": 1.2781581234064256e-06,
+      "loss": 0.1504,
+      "step": 4339
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 6.770807504782781,
+      "learning_rate": 1.276228792334756e-06,
+      "loss": 0.1121,
+      "step": 4340
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 6.086051023648102,
+      "learning_rate": 1.274300705444262e-06,
+      "loss": 0.1239,
+      "step": 4341
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 8.608147481251128,
+      "learning_rate": 1.2723738633791538e-06,
+      "loss": 0.2008,
+      "step": 4342
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.898067540734754,
+      "learning_rate": 1.2704482667832218e-06,
+      "loss": 0.1062,
+      "step": 4343
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 7.497203187118496,
+      "learning_rate": 1.2685239162998485e-06,
+      "loss": 0.1232,
+      "step": 4344
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 19.126104586121958,
+      "learning_rate": 1.2666008125719904e-06,
+      "loss": 0.1772,
+      "step": 4345
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 8.12284642374614,
+      "learning_rate": 1.2646789562421975e-06,
+      "loss": 0.1339,
+      "step": 4346
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 8.079010398172034,
+      "learning_rate": 1.2627583479525913e-06,
+      "loss": 0.1396,
+      "step": 4347
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.1620164187548765,
+      "learning_rate": 1.2608389883448896e-06,
+      "loss": 0.1212,
+      "step": 4348
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 7.859990849012603,
+      "learning_rate": 1.2589208780603795e-06,
+      "loss": 0.1313,
+      "step": 4349
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 7.106346065393317,
+      "learning_rate": 1.2570040177399435e-06,
+      "loss": 0.1424,
+      "step": 4350
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.06095769329329,
+      "learning_rate": 1.255088408024036e-06,
+      "loss": 0.1101,
+      "step": 4351
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.6470216837778135,
+      "learning_rate": 1.2531740495526989e-06,
+      "loss": 0.1184,
+      "step": 4352
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 10.158476286310746,
+      "learning_rate": 1.2512609429655553e-06,
+      "loss": 0.1348,
+      "step": 4353
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.516757987633382,
+      "learning_rate": 1.249349088901809e-06,
+      "loss": 0.1202,
+      "step": 4354
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.6383151746155065,
+      "learning_rate": 1.247438488000247e-06,
+      "loss": 0.0861,
+      "step": 4355
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.12153217876157,
+      "learning_rate": 1.245529140899236e-06,
+      "loss": 0.163,
+      "step": 4356
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.4984484754078276,
+      "learning_rate": 1.2436210482367245e-06,
+      "loss": 0.0786,
+      "step": 4357
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.595141893853033,
+      "learning_rate": 1.2417142106502418e-06,
+      "loss": 0.0952,
+      "step": 4358
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.398814994664079,
+      "learning_rate": 1.2398086287768969e-06,
+      "loss": 0.1138,
+      "step": 4359
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.727230258964825,
+      "learning_rate": 1.237904303253381e-06,
+      "loss": 0.1183,
+      "step": 4360
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.279257418876605,
+      "learning_rate": 1.236001234715965e-06,
+      "loss": 0.1007,
+      "step": 4361
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.402366707274822,
+      "learning_rate": 1.2340994238004987e-06,
+      "loss": 0.1294,
+      "step": 4362
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.776850831934081,
+      "learning_rate": 1.2321988711424132e-06,
+      "loss": 0.1176,
+      "step": 4363
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.872878484333298,
+      "learning_rate": 1.2302995773767174e-06,
+      "loss": 0.1392,
+      "step": 4364
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 10.12314710750376,
+      "learning_rate": 1.2284015431380015e-06,
+      "loss": 0.1311,
+      "step": 4365
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 8.469770250969509,
+      "learning_rate": 1.2265047690604354e-06,
+      "loss": 0.1562,
+      "step": 4366
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 7.966300776911517,
+      "learning_rate": 1.2246092557777633e-06,
+      "loss": 0.1357,
+      "step": 4367
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 7.398989033867051,
+      "learning_rate": 1.2227150039233132e-06,
+      "loss": 0.157,
+      "step": 4368
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 12.612083517699997,
+      "learning_rate": 1.2208220141299893e-06,
+      "loss": 0.1385,
+      "step": 4369
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.966421949471924,
+      "learning_rate": 1.2189302870302755e-06,
+      "loss": 0.1266,
+      "step": 4370
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.641265236099781,
+      "learning_rate": 1.2170398232562324e-06,
+      "loss": 0.1183,
+      "step": 4371
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 20.375633559108344,
+      "learning_rate": 1.2151506234395e-06,
+      "loss": 0.2162,
+      "step": 4372
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 10.58226033610902,
+      "learning_rate": 1.2132626882112935e-06,
+      "loss": 0.1302,
+      "step": 4373
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 9.180226888360846,
+      "learning_rate": 1.211376018202408e-06,
+      "loss": 0.1749,
+      "step": 4374
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 6.08472849567162,
+      "learning_rate": 1.2094906140432155e-06,
+      "loss": 0.1486,
+      "step": 4375
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.74121178158958,
+      "learning_rate": 1.2076064763636641e-06,
+      "loss": 0.0946,
+      "step": 4376
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 8.286183532915876,
+      "learning_rate": 1.205723605793279e-06,
+      "loss": 0.1323,
+      "step": 4377
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 6.530368951438912,
+      "learning_rate": 1.2038420029611625e-06,
+      "loss": 0.1406,
+      "step": 4378
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 9.283823696986877,
+      "learning_rate": 1.2019616684959934e-06,
+      "loss": 0.1354,
+      "step": 4379
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.768580237770132,
+      "learning_rate": 1.2000826030260254e-06,
+      "loss": 0.1118,
+      "step": 4380
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 7.393517881119877,
+      "learning_rate": 1.1982048071790903e-06,
+      "loss": 0.15,
+      "step": 4381
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 7.074275685681868,
+      "learning_rate": 1.1963282815825938e-06,
+      "loss": 0.135,
+      "step": 4382
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.8588043109415,
+      "learning_rate": 1.194453026863519e-06,
+      "loss": 0.0863,
+      "step": 4383
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 8.252158205024786,
+      "learning_rate": 1.1925790436484219e-06,
+      "loss": 0.1028,
+      "step": 4384
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.4442688489776465,
+      "learning_rate": 1.1907063325634376e-06,
+      "loss": 0.0718,
+      "step": 4385
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 6.31765034734141,
+      "learning_rate": 1.1888348942342697e-06,
+      "loss": 0.1121,
+      "step": 4386
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 8.438393240156737,
+      "learning_rate": 1.1869647292862051e-06,
+      "loss": 0.1555,
+      "step": 4387
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 11.366731513055292,
+      "learning_rate": 1.1850958383440957e-06,
+      "loss": 0.1699,
+      "step": 4388
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.652295308843194,
+      "learning_rate": 1.183228222032378e-06,
+      "loss": 0.1521,
+      "step": 4389
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 6.177913643706088,
+      "learning_rate": 1.181361880975052e-06,
+      "loss": 0.1331,
+      "step": 4390
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.520601566917642,
+      "learning_rate": 1.1794968157957026e-06,
+      "loss": 0.0956,
+      "step": 4391
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.836669023955266,
+      "learning_rate": 1.1776330271174786e-06,
+      "loss": 0.1168,
+      "step": 4392
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 5.603939921185731,
+      "learning_rate": 1.1757705155631072e-06,
+      "loss": 0.1146,
+      "step": 4393
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.98450834534631,
+      "learning_rate": 1.1739092817548887e-06,
+      "loss": 0.1356,
+      "step": 4394
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 5.048629847874484,
+      "learning_rate": 1.172049326314696e-06,
+      "loss": 0.1296,
+      "step": 4395
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.414733074091697,
+      "learning_rate": 1.1701906498639741e-06,
+      "loss": 0.0904,
+      "step": 4396
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.206320184441604,
+      "learning_rate": 1.1683332530237423e-06,
+      "loss": 0.0958,
+      "step": 4397
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 8.848301273909941,
+      "learning_rate": 1.1664771364145905e-06,
+      "loss": 0.182,
+      "step": 4398
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 18.28608439063244,
+      "learning_rate": 1.1646223006566827e-06,
+      "loss": 0.3428,
+      "step": 4399
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 5.770307322322705,
+      "learning_rate": 1.162768746369753e-06,
+      "loss": 0.1066,
+      "step": 4400
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 12.372434701004785,
+      "learning_rate": 1.1609164741731105e-06,
+      "loss": 0.2958,
+      "step": 4401
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 5.22763139963171,
+      "learning_rate": 1.1590654846856291e-06,
+      "loss": 0.1393,
+      "step": 4402
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.248723078530217,
+      "learning_rate": 1.1572157785257643e-06,
+      "loss": 0.0935,
+      "step": 4403
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.694964132694257,
+      "learning_rate": 1.1553673563115325e-06,
+      "loss": 0.1555,
+      "step": 4404
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 8.188557172066655,
+      "learning_rate": 1.153520218660531e-06,
+      "loss": 0.1984,
+      "step": 4405
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 6.541895242637127,
+      "learning_rate": 1.1516743661899172e-06,
+      "loss": 0.1124,
+      "step": 4406
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 10.548845679168716,
+      "learning_rate": 1.1498297995164305e-06,
+      "loss": 0.186,
+      "step": 4407
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 8.565033160611465,
+      "learning_rate": 1.1479865192563683e-06,
+      "loss": 0.1838,
+      "step": 4408
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.752373240136069,
+      "learning_rate": 1.146144526025612e-06,
+      "loss": 0.0936,
+      "step": 4409
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 7.942694309672477,
+      "learning_rate": 1.1443038204396007e-06,
+      "loss": 0.1306,
+      "step": 4410
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 6.89945116790705,
+      "learning_rate": 1.1424644031133502e-06,
+      "loss": 0.1313,
+      "step": 4411
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 9.26264775999036,
+      "learning_rate": 1.1406262746614433e-06,
+      "loss": 0.1857,
+      "step": 4412
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 5.4397623406990405,
+      "learning_rate": 1.1387894356980334e-06,
+      "loss": 0.1163,
+      "step": 4413
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 8.491939243996956,
+      "learning_rate": 1.1369538868368424e-06,
+      "loss": 0.1293,
+      "step": 4414
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 5.750098964300776,
+      "learning_rate": 1.1351196286911615e-06,
+      "loss": 0.1233,
+      "step": 4415
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 6.21545036806548,
+      "learning_rate": 1.1332866618738498e-06,
+      "loss": 0.1461,
+      "step": 4416
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.539810649852684,
+      "learning_rate": 1.1314549869973363e-06,
+      "loss": 0.1044,
+      "step": 4417
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.573431615715713,
+      "learning_rate": 1.1296246046736176e-06,
+      "loss": 0.0757,
+      "step": 4418
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 11.731198672608688,
+      "learning_rate": 1.1277955155142578e-06,
+      "loss": 0.1499,
+      "step": 4419
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 5.893278961301827,
+      "learning_rate": 1.1259677201303905e-06,
+      "loss": 0.1198,
+      "step": 4420
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.0550091247501125,
+      "learning_rate": 1.1241412191327155e-06,
+      "loss": 0.1017,
+      "step": 4421
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.282672874051473,
+      "learning_rate": 1.1223160131315008e-06,
+      "loss": 0.1163,
+      "step": 4422
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 12.271735935923063,
+      "learning_rate": 1.1204921027365818e-06,
+      "loss": 0.1686,
+      "step": 4423
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.60070124272832,
+      "learning_rate": 1.1186694885573602e-06,
+      "loss": 0.1394,
+      "step": 4424
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.119314793118176,
+      "learning_rate": 1.1168481712028061e-06,
+      "loss": 0.1188,
+      "step": 4425
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.845012516399945,
+      "learning_rate": 1.115028151281457e-06,
+      "loss": 0.0951,
+      "step": 4426
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 4.51100338272369,
+      "learning_rate": 1.1132094294014106e-06,
+      "loss": 0.0892,
+      "step": 4427
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 8.698616071989141,
+      "learning_rate": 1.1113920061703416e-06,
+      "loss": 0.1289,
+      "step": 4428
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.597902292498433,
+      "learning_rate": 1.1095758821954788e-06,
+      "loss": 0.1199,
+      "step": 4429
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 7.13322096070308,
+      "learning_rate": 1.107761058083629e-06,
+      "loss": 0.1552,
+      "step": 4430
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.3313785556010505,
+      "learning_rate": 1.1059475344411535e-06,
+      "loss": 0.0945,
+      "step": 4431
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 8.065704209387357,
+      "learning_rate": 1.104135311873989e-06,
+      "loss": 0.1471,
+      "step": 4432
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 7.85943646738916,
+      "learning_rate": 1.1023243909876275e-06,
+      "loss": 0.1545,
+      "step": 4433
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 5.066634102847706,
+      "learning_rate": 1.1005147723871374e-06,
+      "loss": 0.0896,
+      "step": 4434
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.8424079666970155,
+      "learning_rate": 1.0987064566771405e-06,
+      "loss": 0.1359,
+      "step": 4435
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.6459659701311695,
+      "learning_rate": 1.0968994444618313e-06,
+      "loss": 0.167,
+      "step": 4436
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 5.500227738741658,
+      "learning_rate": 1.0950937363449659e-06,
+      "loss": 0.1115,
+      "step": 4437
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.2102276882225445,
+      "learning_rate": 1.0932893329298643e-06,
+      "loss": 0.1222,
+      "step": 4438
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.847799303012751,
+      "learning_rate": 1.0914862348194121e-06,
+      "loss": 0.1068,
+      "step": 4439
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 6.851543694579042,
+      "learning_rate": 1.0896844426160575e-06,
+      "loss": 0.1298,
+      "step": 4440
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 6.889551647224606,
+      "learning_rate": 1.0878839569218124e-06,
+      "loss": 0.116,
+      "step": 4441
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.152322142099854,
+      "learning_rate": 1.0860847783382534e-06,
+      "loss": 0.1624,
+      "step": 4442
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.011318397121267,
+      "learning_rate": 1.0842869074665186e-06,
+      "loss": 0.1464,
+      "step": 4443
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 8.68536297533047,
+      "learning_rate": 1.0824903449073115e-06,
+      "loss": 0.2058,
+      "step": 4444
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 8.256267124703266,
+      "learning_rate": 1.0806950912608937e-06,
+      "loss": 0.1498,
+      "step": 4445
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 9.177033572690247,
+      "learning_rate": 1.0789011471270983e-06,
+      "loss": 0.1533,
+      "step": 4446
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 9.42342228155218,
+      "learning_rate": 1.0771085131053087e-06,
+      "loss": 0.1993,
+      "step": 4447
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 5.096230095880606,
+      "learning_rate": 1.0753171897944835e-06,
+      "loss": 0.1036,
+      "step": 4448
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 8.154504511502616,
+      "learning_rate": 1.0735271777931322e-06,
+      "loss": 0.1755,
+      "step": 4449
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 7.04583975692834,
+      "learning_rate": 1.0717384776993356e-06,
+      "loss": 0.2039,
+      "step": 4450
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 6.7824108907953224,
+      "learning_rate": 1.069951090110728e-06,
+      "loss": 0.1451,
+      "step": 4451
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 5.187754014764868,
+      "learning_rate": 1.06816501562451e-06,
+      "loss": 0.0834,
+      "step": 4452
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 7.278261677229582,
+      "learning_rate": 1.0663802548374424e-06,
+      "loss": 0.1149,
+      "step": 4453
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 4.340393440241648,
+      "learning_rate": 1.064596808345847e-06,
+      "loss": 0.1205,
+      "step": 4454
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 7.746103373611341,
+      "learning_rate": 1.0628146767456066e-06,
+      "loss": 0.166,
+      "step": 4455
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 6.3971573426174295,
+      "learning_rate": 1.061033860632164e-06,
+      "loss": 0.1436,
+      "step": 4456
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 9.242403232576306,
+      "learning_rate": 1.0592543606005235e-06,
+      "loss": 0.181,
+      "step": 4457
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 10.503336807115314,
+      "learning_rate": 1.0574761772452486e-06,
+      "loss": 0.2029,
+      "step": 4458
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 7.744097047628762,
+      "learning_rate": 1.0556993111604635e-06,
+      "loss": 0.1935,
+      "step": 4459
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 9.014041823353091,
+      "learning_rate": 1.0539237629398536e-06,
+      "loss": 0.1501,
+      "step": 4460
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 8.4496955488824,
+      "learning_rate": 1.052149533176659e-06,
+      "loss": 0.1421,
+      "step": 4461
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 6.407043789529695,
+      "learning_rate": 1.050376622463688e-06,
+      "loss": 0.1115,
+      "step": 4462
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 5.934429129123148,
+      "learning_rate": 1.0486050313932972e-06,
+      "loss": 0.1068,
+      "step": 4463
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.89642147430061,
+      "learning_rate": 1.0468347605574137e-06,
+      "loss": 0.0684,
+      "step": 4464
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 6.884801342876653,
+      "learning_rate": 1.0450658105475126e-06,
+      "loss": 0.1141,
+      "step": 4465
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 5.913543630708609,
+      "learning_rate": 1.0432981819546384e-06,
+      "loss": 0.1215,
+      "step": 4466
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 7.435283990174136,
+      "learning_rate": 1.0415318753693837e-06,
+      "loss": 0.1531,
+      "step": 4467
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 9.051636352276448,
+      "learning_rate": 1.0397668913819086e-06,
+      "loss": 0.1371,
+      "step": 4468
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 6.3128900776674755,
+      "learning_rate": 1.0380032305819243e-06,
+      "loss": 0.1064,
+      "step": 4469
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 8.142204214879126,
+      "learning_rate": 1.0362408935587026e-06,
+      "loss": 0.1516,
+      "step": 4470
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 7.661871841618407,
+      "learning_rate": 1.0344798809010748e-06,
+      "loss": 0.1476,
+      "step": 4471
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 11.682682233942165,
+      "learning_rate": 1.0327201931974262e-06,
+      "loss": 0.188,
+      "step": 4472
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.381237196793194,
+      "learning_rate": 1.0309618310357023e-06,
+      "loss": 0.1077,
+      "step": 4473
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 7.058241319504342,
+      "learning_rate": 1.0292047950034046e-06,
+      "loss": 0.1934,
+      "step": 4474
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.707890326892382,
+      "learning_rate": 1.0274490856875908e-06,
+      "loss": 0.0759,
+      "step": 4475
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 6.573092415383323,
+      "learning_rate": 1.0256947036748766e-06,
+      "loss": 0.1527,
+      "step": 4476
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.479401416008466,
+      "learning_rate": 1.0239416495514331e-06,
+      "loss": 0.124,
+      "step": 4477
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.404838785894077,
+      "learning_rate": 1.0221899239029887e-06,
+      "loss": 0.0833,
+      "step": 4478
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.3591410165394775,
+      "learning_rate": 1.0204395273148277e-06,
+      "loss": 0.0904,
+      "step": 4479
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 8.06517124521747,
+      "learning_rate": 1.0186904603717894e-06,
+      "loss": 0.1272,
+      "step": 4480
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 8.295034878534254,
+      "learning_rate": 1.0169427236582702e-06,
+      "loss": 0.1395,
+      "step": 4481
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.6017148138313075,
+      "learning_rate": 1.0151963177582208e-06,
+      "loss": 0.0892,
+      "step": 4482
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.168505372280764,
+      "learning_rate": 1.0134512432551492e-06,
+      "loss": 0.0732,
+      "step": 4483
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.844662307023592,
+      "learning_rate": 1.0117075007321152e-06,
+      "loss": 0.0717,
+      "step": 4484
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.47237204481135,
+      "learning_rate": 1.009965090771739e-06,
+      "loss": 0.1017,
+      "step": 4485
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 8.845533470002579,
+      "learning_rate": 1.0082240139561866e-06,
+      "loss": 0.1592,
+      "step": 4486
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 8.253239344807614,
+      "learning_rate": 1.0064842708671908e-06,
+      "loss": 0.141,
+      "step": 4487
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 6.3375223471928095,
+      "learning_rate": 1.0047458620860251e-06,
+      "loss": 0.1008,
+      "step": 4488
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.7321531585346435,
+      "learning_rate": 1.0030087881935308e-06,
+      "loss": 0.0897,
+      "step": 4489
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 7.816882454233154,
+      "learning_rate": 1.0012730497700912e-06,
+      "loss": 0.1459,
+      "step": 4490
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 11.259126205309581,
+      "learning_rate": 9.995386473956531e-07,
+      "loss": 0.1963,
+      "step": 4491
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 4.533563851322039,
+      "learning_rate": 9.978055816497084e-07,
+      "loss": 0.0829,
+      "step": 4492
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 8.880117960445354,
+      "learning_rate": 9.960738531113118e-07,
+      "loss": 0.1495,
+      "step": 4493
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 4.388909654201851,
+      "learning_rate": 9.94343462359061e-07,
+      "loss": 0.1064,
+      "step": 4494
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 5.589157030071463,
+      "learning_rate": 9.926144099711138e-07,
+      "loss": 0.1085,
+      "step": 4495
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 6.486877158713429,
+      "learning_rate": 9.90886696525179e-07,
+      "loss": 0.1301,
+      "step": 4496
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 9.916504210295873,
+      "learning_rate": 9.89160322598517e-07,
+      "loss": 0.1409,
+      "step": 4497
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.150060234978749,
+      "learning_rate": 9.874352887679416e-07,
+      "loss": 0.1315,
+      "step": 4498
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.447202722521027,
+      "learning_rate": 9.857115956098196e-07,
+      "loss": 0.1461,
+      "step": 4499
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.592805113175272,
+      "learning_rate": 9.839892437000675e-07,
+      "loss": 0.0893,
+      "step": 4500
+    },
+    {
+      "epoch": 3.21,
+      "eval_avg_AUC": 0.7918247035704028,
+      "eval_avg_Accuracy": 0.711414124668435,
+      "eval_avg_Accuracy-right": 0.8560062605973653,
+      "eval_avg_Accuracy-wrong": 0.45929042528997044,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6577722838853147,
+      "eval_last_AUC": 0.8121226339774985,
+      "eval_last_Accuracy": 0.7365716180371353,
+      "eval_last_Accuracy-right": 0.7942480761706012,
+      "eval_last_Accuracy-wrong": 0.6360018194223334,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6811064518145272,
+      "eval_max_AUC": 0.766090631206006,
+      "eval_max_Accuracy": 0.6466760610079576,
+      "eval_max_Accuracy-right": 0.9701317334028955,
+      "eval_max_Accuracy-wrong": 0.08267000227427791,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6136546806309608,
+      "eval_min_AUC": 0.7982965177859538,
+      "eval_min_Accuracy": 0.7182940981432361,
+      "eval_min_Accuracy-right": 0.6972088170079562,
+      "eval_min_Accuracy-wrong": 0.7550602683647942,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6616380620169989,
+      "eval_prod_AUC": 0.7992793561684441,
+      "eval_prod_Accuracy": 0.7075596816976127,
+      "eval_prod_Accuracy-right": 0.6415155862788574,
+      "eval_prod_Accuracy-wrong": 0.8227200363884467,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6610387644238402,
+      "eval_runtime": 246.5009,
+      "eval_samples_per_second": 97.882,
+      "eval_steps_per_second": 3.059,
+      "eval_sum_AUC": 0.6820948876089348,
+      "eval_sum_Accuracy": 0.6409565649867374,
+      "eval_sum_Accuracy-right": 0.9834355028042259,
+      "eval_sum_Accuracy-wrong": 0.04377984989765749,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6492510686816373,
+      "step": 4500
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 6.017970370822423,
+      "learning_rate": 9.822682336141558e-07,
+      "loss": 0.1088,
+      "step": 4501
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 8.347395956475857,
+      "learning_rate": 9.805485659271064e-07,
+      "loss": 0.1423,
+      "step": 4502
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 6.147141262858037,
+      "learning_rate": 9.788302412134931e-07,
+      "loss": 0.1143,
+      "step": 4503
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 5.992507716204142,
+      "learning_rate": 9.77113260047436e-07,
+      "loss": 0.1105,
+      "step": 4504
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.896232384462128,
+      "learning_rate": 9.753976230026158e-07,
+      "loss": 0.1024,
+      "step": 4505
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.730396170031993,
+      "learning_rate": 9.736833306522537e-07,
+      "loss": 0.1816,
+      "step": 4506
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 9.87791454318401,
+      "learning_rate": 9.719703835691314e-07,
+      "loss": 0.1908,
+      "step": 4507
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 9.305401921388526,
+      "learning_rate": 9.702587823255715e-07,
+      "loss": 0.1743,
+      "step": 4508
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 12.725995171162081,
+      "learning_rate": 9.685485274934576e-07,
+      "loss": 0.2882,
+      "step": 4509
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 9.06907116647378,
+      "learning_rate": 9.66839619644211e-07,
+      "loss": 0.1427,
+      "step": 4510
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.154694219463353,
+      "learning_rate": 9.651320593488162e-07,
+      "loss": 0.1312,
+      "step": 4511
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.18461317267077,
+      "learning_rate": 9.634258471777958e-07,
+      "loss": 0.1302,
+      "step": 4512
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 8.304539402084123,
+      "learning_rate": 9.617209837012287e-07,
+      "loss": 0.129,
+      "step": 4513
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.210423547955145,
+      "learning_rate": 9.600174694887421e-07,
+      "loss": 0.1196,
+      "step": 4514
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 6.6024670811500705,
+      "learning_rate": 9.583153051095107e-07,
+      "loss": 0.113,
+      "step": 4515
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.713142607529561,
+      "learning_rate": 9.5661449113226e-07,
+      "loss": 0.1359,
+      "step": 4516
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 6.483198577305126,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.1104,
+      "step": 4517
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.5373377619248,
+      "learning_rate": 9.532169166563426e-07,
+      "loss": 0.1328,
+      "step": 4518
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.18316197627452,
+      "learning_rate": 9.515201572928689e-07,
+      "loss": 0.0995,
+      "step": 4519
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 8.12338317679702,
+      "learning_rate": 9.49824750601761e-07,
+      "loss": 0.1083,
+      "step": 4520
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 7.602592282544346,
+      "learning_rate": 9.481306971494858e-07,
+      "loss": 0.1244,
+      "step": 4521
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.777997231167417,
+      "learning_rate": 9.464379975020576e-07,
+      "loss": 0.1606,
+      "step": 4522
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.124881227801997,
+      "learning_rate": 9.447466522250393e-07,
+      "loss": 0.1326,
+      "step": 4523
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.658499339226998,
+      "learning_rate": 9.430566618835407e-07,
+      "loss": 0.1031,
+      "step": 4524
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 8.542957014977226,
+      "learning_rate": 9.413680270422187e-07,
+      "loss": 0.1439,
+      "step": 4525
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 8.15924583194008,
+      "learning_rate": 9.396807482652775e-07,
+      "loss": 0.1235,
+      "step": 4526
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.153176455458262,
+      "learning_rate": 9.3799482611647e-07,
+      "loss": 0.14,
+      "step": 4527
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.857510589303771,
+      "learning_rate": 9.363102611590918e-07,
+      "loss": 0.1151,
+      "step": 4528
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.841608683622035,
+      "learning_rate": 9.346270539559882e-07,
+      "loss": 0.1255,
+      "step": 4529
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.349487702065171,
+      "learning_rate": 9.329452050695497e-07,
+      "loss": 0.1427,
+      "step": 4530
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.185775397201693,
+      "learning_rate": 9.312647150617144e-07,
+      "loss": 0.1322,
+      "step": 4531
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 5.993422226179697,
+      "learning_rate": 9.295855844939639e-07,
+      "loss": 0.1133,
+      "step": 4532
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 6.877262751498541,
+      "learning_rate": 9.279078139273279e-07,
+      "loss": 0.1621,
+      "step": 4533
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 10.9101913835771,
+      "learning_rate": 9.262314039223802e-07,
+      "loss": 0.1537,
+      "step": 4534
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 8.089060660483195,
+      "learning_rate": 9.245563550392406e-07,
+      "loss": 0.1165,
+      "step": 4535
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.9168601004208625,
+      "learning_rate": 9.22882667837574e-07,
+      "loss": 0.1515,
+      "step": 4536
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 3.366227246368016,
+      "learning_rate": 9.212103428765912e-07,
+      "loss": 0.0607,
+      "step": 4537
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 6.537651822483652,
+      "learning_rate": 9.19539380715046e-07,
+      "loss": 0.1354,
+      "step": 4538
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.770155298405159,
+      "learning_rate": 9.178697819112381e-07,
+      "loss": 0.1273,
+      "step": 4539
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 9.281594784034297,
+      "learning_rate": 9.162015470230123e-07,
+      "loss": 0.1356,
+      "step": 4540
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 8.028631915325574,
+      "learning_rate": 9.145346766077562e-07,
+      "loss": 0.1834,
+      "step": 4541
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 4.045279042279176,
+      "learning_rate": 9.128691712224025e-07,
+      "loss": 0.1058,
+      "step": 4542
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 6.339980352393081,
+      "learning_rate": 9.112050314234272e-07,
+      "loss": 0.1326,
+      "step": 4543
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 3.1982368923211744,
+      "learning_rate": 9.0954225776685e-07,
+      "loss": 0.0649,
+      "step": 4544
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 5.253249493987398,
+      "learning_rate": 9.078808508082354e-07,
+      "loss": 0.0817,
+      "step": 4545
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 7.390707096657996,
+      "learning_rate": 9.06220811102691e-07,
+      "loss": 0.1368,
+      "step": 4546
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 15.032324120582194,
+      "learning_rate": 9.045621392048637e-07,
+      "loss": 0.2621,
+      "step": 4547
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 8.212400030385552,
+      "learning_rate": 9.029048356689507e-07,
+      "loss": 0.1085,
+      "step": 4548
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.560047001509493,
+      "learning_rate": 9.012489010486835e-07,
+      "loss": 0.1552,
+      "step": 4549
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 9.24670417792784,
+      "learning_rate": 8.995943358973463e-07,
+      "loss": 0.134,
+      "step": 4550
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 8.061594801279545,
+      "learning_rate": 8.979411407677535e-07,
+      "loss": 0.1624,
+      "step": 4551
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.841536214499788,
+      "learning_rate": 8.962893162122749e-07,
+      "loss": 0.1432,
+      "step": 4552
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.80332838186666,
+      "learning_rate": 8.946388627828106e-07,
+      "loss": 0.1351,
+      "step": 4553
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.577103709697863,
+      "learning_rate": 8.929897810308102e-07,
+      "loss": 0.1639,
+      "step": 4554
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 6.890676018724506,
+      "learning_rate": 8.913420715072619e-07,
+      "loss": 0.1366,
+      "step": 4555
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.615943828764504,
+      "learning_rate": 8.896957347626966e-07,
+      "loss": 0.1583,
+      "step": 4556
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.849886192759192,
+      "learning_rate": 8.880507713471853e-07,
+      "loss": 0.1052,
+      "step": 4557
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 6.021066135784146,
+      "learning_rate": 8.864071818103415e-07,
+      "loss": 0.1222,
+      "step": 4558
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.485990965887972,
+      "learning_rate": 8.847649667013187e-07,
+      "loss": 0.12,
+      "step": 4559
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.798179875468671,
+      "learning_rate": 8.831241265688112e-07,
+      "loss": 0.145,
+      "step": 4560
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.884685397021309,
+      "learning_rate": 8.814846619610545e-07,
+      "loss": 0.1251,
+      "step": 4561
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 9.08712853808379,
+      "learning_rate": 8.79846573425826e-07,
+      "loss": 0.1326,
+      "step": 4562
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.151088191408366,
+      "learning_rate": 8.782098615104373e-07,
+      "loss": 0.1555,
+      "step": 4563
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.585973402310389,
+      "learning_rate": 8.765745267617487e-07,
+      "loss": 0.1342,
+      "step": 4564
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.157889175348612,
+      "learning_rate": 8.749405697261515e-07,
+      "loss": 0.1348,
+      "step": 4565
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.384538264495767,
+      "learning_rate": 8.733079909495868e-07,
+      "loss": 0.1028,
+      "step": 4566
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.53422431942535,
+      "learning_rate": 8.716767909775231e-07,
+      "loss": 0.1035,
+      "step": 4567
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 5.424476627599804,
+      "learning_rate": 8.700469703549802e-07,
+      "loss": 0.0785,
+      "step": 4568
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 8.037680257428233,
+      "learning_rate": 8.684185296265074e-07,
+      "loss": 0.1572,
+      "step": 4569
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.581140538771417,
+      "learning_rate": 8.667914693362006e-07,
+      "loss": 0.1245,
+      "step": 4570
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.29417094424049,
+      "learning_rate": 8.651657900276878e-07,
+      "loss": 0.1288,
+      "step": 4571
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 5.612495200486463,
+      "learning_rate": 8.635414922441398e-07,
+      "loss": 0.1091,
+      "step": 4572
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 5.7330667780604205,
+      "learning_rate": 8.61918576528265e-07,
+      "loss": 0.147,
+      "step": 4573
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.977158518523487,
+      "learning_rate": 8.60297043422309e-07,
+      "loss": 0.1433,
+      "step": 4574
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 9.135644940015077,
+      "learning_rate": 8.586768934680572e-07,
+      "loss": 0.1904,
+      "step": 4575
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.314051493515507,
+      "learning_rate": 8.570581272068307e-07,
+      "loss": 0.0915,
+      "step": 4576
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.336134974937514,
+      "learning_rate": 8.554407451794905e-07,
+      "loss": 0.0985,
+      "step": 4577
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 7.694521064289034,
+      "learning_rate": 8.538247479264327e-07,
+      "loss": 0.1609,
+      "step": 4578
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 7.528554460208515,
+      "learning_rate": 8.522101359875934e-07,
+      "loss": 0.1183,
+      "step": 4579
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 9.875424450214489,
+      "learning_rate": 8.505969099024436e-07,
+      "loss": 0.1346,
+      "step": 4580
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.967351372296031,
+      "learning_rate": 8.489850702099922e-07,
+      "loss": 0.1478,
+      "step": 4581
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.9745850342812785,
+      "learning_rate": 8.473746174487846e-07,
+      "loss": 0.1324,
+      "step": 4582
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 9.59845328096796,
+      "learning_rate": 8.457655521569036e-07,
+      "loss": 0.1688,
+      "step": 4583
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.193901791217303,
+      "learning_rate": 8.441578748719676e-07,
+      "loss": 0.1298,
+      "step": 4584
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.411916620325072,
+      "learning_rate": 8.425515861311312e-07,
+      "loss": 0.1118,
+      "step": 4585
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 11.622734989607526,
+      "learning_rate": 8.409466864710858e-07,
+      "loss": 0.1835,
+      "step": 4586
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 7.586682013277674,
+      "learning_rate": 8.393431764280591e-07,
+      "loss": 0.123,
+      "step": 4587
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 4.8412719082832485,
+      "learning_rate": 8.377410565378097e-07,
+      "loss": 0.1078,
+      "step": 4588
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 8.79076733178493,
+      "learning_rate": 8.361403273356411e-07,
+      "loss": 0.1606,
+      "step": 4589
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 6.946585563110569,
+      "learning_rate": 8.345409893563816e-07,
+      "loss": 0.1257,
+      "step": 4590
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 11.064964839267628,
+      "learning_rate": 8.329430431344043e-07,
+      "loss": 0.174,
+      "step": 4591
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 6.590378379957648,
+      "learning_rate": 8.313464892036083e-07,
+      "loss": 0.0907,
+      "step": 4592
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.76238767588108,
+      "learning_rate": 8.297513280974362e-07,
+      "loss": 0.1501,
+      "step": 4593
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 6.591967597672873,
+      "learning_rate": 8.281575603488573e-07,
+      "loss": 0.1106,
+      "step": 4594
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.372299842456041,
+      "learning_rate": 8.265651864903823e-07,
+      "loss": 0.1545,
+      "step": 4595
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.960406328039799,
+      "learning_rate": 8.249742070540506e-07,
+      "loss": 0.1243,
+      "step": 4596
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 10.592265053976755,
+      "learning_rate": 8.233846225714386e-07,
+      "loss": 0.1826,
+      "step": 4597
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 8.888448431893492,
+      "learning_rate": 8.217964335736556e-07,
+      "loss": 0.1232,
+      "step": 4598
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.383633290511777,
+      "learning_rate": 8.202096405913462e-07,
+      "loss": 0.0842,
+      "step": 4599
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.216804849791828,
+      "learning_rate": 8.186242441546866e-07,
+      "loss": 0.0851,
+      "step": 4600
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 10.806207876507727,
+      "learning_rate": 8.170402447933873e-07,
+      "loss": 0.1648,
+      "step": 4601
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 6.586285027658275,
+      "learning_rate": 8.154576430366922e-07,
+      "loss": 0.0989,
+      "step": 4602
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 7.464315850177548,
+      "learning_rate": 8.13876439413378e-07,
+      "loss": 0.1607,
+      "step": 4603
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 11.060642250176604,
+      "learning_rate": 8.122966344517536e-07,
+      "loss": 0.1982,
+      "step": 4604
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 5.741654866635015,
+      "learning_rate": 8.107182286796633e-07,
+      "loss": 0.0873,
+      "step": 4605
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 7.245045678499837,
+      "learning_rate": 8.091412226244771e-07,
+      "loss": 0.1475,
+      "step": 4606
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 6.950428201411654,
+      "learning_rate": 8.07565616813108e-07,
+      "loss": 0.145,
+      "step": 4607
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 10.063829105184517,
+      "learning_rate": 8.059914117719897e-07,
+      "loss": 0.1516,
+      "step": 4608
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 7.472916859631381,
+      "learning_rate": 8.044186080270983e-07,
+      "loss": 0.1149,
+      "step": 4609
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 6.777826055951287,
+      "learning_rate": 8.028472061039322e-07,
+      "loss": 0.1243,
+      "step": 4610
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 10.824941387370806,
+      "learning_rate": 8.012772065275304e-07,
+      "loss": 0.1042,
+      "step": 4611
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 9.813588704435642,
+      "learning_rate": 7.997086098224555e-07,
+      "loss": 0.1234,
+      "step": 4612
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 9.105923319287843,
+      "learning_rate": 7.981414165128065e-07,
+      "loss": 0.1295,
+      "step": 4613
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 6.18452983325367,
+      "learning_rate": 7.965756271222108e-07,
+      "loss": 0.1241,
+      "step": 4614
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 5.555813455964336,
+      "learning_rate": 7.950112421738282e-07,
+      "loss": 0.101,
+      "step": 4615
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 7.138099022811065,
+      "learning_rate": 7.934482621903494e-07,
+      "loss": 0.153,
+      "step": 4616
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.029921247555546,
+      "learning_rate": 7.91886687693994e-07,
+      "loss": 0.103,
+      "step": 4617
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.58503608433331,
+      "learning_rate": 7.903265192065141e-07,
+      "loss": 0.0841,
+      "step": 4618
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.5716326952600115,
+      "learning_rate": 7.887677572491903e-07,
+      "loss": 0.1713,
+      "step": 4619
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 6.81055005735616,
+      "learning_rate": 7.872104023428339e-07,
+      "loss": 0.1244,
+      "step": 4620
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 4.568296952845966,
+      "learning_rate": 7.856544550077883e-07,
+      "loss": 0.088,
+      "step": 4621
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 6.472127459738777,
+      "learning_rate": 7.840999157639195e-07,
+      "loss": 0.1244,
+      "step": 4622
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 20.818327230629205,
+      "learning_rate": 7.825467851306335e-07,
+      "loss": 0.1603,
+      "step": 4623
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.986882546336097,
+      "learning_rate": 7.809950636268554e-07,
+      "loss": 0.2106,
+      "step": 4624
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.765351049579674,
+      "learning_rate": 7.794447517710485e-07,
+      "loss": 0.1481,
+      "step": 4625
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 13.522443356939034,
+      "learning_rate": 7.778958500811961e-07,
+      "loss": 0.1749,
+      "step": 4626
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.920109695380363,
+      "learning_rate": 7.7634835907482e-07,
+      "loss": 0.1223,
+      "step": 4627
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.115723186086778,
+      "learning_rate": 7.748022792689613e-07,
+      "loss": 0.0972,
+      "step": 4628
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 4.771496666945104,
+      "learning_rate": 7.732576111801982e-07,
+      "loss": 0.0928,
+      "step": 4629
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 11.926008165260582,
+      "learning_rate": 7.717143553246298e-07,
+      "loss": 0.1371,
+      "step": 4630
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 7.99102315789066,
+      "learning_rate": 7.701725122178871e-07,
+      "loss": 0.1481,
+      "step": 4631
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 12.122142585290227,
+      "learning_rate": 7.686320823751298e-07,
+      "loss": 0.146,
+      "step": 4632
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 8.408071273100916,
+      "learning_rate": 7.670930663110426e-07,
+      "loss": 0.1442,
+      "step": 4633
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.679326638442814,
+      "learning_rate": 7.655554645398405e-07,
+      "loss": 0.1382,
+      "step": 4634
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 25.188966590030155,
+      "learning_rate": 7.640192775752647e-07,
+      "loss": 0.129,
+      "step": 4635
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 7.961221286961985,
+      "learning_rate": 7.624845059305836e-07,
+      "loss": 0.1493,
+      "step": 4636
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 9.194206304233527,
+      "learning_rate": 7.609511501185929e-07,
+      "loss": 0.139,
+      "step": 4637
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.073454340195965,
+      "learning_rate": 7.594192106516151e-07,
+      "loss": 0.1255,
+      "step": 4638
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.415599395063826,
+      "learning_rate": 7.578886880414999e-07,
+      "loss": 0.1212,
+      "step": 4639
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 5.17906997606143,
+      "learning_rate": 7.563595827996235e-07,
+      "loss": 0.1142,
+      "step": 4640
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 9.769899102967289,
+      "learning_rate": 7.548318954368883e-07,
+      "loss": 0.1178,
+      "step": 4641
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.54723718176854,
+      "learning_rate": 7.533056264637228e-07,
+      "loss": 0.1173,
+      "step": 4642
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 5.078416054162937,
+      "learning_rate": 7.51780776390082e-07,
+      "loss": 0.1028,
+      "step": 4643
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 10.329231158209321,
+      "learning_rate": 7.50257345725447e-07,
+      "loss": 0.1575,
+      "step": 4644
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.934461236117057,
+      "learning_rate": 7.487353349788234e-07,
+      "loss": 0.1179,
+      "step": 4645
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 11.00570511198538,
+      "learning_rate": 7.472147446587452e-07,
+      "loss": 0.1691,
+      "step": 4646
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 6.461797158871515,
+      "learning_rate": 7.456955752732659e-07,
+      "loss": 0.1371,
+      "step": 4647
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.756284807242618,
+      "learning_rate": 7.441778273299738e-07,
+      "loss": 0.0889,
+      "step": 4648
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.33097177974351,
+      "learning_rate": 7.426615013359706e-07,
+      "loss": 0.1659,
+      "step": 4649
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.974735983581375,
+      "learning_rate": 7.411465977978949e-07,
+      "loss": 0.1,
+      "step": 4650
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 11.534506781082133,
+      "learning_rate": 7.396331172218996e-07,
+      "loss": 0.1552,
+      "step": 4651
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.406433437663972,
+      "learning_rate": 7.381210601136702e-07,
+      "loss": 0.1144,
+      "step": 4652
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.408309239317868,
+      "learning_rate": 7.366104269784086e-07,
+      "loss": 0.1577,
+      "step": 4653
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 11.29949891157189,
+      "learning_rate": 7.351012183208511e-07,
+      "loss": 0.2051,
+      "step": 4654
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.593810512608232,
+      "learning_rate": 7.335934346452484e-07,
+      "loss": 0.1118,
+      "step": 4655
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.302955662252433,
+      "learning_rate": 7.320870764553795e-07,
+      "loss": 0.1782,
+      "step": 4656
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 8.661085471989326,
+      "learning_rate": 7.305821442545474e-07,
+      "loss": 0.1105,
+      "step": 4657
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 6.860609598218582,
+      "learning_rate": 7.290786385455778e-07,
+      "loss": 0.1268,
+      "step": 4658
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 9.255949888842219,
+      "learning_rate": 7.275765598308199e-07,
+      "loss": 0.1941,
+      "step": 4659
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 8.14943801715935,
+      "learning_rate": 7.26075908612146e-07,
+      "loss": 0.1467,
+      "step": 4660
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 8.30112212336164,
+      "learning_rate": 7.245766853909519e-07,
+      "loss": 0.1393,
+      "step": 4661
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 6.7267328951619785,
+      "learning_rate": 7.230788906681558e-07,
+      "loss": 0.1006,
+      "step": 4662
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 9.21055385244117,
+      "learning_rate": 7.215825249441982e-07,
+      "loss": 0.1885,
+      "step": 4663
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 12.814543386192565,
+      "learning_rate": 7.200875887190445e-07,
+      "loss": 0.1476,
+      "step": 4664
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 9.004947303321043,
+      "learning_rate": 7.185940824921772e-07,
+      "loss": 0.1779,
+      "step": 4665
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 7.554539232706164,
+      "learning_rate": 7.171020067626089e-07,
+      "loss": 0.1388,
+      "step": 4666
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 8.517269227462895,
+      "learning_rate": 7.156113620288646e-07,
+      "loss": 0.2025,
+      "step": 4667
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 6.16280886352939,
+      "learning_rate": 7.141221487890027e-07,
+      "loss": 0.1512,
+      "step": 4668
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 6.039791042262847,
+      "learning_rate": 7.126343675405905e-07,
+      "loss": 0.084,
+      "step": 4669
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 7.185117749237396,
+      "learning_rate": 7.111480187807296e-07,
+      "loss": 0.1752,
+      "step": 4670
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 4.745690408307122,
+      "learning_rate": 7.096631030060308e-07,
+      "loss": 0.0933,
+      "step": 4671
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 3.9267132250498196,
+      "learning_rate": 7.081796207126373e-07,
+      "loss": 0.0792,
+      "step": 4672
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.914207317636215,
+      "learning_rate": 7.06697572396205e-07,
+      "loss": 0.0667,
+      "step": 4673
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.048328757763309,
+      "learning_rate": 7.052169585519142e-07,
+      "loss": 0.0873,
+      "step": 4674
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.948918663582835,
+      "learning_rate": 7.037377796744666e-07,
+      "loss": 0.1484,
+      "step": 4675
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.111718013258758,
+      "learning_rate": 7.022600362580817e-07,
+      "loss": 0.1198,
+      "step": 4676
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.507791776880594,
+      "learning_rate": 7.007837287965024e-07,
+      "loss": 0.1039,
+      "step": 4677
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 7.7886653073378005,
+      "learning_rate": 6.993088577829904e-07,
+      "loss": 0.1141,
+      "step": 4678
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 11.19723147804638,
+      "learning_rate": 6.978354237103264e-07,
+      "loss": 0.1788,
+      "step": 4679
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.665500446115932,
+      "learning_rate": 6.963634270708137e-07,
+      "loss": 0.1144,
+      "step": 4680
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.1211146195001245,
+      "learning_rate": 6.948928683562722e-07,
+      "loss": 0.0544,
+      "step": 4681
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.297345133083735,
+      "learning_rate": 6.934237480580435e-07,
+      "loss": 0.0996,
+      "step": 4682
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.441164589598402,
+      "learning_rate": 6.919560666669889e-07,
+      "loss": 0.1211,
+      "step": 4683
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 7.607568415682107,
+      "learning_rate": 6.904898246734864e-07,
+      "loss": 0.1151,
+      "step": 4684
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 7.063782633192273,
+      "learning_rate": 6.890250225674361e-07,
+      "loss": 0.134,
+      "step": 4685
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 7.207873640293048,
+      "learning_rate": 6.875616608382562e-07,
+      "loss": 0.153,
+      "step": 4686
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.1416823543894346,
+      "learning_rate": 6.860997399748792e-07,
+      "loss": 0.1389,
+      "step": 4687
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 4.1354245919128685,
+      "learning_rate": 6.846392604657653e-07,
+      "loss": 0.0589,
+      "step": 4688
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.418657085737841,
+      "learning_rate": 6.831802227988843e-07,
+      "loss": 0.0791,
+      "step": 4689
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.212588473079516,
+      "learning_rate": 6.817226274617283e-07,
+      "loss": 0.0958,
+      "step": 4690
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.918837142437118,
+      "learning_rate": 6.802664749413079e-07,
+      "loss": 0.1281,
+      "step": 4691
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 8.434507021495676,
+      "learning_rate": 6.788117657241506e-07,
+      "loss": 0.1213,
+      "step": 4692
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 7.487469942524017,
+      "learning_rate": 6.773585002963007e-07,
+      "loss": 0.1587,
+      "step": 4693
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 7.914773574255319,
+      "learning_rate": 6.759066791433228e-07,
+      "loss": 0.1387,
+      "step": 4694
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 9.230097938274037,
+      "learning_rate": 6.744563027502959e-07,
+      "loss": 0.139,
+      "step": 4695
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 7.406132054228652,
+      "learning_rate": 6.730073716018187e-07,
+      "loss": 0.1562,
+      "step": 4696
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.54202900315914,
+      "learning_rate": 6.715598861820055e-07,
+      "loss": 0.1226,
+      "step": 4697
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.922145967208217,
+      "learning_rate": 6.701138469744883e-07,
+      "loss": 0.111,
+      "step": 4698
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.397305802789072,
+      "learning_rate": 6.686692544624157e-07,
+      "loss": 0.1105,
+      "step": 4699
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.660593338022077,
+      "learning_rate": 6.672261091284526e-07,
+      "loss": 0.1178,
+      "step": 4700
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 8.2651428920971,
+      "learning_rate": 6.657844114547812e-07,
+      "loss": 0.1681,
+      "step": 4701
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.319009621262608,
+      "learning_rate": 6.643441619230989e-07,
+      "loss": 0.094,
+      "step": 4702
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 6.571037009481898,
+      "learning_rate": 6.629053610146202e-07,
+      "loss": 0.1138,
+      "step": 4703
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 7.379231512952576,
+      "learning_rate": 6.61468009210075e-07,
+      "loss": 0.1219,
+      "step": 4704
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 6.991505241069526,
+      "learning_rate": 6.600321069897097e-07,
+      "loss": 0.1466,
+      "step": 4705
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 11.1500655678584,
+      "learning_rate": 6.585976548332856e-07,
+      "loss": 0.2023,
+      "step": 4706
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 7.836974094823008,
+      "learning_rate": 6.571646532200815e-07,
+      "loss": 0.1,
+      "step": 4707
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.2192067894980907,
+      "learning_rate": 6.557331026288855e-07,
+      "loss": 0.0592,
+      "step": 4708
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.990968911316154,
+      "learning_rate": 6.543030035380099e-07,
+      "loss": 0.1028,
+      "step": 4709
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.338984666872942,
+      "learning_rate": 6.528743564252737e-07,
+      "loss": 0.1162,
+      "step": 4710
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.645866648570372,
+      "learning_rate": 6.514471617680184e-07,
+      "loss": 0.1099,
+      "step": 4711
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.5825772199128085,
+      "learning_rate": 6.500214200430921e-07,
+      "loss": 0.1029,
+      "step": 4712
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.144375515006759,
+      "learning_rate": 6.485971317268658e-07,
+      "loss": 0.1123,
+      "step": 4713
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 5.7099658296260705,
+      "learning_rate": 6.471742972952172e-07,
+      "loss": 0.1021,
+      "step": 4714
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.454486821044744,
+      "learning_rate": 6.457529172235427e-07,
+      "loss": 0.1027,
+      "step": 4715
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.6424283925645256,
+      "learning_rate": 6.44332991986753e-07,
+      "loss": 0.1061,
+      "step": 4716
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.0353165182692345,
+      "learning_rate": 6.429145220592703e-07,
+      "loss": 0.1191,
+      "step": 4717
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.423744958784793,
+      "learning_rate": 6.414975079150321e-07,
+      "loss": 0.1349,
+      "step": 4718
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.998700076551767,
+      "learning_rate": 6.400819500274891e-07,
+      "loss": 0.1064,
+      "step": 4719
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.338283445411675,
+      "learning_rate": 6.386678488696057e-07,
+      "loss": 0.1465,
+      "step": 4720
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.025518446998439,
+      "learning_rate": 6.372552049138591e-07,
+      "loss": 0.1173,
+      "step": 4721
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.876253982615516,
+      "learning_rate": 6.358440186322401e-07,
+      "loss": 0.092,
+      "step": 4722
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 8.410474264470912,
+      "learning_rate": 6.344342904962536e-07,
+      "loss": 0.0841,
+      "step": 4723
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.331845239953185,
+      "learning_rate": 6.330260209769124e-07,
+      "loss": 0.1038,
+      "step": 4724
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.044163364551721,
+      "learning_rate": 6.316192105447499e-07,
+      "loss": 0.1189,
+      "step": 4725
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 4.123656229134195,
+      "learning_rate": 6.302138596698032e-07,
+      "loss": 0.0798,
+      "step": 4726
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.533338908312375,
+      "learning_rate": 6.288099688216309e-07,
+      "loss": 0.1392,
+      "step": 4727
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 7.886494343160055,
+      "learning_rate": 6.27407538469294e-07,
+      "loss": 0.1273,
+      "step": 4728
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 7.451374710261631,
+      "learning_rate": 6.260065690813754e-07,
+      "loss": 0.163,
+      "step": 4729
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 5.280301091354748,
+      "learning_rate": 6.246070611259603e-07,
+      "loss": 0.0771,
+      "step": 4730
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.950197928743349,
+      "learning_rate": 6.232090150706555e-07,
+      "loss": 0.1183,
+      "step": 4731
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 7.311848775780406,
+      "learning_rate": 6.218124313825696e-07,
+      "loss": 0.1131,
+      "step": 4732
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.961362394740585,
+      "learning_rate": 6.204173105283295e-07,
+      "loss": 0.1559,
+      "step": 4733
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.77422676116153,
+      "learning_rate": 6.190236529740701e-07,
+      "loss": 0.1699,
+      "step": 4734
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 8.929780803606354,
+      "learning_rate": 6.176314591854388e-07,
+      "loss": 0.168,
+      "step": 4735
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.220542108817106,
+      "learning_rate": 6.162407296275936e-07,
+      "loss": 0.1192,
+      "step": 4736
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 20.56705499979564,
+      "learning_rate": 6.148514647652026e-07,
+      "loss": 0.149,
+      "step": 4737
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 9.716608379449566,
+      "learning_rate": 6.134636650624448e-07,
+      "loss": 0.1365,
+      "step": 4738
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 5.082791614875223,
+      "learning_rate": 6.120773309830108e-07,
+      "loss": 0.1092,
+      "step": 4739
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.58955029935693,
+      "learning_rate": 6.106924629900996e-07,
+      "loss": 0.1178,
+      "step": 4740
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 7.309633988207233,
+      "learning_rate": 6.09309061546422e-07,
+      "loss": 0.0879,
+      "step": 4741
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 8.559534602189101,
+      "learning_rate": 6.079271271141979e-07,
+      "loss": 0.1665,
+      "step": 4742
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 6.906477266849566,
+      "learning_rate": 6.065466601551578e-07,
+      "loss": 0.1103,
+      "step": 4743
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 9.231629348906024,
+      "learning_rate": 6.051676611305401e-07,
+      "loss": 0.1409,
+      "step": 4744
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 5.120806072520015,
+      "learning_rate": 6.037901305010951e-07,
+      "loss": 0.1036,
+      "step": 4745
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 7.357155083012127,
+      "learning_rate": 6.024140687270813e-07,
+      "loss": 0.1104,
+      "step": 4746
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 5.399708029837064,
+      "learning_rate": 6.010394762682659e-07,
+      "loss": 0.0915,
+      "step": 4747
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 5.7543160484824085,
+      "learning_rate": 5.996663535839275e-07,
+      "loss": 0.0948,
+      "step": 4748
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 10.87274696278645,
+      "learning_rate": 5.982947011328489e-07,
+      "loss": 0.1655,
+      "step": 4749
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 6.79729828235239,
+      "learning_rate": 5.969245193733275e-07,
+      "loss": 0.0967,
+      "step": 4750
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 7.446921297838255,
+      "learning_rate": 5.955558087631641e-07,
+      "loss": 0.1556,
+      "step": 4751
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 6.479713414080842,
+      "learning_rate": 5.941885697596734e-07,
+      "loss": 0.1226,
+      "step": 4752
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 8.37872487494799,
+      "learning_rate": 5.928228028196714e-07,
+      "loss": 0.1355,
+      "step": 4753
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 7.936869826658227,
+      "learning_rate": 5.914585083994906e-07,
+      "loss": 0.1339,
+      "step": 4754
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 6.964322624359045,
+      "learning_rate": 5.900956869549629e-07,
+      "loss": 0.1367,
+      "step": 4755
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 7.650251277890639,
+      "learning_rate": 5.887343389414363e-07,
+      "loss": 0.1934,
+      "step": 4756
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 9.021005113566153,
+      "learning_rate": 5.873744648137592e-07,
+      "loss": 0.1357,
+      "step": 4757
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 11.186719406066896,
+      "learning_rate": 5.860160650262925e-07,
+      "loss": 0.1373,
+      "step": 4758
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.9630385462474855,
+      "learning_rate": 5.846591400329021e-07,
+      "loss": 0.0551,
+      "step": 4759
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 7.011343925108888,
+      "learning_rate": 5.833036902869626e-07,
+      "loss": 0.0994,
+      "step": 4760
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 8.396161662969257,
+      "learning_rate": 5.81949716241354e-07,
+      "loss": 0.1549,
+      "step": 4761
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 7.423724973182227,
+      "learning_rate": 5.805972183484654e-07,
+      "loss": 0.1567,
+      "step": 4762
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 6.139312833169541,
+      "learning_rate": 5.792461970601903e-07,
+      "loss": 0.1575,
+      "step": 4763
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.275973947423691,
+      "learning_rate": 5.778966528279306e-07,
+      "loss": 0.0974,
+      "step": 4764
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 6.929297521573184,
+      "learning_rate": 5.765485861025944e-07,
+      "loss": 0.1428,
+      "step": 4765
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 9.094874672404059,
+      "learning_rate": 5.752019973345963e-07,
+      "loss": 0.1667,
+      "step": 4766
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.088141421563402,
+      "learning_rate": 5.738568869738537e-07,
+      "loss": 0.1113,
+      "step": 4767
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 6.680419138288556,
+      "learning_rate": 5.725132554697971e-07,
+      "loss": 0.1432,
+      "step": 4768
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 5.9210508941282285,
+      "learning_rate": 5.711711032713547e-07,
+      "loss": 0.1071,
+      "step": 4769
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 7.310725061193263,
+      "learning_rate": 5.698304308269686e-07,
+      "loss": 0.1504,
+      "step": 4770
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.741656309899917,
+      "learning_rate": 5.684912385845786e-07,
+      "loss": 0.1337,
+      "step": 4771
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 8.428815406923771,
+      "learning_rate": 5.671535269916373e-07,
+      "loss": 0.1229,
+      "step": 4772
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.190650489682488,
+      "learning_rate": 5.658172964950953e-07,
+      "loss": 0.1238,
+      "step": 4773
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.782034891009411,
+      "learning_rate": 5.644825475414162e-07,
+      "loss": 0.0731,
+      "step": 4774
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.444261645843906,
+      "learning_rate": 5.631492805765609e-07,
+      "loss": 0.1149,
+      "step": 4775
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.550284656212252,
+      "learning_rate": 5.618174960459999e-07,
+      "loss": 0.1343,
+      "step": 4776
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.641197272560726,
+      "learning_rate": 5.604871943947071e-07,
+      "loss": 0.1244,
+      "step": 4777
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 16.192125850606384,
+      "learning_rate": 5.591583760671609e-07,
+      "loss": 0.175,
+      "step": 4778
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.770239360221412,
+      "learning_rate": 5.578310415073451e-07,
+      "loss": 0.0812,
+      "step": 4779
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 9.519992706791493,
+      "learning_rate": 5.565051911587455e-07,
+      "loss": 0.156,
+      "step": 4780
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.899766338731636,
+      "learning_rate": 5.551808254643543e-07,
+      "loss": 0.1325,
+      "step": 4781
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.377638893489806,
+      "learning_rate": 5.538579448666675e-07,
+      "loss": 0.1172,
+      "step": 4782
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.657629714932012,
+      "learning_rate": 5.525365498076807e-07,
+      "loss": 0.0878,
+      "step": 4783
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.999669463350998,
+      "learning_rate": 5.51216640728901e-07,
+      "loss": 0.1361,
+      "step": 4784
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.528022836641468,
+      "learning_rate": 5.498982180713308e-07,
+      "loss": 0.161,
+      "step": 4785
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 13.981093221771234,
+      "learning_rate": 5.485812822754826e-07,
+      "loss": 0.2415,
+      "step": 4786
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 6.204609953884755,
+      "learning_rate": 5.472658337813664e-07,
+      "loss": 0.0935,
+      "step": 4787
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 9.851833696225833,
+      "learning_rate": 5.459518730285007e-07,
+      "loss": 0.1015,
+      "step": 4788
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 8.172761495859053,
+      "learning_rate": 5.446394004559008e-07,
+      "loss": 0.1106,
+      "step": 4789
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 4.069756351444801,
+      "learning_rate": 5.43328416502093e-07,
+      "loss": 0.0833,
+      "step": 4790
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.754921719414479,
+      "learning_rate": 5.420189216050969e-07,
+      "loss": 0.1276,
+      "step": 4791
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 6.097853307500944,
+      "learning_rate": 5.407109162024409e-07,
+      "loss": 0.1288,
+      "step": 4792
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.6596367143990065,
+      "learning_rate": 5.394044007311544e-07,
+      "loss": 0.1259,
+      "step": 4793
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 9.109859513949639,
+      "learning_rate": 5.380993756277675e-07,
+      "loss": 0.1785,
+      "step": 4794
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.758617786067658,
+      "learning_rate": 5.367958413283141e-07,
+      "loss": 0.1019,
+      "step": 4795
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 10.073887412643687,
+      "learning_rate": 5.354937982683283e-07,
+      "loss": 0.1434,
+      "step": 4796
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 20.106138254105133,
+      "learning_rate": 5.341932468828481e-07,
+      "loss": 0.0942,
+      "step": 4797
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.547332850402219,
+      "learning_rate": 5.328941876064114e-07,
+      "loss": 0.1196,
+      "step": 4798
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 9.800610244619884,
+      "learning_rate": 5.315966208730578e-07,
+      "loss": 0.1415,
+      "step": 4799
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 5.735215749717641,
+      "learning_rate": 5.30300547116328e-07,
+      "loss": 0.1469,
+      "step": 4800
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.803423115653081,
+      "learning_rate": 5.290059667692655e-07,
+      "loss": 0.1031,
+      "step": 4801
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 7.114085098454165,
+      "learning_rate": 5.277128802644133e-07,
+      "loss": 0.1191,
+      "step": 4802
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 10.514855529211154,
+      "learning_rate": 5.264212880338154e-07,
+      "loss": 0.1995,
+      "step": 4803
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 7.725958167463443,
+      "learning_rate": 5.251311905090167e-07,
+      "loss": 0.0878,
+      "step": 4804
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.131421273398189,
+      "learning_rate": 5.238425881210624e-07,
+      "loss": 0.0955,
+      "step": 4805
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 9.949154525113872,
+      "learning_rate": 5.225554813004996e-07,
+      "loss": 0.1548,
+      "step": 4806
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 8.331347956438186,
+      "learning_rate": 5.21269870477375e-07,
+      "loss": 0.1467,
+      "step": 4807
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.448868593975817,
+      "learning_rate": 5.199857560812316e-07,
+      "loss": 0.1307,
+      "step": 4808
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 9.401074138435595,
+      "learning_rate": 5.187031385411206e-07,
+      "loss": 0.1598,
+      "step": 4809
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 8.997603436753522,
+      "learning_rate": 5.174220182855844e-07,
+      "loss": 0.1896,
+      "step": 4810
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 5.9623142347970015,
+      "learning_rate": 5.161423957426725e-07,
+      "loss": 0.1287,
+      "step": 4811
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 7.338573874367029,
+      "learning_rate": 5.148642713399272e-07,
+      "loss": 0.112,
+      "step": 4812
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 7.146812781041527,
+      "learning_rate": 5.13587645504397e-07,
+      "loss": 0.1658,
+      "step": 4813
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 9.006764590662835,
+      "learning_rate": 5.123125186626227e-07,
+      "loss": 0.2671,
+      "step": 4814
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.677054214930086,
+      "learning_rate": 5.110388912406517e-07,
+      "loss": 0.1119,
+      "step": 4815
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 5.786888005642126,
+      "learning_rate": 5.097667636640241e-07,
+      "loss": 0.117,
+      "step": 4816
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 8.566924141847709,
+      "learning_rate": 5.084961363577817e-07,
+      "loss": 0.1681,
+      "step": 4817
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.283648205807579,
+      "learning_rate": 5.072270097464649e-07,
+      "loss": 0.1382,
+      "step": 4818
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.70248104170661,
+      "learning_rate": 5.059593842541127e-07,
+      "loss": 0.1403,
+      "step": 4819
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.145804986104878,
+      "learning_rate": 5.04693260304262e-07,
+      "loss": 0.0723,
+      "step": 4820
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 7.58725175311385,
+      "learning_rate": 5.034286383199488e-07,
+      "loss": 0.1171,
+      "step": 4821
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 5.057300916102727,
+      "learning_rate": 5.021655187237067e-07,
+      "loss": 0.1077,
+      "step": 4822
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.823300742743953,
+      "learning_rate": 5.009039019375672e-07,
+      "loss": 0.1105,
+      "step": 4823
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.406379850384206,
+      "learning_rate": 4.996437883830596e-07,
+      "loss": 0.1053,
+      "step": 4824
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 7.476297609589221,
+      "learning_rate": 4.983851784812127e-07,
+      "loss": 0.1573,
+      "step": 4825
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 10.028708462150261,
+      "learning_rate": 4.97128072652549e-07,
+      "loss": 0.0988,
+      "step": 4826
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 6.70771060659306,
+      "learning_rate": 4.958724713170943e-07,
+      "loss": 0.121,
+      "step": 4827
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 7.550234303625725,
+      "learning_rate": 4.946183748943639e-07,
+      "loss": 0.1516,
+      "step": 4828
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 4.250072350581676,
+      "learning_rate": 4.933657838033795e-07,
+      "loss": 0.0782,
+      "step": 4829
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 14.901528479240605,
+      "learning_rate": 4.921146984626507e-07,
+      "loss": 0.2083,
+      "step": 4830
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 8.861570669785552,
+      "learning_rate": 4.908651192901926e-07,
+      "loss": 0.1427,
+      "step": 4831
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 7.774296537989752,
+      "learning_rate": 4.896170467035089e-07,
+      "loss": 0.1311,
+      "step": 4832
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 6.097843488270496,
+      "learning_rate": 4.883704811196072e-07,
+      "loss": 0.1015,
+      "step": 4833
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.944954993942699,
+      "learning_rate": 4.871254229549855e-07,
+      "loss": 0.1368,
+      "step": 4834
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 6.610591286434344,
+      "learning_rate": 4.858818726256425e-07,
+      "loss": 0.1675,
+      "step": 4835
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 7.955225527923404,
+      "learning_rate": 4.846398305470712e-07,
+      "loss": 0.1721,
+      "step": 4836
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 10.896382771601361,
+      "learning_rate": 4.833992971342604e-07,
+      "loss": 0.1346,
+      "step": 4837
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 4.712058505532531,
+      "learning_rate": 4.821602728016955e-07,
+      "loss": 0.0718,
+      "step": 4838
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 4.889206680905457,
+      "learning_rate": 4.809227579633585e-07,
+      "loss": 0.0939,
+      "step": 4839
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 6.531400313220989,
+      "learning_rate": 4.796867530327249e-07,
+      "loss": 0.1354,
+      "step": 4840
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 6.121051370381773,
+      "learning_rate": 4.784522584227675e-07,
+      "loss": 0.1246,
+      "step": 4841
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.447583991391642,
+      "learning_rate": 4.772192745459536e-07,
+      "loss": 0.1349,
+      "step": 4842
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.9717320548571236,
+      "learning_rate": 4.7598780181424666e-07,
+      "loss": 0.1399,
+      "step": 4843
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.189759067456543,
+      "learning_rate": 4.7475784063910404e-07,
+      "loss": 0.1154,
+      "step": 4844
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.200052811630355,
+      "learning_rate": 4.7352939143147927e-07,
+      "loss": 0.1024,
+      "step": 4845
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.327313296684159,
+      "learning_rate": 4.72302454601819e-07,
+      "loss": 0.1127,
+      "step": 4846
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.821690226987369,
+      "learning_rate": 4.7107703056006706e-07,
+      "loss": 0.1036,
+      "step": 4847
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.37306344284546,
+      "learning_rate": 4.6985311971565806e-07,
+      "loss": 0.0723,
+      "step": 4848
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.177107325600691,
+      "learning_rate": 4.6863072247752664e-07,
+      "loss": 0.0664,
+      "step": 4849
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 6.044564249735391,
+      "learning_rate": 4.67409839254096e-07,
+      "loss": 0.1136,
+      "step": 4850
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.84423169120056,
+      "learning_rate": 4.66190470453286e-07,
+      "loss": 0.1257,
+      "step": 4851
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 6.516029681667653,
+      "learning_rate": 4.6497261648251134e-07,
+      "loss": 0.095,
+      "step": 4852
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.602034097877229,
+      "learning_rate": 4.6375627774867925e-07,
+      "loss": 0.1377,
+      "step": 4853
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 8.561475016663382,
+      "learning_rate": 4.6254145465819134e-07,
+      "loss": 0.1226,
+      "step": 4854
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 6.043790120678902,
+      "learning_rate": 4.6132814761694234e-07,
+      "loss": 0.0833,
+      "step": 4855
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 5.934699069054085,
+      "learning_rate": 4.6011635703032075e-07,
+      "loss": 0.0867,
+      "step": 4856
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 8.280932206636068,
+      "learning_rate": 4.589060833032083e-07,
+      "loss": 0.1135,
+      "step": 4857
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 10.11043341804737,
+      "learning_rate": 4.5769732683997983e-07,
+      "loss": 0.1636,
+      "step": 4858
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 4.578783922415493,
+      "learning_rate": 4.564900880445039e-07,
+      "loss": 0.1052,
+      "step": 4859
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 6.231779936962509,
+      "learning_rate": 4.552843673201407e-07,
+      "loss": 0.105,
+      "step": 4860
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 6.714870031308601,
+      "learning_rate": 4.540801650697446e-07,
+      "loss": 0.176,
+      "step": 4861
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 5.82521739237652,
+      "learning_rate": 4.528774816956616e-07,
+      "loss": 0.0887,
+      "step": 4862
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 7.575021134589715,
+      "learning_rate": 4.516763175997302e-07,
+      "loss": 0.1743,
+      "step": 4863
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 7.637041020100914,
+      "learning_rate": 4.5047667318328215e-07,
+      "loss": 0.0961,
+      "step": 4864
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 5.555943353432021,
+      "learning_rate": 4.492785488471413e-07,
+      "loss": 0.1068,
+      "step": 4865
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 6.394841605960666,
+      "learning_rate": 4.480819449916224e-07,
+      "loss": 0.1062,
+      "step": 4866
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 6.625874447124118,
+      "learning_rate": 4.468868620165334e-07,
+      "loss": 0.1735,
+      "step": 4867
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 9.268464740985305,
+      "learning_rate": 4.4569330032117496e-07,
+      "loss": 0.1744,
+      "step": 4868
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 7.161236160543748,
+      "learning_rate": 4.445012603043347e-07,
+      "loss": 0.158,
+      "step": 4869
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 6.150201890120088,
+      "learning_rate": 4.4331074236430014e-07,
+      "loss": 0.1084,
+      "step": 4870
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.604703088860564,
+      "learning_rate": 4.421217468988409e-07,
+      "loss": 0.1189,
+      "step": 4871
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 6.470583792177313,
+      "learning_rate": 4.409342743052264e-07,
+      "loss": 0.1427,
+      "step": 4872
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 6.157097944965909,
+      "learning_rate": 4.3974832498020983e-07,
+      "loss": 0.1149,
+      "step": 4873
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 7.227392034283128,
+      "learning_rate": 4.385638993200425e-07,
+      "loss": 0.1059,
+      "step": 4874
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.359135702963531,
+      "learning_rate": 4.3738099772045963e-07,
+      "loss": 0.0933,
+      "step": 4875
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.540963344549303,
+      "learning_rate": 4.3619962057669216e-07,
+      "loss": 0.1465,
+      "step": 4876
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 8.921949783265053,
+      "learning_rate": 4.350197682834606e-07,
+      "loss": 0.1624,
+      "step": 4877
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 6.8268412987785485,
+      "learning_rate": 4.338414412349745e-07,
+      "loss": 0.1013,
+      "step": 4878
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 7.046004451519703,
+      "learning_rate": 4.3266463982493566e-07,
+      "loss": 0.1239,
+      "step": 4879
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 6.6020244320462504,
+      "learning_rate": 4.314893644465351e-07,
+      "loss": 0.1201,
+      "step": 4880
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.97689046629764,
+      "learning_rate": 4.303156154924537e-07,
+      "loss": 0.1025,
+      "step": 4881
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 16.85592779467006,
+      "learning_rate": 4.291433933548633e-07,
+      "loss": 0.1746,
+      "step": 4882
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 20.986355936055386,
+      "learning_rate": 4.279726984254251e-07,
+      "loss": 0.2146,
+      "step": 4883
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 8.537459210196602,
+      "learning_rate": 4.268035310952906e-07,
+      "loss": 0.1295,
+      "step": 4884
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.993488754974167,
+      "learning_rate": 4.256358917550979e-07,
+      "loss": 0.0913,
+      "step": 4885
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.377518655838321,
+      "learning_rate": 4.244697807949805e-07,
+      "loss": 0.0779,
+      "step": 4886
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 6.679235808102704,
+      "learning_rate": 4.2330519860455446e-07,
+      "loss": 0.1,
+      "step": 4887
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 6.926326752680834,
+      "learning_rate": 4.2214214557293133e-07,
+      "loss": 0.1694,
+      "step": 4888
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 10.173919781648312,
+      "learning_rate": 4.209806220887053e-07,
+      "loss": 0.1553,
+      "step": 4889
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 8.716790283479154,
+      "learning_rate": 4.1982062853996695e-07,
+      "loss": 0.1282,
+      "step": 4890
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.741222580663796,
+      "learning_rate": 4.1866216531428806e-07,
+      "loss": 0.0996,
+      "step": 4891
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 8.604221773949565,
+      "learning_rate": 4.1750523279873613e-07,
+      "loss": 0.119,
+      "step": 4892
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 6.081160173256382,
+      "learning_rate": 4.1634983137986083e-07,
+      "loss": 0.1091,
+      "step": 4893
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 7.791749150013185,
+      "learning_rate": 4.151959614437046e-07,
+      "loss": 0.1576,
+      "step": 4894
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 7.986445225209414,
+      "learning_rate": 4.1404362337579716e-07,
+      "loss": 0.1707,
+      "step": 4895
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 8.400952391001237,
+      "learning_rate": 4.128928175611546e-07,
+      "loss": 0.1184,
+      "step": 4896
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.741392906608111,
+      "learning_rate": 4.1174354438428434e-07,
+      "loss": 0.1042,
+      "step": 4897
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 7.079323754898526,
+      "learning_rate": 4.105958042291791e-07,
+      "loss": 0.1346,
+      "step": 4898
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 8.395224511829195,
+      "learning_rate": 4.0944959747931945e-07,
+      "loss": 0.1548,
+      "step": 4899
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 8.677897205722589,
+      "learning_rate": 4.0830492451767566e-07,
+      "loss": 0.1658,
+      "step": 4900
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.826376865813371,
+      "learning_rate": 4.0716178572670405e-07,
+      "loss": 0.1022,
+      "step": 4901
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 7.570457325496883,
+      "learning_rate": 4.060201814883474e-07,
+      "loss": 0.1015,
+      "step": 4902
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 6.666909786473492,
+      "learning_rate": 4.0488011218403844e-07,
+      "loss": 0.1423,
+      "step": 4903
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 10.844528241311131,
+      "learning_rate": 4.0374157819469406e-07,
+      "loss": 0.1428,
+      "step": 4904
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.958351896197906,
+      "learning_rate": 4.0260457990072113e-07,
+      "loss": 0.106,
+      "step": 4905
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 6.199859560189758,
+      "learning_rate": 4.014691176820107e-07,
+      "loss": 0.1133,
+      "step": 4906
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 6.634470604062944,
+      "learning_rate": 4.003351919179421e-07,
+      "loss": 0.1403,
+      "step": 4907
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 7.633691260784369,
+      "learning_rate": 3.9920280298738125e-07,
+      "loss": 0.2009,
+      "step": 4908
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.181243123537604,
+      "learning_rate": 3.980719512686809e-07,
+      "loss": 0.1056,
+      "step": 4909
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 5.055526650996727,
+      "learning_rate": 3.969426371396773e-07,
+      "loss": 0.0876,
+      "step": 4910
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 9.38989387103326,
+      "learning_rate": 3.9581486097769905e-07,
+      "loss": 0.1478,
+      "step": 4911
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 6.137927650461356,
+      "learning_rate": 3.946886231595526e-07,
+      "loss": 0.1445,
+      "step": 4912
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.845052725437297,
+      "learning_rate": 3.935639240615396e-07,
+      "loss": 0.1101,
+      "step": 4913
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 9.620807970425544,
+      "learning_rate": 3.924407640594391e-07,
+      "loss": 0.1301,
+      "step": 4914
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 5.003315815491063,
+      "learning_rate": 3.913191435285224e-07,
+      "loss": 0.0854,
+      "step": 4915
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 10.51854542922581,
+      "learning_rate": 3.9019906284354145e-07,
+      "loss": 0.1211,
+      "step": 4916
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 11.27692744405923,
+      "learning_rate": 3.8908052237873863e-07,
+      "loss": 0.1339,
+      "step": 4917
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 6.778719813958779,
+      "learning_rate": 3.879635225078371e-07,
+      "loss": 0.1556,
+      "step": 4918
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 6.5589312805748445,
+      "learning_rate": 3.868480636040484e-07,
+      "loss": 0.094,
+      "step": 4919
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 11.230659542738735,
+      "learning_rate": 3.857341460400665e-07,
+      "loss": 0.1584,
+      "step": 4920
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 6.838176925618872,
+      "learning_rate": 3.846217701880739e-07,
+      "loss": 0.1112,
+      "step": 4921
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.807557680870521,
+      "learning_rate": 3.835109364197348e-07,
+      "loss": 0.0952,
+      "step": 4922
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 5.99047386603123,
+      "learning_rate": 3.8240164510620017e-07,
+      "loss": 0.0955,
+      "step": 4923
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 5.921585109056007,
+      "learning_rate": 3.81293896618104e-07,
+      "loss": 0.1046,
+      "step": 4924
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.71667528673715,
+      "learning_rate": 3.8018769132556644e-07,
+      "loss": 0.0726,
+      "step": 4925
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 6.07125569493634,
+      "learning_rate": 3.790830295981912e-07,
+      "loss": 0.1105,
+      "step": 4926
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.207268748077647,
+      "learning_rate": 3.7797991180506643e-07,
+      "loss": 0.0854,
+      "step": 4927
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 6.1493317988743845,
+      "learning_rate": 3.768783383147623e-07,
+      "loss": 0.0932,
+      "step": 4928
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 7.303017501676457,
+      "learning_rate": 3.757783094953382e-07,
+      "loss": 0.1523,
+      "step": 4929
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 10.291370181214797,
+      "learning_rate": 3.746798257143314e-07,
+      "loss": 0.1628,
+      "step": 4930
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.898999805060367,
+      "learning_rate": 3.735828873387681e-07,
+      "loss": 0.0839,
+      "step": 4931
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.965168117910225,
+      "learning_rate": 3.724874947351531e-07,
+      "loss": 0.1173,
+      "step": 4932
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 5.063017750866719,
+      "learning_rate": 3.7139364826948077e-07,
+      "loss": 0.0971,
+      "step": 4933
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 5.753008355329342,
+      "learning_rate": 3.7030134830722207e-07,
+      "loss": 0.0903,
+      "step": 4934
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.717973246272266,
+      "learning_rate": 3.692105952133379e-07,
+      "loss": 0.0975,
+      "step": 4935
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 8.333430271233516,
+      "learning_rate": 3.681213893522667e-07,
+      "loss": 0.1337,
+      "step": 4936
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 10.107222632655112,
+      "learning_rate": 3.670337310879335e-07,
+      "loss": 0.1675,
+      "step": 4937
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 9.419377853290698,
+      "learning_rate": 3.6594762078374536e-07,
+      "loss": 0.1014,
+      "step": 4938
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 12.43896448622122,
+      "learning_rate": 3.6486305880259085e-07,
+      "loss": 0.2437,
+      "step": 4939
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.616562037480525,
+      "learning_rate": 3.6378004550684355e-07,
+      "loss": 0.0884,
+      "step": 4940
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 6.338766725755504,
+      "learning_rate": 3.626985812583572e-07,
+      "loss": 0.1049,
+      "step": 4941
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.202111188217427,
+      "learning_rate": 3.6161866641847007e-07,
+      "loss": 0.1033,
+      "step": 4942
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 7.316758540310851,
+      "learning_rate": 3.6054030134800243e-07,
+      "loss": 0.1517,
+      "step": 4943
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 7.275565880480988,
+      "learning_rate": 3.594634864072527e-07,
+      "loss": 0.1464,
+      "step": 4944
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 10.317802181069961,
+      "learning_rate": 3.583882219560092e-07,
+      "loss": 0.2065,
+      "step": 4945
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 8.125843837992509,
+      "learning_rate": 3.57314508353534e-07,
+      "loss": 0.1593,
+      "step": 4946
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.59942806326855,
+      "learning_rate": 3.5624234595857787e-07,
+      "loss": 0.1151,
+      "step": 4947
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 7.182502216677451,
+      "learning_rate": 3.551717351293676e-07,
+      "loss": 0.1285,
+      "step": 4948
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 8.035563717929566,
+      "learning_rate": 3.541026762236166e-07,
+      "loss": 0.1669,
+      "step": 4949
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.790990792314284,
+      "learning_rate": 3.5303516959851405e-07,
+      "loss": 0.1127,
+      "step": 4950
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 4.881110923975365,
+      "learning_rate": 3.519692156107379e-07,
+      "loss": 0.1031,
+      "step": 4951
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 9.586597575490307,
+      "learning_rate": 3.509048146164401e-07,
+      "loss": 0.1537,
+      "step": 4952
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.858555298946315,
+      "learning_rate": 3.4984196697125827e-07,
+      "loss": 0.0942,
+      "step": 4953
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.355412324174652,
+      "learning_rate": 3.4878067303030836e-07,
+      "loss": 0.0513,
+      "step": 4954
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 9.169699928002089,
+      "learning_rate": 3.4772093314818957e-07,
+      "loss": 0.1544,
+      "step": 4955
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.753478969791648,
+      "learning_rate": 3.4666274767897967e-07,
+      "loss": 0.1509,
+      "step": 4956
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.529002063527844,
+      "learning_rate": 3.456061169762392e-07,
+      "loss": 0.1528,
+      "step": 4957
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.793947603721516,
+      "learning_rate": 3.44551041393007e-07,
+      "loss": 0.1267,
+      "step": 4958
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.569499505899291,
+      "learning_rate": 3.434975212818048e-07,
+      "loss": 0.1304,
+      "step": 4959
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.899698816129731,
+      "learning_rate": 3.424455569946317e-07,
+      "loss": 0.1163,
+      "step": 4960
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.341545137728287,
+      "learning_rate": 3.4139514888296975e-07,
+      "loss": 0.1301,
+      "step": 4961
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 10.0706321470383,
+      "learning_rate": 3.403462972977789e-07,
+      "loss": 0.1243,
+      "step": 4962
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.231489025292407,
+      "learning_rate": 3.392990025895004e-07,
+      "loss": 0.1277,
+      "step": 4963
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 10.311986109419417,
+      "learning_rate": 3.3825326510805556e-07,
+      "loss": 0.1787,
+      "step": 4964
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 7.1396184042409105,
+      "learning_rate": 3.372090852028437e-07,
+      "loss": 0.1366,
+      "step": 4965
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.781959199435736,
+      "learning_rate": 3.361664632227446e-07,
+      "loss": 0.0825,
+      "step": 4966
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 6.2875082133414395,
+      "learning_rate": 3.3512539951611856e-07,
+      "loss": 0.147,
+      "step": 4967
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.914439645723257,
+      "learning_rate": 3.3408589443080395e-07,
+      "loss": 0.1083,
+      "step": 4968
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 7.325977047729606,
+      "learning_rate": 3.3304794831411804e-07,
+      "loss": 0.1431,
+      "step": 4969
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 8.27154445640713,
+      "learning_rate": 3.3201156151285994e-07,
+      "loss": 0.1476,
+      "step": 4970
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 6.2873062920787355,
+      "learning_rate": 3.309767343733028e-07,
+      "loss": 0.0926,
+      "step": 4971
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 3.7912487931444017,
+      "learning_rate": 3.299434672412044e-07,
+      "loss": 0.0831,
+      "step": 4972
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 6.907191593542313,
+      "learning_rate": 3.2891176046179583e-07,
+      "loss": 0.1079,
+      "step": 4973
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 7.634784204009729,
+      "learning_rate": 3.278816143797919e-07,
+      "loss": 0.1184,
+      "step": 4974
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 8.738231719908597,
+      "learning_rate": 3.2685302933938177e-07,
+      "loss": 0.1627,
+      "step": 4975
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 7.051208068222329,
+      "learning_rate": 3.2582600568423715e-07,
+      "loss": 0.1112,
+      "step": 4976
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.633258071747942,
+      "learning_rate": 3.2480054375750305e-07,
+      "loss": 0.0843,
+      "step": 4977
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 6.457146068070933,
+      "learning_rate": 3.237766439018064e-07,
+      "loss": 0.1165,
+      "step": 4978
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 4.686262096507274,
+      "learning_rate": 3.227543064592514e-07,
+      "loss": 0.0764,
+      "step": 4979
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 4.730731801506817,
+      "learning_rate": 3.2173353177142044e-07,
+      "loss": 0.0969,
+      "step": 4980
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.936171738164148,
+      "learning_rate": 3.207143201793722e-07,
+      "loss": 0.0854,
+      "step": 4981
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 5.8137612074069,
+      "learning_rate": 3.1969667202364496e-07,
+      "loss": 0.1029,
+      "step": 4982
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 9.229282590022793,
+      "learning_rate": 3.1868058764425337e-07,
+      "loss": 0.1616,
+      "step": 4983
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.105331418972408,
+      "learning_rate": 3.1766606738069084e-07,
+      "loss": 0.1034,
+      "step": 4984
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.807312628779206,
+      "learning_rate": 3.166531115719268e-07,
+      "loss": 0.144,
+      "step": 4985
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.869038868344821,
+      "learning_rate": 3.1564172055640994e-07,
+      "loss": 0.0964,
+      "step": 4986
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 8.875331548761368,
+      "learning_rate": 3.1463189467206166e-07,
+      "loss": 0.1221,
+      "step": 4987
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.938849961784851,
+      "learning_rate": 3.1362363425628763e-07,
+      "loss": 0.0954,
+      "step": 4988
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 10.180385680278965,
+      "learning_rate": 3.1261693964596275e-07,
+      "loss": 0.1552,
+      "step": 4989
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 6.039383569321269,
+      "learning_rate": 3.116118111774452e-07,
+      "loss": 0.0889,
+      "step": 4990
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 5.068556225143177,
+      "learning_rate": 3.106082491865647e-07,
+      "loss": 0.093,
+      "step": 4991
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.284668219658361,
+      "learning_rate": 3.0960625400863253e-07,
+      "loss": 0.1415,
+      "step": 4992
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 14.454110851656004,
+      "learning_rate": 3.0860582597843137e-07,
+      "loss": 0.2096,
+      "step": 4993
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 6.243067078626986,
+      "learning_rate": 3.0760696543022496e-07,
+      "loss": 0.0966,
+      "step": 4994
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 5.398929947949957,
+      "learning_rate": 3.066096726977502e-07,
+      "loss": 0.0952,
+      "step": 4995
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 6.239239492157707,
+      "learning_rate": 3.056139481142206e-07,
+      "loss": 0.101,
+      "step": 4996
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 9.023044388173963,
+      "learning_rate": 3.0461979201232674e-07,
+      "loss": 0.1794,
+      "step": 4997
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 7.570504247937259,
+      "learning_rate": 3.0362720472423503e-07,
+      "loss": 0.1161,
+      "step": 4998
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 10.355018381146655,
+      "learning_rate": 3.026361865815869e-07,
+      "loss": 0.1808,
+      "step": 4999
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 6.25080861497107,
+      "learning_rate": 3.016467379154997e-07,
+      "loss": 0.1207,
+      "step": 5000
+    },
+    {
+      "epoch": 3.57,
+      "eval_avg_AUC": 0.7939005198150356,
+      "eval_avg_Accuracy": 0.7027105437665783,
+      "eval_avg_Accuracy-right": 0.8833963740706926,
+      "eval_avg_Accuracy-wrong": 0.38765067091198546,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6581912384379904,
+      "eval_last_AUC": 0.8149474201895899,
+      "eval_last_Accuracy": 0.738395225464191,
+      "eval_last_Accuracy-right": 0.8196817529672623,
+      "eval_last_Accuracy-wrong": 0.5966568114623607,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6804592176400829,
+      "eval_max_AUC": 0.7778448578196614,
+      "eval_max_Accuracy": 0.6441893236074271,
+      "eval_max_Accuracy-right": 0.9792617712273379,
+      "eval_max_Accuracy-wrong": 0.05992722310666363,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6216123178884508,
+      "eval_min_AUC": 0.7975842894056173,
+      "eval_min_Accuracy": 0.7238478116710876,
+      "eval_min_Accuracy-right": 0.7295552367288379,
+      "eval_min_Accuracy-wrong": 0.7138958380714123,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6586734622814718,
+      "eval_prod_AUC": 0.8002513229863527,
+      "eval_prod_Accuracy": 0.7148955570291777,
+      "eval_prod_Accuracy-right": 0.6743185078909613,
+      "eval_prod_Accuracy-wrong": 0.7856493063452354,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.661249401851051,
+      "eval_runtime": 251.6637,
+      "eval_samples_per_second": 95.874,
+      "eval_steps_per_second": 2.996,
+      "eval_sum_AUC": 0.6658840128941205,
+      "eval_sum_Accuracy": 0.638967175066313,
+      "eval_sum_Accuracy-right": 0.9868918742663363,
+      "eval_sum_Accuracy-wrong": 0.032294746418012284,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6474363870292106,
+      "step": 5000
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 8.039699710822118,
+      "learning_rate": 3.0065885905656733e-07,
+      "loss": 0.139,
+      "step": 5001
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.561175020518558,
+      "learning_rate": 2.99672550334858e-07,
+      "loss": 0.0732,
+      "step": 5002
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 5.489308809958234,
+      "learning_rate": 2.986878120799158e-07,
+      "loss": 0.1164,
+      "step": 5003
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 7.287913363035142,
+      "learning_rate": 2.977046446207604e-07,
+      "loss": 0.0874,
+      "step": 5004
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 7.221934125161167,
+      "learning_rate": 2.967230482858863e-07,
+      "loss": 0.1486,
+      "step": 5005
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 8.007176575502351,
+      "learning_rate": 2.957430234032627e-07,
+      "loss": 0.1381,
+      "step": 5006
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 9.070326449400396,
+      "learning_rate": 2.947645703003338e-07,
+      "loss": 0.1473,
+      "step": 5007
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 10.242941718824545,
+      "learning_rate": 2.937876893040209e-07,
+      "loss": 0.1548,
+      "step": 5008
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.051438057132722,
+      "learning_rate": 2.9281238074071463e-07,
+      "loss": 0.1425,
+      "step": 5009
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 8.348048515530829,
+      "learning_rate": 2.9183864493628756e-07,
+      "loss": 0.1549,
+      "step": 5010
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 9.348540506356116,
+      "learning_rate": 2.908664822160806e-07,
+      "loss": 0.1504,
+      "step": 5011
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 6.608695626483558,
+      "learning_rate": 2.898958929049117e-07,
+      "loss": 0.0918,
+      "step": 5012
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.522212380654753,
+      "learning_rate": 2.889268773270731e-07,
+      "loss": 0.166,
+      "step": 5013
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 6.086064311100961,
+      "learning_rate": 2.879594358063303e-07,
+      "loss": 0.117,
+      "step": 5014
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 6.2796926492958995,
+      "learning_rate": 2.869935686659248e-07,
+      "loss": 0.0981,
+      "step": 5015
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 6.291532974607174,
+      "learning_rate": 2.8602927622856935e-07,
+      "loss": 0.1378,
+      "step": 5016
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.3972265354997235,
+      "learning_rate": 2.8506655881645305e-07,
+      "loss": 0.0914,
+      "step": 5017
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.768331165784399,
+      "learning_rate": 2.841054167512369e-07,
+      "loss": 0.1865,
+      "step": 5018
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.9944003757927975,
+      "learning_rate": 2.8314585035405683e-07,
+      "loss": 0.1632,
+      "step": 5019
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 6.256066146658917,
+      "learning_rate": 2.8218785994552136e-07,
+      "loss": 0.106,
+      "step": 5020
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 14.01419943071183,
+      "learning_rate": 2.8123144584571326e-07,
+      "loss": 0.1823,
+      "step": 5021
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.421782138203785,
+      "learning_rate": 2.8027660837418813e-07,
+      "loss": 0.1256,
+      "step": 5022
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 6.145061257805026,
+      "learning_rate": 2.793233478499752e-07,
+      "loss": 0.1068,
+      "step": 5023
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 7.773542257086725,
+      "learning_rate": 2.7837166459157625e-07,
+      "loss": 0.1125,
+      "step": 5024
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.764175341048938,
+      "learning_rate": 2.77421558916966e-07,
+      "loss": 0.1198,
+      "step": 5025
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.684493454442424,
+      "learning_rate": 2.764730311435931e-07,
+      "loss": 0.0885,
+      "step": 5026
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 7.440949476639598,
+      "learning_rate": 2.755260815883781e-07,
+      "loss": 0.1259,
+      "step": 5027
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 9.466050890106064,
+      "learning_rate": 2.745807105677145e-07,
+      "loss": 0.1307,
+      "step": 5028
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 5.8345255587512765,
+      "learning_rate": 2.736369183974685e-07,
+      "loss": 0.0985,
+      "step": 5029
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 7.466645169078422,
+      "learning_rate": 2.726947053929768e-07,
+      "loss": 0.1625,
+      "step": 5030
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 13.413102283651124,
+      "learning_rate": 2.7175407186905367e-07,
+      "loss": 0.1526,
+      "step": 5031
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 9.913384538686545,
+      "learning_rate": 2.708150181399788e-07,
+      "loss": 0.1227,
+      "step": 5032
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 29.713012887802034,
+      "learning_rate": 2.698775445195101e-07,
+      "loss": 0.1667,
+      "step": 5033
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 5.012468307125459,
+      "learning_rate": 2.689416513208726e-07,
+      "loss": 0.0981,
+      "step": 5034
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 9.259889616367115,
+      "learning_rate": 2.6800733885676833e-07,
+      "loss": 0.1725,
+      "step": 5035
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 8.444744690254245,
+      "learning_rate": 2.6707460743936653e-07,
+      "loss": 0.1523,
+      "step": 5036
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 10.368928285325568,
+      "learning_rate": 2.6614345738031014e-07,
+      "loss": 0.1543,
+      "step": 5037
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 7.8762410419001005,
+      "learning_rate": 2.6521388899071467e-07,
+      "loss": 0.1006,
+      "step": 5038
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 7.032417525664705,
+      "learning_rate": 2.642859025811656e-07,
+      "loss": 0.1088,
+      "step": 5039
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 8.871769638082888,
+      "learning_rate": 2.633594984617199e-07,
+      "loss": 0.1471,
+      "step": 5040
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 6.930037290213703,
+      "learning_rate": 2.624346769419078e-07,
+      "loss": 0.1376,
+      "step": 5041
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 7.1968598150909955,
+      "learning_rate": 2.6151143833072824e-07,
+      "loss": 0.1517,
+      "step": 5042
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 5.261682987100089,
+      "learning_rate": 2.605897829366527e-07,
+      "loss": 0.0737,
+      "step": 5043
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 6.097460825799445,
+      "learning_rate": 2.596697110676233e-07,
+      "loss": 0.0956,
+      "step": 5044
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 8.016509501397028,
+      "learning_rate": 2.5875122303105403e-07,
+      "loss": 0.1566,
+      "step": 5045
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 6.711365280336349,
+      "learning_rate": 2.5783431913382673e-07,
+      "loss": 0.1152,
+      "step": 5046
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 5.2815052281868216,
+      "learning_rate": 2.5691899968229904e-07,
+      "loss": 0.1055,
+      "step": 5047
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.197556171742845,
+      "learning_rate": 2.560052649822925e-07,
+      "loss": 0.0975,
+      "step": 5048
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 12.246007633872518,
+      "learning_rate": 2.5509311533910674e-07,
+      "loss": 0.1521,
+      "step": 5049
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 7.104104832857858,
+      "learning_rate": 2.5418255105750465e-07,
+      "loss": 0.1245,
+      "step": 5050
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.889736914067923,
+      "learning_rate": 2.532735724417251e-07,
+      "loss": 0.1339,
+      "step": 5051
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 9.149482943543847,
+      "learning_rate": 2.52366179795473e-07,
+      "loss": 0.1666,
+      "step": 5052
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 11.841217531696419,
+      "learning_rate": 2.5146037342192673e-07,
+      "loss": 0.2165,
+      "step": 5053
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 8.367402162775445,
+      "learning_rate": 2.505561536237311e-07,
+      "loss": 0.158,
+      "step": 5054
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.577698639400431,
+      "learning_rate": 2.496535207030043e-07,
+      "loss": 0.0587,
+      "step": 5055
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 9.757830561055071,
+      "learning_rate": 2.4875247496133234e-07,
+      "loss": 0.1003,
+      "step": 5056
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 10.30190449595356,
+      "learning_rate": 2.4785301669977116e-07,
+      "loss": 0.1698,
+      "step": 5057
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.289239425103564,
+      "learning_rate": 2.469551462188463e-07,
+      "loss": 0.1116,
+      "step": 5058
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 6.916398851164515,
+      "learning_rate": 2.460588638185535e-07,
+      "loss": 0.1403,
+      "step": 5059
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 8.156390119211379,
+      "learning_rate": 2.45164169798357e-07,
+      "loss": 0.1232,
+      "step": 5060
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.444029203628988,
+      "learning_rate": 2.4427106445719053e-07,
+      "loss": 0.1094,
+      "step": 5061
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.602904335530537,
+      "learning_rate": 2.4337954809345807e-07,
+      "loss": 0.1433,
+      "step": 5062
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 5.715241288214859,
+      "learning_rate": 2.4248962100503095e-07,
+      "loss": 0.0869,
+      "step": 5063
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 4.962640828966288,
+      "learning_rate": 2.416012834892506e-07,
+      "loss": 0.1055,
+      "step": 5064
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 6.660676204308661,
+      "learning_rate": 2.4071453584292693e-07,
+      "loss": 0.1177,
+      "step": 5065
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 7.962611371474207,
+      "learning_rate": 2.3982937836233954e-07,
+      "loss": 0.1488,
+      "step": 5066
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 7.174907994976234,
+      "learning_rate": 2.389458113432347e-07,
+      "loss": 0.1106,
+      "step": 5067
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 5.848634135676804,
+      "learning_rate": 2.380638350808301e-07,
+      "loss": 0.1166,
+      "step": 5068
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.030302418548135,
+      "learning_rate": 2.371834498698089e-07,
+      "loss": 0.0978,
+      "step": 5069
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 6.902607563471164,
+      "learning_rate": 2.363046560043264e-07,
+      "loss": 0.1177,
+      "step": 5070
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.432084198181119,
+      "learning_rate": 2.3542745377800046e-07,
+      "loss": 0.0757,
+      "step": 5071
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.034887652754706,
+      "learning_rate": 2.3455184348392446e-07,
+      "loss": 0.0648,
+      "step": 5072
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 6.045491895151377,
+      "learning_rate": 2.3367782541465268e-07,
+      "loss": 0.1531,
+      "step": 5073
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 10.244042229911464,
+      "learning_rate": 2.3280539986221317e-07,
+      "loss": 0.1196,
+      "step": 5074
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.67069532901348,
+      "learning_rate": 2.3193456711809837e-07,
+      "loss": 0.1213,
+      "step": 5075
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 5.294449192222114,
+      "learning_rate": 2.3106532747327104e-07,
+      "loss": 0.0883,
+      "step": 5076
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 9.311678585887437,
+      "learning_rate": 2.3019768121815777e-07,
+      "loss": 0.1554,
+      "step": 5077
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 20.638914783602576,
+      "learning_rate": 2.2933162864265836e-07,
+      "loss": 0.1319,
+      "step": 5078
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 9.5037044607945,
+      "learning_rate": 2.2846717003613462e-07,
+      "loss": 0.1365,
+      "step": 5079
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.275015466909017,
+      "learning_rate": 2.2760430568741943e-07,
+      "loss": 0.0837,
+      "step": 5080
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 8.731705867976217,
+      "learning_rate": 2.2674303588481162e-07,
+      "loss": 0.1273,
+      "step": 5081
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 9.145995314417187,
+      "learning_rate": 2.258833609160771e-07,
+      "loss": 0.1223,
+      "step": 5082
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 6.35548689524532,
+      "learning_rate": 2.2502528106845e-07,
+      "loss": 0.1062,
+      "step": 5083
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 8.025136625142972,
+      "learning_rate": 2.241687966286299e-07,
+      "loss": 0.1373,
+      "step": 5084
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.627138376379124,
+      "learning_rate": 2.233139078827845e-07,
+      "loss": 0.0949,
+      "step": 5085
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.550861440722993,
+      "learning_rate": 2.2246061511654816e-07,
+      "loss": 0.0856,
+      "step": 5086
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 7.0832967651621646,
+      "learning_rate": 2.2160891861502165e-07,
+      "loss": 0.1456,
+      "step": 5087
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.494887892207766,
+      "learning_rate": 2.2075881866277348e-07,
+      "loss": 0.1138,
+      "step": 5088
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.653845807621733,
+      "learning_rate": 2.199103155438359e-07,
+      "loss": 0.149,
+      "step": 5089
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 7.533091223076104,
+      "learning_rate": 2.1906340954171212e-07,
+      "loss": 0.1851,
+      "step": 5090
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 6.574364381103182,
+      "learning_rate": 2.1821810093936636e-07,
+      "loss": 0.1305,
+      "step": 5091
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.576468541271489,
+      "learning_rate": 2.1737439001923488e-07,
+      "loss": 0.0978,
+      "step": 5092
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 10.54344185381927,
+      "learning_rate": 2.1653227706321388e-07,
+      "loss": 0.2559,
+      "step": 5093
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.76013310329922,
+      "learning_rate": 2.156917623526722e-07,
+      "loss": 0.1378,
+      "step": 5094
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.096648775601011,
+      "learning_rate": 2.1485284616843904e-07,
+      "loss": 0.0791,
+      "step": 5095
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 7.449715752312049,
+      "learning_rate": 2.140155287908141e-07,
+      "loss": 0.1338,
+      "step": 5096
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.19787476290784,
+      "learning_rate": 2.131798104995586e-07,
+      "loss": 0.1239,
+      "step": 5097
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.365081910281841,
+      "learning_rate": 2.123456915739025e-07,
+      "loss": 0.0817,
+      "step": 5098
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.7839621749109735,
+      "learning_rate": 2.115131722925401e-07,
+      "loss": 0.1301,
+      "step": 5099
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 7.328264787341496,
+      "learning_rate": 2.1068225293363166e-07,
+      "loss": 0.1644,
+      "step": 5100
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.248602303859766,
+      "learning_rate": 2.0985293377480342e-07,
+      "loss": 0.0967,
+      "step": 5101
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.184726044484055,
+      "learning_rate": 2.0902521509314543e-07,
+      "loss": 0.1346,
+      "step": 5102
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.470435895771347,
+      "learning_rate": 2.0819909716521426e-07,
+      "loss": 0.1361,
+      "step": 5103
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.351606542505066,
+      "learning_rate": 2.0737458026703182e-07,
+      "loss": 0.113,
+      "step": 5104
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.701711181172379,
+      "learning_rate": 2.0655166467408283e-07,
+      "loss": 0.1284,
+      "step": 5105
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.303422896520405,
+      "learning_rate": 2.057303506613212e-07,
+      "loss": 0.1362,
+      "step": 5106
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 6.235708485944981,
+      "learning_rate": 2.049106385031602e-07,
+      "loss": 0.131,
+      "step": 5107
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 8.200959008051413,
+      "learning_rate": 2.0409252847348404e-07,
+      "loss": 0.1003,
+      "step": 5108
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 6.76348332083974,
+      "learning_rate": 2.032760208456358e-07,
+      "loss": 0.0938,
+      "step": 5109
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 10.725542818280745,
+      "learning_rate": 2.0246111589242835e-07,
+      "loss": 0.1349,
+      "step": 5110
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 6.858265680340651,
+      "learning_rate": 2.0164781388613386e-07,
+      "loss": 0.1703,
+      "step": 5111
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 7.221302184416432,
+      "learning_rate": 2.0083611509849443e-07,
+      "loss": 0.0912,
+      "step": 5112
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 6.110379248791678,
+      "learning_rate": 2.0002601980071145e-07,
+      "loss": 0.1471,
+      "step": 5113
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 9.264511652438255,
+      "learning_rate": 1.9921752826345397e-07,
+      "loss": 0.1936,
+      "step": 5114
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 9.3500365337071,
+      "learning_rate": 1.9841064075685367e-07,
+      "loss": 0.1804,
+      "step": 5115
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 7.0264249501411795,
+      "learning_rate": 1.9760535755050715e-07,
+      "loss": 0.0991,
+      "step": 5116
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.367028898943652,
+      "learning_rate": 1.9680167891347356e-07,
+      "loss": 0.0659,
+      "step": 5117
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 6.694350207237627,
+      "learning_rate": 1.9599960511427761e-07,
+      "loss": 0.1215,
+      "step": 5118
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 5.01614665495544,
+      "learning_rate": 1.9519913642090715e-07,
+      "loss": 0.0903,
+      "step": 5119
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 7.6950343847045115,
+      "learning_rate": 1.9440027310081323e-07,
+      "loss": 0.1492,
+      "step": 5120
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 13.035091459384203,
+      "learning_rate": 1.9360301542091065e-07,
+      "loss": 0.1222,
+      "step": 5121
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 8.006794034185171,
+      "learning_rate": 1.9280736364757912e-07,
+      "loss": 0.1417,
+      "step": 5122
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.606990746506003,
+      "learning_rate": 1.9201331804665934e-07,
+      "loss": 0.0756,
+      "step": 5123
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.174082855551437,
+      "learning_rate": 1.9122087888345798e-07,
+      "loss": 0.1083,
+      "step": 5124
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.662992780244101,
+      "learning_rate": 1.9043004642274266e-07,
+      "loss": 0.1226,
+      "step": 5125
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 7.587789426133568,
+      "learning_rate": 1.896408209287459e-07,
+      "loss": 0.14,
+      "step": 5126
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 11.678374790231842,
+      "learning_rate": 1.888532026651624e-07,
+      "loss": 0.1396,
+      "step": 5127
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.977244116302907,
+      "learning_rate": 1.880671918951499e-07,
+      "loss": 0.0615,
+      "step": 5128
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.1018897296285175,
+      "learning_rate": 1.8728278888132944e-07,
+      "loss": 0.1375,
+      "step": 5129
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 7.569430217114574,
+      "learning_rate": 1.864999938857842e-07,
+      "loss": 0.1296,
+      "step": 5130
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.084231927095735,
+      "learning_rate": 1.8571880717006218e-07,
+      "loss": 0.0961,
+      "step": 5131
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 7.006697709775796,
+      "learning_rate": 1.8493922899516902e-07,
+      "loss": 0.1226,
+      "step": 5132
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 7.603508544129782,
+      "learning_rate": 1.8416125962157971e-07,
+      "loss": 0.1912,
+      "step": 5133
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.0061495778600555,
+      "learning_rate": 1.8338489930922632e-07,
+      "loss": 0.1002,
+      "step": 5134
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 6.034224352944826,
+      "learning_rate": 1.8261014831750633e-07,
+      "loss": 0.105,
+      "step": 5135
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 6.209079940714725,
+      "learning_rate": 1.8183700690527717e-07,
+      "loss": 0.0973,
+      "step": 5136
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 7.222408389143445,
+      "learning_rate": 1.810654753308616e-07,
+      "loss": 0.1681,
+      "step": 5137
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 8.624917976457603,
+      "learning_rate": 1.8029555385204067e-07,
+      "loss": 0.1587,
+      "step": 5138
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 6.111127472026909,
+      "learning_rate": 1.795272427260608e-07,
+      "loss": 0.0766,
+      "step": 5139
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.316408256721683,
+      "learning_rate": 1.7876054220962835e-07,
+      "loss": 0.0811,
+      "step": 5140
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 6.678820447902406,
+      "learning_rate": 1.779954525589128e-07,
+      "loss": 0.165,
+      "step": 5141
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.925332943931374,
+      "learning_rate": 1.7723197402954419e-07,
+      "loss": 0.0615,
+      "step": 5142
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 8.13593332874638,
+      "learning_rate": 1.7647010687661558e-07,
+      "loss": 0.101,
+      "step": 5143
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.418752250867253,
+      "learning_rate": 1.757098513546801e-07,
+      "loss": 0.0764,
+      "step": 5144
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 13.33740238830133,
+      "learning_rate": 1.74951207717754e-07,
+      "loss": 0.1919,
+      "step": 5145
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.192594742264245,
+      "learning_rate": 1.7419417621931388e-07,
+      "loss": 0.0905,
+      "step": 5146
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 6.938927349266944,
+      "learning_rate": 1.7343875711229864e-07,
+      "loss": 0.1296,
+      "step": 5147
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.150904395129445,
+      "learning_rate": 1.7268495064910574e-07,
+      "loss": 0.0952,
+      "step": 5148
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.645643287522623,
+      "learning_rate": 1.719327570815993e-07,
+      "loss": 0.123,
+      "step": 5149
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 8.033024279828727,
+      "learning_rate": 1.711821766610977e-07,
+      "loss": 0.1221,
+      "step": 5150
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 7.151014646433906,
+      "learning_rate": 1.704332096383865e-07,
+      "loss": 0.144,
+      "step": 5151
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.751515538072794,
+      "learning_rate": 1.696858562637077e-07,
+      "loss": 0.1128,
+      "step": 5152
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 7.3260407768072815,
+      "learning_rate": 1.689401167867677e-07,
+      "loss": 0.137,
+      "step": 5153
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.967392348291816,
+      "learning_rate": 1.6819599145672993e-07,
+      "loss": 0.1198,
+      "step": 5154
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 9.119755770264783,
+      "learning_rate": 1.674534805222222e-07,
+      "loss": 0.1946,
+      "step": 5155
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 5.9494333508876265,
+      "learning_rate": 1.667125842313305e-07,
+      "loss": 0.0823,
+      "step": 5156
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 3.699974499969946,
+      "learning_rate": 1.6597330283160184e-07,
+      "loss": 0.0596,
+      "step": 5157
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 9.275525343827553,
+      "learning_rate": 1.6523563657004416e-07,
+      "loss": 0.1776,
+      "step": 5158
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 7.8398354915929325,
+      "learning_rate": 1.644995856931253e-07,
+      "loss": 0.1151,
+      "step": 5159
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 7.010095295632242,
+      "learning_rate": 1.6376515044677354e-07,
+      "loss": 0.1351,
+      "step": 5160
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 10.561996804977277,
+      "learning_rate": 1.630323310763776e-07,
+      "loss": 0.14,
+      "step": 5161
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 6.3885597894295945,
+      "learning_rate": 1.6230112782678608e-07,
+      "loss": 0.1088,
+      "step": 5162
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 7.464820972006126,
+      "learning_rate": 1.6157154094230744e-07,
+      "loss": 0.1079,
+      "step": 5163
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.489218484166322,
+      "learning_rate": 1.6084357066670997e-07,
+      "loss": 0.1548,
+      "step": 5164
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 7.166544166806824,
+      "learning_rate": 1.601172172432225e-07,
+      "loss": 0.0991,
+      "step": 5165
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 6.044756449618914,
+      "learning_rate": 1.5939248091453252e-07,
+      "loss": 0.1147,
+      "step": 5166
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 7.407020648139834,
+      "learning_rate": 1.5866936192278915e-07,
+      "loss": 0.1702,
+      "step": 5167
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 6.543503830650654,
+      "learning_rate": 1.5794786050959797e-07,
+      "loss": 0.1196,
+      "step": 5168
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.94521426336034,
+      "learning_rate": 1.5722797691602842e-07,
+      "loss": 0.1217,
+      "step": 5169
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 5.678678949600126,
+      "learning_rate": 1.5650971138260473e-07,
+      "loss": 0.1138,
+      "step": 5170
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 6.541004043762944,
+      "learning_rate": 1.5579306414931493e-07,
+      "loss": 0.1263,
+      "step": 5171
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.703199754397636,
+      "learning_rate": 1.5507803545560195e-07,
+      "loss": 0.1287,
+      "step": 5172
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 6.079487966585593,
+      "learning_rate": 1.543646255403719e-07,
+      "loss": 0.1176,
+      "step": 5173
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.247351544243093,
+      "learning_rate": 1.5365283464198743e-07,
+      "loss": 0.1897,
+      "step": 5174
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 9.843355071661746,
+      "learning_rate": 1.529426629982711e-07,
+      "loss": 0.1487,
+      "step": 5175
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 7.88124033472976,
+      "learning_rate": 1.5223411084650476e-07,
+      "loss": 0.1071,
+      "step": 5176
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 8.117452681694125,
+      "learning_rate": 1.5152717842342845e-07,
+      "loss": 0.1221,
+      "step": 5177
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 6.399799739093315,
+      "learning_rate": 1.5082186596524218e-07,
+      "loss": 0.124,
+      "step": 5178
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 8.196187923853193,
+      "learning_rate": 1.501181737076035e-07,
+      "loss": 0.1299,
+      "step": 5179
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 8.07108956974622,
+      "learning_rate": 1.4941610188562884e-07,
+      "loss": 0.1139,
+      "step": 5180
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 9.192620314889446,
+      "learning_rate": 1.4871565073389382e-07,
+      "loss": 0.1345,
+      "step": 5181
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 7.048174028646196,
+      "learning_rate": 1.4801682048643183e-07,
+      "loss": 0.1466,
+      "step": 5182
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 6.816161354114099,
+      "learning_rate": 1.4731961137673555e-07,
+      "loss": 0.1024,
+      "step": 5183
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 16.351201920865254,
+      "learning_rate": 1.466240236377553e-07,
+      "loss": 0.1097,
+      "step": 5184
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 7.299762715915274,
+      "learning_rate": 1.4593005750189958e-07,
+      "loss": 0.1072,
+      "step": 5185
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.167377084530258,
+      "learning_rate": 1.4523771320103574e-07,
+      "loss": 0.0734,
+      "step": 5186
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 6.275327813941812,
+      "learning_rate": 1.4454699096648873e-07,
+      "loss": 0.1506,
+      "step": 5187
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.996799451568316,
+      "learning_rate": 1.4385789102904168e-07,
+      "loss": 0.111,
+      "step": 5188
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 7.7803406635382215,
+      "learning_rate": 1.4317041361893546e-07,
+      "loss": 0.1682,
+      "step": 5189
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.665189120267458,
+      "learning_rate": 1.4248455896587022e-07,
+      "loss": 0.0935,
+      "step": 5190
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 8.005633628165455,
+      "learning_rate": 1.418003272990004e-07,
+      "loss": 0.1013,
+      "step": 5191
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 6.9506037573191515,
+      "learning_rate": 1.4111771884694315e-07,
+      "loss": 0.1016,
+      "step": 5192
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.8278679045020185,
+      "learning_rate": 1.4043673383776825e-07,
+      "loss": 0.0898,
+      "step": 5193
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 7.641464266293792,
+      "learning_rate": 1.3975737249900812e-07,
+      "loss": 0.1395,
+      "step": 5194
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 10.172133880539453,
+      "learning_rate": 1.3907963505764731e-07,
+      "loss": 0.1418,
+      "step": 5195
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 8.166463290616882,
+      "learning_rate": 1.384035217401325e-07,
+      "loss": 0.1249,
+      "step": 5196
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 7.385868049756229,
+      "learning_rate": 1.3772903277236404e-07,
+      "loss": 0.1636,
+      "step": 5197
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 7.952994717184201,
+      "learning_rate": 1.370561683797028e-07,
+      "loss": 0.1356,
+      "step": 5198
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 10.124774749021338,
+      "learning_rate": 1.363849287869645e-07,
+      "loss": 0.1481,
+      "step": 5199
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 6.320488348773809,
+      "learning_rate": 1.3571531421842256e-07,
+      "loss": 0.106,
+      "step": 5200
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 6.804058988616448,
+      "learning_rate": 1.3504732489780849e-07,
+      "loss": 0.114,
+      "step": 5201
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 7.514633914232519,
+      "learning_rate": 1.3438096104830879e-07,
+      "loss": 0.1368,
+      "step": 5202
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.702232298090348,
+      "learning_rate": 1.3371622289256869e-07,
+      "loss": 0.1162,
+      "step": 5203
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 4.859181618551624,
+      "learning_rate": 1.3305311065269e-07,
+      "loss": 0.0895,
+      "step": 5204
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 10.143736353282302,
+      "learning_rate": 1.323916245502299e-07,
+      "loss": 0.1407,
+      "step": 5205
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.899522459445874,
+      "learning_rate": 1.3173176480620442e-07,
+      "loss": 0.1295,
+      "step": 5206
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.406089544140982,
+      "learning_rate": 1.3107353164108273e-07,
+      "loss": 0.094,
+      "step": 5207
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.286668556628162,
+      "learning_rate": 1.3041692527479556e-07,
+      "loss": 0.125,
+      "step": 5208
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.2593888521125,
+      "learning_rate": 1.2976194592672465e-07,
+      "loss": 0.1161,
+      "step": 5209
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 10.413500581443682,
+      "learning_rate": 1.2910859381571327e-07,
+      "loss": 0.1522,
+      "step": 5210
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 7.838444174981346,
+      "learning_rate": 1.284568691600563e-07,
+      "loss": 0.1145,
+      "step": 5211
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.995371120181053,
+      "learning_rate": 1.2780677217750949e-07,
+      "loss": 0.1064,
+      "step": 5212
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 7.572479545934153,
+      "learning_rate": 1.271583030852791e-07,
+      "loss": 0.1183,
+      "step": 5213
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.113754497506733,
+      "learning_rate": 1.2651146210003406e-07,
+      "loss": 0.1029,
+      "step": 5214
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.198270804458642,
+      "learning_rate": 1.2586624943789372e-07,
+      "loss": 0.1324,
+      "step": 5215
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 7.473074709137638,
+      "learning_rate": 1.2522266531443616e-07,
+      "loss": 0.1241,
+      "step": 5216
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 13.207193018450559,
+      "learning_rate": 1.245807099446955e-07,
+      "loss": 0.1882,
+      "step": 5217
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.335816731636773,
+      "learning_rate": 1.239403835431602e-07,
+      "loss": 0.0775,
+      "step": 5218
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 10.620746803482923,
+      "learning_rate": 1.2330168632377514e-07,
+      "loss": 0.1576,
+      "step": 5219
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.84388438837485,
+      "learning_rate": 1.2266461849994138e-07,
+      "loss": 0.1257,
+      "step": 5220
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.3623646823821565,
+      "learning_rate": 1.2202918028451527e-07,
+      "loss": 0.1428,
+      "step": 5221
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.362173666685328,
+      "learning_rate": 1.2139537188980753e-07,
+      "loss": 0.1882,
+      "step": 5222
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 5.891116663875299,
+      "learning_rate": 1.207631935275866e-07,
+      "loss": 0.1367,
+      "step": 5223
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 5.184386123808496,
+      "learning_rate": 1.2013264540907455e-07,
+      "loss": 0.1078,
+      "step": 5224
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 6.019083925984621,
+      "learning_rate": 1.1950372774494846e-07,
+      "loss": 0.0882,
+      "step": 5225
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 6.246318195836614,
+      "learning_rate": 1.1887644074534244e-07,
+      "loss": 0.1174,
+      "step": 5226
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.847553205535199,
+      "learning_rate": 1.182507846198444e-07,
+      "loss": 0.1487,
+      "step": 5227
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 11.611646202895807,
+      "learning_rate": 1.1762675957749769e-07,
+      "loss": 0.2279,
+      "step": 5228
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 5.737661801161129,
+      "learning_rate": 1.1700436582680108e-07,
+      "loss": 0.1067,
+      "step": 5229
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 8.299874718836872,
+      "learning_rate": 1.1638360357570654e-07,
+      "loss": 0.1802,
+      "step": 5230
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.031326583486857,
+      "learning_rate": 1.157644730316243e-07,
+      "loss": 0.115,
+      "step": 5231
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 6.615993043157198,
+      "learning_rate": 1.1514697440141498e-07,
+      "loss": 0.116,
+      "step": 5232
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.184865547249952,
+      "learning_rate": 1.1453110789139855e-07,
+      "loss": 0.1007,
+      "step": 5233
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.274144565026902,
+      "learning_rate": 1.1391687370734594e-07,
+      "loss": 0.1247,
+      "step": 5234
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 8.242541318513114,
+      "learning_rate": 1.1330427205448579e-07,
+      "loss": 0.1362,
+      "step": 5235
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 12.150162498817286,
+      "learning_rate": 1.1269330313749715e-07,
+      "loss": 0.1766,
+      "step": 5236
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.308670112443291,
+      "learning_rate": 1.1208396716051895e-07,
+      "loss": 0.1266,
+      "step": 5237
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.6392974652541605,
+      "learning_rate": 1.1147626432713943e-07,
+      "loss": 0.0908,
+      "step": 5238
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 7.564703684509811,
+      "learning_rate": 1.1087019484040562e-07,
+      "loss": 0.1458,
+      "step": 5239
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 7.470040083902322,
+      "learning_rate": 1.1026575890281443e-07,
+      "loss": 0.1246,
+      "step": 5240
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.086766086218585,
+      "learning_rate": 1.0966295671632043e-07,
+      "loss": 0.1151,
+      "step": 5241
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.4424206168832585,
+      "learning_rate": 1.0906178848233029e-07,
+      "loss": 0.0956,
+      "step": 5242
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 7.511100671159325,
+      "learning_rate": 1.0846225440170611e-07,
+      "loss": 0.1285,
+      "step": 5243
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 5.8044703706351815,
+      "learning_rate": 1.0786435467476264e-07,
+      "loss": 0.1116,
+      "step": 5244
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 5.525857288049917,
+      "learning_rate": 1.072680895012701e-07,
+      "loss": 0.1162,
+      "step": 5245
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.722086445426205,
+      "learning_rate": 1.0667345908045135e-07,
+      "loss": 0.1187,
+      "step": 5246
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.848587661746584,
+      "learning_rate": 1.0608046361098356e-07,
+      "loss": 0.1238,
+      "step": 5247
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 8.613934378361632,
+      "learning_rate": 1.0548910329099771e-07,
+      "loss": 0.1035,
+      "step": 5248
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.758696950894796,
+      "learning_rate": 1.048993783180785e-07,
+      "loss": 0.1001,
+      "step": 5249
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 11.99570441487423,
+      "learning_rate": 1.0431128888926222e-07,
+      "loss": 0.1798,
+      "step": 5250
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 9.68337979709574,
+      "learning_rate": 1.0372483520104337e-07,
+      "loss": 0.2222,
+      "step": 5251
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 9.159716336335311,
+      "learning_rate": 1.0314001744936409e-07,
+      "loss": 0.1133,
+      "step": 5252
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 12.181529893720981,
+      "learning_rate": 1.0255683582962583e-07,
+      "loss": 0.1626,
+      "step": 5253
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.799677606632576,
+      "learning_rate": 1.0197529053667721e-07,
+      "loss": 0.0961,
+      "step": 5254
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 6.57621765664244,
+      "learning_rate": 1.013953817648261e-07,
+      "loss": 0.1383,
+      "step": 5255
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 6.189186863181362,
+      "learning_rate": 1.008171097078292e-07,
+      "loss": 0.1518,
+      "step": 5256
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 12.319055223640666,
+      "learning_rate": 1.0024047455889918e-07,
+      "loss": 0.1925,
+      "step": 5257
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 4.369429730589778,
+      "learning_rate": 9.966547651069913e-08,
+      "loss": 0.0916,
+      "step": 5258
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 7.524207259807521,
+      "learning_rate": 9.909211575534705e-08,
+      "loss": 0.132,
+      "step": 5259
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 7.236946977446954,
+      "learning_rate": 9.852039248441414e-08,
+      "loss": 0.1023,
+      "step": 5260
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 9.47859230035742,
+      "learning_rate": 9.79503068889226e-08,
+      "loss": 0.1814,
+      "step": 5261
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 8.758267109966468,
+      "learning_rate": 9.738185915935005e-08,
+      "loss": 0.175,
+      "step": 5262
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 9.27801847977357,
+      "learning_rate": 9.681504948562403e-08,
+      "loss": 0.0741,
+      "step": 5263
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.229611351433055,
+      "learning_rate": 9.624987805712749e-08,
+      "loss": 0.1229,
+      "step": 5264
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 7.05122561410525,
+      "learning_rate": 9.568634506269381e-08,
+      "loss": 0.1503,
+      "step": 5265
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 6.437579163294752,
+      "learning_rate": 9.51244506906096e-08,
+      "loss": 0.0978,
+      "step": 5266
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.629315173299733,
+      "learning_rate": 9.45641951286158e-08,
+      "loss": 0.076,
+      "step": 5267
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 12.414008878645971,
+      "learning_rate": 9.400557856390158e-08,
+      "loss": 0.1239,
+      "step": 5268
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 7.86659051015309,
+      "learning_rate": 9.344860118311427e-08,
+      "loss": 0.1056,
+      "step": 5269
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.829098697103711,
+      "learning_rate": 9.289326317234726e-08,
+      "loss": 0.114,
+      "step": 5270
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.748563587949899,
+      "learning_rate": 9.23395647171521e-08,
+      "loss": 0.1136,
+      "step": 5271
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 9.201838202910588,
+      "learning_rate": 9.178750600252695e-08,
+      "loss": 0.1781,
+      "step": 5272
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 10.329283760051627,
+      "learning_rate": 9.123708721292756e-08,
+      "loss": 0.1503,
+      "step": 5273
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.90976125143801,
+      "learning_rate": 9.06883085322574e-08,
+      "loss": 0.0987,
+      "step": 5274
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 9.31051888219555,
+      "learning_rate": 9.014117014387424e-08,
+      "loss": 0.1438,
+      "step": 5275
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 5.3011211156186935,
+      "learning_rate": 8.95956722305874e-08,
+      "loss": 0.1042,
+      "step": 5276
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 5.25084946851794,
+      "learning_rate": 8.905181497465664e-08,
+      "loss": 0.1144,
+      "step": 5277
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 12.014268472267695,
+      "learning_rate": 8.850959855779662e-08,
+      "loss": 0.1437,
+      "step": 5278
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 6.3680562353578,
+      "learning_rate": 8.796902316117018e-08,
+      "loss": 0.0974,
+      "step": 5279
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 8.104205605705294,
+      "learning_rate": 8.743008896539451e-08,
+      "loss": 0.1185,
+      "step": 5280
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 5.822397728872039,
+      "learning_rate": 8.68927961505378e-08,
+      "loss": 0.0966,
+      "step": 5281
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 6.962525356143755,
+      "learning_rate": 8.635714489611868e-08,
+      "loss": 0.1772,
+      "step": 5282
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.4875000040030812,
+      "learning_rate": 8.582313538110898e-08,
+      "loss": 0.0679,
+      "step": 5283
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 5.751258070965259,
+      "learning_rate": 8.529076778393097e-08,
+      "loss": 0.1108,
+      "step": 5284
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 22.636631698978753,
+      "learning_rate": 8.476004228245848e-08,
+      "loss": 0.2013,
+      "step": 5285
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 6.386527686790105,
+      "learning_rate": 8.42309590540169e-08,
+      "loss": 0.1111,
+      "step": 5286
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 7.245454822317008,
+      "learning_rate": 8.370351827538259e-08,
+      "loss": 0.1567,
+      "step": 5287
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 7.696534926561681,
+      "learning_rate": 8.317772012278347e-08,
+      "loss": 0.0975,
+      "step": 5288
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 7.159676601821471,
+      "learning_rate": 8.26535647718979e-08,
+      "loss": 0.1014,
+      "step": 5289
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 7.323349288176798,
+      "learning_rate": 8.213105239785691e-08,
+      "loss": 0.121,
+      "step": 5290
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 6.626972303043425,
+      "learning_rate": 8.161018317524139e-08,
+      "loss": 0.1101,
+      "step": 5291
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 9.572853627923612,
+      "learning_rate": 8.109095727808269e-08,
+      "loss": 0.1646,
+      "step": 5292
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 6.1048890627998045,
+      "learning_rate": 8.057337487986427e-08,
+      "loss": 0.1073,
+      "step": 5293
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.486123430984657,
+      "learning_rate": 8.005743615352057e-08,
+      "loss": 0.1229,
+      "step": 5294
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 10.624680701422056,
+      "learning_rate": 7.954314127143481e-08,
+      "loss": 0.1277,
+      "step": 5295
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 9.68142575432856,
+      "learning_rate": 7.903049040544453e-08,
+      "loss": 0.1707,
+      "step": 5296
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 5.035336991397328,
+      "learning_rate": 7.851948372683382e-08,
+      "loss": 0.1169,
+      "step": 5297
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 6.927256106603059,
+      "learning_rate": 7.801012140634167e-08,
+      "loss": 0.1127,
+      "step": 5298
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.320357114365319,
+      "learning_rate": 7.750240361415362e-08,
+      "loss": 0.1432,
+      "step": 5299
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.895191827361518,
+      "learning_rate": 7.69963305199084e-08,
+      "loss": 0.1588,
+      "step": 5300
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.444645977840397,
+      "learning_rate": 7.64919022926941e-08,
+      "loss": 0.1382,
+      "step": 5301
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.356571405819423,
+      "learning_rate": 7.598911910105033e-08,
+      "loss": 0.1312,
+      "step": 5302
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 9.29974473920337,
+      "learning_rate": 7.548798111296552e-08,
+      "loss": 0.1755,
+      "step": 5303
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.36759012920459,
+      "learning_rate": 7.498848849588015e-08,
+      "loss": 0.1384,
+      "step": 5304
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.686410092064511,
+      "learning_rate": 7.449064141668238e-08,
+      "loss": 0.0949,
+      "step": 5305
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.1045580573583695,
+      "learning_rate": 7.399444004171364e-08,
+      "loss": 0.1318,
+      "step": 5306
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.181571610180925,
+      "learning_rate": 7.349988453676349e-08,
+      "loss": 0.1101,
+      "step": 5307
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.948354119093815,
+      "learning_rate": 7.300697506707254e-08,
+      "loss": 0.1398,
+      "step": 5308
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.1083301561778685,
+      "learning_rate": 7.251571179732963e-08,
+      "loss": 0.1018,
+      "step": 5309
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.193753633726565,
+      "learning_rate": 7.202609489167734e-08,
+      "loss": 0.1401,
+      "step": 5310
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 6.971897988354667,
+      "learning_rate": 7.153812451370312e-08,
+      "loss": 0.1178,
+      "step": 5311
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 8.547983835503912,
+      "learning_rate": 7.10518008264488e-08,
+      "loss": 0.1567,
+      "step": 5312
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.901556010148924,
+      "learning_rate": 7.056712399240274e-08,
+      "loss": 0.0923,
+      "step": 5313
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.1229350172421615,
+      "learning_rate": 7.008409417350648e-08,
+      "loss": 0.1046,
+      "step": 5314
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 21.260270101970896,
+      "learning_rate": 6.960271153114706e-08,
+      "loss": 0.279,
+      "step": 5315
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.805978363368682,
+      "learning_rate": 6.912297622616526e-08,
+      "loss": 0.0815,
+      "step": 5316
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 11.926823320423209,
+      "learning_rate": 6.864488841884786e-08,
+      "loss": 0.1357,
+      "step": 5317
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.122540845339858,
+      "learning_rate": 6.816844826893431e-08,
+      "loss": 0.1118,
+      "step": 5318
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 8.90771512298503,
+      "learning_rate": 6.769365593561117e-08,
+      "loss": 0.1603,
+      "step": 5319
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.151048217047981,
+      "learning_rate": 6.722051157751597e-08,
+      "loss": 0.0963,
+      "step": 5320
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 7.657265995185122,
+      "learning_rate": 6.674901535273448e-08,
+      "loss": 0.101,
+      "step": 5321
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.07689362939208,
+      "learning_rate": 6.627916741880291e-08,
+      "loss": 0.1242,
+      "step": 5322
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.1248540676091565,
+      "learning_rate": 6.581096793270625e-08,
+      "loss": 0.073,
+      "step": 5323
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 9.295501650534323,
+      "learning_rate": 6.534441705087768e-08,
+      "loss": 0.1558,
+      "step": 5324
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 14.16116865988068,
+      "learning_rate": 6.487951492920141e-08,
+      "loss": 0.1665,
+      "step": 5325
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.979231240584932,
+      "learning_rate": 6.441626172300986e-08,
+      "loss": 0.149,
+      "step": 5326
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 8.526966234599504,
+      "learning_rate": 6.395465758708419e-08,
+      "loss": 0.1081,
+      "step": 5327
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 13.410360679446114,
+      "learning_rate": 6.349470267565549e-08,
+      "loss": 0.1249,
+      "step": 5328
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.87362720959621,
+      "learning_rate": 6.303639714240196e-08,
+      "loss": 0.1234,
+      "step": 5329
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.4425196121684865,
+      "learning_rate": 6.257974114045385e-08,
+      "loss": 0.0964,
+      "step": 5330
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.646662279495886,
+      "learning_rate": 6.212473482238635e-08,
+      "loss": 0.1497,
+      "step": 5331
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.057643625550772,
+      "learning_rate": 6.167137834022785e-08,
+      "loss": 0.0977,
+      "step": 5332
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.539793036897762,
+      "learning_rate": 6.121967184545107e-08,
+      "loss": 0.1141,
+      "step": 5333
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.761339793280051,
+      "learning_rate": 6.076961548898086e-08,
+      "loss": 0.0969,
+      "step": 5334
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.800019097577064,
+      "learning_rate": 6.032120942118858e-08,
+      "loss": 0.1007,
+      "step": 5335
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.289423884200621,
+      "learning_rate": 5.98744537918955e-08,
+      "loss": 0.0994,
+      "step": 5336
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.500424630787519,
+      "learning_rate": 5.9429348750371097e-08,
+      "loss": 0.1061,
+      "step": 5337
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 7.843475544162,
+      "learning_rate": 5.898589444533254e-08,
+      "loss": 0.146,
+      "step": 5338
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 11.118806698310935,
+      "learning_rate": 5.85440910249474e-08,
+      "loss": 0.1675,
+      "step": 5339
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 7.320211335379985,
+      "learning_rate": 5.810393863682873e-08,
+      "loss": 0.1436,
+      "step": 5340
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.079462595694946,
+      "learning_rate": 5.7665437428041096e-08,
+      "loss": 0.1572,
+      "step": 5341
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 7.477284639355565,
+      "learning_rate": 5.722858754509564e-08,
+      "loss": 0.1337,
+      "step": 5342
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 5.941412410817457,
+      "learning_rate": 5.679338913395116e-08,
+      "loss": 0.0836,
+      "step": 5343
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 15.49462238075577,
+      "learning_rate": 5.6359842340016904e-08,
+      "loss": 0.168,
+      "step": 5344
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 7.814014540138655,
+      "learning_rate": 5.5927947308147545e-08,
+      "loss": 0.0989,
+      "step": 5345
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 6.40539018080863,
+      "learning_rate": 5.549770418264766e-08,
+      "loss": 0.0836,
+      "step": 5346
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 8.444026783490271,
+      "learning_rate": 5.5069113107270034e-08,
+      "loss": 0.1371,
+      "step": 5347
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 6.902013975947065,
+      "learning_rate": 5.464217422521456e-08,
+      "loss": 0.0989,
+      "step": 5348
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.274615043574269,
+      "learning_rate": 5.421688767912936e-08,
+      "loss": 0.0985,
+      "step": 5349
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 5.836113346579141,
+      "learning_rate": 5.3793253611110206e-08,
+      "loss": 0.0904,
+      "step": 5350
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 10.277038370643657,
+      "learning_rate": 5.3371272162702214e-08,
+      "loss": 0.1272,
+      "step": 5351
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 7.427580721132972,
+      "learning_rate": 5.295094347489593e-08,
+      "loss": 0.1472,
+      "step": 5352
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 5.647665876407714,
+      "learning_rate": 5.253226768813235e-08,
+      "loss": 0.0901,
+      "step": 5353
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 9.858056079098924,
+      "learning_rate": 5.211524494229736e-08,
+      "loss": 0.1442,
+      "step": 5354
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 7.289792068781541,
+      "learning_rate": 5.169987537672727e-08,
+      "loss": 0.135,
+      "step": 5355
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 8.232332580344814,
+      "learning_rate": 5.128615913020385e-08,
+      "loss": 0.1414,
+      "step": 5356
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 6.759511421442111,
+      "learning_rate": 5.087409634095819e-08,
+      "loss": 0.1025,
+      "step": 5357
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 9.901571938672475,
+      "learning_rate": 5.046368714666683e-08,
+      "loss": 0.1069,
+      "step": 5358
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 6.677195791425341,
+      "learning_rate": 5.0054931684457296e-08,
+      "loss": 0.1721,
+      "step": 5359
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.5234884387987,
+      "learning_rate": 4.964783009090035e-08,
+      "loss": 0.1272,
+      "step": 5360
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.1565601746836665,
+      "learning_rate": 4.9242382502017185e-08,
+      "loss": 0.1442,
+      "step": 5361
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 6.612248876782069,
+      "learning_rate": 4.883858905327499e-08,
+      "loss": 0.1156,
+      "step": 5362
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 6.653265623648825,
+      "learning_rate": 4.843644987958862e-08,
+      "loss": 0.109,
+      "step": 5363
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.539454475937336,
+      "learning_rate": 4.8035965115320604e-08,
+      "loss": 0.1241,
+      "step": 5364
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 9.420747332016814,
+      "learning_rate": 4.763713489428001e-08,
+      "loss": 0.121,
+      "step": 5365
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.7408741621782,
+      "learning_rate": 4.723995934972414e-08,
+      "loss": 0.1077,
+      "step": 5366
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.081059176994714,
+      "learning_rate": 4.684443861435572e-08,
+      "loss": 0.1456,
+      "step": 5367
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 12.657636446362256,
+      "learning_rate": 4.6450572820325727e-08,
+      "loss": 0.1986,
+      "step": 5368
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.913660788091783,
+      "learning_rate": 4.605836209923331e-08,
+      "loss": 0.1097,
+      "step": 5369
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.210402941111909,
+      "learning_rate": 4.566780658212144e-08,
+      "loss": 0.1443,
+      "step": 5370
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 5.194586966451776,
+      "learning_rate": 4.5278906399483516e-08,
+      "loss": 0.0763,
+      "step": 5371
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 7.14924387954306,
+      "learning_rate": 4.489166168125725e-08,
+      "loss": 0.1792,
+      "step": 5372
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 9.147775379876485,
+      "learning_rate": 4.4506072556829704e-08,
+      "loss": 0.1053,
+      "step": 5373
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 10.730163857110522,
+      "learning_rate": 4.4122139155031717e-08,
+      "loss": 0.0945,
+      "step": 5374
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 9.80794922773675,
+      "learning_rate": 4.373986160414345e-08,
+      "loss": 0.117,
+      "step": 5375
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 5.52024439305837,
+      "learning_rate": 4.335924003189107e-08,
+      "loss": 0.0768,
+      "step": 5376
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 6.642666235134706,
+      "learning_rate": 4.298027456544674e-08,
+      "loss": 0.0894,
+      "step": 5377
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 8.033102864873747,
+      "learning_rate": 4.260296533143027e-08,
+      "loss": 0.1226,
+      "step": 5378
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 7.958363936548108,
+      "learning_rate": 4.22273124559075e-08,
+      "loss": 0.1164,
+      "step": 5379
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 6.064455639298206,
+      "learning_rate": 4.185331606439136e-08,
+      "loss": 0.116,
+      "step": 5380
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 7.2571709246078715,
+      "learning_rate": 4.148097628184078e-08,
+      "loss": 0.1587,
+      "step": 5381
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 5.531175996724435,
+      "learning_rate": 4.111029323266125e-08,
+      "loss": 0.1157,
+      "step": 5382
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.8346388590651186,
+      "learning_rate": 4.07412670407048e-08,
+      "loss": 0.0787,
+      "step": 5383
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 15.590850711055705,
+      "learning_rate": 4.037389782927059e-08,
+      "loss": 0.222,
+      "step": 5384
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 9.156488520749987,
+      "learning_rate": 4.000818572110265e-08,
+      "loss": 0.1475,
+      "step": 5385
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 6.095122261021763,
+      "learning_rate": 3.964413083839269e-08,
+      "loss": 0.0922,
+      "step": 5386
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 7.495658799999846,
+      "learning_rate": 3.9281733302778404e-08,
+      "loss": 0.099,
+      "step": 5387
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 6.902837764596447,
+      "learning_rate": 3.892099323534293e-08,
+      "loss": 0.1339,
+      "step": 5388
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 5.998154833344351,
+      "learning_rate": 3.856191075661708e-08,
+      "loss": 0.0916,
+      "step": 5389
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.054604203914527,
+      "learning_rate": 3.8204485986576e-08,
+      "loss": 0.1375,
+      "step": 5390
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 5.718326581624709,
+      "learning_rate": 3.784871904464249e-08,
+      "loss": 0.1193,
+      "step": 5391
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 6.0495502960815655,
+      "learning_rate": 3.7494610049684796e-08,
+      "loss": 0.1162,
+      "step": 5392
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 10.193965675630396,
+      "learning_rate": 3.714215912001773e-08,
+      "loss": 0.1565,
+      "step": 5393
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 5.735370222498481,
+      "learning_rate": 3.6791366373400974e-08,
+      "loss": 0.0824,
+      "step": 5394
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.796648636559523,
+      "learning_rate": 3.6442231927041324e-08,
+      "loss": 0.118,
+      "step": 5395
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.101682970617409,
+      "learning_rate": 3.609475589759104e-08,
+      "loss": 0.1078,
+      "step": 5396
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 6.642030857014743,
+      "learning_rate": 3.574893840114835e-08,
+      "loss": 0.1002,
+      "step": 5397
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 6.008978983404813,
+      "learning_rate": 3.5404779553257494e-08,
+      "loss": 0.1146,
+      "step": 5398
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.787617973801435,
+      "learning_rate": 3.506227946890761e-08,
+      "loss": 0.209,
+      "step": 5399
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 5.136877676664234,
+      "learning_rate": 3.4721438262534935e-08,
+      "loss": 0.0845,
+      "step": 5400
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.842953334692523,
+      "learning_rate": 3.438225604802115e-08,
+      "loss": 0.1489,
+      "step": 5401
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 5.907887079400435,
+      "learning_rate": 3.404473293869226e-08,
+      "loss": 0.1051,
+      "step": 5402
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 8.675482010996637,
+      "learning_rate": 3.370886904732196e-08,
+      "loss": 0.1149,
+      "step": 5403
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 5.556595931410459,
+      "learning_rate": 3.33746644861288e-08,
+      "loss": 0.0774,
+      "step": 5404
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 7.278531821366081,
+      "learning_rate": 3.30421193667757e-08,
+      "loss": 0.1306,
+      "step": 5405
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 5.349133272872112,
+      "learning_rate": 3.271123380037322e-08,
+      "loss": 0.0956,
+      "step": 5406
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.334550501760739,
+      "learning_rate": 3.2382007897475695e-08,
+      "loss": 0.0842,
+      "step": 5407
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.826027471570106,
+      "learning_rate": 3.2054441768083477e-08,
+      "loss": 0.1401,
+      "step": 5408
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.591080668686281,
+      "learning_rate": 3.1728535521643454e-08,
+      "loss": 0.1153,
+      "step": 5409
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 4.910777250160643,
+      "learning_rate": 3.1404289267046305e-08,
+      "loss": 0.0955,
+      "step": 5410
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.193743146559833,
+      "learning_rate": 3.1081703112628146e-08,
+      "loss": 0.1156,
+      "step": 5411
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 4.131743099975989,
+      "learning_rate": 3.0760777166172206e-08,
+      "loss": 0.0914,
+      "step": 5412
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 9.147186309625447,
+      "learning_rate": 3.0441511534904934e-08,
+      "loss": 0.186,
+      "step": 5413
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 16.89801450783667,
+      "learning_rate": 3.012390632549933e-08,
+      "loss": 0.2162,
+      "step": 5414
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 7.065640199939804,
+      "learning_rate": 2.9807961644073294e-08,
+      "loss": 0.1641,
+      "step": 5415
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 5.507514754630227,
+      "learning_rate": 2.9493677596189595e-08,
+      "loss": 0.1292,
+      "step": 5416
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.253875594598688,
+      "learning_rate": 2.9181054286855916e-08,
+      "loss": 0.1044,
+      "step": 5417
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 9.876290985174975,
+      "learning_rate": 2.887009182052647e-08,
+      "loss": 0.2092,
+      "step": 5418
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 8.262022252186037,
+      "learning_rate": 2.8560790301098705e-08,
+      "loss": 0.1256,
+      "step": 5419
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.444790652855763,
+      "learning_rate": 2.825314983191718e-08,
+      "loss": 0.0795,
+      "step": 5420
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 5.034772326384623,
+      "learning_rate": 2.7947170515768562e-08,
+      "loss": 0.1035,
+      "step": 5421
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 8.398317634313441,
+      "learning_rate": 2.7642852454887736e-08,
+      "loss": 0.1326,
+      "step": 5422
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.212256085984667,
+      "learning_rate": 2.7340195750952813e-08,
+      "loss": 0.0872,
+      "step": 5423
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.414797855257343,
+      "learning_rate": 2.703920050508624e-08,
+      "loss": 0.1196,
+      "step": 5424
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.604779209237305,
+      "learning_rate": 2.673986681785645e-08,
+      "loss": 0.1197,
+      "step": 5425
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.618675757820333,
+      "learning_rate": 2.6442194789277342e-08,
+      "loss": 0.0822,
+      "step": 5426
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 10.808901861160757,
+      "learning_rate": 2.6146184518804908e-08,
+      "loss": 0.1245,
+      "step": 5427
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.610150315103935,
+      "learning_rate": 2.5851836105343363e-08,
+      "loss": 0.098,
+      "step": 5428
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 9.495465719612765,
+      "learning_rate": 2.555914964723849e-08,
+      "loss": 0.1179,
+      "step": 5429
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.207059094738852,
+      "learning_rate": 2.5268125242283724e-08,
+      "loss": 0.0848,
+      "step": 5430
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.450983131869945,
+      "learning_rate": 2.4978762987714067e-08,
+      "loss": 0.123,
+      "step": 5431
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 10.375350232266669,
+      "learning_rate": 2.469106298021273e-08,
+      "loss": 0.165,
+      "step": 5432
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.503010938323237,
+      "learning_rate": 2.4405025315904495e-08,
+      "loss": 0.1229,
+      "step": 5433
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 8.816882888336334,
+      "learning_rate": 2.412065009036013e-08,
+      "loss": 0.1213,
+      "step": 5434
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 5.889173798463627,
+      "learning_rate": 2.3837937398594747e-08,
+      "loss": 0.1117,
+      "step": 5435
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 8.450886362130994,
+      "learning_rate": 2.3556887335067223e-08,
+      "loss": 0.1573,
+      "step": 5436
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 7.633750300186745,
+      "learning_rate": 2.3277499993682452e-08,
+      "loss": 0.1022,
+      "step": 5437
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 5.677983573724589,
+      "learning_rate": 2.2999775467788532e-08,
+      "loss": 0.0836,
+      "step": 5438
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.121189112026573,
+      "learning_rate": 2.272371385017902e-08,
+      "loss": 0.1059,
+      "step": 5439
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.797540986482973,
+      "learning_rate": 2.244931523309013e-08,
+      "loss": 0.1263,
+      "step": 5440
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.379117284864385,
+      "learning_rate": 2.2176579708204636e-08,
+      "loss": 0.1324,
+      "step": 5441
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 10.852701175072118,
+      "learning_rate": 2.190550736664798e-08,
+      "loss": 0.2468,
+      "step": 5442
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 7.513796502846481,
+      "learning_rate": 2.163609829898994e-08,
+      "loss": 0.1409,
+      "step": 5443
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 9.610713700051518,
+      "learning_rate": 2.136835259524628e-08,
+      "loss": 0.1302,
+      "step": 5444
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 7.7294639743377545,
+      "learning_rate": 2.1102270344874887e-08,
+      "loss": 0.1514,
+      "step": 5445
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.143827292567964,
+      "learning_rate": 2.083785163677965e-08,
+      "loss": 0.0825,
+      "step": 5446
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.874115522509679,
+      "learning_rate": 2.0575096559306564e-08,
+      "loss": 0.1707,
+      "step": 5447
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.5660499922054685,
+      "learning_rate": 2.0314005200248178e-08,
+      "loss": 0.0833,
+      "step": 5448
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 6.291346107187886,
+      "learning_rate": 2.0054577646839156e-08,
+      "loss": 0.104,
+      "step": 5449
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 7.101207817053639,
+      "learning_rate": 1.979681398575961e-08,
+      "loss": 0.1224,
+      "step": 5450
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 8.288766121361911,
+      "learning_rate": 1.954071430313287e-08,
+      "loss": 0.1025,
+      "step": 5451
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 7.137866886658638,
+      "learning_rate": 1.9286278684526593e-08,
+      "loss": 0.1288,
+      "step": 5452
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 8.577387327531774,
+      "learning_rate": 1.9033507214952784e-08,
+      "loss": 0.0977,
+      "step": 5453
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 7.789300210080421,
+      "learning_rate": 1.878239997886666e-08,
+      "loss": 0.1188,
+      "step": 5454
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 8.261173115162729,
+      "learning_rate": 1.853295706016778e-08,
+      "loss": 0.145,
+      "step": 5455
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 8.57419604447386,
+      "learning_rate": 1.8285178542200022e-08,
+      "loss": 0.1451,
+      "step": 5456
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 8.624820178787198,
+      "learning_rate": 1.8039064507750503e-08,
+      "loss": 0.0851,
+      "step": 5457
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.3281446819414855,
+      "learning_rate": 1.7794615039050665e-08,
+      "loss": 0.1077,
+      "step": 5458
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 6.719079268451326,
+      "learning_rate": 1.7551830217775734e-08,
+      "loss": 0.1064,
+      "step": 5459
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 6.361150677660455,
+      "learning_rate": 1.7310710125044707e-08,
+      "loss": 0.1342,
+      "step": 5460
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 9.934178727478375,
+      "learning_rate": 1.7071254841419805e-08,
+      "loss": 0.188,
+      "step": 5461
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.961703433504403,
+      "learning_rate": 1.6833464446907588e-08,
+      "loss": 0.126,
+      "step": 5462
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 4.709967478366248,
+      "learning_rate": 1.6597339020958393e-08,
+      "loss": 0.0972,
+      "step": 5463
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 7.621860983040733,
+      "learning_rate": 1.6362878642466328e-08,
+      "loss": 0.1342,
+      "step": 5464
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 10.209473650410427,
+      "learning_rate": 1.6130083389768735e-08,
+      "loss": 0.1364,
+      "step": 5465
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 6.139084520540914,
+      "learning_rate": 1.5898953340646728e-08,
+      "loss": 0.1218,
+      "step": 5466
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 9.71994860554082,
+      "learning_rate": 1.5669488572325197e-08,
+      "loss": 0.1554,
+      "step": 5467
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 12.086414770955265,
+      "learning_rate": 1.5441689161472816e-08,
+      "loss": 0.1787,
+      "step": 5468
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 24.407807669084864,
+      "learning_rate": 1.521555518420148e-08,
+      "loss": 0.1927,
+      "step": 5469
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 6.650159393488484,
+      "learning_rate": 1.499108671606686e-08,
+      "loss": 0.1473,
+      "step": 5470
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 6.438063211825578,
+      "learning_rate": 1.4768283832067853e-08,
+      "loss": 0.1339,
+      "step": 5471
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 15.259887152827618,
+      "learning_rate": 1.4547146606646578e-08,
+      "loss": 0.2236,
+      "step": 5472
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.182419293842352,
+      "learning_rate": 1.4327675113690598e-08,
+      "loss": 0.0764,
+      "step": 5473
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 9.2094503251359,
+      "learning_rate": 1.4109869426527368e-08,
+      "loss": 0.1317,
+      "step": 5474
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 6.283356876259326,
+      "learning_rate": 1.3893729617931451e-08,
+      "loss": 0.1119,
+      "step": 5475
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 6.905426440583543,
+      "learning_rate": 1.3679255760118415e-08,
+      "loss": 0.1092,
+      "step": 5476
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.641634779840197,
+      "learning_rate": 1.3466447924748716e-08,
+      "loss": 0.099,
+      "step": 5477
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 4.588743851538649,
+      "learning_rate": 1.3255306182924365e-08,
+      "loss": 0.0688,
+      "step": 5478
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 8.406396921378816,
+      "learning_rate": 1.3045830605192266e-08,
+      "loss": 0.1873,
+      "step": 5479
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.491021032175865,
+      "learning_rate": 1.2838021261541988e-08,
+      "loss": 0.134,
+      "step": 5480
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 7.735288437189416,
+      "learning_rate": 1.263187822140688e-08,
+      "loss": 0.1151,
+      "step": 5481
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 9.538077504415023,
+      "learning_rate": 1.2427401553662955e-08,
+      "loss": 0.1193,
+      "step": 5482
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 13.617139287693995,
+      "learning_rate": 1.2224591326628898e-08,
+      "loss": 0.2377,
+      "step": 5483
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 7.037545116460254,
+      "learning_rate": 1.2023447608068283e-08,
+      "loss": 0.1289,
+      "step": 5484
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.051654601622574,
+      "learning_rate": 1.182397046518735e-08,
+      "loss": 0.0864,
+      "step": 5485
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.660167436446467,
+      "learning_rate": 1.1626159964633899e-08,
+      "loss": 0.1224,
+      "step": 5486
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 5.448552963768357,
+      "learning_rate": 1.1430016172501169e-08,
+      "loss": 0.0796,
+      "step": 5487
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 5.531596088000746,
+      "learning_rate": 1.1235539154323405e-08,
+      "loss": 0.1172,
+      "step": 5488
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.503280936817476,
+      "learning_rate": 1.1042728975079741e-08,
+      "loss": 0.1206,
+      "step": 5489
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 13.364423091479154,
+      "learning_rate": 1.0851585699191425e-08,
+      "loss": 0.1788,
+      "step": 5490
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 5.933243904552928,
+      "learning_rate": 1.0662109390522924e-08,
+      "loss": 0.1199,
+      "step": 5491
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.257081865184139,
+      "learning_rate": 1.047430011238193e-08,
+      "loss": 0.1149,
+      "step": 5492
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.065807785440134,
+      "learning_rate": 1.028815792751936e-08,
+      "loss": 0.11,
+      "step": 5493
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 4.983139579728731,
+      "learning_rate": 1.0103682898128241e-08,
+      "loss": 0.075,
+      "step": 5494
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 10.618784716584416,
+      "learning_rate": 9.920875085845383e-09,
+      "loss": 0.1307,
+      "step": 5495
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.388645197421817,
+      "learning_rate": 9.739734551749703e-09,
+      "loss": 0.1406,
+      "step": 5496
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 10.393540666994594,
+      "learning_rate": 9.560261356364452e-09,
+      "loss": 0.1676,
+      "step": 5497
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.166647278650001,
+      "learning_rate": 9.382455559654446e-09,
+      "loss": 0.1536,
+      "step": 5498
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.81742399868161,
+      "learning_rate": 9.206317221027717e-09,
+      "loss": 0.0848,
+      "step": 5499
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 7.5698480294633415,
+      "learning_rate": 9.031846399336075e-09,
+      "loss": 0.1356,
+      "step": 5500
+    },
+    {
+      "epoch": 3.93,
+      "eval_avg_AUC": 0.7950252002273744,
+      "eval_avg_Accuracy": 0.7066893236074271,
+      "eval_avg_Accuracy-right": 0.8786357114908048,
+      "eval_avg_Accuracy-wrong": 0.4068683193086195,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6590350719883701,
+      "eval_last_AUC": 0.8161622767447089,
+      "eval_last_Accuracy": 0.741959549071618,
+      "eval_last_Accuracy-right": 0.8153123777227077,
+      "eval_last_Accuracy-wrong": 0.6140550375255857,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6823020710938568,
+      "eval_max_AUC": 0.77593042794898,
+      "eval_max_Accuracy": 0.6453083554376657,
+      "eval_max_Accuracy-right": 0.9788704838920047,
+      "eval_max_Accuracy-wrong": 0.06367978166932,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6221345468988383,
+      "eval_min_AUC": 0.8001189509346562,
+      "eval_min_Accuracy": 0.7241379310344828,
+      "eval_min_Accuracy-right": 0.7241424285900613,
+      "eval_min_Accuracy-wrong": 0.7241300886968387,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6609511576435902,
+      "eval_prod_AUC": 0.802284305563165,
+      "eval_prod_Accuracy": 0.7134864058355438,
+      "eval_prod_Accuracy-right": 0.6670796921872962,
+      "eval_prod_Accuracy-wrong": 0.7944052763247669,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6623692613743422,
+      "eval_runtime": 248.6606,
+      "eval_samples_per_second": 97.032,
+      "eval_steps_per_second": 3.032,
+      "eval_sum_AUC": 0.669465043974663,
+      "eval_sum_Accuracy": 0.6391329575596817,
+      "eval_sum_Accuracy-right": 0.9867614451545585,
+      "eval_sum_Accuracy-wrong": 0.03297702979304071,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6489398581558867,
+      "step": 5500
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 8.429106410555258,
+      "learning_rate": 8.859043152872892e-09,
+      "loss": 0.1237,
+      "step": 5501
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 8.009167896999422,
+      "learning_rate": 8.687907539375318e-09,
+      "loss": 0.1164,
+      "step": 5502
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 5.780427122348584,
+      "learning_rate": 8.518439616022057e-09,
+      "loss": 0.1174,
+      "step": 5503
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 7.547785765991373,
+      "learning_rate": 8.350639439436703e-09,
+      "loss": 0.1143,
+      "step": 5504
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 6.131647170129058,
+      "learning_rate": 8.184507065683855e-09,
+      "loss": 0.0824,
+      "step": 5505
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 11.483331590260356,
+      "learning_rate": 8.020042550271889e-09,
+      "loss": 0.1688,
+      "step": 5506
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.2724277741797625,
+      "learning_rate": 7.857245948150183e-09,
+      "loss": 0.1176,
+      "step": 5507
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 5.573181505757999,
+      "learning_rate": 7.696117313713559e-09,
+      "loss": 0.1044,
+      "step": 5508
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.598852408301717,
+      "learning_rate": 7.536656700797284e-09,
+      "loss": 0.0922,
+      "step": 5509
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 10.981912827262958,
+      "learning_rate": 7.37886416268041e-09,
+      "loss": 0.1188,
+      "step": 5510
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 10.405947394175044,
+      "learning_rate": 7.222739752084096e-09,
+      "loss": 0.1179,
+      "step": 5511
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 6.540864550860748,
+      "learning_rate": 7.068283521172725e-09,
+      "loss": 0.1253,
+      "step": 5512
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.525882733341714,
+      "learning_rate": 6.915495521552795e-09,
+      "loss": 0.1077,
+      "step": 5513
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 6.164782248488719,
+      "learning_rate": 6.764375804274026e-09,
+      "loss": 0.1436,
+      "step": 5514
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.444440508428989,
+      "learning_rate": 6.61492441982714e-09,
+      "loss": 0.1319,
+      "step": 5515
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.812978938104751,
+      "learning_rate": 6.467141418147748e-09,
+      "loss": 0.0881,
+      "step": 5516
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 5.184902484678889,
+      "learning_rate": 6.321026848613021e-09,
+      "loss": 0.1021,
+      "step": 5517
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.699439697594633,
+      "learning_rate": 6.176580760041684e-09,
+      "loss": 0.0901,
+      "step": 5518
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.963371822825025,
+      "learning_rate": 6.033803200696242e-09,
+      "loss": 0.1211,
+      "step": 5519
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 8.425984631270113,
+      "learning_rate": 5.892694218281869e-09,
+      "loss": 0.1633,
+      "step": 5520
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.949188213453105,
+      "learning_rate": 5.753253859944741e-09,
+      "loss": 0.1158,
+      "step": 5521
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 9.136952727452089,
+      "learning_rate": 5.615482172275366e-09,
+      "loss": 0.1742,
+      "step": 5522
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.699454602256628,
+      "learning_rate": 5.479379201305257e-09,
+      "loss": 0.1188,
+      "step": 5523
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 7.953008550854942,
+      "learning_rate": 5.344944992509149e-09,
+      "loss": 0.1521,
+      "step": 5524
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 6.618564531824671,
+      "learning_rate": 5.212179590803335e-09,
+      "loss": 0.1507,
+      "step": 5525
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 9.999104323435565,
+      "learning_rate": 5.08108304054844e-09,
+      "loss": 0.1151,
+      "step": 5526
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 15.707971260259198,
+      "learning_rate": 4.9516553855455395e-09,
+      "loss": 0.1814,
+      "step": 5527
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 8.642743845716193,
+      "learning_rate": 4.82389666903893e-09,
+      "loss": 0.116,
+      "step": 5528
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 7.118772631160384,
+      "learning_rate": 4.697806933715021e-09,
+      "loss": 0.125,
+      "step": 5529
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 9.23091508452254,
+      "learning_rate": 4.573386221703446e-09,
+      "loss": 0.1327,
+      "step": 5530
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 13.387861094338058,
+      "learning_rate": 4.450634574574286e-09,
+      "loss": 0.1758,
+      "step": 5531
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.407477037058821,
+      "learning_rate": 4.329552033341955e-09,
+      "loss": 0.0692,
+      "step": 5532
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.614578814740924,
+      "learning_rate": 4.210138638462424e-09,
+      "loss": 0.0916,
+      "step": 5533
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 10.253247016948384,
+      "learning_rate": 4.0923944298337796e-09,
+      "loss": 0.1483,
+      "step": 5534
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 6.932631685042542,
+      "learning_rate": 3.976319446795662e-09,
+      "loss": 0.1497,
+      "step": 5535
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 7.467510399283004,
+      "learning_rate": 3.8619137281326044e-09,
+      "loss": 0.1044,
+      "step": 5536
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 8.923637692156998,
+      "learning_rate": 3.749177312068475e-09,
+      "loss": 0.1453,
+      "step": 5537
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 13.938583660145987,
+      "learning_rate": 3.63811023627092e-09,
+      "loss": 0.183,
+      "step": 5538
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 10.314501087676051,
+      "learning_rate": 3.528712537849144e-09,
+      "loss": 0.1874,
+      "step": 5539
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 7.711784952194379,
+      "learning_rate": 3.42098425335613e-09,
+      "loss": 0.1631,
+      "step": 5540
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 7.6746819815164455,
+      "learning_rate": 3.3149254187841985e-09,
+      "loss": 0.1301,
+      "step": 5541
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.077385997300826,
+      "learning_rate": 3.210536069571113e-09,
+      "loss": 0.1223,
+      "step": 5542
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.60490536386744,
+      "learning_rate": 3.1078162405939747e-09,
+      "loss": 0.1371,
+      "step": 5543
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.546314650170883,
+      "learning_rate": 3.006765966174774e-09,
+      "loss": 0.1312,
+      "step": 5544
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.402040317933734,
+      "learning_rate": 2.907385280075392e-09,
+      "loss": 0.1078,
+      "step": 5545
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 7.1113178071259355,
+      "learning_rate": 2.80967421550038e-09,
+      "loss": 0.124,
+      "step": 5546
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 9.882733207585435,
+      "learning_rate": 2.7136328050980654e-09,
+      "loss": 0.2028,
+      "step": 5547
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 8.22923561778551,
+      "learning_rate": 2.6192610809566697e-09,
+      "loss": 0.1747,
+      "step": 5548
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 9.744168083139122,
+      "learning_rate": 2.5265590746076373e-09,
+      "loss": 0.0875,
+      "step": 5549
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 7.470206404886737,
+      "learning_rate": 2.43552681702508e-09,
+      "loss": 0.1339,
+      "step": 5550
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 5.8082392348378225,
+      "learning_rate": 2.346164338624113e-09,
+      "loss": 0.1084,
+      "step": 5551
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.477160650962435,
+      "learning_rate": 2.2584716692619636e-09,
+      "loss": 0.1305,
+      "step": 5552
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.159335395880286,
+      "learning_rate": 2.172448838239083e-09,
+      "loss": 0.1267,
+      "step": 5553
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 5.573462560585101,
+      "learning_rate": 2.08809587429748e-09,
+      "loss": 0.0966,
+      "step": 5554
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.343642412756653,
+      "learning_rate": 2.0054128056201662e-09,
+      "loss": 0.0584,
+      "step": 5555
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 8.254147958795857,
+      "learning_rate": 1.924399659833376e-09,
+      "loss": 0.139,
+      "step": 5556
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 10.092190845622381,
+      "learning_rate": 1.8450564640054569e-09,
+      "loss": 0.1674,
+      "step": 5557
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 8.035445013347708,
+      "learning_rate": 1.7673832446463146e-09,
+      "loss": 0.1506,
+      "step": 5558
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 8.832099606863865,
+      "learning_rate": 1.6913800277085225e-09,
+      "loss": 0.1419,
+      "step": 5559
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 9.404074142401095,
+      "learning_rate": 1.6170468385845462e-09,
+      "loss": 0.1946,
+      "step": 5560
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.577566022959948,
+      "learning_rate": 1.5443837021122954e-09,
+      "loss": 0.0952,
+      "step": 5561
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 7.343705255356882,
+      "learning_rate": 1.473390642569017e-09,
+      "loss": 0.1126,
+      "step": 5562
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 7.2585089112143795,
+      "learning_rate": 1.4040676836746259e-09,
+      "loss": 0.1133,
+      "step": 5563
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 6.784254763986543,
+      "learning_rate": 1.336414848591705e-09,
+      "loss": 0.1276,
+      "step": 5564
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 5.854702420201817,
+      "learning_rate": 1.2704321599243951e-09,
+      "loss": 0.0994,
+      "step": 5565
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 5.753710545202959,
+      "learning_rate": 1.206119639718395e-09,
+      "loss": 0.1279,
+      "step": 5566
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 7.307788810185479,
+      "learning_rate": 1.1434773094615158e-09,
+      "loss": 0.1118,
+      "step": 5567
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 16.079904243512647,
+      "learning_rate": 1.0825051900842377e-09,
+      "loss": 0.1475,
+      "step": 5568
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.869485387673937,
+      "learning_rate": 1.0232033019580423e-09,
+      "loss": 0.1205,
+      "step": 5569
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 9.561539939824815,
+      "learning_rate": 9.655716648970804e-10,
+      "loss": 0.1674,
+      "step": 5570
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 8.986453072022883,
+      "learning_rate": 9.096102981570598e-10,
+      "loss": 0.166,
+      "step": 5571
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.372199399777608,
+      "learning_rate": 8.553192204358018e-10,
+      "loss": 0.0487,
+      "step": 5572
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 8.885464213647072,
+      "learning_rate": 8.026984498726853e-10,
+      "loss": 0.1139,
+      "step": 5573
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.969673215125041,
+      "learning_rate": 7.517480040497572e-10,
+      "loss": 0.0748,
+      "step": 5574
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 8.019389192351877,
+      "learning_rate": 7.024678999900669e-10,
+      "loss": 0.1306,
+      "step": 5575
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.295109377091253,
+      "learning_rate": 6.548581541593324e-10,
+      "loss": 0.1327,
+      "step": 5576
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 8.69637612538181,
+      "learning_rate": 6.08918782464829e-10,
+      "loss": 0.1021,
+      "step": 5577
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.261474454096343,
+      "learning_rate": 5.646498002553902e-10,
+      "loss": 0.0916,
+      "step": 5578
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 16.482766831647567,
+      "learning_rate": 5.220512223219621e-10,
+      "loss": 0.2062,
+      "step": 5579
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 7.218560157195067,
+      "learning_rate": 4.81123062898714e-10,
+      "loss": 0.1332,
+      "step": 5580
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.056314731703903,
+      "learning_rate": 4.4186533565915293e-10,
+      "loss": 0.0832,
+      "step": 5581
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 6.337018112627061,
+      "learning_rate": 4.042780537205637e-10,
+      "loss": 0.1094,
+      "step": 5582
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 8.678745644886078,
+      "learning_rate": 3.6836122964178934e-10,
+      "loss": 0.1344,
+      "step": 5583
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 8.323983732961677,
+      "learning_rate": 3.341148754232304e-10,
+      "loss": 0.112,
+      "step": 5584
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.730233285070333,
+      "learning_rate": 3.015390025068454e-10,
+      "loss": 0.1141,
+      "step": 5585
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 5.2887206705015775,
+      "learning_rate": 2.706336217767058e-10,
+      "loss": 0.0966,
+      "step": 5586
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 10.523107370264057,
+      "learning_rate": 2.4139874355955105e-10,
+      "loss": 0.1647,
+      "step": 5587
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 7.034757136628954,
+      "learning_rate": 2.138343776231233e-10,
+      "loss": 0.1074,
+      "step": 5588
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.404901800476969,
+      "learning_rate": 1.8794053317672255e-10,
+      "loss": 0.1074,
+      "step": 5589
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.157714487346645,
+      "learning_rate": 1.6371721887287196e-10,
+      "loss": 0.1122,
+      "step": 5590
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 8.352229011291241,
+      "learning_rate": 1.4116444280398711e-10,
+      "loss": 0.1099,
+      "step": 5591
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 5.989421477649134,
+      "learning_rate": 1.2028221250570683e-10,
+      "loss": 0.0992,
+      "step": 5592
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 9.9800678266692,
+      "learning_rate": 1.0107053495522767e-10,
+      "loss": 0.1093,
+      "step": 5593
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 11.303595468417102,
+      "learning_rate": 8.35294165718592e-11,
+      "loss": 0.15,
+      "step": 5594
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 9.776566003654384,
+      "learning_rate": 6.765886321646874e-11,
+      "loss": 0.1285,
+      "step": 5595
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 9.512879527949313,
+      "learning_rate": 5.345888019092638e-11,
+      "loss": 0.1599,
+      "step": 5596
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 9.328733178787845,
+      "learning_rate": 4.092947224032529e-11,
+      "loss": 0.1466,
+      "step": 5597
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 12.1455101579189,
+      "learning_rate": 3.007064355076139e-11,
+      "loss": 0.1296,
+      "step": 5598
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 6.318880612200429,
+      "learning_rate": 2.088239775044354e-11,
+      "loss": 0.0999,
+      "step": 5599
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 4.899339012947326,
+      "learning_rate": 1.3364737909138392e-11,
+      "loss": 0.0953,
+      "step": 5600
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 7.1594901159283815,
+      "learning_rate": 7.517666539280654e-12,
+      "loss": 0.1243,
+      "step": 5601
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 9.295531250814186,
+      "learning_rate": 3.3411855937526273e-12,
+      "loss": 0.1388,
+      "step": 5602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 9.718829953860425,
+      "learning_rate": 8.352964681046516e-13,
+      "loss": 0.1476,
+      "step": 5603
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 5.714918142396491,
+      "learning_rate": 0.0,
+      "loss": 0.1135,
+      "step": 5604
+    },
+    {
+      "epoch": 4.0,
+      "step": 5604,
+      "total_flos": 750239094767616.0,
+      "train_loss": 0.36047464298572307,
+      "train_runtime": 15643.8891,
+      "train_samples_per_second": 22.919,
+      "train_steps_per_second": 0.358
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5604,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 24000,
+  "total_flos": 750239094767616.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}