diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10969 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 7.0,
+  "eval_steps": 500,
+  "global_step": 1561,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004484304932735426,
+      "grad_norm": 2.813812017440796,
+      "learning_rate": 6.369426751592358e-08,
+      "loss": 0.747,
+      "step": 1
+    },
+    {
+      "epoch": 0.008968609865470852,
+      "grad_norm": 2.828036069869995,
+      "learning_rate": 1.2738853503184715e-07,
+      "loss": 0.7779,
+      "step": 2
+    },
+    {
+      "epoch": 0.013452914798206279,
+      "grad_norm": 2.9274673461914062,
+      "learning_rate": 1.9108280254777072e-07,
+      "loss": 0.7782,
+      "step": 3
+    },
+    {
+      "epoch": 0.017937219730941704,
+      "grad_norm": 2.98175048828125,
+      "learning_rate": 2.547770700636943e-07,
+      "loss": 0.7817,
+      "step": 4
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 2.9156811237335205,
+      "learning_rate": 3.1847133757961787e-07,
+      "loss": 0.7919,
+      "step": 5
+    },
+    {
+      "epoch": 0.026905829596412557,
+      "grad_norm": 2.9519894123077393,
+      "learning_rate": 3.8216560509554143e-07,
+      "loss": 0.7811,
+      "step": 6
+    },
+    {
+      "epoch": 0.03139013452914798,
+      "grad_norm": 2.877030849456787,
+      "learning_rate": 4.45859872611465e-07,
+      "loss": 0.772,
+      "step": 7
+    },
+    {
+      "epoch": 0.03587443946188341,
+      "grad_norm": 2.885763168334961,
+      "learning_rate": 5.095541401273886e-07,
+      "loss": 0.7715,
+      "step": 8
+    },
+    {
+      "epoch": 0.04035874439461883,
+      "grad_norm": 2.8665950298309326,
+      "learning_rate": 5.732484076433121e-07,
+      "loss": 0.7729,
+      "step": 9
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 2.7542173862457275,
+      "learning_rate": 6.369426751592357e-07,
+      "loss": 0.7628,
+      "step": 10
+    },
+    {
+      "epoch": 0.04932735426008968,
+      "grad_norm": 2.6579983234405518,
+      "learning_rate": 7.006369426751592e-07,
+      "loss": 0.7616,
+      "step": 11
+    },
+    {
+      "epoch": 0.053811659192825115,
+      "grad_norm": 2.7461585998535156,
+      "learning_rate": 7.643312101910829e-07,
+      "loss": 0.7611,
+      "step": 12
+    },
+    {
+      "epoch": 0.05829596412556054,
+      "grad_norm": 2.2442593574523926,
+      "learning_rate": 8.280254777070064e-07,
+      "loss": 0.7334,
+      "step": 13
+    },
+    {
+      "epoch": 0.06278026905829596,
+      "grad_norm": 2.1680378913879395,
+      "learning_rate": 8.9171974522293e-07,
+      "loss": 0.734,
+      "step": 14
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 2.089891195297241,
+      "learning_rate": 9.554140127388537e-07,
+      "loss": 0.7223,
+      "step": 15
+    },
+    {
+      "epoch": 0.07174887892376682,
+      "grad_norm": 2.0968801975250244,
+      "learning_rate": 1.0191082802547772e-06,
+      "loss": 0.7254,
+      "step": 16
+    },
+    {
+      "epoch": 0.07623318385650224,
+      "grad_norm": 1.8303896188735962,
+      "learning_rate": 1.0828025477707007e-06,
+      "loss": 0.7242,
+      "step": 17
+    },
+    {
+      "epoch": 0.08071748878923767,
+      "grad_norm": 1.4840261936187744,
+      "learning_rate": 1.1464968152866242e-06,
+      "loss": 0.6959,
+      "step": 18
+    },
+    {
+      "epoch": 0.08520179372197309,
+      "grad_norm": 1.4487215280532837,
+      "learning_rate": 1.210191082802548e-06,
+      "loss": 0.6913,
+      "step": 19
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 1.415422797203064,
+      "learning_rate": 1.2738853503184715e-06,
+      "loss": 0.7022,
+      "step": 20
+    },
+    {
+      "epoch": 0.09417040358744394,
+      "grad_norm": 1.384924292564392,
+      "learning_rate": 1.337579617834395e-06,
+      "loss": 0.6886,
+      "step": 21
+    },
+    {
+      "epoch": 0.09865470852017937,
+      "grad_norm": 1.3160287141799927,
+      "learning_rate": 1.4012738853503185e-06,
+      "loss": 0.6826,
+      "step": 22
+    },
+    {
+      "epoch": 0.1031390134529148,
+      "grad_norm": 1.18776273727417,
+      "learning_rate": 1.4649681528662422e-06,
+      "loss": 0.6766,
+      "step": 23
+    },
+    {
+      "epoch": 0.10762331838565023,
+      "grad_norm": 1.0047998428344727,
+      "learning_rate": 1.5286624203821657e-06,
+      "loss": 0.6473,
+      "step": 24
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 0.9870753884315491,
+      "learning_rate": 1.5923566878980892e-06,
+      "loss": 0.6532,
+      "step": 25
+    },
+    {
+      "epoch": 0.11659192825112108,
+      "grad_norm": 0.9578473567962646,
+      "learning_rate": 1.6560509554140127e-06,
+      "loss": 0.6351,
+      "step": 26
+    },
+    {
+      "epoch": 0.1210762331838565,
+      "grad_norm": 0.9310482740402222,
+      "learning_rate": 1.7197452229299363e-06,
+      "loss": 0.6349,
+      "step": 27
+    },
+    {
+      "epoch": 0.12556053811659193,
+      "grad_norm": 0.8986713886260986,
+      "learning_rate": 1.78343949044586e-06,
+      "loss": 0.6453,
+      "step": 28
+    },
+    {
+      "epoch": 0.13004484304932734,
+      "grad_norm": 0.8593706488609314,
+      "learning_rate": 1.8471337579617835e-06,
+      "loss": 0.6248,
+      "step": 29
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 0.8357294201850891,
+      "learning_rate": 1.9108280254777074e-06,
+      "loss": 0.636,
+      "step": 30
+    },
+    {
+      "epoch": 0.13901345291479822,
+      "grad_norm": 0.7058345079421997,
+      "learning_rate": 1.974522292993631e-06,
+      "loss": 0.6204,
+      "step": 31
+    },
+    {
+      "epoch": 0.14349775784753363,
+      "grad_norm": 0.6469331979751587,
+      "learning_rate": 2.0382165605095544e-06,
+      "loss": 0.6105,
+      "step": 32
+    },
+    {
+      "epoch": 0.14798206278026907,
+      "grad_norm": 0.6891763806343079,
+      "learning_rate": 2.101910828025478e-06,
+      "loss": 0.595,
+      "step": 33
+    },
+    {
+      "epoch": 0.15246636771300448,
+      "grad_norm": 0.7042757868766785,
+      "learning_rate": 2.1656050955414015e-06,
+      "loss": 0.5911,
+      "step": 34
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 0.6550124883651733,
+      "learning_rate": 2.229299363057325e-06,
+      "loss": 0.5925,
+      "step": 35
+    },
+    {
+      "epoch": 0.16143497757847533,
+      "grad_norm": 0.5893523097038269,
+      "learning_rate": 2.2929936305732485e-06,
+      "loss": 0.5752,
+      "step": 36
+    },
+    {
+      "epoch": 0.16591928251121077,
+      "grad_norm": 0.49922969937324524,
+      "learning_rate": 2.356687898089172e-06,
+      "loss": 0.5624,
+      "step": 37
+    },
+    {
+      "epoch": 0.17040358744394618,
+      "grad_norm": 0.4881310760974884,
+      "learning_rate": 2.420382165605096e-06,
+      "loss": 0.5724,
+      "step": 38
+    },
+    {
+      "epoch": 0.17488789237668162,
+      "grad_norm": 0.4961792826652527,
+      "learning_rate": 2.4840764331210194e-06,
+      "loss": 0.5633,
+      "step": 39
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 0.5010235905647278,
+      "learning_rate": 2.547770700636943e-06,
+      "loss": 0.561,
+      "step": 40
+    },
+    {
+      "epoch": 0.18385650224215247,
+      "grad_norm": 0.46078333258628845,
+      "learning_rate": 2.6114649681528665e-06,
+      "loss": 0.5625,
+      "step": 41
+    },
+    {
+      "epoch": 0.18834080717488788,
+      "grad_norm": 0.4454447627067566,
+      "learning_rate": 2.67515923566879e-06,
+      "loss": 0.5477,
+      "step": 42
+    },
+    {
+      "epoch": 0.19282511210762332,
+      "grad_norm": 0.40196821093559265,
+      "learning_rate": 2.7388535031847135e-06,
+      "loss": 0.5392,
+      "step": 43
+    },
+    {
+      "epoch": 0.19730941704035873,
+      "grad_norm": 0.34942522644996643,
+      "learning_rate": 2.802547770700637e-06,
+      "loss": 0.5405,
+      "step": 44
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 0.3442065119743347,
+      "learning_rate": 2.8662420382165605e-06,
+      "loss": 0.532,
+      "step": 45
+    },
+    {
+      "epoch": 0.2062780269058296,
+      "grad_norm": 0.3465350270271301,
+      "learning_rate": 2.9299363057324844e-06,
+      "loss": 0.5368,
+      "step": 46
+    },
+    {
+      "epoch": 0.21076233183856502,
+      "grad_norm": 0.3627440333366394,
+      "learning_rate": 2.993630573248408e-06,
+      "loss": 0.5258,
+      "step": 47
+    },
+    {
+      "epoch": 0.21524663677130046,
+      "grad_norm": 0.3505997955799103,
+      "learning_rate": 3.0573248407643314e-06,
+      "loss": 0.5218,
+      "step": 48
+    },
+    {
+      "epoch": 0.21973094170403587,
+      "grad_norm": 0.34063127636909485,
+      "learning_rate": 3.121019108280255e-06,
+      "loss": 0.5145,
+      "step": 49
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 0.34885120391845703,
+      "learning_rate": 3.1847133757961785e-06,
+      "loss": 0.528,
+      "step": 50
+    },
+    {
+      "epoch": 0.22869955156950672,
+      "grad_norm": 0.32321998476982117,
+      "learning_rate": 3.248407643312102e-06,
+      "loss": 0.5101,
+      "step": 51
+    },
+    {
+      "epoch": 0.23318385650224216,
+      "grad_norm": 0.3193144202232361,
+      "learning_rate": 3.3121019108280255e-06,
+      "loss": 0.5108,
+      "step": 52
+    },
+    {
+      "epoch": 0.23766816143497757,
+      "grad_norm": 0.2741270363330841,
+      "learning_rate": 3.375796178343949e-06,
+      "loss": 0.5105,
+      "step": 53
+    },
+    {
+      "epoch": 0.242152466367713,
+      "grad_norm": 0.267229288816452,
+      "learning_rate": 3.4394904458598725e-06,
+      "loss": 0.5089,
+      "step": 54
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 0.25306186079978943,
+      "learning_rate": 3.5031847133757964e-06,
+      "loss": 0.4919,
+      "step": 55
+    },
+    {
+      "epoch": 0.25112107623318386,
+      "grad_norm": 0.2585896849632263,
+      "learning_rate": 3.56687898089172e-06,
+      "loss": 0.4997,
+      "step": 56
+    },
+    {
+      "epoch": 0.2556053811659193,
+      "grad_norm": 0.23930947482585907,
+      "learning_rate": 3.6305732484076435e-06,
+      "loss": 0.4974,
+      "step": 57
+    },
+    {
+      "epoch": 0.2600896860986547,
+      "grad_norm": 0.2372271716594696,
+      "learning_rate": 3.694267515923567e-06,
+      "loss": 0.4933,
+      "step": 58
+    },
+    {
+      "epoch": 0.2645739910313901,
+      "grad_norm": 0.21972738206386566,
+      "learning_rate": 3.757961783439491e-06,
+      "loss": 0.4991,
+      "step": 59
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 0.20193585753440857,
+      "learning_rate": 3.821656050955415e-06,
+      "loss": 0.4852,
+      "step": 60
+    },
+    {
+      "epoch": 0.273542600896861,
+      "grad_norm": 0.19289743900299072,
+      "learning_rate": 3.885350318471338e-06,
+      "loss": 0.4814,
+      "step": 61
+    },
+    {
+      "epoch": 0.27802690582959644,
+      "grad_norm": 0.18357662856578827,
+      "learning_rate": 3.949044585987262e-06,
+      "loss": 0.4848,
+      "step": 62
+    },
+    {
+      "epoch": 0.2825112107623318,
+      "grad_norm": 0.19417670369148254,
+      "learning_rate": 4.012738853503185e-06,
+      "loss": 0.4979,
+      "step": 63
+    },
+    {
+      "epoch": 0.28699551569506726,
+      "grad_norm": 0.18120144307613373,
+      "learning_rate": 4.076433121019109e-06,
+      "loss": 0.4802,
+      "step": 64
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 0.18545961380004883,
+      "learning_rate": 4.140127388535032e-06,
+      "loss": 0.4822,
+      "step": 65
+    },
+    {
+      "epoch": 0.29596412556053814,
+      "grad_norm": 0.18592469394207,
+      "learning_rate": 4.203821656050956e-06,
+      "loss": 0.47,
+      "step": 66
+    },
+    {
+      "epoch": 0.3004484304932735,
+      "grad_norm": 0.1931634098291397,
+      "learning_rate": 4.26751592356688e-06,
+      "loss": 0.4861,
+      "step": 67
+    },
+    {
+      "epoch": 0.30493273542600896,
+      "grad_norm": 0.18863962590694427,
+      "learning_rate": 4.331210191082803e-06,
+      "loss": 0.4748,
+      "step": 68
+    },
+    {
+      "epoch": 0.3094170403587444,
+      "grad_norm": 0.1812104433774948,
+      "learning_rate": 4.394904458598727e-06,
+      "loss": 0.475,
+      "step": 69
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 0.1759832799434662,
+      "learning_rate": 4.45859872611465e-06,
+      "loss": 0.4667,
+      "step": 70
+    },
+    {
+      "epoch": 0.3183856502242152,
+      "grad_norm": 0.15467405319213867,
+      "learning_rate": 4.522292993630574e-06,
+      "loss": 0.4704,
+      "step": 71
+    },
+    {
+      "epoch": 0.32286995515695066,
+      "grad_norm": 0.1641678512096405,
+      "learning_rate": 4.585987261146497e-06,
+      "loss": 0.4818,
+      "step": 72
+    },
+    {
+      "epoch": 0.3273542600896861,
+      "grad_norm": 0.15334247052669525,
+      "learning_rate": 4.649681528662421e-06,
+      "loss": 0.4661,
+      "step": 73
+    },
+    {
+      "epoch": 0.33183856502242154,
+      "grad_norm": 0.1616121232509613,
+      "learning_rate": 4.713375796178344e-06,
+      "loss": 0.4698,
+      "step": 74
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 0.16387605667114258,
+      "learning_rate": 4.777070063694268e-06,
+      "loss": 0.4804,
+      "step": 75
+    },
+    {
+      "epoch": 0.34080717488789236,
+      "grad_norm": 0.1563800573348999,
+      "learning_rate": 4.840764331210192e-06,
+      "loss": 0.4615,
+      "step": 76
+    },
+    {
+      "epoch": 0.3452914798206278,
+      "grad_norm": 0.1630326360464096,
+      "learning_rate": 4.904458598726115e-06,
+      "loss": 0.4667,
+      "step": 77
+    },
+    {
+      "epoch": 0.34977578475336324,
+      "grad_norm": 0.17320983111858368,
+      "learning_rate": 4.968152866242039e-06,
+      "loss": 0.4599,
+      "step": 78
+    },
+    {
+      "epoch": 0.3542600896860987,
+      "grad_norm": 0.13915883004665375,
+      "learning_rate": 5.031847133757962e-06,
+      "loss": 0.4642,
+      "step": 79
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 0.15249691903591156,
+      "learning_rate": 5.095541401273886e-06,
+      "loss": 0.4669,
+      "step": 80
+    },
+    {
+      "epoch": 0.3632286995515695,
+      "grad_norm": 0.15934018790721893,
+      "learning_rate": 5.159235668789809e-06,
+      "loss": 0.4621,
+      "step": 81
+    },
+    {
+      "epoch": 0.36771300448430494,
+      "grad_norm": 0.16763532161712646,
+      "learning_rate": 5.222929936305733e-06,
+      "loss": 0.472,
+      "step": 82
+    },
+    {
+      "epoch": 0.3721973094170404,
+      "grad_norm": 0.15078717470169067,
+      "learning_rate": 5.286624203821657e-06,
+      "loss": 0.4627,
+      "step": 83
+    },
+    {
+      "epoch": 0.37668161434977576,
+      "grad_norm": 0.1679476499557495,
+      "learning_rate": 5.35031847133758e-06,
+      "loss": 0.4616,
+      "step": 84
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 0.14166119694709778,
+      "learning_rate": 5.414012738853504e-06,
+      "loss": 0.4568,
+      "step": 85
+    },
+    {
+      "epoch": 0.38565022421524664,
+      "grad_norm": 0.17618639767169952,
+      "learning_rate": 5.477707006369427e-06,
+      "loss": 0.4505,
+      "step": 86
+    },
+    {
+      "epoch": 0.3901345291479821,
+      "grad_norm": 0.14954672753810883,
+      "learning_rate": 5.541401273885351e-06,
+      "loss": 0.4528,
+      "step": 87
+    },
+    {
+      "epoch": 0.39461883408071746,
+      "grad_norm": 0.14634643495082855,
+      "learning_rate": 5.605095541401274e-06,
+      "loss": 0.4448,
+      "step": 88
+    },
+    {
+      "epoch": 0.3991031390134529,
+      "grad_norm": 0.1607850193977356,
+      "learning_rate": 5.668789808917198e-06,
+      "loss": 0.4564,
+      "step": 89
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 0.14851216971874237,
+      "learning_rate": 5.732484076433121e-06,
+      "loss": 0.4572,
+      "step": 90
+    },
+    {
+      "epoch": 0.4080717488789238,
+      "grad_norm": 0.14807498455047607,
+      "learning_rate": 5.796178343949045e-06,
+      "loss": 0.4491,
+      "step": 91
+    },
+    {
+      "epoch": 0.4125560538116592,
+      "grad_norm": 0.1566411703824997,
+      "learning_rate": 5.859872611464969e-06,
+      "loss": 0.4494,
+      "step": 92
+    },
+    {
+      "epoch": 0.4170403587443946,
+      "grad_norm": 0.17059752345085144,
+      "learning_rate": 5.923566878980892e-06,
+      "loss": 0.4457,
+      "step": 93
+    },
+    {
+      "epoch": 0.42152466367713004,
+      "grad_norm": 0.15800048410892487,
+      "learning_rate": 5.987261146496816e-06,
+      "loss": 0.4519,
+      "step": 94
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 0.15249966084957123,
+      "learning_rate": 6.050955414012739e-06,
+      "loss": 0.4556,
+      "step": 95
+    },
+    {
+      "epoch": 0.4304932735426009,
+      "grad_norm": 0.1626468151807785,
+      "learning_rate": 6.114649681528663e-06,
+      "loss": 0.436,
+      "step": 96
+    },
+    {
+      "epoch": 0.4349775784753363,
+      "grad_norm": 0.13890047371387482,
+      "learning_rate": 6.178343949044586e-06,
+      "loss": 0.4494,
+      "step": 97
+    },
+    {
+      "epoch": 0.43946188340807174,
+      "grad_norm": 0.15467026829719543,
+      "learning_rate": 6.24203821656051e-06,
+      "loss": 0.4424,
+      "step": 98
+    },
+    {
+      "epoch": 0.4439461883408072,
+      "grad_norm": 0.146198108792305,
+      "learning_rate": 6.305732484076433e-06,
+      "loss": 0.4365,
+      "step": 99
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 0.14116433262825012,
+      "learning_rate": 6.369426751592357e-06,
+      "loss": 0.4347,
+      "step": 100
+    },
+    {
+      "epoch": 0.452914798206278,
+      "grad_norm": 0.1378544718027115,
+      "learning_rate": 6.433121019108281e-06,
+      "loss": 0.448,
+      "step": 101
+    },
+    {
+      "epoch": 0.45739910313901344,
+      "grad_norm": 0.15722863376140594,
+      "learning_rate": 6.496815286624204e-06,
+      "loss": 0.4443,
+      "step": 102
+    },
+    {
+      "epoch": 0.4618834080717489,
+      "grad_norm": 0.15835310518741608,
+      "learning_rate": 6.560509554140128e-06,
+      "loss": 0.4441,
+      "step": 103
+    },
+    {
+      "epoch": 0.4663677130044843,
+      "grad_norm": 0.1398308426141739,
+      "learning_rate": 6.624203821656051e-06,
+      "loss": 0.4312,
+      "step": 104
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 0.15488538146018982,
+      "learning_rate": 6.687898089171975e-06,
+      "loss": 0.429,
+      "step": 105
+    },
+    {
+      "epoch": 0.47533632286995514,
+      "grad_norm": 0.1607794165611267,
+      "learning_rate": 6.751592356687898e-06,
+      "loss": 0.4364,
+      "step": 106
+    },
+    {
+      "epoch": 0.4798206278026906,
+      "grad_norm": 0.1474023461341858,
+      "learning_rate": 6.815286624203822e-06,
+      "loss": 0.4339,
+      "step": 107
+    },
+    {
+      "epoch": 0.484304932735426,
+      "grad_norm": 0.14334262907505035,
+      "learning_rate": 6.878980891719745e-06,
+      "loss": 0.438,
+      "step": 108
+    },
+    {
+      "epoch": 0.48878923766816146,
+      "grad_norm": 0.157660573720932,
+      "learning_rate": 6.942675159235669e-06,
+      "loss": 0.4386,
+      "step": 109
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 0.1440911740064621,
+      "learning_rate": 7.006369426751593e-06,
+      "loss": 0.4363,
+      "step": 110
+    },
+    {
+      "epoch": 0.4977578475336323,
+      "grad_norm": 0.14317628741264343,
+      "learning_rate": 7.070063694267516e-06,
+      "loss": 0.4349,
+      "step": 111
+    },
+    {
+      "epoch": 0.5022421524663677,
+      "grad_norm": 0.1534879207611084,
+      "learning_rate": 7.13375796178344e-06,
+      "loss": 0.4304,
+      "step": 112
+    },
+    {
+      "epoch": 0.5067264573991032,
+      "grad_norm": 0.17746873199939728,
+      "learning_rate": 7.197452229299363e-06,
+      "loss": 0.4415,
+      "step": 113
+    },
+    {
+      "epoch": 0.5112107623318386,
+      "grad_norm": 0.15792234241962433,
+      "learning_rate": 7.261146496815287e-06,
+      "loss": 0.4392,
+      "step": 114
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 0.1436270773410797,
+      "learning_rate": 7.32484076433121e-06,
+      "loss": 0.4267,
+      "step": 115
+    },
+    {
+      "epoch": 0.5201793721973094,
+      "grad_norm": 0.16041822731494904,
+      "learning_rate": 7.388535031847134e-06,
+      "loss": 0.4407,
+      "step": 116
+    },
+    {
+      "epoch": 0.5246636771300448,
+      "grad_norm": 0.17744363844394684,
+      "learning_rate": 7.452229299363057e-06,
+      "loss": 0.4325,
+      "step": 117
+    },
+    {
+      "epoch": 0.5291479820627802,
+      "grad_norm": 0.15530529618263245,
+      "learning_rate": 7.515923566878982e-06,
+      "loss": 0.4349,
+      "step": 118
+    },
+    {
+      "epoch": 0.5336322869955157,
+      "grad_norm": 0.16029410064220428,
+      "learning_rate": 7.579617834394906e-06,
+      "loss": 0.4266,
+      "step": 119
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 0.14973820745944977,
+      "learning_rate": 7.64331210191083e-06,
+      "loss": 0.4304,
+      "step": 120
+    },
+    {
+      "epoch": 0.5426008968609866,
+      "grad_norm": 0.1501941680908203,
+      "learning_rate": 7.707006369426753e-06,
+      "loss": 0.4365,
+      "step": 121
+    },
+    {
+      "epoch": 0.547085201793722,
+      "grad_norm": 0.15940609574317932,
+      "learning_rate": 7.770700636942676e-06,
+      "loss": 0.4266,
+      "step": 122
+    },
+    {
+      "epoch": 0.5515695067264574,
+      "grad_norm": 0.15340670943260193,
+      "learning_rate": 7.8343949044586e-06,
+      "loss": 0.4234,
+      "step": 123
+    },
+    {
+      "epoch": 0.5560538116591929,
+      "grad_norm": 0.14284458756446838,
+      "learning_rate": 7.898089171974524e-06,
+      "loss": 0.4337,
+      "step": 124
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 0.15405072271823883,
+      "learning_rate": 7.961783439490447e-06,
+      "loss": 0.4297,
+      "step": 125
+    },
+    {
+      "epoch": 0.5650224215246636,
+      "grad_norm": 0.15733851492404938,
+      "learning_rate": 8.02547770700637e-06,
+      "loss": 0.4238,
+      "step": 126
+    },
+    {
+      "epoch": 0.5695067264573991,
+      "grad_norm": 0.14362181723117828,
+      "learning_rate": 8.089171974522295e-06,
+      "loss": 0.4235,
+      "step": 127
+    },
+    {
+      "epoch": 0.5739910313901345,
+      "grad_norm": 0.1469912976026535,
+      "learning_rate": 8.152866242038218e-06,
+      "loss": 0.4293,
+      "step": 128
+    },
+    {
+      "epoch": 0.57847533632287,
+      "grad_norm": 0.14993524551391602,
+      "learning_rate": 8.21656050955414e-06,
+      "loss": 0.425,
+      "step": 129
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 0.13857269287109375,
+      "learning_rate": 8.280254777070064e-06,
+      "loss": 0.415,
+      "step": 130
+    },
+    {
+      "epoch": 0.5874439461883408,
+      "grad_norm": 0.13461801409721375,
+      "learning_rate": 8.343949044585989e-06,
+      "loss": 0.4306,
+      "step": 131
+    },
+    {
+      "epoch": 0.5919282511210763,
+      "grad_norm": 0.16194359958171844,
+      "learning_rate": 8.407643312101912e-06,
+      "loss": 0.4129,
+      "step": 132
+    },
+    {
+      "epoch": 0.5964125560538116,
+      "grad_norm": 0.13580960035324097,
+      "learning_rate": 8.471337579617835e-06,
+      "loss": 0.4194,
+      "step": 133
+    },
+    {
+      "epoch": 0.600896860986547,
+      "grad_norm": 0.16961075365543365,
+      "learning_rate": 8.53503184713376e-06,
+      "loss": 0.4301,
+      "step": 134
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 0.1458025574684143,
+      "learning_rate": 8.598726114649683e-06,
+      "loss": 0.4242,
+      "step": 135
+    },
+    {
+      "epoch": 0.6098654708520179,
+      "grad_norm": 0.15790210664272308,
+      "learning_rate": 8.662420382165606e-06,
+      "loss": 0.4293,
+      "step": 136
+    },
+    {
+      "epoch": 0.6143497757847534,
+      "grad_norm": 0.17104558646678925,
+      "learning_rate": 8.726114649681529e-06,
+      "loss": 0.4203,
+      "step": 137
+    },
+    {
+      "epoch": 0.6188340807174888,
+      "grad_norm": 0.16867613792419434,
+      "learning_rate": 8.789808917197454e-06,
+      "loss": 0.4187,
+      "step": 138
+    },
+    {
+      "epoch": 0.6233183856502242,
+      "grad_norm": 0.1832217127084732,
+      "learning_rate": 8.853503184713377e-06,
+      "loss": 0.4229,
+      "step": 139
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 0.15996277332305908,
+      "learning_rate": 8.9171974522293e-06,
+      "loss": 0.4156,
+      "step": 140
+    },
+    {
+      "epoch": 0.6322869955156951,
+      "grad_norm": 0.17733162641525269,
+      "learning_rate": 8.980891719745225e-06,
+      "loss": 0.4281,
+      "step": 141
+    },
+    {
+      "epoch": 0.6367713004484304,
+      "grad_norm": 0.16866809129714966,
+      "learning_rate": 9.044585987261148e-06,
+      "loss": 0.4159,
+      "step": 142
+    },
+    {
+      "epoch": 0.6412556053811659,
+      "grad_norm": 0.16075658798217773,
+      "learning_rate": 9.10828025477707e-06,
+      "loss": 0.4101,
+      "step": 143
+    },
+    {
+      "epoch": 0.6457399103139013,
+      "grad_norm": 0.17321372032165527,
+      "learning_rate": 9.171974522292994e-06,
+      "loss": 0.4158,
+      "step": 144
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 0.18049512803554535,
+      "learning_rate": 9.235668789808919e-06,
+      "loss": 0.4278,
+      "step": 145
+    },
+    {
+      "epoch": 0.6547085201793722,
+      "grad_norm": 0.177161306142807,
+      "learning_rate": 9.299363057324842e-06,
+      "loss": 0.4193,
+      "step": 146
+    },
+    {
+      "epoch": 0.6591928251121076,
+      "grad_norm": 0.17711852490901947,
+      "learning_rate": 9.363057324840765e-06,
+      "loss": 0.4123,
+      "step": 147
+    },
+    {
+      "epoch": 0.6636771300448431,
+      "grad_norm": 0.15281686186790466,
+      "learning_rate": 9.426751592356688e-06,
+      "loss": 0.4177,
+      "step": 148
+    },
+    {
+      "epoch": 0.6681614349775785,
+      "grad_norm": 0.13765569031238556,
+      "learning_rate": 9.490445859872613e-06,
+      "loss": 0.418,
+      "step": 149
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 0.15672436356544495,
+      "learning_rate": 9.554140127388536e-06,
+      "loss": 0.414,
+      "step": 150
+    },
+    {
+      "epoch": 0.6771300448430493,
+      "grad_norm": 0.18714818358421326,
+      "learning_rate": 9.617834394904459e-06,
+      "loss": 0.43,
+      "step": 151
+    },
+    {
+      "epoch": 0.6816143497757847,
+      "grad_norm": 0.15411153435707092,
+      "learning_rate": 9.681528662420384e-06,
+      "loss": 0.4299,
+      "step": 152
+    },
+    {
+      "epoch": 0.6860986547085202,
+      "grad_norm": 0.18314579129219055,
+      "learning_rate": 9.745222929936307e-06,
+      "loss": 0.4251,
+      "step": 153
+    },
+    {
+      "epoch": 0.6905829596412556,
+      "grad_norm": 0.15219919383525848,
+      "learning_rate": 9.80891719745223e-06,
+      "loss": 0.421,
+      "step": 154
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 0.17707888782024384,
+      "learning_rate": 9.872611464968153e-06,
+      "loss": 0.4227,
+      "step": 155
+    },
+    {
+      "epoch": 0.6995515695067265,
+      "grad_norm": 0.16859078407287598,
+      "learning_rate": 9.936305732484078e-06,
+      "loss": 0.4118,
+      "step": 156
+    },
+    {
+      "epoch": 0.7040358744394619,
+      "grad_norm": 0.17086082696914673,
+      "learning_rate": 1e-05,
+      "loss": 0.413,
+      "step": 157
+    },
+    {
+      "epoch": 0.7085201793721974,
+      "grad_norm": 0.17247937619686127,
+      "learning_rate": 9.999987482852865e-06,
+      "loss": 0.4183,
+      "step": 158
+    },
+    {
+      "epoch": 0.7130044843049327,
+      "grad_norm": 0.15963077545166016,
+      "learning_rate": 9.99994993147413e-06,
+      "loss": 0.4204,
+      "step": 159
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 0.18011464178562164,
+      "learning_rate": 9.99988734605181e-06,
+      "loss": 0.4225,
+      "step": 160
+    },
+    {
+      "epoch": 0.7219730941704036,
+      "grad_norm": 0.186809703707695,
+      "learning_rate": 9.999799726899261e-06,
+      "loss": 0.4115,
+      "step": 161
+    },
+    {
+      "epoch": 0.726457399103139,
+      "grad_norm": 0.20191289484500885,
+      "learning_rate": 9.999687074455182e-06,
+      "loss": 0.4149,
+      "step": 162
+    },
+    {
+      "epoch": 0.7309417040358744,
+      "grad_norm": 0.21166710555553436,
+      "learning_rate": 9.999549389283605e-06,
+      "loss": 0.4054,
+      "step": 163
+    },
+    {
+      "epoch": 0.7354260089686099,
+      "grad_norm": 0.1632290780544281,
+      "learning_rate": 9.999386672073902e-06,
+      "loss": 0.4094,
+      "step": 164
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 0.19101959466934204,
+      "learning_rate": 9.999198923640774e-06,
+      "loss": 0.4156,
+      "step": 165
+    },
+    {
+      "epoch": 0.7443946188340808,
+      "grad_norm": 0.1710662543773651,
+      "learning_rate": 9.998986144924253e-06,
+      "loss": 0.4183,
+      "step": 166
+    },
+    {
+      "epoch": 0.7488789237668162,
+      "grad_norm": 0.1906743198633194,
+      "learning_rate": 9.998748336989687e-06,
+      "loss": 0.4169,
+      "step": 167
+    },
+    {
+      "epoch": 0.7533632286995515,
+      "grad_norm": 0.190553680062294,
+      "learning_rate": 9.998485501027752e-06,
+      "loss": 0.4112,
+      "step": 168
+    },
+    {
+      "epoch": 0.757847533632287,
+      "grad_norm": 0.20463505387306213,
+      "learning_rate": 9.998197638354428e-06,
+      "loss": 0.4081,
+      "step": 169
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 0.17837244272232056,
+      "learning_rate": 9.997884750411004e-06,
+      "loss": 0.4105,
+      "step": 170
+    },
+    {
+      "epoch": 0.7668161434977578,
+      "grad_norm": 0.1713983714580536,
+      "learning_rate": 9.997546838764066e-06,
+      "loss": 0.414,
+      "step": 171
+    },
+    {
+      "epoch": 0.7713004484304933,
+      "grad_norm": 0.22886541485786438,
+      "learning_rate": 9.997183905105487e-06,
+      "loss": 0.417,
+      "step": 172
+    },
+    {
+      "epoch": 0.7757847533632287,
+      "grad_norm": 0.208256334066391,
+      "learning_rate": 9.996795951252427e-06,
+      "loss": 0.4032,
+      "step": 173
+    },
+    {
+      "epoch": 0.7802690582959642,
+      "grad_norm": 0.18991126120090485,
+      "learning_rate": 9.996382979147316e-06,
+      "loss": 0.4136,
+      "step": 174
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 0.17698687314987183,
+      "learning_rate": 9.995944990857848e-06,
+      "loss": 0.4172,
+      "step": 175
+    },
+    {
+      "epoch": 0.7892376681614349,
+      "grad_norm": 0.19628863036632538,
+      "learning_rate": 9.995481988576968e-06,
+      "loss": 0.4129,
+      "step": 176
+    },
+    {
+      "epoch": 0.7937219730941704,
+      "grad_norm": 0.19764673709869385,
+      "learning_rate": 9.994993974622863e-06,
+      "loss": 0.4001,
+      "step": 177
+    },
+    {
+      "epoch": 0.7982062780269058,
+      "grad_norm": 0.16025379300117493,
+      "learning_rate": 9.994480951438948e-06,
+      "loss": 0.407,
+      "step": 178
+    },
+    {
+      "epoch": 0.8026905829596412,
+      "grad_norm": 0.16668112576007843,
+      "learning_rate": 9.993942921593858e-06,
+      "loss": 0.4076,
+      "step": 179
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 0.18416959047317505,
+      "learning_rate": 9.993379887781436e-06,
+      "loss": 0.4149,
+      "step": 180
+    },
+    {
+      "epoch": 0.8116591928251121,
+      "grad_norm": 0.18400655686855316,
+      "learning_rate": 9.992791852820709e-06,
+      "loss": 0.4213,
+      "step": 181
+    },
+    {
+      "epoch": 0.8161434977578476,
+      "grad_norm": 0.19249750673770905,
+      "learning_rate": 9.992178819655889e-06,
+      "loss": 0.4113,
+      "step": 182
+    },
+    {
+      "epoch": 0.820627802690583,
+      "grad_norm": 0.17832814157009125,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.4133,
+      "step": 183
+    },
+    {
+      "epoch": 0.8251121076233184,
+      "grad_norm": 0.19760073721408844,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.4111,
+      "step": 184
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 0.15243801474571228,
+      "learning_rate": 9.990189762256275e-06,
+      "loss": 0.4086,
+      "step": 185
+    },
+    {
+      "epoch": 0.8340807174887892,
+      "grad_norm": 0.19544097781181335,
+      "learning_rate": 9.98947676822017e-06,
+      "loss": 0.4116,
+      "step": 186
+    },
+    {
+      "epoch": 0.8385650224215246,
+      "grad_norm": 0.15919189155101776,
+      "learning_rate": 9.988738792578126e-06,
+      "loss": 0.4088,
+      "step": 187
+    },
+    {
+      "epoch": 0.8430493273542601,
+      "grad_norm": 0.1675063967704773,
+      "learning_rate": 9.987975839025091e-06,
+      "loss": 0.4089,
+      "step": 188
+    },
+    {
+      "epoch": 0.8475336322869955,
+      "grad_norm": 0.16887347400188446,
+      "learning_rate": 9.987187911381059e-06,
+      "loss": 0.4074,
+      "step": 189
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 0.1690843552350998,
+      "learning_rate": 9.986375013591077e-06,
+      "loss": 0.4199,
+      "step": 190
+    },
+    {
+      "epoch": 0.8565022421524664,
+      "grad_norm": 0.18787801265716553,
+      "learning_rate": 9.985537149725207e-06,
+      "loss": 0.4147,
+      "step": 191
+    },
+    {
+      "epoch": 0.8609865470852018,
+      "grad_norm": 0.16457174718379974,
+      "learning_rate": 9.984674323978517e-06,
+      "loss": 0.4127,
+      "step": 192
+    },
+    {
+      "epoch": 0.8654708520179372,
+      "grad_norm": 0.2218443602323532,
+      "learning_rate": 9.983786540671052e-06,
+      "loss": 0.4118,
+      "step": 193
+    },
+    {
+      "epoch": 0.8699551569506726,
+      "grad_norm": 0.22543732821941376,
+      "learning_rate": 9.982873804247817e-06,
+      "loss": 0.4136,
+      "step": 194
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 0.1941734105348587,
+      "learning_rate": 9.981936119278758e-06,
+      "loss": 0.3951,
+      "step": 195
+    },
+    {
+      "epoch": 0.8789237668161435,
+      "grad_norm": 0.2310221642255783,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.4019,
+      "step": 196
+    },
+    {
+      "epoch": 0.8834080717488789,
+      "grad_norm": 0.24600385129451752,
+      "learning_rate": 9.979985922607476e-06,
+      "loss": 0.4048,
+      "step": 197
+    },
+    {
+      "epoch": 0.8878923766816144,
+      "grad_norm": 0.23484165966510773,
+      "learning_rate": 9.978973420669613e-06,
+      "loss": 0.4195,
+      "step": 198
+    },
+    {
+      "epoch": 0.8923766816143498,
+      "grad_norm": 0.1665085256099701,
+      "learning_rate": 9.977935989714594e-06,
+      "loss": 0.4065,
+      "step": 199
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 0.24713504314422607,
+      "learning_rate": 9.97687363493669e-06,
+      "loss": 0.4025,
+      "step": 200
+    },
+    {
+      "epoch": 0.9013452914798207,
+      "grad_norm": 0.20779001712799072,
+      "learning_rate": 9.975786361654959e-06,
+      "loss": 0.4065,
+      "step": 201
+    },
+    {
+      "epoch": 0.905829596412556,
+      "grad_norm": 0.16626589000225067,
+      "learning_rate": 9.974674175313228e-06,
+      "loss": 0.4145,
+      "step": 202
+    },
+    {
+      "epoch": 0.9103139013452914,
+      "grad_norm": 0.20812493562698364,
+      "learning_rate": 9.973537081480056e-06,
+      "loss": 0.3997,
+      "step": 203
+    },
+    {
+      "epoch": 0.9147982062780269,
+      "grad_norm": 0.15901827812194824,
+      "learning_rate": 9.972375085848712e-06,
+      "loss": 0.4026,
+      "step": 204
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 0.1733086109161377,
+      "learning_rate": 9.971188194237141e-06,
+      "loss": 0.4119,
+      "step": 205
+    },
+    {
+      "epoch": 0.9237668161434978,
+      "grad_norm": 0.213147833943367,
+      "learning_rate": 9.969976412587943e-06,
+      "loss": 0.413,
+      "step": 206
+    },
+    {
+      "epoch": 0.9282511210762332,
+      "grad_norm": 0.17450547218322754,
+      "learning_rate": 9.96873974696834e-06,
+      "loss": 0.3994,
+      "step": 207
+    },
+    {
+      "epoch": 0.9327354260089686,
+      "grad_norm": 0.171011820435524,
+      "learning_rate": 9.967478203570143e-06,
+      "loss": 0.4024,
+      "step": 208
+    },
+    {
+      "epoch": 0.9372197309417041,
+      "grad_norm": 0.19780731201171875,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.4083,
+      "step": 209
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 0.16172127425670624,
+      "learning_rate": 9.96488050882796e-06,
+      "loss": 0.4108,
+      "step": 210
+    },
+    {
+      "epoch": 0.9461883408071748,
+      "grad_norm": 0.1879730522632599,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.4042,
+      "step": 211
+    },
+    {
+      "epoch": 0.9506726457399103,
+      "grad_norm": 0.2026575803756714,
+      "learning_rate": 9.9621833803865e-06,
+      "loss": 0.4041,
+      "step": 212
+    },
+    {
+      "epoch": 0.9551569506726457,
+      "grad_norm": 0.17352183163166046,
+      "learning_rate": 9.960797545330936e-06,
+      "loss": 0.3989,
+      "step": 213
+    },
+    {
+      "epoch": 0.9596412556053812,
+      "grad_norm": 0.17416566610336304,
+      "learning_rate": 9.95938687226226e-06,
+      "loss": 0.4068,
+      "step": 214
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 0.20091062784194946,
+      "learning_rate": 9.95795136824351e-06,
+      "loss": 0.3932,
+      "step": 215
+    },
+    {
+      "epoch": 0.968609865470852,
+      "grad_norm": 0.15297354757785797,
+      "learning_rate": 9.956491040462051e-06,
+      "loss": 0.3999,
+      "step": 216
+    },
+    {
+      "epoch": 0.9730941704035875,
+      "grad_norm": 0.17038391530513763,
+      "learning_rate": 9.955005896229543e-06,
+      "loss": 0.402,
+      "step": 217
+    },
+    {
+      "epoch": 0.9775784753363229,
+      "grad_norm": 0.16770429909229279,
+      "learning_rate": 9.95349594298189e-06,
+      "loss": 0.4075,
+      "step": 218
+    },
+    {
+      "epoch": 0.9820627802690582,
+      "grad_norm": 0.1569359004497528,
+      "learning_rate": 9.951961188279216e-06,
+      "loss": 0.3993,
+      "step": 219
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 0.16684497892856598,
+      "learning_rate": 9.950401639805822e-06,
+      "loss": 0.4072,
+      "step": 220
+    },
+    {
+      "epoch": 0.9910313901345291,
+      "grad_norm": 0.174079030752182,
+      "learning_rate": 9.948817305370145e-06,
+      "loss": 0.4031,
+      "step": 221
+    },
+    {
+      "epoch": 0.9955156950672646,
+      "grad_norm": 0.15056130290031433,
+      "learning_rate": 9.947208192904722e-06,
+      "loss": 0.3997,
+      "step": 222
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1775795966386795,
+      "learning_rate": 9.94557431046616e-06,
+      "loss": 0.3891,
+      "step": 223
+    },
+    {
+      "epoch": 1.0044843049327354,
+      "grad_norm": 0.1755944937467575,
+      "learning_rate": 9.943915666235068e-06,
+      "loss": 0.3913,
+      "step": 224
+    },
+    {
+      "epoch": 1.0089686098654709,
+      "grad_norm": 0.16261902451515198,
+      "learning_rate": 9.942232268516051e-06,
+      "loss": 0.3901,
+      "step": 225
+    },
+    {
+      "epoch": 1.0134529147982063,
+      "grad_norm": 0.16443823277950287,
+      "learning_rate": 9.940524125737641e-06,
+      "loss": 0.3955,
+      "step": 226
+    },
+    {
+      "epoch": 1.0179372197309418,
+      "grad_norm": 0.1699223667383194,
+      "learning_rate": 9.938791246452267e-06,
+      "loss": 0.3773,
+      "step": 227
+    },
+    {
+      "epoch": 1.0224215246636772,
+      "grad_norm": 0.18911561369895935,
+      "learning_rate": 9.937033639336212e-06,
+      "loss": 0.3897,
+      "step": 228
+    },
+    {
+      "epoch": 1.0269058295964126,
+      "grad_norm": 0.17428790032863617,
+      "learning_rate": 9.935251313189564e-06,
+      "loss": 0.3939,
+      "step": 229
+    },
+    {
+      "epoch": 1.031390134529148,
+      "grad_norm": 0.1741434931755066,
+      "learning_rate": 9.933444276936185e-06,
+      "loss": 0.3959,
+      "step": 230
+    },
+    {
+      "epoch": 1.0358744394618835,
+      "grad_norm": 0.1734904795885086,
+      "learning_rate": 9.931612539623643e-06,
+      "loss": 0.3963,
+      "step": 231
+    },
+    {
+      "epoch": 1.0403587443946187,
+      "grad_norm": 0.16929706931114197,
+      "learning_rate": 9.929756110423195e-06,
+      "loss": 0.3955,
+      "step": 232
+    },
+    {
+      "epoch": 1.0448430493273542,
+      "grad_norm": 0.20600971579551697,
+      "learning_rate": 9.927874998629714e-06,
+      "loss": 0.3902,
+      "step": 233
+    },
+    {
+      "epoch": 1.0493273542600896,
+      "grad_norm": 0.16491708159446716,
+      "learning_rate": 9.925969213661664e-06,
+      "loss": 0.3854,
+      "step": 234
+    },
+    {
+      "epoch": 1.053811659192825,
+      "grad_norm": 0.18594156205654144,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.3971,
+      "step": 235
+    },
+    {
+      "epoch": 1.0582959641255605,
+      "grad_norm": 0.19441187381744385,
+      "learning_rate": 9.92208366249333e-06,
+      "loss": 0.3872,
+      "step": 236
+    },
+    {
+      "epoch": 1.062780269058296,
+      "grad_norm": 0.16155117750167847,
+      "learning_rate": 9.920103915747452e-06,
+      "loss": 0.3917,
+      "step": 237
+    },
+    {
+      "epoch": 1.0672645739910314,
+      "grad_norm": 0.1920769065618515,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.3864,
+      "step": 238
+    },
+    {
+      "epoch": 1.0717488789237668,
+      "grad_norm": 0.15873126685619354,
+      "learning_rate": 9.916070529493785e-06,
+      "loss": 0.3893,
+      "step": 239
+    },
+    {
+      "epoch": 1.0762331838565022,
+      "grad_norm": 0.2126314491033554,
+      "learning_rate": 9.914016910180593e-06,
+      "loss": 0.3938,
+      "step": 240
+    },
+    {
+      "epoch": 1.0807174887892377,
+      "grad_norm": 0.17595785856246948,
+      "learning_rate": 9.911938687078324e-06,
+      "loss": 0.3889,
+      "step": 241
+    },
+    {
+      "epoch": 1.0852017937219731,
+      "grad_norm": 0.17708002030849457,
+      "learning_rate": 9.90983587059235e-06,
+      "loss": 0.391,
+      "step": 242
+    },
+    {
+      "epoch": 1.0896860986547086,
+      "grad_norm": 0.1790732443332672,
+      "learning_rate": 9.907708471251173e-06,
+      "loss": 0.3904,
+      "step": 243
+    },
+    {
+      "epoch": 1.094170403587444,
+      "grad_norm": 0.1829553246498108,
+      "learning_rate": 9.905556499706384e-06,
+      "loss": 0.3888,
+      "step": 244
+    },
+    {
+      "epoch": 1.0986547085201794,
+      "grad_norm": 0.15539512038230896,
+      "learning_rate": 9.9033799667326e-06,
+      "loss": 0.3836,
+      "step": 245
+    },
+    {
+      "epoch": 1.1031390134529149,
+      "grad_norm": 0.1803758591413498,
+      "learning_rate": 9.901178883227413e-06,
+      "loss": 0.3965,
+      "step": 246
+    },
+    {
+      "epoch": 1.1076233183856503,
+      "grad_norm": 0.1554258018732071,
+      "learning_rate": 9.89895326021134e-06,
+      "loss": 0.395,
+      "step": 247
+    },
+    {
+      "epoch": 1.1121076233183858,
+      "grad_norm": 0.17221762239933014,
+      "learning_rate": 9.896703108827758e-06,
+      "loss": 0.3794,
+      "step": 248
+    },
+    {
+      "epoch": 1.116591928251121,
+      "grad_norm": 0.1603137105703354,
+      "learning_rate": 9.89442844034286e-06,
+      "loss": 0.393,
+      "step": 249
+    },
+    {
+      "epoch": 1.1210762331838564,
+      "grad_norm": 0.15259714424610138,
+      "learning_rate": 9.89212926614559e-06,
+      "loss": 0.3907,
+      "step": 250
+    },
+    {
+      "epoch": 1.1255605381165918,
+      "grad_norm": 0.15605948865413666,
+      "learning_rate": 9.889805597747588e-06,
+      "loss": 0.3971,
+      "step": 251
+    },
+    {
+      "epoch": 1.1300448430493273,
+      "grad_norm": 0.1493992805480957,
+      "learning_rate": 9.887457446783133e-06,
+      "loss": 0.3928,
+      "step": 252
+    },
+    {
+      "epoch": 1.1345291479820627,
+      "grad_norm": 0.17648091912269592,
+      "learning_rate": 9.885084825009085e-06,
+      "loss": 0.3956,
+      "step": 253
+    },
+    {
+      "epoch": 1.1390134529147982,
+      "grad_norm": 0.15433448553085327,
+      "learning_rate": 9.88268774430483e-06,
+      "loss": 0.3981,
+      "step": 254
+    },
+    {
+      "epoch": 1.1434977578475336,
+      "grad_norm": 0.1576976329088211,
+      "learning_rate": 9.88026621667221e-06,
+      "loss": 0.3797,
+      "step": 255
+    },
+    {
+      "epoch": 1.147982062780269,
+      "grad_norm": 0.1615358144044876,
+      "learning_rate": 9.87782025423547e-06,
+      "loss": 0.3907,
+      "step": 256
+    },
+    {
+      "epoch": 1.1524663677130045,
+      "grad_norm": 0.17831331491470337,
+      "learning_rate": 9.875349869241202e-06,
+      "loss": 0.3877,
+      "step": 257
+    },
+    {
+      "epoch": 1.15695067264574,
+      "grad_norm": 0.18192827701568604,
+      "learning_rate": 9.872855074058274e-06,
+      "loss": 0.3876,
+      "step": 258
+    },
+    {
+      "epoch": 1.1614349775784754,
+      "grad_norm": 0.16451925039291382,
+      "learning_rate": 9.870335881177774e-06,
+      "loss": 0.3992,
+      "step": 259
+    },
+    {
+      "epoch": 1.1659192825112108,
+      "grad_norm": 0.22533132135868073,
+      "learning_rate": 9.867792303212944e-06,
+      "loss": 0.3916,
+      "step": 260
+    },
+    {
+      "epoch": 1.1704035874439462,
+      "grad_norm": 0.15392068028450012,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.3789,
+      "step": 261
+    },
+    {
+      "epoch": 1.1748878923766817,
+      "grad_norm": 0.18406599760055542,
+      "learning_rate": 9.862632043093667e-06,
+      "loss": 0.3919,
+      "step": 262
+    },
+    {
+      "epoch": 1.1793721973094171,
+      "grad_norm": 0.16980576515197754,
+      "learning_rate": 9.860015386775915e-06,
+      "loss": 0.4031,
+      "step": 263
+    },
+    {
+      "epoch": 1.1838565022421526,
+      "grad_norm": 0.19829297065734863,
+      "learning_rate": 9.857374397047093e-06,
+      "loss": 0.3939,
+      "step": 264
+    },
+    {
+      "epoch": 1.188340807174888,
+      "grad_norm": 0.19180873036384583,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.3878,
+      "step": 265
+    },
+    {
+      "epoch": 1.1928251121076232,
+      "grad_norm": 0.18134139478206635,
+      "learning_rate": 9.852019470370254e-06,
+      "loss": 0.3935,
+      "step": 266
+    },
+    {
+      "epoch": 1.1973094170403586,
+      "grad_norm": 0.20495562255382538,
+      "learning_rate": 9.849305560233598e-06,
+      "loss": 0.3901,
+      "step": 267
+    },
+    {
+      "epoch": 1.201793721973094,
+      "grad_norm": 0.17293596267700195,
+      "learning_rate": 9.846567370308462e-06,
+      "loss": 0.3881,
+      "step": 268
+    },
+    {
+      "epoch": 1.2062780269058295,
+      "grad_norm": 0.1871306300163269,
+      "learning_rate": 9.843804914304578e-06,
+      "loss": 0.384,
+      "step": 269
+    },
+    {
+      "epoch": 1.210762331838565,
+      "grad_norm": 0.16608139872550964,
+      "learning_rate": 9.841018206053167e-06,
+      "loss": 0.3884,
+      "step": 270
+    },
+    {
+      "epoch": 1.2152466367713004,
+      "grad_norm": 0.18193687498569489,
+      "learning_rate": 9.838207259506891e-06,
+      "loss": 0.3893,
+      "step": 271
+    },
+    {
+      "epoch": 1.2197309417040358,
+      "grad_norm": 0.17928655445575714,
+      "learning_rate": 9.835372088739758e-06,
+      "loss": 0.3874,
+      "step": 272
+    },
+    {
+      "epoch": 1.2242152466367713,
+      "grad_norm": 0.16808190941810608,
+      "learning_rate": 9.83251270794707e-06,
+      "loss": 0.3902,
+      "step": 273
+    },
+    {
+      "epoch": 1.2286995515695067,
+      "grad_norm": 0.20220190286636353,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.3855,
+      "step": 274
+    },
+    {
+      "epoch": 1.2331838565022422,
+      "grad_norm": 0.16028372943401337,
+      "learning_rate": 9.826721373672235e-06,
+      "loss": 0.3863,
+      "step": 275
+    },
+    {
+      "epoch": 1.2376681614349776,
+      "grad_norm": 0.2248627245426178,
+      "learning_rate": 9.82378944918648e-06,
+      "loss": 0.399,
+      "step": 276
+    },
+    {
+      "epoch": 1.242152466367713,
+      "grad_norm": 0.19712041318416595,
+      "learning_rate": 9.820833372667813e-06,
+      "loss": 0.3825,
+      "step": 277
+    },
+    {
+      "epoch": 1.2466367713004485,
+      "grad_norm": 0.18009068071842194,
+      "learning_rate": 9.817853158916889e-06,
+      "loss": 0.3877,
+      "step": 278
+    },
+    {
+      "epoch": 1.251121076233184,
+      "grad_norm": 0.1869373470544815,
+      "learning_rate": 9.814848822855216e-06,
+      "loss": 0.3856,
+      "step": 279
+    },
+    {
+      "epoch": 1.2556053811659194,
+      "grad_norm": 0.16159535944461823,
+      "learning_rate": 9.811820379525085e-06,
+      "loss": 0.385,
+      "step": 280
+    },
+    {
+      "epoch": 1.2600896860986546,
+      "grad_norm": 0.19267477095127106,
+      "learning_rate": 9.80876784408948e-06,
+      "loss": 0.3813,
+      "step": 281
+    },
+    {
+      "epoch": 1.2645739910313902,
+      "grad_norm": 0.1528576761484146,
+      "learning_rate": 9.805691231832018e-06,
+      "loss": 0.3855,
+      "step": 282
+    },
+    {
+      "epoch": 1.2690582959641254,
+      "grad_norm": 0.21759846806526184,
+      "learning_rate": 9.802590558156863e-06,
+      "loss": 0.3895,
+      "step": 283
+    },
+    {
+      "epoch": 1.273542600896861,
+      "grad_norm": 0.17682035267353058,
+      "learning_rate": 9.799465838588646e-06,
+      "loss": 0.3925,
+      "step": 284
+    },
+    {
+      "epoch": 1.2780269058295963,
+      "grad_norm": 0.21090620756149292,
+      "learning_rate": 9.796317088772402e-06,
+      "loss": 0.3866,
+      "step": 285
+    },
+    {
+      "epoch": 1.2825112107623318,
+      "grad_norm": 0.20534247159957886,
+      "learning_rate": 9.793144324473473e-06,
+      "loss": 0.3927,
+      "step": 286
+    },
+    {
+      "epoch": 1.2869955156950672,
+      "grad_norm": 0.18179374933242798,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.3908,
+      "step": 287
+    },
+    {
+      "epoch": 1.2914798206278026,
+      "grad_norm": 0.19916561245918274,
+      "learning_rate": 9.786726816090058e-06,
+      "loss": 0.3866,
+      "step": 288
+    },
+    {
+      "epoch": 1.295964125560538,
+      "grad_norm": 0.22441346943378448,
+      "learning_rate": 9.783482104137127e-06,
+      "loss": 0.3841,
+      "step": 289
+    },
+    {
+      "epoch": 1.3004484304932735,
+      "grad_norm": 0.19861099123954773,
+      "learning_rate": 9.78021344196447e-06,
+      "loss": 0.3847,
+      "step": 290
+    },
+    {
+      "epoch": 1.304932735426009,
+      "grad_norm": 0.2116602659225464,
+      "learning_rate": 9.776920845937816e-06,
+      "loss": 0.3892,
+      "step": 291
+    },
+    {
+      "epoch": 1.3094170403587444,
+      "grad_norm": 0.20515024662017822,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.3858,
+      "step": 292
+    },
+    {
+      "epoch": 1.3139013452914798,
+      "grad_norm": 0.21357493102550507,
+      "learning_rate": 9.770263918384523e-06,
+      "loss": 0.3808,
+      "step": 293
+    },
+    {
+      "epoch": 1.3183856502242153,
+      "grad_norm": 0.1979387253522873,
+      "learning_rate": 9.766899620188181e-06,
+      "loss": 0.3854,
+      "step": 294
+    },
+    {
+      "epoch": 1.3228699551569507,
+      "grad_norm": 0.20448064804077148,
+      "learning_rate": 9.763511454798268e-06,
+      "loss": 0.3885,
+      "step": 295
+    },
+    {
+      "epoch": 1.3273542600896862,
+      "grad_norm": 0.22471775114536285,
+      "learning_rate": 9.760099439178852e-06,
+      "loss": 0.397,
+      "step": 296
+    },
+    {
+      "epoch": 1.3318385650224216,
+      "grad_norm": 0.1656077355146408,
+      "learning_rate": 9.75666359041341e-06,
+      "loss": 0.3852,
+      "step": 297
+    },
+    {
+      "epoch": 1.336322869955157,
+      "grad_norm": 0.18242590129375458,
+      "learning_rate": 9.753203925704756e-06,
+      "loss": 0.3852,
+      "step": 298
+    },
+    {
+      "epoch": 1.3408071748878925,
+      "grad_norm": 0.1829691082239151,
+      "learning_rate": 9.749720462374939e-06,
+      "loss": 0.3857,
+      "step": 299
+    },
+    {
+      "epoch": 1.3452914798206277,
+      "grad_norm": 0.13810646533966064,
+      "learning_rate": 9.74621321786517e-06,
+      "loss": 0.3854,
+      "step": 300
+    },
+    {
+      "epoch": 1.3497757847533634,
+      "grad_norm": 0.1642601191997528,
+      "learning_rate": 9.742682209735727e-06,
+      "loss": 0.3854,
+      "step": 301
+    },
+    {
+      "epoch": 1.3542600896860986,
+      "grad_norm": 0.15604956448078156,
+      "learning_rate": 9.739127455665872e-06,
+      "loss": 0.3892,
+      "step": 302
+    },
+    {
+      "epoch": 1.358744394618834,
+      "grad_norm": 0.15881405770778656,
+      "learning_rate": 9.735548973453753e-06,
+      "loss": 0.381,
+      "step": 303
+    },
+    {
+      "epoch": 1.3632286995515694,
+      "grad_norm": 0.18863636255264282,
+      "learning_rate": 9.731946781016328e-06,
+      "loss": 0.3841,
+      "step": 304
+    },
+    {
+      "epoch": 1.3677130044843049,
+      "grad_norm": 0.1484975814819336,
+      "learning_rate": 9.728320896389263e-06,
+      "loss": 0.375,
+      "step": 305
+    },
+    {
+      "epoch": 1.3721973094170403,
+      "grad_norm": 0.15026456117630005,
+      "learning_rate": 9.724671337726855e-06,
+      "loss": 0.3825,
+      "step": 306
+    },
+    {
+      "epoch": 1.3766816143497758,
+      "grad_norm": 0.14361456036567688,
+      "learning_rate": 9.720998123301924e-06,
+      "loss": 0.3773,
+      "step": 307
+    },
+    {
+      "epoch": 1.3811659192825112,
+      "grad_norm": 0.15960386395454407,
+      "learning_rate": 9.717301271505739e-06,
+      "loss": 0.3837,
+      "step": 308
+    },
+    {
+      "epoch": 1.3856502242152466,
+      "grad_norm": 0.14065828919410706,
+      "learning_rate": 9.713580800847917e-06,
+      "loss": 0.3881,
+      "step": 309
+    },
+    {
+      "epoch": 1.390134529147982,
+      "grad_norm": 0.14989614486694336,
+      "learning_rate": 9.709836729956326e-06,
+      "loss": 0.3749,
+      "step": 310
+    },
+    {
+      "epoch": 1.3946188340807175,
+      "grad_norm": 0.1599772423505783,
+      "learning_rate": 9.706069077577e-06,
+      "loss": 0.3835,
+      "step": 311
+    },
+    {
+      "epoch": 1.399103139013453,
+      "grad_norm": 0.13850939273834229,
+      "learning_rate": 9.702277862574048e-06,
+      "loss": 0.3778,
+      "step": 312
+    },
+    {
+      "epoch": 1.4035874439461884,
+      "grad_norm": 0.15607166290283203,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3858,
+      "step": 313
+    },
+    {
+      "epoch": 1.4080717488789238,
+      "grad_norm": 0.1551503986120224,
+      "learning_rate": 9.694624820743446e-06,
+      "loss": 0.3849,
+      "step": 314
+    },
+    {
+      "epoch": 1.4125560538116593,
+      "grad_norm": 0.13863934576511383,
+      "learning_rate": 9.690763032233498e-06,
+      "loss": 0.3902,
+      "step": 315
+    },
+    {
+      "epoch": 1.4170403587443947,
+      "grad_norm": 0.14638298749923706,
+      "learning_rate": 9.686877757735126e-06,
+      "loss": 0.3728,
+      "step": 316
+    },
+    {
+      "epoch": 1.42152466367713,
+      "grad_norm": 0.14968304336071014,
+      "learning_rate": 9.682969016701357e-06,
+      "loss": 0.3835,
+      "step": 317
+    },
+    {
+      "epoch": 1.4260089686098656,
+      "grad_norm": 0.15045005083084106,
+      "learning_rate": 9.679036828702704e-06,
+      "loss": 0.3868,
+      "step": 318
+    },
+    {
+      "epoch": 1.4304932735426008,
+      "grad_norm": 0.15328407287597656,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.3781,
+      "step": 319
+    },
+    {
+      "epoch": 1.4349775784753362,
+      "grad_norm": 0.15265707671642303,
+      "learning_rate": 9.671102190679678e-06,
+      "loss": 0.3957,
+      "step": 320
+    },
+    {
+      "epoch": 1.4394618834080717,
+      "grad_norm": 0.15088792145252228,
+      "learning_rate": 9.66709978038292e-06,
+      "loss": 0.3817,
+      "step": 321
+    },
+    {
+      "epoch": 1.4439461883408071,
+      "grad_norm": 0.16192200779914856,
+      "learning_rate": 9.663074002576303e-06,
+      "loss": 0.3848,
+      "step": 322
+    },
+    {
+      "epoch": 1.4484304932735426,
+      "grad_norm": 0.14499373733997345,
+      "learning_rate": 9.659024877416328e-06,
+      "loss": 0.3883,
+      "step": 323
+    },
+    {
+      "epoch": 1.452914798206278,
+      "grad_norm": 0.16166694462299347,
+      "learning_rate": 9.654952425176392e-06,
+      "loss": 0.3918,
+      "step": 324
+    },
+    {
+      "epoch": 1.4573991031390134,
+      "grad_norm": 0.15286611020565033,
+      "learning_rate": 9.650856666246693e-06,
+      "loss": 0.387,
+      "step": 325
+    },
+    {
+      "epoch": 1.4618834080717489,
+      "grad_norm": 0.13531741499900818,
+      "learning_rate": 9.646737621134112e-06,
+      "loss": 0.3813,
+      "step": 326
+    },
+    {
+      "epoch": 1.4663677130044843,
+      "grad_norm": 0.15205958485603333,
+      "learning_rate": 9.642595310462133e-06,
+      "loss": 0.387,
+      "step": 327
+    },
+    {
+      "epoch": 1.4708520179372198,
+      "grad_norm": 0.15055085718631744,
+      "learning_rate": 9.638429754970715e-06,
+      "loss": 0.3799,
+      "step": 328
+    },
+    {
+      "epoch": 1.4753363228699552,
+      "grad_norm": 0.14826399087905884,
+      "learning_rate": 9.63424097551621e-06,
+      "loss": 0.3937,
+      "step": 329
+    },
+    {
+      "epoch": 1.4798206278026906,
+      "grad_norm": 0.14727114140987396,
+      "learning_rate": 9.630028993071244e-06,
+      "loss": 0.3817,
+      "step": 330
+    },
+    {
+      "epoch": 1.484304932735426,
+      "grad_norm": 0.15946896374225616,
+      "learning_rate": 9.62579382872462e-06,
+      "loss": 0.3846,
+      "step": 331
+    },
+    {
+      "epoch": 1.4887892376681615,
+      "grad_norm": 0.15146270394325256,
+      "learning_rate": 9.621535503681205e-06,
+      "loss": 0.3826,
+      "step": 332
+    },
+    {
+      "epoch": 1.493273542600897,
+      "grad_norm": 0.17063474655151367,
+      "learning_rate": 9.617254039261835e-06,
+      "loss": 0.3751,
+      "step": 333
+    },
+    {
+      "epoch": 1.4977578475336322,
+      "grad_norm": 0.1559085249900818,
+      "learning_rate": 9.612949456903195e-06,
+      "loss": 0.3846,
+      "step": 334
+    },
+    {
+      "epoch": 1.5022421524663678,
+      "grad_norm": 0.16825082898139954,
+      "learning_rate": 9.608621778157722e-06,
+      "loss": 0.388,
+      "step": 335
+    },
+    {
+      "epoch": 1.506726457399103,
+      "grad_norm": 0.1682816594839096,
+      "learning_rate": 9.604271024693495e-06,
+      "loss": 0.3889,
+      "step": 336
+    },
+    {
+      "epoch": 1.5112107623318387,
+      "grad_norm": 0.1359148770570755,
+      "learning_rate": 9.599897218294122e-06,
+      "loss": 0.3809,
+      "step": 337
+    },
+    {
+      "epoch": 1.515695067264574,
+      "grad_norm": 0.14568015933036804,
+      "learning_rate": 9.595500380858632e-06,
+      "loss": 0.3815,
+      "step": 338
+    },
+    {
+      "epoch": 1.5201793721973094,
+      "grad_norm": 0.1531810760498047,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.387,
+      "step": 339
+    },
+    {
+      "epoch": 1.5246636771300448,
+      "grad_norm": 0.14324067533016205,
+      "learning_rate": 9.586637701051886e-06,
+      "loss": 0.3859,
+      "step": 340
+    },
+    {
+      "epoch": 1.5291479820627802,
+      "grad_norm": 0.15393510460853577,
+      "learning_rate": 9.582171903054815e-06,
+      "loss": 0.3881,
+      "step": 341
+    },
+    {
+      "epoch": 1.5336322869955157,
+      "grad_norm": 0.1713324636220932,
+      "learning_rate": 9.577683162769781e-06,
+      "loss": 0.3916,
+      "step": 342
+    },
+    {
+      "epoch": 1.5381165919282511,
+      "grad_norm": 0.12945277988910675,
+      "learning_rate": 9.573171502671273e-06,
+      "loss": 0.38,
+      "step": 343
+    },
+    {
+      "epoch": 1.5426008968609866,
+      "grad_norm": 0.17458106577396393,
+      "learning_rate": 9.568636945348534e-06,
+      "loss": 0.3905,
+      "step": 344
+    },
+    {
+      "epoch": 1.547085201793722,
+      "grad_norm": 0.17422664165496826,
+      "learning_rate": 9.564079513505455e-06,
+      "loss": 0.3909,
+      "step": 345
+    },
+    {
+      "epoch": 1.5515695067264574,
+      "grad_norm": 0.15175458788871765,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.3833,
+      "step": 346
+    },
+    {
+      "epoch": 1.5560538116591929,
+      "grad_norm": 0.16197986900806427,
+      "learning_rate": 9.554896117646357e-06,
+      "loss": 0.381,
+      "step": 347
+    },
+    {
+      "epoch": 1.5605381165919283,
+      "grad_norm": 0.1600109040737152,
+      "learning_rate": 9.550270199610307e-06,
+      "loss": 0.3816,
+      "step": 348
+    },
+    {
+      "epoch": 1.5650224215246635,
+      "grad_norm": 0.1753740906715393,
+      "learning_rate": 9.54562149901362e-06,
+      "loss": 0.3794,
+      "step": 349
+    },
+    {
+      "epoch": 1.5695067264573992,
+      "grad_norm": 0.15255777537822723,
+      "learning_rate": 9.54095003913168e-06,
+      "loss": 0.3872,
+      "step": 350
+    },
+    {
+      "epoch": 1.5739910313901344,
+      "grad_norm": 0.1964779943227768,
+      "learning_rate": 9.536255843353832e-06,
+      "loss": 0.3809,
+      "step": 351
+    },
+    {
+      "epoch": 1.57847533632287,
+      "grad_norm": 0.15007811784744263,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.3813,
+      "step": 352
+    },
+    {
+      "epoch": 1.5829596412556053,
+      "grad_norm": 0.1757270246744156,
+      "learning_rate": 9.526799338236828e-06,
+      "loss": 0.3796,
+      "step": 353
+    },
+    {
+      "epoch": 1.587443946188341,
+      "grad_norm": 0.16849809885025024,
+      "learning_rate": 9.522037076245057e-06,
+      "loss": 0.3774,
+      "step": 354
+    },
+    {
+      "epoch": 1.5919282511210762,
+      "grad_norm": 0.15899567306041718,
+      "learning_rate": 9.517252173051912e-06,
+      "loss": 0.3865,
+      "step": 355
+    },
+    {
+      "epoch": 1.5964125560538116,
+      "grad_norm": 0.19069278240203857,
+      "learning_rate": 9.512444652614728e-06,
+      "loss": 0.381,
+      "step": 356
+    },
+    {
+      "epoch": 1.600896860986547,
+      "grad_norm": 0.15457352995872498,
+      "learning_rate": 9.507614539004082e-06,
+      "loss": 0.376,
+      "step": 357
+    },
+    {
+      "epoch": 1.6053811659192825,
+      "grad_norm": 0.17245250940322876,
+      "learning_rate": 9.502761856403668e-06,
+      "loss": 0.3841,
+      "step": 358
+    },
+    {
+      "epoch": 1.609865470852018,
+      "grad_norm": 0.15316694974899292,
+      "learning_rate": 9.497886629110187e-06,
+      "loss": 0.3887,
+      "step": 359
+    },
+    {
+      "epoch": 1.6143497757847534,
+      "grad_norm": 0.15494902431964874,
+      "learning_rate": 9.49298888153321e-06,
+      "loss": 0.3794,
+      "step": 360
+    },
+    {
+      "epoch": 1.6188340807174888,
+      "grad_norm": 0.17157073318958282,
+      "learning_rate": 9.488068638195072e-06,
+      "loss": 0.3802,
+      "step": 361
+    },
+    {
+      "epoch": 1.6233183856502242,
+      "grad_norm": 0.16301631927490234,
+      "learning_rate": 9.483125923730731e-06,
+      "loss": 0.3773,
+      "step": 362
+    },
+    {
+      "epoch": 1.6278026905829597,
+      "grad_norm": 0.19093914330005646,
+      "learning_rate": 9.478160762887668e-06,
+      "loss": 0.3839,
+      "step": 363
+    },
+    {
+      "epoch": 1.6322869955156951,
+      "grad_norm": 0.166624516248703,
+      "learning_rate": 9.473173180525737e-06,
+      "loss": 0.3731,
+      "step": 364
+    },
+    {
+      "epoch": 1.6367713004484306,
+      "grad_norm": 0.19319555163383484,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.3786,
+      "step": 365
+    },
+    {
+      "epoch": 1.6412556053811658,
+      "grad_norm": 0.17385593056678772,
+      "learning_rate": 9.463130851245899e-06,
+      "loss": 0.3791,
+      "step": 366
+    },
+    {
+      "epoch": 1.6457399103139014,
+      "grad_norm": 0.1947043389081955,
+      "learning_rate": 9.458076154608515e-06,
+      "loss": 0.372,
+      "step": 367
+    },
+    {
+      "epoch": 1.6502242152466366,
+      "grad_norm": 0.19737760722637177,
+      "learning_rate": 9.452999137013064e-06,
+      "loss": 0.3863,
+      "step": 368
+    },
+    {
+      "epoch": 1.6547085201793723,
+      "grad_norm": 0.16445475816726685,
+      "learning_rate": 9.447899823879456e-06,
+      "loss": 0.374,
+      "step": 369
+    },
+    {
+      "epoch": 1.6591928251121075,
+      "grad_norm": 0.19131121039390564,
+      "learning_rate": 9.442778240739234e-06,
+      "loss": 0.3778,
+      "step": 370
+    },
+    {
+      "epoch": 1.6636771300448432,
+      "grad_norm": 0.2015213668346405,
+      "learning_rate": 9.437634413235437e-06,
+      "loss": 0.3763,
+      "step": 371
+    },
+    {
+      "epoch": 1.6681614349775784,
+      "grad_norm": 0.15377578139305115,
+      "learning_rate": 9.43246836712249e-06,
+      "loss": 0.377,
+      "step": 372
+    },
+    {
+      "epoch": 1.672645739910314,
+      "grad_norm": 0.1949223130941391,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.3847,
+      "step": 373
+    },
+    {
+      "epoch": 1.6771300448430493,
+      "grad_norm": 0.17129598557949066,
+      "learning_rate": 9.4220697226429e-06,
+      "loss": 0.3792,
+      "step": 374
+    },
+    {
+      "epoch": 1.6816143497757847,
+      "grad_norm": 0.17037433385849,
+      "learning_rate": 9.416837176340807e-06,
+      "loss": 0.3739,
+      "step": 375
+    },
+    {
+      "epoch": 1.6860986547085202,
+      "grad_norm": 0.21392875909805298,
+      "learning_rate": 9.411582515558391e-06,
+      "loss": 0.373,
+      "step": 376
+    },
+    {
+      "epoch": 1.6905829596412556,
+      "grad_norm": 0.1607356071472168,
+      "learning_rate": 9.406305766604996e-06,
+      "loss": 0.3776,
+      "step": 377
+    },
+    {
+      "epoch": 1.695067264573991,
+      "grad_norm": 0.22532962262630463,
+      "learning_rate": 9.401006955900555e-06,
+      "loss": 0.3847,
+      "step": 378
+    },
+    {
+      "epoch": 1.6995515695067265,
+      "grad_norm": 0.1926848441362381,
+      "learning_rate": 9.395686109975475e-06,
+      "loss": 0.3795,
+      "step": 379
+    },
+    {
+      "epoch": 1.704035874439462,
+      "grad_norm": 0.16258005797863007,
+      "learning_rate": 9.390343255470471e-06,
+      "loss": 0.3799,
+      "step": 380
+    },
+    {
+      "epoch": 1.7085201793721974,
+      "grad_norm": 0.19488303363323212,
+      "learning_rate": 9.384978419136469e-06,
+      "loss": 0.3734,
+      "step": 381
+    },
+    {
+      "epoch": 1.7130044843049328,
+      "grad_norm": 0.20306065678596497,
+      "learning_rate": 9.37959162783444e-06,
+      "loss": 0.3866,
+      "step": 382
+    },
+    {
+      "epoch": 1.717488789237668,
+      "grad_norm": 0.20845548808574677,
+      "learning_rate": 9.374182908535293e-06,
+      "loss": 0.381,
+      "step": 383
+    },
+    {
+      "epoch": 1.7219730941704037,
+      "grad_norm": 0.18295307457447052,
+      "learning_rate": 9.368752288319722e-06,
+      "loss": 0.3926,
+      "step": 384
+    },
+    {
+      "epoch": 1.726457399103139,
+      "grad_norm": 0.17709247767925262,
+      "learning_rate": 9.363299794378072e-06,
+      "loss": 0.3788,
+      "step": 385
+    },
+    {
+      "epoch": 1.7309417040358746,
+      "grad_norm": 0.171927809715271,
+      "learning_rate": 9.357825454010214e-06,
+      "loss": 0.3818,
+      "step": 386
+    },
+    {
+      "epoch": 1.7354260089686098,
+      "grad_norm": 0.18400821089744568,
+      "learning_rate": 9.352329294625397e-06,
+      "loss": 0.3884,
+      "step": 387
+    },
+    {
+      "epoch": 1.7399103139013454,
+      "grad_norm": 0.17629793286323547,
+      "learning_rate": 9.346811343742115e-06,
+      "loss": 0.3778,
+      "step": 388
+    },
+    {
+      "epoch": 1.7443946188340806,
+      "grad_norm": 0.17265784740447998,
+      "learning_rate": 9.34127162898797e-06,
+      "loss": 0.3779,
+      "step": 389
+    },
+    {
+      "epoch": 1.7488789237668163,
+      "grad_norm": 0.1599327027797699,
+      "learning_rate": 9.335710178099528e-06,
+      "loss": 0.3793,
+      "step": 390
+    },
+    {
+      "epoch": 1.7533632286995515,
+      "grad_norm": 0.145245760679245,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.3778,
+      "step": 391
+    },
+    {
+      "epoch": 1.757847533632287,
+      "grad_norm": 0.19388413429260254,
+      "learning_rate": 9.324522179410054e-06,
+      "loss": 0.3864,
+      "step": 392
+    },
+    {
+      "epoch": 1.7623318385650224,
+      "grad_norm": 0.17556312680244446,
+      "learning_rate": 9.318895687625752e-06,
+      "loss": 0.3804,
+      "step": 393
+    },
+    {
+      "epoch": 1.7668161434977578,
+      "grad_norm": 0.17071861028671265,
+      "learning_rate": 9.313247571740336e-06,
+      "loss": 0.3836,
+      "step": 394
+    },
+    {
+      "epoch": 1.7713004484304933,
+      "grad_norm": 0.18459686636924744,
+      "learning_rate": 9.307577860033123e-06,
+      "loss": 0.3643,
+      "step": 395
+    },
+    {
+      "epoch": 1.7757847533632287,
+      "grad_norm": 0.16499431431293488,
+      "learning_rate": 9.301886580891563e-06,
+      "loss": 0.3779,
+      "step": 396
+    },
+    {
+      "epoch": 1.7802690582959642,
+      "grad_norm": 0.17812196910381317,
+      "learning_rate": 9.296173762811084e-06,
+      "loss": 0.3709,
+      "step": 397
+    },
+    {
+      "epoch": 1.7847533632286996,
+      "grad_norm": 0.17254163324832916,
+      "learning_rate": 9.290439434394965e-06,
+      "loss": 0.3782,
+      "step": 398
+    },
+    {
+      "epoch": 1.789237668161435,
+      "grad_norm": 0.17717133462429047,
+      "learning_rate": 9.284683624354172e-06,
+      "loss": 0.3829,
+      "step": 399
+    },
+    {
+      "epoch": 1.7937219730941703,
+      "grad_norm": 0.19152995944023132,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.3785,
+      "step": 400
+    },
+    {
+      "epoch": 1.798206278026906,
+      "grad_norm": 0.1961972564458847,
+      "learning_rate": 9.273107674780102e-06,
+      "loss": 0.379,
+      "step": 401
+    },
+    {
+      "epoch": 1.8026905829596411,
+      "grad_norm": 0.2076851725578308,
+      "learning_rate": 9.26728759320597e-06,
+      "loss": 0.3846,
+      "step": 402
+    },
+    {
+      "epoch": 1.8071748878923768,
+      "grad_norm": 0.2117864042520523,
+      "learning_rate": 9.261446145925167e-06,
+      "loss": 0.3775,
+      "step": 403
+    },
+    {
+      "epoch": 1.811659192825112,
+      "grad_norm": 0.20487642288208008,
+      "learning_rate": 9.255583362184998e-06,
+      "loss": 0.3843,
+      "step": 404
+    },
+    {
+      "epoch": 1.8161434977578477,
+      "grad_norm": 0.18826477229595184,
+      "learning_rate": 9.249699271339594e-06,
+      "loss": 0.3789,
+      "step": 405
+    },
+    {
+      "epoch": 1.8206278026905829,
+      "grad_norm": 0.19905836880207062,
+      "learning_rate": 9.243793902849764e-06,
+      "loss": 0.3766,
+      "step": 406
+    },
+    {
+      "epoch": 1.8251121076233185,
+      "grad_norm": 0.21211640536785126,
+      "learning_rate": 9.237867286282855e-06,
+      "loss": 0.3748,
+      "step": 407
+    },
+    {
+      "epoch": 1.8295964125560538,
+      "grad_norm": 0.1588459610939026,
+      "learning_rate": 9.231919451312603e-06,
+      "loss": 0.3776,
+      "step": 408
+    },
+    {
+      "epoch": 1.8340807174887892,
+      "grad_norm": 0.2558506727218628,
+      "learning_rate": 9.225950427718974e-06,
+      "loss": 0.3696,
+      "step": 409
+    },
+    {
+      "epoch": 1.8385650224215246,
+      "grad_norm": 0.16340257227420807,
+      "learning_rate": 9.21996024538803e-06,
+      "loss": 0.3833,
+      "step": 410
+    },
+    {
+      "epoch": 1.84304932735426,
+      "grad_norm": 0.22385142743587494,
+      "learning_rate": 9.213948934311767e-06,
+      "loss": 0.3833,
+      "step": 411
+    },
+    {
+      "epoch": 1.8475336322869955,
+      "grad_norm": 0.19165366888046265,
+      "learning_rate": 9.207916524587971e-06,
+      "loss": 0.3757,
+      "step": 412
+    },
+    {
+      "epoch": 1.852017937219731,
+      "grad_norm": 0.18462036550045013,
+      "learning_rate": 9.201863046420065e-06,
+      "loss": 0.3688,
+      "step": 413
+    },
+    {
+      "epoch": 1.8565022421524664,
+      "grad_norm": 0.23066729307174683,
+      "learning_rate": 9.195788530116962e-06,
+      "loss": 0.3785,
+      "step": 414
+    },
+    {
+      "epoch": 1.8609865470852018,
+      "grad_norm": 0.16391463577747345,
+      "learning_rate": 9.189693006092907e-06,
+      "loss": 0.381,
+      "step": 415
+    },
+    {
+      "epoch": 1.8654708520179373,
+      "grad_norm": 0.24376490712165833,
+      "learning_rate": 9.183576504867328e-06,
+      "loss": 0.3806,
+      "step": 416
+    },
+    {
+      "epoch": 1.8699551569506725,
+      "grad_norm": 0.17985792458057404,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.3821,
+      "step": 417
+    },
+    {
+      "epoch": 1.8744394618834082,
+      "grad_norm": 0.18049661815166473,
+      "learning_rate": 9.171280693414307e-06,
+      "loss": 0.3757,
+      "step": 418
+    },
+    {
+      "epoch": 1.8789237668161434,
+      "grad_norm": 0.21507222950458527,
+      "learning_rate": 9.165101444750259e-06,
+      "loss": 0.3762,
+      "step": 419
+    },
+    {
+      "epoch": 1.883408071748879,
+      "grad_norm": 0.18147873878479004,
+      "learning_rate": 9.158901342011163e-06,
+      "loss": 0.3769,
+      "step": 420
+    },
+    {
+      "epoch": 1.8878923766816142,
+      "grad_norm": 0.1572391539812088,
+      "learning_rate": 9.152680416240059e-06,
+      "loss": 0.3773,
+      "step": 421
+    },
+    {
+      "epoch": 1.89237668161435,
+      "grad_norm": 0.17494268715381622,
+      "learning_rate": 9.146438698584245e-06,
+      "loss": 0.3808,
+      "step": 422
+    },
+    {
+      "epoch": 1.8968609865470851,
+      "grad_norm": 0.1536751389503479,
+      "learning_rate": 9.140176220295118e-06,
+      "loss": 0.3785,
+      "step": 423
+    },
+    {
+      "epoch": 1.9013452914798208,
+      "grad_norm": 0.29661592841148376,
+      "learning_rate": 9.133893012728027e-06,
+      "loss": 0.3736,
+      "step": 424
+    },
+    {
+      "epoch": 1.905829596412556,
+      "grad_norm": 0.14534755051136017,
+      "learning_rate": 9.1275891073421e-06,
+      "loss": 0.3761,
+      "step": 425
+    },
+    {
+      "epoch": 1.9103139013452914,
+      "grad_norm": 0.1410803347826004,
+      "learning_rate": 9.121264535700107e-06,
+      "loss": 0.3735,
+      "step": 426
+    },
+    {
+      "epoch": 1.9147982062780269,
+      "grad_norm": 0.15303485095500946,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.3765,
+      "step": 427
+    },
+    {
+      "epoch": 1.9192825112107623,
+      "grad_norm": 0.1633804738521576,
+      "learning_rate": 9.10855352041618e-06,
+      "loss": 0.38,
+      "step": 428
+    },
+    {
+      "epoch": 1.9237668161434978,
+      "grad_norm": 0.13735759258270264,
+      "learning_rate": 9.102167140416503e-06,
+      "loss": 0.3797,
+      "step": 429
+    },
+    {
+      "epoch": 1.9282511210762332,
+      "grad_norm": 0.17561876773834229,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.3816,
+      "step": 430
+    },
+    {
+      "epoch": 1.9327354260089686,
+      "grad_norm": 0.13969717919826508,
+      "learning_rate": 9.089332795580085e-06,
+      "loss": 0.3759,
+      "step": 431
+    },
+    {
+      "epoch": 1.937219730941704,
+      "grad_norm": 0.16025178134441376,
+      "learning_rate": 9.082884895003098e-06,
+      "loss": 0.3805,
+      "step": 432
+    },
+    {
+      "epoch": 1.9417040358744395,
+      "grad_norm": 0.1353815644979477,
+      "learning_rate": 9.076416551997721e-06,
+      "loss": 0.3816,
+      "step": 433
+    },
+    {
+      "epoch": 1.9461883408071747,
+      "grad_norm": 0.15468056499958038,
+      "learning_rate": 9.06992779895004e-06,
+      "loss": 0.3701,
+      "step": 434
+    },
+    {
+      "epoch": 1.9506726457399104,
+      "grad_norm": 0.13822025060653687,
+      "learning_rate": 9.063418668348323e-06,
+      "loss": 0.3772,
+      "step": 435
+    },
+    {
+      "epoch": 1.9551569506726456,
+      "grad_norm": 0.16018016636371613,
+      "learning_rate": 9.056889192782865e-06,
+      "loss": 0.3771,
+      "step": 436
+    },
+    {
+      "epoch": 1.9596412556053813,
+      "grad_norm": 0.1503971666097641,
+      "learning_rate": 9.050339404945834e-06,
+      "loss": 0.3757,
+      "step": 437
+    },
+    {
+      "epoch": 1.9641255605381165,
+      "grad_norm": 0.14734366536140442,
+      "learning_rate": 9.04376933763109e-06,
+      "loss": 0.3799,
+      "step": 438
+    },
+    {
+      "epoch": 1.9686098654708521,
+      "grad_norm": 0.16386902332305908,
+      "learning_rate": 9.037179023734036e-06,
+      "loss": 0.3781,
+      "step": 439
+    },
+    {
+      "epoch": 1.9730941704035874,
+      "grad_norm": 0.17295579612255096,
+      "learning_rate": 9.030568496251437e-06,
+      "loss": 0.3698,
+      "step": 440
+    },
+    {
+      "epoch": 1.977578475336323,
+      "grad_norm": 0.1508246809244156,
+      "learning_rate": 9.023937788281278e-06,
+      "loss": 0.3824,
+      "step": 441
+    },
+    {
+      "epoch": 1.9820627802690582,
+      "grad_norm": 0.17555788159370422,
+      "learning_rate": 9.017286933022573e-06,
+      "loss": 0.3791,
+      "step": 442
+    },
+    {
+      "epoch": 1.9865470852017937,
+      "grad_norm": 0.16138505935668945,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.3762,
+      "step": 443
+    },
+    {
+      "epoch": 1.9910313901345291,
+      "grad_norm": 0.1586380898952484,
+      "learning_rate": 9.003924913939816e-06,
+      "loss": 0.3818,
+      "step": 444
+    },
+    {
+      "epoch": 1.9955156950672646,
+      "grad_norm": 0.1761392503976822,
+      "learning_rate": 8.997213817017508e-06,
+      "loss": 0.3796,
+      "step": 445
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.15980252623558044,
+      "learning_rate": 8.990482706609805e-06,
+      "loss": 0.3739,
+      "step": 446
+    },
+    {
+      "epoch": 2.004484304932735,
+      "grad_norm": 0.16549459099769592,
+      "learning_rate": 8.98373161641843e-06,
+      "loss": 0.3582,
+      "step": 447
+    },
+    {
+      "epoch": 2.008968609865471,
+      "grad_norm": 0.152572900056839,
+      "learning_rate": 8.97696058024514e-06,
+      "loss": 0.3624,
+      "step": 448
+    },
+    {
+      "epoch": 2.013452914798206,
+      "grad_norm": 0.17323534190654755,
+      "learning_rate": 8.970169631991556e-06,
+      "loss": 0.36,
+      "step": 449
+    },
+    {
+      "epoch": 2.0179372197309418,
+      "grad_norm": 0.15496471524238586,
+      "learning_rate": 8.963358805658998e-06,
+      "loss": 0.3667,
+      "step": 450
+    },
+    {
+      "epoch": 2.022421524663677,
+      "grad_norm": 0.17136932909488678,
+      "learning_rate": 8.95652813534831e-06,
+      "loss": 0.3585,
+      "step": 451
+    },
+    {
+      "epoch": 2.0269058295964126,
+      "grad_norm": 0.1474735140800476,
+      "learning_rate": 8.949677655259696e-06,
+      "loss": 0.3471,
+      "step": 452
+    },
+    {
+      "epoch": 2.031390134529148,
+      "grad_norm": 0.17043280601501465,
+      "learning_rate": 8.942807399692543e-06,
+      "loss": 0.3657,
+      "step": 453
+    },
+    {
+      "epoch": 2.0358744394618835,
+      "grad_norm": 0.1686600148677826,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.3666,
+      "step": 454
+    },
+    {
+      "epoch": 2.0403587443946187,
+      "grad_norm": 0.1648789793252945,
+      "learning_rate": 8.92900769981506e-06,
+      "loss": 0.3644,
+      "step": 455
+    },
+    {
+      "epoch": 2.0448430493273544,
+      "grad_norm": 0.17585590481758118,
+      "learning_rate": 8.92207832459788e-06,
+      "loss": 0.3516,
+      "step": 456
+    },
+    {
+      "epoch": 2.0493273542600896,
+      "grad_norm": 0.17197641730308533,
+      "learning_rate": 8.915129312088112e-06,
+      "loss": 0.3662,
+      "step": 457
+    },
+    {
+      "epoch": 2.0538116591928253,
+      "grad_norm": 0.17358921468257904,
+      "learning_rate": 8.908160697078482e-06,
+      "loss": 0.3635,
+      "step": 458
+    },
+    {
+      "epoch": 2.0582959641255605,
+      "grad_norm": 0.19035573303699493,
+      "learning_rate": 8.901172514459864e-06,
+      "loss": 0.3624,
+      "step": 459
+    },
+    {
+      "epoch": 2.062780269058296,
+      "grad_norm": 0.1628907471895218,
+      "learning_rate": 8.894164799221099e-06,
+      "loss": 0.36,
+      "step": 460
+    },
+    {
+      "epoch": 2.0672645739910314,
+      "grad_norm": 0.1739048957824707,
+      "learning_rate": 8.88713758644883e-06,
+      "loss": 0.3612,
+      "step": 461
+    },
+    {
+      "epoch": 2.071748878923767,
+      "grad_norm": 0.17438630759716034,
+      "learning_rate": 8.88009091132732e-06,
+      "loss": 0.3515,
+      "step": 462
+    },
+    {
+      "epoch": 2.0762331838565022,
+      "grad_norm": 0.1770601123571396,
+      "learning_rate": 8.873024809138272e-06,
+      "loss": 0.3611,
+      "step": 463
+    },
+    {
+      "epoch": 2.0807174887892375,
+      "grad_norm": 0.16999252140522003,
+      "learning_rate": 8.86593931526067e-06,
+      "loss": 0.3518,
+      "step": 464
+    },
+    {
+      "epoch": 2.085201793721973,
+      "grad_norm": 0.1970926821231842,
+      "learning_rate": 8.858834465170576e-06,
+      "loss": 0.3545,
+      "step": 465
+    },
+    {
+      "epoch": 2.0896860986547083,
+      "grad_norm": 0.18241076171398163,
+      "learning_rate": 8.851710294440974e-06,
+      "loss": 0.3584,
+      "step": 466
+    },
+    {
+      "epoch": 2.094170403587444,
+      "grad_norm": 0.1480012685060501,
+      "learning_rate": 8.84456683874158e-06,
+      "loss": 0.3518,
+      "step": 467
+    },
+    {
+      "epoch": 2.098654708520179,
+      "grad_norm": 0.18665406107902527,
+      "learning_rate": 8.837404133838667e-06,
+      "loss": 0.3559,
+      "step": 468
+    },
+    {
+      "epoch": 2.103139013452915,
+      "grad_norm": 0.15539565682411194,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3569,
+      "step": 469
+    },
+    {
+      "epoch": 2.10762331838565,
+      "grad_norm": 0.16088882088661194,
+      "learning_rate": 8.823021119969102e-06,
+      "loss": 0.3597,
+      "step": 470
+    },
+    {
+      "epoch": 2.1121076233183858,
+      "grad_norm": 0.18173345923423767,
+      "learning_rate": 8.815800883016168e-06,
+      "loss": 0.361,
+      "step": 471
+    },
+    {
+      "epoch": 2.116591928251121,
+      "grad_norm": 0.1613433063030243,
+      "learning_rate": 8.808561540886796e-06,
+      "loss": 0.3623,
+      "step": 472
+    },
+    {
+      "epoch": 2.1210762331838566,
+      "grad_norm": 0.1875883787870407,
+      "learning_rate": 8.801303129827352e-06,
+      "loss": 0.3618,
+      "step": 473
+    },
+    {
+      "epoch": 2.125560538116592,
+      "grad_norm": 0.1790507435798645,
+      "learning_rate": 8.794025686179677e-06,
+      "loss": 0.3642,
+      "step": 474
+    },
+    {
+      "epoch": 2.1300448430493275,
+      "grad_norm": 0.21571233868598938,
+      "learning_rate": 8.786729246380901e-06,
+      "loss": 0.3533,
+      "step": 475
+    },
+    {
+      "epoch": 2.1345291479820627,
+      "grad_norm": 0.14037518203258514,
+      "learning_rate": 8.779413846963267e-06,
+      "loss": 0.356,
+      "step": 476
+    },
+    {
+      "epoch": 2.1390134529147984,
+      "grad_norm": 0.19560834765434265,
+      "learning_rate": 8.772079524553951e-06,
+      "loss": 0.3557,
+      "step": 477
+    },
+    {
+      "epoch": 2.1434977578475336,
+      "grad_norm": 0.20778882503509521,
+      "learning_rate": 8.764726315874872e-06,
+      "loss": 0.361,
+      "step": 478
+    },
+    {
+      "epoch": 2.1479820627802693,
+      "grad_norm": 0.1969592422246933,
+      "learning_rate": 8.757354257742501e-06,
+      "loss": 0.3632,
+      "step": 479
+    },
+    {
+      "epoch": 2.1524663677130045,
+      "grad_norm": 0.1780160367488861,
+      "learning_rate": 8.749963387067697e-06,
+      "loss": 0.3571,
+      "step": 480
+    },
+    {
+      "epoch": 2.15695067264574,
+      "grad_norm": 0.2403516322374344,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.3565,
+      "step": 481
+    },
+    {
+      "epoch": 2.1614349775784754,
+      "grad_norm": 0.16860368847846985,
+      "learning_rate": 8.735125356204982e-06,
+      "loss": 0.3592,
+      "step": 482
+    },
+    {
+      "epoch": 2.1659192825112106,
+      "grad_norm": 0.1759265810251236,
+      "learning_rate": 8.727678270308994e-06,
+      "loss": 0.3594,
+      "step": 483
+    },
+    {
+      "epoch": 2.1704035874439462,
+      "grad_norm": 0.1871018260717392,
+      "learning_rate": 8.720212520454054e-06,
+      "loss": 0.3595,
+      "step": 484
+    },
+    {
+      "epoch": 2.1748878923766815,
+      "grad_norm": 0.1505175530910492,
+      "learning_rate": 8.712728144020118e-06,
+      "loss": 0.3614,
+      "step": 485
+    },
+    {
+      "epoch": 2.179372197309417,
+      "grad_norm": 0.174392431974411,
+      "learning_rate": 8.705225178480397e-06,
+      "loss": 0.3662,
+      "step": 486
+    },
+    {
+      "epoch": 2.1838565022421523,
+      "grad_norm": 0.1607576608657837,
+      "learning_rate": 8.697703661401187e-06,
+      "loss": 0.3568,
+      "step": 487
+    },
+    {
+      "epoch": 2.188340807174888,
+      "grad_norm": 0.16113710403442383,
+      "learning_rate": 8.69016363044166e-06,
+      "loss": 0.3537,
+      "step": 488
+    },
+    {
+      "epoch": 2.192825112107623,
+      "grad_norm": 0.16741614043712616,
+      "learning_rate": 8.682605123353685e-06,
+      "loss": 0.3576,
+      "step": 489
+    },
+    {
+      "epoch": 2.197309417040359,
+      "grad_norm": 0.16952870786190033,
+      "learning_rate": 8.675028177981643e-06,
+      "loss": 0.368,
+      "step": 490
+    },
+    {
+      "epoch": 2.201793721973094,
+      "grad_norm": 0.16490712761878967,
+      "learning_rate": 8.66743283226223e-06,
+      "loss": 0.3532,
+      "step": 491
+    },
+    {
+      "epoch": 2.2062780269058297,
+      "grad_norm": 0.16877040266990662,
+      "learning_rate": 8.659819124224266e-06,
+      "loss": 0.3553,
+      "step": 492
+    },
+    {
+      "epoch": 2.210762331838565,
+      "grad_norm": 0.1317160725593567,
+      "learning_rate": 8.652187091988516e-06,
+      "loss": 0.3536,
+      "step": 493
+    },
+    {
+      "epoch": 2.2152466367713006,
+      "grad_norm": 0.17306114733219147,
+      "learning_rate": 8.644536773767488e-06,
+      "loss": 0.3605,
+      "step": 494
+    },
+    {
+      "epoch": 2.219730941704036,
+      "grad_norm": 0.15019871294498444,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.3531,
+      "step": 495
+    },
+    {
+      "epoch": 2.2242152466367715,
+      "grad_norm": 0.17941038310527802,
+      "learning_rate": 8.629181432677213e-06,
+      "loss": 0.3561,
+      "step": 496
+    },
+    {
+      "epoch": 2.2286995515695067,
+      "grad_norm": 0.15172863006591797,
+      "learning_rate": 8.621476486689991e-06,
+      "loss": 0.3555,
+      "step": 497
+    },
+    {
+      "epoch": 2.233183856502242,
+      "grad_norm": 0.1934608668088913,
+      "learning_rate": 8.613753408481158e-06,
+      "loss": 0.3478,
+      "step": 498
+    },
+    {
+      "epoch": 2.2376681614349776,
+      "grad_norm": 0.16764213144779205,
+      "learning_rate": 8.606012236719073e-06,
+      "loss": 0.3598,
+      "step": 499
+    },
+    {
+      "epoch": 2.242152466367713,
+      "grad_norm": 0.210286945104599,
+      "learning_rate": 8.598253010162693e-06,
+      "loss": 0.3643,
+      "step": 500
+    },
+    {
+      "epoch": 2.2466367713004485,
+      "grad_norm": 0.1573864221572876,
+      "learning_rate": 8.590475767661371e-06,
+      "loss": 0.3493,
+      "step": 501
+    },
+    {
+      "epoch": 2.2511210762331837,
+      "grad_norm": 0.18853873014450073,
+      "learning_rate": 8.58268054815466e-06,
+      "loss": 0.3585,
+      "step": 502
+    },
+    {
+      "epoch": 2.2556053811659194,
+      "grad_norm": 0.1775168776512146,
+      "learning_rate": 8.574867390672124e-06,
+      "loss": 0.3613,
+      "step": 503
+    },
+    {
+      "epoch": 2.2600896860986546,
+      "grad_norm": 0.1714695543050766,
+      "learning_rate": 8.567036334333142e-06,
+      "loss": 0.3687,
+      "step": 504
+    },
+    {
+      "epoch": 2.2645739910313902,
+      "grad_norm": 0.1601528525352478,
+      "learning_rate": 8.559187418346703e-06,
+      "loss": 0.36,
+      "step": 505
+    },
+    {
+      "epoch": 2.2690582959641254,
+      "grad_norm": 0.15437829494476318,
+      "learning_rate": 8.551320682011227e-06,
+      "loss": 0.3602,
+      "step": 506
+    },
+    {
+      "epoch": 2.273542600896861,
+      "grad_norm": 0.14641445875167847,
+      "learning_rate": 8.543436164714351e-06,
+      "loss": 0.358,
+      "step": 507
+    },
+    {
+      "epoch": 2.2780269058295963,
+      "grad_norm": 0.15655145049095154,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.3594,
+      "step": 508
+    },
+    {
+      "epoch": 2.282511210762332,
+      "grad_norm": 0.1548921763896942,
+      "learning_rate": 8.527613945231886e-06,
+      "loss": 0.3683,
+      "step": 509
+    },
+    {
+      "epoch": 2.286995515695067,
+      "grad_norm": 0.14546792209148407,
+      "learning_rate": 8.519676322265914e-06,
+      "loss": 0.3663,
+      "step": 510
+    },
+    {
+      "epoch": 2.291479820627803,
+      "grad_norm": 0.1500290334224701,
+      "learning_rate": 8.511721076777388e-06,
+      "loss": 0.3639,
+      "step": 511
+    },
+    {
+      "epoch": 2.295964125560538,
+      "grad_norm": 0.16159754991531372,
+      "learning_rate": 8.503748248597095e-06,
+      "loss": 0.3637,
+      "step": 512
+    },
+    {
+      "epoch": 2.3004484304932733,
+      "grad_norm": 0.13845127820968628,
+      "learning_rate": 8.495757877643857e-06,
+      "loss": 0.3581,
+      "step": 513
+    },
+    {
+      "epoch": 2.304932735426009,
+      "grad_norm": 0.18970300257205963,
+      "learning_rate": 8.487750003924341e-06,
+      "loss": 0.3616,
+      "step": 514
+    },
+    {
+      "epoch": 2.3094170403587446,
+      "grad_norm": 0.17019912600517273,
+      "learning_rate": 8.479724667532836e-06,
+      "loss": 0.3634,
+      "step": 515
+    },
+    {
+      "epoch": 2.31390134529148,
+      "grad_norm": 0.17842921614646912,
+      "learning_rate": 8.471681908651067e-06,
+      "loss": 0.3559,
+      "step": 516
+    },
+    {
+      "epoch": 2.318385650224215,
+      "grad_norm": 0.1670839786529541,
+      "learning_rate": 8.463621767547998e-06,
+      "loss": 0.3465,
+      "step": 517
+    },
+    {
+      "epoch": 2.3228699551569507,
+      "grad_norm": 0.18545055389404297,
+      "learning_rate": 8.455544284579614e-06,
+      "loss": 0.3607,
+      "step": 518
+    },
+    {
+      "epoch": 2.327354260089686,
+      "grad_norm": 0.18138788640499115,
+      "learning_rate": 8.447449500188731e-06,
+      "loss": 0.3631,
+      "step": 519
+    },
+    {
+      "epoch": 2.3318385650224216,
+      "grad_norm": 0.17482097446918488,
+      "learning_rate": 8.439337454904794e-06,
+      "loss": 0.3521,
+      "step": 520
+    },
+    {
+      "epoch": 2.336322869955157,
+      "grad_norm": 0.15842925012111664,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.3634,
+      "step": 521
+    },
+    {
+      "epoch": 2.3408071748878925,
+      "grad_norm": 0.18594925105571747,
+      "learning_rate": 8.42306174420744e-06,
+      "loss": 0.3614,
+      "step": 522
+    },
+    {
+      "epoch": 2.3452914798206277,
+      "grad_norm": 0.1544816493988037,
+      "learning_rate": 8.414898160284208e-06,
+      "loss": 0.3649,
+      "step": 523
+    },
+    {
+      "epoch": 2.3497757847533634,
+      "grad_norm": 0.16842401027679443,
+      "learning_rate": 8.406717478447889e-06,
+      "loss": 0.3616,
+      "step": 524
+    },
+    {
+      "epoch": 2.3542600896860986,
+      "grad_norm": 0.15589796006679535,
+      "learning_rate": 8.398519739657997e-06,
+      "loss": 0.3649,
+      "step": 525
+    },
+    {
+      "epoch": 2.3587443946188342,
+      "grad_norm": 0.16493292152881622,
+      "learning_rate": 8.390304984959455e-06,
+      "loss": 0.3589,
+      "step": 526
+    },
+    {
+      "epoch": 2.3632286995515694,
+      "grad_norm": 0.17189329862594604,
+      "learning_rate": 8.382073255482381e-06,
+      "loss": 0.3581,
+      "step": 527
+    },
+    {
+      "epoch": 2.367713004484305,
+      "grad_norm": 0.16451218724250793,
+      "learning_rate": 8.373824592441884e-06,
+      "loss": 0.3591,
+      "step": 528
+    },
+    {
+      "epoch": 2.3721973094170403,
+      "grad_norm": 0.15957431495189667,
+      "learning_rate": 8.36555903713785e-06,
+      "loss": 0.357,
+      "step": 529
+    },
+    {
+      "epoch": 2.376681614349776,
+      "grad_norm": 0.16711632907390594,
+      "learning_rate": 8.357276630954756e-06,
+      "loss": 0.3563,
+      "step": 530
+    },
+    {
+      "epoch": 2.381165919282511,
+      "grad_norm": 0.1526818871498108,
+      "learning_rate": 8.348977415361435e-06,
+      "loss": 0.3603,
+      "step": 531
+    },
+    {
+      "epoch": 2.3856502242152464,
+      "grad_norm": 0.1482716053724289,
+      "learning_rate": 8.340661431910888e-06,
+      "loss": 0.3584,
+      "step": 532
+    },
+    {
+      "epoch": 2.390134529147982,
+      "grad_norm": 0.175686314702034,
+      "learning_rate": 8.332328722240072e-06,
+      "loss": 0.37,
+      "step": 533
+    },
+    {
+      "epoch": 2.3946188340807173,
+      "grad_norm": 0.14500749111175537,
+      "learning_rate": 8.323979328069689e-06,
+      "loss": 0.3635,
+      "step": 534
+    },
+    {
+      "epoch": 2.399103139013453,
+      "grad_norm": 0.16548028588294983,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.3582,
+      "step": 535
+    },
+    {
+      "epoch": 2.403587443946188,
+      "grad_norm": 0.1462414413690567,
+      "learning_rate": 8.307230653530501e-06,
+      "loss": 0.3608,
+      "step": 536
+    },
+    {
+      "epoch": 2.408071748878924,
+      "grad_norm": 0.1521567851305008,
+      "learning_rate": 8.298831457019943e-06,
+      "loss": 0.3562,
+      "step": 537
+    },
+    {
+      "epoch": 2.412556053811659,
+      "grad_norm": 0.18554925918579102,
+      "learning_rate": 8.290415743725894e-06,
+      "loss": 0.3601,
+      "step": 538
+    },
+    {
+      "epoch": 2.4170403587443947,
+      "grad_norm": 0.14416217803955078,
+      "learning_rate": 8.28198355578465e-06,
+      "loss": 0.359,
+      "step": 539
+    },
+    {
+      "epoch": 2.42152466367713,
+      "grad_norm": 0.16843508183956146,
+      "learning_rate": 8.273534935414975e-06,
+      "loss": 0.3564,
+      "step": 540
+    },
+    {
+      "epoch": 2.4260089686098656,
+      "grad_norm": 0.15784554183483124,
+      "learning_rate": 8.265069924917925e-06,
+      "loss": 0.3578,
+      "step": 541
+    },
+    {
+      "epoch": 2.430493273542601,
+      "grad_norm": 0.14942175149917603,
+      "learning_rate": 8.256588566676616e-06,
+      "loss": 0.3615,
+      "step": 542
+    },
+    {
+      "epoch": 2.4349775784753365,
+      "grad_norm": 0.1575714498758316,
+      "learning_rate": 8.248090903156003e-06,
+      "loss": 0.3513,
+      "step": 543
+    },
+    {
+      "epoch": 2.4394618834080717,
+      "grad_norm": 0.1543831080198288,
+      "learning_rate": 8.239576976902694e-06,
+      "loss": 0.365,
+      "step": 544
+    },
+    {
+      "epoch": 2.4439461883408073,
+      "grad_norm": 0.14899148046970367,
+      "learning_rate": 8.231046830544716e-06,
+      "loss": 0.3433,
+      "step": 545
+    },
+    {
+      "epoch": 2.4484304932735426,
+      "grad_norm": 0.13672266900539398,
+      "learning_rate": 8.222500506791305e-06,
+      "loss": 0.3563,
+      "step": 546
+    },
+    {
+      "epoch": 2.452914798206278,
+      "grad_norm": 0.16056783497333527,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.362,
+      "step": 547
+    },
+    {
+      "epoch": 2.4573991031390134,
+      "grad_norm": 0.15055257081985474,
+      "learning_rate": 8.205359498339917e-06,
+      "loss": 0.3575,
+      "step": 548
+    },
+    {
+      "epoch": 2.461883408071749,
+      "grad_norm": 0.14585907757282257,
+      "learning_rate": 8.196764899464552e-06,
+      "loss": 0.3639,
+      "step": 549
+    },
+    {
+      "epoch": 2.4663677130044843,
+      "grad_norm": 0.15431244671344757,
+      "learning_rate": 8.188154294838542e-06,
+      "loss": 0.352,
+      "step": 550
+    },
+    {
+      "epoch": 2.4708520179372195,
+      "grad_norm": 0.14333587884902954,
+      "learning_rate": 8.179527727573975e-06,
+      "loss": 0.3622,
+      "step": 551
+    },
+    {
+      "epoch": 2.475336322869955,
+      "grad_norm": 0.1617281585931778,
+      "learning_rate": 8.170885240862854e-06,
+      "loss": 0.3692,
+      "step": 552
+    },
+    {
+      "epoch": 2.4798206278026904,
+      "grad_norm": 0.1339455395936966,
+      "learning_rate": 8.162226877976886e-06,
+      "loss": 0.369,
+      "step": 553
+    },
+    {
+      "epoch": 2.484304932735426,
+      "grad_norm": 0.16107149422168732,
+      "learning_rate": 8.153552682267278e-06,
+      "loss": 0.3586,
+      "step": 554
+    },
+    {
+      "epoch": 2.4887892376681613,
+      "grad_norm": 0.14115692675113678,
+      "learning_rate": 8.144862697164499e-06,
+      "loss": 0.3524,
+      "step": 555
+    },
+    {
+      "epoch": 2.493273542600897,
+      "grad_norm": 0.1845618635416031,
+      "learning_rate": 8.136156966178082e-06,
+      "loss": 0.3612,
+      "step": 556
+    },
+    {
+      "epoch": 2.497757847533632,
+      "grad_norm": 0.14242452383041382,
+      "learning_rate": 8.127435532896388e-06,
+      "loss": 0.3523,
+      "step": 557
+    },
+    {
+      "epoch": 2.502242152466368,
+      "grad_norm": 0.17012883722782135,
+      "learning_rate": 8.118698440986405e-06,
+      "loss": 0.3622,
+      "step": 558
+    },
+    {
+      "epoch": 2.506726457399103,
+      "grad_norm": 0.19423829019069672,
+      "learning_rate": 8.10994573419352e-06,
+      "loss": 0.3448,
+      "step": 559
+    },
+    {
+      "epoch": 2.5112107623318387,
+      "grad_norm": 0.16720794141292572,
+      "learning_rate": 8.101177456341301e-06,
+      "loss": 0.3518,
+      "step": 560
+    },
+    {
+      "epoch": 2.515695067264574,
+      "grad_norm": 0.2052316814661026,
+      "learning_rate": 8.092393651331275e-06,
+      "loss": 0.3552,
+      "step": 561
+    },
+    {
+      "epoch": 2.520179372197309,
+      "grad_norm": 0.17254412174224854,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.353,
+      "step": 562
+    },
+    {
+      "epoch": 2.524663677130045,
+      "grad_norm": 0.17825813591480255,
+      "learning_rate": 8.074779635832417e-06,
+      "loss": 0.3497,
+      "step": 563
+    },
+    {
+      "epoch": 2.5291479820627805,
+      "grad_norm": 0.18631678819656372,
+      "learning_rate": 8.065949513534474e-06,
+      "loss": 0.3623,
+      "step": 564
+    },
+    {
+      "epoch": 2.5336322869955157,
+      "grad_norm": 0.19206692278385162,
+      "learning_rate": 8.057104040460062e-06,
+      "loss": 0.3638,
+      "step": 565
+    },
+    {
+      "epoch": 2.538116591928251,
+      "grad_norm": 0.14845989644527435,
+      "learning_rate": 8.048243260897216e-06,
+      "loss": 0.3683,
+      "step": 566
+    },
+    {
+      "epoch": 2.5426008968609866,
+      "grad_norm": 0.17347513139247894,
+      "learning_rate": 8.03936721921061e-06,
+      "loss": 0.3626,
+      "step": 567
+    },
+    {
+      "epoch": 2.547085201793722,
+      "grad_norm": 0.15112243592739105,
+      "learning_rate": 8.030475959841333e-06,
+      "loss": 0.3557,
+      "step": 568
+    },
+    {
+      "epoch": 2.5515695067264574,
+      "grad_norm": 0.15691283345222473,
+      "learning_rate": 8.021569527306663e-06,
+      "loss": 0.3534,
+      "step": 569
+    },
+    {
+      "epoch": 2.5560538116591927,
+      "grad_norm": 0.14322233200073242,
+      "learning_rate": 8.012647966199851e-06,
+      "loss": 0.366,
+      "step": 570
+    },
+    {
+      "epoch": 2.5605381165919283,
+      "grad_norm": 0.15429705381393433,
+      "learning_rate": 8.003711321189895e-06,
+      "loss": 0.3682,
+      "step": 571
+    },
+    {
+      "epoch": 2.5650224215246635,
+      "grad_norm": 0.15206705033779144,
+      "learning_rate": 7.994759637021316e-06,
+      "loss": 0.3622,
+      "step": 572
+    },
+    {
+      "epoch": 2.569506726457399,
+      "grad_norm": 0.14171403646469116,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.3549,
+      "step": 573
+    },
+    {
+      "epoch": 2.5739910313901344,
+      "grad_norm": 0.147097647190094,
+      "learning_rate": 7.976811330562637e-06,
+      "loss": 0.3577,
+      "step": 574
+    },
+    {
+      "epoch": 2.57847533632287,
+      "grad_norm": 0.1507255584001541,
+      "learning_rate": 7.967814798137173e-06,
+      "loss": 0.3549,
+      "step": 575
+    },
+    {
+      "epoch": 2.5829596412556053,
+      "grad_norm": 0.13839322328567505,
+      "learning_rate": 7.95880340628191e-06,
+      "loss": 0.3553,
+      "step": 576
+    },
+    {
+      "epoch": 2.587443946188341,
+      "grad_norm": 0.1717439442873001,
+      "learning_rate": 7.949777200115617e-06,
+      "loss": 0.3527,
+      "step": 577
+    },
+    {
+      "epoch": 2.591928251121076,
+      "grad_norm": 0.14832676947116852,
+      "learning_rate": 7.940736224831227e-06,
+      "loss": 0.3568,
+      "step": 578
+    },
+    {
+      "epoch": 2.596412556053812,
+      "grad_norm": 0.17640292644500732,
+      "learning_rate": 7.931680525695634e-06,
+      "loss": 0.3534,
+      "step": 579
+    },
+    {
+      "epoch": 2.600896860986547,
+      "grad_norm": 0.19895848631858826,
+      "learning_rate": 7.922610148049445e-06,
+      "loss": 0.3582,
+      "step": 580
+    },
+    {
+      "epoch": 2.6053811659192823,
+      "grad_norm": 0.16366077959537506,
+      "learning_rate": 7.913525137306756e-06,
+      "loss": 0.3547,
+      "step": 581
+    },
+    {
+      "epoch": 2.609865470852018,
+      "grad_norm": 0.16760778427124023,
+      "learning_rate": 7.90442553895494e-06,
+      "loss": 0.3624,
+      "step": 582
+    },
+    {
+      "epoch": 2.6143497757847536,
+      "grad_norm": 0.16546988487243652,
+      "learning_rate": 7.895311398554395e-06,
+      "loss": 0.352,
+      "step": 583
+    },
+    {
+      "epoch": 2.618834080717489,
+      "grad_norm": 0.16307644546031952,
+      "learning_rate": 7.886182761738339e-06,
+      "loss": 0.3611,
+      "step": 584
+    },
+    {
+      "epoch": 2.623318385650224,
+      "grad_norm": 0.16867242753505707,
+      "learning_rate": 7.877039674212569e-06,
+      "loss": 0.3575,
+      "step": 585
+    },
+    {
+      "epoch": 2.6278026905829597,
+      "grad_norm": 0.14333225786685944,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.3627,
+      "step": 586
+    },
+    {
+      "epoch": 2.6322869955156953,
+      "grad_norm": 0.18009604513645172,
+      "learning_rate": 7.8587103302166e-06,
+      "loss": 0.3617,
+      "step": 587
+    },
+    {
+      "epoch": 2.6367713004484306,
+      "grad_norm": 0.1594720482826233,
+      "learning_rate": 7.849524165518838e-06,
+      "loss": 0.3556,
+      "step": 588
+    },
+    {
+      "epoch": 2.6412556053811658,
+      "grad_norm": 0.15891732275485992,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.3589,
+      "step": 589
+    },
+    {
+      "epoch": 2.6457399103139014,
+      "grad_norm": 0.1878863275051117,
+      "learning_rate": 7.831109080692687e-06,
+      "loss": 0.3577,
+      "step": 590
+    },
+    {
+      "epoch": 2.6502242152466366,
+      "grad_norm": 0.15169432759284973,
+      "learning_rate": 7.821880252766025e-06,
+      "loss": 0.3584,
+      "step": 591
+    },
+    {
+      "epoch": 2.6547085201793723,
+      "grad_norm": 0.15574221312999725,
+      "learning_rate": 7.812637296083235e-06,
+      "loss": 0.3665,
+      "step": 592
+    },
+    {
+      "epoch": 2.6591928251121075,
+      "grad_norm": 0.18458415567874908,
+      "learning_rate": 7.803380256922495e-06,
+      "loss": 0.3564,
+      "step": 593
+    },
+    {
+      "epoch": 2.663677130044843,
+      "grad_norm": 0.16051502525806427,
+      "learning_rate": 7.794109181632494e-06,
+      "loss": 0.3593,
+      "step": 594
+    },
+    {
+      "epoch": 2.6681614349775784,
+      "grad_norm": 0.16088628768920898,
+      "learning_rate": 7.784824116632198e-06,
+      "loss": 0.3612,
+      "step": 595
+    },
+    {
+      "epoch": 2.672645739910314,
+      "grad_norm": 0.15537729859352112,
+      "learning_rate": 7.775525108410616e-06,
+      "loss": 0.361,
+      "step": 596
+    },
+    {
+      "epoch": 2.6771300448430493,
+      "grad_norm": 0.1465197205543518,
+      "learning_rate": 7.76621220352657e-06,
+      "loss": 0.3538,
+      "step": 597
+    },
+    {
+      "epoch": 2.681614349775785,
+      "grad_norm": 0.16095510125160217,
+      "learning_rate": 7.75688544860846e-06,
+      "loss": 0.3592,
+      "step": 598
+    },
+    {
+      "epoch": 2.68609865470852,
+      "grad_norm": 0.14529503881931305,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.3545,
+      "step": 599
+    },
+    {
+      "epoch": 2.6905829596412554,
+      "grad_norm": 0.1660734862089157,
+      "learning_rate": 7.73819057553014e-06,
+      "loss": 0.3525,
+      "step": 600
+    },
+    {
+      "epoch": 2.695067264573991,
+      "grad_norm": 0.16490519046783447,
+      "learning_rate": 7.728822550972523e-06,
+      "loss": 0.3554,
+      "step": 601
+    },
+    {
+      "epoch": 2.6995515695067267,
+      "grad_norm": 0.14031031727790833,
+      "learning_rate": 7.719440863585555e-06,
+      "loss": 0.3537,
+      "step": 602
+    },
+    {
+      "epoch": 2.704035874439462,
+      "grad_norm": 0.1618361622095108,
+      "learning_rate": 7.710045560342021e-06,
+      "loss": 0.3563,
+      "step": 603
+    },
+    {
+      "epoch": 2.708520179372197,
+      "grad_norm": 0.14599907398223877,
+      "learning_rate": 7.700636688282878e-06,
+      "loss": 0.3528,
+      "step": 604
+    },
+    {
+      "epoch": 2.713004484304933,
+      "grad_norm": 0.14423228800296783,
+      "learning_rate": 7.69121429451702e-06,
+      "loss": 0.3657,
+      "step": 605
+    },
+    {
+      "epoch": 2.717488789237668,
+      "grad_norm": 0.16366732120513916,
+      "learning_rate": 7.681778426221043e-06,
+      "loss": 0.3571,
+      "step": 606
+    },
+    {
+      "epoch": 2.7219730941704037,
+      "grad_norm": 0.1591261476278305,
+      "learning_rate": 7.672329130639007e-06,
+      "loss": 0.356,
+      "step": 607
+    },
+    {
+      "epoch": 2.726457399103139,
+      "grad_norm": 0.15975883603096008,
+      "learning_rate": 7.662866455082201e-06,
+      "loss": 0.35,
+      "step": 608
+    },
+    {
+      "epoch": 2.7309417040358746,
+      "grad_norm": 0.15265345573425293,
+      "learning_rate": 7.65339044692891e-06,
+      "loss": 0.3572,
+      "step": 609
+    },
+    {
+      "epoch": 2.7354260089686098,
+      "grad_norm": 0.15803271532058716,
+      "learning_rate": 7.643901153624163e-06,
+      "loss": 0.35,
+      "step": 610
+    },
+    {
+      "epoch": 2.7399103139013454,
+      "grad_norm": 0.15297670662403107,
+      "learning_rate": 7.634398622679517e-06,
+      "loss": 0.3611,
+      "step": 611
+    },
+    {
+      "epoch": 2.7443946188340806,
+      "grad_norm": 0.15764224529266357,
+      "learning_rate": 7.624882901672801e-06,
+      "loss": 0.3546,
+      "step": 612
+    },
+    {
+      "epoch": 2.7488789237668163,
+      "grad_norm": 0.16480886936187744,
+      "learning_rate": 7.615354038247889e-06,
+      "loss": 0.3535,
+      "step": 613
+    },
+    {
+      "epoch": 2.7533632286995515,
+      "grad_norm": 0.1803409308195114,
+      "learning_rate": 7.605812080114453e-06,
+      "loss": 0.3559,
+      "step": 614
+    },
+    {
+      "epoch": 2.7578475336322867,
+      "grad_norm": 0.18355964124202728,
+      "learning_rate": 7.59625707504773e-06,
+      "loss": 0.3537,
+      "step": 615
+    },
+    {
+      "epoch": 2.7623318385650224,
+      "grad_norm": 0.16814208030700684,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.3615,
+      "step": 616
+    },
+    {
+      "epoch": 2.766816143497758,
+      "grad_norm": 0.16023153066635132,
+      "learning_rate": 7.577108115541761e-06,
+      "loss": 0.3565,
+      "step": 617
+    },
+    {
+      "epoch": 2.7713004484304933,
+      "grad_norm": 0.16750356554985046,
+      "learning_rate": 7.567514256978652e-06,
+      "loss": 0.3586,
+      "step": 618
+    },
+    {
+      "epoch": 2.7757847533632285,
+      "grad_norm": 0.1632583886384964,
+      "learning_rate": 7.557907543234051e-06,
+      "loss": 0.3598,
+      "step": 619
+    },
+    {
+      "epoch": 2.780269058295964,
+      "grad_norm": 0.146370530128479,
+      "learning_rate": 7.54828802240742e-06,
+      "loss": 0.3563,
+      "step": 620
+    },
+    {
+      "epoch": 2.7847533632287,
+      "grad_norm": 0.18041367828845978,
+      "learning_rate": 7.53865574266234e-06,
+      "loss": 0.3558,
+      "step": 621
+    },
+    {
+      "epoch": 2.789237668161435,
+      "grad_norm": 0.15324024856090546,
+      "learning_rate": 7.529010752226277e-06,
+      "loss": 0.3501,
+      "step": 622
+    },
+    {
+      "epoch": 2.7937219730941703,
+      "grad_norm": 0.16382090747356415,
+      "learning_rate": 7.519353099390336e-06,
+      "loss": 0.3578,
+      "step": 623
+    },
+    {
+      "epoch": 2.798206278026906,
+      "grad_norm": 0.17279858887195587,
+      "learning_rate": 7.509682832509024e-06,
+      "loss": 0.3509,
+      "step": 624
+    },
+    {
+      "epoch": 2.802690582959641,
+      "grad_norm": 0.16233599185943604,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3617,
+      "step": 625
+    },
+    {
+      "epoch": 2.807174887892377,
+      "grad_norm": 0.17482459545135498,
+      "learning_rate": 7.490304650343842e-06,
+      "loss": 0.3594,
+      "step": 626
+    },
+    {
+      "epoch": 2.811659192825112,
+      "grad_norm": 0.15437589585781097,
+      "learning_rate": 7.480596832083795e-06,
+      "loss": 0.3523,
+      "step": 627
+    },
+    {
+      "epoch": 2.8161434977578477,
+      "grad_norm": 0.17096751928329468,
+      "learning_rate": 7.470876593825535e-06,
+      "loss": 0.3604,
+      "step": 628
+    },
+    {
+      "epoch": 2.820627802690583,
+      "grad_norm": 0.17550338804721832,
+      "learning_rate": 7.461143984236925e-06,
+      "loss": 0.3606,
+      "step": 629
+    },
+    {
+      "epoch": 2.8251121076233185,
+      "grad_norm": 0.16155405342578888,
+      "learning_rate": 7.451399052047764e-06,
+      "loss": 0.351,
+      "step": 630
+    },
+    {
+      "epoch": 2.8295964125560538,
+      "grad_norm": 0.18060077726840973,
+      "learning_rate": 7.441641846049557e-06,
+      "loss": 0.3528,
+      "step": 631
+    },
+    {
+      "epoch": 2.8340807174887894,
+      "grad_norm": 0.18699073791503906,
+      "learning_rate": 7.431872415095252e-06,
+      "loss": 0.361,
+      "step": 632
+    },
+    {
+      "epoch": 2.8385650224215246,
+      "grad_norm": 0.16657650470733643,
+      "learning_rate": 7.422090808099014e-06,
+      "loss": 0.3631,
+      "step": 633
+    },
+    {
+      "epoch": 2.84304932735426,
+      "grad_norm": 0.1745719313621521,
+      "learning_rate": 7.412297074035968e-06,
+      "loss": 0.357,
+      "step": 634
+    },
+    {
+      "epoch": 2.8475336322869955,
+      "grad_norm": 0.1742197722196579,
+      "learning_rate": 7.402491261941958e-06,
+      "loss": 0.3591,
+      "step": 635
+    },
+    {
+      "epoch": 2.852017937219731,
+      "grad_norm": 0.1817382425069809,
+      "learning_rate": 7.3926734209133e-06,
+      "loss": 0.3648,
+      "step": 636
+    },
+    {
+      "epoch": 2.8565022421524664,
+      "grad_norm": 0.1701369285583496,
+      "learning_rate": 7.382843600106539e-06,
+      "loss": 0.3552,
+      "step": 637
+    },
+    {
+      "epoch": 2.8609865470852016,
+      "grad_norm": 0.17051516473293304,
+      "learning_rate": 7.373001848738203e-06,
+      "loss": 0.357,
+      "step": 638
+    },
+    {
+      "epoch": 2.8654708520179373,
+      "grad_norm": 0.16486294567584991,
+      "learning_rate": 7.363148216084548e-06,
+      "loss": 0.3638,
+      "step": 639
+    },
+    {
+      "epoch": 2.8699551569506725,
+      "grad_norm": 0.15621639788150787,
+      "learning_rate": 7.353282751481324e-06,
+      "loss": 0.3523,
+      "step": 640
+    },
+    {
+      "epoch": 2.874439461883408,
+      "grad_norm": 0.15908995270729065,
+      "learning_rate": 7.343405504323519e-06,
+      "loss": 0.3541,
+      "step": 641
+    },
+    {
+      "epoch": 2.8789237668161434,
+      "grad_norm": 0.16896823048591614,
+      "learning_rate": 7.333516524065116e-06,
+      "loss": 0.3557,
+      "step": 642
+    },
+    {
+      "epoch": 2.883408071748879,
+      "grad_norm": 0.16952857375144958,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.3555,
+      "step": 643
+    },
+    {
+      "epoch": 2.8878923766816142,
+      "grad_norm": 0.15454255044460297,
+      "learning_rate": 7.3137035623559274e-06,
+      "loss": 0.3647,
+      "step": 644
+    },
+    {
+      "epoch": 2.89237668161435,
+      "grad_norm": 0.18705131113529205,
+      "learning_rate": 7.303779680105844e-06,
+      "loss": 0.3579,
+      "step": 645
+    },
+    {
+      "epoch": 2.896860986547085,
+      "grad_norm": 0.15942417085170746,
+      "learning_rate": 7.2938442631560714e-06,
+      "loss": 0.3604,
+      "step": 646
+    },
+    {
+      "epoch": 2.901345291479821,
+      "grad_norm": 0.17414377629756927,
+      "learning_rate": 7.28389736125184e-06,
+      "loss": 0.3552,
+      "step": 647
+    },
+    {
+      "epoch": 2.905829596412556,
+      "grad_norm": 0.1664162427186966,
+      "learning_rate": 7.273939024195881e-06,
+      "loss": 0.3523,
+      "step": 648
+    },
+    {
+      "epoch": 2.910313901345291,
+      "grad_norm": 0.18107785284519196,
+      "learning_rate": 7.263969301848188e-06,
+      "loss": 0.3567,
+      "step": 649
+    },
+    {
+      "epoch": 2.914798206278027,
+      "grad_norm": 0.15371252596378326,
+      "learning_rate": 7.2539882441257484e-06,
+      "loss": 0.3501,
+      "step": 650
+    },
+    {
+      "epoch": 2.9192825112107625,
+      "grad_norm": 0.1648615598678589,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.3712,
+      "step": 651
+    },
+    {
+      "epoch": 2.9237668161434978,
+      "grad_norm": 0.19698192179203033,
+      "learning_rate": 7.2339923225081296e-06,
+      "loss": 0.3601,
+      "step": 652
+    },
+    {
+      "epoch": 2.928251121076233,
+      "grad_norm": 0.16851653158664703,
+      "learning_rate": 7.223977558729707e-06,
+      "loss": 0.354,
+      "step": 653
+    },
+    {
+      "epoch": 2.9327354260089686,
+      "grad_norm": 0.18450665473937988,
+      "learning_rate": 7.213951659809553e-06,
+      "loss": 0.3547,
+      "step": 654
+    },
+    {
+      "epoch": 2.9372197309417043,
+      "grad_norm": 0.20189598202705383,
+      "learning_rate": 7.203914675945929e-06,
+      "loss": 0.3513,
+      "step": 655
+    },
+    {
+      "epoch": 2.9417040358744395,
+      "grad_norm": 0.1727120578289032,
+      "learning_rate": 7.193866657392597e-06,
+      "loss": 0.3526,
+      "step": 656
+    },
+    {
+      "epoch": 2.9461883408071747,
+      "grad_norm": 0.18441489338874817,
+      "learning_rate": 7.183807654458565e-06,
+      "loss": 0.3529,
+      "step": 657
+    },
+    {
+      "epoch": 2.9506726457399104,
+      "grad_norm": 0.17119796574115753,
+      "learning_rate": 7.173737717507843e-06,
+      "loss": 0.3604,
+      "step": 658
+    },
+    {
+      "epoch": 2.9551569506726456,
+      "grad_norm": 0.18862298130989075,
+      "learning_rate": 7.163656896959181e-06,
+      "loss": 0.3632,
+      "step": 659
+    },
+    {
+      "epoch": 2.9596412556053813,
+      "grad_norm": 0.17836743593215942,
+      "learning_rate": 7.153565243285829e-06,
+      "loss": 0.3584,
+      "step": 660
+    },
+    {
+      "epoch": 2.9641255605381165,
+      "grad_norm": 0.16732634603977203,
+      "learning_rate": 7.143462807015271e-06,
+      "loss": 0.3558,
+      "step": 661
+    },
+    {
+      "epoch": 2.968609865470852,
+      "grad_norm": 0.16331027448177338,
+      "learning_rate": 7.133349638728979e-06,
+      "loss": 0.353,
+      "step": 662
+    },
+    {
+      "epoch": 2.9730941704035874,
+      "grad_norm": 0.15720780193805695,
+      "learning_rate": 7.1232257890621605e-06,
+      "loss": 0.3564,
+      "step": 663
+    },
+    {
+      "epoch": 2.977578475336323,
+      "grad_norm": 0.1624830663204193,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.3577,
+      "step": 664
+    },
+    {
+      "epoch": 2.9820627802690582,
+      "grad_norm": 0.14818091690540314,
+      "learning_rate": 7.102946248394908e-06,
+      "loss": 0.3556,
+      "step": 665
+    },
+    {
+      "epoch": 2.986547085201794,
+      "grad_norm": 0.16482090950012207,
+      "learning_rate": 7.092790658931274e-06,
+      "loss": 0.3535,
+      "step": 666
+    },
+    {
+      "epoch": 2.991031390134529,
+      "grad_norm": 0.17528152465820312,
+      "learning_rate": 7.082624591160201e-06,
+      "loss": 0.359,
+      "step": 667
+    },
+    {
+      "epoch": 2.9955156950672643,
+      "grad_norm": 0.13651765882968903,
+      "learning_rate": 7.072448095981751e-06,
+      "loss": 0.353,
+      "step": 668
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.1478780210018158,
+      "learning_rate": 7.0622612243482035e-06,
+      "loss": 0.3495,
+      "step": 669
+    },
+    {
+      "epoch": 3.004484304932735,
+      "grad_norm": 0.16963088512420654,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.3439,
+      "step": 670
+    },
+    {
+      "epoch": 3.008968609865471,
+      "grad_norm": 0.17842049896717072,
+      "learning_rate": 7.041856555784421e-06,
+      "loss": 0.3343,
+      "step": 671
+    },
+    {
+      "epoch": 3.013452914798206,
+      "grad_norm": 0.18344755470752716,
+      "learning_rate": 7.031638861017484e-06,
+      "loss": 0.338,
+      "step": 672
+    },
+    {
+      "epoch": 3.0179372197309418,
+      "grad_norm": 0.200805202126503,
+      "learning_rate": 7.021410994121525e-06,
+      "loss": 0.3408,
+      "step": 673
+    },
+    {
+      "epoch": 3.022421524663677,
+      "grad_norm": 0.22763468325138092,
+      "learning_rate": 7.011173006306034e-06,
+      "loss": 0.3397,
+      "step": 674
+    },
+    {
+      "epoch": 3.0269058295964126,
+      "grad_norm": 0.18602415919303894,
+      "learning_rate": 7.0009249488311685e-06,
+      "loss": 0.3404,
+      "step": 675
+    },
+    {
+      "epoch": 3.031390134529148,
+      "grad_norm": 0.1775357872247696,
+      "learning_rate": 6.990666873007506e-06,
+      "loss": 0.337,
+      "step": 676
+    },
+    {
+      "epoch": 3.0358744394618835,
+      "grad_norm": 0.19117063283920288,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.3291,
+      "step": 677
+    },
+    {
+      "epoch": 3.0403587443946187,
+      "grad_norm": 0.19715914130210876,
+      "learning_rate": 6.970120871806647e-06,
+      "loss": 0.3395,
+      "step": 678
+    },
+    {
+      "epoch": 3.0448430493273544,
+      "grad_norm": 0.18218578398227692,
+      "learning_rate": 6.959833049300376e-06,
+      "loss": 0.3385,
+      "step": 679
+    },
+    {
+      "epoch": 3.0493273542600896,
+      "grad_norm": 0.18100160360336304,
+      "learning_rate": 6.949535414186651e-06,
+      "loss": 0.3305,
+      "step": 680
+    },
+    {
+      "epoch": 3.0538116591928253,
+      "grad_norm": 0.190889373421669,
+      "learning_rate": 6.939228018024275e-06,
+      "loss": 0.3436,
+      "step": 681
+    },
+    {
+      "epoch": 3.0582959641255605,
+      "grad_norm": 0.19537365436553955,
+      "learning_rate": 6.9289109124209276e-06,
+      "loss": 0.3393,
+      "step": 682
+    },
+    {
+      "epoch": 3.062780269058296,
+      "grad_norm": 0.16741010546684265,
+      "learning_rate": 6.918584149032899e-06,
+      "loss": 0.3308,
+      "step": 683
+    },
+    {
+      "epoch": 3.0672645739910314,
+      "grad_norm": 0.19638636708259583,
+      "learning_rate": 6.908247779564836e-06,
+      "loss": 0.3396,
+      "step": 684
+    },
+    {
+      "epoch": 3.071748878923767,
+      "grad_norm": 0.19751055538654327,
+      "learning_rate": 6.897901855769483e-06,
+      "loss": 0.3306,
+      "step": 685
+    },
+    {
+      "epoch": 3.0762331838565022,
+      "grad_norm": 0.19639840722084045,
+      "learning_rate": 6.887546429447418e-06,
+      "loss": 0.3347,
+      "step": 686
+    },
+    {
+      "epoch": 3.0807174887892375,
+      "grad_norm": 0.18278853595256805,
+      "learning_rate": 6.8771815524468e-06,
+      "loss": 0.3191,
+      "step": 687
+    },
+    {
+      "epoch": 3.085201793721973,
+      "grad_norm": 0.2162567377090454,
+      "learning_rate": 6.8668072766631054e-06,
+      "loss": 0.3352,
+      "step": 688
+    },
+    {
+      "epoch": 3.0896860986547083,
+      "grad_norm": 0.21039368212223053,
+      "learning_rate": 6.856423654038868e-06,
+      "loss": 0.3393,
+      "step": 689
+    },
+    {
+      "epoch": 3.094170403587444,
+      "grad_norm": 0.1685999482870102,
+      "learning_rate": 6.8460307365634225e-06,
+      "loss": 0.3371,
+      "step": 690
+    },
+    {
+      "epoch": 3.098654708520179,
+      "grad_norm": 0.19002294540405273,
+      "learning_rate": 6.8356285762726385e-06,
+      "loss": 0.3403,
+      "step": 691
+    },
+    {
+      "epoch": 3.103139013452915,
+      "grad_norm": 0.16503164172172546,
+      "learning_rate": 6.825217225248664e-06,
+      "loss": 0.3323,
+      "step": 692
+    },
+    {
+      "epoch": 3.10762331838565,
+      "grad_norm": 0.16635221242904663,
+      "learning_rate": 6.814796735619664e-06,
+      "loss": 0.3378,
+      "step": 693
+    },
+    {
+      "epoch": 3.1121076233183858,
+      "grad_norm": 0.1685560941696167,
+      "learning_rate": 6.804367159559561e-06,
+      "loss": 0.3356,
+      "step": 694
+    },
+    {
+      "epoch": 3.116591928251121,
+      "grad_norm": 0.18063698709011078,
+      "learning_rate": 6.793928549287767e-06,
+      "loss": 0.3351,
+      "step": 695
+    },
+    {
+      "epoch": 3.1210762331838566,
+      "grad_norm": 0.16821306943893433,
+      "learning_rate": 6.783480957068934e-06,
+      "loss": 0.3368,
+      "step": 696
+    },
+    {
+      "epoch": 3.125560538116592,
+      "grad_norm": 0.17041970789432526,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.3278,
+      "step": 697
+    },
+    {
+      "epoch": 3.1300448430493275,
+      "grad_norm": 0.21547262370586395,
+      "learning_rate": 6.762559036073331e-06,
+      "loss": 0.3447,
+      "step": 698
+    },
+    {
+      "epoch": 3.1345291479820627,
+      "grad_norm": 0.16423679888248444,
+      "learning_rate": 6.75208481204967e-06,
+      "loss": 0.3397,
+      "step": 699
+    },
+    {
+      "epoch": 3.1390134529147984,
+      "grad_norm": 0.17103219032287598,
+      "learning_rate": 6.74160181558465e-06,
+      "loss": 0.3279,
+      "step": 700
+    },
+    {
+      "epoch": 3.1434977578475336,
+      "grad_norm": 0.16891707479953766,
+      "learning_rate": 6.731110099165165e-06,
+      "loss": 0.3376,
+      "step": 701
+    },
+    {
+      "epoch": 3.1479820627802693,
+      "grad_norm": 0.21367736160755157,
+      "learning_rate": 6.7206097153217474e-06,
+      "loss": 0.3412,
+      "step": 702
+    },
+    {
+      "epoch": 3.1524663677130045,
+      "grad_norm": 0.15597198903560638,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.3357,
+      "step": 703
+    },
+    {
+      "epoch": 3.15695067264574,
+      "grad_norm": 0.20945321023464203,
+      "learning_rate": 6.699583155702025e-06,
+      "loss": 0.3386,
+      "step": 704
+    },
+    {
+      "epoch": 3.1614349775784754,
+      "grad_norm": 0.18091322481632233,
+      "learning_rate": 6.689057085202737e-06,
+      "loss": 0.334,
+      "step": 705
+    },
+    {
+      "epoch": 3.1659192825112106,
+      "grad_norm": 0.18464356660842896,
+      "learning_rate": 6.678522557833025e-06,
+      "loss": 0.3377,
+      "step": 706
+    },
+    {
+      "epoch": 3.1704035874439462,
+      "grad_norm": 0.18079009652137756,
+      "learning_rate": 6.667979626337783e-06,
+      "loss": 0.3237,
+      "step": 707
+    },
+    {
+      "epoch": 3.1748878923766815,
+      "grad_norm": 0.19078385829925537,
+      "learning_rate": 6.6574283435039795e-06,
+      "loss": 0.334,
+      "step": 708
+    },
+    {
+      "epoch": 3.179372197309417,
+      "grad_norm": 0.17668864130973816,
+      "learning_rate": 6.646868762160399e-06,
+      "loss": 0.3317,
+      "step": 709
+    },
+    {
+      "epoch": 3.1838565022421523,
+      "grad_norm": 0.1615522801876068,
+      "learning_rate": 6.6363009351773755e-06,
+      "loss": 0.3352,
+      "step": 710
+    },
+    {
+      "epoch": 3.188340807174888,
+      "grad_norm": 0.1773214340209961,
+      "learning_rate": 6.625724915466526e-06,
+      "loss": 0.3426,
+      "step": 711
+    },
+    {
+      "epoch": 3.192825112107623,
+      "grad_norm": 0.16128070652484894,
+      "learning_rate": 6.615140755980491e-06,
+      "loss": 0.3306,
+      "step": 712
+    },
+    {
+      "epoch": 3.197309417040359,
+      "grad_norm": 0.16551382839679718,
+      "learning_rate": 6.6045485097126585e-06,
+      "loss": 0.3397,
+      "step": 713
+    },
+    {
+      "epoch": 3.201793721973094,
+      "grad_norm": 0.1689879298210144,
+      "learning_rate": 6.593948229696915e-06,
+      "loss": 0.343,
+      "step": 714
+    },
+    {
+      "epoch": 3.2062780269058297,
+      "grad_norm": 0.16034471988677979,
+      "learning_rate": 6.583339969007364e-06,
+      "loss": 0.3399,
+      "step": 715
+    },
+    {
+      "epoch": 3.210762331838565,
+      "grad_norm": 0.16412587463855743,
+      "learning_rate": 6.572723780758069e-06,
+      "loss": 0.3313,
+      "step": 716
+    },
+    {
+      "epoch": 3.2152466367713006,
+      "grad_norm": 0.16765817999839783,
+      "learning_rate": 6.562099718102788e-06,
+      "loss": 0.329,
+      "step": 717
+    },
+    {
+      "epoch": 3.219730941704036,
+      "grad_norm": 0.17685356736183167,
+      "learning_rate": 6.551467834234702e-06,
+      "loss": 0.3322,
+      "step": 718
+    },
+    {
+      "epoch": 3.2242152466367715,
+      "grad_norm": 0.17262746393680573,
+      "learning_rate": 6.540828182386154e-06,
+      "loss": 0.3365,
+      "step": 719
+    },
+    {
+      "epoch": 3.2286995515695067,
+      "grad_norm": 0.1485939919948578,
+      "learning_rate": 6.530180815828377e-06,
+      "loss": 0.3434,
+      "step": 720
+    },
+    {
+      "epoch": 3.233183856502242,
+      "grad_norm": 0.18364308774471283,
+      "learning_rate": 6.519525787871235e-06,
+      "loss": 0.3335,
+      "step": 721
+    },
+    {
+      "epoch": 3.2376681614349776,
+      "grad_norm": 0.17140884697437286,
+      "learning_rate": 6.508863151862947e-06,
+      "loss": 0.3468,
+      "step": 722
+    },
+    {
+      "epoch": 3.242152466367713,
+      "grad_norm": 0.15669117867946625,
+      "learning_rate": 6.498192961189829e-06,
+      "loss": 0.338,
+      "step": 723
+    },
+    {
+      "epoch": 3.2466367713004485,
+      "grad_norm": 0.1569780856370926,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.3365,
+      "step": 724
+    },
+    {
+      "epoch": 3.2511210762331837,
+      "grad_norm": 0.1523321568965912,
+      "learning_rate": 6.476830129583207e-06,
+      "loss": 0.3383,
+      "step": 725
+    },
+    {
+      "epoch": 3.2556053811659194,
+      "grad_norm": 0.16352108120918274,
+      "learning_rate": 6.466137595610388e-06,
+      "loss": 0.3329,
+      "step": 726
+    },
+    {
+      "epoch": 3.2600896860986546,
+      "grad_norm": 0.165395587682724,
+      "learning_rate": 6.455437720893565e-06,
+      "loss": 0.3353,
+      "step": 727
+    },
+    {
+      "epoch": 3.2645739910313902,
+      "grad_norm": 0.17201630771160126,
+      "learning_rate": 6.444730559005504e-06,
+      "loss": 0.3385,
+      "step": 728
+    },
+    {
+      "epoch": 3.2690582959641254,
+      "grad_norm": 0.1399305909872055,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.3323,
+      "step": 729
+    },
+    {
+      "epoch": 3.273542600896861,
+      "grad_norm": 0.15800097584724426,
+      "learning_rate": 6.423294588188874e-06,
+      "loss": 0.3425,
+      "step": 730
+    },
+    {
+      "epoch": 3.2780269058295963,
+      "grad_norm": 0.15714004635810852,
+      "learning_rate": 6.412565886587186e-06,
+      "loss": 0.3361,
+      "step": 731
+    },
+    {
+      "epoch": 3.282511210762332,
+      "grad_norm": 0.17159472405910492,
+      "learning_rate": 6.401830112467479e-06,
+      "loss": 0.3312,
+      "step": 732
+    },
+    {
+      "epoch": 3.286995515695067,
+      "grad_norm": 0.148673415184021,
+      "learning_rate": 6.391087319582264e-06,
+      "loss": 0.3414,
+      "step": 733
+    },
+    {
+      "epoch": 3.291479820627803,
+      "grad_norm": 0.17917297780513763,
+      "learning_rate": 6.380337561719184e-06,
+      "loss": 0.3281,
+      "step": 734
+    },
+    {
+      "epoch": 3.295964125560538,
+      "grad_norm": 0.15738297998905182,
+      "learning_rate": 6.369580892700763e-06,
+      "loss": 0.3368,
+      "step": 735
+    },
+    {
+      "epoch": 3.3004484304932733,
+      "grad_norm": 0.16321231424808502,
+      "learning_rate": 6.358817366384122e-06,
+      "loss": 0.3368,
+      "step": 736
+    },
+    {
+      "epoch": 3.304932735426009,
+      "grad_norm": 0.1804312765598297,
+      "learning_rate": 6.34804703666072e-06,
+      "loss": 0.3388,
+      "step": 737
+    },
+    {
+      "epoch": 3.3094170403587446,
+      "grad_norm": 0.17409038543701172,
+      "learning_rate": 6.337269957456074e-06,
+      "loss": 0.3341,
+      "step": 738
+    },
+    {
+      "epoch": 3.31390134529148,
+      "grad_norm": 0.16761574149131775,
+      "learning_rate": 6.326486182729504e-06,
+      "loss": 0.3383,
+      "step": 739
+    },
+    {
+      "epoch": 3.318385650224215,
+      "grad_norm": 0.20168949663639069,
+      "learning_rate": 6.315695766473844e-06,
+      "loss": 0.3392,
+      "step": 740
+    },
+    {
+      "epoch": 3.3228699551569507,
+      "grad_norm": 0.17255443334579468,
+      "learning_rate": 6.304898762715187e-06,
+      "loss": 0.3421,
+      "step": 741
+    },
+    {
+      "epoch": 3.327354260089686,
+      "grad_norm": 0.20463934540748596,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3345,
+      "step": 742
+    },
+    {
+      "epoch": 3.3318385650224216,
+      "grad_norm": 0.19216547906398773,
+      "learning_rate": 6.283285208957884e-06,
+      "loss": 0.3261,
+      "step": 743
+    },
+    {
+      "epoch": 3.336322869955157,
+      "grad_norm": 0.19602860510349274,
+      "learning_rate": 6.272468767175255e-06,
+      "loss": 0.3396,
+      "step": 744
+    },
+    {
+      "epoch": 3.3408071748878925,
+      "grad_norm": 0.2123761922121048,
+      "learning_rate": 6.261645954321109e-06,
+      "loss": 0.3311,
+      "step": 745
+    },
+    {
+      "epoch": 3.3452914798206277,
+      "grad_norm": 0.16347439587116241,
+      "learning_rate": 6.2508168245837476e-06,
+      "loss": 0.3431,
+      "step": 746
+    },
+    {
+      "epoch": 3.3497757847533634,
+      "grad_norm": 0.20546086132526398,
+      "learning_rate": 6.239981432183093e-06,
+      "loss": 0.3408,
+      "step": 747
+    },
+    {
+      "epoch": 3.3542600896860986,
+      "grad_norm": 0.180713951587677,
+      "learning_rate": 6.2291398313704265e-06,
+      "loss": 0.3389,
+      "step": 748
+    },
+    {
+      "epoch": 3.3587443946188342,
+      "grad_norm": 0.18377161026000977,
+      "learning_rate": 6.21829207642811e-06,
+      "loss": 0.3432,
+      "step": 749
+    },
+    {
+      "epoch": 3.3632286995515694,
+      "grad_norm": 0.18848901987075806,
+      "learning_rate": 6.207438221669325e-06,
+      "loss": 0.3366,
+      "step": 750
+    },
+    {
+      "epoch": 3.367713004484305,
+      "grad_norm": 0.20928214490413666,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.3318,
+      "step": 751
+    },
+    {
+      "epoch": 3.3721973094170403,
+      "grad_norm": 0.14759661257266998,
+      "learning_rate": 6.185712430107489e-06,
+      "loss": 0.3386,
+      "step": 752
+    },
+    {
+      "epoch": 3.376681614349776,
+      "grad_norm": 0.1903652846813202,
+      "learning_rate": 6.1748406020824115e-06,
+      "loss": 0.3375,
+      "step": 753
+    },
+    {
+      "epoch": 3.381165919282511,
+      "grad_norm": 0.16853877902030945,
+      "learning_rate": 6.163962891796261e-06,
+      "loss": 0.3379,
+      "step": 754
+    },
+    {
+      "epoch": 3.3856502242152464,
+      "grad_norm": 0.161412313580513,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.3419,
+      "step": 755
+    },
+    {
+      "epoch": 3.390134529147982,
+      "grad_norm": 0.17498178780078888,
+      "learning_rate": 6.142190042322569e-06,
+      "loss": 0.3355,
+      "step": 756
+    },
+    {
+      "epoch": 3.3946188340807173,
+      "grad_norm": 0.17828604578971863,
+      "learning_rate": 6.131295012148613e-06,
+      "loss": 0.3383,
+      "step": 757
+    },
+    {
+      "epoch": 3.399103139013453,
+      "grad_norm": 0.19606079161167145,
+      "learning_rate": 6.120394317740205e-06,
+      "loss": 0.3249,
+      "step": 758
+    },
+    {
+      "epoch": 3.403587443946188,
+      "grad_norm": 0.17446884512901306,
+      "learning_rate": 6.1094880136755886e-06,
+      "loss": 0.343,
+      "step": 759
+    },
+    {
+      "epoch": 3.408071748878924,
+      "grad_norm": 0.1586799919605255,
+      "learning_rate": 6.0985761545610865e-06,
+      "loss": 0.3344,
+      "step": 760
+    },
+    {
+      "epoch": 3.412556053811659,
+      "grad_norm": 0.17535468935966492,
+      "learning_rate": 6.087658795030838e-06,
+      "loss": 0.3367,
+      "step": 761
+    },
+    {
+      "epoch": 3.4170403587443947,
+      "grad_norm": 0.16907069087028503,
+      "learning_rate": 6.076735989746522e-06,
+      "loss": 0.3347,
+      "step": 762
+    },
+    {
+      "epoch": 3.42152466367713,
+      "grad_norm": 0.16624757647514343,
+      "learning_rate": 6.065807793397081e-06,
+      "loss": 0.3362,
+      "step": 763
+    },
+    {
+      "epoch": 3.4260089686098656,
+      "grad_norm": 0.14994627237319946,
+      "learning_rate": 6.0548742606984545e-06,
+      "loss": 0.3389,
+      "step": 764
+    },
+    {
+      "epoch": 3.430493273542601,
+      "grad_norm": 0.1712806075811386,
+      "learning_rate": 6.043935446393294e-06,
+      "loss": 0.3302,
+      "step": 765
+    },
+    {
+      "epoch": 3.4349775784753365,
+      "grad_norm": 0.17839433252811432,
+      "learning_rate": 6.032991405250702e-06,
+      "loss": 0.3351,
+      "step": 766
+    },
+    {
+      "epoch": 3.4394618834080717,
+      "grad_norm": 0.14378862082958221,
+      "learning_rate": 6.022042192065946e-06,
+      "loss": 0.3366,
+      "step": 767
+    },
+    {
+      "epoch": 3.4439461883408073,
+      "grad_norm": 0.15560346841812134,
+      "learning_rate": 6.011087861660191e-06,
+      "loss": 0.3401,
+      "step": 768
+    },
+    {
+      "epoch": 3.4484304932735426,
+      "grad_norm": 0.17642395198345184,
+      "learning_rate": 6.000128468880223e-06,
+      "loss": 0.3387,
+      "step": 769
+    },
+    {
+      "epoch": 3.452914798206278,
+      "grad_norm": 0.1433219462633133,
+      "learning_rate": 5.989164068598175e-06,
+      "loss": 0.3335,
+      "step": 770
+    },
+    {
+      "epoch": 3.4573991031390134,
+      "grad_norm": 0.14371176064014435,
+      "learning_rate": 5.978194715711254e-06,
+      "loss": 0.3377,
+      "step": 771
+    },
+    {
+      "epoch": 3.461883408071749,
+      "grad_norm": 0.14841391146183014,
+      "learning_rate": 5.9672204651414564e-06,
+      "loss": 0.3462,
+      "step": 772
+    },
+    {
+      "epoch": 3.4663677130044843,
+      "grad_norm": 0.15835507214069366,
+      "learning_rate": 5.956241371835312e-06,
+      "loss": 0.3342,
+      "step": 773
+    },
+    {
+      "epoch": 3.4708520179372195,
+      "grad_norm": 0.16098253428936005,
+      "learning_rate": 5.945257490763588e-06,
+      "loss": 0.3289,
+      "step": 774
+    },
+    {
+      "epoch": 3.475336322869955,
+      "grad_norm": 0.15453940629959106,
+      "learning_rate": 5.934268876921026e-06,
+      "loss": 0.3407,
+      "step": 775
+    },
+    {
+      "epoch": 3.4798206278026904,
+      "grad_norm": 0.16573141515254974,
+      "learning_rate": 5.923275585326064e-06,
+      "loss": 0.3337,
+      "step": 776
+    },
+    {
+      "epoch": 3.484304932735426,
+      "grad_norm": 0.15477514266967773,
+      "learning_rate": 5.912277671020564e-06,
+      "loss": 0.342,
+      "step": 777
+    },
+    {
+      "epoch": 3.4887892376681613,
+      "grad_norm": 0.16412782669067383,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.3399,
+      "step": 778
+    },
+    {
+      "epoch": 3.493273542600897,
+      "grad_norm": 0.16110502183437347,
+      "learning_rate": 5.890268194560834e-06,
+      "loss": 0.339,
+      "step": 779
+    },
+    {
+      "epoch": 3.497757847533632,
+      "grad_norm": 0.16668060421943665,
+      "learning_rate": 5.8792567426049465e-06,
+      "loss": 0.3381,
+      "step": 780
+    },
+    {
+      "epoch": 3.502242152466368,
+      "grad_norm": 0.16138461232185364,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.3379,
+      "step": 781
+    },
+    {
+      "epoch": 3.506726457399103,
+      "grad_norm": 0.16831059753894806,
+      "learning_rate": 5.857220686904779e-06,
+      "loss": 0.3358,
+      "step": 782
+    },
+    {
+      "epoch": 3.5112107623318387,
+      "grad_norm": 0.18557177484035492,
+      "learning_rate": 5.846196193491919e-06,
+      "loss": 0.3389,
+      "step": 783
+    },
+    {
+      "epoch": 3.515695067264574,
+      "grad_norm": 0.15113940834999084,
+      "learning_rate": 5.835167463294155e-06,
+      "loss": 0.3394,
+      "step": 784
+    },
+    {
+      "epoch": 3.520179372197309,
+      "grad_norm": 0.1885959357023239,
+      "learning_rate": 5.824134551530783e-06,
+      "loss": 0.3334,
+      "step": 785
+    },
+    {
+      "epoch": 3.524663677130045,
+      "grad_norm": 0.1717650145292282,
+      "learning_rate": 5.813097513442035e-06,
+      "loss": 0.3294,
+      "step": 786
+    },
+    {
+      "epoch": 3.5291479820627805,
+      "grad_norm": 0.20333001017570496,
+      "learning_rate": 5.8020564042888015e-06,
+      "loss": 0.3354,
+      "step": 787
+    },
+    {
+      "epoch": 3.5336322869955157,
+      "grad_norm": 0.16548173129558563,
+      "learning_rate": 5.79101127935236e-06,
+      "loss": 0.3294,
+      "step": 788
+    },
+    {
+      "epoch": 3.538116591928251,
+      "grad_norm": 0.21434767544269562,
+      "learning_rate": 5.77996219393409e-06,
+      "loss": 0.3333,
+      "step": 789
+    },
+    {
+      "epoch": 3.5426008968609866,
+      "grad_norm": 0.16596321761608124,
+      "learning_rate": 5.768909203355203e-06,
+      "loss": 0.3307,
+      "step": 790
+    },
+    {
+      "epoch": 3.547085201793722,
+      "grad_norm": 0.19439367949962616,
+      "learning_rate": 5.757852362956463e-06,
+      "loss": 0.3387,
+      "step": 791
+    },
+    {
+      "epoch": 3.5515695067264574,
+      "grad_norm": 0.19180350005626678,
+      "learning_rate": 5.7467917280979105e-06,
+      "loss": 0.3348,
+      "step": 792
+    },
+    {
+      "epoch": 3.5560538116591927,
+      "grad_norm": 0.205781489610672,
+      "learning_rate": 5.735727354158581e-06,
+      "loss": 0.3461,
+      "step": 793
+    },
+    {
+      "epoch": 3.5605381165919283,
+      "grad_norm": 0.19090770184993744,
+      "learning_rate": 5.724659296536234e-06,
+      "loss": 0.346,
+      "step": 794
+    },
+    {
+      "epoch": 3.5650224215246635,
+      "grad_norm": 0.2100742757320404,
+      "learning_rate": 5.713587610647073e-06,
+      "loss": 0.3353,
+      "step": 795
+    },
+    {
+      "epoch": 3.569506726457399,
+      "grad_norm": 0.17336231470108032,
+      "learning_rate": 5.7025123519254644e-06,
+      "loss": 0.3384,
+      "step": 796
+    },
+    {
+      "epoch": 3.5739910313901344,
+      "grad_norm": 0.20998361706733704,
+      "learning_rate": 5.6914335758236665e-06,
+      "loss": 0.3368,
+      "step": 797
+    },
+    {
+      "epoch": 3.57847533632287,
+      "grad_norm": 0.19486036896705627,
+      "learning_rate": 5.680351337811547e-06,
+      "loss": 0.3414,
+      "step": 798
+    },
+    {
+      "epoch": 3.5829596412556053,
+      "grad_norm": 0.20958256721496582,
+      "learning_rate": 5.669265693376309e-06,
+      "loss": 0.3342,
+      "step": 799
+    },
+    {
+      "epoch": 3.587443946188341,
+      "grad_norm": 0.17585420608520508,
+      "learning_rate": 5.658176698022208e-06,
+      "loss": 0.3349,
+      "step": 800
+    },
+    {
+      "epoch": 3.591928251121076,
+      "grad_norm": 0.20159097015857697,
+      "learning_rate": 5.647084407270277e-06,
+      "loss": 0.3424,
+      "step": 801
+    },
+    {
+      "epoch": 3.596412556053812,
+      "grad_norm": 0.16286209225654602,
+      "learning_rate": 5.6359888766580555e-06,
+      "loss": 0.3397,
+      "step": 802
+    },
+    {
+      "epoch": 3.600896860986547,
+      "grad_norm": 0.16896358132362366,
+      "learning_rate": 5.624890161739292e-06,
+      "loss": 0.3365,
+      "step": 803
+    },
+    {
+      "epoch": 3.6053811659192823,
+      "grad_norm": 0.14597639441490173,
+      "learning_rate": 5.6137883180836925e-06,
+      "loss": 0.3355,
+      "step": 804
+    },
+    {
+      "epoch": 3.609865470852018,
+      "grad_norm": 0.16035997867584229,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.3371,
+      "step": 805
+    },
+    {
+      "epoch": 3.6143497757847536,
+      "grad_norm": 0.15880902111530304,
+      "learning_rate": 5.591575466918816e-06,
+      "loss": 0.3379,
+      "step": 806
+    },
+    {
+      "epoch": 3.618834080717489,
+      "grad_norm": 0.15364399552345276,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.3367,
+      "step": 807
+    },
+    {
+      "epoch": 3.623318385650224,
+      "grad_norm": 0.1567133665084839,
+      "learning_rate": 5.569350768029312e-06,
+      "loss": 0.3419,
+      "step": 808
+    },
+    {
+      "epoch": 3.6278026905829597,
+      "grad_norm": 0.15572631359100342,
+      "learning_rate": 5.5582341147735396e-06,
+      "loss": 0.3382,
+      "step": 809
+    },
+    {
+      "epoch": 3.6322869955156953,
+      "grad_norm": 0.14985933899879456,
+      "learning_rate": 5.5471146665183455e-06,
+      "loss": 0.336,
+      "step": 810
+    },
+    {
+      "epoch": 3.6367713004484306,
+      "grad_norm": 0.15054303407669067,
+      "learning_rate": 5.53599247893724e-06,
+      "loss": 0.3355,
+      "step": 811
+    },
+    {
+      "epoch": 3.6412556053811658,
+      "grad_norm": 0.16496680676937103,
+      "learning_rate": 5.524867607717445e-06,
+      "loss": 0.3363,
+      "step": 812
+    },
+    {
+      "epoch": 3.6457399103139014,
+      "grad_norm": 0.15003879368305206,
+      "learning_rate": 5.5137401085596224e-06,
+      "loss": 0.3343,
+      "step": 813
+    },
+    {
+      "epoch": 3.6502242152466366,
+      "grad_norm": 0.15723095834255219,
+      "learning_rate": 5.502610037177586e-06,
+      "loss": 0.3334,
+      "step": 814
+    },
+    {
+      "epoch": 3.6547085201793723,
+      "grad_norm": 0.1780412644147873,
+      "learning_rate": 5.491477449298036e-06,
+      "loss": 0.3386,
+      "step": 815
+    },
+    {
+      "epoch": 3.6591928251121075,
+      "grad_norm": 0.14049941301345825,
+      "learning_rate": 5.480342400660268e-06,
+      "loss": 0.3397,
+      "step": 816
+    },
+    {
+      "epoch": 3.663677130044843,
+      "grad_norm": 0.16899578273296356,
+      "learning_rate": 5.469204947015897e-06,
+      "loss": 0.3377,
+      "step": 817
+    },
+    {
+      "epoch": 3.6681614349775784,
+      "grad_norm": 0.15750017762184143,
+      "learning_rate": 5.458065144128584e-06,
+      "loss": 0.3423,
+      "step": 818
+    },
+    {
+      "epoch": 3.672645739910314,
+      "grad_norm": 0.17824910581111908,
+      "learning_rate": 5.4469230477737466e-06,
+      "loss": 0.3356,
+      "step": 819
+    },
+    {
+      "epoch": 3.6771300448430493,
+      "grad_norm": 0.14895974099636078,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.3323,
+      "step": 820
+    },
+    {
+      "epoch": 3.681614349775785,
+      "grad_norm": 0.1680319458246231,
+      "learning_rate": 5.424632197820325e-06,
+      "loss": 0.341,
+      "step": 821
+    },
+    {
+      "epoch": 3.68609865470852,
+      "grad_norm": 0.16378071904182434,
+      "learning_rate": 5.413483555828879e-06,
+      "loss": 0.3398,
+      "step": 822
+    },
+    {
+      "epoch": 3.6905829596412554,
+      "grad_norm": 0.14856277406215668,
+      "learning_rate": 5.402332843583631e-06,
+      "loss": 0.3387,
+      "step": 823
+    },
+    {
+      "epoch": 3.695067264573991,
+      "grad_norm": 0.18501706421375275,
+      "learning_rate": 5.391180116914621e-06,
+      "loss": 0.3263,
+      "step": 824
+    },
+    {
+      "epoch": 3.6995515695067267,
+      "grad_norm": 0.17995493113994598,
+      "learning_rate": 5.380025431661981e-06,
+      "loss": 0.3468,
+      "step": 825
+    },
+    {
+      "epoch": 3.704035874439462,
+      "grad_norm": 0.1680450141429901,
+      "learning_rate": 5.368868843675642e-06,
+      "loss": 0.3381,
+      "step": 826
+    },
+    {
+      "epoch": 3.708520179372197,
+      "grad_norm": 0.15734758973121643,
+      "learning_rate": 5.3577104088150685e-06,
+      "loss": 0.337,
+      "step": 827
+    },
+    {
+      "epoch": 3.713004484304933,
+      "grad_norm": 0.1911763697862625,
+      "learning_rate": 5.346550182948966e-06,
+      "loss": 0.3343,
+      "step": 828
+    },
+    {
+      "epoch": 3.717488789237668,
+      "grad_norm": 0.16121938824653625,
+      "learning_rate": 5.335388221955012e-06,
+      "loss": 0.332,
+      "step": 829
+    },
+    {
+      "epoch": 3.7219730941704037,
+      "grad_norm": 0.16035863757133484,
+      "learning_rate": 5.3242245817195705e-06,
+      "loss": 0.3428,
+      "step": 830
+    },
+    {
+      "epoch": 3.726457399103139,
+      "grad_norm": 0.18414200842380524,
+      "learning_rate": 5.31305931813741e-06,
+      "loss": 0.3449,
+      "step": 831
+    },
+    {
+      "epoch": 3.7309417040358746,
+      "grad_norm": 0.1640741378068924,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.3375,
+      "step": 832
+    },
+    {
+      "epoch": 3.7354260089686098,
+      "grad_norm": 0.17912285029888153,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.3393,
+      "step": 833
+    },
+    {
+      "epoch": 3.7399103139013454,
+      "grad_norm": 0.19356444478034973,
+      "learning_rate": 5.279554346378572e-06,
+      "loss": 0.3287,
+      "step": 834
+    },
+    {
+      "epoch": 3.7443946188340806,
+      "grad_norm": 0.1706717163324356,
+      "learning_rate": 5.268383148515608e-06,
+      "loss": 0.3387,
+      "step": 835
+    },
+    {
+      "epoch": 3.7488789237668163,
+      "grad_norm": 0.20010025799274445,
+      "learning_rate": 5.257210606896102e-06,
+      "loss": 0.3339,
+      "step": 836
+    },
+    {
+      "epoch": 3.7533632286995515,
+      "grad_norm": 0.18915040791034698,
+      "learning_rate": 5.246036777459391e-06,
+      "loss": 0.3254,
+      "step": 837
+    },
+    {
+      "epoch": 3.7578475336322867,
+      "grad_norm": 0.17235472798347473,
+      "learning_rate": 5.234861716151264e-06,
+      "loss": 0.3429,
+      "step": 838
+    },
+    {
+      "epoch": 3.7623318385650224,
+      "grad_norm": 0.19603897631168365,
+      "learning_rate": 5.223685478923671e-06,
+      "loss": 0.3356,
+      "step": 839
+    },
+    {
+      "epoch": 3.766816143497758,
+      "grad_norm": 0.19278575479984283,
+      "learning_rate": 5.2125081217344595e-06,
+      "loss": 0.3379,
+      "step": 840
+    },
+    {
+      "epoch": 3.7713004484304933,
+      "grad_norm": 0.18289639055728912,
+      "learning_rate": 5.201329700547077e-06,
+      "loss": 0.3317,
+      "step": 841
+    },
+    {
+      "epoch": 3.7757847533632285,
+      "grad_norm": 0.17249856889247894,
+      "learning_rate": 5.190150271330298e-06,
+      "loss": 0.3418,
+      "step": 842
+    },
+    {
+      "epoch": 3.780269058295964,
+      "grad_norm": 0.15763366222381592,
+      "learning_rate": 5.178969890057953e-06,
+      "loss": 0.3289,
+      "step": 843
+    },
+    {
+      "epoch": 3.7847533632287,
+      "grad_norm": 0.20511172711849213,
+      "learning_rate": 5.167788612708627e-06,
+      "loss": 0.3338,
+      "step": 844
+    },
+    {
+      "epoch": 3.789237668161435,
+      "grad_norm": 0.13971680402755737,
+      "learning_rate": 5.156606495265402e-06,
+      "loss": 0.3351,
+      "step": 845
+    },
+    {
+      "epoch": 3.7937219730941703,
+      "grad_norm": 0.20679746568202972,
+      "learning_rate": 5.145423593715558e-06,
+      "loss": 0.3348,
+      "step": 846
+    },
+    {
+      "epoch": 3.798206278026906,
+      "grad_norm": 0.19689729809761047,
+      "learning_rate": 5.1342399640503074e-06,
+      "loss": 0.3407,
+      "step": 847
+    },
+    {
+      "epoch": 3.802690582959641,
+      "grad_norm": 0.16386425495147705,
+      "learning_rate": 5.1230556622645026e-06,
+      "loss": 0.3405,
+      "step": 848
+    },
+    {
+      "epoch": 3.807174887892377,
+      "grad_norm": 0.16386070847511292,
+      "learning_rate": 5.111870744356366e-06,
+      "loss": 0.3384,
+      "step": 849
+    },
+    {
+      "epoch": 3.811659192825112,
+      "grad_norm": 0.16335049271583557,
+      "learning_rate": 5.100685266327202e-06,
+      "loss": 0.3453,
+      "step": 850
+    },
+    {
+      "epoch": 3.8161434977578477,
+      "grad_norm": 0.1500851958990097,
+      "learning_rate": 5.0894992841811216e-06,
+      "loss": 0.3412,
+      "step": 851
+    },
+    {
+      "epoch": 3.820627802690583,
+      "grad_norm": 0.1849871426820755,
+      "learning_rate": 5.0783128539247585e-06,
+      "loss": 0.341,
+      "step": 852
+    },
+    {
+      "epoch": 3.8251121076233185,
+      "grad_norm": 0.15688921511173248,
+      "learning_rate": 5.067126031566988e-06,
+      "loss": 0.3349,
+      "step": 853
+    },
+    {
+      "epoch": 3.8295964125560538,
+      "grad_norm": 0.16467468440532684,
+      "learning_rate": 5.055938873118653e-06,
+      "loss": 0.3299,
+      "step": 854
+    },
+    {
+      "epoch": 3.8340807174887894,
+      "grad_norm": 0.15543334186077118,
+      "learning_rate": 5.044751434592274e-06,
+      "loss": 0.3265,
+      "step": 855
+    },
+    {
+      "epoch": 3.8385650224215246,
+      "grad_norm": 0.16785022616386414,
+      "learning_rate": 5.033563772001782e-06,
+      "loss": 0.3399,
+      "step": 856
+    },
+    {
+      "epoch": 3.84304932735426,
+      "grad_norm": 0.1725115180015564,
+      "learning_rate": 5.022375941362218e-06,
+      "loss": 0.3351,
+      "step": 857
+    },
+    {
+      "epoch": 3.8475336322869955,
+      "grad_norm": 0.16734442114830017,
+      "learning_rate": 5.011187998689474e-06,
+      "loss": 0.33,
+      "step": 858
+    },
+    {
+      "epoch": 3.852017937219731,
+      "grad_norm": 0.18463879823684692,
+      "learning_rate": 5e-06,
+      "loss": 0.3414,
+      "step": 859
+    },
+    {
+      "epoch": 3.8565022421524664,
+      "grad_norm": 0.16997478902339935,
+      "learning_rate": 4.988812001310528e-06,
+      "loss": 0.3406,
+      "step": 860
+    },
+    {
+      "epoch": 3.8609865470852016,
+      "grad_norm": 0.16660833358764648,
+      "learning_rate": 4.977624058637783e-06,
+      "loss": 0.3394,
+      "step": 861
+    },
+    {
+      "epoch": 3.8654708520179373,
+      "grad_norm": 0.18688718974590302,
+      "learning_rate": 4.9664362279982205e-06,
+      "loss": 0.3412,
+      "step": 862
+    },
+    {
+      "epoch": 3.8699551569506725,
+      "grad_norm": 0.15582148730754852,
+      "learning_rate": 4.955248565407727e-06,
+      "loss": 0.3444,
+      "step": 863
+    },
+    {
+      "epoch": 3.874439461883408,
+      "grad_norm": 0.16194719076156616,
+      "learning_rate": 4.944061126881348e-06,
+      "loss": 0.3389,
+      "step": 864
+    },
+    {
+      "epoch": 3.8789237668161434,
+      "grad_norm": 0.15295027196407318,
+      "learning_rate": 4.932873968433014e-06,
+      "loss": 0.3329,
+      "step": 865
+    },
+    {
+      "epoch": 3.883408071748879,
+      "grad_norm": 0.17698495090007782,
+      "learning_rate": 4.921687146075244e-06,
+      "loss": 0.3388,
+      "step": 866
+    },
+    {
+      "epoch": 3.8878923766816142,
+      "grad_norm": 0.16384091973304749,
+      "learning_rate": 4.910500715818879e-06,
+      "loss": 0.3296,
+      "step": 867
+    },
+    {
+      "epoch": 3.89237668161435,
+      "grad_norm": 0.177535280585289,
+      "learning_rate": 4.8993147336728e-06,
+      "loss": 0.3345,
+      "step": 868
+    },
+    {
+      "epoch": 3.896860986547085,
+      "grad_norm": 0.19697459042072296,
+      "learning_rate": 4.8881292556436355e-06,
+      "loss": 0.3369,
+      "step": 869
+    },
+    {
+      "epoch": 3.901345291479821,
+      "grad_norm": 0.1570047289133072,
+      "learning_rate": 4.876944337735499e-06,
+      "loss": 0.3329,
+      "step": 870
+    },
+    {
+      "epoch": 3.905829596412556,
+      "grad_norm": 0.17538248002529144,
+      "learning_rate": 4.865760035949695e-06,
+      "loss": 0.3311,
+      "step": 871
+    },
+    {
+      "epoch": 3.910313901345291,
+      "grad_norm": 0.19360147416591644,
+      "learning_rate": 4.854576406284443e-06,
+      "loss": 0.3293,
+      "step": 872
+    },
+    {
+      "epoch": 3.914798206278027,
+      "grad_norm": 0.15960246324539185,
+      "learning_rate": 4.8433935047346e-06,
+      "loss": 0.3361,
+      "step": 873
+    },
+    {
+      "epoch": 3.9192825112107625,
+      "grad_norm": 0.20789240300655365,
+      "learning_rate": 4.832211387291374e-06,
+      "loss": 0.3358,
+      "step": 874
+    },
+    {
+      "epoch": 3.9237668161434978,
+      "grad_norm": 0.2317878156900406,
+      "learning_rate": 4.821030109942048e-06,
+      "loss": 0.3361,
+      "step": 875
+    },
+    {
+      "epoch": 3.928251121076233,
+      "grad_norm": 0.15479105710983276,
+      "learning_rate": 4.8098497286697024e-06,
+      "loss": 0.341,
+      "step": 876
+    },
+    {
+      "epoch": 3.9327354260089686,
+      "grad_norm": 0.22216734290122986,
+      "learning_rate": 4.798670299452926e-06,
+      "loss": 0.3314,
+      "step": 877
+    },
+    {
+      "epoch": 3.9372197309417043,
+      "grad_norm": 0.21352030336856842,
+      "learning_rate": 4.787491878265542e-06,
+      "loss": 0.3276,
+      "step": 878
+    },
+    {
+      "epoch": 3.9417040358744395,
+      "grad_norm": 0.1807023584842682,
+      "learning_rate": 4.77631452107633e-06,
+      "loss": 0.3405,
+      "step": 879
+    },
+    {
+      "epoch": 3.9461883408071747,
+      "grad_norm": 0.19925130903720856,
+      "learning_rate": 4.765138283848739e-06,
+      "loss": 0.3332,
+      "step": 880
+    },
+    {
+      "epoch": 3.9506726457399104,
+      "grad_norm": 0.23317581415176392,
+      "learning_rate": 4.75396322254061e-06,
+      "loss": 0.3376,
+      "step": 881
+    },
+    {
+      "epoch": 3.9551569506726456,
+      "grad_norm": 0.17920540273189545,
+      "learning_rate": 4.742789393103899e-06,
+      "loss": 0.3335,
+      "step": 882
+    },
+    {
+      "epoch": 3.9596412556053813,
+      "grad_norm": 0.17569604516029358,
+      "learning_rate": 4.731616851484392e-06,
+      "loss": 0.3378,
+      "step": 883
+    },
+    {
+      "epoch": 3.9641255605381165,
+      "grad_norm": 0.21742191910743713,
+      "learning_rate": 4.720445653621429e-06,
+      "loss": 0.3433,
+      "step": 884
+    },
+    {
+      "epoch": 3.968609865470852,
+      "grad_norm": 0.1490844339132309,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.3367,
+      "step": 885
+    },
+    {
+      "epoch": 3.9730941704035874,
+      "grad_norm": 0.17976991832256317,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.3374,
+      "step": 886
+    },
+    {
+      "epoch": 3.977578475336323,
+      "grad_norm": 0.191218763589859,
+      "learning_rate": 4.686940681862591e-06,
+      "loss": 0.3381,
+      "step": 887
+    },
+    {
+      "epoch": 3.9820627802690582,
+      "grad_norm": 0.1576366275548935,
+      "learning_rate": 4.675775418280432e-06,
+      "loss": 0.3422,
+      "step": 888
+    },
+    {
+      "epoch": 3.986547085201794,
+      "grad_norm": 0.17911867797374725,
+      "learning_rate": 4.664611778044988e-06,
+      "loss": 0.3341,
+      "step": 889
+    },
+    {
+      "epoch": 3.991031390134529,
+      "grad_norm": 0.19617079198360443,
+      "learning_rate": 4.653449817051035e-06,
+      "loss": 0.3359,
+      "step": 890
+    },
+    {
+      "epoch": 3.9955156950672643,
+      "grad_norm": 0.15171176195144653,
+      "learning_rate": 4.642289591184934e-06,
+      "loss": 0.3383,
+      "step": 891
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.20553705096244812,
+      "learning_rate": 4.631131156324359e-06,
+      "loss": 0.3341,
+      "step": 892
+    },
+    {
+      "epoch": 4.004484304932736,
+      "grad_norm": 0.24108295142650604,
+      "learning_rate": 4.619974568338021e-06,
+      "loss": 0.3053,
+      "step": 893
+    },
+    {
+      "epoch": 4.00896860986547,
+      "grad_norm": 0.22133147716522217,
+      "learning_rate": 4.60881988308538e-06,
+      "loss": 0.3138,
+      "step": 894
+    },
+    {
+      "epoch": 4.013452914798206,
+      "grad_norm": 0.2606760859489441,
+      "learning_rate": 4.597667156416371e-06,
+      "loss": 0.3109,
+      "step": 895
+    },
+    {
+      "epoch": 4.017937219730942,
+      "grad_norm": 0.2504677474498749,
+      "learning_rate": 4.586516444171123e-06,
+      "loss": 0.3159,
+      "step": 896
+    },
+    {
+      "epoch": 4.022421524663677,
+      "grad_norm": 0.20953048765659332,
+      "learning_rate": 4.575367802179675e-06,
+      "loss": 0.3185,
+      "step": 897
+    },
+    {
+      "epoch": 4.026905829596412,
+      "grad_norm": 0.20957928895950317,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.311,
+      "step": 898
+    },
+    {
+      "epoch": 4.031390134529148,
+      "grad_norm": 0.22497303783893585,
+      "learning_rate": 4.553076952226255e-06,
+      "loss": 0.3204,
+      "step": 899
+    },
+    {
+      "epoch": 4.0358744394618835,
+      "grad_norm": 0.2039749026298523,
+      "learning_rate": 4.541934855871417e-06,
+      "loss": 0.3153,
+      "step": 900
+    },
+    {
+      "epoch": 4.040358744394619,
+      "grad_norm": 0.2252999246120453,
+      "learning_rate": 4.530795052984104e-06,
+      "loss": 0.315,
+      "step": 901
+    },
+    {
+      "epoch": 4.044843049327354,
+      "grad_norm": 0.21373526751995087,
+      "learning_rate": 4.519657599339735e-06,
+      "loss": 0.3127,
+      "step": 902
+    },
+    {
+      "epoch": 4.04932735426009,
+      "grad_norm": 0.24079857766628265,
+      "learning_rate": 4.508522550701965e-06,
+      "loss": 0.3035,
+      "step": 903
+    },
+    {
+      "epoch": 4.053811659192825,
+      "grad_norm": 0.16769376397132874,
+      "learning_rate": 4.497389962822416e-06,
+      "loss": 0.308,
+      "step": 904
+    },
+    {
+      "epoch": 4.058295964125561,
+      "grad_norm": 0.2299014925956726,
+      "learning_rate": 4.48625989144038e-06,
+      "loss": 0.3117,
+      "step": 905
+    },
+    {
+      "epoch": 4.062780269058296,
+      "grad_norm": 0.22668282687664032,
+      "learning_rate": 4.475132392282556e-06,
+      "loss": 0.3114,
+      "step": 906
+    },
+    {
+      "epoch": 4.067264573991031,
+      "grad_norm": 0.17460782825946808,
+      "learning_rate": 4.464007521062761e-06,
+      "loss": 0.3149,
+      "step": 907
+    },
+    {
+      "epoch": 4.071748878923767,
+      "grad_norm": 0.20387382805347443,
+      "learning_rate": 4.452885333481657e-06,
+      "loss": 0.3185,
+      "step": 908
+    },
+    {
+      "epoch": 4.076233183856502,
+      "grad_norm": 0.20579856634140015,
+      "learning_rate": 4.441765885226462e-06,
+      "loss": 0.3138,
+      "step": 909
+    },
+    {
+      "epoch": 4.0807174887892375,
+      "grad_norm": 0.1666853427886963,
+      "learning_rate": 4.4306492319706895e-06,
+      "loss": 0.3236,
+      "step": 910
+    },
+    {
+      "epoch": 4.085201793721973,
+      "grad_norm": 0.20942482352256775,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.3118,
+      "step": 911
+    },
+    {
+      "epoch": 4.089686098654709,
+      "grad_norm": 0.20370063185691833,
+      "learning_rate": 4.4084245330811855e-06,
+      "loss": 0.3171,
+      "step": 912
+    },
+    {
+      "epoch": 4.0941704035874436,
+      "grad_norm": 0.19959072768688202,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.3155,
+      "step": 913
+    },
+    {
+      "epoch": 4.098654708520179,
+      "grad_norm": 0.19464878737926483,
+      "learning_rate": 4.386211681916309e-06,
+      "loss": 0.3191,
+      "step": 914
+    },
+    {
+      "epoch": 4.103139013452915,
+      "grad_norm": 0.17180007696151733,
+      "learning_rate": 4.3751098382607084e-06,
+      "loss": 0.3249,
+      "step": 915
+    },
+    {
+      "epoch": 4.1076233183856505,
+      "grad_norm": 0.2266189604997635,
+      "learning_rate": 4.364011123341947e-06,
+      "loss": 0.3128,
+      "step": 916
+    },
+    {
+      "epoch": 4.112107623318385,
+      "grad_norm": 0.16010649502277374,
+      "learning_rate": 4.352915592729723e-06,
+      "loss": 0.3214,
+      "step": 917
+    },
+    {
+      "epoch": 4.116591928251121,
+      "grad_norm": 0.17680487036705017,
+      "learning_rate": 4.341823301977794e-06,
+      "loss": 0.3112,
+      "step": 918
+    },
+    {
+      "epoch": 4.121076233183857,
+      "grad_norm": 0.17622031271457672,
+      "learning_rate": 4.330734306623694e-06,
+      "loss": 0.3089,
+      "step": 919
+    },
+    {
+      "epoch": 4.125560538116592,
+      "grad_norm": 0.1706295609474182,
+      "learning_rate": 4.319648662188453e-06,
+      "loss": 0.3211,
+      "step": 920
+    },
+    {
+      "epoch": 4.130044843049327,
+      "grad_norm": 0.1623934656381607,
+      "learning_rate": 4.308566424176336e-06,
+      "loss": 0.3157,
+      "step": 921
+    },
+    {
+      "epoch": 4.134529147982063,
+      "grad_norm": 0.16201253235340118,
+      "learning_rate": 4.297487648074538e-06,
+      "loss": 0.3065,
+      "step": 922
+    },
+    {
+      "epoch": 4.139013452914798,
+      "grad_norm": 0.16154931485652924,
+      "learning_rate": 4.286412389352929e-06,
+      "loss": 0.3167,
+      "step": 923
+    },
+    {
+      "epoch": 4.143497757847534,
+      "grad_norm": 0.1697588711977005,
+      "learning_rate": 4.275340703463767e-06,
+      "loss": 0.3078,
+      "step": 924
+    },
+    {
+      "epoch": 4.147982062780269,
+      "grad_norm": 0.20315028727054596,
+      "learning_rate": 4.264272645841419e-06,
+      "loss": 0.316,
+      "step": 925
+    },
+    {
+      "epoch": 4.1524663677130045,
+      "grad_norm": 0.1631268411874771,
+      "learning_rate": 4.253208271902091e-06,
+      "loss": 0.307,
+      "step": 926
+    },
+    {
+      "epoch": 4.15695067264574,
+      "grad_norm": 0.18198901414871216,
+      "learning_rate": 4.242147637043539e-06,
+      "loss": 0.3201,
+      "step": 927
+    },
+    {
+      "epoch": 4.161434977578475,
+      "grad_norm": 0.18274594843387604,
+      "learning_rate": 4.231090796644798e-06,
+      "loss": 0.31,
+      "step": 928
+    },
+    {
+      "epoch": 4.165919282511211,
+      "grad_norm": 0.18858183920383453,
+      "learning_rate": 4.220037806065911e-06,
+      "loss": 0.3145,
+      "step": 929
+    },
+    {
+      "epoch": 4.170403587443946,
+      "grad_norm": 0.18607163429260254,
+      "learning_rate": 4.208988720647642e-06,
+      "loss": 0.319,
+      "step": 930
+    },
+    {
+      "epoch": 4.174887892376682,
+      "grad_norm": 0.19816173613071442,
+      "learning_rate": 4.1979435957111984e-06,
+      "loss": 0.3092,
+      "step": 931
+    },
+    {
+      "epoch": 4.179372197309417,
+      "grad_norm": 0.142039492726326,
+      "learning_rate": 4.1869024865579664e-06,
+      "loss": 0.3107,
+      "step": 932
+    },
+    {
+      "epoch": 4.183856502242152,
+      "grad_norm": 0.18035510182380676,
+      "learning_rate": 4.175865448469219e-06,
+      "loss": 0.3184,
+      "step": 933
+    },
+    {
+      "epoch": 4.188340807174888,
+      "grad_norm": 0.14774839580059052,
+      "learning_rate": 4.164832536705845e-06,
+      "loss": 0.3074,
+      "step": 934
+    },
+    {
+      "epoch": 4.192825112107624,
+      "grad_norm": 0.19417749345302582,
+      "learning_rate": 4.153803806508083e-06,
+      "loss": 0.312,
+      "step": 935
+    },
+    {
+      "epoch": 4.197309417040358,
+      "grad_norm": 0.17361126840114594,
+      "learning_rate": 4.142779313095223e-06,
+      "loss": 0.3121,
+      "step": 936
+    },
+    {
+      "epoch": 4.201793721973094,
+      "grad_norm": 0.1730654239654541,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.3192,
+      "step": 937
+    },
+    {
+      "epoch": 4.20627802690583,
+      "grad_norm": 0.16994503140449524,
+      "learning_rate": 4.120743257395054e-06,
+      "loss": 0.3044,
+      "step": 938
+    },
+    {
+      "epoch": 4.210762331838565,
+      "grad_norm": 0.17273321747779846,
+      "learning_rate": 4.109731805439168e-06,
+      "loss": 0.3209,
+      "step": 939
+    },
+    {
+      "epoch": 4.2152466367713,
+      "grad_norm": 0.19757074117660522,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.3121,
+      "step": 940
+    },
+    {
+      "epoch": 4.219730941704036,
+      "grad_norm": 0.19798298180103302,
+      "learning_rate": 4.087722328979437e-06,
+      "loss": 0.3106,
+      "step": 941
+    },
+    {
+      "epoch": 4.2242152466367715,
+      "grad_norm": 0.17775893211364746,
+      "learning_rate": 4.076724414673937e-06,
+      "loss": 0.3107,
+      "step": 942
+    },
+    {
+      "epoch": 4.228699551569507,
+      "grad_norm": 0.16296343505382538,
+      "learning_rate": 4.065731123078977e-06,
+      "loss": 0.3017,
+      "step": 943
+    },
+    {
+      "epoch": 4.233183856502242,
+      "grad_norm": 0.18222899734973907,
+      "learning_rate": 4.054742509236416e-06,
+      "loss": 0.3197,
+      "step": 944
+    },
+    {
+      "epoch": 4.237668161434978,
+      "grad_norm": 0.1799403876066208,
+      "learning_rate": 4.043758628164688e-06,
+      "loss": 0.3183,
+      "step": 945
+    },
+    {
+      "epoch": 4.242152466367713,
+      "grad_norm": 0.17246119678020477,
+      "learning_rate": 4.032779534858544e-06,
+      "loss": 0.3086,
+      "step": 946
+    },
+    {
+      "epoch": 4.246636771300448,
+      "grad_norm": 0.1754005253314972,
+      "learning_rate": 4.021805284288749e-06,
+      "loss": 0.3208,
+      "step": 947
+    },
+    {
+      "epoch": 4.251121076233184,
+      "grad_norm": 0.1725863665342331,
+      "learning_rate": 4.0108359314018254e-06,
+      "loss": 0.3126,
+      "step": 948
+    },
+    {
+      "epoch": 4.255605381165919,
+      "grad_norm": 0.18075457215309143,
+      "learning_rate": 3.999871531119779e-06,
+      "loss": 0.3131,
+      "step": 949
+    },
+    {
+      "epoch": 4.260089686098655,
+      "grad_norm": 0.16615353524684906,
+      "learning_rate": 3.988912138339812e-06,
+      "loss": 0.3196,
+      "step": 950
+    },
+    {
+      "epoch": 4.26457399103139,
+      "grad_norm": 0.19933255016803741,
+      "learning_rate": 3.977957807934055e-06,
+      "loss": 0.3105,
+      "step": 951
+    },
+    {
+      "epoch": 4.2690582959641254,
+      "grad_norm": 0.17227207124233246,
+      "learning_rate": 3.9670085947493e-06,
+      "loss": 0.3168,
+      "step": 952
+    },
+    {
+      "epoch": 4.273542600896861,
+      "grad_norm": 0.18315184116363525,
+      "learning_rate": 3.956064553606708e-06,
+      "loss": 0.314,
+      "step": 953
+    },
+    {
+      "epoch": 4.278026905829597,
+      "grad_norm": 0.17916379868984222,
+      "learning_rate": 3.945125739301547e-06,
+      "loss": 0.3127,
+      "step": 954
+    },
+    {
+      "epoch": 4.2825112107623315,
+      "grad_norm": 0.17832358181476593,
+      "learning_rate": 3.934192206602921e-06,
+      "loss": 0.3099,
+      "step": 955
+    },
+    {
+      "epoch": 4.286995515695067,
+      "grad_norm": 0.16912692785263062,
+      "learning_rate": 3.923264010253479e-06,
+      "loss": 0.3198,
+      "step": 956
+    },
+    {
+      "epoch": 4.291479820627803,
+      "grad_norm": 0.1579030454158783,
+      "learning_rate": 3.912341204969164e-06,
+      "loss": 0.307,
+      "step": 957
+    },
+    {
+      "epoch": 4.2959641255605385,
+      "grad_norm": 0.17374159395694733,
+      "learning_rate": 3.901423845438916e-06,
+      "loss": 0.3148,
+      "step": 958
+    },
+    {
+      "epoch": 4.300448430493273,
+      "grad_norm": 0.19913436472415924,
+      "learning_rate": 3.890511986324413e-06,
+      "loss": 0.3093,
+      "step": 959
+    },
+    {
+      "epoch": 4.304932735426009,
+      "grad_norm": 0.14966368675231934,
+      "learning_rate": 3.879605682259797e-06,
+      "loss": 0.3123,
+      "step": 960
+    },
+    {
+      "epoch": 4.309417040358745,
+      "grad_norm": 0.1593165248632431,
+      "learning_rate": 3.86870498785139e-06,
+      "loss": 0.3126,
+      "step": 961
+    },
+    {
+      "epoch": 4.31390134529148,
+      "grad_norm": 0.1794804036617279,
+      "learning_rate": 3.857809957677432e-06,
+      "loss": 0.3171,
+      "step": 962
+    },
+    {
+      "epoch": 4.318385650224215,
+      "grad_norm": 0.14832071959972382,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.317,
+      "step": 963
+    },
+    {
+      "epoch": 4.322869955156951,
+      "grad_norm": 0.14880065619945526,
+      "learning_rate": 3.83603710820374e-06,
+      "loss": 0.3204,
+      "step": 964
+    },
+    {
+      "epoch": 4.327354260089686,
+      "grad_norm": 0.14962953329086304,
+      "learning_rate": 3.825159397917589e-06,
+      "loss": 0.3096,
+      "step": 965
+    },
+    {
+      "epoch": 4.331838565022421,
+      "grad_norm": 0.14541302621364594,
+      "learning_rate": 3.814287569892512e-06,
+      "loss": 0.3153,
+      "step": 966
+    },
+    {
+      "epoch": 4.336322869955157,
+      "grad_norm": 0.15341897308826447,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.3199,
+      "step": 967
+    },
+    {
+      "epoch": 4.3408071748878925,
+      "grad_norm": 0.16074025630950928,
+      "learning_rate": 3.7925617783306757e-06,
+      "loss": 0.3143,
+      "step": 968
+    },
+    {
+      "epoch": 4.345291479820628,
+      "grad_norm": 0.1580256223678589,
+      "learning_rate": 3.781707923571891e-06,
+      "loss": 0.3057,
+      "step": 969
+    },
+    {
+      "epoch": 4.349775784753363,
+      "grad_norm": 0.16166260838508606,
+      "learning_rate": 3.7708601686295756e-06,
+      "loss": 0.3138,
+      "step": 970
+    },
+    {
+      "epoch": 4.354260089686099,
+      "grad_norm": 0.17629729211330414,
+      "learning_rate": 3.7600185678169083e-06,
+      "loss": 0.323,
+      "step": 971
+    },
+    {
+      "epoch": 4.358744394618834,
+      "grad_norm": 0.14726987481117249,
+      "learning_rate": 3.7491831754162533e-06,
+      "loss": 0.3076,
+      "step": 972
+    },
+    {
+      "epoch": 4.36322869955157,
+      "grad_norm": 0.19199849665164948,
+      "learning_rate": 3.7383540456788915e-06,
+      "loss": 0.3156,
+      "step": 973
+    },
+    {
+      "epoch": 4.367713004484305,
+      "grad_norm": 0.17682954668998718,
+      "learning_rate": 3.727531232824747e-06,
+      "loss": 0.3184,
+      "step": 974
+    },
+    {
+      "epoch": 4.37219730941704,
+      "grad_norm": 0.16782739758491516,
+      "learning_rate": 3.7167147910421165e-06,
+      "loss": 0.3172,
+      "step": 975
+    },
+    {
+      "epoch": 4.376681614349776,
+      "grad_norm": 0.15807676315307617,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.3195,
+      "step": 976
+    },
+    {
+      "epoch": 4.381165919282511,
+      "grad_norm": 0.18422642350196838,
+      "learning_rate": 3.695101237284815e-06,
+      "loss": 0.3247,
+      "step": 977
+    },
+    {
+      "epoch": 4.385650224215246,
+      "grad_norm": 0.15524937212467194,
+      "learning_rate": 3.6843042335261583e-06,
+      "loss": 0.3184,
+      "step": 978
+    },
+    {
+      "epoch": 4.390134529147982,
+      "grad_norm": 0.16836954653263092,
+      "learning_rate": 3.6735138172704967e-06,
+      "loss": 0.3163,
+      "step": 979
+    },
+    {
+      "epoch": 4.394618834080718,
+      "grad_norm": 0.18054649233818054,
+      "learning_rate": 3.662730042543926e-06,
+      "loss": 0.3093,
+      "step": 980
+    },
+    {
+      "epoch": 4.3991031390134525,
+      "grad_norm": 0.16283389925956726,
+      "learning_rate": 3.6519529633392825e-06,
+      "loss": 0.3148,
+      "step": 981
+    },
+    {
+      "epoch": 4.403587443946188,
+      "grad_norm": 0.17240117490291595,
+      "learning_rate": 3.6411826336158785e-06,
+      "loss": 0.3146,
+      "step": 982
+    },
+    {
+      "epoch": 4.408071748878924,
+      "grad_norm": 0.16425174474716187,
+      "learning_rate": 3.6304191072992376e-06,
+      "loss": 0.311,
+      "step": 983
+    },
+    {
+      "epoch": 4.4125560538116595,
+      "grad_norm": 0.17400246858596802,
+      "learning_rate": 3.619662438280816e-06,
+      "loss": 0.3095,
+      "step": 984
+    },
+    {
+      "epoch": 4.417040358744394,
+      "grad_norm": 0.18344831466674805,
+      "learning_rate": 3.6089126804177373e-06,
+      "loss": 0.3171,
+      "step": 985
+    },
+    {
+      "epoch": 4.42152466367713,
+      "grad_norm": 0.18111863732337952,
+      "learning_rate": 3.5981698875325214e-06,
+      "loss": 0.3238,
+      "step": 986
+    },
+    {
+      "epoch": 4.426008968609866,
+      "grad_norm": 0.1845109462738037,
+      "learning_rate": 3.5874341134128156e-06,
+      "loss": 0.3139,
+      "step": 987
+    },
+    {
+      "epoch": 4.430493273542601,
+      "grad_norm": 0.20454421639442444,
+      "learning_rate": 3.5767054118111266e-06,
+      "loss": 0.3093,
+      "step": 988
+    },
+    {
+      "epoch": 4.434977578475336,
+      "grad_norm": 0.17412368953227997,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.3149,
+      "step": 989
+    },
+    {
+      "epoch": 4.439461883408072,
+      "grad_norm": 0.17388635873794556,
+      "learning_rate": 3.555269440994496e-06,
+      "loss": 0.3119,
+      "step": 990
+    },
+    {
+      "epoch": 4.443946188340807,
+      "grad_norm": 0.1891017109155655,
+      "learning_rate": 3.544562279106436e-06,
+      "loss": 0.3165,
+      "step": 991
+    },
+    {
+      "epoch": 4.448430493273543,
+      "grad_norm": 0.15251921117305756,
+      "learning_rate": 3.5338624043896154e-06,
+      "loss": 0.3173,
+      "step": 992
+    },
+    {
+      "epoch": 4.452914798206278,
+      "grad_norm": 0.15604549646377563,
+      "learning_rate": 3.523169870416795e-06,
+      "loss": 0.3204,
+      "step": 993
+    },
+    {
+      "epoch": 4.457399103139013,
+      "grad_norm": 0.15084540843963623,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.3154,
+      "step": 994
+    },
+    {
+      "epoch": 4.461883408071749,
+      "grad_norm": 0.15310721099376678,
+      "learning_rate": 3.501807038810174e-06,
+      "loss": 0.3164,
+      "step": 995
+    },
+    {
+      "epoch": 4.466367713004484,
+      "grad_norm": 0.1659664511680603,
+      "learning_rate": 3.4911368481370535e-06,
+      "loss": 0.3081,
+      "step": 996
+    },
+    {
+      "epoch": 4.4708520179372195,
+      "grad_norm": 0.18596197664737701,
+      "learning_rate": 3.480474212128766e-06,
+      "loss": 0.3175,
+      "step": 997
+    },
+    {
+      "epoch": 4.475336322869955,
+      "grad_norm": 0.16384641826152802,
+      "learning_rate": 3.469819184171623e-06,
+      "loss": 0.3115,
+      "step": 998
+    },
+    {
+      "epoch": 4.479820627802691,
+      "grad_norm": 0.15781474113464355,
+      "learning_rate": 3.459171817613847e-06,
+      "loss": 0.3196,
+      "step": 999
+    },
+    {
+      "epoch": 4.484304932735426,
+      "grad_norm": 0.16845561563968658,
+      "learning_rate": 3.4485321657653e-06,
+      "loss": 0.316,
+      "step": 1000
+    },
+    {
+      "epoch": 4.488789237668161,
+      "grad_norm": 0.15221434831619263,
+      "learning_rate": 3.4379002818972122e-06,
+      "loss": 0.3242,
+      "step": 1001
+    },
+    {
+      "epoch": 4.493273542600897,
+      "grad_norm": 0.15872353315353394,
+      "learning_rate": 3.427276219241933e-06,
+      "loss": 0.313,
+      "step": 1002
+    },
+    {
+      "epoch": 4.497757847533633,
+      "grad_norm": 0.19064044952392578,
+      "learning_rate": 3.416660030992639e-06,
+      "loss": 0.3158,
+      "step": 1003
+    },
+    {
+      "epoch": 4.502242152466367,
+      "grad_norm": 0.16245710849761963,
+      "learning_rate": 3.406051770303087e-06,
+      "loss": 0.3159,
+      "step": 1004
+    },
+    {
+      "epoch": 4.506726457399103,
+      "grad_norm": 0.17834170162677765,
+      "learning_rate": 3.3954514902873427e-06,
+      "loss": 0.327,
+      "step": 1005
+    },
+    {
+      "epoch": 4.511210762331839,
+      "grad_norm": 0.1624448597431183,
+      "learning_rate": 3.3848592440195118e-06,
+      "loss": 0.3142,
+      "step": 1006
+    },
+    {
+      "epoch": 4.515695067264574,
+      "grad_norm": 0.17303700745105743,
+      "learning_rate": 3.3742750845334748e-06,
+      "loss": 0.3111,
+      "step": 1007
+    },
+    {
+      "epoch": 4.520179372197309,
+      "grad_norm": 0.1484447717666626,
+      "learning_rate": 3.3636990648226258e-06,
+      "loss": 0.312,
+      "step": 1008
+    },
+    {
+      "epoch": 4.524663677130045,
+      "grad_norm": 0.15595373511314392,
+      "learning_rate": 3.3531312378396026e-06,
+      "loss": 0.309,
+      "step": 1009
+    },
+    {
+      "epoch": 4.5291479820627805,
+      "grad_norm": 0.1732824444770813,
+      "learning_rate": 3.342571656496022e-06,
+      "loss": 0.3241,
+      "step": 1010
+    },
+    {
+      "epoch": 4.533632286995516,
+      "grad_norm": 0.15389838814735413,
+      "learning_rate": 3.3320203736622185e-06,
+      "loss": 0.3209,
+      "step": 1011
+    },
+    {
+      "epoch": 4.538116591928251,
+      "grad_norm": 0.16666561365127563,
+      "learning_rate": 3.3214774421669777e-06,
+      "loss": 0.3121,
+      "step": 1012
+    },
+    {
+      "epoch": 4.542600896860987,
+      "grad_norm": 0.17681841552257538,
+      "learning_rate": 3.310942914797265e-06,
+      "loss": 0.3239,
+      "step": 1013
+    },
+    {
+      "epoch": 4.547085201793722,
+      "grad_norm": 0.18232913315296173,
+      "learning_rate": 3.3004168442979755e-06,
+      "loss": 0.3126,
+      "step": 1014
+    },
+    {
+      "epoch": 4.551569506726457,
+      "grad_norm": 0.15436193346977234,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.3184,
+      "step": 1015
+    },
+    {
+      "epoch": 4.556053811659193,
+      "grad_norm": 0.17936591804027557,
+      "learning_rate": 3.2793902846782534e-06,
+      "loss": 0.318,
+      "step": 1016
+    },
+    {
+      "epoch": 4.560538116591928,
+      "grad_norm": 0.15491776168346405,
+      "learning_rate": 3.2688899008348386e-06,
+      "loss": 0.3151,
+      "step": 1017
+    },
+    {
+      "epoch": 4.565022421524664,
+      "grad_norm": 0.17885245382785797,
+      "learning_rate": 3.2583981844153487e-06,
+      "loss": 0.3165,
+      "step": 1018
+    },
+    {
+      "epoch": 4.569506726457399,
+      "grad_norm": 0.15264320373535156,
+      "learning_rate": 3.2479151879503324e-06,
+      "loss": 0.3158,
+      "step": 1019
+    },
+    {
+      "epoch": 4.573991031390134,
+      "grad_norm": 0.17361338436603546,
+      "learning_rate": 3.2374409639266695e-06,
+      "loss": 0.3194,
+      "step": 1020
+    },
+    {
+      "epoch": 4.57847533632287,
+      "grad_norm": 0.16670067608356476,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.3133,
+      "step": 1021
+    },
+    {
+      "epoch": 4.582959641255606,
+      "grad_norm": 0.19178202748298645,
+      "learning_rate": 3.2165190429310674e-06,
+      "loss": 0.3265,
+      "step": 1022
+    },
+    {
+      "epoch": 4.5874439461883405,
+      "grad_norm": 0.15275251865386963,
+      "learning_rate": 3.206071450712235e-06,
+      "loss": 0.3164,
+      "step": 1023
+    },
+    {
+      "epoch": 4.591928251121076,
+      "grad_norm": 0.18845996260643005,
+      "learning_rate": 3.1956328404404403e-06,
+      "loss": 0.3106,
+      "step": 1024
+    },
+    {
+      "epoch": 4.596412556053812,
+      "grad_norm": 0.1484411358833313,
+      "learning_rate": 3.1852032643803377e-06,
+      "loss": 0.3163,
+      "step": 1025
+    },
+    {
+      "epoch": 4.600896860986547,
+      "grad_norm": 0.1562689244747162,
+      "learning_rate": 3.174782774751338e-06,
+      "loss": 0.3123,
+      "step": 1026
+    },
+    {
+      "epoch": 4.605381165919282,
+      "grad_norm": 0.17432962357997894,
+      "learning_rate": 3.1643714237273628e-06,
+      "loss": 0.3161,
+      "step": 1027
+    },
+    {
+      "epoch": 4.609865470852018,
+      "grad_norm": 0.16095484793186188,
+      "learning_rate": 3.1539692634365788e-06,
+      "loss": 0.3187,
+      "step": 1028
+    },
+    {
+      "epoch": 4.614349775784754,
+      "grad_norm": 0.15901044011116028,
+      "learning_rate": 3.143576345961132e-06,
+      "loss": 0.3226,
+      "step": 1029
+    },
+    {
+      "epoch": 4.618834080717489,
+      "grad_norm": 0.18347200751304626,
+      "learning_rate": 3.1331927233368954e-06,
+      "loss": 0.3172,
+      "step": 1030
+    },
+    {
+      "epoch": 4.623318385650224,
+      "grad_norm": 0.1799110472202301,
+      "learning_rate": 3.1228184475532015e-06,
+      "loss": 0.321,
+      "step": 1031
+    },
+    {
+      "epoch": 4.62780269058296,
+      "grad_norm": 0.1431790590286255,
+      "learning_rate": 3.112453570552583e-06,
+      "loss": 0.3217,
+      "step": 1032
+    },
+    {
+      "epoch": 4.632286995515695,
+      "grad_norm": 0.1613750010728836,
+      "learning_rate": 3.1020981442305187e-06,
+      "loss": 0.3158,
+      "step": 1033
+    },
+    {
+      "epoch": 4.63677130044843,
+      "grad_norm": 0.15131519734859467,
+      "learning_rate": 3.091752220435166e-06,
+      "loss": 0.3156,
+      "step": 1034
+    },
+    {
+      "epoch": 4.641255605381166,
+      "grad_norm": 0.17270632088184357,
+      "learning_rate": 3.0814158509671015e-06,
+      "loss": 0.3109,
+      "step": 1035
+    },
+    {
+      "epoch": 4.645739910313901,
+      "grad_norm": 0.15472306311130524,
+      "learning_rate": 3.0710890875790745e-06,
+      "loss": 0.3201,
+      "step": 1036
+    },
+    {
+      "epoch": 4.650224215246637,
+      "grad_norm": 0.14837723970413208,
+      "learning_rate": 3.0607719819757264e-06,
+      "loss": 0.3281,
+      "step": 1037
+    },
+    {
+      "epoch": 4.654708520179372,
+      "grad_norm": 0.15884482860565186,
+      "learning_rate": 3.0504645858133507e-06,
+      "loss": 0.304,
+      "step": 1038
+    },
+    {
+      "epoch": 4.6591928251121075,
+      "grad_norm": 0.14489606022834778,
+      "learning_rate": 3.040166950699626e-06,
+      "loss": 0.3207,
+      "step": 1039
+    },
+    {
+      "epoch": 4.663677130044843,
+      "grad_norm": 0.15721720457077026,
+      "learning_rate": 3.029879128193356e-06,
+      "loss": 0.3199,
+      "step": 1040
+    },
+    {
+      "epoch": 4.668161434977579,
+      "grad_norm": 0.14078550040721893,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.3129,
+      "step": 1041
+    },
+    {
+      "epoch": 4.672645739910314,
+      "grad_norm": 0.15547266602516174,
+      "learning_rate": 3.0093331269924954e-06,
+      "loss": 0.3179,
+      "step": 1042
+    },
+    {
+      "epoch": 4.677130044843049,
+      "grad_norm": 0.15227945148944855,
+      "learning_rate": 2.9990750511688323e-06,
+      "loss": 0.314,
+      "step": 1043
+    },
+    {
+      "epoch": 4.681614349775785,
+      "grad_norm": 0.15421082079410553,
+      "learning_rate": 2.988826993693967e-06,
+      "loss": 0.3138,
+      "step": 1044
+    },
+    {
+      "epoch": 4.68609865470852,
+      "grad_norm": 0.15077437460422516,
+      "learning_rate": 2.978589005878476e-06,
+      "loss": 0.3153,
+      "step": 1045
+    },
+    {
+      "epoch": 4.690582959641255,
+      "grad_norm": 0.1700594425201416,
+      "learning_rate": 2.968361138982517e-06,
+      "loss": 0.3275,
+      "step": 1046
+    },
+    {
+      "epoch": 4.695067264573991,
+      "grad_norm": 0.1545550525188446,
+      "learning_rate": 2.9581434442155798e-06,
+      "loss": 0.3184,
+      "step": 1047
+    },
+    {
+      "epoch": 4.699551569506727,
+      "grad_norm": 0.17725931107997894,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.3248,
+      "step": 1048
+    },
+    {
+      "epoch": 4.704035874439462,
+      "grad_norm": 0.174936443567276,
+      "learning_rate": 2.937738775651798e-06,
+      "loss": 0.3125,
+      "step": 1049
+    },
+    {
+      "epoch": 4.708520179372197,
+      "grad_norm": 0.16865046322345734,
+      "learning_rate": 2.9275519040182503e-06,
+      "loss": 0.3189,
+      "step": 1050
+    },
+    {
+      "epoch": 4.713004484304933,
+      "grad_norm": 0.1909942477941513,
+      "learning_rate": 2.917375408839803e-06,
+      "loss": 0.3169,
+      "step": 1051
+    },
+    {
+      "epoch": 4.7174887892376685,
+      "grad_norm": 0.1597238928079605,
+      "learning_rate": 2.9072093410687268e-06,
+      "loss": 0.311,
+      "step": 1052
+    },
+    {
+      "epoch": 4.721973094170403,
+      "grad_norm": 0.1816689521074295,
+      "learning_rate": 2.8970537516050935e-06,
+      "loss": 0.3141,
+      "step": 1053
+    },
+    {
+      "epoch": 4.726457399103139,
+      "grad_norm": 0.1508946716785431,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.3186,
+      "step": 1054
+    },
+    {
+      "epoch": 4.7309417040358746,
+      "grad_norm": 0.1811760514974594,
+      "learning_rate": 2.876774210937843e-06,
+      "loss": 0.3116,
+      "step": 1055
+    },
+    {
+      "epoch": 4.73542600896861,
+      "grad_norm": 0.18800127506256104,
+      "learning_rate": 2.866650361271023e-06,
+      "loss": 0.3154,
+      "step": 1056
+    },
+    {
+      "epoch": 4.739910313901345,
+      "grad_norm": 0.14846614003181458,
+      "learning_rate": 2.8565371929847286e-06,
+      "loss": 0.3212,
+      "step": 1057
+    },
+    {
+      "epoch": 4.744394618834081,
+      "grad_norm": 0.18500037491321564,
+      "learning_rate": 2.84643475671417e-06,
+      "loss": 0.3212,
+      "step": 1058
+    },
+    {
+      "epoch": 4.748878923766816,
+      "grad_norm": 0.18749377131462097,
+      "learning_rate": 2.836343103040819e-06,
+      "loss": 0.3284,
+      "step": 1059
+    },
+    {
+      "epoch": 4.753363228699552,
+      "grad_norm": 0.18528147041797638,
+      "learning_rate": 2.8262622824921593e-06,
+      "loss": 0.3166,
+      "step": 1060
+    },
+    {
+      "epoch": 4.757847533632287,
+      "grad_norm": 0.1705624908208847,
+      "learning_rate": 2.816192345541437e-06,
+      "loss": 0.3105,
+      "step": 1061
+    },
+    {
+      "epoch": 4.762331838565022,
+      "grad_norm": 0.1956166923046112,
+      "learning_rate": 2.8061333426074054e-06,
+      "loss": 0.3087,
+      "step": 1062
+    },
+    {
+      "epoch": 4.766816143497758,
+      "grad_norm": 0.17022459208965302,
+      "learning_rate": 2.7960853240540703e-06,
+      "loss": 0.3264,
+      "step": 1063
+    },
+    {
+      "epoch": 4.771300448430493,
+      "grad_norm": 0.17973558604717255,
+      "learning_rate": 2.7860483401904483e-06,
+      "loss": 0.3145,
+      "step": 1064
+    },
+    {
+      "epoch": 4.7757847533632285,
+      "grad_norm": 0.15205493569374084,
+      "learning_rate": 2.776022441270295e-06,
+      "loss": 0.3096,
+      "step": 1065
+    },
+    {
+      "epoch": 4.780269058295964,
+      "grad_norm": 0.16868656873703003,
+      "learning_rate": 2.766007677491871e-06,
+      "loss": 0.3181,
+      "step": 1066
+    },
+    {
+      "epoch": 4.7847533632287,
+      "grad_norm": 0.16023947298526764,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.3132,
+      "step": 1067
+    },
+    {
+      "epoch": 4.789237668161435,
+      "grad_norm": 0.14992086589336395,
+      "learning_rate": 2.7460117558742532e-06,
+      "loss": 0.3155,
+      "step": 1068
+    },
+    {
+      "epoch": 4.79372197309417,
+      "grad_norm": 0.157132089138031,
+      "learning_rate": 2.736030698151815e-06,
+      "loss": 0.3165,
+      "step": 1069
+    },
+    {
+      "epoch": 4.798206278026906,
+      "grad_norm": 0.16526955366134644,
+      "learning_rate": 2.72606097580412e-06,
+      "loss": 0.3159,
+      "step": 1070
+    },
+    {
+      "epoch": 4.802690582959642,
+      "grad_norm": 0.14836502075195312,
+      "learning_rate": 2.7161026387481636e-06,
+      "loss": 0.3059,
+      "step": 1071
+    },
+    {
+      "epoch": 4.807174887892376,
+      "grad_norm": 0.16198782622814178,
+      "learning_rate": 2.7061557368439294e-06,
+      "loss": 0.3177,
+      "step": 1072
+    },
+    {
+      "epoch": 4.811659192825112,
+      "grad_norm": 0.16490371525287628,
+      "learning_rate": 2.6962203198941587e-06,
+      "loss": 0.3184,
+      "step": 1073
+    },
+    {
+      "epoch": 4.816143497757848,
+      "grad_norm": 0.14605267345905304,
+      "learning_rate": 2.686296437644074e-06,
+      "loss": 0.318,
+      "step": 1074
+    },
+    {
+      "epoch": 4.820627802690583,
+      "grad_norm": 0.14928270876407623,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.3124,
+      "step": 1075
+    },
+    {
+      "epoch": 4.825112107623318,
+      "grad_norm": 0.14538247883319855,
+      "learning_rate": 2.666483475934885e-06,
+      "loss": 0.3128,
+      "step": 1076
+    },
+    {
+      "epoch": 4.829596412556054,
+      "grad_norm": 0.13825848698616028,
+      "learning_rate": 2.656594495676482e-06,
+      "loss": 0.3151,
+      "step": 1077
+    },
+    {
+      "epoch": 4.834080717488789,
+      "grad_norm": 0.1498936414718628,
+      "learning_rate": 2.6467172485186775e-06,
+      "loss": 0.3191,
+      "step": 1078
+    },
+    {
+      "epoch": 4.838565022421525,
+      "grad_norm": 0.15027591586112976,
+      "learning_rate": 2.636851783915454e-06,
+      "loss": 0.3231,
+      "step": 1079
+    },
+    {
+      "epoch": 4.84304932735426,
+      "grad_norm": 0.1482807993888855,
+      "learning_rate": 2.626998151261798e-06,
+      "loss": 0.3222,
+      "step": 1080
+    },
+    {
+      "epoch": 4.8475336322869955,
+      "grad_norm": 0.15227673947811127,
+      "learning_rate": 2.6171563998934605e-06,
+      "loss": 0.318,
+      "step": 1081
+    },
+    {
+      "epoch": 4.852017937219731,
+      "grad_norm": 0.17630086839199066,
+      "learning_rate": 2.607326579086701e-06,
+      "loss": 0.3132,
+      "step": 1082
+    },
+    {
+      "epoch": 4.856502242152466,
+      "grad_norm": 0.16713126003742218,
+      "learning_rate": 2.5975087380580445e-06,
+      "loss": 0.3134,
+      "step": 1083
+    },
+    {
+      "epoch": 4.860986547085202,
+      "grad_norm": 0.1395546793937683,
+      "learning_rate": 2.587702925964034e-06,
+      "loss": 0.3201,
+      "step": 1084
+    },
+    {
+      "epoch": 4.865470852017937,
+      "grad_norm": 0.1659357100725174,
+      "learning_rate": 2.577909191900988e-06,
+      "loss": 0.3182,
+      "step": 1085
+    },
+    {
+      "epoch": 4.869955156950673,
+      "grad_norm": 0.13949145376682281,
+      "learning_rate": 2.5681275849047482e-06,
+      "loss": 0.323,
+      "step": 1086
+    },
+    {
+      "epoch": 4.874439461883408,
+      "grad_norm": 0.1749998927116394,
+      "learning_rate": 2.5583581539504464e-06,
+      "loss": 0.3174,
+      "step": 1087
+    },
+    {
+      "epoch": 4.878923766816143,
+      "grad_norm": 0.16572751104831696,
+      "learning_rate": 2.5486009479522355e-06,
+      "loss": 0.3184,
+      "step": 1088
+    },
+    {
+      "epoch": 4.883408071748879,
+      "grad_norm": 0.18375401198863983,
+      "learning_rate": 2.5388560157630765e-06,
+      "loss": 0.3085,
+      "step": 1089
+    },
+    {
+      "epoch": 4.887892376681615,
+      "grad_norm": 0.17220883071422577,
+      "learning_rate": 2.5291234061744655e-06,
+      "loss": 0.3147,
+      "step": 1090
+    },
+    {
+      "epoch": 4.8923766816143495,
+      "grad_norm": 0.17538686096668243,
+      "learning_rate": 2.519403167916207e-06,
+      "loss": 0.3164,
+      "step": 1091
+    },
+    {
+      "epoch": 4.896860986547085,
+      "grad_norm": 0.1821119338274002,
+      "learning_rate": 2.50969534965616e-06,
+      "loss": 0.325,
+      "step": 1092
+    },
+    {
+      "epoch": 4.901345291479821,
+      "grad_norm": 0.16272936761379242,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.3201,
+      "step": 1093
+    },
+    {
+      "epoch": 4.905829596412556,
+      "grad_norm": 0.17036178708076477,
+      "learning_rate": 2.490317167490976e-06,
+      "loss": 0.3124,
+      "step": 1094
+    },
+    {
+      "epoch": 4.910313901345291,
+      "grad_norm": 0.1778683364391327,
+      "learning_rate": 2.480646900609664e-06,
+      "loss": 0.3058,
+      "step": 1095
+    },
+    {
+      "epoch": 4.914798206278027,
+      "grad_norm": 0.1492152214050293,
+      "learning_rate": 2.4709892477737263e-06,
+      "loss": 0.3207,
+      "step": 1096
+    },
+    {
+      "epoch": 4.9192825112107625,
+      "grad_norm": 0.19235937297344208,
+      "learning_rate": 2.4613442573376625e-06,
+      "loss": 0.3135,
+      "step": 1097
+    },
+    {
+      "epoch": 4.923766816143498,
+      "grad_norm": 0.20054765045642853,
+      "learning_rate": 2.4517119775925824e-06,
+      "loss": 0.3164,
+      "step": 1098
+    },
+    {
+      "epoch": 4.928251121076233,
+      "grad_norm": 0.16830937564373016,
+      "learning_rate": 2.4420924567659508e-06,
+      "loss": 0.31,
+      "step": 1099
+    },
+    {
+      "epoch": 4.932735426008969,
+      "grad_norm": 0.17535680532455444,
+      "learning_rate": 2.4324857430213504e-06,
+      "loss": 0.3215,
+      "step": 1100
+    },
+    {
+      "epoch": 4.937219730941704,
+      "grad_norm": 0.1841764897108078,
+      "learning_rate": 2.422891884458241e-06,
+      "loss": 0.3177,
+      "step": 1101
+    },
+    {
+      "epoch": 4.941704035874439,
+      "grad_norm": 0.1930045485496521,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.3167,
+      "step": 1102
+    },
+    {
+      "epoch": 4.946188340807175,
+      "grad_norm": 0.14761486649513245,
+      "learning_rate": 2.4037429249522702e-06,
+      "loss": 0.3148,
+      "step": 1103
+    },
+    {
+      "epoch": 4.95067264573991,
+      "grad_norm": 0.19395416975021362,
+      "learning_rate": 2.394187919885548e-06,
+      "loss": 0.3209,
+      "step": 1104
+    },
+    {
+      "epoch": 4.955156950672646,
+      "grad_norm": 0.17759071290493011,
+      "learning_rate": 2.384645961752113e-06,
+      "loss": 0.3238,
+      "step": 1105
+    },
+    {
+      "epoch": 4.959641255605381,
+      "grad_norm": 0.14956034719944,
+      "learning_rate": 2.3751170983272e-06,
+      "loss": 0.3145,
+      "step": 1106
+    },
+    {
+      "epoch": 4.9641255605381165,
+      "grad_norm": 0.19400863349437714,
+      "learning_rate": 2.3656013773204843e-06,
+      "loss": 0.316,
+      "step": 1107
+    },
+    {
+      "epoch": 4.968609865470852,
+      "grad_norm": 0.16733893752098083,
+      "learning_rate": 2.3560988463758366e-06,
+      "loss": 0.3146,
+      "step": 1108
+    },
+    {
+      "epoch": 4.973094170403588,
+      "grad_norm": 0.16470779478549957,
+      "learning_rate": 2.346609553071093e-06,
+      "loss": 0.3098,
+      "step": 1109
+    },
+    {
+      "epoch": 4.977578475336323,
+      "grad_norm": 0.19020842015743256,
+      "learning_rate": 2.3371335449178006e-06,
+      "loss": 0.321,
+      "step": 1110
+    },
+    {
+      "epoch": 4.982062780269058,
+      "grad_norm": 0.19639213383197784,
+      "learning_rate": 2.3276708693609947e-06,
+      "loss": 0.3072,
+      "step": 1111
+    },
+    {
+      "epoch": 4.986547085201794,
+      "grad_norm": 0.15658259391784668,
+      "learning_rate": 2.3182215737789593e-06,
+      "loss": 0.325,
+      "step": 1112
+    },
+    {
+      "epoch": 4.991031390134529,
+      "grad_norm": 0.1946086287498474,
+      "learning_rate": 2.308785705482982e-06,
+      "loss": 0.3245,
+      "step": 1113
+    },
+    {
+      "epoch": 4.995515695067264,
+      "grad_norm": 0.1552799642086029,
+      "learning_rate": 2.2993633117171243e-06,
+      "loss": 0.3158,
+      "step": 1114
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.20891344547271729,
+      "learning_rate": 2.289954439657981e-06,
+      "loss": 0.3027,
+      "step": 1115
+    },
+    {
+      "epoch": 5.004484304932736,
+      "grad_norm": 0.2603083550930023,
+      "learning_rate": 2.2805591364144446e-06,
+      "loss": 0.3004,
+      "step": 1116
+    },
+    {
+      "epoch": 5.00896860986547,
+      "grad_norm": 0.33144938945770264,
+      "learning_rate": 2.2711774490274767e-06,
+      "loss": 0.2926,
+      "step": 1117
+    },
+    {
+      "epoch": 5.013452914798206,
+      "grad_norm": 0.29076963663101196,
+      "learning_rate": 2.2618094244698614e-06,
+      "loss": 0.2946,
+      "step": 1118
+    },
+    {
+      "epoch": 5.017937219730942,
+      "grad_norm": 0.2731541097164154,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.2925,
+      "step": 1119
+    },
+    {
+      "epoch": 5.022421524663677,
+      "grad_norm": 0.27501246333122253,
+      "learning_rate": 2.243114551391542e-06,
+      "loss": 0.2928,
+      "step": 1120
+    },
+    {
+      "epoch": 5.026905829596412,
+      "grad_norm": 0.28016960620880127,
+      "learning_rate": 2.2337877964734324e-06,
+      "loss": 0.2845,
+      "step": 1121
+    },
+    {
+      "epoch": 5.031390134529148,
+      "grad_norm": 0.2584560811519623,
+      "learning_rate": 2.224474891589386e-06,
+      "loss": 0.2891,
+      "step": 1122
+    },
+    {
+      "epoch": 5.0358744394618835,
+      "grad_norm": 0.291229248046875,
+      "learning_rate": 2.2151758833678044e-06,
+      "loss": 0.293,
+      "step": 1123
+    },
+    {
+      "epoch": 5.040358744394619,
+      "grad_norm": 0.21028335392475128,
+      "learning_rate": 2.205890818367508e-06,
+      "loss": 0.2924,
+      "step": 1124
+    },
+    {
+      "epoch": 5.044843049327354,
+      "grad_norm": 0.21209950745105743,
+      "learning_rate": 2.1966197430775056e-06,
+      "loss": 0.2952,
+      "step": 1125
+    },
+    {
+      "epoch": 5.04932735426009,
+      "grad_norm": 0.2539127767086029,
+      "learning_rate": 2.187362703916766e-06,
+      "loss": 0.2947,
+      "step": 1126
+    },
+    {
+      "epoch": 5.053811659192825,
+      "grad_norm": 0.18178872764110565,
+      "learning_rate": 2.178119747233976e-06,
+      "loss": 0.2895,
+      "step": 1127
+    },
+    {
+      "epoch": 5.058295964125561,
+      "grad_norm": 0.2011086791753769,
+      "learning_rate": 2.168890919307315e-06,
+      "loss": 0.2884,
+      "step": 1128
+    },
+    {
+      "epoch": 5.062780269058296,
+      "grad_norm": 0.20836083590984344,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.2937,
+      "step": 1129
+    },
+    {
+      "epoch": 5.067264573991031,
+      "grad_norm": 0.18044930696487427,
+      "learning_rate": 2.1504758344811615e-06,
+      "loss": 0.2893,
+      "step": 1130
+    },
+    {
+      "epoch": 5.071748878923767,
+      "grad_norm": 0.17764072120189667,
+      "learning_rate": 2.141289669783401e-06,
+      "loss": 0.292,
+      "step": 1131
+    },
+    {
+      "epoch": 5.076233183856502,
+      "grad_norm": 0.2105308175086975,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.2962,
+      "step": 1132
+    },
+    {
+      "epoch": 5.0807174887892375,
+      "grad_norm": 0.1633826196193695,
+      "learning_rate": 2.122960325787432e-06,
+      "loss": 0.2962,
+      "step": 1133
+    },
+    {
+      "epoch": 5.085201793721973,
+      "grad_norm": 0.185581237077713,
+      "learning_rate": 2.1138172382616612e-06,
+      "loss": 0.2897,
+      "step": 1134
+    },
+    {
+      "epoch": 5.089686098654709,
+      "grad_norm": 0.19294404983520508,
+      "learning_rate": 2.104688601445606e-06,
+      "loss": 0.2981,
+      "step": 1135
+    },
+    {
+      "epoch": 5.0941704035874436,
+      "grad_norm": 0.1717328429222107,
+      "learning_rate": 2.0955744610450616e-06,
+      "loss": 0.3039,
+      "step": 1136
+    },
+    {
+      "epoch": 5.098654708520179,
+      "grad_norm": 0.19071757793426514,
+      "learning_rate": 2.086474862693244e-06,
+      "loss": 0.2941,
+      "step": 1137
+    },
+    {
+      "epoch": 5.103139013452915,
+      "grad_norm": 0.17715370655059814,
+      "learning_rate": 2.077389851950557e-06,
+      "loss": 0.2861,
+      "step": 1138
+    },
+    {
+      "epoch": 5.1076233183856505,
+      "grad_norm": 0.18803030252456665,
+      "learning_rate": 2.068319474304365e-06,
+      "loss": 0.2878,
+      "step": 1139
+    },
+    {
+      "epoch": 5.112107623318385,
+      "grad_norm": 0.17165349423885345,
+      "learning_rate": 2.059263775168773e-06,
+      "loss": 0.2959,
+      "step": 1140
+    },
+    {
+      "epoch": 5.116591928251121,
+      "grad_norm": 0.199208602309227,
+      "learning_rate": 2.050222799884387e-06,
+      "loss": 0.2902,
+      "step": 1141
+    },
+    {
+      "epoch": 5.121076233183857,
+      "grad_norm": 0.15911629796028137,
+      "learning_rate": 2.0411965937180907e-06,
+      "loss": 0.3003,
+      "step": 1142
+    },
+    {
+      "epoch": 5.125560538116592,
+      "grad_norm": 0.16267263889312744,
+      "learning_rate": 2.0321852018628278e-06,
+      "loss": 0.2937,
+      "step": 1143
+    },
+    {
+      "epoch": 5.130044843049327,
+      "grad_norm": 0.14909690618515015,
+      "learning_rate": 2.0231886694373653e-06,
+      "loss": 0.2986,
+      "step": 1144
+    },
+    {
+      "epoch": 5.134529147982063,
+      "grad_norm": 0.1591769903898239,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.303,
+      "step": 1145
+    },
+    {
+      "epoch": 5.139013452914798,
+      "grad_norm": 0.15179646015167236,
+      "learning_rate": 2.005240362978686e-06,
+      "loss": 0.2933,
+      "step": 1146
+    },
+    {
+      "epoch": 5.143497757847534,
+      "grad_norm": 0.16448916494846344,
+      "learning_rate": 1.996288678810105e-06,
+      "loss": 0.2896,
+      "step": 1147
+    },
+    {
+      "epoch": 5.147982062780269,
+      "grad_norm": 0.15320870280265808,
+      "learning_rate": 1.98735203380015e-06,
+      "loss": 0.2889,
+      "step": 1148
+    },
+    {
+      "epoch": 5.1524663677130045,
+      "grad_norm": 0.1648191213607788,
+      "learning_rate": 1.9784304726933384e-06,
+      "loss": 0.2919,
+      "step": 1149
+    },
+    {
+      "epoch": 5.15695067264574,
+      "grad_norm": 0.17357154190540314,
+      "learning_rate": 1.9695240401586687e-06,
+      "loss": 0.2809,
+      "step": 1150
+    },
+    {
+      "epoch": 5.161434977578475,
+      "grad_norm": 0.17532412707805634,
+      "learning_rate": 1.9606327807893905e-06,
+      "loss": 0.3006,
+      "step": 1151
+    },
+    {
+      "epoch": 5.165919282511211,
+      "grad_norm": 0.15751861035823822,
+      "learning_rate": 1.9517567391027853e-06,
+      "loss": 0.2908,
+      "step": 1152
+    },
+    {
+      "epoch": 5.170403587443946,
+      "grad_norm": 0.15868006646633148,
+      "learning_rate": 1.942895959539939e-06,
+      "loss": 0.2925,
+      "step": 1153
+    },
+    {
+      "epoch": 5.174887892376682,
+      "grad_norm": 0.20905593037605286,
+      "learning_rate": 1.934050486465529e-06,
+      "loss": 0.286,
+      "step": 1154
+    },
+    {
+      "epoch": 5.179372197309417,
+      "grad_norm": 0.1699497550725937,
+      "learning_rate": 1.9252203641675854e-06,
+      "loss": 0.2967,
+      "step": 1155
+    },
+    {
+      "epoch": 5.183856502242152,
+      "grad_norm": 0.15313996374607086,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.2955,
+      "step": 1156
+    },
+    {
+      "epoch": 5.188340807174888,
+      "grad_norm": 0.17798364162445068,
+      "learning_rate": 1.9076063486687256e-06,
+      "loss": 0.2907,
+      "step": 1157
+    },
+    {
+      "epoch": 5.192825112107624,
+      "grad_norm": 0.1971624195575714,
+      "learning_rate": 1.8988225436587005e-06,
+      "loss": 0.2831,
+      "step": 1158
+    },
+    {
+      "epoch": 5.197309417040358,
+      "grad_norm": 0.18058791756629944,
+      "learning_rate": 1.8900542658064807e-06,
+      "loss": 0.2963,
+      "step": 1159
+    },
+    {
+      "epoch": 5.201793721973094,
+      "grad_norm": 0.1599760204553604,
+      "learning_rate": 1.8813015590135963e-06,
+      "loss": 0.2867,
+      "step": 1160
+    },
+    {
+      "epoch": 5.20627802690583,
+      "grad_norm": 0.18586428463459015,
+      "learning_rate": 1.8725644671036125e-06,
+      "loss": 0.2836,
+      "step": 1161
+    },
+    {
+      "epoch": 5.210762331838565,
+      "grad_norm": 0.17721138894557953,
+      "learning_rate": 1.8638430338219199e-06,
+      "loss": 0.2989,
+      "step": 1162
+    },
+    {
+      "epoch": 5.2152466367713,
+      "grad_norm": 0.17359228432178497,
+      "learning_rate": 1.8551373028355013e-06,
+      "loss": 0.3026,
+      "step": 1163
+    },
+    {
+      "epoch": 5.219730941704036,
+      "grad_norm": 0.186914324760437,
+      "learning_rate": 1.846447317732723e-06,
+      "loss": 0.293,
+      "step": 1164
+    },
+    {
+      "epoch": 5.2242152466367715,
+      "grad_norm": 0.16008788347244263,
+      "learning_rate": 1.8377731220231144e-06,
+      "loss": 0.3087,
+      "step": 1165
+    },
+    {
+      "epoch": 5.228699551569507,
+      "grad_norm": 0.18430083990097046,
+      "learning_rate": 1.8291147591371482e-06,
+      "loss": 0.2973,
+      "step": 1166
+    },
+    {
+      "epoch": 5.233183856502242,
+      "grad_norm": 0.19583287835121155,
+      "learning_rate": 1.8204722724260266e-06,
+      "loss": 0.2962,
+      "step": 1167
+    },
+    {
+      "epoch": 5.237668161434978,
+      "grad_norm": 0.22866091132164001,
+      "learning_rate": 1.8118457051614591e-06,
+      "loss": 0.2899,
+      "step": 1168
+    },
+    {
+      "epoch": 5.242152466367713,
+      "grad_norm": 0.19682328402996063,
+      "learning_rate": 1.803235100535452e-06,
+      "loss": 0.296,
+      "step": 1169
+    },
+    {
+      "epoch": 5.246636771300448,
+      "grad_norm": 0.17415054142475128,
+      "learning_rate": 1.7946405016600843e-06,
+      "loss": 0.2961,
+      "step": 1170
+    },
+    {
+      "epoch": 5.251121076233184,
+      "grad_norm": 0.1782132238149643,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.2958,
+      "step": 1171
+    },
+    {
+      "epoch": 5.255605381165919,
+      "grad_norm": 0.16123783588409424,
+      "learning_rate": 1.7774994932086976e-06,
+      "loss": 0.2989,
+      "step": 1172
+    },
+    {
+      "epoch": 5.260089686098655,
+      "grad_norm": 0.17435431480407715,
+      "learning_rate": 1.7689531694552863e-06,
+      "loss": 0.2885,
+      "step": 1173
+    },
+    {
+      "epoch": 5.26457399103139,
+      "grad_norm": 0.16294150054454803,
+      "learning_rate": 1.7604230230973068e-06,
+      "loss": 0.2908,
+      "step": 1174
+    },
+    {
+      "epoch": 5.2690582959641254,
+      "grad_norm": 0.16784927248954773,
+      "learning_rate": 1.7519090968439966e-06,
+      "loss": 0.2904,
+      "step": 1175
+    },
+    {
+      "epoch": 5.273542600896861,
+      "grad_norm": 0.14645451307296753,
+      "learning_rate": 1.7434114333233853e-06,
+      "loss": 0.296,
+      "step": 1176
+    },
+    {
+      "epoch": 5.278026905829597,
+      "grad_norm": 0.1862608939409256,
+      "learning_rate": 1.7349300750820758e-06,
+      "loss": 0.2972,
+      "step": 1177
+    },
+    {
+      "epoch": 5.2825112107623315,
+      "grad_norm": 0.16295063495635986,
+      "learning_rate": 1.7264650645850256e-06,
+      "loss": 0.3063,
+      "step": 1178
+    },
+    {
+      "epoch": 5.286995515695067,
+      "grad_norm": 0.16478970646858215,
+      "learning_rate": 1.7180164442153529e-06,
+      "loss": 0.2884,
+      "step": 1179
+    },
+    {
+      "epoch": 5.291479820627803,
+      "grad_norm": 0.163120836019516,
+      "learning_rate": 1.709584256274106e-06,
+      "loss": 0.2998,
+      "step": 1180
+    },
+    {
+      "epoch": 5.2959641255605385,
+      "grad_norm": 0.18035657703876495,
+      "learning_rate": 1.7011685429800596e-06,
+      "loss": 0.2886,
+      "step": 1181
+    },
+    {
+      "epoch": 5.300448430493273,
+      "grad_norm": 0.16473722457885742,
+      "learning_rate": 1.6927693464695022e-06,
+      "loss": 0.2881,
+      "step": 1182
+    },
+    {
+      "epoch": 5.304932735426009,
+      "grad_norm": 0.1668347269296646,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.2918,
+      "step": 1183
+    },
+    {
+      "epoch": 5.309417040358745,
+      "grad_norm": 0.15385517477989197,
+      "learning_rate": 1.6760206719303107e-06,
+      "loss": 0.304,
+      "step": 1184
+    },
+    {
+      "epoch": 5.31390134529148,
+      "grad_norm": 0.27504652738571167,
+      "learning_rate": 1.6676712777599275e-06,
+      "loss": 0.2906,
+      "step": 1185
+    },
+    {
+      "epoch": 5.318385650224215,
+      "grad_norm": 0.1703193187713623,
+      "learning_rate": 1.6593385680891139e-06,
+      "loss": 0.2915,
+      "step": 1186
+    },
+    {
+      "epoch": 5.322869955156951,
+      "grad_norm": 0.1831894963979721,
+      "learning_rate": 1.6510225846385668e-06,
+      "loss": 0.2983,
+      "step": 1187
+    },
+    {
+      "epoch": 5.327354260089686,
+      "grad_norm": 0.16072483360767365,
+      "learning_rate": 1.6427233690452455e-06,
+      "loss": 0.2955,
+      "step": 1188
+    },
+    {
+      "epoch": 5.331838565022421,
+      "grad_norm": 0.19483248889446259,
+      "learning_rate": 1.6344409628621482e-06,
+      "loss": 0.2933,
+      "step": 1189
+    },
+    {
+      "epoch": 5.336322869955157,
+      "grad_norm": 0.16942749917507172,
+      "learning_rate": 1.6261754075581187e-06,
+      "loss": 0.2811,
+      "step": 1190
+    },
+    {
+      "epoch": 5.3408071748878925,
+      "grad_norm": 0.20586992800235748,
+      "learning_rate": 1.6179267445176206e-06,
+      "loss": 0.3037,
+      "step": 1191
+    },
+    {
+      "epoch": 5.345291479820628,
+      "grad_norm": 0.21280357241630554,
+      "learning_rate": 1.6096950150405454e-06,
+      "loss": 0.2912,
+      "step": 1192
+    },
+    {
+      "epoch": 5.349775784753363,
+      "grad_norm": 0.1741655170917511,
+      "learning_rate": 1.6014802603420044e-06,
+      "loss": 0.2986,
+      "step": 1193
+    },
+    {
+      "epoch": 5.354260089686099,
+      "grad_norm": 0.2300104945898056,
+      "learning_rate": 1.593282521552113e-06,
+      "loss": 0.2984,
+      "step": 1194
+    },
+    {
+      "epoch": 5.358744394618834,
+      "grad_norm": 0.20863893628120422,
+      "learning_rate": 1.5851018397157918e-06,
+      "loss": 0.2936,
+      "step": 1195
+    },
+    {
+      "epoch": 5.36322869955157,
+      "grad_norm": 0.18318365514278412,
+      "learning_rate": 1.5769382557925612e-06,
+      "loss": 0.2952,
+      "step": 1196
+    },
+    {
+      "epoch": 5.367713004484305,
+      "grad_norm": 0.21433617174625397,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.291,
+      "step": 1197
+    },
+    {
+      "epoch": 5.37219730941704,
+      "grad_norm": 0.2031414657831192,
+      "learning_rate": 1.5606625450952062e-06,
+      "loss": 0.2905,
+      "step": 1198
+    },
+    {
+      "epoch": 5.376681614349776,
+      "grad_norm": 0.19721800088882446,
+      "learning_rate": 1.5525504998112717e-06,
+      "loss": 0.3003,
+      "step": 1199
+    },
+    {
+      "epoch": 5.381165919282511,
+      "grad_norm": 0.20569628477096558,
+      "learning_rate": 1.5444557154203892e-06,
+      "loss": 0.2838,
+      "step": 1200
+    },
+    {
+      "epoch": 5.385650224215246,
+      "grad_norm": 0.21968922019004822,
+      "learning_rate": 1.5363782324520033e-06,
+      "loss": 0.296,
+      "step": 1201
+    },
+    {
+      "epoch": 5.390134529147982,
+      "grad_norm": 0.1700487583875656,
+      "learning_rate": 1.5283180913489326e-06,
+      "loss": 0.2867,
+      "step": 1202
+    },
+    {
+      "epoch": 5.394618834080718,
+      "grad_norm": 0.21262729167938232,
+      "learning_rate": 1.520275332467166e-06,
+      "loss": 0.2932,
+      "step": 1203
+    },
+    {
+      "epoch": 5.3991031390134525,
+      "grad_norm": 0.22031579911708832,
+      "learning_rate": 1.5122499960756604e-06,
+      "loss": 0.2913,
+      "step": 1204
+    },
+    {
+      "epoch": 5.403587443946188,
+      "grad_norm": 0.20516711473464966,
+      "learning_rate": 1.504242122356143e-06,
+      "loss": 0.2981,
+      "step": 1205
+    },
+    {
+      "epoch": 5.408071748878924,
+      "grad_norm": 0.18187974393367767,
+      "learning_rate": 1.4962517514029069e-06,
+      "loss": 0.293,
+      "step": 1206
+    },
+    {
+      "epoch": 5.4125560538116595,
+      "grad_norm": 0.20170848071575165,
+      "learning_rate": 1.4882789232226124e-06,
+      "loss": 0.2879,
+      "step": 1207
+    },
+    {
+      "epoch": 5.417040358744394,
+      "grad_norm": 0.23348811268806458,
+      "learning_rate": 1.4803236777340856e-06,
+      "loss": 0.2997,
+      "step": 1208
+    },
+    {
+      "epoch": 5.42152466367713,
+      "grad_norm": 0.21533270180225372,
+      "learning_rate": 1.4723860547681163e-06,
+      "loss": 0.2856,
+      "step": 1209
+    },
+    {
+      "epoch": 5.426008968609866,
+      "grad_norm": 0.21855808794498444,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.2862,
+      "step": 1210
+    },
+    {
+      "epoch": 5.430493273542601,
+      "grad_norm": 0.21374934911727905,
+      "learning_rate": 1.4565638352856504e-06,
+      "loss": 0.2944,
+      "step": 1211
+    },
+    {
+      "epoch": 5.434977578475336,
+      "grad_norm": 0.18907372653484344,
+      "learning_rate": 1.4486793179887738e-06,
+      "loss": 0.2892,
+      "step": 1212
+    },
+    {
+      "epoch": 5.439461883408072,
+      "grad_norm": 0.2285439670085907,
+      "learning_rate": 1.4408125816532981e-06,
+      "loss": 0.2953,
+      "step": 1213
+    },
+    {
+      "epoch": 5.443946188340807,
+      "grad_norm": 0.18237270414829254,
+      "learning_rate": 1.4329636656668617e-06,
+      "loss": 0.2874,
+      "step": 1214
+    },
+    {
+      "epoch": 5.448430493273543,
+      "grad_norm": 0.18567265570163727,
+      "learning_rate": 1.4251326093278773e-06,
+      "loss": 0.2986,
+      "step": 1215
+    },
+    {
+      "epoch": 5.452914798206278,
+      "grad_norm": 0.20102040469646454,
+      "learning_rate": 1.4173194518453415e-06,
+      "loss": 0.2971,
+      "step": 1216
+    },
+    {
+      "epoch": 5.457399103139013,
+      "grad_norm": 0.17866244912147522,
+      "learning_rate": 1.4095242323386305e-06,
+      "loss": 0.2966,
+      "step": 1217
+    },
+    {
+      "epoch": 5.461883408071749,
+      "grad_norm": 0.17779867351055145,
+      "learning_rate": 1.4017469898373077e-06,
+      "loss": 0.3003,
+      "step": 1218
+    },
+    {
+      "epoch": 5.466367713004484,
+      "grad_norm": 0.15429572761058807,
+      "learning_rate": 1.3939877632809279e-06,
+      "loss": 0.2935,
+      "step": 1219
+    },
+    {
+      "epoch": 5.4708520179372195,
+      "grad_norm": 0.20106808841228485,
+      "learning_rate": 1.3862465915188427e-06,
+      "loss": 0.2953,
+      "step": 1220
+    },
+    {
+      "epoch": 5.475336322869955,
+      "grad_norm": 0.15430030226707458,
+      "learning_rate": 1.3785235133100088e-06,
+      "loss": 0.2966,
+      "step": 1221
+    },
+    {
+      "epoch": 5.479820627802691,
+      "grad_norm": 0.17439325153827667,
+      "learning_rate": 1.3708185673227896e-06,
+      "loss": 0.2964,
+      "step": 1222
+    },
+    {
+      "epoch": 5.484304932735426,
+      "grad_norm": 0.18226686120033264,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.2978,
+      "step": 1223
+    },
+    {
+      "epoch": 5.488789237668161,
+      "grad_norm": 0.18301521241664886,
+      "learning_rate": 1.3554632262325129e-06,
+      "loss": 0.2985,
+      "step": 1224
+    },
+    {
+      "epoch": 5.493273542600897,
+      "grad_norm": 0.20829766988754272,
+      "learning_rate": 1.347812908011485e-06,
+      "loss": 0.2928,
+      "step": 1225
+    },
+    {
+      "epoch": 5.497757847533633,
+      "grad_norm": 0.1628883332014084,
+      "learning_rate": 1.340180875775735e-06,
+      "loss": 0.2956,
+      "step": 1226
+    },
+    {
+      "epoch": 5.502242152466367,
+      "grad_norm": 0.19676436483860016,
+      "learning_rate": 1.3325671677377727e-06,
+      "loss": 0.2977,
+      "step": 1227
+    },
+    {
+      "epoch": 5.506726457399103,
+      "grad_norm": 0.1734146773815155,
+      "learning_rate": 1.3249718220183583e-06,
+      "loss": 0.2925,
+      "step": 1228
+    },
+    {
+      "epoch": 5.511210762331839,
+      "grad_norm": 0.15616276860237122,
+      "learning_rate": 1.3173948766463146e-06,
+      "loss": 0.2964,
+      "step": 1229
+    },
+    {
+      "epoch": 5.515695067264574,
+      "grad_norm": 0.2004658728837967,
+      "learning_rate": 1.309836369558341e-06,
+      "loss": 0.3023,
+      "step": 1230
+    },
+    {
+      "epoch": 5.520179372197309,
+      "grad_norm": 0.16648982465267181,
+      "learning_rate": 1.3022963385988153e-06,
+      "loss": 0.2876,
+      "step": 1231
+    },
+    {
+      "epoch": 5.524663677130045,
+      "grad_norm": 0.17972449958324432,
+      "learning_rate": 1.2947748215196038e-06,
+      "loss": 0.3006,
+      "step": 1232
+    },
+    {
+      "epoch": 5.5291479820627805,
+      "grad_norm": 0.17453402280807495,
+      "learning_rate": 1.2872718559798852e-06,
+      "loss": 0.2992,
+      "step": 1233
+    },
+    {
+      "epoch": 5.533632286995516,
+      "grad_norm": 0.1862003356218338,
+      "learning_rate": 1.2797874795459464e-06,
+      "loss": 0.2904,
+      "step": 1234
+    },
+    {
+      "epoch": 5.538116591928251,
+      "grad_norm": 0.19105146825313568,
+      "learning_rate": 1.2723217296910078e-06,
+      "loss": 0.2912,
+      "step": 1235
+    },
+    {
+      "epoch": 5.542600896860987,
+      "grad_norm": 0.16527463495731354,
+      "learning_rate": 1.264874643795021e-06,
+      "loss": 0.2917,
+      "step": 1236
+    },
+    {
+      "epoch": 5.547085201793722,
+      "grad_norm": 0.18351060152053833,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.2984,
+      "step": 1237
+    },
+    {
+      "epoch": 5.551569506726457,
+      "grad_norm": 0.19402530789375305,
+      "learning_rate": 1.2500366129323039e-06,
+      "loss": 0.3032,
+      "step": 1238
+    },
+    {
+      "epoch": 5.556053811659193,
+      "grad_norm": 0.1909315437078476,
+      "learning_rate": 1.2426457422575e-06,
+      "loss": 0.2971,
+      "step": 1239
+    },
+    {
+      "epoch": 5.560538116591928,
+      "grad_norm": 0.17255248129367828,
+      "learning_rate": 1.2352736841251306e-06,
+      "loss": 0.2978,
+      "step": 1240
+    },
+    {
+      "epoch": 5.565022421524664,
+      "grad_norm": 0.20259150862693787,
+      "learning_rate": 1.2279204754460494e-06,
+      "loss": 0.2974,
+      "step": 1241
+    },
+    {
+      "epoch": 5.569506726457399,
+      "grad_norm": 0.18866488337516785,
+      "learning_rate": 1.2205861530367342e-06,
+      "loss": 0.2973,
+      "step": 1242
+    },
+    {
+      "epoch": 5.573991031390134,
+      "grad_norm": 0.1950327455997467,
+      "learning_rate": 1.2132707536191008e-06,
+      "loss": 0.2984,
+      "step": 1243
+    },
+    {
+      "epoch": 5.57847533632287,
+      "grad_norm": 0.20055194199085236,
+      "learning_rate": 1.2059743138203256e-06,
+      "loss": 0.289,
+      "step": 1244
+    },
+    {
+      "epoch": 5.582959641255606,
+      "grad_norm": 0.2066355049610138,
+      "learning_rate": 1.1986968701726492e-06,
+      "loss": 0.2957,
+      "step": 1245
+    },
+    {
+      "epoch": 5.5874439461883405,
+      "grad_norm": 0.19331777095794678,
+      "learning_rate": 1.1914384591132045e-06,
+      "loss": 0.2871,
+      "step": 1246
+    },
+    {
+      "epoch": 5.591928251121076,
+      "grad_norm": 0.21093232929706573,
+      "learning_rate": 1.184199116983834e-06,
+      "loss": 0.2922,
+      "step": 1247
+    },
+    {
+      "epoch": 5.596412556053812,
+      "grad_norm": 0.16054339706897736,
+      "learning_rate": 1.1769788800309001e-06,
+      "loss": 0.3006,
+      "step": 1248
+    },
+    {
+      "epoch": 5.600896860986547,
+      "grad_norm": 0.1877669394016266,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.2908,
+      "step": 1249
+    },
+    {
+      "epoch": 5.605381165919282,
+      "grad_norm": 0.16942396759986877,
+      "learning_rate": 1.1625958661613345e-06,
+      "loss": 0.2896,
+      "step": 1250
+    },
+    {
+      "epoch": 5.609865470852018,
+      "grad_norm": 0.18062448501586914,
+      "learning_rate": 1.1554331612584218e-06,
+      "loss": 0.2955,
+      "step": 1251
+    },
+    {
+      "epoch": 5.614349775784754,
+      "grad_norm": 0.1802215278148651,
+      "learning_rate": 1.1482897055590275e-06,
+      "loss": 0.2886,
+      "step": 1252
+    },
+    {
+      "epoch": 5.618834080717489,
+      "grad_norm": 0.16048458218574524,
+      "learning_rate": 1.141165534829425e-06,
+      "loss": 0.2887,
+      "step": 1253
+    },
+    {
+      "epoch": 5.623318385650224,
+      "grad_norm": 0.1839679479598999,
+      "learning_rate": 1.134060684739331e-06,
+      "loss": 0.2882,
+      "step": 1254
+    },
+    {
+      "epoch": 5.62780269058296,
+      "grad_norm": 0.1330212950706482,
+      "learning_rate": 1.1269751908617277e-06,
+      "loss": 0.2908,
+      "step": 1255
+    },
+    {
+      "epoch": 5.632286995515695,
+      "grad_norm": 0.19241786003112793,
+      "learning_rate": 1.119909088672682e-06,
+      "loss": 0.2919,
+      "step": 1256
+    },
+    {
+      "epoch": 5.63677130044843,
+      "grad_norm": 0.1474059522151947,
+      "learning_rate": 1.1128624135511712e-06,
+      "loss": 0.2805,
+      "step": 1257
+    },
+    {
+      "epoch": 5.641255605381166,
+      "grad_norm": 0.19253790378570557,
+      "learning_rate": 1.105835200778902e-06,
+      "loss": 0.3069,
+      "step": 1258
+    },
+    {
+      "epoch": 5.645739910313901,
+      "grad_norm": 0.1764485388994217,
+      "learning_rate": 1.0988274855401377e-06,
+      "loss": 0.2933,
+      "step": 1259
+    },
+    {
+      "epoch": 5.650224215246637,
+      "grad_norm": 0.17042991518974304,
+      "learning_rate": 1.091839302921518e-06,
+      "loss": 0.2927,
+      "step": 1260
+    },
+    {
+      "epoch": 5.654708520179372,
+      "grad_norm": 0.16894523799419403,
+      "learning_rate": 1.0848706879118893e-06,
+      "loss": 0.2847,
+      "step": 1261
+    },
+    {
+      "epoch": 5.6591928251121075,
+      "grad_norm": 0.17675840854644775,
+      "learning_rate": 1.0779216754021215e-06,
+      "loss": 0.2915,
+      "step": 1262
+    },
+    {
+      "epoch": 5.663677130044843,
+      "grad_norm": 0.15621794760227203,
+      "learning_rate": 1.070992300184941e-06,
+      "loss": 0.2898,
+      "step": 1263
+    },
+    {
+      "epoch": 5.668161434977579,
+      "grad_norm": 0.15160402655601501,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.2905,
+      "step": 1264
+    },
+    {
+      "epoch": 5.672645739910314,
+      "grad_norm": 0.16522353887557983,
+      "learning_rate": 1.057192600307456e-06,
+      "loss": 0.3029,
+      "step": 1265
+    },
+    {
+      "epoch": 5.677130044843049,
+      "grad_norm": 0.17653745412826538,
+      "learning_rate": 1.0503223447403033e-06,
+      "loss": 0.2911,
+      "step": 1266
+    },
+    {
+      "epoch": 5.681614349775785,
+      "grad_norm": 0.18189676105976105,
+      "learning_rate": 1.0434718646516917e-06,
+      "loss": 0.2949,
+      "step": 1267
+    },
+    {
+      "epoch": 5.68609865470852,
+      "grad_norm": 0.17526008188724518,
+      "learning_rate": 1.0366411943410033e-06,
+      "loss": 0.2915,
+      "step": 1268
+    },
+    {
+      "epoch": 5.690582959641255,
+      "grad_norm": 0.17741206288337708,
+      "learning_rate": 1.0298303680084448e-06,
+      "loss": 0.2951,
+      "step": 1269
+    },
+    {
+      "epoch": 5.695067264573991,
+      "grad_norm": 0.15074872970581055,
+      "learning_rate": 1.0230394197548605e-06,
+      "loss": 0.2924,
+      "step": 1270
+    },
+    {
+      "epoch": 5.699551569506727,
+      "grad_norm": 0.17078417539596558,
+      "learning_rate": 1.0162683835815706e-06,
+      "loss": 0.2933,
+      "step": 1271
+    },
+    {
+      "epoch": 5.704035874439462,
+      "grad_norm": 0.1575659066438675,
+      "learning_rate": 1.009517293390197e-06,
+      "loss": 0.2973,
+      "step": 1272
+    },
+    {
+      "epoch": 5.708520179372197,
+      "grad_norm": 0.1632554531097412,
+      "learning_rate": 1.0027861829824953e-06,
+      "loss": 0.3075,
+      "step": 1273
+    },
+    {
+      "epoch": 5.713004484304933,
+      "grad_norm": 0.192729189991951,
+      "learning_rate": 9.960750860601842e-07,
+      "loss": 0.2896,
+      "step": 1274
+    },
+    {
+      "epoch": 5.7174887892376685,
+      "grad_norm": 0.19252151250839233,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.2886,
+      "step": 1275
+    },
+    {
+      "epoch": 5.721973094170403,
+      "grad_norm": 0.2064376175403595,
+      "learning_rate": 9.82713066977427e-07,
+      "loss": 0.3001,
+      "step": 1276
+    },
+    {
+      "epoch": 5.726457399103139,
+      "grad_norm": 0.15359379351139069,
+      "learning_rate": 9.760622117187234e-07,
+      "loss": 0.3037,
+      "step": 1277
+    },
+    {
+      "epoch": 5.7309417040358746,
+      "grad_norm": 0.20079933106899261,
+      "learning_rate": 9.694315037485635e-07,
+      "loss": 0.2969,
+      "step": 1278
+    },
+    {
+      "epoch": 5.73542600896861,
+      "grad_norm": 0.15164582431316376,
+      "learning_rate": 9.628209762659658e-07,
+      "loss": 0.2945,
+      "step": 1279
+    },
+    {
+      "epoch": 5.739910313901345,
+      "grad_norm": 0.17495983839035034,
+      "learning_rate": 9.562306623689111e-07,
+      "loss": 0.297,
+      "step": 1280
+    },
+    {
+      "epoch": 5.744394618834081,
+      "grad_norm": 0.15978676080703735,
+      "learning_rate": 9.496605950541676e-07,
+      "loss": 0.3029,
+      "step": 1281
+    },
+    {
+      "epoch": 5.748878923766816,
+      "grad_norm": 0.17553512752056122,
+      "learning_rate": 9.431108072171346e-07,
+      "loss": 0.2914,
+      "step": 1282
+    },
+    {
+      "epoch": 5.753363228699552,
+      "grad_norm": 0.1696394830942154,
+      "learning_rate": 9.365813316516787e-07,
+      "loss": 0.2988,
+      "step": 1283
+    },
+    {
+      "epoch": 5.757847533632287,
+      "grad_norm": 0.17175541818141937,
+      "learning_rate": 9.300722010499608e-07,
+      "loss": 0.2991,
+      "step": 1284
+    },
+    {
+      "epoch": 5.762331838565022,
+      "grad_norm": 0.16820929944515228,
+      "learning_rate": 9.235834480022788e-07,
+      "loss": 0.2985,
+      "step": 1285
+    },
+    {
+      "epoch": 5.766816143497758,
+      "grad_norm": 0.1861841380596161,
+      "learning_rate": 9.17115104996903e-07,
+      "loss": 0.302,
+      "step": 1286
+    },
+    {
+      "epoch": 5.771300448430493,
+      "grad_norm": 0.1937403380870819,
+      "learning_rate": 9.10667204419915e-07,
+      "loss": 0.2961,
+      "step": 1287
+    },
+    {
+      "epoch": 5.7757847533632285,
+      "grad_norm": 0.13993488252162933,
+      "learning_rate": 9.042397785550405e-07,
+      "loss": 0.3026,
+      "step": 1288
+    },
+    {
+      "epoch": 5.780269058295964,
+      "grad_norm": 0.20915843546390533,
+      "learning_rate": 8.978328595834984e-07,
+      "loss": 0.2862,
+      "step": 1289
+    },
+    {
+      "epoch": 5.7847533632287,
+      "grad_norm": 0.159992977976799,
+      "learning_rate": 8.91446479583823e-07,
+      "loss": 0.2949,
+      "step": 1290
+    },
+    {
+      "epoch": 5.789237668161435,
+      "grad_norm": 0.20023779571056366,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.3022,
+      "step": 1291
+    },
+    {
+      "epoch": 5.79372197309417,
+      "grad_norm": 0.16027899086475372,
+      "learning_rate": 8.787354642998936e-07,
+      "loss": 0.2921,
+      "step": 1292
+    },
+    {
+      "epoch": 5.798206278026906,
+      "grad_norm": 0.1833188831806183,
+      "learning_rate": 8.724108926579e-07,
+      "loss": 0.3002,
+      "step": 1293
+    },
+    {
+      "epoch": 5.802690582959642,
+      "grad_norm": 0.16385091841220856,
+      "learning_rate": 8.661069872719746e-07,
+      "loss": 0.3017,
+      "step": 1294
+    },
+    {
+      "epoch": 5.807174887892376,
+      "grad_norm": 0.16897137463092804,
+      "learning_rate": 8.598237797048825e-07,
+      "loss": 0.3028,
+      "step": 1295
+    },
+    {
+      "epoch": 5.811659192825112,
+      "grad_norm": 0.1609504520893097,
+      "learning_rate": 8.535613014157556e-07,
+      "loss": 0.3041,
+      "step": 1296
+    },
+    {
+      "epoch": 5.816143497757848,
+      "grad_norm": 0.1717527061700821,
+      "learning_rate": 8.473195837599419e-07,
+      "loss": 0.3039,
+      "step": 1297
+    },
+    {
+      "epoch": 5.820627802690583,
+      "grad_norm": 0.1589074432849884,
+      "learning_rate": 8.410986579888381e-07,
+      "loss": 0.2888,
+      "step": 1298
+    },
+    {
+      "epoch": 5.825112107623318,
+      "grad_norm": 0.18116521835327148,
+      "learning_rate": 8.348985552497424e-07,
+      "loss": 0.3028,
+      "step": 1299
+    },
+    {
+      "epoch": 5.829596412556054,
+      "grad_norm": 0.17326034605503082,
+      "learning_rate": 8.287193065856936e-07,
+      "loss": 0.299,
+      "step": 1300
+    },
+    {
+      "epoch": 5.834080717488789,
+      "grad_norm": 0.1643850952386856,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.2907,
+      "step": 1301
+    },
+    {
+      "epoch": 5.838565022421525,
+      "grad_norm": 0.17930595576763153,
+      "learning_rate": 8.164234951326727e-07,
+      "loss": 0.3086,
+      "step": 1302
+    },
+    {
+      "epoch": 5.84304932735426,
+      "grad_norm": 0.1675894856452942,
+      "learning_rate": 8.103069939070945e-07,
+      "loss": 0.2972,
+      "step": 1303
+    },
+    {
+      "epoch": 5.8475336322869955,
+      "grad_norm": 0.16618181765079498,
+      "learning_rate": 8.042114698830394e-07,
+      "loss": 0.2848,
+      "step": 1304
+    },
+    {
+      "epoch": 5.852017937219731,
+      "grad_norm": 0.17379313707351685,
+      "learning_rate": 7.981369535799354e-07,
+      "loss": 0.2847,
+      "step": 1305
+    },
+    {
+      "epoch": 5.856502242152466,
+      "grad_norm": 0.1985907405614853,
+      "learning_rate": 7.920834754120305e-07,
+      "loss": 0.2876,
+      "step": 1306
+    },
+    {
+      "epoch": 5.860986547085202,
+      "grad_norm": 0.1465163379907608,
+      "learning_rate": 7.860510656882342e-07,
+      "loss": 0.2934,
+      "step": 1307
+    },
+    {
+      "epoch": 5.865470852017937,
+      "grad_norm": 0.1788966804742813,
+      "learning_rate": 7.800397546119709e-07,
+      "loss": 0.2836,
+      "step": 1308
+    },
+    {
+      "epoch": 5.869955156950673,
+      "grad_norm": 0.19033491611480713,
+      "learning_rate": 7.740495722810271e-07,
+      "loss": 0.2877,
+      "step": 1309
+    },
+    {
+      "epoch": 5.874439461883408,
+      "grad_norm": 0.18502689898014069,
+      "learning_rate": 7.680805486873977e-07,
+      "loss": 0.3009,
+      "step": 1310
+    },
+    {
+      "epoch": 5.878923766816143,
+      "grad_norm": 0.15868401527404785,
+      "learning_rate": 7.621327137171447e-07,
+      "loss": 0.2941,
+      "step": 1311
+    },
+    {
+      "epoch": 5.883408071748879,
+      "grad_norm": 0.19690321385860443,
+      "learning_rate": 7.562060971502383e-07,
+      "loss": 0.2997,
+      "step": 1312
+    },
+    {
+      "epoch": 5.887892376681615,
+      "grad_norm": 0.19460409879684448,
+      "learning_rate": 7.50300728660407e-07,
+      "loss": 0.2931,
+      "step": 1313
+    },
+    {
+      "epoch": 5.8923766816143495,
+      "grad_norm": 0.16974405944347382,
+      "learning_rate": 7.444166378150014e-07,
+      "loss": 0.297,
+      "step": 1314
+    },
+    {
+      "epoch": 5.896860986547085,
+      "grad_norm": 0.1966310441493988,
+      "learning_rate": 7.385538540748327e-07,
+      "loss": 0.2891,
+      "step": 1315
+    },
+    {
+      "epoch": 5.901345291479821,
+      "grad_norm": 0.1836744248867035,
+      "learning_rate": 7.327124067940311e-07,
+      "loss": 0.2992,
+      "step": 1316
+    },
+    {
+      "epoch": 5.905829596412556,
+      "grad_norm": 0.16646960377693176,
+      "learning_rate": 7.26892325219899e-07,
+      "loss": 0.2946,
+      "step": 1317
+    },
+    {
+      "epoch": 5.910313901345291,
+      "grad_norm": 0.1922316700220108,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.2991,
+      "step": 1318
+    },
+    {
+      "epoch": 5.914798206278027,
+      "grad_norm": 0.1692991852760315,
+      "learning_rate": 7.153163756458287e-07,
+      "loss": 0.2944,
+      "step": 1319
+    },
+    {
+      "epoch": 5.9192825112107625,
+      "grad_norm": 0.1410399228334427,
+      "learning_rate": 7.09560565605037e-07,
+      "loss": 0.2876,
+      "step": 1320
+    },
+    {
+      "epoch": 5.923766816143498,
+      "grad_norm": 0.1984073370695114,
+      "learning_rate": 7.03826237188916e-07,
+      "loss": 0.2998,
+      "step": 1321
+    },
+    {
+      "epoch": 5.928251121076233,
+      "grad_norm": 0.16254493594169617,
+      "learning_rate": 6.981134191084388e-07,
+      "loss": 0.298,
+      "step": 1322
+    },
+    {
+      "epoch": 5.932735426008969,
+      "grad_norm": 0.16346454620361328,
+      "learning_rate": 6.924221399668785e-07,
+      "loss": 0.2995,
+      "step": 1323
+    },
+    {
+      "epoch": 5.937219730941704,
+      "grad_norm": 0.15471668541431427,
+      "learning_rate": 6.867524282596655e-07,
+      "loss": 0.3005,
+      "step": 1324
+    },
+    {
+      "epoch": 5.941704035874439,
+      "grad_norm": 0.14939270913600922,
+      "learning_rate": 6.811043123742494e-07,
+      "loss": 0.3006,
+      "step": 1325
+    },
+    {
+      "epoch": 5.946188340807175,
+      "grad_norm": 0.15220074355602264,
+      "learning_rate": 6.754778205899465e-07,
+      "loss": 0.2871,
+      "step": 1326
+    },
+    {
+      "epoch": 5.95067264573991,
+      "grad_norm": 0.17014791071414948,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.2984,
+      "step": 1327
+    },
+    {
+      "epoch": 5.955156950672646,
+      "grad_norm": 0.16732892394065857,
+      "learning_rate": 6.642898219004723e-07,
+      "loss": 0.3008,
+      "step": 1328
+    },
+    {
+      "epoch": 5.959641255605381,
+      "grad_norm": 0.16991640627384186,
+      "learning_rate": 6.587283710120324e-07,
+      "loss": 0.2934,
+      "step": 1329
+    },
+    {
+      "epoch": 5.9641255605381165,
+      "grad_norm": 0.17150558531284332,
+      "learning_rate": 6.531886562578859e-07,
+      "loss": 0.2935,
+      "step": 1330
+    },
+    {
+      "epoch": 5.968609865470852,
+      "grad_norm": 0.16050690412521362,
+      "learning_rate": 6.47670705374604e-07,
+      "loss": 0.2948,
+      "step": 1331
+    },
+    {
+      "epoch": 5.973094170403588,
+      "grad_norm": 0.16353878378868103,
+      "learning_rate": 6.421745459897871e-07,
+      "loss": 0.2948,
+      "step": 1332
+    },
+    {
+      "epoch": 5.977578475336323,
+      "grad_norm": 0.15366114675998688,
+      "learning_rate": 6.367002056219285e-07,
+      "loss": 0.2922,
+      "step": 1333
+    },
+    {
+      "epoch": 5.982062780269058,
+      "grad_norm": 0.16573555767536163,
+      "learning_rate": 6.312477116802807e-07,
+      "loss": 0.296,
+      "step": 1334
+    },
+    {
+      "epoch": 5.986547085201794,
+      "grad_norm": 0.14863835275173187,
+      "learning_rate": 6.258170914647077e-07,
+      "loss": 0.2933,
+      "step": 1335
+    },
+    {
+      "epoch": 5.991031390134529,
+      "grad_norm": 0.13606324791908264,
+      "learning_rate": 6.204083721655607e-07,
+      "loss": 0.2928,
+      "step": 1336
+    },
+    {
+      "epoch": 5.995515695067264,
+      "grad_norm": 0.15118740499019623,
+      "learning_rate": 6.150215808635334e-07,
+      "loss": 0.2964,
+      "step": 1337
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.17073002457618713,
+      "learning_rate": 6.096567445295298e-07,
+      "loss": 0.2852,
+      "step": 1338
+    },
+    {
+      "epoch": 6.004484304932736,
+      "grad_norm": 0.21920576691627502,
+      "learning_rate": 6.043138900245277e-07,
+      "loss": 0.2712,
+      "step": 1339
+    },
+    {
+      "epoch": 6.00896860986547,
+      "grad_norm": 0.321831077337265,
+      "learning_rate": 5.989930440994451e-07,
+      "loss": 0.2703,
+      "step": 1340
+    },
+    {
+      "epoch": 6.013452914798206,
+      "grad_norm": 0.23883455991744995,
+      "learning_rate": 5.936942333950063e-07,
+      "loss": 0.2636,
+      "step": 1341
+    },
+    {
+      "epoch": 6.017937219730942,
+      "grad_norm": 0.24963407218456268,
+      "learning_rate": 5.884174844416102e-07,
+      "loss": 0.2783,
+      "step": 1342
+    },
+    {
+      "epoch": 6.022421524663677,
+      "grad_norm": 0.25278985500335693,
+      "learning_rate": 5.831628236591929e-07,
+      "loss": 0.2659,
+      "step": 1343
+    },
+    {
+      "epoch": 6.026905829596412,
+      "grad_norm": 0.22070910036563873,
+      "learning_rate": 5.779302773570994e-07,
+      "loss": 0.2898,
+      "step": 1344
+    },
+    {
+      "epoch": 6.031390134529148,
+      "grad_norm": 0.26762011647224426,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.262,
+      "step": 1345
+    },
+    {
+      "epoch": 6.0358744394618835,
+      "grad_norm": 0.2077677845954895,
+      "learning_rate": 5.675316328775126e-07,
+      "loss": 0.2712,
+      "step": 1346
+    },
+    {
+      "epoch": 6.040358744394619,
+      "grad_norm": 0.2524491548538208,
+      "learning_rate": 5.623655867645628e-07,
+      "loss": 0.2648,
+      "step": 1347
+    },
+    {
+      "epoch": 6.044843049327354,
+      "grad_norm": 0.22056891024112701,
+      "learning_rate": 5.572217592607687e-07,
+      "loss": 0.2665,
+      "step": 1348
+    },
+    {
+      "epoch": 6.04932735426009,
+      "grad_norm": 0.24403709173202515,
+      "learning_rate": 5.521001761205441e-07,
+      "loss": 0.2611,
+      "step": 1349
+    },
+    {
+      "epoch": 6.053811659192825,
+      "grad_norm": 0.19849088788032532,
+      "learning_rate": 5.470008629869367e-07,
+      "loss": 0.2796,
+      "step": 1350
+    },
+    {
+      "epoch": 6.058295964125561,
+      "grad_norm": 0.24553105235099792,
+      "learning_rate": 5.41923845391486e-07,
+      "loss": 0.2755,
+      "step": 1351
+    },
+    {
+      "epoch": 6.062780269058296,
+      "grad_norm": 0.22280731797218323,
+      "learning_rate": 5.368691487541027e-07,
+      "loss": 0.2718,
+      "step": 1352
+    },
+    {
+      "epoch": 6.067264573991031,
+      "grad_norm": 0.2111351639032364,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.2724,
+      "step": 1353
+    },
+    {
+      "epoch": 6.071748878923767,
+      "grad_norm": 0.19508400559425354,
+      "learning_rate": 5.268268194742638e-07,
+      "loss": 0.2634,
+      "step": 1354
+    },
+    {
+      "epoch": 6.076233183856502,
+      "grad_norm": 0.21231617033481598,
+      "learning_rate": 5.218392371123326e-07,
+      "loss": 0.2677,
+      "step": 1355
+    },
+    {
+      "epoch": 6.0807174887892375,
+      "grad_norm": 0.18558081984519958,
+      "learning_rate": 5.168740762692681e-07,
+      "loss": 0.2667,
+      "step": 1356
+    },
+    {
+      "epoch": 6.085201793721973,
+      "grad_norm": 0.196720153093338,
+      "learning_rate": 5.119313618049309e-07,
+      "loss": 0.2754,
+      "step": 1357
+    },
+    {
+      "epoch": 6.089686098654709,
+      "grad_norm": 0.2096048891544342,
+      "learning_rate": 5.070111184667908e-07,
+      "loss": 0.2763,
+      "step": 1358
+    },
+    {
+      "epoch": 6.0941704035874436,
+      "grad_norm": 0.19278304278850555,
+      "learning_rate": 5.021133708898146e-07,
+      "loss": 0.2652,
+      "step": 1359
+    },
+    {
+      "epoch": 6.098654708520179,
+      "grad_norm": 0.18683874607086182,
+      "learning_rate": 4.972381435963331e-07,
+      "loss": 0.2751,
+      "step": 1360
+    },
+    {
+      "epoch": 6.103139013452915,
+      "grad_norm": 0.18790559470653534,
+      "learning_rate": 4.9238546099592e-07,
+      "loss": 0.2651,
+      "step": 1361
+    },
+    {
+      "epoch": 6.1076233183856505,
+      "grad_norm": 0.18582111597061157,
+      "learning_rate": 4.875553473852735e-07,
+      "loss": 0.2769,
+      "step": 1362
+    },
+    {
+      "epoch": 6.112107623318385,
+      "grad_norm": 0.1966266930103302,
+      "learning_rate": 4.827478269480895e-07,
+      "loss": 0.2705,
+      "step": 1363
+    },
+    {
+      "epoch": 6.116591928251121,
+      "grad_norm": 0.18204768002033234,
+      "learning_rate": 4.779629237549438e-07,
+      "loss": 0.2734,
+      "step": 1364
+    },
+    {
+      "epoch": 6.121076233183857,
+      "grad_norm": 0.19536817073822021,
+      "learning_rate": 4.732006617631729e-07,
+      "loss": 0.273,
+      "step": 1365
+    },
+    {
+      "epoch": 6.125560538116592,
+      "grad_norm": 0.15858043730258942,
+      "learning_rate": 4.6846106481675035e-07,
+      "loss": 0.2658,
+      "step": 1366
+    },
+    {
+      "epoch": 6.130044843049327,
+      "grad_norm": 0.19480106234550476,
+      "learning_rate": 4.637441566461681e-07,
+      "loss": 0.2728,
+      "step": 1367
+    },
+    {
+      "epoch": 6.134529147982063,
+      "grad_norm": 0.18453164398670197,
+      "learning_rate": 4.590499608683202e-07,
+      "loss": 0.2636,
+      "step": 1368
+    },
+    {
+      "epoch": 6.139013452914798,
+      "grad_norm": 0.17493170499801636,
+      "learning_rate": 4.54378500986381e-07,
+      "loss": 0.2749,
+      "step": 1369
+    },
+    {
+      "epoch": 6.143497757847534,
+      "grad_norm": 0.2132866531610489,
+      "learning_rate": 4.497298003896944e-07,
+      "loss": 0.2726,
+      "step": 1370
+    },
+    {
+      "epoch": 6.147982062780269,
+      "grad_norm": 0.18004950881004333,
+      "learning_rate": 4.451038823536441e-07,
+      "loss": 0.2718,
+      "step": 1371
+    },
+    {
+      "epoch": 6.1524663677130045,
+      "grad_norm": 0.1675051748752594,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.2699,
+      "step": 1372
+    },
+    {
+      "epoch": 6.15695067264574,
+      "grad_norm": 0.17843112349510193,
+      "learning_rate": 4.35920486494546e-07,
+      "loss": 0.2766,
+      "step": 1373
+    },
+    {
+      "epoch": 6.161434977578475,
+      "grad_norm": 0.1706741452217102,
+      "learning_rate": 4.313630546514663e-07,
+      "loss": 0.2654,
+      "step": 1374
+    },
+    {
+      "epoch": 6.165919282511211,
+      "grad_norm": 0.20056675374507904,
+      "learning_rate": 4.268284973287273e-07,
+      "loss": 0.273,
+      "step": 1375
+    },
+    {
+      "epoch": 6.170403587443946,
+      "grad_norm": 0.149019256234169,
+      "learning_rate": 4.223168372302189e-07,
+      "loss": 0.2675,
+      "step": 1376
+    },
+    {
+      "epoch": 6.174887892376682,
+      "grad_norm": 0.18899179995059967,
+      "learning_rate": 4.1782809694518533e-07,
+      "loss": 0.2727,
+      "step": 1377
+    },
+    {
+      "epoch": 6.179372197309417,
+      "grad_norm": 0.1483808159828186,
+      "learning_rate": 4.1336229894811454e-07,
+      "loss": 0.2813,
+      "step": 1378
+    },
+    {
+      "epoch": 6.183856502242152,
+      "grad_norm": 0.1846124678850174,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.2749,
+      "step": 1379
+    },
+    {
+      "epoch": 6.188340807174888,
+      "grad_norm": 0.15984001755714417,
+      "learning_rate": 4.044996191413686e-07,
+      "loss": 0.2678,
+      "step": 1380
+    },
+    {
+      "epoch": 6.192825112107624,
+      "grad_norm": 0.17086835205554962,
+      "learning_rate": 4.001027817058789e-07,
+      "loss": 0.2778,
+      "step": 1381
+    },
+    {
+      "epoch": 6.197309417040358,
+      "grad_norm": 0.17133773863315582,
+      "learning_rate": 3.957289753065052e-07,
+      "loss": 0.273,
+      "step": 1382
+    },
+    {
+      "epoch": 6.201793721973094,
+      "grad_norm": 0.17212380468845367,
+      "learning_rate": 3.9137822184227845e-07,
+      "loss": 0.2681,
+      "step": 1383
+    },
+    {
+      "epoch": 6.20627802690583,
+      "grad_norm": 0.1615939736366272,
+      "learning_rate": 3.870505430968069e-07,
+      "loss": 0.2728,
+      "step": 1384
+    },
+    {
+      "epoch": 6.210762331838565,
+      "grad_norm": 0.1794346570968628,
+      "learning_rate": 3.8274596073816784e-07,
+      "loss": 0.2833,
+      "step": 1385
+    },
+    {
+      "epoch": 6.2152466367713,
+      "grad_norm": 0.15747982263565063,
+      "learning_rate": 3.7846449631879664e-07,
+      "loss": 0.2742,
+      "step": 1386
+    },
+    {
+      "epoch": 6.219730941704036,
+      "grad_norm": 0.18398547172546387,
+      "learning_rate": 3.742061712753825e-07,
+      "loss": 0.2715,
+      "step": 1387
+    },
+    {
+      "epoch": 6.2242152466367715,
+      "grad_norm": 0.17561475932598114,
+      "learning_rate": 3.699710069287571e-07,
+      "loss": 0.2655,
+      "step": 1388
+    },
+    {
+      "epoch": 6.228699551569507,
+      "grad_norm": 0.1716562956571579,
+      "learning_rate": 3.657590244837911e-07,
+      "loss": 0.2772,
+      "step": 1389
+    },
+    {
+      "epoch": 6.233183856502242,
+      "grad_norm": 0.14322857558727264,
+      "learning_rate": 3.615702450292857e-07,
+      "loss": 0.2656,
+      "step": 1390
+    },
+    {
+      "epoch": 6.237668161434978,
+      "grad_norm": 0.18851330876350403,
+      "learning_rate": 3.5740468953786854e-07,
+      "loss": 0.2653,
+      "step": 1391
+    },
+    {
+      "epoch": 6.242152466367713,
+      "grad_norm": 0.14088350534439087,
+      "learning_rate": 3.5326237886588734e-07,
+      "loss": 0.2724,
+      "step": 1392
+    },
+    {
+      "epoch": 6.246636771300448,
+      "grad_norm": 0.17594043910503387,
+      "learning_rate": 3.49143333753309e-07,
+      "loss": 0.273,
+      "step": 1393
+    },
+    {
+      "epoch": 6.251121076233184,
+      "grad_norm": 0.1615719050168991,
+      "learning_rate": 3.4504757482360817e-07,
+      "loss": 0.2781,
+      "step": 1394
+    },
+    {
+      "epoch": 6.255605381165919,
+      "grad_norm": 0.16274654865264893,
+      "learning_rate": 3.4097512258367385e-07,
+      "loss": 0.2747,
+      "step": 1395
+    },
+    {
+      "epoch": 6.260089686098655,
+      "grad_norm": 0.1752977967262268,
+      "learning_rate": 3.369259974236988e-07,
+      "loss": 0.275,
+      "step": 1396
+    },
+    {
+      "epoch": 6.26457399103139,
+      "grad_norm": 0.1652955263853073,
+      "learning_rate": 3.3290021961708163e-07,
+      "loss": 0.2731,
+      "step": 1397
+    },
+    {
+      "epoch": 6.2690582959641254,
+      "grad_norm": 0.18381209671497345,
+      "learning_rate": 3.288978093203227e-07,
+      "loss": 0.2754,
+      "step": 1398
+    },
+    {
+      "epoch": 6.273542600896861,
+      "grad_norm": 0.1642630249261856,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.2758,
+      "step": 1399
+    },
+    {
+      "epoch": 6.278026905829597,
+      "grad_norm": 0.17316223680973053,
+      "learning_rate": 3.209631712972966e-07,
+      "loss": 0.2735,
+      "step": 1400
+    },
+    {
+      "epoch": 6.2825112107623315,
+      "grad_norm": 0.1626156121492386,
+      "learning_rate": 3.1703098329864237e-07,
+      "loss": 0.2627,
+      "step": 1401
+    },
+    {
+      "epoch": 6.286995515695067,
+      "grad_norm": 0.17561137676239014,
+      "learning_rate": 3.131222422648744e-07,
+      "loss": 0.2837,
+      "step": 1402
+    },
+    {
+      "epoch": 6.291479820627803,
+      "grad_norm": 0.16027145087718964,
+      "learning_rate": 3.0923696776650414e-07,
+      "loss": 0.2723,
+      "step": 1403
+    },
+    {
+      "epoch": 6.2959641255605385,
+      "grad_norm": 0.19332273304462433,
+      "learning_rate": 3.0537517925655567e-07,
+      "loss": 0.2701,
+      "step": 1404
+    },
+    {
+      "epoch": 6.300448430493273,
+      "grad_norm": 0.18537260591983795,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.2695,
+      "step": 1405
+    },
+    {
+      "epoch": 6.304932735426009,
+      "grad_norm": 0.1527324765920639,
+      "learning_rate": 2.9772213742595367e-07,
+      "loss": 0.2733,
+      "step": 1406
+    },
+    {
+      "epoch": 6.309417040358745,
+      "grad_norm": 0.19918321073055267,
+      "learning_rate": 2.9393092242300026e-07,
+      "loss": 0.2593,
+      "step": 1407
+    },
+    {
+      "epoch": 6.31390134529148,
+      "grad_norm": 0.16420800983905792,
+      "learning_rate": 2.901632700436757e-07,
+      "loss": 0.2794,
+      "step": 1408
+    },
+    {
+      "epoch": 6.318385650224215,
+      "grad_norm": 0.15757115185260773,
+      "learning_rate": 2.864191991520848e-07,
+      "loss": 0.2681,
+      "step": 1409
+    },
+    {
+      "epoch": 6.322869955156951,
+      "grad_norm": 0.1807360202074051,
+      "learning_rate": 2.8269872849426114e-07,
+      "loss": 0.2786,
+      "step": 1410
+    },
+    {
+      "epoch": 6.327354260089686,
+      "grad_norm": 0.162303626537323,
+      "learning_rate": 2.790018766980773e-07,
+      "loss": 0.2786,
+      "step": 1411
+    },
+    {
+      "epoch": 6.331838565022421,
+      "grad_norm": 0.18337154388427734,
+      "learning_rate": 2.7532866227314714e-07,
+      "loss": 0.2705,
+      "step": 1412
+    },
+    {
+      "epoch": 6.336322869955157,
+      "grad_norm": 0.16891908645629883,
+      "learning_rate": 2.71679103610738e-07,
+      "loss": 0.2745,
+      "step": 1413
+    },
+    {
+      "epoch": 6.3408071748878925,
+      "grad_norm": 0.18970470130443573,
+      "learning_rate": 2.6805321898367323e-07,
+      "loss": 0.2827,
+      "step": 1414
+    },
+    {
+      "epoch": 6.345291479820628,
+      "grad_norm": 0.1631837785243988,
+      "learning_rate": 2.64451026546248e-07,
+      "loss": 0.2732,
+      "step": 1415
+    },
+    {
+      "epoch": 6.349775784753363,
+      "grad_norm": 0.21060024201869965,
+      "learning_rate": 2.6087254433412924e-07,
+      "loss": 0.2839,
+      "step": 1416
+    },
+    {
+      "epoch": 6.354260089686099,
+      "grad_norm": 0.16917921602725983,
+      "learning_rate": 2.573177902642726e-07,
+      "loss": 0.279,
+      "step": 1417
+    },
+    {
+      "epoch": 6.358744394618834,
+      "grad_norm": 0.17150048911571503,
+      "learning_rate": 2.5378678213483057e-07,
+      "loss": 0.2786,
+      "step": 1418
+    },
+    {
+      "epoch": 6.36322869955157,
+      "grad_norm": 0.17799870669841766,
+      "learning_rate": 2.502795376250622e-07,
+      "loss": 0.2691,
+      "step": 1419
+    },
+    {
+      "epoch": 6.367713004484305,
+      "grad_norm": 0.17664732038974762,
+      "learning_rate": 2.467960742952463e-07,
+      "loss": 0.276,
+      "step": 1420
+    },
+    {
+      "epoch": 6.37219730941704,
+      "grad_norm": 0.1755521297454834,
+      "learning_rate": 2.4333640958659144e-07,
+      "loss": 0.2759,
+      "step": 1421
+    },
+    {
+      "epoch": 6.376681614349776,
+      "grad_norm": 0.15604710578918457,
+      "learning_rate": 2.399005608211502e-07,
+      "loss": 0.2731,
+      "step": 1422
+    },
+    {
+      "epoch": 6.381165919282511,
+      "grad_norm": 0.14964698255062103,
+      "learning_rate": 2.3648854520173237e-07,
+      "loss": 0.2738,
+      "step": 1423
+    },
+    {
+      "epoch": 6.385650224215246,
+      "grad_norm": 0.17770838737487793,
+      "learning_rate": 2.3310037981182088e-07,
+      "loss": 0.2641,
+      "step": 1424
+    },
+    {
+      "epoch": 6.390134529147982,
+      "grad_norm": 0.15091992914676666,
+      "learning_rate": 2.2973608161547755e-07,
+      "loss": 0.2724,
+      "step": 1425
+    },
+    {
+      "epoch": 6.394618834080718,
+      "grad_norm": 0.1676618456840515,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.2682,
+      "step": 1426
+    },
+    {
+      "epoch": 6.3991031390134525,
+      "grad_norm": 0.17938712239265442,
+      "learning_rate": 2.2307915406218517e-07,
+      "loss": 0.267,
+      "step": 1427
+    },
+    {
+      "epoch": 6.403587443946188,
+      "grad_norm": 0.16242460906505585,
+      "learning_rate": 2.1978655803553128e-07,
+      "loss": 0.2771,
+      "step": 1428
+    },
+    {
+      "epoch": 6.408071748878924,
+      "grad_norm": 0.19494880735874176,
+      "learning_rate": 2.1651789586287442e-07,
+      "loss": 0.2688,
+      "step": 1429
+    },
+    {
+      "epoch": 6.4125560538116595,
+      "grad_norm": 0.16317808628082275,
+      "learning_rate": 2.1327318390994445e-07,
+      "loss": 0.274,
+      "step": 1430
+    },
+    {
+      "epoch": 6.417040358744394,
+      "grad_norm": 0.16894695162773132,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.2769,
+      "step": 1431
+    },
+    {
+      "epoch": 6.42152466367713,
+      "grad_norm": 0.15520493686199188,
+      "learning_rate": 2.068556755265272e-07,
+      "loss": 0.2715,
+      "step": 1432
+    },
+    {
+      "epoch": 6.426008968609866,
+      "grad_norm": 0.18025784194469452,
+      "learning_rate": 2.0368291122759898e-07,
+      "loss": 0.2711,
+      "step": 1433
+    },
+    {
+      "epoch": 6.430493273542601,
+      "grad_norm": 0.15909095108509064,
+      "learning_rate": 2.005341614113543e-07,
+      "loss": 0.2686,
+      "step": 1434
+    },
+    {
+      "epoch": 6.434977578475336,
+      "grad_norm": 0.1772451549768448,
+      "learning_rate": 1.9740944184313882e-07,
+      "loss": 0.2691,
+      "step": 1435
+    },
+    {
+      "epoch": 6.439461883408072,
+      "grad_norm": 0.16724106669425964,
+      "learning_rate": 1.9430876816798228e-07,
+      "loss": 0.2758,
+      "step": 1436
+    },
+    {
+      "epoch": 6.443946188340807,
+      "grad_norm": 0.18545377254486084,
+      "learning_rate": 1.9123215591052014e-07,
+      "loss": 0.268,
+      "step": 1437
+    },
+    {
+      "epoch": 6.448430493273543,
+      "grad_norm": 0.17424210906028748,
+      "learning_rate": 1.88179620474917e-07,
+      "loss": 0.2652,
+      "step": 1438
+    },
+    {
+      "epoch": 6.452914798206278,
+      "grad_norm": 0.18465006351470947,
+      "learning_rate": 1.8515117714478447e-07,
+      "loss": 0.2634,
+      "step": 1439
+    },
+    {
+      "epoch": 6.457399103139013,
+      "grad_norm": 0.2112896889448166,
+      "learning_rate": 1.8214684108311286e-07,
+      "loss": 0.2711,
+      "step": 1440
+    },
+    {
+      "epoch": 6.461883408071749,
+      "grad_norm": 0.1514737755060196,
+      "learning_rate": 1.7916662733218848e-07,
+      "loss": 0.2627,
+      "step": 1441
+    },
+    {
+      "epoch": 6.466367713004484,
+      "grad_norm": 0.2109755128622055,
+      "learning_rate": 1.762105508135198e-07,
+      "loss": 0.2787,
+      "step": 1442
+    },
+    {
+      "epoch": 6.4708520179372195,
+      "grad_norm": 0.18746790289878845,
+      "learning_rate": 1.7327862632776638e-07,
+      "loss": 0.267,
+      "step": 1443
+    },
+    {
+      "epoch": 6.475336322869955,
+      "grad_norm": 0.1652342975139618,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.274,
+      "step": 1444
+    },
+    {
+      "epoch": 6.479820627802691,
+      "grad_norm": 0.17727355659008026,
+      "learning_rate": 1.6748729205293024e-07,
+      "loss": 0.2714,
+      "step": 1445
+    },
+    {
+      "epoch": 6.484304932735426,
+      "grad_norm": 0.18117089569568634,
+      "learning_rate": 1.646279112602417e-07,
+      "loss": 0.2797,
+      "step": 1446
+    },
+    {
+      "epoch": 6.488789237668161,
+      "grad_norm": 0.17036432027816772,
+      "learning_rate": 1.6179274049310966e-07,
+      "loss": 0.264,
+      "step": 1447
+    },
+    {
+      "epoch": 6.493273542600897,
+      "grad_norm": 0.1743333488702774,
+      "learning_rate": 1.5898179394683244e-07,
+      "loss": 0.2731,
+      "step": 1448
+    },
+    {
+      "epoch": 6.497757847533633,
+      "grad_norm": 0.16794371604919434,
+      "learning_rate": 1.5619508569542363e-07,
+      "loss": 0.2786,
+      "step": 1449
+    },
+    {
+      "epoch": 6.502242152466367,
+      "grad_norm": 0.1611374020576477,
+      "learning_rate": 1.5343262969153781e-07,
+      "loss": 0.2843,
+      "step": 1450
+    },
+    {
+      "epoch": 6.506726457399103,
+      "grad_norm": 0.17746052145957947,
+      "learning_rate": 1.5069443976640287e-07,
+      "loss": 0.2797,
+      "step": 1451
+    },
+    {
+      "epoch": 6.511210762331839,
+      "grad_norm": 0.16705839335918427,
+      "learning_rate": 1.4798052962974874e-07,
+      "loss": 0.2693,
+      "step": 1452
+    },
+    {
+      "epoch": 6.515695067264574,
+      "grad_norm": 0.17691704630851746,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.2714,
+      "step": 1453
+    },
+    {
+      "epoch": 6.520179372197309,
+      "grad_norm": 0.16551019251346588,
+      "learning_rate": 1.4262560295290884e-07,
+      "loss": 0.2782,
+      "step": 1454
+    },
+    {
+      "epoch": 6.524663677130045,
+      "grad_norm": 0.1659432202577591,
+      "learning_rate": 1.3998461322408563e-07,
+      "loss": 0.2872,
+      "step": 1455
+    },
+    {
+      "epoch": 6.5291479820627805,
+      "grad_norm": 0.16633793711662292,
+      "learning_rate": 1.3736795690633353e-07,
+      "loss": 0.2728,
+      "step": 1456
+    },
+    {
+      "epoch": 6.533632286995516,
+      "grad_norm": 0.18562401831150055,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.2659,
+      "step": 1457
+    },
+    {
+      "epoch": 6.538116591928251,
+      "grad_norm": 0.1541408896446228,
+      "learning_rate": 1.3220769678705724e-07,
+      "loss": 0.2781,
+      "step": 1458
+    },
+    {
+      "epoch": 6.542600896860987,
+      "grad_norm": 0.16415201127529144,
+      "learning_rate": 1.2966411882222695e-07,
+      "loss": 0.2733,
+      "step": 1459
+    },
+    {
+      "epoch": 6.547085201793722,
+      "grad_norm": 0.1674622744321823,
+      "learning_rate": 1.271449259417268e-07,
+      "loss": 0.2712,
+      "step": 1460
+    },
+    {
+      "epoch": 6.551569506726457,
+      "grad_norm": 0.16234423220157623,
+      "learning_rate": 1.2465013075879884e-07,
+      "loss": 0.2728,
+      "step": 1461
+    },
+    {
+      "epoch": 6.556053811659193,
+      "grad_norm": 0.1773276925086975,
+      "learning_rate": 1.2217974576453072e-07,
+      "loss": 0.2756,
+      "step": 1462
+    },
+    {
+      "epoch": 6.560538116591928,
+      "grad_norm": 0.1632678210735321,
+      "learning_rate": 1.1973378332779229e-07,
+      "loss": 0.2732,
+      "step": 1463
+    },
+    {
+      "epoch": 6.565022421524664,
+      "grad_norm": 0.1739519238471985,
+      "learning_rate": 1.1731225569517113e-07,
+      "loss": 0.2753,
+      "step": 1464
+    },
+    {
+      "epoch": 6.569506726457399,
+      "grad_norm": 0.15234674513339996,
+      "learning_rate": 1.1491517499091498e-07,
+      "loss": 0.2733,
+      "step": 1465
+    },
+    {
+      "epoch": 6.573991031390134,
+      "grad_norm": 0.16561099886894226,
+      "learning_rate": 1.1254255321686836e-07,
+      "loss": 0.2748,
+      "step": 1466
+    },
+    {
+      "epoch": 6.57847533632287,
+      "grad_norm": 0.16730961203575134,
+      "learning_rate": 1.1019440225241317e-07,
+      "loss": 0.2715,
+      "step": 1467
+    },
+    {
+      "epoch": 6.582959641255606,
+      "grad_norm": 0.17169523239135742,
+      "learning_rate": 1.078707338544105e-07,
+      "loss": 0.2722,
+      "step": 1468
+    },
+    {
+      "epoch": 6.5874439461883405,
+      "grad_norm": 0.1718924641609192,
+      "learning_rate": 1.055715596571405e-07,
+      "loss": 0.277,
+      "step": 1469
+    },
+    {
+      "epoch": 6.591928251121076,
+      "grad_norm": 0.16631938517093658,
+      "learning_rate": 1.0329689117224262e-07,
+      "loss": 0.2747,
+      "step": 1470
+    },
+    {
+      "epoch": 6.596412556053812,
+      "grad_norm": 0.17209939658641815,
+      "learning_rate": 1.0104673978866164e-07,
+      "loss": 0.268,
+      "step": 1471
+    },
+    {
+      "epoch": 6.600896860986547,
+      "grad_norm": 0.17010393738746643,
+      "learning_rate": 9.882111677258777e-08,
+      "loss": 0.2747,
+      "step": 1472
+    },
+    {
+      "epoch": 6.605381165919282,
+      "grad_norm": 0.18056197464466095,
+      "learning_rate": 9.662003326740166e-08,
+      "loss": 0.2662,
+      "step": 1473
+    },
+    {
+      "epoch": 6.609865470852018,
+      "grad_norm": 0.1678396612405777,
+      "learning_rate": 9.444350029361671e-08,
+      "loss": 0.2655,
+      "step": 1474
+    },
+    {
+      "epoch": 6.614349775784754,
+      "grad_norm": 0.19450931251049042,
+      "learning_rate": 9.22915287488274e-08,
+      "loss": 0.2723,
+      "step": 1475
+    },
+    {
+      "epoch": 6.618834080717489,
+      "grad_norm": 0.21749603748321533,
+      "learning_rate": 9.016412940765107e-08,
+      "loss": 0.2633,
+      "step": 1476
+    },
+    {
+      "epoch": 6.623318385650224,
+      "grad_norm": 0.17477911710739136,
+      "learning_rate": 8.80613129216762e-08,
+      "loss": 0.2864,
+      "step": 1477
+    },
+    {
+      "epoch": 6.62780269058296,
+      "grad_norm": 0.1605025827884674,
+      "learning_rate": 8.598308981940751e-08,
+      "loss": 0.2817,
+      "step": 1478
+    },
+    {
+      "epoch": 6.632286995515695,
+      "grad_norm": 0.15158961713314056,
+      "learning_rate": 8.392947050621603e-08,
+      "loss": 0.275,
+      "step": 1479
+    },
+    {
+      "epoch": 6.63677130044843,
+      "grad_norm": 0.1628490537405014,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.2649,
+      "step": 1480
+    },
+    {
+      "epoch": 6.641255605381166,
+      "grad_norm": 0.15091273188591003,
+      "learning_rate": 7.989608425254924e-08,
+      "loss": 0.2778,
+      "step": 1481
+    },
+    {
+      "epoch": 6.645739910313901,
+      "grad_norm": 0.1478932797908783,
+      "learning_rate": 7.791633750667105e-08,
+      "loss": 0.2855,
+      "step": 1482
+    },
+    {
+      "epoch": 6.650224215246637,
+      "grad_norm": 0.17141316831111908,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.2737,
+      "step": 1483
+    },
+    {
+      "epoch": 6.654708520179372,
+      "grad_norm": 0.1544095277786255,
+      "learning_rate": 7.403078633833716e-08,
+      "loss": 0.278,
+      "step": 1484
+    },
+    {
+      "epoch": 6.6591928251121075,
+      "grad_norm": 0.16661569476127625,
+      "learning_rate": 7.212500137028789e-08,
+      "loss": 0.2766,
+      "step": 1485
+    },
+    {
+      "epoch": 6.663677130044843,
+      "grad_norm": 0.18654701113700867,
+      "learning_rate": 7.024388957680705e-08,
+      "loss": 0.2801,
+      "step": 1486
+    },
+    {
+      "epoch": 6.668161434977579,
+      "grad_norm": 0.16779200732707977,
+      "learning_rate": 6.838746037635735e-08,
+      "loss": 0.2859,
+      "step": 1487
+    },
+    {
+      "epoch": 6.672645739910314,
+      "grad_norm": 0.13596783578395844,
+      "learning_rate": 6.655572306381696e-08,
+      "loss": 0.2772,
+      "step": 1488
+    },
+    {
+      "epoch": 6.677130044843049,
+      "grad_norm": 0.1895265281200409,
+      "learning_rate": 6.474868681043578e-08,
+      "loss": 0.2672,
+      "step": 1489
+    },
+    {
+      "epoch": 6.681614349775785,
+      "grad_norm": 0.15377727150917053,
+      "learning_rate": 6.29663606637898e-08,
+      "loss": 0.2698,
+      "step": 1490
+    },
+    {
+      "epoch": 6.68609865470852,
+      "grad_norm": 0.15751516819000244,
+      "learning_rate": 6.120875354773459e-08,
+      "loss": 0.2765,
+      "step": 1491
+    },
+    {
+      "epoch": 6.690582959641255,
+      "grad_norm": 0.17083840072155,
+      "learning_rate": 5.947587426236079e-08,
+      "loss": 0.2681,
+      "step": 1492
+    },
+    {
+      "epoch": 6.695067264573991,
+      "grad_norm": 0.14098629355430603,
+      "learning_rate": 5.776773148394976e-08,
+      "loss": 0.2752,
+      "step": 1493
+    },
+    {
+      "epoch": 6.699551569506727,
+      "grad_norm": 0.16371020674705505,
+      "learning_rate": 5.608433376493194e-08,
+      "loss": 0.2752,
+      "step": 1494
+    },
+    {
+      "epoch": 6.704035874439462,
+      "grad_norm": 0.16219350695610046,
+      "learning_rate": 5.4425689533841864e-08,
+      "loss": 0.2689,
+      "step": 1495
+    },
+    {
+      "epoch": 6.708520179372197,
+      "grad_norm": 0.15417857468128204,
+      "learning_rate": 5.279180709527765e-08,
+      "loss": 0.2792,
+      "step": 1496
+    },
+    {
+      "epoch": 6.713004484304933,
+      "grad_norm": 0.17875702679157257,
+      "learning_rate": 5.1182694629857145e-08,
+      "loss": 0.2613,
+      "step": 1497
+    },
+    {
+      "epoch": 6.7174887892376685,
+      "grad_norm": 0.17640921473503113,
+      "learning_rate": 4.959836019417963e-08,
+      "loss": 0.2755,
+      "step": 1498
+    },
+    {
+      "epoch": 6.721973094170403,
+      "grad_norm": 0.16516731679439545,
+      "learning_rate": 4.803881172078473e-08,
+      "loss": 0.2791,
+      "step": 1499
+    },
+    {
+      "epoch": 6.726457399103139,
+      "grad_norm": 0.17578543722629547,
+      "learning_rate": 4.650405701811078e-08,
+      "loss": 0.273,
+      "step": 1500
+    },
+    {
+      "epoch": 6.7309417040358746,
+      "grad_norm": 0.1459975242614746,
+      "learning_rate": 4.499410377045765e-08,
+      "loss": 0.2831,
+      "step": 1501
+    },
+    {
+      "epoch": 6.73542600896861,
+      "grad_norm": 0.149170383810997,
+      "learning_rate": 4.350895953794898e-08,
+      "loss": 0.2733,
+      "step": 1502
+    },
+    {
+      "epoch": 6.739910313901345,
+      "grad_norm": 0.1679002195596695,
+      "learning_rate": 4.2048631756492206e-08,
+      "loss": 0.2823,
+      "step": 1503
+    },
+    {
+      "epoch": 6.744394618834081,
+      "grad_norm": 0.14103658497333527,
+      "learning_rate": 4.0613127737741396e-08,
+      "loss": 0.2762,
+      "step": 1504
+    },
+    {
+      "epoch": 6.748878923766816,
+      "grad_norm": 0.16264501214027405,
+      "learning_rate": 3.9202454669063915e-08,
+      "loss": 0.2733,
+      "step": 1505
+    },
+    {
+      "epoch": 6.753363228699552,
+      "grad_norm": 0.1775323450565338,
+      "learning_rate": 3.781661961349992e-08,
+      "loss": 0.279,
+      "step": 1506
+    },
+    {
+      "epoch": 6.757847533632287,
+      "grad_norm": 0.17399895191192627,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.2749,
+      "step": 1507
+    },
+    {
+      "epoch": 6.762331838565022,
+      "grad_norm": 0.15998288989067078,
+      "learning_rate": 3.5119491172039836e-08,
+      "loss": 0.2717,
+      "step": 1508
+    },
+    {
+      "epoch": 6.766816143497758,
+      "grad_norm": 0.14236798882484436,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.275,
+      "step": 1509
+    },
+    {
+      "epoch": 6.771300448430493,
+      "grad_norm": 0.14472728967666626,
+      "learning_rate": 3.252179642985909e-08,
+      "loss": 0.2736,
+      "step": 1510
+    },
+    {
+      "epoch": 6.7757847533632285,
+      "grad_norm": 0.15147371590137482,
+      "learning_rate": 3.1260253031660247e-08,
+      "loss": 0.2773,
+      "step": 1511
+    },
+    {
+      "epoch": 6.780269058295964,
+      "grad_norm": 0.13943013548851013,
+      "learning_rate": 3.002358741205691e-08,
+      "loss": 0.2684,
+      "step": 1512
+    },
+    {
+      "epoch": 6.7847533632287,
+      "grad_norm": 0.1565491259098053,
+      "learning_rate": 2.8811805762860578e-08,
+      "loss": 0.2776,
+      "step": 1513
+    },
+    {
+      "epoch": 6.789237668161435,
+      "grad_norm": 0.15105099976062775,
+      "learning_rate": 2.762491415128965e-08,
+      "loss": 0.2584,
+      "step": 1514
+    },
+    {
+      "epoch": 6.79372197309417,
+      "grad_norm": 0.1686726063489914,
+      "learning_rate": 2.6462918519944425e-08,
+      "loss": 0.2732,
+      "step": 1515
+    },
+    {
+      "epoch": 6.798206278026906,
+      "grad_norm": 0.1443524807691574,
+      "learning_rate": 2.5325824686772138e-08,
+      "loss": 0.2737,
+      "step": 1516
+    },
+    {
+      "epoch": 6.802690582959642,
+      "grad_norm": 0.16183225810527802,
+      "learning_rate": 2.4213638345040868e-08,
+      "loss": 0.2662,
+      "step": 1517
+    },
+    {
+      "epoch": 6.807174887892376,
+      "grad_norm": 0.16884076595306396,
+      "learning_rate": 2.3126365063311228e-08,
+      "loss": 0.2805,
+      "step": 1518
+    },
+    {
+      "epoch": 6.811659192825112,
+      "grad_norm": 0.15632623434066772,
+      "learning_rate": 2.206401028540639e-08,
+      "loss": 0.2745,
+      "step": 1519
+    },
+    {
+      "epoch": 6.816143497757848,
+      "grad_norm": 0.18027900159358978,
+      "learning_rate": 2.1026579330387655e-08,
+      "loss": 0.2709,
+      "step": 1520
+    },
+    {
+      "epoch": 6.820627802690583,
+      "grad_norm": 0.20842447876930237,
+      "learning_rate": 2.0014077392525035e-08,
+      "loss": 0.2823,
+      "step": 1521
+    },
+    {
+      "epoch": 6.825112107623318,
+      "grad_norm": 0.15945984423160553,
+      "learning_rate": 1.9026509541272276e-08,
+      "loss": 0.2702,
+      "step": 1522
+    },
+    {
+      "epoch": 6.829596412556054,
+      "grad_norm": 0.18381254374980927,
+      "learning_rate": 1.8063880721242986e-08,
+      "loss": 0.2736,
+      "step": 1523
+    },
+    {
+      "epoch": 6.834080717488789,
+      "grad_norm": 0.22747130692005157,
+      "learning_rate": 1.712619575218344e-08,
+      "loss": 0.2759,
+      "step": 1524
+    },
+    {
+      "epoch": 6.838565022421525,
+      "grad_norm": 0.17451879382133484,
+      "learning_rate": 1.6213459328950355e-08,
+      "loss": 0.2739,
+      "step": 1525
+    },
+    {
+      "epoch": 6.84304932735426,
+      "grad_norm": 0.19969181716442108,
+      "learning_rate": 1.5325676021484826e-08,
+      "loss": 0.2737,
+      "step": 1526
+    },
+    {
+      "epoch": 6.8475336322869955,
+      "grad_norm": 0.22152204811573029,
+      "learning_rate": 1.4462850274794548e-08,
+      "loss": 0.2821,
+      "step": 1527
+    },
+    {
+      "epoch": 6.852017937219731,
+      "grad_norm": 0.2019326239824295,
+      "learning_rate": 1.3624986408924956e-08,
+      "loss": 0.2862,
+      "step": 1528
+    },
+    {
+      "epoch": 6.856502242152466,
+      "grad_norm": 0.173894464969635,
+      "learning_rate": 1.2812088618942009e-08,
+      "loss": 0.2766,
+      "step": 1529
+    },
+    {
+      "epoch": 6.860986547085202,
+      "grad_norm": 0.2000262290239334,
+      "learning_rate": 1.2024160974911103e-08,
+      "loss": 0.2835,
+      "step": 1530
+    },
+    {
+      "epoch": 6.865470852017937,
+      "grad_norm": 0.17501193284988403,
+      "learning_rate": 1.1261207421874309e-08,
+      "loss": 0.2731,
+      "step": 1531
+    },
+    {
+      "epoch": 6.869955156950673,
+      "grad_norm": 0.2013264298439026,
+      "learning_rate": 1.0523231779832055e-08,
+      "loss": 0.275,
+      "step": 1532
+    },
+    {
+      "epoch": 6.874439461883408,
+      "grad_norm": 0.17029717564582825,
+      "learning_rate": 9.810237743724805e-09,
+      "loss": 0.2817,
+      "step": 1533
+    },
+    {
+      "epoch": 6.878923766816143,
+      "grad_norm": 0.17487560212612152,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.2779,
+      "step": 1534
+    },
+    {
+      "epoch": 6.883408071748879,
+      "grad_norm": 0.2024119347333908,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.2769,
+      "step": 1535
+    },
+    {
+      "epoch": 6.887892376681615,
+      "grad_norm": 0.14625321328639984,
+      "learning_rate": 7.8211803441125e-09,
+      "loss": 0.2634,
+      "step": 1536
+    },
+    {
+      "epoch": 6.8923766816143495,
+      "grad_norm": 0.17858166992664337,
+      "learning_rate": 7.2081471792911914e-09,
+      "loss": 0.2707,
+      "step": 1537
+    },
+    {
+      "epoch": 6.896860986547085,
+      "grad_norm": 0.17319679260253906,
+      "learning_rate": 6.6201122185649555e-09,
+      "loss": 0.2724,
+      "step": 1538
+    },
+    {
+      "epoch": 6.901345291479821,
+      "grad_norm": 0.15241561830043793,
+      "learning_rate": 6.057078406142003e-09,
+      "loss": 0.2738,
+      "step": 1539
+    },
+    {
+      "epoch": 6.905829596412556,
+      "grad_norm": 0.22690723836421967,
+      "learning_rate": 5.519048561053453e-09,
+      "loss": 0.2765,
+      "step": 1540
+    },
+    {
+      "epoch": 6.910313901345291,
+      "grad_norm": 0.14751440286636353,
+      "learning_rate": 5.006025377138901e-09,
+      "loss": 0.2784,
+      "step": 1541
+    },
+    {
+      "epoch": 6.914798206278027,
+      "grad_norm": 0.16379216313362122,
+      "learning_rate": 4.518011423032542e-09,
+      "loss": 0.2856,
+      "step": 1542
+    },
+    {
+      "epoch": 6.9192825112107625,
+      "grad_norm": 0.17795665562152863,
+      "learning_rate": 4.055009142152066e-09,
+      "loss": 0.2862,
+      "step": 1543
+    },
+    {
+      "epoch": 6.923766816143498,
+      "grad_norm": 0.15283875167369843,
+      "learning_rate": 3.6170208526836724e-09,
+      "loss": 0.2691,
+      "step": 1544
+    },
+    {
+      "epoch": 6.928251121076233,
+      "grad_norm": 0.15915773808956146,
+      "learning_rate": 3.204048747573185e-09,
+      "loss": 0.2755,
+      "step": 1545
+    },
+    {
+      "epoch": 6.932735426008969,
+      "grad_norm": 0.17655691504478455,
+      "learning_rate": 2.816094894513843e-09,
+      "loss": 0.2751,
+      "step": 1546
+    },
+    {
+      "epoch": 6.937219730941704,
+      "grad_norm": 0.1536806970834732,
+      "learning_rate": 2.4531612359363077e-09,
+      "loss": 0.2781,
+      "step": 1547
+    },
+    {
+      "epoch": 6.941704035874439,
+      "grad_norm": 0.21177127957344055,
+      "learning_rate": 2.1152495889970035e-09,
+      "loss": 0.28,
+      "step": 1548
+    },
+    {
+      "epoch": 6.946188340807175,
+      "grad_norm": 0.20526771247386932,
+      "learning_rate": 1.8023616455731253e-09,
+      "loss": 0.271,
+      "step": 1549
+    },
+    {
+      "epoch": 6.95067264573991,
+      "grad_norm": 0.18504661321640015,
+      "learning_rate": 1.514498972249312e-09,
+      "loss": 0.2738,
+      "step": 1550
+    },
+    {
+      "epoch": 6.955156950672646,
+      "grad_norm": 0.16708187758922577,
+      "learning_rate": 1.2516630103137638e-09,
+      "loss": 0.273,
+      "step": 1551
+    },
+    {
+      "epoch": 6.959641255605381,
+      "grad_norm": 0.1712672859430313,
+      "learning_rate": 1.0138550757493592e-09,
+      "loss": 0.2752,
+      "step": 1552
+    },
+    {
+      "epoch": 6.9641255605381165,
+      "grad_norm": 0.16770803928375244,
+      "learning_rate": 8.010763592264381e-10,
+      "loss": 0.2765,
+      "step": 1553
+    },
+    {
+      "epoch": 6.968609865470852,
+      "grad_norm": 0.16077080368995667,
+      "learning_rate": 6.133279260983616e-10,
+      "loss": 0.273,
+      "step": 1554
+    },
+    {
+      "epoch": 6.973094170403588,
+      "grad_norm": 0.20309525728225708,
+      "learning_rate": 4.506107163948503e-10,
+      "loss": 0.2685,
+      "step": 1555
+    },
+    {
+      "epoch": 6.977578475336323,
+      "grad_norm": 0.18380437791347504,
+      "learning_rate": 3.12925544818099e-10,
+      "loss": 0.2764,
+      "step": 1556
+    },
+    {
+      "epoch": 6.982062780269058,
+      "grad_norm": 0.18821093440055847,
+      "learning_rate": 2.0027310073833516e-10,
+      "loss": 0.2748,
+      "step": 1557
+    },
+    {
+      "epoch": 6.986547085201794,
+      "grad_norm": 0.15693610906600952,
+      "learning_rate": 1.1265394818993358e-10,
+      "loss": 0.2711,
+      "step": 1558
+    },
+    {
+      "epoch": 6.991031390134529,
+      "grad_norm": 0.197928786277771,
+      "learning_rate": 5.0068525870305974e-11,
+      "loss": 0.2608,
+      "step": 1559
+    },
+    {
+      "epoch": 6.995515695067264,
+      "grad_norm": 0.2086964100599289,
+      "learning_rate": 1.251714713546015e-11,
+      "loss": 0.2761,
+      "step": 1560
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.20751763880252838,
+      "learning_rate": 0.0,
+      "loss": 0.2595,
+      "step": 1561
+    },
+    {
+      "epoch": 7.0,
+      "step": 1561,
+      "total_flos": 3.626520802649702e+16,
+      "train_loss": 0.26103385187050687,
+      "train_runtime": 81858.0264,
+      "train_samples_per_second": 9.745,
+      "train_steps_per_second": 0.019
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1561,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.626520802649702e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}