diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,47889 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999023866464932,
+  "eval_steps": 900,
+  "global_step": 6829,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.867297319395897,
+      "learning_rate": 8.99868209108215e-05,
+      "loss": 3.2458,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.122135116455087,
+      "learning_rate": 8.9973641821643e-05,
+      "loss": 6.8269,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.493605611332246,
+      "learning_rate": 8.99604627324645e-05,
+      "loss": 6.1526,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.993597101446063,
+      "learning_rate": 8.994728364328598e-05,
+      "loss": 6.2791,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.994941997336733,
+      "learning_rate": 8.993410455410749e-05,
+      "loss": 5.5069,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.044801050606559,
+      "learning_rate": 8.992092546492898e-05,
+      "loss": 5.2917,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.371025975927372,
+      "learning_rate": 8.990774637575048e-05,
+      "loss": 5.0841,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.095134034689849,
+      "learning_rate": 8.989456728657198e-05,
+      "loss": 4.6676,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.727749907194312,
+      "learning_rate": 8.988138819739348e-05,
+      "loss": 4.4157,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1785847861912533,
+      "learning_rate": 8.986820910821497e-05,
+      "loss": 4.3083,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5890441000530777,
+      "learning_rate": 8.985503001903648e-05,
+      "loss": 4.0652,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7851888183485998,
+      "learning_rate": 8.984185092985796e-05,
+      "loss": 4.0157,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.177578439182146,
+      "learning_rate": 8.982867184067945e-05,
+      "loss": 3.8923,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.189312958133593,
+      "learning_rate": 8.981549275150096e-05,
+      "loss": 3.8165,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.135971084370728,
+      "learning_rate": 8.980231366232245e-05,
+      "loss": 3.7407,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6742409253843133,
+      "learning_rate": 8.978913457314395e-05,
+      "loss": 3.6203,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.740647923231835,
+      "learning_rate": 8.977595548396545e-05,
+      "loss": 3.5358,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3386504358864526,
+      "learning_rate": 8.976277639478695e-05,
+      "loss": 3.4649,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4146027388570108,
+      "learning_rate": 8.974959730560844e-05,
+      "loss": 3.4096,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3101076465055932,
+      "learning_rate": 8.973641821642993e-05,
+      "loss": 3.3495,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3979232652764932,
+      "learning_rate": 8.972323912725143e-05,
+      "loss": 3.2864,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.112058959017875,
+      "learning_rate": 8.971006003807292e-05,
+      "loss": 3.2272,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5723342224397572,
+      "learning_rate": 8.969688094889443e-05,
+      "loss": 3.202,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6929219679520722,
+      "learning_rate": 8.968370185971592e-05,
+      "loss": 3.1556,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0947475079787117,
+      "learning_rate": 8.967052277053742e-05,
+      "loss": 3.1127,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0263062806535903,
+      "learning_rate": 8.965734368135892e-05,
+      "loss": 3.07,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9314787514966034,
+      "learning_rate": 8.964416459218042e-05,
+      "loss": 3.0321,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8258215193640202,
+      "learning_rate": 8.96309855030019e-05,
+      "loss": 2.9997,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8854426518873304,
+      "learning_rate": 8.96178064138234e-05,
+      "loss": 2.9693,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7136563954565109,
+      "learning_rate": 8.96046273246449e-05,
+      "loss": 2.9337,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8520936236031108,
+      "learning_rate": 8.959144823546639e-05,
+      "loss": 2.8958,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7069687542653055,
+      "learning_rate": 8.95782691462879e-05,
+      "loss": 2.8893,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6112456126798674,
+      "learning_rate": 8.95650900571094e-05,
+      "loss": 2.8452,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6182491443128697,
+      "learning_rate": 8.955191096793089e-05,
+      "loss": 2.8375,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6354831752982139,
+      "learning_rate": 8.95387318787524e-05,
+      "loss": 2.8036,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5551758024561351,
+      "learning_rate": 8.952555278957388e-05,
+      "loss": 2.7898,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5385458807945395,
+      "learning_rate": 8.951237370039537e-05,
+      "loss": 2.7654,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.611312479127073,
+      "learning_rate": 8.949919461121688e-05,
+      "loss": 2.7404,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.41994683299580327,
+      "learning_rate": 8.948601552203837e-05,
+      "loss": 2.727,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4844003990796705,
+      "learning_rate": 8.947283643285986e-05,
+      "loss": 2.7055,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5096807026587409,
+      "learning_rate": 8.945965734368137e-05,
+      "loss": 2.6912,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7522366060460143,
+      "learning_rate": 8.944647825450286e-05,
+      "loss": 2.681,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9331119398395223,
+      "learning_rate": 8.943329916532436e-05,
+      "loss": 2.6557,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.012546232060092,
+      "learning_rate": 8.942012007614585e-05,
+      "loss": 2.6533,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8636983952962917,
+      "learning_rate": 8.940694098696735e-05,
+      "loss": 2.6397,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7226119626951081,
+      "learning_rate": 8.939376189778884e-05,
+      "loss": 2.6172,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8159190836781818,
+      "learning_rate": 8.938058280861035e-05,
+      "loss": 2.601,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8397723356140642,
+      "learning_rate": 8.936740371943184e-05,
+      "loss": 2.5913,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9302397602027757,
+      "learning_rate": 8.935422463025333e-05,
+      "loss": 2.5779,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9294815366036785,
+      "learning_rate": 8.934104554107484e-05,
+      "loss": 2.5666,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9876017326778281,
+      "learning_rate": 8.932786645189634e-05,
+      "loss": 2.559,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0531686104317621,
+      "learning_rate": 8.931468736271782e-05,
+      "loss": 2.5426,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9432926643235091,
+      "learning_rate": 8.930150827353932e-05,
+      "loss": 2.5324,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0147959053073372,
+      "learning_rate": 8.928832918436082e-05,
+      "loss": 2.5166,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8370421203774538,
+      "learning_rate": 8.927515009518231e-05,
+      "loss": 2.5035,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7658170492109742,
+      "learning_rate": 8.926197100600382e-05,
+      "loss": 2.5064,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5836453359341242,
+      "learning_rate": 8.924879191682531e-05,
+      "loss": 2.4889,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6774145346502838,
+      "learning_rate": 8.92356128276468e-05,
+      "loss": 2.4773,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6303115552198764,
+      "learning_rate": 8.922243373846831e-05,
+      "loss": 2.4746,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4733727246322016,
+      "learning_rate": 8.920925464928979e-05,
+      "loss": 2.4583,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66470201831204,
+      "learning_rate": 8.919607556011129e-05,
+      "loss": 2.4509,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6356308149108085,
+      "learning_rate": 8.91828964709328e-05,
+      "loss": 2.4375,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46831466582575737,
+      "learning_rate": 8.916971738175429e-05,
+      "loss": 2.4314,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5711699065497955,
+      "learning_rate": 8.915653829257578e-05,
+      "loss": 2.42,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5389976605269295,
+      "learning_rate": 8.914335920339729e-05,
+      "loss": 2.4161,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5007278366304382,
+      "learning_rate": 8.913018011421878e-05,
+      "loss": 2.3977,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.487467802029198,
+      "learning_rate": 8.911700102504028e-05,
+      "loss": 2.4007,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4113420784339612,
+      "learning_rate": 8.910382193586177e-05,
+      "loss": 2.3966,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4504091272252521,
+      "learning_rate": 8.909064284668326e-05,
+      "loss": 2.3862,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34513031845084635,
+      "learning_rate": 8.907746375750476e-05,
+      "loss": 2.3712,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5047955438477213,
+      "learning_rate": 8.906428466832626e-05,
+      "loss": 2.3695,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5024769592440181,
+      "learning_rate": 8.905110557914776e-05,
+      "loss": 2.3644,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4721080086046586,
+      "learning_rate": 8.903792648996925e-05,
+      "loss": 2.3555,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4591297268523745,
+      "learning_rate": 8.902474740079076e-05,
+      "loss": 2.3376,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5656344865220397,
+      "learning_rate": 8.901156831161225e-05,
+      "loss": 2.3398,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6500485453460426,
+      "learning_rate": 8.899838922243373e-05,
+      "loss": 2.3325,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6536690673884828,
+      "learning_rate": 8.898521013325524e-05,
+      "loss": 2.3345,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5143484723108316,
+      "learning_rate": 8.897203104407673e-05,
+      "loss": 2.3139,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.37322968848559845,
+      "learning_rate": 8.895885195489823e-05,
+      "loss": 2.3079,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34462618220763114,
+      "learning_rate": 8.894567286571974e-05,
+      "loss": 2.2973,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38708375377202425,
+      "learning_rate": 8.893249377654123e-05,
+      "loss": 2.2951,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.421171485688958,
+      "learning_rate": 8.891931468736272e-05,
+      "loss": 2.2951,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4465155129873313,
+      "learning_rate": 8.890613559818423e-05,
+      "loss": 2.2908,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.42100671498868236,
+      "learning_rate": 8.889295650900571e-05,
+      "loss": 2.2862,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4308804247735972,
+      "learning_rate": 8.88797774198272e-05,
+      "loss": 2.2868,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4490696078386771,
+      "learning_rate": 8.886659833064871e-05,
+      "loss": 2.2784,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4720132579092968,
+      "learning_rate": 8.88534192414702e-05,
+      "loss": 2.2596,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5051381693356659,
+      "learning_rate": 8.88402401522917e-05,
+      "loss": 2.2644,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5161793827955174,
+      "learning_rate": 8.88270610631132e-05,
+      "loss": 2.2561,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5947828202225921,
+      "learning_rate": 8.88138819739347e-05,
+      "loss": 2.2485,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7363411963455597,
+      "learning_rate": 8.88007028847562e-05,
+      "loss": 2.256,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7827586632222484,
+      "learning_rate": 8.878752379557769e-05,
+      "loss": 2.2498,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5518918442765439,
+      "learning_rate": 8.877434470639918e-05,
+      "loss": 2.242,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47976621209079673,
+      "learning_rate": 8.876116561722067e-05,
+      "loss": 2.2336,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6515602255953327,
+      "learning_rate": 8.874798652804218e-05,
+      "loss": 2.2348,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6536808073233461,
+      "learning_rate": 8.873480743886368e-05,
+      "loss": 2.2269,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.45429917400178504,
+      "learning_rate": 8.872162834968517e-05,
+      "loss": 2.2215,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6059083408950834,
+      "learning_rate": 8.870844926050668e-05,
+      "loss": 2.2141,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6153517456337565,
+      "learning_rate": 8.869527017132817e-05,
+      "loss": 2.2135,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4350225773043892,
+      "learning_rate": 8.868209108214965e-05,
+      "loss": 2.2046,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7538851308577357,
+      "learning_rate": 8.866891199297116e-05,
+      "loss": 2.1987,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7783668567986096,
+      "learning_rate": 8.865573290379265e-05,
+      "loss": 2.1944,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6152937329119922,
+      "learning_rate": 8.864255381461415e-05,
+      "loss": 2.1976,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8129985392759269,
+      "learning_rate": 8.862937472543565e-05,
+      "loss": 2.1969,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.658961787535577,
+      "learning_rate": 8.861619563625715e-05,
+      "loss": 2.1916,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5694890400136322,
+      "learning_rate": 8.860301654707864e-05,
+      "loss": 2.1853,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8449981372046623,
+      "learning_rate": 8.858983745790015e-05,
+      "loss": 2.1918,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6504754487654488,
+      "learning_rate": 8.857665836872163e-05,
+      "loss": 2.18,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5770328719385773,
+      "learning_rate": 8.856347927954312e-05,
+      "loss": 2.1726,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.876682475915139,
+      "learning_rate": 8.855030019036463e-05,
+      "loss": 2.1697,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6193197639347029,
+      "learning_rate": 8.853712110118612e-05,
+      "loss": 2.1762,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5205430357287273,
+      "learning_rate": 8.852394201200762e-05,
+      "loss": 2.1643,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6357317276650712,
+      "learning_rate": 8.851076292282912e-05,
+      "loss": 2.1698,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5337930185336511,
+      "learning_rate": 8.849758383365062e-05,
+      "loss": 2.1533,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5062081503044676,
+      "learning_rate": 8.848440474447211e-05,
+      "loss": 2.1554,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4764632463998196,
+      "learning_rate": 8.84712256552936e-05,
+      "loss": 2.1505,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.41457257334024383,
+      "learning_rate": 8.84580465661151e-05,
+      "loss": 2.1467,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5065884475761684,
+      "learning_rate": 8.844486747693659e-05,
+      "loss": 2.1422,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4564614277589642,
+      "learning_rate": 8.84316883877581e-05,
+      "loss": 2.1412,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3554317111825436,
+      "learning_rate": 8.84185092985796e-05,
+      "loss": 2.1455,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42806997535938857,
+      "learning_rate": 8.840533020940109e-05,
+      "loss": 2.1275,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42000313047228083,
+      "learning_rate": 8.83921511202226e-05,
+      "loss": 2.1342,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3885632469021741,
+      "learning_rate": 8.837897203104409e-05,
+      "loss": 2.1307,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47343669511945285,
+      "learning_rate": 8.836579294186557e-05,
+      "loss": 2.1317,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48475235283422596,
+      "learning_rate": 8.835261385268708e-05,
+      "loss": 2.1161,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.566942141589765,
+      "learning_rate": 8.833943476350857e-05,
+      "loss": 2.1182,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6962692105838626,
+      "learning_rate": 8.832625567433006e-05,
+      "loss": 2.1201,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8872705283662579,
+      "learning_rate": 8.831307658515157e-05,
+      "loss": 2.1125,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0040048160020307,
+      "learning_rate": 8.829989749597306e-05,
+      "loss": 2.1243,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8763157343936164,
+      "learning_rate": 8.828671840679456e-05,
+      "loss": 2.135,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4928878082025761,
+      "learning_rate": 8.827353931761607e-05,
+      "loss": 2.1159,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5070687239013271,
+      "learning_rate": 8.826036022843755e-05,
+      "loss": 2.1048,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7840409646534107,
+      "learning_rate": 8.824718113925904e-05,
+      "loss": 2.1037,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5221585439101886,
+      "learning_rate": 8.823400205008055e-05,
+      "loss": 2.0959,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4568661842831729,
+      "learning_rate": 8.822082296090204e-05,
+      "loss": 2.0918,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5841088902261057,
+      "learning_rate": 8.820764387172353e-05,
+      "loss": 2.0916,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5321596782302892,
+      "learning_rate": 8.819446478254504e-05,
+      "loss": 2.0897,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4811464623013186,
+      "learning_rate": 8.818128569336654e-05,
+      "loss": 2.0797,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45409201842843017,
+      "learning_rate": 8.816810660418803e-05,
+      "loss": 2.0809,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5436744181229896,
+      "learning_rate": 8.815492751500952e-05,
+      "loss": 2.0768,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.465625799362755,
+      "learning_rate": 8.814174842583102e-05,
+      "loss": 2.0801,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4631920693131495,
+      "learning_rate": 8.812856933665251e-05,
+      "loss": 2.0729,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4973179737348441,
+      "learning_rate": 8.811539024747402e-05,
+      "loss": 2.0765,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3900332148198551,
+      "learning_rate": 8.810221115829551e-05,
+      "loss": 2.0688,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45058451016387613,
+      "learning_rate": 8.8089032069117e-05,
+      "loss": 2.0684,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35118754238330746,
+      "learning_rate": 8.807585297993851e-05,
+      "loss": 2.0681,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33942242720035193,
+      "learning_rate": 8.806267389076e-05,
+      "loss": 2.0626,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34835606297437544,
+      "learning_rate": 8.804949480158149e-05,
+      "loss": 2.0611,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33374548138193794,
+      "learning_rate": 8.8036315712403e-05,
+      "loss": 2.0589,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3554684609547751,
+      "learning_rate": 8.802313662322449e-05,
+      "loss": 2.0531,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30356995848645646,
+      "learning_rate": 8.800995753404598e-05,
+      "loss": 2.0527,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3749048670516549,
+      "learning_rate": 8.799677844486749e-05,
+      "loss": 2.0409,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44906357358681387,
+      "learning_rate": 8.798359935568898e-05,
+      "loss": 2.0455,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4937506907032551,
+      "learning_rate": 8.797042026651048e-05,
+      "loss": 2.0593,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6302087869169094,
+      "learning_rate": 8.795724117733198e-05,
+      "loss": 2.0444,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7666465190058978,
+      "learning_rate": 8.794406208815346e-05,
+      "loss": 2.0482,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8415113455563511,
+      "learning_rate": 8.793088299897496e-05,
+      "loss": 2.0553,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6599046675210126,
+      "learning_rate": 8.791770390979646e-05,
+      "loss": 2.0346,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4286698284091557,
+      "learning_rate": 8.790452482061796e-05,
+      "loss": 2.0364,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4636961335954888,
+      "learning_rate": 8.789134573143945e-05,
+      "loss": 2.027,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4729809233526084,
+      "learning_rate": 8.787816664226096e-05,
+      "loss": 2.0344,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5102897103629896,
+      "learning_rate": 8.786498755308245e-05,
+      "loss": 2.0313,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3821986499935569,
+      "learning_rate": 8.785180846390395e-05,
+      "loss": 2.0281,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31922482573170025,
+      "learning_rate": 8.783862937472544e-05,
+      "loss": 2.0246,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4125805076448191,
+      "learning_rate": 8.782545028554693e-05,
+      "loss": 2.0209,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3584157670577966,
+      "learning_rate": 8.781227119636843e-05,
+      "loss": 2.0201,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3732188737105076,
+      "learning_rate": 8.779909210718993e-05,
+      "loss": 2.0181,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34438987877017796,
+      "learning_rate": 8.778591301801143e-05,
+      "loss": 2.0155,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.322936068818923,
+      "learning_rate": 8.777273392883292e-05,
+      "loss": 2.0206,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3671965249502594,
+      "learning_rate": 8.775955483965443e-05,
+      "loss": 2.0215,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3303955763338707,
+      "learning_rate": 8.774637575047592e-05,
+      "loss": 2.0134,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26010553136423376,
+      "learning_rate": 8.77331966612974e-05,
+      "loss": 1.9998,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28273036383624667,
+      "learning_rate": 8.772001757211891e-05,
+      "loss": 2.008,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3493466030437232,
+      "learning_rate": 8.77068384829404e-05,
+      "loss": 2.0077,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3240073311732844,
+      "learning_rate": 8.76936593937619e-05,
+      "loss": 2.0005,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37871862295109215,
+      "learning_rate": 8.76804803045834e-05,
+      "loss": 1.9947,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36425872526816194,
+      "learning_rate": 8.76673012154049e-05,
+      "loss": 2.0012,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3146144912051344,
+      "learning_rate": 8.765412212622639e-05,
+      "loss": 1.9888,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32672427059297815,
+      "learning_rate": 8.764094303704789e-05,
+      "loss": 1.9914,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2874426352457461,
+      "learning_rate": 8.762776394786938e-05,
+      "loss": 1.991,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.304367486735733,
+      "learning_rate": 8.761458485869087e-05,
+      "loss": 1.9806,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37875831866804927,
+      "learning_rate": 8.760140576951238e-05,
+      "loss": 1.9924,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4835755191486126,
+      "learning_rate": 8.758822668033388e-05,
+      "loss": 1.9921,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5565270140077677,
+      "learning_rate": 8.757504759115537e-05,
+      "loss": 1.9822,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5464855527172483,
+      "learning_rate": 8.756186850197688e-05,
+      "loss": 1.9839,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4989777454659821,
+      "learning_rate": 8.754868941279837e-05,
+      "loss": 1.9867,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4574274365674215,
+      "learning_rate": 8.753551032361986e-05,
+      "loss": 1.9837,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47786281828321875,
+      "learning_rate": 8.752233123444136e-05,
+      "loss": 1.989,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5193952650470612,
+      "learning_rate": 8.750915214526285e-05,
+      "loss": 1.9721,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46706899407183416,
+      "learning_rate": 8.749597305608435e-05,
+      "loss": 1.9721,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41004377608522635,
+      "learning_rate": 8.748279396690585e-05,
+      "loss": 1.9736,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4094480903013732,
+      "learning_rate": 8.746961487772735e-05,
+      "loss": 1.9714,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5224817485503537,
+      "learning_rate": 8.745643578854884e-05,
+      "loss": 1.9775,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5763427830733111,
+      "learning_rate": 8.744325669937035e-05,
+      "loss": 1.9691,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45154944484866116,
+      "learning_rate": 8.743007761019184e-05,
+      "loss": 1.9685,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29409144127731623,
+      "learning_rate": 8.741689852101332e-05,
+      "loss": 1.9665,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3391706172784746,
+      "learning_rate": 8.740371943183483e-05,
+      "loss": 1.9657,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3584680315892078,
+      "learning_rate": 8.739054034265632e-05,
+      "loss": 1.9662,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32480662203931654,
+      "learning_rate": 8.737736125347782e-05,
+      "loss": 1.967,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2678893712664796,
+      "learning_rate": 8.736418216429932e-05,
+      "loss": 1.9667,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3034157270439372,
+      "learning_rate": 8.735100307512082e-05,
+      "loss": 1.9639,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38638946729313517,
+      "learning_rate": 8.733782398594231e-05,
+      "loss": 1.9654,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44116092753081665,
+      "learning_rate": 8.73246448967638e-05,
+      "loss": 1.9662,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5599610453487487,
+      "learning_rate": 8.73114658075853e-05,
+      "loss": 1.9589,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6813838919621874,
+      "learning_rate": 8.729828671840679e-05,
+      "loss": 1.96,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6749810768948219,
+      "learning_rate": 8.72851076292283e-05,
+      "loss": 1.9669,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5474380551907909,
+      "learning_rate": 8.727192854004979e-05,
+      "loss": 1.9479,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39912501114471677,
+      "learning_rate": 8.725874945087129e-05,
+      "loss": 1.9591,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3704777650253716,
+      "learning_rate": 8.72455703616928e-05,
+      "loss": 1.9535,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45420800985949406,
+      "learning_rate": 8.723239127251429e-05,
+      "loss": 1.9476,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5106582102448923,
+      "learning_rate": 8.721921218333578e-05,
+      "loss": 1.954,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.648206578856157,
+      "learning_rate": 8.720603309415728e-05,
+      "loss": 1.963,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.35693125434587636,
+      "learning_rate": 8.719285400497877e-05,
+      "loss": 1.9419,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6519985094603863,
+      "learning_rate": 8.717967491580026e-05,
+      "loss": 1.9445,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6567675718025047,
+      "learning_rate": 8.716649582662177e-05,
+      "loss": 1.9469,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4872034695801568,
+      "learning_rate": 8.715331673744326e-05,
+      "loss": 1.9495,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4145209119308714,
+      "learning_rate": 8.714013764826476e-05,
+      "loss": 1.9486,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4775886307734156,
+      "learning_rate": 8.712695855908626e-05,
+      "loss": 1.9496,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40804530533441385,
+      "learning_rate": 8.711377946990776e-05,
+      "loss": 1.9419,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3313950875913946,
+      "learning_rate": 8.710060038072924e-05,
+      "loss": 1.9372,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3887501325873841,
+      "learning_rate": 8.708742129155075e-05,
+      "loss": 1.9421,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4319364449745349,
+      "learning_rate": 8.707424220237224e-05,
+      "loss": 1.9385,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39127832437474,
+      "learning_rate": 8.706106311319373e-05,
+      "loss": 1.935,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36405269202971846,
+      "learning_rate": 8.704788402401524e-05,
+      "loss": 1.9395,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3815614792943797,
+      "learning_rate": 8.703470493483673e-05,
+      "loss": 1.9351,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43828706493068725,
+      "learning_rate": 8.702152584565823e-05,
+      "loss": 1.9217,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39922403213927826,
+      "learning_rate": 8.700834675647972e-05,
+      "loss": 1.9289,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3844142284075385,
+      "learning_rate": 8.699516766730122e-05,
+      "loss": 1.9248,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38881007783639865,
+      "learning_rate": 8.698198857812271e-05,
+      "loss": 1.9201,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2975403535899065,
+      "learning_rate": 8.696880948894422e-05,
+      "loss": 1.928,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2832203830943493,
+      "learning_rate": 8.695563039976571e-05,
+      "loss": 1.9162,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2979487231657903,
+      "learning_rate": 8.69424513105872e-05,
+      "loss": 1.9271,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3621148892850887,
+      "learning_rate": 8.692927222140871e-05,
+      "loss": 1.9156,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3252924764609856,
+      "learning_rate": 8.69160931322302e-05,
+      "loss": 1.9208,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38778060232379175,
+      "learning_rate": 8.69029140430517e-05,
+      "loss": 1.9137,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5259448310725102,
+      "learning_rate": 8.688973495387319e-05,
+      "loss": 1.915,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6748649230965568,
+      "learning_rate": 8.687655586469469e-05,
+      "loss": 1.9335,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7329646137433606,
+      "learning_rate": 8.686337677551618e-05,
+      "loss": 1.921,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6477306709245009,
+      "learning_rate": 8.685019768633769e-05,
+      "loss": 1.9107,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4249940595660931,
+      "learning_rate": 8.683701859715918e-05,
+      "loss": 1.9074,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40780487506439317,
+      "learning_rate": 8.682383950798068e-05,
+      "loss": 1.9121,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5063150515112605,
+      "learning_rate": 8.681066041880218e-05,
+      "loss": 1.9122,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43654436775464556,
+      "learning_rate": 8.679748132962368e-05,
+      "loss": 1.9101,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28693671377727636,
+      "learning_rate": 8.678430224044516e-05,
+      "loss": 1.9071,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35062666021736943,
+      "learning_rate": 8.677112315126666e-05,
+      "loss": 1.9052,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4571466841247422,
+      "learning_rate": 8.675794406208816e-05,
+      "loss": 1.9037,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3856445615558911,
+      "learning_rate": 8.674476497290965e-05,
+      "loss": 1.9033,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3233514425302591,
+      "learning_rate": 8.673158588373116e-05,
+      "loss": 1.904,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4212339592559155,
+      "learning_rate": 8.671840679455265e-05,
+      "loss": 1.9029,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4275588677912957,
+      "learning_rate": 8.670522770537415e-05,
+      "loss": 1.8994,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29035057566401584,
+      "learning_rate": 8.669204861619564e-05,
+      "loss": 1.8968,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3292457727990474,
+      "learning_rate": 8.667886952701713e-05,
+      "loss": 1.8971,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39002632439582646,
+      "learning_rate": 8.666569043783863e-05,
+      "loss": 1.9019,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3058384482996071,
+      "learning_rate": 8.665251134866013e-05,
+      "loss": 1.8991,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33747054066822696,
+      "learning_rate": 8.663933225948163e-05,
+      "loss": 1.8891,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35931140472197654,
+      "learning_rate": 8.662615317030312e-05,
+      "loss": 1.9025,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.312755193621967,
+      "learning_rate": 8.661297408112463e-05,
+      "loss": 1.8904,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.24759800889452924,
+      "learning_rate": 8.659979499194612e-05,
+      "loss": 1.8957,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23365570528449237,
+      "learning_rate": 8.658661590276762e-05,
+      "loss": 1.8895,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26012031059809776,
+      "learning_rate": 8.657343681358911e-05,
+      "loss": 1.874,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2369419439744028,
+      "learning_rate": 8.65602577244106e-05,
+      "loss": 1.8888,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23737605818719734,
+      "learning_rate": 8.65470786352321e-05,
+      "loss": 1.892,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20560803394813162,
+      "learning_rate": 8.65338995460536e-05,
+      "loss": 1.8902,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23756980446551546,
+      "learning_rate": 8.65207204568751e-05,
+      "loss": 1.8914,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2955355993564412,
+      "learning_rate": 8.650754136769659e-05,
+      "loss": 1.8741,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36625312284796796,
+      "learning_rate": 8.649436227851809e-05,
+      "loss": 1.8849,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4377187791569674,
+      "learning_rate": 8.64811831893396e-05,
+      "loss": 1.875,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4751457670664482,
+      "learning_rate": 8.646800410016107e-05,
+      "loss": 1.887,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5124162062393218,
+      "learning_rate": 8.645482501098258e-05,
+      "loss": 1.8759,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5128407598622339,
+      "learning_rate": 8.644164592180407e-05,
+      "loss": 1.884,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4645016436292238,
+      "learning_rate": 8.642846683262557e-05,
+      "loss": 1.8748,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42963082194049396,
+      "learning_rate": 8.641528774344708e-05,
+      "loss": 1.8725,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41218244280637906,
+      "learning_rate": 8.640210865426857e-05,
+      "loss": 1.8791,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34726276152467556,
+      "learning_rate": 8.638892956509006e-05,
+      "loss": 1.8742,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.328823655826036,
+      "learning_rate": 8.637575047591156e-05,
+      "loss": 1.8721,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3686461464930126,
+      "learning_rate": 8.636257138673305e-05,
+      "loss": 1.8687,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3656673874029743,
+      "learning_rate": 8.634939229755454e-05,
+      "loss": 1.8654,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30525439673036175,
+      "learning_rate": 8.633621320837605e-05,
+      "loss": 1.8746,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3007226581860427,
+      "learning_rate": 8.632303411919755e-05,
+      "loss": 1.8664,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3849279004150829,
+      "learning_rate": 8.630985503001904e-05,
+      "loss": 1.8717,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43777228984264327,
+      "learning_rate": 8.629667594084055e-05,
+      "loss": 1.8776,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4220356221423105,
+      "learning_rate": 8.628349685166204e-05,
+      "loss": 1.8705,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4049297705769528,
+      "learning_rate": 8.627031776248353e-05,
+      "loss": 1.8702,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3919834421059084,
+      "learning_rate": 8.625713867330503e-05,
+      "loss": 1.8672,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3713089483320666,
+      "learning_rate": 8.624395958412652e-05,
+      "loss": 1.8626,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3604341363789889,
+      "learning_rate": 8.623078049494802e-05,
+      "loss": 1.8724,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39717353573166714,
+      "learning_rate": 8.621760140576952e-05,
+      "loss": 1.8646,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44719078220357295,
+      "learning_rate": 8.620442231659102e-05,
+      "loss": 1.8582,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4513680274672226,
+      "learning_rate": 8.619124322741251e-05,
+      "loss": 1.8696,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43023542467375564,
+      "learning_rate": 8.6178064138234e-05,
+      "loss": 1.8699,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43858266418234654,
+      "learning_rate": 8.616488504905551e-05,
+      "loss": 1.86,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5064425360850751,
+      "learning_rate": 8.615170595987699e-05,
+      "loss": 1.858,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48407682623823045,
+      "learning_rate": 8.61385268706985e-05,
+      "loss": 1.853,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46114150689754957,
+      "learning_rate": 8.612534778151999e-05,
+      "loss": 1.8598,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44714939482574123,
+      "learning_rate": 8.611216869234149e-05,
+      "loss": 1.8571,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29499856062245083,
+      "learning_rate": 8.6098989603163e-05,
+      "loss": 1.85,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.27237302520628326,
+      "learning_rate": 8.608581051398449e-05,
+      "loss": 1.8633,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3852560547170149,
+      "learning_rate": 8.607263142480598e-05,
+      "loss": 1.84,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4338967169702793,
+      "learning_rate": 8.605945233562747e-05,
+      "loss": 1.86,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3952031517559691,
+      "learning_rate": 8.604627324644897e-05,
+      "loss": 1.8473,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25514061111362424,
+      "learning_rate": 8.603309415727046e-05,
+      "loss": 1.8464,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3078510816972887,
+      "learning_rate": 8.601991506809197e-05,
+      "loss": 1.8518,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38406564873200627,
+      "learning_rate": 8.600673597891346e-05,
+      "loss": 1.8453,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29032703723109443,
+      "learning_rate": 8.599355688973496e-05,
+      "loss": 1.8426,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21551221995312625,
+      "learning_rate": 8.598037780055646e-05,
+      "loss": 1.8407,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25104013568498373,
+      "learning_rate": 8.596719871137796e-05,
+      "loss": 1.8412,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2753448788479235,
+      "learning_rate": 8.595401962219945e-05,
+      "loss": 1.85,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28287628219754707,
+      "learning_rate": 8.594084053302095e-05,
+      "loss": 1.8393,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29153292982905893,
+      "learning_rate": 8.592766144384244e-05,
+      "loss": 1.8436,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3356468622059691,
+      "learning_rate": 8.591448235466393e-05,
+      "loss": 1.8501,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3668022644305236,
+      "learning_rate": 8.590130326548544e-05,
+      "loss": 1.8442,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35131341619593265,
+      "learning_rate": 8.588812417630693e-05,
+      "loss": 1.8385,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32173508052581773,
+      "learning_rate": 8.587494508712843e-05,
+      "loss": 1.8436,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.300778040243218,
+      "learning_rate": 8.586176599794992e-05,
+      "loss": 1.8382,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30861066996157027,
+      "learning_rate": 8.584858690877143e-05,
+      "loss": 1.8351,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34553615609293836,
+      "learning_rate": 8.583540781959291e-05,
+      "loss": 1.8326,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3945637083333217,
+      "learning_rate": 8.582222873041442e-05,
+      "loss": 1.8439,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4185609027500278,
+      "learning_rate": 8.580904964123591e-05,
+      "loss": 1.836,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4737109913412694,
+      "learning_rate": 8.57958705520574e-05,
+      "loss": 1.8349,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.542744291036962,
+      "learning_rate": 8.578269146287891e-05,
+      "loss": 1.8369,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5698184063084959,
+      "learning_rate": 8.57695123737004e-05,
+      "loss": 1.8364,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5216244389865377,
+      "learning_rate": 8.57563332845219e-05,
+      "loss": 1.8391,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4349460582065712,
+      "learning_rate": 8.574315419534339e-05,
+      "loss": 1.8316,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4060439023599905,
+      "learning_rate": 8.572997510616489e-05,
+      "loss": 1.829,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4006651254790203,
+      "learning_rate": 8.571679601698638e-05,
+      "loss": 1.8289,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35982668295004727,
+      "learning_rate": 8.570361692780789e-05,
+      "loss": 1.8345,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34270943020191247,
+      "learning_rate": 8.569043783862938e-05,
+      "loss": 1.8244,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35208015052905844,
+      "learning_rate": 8.567725874945087e-05,
+      "loss": 1.8395,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3560007421719364,
+      "learning_rate": 8.566407966027238e-05,
+      "loss": 1.8244,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36632340017442266,
+      "learning_rate": 8.565090057109388e-05,
+      "loss": 1.8225,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33466423040979987,
+      "learning_rate": 8.563772148191537e-05,
+      "loss": 1.829,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2716882763297445,
+      "learning_rate": 8.562454239273686e-05,
+      "loss": 1.8341,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26453069945784297,
+      "learning_rate": 8.561136330355836e-05,
+      "loss": 1.834,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2678715963585482,
+      "learning_rate": 8.559818421437985e-05,
+      "loss": 1.8246,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28963443783737886,
+      "learning_rate": 8.558500512520136e-05,
+      "loss": 1.8217,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2799148037129182,
+      "learning_rate": 8.557182603602285e-05,
+      "loss": 1.8232,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25449424333816384,
+      "learning_rate": 8.555864694684435e-05,
+      "loss": 1.8211,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29873070853220085,
+      "learning_rate": 8.554546785766584e-05,
+      "loss": 1.8218,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31256528253176236,
+      "learning_rate": 8.553228876848735e-05,
+      "loss": 1.8184,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2579289256497001,
+      "learning_rate": 8.551910967930883e-05,
+      "loss": 1.8285,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2692223001749751,
+      "learning_rate": 8.550593059013033e-05,
+      "loss": 1.8199,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2828285551419333,
+      "learning_rate": 8.549275150095183e-05,
+      "loss": 1.812,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25760648063355074,
+      "learning_rate": 8.547957241177332e-05,
+      "loss": 1.8196,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3229138587412795,
+      "learning_rate": 8.546639332259483e-05,
+      "loss": 1.8168,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32584515922858565,
+      "learning_rate": 8.545321423341632e-05,
+      "loss": 1.821,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2785648276363959,
+      "learning_rate": 8.544003514423782e-05,
+      "loss": 1.8226,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32332023531594767,
+      "learning_rate": 8.542685605505931e-05,
+      "loss": 1.8067,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30635180846369325,
+      "learning_rate": 8.54136769658808e-05,
+      "loss": 1.8132,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29630729064298555,
+      "learning_rate": 8.54004978767023e-05,
+      "loss": 1.8101,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34621816808581285,
+      "learning_rate": 8.53873187875238e-05,
+      "loss": 1.8176,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37813665939757496,
+      "learning_rate": 8.53741396983453e-05,
+      "loss": 1.8118,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4024072071879195,
+      "learning_rate": 8.536096060916679e-05,
+      "loss": 1.8151,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4804782375915566,
+      "learning_rate": 8.534778151998829e-05,
+      "loss": 1.8163,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5127084558984112,
+      "learning_rate": 8.533460243080979e-05,
+      "loss": 1.811,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4388610896297329,
+      "learning_rate": 8.532142334163129e-05,
+      "loss": 1.8137,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3221135161434642,
+      "learning_rate": 8.530824425245278e-05,
+      "loss": 1.7968,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.340563157991881,
+      "learning_rate": 8.529506516327427e-05,
+      "loss": 1.8103,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39991989603879213,
+      "learning_rate": 8.528188607409577e-05,
+      "loss": 1.8087,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4120407943765869,
+      "learning_rate": 8.526870698491728e-05,
+      "loss": 1.805,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34596624387275093,
+      "learning_rate": 8.525552789573877e-05,
+      "loss": 1.8035,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25789497090466024,
+      "learning_rate": 8.524234880656026e-05,
+      "loss": 1.8058,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30263266008859196,
+      "learning_rate": 8.522916971738176e-05,
+      "loss": 1.8084,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3562148440043474,
+      "learning_rate": 8.521599062820326e-05,
+      "loss": 1.799,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30587484606473564,
+      "learning_rate": 8.520281153902474e-05,
+      "loss": 1.8003,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27973522082893804,
+      "learning_rate": 8.518963244984625e-05,
+      "loss": 1.8153,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28180525427229874,
+      "learning_rate": 8.517645336066775e-05,
+      "loss": 1.7987,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.24578221711398116,
+      "learning_rate": 8.516327427148924e-05,
+      "loss": 1.8116,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.265330003846191,
+      "learning_rate": 8.515009518231075e-05,
+      "loss": 1.8034,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2777372207789329,
+      "learning_rate": 8.513691609313224e-05,
+      "loss": 1.793,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.22568033781908964,
+      "learning_rate": 8.512373700395373e-05,
+      "loss": 1.8029,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25601857440834863,
+      "learning_rate": 8.511055791477523e-05,
+      "loss": 1.7958,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31701347224825027,
+      "learning_rate": 8.509737882559672e-05,
+      "loss": 1.8006,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2911487871418614,
+      "learning_rate": 8.508419973641821e-05,
+      "loss": 1.7955,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2082178563597066,
+      "learning_rate": 8.507102064723972e-05,
+      "loss": 1.797,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25022339272015853,
+      "learning_rate": 8.505784155806122e-05,
+      "loss": 1.7918,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2857289749098349,
+      "learning_rate": 8.504466246888271e-05,
+      "loss": 1.7919,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2982765315572785,
+      "learning_rate": 8.50314833797042e-05,
+      "loss": 1.7945,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3632832111118879,
+      "learning_rate": 8.501830429052571e-05,
+      "loss": 1.7891,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4582525250106023,
+      "learning_rate": 8.50051252013472e-05,
+      "loss": 1.7909,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49699840172624643,
+      "learning_rate": 8.49919461121687e-05,
+      "loss": 1.7971,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4504572316819271,
+      "learning_rate": 8.497876702299019e-05,
+      "loss": 1.8089,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3849459017582572,
+      "learning_rate": 8.496558793381169e-05,
+      "loss": 1.8009,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3899357827016154,
+      "learning_rate": 8.495240884463319e-05,
+      "loss": 1.7915,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3914658024096451,
+      "learning_rate": 8.493922975545469e-05,
+      "loss": 1.7883,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3709299103461753,
+      "learning_rate": 8.492605066627618e-05,
+      "loss": 1.7951,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3506925098949618,
+      "learning_rate": 8.491287157709767e-05,
+      "loss": 1.7889,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29624563115275554,
+      "learning_rate": 8.489969248791918e-05,
+      "loss": 1.7831,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25682264450924897,
+      "learning_rate": 8.488651339874066e-05,
+      "loss": 1.7874,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2460261750961129,
+      "learning_rate": 8.487333430956217e-05,
+      "loss": 1.7828,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.24758082633885842,
+      "learning_rate": 8.486015522038366e-05,
+      "loss": 1.7794,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33054734116446843,
+      "learning_rate": 8.484697613120516e-05,
+      "loss": 1.7818,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4561811736514614,
+      "learning_rate": 8.483379704202666e-05,
+      "loss": 1.7922,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49921174238421934,
+      "learning_rate": 8.482061795284816e-05,
+      "loss": 1.7798,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4554136076143871,
+      "learning_rate": 8.480743886366965e-05,
+      "loss": 1.7822,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41157410101274916,
+      "learning_rate": 8.479425977449115e-05,
+      "loss": 1.7798,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40589712484793644,
+      "learning_rate": 8.478108068531264e-05,
+      "loss": 1.7865,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41969602312949894,
+      "learning_rate": 8.476790159613413e-05,
+      "loss": 1.7846,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3716209700804159,
+      "learning_rate": 8.475472250695564e-05,
+      "loss": 1.7831,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2807622431725275,
+      "learning_rate": 8.474154341777713e-05,
+      "loss": 1.7785,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2882135171870514,
+      "learning_rate": 8.472836432859863e-05,
+      "loss": 1.7955,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33929519699606947,
+      "learning_rate": 8.471518523942012e-05,
+      "loss": 1.7821,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36786610502124734,
+      "learning_rate": 8.470200615024163e-05,
+      "loss": 1.776,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3364252349308489,
+      "learning_rate": 8.468882706106312e-05,
+      "loss": 1.7886,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2402932781772499,
+      "learning_rate": 8.467564797188462e-05,
+      "loss": 1.7841,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25361231156390895,
+      "learning_rate": 8.466246888270611e-05,
+      "loss": 1.7807,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2615629884646646,
+      "learning_rate": 8.46492897935276e-05,
+      "loss": 1.7791,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28499785308119957,
+      "learning_rate": 8.463611070434911e-05,
+      "loss": 1.7755,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3356974383461246,
+      "learning_rate": 8.46229316151706e-05,
+      "loss": 1.7711,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30749102123781535,
+      "learning_rate": 8.46097525259921e-05,
+      "loss": 1.7825,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2672389622630441,
+      "learning_rate": 8.459657343681359e-05,
+      "loss": 1.7726,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2658653893968794,
+      "learning_rate": 8.45833943476351e-05,
+      "loss": 1.7919,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28816762727415146,
+      "learning_rate": 8.457021525845658e-05,
+      "loss": 1.774,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3546252508516257,
+      "learning_rate": 8.455703616927809e-05,
+      "loss": 1.7764,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3982229064694483,
+      "learning_rate": 8.454385708009958e-05,
+      "loss": 1.7797,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39743403855878273,
+      "learning_rate": 8.453067799092107e-05,
+      "loss": 1.7818,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3565095947734143,
+      "learning_rate": 8.451749890174258e-05,
+      "loss": 1.7689,
+      "step": 416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33445086643446986,
+      "learning_rate": 8.450431981256408e-05,
+      "loss": 1.7703,
+      "step": 417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3315820482846953,
+      "learning_rate": 8.449114072338557e-05,
+      "loss": 1.7695,
+      "step": 418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28800691098886155,
+      "learning_rate": 8.447796163420706e-05,
+      "loss": 1.7735,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2801123597101435,
+      "learning_rate": 8.446478254502856e-05,
+      "loss": 1.7727,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3002418032943439,
+      "learning_rate": 8.445160345585005e-05,
+      "loss": 1.7694,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3105457712936389,
+      "learning_rate": 8.443842436667156e-05,
+      "loss": 1.7743,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31681589803189375,
+      "learning_rate": 8.442524527749305e-05,
+      "loss": 1.7754,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29540052374128717,
+      "learning_rate": 8.441206618831454e-05,
+      "loss": 1.7676,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2786507071219561,
+      "learning_rate": 8.439888709913604e-05,
+      "loss": 1.7768,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2882558823750487,
+      "learning_rate": 8.438570800995755e-05,
+      "loss": 1.7559,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3231073755842465,
+      "learning_rate": 8.437252892077904e-05,
+      "loss": 1.7718,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3009918727669168,
+      "learning_rate": 8.435934983160053e-05,
+      "loss": 1.7558,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2628726596154197,
+      "learning_rate": 8.434617074242203e-05,
+      "loss": 1.7592,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28399571790394734,
+      "learning_rate": 8.433299165324352e-05,
+      "loss": 1.7671,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25544475849265763,
+      "learning_rate": 8.431981256406503e-05,
+      "loss": 1.7663,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.23553288100790656,
+      "learning_rate": 8.430663347488652e-05,
+      "loss": 1.7739,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2775059913303605,
+      "learning_rate": 8.429345438570802e-05,
+      "loss": 1.7679,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3075973810425004,
+      "learning_rate": 8.428027529652951e-05,
+      "loss": 1.7615,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2930374223294978,
+      "learning_rate": 8.426709620735102e-05,
+      "loss": 1.7666,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2579433797280727,
+      "learning_rate": 8.42539171181725e-05,
+      "loss": 1.767,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.24664576276610475,
+      "learning_rate": 8.4240738028994e-05,
+      "loss": 1.7673,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2369430249499012,
+      "learning_rate": 8.42275589398155e-05,
+      "loss": 1.7651,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.22260512529991971,
+      "learning_rate": 8.421437985063699e-05,
+      "loss": 1.7656,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2500300140196859,
+      "learning_rate": 8.420120076145849e-05,
+      "loss": 1.7563,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2973553354592089,
+      "learning_rate": 8.418802167227999e-05,
+      "loss": 1.7686,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3473837490426601,
+      "learning_rate": 8.417484258310149e-05,
+      "loss": 1.766,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37681005189169203,
+      "learning_rate": 8.416166349392298e-05,
+      "loss": 1.7644,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43730723219972917,
+      "learning_rate": 8.414848440474447e-05,
+      "loss": 1.755,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48972509251420343,
+      "learning_rate": 8.413530531556597e-05,
+      "loss": 1.7574,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4405993959888919,
+      "learning_rate": 8.412212622638747e-05,
+      "loss": 1.7486,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3363658300532759,
+      "learning_rate": 8.410894713720897e-05,
+      "loss": 1.7578,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2582139601078949,
+      "learning_rate": 8.409576804803046e-05,
+      "loss": 1.7612,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2811740539218468,
+      "learning_rate": 8.408258895885196e-05,
+      "loss": 1.7597,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2754375286092553,
+      "learning_rate": 8.406940986967346e-05,
+      "loss": 1.7618,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2708471785317565,
+      "learning_rate": 8.405623078049496e-05,
+      "loss": 1.7566,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27721812968333487,
+      "learning_rate": 8.404305169131645e-05,
+      "loss": 1.7546,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23002196109659534,
+      "learning_rate": 8.402987260213794e-05,
+      "loss": 1.7563,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26623499924803207,
+      "learning_rate": 8.401669351295944e-05,
+      "loss": 1.7632,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31058712725330645,
+      "learning_rate": 8.400351442378095e-05,
+      "loss": 1.7554,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2565651528940601,
+      "learning_rate": 8.399033533460244e-05,
+      "loss": 1.7495,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22030881745923991,
+      "learning_rate": 8.397715624542393e-05,
+      "loss": 1.7516,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22786327950251026,
+      "learning_rate": 8.396397715624543e-05,
+      "loss": 1.7545,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22992085421490108,
+      "learning_rate": 8.395079806706693e-05,
+      "loss": 1.7524,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2745377165784387,
+      "learning_rate": 8.393761897788841e-05,
+      "loss": 1.757,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27034688441148863,
+      "learning_rate": 8.392443988870992e-05,
+      "loss": 1.7514,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21431031046245713,
+      "learning_rate": 8.391126079953142e-05,
+      "loss": 1.7516,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.24404100887840613,
+      "learning_rate": 8.389808171035291e-05,
+      "loss": 1.7446,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25123054377636167,
+      "learning_rate": 8.38849026211744e-05,
+      "loss": 1.7375,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22147497464193822,
+      "learning_rate": 8.387172353199591e-05,
+      "loss": 1.7531,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2719165207073828,
+      "learning_rate": 8.38585444428174e-05,
+      "loss": 1.7427,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3072701977020835,
+      "learning_rate": 8.38453653536389e-05,
+      "loss": 1.7554,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3465803271567775,
+      "learning_rate": 8.383218626446039e-05,
+      "loss": 1.7362,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.401606701101308,
+      "learning_rate": 8.381900717528189e-05,
+      "loss": 1.7487,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3606552145593931,
+      "learning_rate": 8.380582808610339e-05,
+      "loss": 1.7401,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31305392679932725,
+      "learning_rate": 8.379264899692489e-05,
+      "loss": 1.7426,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3283966913328143,
+      "learning_rate": 8.377946990774638e-05,
+      "loss": 1.745,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29719539854668975,
+      "learning_rate": 8.376629081856787e-05,
+      "loss": 1.7479,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2858720938289175,
+      "learning_rate": 8.375311172938938e-05,
+      "loss": 1.7471,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30869366396700376,
+      "learning_rate": 8.373993264021087e-05,
+      "loss": 1.7426,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2970917996723054,
+      "learning_rate": 8.372675355103237e-05,
+      "loss": 1.748,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31796639499576057,
+      "learning_rate": 8.371357446185386e-05,
+      "loss": 1.7482,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3755211025212181,
+      "learning_rate": 8.370039537267536e-05,
+      "loss": 1.7441,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4281842470615676,
+      "learning_rate": 8.368721628349686e-05,
+      "loss": 1.7413,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4094549002505819,
+      "learning_rate": 8.367403719431836e-05,
+      "loss": 1.7509,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3581404361289346,
+      "learning_rate": 8.366085810513985e-05,
+      "loss": 1.7416,
+      "step": 481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3304524068236931,
+      "learning_rate": 8.364767901596134e-05,
+      "loss": 1.7444,
+      "step": 482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35489265098811323,
+      "learning_rate": 8.363449992678285e-05,
+      "loss": 1.7506,
+      "step": 483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3714282179236144,
+      "learning_rate": 8.362132083760433e-05,
+      "loss": 1.745,
+      "step": 484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3586675337061257,
+      "learning_rate": 8.360814174842584e-05,
+      "loss": 1.7425,
+      "step": 485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2345396435881697,
+      "learning_rate": 8.359496265924733e-05,
+      "loss": 1.7367,
+      "step": 486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2053682387601687,
+      "learning_rate": 8.358178357006883e-05,
+      "loss": 1.7386,
+      "step": 487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2711812966223324,
+      "learning_rate": 8.356860448089032e-05,
+      "loss": 1.7337,
+      "step": 488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.258724050629809,
+      "learning_rate": 8.355542539171183e-05,
+      "loss": 1.7356,
+      "step": 489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.19885527922839485,
+      "learning_rate": 8.354224630253332e-05,
+      "loss": 1.7313,
+      "step": 490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23059280089577702,
+      "learning_rate": 8.352906721335482e-05,
+      "loss": 1.737,
+      "step": 491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2870358462685661,
+      "learning_rate": 8.351588812417631e-05,
+      "loss": 1.7349,
+      "step": 492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2633075013669381,
+      "learning_rate": 8.35027090349978e-05,
+      "loss": 1.7353,
+      "step": 493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3049199627731043,
+      "learning_rate": 8.348952994581931e-05,
+      "loss": 1.7384,
+      "step": 494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34926345039498624,
+      "learning_rate": 8.34763508566408e-05,
+      "loss": 1.7287,
+      "step": 495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31033645995550346,
+      "learning_rate": 8.34631717674623e-05,
+      "loss": 1.7386,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23708657622061127,
+      "learning_rate": 8.344999267828379e-05,
+      "loss": 1.7376,
+      "step": 497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2503644410759697,
+      "learning_rate": 8.34368135891053e-05,
+      "loss": 1.7286,
+      "step": 498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30108885609906977,
+      "learning_rate": 8.342363449992679e-05,
+      "loss": 1.7424,
+      "step": 499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3640305222721408,
+      "learning_rate": 8.341045541074829e-05,
+      "loss": 1.7365,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3862444666501204,
+      "learning_rate": 8.339727632156978e-05,
+      "loss": 1.7388,
+      "step": 501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33565159533458905,
+      "learning_rate": 8.338409723239127e-05,
+      "loss": 1.725,
+      "step": 502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.24120380275839087,
+      "learning_rate": 8.337091814321278e-05,
+      "loss": 1.7386,
+      "step": 503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2019440056977102,
+      "learning_rate": 8.335773905403427e-05,
+      "loss": 1.7364,
+      "step": 504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.19665207193807027,
+      "learning_rate": 8.334455996485577e-05,
+      "loss": 1.7384,
+      "step": 505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.20041439367425065,
+      "learning_rate": 8.333138087567726e-05,
+      "loss": 1.7298,
+      "step": 506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22759855658411884,
+      "learning_rate": 8.331820178649877e-05,
+      "loss": 1.7374,
+      "step": 507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2237449007183391,
+      "learning_rate": 8.330502269732025e-05,
+      "loss": 1.736,
+      "step": 508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.19892064401379644,
+      "learning_rate": 8.329184360814176e-05,
+      "loss": 1.7309,
+      "step": 509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21780921689097468,
+      "learning_rate": 8.327866451896325e-05,
+      "loss": 1.7247,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22223023552216575,
+      "learning_rate": 8.326548542978474e-05,
+      "loss": 1.7352,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23435078161758566,
+      "learning_rate": 8.325230634060624e-05,
+      "loss": 1.727,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25436357491660655,
+      "learning_rate": 8.323912725142775e-05,
+      "loss": 1.7182,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2542313172961627,
+      "learning_rate": 8.322594816224924e-05,
+      "loss": 1.7259,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25353563544590524,
+      "learning_rate": 8.321276907307073e-05,
+      "loss": 1.7131,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25178826296170725,
+      "learning_rate": 8.319958998389223e-05,
+      "loss": 1.7303,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24967442898411737,
+      "learning_rate": 8.318641089471372e-05,
+      "loss": 1.7156,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29093688012404956,
+      "learning_rate": 8.317323180553523e-05,
+      "loss": 1.7273,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3408463856901383,
+      "learning_rate": 8.316005271635672e-05,
+      "loss": 1.731,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3654511204209073,
+      "learning_rate": 8.314687362717822e-05,
+      "loss": 1.732,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3271746533985564,
+      "learning_rate": 8.313369453799971e-05,
+      "loss": 1.7295,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2717225234340834,
+      "learning_rate": 8.312051544882122e-05,
+      "loss": 1.7325,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3081217924018755,
+      "learning_rate": 8.310733635964271e-05,
+      "loss": 1.7246,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3527075080404129,
+      "learning_rate": 8.30941572704642e-05,
+      "loss": 1.7216,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4154842597521178,
+      "learning_rate": 8.30809781812857e-05,
+      "loss": 1.7255,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.443067725739395,
+      "learning_rate": 8.306779909210719e-05,
+      "loss": 1.7281,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4106745425675618,
+      "learning_rate": 8.305462000292868e-05,
+      "loss": 1.7302,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33999747867070007,
+      "learning_rate": 8.304144091375019e-05,
+      "loss": 1.7267,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28901846740062775,
+      "learning_rate": 8.302826182457169e-05,
+      "loss": 1.7262,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29966915847542663,
+      "learning_rate": 8.301508273539318e-05,
+      "loss": 1.7313,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31328357913945787,
+      "learning_rate": 8.300190364621469e-05,
+      "loss": 1.7218,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2814137310527097,
+      "learning_rate": 8.298872455703617e-05,
+      "loss": 1.7215,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24123503809014038,
+      "learning_rate": 8.297554546785767e-05,
+      "loss": 1.7188,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28016429302641516,
+      "learning_rate": 8.296236637867917e-05,
+      "loss": 1.7198,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3117115191862107,
+      "learning_rate": 8.294918728950066e-05,
+      "loss": 1.7262,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.278045134177961,
+      "learning_rate": 8.293600820032216e-05,
+      "loss": 1.723,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23769710910310898,
+      "learning_rate": 8.292282911114366e-05,
+      "loss": 1.723,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24511433211495529,
+      "learning_rate": 8.290965002196516e-05,
+      "loss": 1.7253,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2384929581877857,
+      "learning_rate": 8.289647093278665e-05,
+      "loss": 1.7122,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.22551550594015082,
+      "learning_rate": 8.288329184360814e-05,
+      "loss": 1.7219,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23506955783092845,
+      "learning_rate": 8.287011275442964e-05,
+      "loss": 1.7233,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21860310689837917,
+      "learning_rate": 8.285693366525115e-05,
+      "loss": 1.7221,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.20398647046618187,
+      "learning_rate": 8.284375457607264e-05,
+      "loss": 1.7213,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24633014333201053,
+      "learning_rate": 8.283057548689413e-05,
+      "loss": 1.7132,
+      "step": 544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2667137895735778,
+      "learning_rate": 8.281739639771563e-05,
+      "loss": 1.726,
+      "step": 545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21858001204347083,
+      "learning_rate": 8.280421730853713e-05,
+      "loss": 1.7112,
+      "step": 546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.19721106291794568,
+      "learning_rate": 8.279103821935863e-05,
+      "loss": 1.722,
+      "step": 547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2051408273401384,
+      "learning_rate": 8.277785913018012e-05,
+      "loss": 1.7206,
+      "step": 548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21682627662474413,
+      "learning_rate": 8.276468004100161e-05,
+      "loss": 1.7151,
+      "step": 549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2438456543254191,
+      "learning_rate": 8.275150095182311e-05,
+      "loss": 1.711,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2629415416448845,
+      "learning_rate": 8.27383218626446e-05,
+      "loss": 1.7155,
+      "step": 551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29381087938708167,
+      "learning_rate": 8.272514277346611e-05,
+      "loss": 1.7184,
+      "step": 552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3068644717057271,
+      "learning_rate": 8.27119636842876e-05,
+      "loss": 1.7071,
+      "step": 553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3266717891225327,
+      "learning_rate": 8.26987845951091e-05,
+      "loss": 1.7144,
+      "step": 554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3342256318259162,
+      "learning_rate": 8.26856055059306e-05,
+      "loss": 1.7132,
+      "step": 555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3507430866157419,
+      "learning_rate": 8.267242641675208e-05,
+      "loss": 1.7222,
+      "step": 556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35401841712709037,
+      "learning_rate": 8.265924732757359e-05,
+      "loss": 1.7163,
+      "step": 557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29152885792134353,
+      "learning_rate": 8.264606823839509e-05,
+      "loss": 1.7021,
+      "step": 558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2321802057678123,
+      "learning_rate": 8.263288914921658e-05,
+      "loss": 1.715,
+      "step": 559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.20521573981201763,
+      "learning_rate": 8.261971006003807e-05,
+      "loss": 1.7136,
+      "step": 560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.22389843673660376,
+      "learning_rate": 8.260653097085958e-05,
+      "loss": 1.7145,
+      "step": 561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26769790258636306,
+      "learning_rate": 8.259335188168107e-05,
+      "loss": 1.7135,
+      "step": 562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.263472778250681,
+      "learning_rate": 8.258017279250257e-05,
+      "loss": 1.7153,
+      "step": 563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.20438761322969645,
+      "learning_rate": 8.256699370332406e-05,
+      "loss": 1.7053,
+      "step": 564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2711791794052338,
+      "learning_rate": 8.255381461414556e-05,
+      "loss": 1.7076,
+      "step": 565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29324151826012396,
+      "learning_rate": 8.254063552496706e-05,
+      "loss": 1.7162,
+      "step": 566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23680109722780296,
+      "learning_rate": 8.252745643578856e-05,
+      "loss": 1.7173,
+      "step": 567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3897963836334016,
+      "learning_rate": 8.251427734661005e-05,
+      "loss": 1.709,
+      "step": 568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3107639335861359,
+      "learning_rate": 8.250109825743154e-05,
+      "loss": 1.7108,
+      "step": 569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31507837210091977,
+      "learning_rate": 8.248791916825305e-05,
+      "loss": 1.7177,
+      "step": 570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.407833593775017,
+      "learning_rate": 8.247474007907455e-05,
+      "loss": 1.7158,
+      "step": 571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33475655675893035,
+      "learning_rate": 8.246156098989604e-05,
+      "loss": 1.7159,
+      "step": 572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30470045477617375,
+      "learning_rate": 8.244838190071753e-05,
+      "loss": 1.7133,
+      "step": 573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2361385908228469,
+      "learning_rate": 8.243520281153903e-05,
+      "loss": 1.7138,
+      "step": 574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3028517511605204,
+      "learning_rate": 8.242202372236052e-05,
+      "loss": 1.7061,
+      "step": 575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26913911701533544,
+      "learning_rate": 8.240884463318203e-05,
+      "loss": 1.7081,
+      "step": 576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24779044758095875,
+      "learning_rate": 8.239566554400352e-05,
+      "loss": 1.708,
+      "step": 577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2233287796665562,
+      "learning_rate": 8.238248645482501e-05,
+      "loss": 1.7112,
+      "step": 578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2712079799980342,
+      "learning_rate": 8.236930736564652e-05,
+      "loss": 1.7124,
+      "step": 579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2094942014575434,
+      "learning_rate": 8.2356128276468e-05,
+      "loss": 1.7054,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.21486524279150165,
+      "learning_rate": 8.234294918728951e-05,
+      "loss": 1.7119,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2090205889232179,
+      "learning_rate": 8.2329770098111e-05,
+      "loss": 1.7059,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.20460593187920675,
+      "learning_rate": 8.23165910089325e-05,
+      "loss": 1.7066,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1767845416143033,
+      "learning_rate": 8.230341191975399e-05,
+      "loss": 1.7024,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.17887391851705126,
+      "learning_rate": 8.22902328305755e-05,
+      "loss": 1.7067,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1695109544582885,
+      "learning_rate": 8.227705374139699e-05,
+      "loss": 1.6895,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2083349170582134,
+      "learning_rate": 8.226387465221849e-05,
+      "loss": 1.6915,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.217567810379028,
+      "learning_rate": 8.225069556303998e-05,
+      "loss": 1.7014,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26578168421783505,
+      "learning_rate": 8.223751647386147e-05,
+      "loss": 1.6959,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26759525206680784,
+      "learning_rate": 8.222433738468297e-05,
+      "loss": 1.7081,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29241320421717,
+      "learning_rate": 8.221115829550447e-05,
+      "loss": 1.7,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2665660324658676,
+      "learning_rate": 8.219797920632597e-05,
+      "loss": 1.7116,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2543632897483254,
+      "learning_rate": 8.218480011714746e-05,
+      "loss": 1.6998,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24422324009786234,
+      "learning_rate": 8.217162102796897e-05,
+      "loss": 1.7032,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2412157212895076,
+      "learning_rate": 8.215844193879046e-05,
+      "loss": 1.7002,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22950526573401658,
+      "learning_rate": 8.214526284961196e-05,
+      "loss": 1.6968,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2405536739297689,
+      "learning_rate": 8.213208376043345e-05,
+      "loss": 1.7067,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22384819760583757,
+      "learning_rate": 8.211890467125494e-05,
+      "loss": 1.7002,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22165394505798217,
+      "learning_rate": 8.210572558207644e-05,
+      "loss": 1.7053,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22325001055054616,
+      "learning_rate": 8.209254649289794e-05,
+      "loss": 1.7076,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24437917178401694,
+      "learning_rate": 8.207936740371944e-05,
+      "loss": 1.6863,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24473623901397126,
+      "learning_rate": 8.206618831454093e-05,
+      "loss": 1.7029,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23793249982285095,
+      "learning_rate": 8.205300922536244e-05,
+      "loss": 1.7003,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27010899475555333,
+      "learning_rate": 8.203983013618392e-05,
+      "loss": 1.7035,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2974971413903659,
+      "learning_rate": 8.202665104700543e-05,
+      "loss": 1.6963,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3391211533028402,
+      "learning_rate": 8.201347195782692e-05,
+      "loss": 1.699,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.401887788096081,
+      "learning_rate": 8.200029286864841e-05,
+      "loss": 1.7001,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4014157913865194,
+      "learning_rate": 8.198711377946991e-05,
+      "loss": 1.7035,
+      "step": 608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34227601583187883,
+      "learning_rate": 8.197393469029142e-05,
+      "loss": 1.7031,
+      "step": 609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2572596615978446,
+      "learning_rate": 8.196075560111291e-05,
+      "loss": 1.691,
+      "step": 610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25436898174940326,
+      "learning_rate": 8.19475765119344e-05,
+      "loss": 1.6983,
+      "step": 611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2966388678401186,
+      "learning_rate": 8.19343974227559e-05,
+      "loss": 1.6939,
+      "step": 612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2609270973789931,
+      "learning_rate": 8.192121833357739e-05,
+      "loss": 1.6882,
+      "step": 613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22900064576600393,
+      "learning_rate": 8.190803924439888e-05,
+      "loss": 1.6962,
+      "step": 614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2157661551046652,
+      "learning_rate": 8.189486015522039e-05,
+      "loss": 1.6889,
+      "step": 615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23909710323754044,
+      "learning_rate": 8.188168106604189e-05,
+      "loss": 1.6969,
+      "step": 616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25769486284072124,
+      "learning_rate": 8.186850197686338e-05,
+      "loss": 1.699,
+      "step": 617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23459236703850359,
+      "learning_rate": 8.185532288768489e-05,
+      "loss": 1.6904,
+      "step": 618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23197840287816318,
+      "learning_rate": 8.184214379850638e-05,
+      "loss": 1.6927,
+      "step": 619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22922928506230492,
+      "learning_rate": 8.182896470932787e-05,
+      "loss": 1.691,
+      "step": 620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.20472350901775296,
+      "learning_rate": 8.181578562014937e-05,
+      "loss": 1.6988,
+      "step": 621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.21293532425438386,
+      "learning_rate": 8.180260653097086e-05,
+      "loss": 1.6983,
+      "step": 622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25813129719442796,
+      "learning_rate": 8.178942744179236e-05,
+      "loss": 1.6977,
+      "step": 623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28731586003331094,
+      "learning_rate": 8.177624835261386e-05,
+      "loss": 1.6917,
+      "step": 624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27778976830351854,
+      "learning_rate": 8.176306926343536e-05,
+      "loss": 1.6904,
+      "step": 625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2667191282626614,
+      "learning_rate": 8.174989017425685e-05,
+      "loss": 1.6906,
+      "step": 626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27135549254051117,
+      "learning_rate": 8.173671108507834e-05,
+      "loss": 1.7077,
+      "step": 627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2737735487643415,
+      "learning_rate": 8.172353199589984e-05,
+      "loss": 1.6919,
+      "step": 628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23702108817197906,
+      "learning_rate": 8.171035290672134e-05,
+      "loss": 1.6833,
+      "step": 629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.21824459389931547,
+      "learning_rate": 8.169717381754284e-05,
+      "loss": 1.6874,
+      "step": 630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2551514671888147,
+      "learning_rate": 8.168399472836433e-05,
+      "loss": 1.6964,
+      "step": 631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2888406849522713,
+      "learning_rate": 8.167081563918583e-05,
+      "loss": 1.686,
+      "step": 632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3043373169879388,
+      "learning_rate": 8.165763655000733e-05,
+      "loss": 1.6909,
+      "step": 633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31851826485165624,
+      "learning_rate": 8.164445746082883e-05,
+      "loss": 1.6955,
+      "step": 634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31677231407393813,
+      "learning_rate": 8.163127837165032e-05,
+      "loss": 1.6931,
+      "step": 635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3235809995826595,
+      "learning_rate": 8.161809928247181e-05,
+      "loss": 1.6919,
+      "step": 636
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31715316802417376,
+      "learning_rate": 8.160492019329331e-05,
+      "loss": 1.6972,
+      "step": 637
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.258710234201555,
+      "learning_rate": 8.15917411041148e-05,
+      "loss": 1.6908,
+      "step": 638
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.19364210951795302,
+      "learning_rate": 8.157856201493631e-05,
+      "loss": 1.6893,
+      "step": 639
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1934820062184886,
+      "learning_rate": 8.15653829257578e-05,
+      "loss": 1.6937,
+      "step": 640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22124564568502275,
+      "learning_rate": 8.15522038365793e-05,
+      "loss": 1.6857,
+      "step": 641
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24527757345647355,
+      "learning_rate": 8.15390247474008e-05,
+      "loss": 1.6964,
+      "step": 642
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31637585588010625,
+      "learning_rate": 8.152584565822228e-05,
+      "loss": 1.6911,
+      "step": 643
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33231300289557464,
+      "learning_rate": 8.151266656904379e-05,
+      "loss": 1.6857,
+      "step": 644
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2979233182409515,
+      "learning_rate": 8.149948747986529e-05,
+      "loss": 1.6907,
+      "step": 645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2637032106008522,
+      "learning_rate": 8.148630839068678e-05,
+      "loss": 1.6922,
+      "step": 646
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2729619019235736,
+      "learning_rate": 8.147312930150827e-05,
+      "loss": 1.6919,
+      "step": 647
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3056180656791934,
+      "learning_rate": 8.145995021232978e-05,
+      "loss": 1.6866,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3025381384222673,
+      "learning_rate": 8.144677112315127e-05,
+      "loss": 1.6895,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26847771781612867,
+      "learning_rate": 8.143359203397277e-05,
+      "loss": 1.6893,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2438844041255715,
+      "learning_rate": 8.142041294479426e-05,
+      "loss": 1.6777,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29212647337890685,
+      "learning_rate": 8.140723385561575e-05,
+      "loss": 1.6962,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3279764517256754,
+      "learning_rate": 8.139405476643726e-05,
+      "loss": 1.686,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3266201901472985,
+      "learning_rate": 8.138087567725876e-05,
+      "loss": 1.6861,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29731852257716146,
+      "learning_rate": 8.136769658808025e-05,
+      "loss": 1.684,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31099064594669873,
+      "learning_rate": 8.135451749890174e-05,
+      "loss": 1.6842,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3060576588797521,
+      "learning_rate": 8.134133840972325e-05,
+      "loss": 1.6891,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26004140101304163,
+      "learning_rate": 8.132815932054474e-05,
+      "loss": 1.6854,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.268187676963691,
+      "learning_rate": 8.131498023136624e-05,
+      "loss": 1.6838,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2726005198375338,
+      "learning_rate": 8.130180114218773e-05,
+      "loss": 1.692,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23864901844548525,
+      "learning_rate": 8.128862205300923e-05,
+      "loss": 1.6849,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2244877830433347,
+      "learning_rate": 8.127544296383072e-05,
+      "loss": 1.6806,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21510789414916953,
+      "learning_rate": 8.126226387465223e-05,
+      "loss": 1.6835,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.207451826116181,
+      "learning_rate": 8.124908478547372e-05,
+      "loss": 1.6731,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2410531205632606,
+      "learning_rate": 8.123590569629521e-05,
+      "loss": 1.679,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2658490081241801,
+      "learning_rate": 8.122272660711672e-05,
+      "loss": 1.683,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24153141494391162,
+      "learning_rate": 8.12095475179382e-05,
+      "loss": 1.6738,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23164870510130753,
+      "learning_rate": 8.119636842875971e-05,
+      "loss": 1.6867,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24233480776179323,
+      "learning_rate": 8.11831893395812e-05,
+      "loss": 1.6773,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2574837382310888,
+      "learning_rate": 8.11700102504027e-05,
+      "loss": 1.6844,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22126722703874208,
+      "learning_rate": 8.115683116122419e-05,
+      "loss": 1.6869,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17423158084927914,
+      "learning_rate": 8.11436520720457e-05,
+      "loss": 1.6826,
+      "step": 672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2011587498434974,
+      "learning_rate": 8.113047298286719e-05,
+      "loss": 1.6932,
+      "step": 673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23419572476137138,
+      "learning_rate": 8.111729389368869e-05,
+      "loss": 1.681,
+      "step": 674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21861982894310164,
+      "learning_rate": 8.110411480451018e-05,
+      "loss": 1.6845,
+      "step": 675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2020605297807059,
+      "learning_rate": 8.109093571533167e-05,
+      "loss": 1.6758,
+      "step": 676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.200339221622581,
+      "learning_rate": 8.107775662615317e-05,
+      "loss": 1.6852,
+      "step": 677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2272004211020317,
+      "learning_rate": 8.106457753697467e-05,
+      "loss": 1.6754,
+      "step": 678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2639656585184034,
+      "learning_rate": 8.105139844779617e-05,
+      "loss": 1.6651,
+      "step": 679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2821270304501907,
+      "learning_rate": 8.103821935861766e-05,
+      "loss": 1.6727,
+      "step": 680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2385914002979489,
+      "learning_rate": 8.102504026943917e-05,
+      "loss": 1.6757,
+      "step": 681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.19689463498619292,
+      "learning_rate": 8.101186118026066e-05,
+      "loss": 1.675,
+      "step": 682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.19388551672385698,
+      "learning_rate": 8.099868209108216e-05,
+      "loss": 1.6865,
+      "step": 683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21717430076929156,
+      "learning_rate": 8.098550300190365e-05,
+      "loss": 1.6786,
+      "step": 684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21911631079722832,
+      "learning_rate": 8.097232391272514e-05,
+      "loss": 1.677,
+      "step": 685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21899413808647583,
+      "learning_rate": 8.095914482354664e-05,
+      "loss": 1.6695,
+      "step": 686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.256376958131802,
+      "learning_rate": 8.094596573436814e-05,
+      "loss": 1.681,
+      "step": 687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30968776120326585,
+      "learning_rate": 8.093278664518964e-05,
+      "loss": 1.6719,
+      "step": 688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3405179169627675,
+      "learning_rate": 8.091960755601113e-05,
+      "loss": 1.6729,
+      "step": 689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3758025068653412,
+      "learning_rate": 8.090642846683264e-05,
+      "loss": 1.6767,
+      "step": 690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37947974517090455,
+      "learning_rate": 8.089324937765412e-05,
+      "loss": 1.6722,
+      "step": 691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3209320177396622,
+      "learning_rate": 8.088007028847563e-05,
+      "loss": 1.6743,
+      "step": 692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2686562575879156,
+      "learning_rate": 8.086689119929712e-05,
+      "loss": 1.6722,
+      "step": 693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24152315432813073,
+      "learning_rate": 8.085371211011861e-05,
+      "loss": 1.6817,
+      "step": 694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25903562203260866,
+      "learning_rate": 8.084053302094011e-05,
+      "loss": 1.6808,
+      "step": 695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27215275116010934,
+      "learning_rate": 8.082735393176162e-05,
+      "loss": 1.6735,
+      "step": 696
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2665073871904065,
+      "learning_rate": 8.081417484258311e-05,
+      "loss": 1.6813,
+      "step": 697
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2595685011638328,
+      "learning_rate": 8.08009957534046e-05,
+      "loss": 1.6813,
+      "step": 698
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2584829914241195,
+      "learning_rate": 8.07878166642261e-05,
+      "loss": 1.6699,
+      "step": 699
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23308896936839194,
+      "learning_rate": 8.077463757504759e-05,
+      "loss": 1.6686,
+      "step": 700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18262768380478986,
+      "learning_rate": 8.076145848586908e-05,
+      "loss": 1.6727,
+      "step": 701
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17402802326250397,
+      "learning_rate": 8.074827939669059e-05,
+      "loss": 1.671,
+      "step": 702
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2079124720731814,
+      "learning_rate": 8.073510030751208e-05,
+      "loss": 1.6718,
+      "step": 703
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21030626632342236,
+      "learning_rate": 8.072192121833358e-05,
+      "loss": 1.675,
+      "step": 704
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.19979944547574685,
+      "learning_rate": 8.070874212915509e-05,
+      "loss": 1.6719,
+      "step": 705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1960619390208134,
+      "learning_rate": 8.069556303997658e-05,
+      "loss": 1.6729,
+      "step": 706
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1800588887952547,
+      "learning_rate": 8.068238395079807e-05,
+      "loss": 1.6742,
+      "step": 707
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.171338131397515,
+      "learning_rate": 8.066920486161957e-05,
+      "loss": 1.6681,
+      "step": 708
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1952527594313872,
+      "learning_rate": 8.065602577244106e-05,
+      "loss": 1.672,
+      "step": 709
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2029111128206527,
+      "learning_rate": 8.064284668326255e-05,
+      "loss": 1.6729,
+      "step": 710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2093511465528856,
+      "learning_rate": 8.062966759408406e-05,
+      "loss": 1.6662,
+      "step": 711
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2355948665280437,
+      "learning_rate": 8.061648850490556e-05,
+      "loss": 1.6691,
+      "step": 712
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25612769652057044,
+      "learning_rate": 8.060330941572705e-05,
+      "loss": 1.6691,
+      "step": 713
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3178891893081287,
+      "learning_rate": 8.059013032654856e-05,
+      "loss": 1.6702,
+      "step": 714
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32928109757980245,
+      "learning_rate": 8.057695123737004e-05,
+      "loss": 1.6804,
+      "step": 715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2732800407347181,
+      "learning_rate": 8.056377214819154e-05,
+      "loss": 1.6799,
+      "step": 716
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24062058611392861,
+      "learning_rate": 8.055059305901304e-05,
+      "loss": 1.6678,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3002907368750325,
+      "learning_rate": 8.053741396983453e-05,
+      "loss": 1.6672,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31553599792720294,
+      "learning_rate": 8.052423488065603e-05,
+      "loss": 1.671,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25035709769683784,
+      "learning_rate": 8.051105579147753e-05,
+      "loss": 1.6746,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16571551955558791,
+      "learning_rate": 8.049787670229903e-05,
+      "loss": 1.6619,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21454683239201158,
+      "learning_rate": 8.048469761312052e-05,
+      "loss": 1.6686,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2310455452470971,
+      "learning_rate": 8.047151852394201e-05,
+      "loss": 1.6771,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1994236882273046,
+      "learning_rate": 8.045833943476351e-05,
+      "loss": 1.671,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18974389362814953,
+      "learning_rate": 8.0445160345585e-05,
+      "loss": 1.6675,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1860500322260672,
+      "learning_rate": 8.043198125640651e-05,
+      "loss": 1.658,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18863809672549492,
+      "learning_rate": 8.0418802167228e-05,
+      "loss": 1.6692,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19261227728615138,
+      "learning_rate": 8.04056230780495e-05,
+      "loss": 1.6722,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19104427920831676,
+      "learning_rate": 8.0392443988871e-05,
+      "loss": 1.6682,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19351874126385604,
+      "learning_rate": 8.03792648996925e-05,
+      "loss": 1.6698,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18561279955951115,
+      "learning_rate": 8.036608581051399e-05,
+      "loss": 1.6587,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20733342786735096,
+      "learning_rate": 8.035290672133548e-05,
+      "loss": 1.6686,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24389803201789031,
+      "learning_rate": 8.033972763215698e-05,
+      "loss": 1.6699,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3303044019383874,
+      "learning_rate": 8.032654854297847e-05,
+      "loss": 1.6685,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4326972184756624,
+      "learning_rate": 8.031336945379998e-05,
+      "loss": 1.671,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4749815372235629,
+      "learning_rate": 8.030019036462147e-05,
+      "loss": 1.6732,
+      "step": 736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4074470713549075,
+      "learning_rate": 8.028701127544297e-05,
+      "loss": 1.6627,
+      "step": 737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2325050031860467,
+      "learning_rate": 8.027383218626447e-05,
+      "loss": 1.6624,
+      "step": 738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26482099963107963,
+      "learning_rate": 8.026065309708595e-05,
+      "loss": 1.6714,
+      "step": 739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36679572880941774,
+      "learning_rate": 8.024747400790746e-05,
+      "loss": 1.6704,
+      "step": 740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26232767280919794,
+      "learning_rate": 8.023429491872896e-05,
+      "loss": 1.663,
+      "step": 741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21445846555925532,
+      "learning_rate": 8.022111582955045e-05,
+      "loss": 1.6525,
+      "step": 742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2532638225714052,
+      "learning_rate": 8.020793674037194e-05,
+      "loss": 1.6642,
+      "step": 743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2042942070651275,
+      "learning_rate": 8.019475765119345e-05,
+      "loss": 1.6598,
+      "step": 744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.171722618385623,
+      "learning_rate": 8.018157856201494e-05,
+      "loss": 1.6588,
+      "step": 745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1858922147873516,
+      "learning_rate": 8.016839947283644e-05,
+      "loss": 1.6644,
+      "step": 746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1800446333017018,
+      "learning_rate": 8.015522038365793e-05,
+      "loss": 1.6596,
+      "step": 747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1803808001678191,
+      "learning_rate": 8.014204129447943e-05,
+      "loss": 1.672,
+      "step": 748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1768537234586101,
+      "learning_rate": 8.012886220530092e-05,
+      "loss": 1.6657,
+      "step": 749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19980216971257497,
+      "learning_rate": 8.011568311612243e-05,
+      "loss": 1.6534,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1832264261976631,
+      "learning_rate": 8.010250402694392e-05,
+      "loss": 1.6591,
+      "step": 751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16513639779797532,
+      "learning_rate": 8.008932493776541e-05,
+      "loss": 1.6648,
+      "step": 752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17022762074604036,
+      "learning_rate": 8.007614584858692e-05,
+      "loss": 1.6591,
+      "step": 753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1681936250352472,
+      "learning_rate": 8.006296675940841e-05,
+      "loss": 1.6571,
+      "step": 754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21312758323221945,
+      "learning_rate": 8.004978767022991e-05,
+      "loss": 1.6562,
+      "step": 755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2523673118943729,
+      "learning_rate": 8.00366085810514e-05,
+      "loss": 1.6663,
+      "step": 756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2623904537775464,
+      "learning_rate": 8.00234294918729e-05,
+      "loss": 1.6674,
+      "step": 757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24341343434762125,
+      "learning_rate": 8.001025040269439e-05,
+      "loss": 1.6713,
+      "step": 758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.217043521795163,
+      "learning_rate": 7.99970713135159e-05,
+      "loss": 1.6706,
+      "step": 759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20902151519730236,
+      "learning_rate": 7.998389222433739e-05,
+      "loss": 1.664,
+      "step": 760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22505288309338028,
+      "learning_rate": 7.997071313515888e-05,
+      "loss": 1.6588,
+      "step": 761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22608418892993373,
+      "learning_rate": 7.995753404598039e-05,
+      "loss": 1.6617,
+      "step": 762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2204380979289179,
+      "learning_rate": 7.994435495680187e-05,
+      "loss": 1.6573,
+      "step": 763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1969471585321503,
+      "learning_rate": 7.993117586762337e-05,
+      "loss": 1.6601,
+      "step": 764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20082564014850096,
+      "learning_rate": 7.991799677844487e-05,
+      "loss": 1.6617,
+      "step": 765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20192437570777552,
+      "learning_rate": 7.990481768926637e-05,
+      "loss": 1.6544,
+      "step": 766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2574830285462354,
+      "learning_rate": 7.989163860008786e-05,
+      "loss": 1.6533,
+      "step": 767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23002776136831127,
+      "learning_rate": 7.987845951090937e-05,
+      "loss": 1.661,
+      "step": 768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22713613412496464,
+      "learning_rate": 7.986528042173086e-05,
+      "loss": 1.6663,
+      "step": 769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2042318478702029,
+      "learning_rate": 7.985210133255236e-05,
+      "loss": 1.6546,
+      "step": 770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20073586989829498,
+      "learning_rate": 7.983892224337385e-05,
+      "loss": 1.6591,
+      "step": 771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22156482786667436,
+      "learning_rate": 7.982574315419534e-05,
+      "loss": 1.6559,
+      "step": 772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23204275860685752,
+      "learning_rate": 7.981256406501684e-05,
+      "loss": 1.654,
+      "step": 773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22578907014380378,
+      "learning_rate": 7.979938497583834e-05,
+      "loss": 1.6598,
+      "step": 774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22278741960348777,
+      "learning_rate": 7.978620588665984e-05,
+      "loss": 1.6521,
+      "step": 775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21685600767937083,
+      "learning_rate": 7.977302679748133e-05,
+      "loss": 1.6607,
+      "step": 776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22266066710897367,
+      "learning_rate": 7.975984770830284e-05,
+      "loss": 1.6487,
+      "step": 777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19632860059659135,
+      "learning_rate": 7.974666861912433e-05,
+      "loss": 1.6613,
+      "step": 778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18234394704028628,
+      "learning_rate": 7.973348952994583e-05,
+      "loss": 1.6574,
+      "step": 779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.177598112664479,
+      "learning_rate": 7.972031044076732e-05,
+      "loss": 1.6529,
+      "step": 780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.19095741463896163,
+      "learning_rate": 7.970713135158881e-05,
+      "loss": 1.654,
+      "step": 781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23354657913440505,
+      "learning_rate": 7.969395226241031e-05,
+      "loss": 1.6568,
+      "step": 782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23167577068773595,
+      "learning_rate": 7.968077317323181e-05,
+      "loss": 1.6554,
+      "step": 783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21367008980339683,
+      "learning_rate": 7.966759408405331e-05,
+      "loss": 1.6495,
+      "step": 784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2207178544949823,
+      "learning_rate": 7.96544149948748e-05,
+      "loss": 1.6548,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21891851522878888,
+      "learning_rate": 7.964123590569631e-05,
+      "loss": 1.657,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24695814979029723,
+      "learning_rate": 7.962805681651779e-05,
+      "loss": 1.6488,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.287912034486528,
+      "learning_rate": 7.961487772733928e-05,
+      "loss": 1.65,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3399743120660956,
+      "learning_rate": 7.960169863816079e-05,
+      "loss": 1.6526,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3616996652020229,
+      "learning_rate": 7.958851954898228e-05,
+      "loss": 1.6602,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3063137729348896,
+      "learning_rate": 7.957534045980378e-05,
+      "loss": 1.6573,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24333897069393048,
+      "learning_rate": 7.956216137062529e-05,
+      "loss": 1.6517,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2536956392375361,
+      "learning_rate": 7.954898228144678e-05,
+      "loss": 1.6548,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2992558453991358,
+      "learning_rate": 7.953580319226827e-05,
+      "loss": 1.6476,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2919938016387346,
+      "learning_rate": 7.952262410308977e-05,
+      "loss": 1.658,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22425446070873226,
+      "learning_rate": 7.950944501391126e-05,
+      "loss": 1.6561,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17979423947699205,
+      "learning_rate": 7.949626592473275e-05,
+      "loss": 1.6511,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2003220647484553,
+      "learning_rate": 7.948308683555426e-05,
+      "loss": 1.6584,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21655839117724143,
+      "learning_rate": 7.946990774637576e-05,
+      "loss": 1.6413,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.19654884015176893,
+      "learning_rate": 7.945672865719725e-05,
+      "loss": 1.6478,
+      "step": 800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.20643622425199437,
+      "learning_rate": 7.944354956801876e-05,
+      "loss": 1.6487,
+      "step": 801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2210426562771846,
+      "learning_rate": 7.943037047884025e-05,
+      "loss": 1.6417,
+      "step": 802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2370399462767824,
+      "learning_rate": 7.941719138966174e-05,
+      "loss": 1.6526,
+      "step": 803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27064884564726144,
+      "learning_rate": 7.940401230048324e-05,
+      "loss": 1.6452,
+      "step": 804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24922057360329597,
+      "learning_rate": 7.939083321130473e-05,
+      "loss": 1.6476,
+      "step": 805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22578777752481574,
+      "learning_rate": 7.937765412212622e-05,
+      "loss": 1.6512,
+      "step": 806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2450981377837511,
+      "learning_rate": 7.936447503294773e-05,
+      "loss": 1.6447,
+      "step": 807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.274010641548682,
+      "learning_rate": 7.935129594376923e-05,
+      "loss": 1.6559,
+      "step": 808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25880479404705736,
+      "learning_rate": 7.933811685459072e-05,
+      "loss": 1.6489,
+      "step": 809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22048806582617783,
+      "learning_rate": 7.932493776541223e-05,
+      "loss": 1.6524,
+      "step": 810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1887738283254872,
+      "learning_rate": 7.931175867623371e-05,
+      "loss": 1.64,
+      "step": 811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18522517056579318,
+      "learning_rate": 7.92985795870552e-05,
+      "loss": 1.6494,
+      "step": 812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22546381379315691,
+      "learning_rate": 7.928540049787671e-05,
+      "loss": 1.6464,
+      "step": 813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2506628844421445,
+      "learning_rate": 7.92722214086982e-05,
+      "loss": 1.6542,
+      "step": 814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23722684616063827,
+      "learning_rate": 7.92590423195197e-05,
+      "loss": 1.6462,
+      "step": 815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21527026226593318,
+      "learning_rate": 7.92458632303412e-05,
+      "loss": 1.6429,
+      "step": 816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.20417560769017126,
+      "learning_rate": 7.92326841411627e-05,
+      "loss": 1.6463,
+      "step": 817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25657787896551143,
+      "learning_rate": 7.921950505198419e-05,
+      "loss": 1.655,
+      "step": 818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.267753342657363,
+      "learning_rate": 7.920632596280568e-05,
+      "loss": 1.6519,
+      "step": 819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24920909128098695,
+      "learning_rate": 7.919314687362718e-05,
+      "loss": 1.6473,
+      "step": 820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2478367552587059,
+      "learning_rate": 7.917996778444867e-05,
+      "loss": 1.6426,
+      "step": 821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2153066175677768,
+      "learning_rate": 7.916678869527018e-05,
+      "loss": 1.6491,
+      "step": 822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2014632227972863,
+      "learning_rate": 7.915360960609167e-05,
+      "loss": 1.6505,
+      "step": 823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2444739546695772,
+      "learning_rate": 7.914043051691317e-05,
+      "loss": 1.6481,
+      "step": 824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24448366789416284,
+      "learning_rate": 7.912725142773467e-05,
+      "loss": 1.6536,
+      "step": 825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18988494902818237,
+      "learning_rate": 7.911407233855617e-05,
+      "loss": 1.6527,
+      "step": 826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1596982906421288,
+      "learning_rate": 7.910089324937766e-05,
+      "loss": 1.6434,
+      "step": 827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18102812476693692,
+      "learning_rate": 7.908771416019915e-05,
+      "loss": 1.6518,
+      "step": 828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2102523140397989,
+      "learning_rate": 7.907453507102065e-05,
+      "loss": 1.6498,
+      "step": 829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23412540024706124,
+      "learning_rate": 7.906135598184214e-05,
+      "loss": 1.641,
+      "step": 830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24005571074787757,
+      "learning_rate": 7.904817689266365e-05,
+      "loss": 1.6431,
+      "step": 831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23238476130237243,
+      "learning_rate": 7.903499780348514e-05,
+      "loss": 1.6413,
+      "step": 832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1970713398858913,
+      "learning_rate": 7.902181871430664e-05,
+      "loss": 1.6423,
+      "step": 833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2242200272612161,
+      "learning_rate": 7.900863962512814e-05,
+      "loss": 1.6517,
+      "step": 834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22807610187834748,
+      "learning_rate": 7.899546053594962e-05,
+      "loss": 1.6406,
+      "step": 835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1972521968320005,
+      "learning_rate": 7.898228144677112e-05,
+      "loss": 1.6507,
+      "step": 836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23394212314333337,
+      "learning_rate": 7.896910235759263e-05,
+      "loss": 1.6534,
+      "step": 837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26703958427818103,
+      "learning_rate": 7.895592326841412e-05,
+      "loss": 1.6324,
+      "step": 838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22223709057066937,
+      "learning_rate": 7.894274417923561e-05,
+      "loss": 1.6472,
+      "step": 839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18015143412151977,
+      "learning_rate": 7.892956509005712e-05,
+      "loss": 1.6523,
+      "step": 840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16776687777891955,
+      "learning_rate": 7.891638600087861e-05,
+      "loss": 1.6408,
+      "step": 841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2110093032771735,
+      "learning_rate": 7.890320691170011e-05,
+      "loss": 1.6475,
+      "step": 842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21006634515360098,
+      "learning_rate": 7.88900278225216e-05,
+      "loss": 1.6495,
+      "step": 843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.19405253109457415,
+      "learning_rate": 7.88768487333431e-05,
+      "loss": 1.6373,
+      "step": 844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17598128896680296,
+      "learning_rate": 7.886366964416459e-05,
+      "loss": 1.6319,
+      "step": 845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1649091128056297,
+      "learning_rate": 7.88504905549861e-05,
+      "loss": 1.6457,
+      "step": 846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15971492084809902,
+      "learning_rate": 7.883731146580759e-05,
+      "loss": 1.6352,
+      "step": 847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.17712277192126102,
+      "learning_rate": 7.882413237662908e-05,
+      "loss": 1.6381,
+      "step": 848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.20906554492964108,
+      "learning_rate": 7.881095328745059e-05,
+      "loss": 1.6437,
+      "step": 849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2142843140002136,
+      "learning_rate": 7.879777419827209e-05,
+      "loss": 1.6434,
+      "step": 850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.21006215293899588,
+      "learning_rate": 7.878459510909357e-05,
+      "loss": 1.6427,
+      "step": 851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23775701204551145,
+      "learning_rate": 7.877141601991507e-05,
+      "loss": 1.6455,
+      "step": 852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2707019175159962,
+      "learning_rate": 7.875823693073657e-05,
+      "loss": 1.6456,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3127403618384618,
+      "learning_rate": 7.874505784155806e-05,
+      "loss": 1.6385,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36233560236813683,
+      "learning_rate": 7.873187875237957e-05,
+      "loss": 1.6392,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36917547935418177,
+      "learning_rate": 7.871869966320106e-05,
+      "loss": 1.643,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2799689100614571,
+      "learning_rate": 7.870552057402255e-05,
+      "loss": 1.6364,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1909288497464617,
+      "learning_rate": 7.869234148484406e-05,
+      "loss": 1.6442,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22049646685021101,
+      "learning_rate": 7.867916239566554e-05,
+      "loss": 1.6381,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2792850134812065,
+      "learning_rate": 7.866598330648704e-05,
+      "loss": 1.6439,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2767586748336834,
+      "learning_rate": 7.865280421730854e-05,
+      "loss": 1.6406,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1964234052122111,
+      "learning_rate": 7.863962512813004e-05,
+      "loss": 1.6403,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21916883082393832,
+      "learning_rate": 7.862644603895153e-05,
+      "loss": 1.6433,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2664436290112629,
+      "learning_rate": 7.861326694977304e-05,
+      "loss": 1.6422,
+      "step": 864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21301889349155526,
+      "learning_rate": 7.860008786059453e-05,
+      "loss": 1.6416,
+      "step": 865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17884914022978088,
+      "learning_rate": 7.858690877141603e-05,
+      "loss": 1.6312,
+      "step": 866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18843102073542442,
+      "learning_rate": 7.857372968223752e-05,
+      "loss": 1.6416,
+      "step": 867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18688924194495246,
+      "learning_rate": 7.856055059305901e-05,
+      "loss": 1.6441,
+      "step": 868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1650145694966954,
+      "learning_rate": 7.854737150388051e-05,
+      "loss": 1.6454,
+      "step": 869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1582803846188149,
+      "learning_rate": 7.853419241470201e-05,
+      "loss": 1.6359,
+      "step": 870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1914936260863352,
+      "learning_rate": 7.852101332552351e-05,
+      "loss": 1.6346,
+      "step": 871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2272665695903623,
+      "learning_rate": 7.8507834236345e-05,
+      "loss": 1.6424,
+      "step": 872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2918408917481599,
+      "learning_rate": 7.849465514716651e-05,
+      "loss": 1.6362,
+      "step": 873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33124958511818076,
+      "learning_rate": 7.8481476057988e-05,
+      "loss": 1.6457,
+      "step": 874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28392389646236854,
+      "learning_rate": 7.846829696880948e-05,
+      "loss": 1.636,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21187264881262416,
+      "learning_rate": 7.845511787963099e-05,
+      "loss": 1.6391,
+      "step": 876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19512382269665135,
+      "learning_rate": 7.844193879045248e-05,
+      "loss": 1.6375,
+      "step": 877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2315133095228095,
+      "learning_rate": 7.842875970127398e-05,
+      "loss": 1.6434,
+      "step": 878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22455542676851648,
+      "learning_rate": 7.841558061209548e-05,
+      "loss": 1.6375,
+      "step": 879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18274445518703517,
+      "learning_rate": 7.840240152291698e-05,
+      "loss": 1.6403,
+      "step": 880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1680652375376494,
+      "learning_rate": 7.838922243373847e-05,
+      "loss": 1.6253,
+      "step": 881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2029462350044378,
+      "learning_rate": 7.837604334455998e-05,
+      "loss": 1.6335,
+      "step": 882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2385785209220654,
+      "learning_rate": 7.836286425538146e-05,
+      "loss": 1.636,
+      "step": 883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22966225461134943,
+      "learning_rate": 7.834968516620295e-05,
+      "loss": 1.6351,
+      "step": 884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20620730487272254,
+      "learning_rate": 7.833650607702446e-05,
+      "loss": 1.64,
+      "step": 885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20815921246674304,
+      "learning_rate": 7.832332698784595e-05,
+      "loss": 1.642,
+      "step": 886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21329180245890958,
+      "learning_rate": 7.831014789866745e-05,
+      "loss": 1.6289,
+      "step": 887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20448951459967843,
+      "learning_rate": 7.829696880948896e-05,
+      "loss": 1.6327,
+      "step": 888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19419304903131157,
+      "learning_rate": 7.828378972031045e-05,
+      "loss": 1.6322,
+      "step": 889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18030150429004935,
+      "learning_rate": 7.827061063113194e-05,
+      "loss": 1.6331,
+      "step": 890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1898733739259939,
+      "learning_rate": 7.825743154195344e-05,
+      "loss": 1.6342,
+      "step": 891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21064065788890746,
+      "learning_rate": 7.824425245277493e-05,
+      "loss": 1.6407,
+      "step": 892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20679616302698337,
+      "learning_rate": 7.823107336359642e-05,
+      "loss": 1.6328,
+      "step": 893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21400981491846863,
+      "learning_rate": 7.821789427441793e-05,
+      "loss": 1.6426,
+      "step": 894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20061186083901789,
+      "learning_rate": 7.820471518523943e-05,
+      "loss": 1.6359,
+      "step": 895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18863815744505064,
+      "learning_rate": 7.819153609606092e-05,
+      "loss": 1.6307,
+      "step": 896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19996719699171972,
+      "learning_rate": 7.817835700688243e-05,
+      "loss": 1.6299,
+      "step": 897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2074855070501711,
+      "learning_rate": 7.816517791770392e-05,
+      "loss": 1.6431,
+      "step": 898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18908061531345277,
+      "learning_rate": 7.81519988285254e-05,
+      "loss": 1.6359,
+      "step": 899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1621341838335611,
+      "learning_rate": 7.813881973934691e-05,
+      "loss": 1.6334,
+      "step": 900
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.5936576128005981,
+      "eval_runtime": 447.3124,
+      "eval_samples_per_second": 40.274,
+      "eval_steps_per_second": 0.06,
+      "step": 900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15865732799100835,
+      "learning_rate": 7.81256406501684e-05,
+      "loss": 1.6265,
+      "step": 901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16466989815596636,
+      "learning_rate": 7.81124615609899e-05,
+      "loss": 1.636,
+      "step": 902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18355978585759908,
+      "learning_rate": 7.80992824718114e-05,
+      "loss": 1.6412,
+      "step": 903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16881096243427599,
+      "learning_rate": 7.80861033826329e-05,
+      "loss": 1.6218,
+      "step": 904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17792146649092913,
+      "learning_rate": 7.807292429345439e-05,
+      "loss": 1.6351,
+      "step": 905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1861296128116092,
+      "learning_rate": 7.80597452042759e-05,
+      "loss": 1.6314,
+      "step": 906
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19805706857044636,
+      "learning_rate": 7.804656611509738e-05,
+      "loss": 1.6274,
+      "step": 907
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2477603308349337,
+      "learning_rate": 7.803338702591887e-05,
+      "loss": 1.6345,
+      "step": 908
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3058009391702576,
+      "learning_rate": 7.802020793674038e-05,
+      "loss": 1.6349,
+      "step": 909
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3285989768622068,
+      "learning_rate": 7.800702884756187e-05,
+      "loss": 1.6354,
+      "step": 910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29327673429046075,
+      "learning_rate": 7.799384975838337e-05,
+      "loss": 1.636,
+      "step": 911
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23773885823846896,
+      "learning_rate": 7.798067066920487e-05,
+      "loss": 1.6264,
+      "step": 912
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19226268966581236,
+      "learning_rate": 7.796749158002637e-05,
+      "loss": 1.635,
+      "step": 913
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20182491167455907,
+      "learning_rate": 7.795431249084786e-05,
+      "loss": 1.638,
+      "step": 914
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23238881306153533,
+      "learning_rate": 7.794113340166935e-05,
+      "loss": 1.6316,
+      "step": 915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23986180803329027,
+      "learning_rate": 7.792795431249085e-05,
+      "loss": 1.6258,
+      "step": 916
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1961225124547818,
+      "learning_rate": 7.791477522331234e-05,
+      "loss": 1.6326,
+      "step": 917
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16968519491312442,
+      "learning_rate": 7.790159613413385e-05,
+      "loss": 1.6357,
+      "step": 918
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.17723689039254287,
+      "learning_rate": 7.788841704495534e-05,
+      "loss": 1.6321,
+      "step": 919
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19952319110339864,
+      "learning_rate": 7.787523795577684e-05,
+      "loss": 1.6294,
+      "step": 920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.20482205553447078,
+      "learning_rate": 7.786205886659834e-05,
+      "loss": 1.6304,
+      "step": 921
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.19973731737429154,
+      "learning_rate": 7.784887977741984e-05,
+      "loss": 1.6305,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19968852921711655,
+      "learning_rate": 7.783570068824132e-05,
+      "loss": 1.6216,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2432499755542469,
+      "learning_rate": 7.782252159906283e-05,
+      "loss": 1.6183,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28904807766758667,
+      "learning_rate": 7.780934250988432e-05,
+      "loss": 1.6331,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30605720879153786,
+      "learning_rate": 7.779616342070581e-05,
+      "loss": 1.6429,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28776303552521276,
+      "learning_rate": 7.778298433152732e-05,
+      "loss": 1.6376,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26692576512866534,
+      "learning_rate": 7.776980524234881e-05,
+      "loss": 1.636,
+      "step": 928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23006966654469613,
+      "learning_rate": 7.775662615317031e-05,
+      "loss": 1.6375,
+      "step": 929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18021078514991914,
+      "learning_rate": 7.774344706399181e-05,
+      "loss": 1.6231,
+      "step": 930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18063501767218804,
+      "learning_rate": 7.77302679748133e-05,
+      "loss": 1.6316,
+      "step": 931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2146429092144611,
+      "learning_rate": 7.771708888563479e-05,
+      "loss": 1.6279,
+      "step": 932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20818499945204852,
+      "learning_rate": 7.77039097964563e-05,
+      "loss": 1.6269,
+      "step": 933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18354993630667119,
+      "learning_rate": 7.769073070727779e-05,
+      "loss": 1.6227,
+      "step": 934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18200851976694746,
+      "learning_rate": 7.767755161809928e-05,
+      "loss": 1.6351,
+      "step": 935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1865682691641124,
+      "learning_rate": 7.766437252892079e-05,
+      "loss": 1.6275,
+      "step": 936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1802592148443168,
+      "learning_rate": 7.765119343974228e-05,
+      "loss": 1.6275,
+      "step": 937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17289003336299882,
+      "learning_rate": 7.763801435056378e-05,
+      "loss": 1.6341,
+      "step": 938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1986135066316666,
+      "learning_rate": 7.762483526138527e-05,
+      "loss": 1.6196,
+      "step": 939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18010570303182816,
+      "learning_rate": 7.761165617220677e-05,
+      "loss": 1.624,
+      "step": 940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1673925458817595,
+      "learning_rate": 7.759847708302826e-05,
+      "loss": 1.6334,
+      "step": 941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17837996599687475,
+      "learning_rate": 7.758529799384977e-05,
+      "loss": 1.6344,
+      "step": 942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.167225580195058,
+      "learning_rate": 7.757211890467126e-05,
+      "loss": 1.6311,
+      "step": 943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16484688881261245,
+      "learning_rate": 7.755893981549275e-05,
+      "loss": 1.633,
+      "step": 944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16970087441567513,
+      "learning_rate": 7.754576072631426e-05,
+      "loss": 1.6277,
+      "step": 945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17706071153067815,
+      "learning_rate": 7.753258163713576e-05,
+      "loss": 1.6276,
+      "step": 946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19185774572084416,
+      "learning_rate": 7.751940254795724e-05,
+      "loss": 1.6282,
+      "step": 947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21534610743277194,
+      "learning_rate": 7.750622345877874e-05,
+      "loss": 1.628,
+      "step": 948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22682207930909348,
+      "learning_rate": 7.749304436960024e-05,
+      "loss": 1.6351,
+      "step": 949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20425437523432943,
+      "learning_rate": 7.747986528042173e-05,
+      "loss": 1.6226,
+      "step": 950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1815592545929023,
+      "learning_rate": 7.746668619124324e-05,
+      "loss": 1.6294,
+      "step": 951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1914304657655877,
+      "learning_rate": 7.745350710206473e-05,
+      "loss": 1.624,
+      "step": 952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19841817505160111,
+      "learning_rate": 7.744032801288623e-05,
+      "loss": 1.626,
+      "step": 953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20513655191404673,
+      "learning_rate": 7.742714892370773e-05,
+      "loss": 1.625,
+      "step": 954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20690976922591042,
+      "learning_rate": 7.741396983452921e-05,
+      "loss": 1.6269,
+      "step": 955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20042970552356906,
+      "learning_rate": 7.74007907453507e-05,
+      "loss": 1.6227,
+      "step": 956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22313249765436127,
+      "learning_rate": 7.738761165617221e-05,
+      "loss": 1.6307,
+      "step": 957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24940862321038032,
+      "learning_rate": 7.737443256699371e-05,
+      "loss": 1.6262,
+      "step": 958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2733600251891239,
+      "learning_rate": 7.73612534778152e-05,
+      "loss": 1.6234,
+      "step": 959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27327679593574355,
+      "learning_rate": 7.734807438863671e-05,
+      "loss": 1.6204,
+      "step": 960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2528084631262215,
+      "learning_rate": 7.73348952994582e-05,
+      "loss": 1.6235,
+      "step": 961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21747127891255372,
+      "learning_rate": 7.73217162102797e-05,
+      "loss": 1.625,
+      "step": 962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1910347418430321,
+      "learning_rate": 7.730853712110119e-05,
+      "loss": 1.6167,
+      "step": 963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17541573212547087,
+      "learning_rate": 7.729535803192268e-05,
+      "loss": 1.627,
+      "step": 964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20823218712713615,
+      "learning_rate": 7.728217894274418e-05,
+      "loss": 1.6303,
+      "step": 965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20802148470127552,
+      "learning_rate": 7.726899985356568e-05,
+      "loss": 1.6278,
+      "step": 966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17924417756170902,
+      "learning_rate": 7.725582076438718e-05,
+      "loss": 1.6321,
+      "step": 967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16948531308406123,
+      "learning_rate": 7.724264167520867e-05,
+      "loss": 1.6243,
+      "step": 968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.167050690087119,
+      "learning_rate": 7.722946258603018e-05,
+      "loss": 1.627,
+      "step": 969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1707462068741091,
+      "learning_rate": 7.721628349685167e-05,
+      "loss": 1.6105,
+      "step": 970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18254429275405454,
+      "learning_rate": 7.720310440767315e-05,
+      "loss": 1.6269,
+      "step": 971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1915398246036987,
+      "learning_rate": 7.718992531849466e-05,
+      "loss": 1.6143,
+      "step": 972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22066508050668615,
+      "learning_rate": 7.717674622931615e-05,
+      "loss": 1.6255,
+      "step": 973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21683723311113615,
+      "learning_rate": 7.716356714013765e-05,
+      "loss": 1.6297,
+      "step": 974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19079856723315286,
+      "learning_rate": 7.715038805095916e-05,
+      "loss": 1.626,
+      "step": 975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20257609739912277,
+      "learning_rate": 7.713720896178065e-05,
+      "loss": 1.6123,
+      "step": 976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.19318811031897568,
+      "learning_rate": 7.712402987260214e-05,
+      "loss": 1.6269,
+      "step": 977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20006255429024558,
+      "learning_rate": 7.711085078342365e-05,
+      "loss": 1.6206,
+      "step": 978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20243419836299764,
+      "learning_rate": 7.709767169424513e-05,
+      "loss": 1.6224,
+      "step": 979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21154717766882877,
+      "learning_rate": 7.708449260506662e-05,
+      "loss": 1.6296,
+      "step": 980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2111145585800829,
+      "learning_rate": 7.707131351588813e-05,
+      "loss": 1.6246,
+      "step": 981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20904551886984418,
+      "learning_rate": 7.705813442670962e-05,
+      "loss": 1.6242,
+      "step": 982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2155821099387234,
+      "learning_rate": 7.704495533753112e-05,
+      "loss": 1.6157,
+      "step": 983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24171139714485312,
+      "learning_rate": 7.703177624835263e-05,
+      "loss": 1.6194,
+      "step": 984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24021769786284733,
+      "learning_rate": 7.701859715917412e-05,
+      "loss": 1.6179,
+      "step": 985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22097810478380417,
+      "learning_rate": 7.700541806999561e-05,
+      "loss": 1.6138,
+      "step": 986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.20419669337349122,
+      "learning_rate": 7.699223898081711e-05,
+      "loss": 1.617,
+      "step": 987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18295944585814536,
+      "learning_rate": 7.69790598916386e-05,
+      "loss": 1.6192,
+      "step": 988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2017143911669201,
+      "learning_rate": 7.69658808024601e-05,
+      "loss": 1.6183,
+      "step": 989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2550706626997802,
+      "learning_rate": 7.69527017132816e-05,
+      "loss": 1.622,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2616391313909151,
+      "learning_rate": 7.69395226241031e-05,
+      "loss": 1.6148,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22476181941965861,
+      "learning_rate": 7.692634353492459e-05,
+      "loss": 1.6303,
+      "step": 992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16958496517690427,
+      "learning_rate": 7.69131644457461e-05,
+      "loss": 1.6289,
+      "step": 993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16535160221593995,
+      "learning_rate": 7.689998535656759e-05,
+      "loss": 1.6205,
+      "step": 994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19695198203233968,
+      "learning_rate": 7.688680626738907e-05,
+      "loss": 1.6244,
+      "step": 995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18773614722297197,
+      "learning_rate": 7.687362717821058e-05,
+      "loss": 1.6206,
+      "step": 996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16768417110529632,
+      "learning_rate": 7.686044808903207e-05,
+      "loss": 1.6164,
+      "step": 997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15543589531360824,
+      "learning_rate": 7.684726899985357e-05,
+      "loss": 1.6168,
+      "step": 998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15781452108271793,
+      "learning_rate": 7.683408991067507e-05,
+      "loss": 1.6169,
+      "step": 999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19721598910764596,
+      "learning_rate": 7.682091082149657e-05,
+      "loss": 1.6159,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.20319788131936403,
+      "learning_rate": 7.680773173231806e-05,
+      "loss": 1.6198,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16718938876894052,
+      "learning_rate": 7.679455264313957e-05,
+      "loss": 1.6151,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1287659352551544,
+      "learning_rate": 7.678137355396105e-05,
+      "loss": 1.6171,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15368443936088208,
+      "learning_rate": 7.676819446478254e-05,
+      "loss": 1.6217,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2007291365676937,
+      "learning_rate": 7.675501537560405e-05,
+      "loss": 1.6168,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18906054405701359,
+      "learning_rate": 7.674183628642554e-05,
+      "loss": 1.621,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1468837292776729,
+      "learning_rate": 7.672865719724704e-05,
+      "loss": 1.6129,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13377111078446055,
+      "learning_rate": 7.671547810806854e-05,
+      "loss": 1.6155,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1426223536761099,
+      "learning_rate": 7.670229901889004e-05,
+      "loss": 1.6115,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17114665015966507,
+      "learning_rate": 7.668911992971153e-05,
+      "loss": 1.6281,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1807116919105783,
+      "learning_rate": 7.667594084053302e-05,
+      "loss": 1.6159,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17057827920089186,
+      "learning_rate": 7.666276175135452e-05,
+      "loss": 1.6246,
+      "step": 1012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18704211275868782,
+      "learning_rate": 7.664958266217601e-05,
+      "loss": 1.6119,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2011380246874562,
+      "learning_rate": 7.663640357299752e-05,
+      "loss": 1.615,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22834646726527497,
+      "learning_rate": 7.662322448381901e-05,
+      "loss": 1.6118,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24388254809300247,
+      "learning_rate": 7.661004539464051e-05,
+      "loss": 1.6147,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23555275478651708,
+      "learning_rate": 7.659686630546201e-05,
+      "loss": 1.6082,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2273124399174075,
+      "learning_rate": 7.658368721628351e-05,
+      "loss": 1.6178,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19649796595180624,
+      "learning_rate": 7.657050812710499e-05,
+      "loss": 1.6146,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.20344279211487082,
+      "learning_rate": 7.65573290379265e-05,
+      "loss": 1.6095,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2203695405176659,
+      "learning_rate": 7.654414994874799e-05,
+      "loss": 1.6153,
+      "step": 1021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.20941131575413013,
+      "learning_rate": 7.653097085956948e-05,
+      "loss": 1.6088,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17696528858927837,
+      "learning_rate": 7.651779177039099e-05,
+      "loss": 1.6186,
+      "step": 1023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17541204603143867,
+      "learning_rate": 7.650461268121248e-05,
+      "loss": 1.6109,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19137693781810725,
+      "learning_rate": 7.649143359203398e-05,
+      "loss": 1.6103,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2104394048421515,
+      "learning_rate": 7.647825450285549e-05,
+      "loss": 1.6104,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23866573734409802,
+      "learning_rate": 7.646507541367697e-05,
+      "loss": 1.6158,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2711289371560318,
+      "learning_rate": 7.645189632449846e-05,
+      "loss": 1.6097,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2903926058136879,
+      "learning_rate": 7.643871723531997e-05,
+      "loss": 1.6146,
+      "step": 1029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578361261344392,
+      "learning_rate": 7.642553814614146e-05,
+      "loss": 1.6119,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.206355859168418,
+      "learning_rate": 7.641235905696295e-05,
+      "loss": 1.617,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18121724499536424,
+      "learning_rate": 7.639917996778446e-05,
+      "loss": 1.617,
+      "step": 1032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1681573451991133,
+      "learning_rate": 7.638600087860595e-05,
+      "loss": 1.619,
+      "step": 1033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1868021605182587,
+      "learning_rate": 7.637282178942745e-05,
+      "loss": 1.6267,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.20048387395095732,
+      "learning_rate": 7.635964270024894e-05,
+      "loss": 1.6093,
+      "step": 1035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1902918229490612,
+      "learning_rate": 7.634646361107044e-05,
+      "loss": 1.6177,
+      "step": 1036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1995438979121453,
+      "learning_rate": 7.633328452189193e-05,
+      "loss": 1.6104,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18509718339047015,
+      "learning_rate": 7.632010543271344e-05,
+      "loss": 1.6085,
+      "step": 1038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16406358483238698,
+      "learning_rate": 7.630692634353493e-05,
+      "loss": 1.6065,
+      "step": 1039
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1686539810400193,
+      "learning_rate": 7.629374725435642e-05,
+      "loss": 1.6075,
+      "step": 1040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1616937354360527,
+      "learning_rate": 7.628056816517793e-05,
+      "loss": 1.6138,
+      "step": 1041
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15587744447247096,
+      "learning_rate": 7.626738907599943e-05,
+      "loss": 1.6141,
+      "step": 1042
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17054924211340525,
+      "learning_rate": 7.62542099868209e-05,
+      "loss": 1.6144,
+      "step": 1043
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17357874700438133,
+      "learning_rate": 7.624103089764241e-05,
+      "loss": 1.6077,
+      "step": 1044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1884137124401568,
+      "learning_rate": 7.622785180846391e-05,
+      "loss": 1.6214,
+      "step": 1045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.20954300727869857,
+      "learning_rate": 7.62146727192854e-05,
+      "loss": 1.6113,
+      "step": 1046
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24486234826133343,
+      "learning_rate": 7.620149363010691e-05,
+      "loss": 1.6105,
+      "step": 1047
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2829288363932686,
+      "learning_rate": 7.61883145409284e-05,
+      "loss": 1.6063,
+      "step": 1048
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26289957468995534,
+      "learning_rate": 7.61751354517499e-05,
+      "loss": 1.6159,
+      "step": 1049
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2063007086026833,
+      "learning_rate": 7.61619563625714e-05,
+      "loss": 1.6203,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17189607819088035,
+      "learning_rate": 7.614877727339288e-05,
+      "loss": 1.6,
+      "step": 1051
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1562689270878245,
+      "learning_rate": 7.613559818421438e-05,
+      "loss": 1.6137,
+      "step": 1052
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.18708728589568419,
+      "learning_rate": 7.612241909503588e-05,
+      "loss": 1.6039,
+      "step": 1053
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.19790043445558433,
+      "learning_rate": 7.610924000585738e-05,
+      "loss": 1.6106,
+      "step": 1054
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17858565211710867,
+      "learning_rate": 7.609606091667887e-05,
+      "loss": 1.6109,
+      "step": 1055
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16704932302988487,
+      "learning_rate": 7.608288182750038e-05,
+      "loss": 1.6118,
+      "step": 1056
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.17947082466270875,
+      "learning_rate": 7.606970273832187e-05,
+      "loss": 1.609,
+      "step": 1057
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.176708713327256,
+      "learning_rate": 7.605652364914337e-05,
+      "loss": 1.6191,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1622950277483614,
+      "learning_rate": 7.604334455996486e-05,
+      "loss": 1.6116,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14488222389166788,
+      "learning_rate": 7.603016547078635e-05,
+      "loss": 1.6051,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14531527208325867,
+      "learning_rate": 7.601698638160785e-05,
+      "loss": 1.606,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17752790968480894,
+      "learning_rate": 7.600380729242935e-05,
+      "loss": 1.6076,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1957495026607104,
+      "learning_rate": 7.599062820325085e-05,
+      "loss": 1.6136,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2202660125439618,
+      "learning_rate": 7.597744911407234e-05,
+      "loss": 1.6114,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21781433216601626,
+      "learning_rate": 7.596427002489385e-05,
+      "loss": 1.6052,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19946411918365636,
+      "learning_rate": 7.595109093571534e-05,
+      "loss": 1.6138,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17221638475638887,
+      "learning_rate": 7.593791184653682e-05,
+      "loss": 1.6129,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1531692080815194,
+      "learning_rate": 7.592473275735833e-05,
+      "loss": 1.6028,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16538560411066186,
+      "learning_rate": 7.591155366817982e-05,
+      "loss": 1.6106,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.18631496746925127,
+      "learning_rate": 7.589837457900132e-05,
+      "loss": 1.5899,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.18424485959963482,
+      "learning_rate": 7.588519548982283e-05,
+      "loss": 1.6202,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15671571674110887,
+      "learning_rate": 7.587201640064432e-05,
+      "loss": 1.6021,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.18127997297391865,
+      "learning_rate": 7.585883731146581e-05,
+      "loss": 1.6159,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2259241643675569,
+      "learning_rate": 7.584565822228732e-05,
+      "loss": 1.6092,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23587302884105796,
+      "learning_rate": 7.58324791331088e-05,
+      "loss": 1.6074,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22149002998233225,
+      "learning_rate": 7.58193000439303e-05,
+      "loss": 1.6002,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20000549277457733,
+      "learning_rate": 7.58061209547518e-05,
+      "loss": 1.6021,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1867293286337856,
+      "learning_rate": 7.57929418655733e-05,
+      "loss": 1.6157,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20901234864931093,
+      "learning_rate": 7.577976277639479e-05,
+      "loss": 1.6106,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2329710725596326,
+      "learning_rate": 7.57665836872163e-05,
+      "loss": 1.5965,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23558076759578483,
+      "learning_rate": 7.575340459803779e-05,
+      "loss": 1.6088,
+      "step": 1081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20255246934932447,
+      "learning_rate": 7.574022550885928e-05,
+      "loss": 1.6118,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17105047596886078,
+      "learning_rate": 7.572704641968078e-05,
+      "loss": 1.6031,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.151575851349542,
+      "learning_rate": 7.571386733050227e-05,
+      "loss": 1.5992,
+      "step": 1084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14457999254055678,
+      "learning_rate": 7.570068824132376e-05,
+      "loss": 1.6028,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17929756036682534,
+      "learning_rate": 7.568750915214527e-05,
+      "loss": 1.6093,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22006099380977615,
+      "learning_rate": 7.567433006296677e-05,
+      "loss": 1.614,
+      "step": 1087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23725118276054816,
+      "learning_rate": 7.566115097378826e-05,
+      "loss": 1.6115,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23038896462149644,
+      "learning_rate": 7.564797188460977e-05,
+      "loss": 1.6046,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22163894578256085,
+      "learning_rate": 7.563479279543126e-05,
+      "loss": 1.6135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2103881958644772,
+      "learning_rate": 7.562161370625274e-05,
+      "loss": 1.6005,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21034951411455005,
+      "learning_rate": 7.560843461707425e-05,
+      "loss": 1.6097,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22289981805998374,
+      "learning_rate": 7.559525552789574e-05,
+      "loss": 1.5967,
+      "step": 1093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2756558475920652,
+      "learning_rate": 7.558207643871724e-05,
+      "loss": 1.606,
+      "step": 1094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24802208075338164,
+      "learning_rate": 7.556889734953874e-05,
+      "loss": 1.5984,
+      "step": 1095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21897014261396455,
+      "learning_rate": 7.555571826036024e-05,
+      "loss": 1.604,
+      "step": 1096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17301569858025576,
+      "learning_rate": 7.554253917118173e-05,
+      "loss": 1.6085,
+      "step": 1097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15443666681450202,
+      "learning_rate": 7.552936008200324e-05,
+      "loss": 1.5974,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20297306903978035,
+      "learning_rate": 7.551618099282472e-05,
+      "loss": 1.6101,
+      "step": 1099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24144721356383153,
+      "learning_rate": 7.550300190364621e-05,
+      "loss": 1.598,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22156745220232862,
+      "learning_rate": 7.548982281446772e-05,
+      "loss": 1.6168,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.193769033449188,
+      "learning_rate": 7.547664372528921e-05,
+      "loss": 1.6096,
+      "step": 1102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19547711850030344,
+      "learning_rate": 7.54634646361107e-05,
+      "loss": 1.6078,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20219810170780822,
+      "learning_rate": 7.545028554693221e-05,
+      "loss": 1.6087,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1656378663366526,
+      "learning_rate": 7.543710645775371e-05,
+      "loss": 1.6047,
+      "step": 1105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12279504679737889,
+      "learning_rate": 7.54239273685752e-05,
+      "loss": 1.608,
+      "step": 1106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14332841985114664,
+      "learning_rate": 7.54107482793967e-05,
+      "loss": 1.6007,
+      "step": 1107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17869073671045163,
+      "learning_rate": 7.539756919021819e-05,
+      "loss": 1.6018,
+      "step": 1108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17562435746125366,
+      "learning_rate": 7.538439010103968e-05,
+      "loss": 1.6085,
+      "step": 1109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15844195620891538,
+      "learning_rate": 7.537121101186119e-05,
+      "loss": 1.6046,
+      "step": 1110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15431885892046604,
+      "learning_rate": 7.535803192268268e-05,
+      "loss": 1.6017,
+      "step": 1111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16010284786126977,
+      "learning_rate": 7.534485283350418e-05,
+      "loss": 1.6078,
+      "step": 1112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14618258638327017,
+      "learning_rate": 7.533167374432568e-05,
+      "loss": 1.6061,
+      "step": 1113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16207811103063444,
+      "learning_rate": 7.531849465514718e-05,
+      "loss": 1.6015,
+      "step": 1114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.195737447043241,
+      "learning_rate": 7.530531556596866e-05,
+      "loss": 1.5978,
+      "step": 1115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.20987994206889682,
+      "learning_rate": 7.529213647679017e-05,
+      "loss": 1.6133,
+      "step": 1116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23671104304648102,
+      "learning_rate": 7.527895738761166e-05,
+      "loss": 1.5964,
+      "step": 1117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26675358723328,
+      "learning_rate": 7.526577829843315e-05,
+      "loss": 1.6112,
+      "step": 1118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26866295102797627,
+      "learning_rate": 7.525259920925466e-05,
+      "loss": 1.6018,
+      "step": 1119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23052925267631813,
+      "learning_rate": 7.523942012007615e-05,
+      "loss": 1.6133,
+      "step": 1120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17178396931122542,
+      "learning_rate": 7.522624103089765e-05,
+      "loss": 1.6071,
+      "step": 1121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16832141500530334,
+      "learning_rate": 7.521306194171914e-05,
+      "loss": 1.6034,
+      "step": 1122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22478896594336525,
+      "learning_rate": 7.519988285254064e-05,
+      "loss": 1.5981,
+      "step": 1123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22109419557694107,
+      "learning_rate": 7.518670376336213e-05,
+      "loss": 1.6036,
+      "step": 1124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2011335701566258,
+      "learning_rate": 7.517352467418364e-05,
+      "loss": 1.606,
+      "step": 1125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.19999382668915064,
+      "learning_rate": 7.516034558500513e-05,
+      "loss": 1.5958,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2087313106488091,
+      "learning_rate": 7.514716649582662e-05,
+      "loss": 1.6041,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25624726246634655,
+      "learning_rate": 7.513398740664813e-05,
+      "loss": 1.6054,
+      "step": 1128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2760442187545006,
+      "learning_rate": 7.512080831746963e-05,
+      "loss": 1.6026,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21828668039689092,
+      "learning_rate": 7.510762922829112e-05,
+      "loss": 1.6058,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15791621829412608,
+      "learning_rate": 7.509445013911261e-05,
+      "loss": 1.6009,
+      "step": 1131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1554563269196245,
+      "learning_rate": 7.50812710499341e-05,
+      "loss": 1.6036,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16523751476714793,
+      "learning_rate": 7.50680919607556e-05,
+      "loss": 1.6029,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17099491590982738,
+      "learning_rate": 7.505491287157711e-05,
+      "loss": 1.5972,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17134299136620068,
+      "learning_rate": 7.50417337823986e-05,
+      "loss": 1.5904,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18427242584450257,
+      "learning_rate": 7.50285546932201e-05,
+      "loss": 1.5935,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17050705423416349,
+      "learning_rate": 7.50153756040416e-05,
+      "loss": 1.5972,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16797826815156477,
+      "learning_rate": 7.50021965148631e-05,
+      "loss": 1.599,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18202925378142923,
+      "learning_rate": 7.498901742568458e-05,
+      "loss": 1.6038,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1986542020413657,
+      "learning_rate": 7.497583833650608e-05,
+      "loss": 1.6116,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18787713595892783,
+      "learning_rate": 7.496265924732758e-05,
+      "loss": 1.6074,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1601110159530077,
+      "learning_rate": 7.494948015814907e-05,
+      "loss": 1.6044,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15871745658510408,
+      "learning_rate": 7.493630106897058e-05,
+      "loss": 1.6032,
+      "step": 1143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1430625947479129,
+      "learning_rate": 7.492312197979207e-05,
+      "loss": 1.5926,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1469464248975521,
+      "learning_rate": 7.490994289061357e-05,
+      "loss": 1.6002,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15258319044288818,
+      "learning_rate": 7.489676380143506e-05,
+      "loss": 1.5978,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16643604102356668,
+      "learning_rate": 7.488358471225655e-05,
+      "loss": 1.599,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16005469141475748,
+      "learning_rate": 7.487040562307805e-05,
+      "loss": 1.5864,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15986595889960914,
+      "learning_rate": 7.485722653389955e-05,
+      "loss": 1.5997,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15784680612918575,
+      "learning_rate": 7.484404744472105e-05,
+      "loss": 1.5986,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15000600634626432,
+      "learning_rate": 7.483086835554254e-05,
+      "loss": 1.599,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17485532780513346,
+      "learning_rate": 7.481768926636405e-05,
+      "loss": 1.6024,
+      "step": 1152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1967341085166345,
+      "learning_rate": 7.480451017718554e-05,
+      "loss": 1.5936,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2344844762884148,
+      "learning_rate": 7.479133108800704e-05,
+      "loss": 1.5998,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2497106189098422,
+      "learning_rate": 7.477815199882853e-05,
+      "loss": 1.596,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2040425471013873,
+      "learning_rate": 7.476497290965002e-05,
+      "loss": 1.595,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17477543215459185,
+      "learning_rate": 7.475179382047152e-05,
+      "loss": 1.5879,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1747166451083869,
+      "learning_rate": 7.473861473129302e-05,
+      "loss": 1.5973,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.19091852317109032,
+      "learning_rate": 7.472543564211452e-05,
+      "loss": 1.5935,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20616285346047855,
+      "learning_rate": 7.471225655293601e-05,
+      "loss": 1.5923,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2014869666468898,
+      "learning_rate": 7.469907746375752e-05,
+      "loss": 1.5957,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16960374414389154,
+      "learning_rate": 7.468589837457901e-05,
+      "loss": 1.5968,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.174537560571573,
+      "learning_rate": 7.46727192854005e-05,
+      "loss": 1.5975,
+      "step": 1163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2042247405656548,
+      "learning_rate": 7.4659540196222e-05,
+      "loss": 1.599,
+      "step": 1164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22065476727458821,
+      "learning_rate": 7.46463611070435e-05,
+      "loss": 1.5954,
+      "step": 1165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.19733900341635885,
+      "learning_rate": 7.463318201786499e-05,
+      "loss": 1.5947,
+      "step": 1166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1531704474147821,
+      "learning_rate": 7.46200029286865e-05,
+      "loss": 1.5997,
+      "step": 1167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15625856015130773,
+      "learning_rate": 7.460682383950799e-05,
+      "loss": 1.605,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18227226464739196,
+      "learning_rate": 7.459364475032948e-05,
+      "loss": 1.5843,
+      "step": 1169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18336918188164913,
+      "learning_rate": 7.458046566115098e-05,
+      "loss": 1.5955,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1653696455586279,
+      "learning_rate": 7.456728657197247e-05,
+      "loss": 1.593,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14145061297298278,
+      "learning_rate": 7.455410748279396e-05,
+      "loss": 1.5984,
+      "step": 1172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15177089703146257,
+      "learning_rate": 7.454092839361547e-05,
+      "loss": 1.5908,
+      "step": 1173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1657687250680247,
+      "learning_rate": 7.452774930443697e-05,
+      "loss": 1.5878,
+      "step": 1174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18104982414107768,
+      "learning_rate": 7.451457021525846e-05,
+      "loss": 1.5896,
+      "step": 1175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18058656545512794,
+      "learning_rate": 7.450139112607997e-05,
+      "loss": 1.5946,
+      "step": 1176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18947505554561722,
+      "learning_rate": 7.448821203690146e-05,
+      "loss": 1.595,
+      "step": 1177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23178548541042568,
+      "learning_rate": 7.447503294772295e-05,
+      "loss": 1.5883,
+      "step": 1178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2650401902604043,
+      "learning_rate": 7.446185385854445e-05,
+      "loss": 1.6039,
+      "step": 1179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2423595493228995,
+      "learning_rate": 7.444867476936594e-05,
+      "loss": 1.5925,
+      "step": 1180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21249249646756155,
+      "learning_rate": 7.443549568018744e-05,
+      "loss": 1.5979,
+      "step": 1181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24250610726543256,
+      "learning_rate": 7.442231659100894e-05,
+      "loss": 1.5969,
+      "step": 1182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2537655199876406,
+      "learning_rate": 7.440913750183044e-05,
+      "loss": 1.5933,
+      "step": 1183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21405956441015037,
+      "learning_rate": 7.439595841265193e-05,
+      "loss": 1.5954,
+      "step": 1184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16174367763094688,
+      "learning_rate": 7.438277932347342e-05,
+      "loss": 1.5946,
+      "step": 1185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17946717348892616,
+      "learning_rate": 7.436960023429493e-05,
+      "loss": 1.5911,
+      "step": 1186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20846968398819296,
+      "learning_rate": 7.435642114511641e-05,
+      "loss": 1.589,
+      "step": 1187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1856354988357816,
+      "learning_rate": 7.434324205593792e-05,
+      "loss": 1.5929,
+      "step": 1188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15578117731766,
+      "learning_rate": 7.433006296675941e-05,
+      "loss": 1.5966,
+      "step": 1189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1842436337816048,
+      "learning_rate": 7.43168838775809e-05,
+      "loss": 1.5932,
+      "step": 1190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20401450420216943,
+      "learning_rate": 7.430370478840241e-05,
+      "loss": 1.589,
+      "step": 1191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1571115083265962,
+      "learning_rate": 7.429052569922391e-05,
+      "loss": 1.5852,
+      "step": 1192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1406605676557017,
+      "learning_rate": 7.42773466100454e-05,
+      "loss": 1.5909,
+      "step": 1193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1581183982669551,
+      "learning_rate": 7.42641675208669e-05,
+      "loss": 1.5847,
+      "step": 1194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15952906110455567,
+      "learning_rate": 7.425098843168839e-05,
+      "loss": 1.5862,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15607263351309267,
+      "learning_rate": 7.423780934250988e-05,
+      "loss": 1.5937,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1310303481341675,
+      "learning_rate": 7.422463025333139e-05,
+      "loss": 1.594,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12984942531622481,
+      "learning_rate": 7.421145116415288e-05,
+      "loss": 1.5939,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14587642352300043,
+      "learning_rate": 7.419827207497438e-05,
+      "loss": 1.5882,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15198871500094838,
+      "learning_rate": 7.418509298579588e-05,
+      "loss": 1.59,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154450062692067,
+      "learning_rate": 7.417191389661738e-05,
+      "loss": 1.5888,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18039937998823308,
+      "learning_rate": 7.415873480743887e-05,
+      "loss": 1.6007,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18845018044860143,
+      "learning_rate": 7.414555571826037e-05,
+      "loss": 1.593,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17172832353041226,
+      "learning_rate": 7.413237662908186e-05,
+      "loss": 1.5918,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1581293047398025,
+      "learning_rate": 7.411919753990335e-05,
+      "loss": 1.6033,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16734747711724823,
+      "learning_rate": 7.410601845072486e-05,
+      "loss": 1.5808,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19019197233208932,
+      "learning_rate": 7.409283936154635e-05,
+      "loss": 1.5904,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19620446308614556,
+      "learning_rate": 7.407966027236785e-05,
+      "loss": 1.5917,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20898204213566193,
+      "learning_rate": 7.406648118318934e-05,
+      "loss": 1.5954,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24614257226734018,
+      "learning_rate": 7.405330209401085e-05,
+      "loss": 1.5927,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2653501164736441,
+      "learning_rate": 7.404012300483233e-05,
+      "loss": 1.5895,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25455637155885896,
+      "learning_rate": 7.402694391565384e-05,
+      "loss": 1.5808,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2219490743301624,
+      "learning_rate": 7.401376482647533e-05,
+      "loss": 1.5923,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14713310813171854,
+      "learning_rate": 7.400058573729682e-05,
+      "loss": 1.5842,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15660737876177702,
+      "learning_rate": 7.398740664811833e-05,
+      "loss": 1.5955,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1811379789536238,
+      "learning_rate": 7.397422755893982e-05,
+      "loss": 1.5913,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1961687660102907,
+      "learning_rate": 7.396104846976132e-05,
+      "loss": 1.5896,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17373689540381426,
+      "learning_rate": 7.394786938058281e-05,
+      "loss": 1.5866,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17773055257272066,
+      "learning_rate": 7.39346902914043e-05,
+      "loss": 1.5888,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16536252481789593,
+      "learning_rate": 7.39215112022258e-05,
+      "loss": 1.5763,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14193680341594905,
+      "learning_rate": 7.390833211304731e-05,
+      "loss": 1.5899,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16528380516967012,
+      "learning_rate": 7.38951530238688e-05,
+      "loss": 1.5827,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1585259135194698,
+      "learning_rate": 7.38819739346903e-05,
+      "loss": 1.5951,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14843359958766764,
+      "learning_rate": 7.38687948455118e-05,
+      "loss": 1.5966,
+      "step": 1224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16795811232997543,
+      "learning_rate": 7.38556157563333e-05,
+      "loss": 1.5822,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17063382746855169,
+      "learning_rate": 7.384243666715479e-05,
+      "loss": 1.5821,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15646631540555336,
+      "learning_rate": 7.382925757797628e-05,
+      "loss": 1.5925,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12663185635907337,
+      "learning_rate": 7.381607848879778e-05,
+      "loss": 1.5755,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13506195484438943,
+      "learning_rate": 7.380289939961927e-05,
+      "loss": 1.5856,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1355655412063337,
+      "learning_rate": 7.378972031044078e-05,
+      "loss": 1.5773,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15651881391255412,
+      "learning_rate": 7.377654122126227e-05,
+      "loss": 1.5865,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1577977275413669,
+      "learning_rate": 7.376336213208377e-05,
+      "loss": 1.5822,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17044005631971684,
+      "learning_rate": 7.375018304290526e-05,
+      "loss": 1.5856,
+      "step": 1233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18593965782849248,
+      "learning_rate": 7.373700395372677e-05,
+      "loss": 1.5838,
+      "step": 1234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1939373178548821,
+      "learning_rate": 7.372382486454825e-05,
+      "loss": 1.5948,
+      "step": 1235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20068033024298984,
+      "learning_rate": 7.371064577536975e-05,
+      "loss": 1.5944,
+      "step": 1236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17164694369691896,
+      "learning_rate": 7.369746668619125e-05,
+      "loss": 1.584,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1436238943749497,
+      "learning_rate": 7.368428759701274e-05,
+      "loss": 1.5878,
+      "step": 1238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15172544100061827,
+      "learning_rate": 7.367110850783425e-05,
+      "loss": 1.5892,
+      "step": 1239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15646725821374724,
+      "learning_rate": 7.365792941865574e-05,
+      "loss": 1.5964,
+      "step": 1240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15865900762765828,
+      "learning_rate": 7.364475032947724e-05,
+      "loss": 1.5952,
+      "step": 1241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.18082519774905434,
+      "learning_rate": 7.363157124029873e-05,
+      "loss": 1.5729,
+      "step": 1242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20299924734651137,
+      "learning_rate": 7.361839215112022e-05,
+      "loss": 1.5865,
+      "step": 1243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19780510869427326,
+      "learning_rate": 7.360521306194172e-05,
+      "loss": 1.5848,
+      "step": 1244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1783130384452761,
+      "learning_rate": 7.359203397276322e-05,
+      "loss": 1.5903,
+      "step": 1245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15261926702540918,
+      "learning_rate": 7.357885488358472e-05,
+      "loss": 1.5879,
+      "step": 1246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13783811582529387,
+      "learning_rate": 7.356567579440621e-05,
+      "loss": 1.5747,
+      "step": 1247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1467529651088208,
+      "learning_rate": 7.355249670522772e-05,
+      "loss": 1.5808,
+      "step": 1248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16377764961345861,
+      "learning_rate": 7.353931761604921e-05,
+      "loss": 1.58,
+      "step": 1249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15000263884955514,
+      "learning_rate": 7.352613852687069e-05,
+      "loss": 1.5845,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14171089245368748,
+      "learning_rate": 7.35129594376922e-05,
+      "loss": 1.5919,
+      "step": 1251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20435307840612746,
+      "learning_rate": 7.34997803485137e-05,
+      "loss": 1.5867,
+      "step": 1252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24045885356362473,
+      "learning_rate": 7.348660125933519e-05,
+      "loss": 1.584,
+      "step": 1253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28740932434468447,
+      "learning_rate": 7.34734221701567e-05,
+      "loss": 1.5936,
+      "step": 1254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3196108046291853,
+      "learning_rate": 7.346024308097819e-05,
+      "loss": 1.5997,
+      "step": 1255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2979197316242365,
+      "learning_rate": 7.344706399179968e-05,
+      "loss": 1.5838,
+      "step": 1256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22072665514336615,
+      "learning_rate": 7.343388490262118e-05,
+      "loss": 1.5819,
+      "step": 1257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15552888234227324,
+      "learning_rate": 7.342070581344267e-05,
+      "loss": 1.5833,
+      "step": 1258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20544194277297625,
+      "learning_rate": 7.340752672426416e-05,
+      "loss": 1.5921,
+      "step": 1259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26303138800180975,
+      "learning_rate": 7.339434763508567e-05,
+      "loss": 1.5859,
+      "step": 1260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22397522066945746,
+      "learning_rate": 7.338116854590716e-05,
+      "loss": 1.5841,
+      "step": 1261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.14991417239420035,
+      "learning_rate": 7.336798945672866e-05,
+      "loss": 1.5841,
+      "step": 1262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16359585502730267,
+      "learning_rate": 7.335481036755017e-05,
+      "loss": 1.5816,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19466431015661523,
+      "learning_rate": 7.334163127837166e-05,
+      "loss": 1.5828,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17089651320335977,
+      "learning_rate": 7.332845218919315e-05,
+      "loss": 1.5876,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13925314072410105,
+      "learning_rate": 7.331527310001465e-05,
+      "loss": 1.5885,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14834175295114693,
+      "learning_rate": 7.330209401083614e-05,
+      "loss": 1.5785,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1699156449333421,
+      "learning_rate": 7.328891492165763e-05,
+      "loss": 1.5804,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15496535500653283,
+      "learning_rate": 7.327573583247914e-05,
+      "loss": 1.5786,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1556764604070721,
+      "learning_rate": 7.326255674330064e-05,
+      "loss": 1.5873,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18129236452074904,
+      "learning_rate": 7.324937765412213e-05,
+      "loss": 1.5908,
+      "step": 1271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1572516060188371,
+      "learning_rate": 7.323619856494362e-05,
+      "loss": 1.5834,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15305720865121053,
+      "learning_rate": 7.322301947576513e-05,
+      "loss": 1.5848,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1551328911651992,
+      "learning_rate": 7.320984038658661e-05,
+      "loss": 1.5902,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1577934048703626,
+      "learning_rate": 7.319666129740812e-05,
+      "loss": 1.5794,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16434049110207993,
+      "learning_rate": 7.318348220822961e-05,
+      "loss": 1.5783,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16776089049627255,
+      "learning_rate": 7.31703031190511e-05,
+      "loss": 1.5768,
+      "step": 1277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16243702089894851,
+      "learning_rate": 7.315712402987261e-05,
+      "loss": 1.5826,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1536195883355736,
+      "learning_rate": 7.31439449406941e-05,
+      "loss": 1.5788,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17100276923275118,
+      "learning_rate": 7.31307658515156e-05,
+      "loss": 1.5801,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17375573905803932,
+      "learning_rate": 7.31175867623371e-05,
+      "loss": 1.5781,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1737765212522266,
+      "learning_rate": 7.310440767315859e-05,
+      "loss": 1.5841,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1557338864747223,
+      "learning_rate": 7.309122858398008e-05,
+      "loss": 1.5839,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1514952769269247,
+      "learning_rate": 7.307804949480159e-05,
+      "loss": 1.5838,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15888225905146014,
+      "learning_rate": 7.306487040562308e-05,
+      "loss": 1.5728,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15322549061770394,
+      "learning_rate": 7.305169131644458e-05,
+      "loss": 1.5807,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16366336204652251,
+      "learning_rate": 7.303851222726608e-05,
+      "loss": 1.5878,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16799314356241987,
+      "learning_rate": 7.302533313808758e-05,
+      "loss": 1.5737,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15569342409500003,
+      "learning_rate": 7.301215404890907e-05,
+      "loss": 1.5863,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13541629448681836,
+      "learning_rate": 7.299897495973056e-05,
+      "loss": 1.5886,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1427857963080412,
+      "learning_rate": 7.298579587055206e-05,
+      "loss": 1.5741,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1485360781927371,
+      "learning_rate": 7.297261678137355e-05,
+      "loss": 1.5666,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15300033477123595,
+      "learning_rate": 7.295943769219506e-05,
+      "loss": 1.5881,
+      "step": 1293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1523533071984983,
+      "learning_rate": 7.294625860301655e-05,
+      "loss": 1.5833,
+      "step": 1294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14366735707085823,
+      "learning_rate": 7.293307951383805e-05,
+      "loss": 1.5852,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15488474790280232,
+      "learning_rate": 7.291990042465954e-05,
+      "loss": 1.5764,
+      "step": 1296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17693604786333558,
+      "learning_rate": 7.290672133548105e-05,
+      "loss": 1.5823,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2165069651574788,
+      "learning_rate": 7.289354224630253e-05,
+      "loss": 1.5885,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2108685441610128,
+      "learning_rate": 7.288036315712404e-05,
+      "loss": 1.5763,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19114050529562934,
+      "learning_rate": 7.286718406794553e-05,
+      "loss": 1.5822,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21363514589823596,
+      "learning_rate": 7.285400497876702e-05,
+      "loss": 1.5893,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2437901279010662,
+      "learning_rate": 7.284082588958853e-05,
+      "loss": 1.5778,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26098162568490296,
+      "learning_rate": 7.282764680041002e-05,
+      "loss": 1.5849,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25540292032332557,
+      "learning_rate": 7.281446771123152e-05,
+      "loss": 1.5831,
+      "step": 1304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21631749808313838,
+      "learning_rate": 7.280128862205301e-05,
+      "loss": 1.5734,
+      "step": 1305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1718013292626418,
+      "learning_rate": 7.27881095328745e-05,
+      "loss": 1.5828,
+      "step": 1306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13719025908171398,
+      "learning_rate": 7.2774930443696e-05,
+      "loss": 1.5808,
+      "step": 1307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1690183237112528,
+      "learning_rate": 7.27617513545175e-05,
+      "loss": 1.5817,
+      "step": 1308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1848978593010119,
+      "learning_rate": 7.2748572265339e-05,
+      "loss": 1.5762,
+      "step": 1309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15635718310334223,
+      "learning_rate": 7.27353931761605e-05,
+      "loss": 1.5757,
+      "step": 1310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14725817433618338,
+      "learning_rate": 7.2722214086982e-05,
+      "loss": 1.591,
+      "step": 1311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1512466403819289,
+      "learning_rate": 7.27090349978035e-05,
+      "loss": 1.5847,
+      "step": 1312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1335054707530398,
+      "learning_rate": 7.269585590862499e-05,
+      "loss": 1.5731,
+      "step": 1313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13882677911813507,
+      "learning_rate": 7.268267681944648e-05,
+      "loss": 1.5797,
+      "step": 1314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15268615715918724,
+      "learning_rate": 7.266949773026798e-05,
+      "loss": 1.5853,
+      "step": 1315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1488126658370855,
+      "learning_rate": 7.265631864108947e-05,
+      "loss": 1.5849,
+      "step": 1316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15007617848273547,
+      "learning_rate": 7.264313955191098e-05,
+      "loss": 1.5705,
+      "step": 1317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1584098535147045,
+      "learning_rate": 7.262996046273247e-05,
+      "loss": 1.5852,
+      "step": 1318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19072950094070457,
+      "learning_rate": 7.261678137355396e-05,
+      "loss": 1.5787,
+      "step": 1319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.18522875418280946,
+      "learning_rate": 7.260360228437546e-05,
+      "loss": 1.586,
+      "step": 1320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16446527308539624,
+      "learning_rate": 7.259042319519697e-05,
+      "loss": 1.5847,
+      "step": 1321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13874396432926306,
+      "learning_rate": 7.257724410601845e-05,
+      "loss": 1.5756,
+      "step": 1322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13928631137367814,
+      "learning_rate": 7.256406501683995e-05,
+      "loss": 1.5738,
+      "step": 1323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1673872997735078,
+      "learning_rate": 7.255088592766145e-05,
+      "loss": 1.5789,
+      "step": 1324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17153496675553456,
+      "learning_rate": 7.253770683848294e-05,
+      "loss": 1.5791,
+      "step": 1325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1465187797607589,
+      "learning_rate": 7.252452774930445e-05,
+      "loss": 1.5703,
+      "step": 1326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1351520271793626,
+      "learning_rate": 7.251134866012594e-05,
+      "loss": 1.5778,
+      "step": 1327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13031490453618944,
+      "learning_rate": 7.249816957094744e-05,
+      "loss": 1.5792,
+      "step": 1328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1262777867945403,
+      "learning_rate": 7.248499048176893e-05,
+      "loss": 1.574,
+      "step": 1329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13212997342047494,
+      "learning_rate": 7.247181139259042e-05,
+      "loss": 1.58,
+      "step": 1330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14500318487143435,
+      "learning_rate": 7.245863230341192e-05,
+      "loss": 1.5681,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17875663213637794,
+      "learning_rate": 7.244545321423342e-05,
+      "loss": 1.5719,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20287564117034,
+      "learning_rate": 7.243227412505492e-05,
+      "loss": 1.5783,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22538319404257154,
+      "learning_rate": 7.241909503587641e-05,
+      "loss": 1.5799,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24510429626641428,
+      "learning_rate": 7.240591594669792e-05,
+      "loss": 1.5688,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.21758774192310035,
+      "learning_rate": 7.239273685751941e-05,
+      "loss": 1.5748,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17046856072667518,
+      "learning_rate": 7.23795577683409e-05,
+      "loss": 1.5744,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14997523013715272,
+      "learning_rate": 7.23663786791624e-05,
+      "loss": 1.5764,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17864699485620184,
+      "learning_rate": 7.23531995899839e-05,
+      "loss": 1.5676,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19732979726685498,
+      "learning_rate": 7.234002050080539e-05,
+      "loss": 1.5708,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17655391742101245,
+      "learning_rate": 7.23268414116269e-05,
+      "loss": 1.5713,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13019863917633345,
+      "learning_rate": 7.231366232244839e-05,
+      "loss": 1.5834,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.163754576536066,
+      "learning_rate": 7.230048323326988e-05,
+      "loss": 1.5819,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22301870470160576,
+      "learning_rate": 7.228730414409138e-05,
+      "loss": 1.5806,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22979909365643392,
+      "learning_rate": 7.227412505491288e-05,
+      "loss": 1.5801,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18015517540086598,
+      "learning_rate": 7.226094596573436e-05,
+      "loss": 1.5767,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1557837416424971,
+      "learning_rate": 7.224776687655587e-05,
+      "loss": 1.5728,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1669198000676956,
+      "learning_rate": 7.223458778737736e-05,
+      "loss": 1.5767,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15973163251561107,
+      "learning_rate": 7.222140869819886e-05,
+      "loss": 1.5817,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16671599564277967,
+      "learning_rate": 7.220822960902037e-05,
+      "loss": 1.5737,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15068265123287375,
+      "learning_rate": 7.219505051984186e-05,
+      "loss": 1.5796,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1463851065191522,
+      "learning_rate": 7.218187143066335e-05,
+      "loss": 1.5666,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14776418821561846,
+      "learning_rate": 7.216869234148485e-05,
+      "loss": 1.5812,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16447130345045846,
+      "learning_rate": 7.215551325230634e-05,
+      "loss": 1.584,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1488498411391017,
+      "learning_rate": 7.214233416312783e-05,
+      "loss": 1.5732,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14942787853682804,
+      "learning_rate": 7.212915507394934e-05,
+      "loss": 1.5822,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.169962466274362,
+      "learning_rate": 7.211597598477084e-05,
+      "loss": 1.5807,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17265910173066412,
+      "learning_rate": 7.210279689559233e-05,
+      "loss": 1.5732,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16447068445486143,
+      "learning_rate": 7.208961780641382e-05,
+      "loss": 1.573,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1382385456580122,
+      "learning_rate": 7.207643871723533e-05,
+      "loss": 1.5773,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15127923245359043,
+      "learning_rate": 7.206325962805682e-05,
+      "loss": 1.575,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14745332967548117,
+      "learning_rate": 7.205008053887832e-05,
+      "loss": 1.572,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1409200003691044,
+      "learning_rate": 7.203690144969981e-05,
+      "loss": 1.5764,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1602133318398135,
+      "learning_rate": 7.20237223605213e-05,
+      "loss": 1.58,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1808418713819235,
+      "learning_rate": 7.201054327134281e-05,
+      "loss": 1.5771,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18480911250072243,
+      "learning_rate": 7.19973641821643e-05,
+      "loss": 1.5766,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19252559378380835,
+      "learning_rate": 7.19841850929858e-05,
+      "loss": 1.5743,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20052408124118393,
+      "learning_rate": 7.19710060038073e-05,
+      "loss": 1.5709,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17837119621041397,
+      "learning_rate": 7.19578269146288e-05,
+      "loss": 1.5734,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14591392612986573,
+      "learning_rate": 7.194464782545028e-05,
+      "loss": 1.5716,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14918642258613896,
+      "learning_rate": 7.193146873627179e-05,
+      "loss": 1.5718,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1433548846051071,
+      "learning_rate": 7.191828964709328e-05,
+      "loss": 1.5796,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.17948405788478852,
+      "learning_rate": 7.190511055791478e-05,
+      "loss": 1.5835,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2310307534670035,
+      "learning_rate": 7.189193146873628e-05,
+      "loss": 1.5771,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25695389114643785,
+      "learning_rate": 7.187875237955778e-05,
+      "loss": 1.569,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25999039315686817,
+      "learning_rate": 7.186557329037927e-05,
+      "loss": 1.5738,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22576403415131382,
+      "learning_rate": 7.185239420120076e-05,
+      "loss": 1.5804,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1845921032220699,
+      "learning_rate": 7.183921511202226e-05,
+      "loss": 1.5703,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18950643124102381,
+      "learning_rate": 7.182603602284375e-05,
+      "loss": 1.5761,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22317918435110773,
+      "learning_rate": 7.181285693366526e-05,
+      "loss": 1.5757,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2454572265669023,
+      "learning_rate": 7.179967784448675e-05,
+      "loss": 1.576,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20188147907095785,
+      "learning_rate": 7.178649875530825e-05,
+      "loss": 1.566,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.140275625579357,
+      "learning_rate": 7.177331966612974e-05,
+      "loss": 1.5719,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18452131194815793,
+      "learning_rate": 7.176014057695125e-05,
+      "loss": 1.5657,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.22651983400307935,
+      "learning_rate": 7.174696148777274e-05,
+      "loss": 1.5739,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.19651377875239404,
+      "learning_rate": 7.173378239859424e-05,
+      "loss": 1.5773,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13581992227920534,
+      "learning_rate": 7.172060330941573e-05,
+      "loss": 1.5669,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1711417241102865,
+      "learning_rate": 7.170742422023722e-05,
+      "loss": 1.5708,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1970811470534912,
+      "learning_rate": 7.169424513105873e-05,
+      "loss": 1.5707,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1730218026826818,
+      "learning_rate": 7.168106604188022e-05,
+      "loss": 1.5753,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14249391458274216,
+      "learning_rate": 7.166788695270172e-05,
+      "loss": 1.5703,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15576625474276548,
+      "learning_rate": 7.165470786352321e-05,
+      "loss": 1.5658,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16844771524906862,
+      "learning_rate": 7.164152877434472e-05,
+      "loss": 1.5638,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14842972901098414,
+      "learning_rate": 7.16283496851662e-05,
+      "loss": 1.5684,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13938539030658934,
+      "learning_rate": 7.16151705959877e-05,
+      "loss": 1.5826,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.150557903672672,
+      "learning_rate": 7.16019915068092e-05,
+      "loss": 1.5641,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15543635281372953,
+      "learning_rate": 7.15888124176307e-05,
+      "loss": 1.5696,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1665289062823907,
+      "learning_rate": 7.15756333284522e-05,
+      "loss": 1.5667,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20038629439420957,
+      "learning_rate": 7.15624542392737e-05,
+      "loss": 1.5738,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.20404987893441098,
+      "learning_rate": 7.154927515009519e-05,
+      "loss": 1.5641,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.19056674005837818,
+      "learning_rate": 7.153609606091668e-05,
+      "loss": 1.5789,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15936990939657109,
+      "learning_rate": 7.152291697173818e-05,
+      "loss": 1.5727,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1327675837335293,
+      "learning_rate": 7.150973788255967e-05,
+      "loss": 1.5784,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1410578829295748,
+      "learning_rate": 7.149655879338118e-05,
+      "loss": 1.5759,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16837905654706764,
+      "learning_rate": 7.148337970420267e-05,
+      "loss": 1.5678,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1576722100754528,
+      "learning_rate": 7.147020061502416e-05,
+      "loss": 1.5652,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15260049321124103,
+      "learning_rate": 7.145702152584566e-05,
+      "loss": 1.5649,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16787848023479865,
+      "learning_rate": 7.144384243666717e-05,
+      "loss": 1.5709,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15351278195144266,
+      "learning_rate": 7.143066334748866e-05,
+      "loss": 1.5652,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.141387994112907,
+      "learning_rate": 7.141748425831015e-05,
+      "loss": 1.5714,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13817472717587612,
+      "learning_rate": 7.140430516913165e-05,
+      "loss": 1.5696,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14708798930634917,
+      "learning_rate": 7.139112607995314e-05,
+      "loss": 1.5721,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16574803886961384,
+      "learning_rate": 7.137794699077465e-05,
+      "loss": 1.5788,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16999064687089108,
+      "learning_rate": 7.136476790159614e-05,
+      "loss": 1.5711,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15401061336166366,
+      "learning_rate": 7.135158881241763e-05,
+      "loss": 1.5694,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1522428976991431,
+      "learning_rate": 7.133840972323913e-05,
+      "loss": 1.5768,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1620006742059713,
+      "learning_rate": 7.132523063406064e-05,
+      "loss": 1.5684,
+      "step": 1417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14184632294364613,
+      "learning_rate": 7.131205154488212e-05,
+      "loss": 1.5677,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12851766257232203,
+      "learning_rate": 7.129887245570362e-05,
+      "loss": 1.566,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13223607719411584,
+      "learning_rate": 7.128569336652512e-05,
+      "loss": 1.5701,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13725326932964607,
+      "learning_rate": 7.127251427734661e-05,
+      "loss": 1.5767,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1402305033389597,
+      "learning_rate": 7.125933518816812e-05,
+      "loss": 1.565,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14563085034581674,
+      "learning_rate": 7.124615609898961e-05,
+      "loss": 1.5727,
+      "step": 1423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15859311190689154,
+      "learning_rate": 7.12329770098111e-05,
+      "loss": 1.5723,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1640105183898524,
+      "learning_rate": 7.12197979206326e-05,
+      "loss": 1.5712,
+      "step": 1425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16108347576592652,
+      "learning_rate": 7.120661883145409e-05,
+      "loss": 1.5634,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14570669209087186,
+      "learning_rate": 7.119343974227559e-05,
+      "loss": 1.5625,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13374217351027243,
+      "learning_rate": 7.11802606530971e-05,
+      "loss": 1.561,
+      "step": 1428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13131168220415707,
+      "learning_rate": 7.116708156391859e-05,
+      "loss": 1.5586,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14256260061186307,
+      "learning_rate": 7.115390247474008e-05,
+      "loss": 1.5623,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13433942314754024,
+      "learning_rate": 7.114072338556158e-05,
+      "loss": 1.5644,
+      "step": 1431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12327503612115467,
+      "learning_rate": 7.112754429638308e-05,
+      "loss": 1.5728,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1304444249984789,
+      "learning_rate": 7.111436520720458e-05,
+      "loss": 1.5717,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1372735123829869,
+      "learning_rate": 7.110118611802607e-05,
+      "loss": 1.5749,
+      "step": 1434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15481059663823463,
+      "learning_rate": 7.108800702884756e-05,
+      "loss": 1.5728,
+      "step": 1435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1727623966582496,
+      "learning_rate": 7.107482793966906e-05,
+      "loss": 1.5639,
+      "step": 1436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.19401076610760798,
+      "learning_rate": 7.106164885049056e-05,
+      "loss": 1.5684,
+      "step": 1437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2156803932249311,
+      "learning_rate": 7.104846976131206e-05,
+      "loss": 1.5714,
+      "step": 1438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24393826691213036,
+      "learning_rate": 7.103529067213355e-05,
+      "loss": 1.5677,
+      "step": 1439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24729171934390767,
+      "learning_rate": 7.102211158295505e-05,
+      "loss": 1.5666,
+      "step": 1440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.20257547505533538,
+      "learning_rate": 7.100893249377655e-05,
+      "loss": 1.5774,
+      "step": 1441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1381135615527873,
+      "learning_rate": 7.099575340459803e-05,
+      "loss": 1.5727,
+      "step": 1442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13699646040366453,
+      "learning_rate": 7.098257431541954e-05,
+      "loss": 1.5675,
+      "step": 1443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1898081051960504,
+      "learning_rate": 7.096939522624103e-05,
+      "loss": 1.5647,
+      "step": 1444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21299109026323154,
+      "learning_rate": 7.095621613706253e-05,
+      "loss": 1.5707,
+      "step": 1445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.20428555413508584,
+      "learning_rate": 7.094303704788402e-05,
+      "loss": 1.5708,
+      "step": 1446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1560001862551339,
+      "learning_rate": 7.092985795870553e-05,
+      "loss": 1.5583,
+      "step": 1447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16279311630157717,
+      "learning_rate": 7.091667886952702e-05,
+      "loss": 1.5666,
+      "step": 1448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2254152694317573,
+      "learning_rate": 7.090349978034852e-05,
+      "loss": 1.5731,
+      "step": 1449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23812366301388427,
+      "learning_rate": 7.089032069117001e-05,
+      "loss": 1.5583,
+      "step": 1450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23617513876866209,
+      "learning_rate": 7.08771416019915e-05,
+      "loss": 1.5721,
+      "step": 1451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2107361838293821,
+      "learning_rate": 7.086396251281301e-05,
+      "loss": 1.5634,
+      "step": 1452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1723002286014294,
+      "learning_rate": 7.08507834236345e-05,
+      "loss": 1.5671,
+      "step": 1453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14627668143930986,
+      "learning_rate": 7.0837604334456e-05,
+      "loss": 1.568,
+      "step": 1454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1336544861801278,
+      "learning_rate": 7.082442524527749e-05,
+      "loss": 1.5683,
+      "step": 1455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14421751195441476,
+      "learning_rate": 7.0811246156099e-05,
+      "loss": 1.5617,
+      "step": 1456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15430975229793042,
+      "learning_rate": 7.07980670669205e-05,
+      "loss": 1.5667,
+      "step": 1457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16660631410876542,
+      "learning_rate": 7.078488797774199e-05,
+      "loss": 1.5632,
+      "step": 1458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15222909256571887,
+      "learning_rate": 7.077170888856348e-05,
+      "loss": 1.564,
+      "step": 1459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14623520946095997,
+      "learning_rate": 7.075852979938498e-05,
+      "loss": 1.5575,
+      "step": 1460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1440011629967226,
+      "learning_rate": 7.074535071020648e-05,
+      "loss": 1.5597,
+      "step": 1461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1469726202146802,
+      "learning_rate": 7.073217162102798e-05,
+      "loss": 1.5662,
+      "step": 1462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13611688819390158,
+      "learning_rate": 7.071899253184947e-05,
+      "loss": 1.5689,
+      "step": 1463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13792597129053707,
+      "learning_rate": 7.070581344267096e-05,
+      "loss": 1.563,
+      "step": 1464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1453080159775326,
+      "learning_rate": 7.069263435349247e-05,
+      "loss": 1.5589,
+      "step": 1465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1483809349019845,
+      "learning_rate": 7.067945526431395e-05,
+      "loss": 1.5655,
+      "step": 1466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16564527303263538,
+      "learning_rate": 7.066627617513546e-05,
+      "loss": 1.5599,
+      "step": 1467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1759936696979058,
+      "learning_rate": 7.065309708595695e-05,
+      "loss": 1.5643,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17115919022752862,
+      "learning_rate": 7.063991799677845e-05,
+      "loss": 1.5635,
+      "step": 1469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15619582944022972,
+      "learning_rate": 7.062673890759994e-05,
+      "loss": 1.5671,
+      "step": 1470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14548556726946826,
+      "learning_rate": 7.061355981842145e-05,
+      "loss": 1.5658,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14857073163622167,
+      "learning_rate": 7.060038072924294e-05,
+      "loss": 1.5647,
+      "step": 1472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15690757349631196,
+      "learning_rate": 7.058720164006443e-05,
+      "loss": 1.5606,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16896840074122582,
+      "learning_rate": 7.057402255088593e-05,
+      "loss": 1.5589,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20333875962129536,
+      "learning_rate": 7.056084346170742e-05,
+      "loss": 1.5637,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20502555973413897,
+      "learning_rate": 7.054766437252893e-05,
+      "loss": 1.5543,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20543728861342644,
+      "learning_rate": 7.053448528335042e-05,
+      "loss": 1.5653,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20358492809509288,
+      "learning_rate": 7.052130619417192e-05,
+      "loss": 1.5601,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1722967696401652,
+      "learning_rate": 7.050812710499341e-05,
+      "loss": 1.5684,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14801484555166847,
+      "learning_rate": 7.049494801581492e-05,
+      "loss": 1.5683,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14707235536972355,
+      "learning_rate": 7.048176892663641e-05,
+      "loss": 1.5618,
+      "step": 1481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17503038537922339,
+      "learning_rate": 7.04685898374579e-05,
+      "loss": 1.5639,
+      "step": 1482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18686745881119146,
+      "learning_rate": 7.04554107482794e-05,
+      "loss": 1.5591,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17060754128859854,
+      "learning_rate": 7.044223165910089e-05,
+      "loss": 1.567,
+      "step": 1484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16148218226752092,
+      "learning_rate": 7.04290525699224e-05,
+      "loss": 1.5624,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12392433723907219,
+      "learning_rate": 7.04158734807439e-05,
+      "loss": 1.563,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12980946855498254,
+      "learning_rate": 7.040269439156539e-05,
+      "loss": 1.5632,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14339395974650324,
+      "learning_rate": 7.038951530238688e-05,
+      "loss": 1.5669,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1448792621127726,
+      "learning_rate": 7.037633621320839e-05,
+      "loss": 1.5609,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13553794582560605,
+      "learning_rate": 7.036315712402987e-05,
+      "loss": 1.5622,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13575873572440034,
+      "learning_rate": 7.034997803485138e-05,
+      "loss": 1.562,
+      "step": 1491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16082133760741707,
+      "learning_rate": 7.033679894567287e-05,
+      "loss": 1.5603,
+      "step": 1492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17242897281518704,
+      "learning_rate": 7.032361985649436e-05,
+      "loss": 1.5694,
+      "step": 1493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18542150498618903,
+      "learning_rate": 7.031044076731586e-05,
+      "loss": 1.5732,
+      "step": 1494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.19823017532318143,
+      "learning_rate": 7.029726167813736e-05,
+      "loss": 1.5587,
+      "step": 1495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1727625831103123,
+      "learning_rate": 7.028408258895886e-05,
+      "loss": 1.5661,
+      "step": 1496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14792025021664829,
+      "learning_rate": 7.027090349978035e-05,
+      "loss": 1.5613,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14369828058714515,
+      "learning_rate": 7.025772441060185e-05,
+      "loss": 1.557,
+      "step": 1498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13415550412624946,
+      "learning_rate": 7.024454532142334e-05,
+      "loss": 1.5696,
+      "step": 1499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12605195451631998,
+      "learning_rate": 7.023136623224485e-05,
+      "loss": 1.5578,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13889230083947712,
+      "learning_rate": 7.021818714306634e-05,
+      "loss": 1.556,
+      "step": 1501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1625570831758341,
+      "learning_rate": 7.020500805388783e-05,
+      "loss": 1.566,
+      "step": 1502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.159051736431182,
+      "learning_rate": 7.019182896470933e-05,
+      "loss": 1.5681,
+      "step": 1503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1391939881062468,
+      "learning_rate": 7.017864987553084e-05,
+      "loss": 1.5636,
+      "step": 1504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12197288438358046,
+      "learning_rate": 7.016547078635233e-05,
+      "loss": 1.5666,
+      "step": 1505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15668727220869816,
+      "learning_rate": 7.015229169717382e-05,
+      "loss": 1.5656,
+      "step": 1506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20314314799885375,
+      "learning_rate": 7.013911260799532e-05,
+      "loss": 1.5607,
+      "step": 1507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22504397291844602,
+      "learning_rate": 7.012593351881681e-05,
+      "loss": 1.5525,
+      "step": 1508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20827134396232042,
+      "learning_rate": 7.011275442963832e-05,
+      "loss": 1.5645,
+      "step": 1509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16572765862670633,
+      "learning_rate": 7.009957534045981e-05,
+      "loss": 1.5626,
+      "step": 1510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14238041254108588,
+      "learning_rate": 7.00863962512813e-05,
+      "loss": 1.5617,
+      "step": 1511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14194615976277775,
+      "learning_rate": 7.00732171621028e-05,
+      "loss": 1.564,
+      "step": 1512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1746649614952692,
+      "learning_rate": 7.00600380729243e-05,
+      "loss": 1.5662,
+      "step": 1513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17817314772801265,
+      "learning_rate": 7.004685898374579e-05,
+      "loss": 1.5567,
+      "step": 1514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1627679458697912,
+      "learning_rate": 7.00336798945673e-05,
+      "loss": 1.5641,
+      "step": 1515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15787514391777924,
+      "learning_rate": 7.002050080538879e-05,
+      "loss": 1.5632,
+      "step": 1516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15767476971003588,
+      "learning_rate": 7.000732171621028e-05,
+      "loss": 1.5689,
+      "step": 1517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1597712748945363,
+      "learning_rate": 6.999414262703177e-05,
+      "loss": 1.5524,
+      "step": 1518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14274879675979987,
+      "learning_rate": 6.998096353785328e-05,
+      "loss": 1.5627,
+      "step": 1519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13862167988922744,
+      "learning_rate": 6.996778444867478e-05,
+      "loss": 1.5619,
+      "step": 1520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1504538492286387,
+      "learning_rate": 6.995460535949627e-05,
+      "loss": 1.5574,
+      "step": 1521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15973123245612725,
+      "learning_rate": 6.994142627031776e-05,
+      "loss": 1.5571,
+      "step": 1522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1451140351390426,
+      "learning_rate": 6.992824718113926e-05,
+      "loss": 1.5565,
+      "step": 1523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14241471019026736,
+      "learning_rate": 6.991506809196076e-05,
+      "loss": 1.5607,
+      "step": 1524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16554926350590574,
+      "learning_rate": 6.990188900278226e-05,
+      "loss": 1.5662,
+      "step": 1525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18782873264693847,
+      "learning_rate": 6.988870991360375e-05,
+      "loss": 1.5574,
+      "step": 1526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21976343522446412,
+      "learning_rate": 6.987553082442525e-05,
+      "loss": 1.5565,
+      "step": 1527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2202794534236301,
+      "learning_rate": 6.986235173524675e-05,
+      "loss": 1.5581,
+      "step": 1528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18906491039358303,
+      "learning_rate": 6.984917264606825e-05,
+      "loss": 1.5616,
+      "step": 1529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1917135397649,
+      "learning_rate": 6.983599355688974e-05,
+      "loss": 1.5685,
+      "step": 1530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22376646705566025,
+      "learning_rate": 6.982281446771123e-05,
+      "loss": 1.5546,
+      "step": 1531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2557629632843123,
+      "learning_rate": 6.980963537853273e-05,
+      "loss": 1.567,
+      "step": 1532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26264014512059436,
+      "learning_rate": 6.979645628935422e-05,
+      "loss": 1.5558,
+      "step": 1533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23630888222603547,
+      "learning_rate": 6.978327720017573e-05,
+      "loss": 1.5639,
+      "step": 1534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18781426785491603,
+      "learning_rate": 6.977009811099722e-05,
+      "loss": 1.5656,
+      "step": 1535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1350962887962464,
+      "learning_rate": 6.975691902181872e-05,
+      "loss": 1.5589,
+      "step": 1536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15163336603393449,
+      "learning_rate": 6.974373993264022e-05,
+      "loss": 1.5632,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18782571169299847,
+      "learning_rate": 6.97305608434617e-05,
+      "loss": 1.5544,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18238881835124143,
+      "learning_rate": 6.971738175428321e-05,
+      "loss": 1.5558,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14843644500290679,
+      "learning_rate": 6.97042026651047e-05,
+      "loss": 1.5525,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14966300924676834,
+      "learning_rate": 6.96910235759262e-05,
+      "loss": 1.5652,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15599352551366935,
+      "learning_rate": 6.967784448674769e-05,
+      "loss": 1.5456,
+      "step": 1542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15698421958524125,
+      "learning_rate": 6.96646653975692e-05,
+      "loss": 1.5608,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13827373140596677,
+      "learning_rate": 6.96514863083907e-05,
+      "loss": 1.5597,
+      "step": 1544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11917501235585805,
+      "learning_rate": 6.963830721921219e-05,
+      "loss": 1.5675,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13790903649987726,
+      "learning_rate": 6.962512813003368e-05,
+      "loss": 1.553,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1388510888327217,
+      "learning_rate": 6.961194904085517e-05,
+      "loss": 1.5513,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14143505987201108,
+      "learning_rate": 6.959876995167668e-05,
+      "loss": 1.5677,
+      "step": 1548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12073834211466171,
+      "learning_rate": 6.958559086249818e-05,
+      "loss": 1.5603,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11587208279760582,
+      "learning_rate": 6.957241177331967e-05,
+      "loss": 1.5601,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13943580533691913,
+      "learning_rate": 6.955923268414116e-05,
+      "loss": 1.555,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13699092615514738,
+      "learning_rate": 6.954605359496267e-05,
+      "loss": 1.5589,
+      "step": 1552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13830957900595023,
+      "learning_rate": 6.953287450578416e-05,
+      "loss": 1.5535,
+      "step": 1553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15082247073447994,
+      "learning_rate": 6.951969541660566e-05,
+      "loss": 1.5615,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15495371687796394,
+      "learning_rate": 6.950651632742715e-05,
+      "loss": 1.5549,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15437640247134612,
+      "learning_rate": 6.949333723824865e-05,
+      "loss": 1.5552,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14173062168904724,
+      "learning_rate": 6.948015814907014e-05,
+      "loss": 1.5542,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13055811798338282,
+      "learning_rate": 6.946697905989165e-05,
+      "loss": 1.5582,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12438424333989716,
+      "learning_rate": 6.945379997071314e-05,
+      "loss": 1.5665,
+      "step": 1559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14003970899149606,
+      "learning_rate": 6.944062088153463e-05,
+      "loss": 1.5609,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13391975312221754,
+      "learning_rate": 6.942744179235614e-05,
+      "loss": 1.5572,
+      "step": 1561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13832258206726172,
+      "learning_rate": 6.941426270317762e-05,
+      "loss": 1.5555,
+      "step": 1562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17016478039826613,
+      "learning_rate": 6.940108361399913e-05,
+      "loss": 1.5565,
+      "step": 1563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1906472940652705,
+      "learning_rate": 6.938790452482062e-05,
+      "loss": 1.5606,
+      "step": 1564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18395224804146684,
+      "learning_rate": 6.937472543564212e-05,
+      "loss": 1.5533,
+      "step": 1565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1670491650995815,
+      "learning_rate": 6.936154634646361e-05,
+      "loss": 1.5517,
+      "step": 1566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15203633399251587,
+      "learning_rate": 6.934836725728512e-05,
+      "loss": 1.5623,
+      "step": 1567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1386523723876345,
+      "learning_rate": 6.933518816810661e-05,
+      "loss": 1.5525,
+      "step": 1568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1397326756350934,
+      "learning_rate": 6.93220090789281e-05,
+      "loss": 1.5544,
+      "step": 1569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17255065922931825,
+      "learning_rate": 6.93088299897496e-05,
+      "loss": 1.559,
+      "step": 1570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20704283710173185,
+      "learning_rate": 6.929565090057109e-05,
+      "loss": 1.5569,
+      "step": 1571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19968959715172496,
+      "learning_rate": 6.92824718113926e-05,
+      "loss": 1.5478,
+      "step": 1572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19230184134704087,
+      "learning_rate": 6.92692927222141e-05,
+      "loss": 1.5503,
+      "step": 1573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18909160244964154,
+      "learning_rate": 6.925611363303559e-05,
+      "loss": 1.5544,
+      "step": 1574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16844303612647166,
+      "learning_rate": 6.924293454385708e-05,
+      "loss": 1.5571,
+      "step": 1575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17285965640836326,
+      "learning_rate": 6.922975545467859e-05,
+      "loss": 1.5548,
+      "step": 1576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17097045311532186,
+      "learning_rate": 6.921657636550008e-05,
+      "loss": 1.5581,
+      "step": 1577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15208574230500754,
+      "learning_rate": 6.920339727632158e-05,
+      "loss": 1.559,
+      "step": 1578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1302108613879427,
+      "learning_rate": 6.919021818714307e-05,
+      "loss": 1.5559,
+      "step": 1579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14151167716760873,
+      "learning_rate": 6.917703909796456e-05,
+      "loss": 1.5617,
+      "step": 1580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13644332890371155,
+      "learning_rate": 6.916386000878606e-05,
+      "loss": 1.5599,
+      "step": 1581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12156665440722327,
+      "learning_rate": 6.915068091960756e-05,
+      "loss": 1.5524,
+      "step": 1582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11782989442486515,
+      "learning_rate": 6.913750183042906e-05,
+      "loss": 1.5557,
+      "step": 1583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10458106948141914,
+      "learning_rate": 6.912432274125055e-05,
+      "loss": 1.5525,
+      "step": 1584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11610078372776365,
+      "learning_rate": 6.911114365207206e-05,
+      "loss": 1.556,
+      "step": 1585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11430857791497404,
+      "learning_rate": 6.909796456289354e-05,
+      "loss": 1.5505,
+      "step": 1586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1312092908514261,
+      "learning_rate": 6.908478547371505e-05,
+      "loss": 1.5501,
+      "step": 1587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12575919071416639,
+      "learning_rate": 6.907160638453654e-05,
+      "loss": 1.554,
+      "step": 1588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12565900631995755,
+      "learning_rate": 6.905842729535803e-05,
+      "loss": 1.5528,
+      "step": 1589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1161327185864553,
+      "learning_rate": 6.904524820617953e-05,
+      "loss": 1.5516,
+      "step": 1590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11177341146138027,
+      "learning_rate": 6.903206911700103e-05,
+      "loss": 1.5486,
+      "step": 1591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11905598040671617,
+      "learning_rate": 6.901889002782253e-05,
+      "loss": 1.5568,
+      "step": 1592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.12616538195900687,
+      "learning_rate": 6.900571093864402e-05,
+      "loss": 1.5502,
+      "step": 1593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17498550189708104,
+      "learning_rate": 6.899253184946552e-05,
+      "loss": 1.5654,
+      "step": 1594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2130377460518065,
+      "learning_rate": 6.897935276028701e-05,
+      "loss": 1.5627,
+      "step": 1595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2423218366499862,
+      "learning_rate": 6.896617367110852e-05,
+      "loss": 1.5502,
+      "step": 1596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3065417363260542,
+      "learning_rate": 6.895299458193001e-05,
+      "loss": 1.5481,
+      "step": 1597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3362504924186193,
+      "learning_rate": 6.89398154927515e-05,
+      "loss": 1.5511,
+      "step": 1598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2602709928671128,
+      "learning_rate": 6.8926636403573e-05,
+      "loss": 1.5533,
+      "step": 1599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1706302921715801,
+      "learning_rate": 6.89134573143945e-05,
+      "loss": 1.5543,
+      "step": 1600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18696585801693164,
+      "learning_rate": 6.8900278225216e-05,
+      "loss": 1.5513,
+      "step": 1601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2425990608578436,
+      "learning_rate": 6.888709913603749e-05,
+      "loss": 1.5568,
+      "step": 1602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25557402179698885,
+      "learning_rate": 6.887392004685899e-05,
+      "loss": 1.5583,
+      "step": 1603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19254643621502684,
+      "learning_rate": 6.886074095768048e-05,
+      "loss": 1.5558,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1501256773970859,
+      "learning_rate": 6.884756186850197e-05,
+      "loss": 1.5483,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18027674327204868,
+      "learning_rate": 6.883438277932348e-05,
+      "loss": 1.5482,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18512324904623811,
+      "learning_rate": 6.882120369014498e-05,
+      "loss": 1.5534,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15350827244461232,
+      "learning_rate": 6.880802460096647e-05,
+      "loss": 1.5546,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12508392755322637,
+      "learning_rate": 6.879484551178798e-05,
+      "loss": 1.5482,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14046849983607992,
+      "learning_rate": 6.878166642260946e-05,
+      "loss": 1.5483,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14867576282964143,
+      "learning_rate": 6.876848733343096e-05,
+      "loss": 1.5476,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1450045504632268,
+      "learning_rate": 6.875530824425246e-05,
+      "loss": 1.5608,
+      "step": 1612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12586038081984513,
+      "learning_rate": 6.874212915507395e-05,
+      "loss": 1.5556,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13415831883493776,
+      "learning_rate": 6.872895006589545e-05,
+      "loss": 1.5563,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1582235400657942,
+      "learning_rate": 6.871577097671695e-05,
+      "loss": 1.5543,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16607748843590825,
+      "learning_rate": 6.870259188753845e-05,
+      "loss": 1.5536,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16098113359203,
+      "learning_rate": 6.868941279835994e-05,
+      "loss": 1.5499,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1533723936762229,
+      "learning_rate": 6.867623370918143e-05,
+      "loss": 1.5559,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1419677865844217,
+      "learning_rate": 6.866305462000293e-05,
+      "loss": 1.5563,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12337679315639093,
+      "learning_rate": 6.864987553082442e-05,
+      "loss": 1.5566,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14008735419243357,
+      "learning_rate": 6.863669644164593e-05,
+      "loss": 1.5504,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15354721125578877,
+      "learning_rate": 6.862351735246742e-05,
+      "loss": 1.5548,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13722359030310177,
+      "learning_rate": 6.861033826328892e-05,
+      "loss": 1.5523,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11942493175369336,
+      "learning_rate": 6.859715917411042e-05,
+      "loss": 1.5501,
+      "step": 1624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12604318775555223,
+      "learning_rate": 6.858398008493192e-05,
+      "loss": 1.5498,
+      "step": 1625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13370239925321337,
+      "learning_rate": 6.857080099575341e-05,
+      "loss": 1.559,
+      "step": 1626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12637856740613754,
+      "learning_rate": 6.85576219065749e-05,
+      "loss": 1.5547,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12061462620393058,
+      "learning_rate": 6.85444428173964e-05,
+      "loss": 1.5455,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13680049491771548,
+      "learning_rate": 6.853126372821789e-05,
+      "loss": 1.5471,
+      "step": 1629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14078602589632402,
+      "learning_rate": 6.85180846390394e-05,
+      "loss": 1.5499,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14022146866088356,
+      "learning_rate": 6.850490554986089e-05,
+      "loss": 1.5475,
+      "step": 1631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15892178667718265,
+      "learning_rate": 6.849172646068239e-05,
+      "loss": 1.5415,
+      "step": 1632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15771824336327828,
+      "learning_rate": 6.84785473715039e-05,
+      "loss": 1.5491,
+      "step": 1633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15431553280773133,
+      "learning_rate": 6.846536828232537e-05,
+      "loss": 1.5538,
+      "step": 1634
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.146165745873896,
+      "learning_rate": 6.845218919314688e-05,
+      "loss": 1.5469,
+      "step": 1635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14787594181143815,
+      "learning_rate": 6.843901010396838e-05,
+      "loss": 1.5432,
+      "step": 1636
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15702479363803623,
+      "learning_rate": 6.842583101478987e-05,
+      "loss": 1.5484,
+      "step": 1637
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16689851495876992,
+      "learning_rate": 6.841265192561136e-05,
+      "loss": 1.552,
+      "step": 1638
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1848572151341898,
+      "learning_rate": 6.839947283643287e-05,
+      "loss": 1.5525,
+      "step": 1639
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1985667992824036,
+      "learning_rate": 6.838629374725436e-05,
+      "loss": 1.5527,
+      "step": 1640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.19685054777178346,
+      "learning_rate": 6.837311465807586e-05,
+      "loss": 1.5566,
+      "step": 1641
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17302806445483987,
+      "learning_rate": 6.835993556889735e-05,
+      "loss": 1.5552,
+      "step": 1642
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1624047549497897,
+      "learning_rate": 6.834675647971884e-05,
+      "loss": 1.551,
+      "step": 1643
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1541815338981589,
+      "learning_rate": 6.833357739054034e-05,
+      "loss": 1.5536,
+      "step": 1644
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14690832457803973,
+      "learning_rate": 6.832039830136185e-05,
+      "loss": 1.5556,
+      "step": 1645
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14114554865450257,
+      "learning_rate": 6.830721921218334e-05,
+      "loss": 1.5574,
+      "step": 1646
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1374618278492853,
+      "learning_rate": 6.829404012300483e-05,
+      "loss": 1.5473,
+      "step": 1647
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16334798339082507,
+      "learning_rate": 6.828086103382634e-05,
+      "loss": 1.5436,
+      "step": 1648
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17823021166295652,
+      "learning_rate": 6.826768194464783e-05,
+      "loss": 1.5526,
+      "step": 1649
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16491821517268068,
+      "learning_rate": 6.825450285546933e-05,
+      "loss": 1.5448,
+      "step": 1650
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14340839944429526,
+      "learning_rate": 6.824132376629082e-05,
+      "loss": 1.5555,
+      "step": 1651
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14459695445995016,
+      "learning_rate": 6.822814467711232e-05,
+      "loss": 1.5491,
+      "step": 1652
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14087318597895018,
+      "learning_rate": 6.821496558793381e-05,
+      "loss": 1.5514,
+      "step": 1653
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13530581550555565,
+      "learning_rate": 6.820178649875532e-05,
+      "loss": 1.5493,
+      "step": 1654
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12914300472200602,
+      "learning_rate": 6.818860740957681e-05,
+      "loss": 1.5502,
+      "step": 1655
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14791075998576286,
+      "learning_rate": 6.81754283203983e-05,
+      "loss": 1.5527,
+      "step": 1656
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15767032298140526,
+      "learning_rate": 6.816224923121981e-05,
+      "loss": 1.5507,
+      "step": 1657
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15762124335406885,
+      "learning_rate": 6.814907014204129e-05,
+      "loss": 1.5472,
+      "step": 1658
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15770761252048401,
+      "learning_rate": 6.81358910528628e-05,
+      "loss": 1.5504,
+      "step": 1659
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14539977323851627,
+      "learning_rate": 6.812271196368429e-05,
+      "loss": 1.5416,
+      "step": 1660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1360839511514339,
+      "learning_rate": 6.810953287450579e-05,
+      "loss": 1.5448,
+      "step": 1661
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14633674426067697,
+      "learning_rate": 6.809635378532728e-05,
+      "loss": 1.5532,
+      "step": 1662
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15741046997567085,
+      "learning_rate": 6.808317469614879e-05,
+      "loss": 1.5496,
+      "step": 1663
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15335271399063555,
+      "learning_rate": 6.806999560697028e-05,
+      "loss": 1.5475,
+      "step": 1664
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17257629000740207,
+      "learning_rate": 6.805681651779178e-05,
+      "loss": 1.5563,
+      "step": 1665
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16040574818475473,
+      "learning_rate": 6.804363742861327e-05,
+      "loss": 1.5491,
+      "step": 1666
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12827580813406658,
+      "learning_rate": 6.803045833943476e-05,
+      "loss": 1.5484,
+      "step": 1667
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.137000939385371,
+      "learning_rate": 6.801727925025626e-05,
+      "loss": 1.5539,
+      "step": 1668
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14792870207776196,
+      "learning_rate": 6.800410016107776e-05,
+      "loss": 1.5398,
+      "step": 1669
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13377235683031535,
+      "learning_rate": 6.799092107189926e-05,
+      "loss": 1.5506,
+      "step": 1670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.12524350688472186,
+      "learning_rate": 6.797774198272075e-05,
+      "loss": 1.5477,
+      "step": 1671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1375908850825439,
+      "learning_rate": 6.796456289354226e-05,
+      "loss": 1.5516,
+      "step": 1672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16872797756026237,
+      "learning_rate": 6.795138380436375e-05,
+      "loss": 1.5524,
+      "step": 1673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17255656413666468,
+      "learning_rate": 6.793820471518525e-05,
+      "loss": 1.5519,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15649581350662647,
+      "learning_rate": 6.792502562600674e-05,
+      "loss": 1.545,
+      "step": 1675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14603954181759182,
+      "learning_rate": 6.791184653682823e-05,
+      "loss": 1.5354,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14996652222064816,
+      "learning_rate": 6.789866744764973e-05,
+      "loss": 1.5462,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14098612021083787,
+      "learning_rate": 6.788548835847123e-05,
+      "loss": 1.5463,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13786176118009635,
+      "learning_rate": 6.787230926929273e-05,
+      "loss": 1.55,
+      "step": 1679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15617176658263263,
+      "learning_rate": 6.785913018011422e-05,
+      "loss": 1.5469,
+      "step": 1680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1879698889780054,
+      "learning_rate": 6.784595109093573e-05,
+      "loss": 1.5485,
+      "step": 1681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22122586221883833,
+      "learning_rate": 6.783277200175721e-05,
+      "loss": 1.5461,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22292103720296774,
+      "learning_rate": 6.781959291257872e-05,
+      "loss": 1.5543,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1736273162608752,
+      "learning_rate": 6.780641382340021e-05,
+      "loss": 1.5501,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11509864784919802,
+      "learning_rate": 6.77932347342217e-05,
+      "loss": 1.5532,
+      "step": 1685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12708499410489232,
+      "learning_rate": 6.77800556450432e-05,
+      "loss": 1.5358,
+      "step": 1686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17299827793930234,
+      "learning_rate": 6.77668765558647e-05,
+      "loss": 1.5481,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1822316133159184,
+      "learning_rate": 6.77536974666862e-05,
+      "loss": 1.5497,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17414788070503254,
+      "learning_rate": 6.774051837750769e-05,
+      "loss": 1.5487,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15984217459846126,
+      "learning_rate": 6.772733928832919e-05,
+      "loss": 1.5434,
+      "step": 1690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1542107562856,
+      "learning_rate": 6.771416019915068e-05,
+      "loss": 1.5508,
+      "step": 1691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16229994239377557,
+      "learning_rate": 6.770098110997217e-05,
+      "loss": 1.5489,
+      "step": 1692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1546526960370925,
+      "learning_rate": 6.768780202079368e-05,
+      "loss": 1.5499,
+      "step": 1693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12279914387612359,
+      "learning_rate": 6.767462293161517e-05,
+      "loss": 1.5427,
+      "step": 1694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12975757335563615,
+      "learning_rate": 6.766144384243667e-05,
+      "loss": 1.5467,
+      "step": 1695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13422122080467072,
+      "learning_rate": 6.764826475325818e-05,
+      "loss": 1.5554,
+      "step": 1696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14192591482733333,
+      "learning_rate": 6.763508566407967e-05,
+      "loss": 1.5408,
+      "step": 1697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13893637523732244,
+      "learning_rate": 6.762190657490116e-05,
+      "loss": 1.5495,
+      "step": 1698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15491412035183788,
+      "learning_rate": 6.760872748572266e-05,
+      "loss": 1.5501,
+      "step": 1699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17883988461434905,
+      "learning_rate": 6.759554839654415e-05,
+      "loss": 1.5503,
+      "step": 1700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20265371705570628,
+      "learning_rate": 6.758236930736564e-05,
+      "loss": 1.5361,
+      "step": 1701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2139359712926301,
+      "learning_rate": 6.756919021818715e-05,
+      "loss": 1.5381,
+      "step": 1702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18614804141636146,
+      "learning_rate": 6.755601112900865e-05,
+      "loss": 1.5482,
+      "step": 1703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15804041466228397,
+      "learning_rate": 6.754283203983014e-05,
+      "loss": 1.5361,
+      "step": 1704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15062690335608622,
+      "learning_rate": 6.752965295065165e-05,
+      "loss": 1.5455,
+      "step": 1705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12443563823484106,
+      "learning_rate": 6.751647386147313e-05,
+      "loss": 1.5435,
+      "step": 1706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14089290268248464,
+      "learning_rate": 6.750329477229462e-05,
+      "loss": 1.5403,
+      "step": 1707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1550833858910459,
+      "learning_rate": 6.749011568311613e-05,
+      "loss": 1.5505,
+      "step": 1708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.149359947538185,
+      "learning_rate": 6.747693659393762e-05,
+      "loss": 1.5469,
+      "step": 1709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15937266282828347,
+      "learning_rate": 6.746375750475912e-05,
+      "loss": 1.5419,
+      "step": 1710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1631750053650106,
+      "learning_rate": 6.745057841558062e-05,
+      "loss": 1.5399,
+      "step": 1711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14972740561419975,
+      "learning_rate": 6.743739932640212e-05,
+      "loss": 1.5528,
+      "step": 1712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1587005414140819,
+      "learning_rate": 6.742422023722361e-05,
+      "loss": 1.5471,
+      "step": 1713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16023087202413147,
+      "learning_rate": 6.74110411480451e-05,
+      "loss": 1.55,
+      "step": 1714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1507360827043738,
+      "learning_rate": 6.73978620588666e-05,
+      "loss": 1.5436,
+      "step": 1715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1447030944327568,
+      "learning_rate": 6.738468296968809e-05,
+      "loss": 1.5498,
+      "step": 1716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12385411311009659,
+      "learning_rate": 6.73715038805096e-05,
+      "loss": 1.5518,
+      "step": 1717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1164550428705913,
+      "learning_rate": 6.735832479133109e-05,
+      "loss": 1.5425,
+      "step": 1718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13629614084783107,
+      "learning_rate": 6.734514570215259e-05,
+      "loss": 1.5511,
+      "step": 1719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1506489926079727,
+      "learning_rate": 6.73319666129741e-05,
+      "loss": 1.5419,
+      "step": 1720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13855334627153476,
+      "learning_rate": 6.731878752379559e-05,
+      "loss": 1.5484,
+      "step": 1721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12832134059259032,
+      "learning_rate": 6.730560843461708e-05,
+      "loss": 1.5398,
+      "step": 1722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12090987219413707,
+      "learning_rate": 6.729242934543857e-05,
+      "loss": 1.5402,
+      "step": 1723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11453651059207949,
+      "learning_rate": 6.727925025626007e-05,
+      "loss": 1.5506,
+      "step": 1724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1196098497984597,
+      "learning_rate": 6.726607116708156e-05,
+      "loss": 1.5451,
+      "step": 1725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11367821126555626,
+      "learning_rate": 6.725289207790307e-05,
+      "loss": 1.5503,
+      "step": 1726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10401603162688644,
+      "learning_rate": 6.723971298872456e-05,
+      "loss": 1.5495,
+      "step": 1727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.10956859510139251,
+      "learning_rate": 6.722653389954606e-05,
+      "loss": 1.5462,
+      "step": 1728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.11932498061521109,
+      "learning_rate": 6.721335481036756e-05,
+      "loss": 1.5457,
+      "step": 1729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12531582645142714,
+      "learning_rate": 6.720017572118904e-05,
+      "loss": 1.5492,
+      "step": 1730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.12833505710933477,
+      "learning_rate": 6.718699663201054e-05,
+      "loss": 1.5456,
+      "step": 1731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14072231726948428,
+      "learning_rate": 6.717381754283205e-05,
+      "loss": 1.5498,
+      "step": 1732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16831541592684482,
+      "learning_rate": 6.716063845365354e-05,
+      "loss": 1.5558,
+      "step": 1733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2093135547333804,
+      "learning_rate": 6.714745936447503e-05,
+      "loss": 1.5606,
+      "step": 1734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22413912594550753,
+      "learning_rate": 6.713428027529654e-05,
+      "loss": 1.5515,
+      "step": 1735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21448251347724995,
+      "learning_rate": 6.712110118611803e-05,
+      "loss": 1.5362,
+      "step": 1736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.20678506449185255,
+      "learning_rate": 6.710792209693953e-05,
+      "loss": 1.5395,
+      "step": 1737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19348081818742874,
+      "learning_rate": 6.709474300776102e-05,
+      "loss": 1.5459,
+      "step": 1738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.19216843244188353,
+      "learning_rate": 6.708156391858252e-05,
+      "loss": 1.5361,
+      "step": 1739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1950956680007854,
+      "learning_rate": 6.706838482940401e-05,
+      "loss": 1.5451,
+      "step": 1740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18065563157921122,
+      "learning_rate": 6.705520574022552e-05,
+      "loss": 1.535,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1724051176877015,
+      "learning_rate": 6.704202665104701e-05,
+      "loss": 1.543,
+      "step": 1742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16492687488657337,
+      "learning_rate": 6.70288475618685e-05,
+      "loss": 1.5386,
+      "step": 1743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1713929297928792,
+      "learning_rate": 6.701566847269001e-05,
+      "loss": 1.541,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17209785450990964,
+      "learning_rate": 6.70024893835115e-05,
+      "loss": 1.5393,
+      "step": 1745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12260311659465224,
+      "learning_rate": 6.6989310294333e-05,
+      "loss": 1.542,
+      "step": 1746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.11796151518037762,
+      "learning_rate": 6.697613120515449e-05,
+      "loss": 1.546,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14676550674762198,
+      "learning_rate": 6.696295211597599e-05,
+      "loss": 1.5407,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14285664313926463,
+      "learning_rate": 6.694977302679748e-05,
+      "loss": 1.5432,
+      "step": 1749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15040983251816017,
+      "learning_rate": 6.693659393761899e-05,
+      "loss": 1.5471,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14155615679807018,
+      "learning_rate": 6.692341484844048e-05,
+      "loss": 1.5444,
+      "step": 1751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12521018590917282,
+      "learning_rate": 6.691023575926197e-05,
+      "loss": 1.5415,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1483724473478806,
+      "learning_rate": 6.689705667008348e-05,
+      "loss": 1.5462,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16302152525308417,
+      "learning_rate": 6.688387758090496e-05,
+      "loss": 1.5435,
+      "step": 1754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1593150408912617,
+      "learning_rate": 6.687069849172646e-05,
+      "loss": 1.5455,
+      "step": 1755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15798766607610612,
+      "learning_rate": 6.685751940254796e-05,
+      "loss": 1.5424,
+      "step": 1756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15509643681103627,
+      "learning_rate": 6.684434031336946e-05,
+      "loss": 1.5474,
+      "step": 1757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15042707266000108,
+      "learning_rate": 6.683116122419095e-05,
+      "loss": 1.5386,
+      "step": 1758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12529483918638012,
+      "learning_rate": 6.681798213501246e-05,
+      "loss": 1.5483,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16020600612445207,
+      "learning_rate": 6.680480304583395e-05,
+      "loss": 1.545,
+      "step": 1760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17365135713154362,
+      "learning_rate": 6.679162395665545e-05,
+      "loss": 1.5399,
+      "step": 1761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15617117207014766,
+      "learning_rate": 6.677844486747694e-05,
+      "loss": 1.5417,
+      "step": 1762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1826268055689078,
+      "learning_rate": 6.676526577829843e-05,
+      "loss": 1.5356,
+      "step": 1763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1792955657441501,
+      "learning_rate": 6.675208668911993e-05,
+      "loss": 1.5409,
+      "step": 1764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1549319116339377,
+      "learning_rate": 6.673890759994143e-05,
+      "loss": 1.5425,
+      "step": 1765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16797264760156416,
+      "learning_rate": 6.672572851076293e-05,
+      "loss": 1.5472,
+      "step": 1766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1689579541348412,
+      "learning_rate": 6.671254942158442e-05,
+      "loss": 1.5333,
+      "step": 1767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1443581888928475,
+      "learning_rate": 6.669937033240593e-05,
+      "loss": 1.549,
+      "step": 1768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12295030770476027,
+      "learning_rate": 6.668619124322742e-05,
+      "loss": 1.5471,
+      "step": 1769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14316536012437356,
+      "learning_rate": 6.66730121540489e-05,
+      "loss": 1.5388,
+      "step": 1770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16946707074936326,
+      "learning_rate": 6.665983306487041e-05,
+      "loss": 1.5449,
+      "step": 1771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19183626388609665,
+      "learning_rate": 6.66466539756919e-05,
+      "loss": 1.5525,
+      "step": 1772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17394581389516967,
+      "learning_rate": 6.66334748865134e-05,
+      "loss": 1.5455,
+      "step": 1773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13796787925409335,
+      "learning_rate": 6.66202957973349e-05,
+      "loss": 1.5453,
+      "step": 1774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14737682661347665,
+      "learning_rate": 6.66071167081564e-05,
+      "loss": 1.5383,
+      "step": 1775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1622449698758281,
+      "learning_rate": 6.659393761897789e-05,
+      "loss": 1.5412,
+      "step": 1776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16064176456627283,
+      "learning_rate": 6.65807585297994e-05,
+      "loss": 1.5452,
+      "step": 1777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15131456902589685,
+      "learning_rate": 6.656757944062088e-05,
+      "loss": 1.5393,
+      "step": 1778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1529436529522335,
+      "learning_rate": 6.655440035144237e-05,
+      "loss": 1.5382,
+      "step": 1779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15466089796538804,
+      "learning_rate": 6.654122126226388e-05,
+      "loss": 1.5321,
+      "step": 1780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13648831163966982,
+      "learning_rate": 6.652804217308537e-05,
+      "loss": 1.5435,
+      "step": 1781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13571229687876288,
+      "learning_rate": 6.651486308390687e-05,
+      "loss": 1.5326,
+      "step": 1782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14664723920130965,
+      "learning_rate": 6.650168399472838e-05,
+      "loss": 1.541,
+      "step": 1783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1322498917434933,
+      "learning_rate": 6.648850490554987e-05,
+      "loss": 1.5401,
+      "step": 1784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1157193538823997,
+      "learning_rate": 6.647532581637136e-05,
+      "loss": 1.5438,
+      "step": 1785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12401343440978063,
+      "learning_rate": 6.646214672719286e-05,
+      "loss": 1.537,
+      "step": 1786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13123152593043114,
+      "learning_rate": 6.644896763801435e-05,
+      "loss": 1.5407,
+      "step": 1787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1471707795413792,
+      "learning_rate": 6.643578854883584e-05,
+      "loss": 1.534,
+      "step": 1788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14848671197739402,
+      "learning_rate": 6.642260945965735e-05,
+      "loss": 1.5409,
+      "step": 1789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14341236634610707,
+      "learning_rate": 6.640943037047885e-05,
+      "loss": 1.5421,
+      "step": 1790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1326356379910537,
+      "learning_rate": 6.639625128130034e-05,
+      "loss": 1.5427,
+      "step": 1791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1223557409627182,
+      "learning_rate": 6.638307219212185e-05,
+      "loss": 1.5473,
+      "step": 1792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13155912571100425,
+      "learning_rate": 6.636989310294334e-05,
+      "loss": 1.5307,
+      "step": 1793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15378406626616598,
+      "learning_rate": 6.635671401376482e-05,
+      "loss": 1.5412,
+      "step": 1794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1479907048479965,
+      "learning_rate": 6.634353492458633e-05,
+      "loss": 1.5376,
+      "step": 1795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13370072014639775,
+      "learning_rate": 6.633035583540782e-05,
+      "loss": 1.5416,
+      "step": 1796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.12079769907305982,
+      "learning_rate": 6.631717674622931e-05,
+      "loss": 1.5436,
+      "step": 1797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13715067412054485,
+      "learning_rate": 6.630399765705082e-05,
+      "loss": 1.5505,
+      "step": 1798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13862894618039118,
+      "learning_rate": 6.629081856787232e-05,
+      "loss": 1.5425,
+      "step": 1799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14258297871949263,
+      "learning_rate": 6.627763947869381e-05,
+      "loss": 1.5469,
+      "step": 1800
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.5050686597824097,
+      "eval_runtime": 446.284,
+      "eval_samples_per_second": 40.367,
+      "eval_steps_per_second": 0.06,
+      "step": 1800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14582146377787855,
+      "learning_rate": 6.626446038951532e-05,
+      "loss": 1.5345,
+      "step": 1801
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14561113226125544,
+      "learning_rate": 6.62512813003368e-05,
+      "loss": 1.5328,
+      "step": 1802
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14448104423230185,
+      "learning_rate": 6.623810221115829e-05,
+      "loss": 1.541,
+      "step": 1803
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1466153044412786,
+      "learning_rate": 6.62249231219798e-05,
+      "loss": 1.5399,
+      "step": 1804
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1437769469124596,
+      "learning_rate": 6.621174403280129e-05,
+      "loss": 1.5415,
+      "step": 1805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1305816646967658,
+      "learning_rate": 6.619856494362279e-05,
+      "loss": 1.5339,
+      "step": 1806
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15306058911770729,
+      "learning_rate": 6.618538585444429e-05,
+      "loss": 1.5322,
+      "step": 1807
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19189215993239692,
+      "learning_rate": 6.617220676526579e-05,
+      "loss": 1.5375,
+      "step": 1808
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.21994635331924403,
+      "learning_rate": 6.615902767608728e-05,
+      "loss": 1.5537,
+      "step": 1809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23545604821603358,
+      "learning_rate": 6.614584858690877e-05,
+      "loss": 1.5448,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.22439299112952654,
+      "learning_rate": 6.613266949773027e-05,
+      "loss": 1.5355,
+      "step": 1811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.19877399470688997,
+      "learning_rate": 6.611949040855176e-05,
+      "loss": 1.5326,
+      "step": 1812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18812797370997456,
+      "learning_rate": 6.610631131937327e-05,
+      "loss": 1.5428,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15978267021428885,
+      "learning_rate": 6.609313223019476e-05,
+      "loss": 1.5326,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13271653534984373,
+      "learning_rate": 6.607995314101626e-05,
+      "loss": 1.5332,
+      "step": 1815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15377296678703303,
+      "learning_rate": 6.606677405183776e-05,
+      "loss": 1.54,
+      "step": 1816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1722532469808815,
+      "learning_rate": 6.605359496265926e-05,
+      "loss": 1.5402,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15469015104803427,
+      "learning_rate": 6.604041587348074e-05,
+      "loss": 1.5391,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14036848908680152,
+      "learning_rate": 6.602723678430225e-05,
+      "loss": 1.5403,
+      "step": 1819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14393452820901584,
+      "learning_rate": 6.601405769512374e-05,
+      "loss": 1.5451,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15707275687607905,
+      "learning_rate": 6.600087860594523e-05,
+      "loss": 1.5424,
+      "step": 1821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13162878232984085,
+      "learning_rate": 6.598769951676674e-05,
+      "loss": 1.5393,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12608100407092024,
+      "learning_rate": 6.597452042758823e-05,
+      "loss": 1.5349,
+      "step": 1823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13233817296489414,
+      "learning_rate": 6.596134133840973e-05,
+      "loss": 1.5347,
+      "step": 1824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13050126657634636,
+      "learning_rate": 6.594816224923123e-05,
+      "loss": 1.5352,
+      "step": 1825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1358397783457529,
+      "learning_rate": 6.593498316005271e-05,
+      "loss": 1.5385,
+      "step": 1826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15120960081930016,
+      "learning_rate": 6.592180407087421e-05,
+      "loss": 1.5563,
+      "step": 1827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15098843641155918,
+      "learning_rate": 6.590862498169572e-05,
+      "loss": 1.54,
+      "step": 1828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13425392810137007,
+      "learning_rate": 6.589544589251721e-05,
+      "loss": 1.5403,
+      "step": 1829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10649385371019933,
+      "learning_rate": 6.58822668033387e-05,
+      "loss": 1.5358,
+      "step": 1830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10730036711120357,
+      "learning_rate": 6.586908771416021e-05,
+      "loss": 1.542,
+      "step": 1831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.11956934945327176,
+      "learning_rate": 6.58559086249817e-05,
+      "loss": 1.5444,
+      "step": 1832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12387275009941737,
+      "learning_rate": 6.58427295358032e-05,
+      "loss": 1.5421,
+      "step": 1833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13502573869963072,
+      "learning_rate": 6.582955044662469e-05,
+      "loss": 1.5344,
+      "step": 1834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14932387703824,
+      "learning_rate": 6.581637135744619e-05,
+      "loss": 1.5391,
+      "step": 1835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15467034804752572,
+      "learning_rate": 6.580319226826768e-05,
+      "loss": 1.5422,
+      "step": 1836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14605819492150904,
+      "learning_rate": 6.579001317908919e-05,
+      "loss": 1.5492,
+      "step": 1837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14117361603242334,
+      "learning_rate": 6.577683408991068e-05,
+      "loss": 1.5444,
+      "step": 1838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15595524076571482,
+      "learning_rate": 6.576365500073217e-05,
+      "loss": 1.536,
+      "step": 1839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16609341046243062,
+      "learning_rate": 6.575047591155368e-05,
+      "loss": 1.5402,
+      "step": 1840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1531555863217873,
+      "learning_rate": 6.573729682237518e-05,
+      "loss": 1.5397,
+      "step": 1841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1262455511117069,
+      "learning_rate": 6.572411773319666e-05,
+      "loss": 1.5365,
+      "step": 1842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12736945248259027,
+      "learning_rate": 6.571093864401816e-05,
+      "loss": 1.5365,
+      "step": 1843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12649547404752876,
+      "learning_rate": 6.569775955483966e-05,
+      "loss": 1.5304,
+      "step": 1844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1301992882193159,
+      "learning_rate": 6.568458046566115e-05,
+      "loss": 1.533,
+      "step": 1845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13173740270605602,
+      "learning_rate": 6.567140137648266e-05,
+      "loss": 1.5453,
+      "step": 1846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1509395261337269,
+      "learning_rate": 6.565822228730415e-05,
+      "loss": 1.5413,
+      "step": 1847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1737928398810218,
+      "learning_rate": 6.564504319812564e-05,
+      "loss": 1.5385,
+      "step": 1848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1825036525936605,
+      "learning_rate": 6.563186410894715e-05,
+      "loss": 1.5343,
+      "step": 1849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14914145249944416,
+      "learning_rate": 6.561868501976863e-05,
+      "loss": 1.5381,
+      "step": 1850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1411401274879375,
+      "learning_rate": 6.560550593059013e-05,
+      "loss": 1.5377,
+      "step": 1851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13190444561775116,
+      "learning_rate": 6.559232684141163e-05,
+      "loss": 1.537,
+      "step": 1852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14727826338528666,
+      "learning_rate": 6.557914775223313e-05,
+      "loss": 1.5342,
+      "step": 1853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18032220911577906,
+      "learning_rate": 6.556596866305462e-05,
+      "loss": 1.5309,
+      "step": 1854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21997773676839405,
+      "learning_rate": 6.555278957387613e-05,
+      "loss": 1.5339,
+      "step": 1855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23939108351438856,
+      "learning_rate": 6.553961048469762e-05,
+      "loss": 1.5335,
+      "step": 1856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21647697752502287,
+      "learning_rate": 6.552643139551912e-05,
+      "loss": 1.5345,
+      "step": 1857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.19365149697676406,
+      "learning_rate": 6.551325230634061e-05,
+      "loss": 1.532,
+      "step": 1858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1923078109676377,
+      "learning_rate": 6.55000732171621e-05,
+      "loss": 1.5311,
+      "step": 1859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18571127464856396,
+      "learning_rate": 6.54868941279836e-05,
+      "loss": 1.5364,
+      "step": 1860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15629991416317407,
+      "learning_rate": 6.54737150388051e-05,
+      "loss": 1.5373,
+      "step": 1861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.12966182497389114,
+      "learning_rate": 6.54605359496266e-05,
+      "loss": 1.5389,
+      "step": 1862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14476534994445542,
+      "learning_rate": 6.544735686044809e-05,
+      "loss": 1.541,
+      "step": 1863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15133948999974933,
+      "learning_rate": 6.54341777712696e-05,
+      "loss": 1.5393,
+      "step": 1864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.11483646251386517,
+      "learning_rate": 6.542099868209108e-05,
+      "loss": 1.5417,
+      "step": 1865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10598444985495092,
+      "learning_rate": 6.540781959291257e-05,
+      "loss": 1.5396,
+      "step": 1866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1474562074661653,
+      "learning_rate": 6.539464050373408e-05,
+      "loss": 1.5364,
+      "step": 1867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16648792502735327,
+      "learning_rate": 6.538146141455557e-05,
+      "loss": 1.5371,
+      "step": 1868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15052276885283872,
+      "learning_rate": 6.536828232537707e-05,
+      "loss": 1.5354,
+      "step": 1869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13771541946283541,
+      "learning_rate": 6.535510323619857e-05,
+      "loss": 1.5289,
+      "step": 1870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14778963249344762,
+      "learning_rate": 6.534192414702007e-05,
+      "loss": 1.5351,
+      "step": 1871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16593630176835306,
+      "learning_rate": 6.532874505784156e-05,
+      "loss": 1.5333,
+      "step": 1872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13651141730237293,
+      "learning_rate": 6.531556596866306e-05,
+      "loss": 1.5325,
+      "step": 1873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13072234951434641,
+      "learning_rate": 6.530238687948455e-05,
+      "loss": 1.5326,
+      "step": 1874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15718747632093522,
+      "learning_rate": 6.528920779030604e-05,
+      "loss": 1.5311,
+      "step": 1875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14102010808330911,
+      "learning_rate": 6.527602870112755e-05,
+      "loss": 1.5349,
+      "step": 1876
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10908492610192105,
+      "learning_rate": 6.526284961194904e-05,
+      "loss": 1.5355,
+      "step": 1877
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13730327480922572,
+      "learning_rate": 6.524967052277054e-05,
+      "loss": 1.5321,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13458348007637888,
+      "learning_rate": 6.523649143359205e-05,
+      "loss": 1.5276,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12531968121607637,
+      "learning_rate": 6.522331234441354e-05,
+      "loss": 1.5402,
+      "step": 1880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15233508794526657,
+      "learning_rate": 6.521013325523502e-05,
+      "loss": 1.5369,
+      "step": 1881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15477101589918488,
+      "learning_rate": 6.519695416605653e-05,
+      "loss": 1.5319,
+      "step": 1882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1542098513682894,
+      "learning_rate": 6.518377507687802e-05,
+      "loss": 1.5459,
+      "step": 1883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15111055672367912,
+      "learning_rate": 6.517059598769951e-05,
+      "loss": 1.5288,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14608804446291268,
+      "learning_rate": 6.515741689852102e-05,
+      "loss": 1.5338,
+      "step": 1885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1320870232295378,
+      "learning_rate": 6.514423780934252e-05,
+      "loss": 1.5387,
+      "step": 1886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15131762156658068,
+      "learning_rate": 6.513105872016401e-05,
+      "loss": 1.5345,
+      "step": 1887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1450449269213637,
+      "learning_rate": 6.511787963098552e-05,
+      "loss": 1.5365,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14651491786758905,
+      "learning_rate": 6.5104700541807e-05,
+      "loss": 1.5346,
+      "step": 1889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13345714227804523,
+      "learning_rate": 6.509152145262849e-05,
+      "loss": 1.527,
+      "step": 1890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1335077440167749,
+      "learning_rate": 6.507834236345e-05,
+      "loss": 1.5376,
+      "step": 1891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14817952051451277,
+      "learning_rate": 6.506516327427149e-05,
+      "loss": 1.5358,
+      "step": 1892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1750873210594401,
+      "learning_rate": 6.505198418509299e-05,
+      "loss": 1.5336,
+      "step": 1893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1492711758300829,
+      "learning_rate": 6.503880509591449e-05,
+      "loss": 1.5353,
+      "step": 1894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15056312201224253,
+      "learning_rate": 6.502562600673599e-05,
+      "loss": 1.5322,
+      "step": 1895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17551303219840247,
+      "learning_rate": 6.501244691755748e-05,
+      "loss": 1.5372,
+      "step": 1896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15525943003519552,
+      "learning_rate": 6.499926782837897e-05,
+      "loss": 1.5318,
+      "step": 1897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1310968681897291,
+      "learning_rate": 6.498608873920047e-05,
+      "loss": 1.525,
+      "step": 1898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14144751702735095,
+      "learning_rate": 6.497290965002196e-05,
+      "loss": 1.5295,
+      "step": 1899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13401032728730045,
+      "learning_rate": 6.495973056084347e-05,
+      "loss": 1.5304,
+      "step": 1900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1390255390245617,
+      "learning_rate": 6.494655147166496e-05,
+      "loss": 1.535,
+      "step": 1901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14276204931284214,
+      "learning_rate": 6.493337238248646e-05,
+      "loss": 1.5354,
+      "step": 1902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12009099730090483,
+      "learning_rate": 6.492019329330796e-05,
+      "loss": 1.5382,
+      "step": 1903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1146216007732983,
+      "learning_rate": 6.490701420412946e-05,
+      "loss": 1.5355,
+      "step": 1904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1415244292925028,
+      "learning_rate": 6.489383511495094e-05,
+      "loss": 1.5347,
+      "step": 1905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.141703112248383,
+      "learning_rate": 6.488065602577244e-05,
+      "loss": 1.529,
+      "step": 1906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14613135640837022,
+      "learning_rate": 6.486747693659394e-05,
+      "loss": 1.5297,
+      "step": 1907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1615341703344672,
+      "learning_rate": 6.485429784741543e-05,
+      "loss": 1.5295,
+      "step": 1908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15831620142565728,
+      "learning_rate": 6.484111875823694e-05,
+      "loss": 1.543,
+      "step": 1909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1689989000098508,
+      "learning_rate": 6.482793966905843e-05,
+      "loss": 1.533,
+      "step": 1910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1813491020601592,
+      "learning_rate": 6.481476057987993e-05,
+      "loss": 1.5284,
+      "step": 1911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15988943299074287,
+      "learning_rate": 6.480158149070143e-05,
+      "loss": 1.5413,
+      "step": 1912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1509888625443986,
+      "learning_rate": 6.478840240152291e-05,
+      "loss": 1.5372,
+      "step": 1913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13757177924900643,
+      "learning_rate": 6.477522331234441e-05,
+      "loss": 1.5319,
+      "step": 1914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.11609129535195285,
+      "learning_rate": 6.476204422316592e-05,
+      "loss": 1.5294,
+      "step": 1915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12476376298090223,
+      "learning_rate": 6.474886513398741e-05,
+      "loss": 1.5324,
+      "step": 1916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14163803747323508,
+      "learning_rate": 6.47356860448089e-05,
+      "loss": 1.5429,
+      "step": 1917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1292210396570364,
+      "learning_rate": 6.472250695563041e-05,
+      "loss": 1.5249,
+      "step": 1918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13565267718513704,
+      "learning_rate": 6.47093278664519e-05,
+      "loss": 1.5295,
+      "step": 1919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12368825930145683,
+      "learning_rate": 6.46961487772734e-05,
+      "loss": 1.5295,
+      "step": 1920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1243358548029078,
+      "learning_rate": 6.468296968809489e-05,
+      "loss": 1.5356,
+      "step": 1921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14279450935765994,
+      "learning_rate": 6.466979059891639e-05,
+      "loss": 1.5273,
+      "step": 1922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15281257618762362,
+      "learning_rate": 6.465661150973788e-05,
+      "loss": 1.5278,
+      "step": 1923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16454769000361816,
+      "learning_rate": 6.464343242055939e-05,
+      "loss": 1.528,
+      "step": 1924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16218346940497222,
+      "learning_rate": 6.463025333138088e-05,
+      "loss": 1.536,
+      "step": 1925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14910211469714882,
+      "learning_rate": 6.461707424220237e-05,
+      "loss": 1.5383,
+      "step": 1926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12283017088886376,
+      "learning_rate": 6.460389515302388e-05,
+      "loss": 1.5244,
+      "step": 1927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1121180101341425,
+      "learning_rate": 6.459071606384537e-05,
+      "loss": 1.5322,
+      "step": 1928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12047061707002905,
+      "learning_rate": 6.457753697466685e-05,
+      "loss": 1.5314,
+      "step": 1929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1422335599916664,
+      "learning_rate": 6.456435788548836e-05,
+      "loss": 1.531,
+      "step": 1930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13298205790569448,
+      "learning_rate": 6.455117879630986e-05,
+      "loss": 1.5358,
+      "step": 1931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12317250095530176,
+      "learning_rate": 6.453799970713135e-05,
+      "loss": 1.5388,
+      "step": 1932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12977704215481517,
+      "learning_rate": 6.452482061795286e-05,
+      "loss": 1.5406,
+      "step": 1933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1434596835970703,
+      "learning_rate": 6.451164152877435e-05,
+      "loss": 1.5303,
+      "step": 1934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14954350641793915,
+      "learning_rate": 6.449846243959584e-05,
+      "loss": 1.5341,
+      "step": 1935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12821086681598565,
+      "learning_rate": 6.448528335041735e-05,
+      "loss": 1.5335,
+      "step": 1936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.12683351157331768,
+      "learning_rate": 6.447210426123883e-05,
+      "loss": 1.5393,
+      "step": 1937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13338056148565824,
+      "learning_rate": 6.445892517206033e-05,
+      "loss": 1.5331,
+      "step": 1938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16736424698526148,
+      "learning_rate": 6.444574608288183e-05,
+      "loss": 1.5258,
+      "step": 1939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.20601644305563926,
+      "learning_rate": 6.443256699370333e-05,
+      "loss": 1.5363,
+      "step": 1940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1933094251208626,
+      "learning_rate": 6.441938790452482e-05,
+      "loss": 1.5311,
+      "step": 1941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13776507684940908,
+      "learning_rate": 6.440620881534633e-05,
+      "loss": 1.5249,
+      "step": 1942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1114087025609903,
+      "learning_rate": 6.439302972616782e-05,
+      "loss": 1.5341,
+      "step": 1943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15488304429854696,
+      "learning_rate": 6.437985063698932e-05,
+      "loss": 1.5279,
+      "step": 1944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17311198474671785,
+      "learning_rate": 6.436667154781081e-05,
+      "loss": 1.5366,
+      "step": 1945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14995971280568912,
+      "learning_rate": 6.43534924586323e-05,
+      "loss": 1.5357,
+      "step": 1946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15508882239348012,
+      "learning_rate": 6.43403133694538e-05,
+      "loss": 1.526,
+      "step": 1947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1658475598721562,
+      "learning_rate": 6.43271342802753e-05,
+      "loss": 1.5313,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17845149328758675,
+      "learning_rate": 6.43139551910968e-05,
+      "loss": 1.5208,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.19482180768743815,
+      "learning_rate": 6.430077610191829e-05,
+      "loss": 1.5409,
+      "step": 1950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18646501396825818,
+      "learning_rate": 6.42875970127398e-05,
+      "loss": 1.528,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16710692161764742,
+      "learning_rate": 6.427441792356129e-05,
+      "loss": 1.5301,
+      "step": 1952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13997945381714386,
+      "learning_rate": 6.426123883438277e-05,
+      "loss": 1.538,
+      "step": 1953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14141775562262965,
+      "learning_rate": 6.424805974520428e-05,
+      "loss": 1.5296,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14290987763175558,
+      "learning_rate": 6.423488065602577e-05,
+      "loss": 1.5298,
+      "step": 1955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14435471045549242,
+      "learning_rate": 6.422170156684727e-05,
+      "loss": 1.5304,
+      "step": 1956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17021696930578623,
+      "learning_rate": 6.420852247766877e-05,
+      "loss": 1.5329,
+      "step": 1957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1722560099387899,
+      "learning_rate": 6.419534338849027e-05,
+      "loss": 1.5348,
+      "step": 1958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13401327015248482,
+      "learning_rate": 6.418216429931176e-05,
+      "loss": 1.5346,
+      "step": 1959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10978970597654364,
+      "learning_rate": 6.416898521013327e-05,
+      "loss": 1.5258,
+      "step": 1960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14928376860258072,
+      "learning_rate": 6.415580612095475e-05,
+      "loss": 1.5343,
+      "step": 1961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17944353179995445,
+      "learning_rate": 6.414262703177624e-05,
+      "loss": 1.5298,
+      "step": 1962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15416066242017137,
+      "learning_rate": 6.412944794259775e-05,
+      "loss": 1.5365,
+      "step": 1963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14815642063077472,
+      "learning_rate": 6.411626885341924e-05,
+      "loss": 1.5225,
+      "step": 1964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14630157311321682,
+      "learning_rate": 6.410308976424074e-05,
+      "loss": 1.5269,
+      "step": 1965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14813934644937285,
+      "learning_rate": 6.408991067506225e-05,
+      "loss": 1.5256,
+      "step": 1966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15068049365535827,
+      "learning_rate": 6.407673158588374e-05,
+      "loss": 1.5282,
+      "step": 1967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13628674203757674,
+      "learning_rate": 6.406355249670523e-05,
+      "loss": 1.5271,
+      "step": 1968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14597609932664182,
+      "learning_rate": 6.405037340752673e-05,
+      "loss": 1.5278,
+      "step": 1969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16404348755929074,
+      "learning_rate": 6.403719431834822e-05,
+      "loss": 1.5414,
+      "step": 1970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14298124778173457,
+      "learning_rate": 6.402401522916971e-05,
+      "loss": 1.5349,
+      "step": 1971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11584516029492996,
+      "learning_rate": 6.401083613999122e-05,
+      "loss": 1.5385,
+      "step": 1972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12721590288881943,
+      "learning_rate": 6.399765705081271e-05,
+      "loss": 1.5259,
+      "step": 1973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1299696878955865,
+      "learning_rate": 6.398447796163421e-05,
+      "loss": 1.532,
+      "step": 1974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13096131559896187,
+      "learning_rate": 6.397129887245572e-05,
+      "loss": 1.5285,
+      "step": 1975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11710021873513451,
+      "learning_rate": 6.395811978327721e-05,
+      "loss": 1.5279,
+      "step": 1976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12442076916158991,
+      "learning_rate": 6.394494069409869e-05,
+      "loss": 1.5294,
+      "step": 1977
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1024723308527326,
+      "learning_rate": 6.39317616049202e-05,
+      "loss": 1.5267,
+      "step": 1978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11165068365611418,
+      "learning_rate": 6.391858251574169e-05,
+      "loss": 1.5274,
+      "step": 1979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.147034446588346,
+      "learning_rate": 6.390540342656318e-05,
+      "loss": 1.5194,
+      "step": 1980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1640015115937058,
+      "learning_rate": 6.389222433738469e-05,
+      "loss": 1.5165,
+      "step": 1981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16856586394524478,
+      "learning_rate": 6.387904524820619e-05,
+      "loss": 1.5248,
+      "step": 1982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1578909602787134,
+      "learning_rate": 6.386586615902768e-05,
+      "loss": 1.531,
+      "step": 1983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1296790414705928,
+      "learning_rate": 6.385268706984919e-05,
+      "loss": 1.5315,
+      "step": 1984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12693440111703813,
+      "learning_rate": 6.383950798067067e-05,
+      "loss": 1.5293,
+      "step": 1985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12488187967920705,
+      "learning_rate": 6.382632889149216e-05,
+      "loss": 1.529,
+      "step": 1986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12255430514995924,
+      "learning_rate": 6.381314980231367e-05,
+      "loss": 1.533,
+      "step": 1987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1283520170754381,
+      "learning_rate": 6.379997071313516e-05,
+      "loss": 1.5308,
+      "step": 1988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12147311544977982,
+      "learning_rate": 6.378679162395666e-05,
+      "loss": 1.5332,
+      "step": 1989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11603357137865991,
+      "learning_rate": 6.377361253477816e-05,
+      "loss": 1.5233,
+      "step": 1990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1174380038288134,
+      "learning_rate": 6.376043344559966e-05,
+      "loss": 1.5303,
+      "step": 1991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11046838544326881,
+      "learning_rate": 6.374725435642115e-05,
+      "loss": 1.5285,
+      "step": 1992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11379676519576967,
+      "learning_rate": 6.373407526724264e-05,
+      "loss": 1.5221,
+      "step": 1993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11086152224760046,
+      "learning_rate": 6.372089617806414e-05,
+      "loss": 1.5283,
+      "step": 1994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12149543614351632,
+      "learning_rate": 6.370771708888563e-05,
+      "loss": 1.5304,
+      "step": 1995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.12589873641900892,
+      "learning_rate": 6.369453799970714e-05,
+      "loss": 1.5175,
+      "step": 1996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13532060586153982,
+      "learning_rate": 6.368135891052863e-05,
+      "loss": 1.5276,
+      "step": 1997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15843934522773848,
+      "learning_rate": 6.366817982135013e-05,
+      "loss": 1.5314,
+      "step": 1998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16436645398371102,
+      "learning_rate": 6.365500073217163e-05,
+      "loss": 1.5337,
+      "step": 1999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16019189514639934,
+      "learning_rate": 6.364182164299313e-05,
+      "loss": 1.5273,
+      "step": 2000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14677216245487681,
+      "learning_rate": 6.362864255381461e-05,
+      "loss": 1.5249,
+      "step": 2001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13885247148542626,
+      "learning_rate": 6.361546346463611e-05,
+      "loss": 1.5235,
+      "step": 2002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1381079419379639,
+      "learning_rate": 6.360228437545761e-05,
+      "loss": 1.5313,
+      "step": 2003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13543367276533416,
+      "learning_rate": 6.35891052862791e-05,
+      "loss": 1.5318,
+      "step": 2004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14667006522729314,
+      "learning_rate": 6.357592619710061e-05,
+      "loss": 1.5302,
+      "step": 2005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15636463113996518,
+      "learning_rate": 6.35627471079221e-05,
+      "loss": 1.5246,
+      "step": 2006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16288154732922017,
+      "learning_rate": 6.35495680187436e-05,
+      "loss": 1.5341,
+      "step": 2007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1399915004877408,
+      "learning_rate": 6.35363889295651e-05,
+      "loss": 1.5257,
+      "step": 2008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13054217167181992,
+      "learning_rate": 6.352320984038658e-05,
+      "loss": 1.5184,
+      "step": 2009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13244813659161758,
+      "learning_rate": 6.351003075120808e-05,
+      "loss": 1.5241,
+      "step": 2010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.13131350747728096,
+      "learning_rate": 6.349685166202959e-05,
+      "loss": 1.5251,
+      "step": 2011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14730544040104143,
+      "learning_rate": 6.348367257285108e-05,
+      "loss": 1.5286,
+      "step": 2012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15135394532562235,
+      "learning_rate": 6.347049348367257e-05,
+      "loss": 1.5225,
+      "step": 2013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14174233246152268,
+      "learning_rate": 6.345731439449408e-05,
+      "loss": 1.5341,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13425377150563123,
+      "learning_rate": 6.344413530531557e-05,
+      "loss": 1.5194,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13735074005602388,
+      "learning_rate": 6.343095621613707e-05,
+      "loss": 1.5321,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15473624121288052,
+      "learning_rate": 6.341777712695856e-05,
+      "loss": 1.5278,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16959113828312192,
+      "learning_rate": 6.340459803778006e-05,
+      "loss": 1.5232,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16364539039813342,
+      "learning_rate": 6.339141894860155e-05,
+      "loss": 1.535,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1554188858293167,
+      "learning_rate": 6.337823985942306e-05,
+      "loss": 1.5298,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12891206482439102,
+      "learning_rate": 6.336506077024455e-05,
+      "loss": 1.5207,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16511871071057288,
+      "learning_rate": 6.335188168106604e-05,
+      "loss": 1.5317,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15123104752778885,
+      "learning_rate": 6.333870259188755e-05,
+      "loss": 1.5316,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1680258555015795,
+      "learning_rate": 6.332552350270904e-05,
+      "loss": 1.5281,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15804994029225125,
+      "learning_rate": 6.331234441353053e-05,
+      "loss": 1.5344,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11312046318528826,
+      "learning_rate": 6.329916532435203e-05,
+      "loss": 1.5234,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11775224240049975,
+      "learning_rate": 6.328598623517353e-05,
+      "loss": 1.5248,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.146637135719089,
+      "learning_rate": 6.327280714599502e-05,
+      "loss": 1.5257,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15149046747220676,
+      "learning_rate": 6.325962805681653e-05,
+      "loss": 1.5287,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15128944270360656,
+      "learning_rate": 6.324644896763802e-05,
+      "loss": 1.5226,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1569931118523244,
+      "learning_rate": 6.323326987845951e-05,
+      "loss": 1.5287,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.144858116749402,
+      "learning_rate": 6.322009078928102e-05,
+      "loss": 1.5236,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14044629501359082,
+      "learning_rate": 6.32069117001025e-05,
+      "loss": 1.5268,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1842119584107408,
+      "learning_rate": 6.3193732610924e-05,
+      "loss": 1.5192,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20988283059149537,
+      "learning_rate": 6.31805535217455e-05,
+      "loss": 1.5313,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.20986909901565357,
+      "learning_rate": 6.3167374432567e-05,
+      "loss": 1.5242,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1677510987702032,
+      "learning_rate": 6.315419534338849e-05,
+      "loss": 1.5258,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11771005464733358,
+      "learning_rate": 6.314101625421e-05,
+      "loss": 1.5361,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11056557991347925,
+      "learning_rate": 6.312783716503149e-05,
+      "loss": 1.5281,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14285556468804533,
+      "learning_rate": 6.311465807585299e-05,
+      "loss": 1.5182,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14920707132294436,
+      "learning_rate": 6.310147898667448e-05,
+      "loss": 1.5268,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12531920177305467,
+      "learning_rate": 6.308829989749597e-05,
+      "loss": 1.5304,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.10916276396215785,
+      "learning_rate": 6.307512080831747e-05,
+      "loss": 1.5301,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11880903156835278,
+      "learning_rate": 6.306194171913897e-05,
+      "loss": 1.5294,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1476200906034467,
+      "learning_rate": 6.304876262996047e-05,
+      "loss": 1.5296,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14903269063748992,
+      "learning_rate": 6.303558354078196e-05,
+      "loss": 1.5249,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15459600834741957,
+      "learning_rate": 6.302240445160347e-05,
+      "loss": 1.5197,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15672779258224045,
+      "learning_rate": 6.300922536242496e-05,
+      "loss": 1.5272,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13938939898386407,
+      "learning_rate": 6.299604627324644e-05,
+      "loss": 1.5235,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1097708027466884,
+      "learning_rate": 6.298286718406795e-05,
+      "loss": 1.5255,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1365426157965458,
+      "learning_rate": 6.296968809488944e-05,
+      "loss": 1.5174,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17341059382962068,
+      "learning_rate": 6.295650900571094e-05,
+      "loss": 1.5316,
+      "step": 2052
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.19809015466837052,
+      "learning_rate": 6.294332991653244e-05,
+      "loss": 1.5285,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17858032782243852,
+      "learning_rate": 6.293015082735394e-05,
+      "loss": 1.527,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1526241802177916,
+      "learning_rate": 6.291697173817543e-05,
+      "loss": 1.5323,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.167139678562081,
+      "learning_rate": 6.290379264899694e-05,
+      "loss": 1.5266,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16257078929642566,
+      "learning_rate": 6.289061355981842e-05,
+      "loss": 1.5288,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14761594393287217,
+      "learning_rate": 6.287743447063991e-05,
+      "loss": 1.5266,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15818684870392483,
+      "learning_rate": 6.286425538146142e-05,
+      "loss": 1.5237,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16913684589848507,
+      "learning_rate": 6.285107629228291e-05,
+      "loss": 1.5212,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13880128501893005,
+      "learning_rate": 6.283789720310441e-05,
+      "loss": 1.5177,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12000133832309363,
+      "learning_rate": 6.282471811392592e-05,
+      "loss": 1.5208,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15034483805846374,
+      "learning_rate": 6.281153902474741e-05,
+      "loss": 1.5161,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1518665293271758,
+      "learning_rate": 6.27983599355689e-05,
+      "loss": 1.5216,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11210002886617655,
+      "learning_rate": 6.27851808463904e-05,
+      "loss": 1.5167,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13764707094949952,
+      "learning_rate": 6.277200175721189e-05,
+      "loss": 1.5192,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15876657522430118,
+      "learning_rate": 6.275882266803338e-05,
+      "loss": 1.5169,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1458913519388383,
+      "learning_rate": 6.274564357885489e-05,
+      "loss": 1.5266,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13936142826589698,
+      "learning_rate": 6.273246448967639e-05,
+      "loss": 1.5206,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14290884117408803,
+      "learning_rate": 6.271928540049788e-05,
+      "loss": 1.5256,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14658526556232296,
+      "learning_rate": 6.270610631131939e-05,
+      "loss": 1.5242,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1318474205962593,
+      "learning_rate": 6.269292722214088e-05,
+      "loss": 1.5222,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13942608972220757,
+      "learning_rate": 6.267974813296236e-05,
+      "loss": 1.5238,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1600106030502008,
+      "learning_rate": 6.266656904378387e-05,
+      "loss": 1.5363,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16540670613012,
+      "learning_rate": 6.265338995460536e-05,
+      "loss": 1.5185,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15660660018644842,
+      "learning_rate": 6.264021086542685e-05,
+      "loss": 1.5213,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14070789360372435,
+      "learning_rate": 6.262703177624836e-05,
+      "loss": 1.5234,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.12650163493219327,
+      "learning_rate": 6.261385268706986e-05,
+      "loss": 1.5204,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14292000982232989,
+      "learning_rate": 6.260067359789135e-05,
+      "loss": 1.5206,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1781726143439805,
+      "learning_rate": 6.258749450871286e-05,
+      "loss": 1.5269,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17244164409599222,
+      "learning_rate": 6.257431541953434e-05,
+      "loss": 1.5243,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13845288598934985,
+      "learning_rate": 6.256113633035583e-05,
+      "loss": 1.5264,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.11816100900783928,
+      "learning_rate": 6.254795724117734e-05,
+      "loss": 1.5324,
+      "step": 2083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10858824455480823,
+      "learning_rate": 6.253477815199883e-05,
+      "loss": 1.5167,
+      "step": 2084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10916423406931997,
+      "learning_rate": 6.252159906282033e-05,
+      "loss": 1.5219,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11799330989933059,
+      "learning_rate": 6.250841997364183e-05,
+      "loss": 1.5223,
+      "step": 2086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12280915579934909,
+      "learning_rate": 6.249524088446333e-05,
+      "loss": 1.5188,
+      "step": 2087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1129198867028369,
+      "learning_rate": 6.248206179528482e-05,
+      "loss": 1.5242,
+      "step": 2088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11075462406399025,
+      "learning_rate": 6.246888270610631e-05,
+      "loss": 1.5194,
+      "step": 2089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12384079974973078,
+      "learning_rate": 6.245570361692781e-05,
+      "loss": 1.5237,
+      "step": 2090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1314566500906056,
+      "learning_rate": 6.24425245277493e-05,
+      "loss": 1.5183,
+      "step": 2091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14104692187892323,
+      "learning_rate": 6.242934543857081e-05,
+      "loss": 1.5327,
+      "step": 2092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1320262214014273,
+      "learning_rate": 6.24161663493923e-05,
+      "loss": 1.5217,
+      "step": 2093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12370053498964839,
+      "learning_rate": 6.24029872602138e-05,
+      "loss": 1.5303,
+      "step": 2094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10601382170389088,
+      "learning_rate": 6.23898081710353e-05,
+      "loss": 1.5236,
+      "step": 2095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10673248431632194,
+      "learning_rate": 6.23766290818568e-05,
+      "loss": 1.5222,
+      "step": 2096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10255670048757975,
+      "learning_rate": 6.236344999267828e-05,
+      "loss": 1.5338,
+      "step": 2097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11300182178978282,
+      "learning_rate": 6.235027090349979e-05,
+      "loss": 1.5169,
+      "step": 2098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13784993946486357,
+      "learning_rate": 6.233709181432128e-05,
+      "loss": 1.5186,
+      "step": 2099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17433167829909155,
+      "learning_rate": 6.232391272514277e-05,
+      "loss": 1.5218,
+      "step": 2100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18480759269312977,
+      "learning_rate": 6.231073363596428e-05,
+      "loss": 1.518,
+      "step": 2101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1610794472745597,
+      "learning_rate": 6.229755454678577e-05,
+      "loss": 1.5216,
+      "step": 2102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1570270132379384,
+      "learning_rate": 6.228437545760727e-05,
+      "loss": 1.53,
+      "step": 2103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14994918598135903,
+      "learning_rate": 6.227119636842877e-05,
+      "loss": 1.5199,
+      "step": 2104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11732884280673465,
+      "learning_rate": 6.225801727925025e-05,
+      "loss": 1.5182,
+      "step": 2105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11551578676040783,
+      "learning_rate": 6.224483819007175e-05,
+      "loss": 1.5254,
+      "step": 2106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11494328541395543,
+      "learning_rate": 6.223165910089326e-05,
+      "loss": 1.5315,
+      "step": 2107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1159629906763021,
+      "learning_rate": 6.221848001171475e-05,
+      "loss": 1.5278,
+      "step": 2108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12482748178892049,
+      "learning_rate": 6.220530092253624e-05,
+      "loss": 1.5288,
+      "step": 2109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11945074090659315,
+      "learning_rate": 6.219212183335775e-05,
+      "loss": 1.526,
+      "step": 2110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1261009538026196,
+      "learning_rate": 6.217894274417924e-05,
+      "loss": 1.5157,
+      "step": 2111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13772208065010083,
+      "learning_rate": 6.216576365500074e-05,
+      "loss": 1.5254,
+      "step": 2112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11745654353857617,
+      "learning_rate": 6.215258456582223e-05,
+      "loss": 1.5252,
+      "step": 2113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11112063557293947,
+      "learning_rate": 6.213940547664373e-05,
+      "loss": 1.5161,
+      "step": 2114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14782301294193873,
+      "learning_rate": 6.212622638746522e-05,
+      "loss": 1.5209,
+      "step": 2115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1426490994014961,
+      "learning_rate": 6.211304729828673e-05,
+      "loss": 1.53,
+      "step": 2116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13353458436953602,
+      "learning_rate": 6.209986820910822e-05,
+      "loss": 1.523,
+      "step": 2117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13384075917485422,
+      "learning_rate": 6.208668911992971e-05,
+      "loss": 1.5214,
+      "step": 2118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1467649180316527,
+      "learning_rate": 6.207351003075122e-05,
+      "loss": 1.5191,
+      "step": 2119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1341940269480434,
+      "learning_rate": 6.206033094157272e-05,
+      "loss": 1.5256,
+      "step": 2120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12150781117978154,
+      "learning_rate": 6.20471518523942e-05,
+      "loss": 1.5268,
+      "step": 2121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13490472077923257,
+      "learning_rate": 6.20339727632157e-05,
+      "loss": 1.523,
+      "step": 2122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15341671024188455,
+      "learning_rate": 6.20207936740372e-05,
+      "loss": 1.527,
+      "step": 2123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15940683740767694,
+      "learning_rate": 6.200761458485869e-05,
+      "loss": 1.5256,
+      "step": 2124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1514136424075867,
+      "learning_rate": 6.19944354956802e-05,
+      "loss": 1.5267,
+      "step": 2125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15394978792425873,
+      "learning_rate": 6.198125640650169e-05,
+      "loss": 1.5206,
+      "step": 2126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14754759110411791,
+      "learning_rate": 6.196807731732318e-05,
+      "loss": 1.5242,
+      "step": 2127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1343463931900218,
+      "learning_rate": 6.195489822814468e-05,
+      "loss": 1.5296,
+      "step": 2128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.12606902242283752,
+      "learning_rate": 6.194171913896617e-05,
+      "loss": 1.5198,
+      "step": 2129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1386164722778369,
+      "learning_rate": 6.192854004978767e-05,
+      "loss": 1.524,
+      "step": 2130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16720177836513764,
+      "learning_rate": 6.191536096060917e-05,
+      "loss": 1.5287,
+      "step": 2131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18036784228072253,
+      "learning_rate": 6.190218187143067e-05,
+      "loss": 1.5276,
+      "step": 2132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.19996670931802857,
+      "learning_rate": 6.188900278225216e-05,
+      "loss": 1.5142,
+      "step": 2133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1798206623135205,
+      "learning_rate": 6.187582369307367e-05,
+      "loss": 1.5281,
+      "step": 2134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1543231328673258,
+      "learning_rate": 6.186264460389516e-05,
+      "loss": 1.5251,
+      "step": 2135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15869241635868248,
+      "learning_rate": 6.184946551471666e-05,
+      "loss": 1.5263,
+      "step": 2136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16892357931889682,
+      "learning_rate": 6.183628642553815e-05,
+      "loss": 1.5171,
+      "step": 2137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17053501592063305,
+      "learning_rate": 6.182310733635964e-05,
+      "loss": 1.52,
+      "step": 2138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14972724557662342,
+      "learning_rate": 6.180992824718114e-05,
+      "loss": 1.5246,
+      "step": 2139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11810819683874355,
+      "learning_rate": 6.179674915800264e-05,
+      "loss": 1.5237,
+      "step": 2140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1341293478568526,
+      "learning_rate": 6.178357006882414e-05,
+      "loss": 1.5162,
+      "step": 2141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1499312699717389,
+      "learning_rate": 6.177039097964563e-05,
+      "loss": 1.519,
+      "step": 2142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15556310858579758,
+      "learning_rate": 6.175721189046714e-05,
+      "loss": 1.5159,
+      "step": 2143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.13429451605235562,
+      "learning_rate": 6.174403280128863e-05,
+      "loss": 1.5171,
+      "step": 2144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.10326832793003027,
+      "learning_rate": 6.173085371211011e-05,
+      "loss": 1.5203,
+      "step": 2145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1278609006530773,
+      "learning_rate": 6.171767462293162e-05,
+      "loss": 1.5147,
+      "step": 2146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14592769814411263,
+      "learning_rate": 6.170449553375311e-05,
+      "loss": 1.5218,
+      "step": 2147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11893169124434697,
+      "learning_rate": 6.169131644457461e-05,
+      "loss": 1.5193,
+      "step": 2148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11218227875376423,
+      "learning_rate": 6.167813735539611e-05,
+      "loss": 1.5191,
+      "step": 2149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1325608606374997,
+      "learning_rate": 6.166495826621761e-05,
+      "loss": 1.5182,
+      "step": 2150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1376841895840266,
+      "learning_rate": 6.16517791770391e-05,
+      "loss": 1.5162,
+      "step": 2151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13468549379655173,
+      "learning_rate": 6.16386000878606e-05,
+      "loss": 1.5166,
+      "step": 2152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11899066473699894,
+      "learning_rate": 6.162542099868209e-05,
+      "loss": 1.5212,
+      "step": 2153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12021969210563359,
+      "learning_rate": 6.161224190950358e-05,
+      "loss": 1.5168,
+      "step": 2154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.130004921563983,
+      "learning_rate": 6.159906282032509e-05,
+      "loss": 1.5125,
+      "step": 2155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11338274254315026,
+      "learning_rate": 6.158588373114658e-05,
+      "loss": 1.5168,
+      "step": 2156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12346004929071033,
+      "learning_rate": 6.157270464196808e-05,
+      "loss": 1.5303,
+      "step": 2157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12537590301914633,
+      "learning_rate": 6.155952555278959e-05,
+      "loss": 1.5335,
+      "step": 2158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11383495976122389,
+      "learning_rate": 6.154634646361108e-05,
+      "loss": 1.5255,
+      "step": 2159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11584594695712067,
+      "learning_rate": 6.153316737443257e-05,
+      "loss": 1.5181,
+      "step": 2160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12418961877395383,
+      "learning_rate": 6.151998828525407e-05,
+      "loss": 1.5219,
+      "step": 2161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11563294528739124,
+      "learning_rate": 6.150680919607556e-05,
+      "loss": 1.512,
+      "step": 2162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11098924022724428,
+      "learning_rate": 6.149363010689705e-05,
+      "loss": 1.5254,
+      "step": 2163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1271328090610676,
+      "learning_rate": 6.148045101771856e-05,
+      "loss": 1.5207,
+      "step": 2164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12682596692685158,
+      "learning_rate": 6.146727192854006e-05,
+      "loss": 1.5216,
+      "step": 2165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1359494269149639,
+      "learning_rate": 6.145409283936155e-05,
+      "loss": 1.5242,
+      "step": 2166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1538581193615903,
+      "learning_rate": 6.144091375018306e-05,
+      "loss": 1.5221,
+      "step": 2167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1438246973540772,
+      "learning_rate": 6.142773466100455e-05,
+      "loss": 1.5052,
+      "step": 2168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12598987092020772,
+      "learning_rate": 6.141455557182603e-05,
+      "loss": 1.5192,
+      "step": 2169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14076273253077945,
+      "learning_rate": 6.140137648264754e-05,
+      "loss": 1.5195,
+      "step": 2170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14192256149988136,
+      "learning_rate": 6.138819739346903e-05,
+      "loss": 1.5189,
+      "step": 2171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14972816528561844,
+      "learning_rate": 6.137501830429053e-05,
+      "loss": 1.5282,
+      "step": 2172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15744887915544908,
+      "learning_rate": 6.136183921511203e-05,
+      "loss": 1.5145,
+      "step": 2173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15426414173763175,
+      "learning_rate": 6.134866012593353e-05,
+      "loss": 1.5129,
+      "step": 2174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14919806560287394,
+      "learning_rate": 6.133548103675502e-05,
+      "loss": 1.5231,
+      "step": 2175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12908110477474147,
+      "learning_rate": 6.132230194757651e-05,
+      "loss": 1.5201,
+      "step": 2176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1212412104157957,
+      "learning_rate": 6.130912285839801e-05,
+      "loss": 1.5244,
+      "step": 2177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11759027894844022,
+      "learning_rate": 6.12959437692195e-05,
+      "loss": 1.5161,
+      "step": 2178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12909687575249174,
+      "learning_rate": 6.128276468004101e-05,
+      "loss": 1.5204,
+      "step": 2179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1246755699299353,
+      "learning_rate": 6.12695855908625e-05,
+      "loss": 1.5155,
+      "step": 2180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12017485686709545,
+      "learning_rate": 6.1256406501684e-05,
+      "loss": 1.519,
+      "step": 2181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12438424819224814,
+      "learning_rate": 6.12432274125055e-05,
+      "loss": 1.5226,
+      "step": 2182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11717613422885555,
+      "learning_rate": 6.1230048323327e-05,
+      "loss": 1.5208,
+      "step": 2183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.11385027050578662,
+      "learning_rate": 6.121686923414849e-05,
+      "loss": 1.5223,
+      "step": 2184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12243790439467431,
+      "learning_rate": 6.120369014496998e-05,
+      "loss": 1.5197,
+      "step": 2185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1535769800043021,
+      "learning_rate": 6.119051105579148e-05,
+      "loss": 1.5119,
+      "step": 2186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1563604709168843,
+      "learning_rate": 6.117733196661297e-05,
+      "loss": 1.5188,
+      "step": 2187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14345600535864092,
+      "learning_rate": 6.116415287743448e-05,
+      "loss": 1.5139,
+      "step": 2188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14401933145453452,
+      "learning_rate": 6.115097378825597e-05,
+      "loss": 1.523,
+      "step": 2189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15050556940892965,
+      "learning_rate": 6.113779469907747e-05,
+      "loss": 1.5215,
+      "step": 2190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14881320020338035,
+      "learning_rate": 6.112461560989897e-05,
+      "loss": 1.5146,
+      "step": 2191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13100690192036496,
+      "learning_rate": 6.111143652072047e-05,
+      "loss": 1.5158,
+      "step": 2192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1492501680035062,
+      "learning_rate": 6.109825743154195e-05,
+      "loss": 1.5246,
+      "step": 2193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12403074189373349,
+      "learning_rate": 6.108507834236346e-05,
+      "loss": 1.5219,
+      "step": 2194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.10735070017391429,
+      "learning_rate": 6.107189925318495e-05,
+      "loss": 1.5147,
+      "step": 2195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1232443082160195,
+      "learning_rate": 6.105872016400644e-05,
+      "loss": 1.5245,
+      "step": 2196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14006600996983623,
+      "learning_rate": 6.104554107482795e-05,
+      "loss": 1.5245,
+      "step": 2197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15768477104096448,
+      "learning_rate": 6.1032361985649444e-05,
+      "loss": 1.5103,
+      "step": 2198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15262229524260412,
+      "learning_rate": 6.101918289647094e-05,
+      "loss": 1.5171,
+      "step": 2199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1338256255742811,
+      "learning_rate": 6.100600380729244e-05,
+      "loss": 1.5262,
+      "step": 2200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12576402768689363,
+      "learning_rate": 6.0992824718113925e-05,
+      "loss": 1.5204,
+      "step": 2201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14227783767708035,
+      "learning_rate": 6.0979645628935426e-05,
+      "loss": 1.5274,
+      "step": 2202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1394660145793301,
+      "learning_rate": 6.096646653975692e-05,
+      "loss": 1.5146,
+      "step": 2203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1549578750169,
+      "learning_rate": 6.095328745057842e-05,
+      "loss": 1.5173,
+      "step": 2204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15966982210385058,
+      "learning_rate": 6.0940108361399914e-05,
+      "loss": 1.5191,
+      "step": 2205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14148208330633738,
+      "learning_rate": 6.0926929272221414e-05,
+      "loss": 1.5194,
+      "step": 2206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13135188439858728,
+      "learning_rate": 6.0913750183042915e-05,
+      "loss": 1.5146,
+      "step": 2207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13726796317359796,
+      "learning_rate": 6.090057109386441e-05,
+      "loss": 1.5178,
+      "step": 2208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14972362616720578,
+      "learning_rate": 6.0887392004685895e-05,
+      "loss": 1.5148,
+      "step": 2209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12931920561764262,
+      "learning_rate": 6.0874212915507396e-05,
+      "loss": 1.5242,
+      "step": 2210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1392816924653237,
+      "learning_rate": 6.0861033826328896e-05,
+      "loss": 1.5203,
+      "step": 2211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14016949656829314,
+      "learning_rate": 6.084785473715039e-05,
+      "loss": 1.512,
+      "step": 2212
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14068070920322853,
+      "learning_rate": 6.083467564797189e-05,
+      "loss": 1.5138,
+      "step": 2213
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1347681189289797,
+      "learning_rate": 6.0821496558793384e-05,
+      "loss": 1.5094,
+      "step": 2214
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1306889685215971,
+      "learning_rate": 6.0808317469614885e-05,
+      "loss": 1.5133,
+      "step": 2215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13787621691363156,
+      "learning_rate": 6.0795138380436385e-05,
+      "loss": 1.5168,
+      "step": 2216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14438672619225318,
+      "learning_rate": 6.078195929125787e-05,
+      "loss": 1.516,
+      "step": 2217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1403438357513446,
+      "learning_rate": 6.0768780202079366e-05,
+      "loss": 1.5144,
+      "step": 2218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.13343726479479032,
+      "learning_rate": 6.075560111290087e-05,
+      "loss": 1.509,
+      "step": 2219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14207319565332172,
+      "learning_rate": 6.074242202372237e-05,
+      "loss": 1.5153,
+      "step": 2220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15018152618851524,
+      "learning_rate": 6.072924293454386e-05,
+      "loss": 1.509,
+      "step": 2221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14253630903469122,
+      "learning_rate": 6.071606384536536e-05,
+      "loss": 1.5063,
+      "step": 2222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11996621425435175,
+      "learning_rate": 6.0702884756186855e-05,
+      "loss": 1.5203,
+      "step": 2223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10632231844499059,
+      "learning_rate": 6.0689705667008356e-05,
+      "loss": 1.5233,
+      "step": 2224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11023809622073384,
+      "learning_rate": 6.067652657782984e-05,
+      "loss": 1.5195,
+      "step": 2225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12987334915415513,
+      "learning_rate": 6.066334748865134e-05,
+      "loss": 1.5223,
+      "step": 2226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15229736794814935,
+      "learning_rate": 6.065016839947284e-05,
+      "loss": 1.5154,
+      "step": 2227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1454958330057587,
+      "learning_rate": 6.063698931029434e-05,
+      "loss": 1.5092,
+      "step": 2228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12754791406716226,
+      "learning_rate": 6.062381022111583e-05,
+      "loss": 1.5133,
+      "step": 2229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11375636252849265,
+      "learning_rate": 6.061063113193733e-05,
+      "loss": 1.5186,
+      "step": 2230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10627831212203363,
+      "learning_rate": 6.059745204275883e-05,
+      "loss": 1.5178,
+      "step": 2231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11476524189766231,
+      "learning_rate": 6.0584272953580326e-05,
+      "loss": 1.518,
+      "step": 2232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13067748752523878,
+      "learning_rate": 6.057109386440181e-05,
+      "loss": 1.5105,
+      "step": 2233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1492409448564739,
+      "learning_rate": 6.0557914775223313e-05,
+      "loss": 1.512,
+      "step": 2234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.154892183214198,
+      "learning_rate": 6.0544735686044814e-05,
+      "loss": 1.5123,
+      "step": 2235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14459205025903116,
+      "learning_rate": 6.053155659686631e-05,
+      "loss": 1.5174,
+      "step": 2236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1377349791022884,
+      "learning_rate": 6.051837750768781e-05,
+      "loss": 1.5191,
+      "step": 2237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15735570348333222,
+      "learning_rate": 6.05051984185093e-05,
+      "loss": 1.5212,
+      "step": 2238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1705552500302857,
+      "learning_rate": 6.04920193293308e-05,
+      "loss": 1.52,
+      "step": 2239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16058361612777433,
+      "learning_rate": 6.04788402401523e-05,
+      "loss": 1.5108,
+      "step": 2240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11862762681033137,
+      "learning_rate": 6.046566115097379e-05,
+      "loss": 1.504,
+      "step": 2241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0980019746128759,
+      "learning_rate": 6.0452482061795284e-05,
+      "loss": 1.5118,
+      "step": 2242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11372989415808786,
+      "learning_rate": 6.0439302972616784e-05,
+      "loss": 1.5166,
+      "step": 2243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14048980444841072,
+      "learning_rate": 6.042612388343828e-05,
+      "loss": 1.5144,
+      "step": 2244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17491367992626988,
+      "learning_rate": 6.041294479425978e-05,
+      "loss": 1.5119,
+      "step": 2245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18790579518418113,
+      "learning_rate": 6.039976570508128e-05,
+      "loss": 1.5094,
+      "step": 2246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16394636100142454,
+      "learning_rate": 6.038658661590277e-05,
+      "loss": 1.5233,
+      "step": 2247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1495500588763787,
+      "learning_rate": 6.037340752672427e-05,
+      "loss": 1.5129,
+      "step": 2248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1422219547478819,
+      "learning_rate": 6.036022843754576e-05,
+      "loss": 1.5161,
+      "step": 2249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13395423795003536,
+      "learning_rate": 6.034704934836726e-05,
+      "loss": 1.5072,
+      "step": 2250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11883866308225731,
+      "learning_rate": 6.0333870259188754e-05,
+      "loss": 1.5251,
+      "step": 2251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13630500334424248,
+      "learning_rate": 6.0320691170010255e-05,
+      "loss": 1.5219,
+      "step": 2252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15083395736754596,
+      "learning_rate": 6.030751208083175e-05,
+      "loss": 1.5179,
+      "step": 2253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12782707662496293,
+      "learning_rate": 6.029433299165325e-05,
+      "loss": 1.5212,
+      "step": 2254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11308165374543254,
+      "learning_rate": 6.028115390247475e-05,
+      "loss": 1.5086,
+      "step": 2255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12842026365312598,
+      "learning_rate": 6.0267974813296243e-05,
+      "loss": 1.5127,
+      "step": 2256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1274407291993604,
+      "learning_rate": 6.025479572411773e-05,
+      "loss": 1.5088,
+      "step": 2257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12737119408607142,
+      "learning_rate": 6.024161663493923e-05,
+      "loss": 1.5147,
+      "step": 2258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1244518278208766,
+      "learning_rate": 6.022843754576073e-05,
+      "loss": 1.5162,
+      "step": 2259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12057188889356055,
+      "learning_rate": 6.0215258456582225e-05,
+      "loss": 1.5124,
+      "step": 2260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12366531835743183,
+      "learning_rate": 6.0202079367403726e-05,
+      "loss": 1.5141,
+      "step": 2261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14221191121841073,
+      "learning_rate": 6.018890027822522e-05,
+      "loss": 1.511,
+      "step": 2262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1643428440713606,
+      "learning_rate": 6.017572118904672e-05,
+      "loss": 1.5116,
+      "step": 2263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16810873599560244,
+      "learning_rate": 6.016254209986822e-05,
+      "loss": 1.5097,
+      "step": 2264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13418500271250552,
+      "learning_rate": 6.014936301068971e-05,
+      "loss": 1.5093,
+      "step": 2265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10842319253428946,
+      "learning_rate": 6.01361839215112e-05,
+      "loss": 1.5153,
+      "step": 2266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1306804214291702,
+      "learning_rate": 6.01230048323327e-05,
+      "loss": 1.5112,
+      "step": 2267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1483384304288999,
+      "learning_rate": 6.0109825743154195e-05,
+      "loss": 1.5105,
+      "step": 2268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15329531745665928,
+      "learning_rate": 6.0096646653975696e-05,
+      "loss": 1.5117,
+      "step": 2269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13330554061078204,
+      "learning_rate": 6.0083467564797197e-05,
+      "loss": 1.513,
+      "step": 2270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12541350675748444,
+      "learning_rate": 6.007028847561869e-05,
+      "loss": 1.5062,
+      "step": 2271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15185901417694483,
+      "learning_rate": 6.005710938644019e-05,
+      "loss": 1.5213,
+      "step": 2272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17323560079697678,
+      "learning_rate": 6.004393029726168e-05,
+      "loss": 1.5168,
+      "step": 2273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17365062133664777,
+      "learning_rate": 6.003075120808318e-05,
+      "loss": 1.5188,
+      "step": 2274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14260803239692582,
+      "learning_rate": 6.001757211890467e-05,
+      "loss": 1.5141,
+      "step": 2275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11856328850877632,
+      "learning_rate": 6.000439302972617e-05,
+      "loss": 1.515,
+      "step": 2276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11614783272823101,
+      "learning_rate": 5.9991213940547666e-05,
+      "loss": 1.5193,
+      "step": 2277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12035431959878432,
+      "learning_rate": 5.997803485136917e-05,
+      "loss": 1.5217,
+      "step": 2278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11826757131298729,
+      "learning_rate": 5.996485576219067e-05,
+      "loss": 1.5196,
+      "step": 2279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10158968047111043,
+      "learning_rate": 5.995167667301216e-05,
+      "loss": 1.5116,
+      "step": 2280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10427228208635114,
+      "learning_rate": 5.993849758383365e-05,
+      "loss": 1.5179,
+      "step": 2281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1226445252814833,
+      "learning_rate": 5.992531849465515e-05,
+      "loss": 1.5094,
+      "step": 2282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11839570582485213,
+      "learning_rate": 5.991213940547665e-05,
+      "loss": 1.5102,
+      "step": 2283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.11155394009657069,
+      "learning_rate": 5.989896031629814e-05,
+      "loss": 1.5218,
+      "step": 2284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.10061147077510676,
+      "learning_rate": 5.988578122711964e-05,
+      "loss": 1.5143,
+      "step": 2285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.12176600975112552,
+      "learning_rate": 5.987260213794114e-05,
+      "loss": 1.5048,
+      "step": 2286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14557431900923776,
+      "learning_rate": 5.985942304876264e-05,
+      "loss": 1.5215,
+      "step": 2287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14220588873925638,
+      "learning_rate": 5.984624395958414e-05,
+      "loss": 1.5103,
+      "step": 2288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13815402662075474,
+      "learning_rate": 5.9833064870405625e-05,
+      "loss": 1.5072,
+      "step": 2289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13598390117939177,
+      "learning_rate": 5.981988578122712e-05,
+      "loss": 1.5099,
+      "step": 2290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14142875591772053,
+      "learning_rate": 5.980670669204862e-05,
+      "loss": 1.512,
+      "step": 2291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11608781189353126,
+      "learning_rate": 5.979352760287011e-05,
+      "loss": 1.5187,
+      "step": 2292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11258987810208357,
+      "learning_rate": 5.9780348513691614e-05,
+      "loss": 1.5179,
+      "step": 2293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1270923916922911,
+      "learning_rate": 5.9767169424513114e-05,
+      "loss": 1.5151,
+      "step": 2294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13133765394517505,
+      "learning_rate": 5.975399033533461e-05,
+      "loss": 1.5069,
+      "step": 2295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11158885718225602,
+      "learning_rate": 5.974081124615611e-05,
+      "loss": 1.5062,
+      "step": 2296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11270342495988682,
+      "learning_rate": 5.9727632156977595e-05,
+      "loss": 1.5042,
+      "step": 2297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1235247552591486,
+      "learning_rate": 5.9714453067799096e-05,
+      "loss": 1.5167,
+      "step": 2298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10741617774758043,
+      "learning_rate": 5.970127397862059e-05,
+      "loss": 1.5125,
+      "step": 2299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1090931812147054,
+      "learning_rate": 5.968809488944209e-05,
+      "loss": 1.5139,
+      "step": 2300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1252055834714194,
+      "learning_rate": 5.9674915800263584e-05,
+      "loss": 1.5083,
+      "step": 2301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11481940130344365,
+      "learning_rate": 5.9661736711085084e-05,
+      "loss": 1.5126,
+      "step": 2302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11815845125507025,
+      "learning_rate": 5.9648557621906585e-05,
+      "loss": 1.5126,
+      "step": 2303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12961524773565936,
+      "learning_rate": 5.963537853272808e-05,
+      "loss": 1.5169,
+      "step": 2304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12218272942945943,
+      "learning_rate": 5.9622199443549566e-05,
+      "loss": 1.5185,
+      "step": 2305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12345677871551518,
+      "learning_rate": 5.9609020354371066e-05,
+      "loss": 1.5128,
+      "step": 2306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14009425097158315,
+      "learning_rate": 5.959584126519257e-05,
+      "loss": 1.503,
+      "step": 2307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1565536909925842,
+      "learning_rate": 5.958266217601406e-05,
+      "loss": 1.5115,
+      "step": 2308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17446475399949485,
+      "learning_rate": 5.956948308683556e-05,
+      "loss": 1.5121,
+      "step": 2309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18241982693618455,
+      "learning_rate": 5.9556303997657055e-05,
+      "loss": 1.5146,
+      "step": 2310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16870708168047874,
+      "learning_rate": 5.9543124908478555e-05,
+      "loss": 1.5136,
+      "step": 2311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14483968346254927,
+      "learning_rate": 5.9529945819300056e-05,
+      "loss": 1.5083,
+      "step": 2312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.136067833805124,
+      "learning_rate": 5.951676673012154e-05,
+      "loss": 1.5175,
+      "step": 2313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16074365590251163,
+      "learning_rate": 5.9503587640943036e-05,
+      "loss": 1.5149,
+      "step": 2314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18442482391346737,
+      "learning_rate": 5.949040855176454e-05,
+      "loss": 1.5045,
+      "step": 2315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17204160786371323,
+      "learning_rate": 5.947722946258603e-05,
+      "loss": 1.5115,
+      "step": 2316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14749100248425168,
+      "learning_rate": 5.946405037340753e-05,
+      "loss": 1.5142,
+      "step": 2317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13741528849507761,
+      "learning_rate": 5.945087128422903e-05,
+      "loss": 1.518,
+      "step": 2318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15953547741827048,
+      "learning_rate": 5.9437692195050525e-05,
+      "loss": 1.5097,
+      "step": 2319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15594698142886393,
+      "learning_rate": 5.9424513105872026e-05,
+      "loss": 1.5105,
+      "step": 2320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12795921343371475,
+      "learning_rate": 5.941133401669351e-05,
+      "loss": 1.5073,
+      "step": 2321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1516510598250895,
+      "learning_rate": 5.9398154927515013e-05,
+      "loss": 1.507,
+      "step": 2322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16801484170839506,
+      "learning_rate": 5.938497583833651e-05,
+      "loss": 1.5108,
+      "step": 2323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1433289099973008,
+      "learning_rate": 5.937179674915801e-05,
+      "loss": 1.5136,
+      "step": 2324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10911239238666354,
+      "learning_rate": 5.93586176599795e-05,
+      "loss": 1.5139,
+      "step": 2325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12109525033515565,
+      "learning_rate": 5.9345438570801e-05,
+      "loss": 1.5133,
+      "step": 2326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13050242986462707,
+      "learning_rate": 5.93322594816225e-05,
+      "loss": 1.5152,
+      "step": 2327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11966664358982611,
+      "learning_rate": 5.9319080392443996e-05,
+      "loss": 1.5033,
+      "step": 2328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12259351327811138,
+      "learning_rate": 5.930590130326548e-05,
+      "loss": 1.5196,
+      "step": 2329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1295117451346283,
+      "learning_rate": 5.9292722214086984e-05,
+      "loss": 1.5124,
+      "step": 2330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13009982710000306,
+      "learning_rate": 5.927954312490848e-05,
+      "loss": 1.5106,
+      "step": 2331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14924677730276928,
+      "learning_rate": 5.926636403572998e-05,
+      "loss": 1.5168,
+      "step": 2332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1643521195010823,
+      "learning_rate": 5.925318494655148e-05,
+      "loss": 1.5171,
+      "step": 2333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14960372601024685,
+      "learning_rate": 5.924000585737297e-05,
+      "loss": 1.511,
+      "step": 2334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12228343418150839,
+      "learning_rate": 5.922682676819447e-05,
+      "loss": 1.5097,
+      "step": 2335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11554949250186168,
+      "learning_rate": 5.9213647679015966e-05,
+      "loss": 1.5117,
+      "step": 2336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12397858855431088,
+      "learning_rate": 5.920046858983746e-05,
+      "loss": 1.5086,
+      "step": 2337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1148775527318414,
+      "learning_rate": 5.9187289500658954e-05,
+      "loss": 1.5134,
+      "step": 2338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12357664880087452,
+      "learning_rate": 5.9174110411480454e-05,
+      "loss": 1.5143,
+      "step": 2339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1254345613825194,
+      "learning_rate": 5.916093132230195e-05,
+      "loss": 1.5117,
+      "step": 2340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11543895487570961,
+      "learning_rate": 5.914775223312345e-05,
+      "loss": 1.5099,
+      "step": 2341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11055522475661987,
+      "learning_rate": 5.913457314394495e-05,
+      "loss": 1.5176,
+      "step": 2342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11661583871800599,
+      "learning_rate": 5.912139405476644e-05,
+      "loss": 1.4989,
+      "step": 2343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14179863617943286,
+      "learning_rate": 5.9108214965587943e-05,
+      "loss": 1.511,
+      "step": 2344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15524190919099576,
+      "learning_rate": 5.909503587640943e-05,
+      "loss": 1.5098,
+      "step": 2345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1664073002111415,
+      "learning_rate": 5.908185678723093e-05,
+      "loss": 1.5067,
+      "step": 2346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18567156023387613,
+      "learning_rate": 5.9068677698052425e-05,
+      "loss": 1.5125,
+      "step": 2347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14915082887518222,
+      "learning_rate": 5.9055498608873925e-05,
+      "loss": 1.5147,
+      "step": 2348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.10131138031405326,
+      "learning_rate": 5.904231951969542e-05,
+      "loss": 1.5077,
+      "step": 2349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.13499286035249544,
+      "learning_rate": 5.902914043051692e-05,
+      "loss": 1.5067,
+      "step": 2350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16757785358469987,
+      "learning_rate": 5.901596134133842e-05,
+      "loss": 1.5263,
+      "step": 2351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.12355971299368526,
+      "learning_rate": 5.9002782252159914e-05,
+      "loss": 1.5062,
+      "step": 2352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.11277402666384855,
+      "learning_rate": 5.89896031629814e-05,
+      "loss": 1.5162,
+      "step": 2353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14851872240131123,
+      "learning_rate": 5.89764240738029e-05,
+      "loss": 1.509,
+      "step": 2354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16035805736946893,
+      "learning_rate": 5.8963244984624395e-05,
+      "loss": 1.5142,
+      "step": 2355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14549309326154064,
+      "learning_rate": 5.8950065895445895e-05,
+      "loss": 1.516,
+      "step": 2356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11823461453541943,
+      "learning_rate": 5.8936886806267396e-05,
+      "loss": 1.5113,
+      "step": 2357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11135648202708422,
+      "learning_rate": 5.892370771708889e-05,
+      "loss": 1.5106,
+      "step": 2358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1280296027026804,
+      "learning_rate": 5.891052862791039e-05,
+      "loss": 1.5171,
+      "step": 2359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13041222793242424,
+      "learning_rate": 5.8897349538731884e-05,
+      "loss": 1.5176,
+      "step": 2360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13265172100585554,
+      "learning_rate": 5.888417044955338e-05,
+      "loss": 1.5089,
+      "step": 2361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11622142253722154,
+      "learning_rate": 5.887099136037487e-05,
+      "loss": 1.5178,
+      "step": 2362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10986139183967045,
+      "learning_rate": 5.885781227119637e-05,
+      "loss": 1.5102,
+      "step": 2363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11959596500157584,
+      "learning_rate": 5.8844633182017866e-05,
+      "loss": 1.5177,
+      "step": 2364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10774005996864272,
+      "learning_rate": 5.8831454092839366e-05,
+      "loss": 1.5139,
+      "step": 2365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11467317828107283,
+      "learning_rate": 5.881827500366087e-05,
+      "loss": 1.5105,
+      "step": 2366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11672584177896994,
+      "learning_rate": 5.880509591448236e-05,
+      "loss": 1.4996,
+      "step": 2367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1179546429265132,
+      "learning_rate": 5.879191682530386e-05,
+      "loss": 1.5073,
+      "step": 2368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11074274602837267,
+      "learning_rate": 5.877873773612535e-05,
+      "loss": 1.5089,
+      "step": 2369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11714311168744063,
+      "learning_rate": 5.876555864694685e-05,
+      "loss": 1.5232,
+      "step": 2370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13930831155271559,
+      "learning_rate": 5.875237955776834e-05,
+      "loss": 1.5041,
+      "step": 2371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1379863245255307,
+      "learning_rate": 5.873920046858984e-05,
+      "loss": 1.4956,
+      "step": 2372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15389324214128705,
+      "learning_rate": 5.8726021379411337e-05,
+      "loss": 1.5125,
+      "step": 2373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17023067534596448,
+      "learning_rate": 5.871284229023284e-05,
+      "loss": 1.5135,
+      "step": 2374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15840504067573727,
+      "learning_rate": 5.869966320105434e-05,
+      "loss": 1.5177,
+      "step": 2375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15449407225936562,
+      "learning_rate": 5.868648411187583e-05,
+      "loss": 1.5112,
+      "step": 2376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1544973344622811,
+      "learning_rate": 5.867330502269732e-05,
+      "loss": 1.509,
+      "step": 2377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15011410916788887,
+      "learning_rate": 5.866012593351882e-05,
+      "loss": 1.5078,
+      "step": 2378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14261796254441736,
+      "learning_rate": 5.864694684434031e-05,
+      "loss": 1.5091,
+      "step": 2379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14803517384307946,
+      "learning_rate": 5.863376775516181e-05,
+      "loss": 1.5034,
+      "step": 2380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1559518266225165,
+      "learning_rate": 5.8620588665983314e-05,
+      "loss": 1.519,
+      "step": 2381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14822233781943328,
+      "learning_rate": 5.860740957680481e-05,
+      "loss": 1.5073,
+      "step": 2382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14977540032924916,
+      "learning_rate": 5.859423048762631e-05,
+      "loss": 1.5158,
+      "step": 2383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12351033347669473,
+      "learning_rate": 5.85810513984478e-05,
+      "loss": 1.5181,
+      "step": 2384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10843185805490717,
+      "learning_rate": 5.8567872309269295e-05,
+      "loss": 1.513,
+      "step": 2385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14290578132206802,
+      "learning_rate": 5.855469322009079e-05,
+      "loss": 1.5108,
+      "step": 2386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13619493438639937,
+      "learning_rate": 5.854151413091229e-05,
+      "loss": 1.5073,
+      "step": 2387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10991519562772543,
+      "learning_rate": 5.852833504173378e-05,
+      "loss": 1.5128,
+      "step": 2388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11496241927988596,
+      "learning_rate": 5.8515155952555284e-05,
+      "loss": 1.5102,
+      "step": 2389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12835498089731012,
+      "learning_rate": 5.8501976863376784e-05,
+      "loss": 1.5104,
+      "step": 2390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12505835808283874,
+      "learning_rate": 5.848879777419828e-05,
+      "loss": 1.5155,
+      "step": 2391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12463706867211648,
+      "learning_rate": 5.847561868501978e-05,
+      "loss": 1.5089,
+      "step": 2392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11959775818340825,
+      "learning_rate": 5.8462439595841266e-05,
+      "loss": 1.5161,
+      "step": 2393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10481289753721347,
+      "learning_rate": 5.8449260506662766e-05,
+      "loss": 1.5138,
+      "step": 2394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10304605714607609,
+      "learning_rate": 5.843608141748426e-05,
+      "loss": 1.5045,
+      "step": 2395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11927925574358386,
+      "learning_rate": 5.842290232830576e-05,
+      "loss": 1.51,
+      "step": 2396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1275850415805112,
+      "learning_rate": 5.8409723239127254e-05,
+      "loss": 1.5092,
+      "step": 2397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1401648820298506,
+      "learning_rate": 5.8396544149948755e-05,
+      "loss": 1.5076,
+      "step": 2398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14484221349086465,
+      "learning_rate": 5.8383365060770255e-05,
+      "loss": 1.5125,
+      "step": 2399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12481174925472197,
+      "learning_rate": 5.837018597159175e-05,
+      "loss": 1.5036,
+      "step": 2400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11378459155159684,
+      "learning_rate": 5.8357006882413236e-05,
+      "loss": 1.513,
+      "step": 2401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11817779556161971,
+      "learning_rate": 5.8343827793234736e-05,
+      "loss": 1.5096,
+      "step": 2402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11981932699231865,
+      "learning_rate": 5.833064870405623e-05,
+      "loss": 1.5104,
+      "step": 2403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1411583140145947,
+      "learning_rate": 5.831746961487773e-05,
+      "loss": 1.5133,
+      "step": 2404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13159232256966769,
+      "learning_rate": 5.830429052569923e-05,
+      "loss": 1.5164,
+      "step": 2405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12295977425119016,
+      "learning_rate": 5.8291111436520725e-05,
+      "loss": 1.5052,
+      "step": 2406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11674656159906896,
+      "learning_rate": 5.8277932347342225e-05,
+      "loss": 1.508,
+      "step": 2407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12008316180768423,
+      "learning_rate": 5.826475325816372e-05,
+      "loss": 1.5084,
+      "step": 2408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11735325558767315,
+      "learning_rate": 5.825157416898521e-05,
+      "loss": 1.5151,
+      "step": 2409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11378939528318825,
+      "learning_rate": 5.823839507980671e-05,
+      "loss": 1.4988,
+      "step": 2410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11997100617151922,
+      "learning_rate": 5.822521599062821e-05,
+      "loss": 1.5019,
+      "step": 2411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11617092606652385,
+      "learning_rate": 5.82120369014497e-05,
+      "loss": 1.5094,
+      "step": 2412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12080799041690947,
+      "learning_rate": 5.81988578122712e-05,
+      "loss": 1.51,
+      "step": 2413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11471316854917507,
+      "learning_rate": 5.81856787230927e-05,
+      "loss": 1.5098,
+      "step": 2414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11346945842745364,
+      "learning_rate": 5.8172499633914196e-05,
+      "loss": 1.5049,
+      "step": 2415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.12073770578593924,
+      "learning_rate": 5.8159320544735696e-05,
+      "loss": 1.5066,
+      "step": 2416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11106492323317937,
+      "learning_rate": 5.814614145555718e-05,
+      "loss": 1.5101,
+      "step": 2417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09836252433227956,
+      "learning_rate": 5.813296236637868e-05,
+      "loss": 1.5128,
+      "step": 2418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09248821060207346,
+      "learning_rate": 5.811978327720018e-05,
+      "loss": 1.5129,
+      "step": 2419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09523718541408982,
+      "learning_rate": 5.810660418802168e-05,
+      "loss": 1.5069,
+      "step": 2420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.09696925867590214,
+      "learning_rate": 5.809342509884317e-05,
+      "loss": 1.5037,
+      "step": 2421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.10283567599721107,
+      "learning_rate": 5.808024600966467e-05,
+      "loss": 1.5082,
+      "step": 2422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.11785020413766834,
+      "learning_rate": 5.8067066920486166e-05,
+      "loss": 1.5035,
+      "step": 2423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13999799881710598,
+      "learning_rate": 5.8053887831307666e-05,
+      "loss": 1.5088,
+      "step": 2424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16154130413391518,
+      "learning_rate": 5.8040708742129153e-05,
+      "loss": 1.5057,
+      "step": 2425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18130568656915494,
+      "learning_rate": 5.8027529652950654e-05,
+      "loss": 1.5025,
+      "step": 2426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1956580439534201,
+      "learning_rate": 5.801435056377215e-05,
+      "loss": 1.496,
+      "step": 2427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16560765815719927,
+      "learning_rate": 5.800117147459365e-05,
+      "loss": 1.5067,
+      "step": 2428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12043801463276158,
+      "learning_rate": 5.798799238541515e-05,
+      "loss": 1.5145,
+      "step": 2429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10867246150111985,
+      "learning_rate": 5.797481329623664e-05,
+      "loss": 1.5093,
+      "step": 2430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15470268924575356,
+      "learning_rate": 5.796163420705814e-05,
+      "loss": 1.5005,
+      "step": 2431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15206877237417046,
+      "learning_rate": 5.794845511787964e-05,
+      "loss": 1.5084,
+      "step": 2432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1310458341309928,
+      "learning_rate": 5.793527602870113e-05,
+      "loss": 1.5014,
+      "step": 2433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12314100234935077,
+      "learning_rate": 5.7922096939522624e-05,
+      "loss": 1.5085,
+      "step": 2434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11452985007064778,
+      "learning_rate": 5.7908917850344125e-05,
+      "loss": 1.5107,
+      "step": 2435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11854208591596396,
+      "learning_rate": 5.789573876116562e-05,
+      "loss": 1.5125,
+      "step": 2436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11965276404539656,
+      "learning_rate": 5.788255967198712e-05,
+      "loss": 1.5064,
+      "step": 2437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12866380957099333,
+      "learning_rate": 5.786938058280862e-05,
+      "loss": 1.5075,
+      "step": 2438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12686309858747769,
+      "learning_rate": 5.785620149363011e-05,
+      "loss": 1.507,
+      "step": 2439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11771028188854153,
+      "learning_rate": 5.7843022404451614e-05,
+      "loss": 1.5021,
+      "step": 2440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1384604099902792,
+      "learning_rate": 5.78298433152731e-05,
+      "loss": 1.5094,
+      "step": 2441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15316745593078296,
+      "learning_rate": 5.7816664226094594e-05,
+      "loss": 1.5069,
+      "step": 2442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15341795983332118,
+      "learning_rate": 5.7803485136916095e-05,
+      "loss": 1.5053,
+      "step": 2443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14732282188209095,
+      "learning_rate": 5.7790306047737596e-05,
+      "loss": 1.5072,
+      "step": 2444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15648459777624665,
+      "learning_rate": 5.777712695855909e-05,
+      "loss": 1.5091,
+      "step": 2445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16909830585277152,
+      "learning_rate": 5.776394786938059e-05,
+      "loss": 1.5093,
+      "step": 2446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15180531223189064,
+      "learning_rate": 5.7750768780202084e-05,
+      "loss": 1.5062,
+      "step": 2447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13546067718873997,
+      "learning_rate": 5.7737589691023584e-05,
+      "loss": 1.5031,
+      "step": 2448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10958066663387739,
+      "learning_rate": 5.772441060184507e-05,
+      "loss": 1.5051,
+      "step": 2449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11804396986433881,
+      "learning_rate": 5.771123151266657e-05,
+      "loss": 1.5039,
+      "step": 2450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13831637087338952,
+      "learning_rate": 5.7698052423488065e-05,
+      "loss": 1.4979,
+      "step": 2451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12560226978140668,
+      "learning_rate": 5.7684873334309566e-05,
+      "loss": 1.5044,
+      "step": 2452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12846162125422758,
+      "learning_rate": 5.7671694245131066e-05,
+      "loss": 1.5079,
+      "step": 2453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12567619306085434,
+      "learning_rate": 5.765851515595256e-05,
+      "loss": 1.514,
+      "step": 2454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10293731595209837,
+      "learning_rate": 5.764533606677406e-05,
+      "loss": 1.5035,
+      "step": 2455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1284706211714589,
+      "learning_rate": 5.7632156977595554e-05,
+      "loss": 1.5151,
+      "step": 2456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1778925274504863,
+      "learning_rate": 5.761897788841705e-05,
+      "loss": 1.4977,
+      "step": 2457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16865835681868824,
+      "learning_rate": 5.760579879923854e-05,
+      "loss": 1.5034,
+      "step": 2458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13522556289276255,
+      "learning_rate": 5.759261971006004e-05,
+      "loss": 1.5067,
+      "step": 2459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1177540358785724,
+      "learning_rate": 5.7579440620881536e-05,
+      "loss": 1.507,
+      "step": 2460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1395411640331523,
+      "learning_rate": 5.7566261531703037e-05,
+      "loss": 1.5083,
+      "step": 2461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15289307191684226,
+      "learning_rate": 5.755308244252454e-05,
+      "loss": 1.5095,
+      "step": 2462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17350249748708887,
+      "learning_rate": 5.753990335334603e-05,
+      "loss": 1.4999,
+      "step": 2463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17403081700873313,
+      "learning_rate": 5.752672426416753e-05,
+      "loss": 1.5034,
+      "step": 2464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13193529752449212,
+      "learning_rate": 5.751354517498902e-05,
+      "loss": 1.5123,
+      "step": 2465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11741191232290907,
+      "learning_rate": 5.750036608581051e-05,
+      "loss": 1.5036,
+      "step": 2466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14065329256724485,
+      "learning_rate": 5.748718699663201e-05,
+      "loss": 1.5001,
+      "step": 2467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1455895606043078,
+      "learning_rate": 5.747400790745351e-05,
+      "loss": 1.5041,
+      "step": 2468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1256410302759854,
+      "learning_rate": 5.746082881827501e-05,
+      "loss": 1.5071,
+      "step": 2469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13768896879598247,
+      "learning_rate": 5.744764972909651e-05,
+      "loss": 1.5058,
+      "step": 2470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14706903287105474,
+      "learning_rate": 5.7434470639918e-05,
+      "loss": 1.5133,
+      "step": 2471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10480490206095323,
+      "learning_rate": 5.74212915507395e-05,
+      "loss": 1.5079,
+      "step": 2472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12319788658440185,
+      "learning_rate": 5.740811246156099e-05,
+      "loss": 1.5166,
+      "step": 2473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13850494669823585,
+      "learning_rate": 5.739493337238249e-05,
+      "loss": 1.5085,
+      "step": 2474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10682940617969337,
+      "learning_rate": 5.738175428320398e-05,
+      "loss": 1.5036,
+      "step": 2475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12310121479794217,
+      "learning_rate": 5.736857519402548e-05,
+      "loss": 1.5052,
+      "step": 2476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12258791950187499,
+      "learning_rate": 5.7355396104846984e-05,
+      "loss": 1.5084,
+      "step": 2477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12073941092998565,
+      "learning_rate": 5.734221701566848e-05,
+      "loss": 1.5048,
+      "step": 2478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14674887514107002,
+      "learning_rate": 5.732903792648998e-05,
+      "loss": 1.5011,
+      "step": 2479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14977158092103976,
+      "learning_rate": 5.731585883731147e-05,
+      "loss": 1.5063,
+      "step": 2480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.12867231796829376,
+      "learning_rate": 5.7302679748132966e-05,
+      "loss": 1.5031,
+      "step": 2481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1278790311928587,
+      "learning_rate": 5.728950065895446e-05,
+      "loss": 1.5025,
+      "step": 2482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1080974292636994,
+      "learning_rate": 5.727632156977596e-05,
+      "loss": 1.5013,
+      "step": 2483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.10853187302457351,
+      "learning_rate": 5.7263142480597454e-05,
+      "loss": 1.5083,
+      "step": 2484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1361375456638516,
+      "learning_rate": 5.7249963391418954e-05,
+      "loss": 1.498,
+      "step": 2485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13032011114142084,
+      "learning_rate": 5.7236784302240455e-05,
+      "loss": 1.5058,
+      "step": 2486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11966347684153905,
+      "learning_rate": 5.722360521306195e-05,
+      "loss": 1.5078,
+      "step": 2487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13379664481727122,
+      "learning_rate": 5.7210426123883435e-05,
+      "loss": 1.4965,
+      "step": 2488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.11980451107298819,
+      "learning_rate": 5.7197247034704936e-05,
+      "loss": 1.5031,
+      "step": 2489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1190297715998218,
+      "learning_rate": 5.718406794552643e-05,
+      "loss": 1.503,
+      "step": 2490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.13253829662204428,
+      "learning_rate": 5.717088885634793e-05,
+      "loss": 1.5091,
+      "step": 2491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1249260827487292,
+      "learning_rate": 5.715770976716943e-05,
+      "loss": 1.5073,
+      "step": 2492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12709344813651424,
+      "learning_rate": 5.7144530677990924e-05,
+      "loss": 1.5103,
+      "step": 2493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12654223602255424,
+      "learning_rate": 5.7131351588812425e-05,
+      "loss": 1.5037,
+      "step": 2494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13189403618703285,
+      "learning_rate": 5.711817249963392e-05,
+      "loss": 1.5104,
+      "step": 2495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1139602098938948,
+      "learning_rate": 5.710499341045541e-05,
+      "loss": 1.5049,
+      "step": 2496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1178111147281891,
+      "learning_rate": 5.7091814321276906e-05,
+      "loss": 1.5001,
+      "step": 2497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13082771454566283,
+      "learning_rate": 5.707863523209841e-05,
+      "loss": 1.5044,
+      "step": 2498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14883173163426452,
+      "learning_rate": 5.70654561429199e-05,
+      "loss": 1.4972,
+      "step": 2499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14737192597637153,
+      "learning_rate": 5.70522770537414e-05,
+      "loss": 1.5056,
+      "step": 2500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13550363556130138,
+      "learning_rate": 5.70390979645629e-05,
+      "loss": 1.4956,
+      "step": 2501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11202242221999903,
+      "learning_rate": 5.7025918875384395e-05,
+      "loss": 1.5168,
+      "step": 2502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10377582185525412,
+      "learning_rate": 5.7012739786205896e-05,
+      "loss": 1.505,
+      "step": 2503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13214612851123936,
+      "learning_rate": 5.699956069702738e-05,
+      "loss": 1.5068,
+      "step": 2504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12295854838396045,
+      "learning_rate": 5.6986381607848876e-05,
+      "loss": 1.5136,
+      "step": 2505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10552791337965739,
+      "learning_rate": 5.697320251867038e-05,
+      "loss": 1.5,
+      "step": 2506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11544630537569452,
+      "learning_rate": 5.696002342949188e-05,
+      "loss": 1.5069,
+      "step": 2507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14245634922277822,
+      "learning_rate": 5.694684434031337e-05,
+      "loss": 1.5083,
+      "step": 2508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11484101930274358,
+      "learning_rate": 5.693366525113487e-05,
+      "loss": 1.5115,
+      "step": 2509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13028964913035684,
+      "learning_rate": 5.6920486161956365e-05,
+      "loss": 1.5,
+      "step": 2510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14343815333502105,
+      "learning_rate": 5.6907307072777866e-05,
+      "loss": 1.5127,
+      "step": 2511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13654196351995743,
+      "learning_rate": 5.689412798359935e-05,
+      "loss": 1.4991,
+      "step": 2512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1264564005450845,
+      "learning_rate": 5.6880948894420853e-05,
+      "loss": 1.497,
+      "step": 2513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12174638357504436,
+      "learning_rate": 5.686776980524235e-05,
+      "loss": 1.5009,
+      "step": 2514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13183257373248772,
+      "learning_rate": 5.685459071606385e-05,
+      "loss": 1.5076,
+      "step": 2515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14572214628246324,
+      "learning_rate": 5.684141162688535e-05,
+      "loss": 1.5068,
+      "step": 2516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13503439901512942,
+      "learning_rate": 5.682823253770684e-05,
+      "loss": 1.4959,
+      "step": 2517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11961992533608186,
+      "learning_rate": 5.681505344852834e-05,
+      "loss": 1.5095,
+      "step": 2518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12728869058651265,
+      "learning_rate": 5.6801874359349836e-05,
+      "loss": 1.5047,
+      "step": 2519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1292137879448682,
+      "learning_rate": 5.678869527017133e-05,
+      "loss": 1.5061,
+      "step": 2520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12587240250927748,
+      "learning_rate": 5.6775516180992824e-05,
+      "loss": 1.5116,
+      "step": 2521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11836560229652372,
+      "learning_rate": 5.6762337091814324e-05,
+      "loss": 1.5022,
+      "step": 2522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11103196939991006,
+      "learning_rate": 5.674915800263582e-05,
+      "loss": 1.5061,
+      "step": 2523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13118218194811898,
+      "learning_rate": 5.673597891345732e-05,
+      "loss": 1.5034,
+      "step": 2524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13747412282257052,
+      "learning_rate": 5.672279982427882e-05,
+      "loss": 1.5048,
+      "step": 2525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11883087531527477,
+      "learning_rate": 5.670962073510031e-05,
+      "loss": 1.4998,
+      "step": 2526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12222090851518656,
+      "learning_rate": 5.669644164592181e-05,
+      "loss": 1.4948,
+      "step": 2527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12021419816001305,
+      "learning_rate": 5.66832625567433e-05,
+      "loss": 1.4971,
+      "step": 2528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1144387240991996,
+      "learning_rate": 5.6670083467564794e-05,
+      "loss": 1.5102,
+      "step": 2529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1071609638383454,
+      "learning_rate": 5.6656904378386294e-05,
+      "loss": 1.5003,
+      "step": 2530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11759516346019797,
+      "learning_rate": 5.6643725289207795e-05,
+      "loss": 1.5065,
+      "step": 2531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11916407032029057,
+      "learning_rate": 5.663054620002929e-05,
+      "loss": 1.5118,
+      "step": 2532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12200129665945093,
+      "learning_rate": 5.661736711085079e-05,
+      "loss": 1.502,
+      "step": 2533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13758281214818374,
+      "learning_rate": 5.660418802167228e-05,
+      "loss": 1.5051,
+      "step": 2534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14276149696355553,
+      "learning_rate": 5.6591008932493784e-05,
+      "loss": 1.5007,
+      "step": 2535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15563411698732094,
+      "learning_rate": 5.657782984331527e-05,
+      "loss": 1.5109,
+      "step": 2536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1507817091746234,
+      "learning_rate": 5.656465075413677e-05,
+      "loss": 1.5001,
+      "step": 2537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13810927636722609,
+      "learning_rate": 5.6551471664958265e-05,
+      "loss": 1.5083,
+      "step": 2538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10805633369998784,
+      "learning_rate": 5.6538292575779765e-05,
+      "loss": 1.5011,
+      "step": 2539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1113973196995999,
+      "learning_rate": 5.6525113486601266e-05,
+      "loss": 1.4947,
+      "step": 2540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11128764103208831,
+      "learning_rate": 5.651193439742276e-05,
+      "loss": 1.5102,
+      "step": 2541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13281956202291673,
+      "learning_rate": 5.649875530824426e-05,
+      "loss": 1.5002,
+      "step": 2542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13328435433238192,
+      "learning_rate": 5.6485576219065754e-05,
+      "loss": 1.5122,
+      "step": 2543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13769853944364763,
+      "learning_rate": 5.647239712988725e-05,
+      "loss": 1.4955,
+      "step": 2544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13306643470227297,
+      "learning_rate": 5.645921804070874e-05,
+      "loss": 1.4969,
+      "step": 2545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11517387028223866,
+      "learning_rate": 5.644603895153024e-05,
+      "loss": 1.505,
+      "step": 2546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09816294734529074,
+      "learning_rate": 5.6432859862351736e-05,
+      "loss": 1.5031,
+      "step": 2547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.08836111713510432,
+      "learning_rate": 5.6419680773173236e-05,
+      "loss": 1.5069,
+      "step": 2548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10003602808407779,
+      "learning_rate": 5.6406501683994737e-05,
+      "loss": 1.5059,
+      "step": 2549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10548629143030344,
+      "learning_rate": 5.639332259481623e-05,
+      "loss": 1.5082,
+      "step": 2550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09213606699499079,
+      "learning_rate": 5.638014350563773e-05,
+      "loss": 1.51,
+      "step": 2551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.10827698747487509,
+      "learning_rate": 5.636696441645922e-05,
+      "loss": 1.502,
+      "step": 2552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.13891534047253862,
+      "learning_rate": 5.635378532728071e-05,
+      "loss": 1.504,
+      "step": 2553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15567491381989368,
+      "learning_rate": 5.634060623810221e-05,
+      "loss": 1.5109,
+      "step": 2554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15670366492958698,
+      "learning_rate": 5.632742714892371e-05,
+      "loss": 1.5004,
+      "step": 2555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15191271379528848,
+      "learning_rate": 5.6314248059745206e-05,
+      "loss": 1.5084,
+      "step": 2556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15684889577871763,
+      "learning_rate": 5.630106897056671e-05,
+      "loss": 1.4985,
+      "step": 2557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15409726512697872,
+      "learning_rate": 5.62878898813882e-05,
+      "loss": 1.4993,
+      "step": 2558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14614945272437785,
+      "learning_rate": 5.62747107922097e-05,
+      "loss": 1.5136,
+      "step": 2559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.11574710376980162,
+      "learning_rate": 5.626153170303119e-05,
+      "loss": 1.509,
+      "step": 2560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.09346181342472516,
+      "learning_rate": 5.624835261385269e-05,
+      "loss": 1.4992,
+      "step": 2561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11611228760291178,
+      "learning_rate": 5.623517352467418e-05,
+      "loss": 1.5004,
+      "step": 2562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11813538263985789,
+      "learning_rate": 5.622199443549568e-05,
+      "loss": 1.4992,
+      "step": 2563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1077153754368645,
+      "learning_rate": 5.620881534631718e-05,
+      "loss": 1.4991,
+      "step": 2564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10100195165188165,
+      "learning_rate": 5.619563625713868e-05,
+      "loss": 1.5058,
+      "step": 2565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13035254225218515,
+      "learning_rate": 5.618245716796018e-05,
+      "loss": 1.4996,
+      "step": 2566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14645268468285905,
+      "learning_rate": 5.616927807878167e-05,
+      "loss": 1.5074,
+      "step": 2567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15496188510702244,
+      "learning_rate": 5.615609898960316e-05,
+      "loss": 1.4985,
+      "step": 2568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13391395782093735,
+      "learning_rate": 5.614291990042466e-05,
+      "loss": 1.5051,
+      "step": 2569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1393555622899728,
+      "learning_rate": 5.612974081124616e-05,
+      "loss": 1.5012,
+      "step": 2570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1669860319328833,
+      "learning_rate": 5.611656172206765e-05,
+      "loss": 1.4984,
+      "step": 2571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18125701587227783,
+      "learning_rate": 5.6103382632889154e-05,
+      "loss": 1.4938,
+      "step": 2572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1282189402907676,
+      "learning_rate": 5.609020354371065e-05,
+      "loss": 1.5018,
+      "step": 2573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10743804874775192,
+      "learning_rate": 5.607702445453215e-05,
+      "loss": 1.5091,
+      "step": 2574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13323467395786334,
+      "learning_rate": 5.606384536535365e-05,
+      "loss": 1.5122,
+      "step": 2575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1346843095510897,
+      "learning_rate": 5.6050666276175135e-05,
+      "loss": 1.5019,
+      "step": 2576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11630831926481625,
+      "learning_rate": 5.603748718699663e-05,
+      "loss": 1.5031,
+      "step": 2577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12983128713033346,
+      "learning_rate": 5.602430809781813e-05,
+      "loss": 1.5081,
+      "step": 2578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13740801339290296,
+      "learning_rate": 5.601112900863963e-05,
+      "loss": 1.5041,
+      "step": 2579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12372202570139572,
+      "learning_rate": 5.5997949919461124e-05,
+      "loss": 1.4892,
+      "step": 2580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11843783466294949,
+      "learning_rate": 5.5984770830282624e-05,
+      "loss": 1.5062,
+      "step": 2581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1345076650706001,
+      "learning_rate": 5.597159174110412e-05,
+      "loss": 1.5019,
+      "step": 2582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14683574635100544,
+      "learning_rate": 5.595841265192562e-05,
+      "loss": 1.4917,
+      "step": 2583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14731540321974523,
+      "learning_rate": 5.5945233562747106e-05,
+      "loss": 1.4996,
+      "step": 2584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13471806588271443,
+      "learning_rate": 5.5932054473568606e-05,
+      "loss": 1.5096,
+      "step": 2585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15587079515639365,
+      "learning_rate": 5.59188753843901e-05,
+      "loss": 1.5082,
+      "step": 2586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14092922425692261,
+      "learning_rate": 5.59056962952116e-05,
+      "loss": 1.4938,
+      "step": 2587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13631810620680743,
+      "learning_rate": 5.58925172060331e-05,
+      "loss": 1.502,
+      "step": 2588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12346609249894205,
+      "learning_rate": 5.5879338116854595e-05,
+      "loss": 1.5022,
+      "step": 2589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11473171543530251,
+      "learning_rate": 5.5866159027676095e-05,
+      "loss": 1.4965,
+      "step": 2590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13197678965689746,
+      "learning_rate": 5.585297993849759e-05,
+      "loss": 1.5075,
+      "step": 2591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12313907879075785,
+      "learning_rate": 5.5839800849319076e-05,
+      "loss": 1.5041,
+      "step": 2592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09977512436770519,
+      "learning_rate": 5.5826621760140576e-05,
+      "loss": 1.4992,
+      "step": 2593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10589650504943571,
+      "learning_rate": 5.581344267096208e-05,
+      "loss": 1.5014,
+      "step": 2594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11691687646119499,
+      "learning_rate": 5.580026358178357e-05,
+      "loss": 1.5097,
+      "step": 2595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10046283739189929,
+      "learning_rate": 5.578708449260507e-05,
+      "loss": 1.4986,
+      "step": 2596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11485025611895172,
+      "learning_rate": 5.5773905403426565e-05,
+      "loss": 1.5045,
+      "step": 2597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12763373999565253,
+      "learning_rate": 5.5760726314248065e-05,
+      "loss": 1.5056,
+      "step": 2598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13414824547044454,
+      "learning_rate": 5.5747547225069566e-05,
+      "loss": 1.505,
+      "step": 2599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12252722326329189,
+      "learning_rate": 5.573436813589105e-05,
+      "loss": 1.5017,
+      "step": 2600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11986181749087547,
+      "learning_rate": 5.572118904671255e-05,
+      "loss": 1.4999,
+      "step": 2601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12389088935794469,
+      "learning_rate": 5.570800995753405e-05,
+      "loss": 1.5098,
+      "step": 2602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.14306639042095648,
+      "learning_rate": 5.569483086835555e-05,
+      "loss": 1.5018,
+      "step": 2603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12519044635560983,
+      "learning_rate": 5.568165177917704e-05,
+      "loss": 1.4978,
+      "step": 2604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11068944957759835,
+      "learning_rate": 5.566847268999854e-05,
+      "loss": 1.4926,
+      "step": 2605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11683023564463355,
+      "learning_rate": 5.5655293600820036e-05,
+      "loss": 1.4979,
+      "step": 2606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13214027337795956,
+      "learning_rate": 5.5642114511641536e-05,
+      "loss": 1.5093,
+      "step": 2607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1305014926295693,
+      "learning_rate": 5.562893542246302e-05,
+      "loss": 1.4991,
+      "step": 2608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1267322226375837,
+      "learning_rate": 5.5615756333284524e-05,
+      "loss": 1.4981,
+      "step": 2609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12036458455064852,
+      "learning_rate": 5.560257724410602e-05,
+      "loss": 1.4946,
+      "step": 2610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09916709571454511,
+      "learning_rate": 5.558939815492752e-05,
+      "loss": 1.504,
+      "step": 2611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.09497818113507324,
+      "learning_rate": 5.557621906574902e-05,
+      "loss": 1.5007,
+      "step": 2612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11613628389416253,
+      "learning_rate": 5.556303997657051e-05,
+      "loss": 1.4963,
+      "step": 2613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1213045537368782,
+      "learning_rate": 5.554986088739201e-05,
+      "loss": 1.5003,
+      "step": 2614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12271703067104583,
+      "learning_rate": 5.5536681798213506e-05,
+      "loss": 1.4988,
+      "step": 2615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1194462428663036,
+      "learning_rate": 5.5523502709034993e-05,
+      "loss": 1.495,
+      "step": 2616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12375857875943118,
+      "learning_rate": 5.5510323619856494e-05,
+      "loss": 1.4994,
+      "step": 2617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11911586076399254,
+      "learning_rate": 5.5497144530677994e-05,
+      "loss": 1.493,
+      "step": 2618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12364111769201645,
+      "learning_rate": 5.548396544149949e-05,
+      "loss": 1.4964,
+      "step": 2619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1154731813331778,
+      "learning_rate": 5.547078635232099e-05,
+      "loss": 1.5107,
+      "step": 2620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.13218078778859196,
+      "learning_rate": 5.545760726314248e-05,
+      "loss": 1.503,
+      "step": 2621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.12865241016517373,
+      "learning_rate": 5.544442817396398e-05,
+      "loss": 1.5052,
+      "step": 2622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11682540751117311,
+      "learning_rate": 5.5431249084785484e-05,
+      "loss": 1.5033,
+      "step": 2623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11249860906586356,
+      "learning_rate": 5.541806999560697e-05,
+      "loss": 1.5027,
+      "step": 2624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11397548783047862,
+      "learning_rate": 5.5404890906428464e-05,
+      "loss": 1.5012,
+      "step": 2625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11948329424996178,
+      "learning_rate": 5.5391711817249965e-05,
+      "loss": 1.4992,
+      "step": 2626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.11063822427792022,
+      "learning_rate": 5.5378532728071465e-05,
+      "loss": 1.5054,
+      "step": 2627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10550876054602105,
+      "learning_rate": 5.536535363889296e-05,
+      "loss": 1.4956,
+      "step": 2628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10221273396187332,
+      "learning_rate": 5.535217454971446e-05,
+      "loss": 1.4917,
+      "step": 2629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10414189579616352,
+      "learning_rate": 5.533899546053595e-05,
+      "loss": 1.4998,
+      "step": 2630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1220441622806211,
+      "learning_rate": 5.5325816371357454e-05,
+      "loss": 1.4902,
+      "step": 2631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14291373301126317,
+      "learning_rate": 5.531263728217894e-05,
+      "loss": 1.5009,
+      "step": 2632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1390524282662258,
+      "learning_rate": 5.529945819300044e-05,
+      "loss": 1.5049,
+      "step": 2633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14963540305639234,
+      "learning_rate": 5.5286279103821935e-05,
+      "loss": 1.5009,
+      "step": 2634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16656734847563046,
+      "learning_rate": 5.5273100014643436e-05,
+      "loss": 1.4976,
+      "step": 2635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14861532731357027,
+      "learning_rate": 5.5259920925464936e-05,
+      "loss": 1.5005,
+      "step": 2636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11228583550937557,
+      "learning_rate": 5.524674183628643e-05,
+      "loss": 1.4928,
+      "step": 2637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12415872288773738,
+      "learning_rate": 5.523356274710793e-05,
+      "loss": 1.4992,
+      "step": 2638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15479119495425844,
+      "learning_rate": 5.5220383657929424e-05,
+      "loss": 1.4944,
+      "step": 2639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16938585400772194,
+      "learning_rate": 5.520720456875091e-05,
+      "loss": 1.5,
+      "step": 2640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14095081853847982,
+      "learning_rate": 5.519402547957241e-05,
+      "loss": 1.4934,
+      "step": 2641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10497175966144232,
+      "learning_rate": 5.518084639039391e-05,
+      "loss": 1.5039,
+      "step": 2642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09629851684546785,
+      "learning_rate": 5.5167667301215406e-05,
+      "loss": 1.494,
+      "step": 2643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11594524644418633,
+      "learning_rate": 5.5154488212036906e-05,
+      "loss": 1.4972,
+      "step": 2644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12195978416625697,
+      "learning_rate": 5.51413091228584e-05,
+      "loss": 1.5014,
+      "step": 2645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11361189837887926,
+      "learning_rate": 5.51281300336799e-05,
+      "loss": 1.4949,
+      "step": 2646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11573868273017067,
+      "learning_rate": 5.51149509445014e-05,
+      "loss": 1.5014,
+      "step": 2647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12187490798303037,
+      "learning_rate": 5.510177185532289e-05,
+      "loss": 1.4973,
+      "step": 2648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.147050026002942,
+      "learning_rate": 5.508859276614438e-05,
+      "loss": 1.5069,
+      "step": 2649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15165977768208538,
+      "learning_rate": 5.507541367696588e-05,
+      "loss": 1.5053,
+      "step": 2650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12494853503516767,
+      "learning_rate": 5.506223458778738e-05,
+      "loss": 1.4943,
+      "step": 2651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11689566265855862,
+      "learning_rate": 5.5049055498608877e-05,
+      "loss": 1.5001,
+      "step": 2652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.09947987642144436,
+      "learning_rate": 5.503587640943038e-05,
+      "loss": 1.5022,
+      "step": 2653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10856460896573135,
+      "learning_rate": 5.502269732025187e-05,
+      "loss": 1.4962,
+      "step": 2654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10045959209970132,
+      "learning_rate": 5.500951823107337e-05,
+      "loss": 1.5043,
+      "step": 2655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1037189538165912,
+      "learning_rate": 5.499633914189486e-05,
+      "loss": 1.4918,
+      "step": 2656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10798503172381145,
+      "learning_rate": 5.498316005271636e-05,
+      "loss": 1.5096,
+      "step": 2657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11179036592113262,
+      "learning_rate": 5.496998096353785e-05,
+      "loss": 1.4981,
+      "step": 2658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10847526188804246,
+      "learning_rate": 5.495680187435935e-05,
+      "loss": 1.5093,
+      "step": 2659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11233554256165626,
+      "learning_rate": 5.494362278518085e-05,
+      "loss": 1.4982,
+      "step": 2660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12740459037475713,
+      "learning_rate": 5.493044369600235e-05,
+      "loss": 1.5062,
+      "step": 2661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1249548287101282,
+      "learning_rate": 5.491726460682385e-05,
+      "loss": 1.4983,
+      "step": 2662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1174804098847243,
+      "learning_rate": 5.490408551764534e-05,
+      "loss": 1.4939,
+      "step": 2663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12524255057298506,
+      "learning_rate": 5.489090642846683e-05,
+      "loss": 1.5068,
+      "step": 2664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1316953358273972,
+      "learning_rate": 5.487772733928833e-05,
+      "loss": 1.4939,
+      "step": 2665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13162786220260977,
+      "learning_rate": 5.486454825010983e-05,
+      "loss": 1.5,
+      "step": 2666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1479393897801665,
+      "learning_rate": 5.485136916093132e-05,
+      "loss": 1.4993,
+      "step": 2667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15487542308436208,
+      "learning_rate": 5.4838190071752824e-05,
+      "loss": 1.5023,
+      "step": 2668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13834275482773858,
+      "learning_rate": 5.482501098257432e-05,
+      "loss": 1.4963,
+      "step": 2669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13697863159561946,
+      "learning_rate": 5.481183189339582e-05,
+      "loss": 1.4962,
+      "step": 2670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13000910593151058,
+      "learning_rate": 5.479865280421732e-05,
+      "loss": 1.4932,
+      "step": 2671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1130563924012113,
+      "learning_rate": 5.4785473715038806e-05,
+      "loss": 1.4902,
+      "step": 2672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11044581390286333,
+      "learning_rate": 5.47722946258603e-05,
+      "loss": 1.4966,
+      "step": 2673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1043677826111573,
+      "learning_rate": 5.47591155366818e-05,
+      "loss": 1.5044,
+      "step": 2674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10802404564665045,
+      "learning_rate": 5.47459364475033e-05,
+      "loss": 1.4959,
+      "step": 2675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1141886435527675,
+      "learning_rate": 5.4732757358324794e-05,
+      "loss": 1.5047,
+      "step": 2676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10256986032590047,
+      "learning_rate": 5.4719578269146295e-05,
+      "loss": 1.4966,
+      "step": 2677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11304002675981925,
+      "learning_rate": 5.470639917996779e-05,
+      "loss": 1.4986,
+      "step": 2678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10107169780696014,
+      "learning_rate": 5.469322009078929e-05,
+      "loss": 1.4956,
+      "step": 2679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12086820901148812,
+      "learning_rate": 5.4680041001610776e-05,
+      "loss": 1.4892,
+      "step": 2680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1236630676977422,
+      "learning_rate": 5.4666861912432276e-05,
+      "loss": 1.5015,
+      "step": 2681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10559484654020818,
+      "learning_rate": 5.465368282325377e-05,
+      "loss": 1.4911,
+      "step": 2682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10775688497589157,
+      "learning_rate": 5.464050373407527e-05,
+      "loss": 1.4971,
+      "step": 2683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1105673404446589,
+      "learning_rate": 5.4627324644896764e-05,
+      "loss": 1.4916,
+      "step": 2684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.105043368076363,
+      "learning_rate": 5.4614145555718265e-05,
+      "loss": 1.4926,
+      "step": 2685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11434234966232248,
+      "learning_rate": 5.4600966466539765e-05,
+      "loss": 1.4856,
+      "step": 2686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.11086061460774911,
+      "learning_rate": 5.458778737736126e-05,
+      "loss": 1.4971,
+      "step": 2687
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13086285922510582,
+      "learning_rate": 5.4574608288182746e-05,
+      "loss": 1.4952,
+      "step": 2688
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15417524871681482,
+      "learning_rate": 5.456142919900425e-05,
+      "loss": 1.4923,
+      "step": 2689
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16790183937916128,
+      "learning_rate": 5.454825010982575e-05,
+      "loss": 1.4925,
+      "step": 2690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1752363618982495,
+      "learning_rate": 5.453507102064724e-05,
+      "loss": 1.5065,
+      "step": 2691
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.14916103514941983,
+      "learning_rate": 5.452189193146874e-05,
+      "loss": 1.497,
+      "step": 2692
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10463169788759112,
+      "learning_rate": 5.4508712842290235e-05,
+      "loss": 1.4922,
+      "step": 2693
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.12563724741395643,
+      "learning_rate": 5.4495533753111736e-05,
+      "loss": 1.4998,
+      "step": 2694
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13639123665421907,
+      "learning_rate": 5.4482354663933236e-05,
+      "loss": 1.4971,
+      "step": 2695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.13502044308709013,
+      "learning_rate": 5.446917557475472e-05,
+      "loss": 1.4988,
+      "step": 2696
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1575934104484493,
+      "learning_rate": 5.445599648557622e-05,
+      "loss": 1.4956,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14727120143151637,
+      "learning_rate": 5.444281739639772e-05,
+      "loss": 1.4987,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12756276904091318,
+      "learning_rate": 5.442963830721922e-05,
+      "loss": 1.4925,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12041115696957813,
+      "learning_rate": 5.441645921804071e-05,
+      "loss": 1.4937,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.462758183479309,
+      "eval_runtime": 465.2813,
+      "eval_samples_per_second": 38.719,
+      "eval_steps_per_second": 0.058,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12020950762480835,
+      "learning_rate": 5.440328012886221e-05,
+      "loss": 1.4946,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12254378276106732,
+      "learning_rate": 5.4390101039683706e-05,
+      "loss": 1.4987,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11428970562789371,
+      "learning_rate": 5.4376921950505206e-05,
+      "loss": 1.4955,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11174316198969847,
+      "learning_rate": 5.4363742861326693e-05,
+      "loss": 1.497,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12640444053224717,
+      "learning_rate": 5.4350563772148194e-05,
+      "loss": 1.4976,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11593515677701371,
+      "learning_rate": 5.433738468296969e-05,
+      "loss": 1.501,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11062625491134905,
+      "learning_rate": 5.432420559379119e-05,
+      "loss": 1.504,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13029018402280657,
+      "learning_rate": 5.431102650461268e-05,
+      "loss": 1.496,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13760013048371714,
+      "learning_rate": 5.429784741543418e-05,
+      "loss": 1.4904,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13090014237423164,
+      "learning_rate": 5.428466832625568e-05,
+      "loss": 1.4837,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11944935981024217,
+      "learning_rate": 5.427148923707718e-05,
+      "loss": 1.4916,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13047391437801112,
+      "learning_rate": 5.4258310147898664e-05,
+      "loss": 1.4936,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1532598160282253,
+      "learning_rate": 5.4245131058720164e-05,
+      "loss": 1.4926,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14980318213028304,
+      "learning_rate": 5.4231951969541665e-05,
+      "loss": 1.5009,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12075883551687398,
+      "learning_rate": 5.421877288036316e-05,
+      "loss": 1.5037,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13645455580342913,
+      "learning_rate": 5.420559379118466e-05,
+      "loss": 1.5007,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15181887872766514,
+      "learning_rate": 5.419241470200615e-05,
+      "loss": 1.5013,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13269069323637778,
+      "learning_rate": 5.417923561282765e-05,
+      "loss": 1.4937,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1152613998408127,
+      "learning_rate": 5.4166056523649154e-05,
+      "loss": 1.4943,
+      "step": 2719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13598829241540927,
+      "learning_rate": 5.415287743447064e-05,
+      "loss": 1.4982,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.14295029910028717,
+      "learning_rate": 5.4139698345292135e-05,
+      "loss": 1.5098,
+      "step": 2721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12740368205381014,
+      "learning_rate": 5.4126519256113635e-05,
+      "loss": 1.4991,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10990320146867595,
+      "learning_rate": 5.4113340166935136e-05,
+      "loss": 1.4927,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11872007626826167,
+      "learning_rate": 5.410016107775663e-05,
+      "loss": 1.4965,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13086661068489666,
+      "learning_rate": 5.408698198857813e-05,
+      "loss": 1.4936,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12895168091348805,
+      "learning_rate": 5.4073802899399624e-05,
+      "loss": 1.4985,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12007572992309656,
+      "learning_rate": 5.4060623810221124e-05,
+      "loss": 1.4851,
+      "step": 2727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10762835781886866,
+      "learning_rate": 5.404744472104261e-05,
+      "loss": 1.4994,
+      "step": 2728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11545125667048321,
+      "learning_rate": 5.403426563186411e-05,
+      "loss": 1.4981,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13426174456839196,
+      "learning_rate": 5.4021086542685605e-05,
+      "loss": 1.4972,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13194306218688562,
+      "learning_rate": 5.4007907453507106e-05,
+      "loss": 1.5014,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11017430693757575,
+      "learning_rate": 5.39947283643286e-05,
+      "loss": 1.5028,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11820320150595556,
+      "learning_rate": 5.39815492751501e-05,
+      "loss": 1.4923,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1226448866967627,
+      "learning_rate": 5.39683701859716e-05,
+      "loss": 1.4948,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12388214949324229,
+      "learning_rate": 5.3955191096793094e-05,
+      "loss": 1.4847,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11479588660445643,
+      "learning_rate": 5.394201200761458e-05,
+      "loss": 1.4964,
+      "step": 2736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11445826007963684,
+      "learning_rate": 5.392883291843608e-05,
+      "loss": 1.497,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.106442710036967,
+      "learning_rate": 5.391565382925758e-05,
+      "loss": 1.4974,
+      "step": 2738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12603679411474158,
+      "learning_rate": 5.3902474740079076e-05,
+      "loss": 1.4973,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12085915108553473,
+      "learning_rate": 5.3889295650900577e-05,
+      "loss": 1.4872,
+      "step": 2740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10875403068490205,
+      "learning_rate": 5.387611656172207e-05,
+      "loss": 1.499,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09988789490437616,
+      "learning_rate": 5.386293747254357e-05,
+      "loss": 1.4953,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11800916436345775,
+      "learning_rate": 5.384975838336507e-05,
+      "loss": 1.4958,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1327836821905416,
+      "learning_rate": 5.383657929418656e-05,
+      "loss": 1.4928,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12492756436125103,
+      "learning_rate": 5.382340020500805e-05,
+      "loss": 1.4927,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11440389708858327,
+      "learning_rate": 5.381022111582955e-05,
+      "loss": 1.4937,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10708241622657064,
+      "learning_rate": 5.3797042026651046e-05,
+      "loss": 1.4899,
+      "step": 2747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11210390000206961,
+      "learning_rate": 5.378386293747255e-05,
+      "loss": 1.5031,
+      "step": 2748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11787802298368286,
+      "learning_rate": 5.377068384829405e-05,
+      "loss": 1.4951,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10859822508099805,
+      "learning_rate": 5.375750475911554e-05,
+      "loss": 1.4953,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10234817779955817,
+      "learning_rate": 5.374432566993704e-05,
+      "loss": 1.5039,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1063917258191718,
+      "learning_rate": 5.373114658075853e-05,
+      "loss": 1.4979,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09985009896110485,
+      "learning_rate": 5.371796749158003e-05,
+      "loss": 1.4964,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11759021809302192,
+      "learning_rate": 5.370478840240152e-05,
+      "loss": 1.4983,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13473355925851077,
+      "learning_rate": 5.369160931322302e-05,
+      "loss": 1.4946,
+      "step": 2755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12050949917213014,
+      "learning_rate": 5.367843022404452e-05,
+      "loss": 1.5007,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1093849450997433,
+      "learning_rate": 5.366525113486602e-05,
+      "loss": 1.5039,
+      "step": 2757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11631927948767745,
+      "learning_rate": 5.365207204568752e-05,
+      "loss": 1.4954,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.13400192664514515,
+      "learning_rate": 5.363889295650901e-05,
+      "loss": 1.4944,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11628817067389305,
+      "learning_rate": 5.36257138673305e-05,
+      "loss": 1.487,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.115213123601247,
+      "learning_rate": 5.3612534778152e-05,
+      "loss": 1.4947,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.10962378124411792,
+      "learning_rate": 5.35993556889735e-05,
+      "loss": 1.4936,
+      "step": 2762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12007404675610911,
+      "learning_rate": 5.3586176599794994e-05,
+      "loss": 1.5018,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11952795870574624,
+      "learning_rate": 5.3572997510616494e-05,
+      "loss": 1.4922,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12577812841780403,
+      "learning_rate": 5.355981842143799e-05,
+      "loss": 1.498,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12470054075387367,
+      "learning_rate": 5.354663933225949e-05,
+      "loss": 1.4948,
+      "step": 2766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11226311571344859,
+      "learning_rate": 5.353346024308099e-05,
+      "loss": 1.4894,
+      "step": 2767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09932442535942661,
+      "learning_rate": 5.3520281153902476e-05,
+      "loss": 1.4947,
+      "step": 2768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1026447485399691,
+      "learning_rate": 5.350710206472397e-05,
+      "loss": 1.5007,
+      "step": 2769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1020612701802325,
+      "learning_rate": 5.349392297554547e-05,
+      "loss": 1.5006,
+      "step": 2770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1091300750985694,
+      "learning_rate": 5.3480743886366964e-05,
+      "loss": 1.497,
+      "step": 2771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12460885902702294,
+      "learning_rate": 5.3467564797188464e-05,
+      "loss": 1.4939,
+      "step": 2772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14957407368554865,
+      "learning_rate": 5.3454385708009965e-05,
+      "loss": 1.4963,
+      "step": 2773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15194096680983524,
+      "learning_rate": 5.344120661883146e-05,
+      "loss": 1.4971,
+      "step": 2774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13186866203569403,
+      "learning_rate": 5.342802752965296e-05,
+      "loss": 1.4943,
+      "step": 2775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1214371603209352,
+      "learning_rate": 5.3414848440474446e-05,
+      "loss": 1.4927,
+      "step": 2776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11456141456936693,
+      "learning_rate": 5.340166935129595e-05,
+      "loss": 1.4952,
+      "step": 2777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11243413547360485,
+      "learning_rate": 5.338849026211744e-05,
+      "loss": 1.4929,
+      "step": 2778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11510190862426196,
+      "learning_rate": 5.337531117293894e-05,
+      "loss": 1.4922,
+      "step": 2779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09659206854781295,
+      "learning_rate": 5.3362132083760435e-05,
+      "loss": 1.4997,
+      "step": 2780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10177532486912877,
+      "learning_rate": 5.3348952994581935e-05,
+      "loss": 1.4951,
+      "step": 2781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12482216190856427,
+      "learning_rate": 5.3335773905403436e-05,
+      "loss": 1.4979,
+      "step": 2782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15196121229213047,
+      "learning_rate": 5.332259481622493e-05,
+      "loss": 1.5014,
+      "step": 2783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14360804932367807,
+      "learning_rate": 5.3309415727046416e-05,
+      "loss": 1.4937,
+      "step": 2784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1255800889383325,
+      "learning_rate": 5.329623663786792e-05,
+      "loss": 1.4992,
+      "step": 2785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10040661137902239,
+      "learning_rate": 5.328305754868942e-05,
+      "loss": 1.4899,
+      "step": 2786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0935679608322714,
+      "learning_rate": 5.326987845951091e-05,
+      "loss": 1.4954,
+      "step": 2787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10643911354121754,
+      "learning_rate": 5.325669937033241e-05,
+      "loss": 1.4954,
+      "step": 2788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10399520065595433,
+      "learning_rate": 5.3243520281153905e-05,
+      "loss": 1.4988,
+      "step": 2789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1042784785278281,
+      "learning_rate": 5.3230341191975406e-05,
+      "loss": 1.4941,
+      "step": 2790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10280408335258448,
+      "learning_rate": 5.3217162102796906e-05,
+      "loss": 1.4878,
+      "step": 2791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10513666717061232,
+      "learning_rate": 5.3203983013618393e-05,
+      "loss": 1.4913,
+      "step": 2792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11669284871397308,
+      "learning_rate": 5.319080392443989e-05,
+      "loss": 1.4898,
+      "step": 2793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11321484214083428,
+      "learning_rate": 5.317762483526139e-05,
+      "loss": 1.4951,
+      "step": 2794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09493581696110297,
+      "learning_rate": 5.316444574608288e-05,
+      "loss": 1.4863,
+      "step": 2795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09545138105447225,
+      "learning_rate": 5.315126665690438e-05,
+      "loss": 1.4979,
+      "step": 2796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10294111101641779,
+      "learning_rate": 5.313808756772588e-05,
+      "loss": 1.484,
+      "step": 2797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11147301544860971,
+      "learning_rate": 5.3124908478547376e-05,
+      "loss": 1.4955,
+      "step": 2798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10598123043730247,
+      "learning_rate": 5.311172938936888e-05,
+      "loss": 1.4917,
+      "step": 2799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11568637707719898,
+      "learning_rate": 5.3098550300190364e-05,
+      "loss": 1.49,
+      "step": 2800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1160692740409863,
+      "learning_rate": 5.3085371211011864e-05,
+      "loss": 1.4904,
+      "step": 2801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1169640248885231,
+      "learning_rate": 5.307219212183336e-05,
+      "loss": 1.4975,
+      "step": 2802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12061371621918753,
+      "learning_rate": 5.305901303265486e-05,
+      "loss": 1.4967,
+      "step": 2803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10843660690388816,
+      "learning_rate": 5.304583394347635e-05,
+      "loss": 1.4921,
+      "step": 2804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11129190927515167,
+      "learning_rate": 5.303265485429785e-05,
+      "loss": 1.4905,
+      "step": 2805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13573488780908907,
+      "learning_rate": 5.301947576511935e-05,
+      "loss": 1.4901,
+      "step": 2806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14519569357814952,
+      "learning_rate": 5.300629667594085e-05,
+      "loss": 1.4893,
+      "step": 2807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1399008801514865,
+      "learning_rate": 5.2993117586762334e-05,
+      "loss": 1.4948,
+      "step": 2808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12834406765147194,
+      "learning_rate": 5.2979938497583835e-05,
+      "loss": 1.4936,
+      "step": 2809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12006376041046078,
+      "learning_rate": 5.2966759408405335e-05,
+      "loss": 1.4928,
+      "step": 2810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12409474821704115,
+      "learning_rate": 5.295358031922683e-05,
+      "loss": 1.4882,
+      "step": 2811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.14294561549894308,
+      "learning_rate": 5.294040123004833e-05,
+      "loss": 1.4831,
+      "step": 2812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13098532563080337,
+      "learning_rate": 5.292722214086982e-05,
+      "loss": 1.4934,
+      "step": 2813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10906509542638045,
+      "learning_rate": 5.2914043051691324e-05,
+      "loss": 1.4897,
+      "step": 2814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1184671267395667,
+      "learning_rate": 5.2900863962512824e-05,
+      "loss": 1.4927,
+      "step": 2815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11814152946794061,
+      "learning_rate": 5.288768487333431e-05,
+      "loss": 1.4976,
+      "step": 2816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10444410519832348,
+      "learning_rate": 5.2874505784155805e-05,
+      "loss": 1.486,
+      "step": 2817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09299889795242428,
+      "learning_rate": 5.2861326694977305e-05,
+      "loss": 1.4892,
+      "step": 2818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11313188858553154,
+      "learning_rate": 5.28481476057988e-05,
+      "loss": 1.4837,
+      "step": 2819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10877945785823535,
+      "learning_rate": 5.28349685166203e-05,
+      "loss": 1.4934,
+      "step": 2820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10061229574269047,
+      "learning_rate": 5.28217894274418e-05,
+      "loss": 1.4979,
+      "step": 2821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09799133563439803,
+      "learning_rate": 5.2808610338263294e-05,
+      "loss": 1.4884,
+      "step": 2822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1069178511070031,
+      "learning_rate": 5.2795431249084794e-05,
+      "loss": 1.4877,
+      "step": 2823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13203382173845846,
+      "learning_rate": 5.278225215990628e-05,
+      "loss": 1.4873,
+      "step": 2824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13199757108591673,
+      "learning_rate": 5.276907307072778e-05,
+      "loss": 1.4959,
+      "step": 2825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1095509360963129,
+      "learning_rate": 5.2755893981549276e-05,
+      "loss": 1.4991,
+      "step": 2826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10220674943380316,
+      "learning_rate": 5.2742714892370776e-05,
+      "loss": 1.5045,
+      "step": 2827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12204298168002654,
+      "learning_rate": 5.272953580319227e-05,
+      "loss": 1.4927,
+      "step": 2828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11639584142528912,
+      "learning_rate": 5.271635671401377e-05,
+      "loss": 1.4921,
+      "step": 2829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10223214749717653,
+      "learning_rate": 5.270317762483527e-05,
+      "loss": 1.487,
+      "step": 2830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12269498185311331,
+      "learning_rate": 5.2689998535656765e-05,
+      "loss": 1.4845,
+      "step": 2831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.13204890799062727,
+      "learning_rate": 5.267681944647825e-05,
+      "loss": 1.4844,
+      "step": 2832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.12191290242573255,
+      "learning_rate": 5.266364035729975e-05,
+      "loss": 1.4887,
+      "step": 2833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.10721675030238649,
+      "learning_rate": 5.2650461268121246e-05,
+      "loss": 1.4971,
+      "step": 2834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12755702344489817,
+      "learning_rate": 5.2637282178942746e-05,
+      "loss": 1.4863,
+      "step": 2835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13294121872047773,
+      "learning_rate": 5.262410308976425e-05,
+      "loss": 1.4967,
+      "step": 2836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10523330103115766,
+      "learning_rate": 5.261092400058574e-05,
+      "loss": 1.4921,
+      "step": 2837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1156500382633159,
+      "learning_rate": 5.259774491140724e-05,
+      "loss": 1.4875,
+      "step": 2838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12598862335721164,
+      "learning_rate": 5.2584565822228735e-05,
+      "loss": 1.4933,
+      "step": 2839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10700025290379674,
+      "learning_rate": 5.257138673305023e-05,
+      "loss": 1.5007,
+      "step": 2840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1323049786463514,
+      "learning_rate": 5.255820764387172e-05,
+      "loss": 1.4972,
+      "step": 2841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14932602488521043,
+      "learning_rate": 5.254502855469322e-05,
+      "loss": 1.492,
+      "step": 2842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13447681533091455,
+      "learning_rate": 5.2531849465514717e-05,
+      "loss": 1.4914,
+      "step": 2843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13670429112620638,
+      "learning_rate": 5.251867037633622e-05,
+      "loss": 1.4854,
+      "step": 2844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.153011536077681,
+      "learning_rate": 5.250549128715772e-05,
+      "loss": 1.499,
+      "step": 2845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1486824720899116,
+      "learning_rate": 5.249231219797921e-05,
+      "loss": 1.4914,
+      "step": 2846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13338485946193251,
+      "learning_rate": 5.247913310880071e-05,
+      "loss": 1.4883,
+      "step": 2847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10065604643950697,
+      "learning_rate": 5.24659540196222e-05,
+      "loss": 1.4904,
+      "step": 2848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09867917443478179,
+      "learning_rate": 5.24527749304437e-05,
+      "loss": 1.4888,
+      "step": 2849
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13440653713110623,
+      "learning_rate": 5.243959584126519e-05,
+      "loss": 1.4908,
+      "step": 2850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12466052866159136,
+      "learning_rate": 5.2426416752086694e-05,
+      "loss": 1.5002,
+      "step": 2851
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09420726497846425,
+      "learning_rate": 5.241323766290819e-05,
+      "loss": 1.4864,
+      "step": 2852
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1082165370549008,
+      "learning_rate": 5.240005857372969e-05,
+      "loss": 1.4973,
+      "step": 2853
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11573384436310881,
+      "learning_rate": 5.238687948455119e-05,
+      "loss": 1.4937,
+      "step": 2854
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0971961911973844,
+      "learning_rate": 5.237370039537268e-05,
+      "loss": 1.4934,
+      "step": 2855
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10313003490716678,
+      "learning_rate": 5.236052130619417e-05,
+      "loss": 1.4859,
+      "step": 2856
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11472144777125268,
+      "learning_rate": 5.234734221701567e-05,
+      "loss": 1.496,
+      "step": 2857
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1331224776538519,
+      "learning_rate": 5.233416312783716e-05,
+      "loss": 1.4997,
+      "step": 2858
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13171744380273845,
+      "learning_rate": 5.2320984038658664e-05,
+      "loss": 1.4954,
+      "step": 2859
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10789171047595526,
+      "learning_rate": 5.2307804949480164e-05,
+      "loss": 1.4874,
+      "step": 2860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12590013016444698,
+      "learning_rate": 5.229462586030166e-05,
+      "loss": 1.4879,
+      "step": 2861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13753423780532073,
+      "learning_rate": 5.228144677112316e-05,
+      "loss": 1.4868,
+      "step": 2862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10039492784205267,
+      "learning_rate": 5.226826768194465e-05,
+      "loss": 1.4834,
+      "step": 2863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1121301906591348,
+      "learning_rate": 5.2255088592766146e-05,
+      "loss": 1.4827,
+      "step": 2864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12445233507829614,
+      "learning_rate": 5.224190950358764e-05,
+      "loss": 1.4911,
+      "step": 2865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1225088509190975,
+      "learning_rate": 5.222873041440914e-05,
+      "loss": 1.4928,
+      "step": 2866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1005764327255583,
+      "learning_rate": 5.2215551325230634e-05,
+      "loss": 1.4878,
+      "step": 2867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1218370743878816,
+      "learning_rate": 5.2202372236052135e-05,
+      "loss": 1.4824,
+      "step": 2868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1332075987738963,
+      "learning_rate": 5.2189193146873635e-05,
+      "loss": 1.4905,
+      "step": 2869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12847345560482715,
+      "learning_rate": 5.217601405769513e-05,
+      "loss": 1.4857,
+      "step": 2870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10684395792615567,
+      "learning_rate": 5.216283496851663e-05,
+      "loss": 1.4838,
+      "step": 2871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09516209738249969,
+      "learning_rate": 5.2149655879338116e-05,
+      "loss": 1.4853,
+      "step": 2872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11518671684433118,
+      "learning_rate": 5.213647679015962e-05,
+      "loss": 1.4928,
+      "step": 2873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11004766651502208,
+      "learning_rate": 5.212329770098111e-05,
+      "loss": 1.5009,
+      "step": 2874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12156408349018782,
+      "learning_rate": 5.211011861180261e-05,
+      "loss": 1.4992,
+      "step": 2875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13194937838537651,
+      "learning_rate": 5.2096939522624105e-05,
+      "loss": 1.4898,
+      "step": 2876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12157628837887174,
+      "learning_rate": 5.2083760433445605e-05,
+      "loss": 1.4955,
+      "step": 2877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11001785681871887,
+      "learning_rate": 5.2070581344267106e-05,
+      "loss": 1.4907,
+      "step": 2878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10919827174080704,
+      "learning_rate": 5.20574022550886e-05,
+      "loss": 1.4988,
+      "step": 2879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1066883515697405,
+      "learning_rate": 5.204422316591009e-05,
+      "loss": 1.4877,
+      "step": 2880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11733174361789249,
+      "learning_rate": 5.203104407673159e-05,
+      "loss": 1.4931,
+      "step": 2881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14031600115321025,
+      "learning_rate": 5.201786498755308e-05,
+      "loss": 1.4948,
+      "step": 2882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12495188684307555,
+      "learning_rate": 5.200468589837458e-05,
+      "loss": 1.4889,
+      "step": 2883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11028132625855562,
+      "learning_rate": 5.199150680919608e-05,
+      "loss": 1.4936,
+      "step": 2884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10801291119859457,
+      "learning_rate": 5.1978327720017576e-05,
+      "loss": 1.4869,
+      "step": 2885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.14069254082292612,
+      "learning_rate": 5.1965148630839076e-05,
+      "loss": 1.4984,
+      "step": 2886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1180925846654035,
+      "learning_rate": 5.195196954166057e-05,
+      "loss": 1.4912,
+      "step": 2887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10325787352746527,
+      "learning_rate": 5.1938790452482064e-05,
+      "loss": 1.4949,
+      "step": 2888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1318553842969829,
+      "learning_rate": 5.192561136330356e-05,
+      "loss": 1.4969,
+      "step": 2889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13700032609108334,
+      "learning_rate": 5.191243227412506e-05,
+      "loss": 1.4925,
+      "step": 2890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12259736958552872,
+      "learning_rate": 5.189925318494655e-05,
+      "loss": 1.491,
+      "step": 2891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10688835808826774,
+      "learning_rate": 5.188607409576805e-05,
+      "loss": 1.4911,
+      "step": 2892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10730508925639033,
+      "learning_rate": 5.187289500658955e-05,
+      "loss": 1.4931,
+      "step": 2893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11833212117678003,
+      "learning_rate": 5.1859715917411047e-05,
+      "loss": 1.4973,
+      "step": 2894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11698236606961583,
+      "learning_rate": 5.184653682823255e-05,
+      "loss": 1.4911,
+      "step": 2895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12805228488556739,
+      "learning_rate": 5.1833357739054034e-05,
+      "loss": 1.5061,
+      "step": 2896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10649907950037944,
+      "learning_rate": 5.1820178649875535e-05,
+      "loss": 1.4927,
+      "step": 2897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09692053616014223,
+      "learning_rate": 5.180699956069703e-05,
+      "loss": 1.492,
+      "step": 2898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.11044975084262314,
+      "learning_rate": 5.179382047151853e-05,
+      "loss": 1.4882,
+      "step": 2899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.10096830609712813,
+      "learning_rate": 5.178064138234002e-05,
+      "loss": 1.4927,
+      "step": 2900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09972036711671924,
+      "learning_rate": 5.176746229316152e-05,
+      "loss": 1.491,
+      "step": 2901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12201635765737624,
+      "learning_rate": 5.1754283203983024e-05,
+      "loss": 1.4998,
+      "step": 2902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13877030491066733,
+      "learning_rate": 5.174110411480452e-05,
+      "loss": 1.4915,
+      "step": 2903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11801029153115562,
+      "learning_rate": 5.1727925025626004e-05,
+      "loss": 1.4966,
+      "step": 2904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12048964771530059,
+      "learning_rate": 5.1714745936447505e-05,
+      "loss": 1.494,
+      "step": 2905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12650273469476536,
+      "learning_rate": 5.1701566847269e-05,
+      "loss": 1.4906,
+      "step": 2906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11769995404398347,
+      "learning_rate": 5.16883877580905e-05,
+      "loss": 1.489,
+      "step": 2907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1033804074536896,
+      "learning_rate": 5.1675208668912e-05,
+      "loss": 1.4881,
+      "step": 2908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10276181779346794,
+      "learning_rate": 5.166202957973349e-05,
+      "loss": 1.4875,
+      "step": 2909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10619644828236466,
+      "learning_rate": 5.1648850490554994e-05,
+      "loss": 1.489,
+      "step": 2910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10136166845391188,
+      "learning_rate": 5.163567140137649e-05,
+      "loss": 1.494,
+      "step": 2911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12249496176463472,
+      "learning_rate": 5.162249231219798e-05,
+      "loss": 1.4907,
+      "step": 2912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11277963809204937,
+      "learning_rate": 5.1609313223019475e-05,
+      "loss": 1.4895,
+      "step": 2913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10926417609585298,
+      "learning_rate": 5.1596134133840976e-05,
+      "loss": 1.4891,
+      "step": 2914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11863117597191604,
+      "learning_rate": 5.158295504466247e-05,
+      "loss": 1.4906,
+      "step": 2915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13340931734432715,
+      "learning_rate": 5.156977595548397e-05,
+      "loss": 1.4903,
+      "step": 2916
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12111168529300594,
+      "learning_rate": 5.155659686630547e-05,
+      "loss": 1.5012,
+      "step": 2917
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1225701475390722,
+      "learning_rate": 5.1543417777126964e-05,
+      "loss": 1.496,
+      "step": 2918
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12785664687211326,
+      "learning_rate": 5.1530238687948465e-05,
+      "loss": 1.4871,
+      "step": 2919
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10939243218690753,
+      "learning_rate": 5.151705959876995e-05,
+      "loss": 1.4921,
+      "step": 2920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10119128779884717,
+      "learning_rate": 5.1503880509591445e-05,
+      "loss": 1.4854,
+      "step": 2921
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11839237379324696,
+      "learning_rate": 5.1490701420412946e-05,
+      "loss": 1.4892,
+      "step": 2922
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09910129986154906,
+      "learning_rate": 5.1477522331234446e-05,
+      "loss": 1.4906,
+      "step": 2923
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1075909215870641,
+      "learning_rate": 5.146434324205594e-05,
+      "loss": 1.4889,
+      "step": 2924
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12299386528236268,
+      "learning_rate": 5.145116415287744e-05,
+      "loss": 1.4784,
+      "step": 2925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1268238091578074,
+      "learning_rate": 5.1437985063698934e-05,
+      "loss": 1.4906,
+      "step": 2926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11716083843974592,
+      "learning_rate": 5.1424805974520435e-05,
+      "loss": 1.4922,
+      "step": 2927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11706933867922932,
+      "learning_rate": 5.141162688534192e-05,
+      "loss": 1.485,
+      "step": 2928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09305892633432755,
+      "learning_rate": 5.139844779616342e-05,
+      "loss": 1.4894,
+      "step": 2929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10195279518492885,
+      "learning_rate": 5.1385268706984916e-05,
+      "loss": 1.4828,
+      "step": 2930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10784380305302292,
+      "learning_rate": 5.1372089617806417e-05,
+      "loss": 1.4888,
+      "step": 2931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12323749539923103,
+      "learning_rate": 5.135891052862792e-05,
+      "loss": 1.4941,
+      "step": 2932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11433218883477038,
+      "learning_rate": 5.134573143944941e-05,
+      "loss": 1.4963,
+      "step": 2933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10215594148173214,
+      "learning_rate": 5.133255235027091e-05,
+      "loss": 1.4927,
+      "step": 2934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11246018513424666,
+      "learning_rate": 5.1319373261092405e-05,
+      "loss": 1.4858,
+      "step": 2935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11195580451436517,
+      "learning_rate": 5.13061941719139e-05,
+      "loss": 1.4971,
+      "step": 2936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1092535547791243,
+      "learning_rate": 5.129301508273539e-05,
+      "loss": 1.4922,
+      "step": 2937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1146004621515018,
+      "learning_rate": 5.127983599355689e-05,
+      "loss": 1.493,
+      "step": 2938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10377787260151529,
+      "learning_rate": 5.126665690437839e-05,
+      "loss": 1.4874,
+      "step": 2939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09026233911023682,
+      "learning_rate": 5.125347781519989e-05,
+      "loss": 1.488,
+      "step": 2940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11576794501907824,
+      "learning_rate": 5.124029872602139e-05,
+      "loss": 1.4839,
+      "step": 2941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09809744590350951,
+      "learning_rate": 5.122711963684288e-05,
+      "loss": 1.4912,
+      "step": 2942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10363043487132922,
+      "learning_rate": 5.121394054766438e-05,
+      "loss": 1.4928,
+      "step": 2943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1011953296555709,
+      "learning_rate": 5.120076145848587e-05,
+      "loss": 1.4781,
+      "step": 2944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10099743580539994,
+      "learning_rate": 5.118758236930736e-05,
+      "loss": 1.4848,
+      "step": 2945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.10488147541496755,
+      "learning_rate": 5.1174403280128863e-05,
+      "loss": 1.4904,
+      "step": 2946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09981249196186916,
+      "learning_rate": 5.1161224190950364e-05,
+      "loss": 1.4897,
+      "step": 2947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08826465802023163,
+      "learning_rate": 5.114804510177186e-05,
+      "loss": 1.4875,
+      "step": 2948
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09573498059939568,
+      "learning_rate": 5.113486601259336e-05,
+      "loss": 1.4886,
+      "step": 2949
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11267604084081835,
+      "learning_rate": 5.112168692341485e-05,
+      "loss": 1.4914,
+      "step": 2950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11889889163664931,
+      "learning_rate": 5.110850783423635e-05,
+      "loss": 1.4911,
+      "step": 2951
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1442179752292733,
+      "learning_rate": 5.109532874505784e-05,
+      "loss": 1.4857,
+      "step": 2952
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15185475526982706,
+      "learning_rate": 5.108214965587934e-05,
+      "loss": 1.4945,
+      "step": 2953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14608792971652135,
+      "learning_rate": 5.1068970566700834e-05,
+      "loss": 1.4796,
+      "step": 2954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13019551604757795,
+      "learning_rate": 5.1055791477522334e-05,
+      "loss": 1.4858,
+      "step": 2955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12557375928462197,
+      "learning_rate": 5.1042612388343835e-05,
+      "loss": 1.4825,
+      "step": 2956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15329701804962378,
+      "learning_rate": 5.102943329916533e-05,
+      "loss": 1.4967,
+      "step": 2957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18025086553678346,
+      "learning_rate": 5.101625420998683e-05,
+      "loss": 1.4894,
+      "step": 2958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1368415672722037,
+      "learning_rate": 5.100307512080832e-05,
+      "loss": 1.4952,
+      "step": 2959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.09189861610771621,
+      "learning_rate": 5.0989896031629816e-05,
+      "loss": 1.4868,
+      "step": 2960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12865859396396095,
+      "learning_rate": 5.097671694245131e-05,
+      "loss": 1.4917,
+      "step": 2961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.14521402154587484,
+      "learning_rate": 5.096353785327281e-05,
+      "loss": 1.4862,
+      "step": 2962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.13118737042239717,
+      "learning_rate": 5.0950358764094304e-05,
+      "loss": 1.4916,
+      "step": 2963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.11416020697795022,
+      "learning_rate": 5.0937179674915805e-05,
+      "loss": 1.486,
+      "step": 2964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.130161089212139,
+      "learning_rate": 5.0924000585737305e-05,
+      "loss": 1.4872,
+      "step": 2965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.12420549440874959,
+      "learning_rate": 5.09108214965588e-05,
+      "loss": 1.4992,
+      "step": 2966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1430832597331034,
+      "learning_rate": 5.08976424073803e-05,
+      "loss": 1.4923,
+      "step": 2967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15610352676118183,
+      "learning_rate": 5.088446331820179e-05,
+      "loss": 1.4932,
+      "step": 2968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1239537042771846,
+      "learning_rate": 5.087128422902328e-05,
+      "loss": 1.4825,
+      "step": 2969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1261314055659946,
+      "learning_rate": 5.085810513984478e-05,
+      "loss": 1.4874,
+      "step": 2970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1505423211666796,
+      "learning_rate": 5.084492605066628e-05,
+      "loss": 1.489,
+      "step": 2971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12532503128131445,
+      "learning_rate": 5.0831746961487775e-05,
+      "loss": 1.4874,
+      "step": 2972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10715609285102758,
+      "learning_rate": 5.0818567872309276e-05,
+      "loss": 1.4823,
+      "step": 2973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11600155388732597,
+      "learning_rate": 5.080538878313077e-05,
+      "loss": 1.4853,
+      "step": 2974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11152182340769132,
+      "learning_rate": 5.079220969395227e-05,
+      "loss": 1.4875,
+      "step": 2975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10331995288301123,
+      "learning_rate": 5.077903060477376e-05,
+      "loss": 1.4915,
+      "step": 2976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10283688603052797,
+      "learning_rate": 5.076585151559526e-05,
+      "loss": 1.4857,
+      "step": 2977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10227281136075494,
+      "learning_rate": 5.075267242641675e-05,
+      "loss": 1.4892,
+      "step": 2978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11054933134589809,
+      "learning_rate": 5.073949333723825e-05,
+      "loss": 1.4929,
+      "step": 2979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09806885209275167,
+      "learning_rate": 5.072631424805975e-05,
+      "loss": 1.4943,
+      "step": 2980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11886481442849188,
+      "learning_rate": 5.0713135158881246e-05,
+      "loss": 1.4799,
+      "step": 2981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12128193491984597,
+      "learning_rate": 5.0699956069702747e-05,
+      "loss": 1.4971,
+      "step": 2982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11315770644139304,
+      "learning_rate": 5.068677698052424e-05,
+      "loss": 1.4944,
+      "step": 2983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10428122260455944,
+      "learning_rate": 5.0673597891345734e-05,
+      "loss": 1.483,
+      "step": 2984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10340822486747624,
+      "learning_rate": 5.066041880216723e-05,
+      "loss": 1.4807,
+      "step": 2985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09418851110288269,
+      "learning_rate": 5.064723971298873e-05,
+      "loss": 1.4811,
+      "step": 2986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10585765051908941,
+      "learning_rate": 5.063406062381022e-05,
+      "loss": 1.4886,
+      "step": 2987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10857075595483152,
+      "learning_rate": 5.062088153463172e-05,
+      "loss": 1.4818,
+      "step": 2988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09889682148389331,
+      "learning_rate": 5.060770244545322e-05,
+      "loss": 1.4811,
+      "step": 2989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10722732229470074,
+      "learning_rate": 5.059452335627472e-05,
+      "loss": 1.4952,
+      "step": 2990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1134827867737587,
+      "learning_rate": 5.058134426709622e-05,
+      "loss": 1.4907,
+      "step": 2991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13420669948967387,
+      "learning_rate": 5.0568165177917704e-05,
+      "loss": 1.4927,
+      "step": 2992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.13452373281251565,
+      "learning_rate": 5.05549860887392e-05,
+      "loss": 1.4891,
+      "step": 2993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10180849396557298,
+      "learning_rate": 5.05418069995607e-05,
+      "loss": 1.4904,
+      "step": 2994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11214222696374145,
+      "learning_rate": 5.05286279103822e-05,
+      "loss": 1.4828,
+      "step": 2995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1308248104254933,
+      "learning_rate": 5.051544882120369e-05,
+      "loss": 1.4837,
+      "step": 2996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14514465517656802,
+      "learning_rate": 5.050226973202519e-05,
+      "loss": 1.491,
+      "step": 2997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12625308027250007,
+      "learning_rate": 5.048909064284669e-05,
+      "loss": 1.4754,
+      "step": 2998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12154642993620479,
+      "learning_rate": 5.047591155366819e-05,
+      "loss": 1.4923,
+      "step": 2999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11402219131263519,
+      "learning_rate": 5.0462732464489675e-05,
+      "loss": 1.4882,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1127525910807034,
+      "learning_rate": 5.0449553375311175e-05,
+      "loss": 1.4907,
+      "step": 3001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09957561008819966,
+      "learning_rate": 5.043637428613267e-05,
+      "loss": 1.4874,
+      "step": 3002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09298138155319488,
+      "learning_rate": 5.042319519695417e-05,
+      "loss": 1.4846,
+      "step": 3003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1041387414336425,
+      "learning_rate": 5.041001610777567e-05,
+      "loss": 1.4825,
+      "step": 3004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10729955376326311,
+      "learning_rate": 5.0396837018597164e-05,
+      "loss": 1.4757,
+      "step": 3005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10123693015672049,
+      "learning_rate": 5.0383657929418664e-05,
+      "loss": 1.4878,
+      "step": 3006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08748049858028269,
+      "learning_rate": 5.037047884024016e-05,
+      "loss": 1.484,
+      "step": 3007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10065118962918254,
+      "learning_rate": 5.0357299751061645e-05,
+      "loss": 1.4895,
+      "step": 3008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11256252249507148,
+      "learning_rate": 5.0344120661883145e-05,
+      "loss": 1.4878,
+      "step": 3009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12123069517515592,
+      "learning_rate": 5.0330941572704646e-05,
+      "loss": 1.4771,
+      "step": 3010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10951033739840839,
+      "learning_rate": 5.031776248352614e-05,
+      "loss": 1.4947,
+      "step": 3011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10791626609031876,
+      "learning_rate": 5.030458339434764e-05,
+      "loss": 1.4877,
+      "step": 3012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11729905813247374,
+      "learning_rate": 5.0291404305169134e-05,
+      "loss": 1.4956,
+      "step": 3013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12243132740398185,
+      "learning_rate": 5.0278225215990634e-05,
+      "loss": 1.4956,
+      "step": 3014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10163163067114893,
+      "learning_rate": 5.0265046126812135e-05,
+      "loss": 1.494,
+      "step": 3015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09927407338343205,
+      "learning_rate": 5.025186703763362e-05,
+      "loss": 1.4866,
+      "step": 3016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1223479211158611,
+      "learning_rate": 5.0238687948455116e-05,
+      "loss": 1.481,
+      "step": 3017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12686664004705292,
+      "learning_rate": 5.0225508859276616e-05,
+      "loss": 1.4854,
+      "step": 3018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11479716569540002,
+      "learning_rate": 5.021232977009812e-05,
+      "loss": 1.4856,
+      "step": 3019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0955342718122527,
+      "learning_rate": 5.019915068091961e-05,
+      "loss": 1.4943,
+      "step": 3020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10026464659142803,
+      "learning_rate": 5.018597159174111e-05,
+      "loss": 1.4812,
+      "step": 3021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11142962278088012,
+      "learning_rate": 5.0172792502562605e-05,
+      "loss": 1.4886,
+      "step": 3022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11158356962420991,
+      "learning_rate": 5.0159613413384105e-05,
+      "loss": 1.4857,
+      "step": 3023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.099021288436058,
+      "learning_rate": 5.014643432420559e-05,
+      "loss": 1.4897,
+      "step": 3024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.14144360698318428,
+      "learning_rate": 5.013325523502709e-05,
+      "loss": 1.4843,
+      "step": 3025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.15847409143756244,
+      "learning_rate": 5.0120076145848586e-05,
+      "loss": 1.4824,
+      "step": 3026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11678394943826928,
+      "learning_rate": 5.010689705667009e-05,
+      "loss": 1.4969,
+      "step": 3027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10735377983094571,
+      "learning_rate": 5.009371796749159e-05,
+      "loss": 1.493,
+      "step": 3028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12948354015602467,
+      "learning_rate": 5.008053887831308e-05,
+      "loss": 1.4782,
+      "step": 3029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09925554542531574,
+      "learning_rate": 5.006735978913458e-05,
+      "loss": 1.4777,
+      "step": 3030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10387612801417175,
+      "learning_rate": 5.0054180699956075e-05,
+      "loss": 1.4913,
+      "step": 3031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.12266478932372457,
+      "learning_rate": 5.004100161077756e-05,
+      "loss": 1.482,
+      "step": 3032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1063204170358716,
+      "learning_rate": 5.002782252159906e-05,
+      "loss": 1.4879,
+      "step": 3033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.10628454890028889,
+      "learning_rate": 5.0014643432420563e-05,
+      "loss": 1.4842,
+      "step": 3034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.09621304421494409,
+      "learning_rate": 5.000146434324206e-05,
+      "loss": 1.4802,
+      "step": 3035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.08934248157042414,
+      "learning_rate": 4.998828525406356e-05,
+      "loss": 1.4827,
+      "step": 3036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1021657195405572,
+      "learning_rate": 4.997510616488505e-05,
+      "loss": 1.4809,
+      "step": 3037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.11727088581785619,
+      "learning_rate": 4.996192707570655e-05,
+      "loss": 1.4825,
+      "step": 3038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1051147033506965,
+      "learning_rate": 4.994874798652805e-05,
+      "loss": 1.4865,
+      "step": 3039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10501693251991778,
+      "learning_rate": 4.993556889734954e-05,
+      "loss": 1.4816,
+      "step": 3040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11643623045623495,
+      "learning_rate": 4.992238980817103e-05,
+      "loss": 1.4903,
+      "step": 3041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10128450014850401,
+      "learning_rate": 4.9909210718992534e-05,
+      "loss": 1.4902,
+      "step": 3042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0974708100934228,
+      "learning_rate": 4.9896031629814034e-05,
+      "loss": 1.4833,
+      "step": 3043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10912619023348644,
+      "learning_rate": 4.988285254063553e-05,
+      "loss": 1.4856,
+      "step": 3044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10586934759224945,
+      "learning_rate": 4.986967345145703e-05,
+      "loss": 1.4848,
+      "step": 3045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11116631070120875,
+      "learning_rate": 4.985649436227852e-05,
+      "loss": 1.4818,
+      "step": 3046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09475249945596483,
+      "learning_rate": 4.984331527310002e-05,
+      "loss": 1.4882,
+      "step": 3047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.093698687517116,
+      "learning_rate": 4.983013618392151e-05,
+      "loss": 1.4911,
+      "step": 3048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0978995526213321,
+      "learning_rate": 4.981695709474301e-05,
+      "loss": 1.4737,
+      "step": 3049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10000145670177514,
+      "learning_rate": 4.9803778005564504e-05,
+      "loss": 1.4881,
+      "step": 3050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09236423135293678,
+      "learning_rate": 4.9790598916386004e-05,
+      "loss": 1.4929,
+      "step": 3051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09982289041624612,
+      "learning_rate": 4.9777419827207505e-05,
+      "loss": 1.4767,
+      "step": 3052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11355334693245564,
+      "learning_rate": 4.9764240738029e-05,
+      "loss": 1.4838,
+      "step": 3053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.123250592930831,
+      "learning_rate": 4.97510616488505e-05,
+      "loss": 1.4867,
+      "step": 3054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13982321047328852,
+      "learning_rate": 4.973788255967199e-05,
+      "loss": 1.4787,
+      "step": 3055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14041586604417855,
+      "learning_rate": 4.972470347049348e-05,
+      "loss": 1.4826,
+      "step": 3056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1377067168262921,
+      "learning_rate": 4.971152438131498e-05,
+      "loss": 1.4939,
+      "step": 3057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13542274696959128,
+      "learning_rate": 4.969834529213648e-05,
+      "loss": 1.4715,
+      "step": 3058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12754081100580697,
+      "learning_rate": 4.9685166202957975e-05,
+      "loss": 1.4855,
+      "step": 3059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13260366003837676,
+      "learning_rate": 4.9671987113779475e-05,
+      "loss": 1.4833,
+      "step": 3060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12547482595686282,
+      "learning_rate": 4.965880802460097e-05,
+      "loss": 1.4768,
+      "step": 3061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12255481196810049,
+      "learning_rate": 4.964562893542247e-05,
+      "loss": 1.4898,
+      "step": 3062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12297387738635002,
+      "learning_rate": 4.963244984624397e-05,
+      "loss": 1.4808,
+      "step": 3063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11178658170803041,
+      "learning_rate": 4.961927075706546e-05,
+      "loss": 1.4795,
+      "step": 3064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10964611968772277,
+      "learning_rate": 4.960609166788695e-05,
+      "loss": 1.4731,
+      "step": 3065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10899645141992698,
+      "learning_rate": 4.959291257870845e-05,
+      "loss": 1.4796,
+      "step": 3066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11043471460459171,
+      "learning_rate": 4.957973348952995e-05,
+      "loss": 1.4892,
+      "step": 3067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09997242789307366,
+      "learning_rate": 4.9566554400351445e-05,
+      "loss": 1.4791,
+      "step": 3068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1042143863626963,
+      "learning_rate": 4.9553375311172946e-05,
+      "loss": 1.4828,
+      "step": 3069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1120717617912667,
+      "learning_rate": 4.954019622199444e-05,
+      "loss": 1.4804,
+      "step": 3070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10909681242484075,
+      "learning_rate": 4.952701713281594e-05,
+      "loss": 1.4792,
+      "step": 3071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09932080542646618,
+      "learning_rate": 4.951383804363743e-05,
+      "loss": 1.4842,
+      "step": 3072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10101090697772276,
+      "learning_rate": 4.950065895445893e-05,
+      "loss": 1.4837,
+      "step": 3073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11410663228926918,
+      "learning_rate": 4.948747986528042e-05,
+      "loss": 1.4913,
+      "step": 3074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13810373692800149,
+      "learning_rate": 4.947430077610192e-05,
+      "loss": 1.4954,
+      "step": 3075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.116019027101331,
+      "learning_rate": 4.946112168692342e-05,
+      "loss": 1.4892,
+      "step": 3076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09326868670120109,
+      "learning_rate": 4.9447942597744916e-05,
+      "loss": 1.4819,
+      "step": 3077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11740128495989403,
+      "learning_rate": 4.943476350856642e-05,
+      "loss": 1.4862,
+      "step": 3078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14162431587825897,
+      "learning_rate": 4.942158441938791e-05,
+      "loss": 1.4873,
+      "step": 3079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14748445382725184,
+      "learning_rate": 4.94084053302094e-05,
+      "loss": 1.4828,
+      "step": 3080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.13464212203462425,
+      "learning_rate": 4.93952262410309e-05,
+      "loss": 1.4863,
+      "step": 3081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10599262879124899,
+      "learning_rate": 4.93820471518524e-05,
+      "loss": 1.4854,
+      "step": 3082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08844975596246288,
+      "learning_rate": 4.936886806267389e-05,
+      "loss": 1.4854,
+      "step": 3083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1079403839312427,
+      "learning_rate": 4.935568897349539e-05,
+      "loss": 1.4894,
+      "step": 3084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10624884015661072,
+      "learning_rate": 4.9342509884316887e-05,
+      "loss": 1.4881,
+      "step": 3085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09246659139557792,
+      "learning_rate": 4.932933079513839e-05,
+      "loss": 1.4827,
+      "step": 3086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1010911610458932,
+      "learning_rate": 4.931615170595989e-05,
+      "loss": 1.4802,
+      "step": 3087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12566895357385868,
+      "learning_rate": 4.9302972616781375e-05,
+      "loss": 1.4795,
+      "step": 3088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1300980832198028,
+      "learning_rate": 4.928979352760287e-05,
+      "loss": 1.4791,
+      "step": 3089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11224377152288904,
+      "learning_rate": 4.927661443842437e-05,
+      "loss": 1.4906,
+      "step": 3090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10741356930662242,
+      "learning_rate": 4.926343534924587e-05,
+      "loss": 1.4782,
+      "step": 3091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11497616237073646,
+      "learning_rate": 4.925025626006736e-05,
+      "loss": 1.4901,
+      "step": 3092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12941106208591396,
+      "learning_rate": 4.9237077170888864e-05,
+      "loss": 1.4842,
+      "step": 3093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.14477350323842206,
+      "learning_rate": 4.922389808171036e-05,
+      "loss": 1.4868,
+      "step": 3094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11733442103430777,
+      "learning_rate": 4.921071899253186e-05,
+      "loss": 1.4833,
+      "step": 3095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09298816882140135,
+      "learning_rate": 4.9197539903353345e-05,
+      "loss": 1.4808,
+      "step": 3096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10026577704660987,
+      "learning_rate": 4.9184360814174845e-05,
+      "loss": 1.4887,
+      "step": 3097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09277244782373945,
+      "learning_rate": 4.917118172499634e-05,
+      "loss": 1.4895,
+      "step": 3098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.08614971580643997,
+      "learning_rate": 4.915800263581784e-05,
+      "loss": 1.4851,
+      "step": 3099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09237499051000185,
+      "learning_rate": 4.914482354663933e-05,
+      "loss": 1.4779,
+      "step": 3100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09207823520166294,
+      "learning_rate": 4.9131644457460834e-05,
+      "loss": 1.4755,
+      "step": 3101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09874973165953531,
+      "learning_rate": 4.9118465368282334e-05,
+      "loss": 1.4805,
+      "step": 3102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.10129801366275476,
+      "learning_rate": 4.910528627910383e-05,
+      "loss": 1.4746,
+      "step": 3103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.11239118637833209,
+      "learning_rate": 4.9092107189925315e-05,
+      "loss": 1.4852,
+      "step": 3104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.112257520691024,
+      "learning_rate": 4.9078928100746816e-05,
+      "loss": 1.4813,
+      "step": 3105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0995665507036368,
+      "learning_rate": 4.9065749011568316e-05,
+      "loss": 1.4866,
+      "step": 3106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09325682214752971,
+      "learning_rate": 4.905256992238981e-05,
+      "loss": 1.4912,
+      "step": 3107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10852855199058442,
+      "learning_rate": 4.903939083321131e-05,
+      "loss": 1.4821,
+      "step": 3108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12137060451908553,
+      "learning_rate": 4.9026211744032804e-05,
+      "loss": 1.4796,
+      "step": 3109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11841652400302279,
+      "learning_rate": 4.9013032654854305e-05,
+      "loss": 1.4788,
+      "step": 3110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10702651467405444,
+      "learning_rate": 4.899985356567579e-05,
+      "loss": 1.4841,
+      "step": 3111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10531153737891703,
+      "learning_rate": 4.898667447649729e-05,
+      "loss": 1.4883,
+      "step": 3112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13816917757411387,
+      "learning_rate": 4.8973495387318786e-05,
+      "loss": 1.4866,
+      "step": 3113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13436495874287557,
+      "learning_rate": 4.8960316298140286e-05,
+      "loss": 1.4844,
+      "step": 3114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11542945655387793,
+      "learning_rate": 4.894713720896179e-05,
+      "loss": 1.4856,
+      "step": 3115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0958186605107752,
+      "learning_rate": 4.893395811978328e-05,
+      "loss": 1.485,
+      "step": 3116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10674820320006848,
+      "learning_rate": 4.892077903060478e-05,
+      "loss": 1.4913,
+      "step": 3117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1398191332262879,
+      "learning_rate": 4.8907599941426275e-05,
+      "loss": 1.4904,
+      "step": 3118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.14460703128937855,
+      "learning_rate": 4.889442085224776e-05,
+      "loss": 1.4805,
+      "step": 3119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11179427362816495,
+      "learning_rate": 4.888124176306926e-05,
+      "loss": 1.489,
+      "step": 3120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10327470729953783,
+      "learning_rate": 4.886806267389076e-05,
+      "loss": 1.4772,
+      "step": 3121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11078418112476414,
+      "learning_rate": 4.885488358471226e-05,
+      "loss": 1.486,
+      "step": 3122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12771938974774716,
+      "learning_rate": 4.884170449553376e-05,
+      "loss": 1.4795,
+      "step": 3123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13062382377152118,
+      "learning_rate": 4.882852540635525e-05,
+      "loss": 1.4945,
+      "step": 3124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11955247858824192,
+      "learning_rate": 4.881534631717675e-05,
+      "loss": 1.4835,
+      "step": 3125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10857399236185736,
+      "learning_rate": 4.880216722799825e-05,
+      "loss": 1.4783,
+      "step": 3126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11606840957023472,
+      "learning_rate": 4.878898813881974e-05,
+      "loss": 1.4783,
+      "step": 3127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12711730349177944,
+      "learning_rate": 4.877580904964123e-05,
+      "loss": 1.4815,
+      "step": 3128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12929953785857756,
+      "learning_rate": 4.876262996046273e-05,
+      "loss": 1.4943,
+      "step": 3129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10574971253158875,
+      "learning_rate": 4.8749450871284234e-05,
+      "loss": 1.4826,
+      "step": 3130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10106330481092327,
+      "learning_rate": 4.873627178210573e-05,
+      "loss": 1.4961,
+      "step": 3131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1306870398287056,
+      "learning_rate": 4.872309269292723e-05,
+      "loss": 1.4789,
+      "step": 3132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10067026048847022,
+      "learning_rate": 4.870991360374872e-05,
+      "loss": 1.4846,
+      "step": 3133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10687764587379149,
+      "learning_rate": 4.869673451457022e-05,
+      "loss": 1.4782,
+      "step": 3134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1369048357900996,
+      "learning_rate": 4.868355542539171e-05,
+      "loss": 1.4833,
+      "step": 3135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1330767149059637,
+      "learning_rate": 4.867037633621321e-05,
+      "loss": 1.4881,
+      "step": 3136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1253343717010431,
+      "learning_rate": 4.8657197247034703e-05,
+      "loss": 1.4745,
+      "step": 3137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1316196748416333,
+      "learning_rate": 4.8644018157856204e-05,
+      "loss": 1.4866,
+      "step": 3138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1280273583124511,
+      "learning_rate": 4.8630839068677704e-05,
+      "loss": 1.4774,
+      "step": 3139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10349381301024442,
+      "learning_rate": 4.86176599794992e-05,
+      "loss": 1.4834,
+      "step": 3140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0880595387861994,
+      "learning_rate": 4.86044808903207e-05,
+      "loss": 1.4814,
+      "step": 3141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1281028605838777,
+      "learning_rate": 4.859130180114219e-05,
+      "loss": 1.4848,
+      "step": 3142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12468311628636931,
+      "learning_rate": 4.857812271196368e-05,
+      "loss": 1.4782,
+      "step": 3143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11247644686710012,
+      "learning_rate": 4.856494362278518e-05,
+      "loss": 1.4857,
+      "step": 3144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10498762979658031,
+      "learning_rate": 4.855176453360668e-05,
+      "loss": 1.4771,
+      "step": 3145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11127330383509841,
+      "learning_rate": 4.8538585444428174e-05,
+      "loss": 1.4742,
+      "step": 3146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10120905635531767,
+      "learning_rate": 4.8525406355249675e-05,
+      "loss": 1.4836,
+      "step": 3147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1092906787443453,
+      "learning_rate": 4.851222726607117e-05,
+      "loss": 1.4854,
+      "step": 3148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10723175837725141,
+      "learning_rate": 4.849904817689267e-05,
+      "loss": 1.4787,
+      "step": 3149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10695691103330428,
+      "learning_rate": 4.848586908771417e-05,
+      "loss": 1.4841,
+      "step": 3150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09272888929256734,
+      "learning_rate": 4.8472689998535656e-05,
+      "loss": 1.4856,
+      "step": 3151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10602848287718197,
+      "learning_rate": 4.845951090935715e-05,
+      "loss": 1.4822,
+      "step": 3152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11800757619184146,
+      "learning_rate": 4.844633182017865e-05,
+      "loss": 1.4774,
+      "step": 3153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12009172303400949,
+      "learning_rate": 4.843315273100015e-05,
+      "loss": 1.4853,
+      "step": 3154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10426472596241779,
+      "learning_rate": 4.8419973641821645e-05,
+      "loss": 1.4841,
+      "step": 3155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09868927633022323,
+      "learning_rate": 4.8406794552643145e-05,
+      "loss": 1.4865,
+      "step": 3156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0999849881870173,
+      "learning_rate": 4.839361546346464e-05,
+      "loss": 1.477,
+      "step": 3157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09860216503988592,
+      "learning_rate": 4.838043637428614e-05,
+      "loss": 1.494,
+      "step": 3158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09641341249030705,
+      "learning_rate": 4.836725728510763e-05,
+      "loss": 1.4791,
+      "step": 3159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09825447311966723,
+      "learning_rate": 4.835407819592913e-05,
+      "loss": 1.4856,
+      "step": 3160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09175859491204413,
+      "learning_rate": 4.834089910675062e-05,
+      "loss": 1.4756,
+      "step": 3161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09445440311293929,
+      "learning_rate": 4.832772001757212e-05,
+      "loss": 1.48,
+      "step": 3162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12344318413713082,
+      "learning_rate": 4.8314540928393615e-05,
+      "loss": 1.4757,
+      "step": 3163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.10932327197823646,
+      "learning_rate": 4.8301361839215116e-05,
+      "loss": 1.4876,
+      "step": 3164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09103664268293242,
+      "learning_rate": 4.8288182750036616e-05,
+      "loss": 1.4815,
+      "step": 3165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11938514756314267,
+      "learning_rate": 4.827500366085811e-05,
+      "loss": 1.4816,
+      "step": 3166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12768058836501153,
+      "learning_rate": 4.82618245716796e-05,
+      "loss": 1.4784,
+      "step": 3167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13798988749605318,
+      "learning_rate": 4.82486454825011e-05,
+      "loss": 1.48,
+      "step": 3168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1387748279744854,
+      "learning_rate": 4.82354663933226e-05,
+      "loss": 1.4784,
+      "step": 3169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1296899337031559,
+      "learning_rate": 4.822228730414409e-05,
+      "loss": 1.4828,
+      "step": 3170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09464731397008698,
+      "learning_rate": 4.820910821496559e-05,
+      "loss": 1.4824,
+      "step": 3171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09966609987785571,
+      "learning_rate": 4.8195929125787086e-05,
+      "loss": 1.4815,
+      "step": 3172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11680368435396138,
+      "learning_rate": 4.8182750036608587e-05,
+      "loss": 1.4815,
+      "step": 3173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12363832079839647,
+      "learning_rate": 4.816957094743009e-05,
+      "loss": 1.4753,
+      "step": 3174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.09640147562551801,
+      "learning_rate": 4.8156391858251574e-05,
+      "loss": 1.4814,
+      "step": 3175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10853646217367363,
+      "learning_rate": 4.814321276907307e-05,
+      "loss": 1.4783,
+      "step": 3176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13664018353279692,
+      "learning_rate": 4.813003367989457e-05,
+      "loss": 1.4825,
+      "step": 3177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12762035468058744,
+      "learning_rate": 4.811685459071607e-05,
+      "loss": 1.4766,
+      "step": 3178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11125045479209568,
+      "learning_rate": 4.810367550153756e-05,
+      "loss": 1.4773,
+      "step": 3179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09600475593558505,
+      "learning_rate": 4.809049641235906e-05,
+      "loss": 1.4818,
+      "step": 3180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09402201515016562,
+      "learning_rate": 4.807731732318056e-05,
+      "loss": 1.4897,
+      "step": 3181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09243781259132346,
+      "learning_rate": 4.806413823400206e-05,
+      "loss": 1.4857,
+      "step": 3182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09531458105659713,
+      "learning_rate": 4.8050959144823544e-05,
+      "loss": 1.4724,
+      "step": 3183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0921157847225877,
+      "learning_rate": 4.8037780055645045e-05,
+      "loss": 1.4793,
+      "step": 3184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09263421776917255,
+      "learning_rate": 4.802460096646654e-05,
+      "loss": 1.489,
+      "step": 3185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11001457061513037,
+      "learning_rate": 4.801142187728804e-05,
+      "loss": 1.4779,
+      "step": 3186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12826485445221494,
+      "learning_rate": 4.799824278810953e-05,
+      "loss": 1.4785,
+      "step": 3187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12748872975348158,
+      "learning_rate": 4.798506369893103e-05,
+      "loss": 1.4783,
+      "step": 3188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.135373462039035,
+      "learning_rate": 4.7971884609752534e-05,
+      "loss": 1.4752,
+      "step": 3189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.13102904266363286,
+      "learning_rate": 4.795870552057403e-05,
+      "loss": 1.4789,
+      "step": 3190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11838466932933965,
+      "learning_rate": 4.7945526431395515e-05,
+      "loss": 1.4682,
+      "step": 3191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12620775732847284,
+      "learning_rate": 4.7932347342217015e-05,
+      "loss": 1.484,
+      "step": 3192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11012609819289197,
+      "learning_rate": 4.7919168253038516e-05,
+      "loss": 1.4755,
+      "step": 3193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09867750490861389,
+      "learning_rate": 4.790598916386001e-05,
+      "loss": 1.4847,
+      "step": 3194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10623889456374866,
+      "learning_rate": 4.789281007468151e-05,
+      "loss": 1.4927,
+      "step": 3195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10329244472563599,
+      "learning_rate": 4.7879630985503004e-05,
+      "loss": 1.476,
+      "step": 3196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1067475556019714,
+      "learning_rate": 4.7866451896324504e-05,
+      "loss": 1.4804,
+      "step": 3197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11093224292498935,
+      "learning_rate": 4.7853272807146005e-05,
+      "loss": 1.4837,
+      "step": 3198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0986108642025366,
+      "learning_rate": 4.784009371796749e-05,
+      "loss": 1.4745,
+      "step": 3199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10664969086590609,
+      "learning_rate": 4.7826914628788985e-05,
+      "loss": 1.4803,
+      "step": 3200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11164165014171445,
+      "learning_rate": 4.7813735539610486e-05,
+      "loss": 1.483,
+      "step": 3201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09531533922281943,
+      "learning_rate": 4.7800556450431986e-05,
+      "loss": 1.4765,
+      "step": 3202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10104940928810928,
+      "learning_rate": 4.778737736125348e-05,
+      "loss": 1.4756,
+      "step": 3203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10042267577362726,
+      "learning_rate": 4.777419827207498e-05,
+      "loss": 1.474,
+      "step": 3204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10936200415632664,
+      "learning_rate": 4.7761019182896474e-05,
+      "loss": 1.4771,
+      "step": 3205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1074770924562413,
+      "learning_rate": 4.7747840093717975e-05,
+      "loss": 1.4886,
+      "step": 3206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10252930887547491,
+      "learning_rate": 4.773466100453946e-05,
+      "loss": 1.4894,
+      "step": 3207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10882930375713758,
+      "learning_rate": 4.772148191536096e-05,
+      "loss": 1.4811,
+      "step": 3208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12080723768486536,
+      "learning_rate": 4.7708302826182456e-05,
+      "loss": 1.4845,
+      "step": 3209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09191871770512014,
+      "learning_rate": 4.769512373700396e-05,
+      "loss": 1.4812,
+      "step": 3210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10465056481801348,
+      "learning_rate": 4.768194464782545e-05,
+      "loss": 1.4835,
+      "step": 3211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10351246615745525,
+      "learning_rate": 4.766876555864695e-05,
+      "loss": 1.4827,
+      "step": 3212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10592654048942896,
+      "learning_rate": 4.765558646946845e-05,
+      "loss": 1.4824,
+      "step": 3213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10229823485214797,
+      "learning_rate": 4.7642407380289945e-05,
+      "loss": 1.4759,
+      "step": 3214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10891301663504581,
+      "learning_rate": 4.762922829111143e-05,
+      "loss": 1.4822,
+      "step": 3215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10298734682867774,
+      "learning_rate": 4.761604920193293e-05,
+      "loss": 1.4706,
+      "step": 3216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11469190464600183,
+      "learning_rate": 4.760287011275443e-05,
+      "loss": 1.4828,
+      "step": 3217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10936274347400049,
+      "learning_rate": 4.758969102357593e-05,
+      "loss": 1.4854,
+      "step": 3218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1203470143305391,
+      "learning_rate": 4.757651193439743e-05,
+      "loss": 1.4814,
+      "step": 3219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1193680575910386,
+      "learning_rate": 4.756333284521892e-05,
+      "loss": 1.4789,
+      "step": 3220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.12486938815256711,
+      "learning_rate": 4.755015375604042e-05,
+      "loss": 1.4808,
+      "step": 3221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11139837424278894,
+      "learning_rate": 4.753697466686192e-05,
+      "loss": 1.4789,
+      "step": 3222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10033682661860187,
+      "learning_rate": 4.752379557768341e-05,
+      "loss": 1.4782,
+      "step": 3223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11584206679323254,
+      "learning_rate": 4.75106164885049e-05,
+      "loss": 1.4871,
+      "step": 3224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11861388845304666,
+      "learning_rate": 4.7497437399326403e-05,
+      "loss": 1.4747,
+      "step": 3225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11138387880870616,
+      "learning_rate": 4.7484258310147904e-05,
+      "loss": 1.4811,
+      "step": 3226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10578100582954846,
+      "learning_rate": 4.74710792209694e-05,
+      "loss": 1.4792,
+      "step": 3227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10523016367731103,
+      "learning_rate": 4.74579001317909e-05,
+      "loss": 1.4862,
+      "step": 3228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09112217103244113,
+      "learning_rate": 4.744472104261239e-05,
+      "loss": 1.4898,
+      "step": 3229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10072218341617614,
+      "learning_rate": 4.743154195343389e-05,
+      "loss": 1.485,
+      "step": 3230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.100413249568223,
+      "learning_rate": 4.741836286425538e-05,
+      "loss": 1.4755,
+      "step": 3231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09813109847127319,
+      "learning_rate": 4.740518377507688e-05,
+      "loss": 1.4761,
+      "step": 3232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09581430719531937,
+      "learning_rate": 4.7392004685898374e-05,
+      "loss": 1.4711,
+      "step": 3233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09740485325649337,
+      "learning_rate": 4.7378825596719874e-05,
+      "loss": 1.488,
+      "step": 3234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11566282829914687,
+      "learning_rate": 4.736564650754137e-05,
+      "loss": 1.4838,
+      "step": 3235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11578362785932901,
+      "learning_rate": 4.735246741836287e-05,
+      "loss": 1.4893,
+      "step": 3236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10808551327710461,
+      "learning_rate": 4.733928832918437e-05,
+      "loss": 1.4802,
+      "step": 3237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11018372159461605,
+      "learning_rate": 4.732610924000586e-05,
+      "loss": 1.4811,
+      "step": 3238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10603068557647549,
+      "learning_rate": 4.731293015082735e-05,
+      "loss": 1.4892,
+      "step": 3239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.09102986814563208,
+      "learning_rate": 4.729975106164885e-05,
+      "loss": 1.4729,
+      "step": 3240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11494250887466681,
+      "learning_rate": 4.728657197247035e-05,
+      "loss": 1.4832,
+      "step": 3241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11295130761675626,
+      "learning_rate": 4.7273392883291844e-05,
+      "loss": 1.4786,
+      "step": 3242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.10551195667763467,
+      "learning_rate": 4.7260213794113345e-05,
+      "loss": 1.4784,
+      "step": 3243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11110558659154983,
+      "learning_rate": 4.724703470493484e-05,
+      "loss": 1.4886,
+      "step": 3244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10952371386144061,
+      "learning_rate": 4.723385561575634e-05,
+      "loss": 1.4776,
+      "step": 3245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10755279738002295,
+      "learning_rate": 4.722067652657784e-05,
+      "loss": 1.4747,
+      "step": 3246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09423096264775213,
+      "learning_rate": 4.720749743739933e-05,
+      "loss": 1.4767,
+      "step": 3247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09108688725075476,
+      "learning_rate": 4.719431834822082e-05,
+      "loss": 1.4707,
+      "step": 3248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11088131223827218,
+      "learning_rate": 4.718113925904232e-05,
+      "loss": 1.4823,
+      "step": 3249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11836863863549656,
+      "learning_rate": 4.7167960169863815e-05,
+      "loss": 1.4806,
+      "step": 3250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10092718890085575,
+      "learning_rate": 4.7154781080685315e-05,
+      "loss": 1.482,
+      "step": 3251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1056637675875258,
+      "learning_rate": 4.7141601991506816e-05,
+      "loss": 1.4699,
+      "step": 3252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12519733395904004,
+      "learning_rate": 4.712842290232831e-05,
+      "loss": 1.4829,
+      "step": 3253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12220591403319689,
+      "learning_rate": 4.711524381314981e-05,
+      "loss": 1.4815,
+      "step": 3254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11675271693683091,
+      "learning_rate": 4.71020647239713e-05,
+      "loss": 1.4846,
+      "step": 3255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10107820371898066,
+      "learning_rate": 4.70888856347928e-05,
+      "loss": 1.4813,
+      "step": 3256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11544145036407617,
+      "learning_rate": 4.707570654561429e-05,
+      "loss": 1.4866,
+      "step": 3257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12849550566078094,
+      "learning_rate": 4.706252745643579e-05,
+      "loss": 1.4802,
+      "step": 3258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11538748953055909,
+      "learning_rate": 4.7049348367257286e-05,
+      "loss": 1.4784,
+      "step": 3259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11424133251258727,
+      "learning_rate": 4.7036169278078786e-05,
+      "loss": 1.4748,
+      "step": 3260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1029449594446688,
+      "learning_rate": 4.7022990188900287e-05,
+      "loss": 1.475,
+      "step": 3261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09467915767103362,
+      "learning_rate": 4.700981109972178e-05,
+      "loss": 1.4751,
+      "step": 3262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1142782219051418,
+      "learning_rate": 4.699663201054327e-05,
+      "loss": 1.4804,
+      "step": 3263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10128777606230481,
+      "learning_rate": 4.698345292136477e-05,
+      "loss": 1.4795,
+      "step": 3264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09797821624231089,
+      "learning_rate": 4.697027383218627e-05,
+      "loss": 1.4769,
+      "step": 3265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10608071974815862,
+      "learning_rate": 4.695709474300776e-05,
+      "loss": 1.4734,
+      "step": 3266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0926266624502676,
+      "learning_rate": 4.694391565382926e-05,
+      "loss": 1.4704,
+      "step": 3267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09740326908519836,
+      "learning_rate": 4.6930736564650756e-05,
+      "loss": 1.4734,
+      "step": 3268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10733876814786933,
+      "learning_rate": 4.691755747547226e-05,
+      "loss": 1.4811,
+      "step": 3269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08580527577850336,
+      "learning_rate": 4.690437838629376e-05,
+      "loss": 1.4857,
+      "step": 3270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10911537746363917,
+      "learning_rate": 4.6891199297115244e-05,
+      "loss": 1.4779,
+      "step": 3271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1180016187703272,
+      "learning_rate": 4.687802020793674e-05,
+      "loss": 1.4825,
+      "step": 3272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10218586327904385,
+      "learning_rate": 4.686484111875824e-05,
+      "loss": 1.4902,
+      "step": 3273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11485853380572177,
+      "learning_rate": 4.685166202957973e-05,
+      "loss": 1.4713,
+      "step": 3274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11504203436289578,
+      "learning_rate": 4.683848294040123e-05,
+      "loss": 1.4735,
+      "step": 3275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09171548676501248,
+      "learning_rate": 4.682530385122273e-05,
+      "loss": 1.4895,
+      "step": 3276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10720078108883631,
+      "learning_rate": 4.681212476204423e-05,
+      "loss": 1.4757,
+      "step": 3277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12795860110744936,
+      "learning_rate": 4.679894567286573e-05,
+      "loss": 1.4836,
+      "step": 3278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10787500650945121,
+      "learning_rate": 4.6785766583687215e-05,
+      "loss": 1.4803,
+      "step": 3279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10185237780407037,
+      "learning_rate": 4.6772587494508715e-05,
+      "loss": 1.4757,
+      "step": 3280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.109256124384179,
+      "learning_rate": 4.675940840533021e-05,
+      "loss": 1.4786,
+      "step": 3281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.08527746540745845,
+      "learning_rate": 4.674622931615171e-05,
+      "loss": 1.4802,
+      "step": 3282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10086147230274332,
+      "learning_rate": 4.67330502269732e-05,
+      "loss": 1.4793,
+      "step": 3283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11217579268399944,
+      "learning_rate": 4.6719871137794704e-05,
+      "loss": 1.4725,
+      "step": 3284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0944747876429426,
+      "learning_rate": 4.6706692048616204e-05,
+      "loss": 1.4793,
+      "step": 3285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10750905230694908,
+      "learning_rate": 4.66935129594377e-05,
+      "loss": 1.4835,
+      "step": 3286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1103425896855546,
+      "learning_rate": 4.6680333870259185e-05,
+      "loss": 1.4755,
+      "step": 3287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0896936031292816,
+      "learning_rate": 4.6667154781080685e-05,
+      "loss": 1.4866,
+      "step": 3288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11019657074717892,
+      "learning_rate": 4.6653975691902186e-05,
+      "loss": 1.4711,
+      "step": 3289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11261601481363177,
+      "learning_rate": 4.664079660272368e-05,
+      "loss": 1.4853,
+      "step": 3290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10270184510308565,
+      "learning_rate": 4.662761751354518e-05,
+      "loss": 1.4812,
+      "step": 3291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10149041231616304,
+      "learning_rate": 4.6614438424366674e-05,
+      "loss": 1.47,
+      "step": 3292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10145233779625921,
+      "learning_rate": 4.6601259335188174e-05,
+      "loss": 1.4692,
+      "step": 3293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10318419949222454,
+      "learning_rate": 4.6588080246009675e-05,
+      "loss": 1.4751,
+      "step": 3294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11334558038833184,
+      "learning_rate": 4.657490115683116e-05,
+      "loss": 1.4835,
+      "step": 3295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10066221937417741,
+      "learning_rate": 4.6561722067652656e-05,
+      "loss": 1.4776,
+      "step": 3296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10332692750997513,
+      "learning_rate": 4.6548542978474156e-05,
+      "loss": 1.4754,
+      "step": 3297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09940580776432253,
+      "learning_rate": 4.653536388929565e-05,
+      "loss": 1.4828,
+      "step": 3298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.09944249962670595,
+      "learning_rate": 4.652218480011715e-05,
+      "loss": 1.475,
+      "step": 3299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10936798902324392,
+      "learning_rate": 4.650900571093865e-05,
+      "loss": 1.4781,
+      "step": 3300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.12672502026045085,
+      "learning_rate": 4.6495826621760145e-05,
+      "loss": 1.4824,
+      "step": 3301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10439200701485132,
+      "learning_rate": 4.6482647532581645e-05,
+      "loss": 1.4718,
+      "step": 3302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11575318136253052,
+      "learning_rate": 4.646946844340313e-05,
+      "loss": 1.4758,
+      "step": 3303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11648066390538812,
+      "learning_rate": 4.645628935422463e-05,
+      "loss": 1.486,
+      "step": 3304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.13240125720951518,
+      "learning_rate": 4.6443110265046126e-05,
+      "loss": 1.481,
+      "step": 3305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1388002079984198,
+      "learning_rate": 4.642993117586763e-05,
+      "loss": 1.4777,
+      "step": 3306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.11900705729193184,
+      "learning_rate": 4.641675208668912e-05,
+      "loss": 1.4802,
+      "step": 3307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10368840720208806,
+      "learning_rate": 4.640357299751062e-05,
+      "loss": 1.4775,
+      "step": 3308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.10331723809666844,
+      "learning_rate": 4.639039390833212e-05,
+      "loss": 1.4785,
+      "step": 3309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14663422202826054,
+      "learning_rate": 4.6377214819153615e-05,
+      "loss": 1.4787,
+      "step": 3310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.15702788873999843,
+      "learning_rate": 4.63640357299751e-05,
+      "loss": 1.4792,
+      "step": 3311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.14064221647717315,
+      "learning_rate": 4.63508566407966e-05,
+      "loss": 1.4755,
+      "step": 3312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10299279111394323,
+      "learning_rate": 4.6337677551618103e-05,
+      "loss": 1.4777,
+      "step": 3313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10945073582612996,
+      "learning_rate": 4.63244984624396e-05,
+      "loss": 1.4725,
+      "step": 3314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.15034747095268505,
+      "learning_rate": 4.63113193732611e-05,
+      "loss": 1.4798,
+      "step": 3315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.13198249006351517,
+      "learning_rate": 4.629814028408259e-05,
+      "loss": 1.4685,
+      "step": 3316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09540427042049546,
+      "learning_rate": 4.628496119490409e-05,
+      "loss": 1.4758,
+      "step": 3317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09680397943406296,
+      "learning_rate": 4.627178210572559e-05,
+      "loss": 1.4755,
+      "step": 3318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11397553390678308,
+      "learning_rate": 4.625860301654708e-05,
+      "loss": 1.4819,
+      "step": 3319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12145665557489105,
+      "learning_rate": 4.624542392736857e-05,
+      "loss": 1.4764,
+      "step": 3320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09634629249049059,
+      "learning_rate": 4.6232244838190074e-05,
+      "loss": 1.4785,
+      "step": 3321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09745283706977684,
+      "learning_rate": 4.621906574901157e-05,
+      "loss": 1.4808,
+      "step": 3322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11618566725337726,
+      "learning_rate": 4.620588665983307e-05,
+      "loss": 1.478,
+      "step": 3323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.13913190078003895,
+      "learning_rate": 4.619270757065457e-05,
+      "loss": 1.478,
+      "step": 3324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12622363146991142,
+      "learning_rate": 4.617952848147606e-05,
+      "loss": 1.4799,
+      "step": 3325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09271286785785915,
+      "learning_rate": 4.616634939229756e-05,
+      "loss": 1.4818,
+      "step": 3326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1159007394018209,
+      "learning_rate": 4.615317030311905e-05,
+      "loss": 1.4736,
+      "step": 3327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11380338317597227,
+      "learning_rate": 4.613999121394055e-05,
+      "loss": 1.4717,
+      "step": 3328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10557449186713147,
+      "learning_rate": 4.6126812124762044e-05,
+      "loss": 1.474,
+      "step": 3329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10618164605696866,
+      "learning_rate": 4.6113633035583544e-05,
+      "loss": 1.4797,
+      "step": 3330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09329340600563593,
+      "learning_rate": 4.610045394640504e-05,
+      "loss": 1.4765,
+      "step": 3331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0984726343718454,
+      "learning_rate": 4.608727485722654e-05,
+      "loss": 1.4848,
+      "step": 3332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11273235194666974,
+      "learning_rate": 4.607409576804804e-05,
+      "loss": 1.4773,
+      "step": 3333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10292721274167747,
+      "learning_rate": 4.606091667886953e-05,
+      "loss": 1.4819,
+      "step": 3334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10979034323100247,
+      "learning_rate": 4.604773758969102e-05,
+      "loss": 1.4742,
+      "step": 3335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12118650319089747,
+      "learning_rate": 4.603455850051252e-05,
+      "loss": 1.4739,
+      "step": 3336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11700685361980508,
+      "learning_rate": 4.6021379411334014e-05,
+      "loss": 1.4815,
+      "step": 3337
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09955691658831177,
+      "learning_rate": 4.6008200322155515e-05,
+      "loss": 1.474,
+      "step": 3338
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12085604495036634,
+      "learning_rate": 4.5995021232977015e-05,
+      "loss": 1.4792,
+      "step": 3339
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.14125362685790047,
+      "learning_rate": 4.598184214379851e-05,
+      "loss": 1.4726,
+      "step": 3340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12589466971628638,
+      "learning_rate": 4.596866305462001e-05,
+      "loss": 1.4799,
+      "step": 3341
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11235699550159596,
+      "learning_rate": 4.59554839654415e-05,
+      "loss": 1.4806,
+      "step": 3342
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1036460528710127,
+      "learning_rate": 4.5942304876263e-05,
+      "loss": 1.4783,
+      "step": 3343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09879570507282796,
+      "learning_rate": 4.592912578708449e-05,
+      "loss": 1.4696,
+      "step": 3344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10618227075496174,
+      "learning_rate": 4.591594669790599e-05,
+      "loss": 1.474,
+      "step": 3345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09907981421417571,
+      "learning_rate": 4.5902767608727485e-05,
+      "loss": 1.4748,
+      "step": 3346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11550340014658293,
+      "learning_rate": 4.5889588519548986e-05,
+      "loss": 1.4718,
+      "step": 3347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1218040192427924,
+      "learning_rate": 4.5876409430370486e-05,
+      "loss": 1.4739,
+      "step": 3348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1042366760775742,
+      "learning_rate": 4.586323034119198e-05,
+      "loss": 1.4634,
+      "step": 3349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10131068538908217,
+      "learning_rate": 4.585005125201348e-05,
+      "loss": 1.4714,
+      "step": 3350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10470661301884261,
+      "learning_rate": 4.583687216283497e-05,
+      "loss": 1.4713,
+      "step": 3351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09561952649041397,
+      "learning_rate": 4.582369307365647e-05,
+      "loss": 1.4762,
+      "step": 3352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11336288216074543,
+      "learning_rate": 4.581051398447796e-05,
+      "loss": 1.4709,
+      "step": 3353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1167599323835074,
+      "learning_rate": 4.579733489529946e-05,
+      "loss": 1.4674,
+      "step": 3354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10395537291594448,
+      "learning_rate": 4.5784155806120956e-05,
+      "loss": 1.4772,
+      "step": 3355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1005855170855811,
+      "learning_rate": 4.5770976716942456e-05,
+      "loss": 1.4812,
+      "step": 3356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11112319758416113,
+      "learning_rate": 4.575779762776396e-05,
+      "loss": 1.473,
+      "step": 3357
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11132040102516061,
+      "learning_rate": 4.574461853858545e-05,
+      "loss": 1.4782,
+      "step": 3358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10628464988780016,
+      "learning_rate": 4.573143944940694e-05,
+      "loss": 1.4718,
+      "step": 3359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11843786356280744,
+      "learning_rate": 4.571826036022844e-05,
+      "loss": 1.4782,
+      "step": 3360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1049201130080045,
+      "learning_rate": 4.570508127104993e-05,
+      "loss": 1.4773,
+      "step": 3361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.12009429151778062,
+      "learning_rate": 4.569190218187143e-05,
+      "loss": 1.4748,
+      "step": 3362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09857192497877801,
+      "learning_rate": 4.567872309269293e-05,
+      "loss": 1.4805,
+      "step": 3363
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10674419632740345,
+      "learning_rate": 4.5665544003514427e-05,
+      "loss": 1.482,
+      "step": 3364
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10050162376057341,
+      "learning_rate": 4.565236491433593e-05,
+      "loss": 1.4778,
+      "step": 3365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0890840865351882,
+      "learning_rate": 4.563918582515742e-05,
+      "loss": 1.4794,
+      "step": 3366
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10068011790604842,
+      "learning_rate": 4.5626006735978915e-05,
+      "loss": 1.4808,
+      "step": 3367
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10142850402148403,
+      "learning_rate": 4.561282764680041e-05,
+      "loss": 1.4819,
+      "step": 3368
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.10480996673214668,
+      "learning_rate": 4.559964855762191e-05,
+      "loss": 1.4734,
+      "step": 3369
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11374958074285867,
+      "learning_rate": 4.55864694684434e-05,
+      "loss": 1.4807,
+      "step": 3370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.11627587542946168,
+      "learning_rate": 4.55732903792649e-05,
+      "loss": 1.4724,
+      "step": 3371
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09418056341723205,
+      "learning_rate": 4.5560111290086404e-05,
+      "loss": 1.476,
+      "step": 3372
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08830397209975521,
+      "learning_rate": 4.55469322009079e-05,
+      "loss": 1.4725,
+      "step": 3373
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08964495568699472,
+      "learning_rate": 4.55337531117294e-05,
+      "loss": 1.473,
+      "step": 3374
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08634027874587423,
+      "learning_rate": 4.5520574022550885e-05,
+      "loss": 1.4916,
+      "step": 3375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08471849821302445,
+      "learning_rate": 4.5507394933372385e-05,
+      "loss": 1.4827,
+      "step": 3376
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08424747397657703,
+      "learning_rate": 4.549421584419388e-05,
+      "loss": 1.477,
+      "step": 3377
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.08901014214254244,
+      "learning_rate": 4.548103675501538e-05,
+      "loss": 1.4771,
+      "step": 3378
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09168148630258029,
+      "learning_rate": 4.546785766583687e-05,
+      "loss": 1.4792,
+      "step": 3379
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09618241395003076,
+      "learning_rate": 4.5454678576658374e-05,
+      "loss": 1.4846,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09155537003950347,
+      "learning_rate": 4.5441499487479874e-05,
+      "loss": 1.4709,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10634705586446547,
+      "learning_rate": 4.542832039830137e-05,
+      "loss": 1.4772,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10040843638353786,
+      "learning_rate": 4.5415141309122855e-05,
+      "loss": 1.4793,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11088400991851863,
+      "learning_rate": 4.5401962219944356e-05,
+      "loss": 1.4852,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11108910470842344,
+      "learning_rate": 4.538878313076585e-05,
+      "loss": 1.4764,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10725501981447991,
+      "learning_rate": 4.537560404158735e-05,
+      "loss": 1.4736,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09147274631917804,
+      "learning_rate": 4.536242495240885e-05,
+      "loss": 1.4708,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0877289030471494,
+      "learning_rate": 4.5349245863230344e-05,
+      "loss": 1.483,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08905133971061531,
+      "learning_rate": 4.5336066774051845e-05,
+      "loss": 1.4773,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09969446615014844,
+      "learning_rate": 4.532288768487334e-05,
+      "loss": 1.481,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10747473753353924,
+      "learning_rate": 4.530970859569483e-05,
+      "loss": 1.4811,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10253924935991407,
+      "learning_rate": 4.5296529506516326e-05,
+      "loss": 1.4819,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09117882037205635,
+      "learning_rate": 4.5283350417337826e-05,
+      "loss": 1.4875,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09287429434057667,
+      "learning_rate": 4.527017132815932e-05,
+      "loss": 1.4802,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11106670529326074,
+      "learning_rate": 4.525699223898082e-05,
+      "loss": 1.4661,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11393260885803627,
+      "learning_rate": 4.524381314980232e-05,
+      "loss": 1.4783,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11319527496259497,
+      "learning_rate": 4.5230634060623815e-05,
+      "loss": 1.4731,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10778257271926264,
+      "learning_rate": 4.5217454971445315e-05,
+      "loss": 1.4801,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09669403729932774,
+      "learning_rate": 4.52042758822668e-05,
+      "loss": 1.4794,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08759385341709595,
+      "learning_rate": 4.51910967930883e-05,
+      "loss": 1.4787,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10394738502575089,
+      "learning_rate": 4.51779177039098e-05,
+      "loss": 1.4788,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1019188929571694,
+      "learning_rate": 4.51647386147313e-05,
+      "loss": 1.4721,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1133374363008348,
+      "learning_rate": 4.515155952555279e-05,
+      "loss": 1.479,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1304122533477421,
+      "learning_rate": 4.513838043637429e-05,
+      "loss": 1.4736,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.13083270263255814,
+      "learning_rate": 4.512520134719579e-05,
+      "loss": 1.4747,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10478918779573601,
+      "learning_rate": 4.5112022258017286e-05,
+      "loss": 1.4766,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09791677405523837,
+      "learning_rate": 4.509884316883877e-05,
+      "loss": 1.4773,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.12315973239997448,
+      "learning_rate": 4.508566407966027e-05,
+      "loss": 1.4686,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.15094844253033543,
+      "learning_rate": 4.507248499048177e-05,
+      "loss": 1.4721,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11894404053960556,
+      "learning_rate": 4.505930590130327e-05,
+      "loss": 1.4801,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09230648374035921,
+      "learning_rate": 4.504612681212477e-05,
+      "loss": 1.4752,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10571930945167998,
+      "learning_rate": 4.503294772294626e-05,
+      "loss": 1.4639,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1119978094471148,
+      "learning_rate": 4.501976863376776e-05,
+      "loss": 1.4746,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10208312406264797,
+      "learning_rate": 4.5006589544589256e-05,
+      "loss": 1.4776,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09650817574671781,
+      "learning_rate": 4.499341045541075e-05,
+      "loss": 1.4794,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10170486210186566,
+      "learning_rate": 4.498023136623225e-05,
+      "loss": 1.4774,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10375988707621675,
+      "learning_rate": 4.4967052277053744e-05,
+      "loss": 1.4773,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09035304247418476,
+      "learning_rate": 4.495387318787524e-05,
+      "loss": 1.4771,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11258001951958191,
+      "learning_rate": 4.494069409869674e-05,
+      "loss": 1.4721,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09764562906349454,
+      "learning_rate": 4.492751500951824e-05,
+      "loss": 1.4779,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10028094203792484,
+      "learning_rate": 4.4914335920339726e-05,
+      "loss": 1.4673,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10150113920159093,
+      "learning_rate": 4.4901156831161226e-05,
+      "loss": 1.4749,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09425614183188119,
+      "learning_rate": 4.488797774198273e-05,
+      "loss": 1.4784,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10294151846143561,
+      "learning_rate": 4.487479865280422e-05,
+      "loss": 1.4719,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11059842865207657,
+      "learning_rate": 4.4861619563625714e-05,
+      "loss": 1.4752,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09398447703315485,
+      "learning_rate": 4.4848440474447215e-05,
+      "loss": 1.4722,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09882127451958048,
+      "learning_rate": 4.483526138526871e-05,
+      "loss": 1.4799,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.13906444661374448,
+      "learning_rate": 4.482208229609021e-05,
+      "loss": 1.4786,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.132390102635166,
+      "learning_rate": 4.48089032069117e-05,
+      "loss": 1.4792,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1154502951589162,
+      "learning_rate": 4.4795724117733196e-05,
+      "loss": 1.4723,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10764584092267204,
+      "learning_rate": 4.47825450285547e-05,
+      "loss": 1.4765,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.12090163810330704,
+      "learning_rate": 4.47693659393762e-05,
+      "loss": 1.4762,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.12574139519141525,
+      "learning_rate": 4.4756186850197684e-05,
+      "loss": 1.474,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.13571557601798295,
+      "learning_rate": 4.4743007761019185e-05,
+      "loss": 1.4843,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.14149248635779987,
+      "learning_rate": 4.4729828671840686e-05,
+      "loss": 1.4799,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1153797644919215,
+      "learning_rate": 4.471664958266218e-05,
+      "loss": 1.4712,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09880944334741432,
+      "learning_rate": 4.470347049348367e-05,
+      "loss": 1.4725,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.11824532023024314,
+      "learning_rate": 4.4690291404305174e-05,
+      "loss": 1.4811,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1161492437373776,
+      "learning_rate": 4.467711231512667e-05,
+      "loss": 1.4767,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09086459365015795,
+      "learning_rate": 4.466393322594817e-05,
+      "loss": 1.4779,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09315302100143384,
+      "learning_rate": 4.465075413676966e-05,
+      "loss": 1.4803,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.08967099452628344,
+      "learning_rate": 4.4637575047591155e-05,
+      "loss": 1.4668,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09103320241788868,
+      "learning_rate": 4.4624395958412656e-05,
+      "loss": 1.4726,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09832523510017122,
+      "learning_rate": 4.4611216869234156e-05,
+      "loss": 1.4755,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09958513661658622,
+      "learning_rate": 4.459803778005564e-05,
+      "loss": 1.4739,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.10864585070517314,
+      "learning_rate": 4.4584858690877144e-05,
+      "loss": 1.476,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09590414692941707,
+      "learning_rate": 4.4571679601698644e-05,
+      "loss": 1.4775,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.09380741771446086,
+      "learning_rate": 4.455850051252014e-05,
+      "loss": 1.4752,
+      "step": 3448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1130135771734522,
+      "learning_rate": 4.454532142334163e-05,
+      "loss": 1.4824,
+      "step": 3449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10844258755304766,
+      "learning_rate": 4.453214233416313e-05,
+      "loss": 1.4792,
+      "step": 3450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10019911522633734,
+      "learning_rate": 4.4518963244984626e-05,
+      "loss": 1.4692,
+      "step": 3451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09233356136812182,
+      "learning_rate": 4.4505784155806127e-05,
+      "loss": 1.4746,
+      "step": 3452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10241763754145078,
+      "learning_rate": 4.449260506662762e-05,
+      "loss": 1.4721,
+      "step": 3453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1017889671102716,
+      "learning_rate": 4.4479425977449114e-05,
+      "loss": 1.4746,
+      "step": 3454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09401719388679293,
+      "learning_rate": 4.4466246888270615e-05,
+      "loss": 1.4673,
+      "step": 3455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.12277981724119107,
+      "learning_rate": 4.4453067799092115e-05,
+      "loss": 1.474,
+      "step": 3456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.121312785852226,
+      "learning_rate": 4.44398887099136e-05,
+      "loss": 1.4743,
+      "step": 3457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10130099222334792,
+      "learning_rate": 4.44267096207351e-05,
+      "loss": 1.4775,
+      "step": 3458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09407281948354373,
+      "learning_rate": 4.44135305315566e-05,
+      "loss": 1.4665,
+      "step": 3459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09840017001554138,
+      "learning_rate": 4.44003514423781e-05,
+      "loss": 1.4687,
+      "step": 3460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08979960578661009,
+      "learning_rate": 4.438717235319959e-05,
+      "loss": 1.479,
+      "step": 3461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09233703543382688,
+      "learning_rate": 4.437399326402109e-05,
+      "loss": 1.4726,
+      "step": 3462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09316988877111991,
+      "learning_rate": 4.4360814174842585e-05,
+      "loss": 1.474,
+      "step": 3463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09236859150291757,
+      "learning_rate": 4.4347635085664085e-05,
+      "loss": 1.4766,
+      "step": 3464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08728063112672506,
+      "learning_rate": 4.433445599648558e-05,
+      "loss": 1.4796,
+      "step": 3465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09059192514462994,
+      "learning_rate": 4.432127690730707e-05,
+      "loss": 1.4757,
+      "step": 3466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08767494853969984,
+      "learning_rate": 4.430809781812857e-05,
+      "loss": 1.4723,
+      "step": 3467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09405181384977164,
+      "learning_rate": 4.4294918728950074e-05,
+      "loss": 1.4723,
+      "step": 3468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09853840759494308,
+      "learning_rate": 4.428173963977156e-05,
+      "loss": 1.4852,
+      "step": 3469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10089471521921023,
+      "learning_rate": 4.426856055059306e-05,
+      "loss": 1.4805,
+      "step": 3470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09215300999017471,
+      "learning_rate": 4.425538146141456e-05,
+      "loss": 1.4704,
+      "step": 3471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10878127045204569,
+      "learning_rate": 4.4242202372236056e-05,
+      "loss": 1.4791,
+      "step": 3472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10258956125379405,
+      "learning_rate": 4.422902328305755e-05,
+      "loss": 1.4688,
+      "step": 3473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09644979819888333,
+      "learning_rate": 4.421584419387905e-05,
+      "loss": 1.4849,
+      "step": 3474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09671125014968235,
+      "learning_rate": 4.4202665104700544e-05,
+      "loss": 1.4678,
+      "step": 3475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10518331455294133,
+      "learning_rate": 4.4189486015522044e-05,
+      "loss": 1.4752,
+      "step": 3476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.122808214228434,
+      "learning_rate": 4.417630692634354e-05,
+      "loss": 1.4772,
+      "step": 3477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.13250527569009143,
+      "learning_rate": 4.416312783716503e-05,
+      "loss": 1.4739,
+      "step": 3478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11388619778929629,
+      "learning_rate": 4.414994874798653e-05,
+      "loss": 1.4718,
+      "step": 3479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10136730384175127,
+      "learning_rate": 4.413676965880803e-05,
+      "loss": 1.4699,
+      "step": 3480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09415147082829967,
+      "learning_rate": 4.412359056962952e-05,
+      "loss": 1.4685,
+      "step": 3481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10611680156863319,
+      "learning_rate": 4.411041148045102e-05,
+      "loss": 1.4756,
+      "step": 3482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09866105429207099,
+      "learning_rate": 4.409723239127252e-05,
+      "loss": 1.4705,
+      "step": 3483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0796366821681927,
+      "learning_rate": 4.4084053302094014e-05,
+      "loss": 1.4779,
+      "step": 3484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09252832743904886,
+      "learning_rate": 4.407087421291551e-05,
+      "loss": 1.4694,
+      "step": 3485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09925224993937147,
+      "learning_rate": 4.405769512373701e-05,
+      "loss": 1.4694,
+      "step": 3486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08782646940890683,
+      "learning_rate": 4.40445160345585e-05,
+      "loss": 1.482,
+      "step": 3487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09015028077851482,
+      "learning_rate": 4.403133694538e-05,
+      "loss": 1.4793,
+      "step": 3488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10694848708864352,
+      "learning_rate": 4.40181578562015e-05,
+      "loss": 1.4722,
+      "step": 3489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10371733642595297,
+      "learning_rate": 4.400497876702299e-05,
+      "loss": 1.4722,
+      "step": 3490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11597641356271447,
+      "learning_rate": 4.399179967784449e-05,
+      "loss": 1.466,
+      "step": 3491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.12303219670604801,
+      "learning_rate": 4.397862058866599e-05,
+      "loss": 1.4641,
+      "step": 3492
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.12430630910825884,
+      "learning_rate": 4.396544149948748e-05,
+      "loss": 1.4667,
+      "step": 3493
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0980953672375995,
+      "learning_rate": 4.395226241030898e-05,
+      "loss": 1.4696,
+      "step": 3494
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08636767703799336,
+      "learning_rate": 4.393908332113048e-05,
+      "loss": 1.4774,
+      "step": 3495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.102617725120901,
+      "learning_rate": 4.392590423195197e-05,
+      "loss": 1.4844,
+      "step": 3496
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10522086512418981,
+      "learning_rate": 4.391272514277347e-05,
+      "loss": 1.4763,
+      "step": 3497
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09894879855360653,
+      "learning_rate": 4.389954605359497e-05,
+      "loss": 1.4691,
+      "step": 3498
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10579517299604888,
+      "learning_rate": 4.388636696441646e-05,
+      "loss": 1.4693,
+      "step": 3499
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10421210435143383,
+      "learning_rate": 4.387318787523796e-05,
+      "loss": 1.4717,
+      "step": 3500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.085522556133978,
+      "learning_rate": 4.3860008786059455e-05,
+      "loss": 1.4785,
+      "step": 3501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10936543703037395,
+      "learning_rate": 4.384682969688095e-05,
+      "loss": 1.4735,
+      "step": 3502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11743472204869046,
+      "learning_rate": 4.383365060770245e-05,
+      "loss": 1.4735,
+      "step": 3503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.08775007027536821,
+      "learning_rate": 4.3820471518523943e-05,
+      "loss": 1.4803,
+      "step": 3504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09683893988409736,
+      "learning_rate": 4.380729242934544e-05,
+      "loss": 1.4778,
+      "step": 3505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10043124107025511,
+      "learning_rate": 4.379411334016694e-05,
+      "loss": 1.4685,
+      "step": 3506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11021582458192951,
+      "learning_rate": 4.378093425098844e-05,
+      "loss": 1.4626,
+      "step": 3507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.13281907882781138,
+      "learning_rate": 4.376775516180993e-05,
+      "loss": 1.467,
+      "step": 3508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.13166651751625358,
+      "learning_rate": 4.3754576072631426e-05,
+      "loss": 1.4732,
+      "step": 3509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11288613987857332,
+      "learning_rate": 4.3741396983452926e-05,
+      "loss": 1.4634,
+      "step": 3510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.11293443419982177,
+      "learning_rate": 4.372821789427442e-05,
+      "loss": 1.4766,
+      "step": 3511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1540681675938617,
+      "learning_rate": 4.371503880509592e-05,
+      "loss": 1.4618,
+      "step": 3512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1227205983761109,
+      "learning_rate": 4.3701859715917414e-05,
+      "loss": 1.4673,
+      "step": 3513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.09707737744087955,
+      "learning_rate": 4.368868062673891e-05,
+      "loss": 1.4808,
+      "step": 3514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.13452863270835638,
+      "learning_rate": 4.367550153756041e-05,
+      "loss": 1.469,
+      "step": 3515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.15273817236930584,
+      "learning_rate": 4.36623224483819e-05,
+      "loss": 1.4667,
+      "step": 3516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.10222554238573596,
+      "learning_rate": 4.3649143359203396e-05,
+      "loss": 1.4715,
+      "step": 3517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09865482988096778,
+      "learning_rate": 4.3635964270024896e-05,
+      "loss": 1.4718,
+      "step": 3518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11705458268492745,
+      "learning_rate": 4.36227851808464e-05,
+      "loss": 1.4726,
+      "step": 3519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09944473165143827,
+      "learning_rate": 4.360960609166789e-05,
+      "loss": 1.4817,
+      "step": 3520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08723355542749019,
+      "learning_rate": 4.3596427002489384e-05,
+      "loss": 1.4811,
+      "step": 3521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09164155470292941,
+      "learning_rate": 4.3583247913310885e-05,
+      "loss": 1.4817,
+      "step": 3522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09381808196902729,
+      "learning_rate": 4.357006882413238e-05,
+      "loss": 1.465,
+      "step": 3523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09222481693571817,
+      "learning_rate": 4.355688973495388e-05,
+      "loss": 1.4761,
+      "step": 3524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10284655926559037,
+      "learning_rate": 4.354371064577537e-05,
+      "loss": 1.473,
+      "step": 3525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09108893271417359,
+      "learning_rate": 4.353053155659687e-05,
+      "loss": 1.4662,
+      "step": 3526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10281336257698209,
+      "learning_rate": 4.351735246741837e-05,
+      "loss": 1.4716,
+      "step": 3527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11359593101188963,
+      "learning_rate": 4.350417337823986e-05,
+      "loss": 1.4674,
+      "step": 3528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10512998725162467,
+      "learning_rate": 4.3490994289061355e-05,
+      "loss": 1.4723,
+      "step": 3529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09236109774248652,
+      "learning_rate": 4.3477815199882855e-05,
+      "loss": 1.47,
+      "step": 3530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11454644776673342,
+      "learning_rate": 4.3464636110704356e-05,
+      "loss": 1.4658,
+      "step": 3531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1031091386758919,
+      "learning_rate": 4.345145702152585e-05,
+      "loss": 1.4674,
+      "step": 3532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.106479675712738,
+      "learning_rate": 4.343827793234734e-05,
+      "loss": 1.4731,
+      "step": 3533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10507810761331945,
+      "learning_rate": 4.3425098843168844e-05,
+      "loss": 1.4741,
+      "step": 3534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10667714354281968,
+      "learning_rate": 4.341191975399034e-05,
+      "loss": 1.4805,
+      "step": 3535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09161838712832583,
+      "learning_rate": 4.339874066481184e-05,
+      "loss": 1.4714,
+      "step": 3536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10663244747356798,
+      "learning_rate": 4.338556157563333e-05,
+      "loss": 1.4743,
+      "step": 3537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10797214170685232,
+      "learning_rate": 4.3372382486454826e-05,
+      "loss": 1.4702,
+      "step": 3538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08802118143253278,
+      "learning_rate": 4.3359203397276326e-05,
+      "loss": 1.4757,
+      "step": 3539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1042537897405221,
+      "learning_rate": 4.334602430809782e-05,
+      "loss": 1.4692,
+      "step": 3540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09142985865609306,
+      "learning_rate": 4.3332845218919314e-05,
+      "loss": 1.4628,
+      "step": 3541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08702010290440296,
+      "learning_rate": 4.3319666129740814e-05,
+      "loss": 1.4678,
+      "step": 3542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08389124221561547,
+      "learning_rate": 4.3306487040562315e-05,
+      "loss": 1.4797,
+      "step": 3543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09051662756611933,
+      "learning_rate": 4.329330795138381e-05,
+      "loss": 1.4674,
+      "step": 3544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0919031900720737,
+      "learning_rate": 4.32801288622053e-05,
+      "loss": 1.4736,
+      "step": 3545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09379315262072051,
+      "learning_rate": 4.32669497730268e-05,
+      "loss": 1.4727,
+      "step": 3546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0888892797353294,
+      "learning_rate": 4.3253770683848296e-05,
+      "loss": 1.475,
+      "step": 3547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08215999820854653,
+      "learning_rate": 4.32405915946698e-05,
+      "loss": 1.4756,
+      "step": 3548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09301194581240517,
+      "learning_rate": 4.322741250549129e-05,
+      "loss": 1.4715,
+      "step": 3549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08199343491057857,
+      "learning_rate": 4.3214233416312784e-05,
+      "loss": 1.4766,
+      "step": 3550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08571417015313777,
+      "learning_rate": 4.3201054327134285e-05,
+      "loss": 1.478,
+      "step": 3551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10732823094954726,
+      "learning_rate": 4.318787523795578e-05,
+      "loss": 1.4666,
+      "step": 3552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10880422296138018,
+      "learning_rate": 4.317469614877727e-05,
+      "loss": 1.4723,
+      "step": 3553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1013726272892448,
+      "learning_rate": 4.316151705959877e-05,
+      "loss": 1.4687,
+      "step": 3554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11156342489371512,
+      "learning_rate": 4.314833797042027e-05,
+      "loss": 1.4838,
+      "step": 3555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11105708599280344,
+      "learning_rate": 4.313515888124177e-05,
+      "loss": 1.4772,
+      "step": 3556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10448188933876922,
+      "learning_rate": 4.312197979206326e-05,
+      "loss": 1.4625,
+      "step": 3557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10068413485288008,
+      "learning_rate": 4.310880070288476e-05,
+      "loss": 1.4774,
+      "step": 3558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08468788029368693,
+      "learning_rate": 4.3095621613706255e-05,
+      "loss": 1.4722,
+      "step": 3559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10457369681979863,
+      "learning_rate": 4.3082442524527756e-05,
+      "loss": 1.4696,
+      "step": 3560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11121716619788886,
+      "learning_rate": 4.306926343534925e-05,
+      "loss": 1.4688,
+      "step": 3561
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09938039898385992,
+      "learning_rate": 4.305608434617074e-05,
+      "loss": 1.4733,
+      "step": 3562
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09539227299315968,
+      "learning_rate": 4.3042905256992244e-05,
+      "loss": 1.4824,
+      "step": 3563
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09564334924908738,
+      "learning_rate": 4.302972616781374e-05,
+      "loss": 1.4728,
+      "step": 3564
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09257643197433613,
+      "learning_rate": 4.301654707863523e-05,
+      "loss": 1.4727,
+      "step": 3565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.101807335860489,
+      "learning_rate": 4.300336798945673e-05,
+      "loss": 1.4776,
+      "step": 3566
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10186389569453638,
+      "learning_rate": 4.299018890027823e-05,
+      "loss": 1.4743,
+      "step": 3567
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09976989391768688,
+      "learning_rate": 4.2977009811099726e-05,
+      "loss": 1.4815,
+      "step": 3568
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10706114934489873,
+      "learning_rate": 4.296383072192122e-05,
+      "loss": 1.4743,
+      "step": 3569
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1023721101312706,
+      "learning_rate": 4.295065163274272e-05,
+      "loss": 1.4694,
+      "step": 3570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0985821012616563,
+      "learning_rate": 4.2937472543564214e-05,
+      "loss": 1.469,
+      "step": 3571
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.12029073791997642,
+      "learning_rate": 4.2924293454385714e-05,
+      "loss": 1.4808,
+      "step": 3572
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.12508779830244057,
+      "learning_rate": 4.291111436520721e-05,
+      "loss": 1.4698,
+      "step": 3573
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.08643463771563238,
+      "learning_rate": 4.28979352760287e-05,
+      "loss": 1.4701,
+      "step": 3574
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10437244114378898,
+      "learning_rate": 4.28847561868502e-05,
+      "loss": 1.476,
+      "step": 3575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.13126920961882035,
+      "learning_rate": 4.2871577097671696e-05,
+      "loss": 1.4755,
+      "step": 3576
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.11256315492403032,
+      "learning_rate": 4.285839800849319e-05,
+      "loss": 1.4764,
+      "step": 3577
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09373919673336804,
+      "learning_rate": 4.284521891931469e-05,
+      "loss": 1.4751,
+      "step": 3578
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09890746123489094,
+      "learning_rate": 4.283203983013619e-05,
+      "loss": 1.4756,
+      "step": 3579
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10232789131959476,
+      "learning_rate": 4.2818860740957685e-05,
+      "loss": 1.4728,
+      "step": 3580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09167958240701786,
+      "learning_rate": 4.280568165177918e-05,
+      "loss": 1.4734,
+      "step": 3581
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.09741530401567992,
+      "learning_rate": 4.279250256260068e-05,
+      "loss": 1.4689,
+      "step": 3582
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10229993138789367,
+      "learning_rate": 4.277932347342217e-05,
+      "loss": 1.4678,
+      "step": 3583
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10478469893304154,
+      "learning_rate": 4.276614438424367e-05,
+      "loss": 1.4725,
+      "step": 3584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.10407510098530612,
+      "learning_rate": 4.275296529506517e-05,
+      "loss": 1.4702,
+      "step": 3585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10189067892590038,
+      "learning_rate": 4.273978620588666e-05,
+      "loss": 1.4697,
+      "step": 3586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10203089270189382,
+      "learning_rate": 4.272660711670816e-05,
+      "loss": 1.4777,
+      "step": 3587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09818285864086697,
+      "learning_rate": 4.2713428027529655e-05,
+      "loss": 1.4698,
+      "step": 3588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09769285560800008,
+      "learning_rate": 4.270024893835115e-05,
+      "loss": 1.4757,
+      "step": 3589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08996444356678641,
+      "learning_rate": 4.268706984917265e-05,
+      "loss": 1.4751,
+      "step": 3590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11247834946840582,
+      "learning_rate": 4.267389075999414e-05,
+      "loss": 1.4738,
+      "step": 3591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1146790465450119,
+      "learning_rate": 4.2660711670815643e-05,
+      "loss": 1.467,
+      "step": 3592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.12096456463143791,
+      "learning_rate": 4.264753258163714e-05,
+      "loss": 1.471,
+      "step": 3593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09472095430135621,
+      "learning_rate": 4.263435349245864e-05,
+      "loss": 1.4745,
+      "step": 3594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08789115754798638,
+      "learning_rate": 4.262117440328013e-05,
+      "loss": 1.4647,
+      "step": 3595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11229278357816025,
+      "learning_rate": 4.260799531410163e-05,
+      "loss": 1.4746,
+      "step": 3596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10893885689368075,
+      "learning_rate": 4.2594816224923126e-05,
+      "loss": 1.4688,
+      "step": 3597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09937958581896507,
+      "learning_rate": 4.258163713574462e-05,
+      "loss": 1.4746,
+      "step": 3598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10025046353605845,
+      "learning_rate": 4.256845804656612e-05,
+      "loss": 1.4735,
+      "step": 3599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09969388082276048,
+      "learning_rate": 4.2555278957387614e-05,
+      "loss": 1.4633,
+      "step": 3600
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.4375040531158447,
+      "eval_runtime": 453.7743,
+      "eval_samples_per_second": 39.7,
+      "eval_steps_per_second": 0.06,
+      "step": 3600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08648901792129295,
+      "learning_rate": 4.254209986820911e-05,
+      "loss": 1.478,
+      "step": 3601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1022017079164071,
+      "learning_rate": 4.252892077903061e-05,
+      "loss": 1.4736,
+      "step": 3602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09682079508637922,
+      "learning_rate": 4.25157416898521e-05,
+      "loss": 1.4696,
+      "step": 3603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09555578061101827,
+      "learning_rate": 4.25025626006736e-05,
+      "loss": 1.4714,
+      "step": 3604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10582721785749112,
+      "learning_rate": 4.2489383511495096e-05,
+      "loss": 1.4688,
+      "step": 3605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10108742141361787,
+      "learning_rate": 4.2476204422316597e-05,
+      "loss": 1.4728,
+      "step": 3606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0999345340737356,
+      "learning_rate": 4.246302533313809e-05,
+      "loss": 1.4828,
+      "step": 3607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11807779148305632,
+      "learning_rate": 4.244984624395959e-05,
+      "loss": 1.4672,
+      "step": 3608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09996560565977411,
+      "learning_rate": 4.2436667154781085e-05,
+      "loss": 1.4775,
+      "step": 3609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09692130862318266,
+      "learning_rate": 4.242348806560258e-05,
+      "loss": 1.4722,
+      "step": 3610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08955380280476279,
+      "learning_rate": 4.241030897642408e-05,
+      "loss": 1.4616,
+      "step": 3611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09143072487081126,
+      "learning_rate": 4.239712988724557e-05,
+      "loss": 1.4767,
+      "step": 3612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10438995124000315,
+      "learning_rate": 4.2383950798067066e-05,
+      "loss": 1.4772,
+      "step": 3613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10664459925428803,
+      "learning_rate": 4.237077170888857e-05,
+      "loss": 1.4776,
+      "step": 3614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09478015538668022,
+      "learning_rate": 4.235759261971006e-05,
+      "loss": 1.469,
+      "step": 3615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09929898773522321,
+      "learning_rate": 4.234441353053156e-05,
+      "loss": 1.4707,
+      "step": 3616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1030953064775713,
+      "learning_rate": 4.2331234441353055e-05,
+      "loss": 1.4709,
+      "step": 3617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11370014020137739,
+      "learning_rate": 4.2318055352174555e-05,
+      "loss": 1.4639,
+      "step": 3618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1065983998160853,
+      "learning_rate": 4.230487626299605e-05,
+      "loss": 1.466,
+      "step": 3619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1076728745944197,
+      "learning_rate": 4.229169717381755e-05,
+      "loss": 1.4726,
+      "step": 3620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09777474611468948,
+      "learning_rate": 4.227851808463904e-05,
+      "loss": 1.4696,
+      "step": 3621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0958611496003724,
+      "learning_rate": 4.226533899546054e-05,
+      "loss": 1.4536,
+      "step": 3622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10262563293259959,
+      "learning_rate": 4.225215990628204e-05,
+      "loss": 1.4729,
+      "step": 3623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10249379862007629,
+      "learning_rate": 4.223898081710353e-05,
+      "loss": 1.4664,
+      "step": 3624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08491163706317213,
+      "learning_rate": 4.2225801727925025e-05,
+      "loss": 1.4575,
+      "step": 3625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09815245417869892,
+      "learning_rate": 4.2212622638746526e-05,
+      "loss": 1.4666,
+      "step": 3626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09900302538552204,
+      "learning_rate": 4.219944354956802e-05,
+      "loss": 1.4752,
+      "step": 3627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09158111055558077,
+      "learning_rate": 4.218626446038952e-05,
+      "loss": 1.4725,
+      "step": 3628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09955926171060665,
+      "learning_rate": 4.2173085371211014e-05,
+      "loss": 1.4699,
+      "step": 3629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.07727876989273587,
+      "learning_rate": 4.2159906282032514e-05,
+      "loss": 1.4715,
+      "step": 3630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08945236714048366,
+      "learning_rate": 4.214672719285401e-05,
+      "loss": 1.4738,
+      "step": 3631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10060918236477862,
+      "learning_rate": 4.213354810367551e-05,
+      "loss": 1.469,
+      "step": 3632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10909351669197022,
+      "learning_rate": 4.2120369014497e-05,
+      "loss": 1.4705,
+      "step": 3633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09669218555434701,
+      "learning_rate": 4.2107189925318496e-05,
+      "loss": 1.4722,
+      "step": 3634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1149274163364703,
+      "learning_rate": 4.2094010836139996e-05,
+      "loss": 1.4715,
+      "step": 3635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11310767162790437,
+      "learning_rate": 4.208083174696149e-05,
+      "loss": 1.467,
+      "step": 3636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10055051796867218,
+      "learning_rate": 4.2067652657782984e-05,
+      "loss": 1.4701,
+      "step": 3637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10844421967458424,
+      "learning_rate": 4.2054473568604484e-05,
+      "loss": 1.4714,
+      "step": 3638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09732407119880346,
+      "learning_rate": 4.204129447942598e-05,
+      "loss": 1.468,
+      "step": 3639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.10659052774094743,
+      "learning_rate": 4.202811539024748e-05,
+      "loss": 1.4701,
+      "step": 3640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11000552287873529,
+      "learning_rate": 4.201493630106897e-05,
+      "loss": 1.4745,
+      "step": 3641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09008122672168044,
+      "learning_rate": 4.200175721189047e-05,
+      "loss": 1.4729,
+      "step": 3642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09222583699242914,
+      "learning_rate": 4.1988578122711967e-05,
+      "loss": 1.4785,
+      "step": 3643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09952381045869083,
+      "learning_rate": 4.197539903353347e-05,
+      "loss": 1.4717,
+      "step": 3644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.08996588610491216,
+      "learning_rate": 4.196221994435496e-05,
+      "loss": 1.4762,
+      "step": 3645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09715077150254911,
+      "learning_rate": 4.1949040855176455e-05,
+      "loss": 1.4711,
+      "step": 3646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11472910905235331,
+      "learning_rate": 4.1935861765997955e-05,
+      "loss": 1.4631,
+      "step": 3647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11541983005198843,
+      "learning_rate": 4.192268267681945e-05,
+      "loss": 1.4689,
+      "step": 3648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09621529753942798,
+      "learning_rate": 4.190950358764094e-05,
+      "loss": 1.4747,
+      "step": 3649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.09972471746294176,
+      "learning_rate": 4.189632449846244e-05,
+      "loss": 1.4754,
+      "step": 3650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11785284977366027,
+      "learning_rate": 4.188314540928394e-05,
+      "loss": 1.4712,
+      "step": 3651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.11787977121441519,
+      "learning_rate": 4.186996632010544e-05,
+      "loss": 1.474,
+      "step": 3652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1189129831948122,
+      "learning_rate": 4.185678723092693e-05,
+      "loss": 1.469,
+      "step": 3653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1047824072854484,
+      "learning_rate": 4.184360814174843e-05,
+      "loss": 1.4746,
+      "step": 3654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0995995985720023,
+      "learning_rate": 4.1830429052569925e-05,
+      "loss": 1.4755,
+      "step": 3655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10624337284435871,
+      "learning_rate": 4.1817249963391426e-05,
+      "loss": 1.4771,
+      "step": 3656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.12038003659862227,
+      "learning_rate": 4.180407087421292e-05,
+      "loss": 1.4657,
+      "step": 3657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1179183509700613,
+      "learning_rate": 4.179089178503441e-05,
+      "loss": 1.4702,
+      "step": 3658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08999519602565301,
+      "learning_rate": 4.1777712695855914e-05,
+      "loss": 1.4782,
+      "step": 3659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10562149490015743,
+      "learning_rate": 4.176453360667741e-05,
+      "loss": 1.4695,
+      "step": 3660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09445967813710512,
+      "learning_rate": 4.17513545174989e-05,
+      "loss": 1.4734,
+      "step": 3661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08840498079163638,
+      "learning_rate": 4.17381754283204e-05,
+      "loss": 1.4664,
+      "step": 3662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09271391926576064,
+      "learning_rate": 4.1724996339141896e-05,
+      "loss": 1.4653,
+      "step": 3663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09330427473495395,
+      "learning_rate": 4.1711817249963396e-05,
+      "loss": 1.4715,
+      "step": 3664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09041772173306714,
+      "learning_rate": 4.169863816078489e-05,
+      "loss": 1.4719,
+      "step": 3665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08835892512960027,
+      "learning_rate": 4.168545907160639e-05,
+      "loss": 1.4648,
+      "step": 3666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08906709756720328,
+      "learning_rate": 4.1672279982427884e-05,
+      "loss": 1.4719,
+      "step": 3667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09064640900735238,
+      "learning_rate": 4.1659100893249385e-05,
+      "loss": 1.4638,
+      "step": 3668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08672129563606673,
+      "learning_rate": 4.164592180407088e-05,
+      "loss": 1.4687,
+      "step": 3669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08797717961068938,
+      "learning_rate": 4.163274271489237e-05,
+      "loss": 1.4821,
+      "step": 3670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08488244231812088,
+      "learning_rate": 4.161956362571387e-05,
+      "loss": 1.4668,
+      "step": 3671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08731864298844223,
+      "learning_rate": 4.1606384536535366e-05,
+      "loss": 1.4723,
+      "step": 3672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09355514008613074,
+      "learning_rate": 4.159320544735686e-05,
+      "loss": 1.4641,
+      "step": 3673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09024856281498474,
+      "learning_rate": 4.158002635817836e-05,
+      "loss": 1.4673,
+      "step": 3674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1019088596838377,
+      "learning_rate": 4.1566847268999854e-05,
+      "loss": 1.4639,
+      "step": 3675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09871676589000505,
+      "learning_rate": 4.1553668179821355e-05,
+      "loss": 1.4665,
+      "step": 3676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1085070345369237,
+      "learning_rate": 4.154048909064285e-05,
+      "loss": 1.4759,
+      "step": 3677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11731688663738626,
+      "learning_rate": 4.152731000146434e-05,
+      "loss": 1.4578,
+      "step": 3678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10921384686125629,
+      "learning_rate": 4.151413091228584e-05,
+      "loss": 1.4698,
+      "step": 3679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08675170777447203,
+      "learning_rate": 4.1500951823107343e-05,
+      "loss": 1.4805,
+      "step": 3680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11579777594568938,
+      "learning_rate": 4.148777273392884e-05,
+      "loss": 1.4722,
+      "step": 3681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1044001315872451,
+      "learning_rate": 4.147459364475033e-05,
+      "loss": 1.4794,
+      "step": 3682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08832368489268708,
+      "learning_rate": 4.146141455557183e-05,
+      "loss": 1.4665,
+      "step": 3683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09593217761506204,
+      "learning_rate": 4.1448235466393325e-05,
+      "loss": 1.4724,
+      "step": 3684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10646214060807353,
+      "learning_rate": 4.143505637721482e-05,
+      "loss": 1.4743,
+      "step": 3685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09032909035838026,
+      "learning_rate": 4.142187728803632e-05,
+      "loss": 1.4685,
+      "step": 3686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09655319124682206,
+      "learning_rate": 4.140869819885781e-05,
+      "loss": 1.467,
+      "step": 3687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09489108817829071,
+      "learning_rate": 4.1395519109679314e-05,
+      "loss": 1.4704,
+      "step": 3688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.08868004664260926,
+      "learning_rate": 4.138234002050081e-05,
+      "loss": 1.4674,
+      "step": 3689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09277167069197975,
+      "learning_rate": 4.13691609313223e-05,
+      "loss": 1.4707,
+      "step": 3690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09502770859407819,
+      "learning_rate": 4.13559818421438e-05,
+      "loss": 1.4709,
+      "step": 3691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11627385628011491,
+      "learning_rate": 4.13428027529653e-05,
+      "loss": 1.4672,
+      "step": 3692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11632786771024482,
+      "learning_rate": 4.1329623663786796e-05,
+      "loss": 1.4702,
+      "step": 3693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.12123998230075475,
+      "learning_rate": 4.131644457460829e-05,
+      "loss": 1.461,
+      "step": 3694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11310647588128707,
+      "learning_rate": 4.130326548542979e-05,
+      "loss": 1.4865,
+      "step": 3695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09282906045364032,
+      "learning_rate": 4.1290086396251284e-05,
+      "loss": 1.4635,
+      "step": 3696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11116432291447755,
+      "learning_rate": 4.127690730707278e-05,
+      "loss": 1.4757,
+      "step": 3697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10795930418002754,
+      "learning_rate": 4.126372821789428e-05,
+      "loss": 1.4679,
+      "step": 3698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10867011074141553,
+      "learning_rate": 4.125054912871577e-05,
+      "loss": 1.4729,
+      "step": 3699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09463498891673043,
+      "learning_rate": 4.123737003953727e-05,
+      "loss": 1.4706,
+      "step": 3700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09584138264144786,
+      "learning_rate": 4.1224190950358766e-05,
+      "loss": 1.4718,
+      "step": 3701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11913362323614457,
+      "learning_rate": 4.121101186118026e-05,
+      "loss": 1.4677,
+      "step": 3702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.11844915465407713,
+      "learning_rate": 4.119783277200176e-05,
+      "loss": 1.4602,
+      "step": 3703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10725626695329547,
+      "learning_rate": 4.118465368282326e-05,
+      "loss": 1.465,
+      "step": 3704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10018314521965994,
+      "learning_rate": 4.1171474593644755e-05,
+      "loss": 1.4652,
+      "step": 3705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09660796504440534,
+      "learning_rate": 4.115829550446625e-05,
+      "loss": 1.4626,
+      "step": 3706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09278848593525377,
+      "learning_rate": 4.114511641528775e-05,
+      "loss": 1.4657,
+      "step": 3707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09663749246671466,
+      "learning_rate": 4.113193732610924e-05,
+      "loss": 1.4644,
+      "step": 3708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10201854007178347,
+      "learning_rate": 4.1118758236930737e-05,
+      "loss": 1.467,
+      "step": 3709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09918319878759886,
+      "learning_rate": 4.110557914775224e-05,
+      "loss": 1.4702,
+      "step": 3710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10410791968982307,
+      "learning_rate": 4.109240005857373e-05,
+      "loss": 1.4677,
+      "step": 3711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10376747240917623,
+      "learning_rate": 4.107922096939523e-05,
+      "loss": 1.4677,
+      "step": 3712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1020781700689239,
+      "learning_rate": 4.1066041880216725e-05,
+      "loss": 1.4687,
+      "step": 3713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09182953749531157,
+      "learning_rate": 4.105286279103822e-05,
+      "loss": 1.4661,
+      "step": 3714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.091912750499163,
+      "learning_rate": 4.103968370185972e-05,
+      "loss": 1.4681,
+      "step": 3715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0829608920704335,
+      "learning_rate": 4.102650461268122e-05,
+      "loss": 1.4653,
+      "step": 3716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.09209183490461377,
+      "learning_rate": 4.1013325523502714e-05,
+      "loss": 1.4649,
+      "step": 3717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0970245966695559,
+      "learning_rate": 4.100014643432421e-05,
+      "loss": 1.462,
+      "step": 3718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10610443106275694,
+      "learning_rate": 4.098696734514571e-05,
+      "loss": 1.4641,
+      "step": 3719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1034360762326821,
+      "learning_rate": 4.09737882559672e-05,
+      "loss": 1.4679,
+      "step": 3720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10240346393151627,
+      "learning_rate": 4.0960609166788695e-05,
+      "loss": 1.4685,
+      "step": 3721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.10427983847000105,
+      "learning_rate": 4.0947430077610196e-05,
+      "loss": 1.4718,
+      "step": 3722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10578127634994525,
+      "learning_rate": 4.093425098843169e-05,
+      "loss": 1.4727,
+      "step": 3723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10697106258054521,
+      "learning_rate": 4.092107189925319e-05,
+      "loss": 1.4685,
+      "step": 3724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11564748267963856,
+      "learning_rate": 4.0907892810074684e-05,
+      "loss": 1.4656,
+      "step": 3725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.129491195962255,
+      "learning_rate": 4.089471372089618e-05,
+      "loss": 1.4663,
+      "step": 3726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10181732194390819,
+      "learning_rate": 4.088153463171768e-05,
+      "loss": 1.4637,
+      "step": 3727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09211480497711666,
+      "learning_rate": 4.086835554253917e-05,
+      "loss": 1.4702,
+      "step": 3728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11880700800363696,
+      "learning_rate": 4.085517645336067e-05,
+      "loss": 1.465,
+      "step": 3729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.12731414061357918,
+      "learning_rate": 4.0841997364182166e-05,
+      "loss": 1.4728,
+      "step": 3730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1201380789050626,
+      "learning_rate": 4.0828818275003667e-05,
+      "loss": 1.4652,
+      "step": 3731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09056872909722997,
+      "learning_rate": 4.081563918582516e-05,
+      "loss": 1.4652,
+      "step": 3732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.12405635144680502,
+      "learning_rate": 4.0802460096646654e-05,
+      "loss": 1.4623,
+      "step": 3733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.12545100504572532,
+      "learning_rate": 4.0789281007468155e-05,
+      "loss": 1.47,
+      "step": 3734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10725061519945361,
+      "learning_rate": 4.077610191828965e-05,
+      "loss": 1.468,
+      "step": 3735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10320794735569412,
+      "learning_rate": 4.076292282911114e-05,
+      "loss": 1.4761,
+      "step": 3736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10875056364775591,
+      "learning_rate": 4.074974373993264e-05,
+      "loss": 1.4633,
+      "step": 3737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09558108340436797,
+      "learning_rate": 4.0736564650754136e-05,
+      "loss": 1.4619,
+      "step": 3738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08520708605043414,
+      "learning_rate": 4.072338556157564e-05,
+      "loss": 1.4772,
+      "step": 3739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10115537682402917,
+      "learning_rate": 4.071020647239713e-05,
+      "loss": 1.4633,
+      "step": 3740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10089652190731489,
+      "learning_rate": 4.069702738321863e-05,
+      "loss": 1.4725,
+      "step": 3741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0851722587188958,
+      "learning_rate": 4.0683848294040125e-05,
+      "loss": 1.4709,
+      "step": 3742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09138417173529959,
+      "learning_rate": 4.0670669204861625e-05,
+      "loss": 1.4716,
+      "step": 3743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09968144711165972,
+      "learning_rate": 4.065749011568312e-05,
+      "loss": 1.4799,
+      "step": 3744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09316303449184586,
+      "learning_rate": 4.064431102650461e-05,
+      "loss": 1.463,
+      "step": 3745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08962141548234415,
+      "learning_rate": 4.0631131937326113e-05,
+      "loss": 1.471,
+      "step": 3746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10127843329459699,
+      "learning_rate": 4.061795284814761e-05,
+      "loss": 1.475,
+      "step": 3747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10703346543284581,
+      "learning_rate": 4.06047737589691e-05,
+      "loss": 1.4596,
+      "step": 3748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0877912558290748,
+      "learning_rate": 4.05915946697906e-05,
+      "loss": 1.4651,
+      "step": 3749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09081056276223874,
+      "learning_rate": 4.0578415580612095e-05,
+      "loss": 1.4594,
+      "step": 3750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10634727694366142,
+      "learning_rate": 4.0565236491433596e-05,
+      "loss": 1.4707,
+      "step": 3751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0992193342728832,
+      "learning_rate": 4.055205740225509e-05,
+      "loss": 1.4625,
+      "step": 3752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.088235005003262,
+      "learning_rate": 4.053887831307658e-05,
+      "loss": 1.4666,
+      "step": 3753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11140057280867308,
+      "learning_rate": 4.0525699223898084e-05,
+      "loss": 1.4705,
+      "step": 3754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10508670334155126,
+      "learning_rate": 4.0512520134719584e-05,
+      "loss": 1.4629,
+      "step": 3755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08264932471229387,
+      "learning_rate": 4.049934104554108e-05,
+      "loss": 1.4606,
+      "step": 3756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09515153723174138,
+      "learning_rate": 4.048616195636257e-05,
+      "loss": 1.4651,
+      "step": 3757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.12012652499980772,
+      "learning_rate": 4.047298286718407e-05,
+      "loss": 1.4666,
+      "step": 3758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09655813489464267,
+      "learning_rate": 4.0459803778005566e-05,
+      "loss": 1.4699,
+      "step": 3759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09055240431979368,
+      "learning_rate": 4.044662468882706e-05,
+      "loss": 1.4673,
+      "step": 3760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10196002544639014,
+      "learning_rate": 4.043344559964856e-05,
+      "loss": 1.4661,
+      "step": 3761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09753412251047927,
+      "learning_rate": 4.0420266510470054e-05,
+      "loss": 1.467,
+      "step": 3762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0858267447058072,
+      "learning_rate": 4.0407087421291554e-05,
+      "loss": 1.4682,
+      "step": 3763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09849664108365207,
+      "learning_rate": 4.039390833211305e-05,
+      "loss": 1.4658,
+      "step": 3764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08955803399865314,
+      "learning_rate": 4.038072924293454e-05,
+      "loss": 1.4727,
+      "step": 3765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.099085737530737,
+      "learning_rate": 4.036755015375604e-05,
+      "loss": 1.4719,
+      "step": 3766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11178739293061442,
+      "learning_rate": 4.035437106457754e-05,
+      "loss": 1.4679,
+      "step": 3767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09904371801877118,
+      "learning_rate": 4.034119197539904e-05,
+      "loss": 1.4643,
+      "step": 3768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.08537189836714962,
+      "learning_rate": 4.032801288622053e-05,
+      "loss": 1.4688,
+      "step": 3769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09887440057792721,
+      "learning_rate": 4.031483379704203e-05,
+      "loss": 1.4662,
+      "step": 3770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1026164705299437,
+      "learning_rate": 4.0301654707863525e-05,
+      "loss": 1.4693,
+      "step": 3771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1034768395013401,
+      "learning_rate": 4.028847561868502e-05,
+      "loss": 1.4682,
+      "step": 3772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09901564281487693,
+      "learning_rate": 4.027529652950652e-05,
+      "loss": 1.4619,
+      "step": 3773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1057424826498627,
+      "learning_rate": 4.026211744032801e-05,
+      "loss": 1.4733,
+      "step": 3774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1093259912369232,
+      "learning_rate": 4.024893835114951e-05,
+      "loss": 1.4661,
+      "step": 3775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10727375286860644,
+      "learning_rate": 4.023575926197101e-05,
+      "loss": 1.4652,
+      "step": 3776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10418859739682694,
+      "learning_rate": 4.02225801727925e-05,
+      "loss": 1.4715,
+      "step": 3777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09643049225460003,
+      "learning_rate": 4.0209401083614e-05,
+      "loss": 1.4653,
+      "step": 3778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10500539008896145,
+      "learning_rate": 4.01962219944355e-05,
+      "loss": 1.4682,
+      "step": 3779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.093970353254377,
+      "learning_rate": 4.0183042905256995e-05,
+      "loss": 1.4661,
+      "step": 3780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10714539689893451,
+      "learning_rate": 4.016986381607849e-05,
+      "loss": 1.4609,
+      "step": 3781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11598433285061151,
+      "learning_rate": 4.015668472689999e-05,
+      "loss": 1.4617,
+      "step": 3782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09826055997523928,
+      "learning_rate": 4.0143505637721483e-05,
+      "loss": 1.4594,
+      "step": 3783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11076613873635796,
+      "learning_rate": 4.013032654854298e-05,
+      "loss": 1.4632,
+      "step": 3784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.12111274236569988,
+      "learning_rate": 4.011714745936448e-05,
+      "loss": 1.4735,
+      "step": 3785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0928339769593714,
+      "learning_rate": 4.010396837018597e-05,
+      "loss": 1.4757,
+      "step": 3786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10888424104679424,
+      "learning_rate": 4.009078928100747e-05,
+      "loss": 1.4696,
+      "step": 3787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11988448390324939,
+      "learning_rate": 4.0077610191828966e-05,
+      "loss": 1.4629,
+      "step": 3788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.10516479781386967,
+      "learning_rate": 4.006443110265046e-05,
+      "loss": 1.4671,
+      "step": 3789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09921012132347778,
+      "learning_rate": 4.005125201347196e-05,
+      "loss": 1.4656,
+      "step": 3790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10698379323787395,
+      "learning_rate": 4.003807292429346e-05,
+      "loss": 1.4678,
+      "step": 3791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11209276449223063,
+      "learning_rate": 4.0024893835114954e-05,
+      "loss": 1.4622,
+      "step": 3792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10359350766774725,
+      "learning_rate": 4.001171474593645e-05,
+      "loss": 1.4773,
+      "step": 3793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08972248060075333,
+      "learning_rate": 3.999853565675795e-05,
+      "loss": 1.4679,
+      "step": 3794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10627658265334344,
+      "learning_rate": 3.998535656757944e-05,
+      "loss": 1.4652,
+      "step": 3795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1048997730517494,
+      "learning_rate": 3.9972177478400936e-05,
+      "loss": 1.4594,
+      "step": 3796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09269357854983028,
+      "learning_rate": 3.9958998389222437e-05,
+      "loss": 1.4605,
+      "step": 3797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09351782249379877,
+      "learning_rate": 3.994581930004393e-05,
+      "loss": 1.4602,
+      "step": 3798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11403232441697181,
+      "learning_rate": 3.993264021086543e-05,
+      "loss": 1.4688,
+      "step": 3799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10322543984479983,
+      "learning_rate": 3.9919461121686925e-05,
+      "loss": 1.4748,
+      "step": 3800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09046187167627655,
+      "learning_rate": 3.990628203250842e-05,
+      "loss": 1.4614,
+      "step": 3801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11238517549410314,
+      "learning_rate": 3.989310294332992e-05,
+      "loss": 1.4634,
+      "step": 3802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10222889670135653,
+      "learning_rate": 3.987992385415142e-05,
+      "loss": 1.4653,
+      "step": 3803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08412056673496504,
+      "learning_rate": 3.986674476497291e-05,
+      "loss": 1.4615,
+      "step": 3804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10329399598539306,
+      "learning_rate": 3.985356567579441e-05,
+      "loss": 1.4558,
+      "step": 3805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09519291706296158,
+      "learning_rate": 3.984038658661591e-05,
+      "loss": 1.465,
+      "step": 3806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10538877047999522,
+      "learning_rate": 3.98272074974374e-05,
+      "loss": 1.4585,
+      "step": 3807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08780210079076763,
+      "learning_rate": 3.9814028408258895e-05,
+      "loss": 1.4699,
+      "step": 3808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08061573961447023,
+      "learning_rate": 3.9800849319080395e-05,
+      "loss": 1.4696,
+      "step": 3809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09280044932020859,
+      "learning_rate": 3.978767022990189e-05,
+      "loss": 1.4612,
+      "step": 3810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.07939478682412891,
+      "learning_rate": 3.977449114072339e-05,
+      "loss": 1.4697,
+      "step": 3811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08581241845214162,
+      "learning_rate": 3.976131205154488e-05,
+      "loss": 1.4675,
+      "step": 3812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08236807318699971,
+      "learning_rate": 3.974813296236638e-05,
+      "loss": 1.4639,
+      "step": 3813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09268012654245848,
+      "learning_rate": 3.973495387318788e-05,
+      "loss": 1.4686,
+      "step": 3814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0890846873310648,
+      "learning_rate": 3.972177478400938e-05,
+      "loss": 1.4661,
+      "step": 3815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09854422858232173,
+      "learning_rate": 3.970859569483087e-05,
+      "loss": 1.4688,
+      "step": 3816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10648232528614711,
+      "learning_rate": 3.9695416605652366e-05,
+      "loss": 1.4609,
+      "step": 3817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1241277248761442,
+      "learning_rate": 3.9682237516473866e-05,
+      "loss": 1.4676,
+      "step": 3818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.12067422881911319,
+      "learning_rate": 3.966905842729536e-05,
+      "loss": 1.4598,
+      "step": 3819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10003978300517215,
+      "learning_rate": 3.9655879338116854e-05,
+      "loss": 1.4731,
+      "step": 3820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09279896159640413,
+      "learning_rate": 3.9642700248938354e-05,
+      "loss": 1.4659,
+      "step": 3821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10898115394994307,
+      "learning_rate": 3.962952115975985e-05,
+      "loss": 1.4588,
+      "step": 3822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11171517257306918,
+      "learning_rate": 3.961634207058135e-05,
+      "loss": 1.4706,
+      "step": 3823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0979637243418282,
+      "learning_rate": 3.960316298140284e-05,
+      "loss": 1.4668,
+      "step": 3824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08430971482196392,
+      "learning_rate": 3.9589983892224336e-05,
+      "loss": 1.4603,
+      "step": 3825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09398289737559436,
+      "learning_rate": 3.9576804803045836e-05,
+      "loss": 1.4634,
+      "step": 3826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09485629898716312,
+      "learning_rate": 3.956362571386734e-05,
+      "loss": 1.4671,
+      "step": 3827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08744753211914777,
+      "learning_rate": 3.955044662468883e-05,
+      "loss": 1.4643,
+      "step": 3828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10587077450032176,
+      "learning_rate": 3.9537267535510324e-05,
+      "loss": 1.4682,
+      "step": 3829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09332632031057607,
+      "learning_rate": 3.9524088446331825e-05,
+      "loss": 1.4556,
+      "step": 3830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10103292123821495,
+      "learning_rate": 3.951090935715332e-05,
+      "loss": 1.469,
+      "step": 3831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08899002660654494,
+      "learning_rate": 3.949773026797481e-05,
+      "loss": 1.4638,
+      "step": 3832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09102755643087831,
+      "learning_rate": 3.948455117879631e-05,
+      "loss": 1.4595,
+      "step": 3833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09605897195550481,
+      "learning_rate": 3.947137208961781e-05,
+      "loss": 1.4705,
+      "step": 3834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08407891539194164,
+      "learning_rate": 3.945819300043931e-05,
+      "loss": 1.4573,
+      "step": 3835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10484775272523146,
+      "learning_rate": 3.94450139112608e-05,
+      "loss": 1.474,
+      "step": 3836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11049776731027153,
+      "learning_rate": 3.9431834822082295e-05,
+      "loss": 1.4703,
+      "step": 3837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10372918457406033,
+      "learning_rate": 3.9418655732903795e-05,
+      "loss": 1.4654,
+      "step": 3838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09351226581681775,
+      "learning_rate": 3.9405476643725296e-05,
+      "loss": 1.4729,
+      "step": 3839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11292798461569499,
+      "learning_rate": 3.939229755454678e-05,
+      "loss": 1.4595,
+      "step": 3840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.12125756979497006,
+      "learning_rate": 3.937911846536828e-05,
+      "loss": 1.4765,
+      "step": 3841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.09208585157830992,
+      "learning_rate": 3.9365939376189784e-05,
+      "loss": 1.4589,
+      "step": 3842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1081016029016093,
+      "learning_rate": 3.935276028701128e-05,
+      "loss": 1.4644,
+      "step": 3843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10676548714435688,
+      "learning_rate": 3.933958119783277e-05,
+      "loss": 1.4556,
+      "step": 3844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10397249284698933,
+      "learning_rate": 3.932640210865427e-05,
+      "loss": 1.4695,
+      "step": 3845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10780758599200632,
+      "learning_rate": 3.9313223019475765e-05,
+      "loss": 1.4658,
+      "step": 3846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11962512419774642,
+      "learning_rate": 3.9300043930297266e-05,
+      "loss": 1.4699,
+      "step": 3847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11103421945595564,
+      "learning_rate": 3.928686484111876e-05,
+      "loss": 1.4704,
+      "step": 3848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11559126430275407,
+      "learning_rate": 3.9273685751940253e-05,
+      "loss": 1.4665,
+      "step": 3849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1278880410889115,
+      "learning_rate": 3.9260506662761754e-05,
+      "loss": 1.4685,
+      "step": 3850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11572478503456345,
+      "learning_rate": 3.9247327573583254e-05,
+      "loss": 1.4684,
+      "step": 3851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11524650877968702,
+      "learning_rate": 3.923414848440474e-05,
+      "loss": 1.4714,
+      "step": 3852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10280664133375979,
+      "learning_rate": 3.922096939522624e-05,
+      "loss": 1.4636,
+      "step": 3853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10277927107975457,
+      "learning_rate": 3.920779030604774e-05,
+      "loss": 1.4673,
+      "step": 3854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.11747773093461299,
+      "learning_rate": 3.9194611216869236e-05,
+      "loss": 1.464,
+      "step": 3855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10469851677726844,
+      "learning_rate": 3.918143212769073e-05,
+      "loss": 1.4659,
+      "step": 3856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.08903431070205006,
+      "learning_rate": 3.916825303851223e-05,
+      "loss": 1.4674,
+      "step": 3857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.10202134388537475,
+      "learning_rate": 3.9155073949333724e-05,
+      "loss": 1.4641,
+      "step": 3858
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09650419847927245,
+      "learning_rate": 3.9141894860155225e-05,
+      "loss": 1.4708,
+      "step": 3859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08479725898014656,
+      "learning_rate": 3.912871577097672e-05,
+      "loss": 1.468,
+      "step": 3860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10000207844096234,
+      "learning_rate": 3.911553668179821e-05,
+      "loss": 1.4633,
+      "step": 3861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08662612059378096,
+      "learning_rate": 3.910235759261971e-05,
+      "loss": 1.4711,
+      "step": 3862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10331633463729314,
+      "learning_rate": 3.908917850344121e-05,
+      "loss": 1.4663,
+      "step": 3863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.12761874943013315,
+      "learning_rate": 3.90759994142627e-05,
+      "loss": 1.4674,
+      "step": 3864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10315530233923854,
+      "learning_rate": 3.90628203250842e-05,
+      "loss": 1.4643,
+      "step": 3865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08569552008383129,
+      "learning_rate": 3.90496412359057e-05,
+      "loss": 1.4698,
+      "step": 3866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.11115530433584912,
+      "learning_rate": 3.9036462146727195e-05,
+      "loss": 1.4683,
+      "step": 3867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.11223510563757794,
+      "learning_rate": 3.902328305754869e-05,
+      "loss": 1.4623,
+      "step": 3868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08861896210024807,
+      "learning_rate": 3.901010396837019e-05,
+      "loss": 1.4694,
+      "step": 3869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0907107688402517,
+      "learning_rate": 3.899692487919168e-05,
+      "loss": 1.4614,
+      "step": 3870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09623263862447581,
+      "learning_rate": 3.8983745790013183e-05,
+      "loss": 1.4635,
+      "step": 3871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07896367708452419,
+      "learning_rate": 3.897056670083468e-05,
+      "loss": 1.4594,
+      "step": 3872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09163551133335276,
+      "learning_rate": 3.895738761165617e-05,
+      "loss": 1.456,
+      "step": 3873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07886331454897709,
+      "learning_rate": 3.894420852247767e-05,
+      "loss": 1.4642,
+      "step": 3874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09611045215738914,
+      "learning_rate": 3.893102943329917e-05,
+      "loss": 1.4617,
+      "step": 3875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08014469718956548,
+      "learning_rate": 3.891785034412066e-05,
+      "loss": 1.4557,
+      "step": 3876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0879470539138674,
+      "learning_rate": 3.890467125494216e-05,
+      "loss": 1.4635,
+      "step": 3877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09566801650102078,
+      "learning_rate": 3.889149216576366e-05,
+      "loss": 1.4693,
+      "step": 3878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08908083607762919,
+      "learning_rate": 3.8878313076585154e-05,
+      "loss": 1.4762,
+      "step": 3879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08964134523974454,
+      "learning_rate": 3.886513398740665e-05,
+      "loss": 1.4695,
+      "step": 3880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09749923202457861,
+      "learning_rate": 3.885195489822815e-05,
+      "loss": 1.4681,
+      "step": 3881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09277297507061086,
+      "learning_rate": 3.883877580904964e-05,
+      "loss": 1.4654,
+      "step": 3882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09558407018308544,
+      "learning_rate": 3.882559671987114e-05,
+      "loss": 1.4599,
+      "step": 3883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10650477564781217,
+      "learning_rate": 3.8812417630692636e-05,
+      "loss": 1.4576,
+      "step": 3884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10613430309400695,
+      "learning_rate": 3.879923854151413e-05,
+      "loss": 1.4676,
+      "step": 3885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0935225964197205,
+      "learning_rate": 3.878605945233563e-05,
+      "loss": 1.4749,
+      "step": 3886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10494117570631585,
+      "learning_rate": 3.877288036315713e-05,
+      "loss": 1.4576,
+      "step": 3887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10846922674236498,
+      "learning_rate": 3.875970127397862e-05,
+      "loss": 1.4616,
+      "step": 3888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08847356455922839,
+      "learning_rate": 3.874652218480012e-05,
+      "loss": 1.462,
+      "step": 3889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09251156542127535,
+      "learning_rate": 3.873334309562162e-05,
+      "loss": 1.4704,
+      "step": 3890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10398847330808851,
+      "learning_rate": 3.872016400644311e-05,
+      "loss": 1.4744,
+      "step": 3891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10056695511919268,
+      "learning_rate": 3.8706984917264606e-05,
+      "loss": 1.4596,
+      "step": 3892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07976810877495029,
+      "learning_rate": 3.869380582808611e-05,
+      "loss": 1.4676,
+      "step": 3893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09576994438838281,
+      "learning_rate": 3.86806267389076e-05,
+      "loss": 1.4661,
+      "step": 3894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10996543745913165,
+      "learning_rate": 3.86674476497291e-05,
+      "loss": 1.4642,
+      "step": 3895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08799714574233845,
+      "learning_rate": 3.8654268560550595e-05,
+      "loss": 1.4661,
+      "step": 3896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08614802460744639,
+      "learning_rate": 3.864108947137209e-05,
+      "loss": 1.4653,
+      "step": 3897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09826216128315832,
+      "learning_rate": 3.862791038219359e-05,
+      "loss": 1.4576,
+      "step": 3898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10360046302933634,
+      "learning_rate": 3.861473129301509e-05,
+      "loss": 1.4621,
+      "step": 3899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09380219217772977,
+      "learning_rate": 3.8601552203836577e-05,
+      "loss": 1.4701,
+      "step": 3900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08524721150860513,
+      "learning_rate": 3.858837311465808e-05,
+      "loss": 1.457,
+      "step": 3901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09049227161844146,
+      "learning_rate": 3.857519402547958e-05,
+      "loss": 1.469,
+      "step": 3902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08414016342844434,
+      "learning_rate": 3.856201493630107e-05,
+      "loss": 1.4598,
+      "step": 3903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08587093273811171,
+      "learning_rate": 3.8548835847122565e-05,
+      "loss": 1.4548,
+      "step": 3904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08379766352901509,
+      "learning_rate": 3.8535656757944066e-05,
+      "loss": 1.458,
+      "step": 3905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08777111202137357,
+      "learning_rate": 3.852247766876556e-05,
+      "loss": 1.4621,
+      "step": 3906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08636497426854393,
+      "learning_rate": 3.850929857958706e-05,
+      "loss": 1.4622,
+      "step": 3907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09023872844399736,
+      "learning_rate": 3.8496119490408554e-05,
+      "loss": 1.4636,
+      "step": 3908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0920414563993299,
+      "learning_rate": 3.848294040123005e-05,
+      "loss": 1.4605,
+      "step": 3909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08289939714421603,
+      "learning_rate": 3.846976131205155e-05,
+      "loss": 1.4544,
+      "step": 3910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08317734279338648,
+      "learning_rate": 3.845658222287305e-05,
+      "loss": 1.4629,
+      "step": 3911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09190350833335617,
+      "learning_rate": 3.8443403133694535e-05,
+      "loss": 1.4629,
+      "step": 3912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07657073060193323,
+      "learning_rate": 3.8430224044516036e-05,
+      "loss": 1.4614,
+      "step": 3913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09185372145030234,
+      "learning_rate": 3.8417044955337536e-05,
+      "loss": 1.4631,
+      "step": 3914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08890726765739557,
+      "learning_rate": 3.840386586615903e-05,
+      "loss": 1.466,
+      "step": 3915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09576434405159742,
+      "learning_rate": 3.8390686776980524e-05,
+      "loss": 1.4642,
+      "step": 3916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0884647945767822,
+      "learning_rate": 3.8377507687802024e-05,
+      "loss": 1.4606,
+      "step": 3917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09153482618334476,
+      "learning_rate": 3.836432859862352e-05,
+      "loss": 1.4628,
+      "step": 3918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08978546140598863,
+      "learning_rate": 3.835114950944502e-05,
+      "loss": 1.4669,
+      "step": 3919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10093000854773596,
+      "learning_rate": 3.833797042026651e-05,
+      "loss": 1.4655,
+      "step": 3920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08707736888903313,
+      "learning_rate": 3.8324791331088006e-05,
+      "loss": 1.4558,
+      "step": 3921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09091569307775783,
+      "learning_rate": 3.831161224190951e-05,
+      "loss": 1.4671,
+      "step": 3922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10158410394124545,
+      "learning_rate": 3.829843315273101e-05,
+      "loss": 1.4699,
+      "step": 3923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09385519693547598,
+      "learning_rate": 3.8285254063552494e-05,
+      "loss": 1.4596,
+      "step": 3924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10084383843905272,
+      "learning_rate": 3.8272074974373995e-05,
+      "loss": 1.4624,
+      "step": 3925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.11305616456394857,
+      "learning_rate": 3.8258895885195495e-05,
+      "loss": 1.4694,
+      "step": 3926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.10390411379359668,
+      "learning_rate": 3.824571679601699e-05,
+      "loss": 1.4685,
+      "step": 3927
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10459884307237476,
+      "learning_rate": 3.823253770683848e-05,
+      "loss": 1.4627,
+      "step": 3928
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09137271463220997,
+      "learning_rate": 3.821935861765998e-05,
+      "loss": 1.4692,
+      "step": 3929
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10022700083900685,
+      "learning_rate": 3.820617952848148e-05,
+      "loss": 1.4658,
+      "step": 3930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08947292849304089,
+      "learning_rate": 3.819300043930298e-05,
+      "loss": 1.4702,
+      "step": 3931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0891525429715699,
+      "learning_rate": 3.817982135012447e-05,
+      "loss": 1.47,
+      "step": 3932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0913961966545852,
+      "learning_rate": 3.8166642260945965e-05,
+      "loss": 1.4667,
+      "step": 3933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08959638903851605,
+      "learning_rate": 3.8153463171767465e-05,
+      "loss": 1.461,
+      "step": 3934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09870675195561772,
+      "learning_rate": 3.8140284082588966e-05,
+      "loss": 1.4588,
+      "step": 3935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08813531244089726,
+      "learning_rate": 3.812710499341045e-05,
+      "loss": 1.4606,
+      "step": 3936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09240055834394416,
+      "learning_rate": 3.8113925904231953e-05,
+      "loss": 1.4633,
+      "step": 3937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08785154608600015,
+      "learning_rate": 3.8100746815053454e-05,
+      "loss": 1.4571,
+      "step": 3938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10647854518913784,
+      "learning_rate": 3.808756772587495e-05,
+      "loss": 1.4595,
+      "step": 3939
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10708809517436578,
+      "learning_rate": 3.807438863669644e-05,
+      "loss": 1.4692,
+      "step": 3940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.11018053838874725,
+      "learning_rate": 3.806120954751794e-05,
+      "loss": 1.4633,
+      "step": 3941
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09506716547113778,
+      "learning_rate": 3.8048030458339436e-05,
+      "loss": 1.4612,
+      "step": 3942
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08800372990813754,
+      "learning_rate": 3.8034851369160936e-05,
+      "loss": 1.4629,
+      "step": 3943
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0971795924701803,
+      "learning_rate": 3.802167227998243e-05,
+      "loss": 1.457,
+      "step": 3944
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08794436467627645,
+      "learning_rate": 3.8008493190803924e-05,
+      "loss": 1.4601,
+      "step": 3945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08889981445717776,
+      "learning_rate": 3.7995314101625424e-05,
+      "loss": 1.4659,
+      "step": 3946
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08689665858240948,
+      "learning_rate": 3.7982135012446925e-05,
+      "loss": 1.4629,
+      "step": 3947
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.07701664581826101,
+      "learning_rate": 3.796895592326841e-05,
+      "loss": 1.4675,
+      "step": 3948
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08054317573776923,
+      "learning_rate": 3.795577683408991e-05,
+      "loss": 1.4646,
+      "step": 3949
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08703135983812776,
+      "learning_rate": 3.794259774491141e-05,
+      "loss": 1.4489,
+      "step": 3950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08915930232426807,
+      "learning_rate": 3.7929418655732906e-05,
+      "loss": 1.457,
+      "step": 3951
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08935086548811516,
+      "learning_rate": 3.79162395665544e-05,
+      "loss": 1.457,
+      "step": 3952
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09537160736194412,
+      "learning_rate": 3.79030604773759e-05,
+      "loss": 1.4577,
+      "step": 3953
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10575620985734344,
+      "learning_rate": 3.7889881388197394e-05,
+      "loss": 1.4615,
+      "step": 3954
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09147783351318296,
+      "learning_rate": 3.7876702299018895e-05,
+      "loss": 1.46,
+      "step": 3955
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08958642661189059,
+      "learning_rate": 3.786352320984039e-05,
+      "loss": 1.4621,
+      "step": 3956
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09498938838632914,
+      "learning_rate": 3.785034412066188e-05,
+      "loss": 1.4578,
+      "step": 3957
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0935481248037015,
+      "learning_rate": 3.783716503148338e-05,
+      "loss": 1.4712,
+      "step": 3958
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08984505015502128,
+      "learning_rate": 3.7823985942304883e-05,
+      "loss": 1.4642,
+      "step": 3959
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08777664958956981,
+      "learning_rate": 3.781080685312637e-05,
+      "loss": 1.4628,
+      "step": 3960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09737737041067886,
+      "learning_rate": 3.779762776394787e-05,
+      "loss": 1.4635,
+      "step": 3961
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10242536487756589,
+      "learning_rate": 3.778444867476937e-05,
+      "loss": 1.4598,
+      "step": 3962
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09751237129939612,
+      "learning_rate": 3.7771269585590865e-05,
+      "loss": 1.4631,
+      "step": 3963
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09305398237717011,
+      "learning_rate": 3.775809049641236e-05,
+      "loss": 1.4613,
+      "step": 3964
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09358261046031106,
+      "learning_rate": 3.774491140723386e-05,
+      "loss": 1.468,
+      "step": 3965
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1122137396241042,
+      "learning_rate": 3.773173231805535e-05,
+      "loss": 1.4613,
+      "step": 3966
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09414112980158076,
+      "learning_rate": 3.7718553228876854e-05,
+      "loss": 1.4603,
+      "step": 3967
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09530398744617836,
+      "learning_rate": 3.770537413969835e-05,
+      "loss": 1.4638,
+      "step": 3968
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10412528326810842,
+      "learning_rate": 3.769219505051984e-05,
+      "loss": 1.4662,
+      "step": 3969
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09233291590029352,
+      "learning_rate": 3.767901596134134e-05,
+      "loss": 1.4628,
+      "step": 3970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09452688663858519,
+      "learning_rate": 3.766583687216284e-05,
+      "loss": 1.4581,
+      "step": 3971
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0883894221971201,
+      "learning_rate": 3.765265778298433e-05,
+      "loss": 1.4695,
+      "step": 3972
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09472674976986184,
+      "learning_rate": 3.763947869380583e-05,
+      "loss": 1.4626,
+      "step": 3973
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.102328474996461,
+      "learning_rate": 3.762629960462733e-05,
+      "loss": 1.4694,
+      "step": 3974
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0883756740079651,
+      "learning_rate": 3.7613120515448824e-05,
+      "loss": 1.4691,
+      "step": 3975
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0915812070710644,
+      "learning_rate": 3.759994142627032e-05,
+      "loss": 1.4612,
+      "step": 3976
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08954396218247782,
+      "learning_rate": 3.758676233709182e-05,
+      "loss": 1.4567,
+      "step": 3977
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08855525117517009,
+      "learning_rate": 3.757358324791331e-05,
+      "loss": 1.4638,
+      "step": 3978
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10288958701320854,
+      "learning_rate": 3.756040415873481e-05,
+      "loss": 1.464,
+      "step": 3979
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08605787322121565,
+      "learning_rate": 3.7547225069556306e-05,
+      "loss": 1.4709,
+      "step": 3980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09050576289698595,
+      "learning_rate": 3.75340459803778e-05,
+      "loss": 1.4574,
+      "step": 3981
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10501427493324042,
+      "learning_rate": 3.75208668911993e-05,
+      "loss": 1.4605,
+      "step": 3982
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.11126970076048205,
+      "learning_rate": 3.75076878020208e-05,
+      "loss": 1.4658,
+      "step": 3983
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08710119807162217,
+      "learning_rate": 3.749450871284229e-05,
+      "loss": 1.4578,
+      "step": 3984
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09322860593416482,
+      "learning_rate": 3.748132962366379e-05,
+      "loss": 1.4764,
+      "step": 3985
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10627254113452746,
+      "learning_rate": 3.746815053448529e-05,
+      "loss": 1.4613,
+      "step": 3986
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09407565529061988,
+      "learning_rate": 3.745497144530678e-05,
+      "loss": 1.4595,
+      "step": 3987
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09611613509907599,
+      "learning_rate": 3.7441792356128277e-05,
+      "loss": 1.4737,
+      "step": 3988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09219124284188378,
+      "learning_rate": 3.742861326694978e-05,
+      "loss": 1.4656,
+      "step": 3989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.08366272101905572,
+      "learning_rate": 3.741543417777127e-05,
+      "loss": 1.463,
+      "step": 3990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10402069254087079,
+      "learning_rate": 3.740225508859277e-05,
+      "loss": 1.4621,
+      "step": 3991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09094014723530416,
+      "learning_rate": 3.7389075999414265e-05,
+      "loss": 1.4703,
+      "step": 3992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10505885081860455,
+      "learning_rate": 3.737589691023576e-05,
+      "loss": 1.4569,
+      "step": 3993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.09874953668288199,
+      "learning_rate": 3.736271782105726e-05,
+      "loss": 1.4621,
+      "step": 3994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.10028719699708984,
+      "learning_rate": 3.734953873187876e-05,
+      "loss": 1.4605,
+      "step": 3995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.100388668298304,
+      "learning_rate": 3.733635964270025e-05,
+      "loss": 1.4607,
+      "step": 3996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09066069351053929,
+      "learning_rate": 3.732318055352175e-05,
+      "loss": 1.4567,
+      "step": 3997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10487327802058376,
+      "learning_rate": 3.731000146434325e-05,
+      "loss": 1.4637,
+      "step": 3998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1010143977995093,
+      "learning_rate": 3.729682237516474e-05,
+      "loss": 1.4607,
+      "step": 3999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08992854035624923,
+      "learning_rate": 3.7283643285986235e-05,
+      "loss": 1.4597,
+      "step": 4000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09463155558447088,
+      "learning_rate": 3.7270464196807736e-05,
+      "loss": 1.4634,
+      "step": 4001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1080010103166352,
+      "learning_rate": 3.725728510762923e-05,
+      "loss": 1.4586,
+      "step": 4002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09346830307752084,
+      "learning_rate": 3.724410601845073e-05,
+      "loss": 1.4502,
+      "step": 4003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09043132447938369,
+      "learning_rate": 3.7230926929272224e-05,
+      "loss": 1.4643,
+      "step": 4004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09568716622318402,
+      "learning_rate": 3.721774784009372e-05,
+      "loss": 1.453,
+      "step": 4005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09442058296668909,
+      "learning_rate": 3.720456875091522e-05,
+      "loss": 1.4602,
+      "step": 4006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0933864132830178,
+      "learning_rate": 3.719138966173671e-05,
+      "loss": 1.4679,
+      "step": 4007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09309419908511357,
+      "learning_rate": 3.7178210572558206e-05,
+      "loss": 1.4633,
+      "step": 4008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09111340834430919,
+      "learning_rate": 3.7165031483379706e-05,
+      "loss": 1.461,
+      "step": 4009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10241035135207124,
+      "learning_rate": 3.715185239420121e-05,
+      "loss": 1.4498,
+      "step": 4010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.11317093007375155,
+      "learning_rate": 3.71386733050227e-05,
+      "loss": 1.4622,
+      "step": 4011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10366163642590323,
+      "learning_rate": 3.7125494215844194e-05,
+      "loss": 1.4737,
+      "step": 4012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07811209659355733,
+      "learning_rate": 3.7112315126665695e-05,
+      "loss": 1.4642,
+      "step": 4013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09517646954173861,
+      "learning_rate": 3.709913603748719e-05,
+      "loss": 1.4588,
+      "step": 4014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09411069985635609,
+      "learning_rate": 3.708595694830869e-05,
+      "loss": 1.4654,
+      "step": 4015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09675517306122015,
+      "learning_rate": 3.707277785913018e-05,
+      "loss": 1.4587,
+      "step": 4016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09841413938255572,
+      "learning_rate": 3.7059598769951676e-05,
+      "loss": 1.4606,
+      "step": 4017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08732991270229419,
+      "learning_rate": 3.704641968077318e-05,
+      "loss": 1.4655,
+      "step": 4018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09349334081974249,
+      "learning_rate": 3.703324059159467e-05,
+      "loss": 1.4572,
+      "step": 4019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.084005373784943,
+      "learning_rate": 3.7020061502416164e-05,
+      "loss": 1.4605,
+      "step": 4020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08782576122314682,
+      "learning_rate": 3.7006882413237665e-05,
+      "loss": 1.4638,
+      "step": 4021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09736161193466197,
+      "learning_rate": 3.6993703324059165e-05,
+      "loss": 1.4624,
+      "step": 4022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0940738158223846,
+      "learning_rate": 3.698052423488066e-05,
+      "loss": 1.4678,
+      "step": 4023
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09213599400601807,
+      "learning_rate": 3.696734514570215e-05,
+      "loss": 1.4666,
+      "step": 4024
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09532899386632171,
+      "learning_rate": 3.6954166056523653e-05,
+      "loss": 1.4554,
+      "step": 4025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09726132759185181,
+      "learning_rate": 3.694098696734515e-05,
+      "loss": 1.4509,
+      "step": 4026
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08895886953867783,
+      "learning_rate": 3.692780787816665e-05,
+      "loss": 1.4707,
+      "step": 4027
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07987199690397902,
+      "learning_rate": 3.691462878898814e-05,
+      "loss": 1.4659,
+      "step": 4028
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08739275629715555,
+      "learning_rate": 3.6901449699809635e-05,
+      "loss": 1.4595,
+      "step": 4029
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09826514462373029,
+      "learning_rate": 3.6888270610631136e-05,
+      "loss": 1.4604,
+      "step": 4030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08836828281568164,
+      "learning_rate": 3.687509152145263e-05,
+      "loss": 1.4611,
+      "step": 4031
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07817998786067745,
+      "learning_rate": 3.686191243227412e-05,
+      "loss": 1.4547,
+      "step": 4032
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08851377841606485,
+      "learning_rate": 3.6848733343095624e-05,
+      "loss": 1.4573,
+      "step": 4033
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0845544905141117,
+      "learning_rate": 3.6835554253917124e-05,
+      "loss": 1.461,
+      "step": 4034
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0789577708441255,
+      "learning_rate": 3.682237516473862e-05,
+      "loss": 1.4613,
+      "step": 4035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09274122923608959,
+      "learning_rate": 3.680919607556011e-05,
+      "loss": 1.4735,
+      "step": 4036
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07887610219111657,
+      "learning_rate": 3.679601698638161e-05,
+      "loss": 1.472,
+      "step": 4037
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10534860140957272,
+      "learning_rate": 3.6782837897203106e-05,
+      "loss": 1.4659,
+      "step": 4038
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10776694913893493,
+      "learning_rate": 3.6769658808024606e-05,
+      "loss": 1.4633,
+      "step": 4039
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09683082428607578,
+      "learning_rate": 3.67564797188461e-05,
+      "loss": 1.4623,
+      "step": 4040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08484577699534351,
+      "learning_rate": 3.6743300629667594e-05,
+      "loss": 1.4619,
+      "step": 4041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09666353711519092,
+      "learning_rate": 3.6730121540489094e-05,
+      "loss": 1.4649,
+      "step": 4042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10010664845724271,
+      "learning_rate": 3.671694245131059e-05,
+      "loss": 1.4658,
+      "step": 4043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08843825975459098,
+      "learning_rate": 3.670376336213208e-05,
+      "loss": 1.4683,
+      "step": 4044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09431755019764375,
+      "learning_rate": 3.669058427295358e-05,
+      "loss": 1.4555,
+      "step": 4045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.10894517359669911,
+      "learning_rate": 3.667740518377508e-05,
+      "loss": 1.4625,
+      "step": 4046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09807119985852776,
+      "learning_rate": 3.666422609459658e-05,
+      "loss": 1.4551,
+      "step": 4047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08805519027856537,
+      "learning_rate": 3.665104700541807e-05,
+      "loss": 1.4663,
+      "step": 4048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09574485522484506,
+      "learning_rate": 3.663786791623957e-05,
+      "loss": 1.459,
+      "step": 4049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09937742521529457,
+      "learning_rate": 3.6624688827061065e-05,
+      "loss": 1.4648,
+      "step": 4050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08985310145729646,
+      "learning_rate": 3.6611509737882565e-05,
+      "loss": 1.4566,
+      "step": 4051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0958675917894909,
+      "learning_rate": 3.659833064870406e-05,
+      "loss": 1.4616,
+      "step": 4052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0835247624342829,
+      "learning_rate": 3.658515155952555e-05,
+      "loss": 1.4633,
+      "step": 4053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08539979918612402,
+      "learning_rate": 3.657197247034705e-05,
+      "loss": 1.4608,
+      "step": 4054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08787080963562453,
+      "learning_rate": 3.655879338116855e-05,
+      "loss": 1.4546,
+      "step": 4055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08255794920386829,
+      "learning_rate": 3.654561429199004e-05,
+      "loss": 1.4575,
+      "step": 4056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07808767673802681,
+      "learning_rate": 3.653243520281154e-05,
+      "loss": 1.4575,
+      "step": 4057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08541585504485273,
+      "learning_rate": 3.651925611363304e-05,
+      "loss": 1.4606,
+      "step": 4058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08675445453812548,
+      "learning_rate": 3.6506077024454536e-05,
+      "loss": 1.4599,
+      "step": 4059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09078252137159491,
+      "learning_rate": 3.649289793527603e-05,
+      "loss": 1.4662,
+      "step": 4060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0802058808459146,
+      "learning_rate": 3.647971884609753e-05,
+      "loss": 1.4584,
+      "step": 4061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.09767542180355214,
+      "learning_rate": 3.6466539756919024e-05,
+      "loss": 1.4604,
+      "step": 4062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08404608286005348,
+      "learning_rate": 3.6453360667740524e-05,
+      "loss": 1.4639,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0844339217142125,
+      "learning_rate": 3.644018157856202e-05,
+      "loss": 1.4603,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09175847433557514,
+      "learning_rate": 3.642700248938351e-05,
+      "loss": 1.463,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09912568176023792,
+      "learning_rate": 3.641382340020501e-05,
+      "loss": 1.4708,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09363082730691985,
+      "learning_rate": 3.6400644311026506e-05,
+      "loss": 1.4557,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09222344583387855,
+      "learning_rate": 3.6387465221848e-05,
+      "loss": 1.4579,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09337475950505614,
+      "learning_rate": 3.63742861326695e-05,
+      "loss": 1.4553,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09435020699599851,
+      "learning_rate": 3.6361107043491e-05,
+      "loss": 1.458,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0880223757166046,
+      "learning_rate": 3.6347927954312494e-05,
+      "loss": 1.4639,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08646384644609066,
+      "learning_rate": 3.633474886513399e-05,
+      "loss": 1.4594,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09557666951409552,
+      "learning_rate": 3.632156977595549e-05,
+      "loss": 1.4599,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10723922105171788,
+      "learning_rate": 3.630839068677698e-05,
+      "loss": 1.4606,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.11316371527528012,
+      "learning_rate": 3.629521159759848e-05,
+      "loss": 1.4586,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09881582989099684,
+      "learning_rate": 3.6282032508419977e-05,
+      "loss": 1.4613,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09484496625154912,
+      "learning_rate": 3.626885341924147e-05,
+      "loss": 1.4632,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09699313120609244,
+      "learning_rate": 3.625567433006297e-05,
+      "loss": 1.4614,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09478948223105405,
+      "learning_rate": 3.6242495240884465e-05,
+      "loss": 1.4651,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1151656355289001,
+      "learning_rate": 3.622931615170596e-05,
+      "loss": 1.4648,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09834468512993322,
+      "learning_rate": 3.621613706252746e-05,
+      "loss": 1.4616,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09100702046515026,
+      "learning_rate": 3.620295797334896e-05,
+      "loss": 1.4535,
+      "step": 4082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09586623311280836,
+      "learning_rate": 3.618977888417045e-05,
+      "loss": 1.458,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1148718485987422,
+      "learning_rate": 3.617659979499195e-05,
+      "loss": 1.4537,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09717195132346668,
+      "learning_rate": 3.616342070581345e-05,
+      "loss": 1.4561,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09537107466133744,
+      "learning_rate": 3.615024161663494e-05,
+      "loss": 1.4544,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.12516315130955358,
+      "learning_rate": 3.613706252745644e-05,
+      "loss": 1.4631,
+      "step": 4087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.11810817525319416,
+      "learning_rate": 3.6123883438277935e-05,
+      "loss": 1.4641,
+      "step": 4088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08673677250759168,
+      "learning_rate": 3.611070434909943e-05,
+      "loss": 1.4592,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10593338149074498,
+      "learning_rate": 3.609752525992093e-05,
+      "loss": 1.4596,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10298888819252361,
+      "learning_rate": 3.608434617074242e-05,
+      "loss": 1.4592,
+      "step": 4091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08318782422144774,
+      "learning_rate": 3.607116708156392e-05,
+      "loss": 1.4613,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09999995831314856,
+      "learning_rate": 3.605798799238542e-05,
+      "loss": 1.4668,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10114205740035632,
+      "learning_rate": 3.604480890320691e-05,
+      "loss": 1.4686,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1010143464185902,
+      "learning_rate": 3.603162981402841e-05,
+      "loss": 1.4511,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09517778683659692,
+      "learning_rate": 3.6018450724849906e-05,
+      "loss": 1.4635,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08607441911760848,
+      "learning_rate": 3.6005271635671406e-05,
+      "loss": 1.4567,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09659849031302806,
+      "learning_rate": 3.59920925464929e-05,
+      "loss": 1.4616,
+      "step": 4098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10162464749785326,
+      "learning_rate": 3.59789134573144e-05,
+      "loss": 1.4624,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09405248276306805,
+      "learning_rate": 3.5965734368135894e-05,
+      "loss": 1.4704,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09019912749352207,
+      "learning_rate": 3.595255527895739e-05,
+      "loss": 1.4584,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09597605296374326,
+      "learning_rate": 3.593937618977889e-05,
+      "loss": 1.4633,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1095314778257035,
+      "learning_rate": 3.592619710060038e-05,
+      "loss": 1.4589,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10557590666026224,
+      "learning_rate": 3.5913018011421876e-05,
+      "loss": 1.4531,
+      "step": 4104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08723868886303736,
+      "learning_rate": 3.5899838922243376e-05,
+      "loss": 1.4621,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.12109617245126036,
+      "learning_rate": 3.588665983306487e-05,
+      "loss": 1.4531,
+      "step": 4106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.12316892110698359,
+      "learning_rate": 3.587348074388637e-05,
+      "loss": 1.4624,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10712697138264361,
+      "learning_rate": 3.5860301654707864e-05,
+      "loss": 1.463,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10105801902435284,
+      "learning_rate": 3.5847122565529365e-05,
+      "loss": 1.4592,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09114740828955901,
+      "learning_rate": 3.583394347635086e-05,
+      "loss": 1.4633,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.11152725327440315,
+      "learning_rate": 3.582076438717236e-05,
+      "loss": 1.4631,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10091540689033697,
+      "learning_rate": 3.580758529799385e-05,
+      "loss": 1.4561,
+      "step": 4112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08784619126497081,
+      "learning_rate": 3.579440620881535e-05,
+      "loss": 1.4669,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09609444706350087,
+      "learning_rate": 3.578122711963685e-05,
+      "loss": 1.4575,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09534468922218033,
+      "learning_rate": 3.576804803045834e-05,
+      "loss": 1.4501,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09621304356341669,
+      "learning_rate": 3.5754868941279835e-05,
+      "loss": 1.4573,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09702427940815883,
+      "learning_rate": 3.5741689852101335e-05,
+      "loss": 1.458,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09706706847028933,
+      "learning_rate": 3.572851076292283e-05,
+      "loss": 1.4557,
+      "step": 4118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09451011417876333,
+      "learning_rate": 3.571533167374433e-05,
+      "loss": 1.463,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09807822556629721,
+      "learning_rate": 3.570215258456582e-05,
+      "loss": 1.4558,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.10140581402310375,
+      "learning_rate": 3.5688973495387324e-05,
+      "loss": 1.4564,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09216655361147295,
+      "learning_rate": 3.567579440620882e-05,
+      "loss": 1.4632,
+      "step": 4122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09480100747420707,
+      "learning_rate": 3.566261531703032e-05,
+      "loss": 1.4627,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08669694036449263,
+      "learning_rate": 3.564943622785181e-05,
+      "loss": 1.4636,
+      "step": 4124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09530263763201641,
+      "learning_rate": 3.5636257138673305e-05,
+      "loss": 1.454,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0910275141045834,
+      "learning_rate": 3.5623078049494806e-05,
+      "loss": 1.459,
+      "step": 4126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.09365422449639091,
+      "learning_rate": 3.56098989603163e-05,
+      "loss": 1.4619,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1017903995211755,
+      "learning_rate": 3.5596719871137793e-05,
+      "loss": 1.4614,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08604677926016661,
+      "learning_rate": 3.5583540781959294e-05,
+      "loss": 1.4639,
+      "step": 4129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.08609879594274088,
+      "learning_rate": 3.557036169278079e-05,
+      "loss": 1.4515,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0831714048136643,
+      "learning_rate": 3.555718260360229e-05,
+      "loss": 1.47,
+      "step": 4131
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08744832294485716,
+      "learning_rate": 3.554400351442378e-05,
+      "loss": 1.4587,
+      "step": 4132
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09096361346160022,
+      "learning_rate": 3.553082442524528e-05,
+      "loss": 1.4668,
+      "step": 4133
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08716635470346414,
+      "learning_rate": 3.5517645336066776e-05,
+      "loss": 1.4657,
+      "step": 4134
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08779376502780945,
+      "learning_rate": 3.550446624688828e-05,
+      "loss": 1.4634,
+      "step": 4135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08510719336968497,
+      "learning_rate": 3.549128715770977e-05,
+      "loss": 1.4553,
+      "step": 4136
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09818091810232923,
+      "learning_rate": 3.5478108068531264e-05,
+      "loss": 1.4638,
+      "step": 4137
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09881882947010434,
+      "learning_rate": 3.5464928979352765e-05,
+      "loss": 1.4613,
+      "step": 4138
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10498446738251806,
+      "learning_rate": 3.545174989017426e-05,
+      "loss": 1.4629,
+      "step": 4139
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0855633736189558,
+      "learning_rate": 3.543857080099575e-05,
+      "loss": 1.4603,
+      "step": 4140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0934182776489761,
+      "learning_rate": 3.542539171181725e-05,
+      "loss": 1.4619,
+      "step": 4141
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10715659534423529,
+      "learning_rate": 3.5412212622638746e-05,
+      "loss": 1.4641,
+      "step": 4142
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0950382468241048,
+      "learning_rate": 3.539903353346025e-05,
+      "loss": 1.4534,
+      "step": 4143
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10100445086521913,
+      "learning_rate": 3.538585444428174e-05,
+      "loss": 1.4571,
+      "step": 4144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.11596733912908533,
+      "learning_rate": 3.537267535510324e-05,
+      "loss": 1.4594,
+      "step": 4145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08732468110678535,
+      "learning_rate": 3.5359496265924735e-05,
+      "loss": 1.4512,
+      "step": 4146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10750021232422902,
+      "learning_rate": 3.5346317176746236e-05,
+      "loss": 1.4594,
+      "step": 4147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10568459750655676,
+      "learning_rate": 3.533313808756773e-05,
+      "loss": 1.4605,
+      "step": 4148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07678035677180127,
+      "learning_rate": 3.531995899838922e-05,
+      "loss": 1.4554,
+      "step": 4149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09648428196247905,
+      "learning_rate": 3.5306779909210724e-05,
+      "loss": 1.4608,
+      "step": 4150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08086926811982705,
+      "learning_rate": 3.529360082003222e-05,
+      "loss": 1.4565,
+      "step": 4151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10602202274756764,
+      "learning_rate": 3.528042173085371e-05,
+      "loss": 1.4672,
+      "step": 4152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09775340731868948,
+      "learning_rate": 3.526724264167521e-05,
+      "loss": 1.4656,
+      "step": 4153
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09355743765175129,
+      "learning_rate": 3.5254063552496705e-05,
+      "loss": 1.4648,
+      "step": 4154
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10264323606993978,
+      "learning_rate": 3.5240884463318206e-05,
+      "loss": 1.4656,
+      "step": 4155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08771810515637585,
+      "learning_rate": 3.52277053741397e-05,
+      "loss": 1.463,
+      "step": 4156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.10532983671855368,
+      "learning_rate": 3.52145262849612e-05,
+      "loss": 1.4621,
+      "step": 4157
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08767900813038022,
+      "learning_rate": 3.5201347195782694e-05,
+      "loss": 1.4641,
+      "step": 4158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09566440221364371,
+      "learning_rate": 3.5188168106604194e-05,
+      "loss": 1.4671,
+      "step": 4159
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08871380020611074,
+      "learning_rate": 3.517498901742569e-05,
+      "loss": 1.4539,
+      "step": 4160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09704192468383226,
+      "learning_rate": 3.516180992824718e-05,
+      "loss": 1.455,
+      "step": 4161
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08500126865742805,
+      "learning_rate": 3.514863083906868e-05,
+      "loss": 1.4492,
+      "step": 4162
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09691252019145281,
+      "learning_rate": 3.5135451749890176e-05,
+      "loss": 1.4656,
+      "step": 4163
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09501921137172378,
+      "learning_rate": 3.512227266071167e-05,
+      "loss": 1.4538,
+      "step": 4164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08567886813886343,
+      "learning_rate": 3.510909357153317e-05,
+      "loss": 1.4526,
+      "step": 4165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0832261920273469,
+      "learning_rate": 3.5095914482354664e-05,
+      "loss": 1.4551,
+      "step": 4166
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08513382463071448,
+      "learning_rate": 3.5082735393176165e-05,
+      "loss": 1.4561,
+      "step": 4167
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08867675133276522,
+      "learning_rate": 3.506955630399766e-05,
+      "loss": 1.4592,
+      "step": 4168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08665041320698454,
+      "learning_rate": 3.505637721481916e-05,
+      "loss": 1.4526,
+      "step": 4169
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1020452437897344,
+      "learning_rate": 3.504319812564065e-05,
+      "loss": 1.4595,
+      "step": 4170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08459339076110312,
+      "learning_rate": 3.503001903646215e-05,
+      "loss": 1.4619,
+      "step": 4171
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09292283609543972,
+      "learning_rate": 3.501683994728365e-05,
+      "loss": 1.4604,
+      "step": 4172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09144762883316097,
+      "learning_rate": 3.500366085810514e-05,
+      "loss": 1.4582,
+      "step": 4173
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09531492748130237,
+      "learning_rate": 3.499048176892664e-05,
+      "loss": 1.457,
+      "step": 4174
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08879625812872735,
+      "learning_rate": 3.4977302679748135e-05,
+      "loss": 1.4575,
+      "step": 4175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09602329324220239,
+      "learning_rate": 3.496412359056963e-05,
+      "loss": 1.4608,
+      "step": 4176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09710265606536271,
+      "learning_rate": 3.495094450139113e-05,
+      "loss": 1.4494,
+      "step": 4177
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09987716233499226,
+      "learning_rate": 3.493776541221262e-05,
+      "loss": 1.4587,
+      "step": 4178
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09902496250782378,
+      "learning_rate": 3.492458632303412e-05,
+      "loss": 1.4652,
+      "step": 4179
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09426048217227102,
+      "learning_rate": 3.491140723385562e-05,
+      "loss": 1.4548,
+      "step": 4180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08909469339050366,
+      "learning_rate": 3.489822814467711e-05,
+      "loss": 1.4553,
+      "step": 4181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08382582296559407,
+      "learning_rate": 3.488504905549861e-05,
+      "loss": 1.4509,
+      "step": 4182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09321130467157476,
+      "learning_rate": 3.487186996632011e-05,
+      "loss": 1.4629,
+      "step": 4183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0884914496946694,
+      "learning_rate": 3.4858690877141606e-05,
+      "loss": 1.4603,
+      "step": 4184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09124253742652434,
+      "learning_rate": 3.48455117879631e-05,
+      "loss": 1.4559,
+      "step": 4185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08730063066454403,
+      "learning_rate": 3.48323326987846e-05,
+      "loss": 1.4598,
+      "step": 4186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08367568342185899,
+      "learning_rate": 3.4819153609606094e-05,
+      "loss": 1.4585,
+      "step": 4187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09157185709129491,
+      "learning_rate": 3.480597452042759e-05,
+      "loss": 1.4526,
+      "step": 4188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09597838851158282,
+      "learning_rate": 3.479279543124909e-05,
+      "loss": 1.4569,
+      "step": 4189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09463833000951115,
+      "learning_rate": 3.477961634207058e-05,
+      "loss": 1.4605,
+      "step": 4190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08226979894152323,
+      "learning_rate": 3.476643725289208e-05,
+      "loss": 1.4601,
+      "step": 4191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08409146859547949,
+      "learning_rate": 3.4753258163713576e-05,
+      "loss": 1.4589,
+      "step": 4192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09192360600817692,
+      "learning_rate": 3.474007907453507e-05,
+      "loss": 1.4517,
+      "step": 4193
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07954442677539894,
+      "learning_rate": 3.472689998535657e-05,
+      "loss": 1.4536,
+      "step": 4194
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08490573932866537,
+      "learning_rate": 3.471372089617807e-05,
+      "loss": 1.4586,
+      "step": 4195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0841417510661447,
+      "learning_rate": 3.4700541806999564e-05,
+      "loss": 1.462,
+      "step": 4196
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.07862316797032978,
+      "learning_rate": 3.468736271782106e-05,
+      "loss": 1.4596,
+      "step": 4197
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08446719314354117,
+      "learning_rate": 3.467418362864256e-05,
+      "loss": 1.4563,
+      "step": 4198
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.08460001983767332,
+      "learning_rate": 3.466100453946405e-05,
+      "loss": 1.4707,
+      "step": 4199
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.09478044004285402,
+      "learning_rate": 3.4647825450285546e-05,
+      "loss": 1.4567,
+      "step": 4200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09528038283637834,
+      "learning_rate": 3.463464636110705e-05,
+      "loss": 1.455,
+      "step": 4201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08638720277565505,
+      "learning_rate": 3.462146727192854e-05,
+      "loss": 1.4618,
+      "step": 4202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09965070204247381,
+      "learning_rate": 3.460828818275004e-05,
+      "loss": 1.4575,
+      "step": 4203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09656828594923779,
+      "learning_rate": 3.4595109093571535e-05,
+      "loss": 1.4584,
+      "step": 4204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08456462636006784,
+      "learning_rate": 3.458193000439303e-05,
+      "loss": 1.4595,
+      "step": 4205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09295115372532725,
+      "learning_rate": 3.456875091521453e-05,
+      "loss": 1.4544,
+      "step": 4206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10675261866639751,
+      "learning_rate": 3.455557182603603e-05,
+      "loss": 1.4547,
+      "step": 4207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10331913104063427,
+      "learning_rate": 3.454239273685752e-05,
+      "loss": 1.4554,
+      "step": 4208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08418910068258378,
+      "learning_rate": 3.452921364767902e-05,
+      "loss": 1.4625,
+      "step": 4209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10472721486464265,
+      "learning_rate": 3.451603455850052e-05,
+      "loss": 1.4498,
+      "step": 4210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10321643505821164,
+      "learning_rate": 3.450285546932201e-05,
+      "loss": 1.4561,
+      "step": 4211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09514163759218197,
+      "learning_rate": 3.4489676380143505e-05,
+      "loss": 1.458,
+      "step": 4212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09085380844769922,
+      "learning_rate": 3.4476497290965005e-05,
+      "loss": 1.4483,
+      "step": 4213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08991553400515429,
+      "learning_rate": 3.44633182017865e-05,
+      "loss": 1.4641,
+      "step": 4214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08924908208885056,
+      "learning_rate": 3.4450139112608e-05,
+      "loss": 1.4563,
+      "step": 4215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07686864028815354,
+      "learning_rate": 3.4436960023429493e-05,
+      "loss": 1.4519,
+      "step": 4216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09873597174320839,
+      "learning_rate": 3.442378093425099e-05,
+      "loss": 1.4524,
+      "step": 4217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09897909167818078,
+      "learning_rate": 3.441060184507249e-05,
+      "loss": 1.4633,
+      "step": 4218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08692550258266177,
+      "learning_rate": 3.439742275589399e-05,
+      "loss": 1.4663,
+      "step": 4219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08568593955958669,
+      "learning_rate": 3.438424366671548e-05,
+      "loss": 1.4466,
+      "step": 4220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0925512386558198,
+      "learning_rate": 3.4371064577536976e-05,
+      "loss": 1.4551,
+      "step": 4221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09582220184539601,
+      "learning_rate": 3.4357885488358476e-05,
+      "loss": 1.4615,
+      "step": 4222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08949392527816077,
+      "learning_rate": 3.434470639917997e-05,
+      "loss": 1.4673,
+      "step": 4223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09237752207844915,
+      "learning_rate": 3.4331527310001464e-05,
+      "loss": 1.4495,
+      "step": 4224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10389115327542622,
+      "learning_rate": 3.4318348220822964e-05,
+      "loss": 1.4579,
+      "step": 4225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09233764197160474,
+      "learning_rate": 3.430516913164446e-05,
+      "loss": 1.4582,
+      "step": 4226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10077240430603676,
+      "learning_rate": 3.429199004246596e-05,
+      "loss": 1.4537,
+      "step": 4227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.10424431601732678,
+      "learning_rate": 3.427881095328745e-05,
+      "loss": 1.4627,
+      "step": 4228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0857968038726068,
+      "learning_rate": 3.4265631864108946e-05,
+      "loss": 1.4582,
+      "step": 4229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09754039675275872,
+      "learning_rate": 3.4252452774930446e-05,
+      "loss": 1.4621,
+      "step": 4230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09657430233825057,
+      "learning_rate": 3.423927368575195e-05,
+      "loss": 1.4606,
+      "step": 4231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09064214778704187,
+      "learning_rate": 3.422609459657344e-05,
+      "loss": 1.4546,
+      "step": 4232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09182667123903558,
+      "learning_rate": 3.4212915507394934e-05,
+      "loss": 1.4539,
+      "step": 4233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0823493050774827,
+      "learning_rate": 3.4199736418216435e-05,
+      "loss": 1.455,
+      "step": 4234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09187945916950739,
+      "learning_rate": 3.418655732903793e-05,
+      "loss": 1.4613,
+      "step": 4235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.082366546266498,
+      "learning_rate": 3.417337823985942e-05,
+      "loss": 1.4523,
+      "step": 4236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0914231680948735,
+      "learning_rate": 3.416019915068092e-05,
+      "loss": 1.4496,
+      "step": 4237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09213085672307499,
+      "learning_rate": 3.414702006150242e-05,
+      "loss": 1.4604,
+      "step": 4238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08522732664227879,
+      "learning_rate": 3.413384097232392e-05,
+      "loss": 1.455,
+      "step": 4239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08868569258613841,
+      "learning_rate": 3.412066188314541e-05,
+      "loss": 1.4477,
+      "step": 4240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09887930587529162,
+      "learning_rate": 3.4107482793966905e-05,
+      "loss": 1.4619,
+      "step": 4241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09874851875877719,
+      "learning_rate": 3.4094303704788405e-05,
+      "loss": 1.4581,
+      "step": 4242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07625820534407421,
+      "learning_rate": 3.4081124615609906e-05,
+      "loss": 1.464,
+      "step": 4243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09957257627618371,
+      "learning_rate": 3.40679455264314e-05,
+      "loss": 1.4608,
+      "step": 4244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09001699907430771,
+      "learning_rate": 3.405476643725289e-05,
+      "loss": 1.4559,
+      "step": 4245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08596879095397073,
+      "learning_rate": 3.4041587348074394e-05,
+      "loss": 1.4597,
+      "step": 4246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.100214996843057,
+      "learning_rate": 3.402840825889589e-05,
+      "loss": 1.4579,
+      "step": 4247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09559651428026898,
+      "learning_rate": 3.401522916971738e-05,
+      "loss": 1.4623,
+      "step": 4248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08496313923480217,
+      "learning_rate": 3.400205008053888e-05,
+      "loss": 1.4586,
+      "step": 4249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09704117375211302,
+      "learning_rate": 3.3988870991360376e-05,
+      "loss": 1.4674,
+      "step": 4250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09060723799135048,
+      "learning_rate": 3.3975691902181876e-05,
+      "loss": 1.4487,
+      "step": 4251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0877314131101142,
+      "learning_rate": 3.396251281300337e-05,
+      "loss": 1.4569,
+      "step": 4252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08732153654490173,
+      "learning_rate": 3.3949333723824864e-05,
+      "loss": 1.4652,
+      "step": 4253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08476989158203263,
+      "learning_rate": 3.3936154634646364e-05,
+      "loss": 1.4588,
+      "step": 4254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09085042185977607,
+      "learning_rate": 3.3922975545467865e-05,
+      "loss": 1.4635,
+      "step": 4255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09013010520868399,
+      "learning_rate": 3.390979645628936e-05,
+      "loss": 1.4454,
+      "step": 4256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08288848581411577,
+      "learning_rate": 3.389661736711085e-05,
+      "loss": 1.4558,
+      "step": 4257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08571755784806216,
+      "learning_rate": 3.388343827793235e-05,
+      "loss": 1.4578,
+      "step": 4258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0927191567806883,
+      "learning_rate": 3.3870259188753846e-05,
+      "loss": 1.4674,
+      "step": 4259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08179070488959674,
+      "learning_rate": 3.385708009957534e-05,
+      "loss": 1.4489,
+      "step": 4260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08041867756005006,
+      "learning_rate": 3.384390101039684e-05,
+      "loss": 1.4588,
+      "step": 4261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07628636664945908,
+      "learning_rate": 3.3830721921218334e-05,
+      "loss": 1.4579,
+      "step": 4262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08936192827459284,
+      "learning_rate": 3.3817542832039835e-05,
+      "loss": 1.4583,
+      "step": 4263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.08464718945439662,
+      "learning_rate": 3.380436374286133e-05,
+      "loss": 1.4557,
+      "step": 4264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.07950421030698213,
+      "learning_rate": 3.379118465368282e-05,
+      "loss": 1.4637,
+      "step": 4265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0859614826910382,
+      "learning_rate": 3.377800556450432e-05,
+      "loss": 1.447,
+      "step": 4266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09521474264133512,
+      "learning_rate": 3.376482647532582e-05,
+      "loss": 1.45,
+      "step": 4267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.09446248846008144,
+      "learning_rate": 3.375164738614731e-05,
+      "loss": 1.4541,
+      "step": 4268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08679772876815729,
+      "learning_rate": 3.373846829696881e-05,
+      "loss": 1.4503,
+      "step": 4269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08838316052538983,
+      "learning_rate": 3.372528920779031e-05,
+      "loss": 1.4519,
+      "step": 4270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08001997789625098,
+      "learning_rate": 3.3712110118611805e-05,
+      "loss": 1.4481,
+      "step": 4271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09784967378242597,
+      "learning_rate": 3.36989310294333e-05,
+      "loss": 1.4506,
+      "step": 4272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08633284833892038,
+      "learning_rate": 3.36857519402548e-05,
+      "loss": 1.4516,
+      "step": 4273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09424268642783219,
+      "learning_rate": 3.367257285107629e-05,
+      "loss": 1.4581,
+      "step": 4274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0931953785572421,
+      "learning_rate": 3.3659393761897794e-05,
+      "loss": 1.4558,
+      "step": 4275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09134670698717626,
+      "learning_rate": 3.364621467271929e-05,
+      "loss": 1.4588,
+      "step": 4276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08333938524433045,
+      "learning_rate": 3.363303558354078e-05,
+      "loss": 1.4625,
+      "step": 4277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08855345200252082,
+      "learning_rate": 3.361985649436228e-05,
+      "loss": 1.4573,
+      "step": 4278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08751907996831547,
+      "learning_rate": 3.360667740518378e-05,
+      "loss": 1.4669,
+      "step": 4279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0865332420586273,
+      "learning_rate": 3.359349831600527e-05,
+      "loss": 1.4578,
+      "step": 4280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07603118958094983,
+      "learning_rate": 3.358031922682677e-05,
+      "loss": 1.4592,
+      "step": 4281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07813064170730598,
+      "learning_rate": 3.356714013764827e-05,
+      "loss": 1.4646,
+      "step": 4282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08007899295883016,
+      "learning_rate": 3.3553961048469764e-05,
+      "loss": 1.4576,
+      "step": 4283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0773654117990903,
+      "learning_rate": 3.354078195929126e-05,
+      "loss": 1.4563,
+      "step": 4284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07551243973385793,
+      "learning_rate": 3.352760287011276e-05,
+      "loss": 1.4588,
+      "step": 4285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07750326112651743,
+      "learning_rate": 3.351442378093425e-05,
+      "loss": 1.457,
+      "step": 4286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07654662252947435,
+      "learning_rate": 3.350124469175575e-05,
+      "loss": 1.462,
+      "step": 4287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08416023823569603,
+      "learning_rate": 3.3488065602577246e-05,
+      "loss": 1.4467,
+      "step": 4288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07627550592279385,
+      "learning_rate": 3.347488651339874e-05,
+      "loss": 1.45,
+      "step": 4289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08196602285700345,
+      "learning_rate": 3.346170742422024e-05,
+      "loss": 1.4575,
+      "step": 4290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08737772941531183,
+      "learning_rate": 3.344852833504174e-05,
+      "loss": 1.4554,
+      "step": 4291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09248855563587018,
+      "learning_rate": 3.343534924586323e-05,
+      "loss": 1.4551,
+      "step": 4292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.095675338322473,
+      "learning_rate": 3.342217015668473e-05,
+      "loss": 1.46,
+      "step": 4293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08581744589546668,
+      "learning_rate": 3.340899106750623e-05,
+      "loss": 1.4721,
+      "step": 4294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08224506286017468,
+      "learning_rate": 3.339581197832772e-05,
+      "loss": 1.4571,
+      "step": 4295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07856995983986073,
+      "learning_rate": 3.3382632889149216e-05,
+      "loss": 1.4668,
+      "step": 4296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0767081530271087,
+      "learning_rate": 3.336945379997072e-05,
+      "loss": 1.4527,
+      "step": 4297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08161646117598118,
+      "learning_rate": 3.335627471079221e-05,
+      "loss": 1.46,
+      "step": 4298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08673377467950283,
+      "learning_rate": 3.334309562161371e-05,
+      "loss": 1.4584,
+      "step": 4299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07790216843887321,
+      "learning_rate": 3.3329916532435205e-05,
+      "loss": 1.452,
+      "step": 4300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07865307856779705,
+      "learning_rate": 3.33167374432567e-05,
+      "loss": 1.4539,
+      "step": 4301
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07853689862257593,
+      "learning_rate": 3.33035583540782e-05,
+      "loss": 1.4514,
+      "step": 4302
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07768700342903873,
+      "learning_rate": 3.32903792648997e-05,
+      "loss": 1.4564,
+      "step": 4303
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0803670196836774,
+      "learning_rate": 3.327720017572119e-05,
+      "loss": 1.4565,
+      "step": 4304
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08057281807620782,
+      "learning_rate": 3.326402108654269e-05,
+      "loss": 1.4421,
+      "step": 4305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0887681169662771,
+      "learning_rate": 3.325084199736419e-05,
+      "loss": 1.4567,
+      "step": 4306
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08387875094353715,
+      "learning_rate": 3.323766290818568e-05,
+      "loss": 1.4547,
+      "step": 4307
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09639484071265624,
+      "learning_rate": 3.3224483819007175e-05,
+      "loss": 1.4651,
+      "step": 4308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09817352793058977,
+      "learning_rate": 3.3211304729828676e-05,
+      "loss": 1.4573,
+      "step": 4309
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09491659170248834,
+      "learning_rate": 3.319812564065017e-05,
+      "loss": 1.4624,
+      "step": 4310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09257259110824614,
+      "learning_rate": 3.318494655147167e-05,
+      "loss": 1.4569,
+      "step": 4311
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1065508192350221,
+      "learning_rate": 3.3171767462293164e-05,
+      "loss": 1.4493,
+      "step": 4312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.10814668130273937,
+      "learning_rate": 3.315858837311466e-05,
+      "loss": 1.4627,
+      "step": 4313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08477825160164412,
+      "learning_rate": 3.314540928393616e-05,
+      "loss": 1.4572,
+      "step": 4314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09815537894808533,
+      "learning_rate": 3.313223019475766e-05,
+      "loss": 1.4525,
+      "step": 4315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.10537939757744769,
+      "learning_rate": 3.3119051105579145e-05,
+      "loss": 1.4629,
+      "step": 4316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08809264935457162,
+      "learning_rate": 3.3105872016400646e-05,
+      "loss": 1.4539,
+      "step": 4317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09023939020937763,
+      "learning_rate": 3.3092692927222146e-05,
+      "loss": 1.462,
+      "step": 4318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.07430729899052729,
+      "learning_rate": 3.307951383804364e-05,
+      "loss": 1.4616,
+      "step": 4319
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08693848581056898,
+      "learning_rate": 3.3066334748865134e-05,
+      "loss": 1.4549,
+      "step": 4320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08298831662986825,
+      "learning_rate": 3.3053155659686634e-05,
+      "loss": 1.4546,
+      "step": 4321
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08024658514610188,
+      "learning_rate": 3.303997657050813e-05,
+      "loss": 1.4535,
+      "step": 4322
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09944019674557723,
+      "learning_rate": 3.302679748132963e-05,
+      "loss": 1.4598,
+      "step": 4323
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08920241116476177,
+      "learning_rate": 3.301361839215112e-05,
+      "loss": 1.4535,
+      "step": 4324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0896300915686245,
+      "learning_rate": 3.3000439302972616e-05,
+      "loss": 1.4555,
+      "step": 4325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08651313392260293,
+      "learning_rate": 3.298726021379412e-05,
+      "loss": 1.4637,
+      "step": 4326
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0955612342213602,
+      "learning_rate": 3.297408112461562e-05,
+      "loss": 1.4502,
+      "step": 4327
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09629969802507145,
+      "learning_rate": 3.2960902035437104e-05,
+      "loss": 1.4473,
+      "step": 4328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0984658208566818,
+      "learning_rate": 3.2947722946258605e-05,
+      "loss": 1.4623,
+      "step": 4329
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08893373314956414,
+      "learning_rate": 3.2934543857080105e-05,
+      "loss": 1.4607,
+      "step": 4330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08339565957448036,
+      "learning_rate": 3.29213647679016e-05,
+      "loss": 1.4492,
+      "step": 4331
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0877337689382627,
+      "learning_rate": 3.290818567872309e-05,
+      "loss": 1.4549,
+      "step": 4332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09664202168664036,
+      "learning_rate": 3.289500658954459e-05,
+      "loss": 1.4627,
+      "step": 4333
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08657791273317984,
+      "learning_rate": 3.288182750036609e-05,
+      "loss": 1.4544,
+      "step": 4334
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08246009939376166,
+      "learning_rate": 3.286864841118759e-05,
+      "loss": 1.4499,
+      "step": 4335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.08631864429463637,
+      "learning_rate": 3.285546932200908e-05,
+      "loss": 1.4547,
+      "step": 4336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08427139824844483,
+      "learning_rate": 3.2842290232830575e-05,
+      "loss": 1.4616,
+      "step": 4337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0866843409199846,
+      "learning_rate": 3.2829111143652076e-05,
+      "loss": 1.4543,
+      "step": 4338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08116430978121252,
+      "learning_rate": 3.2815932054473576e-05,
+      "loss": 1.4588,
+      "step": 4339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08991239691779003,
+      "learning_rate": 3.280275296529506e-05,
+      "loss": 1.4567,
+      "step": 4340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07753475246097835,
+      "learning_rate": 3.2789573876116564e-05,
+      "loss": 1.4617,
+      "step": 4341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07960901507398974,
+      "learning_rate": 3.2776394786938064e-05,
+      "loss": 1.4604,
+      "step": 4342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09834720332600227,
+      "learning_rate": 3.276321569775956e-05,
+      "loss": 1.4582,
+      "step": 4343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08936261354668502,
+      "learning_rate": 3.275003660858105e-05,
+      "loss": 1.4583,
+      "step": 4344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09341086727961298,
+      "learning_rate": 3.273685751940255e-05,
+      "loss": 1.4559,
+      "step": 4345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09866375608125309,
+      "learning_rate": 3.2723678430224046e-05,
+      "loss": 1.4528,
+      "step": 4346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09010861799857944,
+      "learning_rate": 3.271049934104554e-05,
+      "loss": 1.4523,
+      "step": 4347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09151963774173763,
+      "learning_rate": 3.269732025186704e-05,
+      "loss": 1.46,
+      "step": 4348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09569553673602202,
+      "learning_rate": 3.2684141162688534e-05,
+      "loss": 1.4583,
+      "step": 4349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.10815440454792727,
+      "learning_rate": 3.2670962073510034e-05,
+      "loss": 1.4594,
+      "step": 4350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08628310907633255,
+      "learning_rate": 3.265778298433153e-05,
+      "loss": 1.4494,
+      "step": 4351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08946319938694107,
+      "learning_rate": 3.264460389515302e-05,
+      "loss": 1.451,
+      "step": 4352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08771798525127145,
+      "learning_rate": 3.263142480597452e-05,
+      "loss": 1.4481,
+      "step": 4353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09818091748770104,
+      "learning_rate": 3.261824571679602e-05,
+      "loss": 1.453,
+      "step": 4354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.10048422101413815,
+      "learning_rate": 3.260506662761751e-05,
+      "loss": 1.4597,
+      "step": 4355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08779324674747291,
+      "learning_rate": 3.259188753843901e-05,
+      "loss": 1.4537,
+      "step": 4356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09807514116526368,
+      "learning_rate": 3.257870844926051e-05,
+      "loss": 1.4506,
+      "step": 4357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08619743266183094,
+      "learning_rate": 3.2565529360082005e-05,
+      "loss": 1.4586,
+      "step": 4358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09174516357734093,
+      "learning_rate": 3.25523502709035e-05,
+      "loss": 1.4481,
+      "step": 4359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09638778393871786,
+      "learning_rate": 3.2539171181725e-05,
+      "loss": 1.4546,
+      "step": 4360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08961514609643316,
+      "learning_rate": 3.252599209254649e-05,
+      "loss": 1.453,
+      "step": 4361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08421577829058954,
+      "learning_rate": 3.251281300336799e-05,
+      "loss": 1.45,
+      "step": 4362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0982086026391868,
+      "learning_rate": 3.249963391418949e-05,
+      "loss": 1.4579,
+      "step": 4363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08260709701335599,
+      "learning_rate": 3.248645482501098e-05,
+      "loss": 1.4621,
+      "step": 4364
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0894679512886158,
+      "learning_rate": 3.247327573583248e-05,
+      "loss": 1.4564,
+      "step": 4365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09120458749897656,
+      "learning_rate": 3.246009664665398e-05,
+      "loss": 1.4536,
+      "step": 4366
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08192769066634709,
+      "learning_rate": 3.244691755747547e-05,
+      "loss": 1.4601,
+      "step": 4367
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09641078934948785,
+      "learning_rate": 3.243373846829697e-05,
+      "loss": 1.4602,
+      "step": 4368
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08634519438984592,
+      "learning_rate": 3.242055937911847e-05,
+      "loss": 1.465,
+      "step": 4369
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09033529236454006,
+      "learning_rate": 3.240738028993996e-05,
+      "loss": 1.4568,
+      "step": 4370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09064346416072408,
+      "learning_rate": 3.239420120076146e-05,
+      "loss": 1.4544,
+      "step": 4371
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.10166571411085032,
+      "learning_rate": 3.238102211158296e-05,
+      "loss": 1.4627,
+      "step": 4372
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09843185353105724,
+      "learning_rate": 3.236784302240445e-05,
+      "loss": 1.4526,
+      "step": 4373
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08933613189870343,
+      "learning_rate": 3.235466393322595e-05,
+      "loss": 1.4554,
+      "step": 4374
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.097660226966949,
+      "learning_rate": 3.2341484844047446e-05,
+      "loss": 1.4501,
+      "step": 4375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09621726251237661,
+      "learning_rate": 3.232830575486894e-05,
+      "loss": 1.4589,
+      "step": 4376
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0816482242565533,
+      "learning_rate": 3.231512666569044e-05,
+      "loss": 1.4542,
+      "step": 4377
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09159067029852458,
+      "learning_rate": 3.230194757651194e-05,
+      "loss": 1.4547,
+      "step": 4378
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08148601982846682,
+      "learning_rate": 3.228876848733343e-05,
+      "loss": 1.4538,
+      "step": 4379
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08484027116123462,
+      "learning_rate": 3.227558939815493e-05,
+      "loss": 1.4532,
+      "step": 4380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09074786491831799,
+      "learning_rate": 3.226241030897643e-05,
+      "loss": 1.4516,
+      "step": 4381
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09046335426930993,
+      "learning_rate": 3.224923121979792e-05,
+      "loss": 1.4542,
+      "step": 4382
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08450361136702746,
+      "learning_rate": 3.2236052130619416e-05,
+      "loss": 1.4502,
+      "step": 4383
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08802128373468956,
+      "learning_rate": 3.2222873041440916e-05,
+      "loss": 1.4522,
+      "step": 4384
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09532114459364328,
+      "learning_rate": 3.220969395226241e-05,
+      "loss": 1.4474,
+      "step": 4385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08624245047540476,
+      "learning_rate": 3.219651486308391e-05,
+      "loss": 1.4546,
+      "step": 4386
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08704600100801405,
+      "learning_rate": 3.2183335773905404e-05,
+      "loss": 1.4561,
+      "step": 4387
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08063793461744394,
+      "learning_rate": 3.21701566847269e-05,
+      "loss": 1.453,
+      "step": 4388
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08381742559928042,
+      "learning_rate": 3.21569775955484e-05,
+      "loss": 1.4451,
+      "step": 4389
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07888359729830437,
+      "learning_rate": 3.21437985063699e-05,
+      "loss": 1.4586,
+      "step": 4390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07823074502277746,
+      "learning_rate": 3.2130619417191386e-05,
+      "loss": 1.4548,
+      "step": 4391
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08446804488309965,
+      "learning_rate": 3.211744032801289e-05,
+      "loss": 1.4521,
+      "step": 4392
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08341184777493726,
+      "learning_rate": 3.210426123883439e-05,
+      "loss": 1.4465,
+      "step": 4393
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07836602821304224,
+      "learning_rate": 3.209108214965588e-05,
+      "loss": 1.4485,
+      "step": 4394
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07788101006106174,
+      "learning_rate": 3.2077903060477375e-05,
+      "loss": 1.4539,
+      "step": 4395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08540681046887158,
+      "learning_rate": 3.2064723971298875e-05,
+      "loss": 1.4474,
+      "step": 4396
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08437387350805144,
+      "learning_rate": 3.205154488212037e-05,
+      "loss": 1.459,
+      "step": 4397
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07422304830887923,
+      "learning_rate": 3.203836579294187e-05,
+      "loss": 1.4589,
+      "step": 4398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07932197459130426,
+      "learning_rate": 3.202518670376336e-05,
+      "loss": 1.4512,
+      "step": 4399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07729185967000474,
+      "learning_rate": 3.201200761458486e-05,
+      "loss": 1.4444,
+      "step": 4400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07888902819142174,
+      "learning_rate": 3.199882852540636e-05,
+      "loss": 1.4508,
+      "step": 4401
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08435603279126659,
+      "learning_rate": 3.198564943622786e-05,
+      "loss": 1.4504,
+      "step": 4402
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.08292014954753203,
+      "learning_rate": 3.1972470347049345e-05,
+      "loss": 1.4583,
+      "step": 4403
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.07483185235182781,
+      "learning_rate": 3.1959291257870845e-05,
+      "loss": 1.4548,
+      "step": 4404
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.09124131161989013,
+      "learning_rate": 3.1946112168692346e-05,
+      "loss": 1.4499,
+      "step": 4405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09154180625954703,
+      "learning_rate": 3.193293307951384e-05,
+      "loss": 1.4506,
+      "step": 4406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08748985481249487,
+      "learning_rate": 3.1919753990335333e-05,
+      "loss": 1.4638,
+      "step": 4407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07618263268043857,
+      "learning_rate": 3.1906574901156834e-05,
+      "loss": 1.457,
+      "step": 4408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07975254625037119,
+      "learning_rate": 3.189339581197833e-05,
+      "loss": 1.4514,
+      "step": 4409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07508101260568421,
+      "learning_rate": 3.188021672279983e-05,
+      "loss": 1.4539,
+      "step": 4410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08396596959928797,
+      "learning_rate": 3.186703763362132e-05,
+      "loss": 1.4524,
+      "step": 4411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09073096440187978,
+      "learning_rate": 3.1853858544442816e-05,
+      "loss": 1.4492,
+      "step": 4412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0804013852044971,
+      "learning_rate": 3.1840679455264316e-05,
+      "loss": 1.4517,
+      "step": 4413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09080845876773613,
+      "learning_rate": 3.182750036608582e-05,
+      "loss": 1.463,
+      "step": 4414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09033906532325708,
+      "learning_rate": 3.1814321276907304e-05,
+      "loss": 1.4486,
+      "step": 4415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08130753188946163,
+      "learning_rate": 3.1801142187728804e-05,
+      "loss": 1.4555,
+      "step": 4416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07478494894425627,
+      "learning_rate": 3.1787963098550305e-05,
+      "loss": 1.4662,
+      "step": 4417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0872633467848335,
+      "learning_rate": 3.17747840093718e-05,
+      "loss": 1.458,
+      "step": 4418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08894803642510618,
+      "learning_rate": 3.176160492019329e-05,
+      "loss": 1.4478,
+      "step": 4419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07666796755237197,
+      "learning_rate": 3.174842583101479e-05,
+      "loss": 1.4465,
+      "step": 4420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08340827441399677,
+      "learning_rate": 3.1735246741836287e-05,
+      "loss": 1.4498,
+      "step": 4421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09065781239156646,
+      "learning_rate": 3.172206765265779e-05,
+      "loss": 1.4591,
+      "step": 4422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09034711009994305,
+      "learning_rate": 3.170888856347928e-05,
+      "loss": 1.4547,
+      "step": 4423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09176972728901148,
+      "learning_rate": 3.1695709474300775e-05,
+      "loss": 1.4541,
+      "step": 4424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08793781492421768,
+      "learning_rate": 3.1682530385122275e-05,
+      "loss": 1.4514,
+      "step": 4425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0817064895774374,
+      "learning_rate": 3.1669351295943776e-05,
+      "loss": 1.4541,
+      "step": 4426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08798349367938448,
+      "learning_rate": 3.165617220676526e-05,
+      "loss": 1.453,
+      "step": 4427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09631957960183928,
+      "learning_rate": 3.164299311758676e-05,
+      "loss": 1.447,
+      "step": 4428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08680659637239242,
+      "learning_rate": 3.1629814028408264e-05,
+      "loss": 1.4515,
+      "step": 4429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0948873770941176,
+      "learning_rate": 3.161663493922976e-05,
+      "loss": 1.4556,
+      "step": 4430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08821566046668772,
+      "learning_rate": 3.160345585005125e-05,
+      "loss": 1.4597,
+      "step": 4431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08399546639138704,
+      "learning_rate": 3.159027676087275e-05,
+      "loss": 1.4518,
+      "step": 4432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08984577954189613,
+      "learning_rate": 3.1577097671694245e-05,
+      "loss": 1.4505,
+      "step": 4433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09140073511693524,
+      "learning_rate": 3.1563918582515746e-05,
+      "loss": 1.4537,
+      "step": 4434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09909010872302737,
+      "learning_rate": 3.155073949333724e-05,
+      "loss": 1.4503,
+      "step": 4435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08382899776823209,
+      "learning_rate": 3.153756040415873e-05,
+      "loss": 1.4466,
+      "step": 4436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09960645084069802,
+      "learning_rate": 3.1524381314980234e-05,
+      "loss": 1.4467,
+      "step": 4437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08026974951449446,
+      "learning_rate": 3.1511202225801734e-05,
+      "loss": 1.4576,
+      "step": 4438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09258234730844396,
+      "learning_rate": 3.149802313662322e-05,
+      "loss": 1.4548,
+      "step": 4439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0916406179809907,
+      "learning_rate": 3.148484404744472e-05,
+      "loss": 1.4527,
+      "step": 4440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07842034598511961,
+      "learning_rate": 3.147166495826622e-05,
+      "loss": 1.458,
+      "step": 4441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08249782076451012,
+      "learning_rate": 3.1458485869087716e-05,
+      "loss": 1.4489,
+      "step": 4442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08562307127326475,
+      "learning_rate": 3.144530677990921e-05,
+      "loss": 1.4532,
+      "step": 4443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08493562597356177,
+      "learning_rate": 3.143212769073071e-05,
+      "loss": 1.4561,
+      "step": 4444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09604424247795006,
+      "learning_rate": 3.1418948601552204e-05,
+      "loss": 1.4561,
+      "step": 4445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09858599662410354,
+      "learning_rate": 3.1405769512373705e-05,
+      "loss": 1.4539,
+      "step": 4446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07934003381025359,
+      "learning_rate": 3.13925904231952e-05,
+      "loss": 1.4545,
+      "step": 4447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0927984197843123,
+      "learning_rate": 3.137941133401669e-05,
+      "loss": 1.4496,
+      "step": 4448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.10013662863653908,
+      "learning_rate": 3.136623224483819e-05,
+      "loss": 1.4455,
+      "step": 4449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0826169895924107,
+      "learning_rate": 3.135305315565969e-05,
+      "loss": 1.4448,
+      "step": 4450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0859689133714226,
+      "learning_rate": 3.133987406648118e-05,
+      "loss": 1.4573,
+      "step": 4451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08910147237692803,
+      "learning_rate": 3.132669497730268e-05,
+      "loss": 1.4529,
+      "step": 4452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08512643036866353,
+      "learning_rate": 3.131351588812418e-05,
+      "loss": 1.4532,
+      "step": 4453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09701218996443961,
+      "learning_rate": 3.1300336798945675e-05,
+      "loss": 1.4514,
+      "step": 4454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09127917047304775,
+      "learning_rate": 3.128715770976717e-05,
+      "loss": 1.4503,
+      "step": 4455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08582123377614034,
+      "learning_rate": 3.127397862058867e-05,
+      "loss": 1.4537,
+      "step": 4456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0820058586878224,
+      "learning_rate": 3.126079953141016e-05,
+      "loss": 1.4515,
+      "step": 4457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09298841953286734,
+      "learning_rate": 3.124762044223166e-05,
+      "loss": 1.4604,
+      "step": 4458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09011970166935124,
+      "learning_rate": 3.123444135305316e-05,
+      "loss": 1.4504,
+      "step": 4459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08294494701357177,
+      "learning_rate": 3.122126226387465e-05,
+      "loss": 1.4539,
+      "step": 4460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08735642411594997,
+      "learning_rate": 3.120808317469615e-05,
+      "loss": 1.4494,
+      "step": 4461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08400494573246266,
+      "learning_rate": 3.119490408551765e-05,
+      "loss": 1.449,
+      "step": 4462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07958500331482186,
+      "learning_rate": 3.118172499633914e-05,
+      "loss": 1.4549,
+      "step": 4463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08475185661613631,
+      "learning_rate": 3.116854590716064e-05,
+      "loss": 1.4554,
+      "step": 4464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08328127532447417,
+      "learning_rate": 3.115536681798214e-05,
+      "loss": 1.4512,
+      "step": 4465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07853224607873865,
+      "learning_rate": 3.1142187728803634e-05,
+      "loss": 1.4655,
+      "step": 4466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08465516960559241,
+      "learning_rate": 3.112900863962513e-05,
+      "loss": 1.4545,
+      "step": 4467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07892448046935134,
+      "learning_rate": 3.111582955044663e-05,
+      "loss": 1.4529,
+      "step": 4468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.07992509811855186,
+      "learning_rate": 3.110265046126812e-05,
+      "loss": 1.4471,
+      "step": 4469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08650071559589635,
+      "learning_rate": 3.108947137208962e-05,
+      "loss": 1.4575,
+      "step": 4470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09586470617712303,
+      "learning_rate": 3.1076292282911116e-05,
+      "loss": 1.4515,
+      "step": 4471
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08386278926017826,
+      "learning_rate": 3.106311319373261e-05,
+      "loss": 1.4532,
+      "step": 4472
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.08734641071319846,
+      "learning_rate": 3.104993410455411e-05,
+      "loss": 1.4511,
+      "step": 4473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0914202344925143,
+      "learning_rate": 3.103675501537561e-05,
+      "loss": 1.447,
+      "step": 4474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0919468786790706,
+      "learning_rate": 3.10235759261971e-05,
+      "loss": 1.4507,
+      "step": 4475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08804802703999227,
+      "learning_rate": 3.10103968370186e-05,
+      "loss": 1.4501,
+      "step": 4476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08375586010510225,
+      "learning_rate": 3.09972177478401e-05,
+      "loss": 1.4482,
+      "step": 4477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0869382192232415,
+      "learning_rate": 3.098403865866159e-05,
+      "loss": 1.4534,
+      "step": 4478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0869973121205379,
+      "learning_rate": 3.0970859569483086e-05,
+      "loss": 1.4518,
+      "step": 4479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08427248492633913,
+      "learning_rate": 3.095768048030459e-05,
+      "loss": 1.451,
+      "step": 4480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08326604798237942,
+      "learning_rate": 3.094450139112608e-05,
+      "loss": 1.4596,
+      "step": 4481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09402945300546883,
+      "learning_rate": 3.093132230194758e-05,
+      "loss": 1.4551,
+      "step": 4482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08607627660481962,
+      "learning_rate": 3.0918143212769075e-05,
+      "loss": 1.4506,
+      "step": 4483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08326816986699975,
+      "learning_rate": 3.090496412359057e-05,
+      "loss": 1.4479,
+      "step": 4484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08020173026357096,
+      "learning_rate": 3.089178503441207e-05,
+      "loss": 1.4492,
+      "step": 4485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08917259908670722,
+      "learning_rate": 3.087860594523357e-05,
+      "loss": 1.4575,
+      "step": 4486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07968654676945207,
+      "learning_rate": 3.0865426856055056e-05,
+      "loss": 1.4551,
+      "step": 4487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09405584589529892,
+      "learning_rate": 3.085224776687656e-05,
+      "loss": 1.4445,
+      "step": 4488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0923050549342724,
+      "learning_rate": 3.083906867769806e-05,
+      "loss": 1.4518,
+      "step": 4489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08308517785856671,
+      "learning_rate": 3.082588958851955e-05,
+      "loss": 1.458,
+      "step": 4490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08652629379749409,
+      "learning_rate": 3.0812710499341045e-05,
+      "loss": 1.4524,
+      "step": 4491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08602562015365464,
+      "learning_rate": 3.0799531410162545e-05,
+      "loss": 1.4492,
+      "step": 4492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08271302136186938,
+      "learning_rate": 3.078635232098404e-05,
+      "loss": 1.4463,
+      "step": 4493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07939421533826087,
+      "learning_rate": 3.077317323180554e-05,
+      "loss": 1.4493,
+      "step": 4494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08217152884128309,
+      "learning_rate": 3.0759994142627033e-05,
+      "loss": 1.4615,
+      "step": 4495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08175155778275828,
+      "learning_rate": 3.074681505344853e-05,
+      "loss": 1.4528,
+      "step": 4496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08774781246318704,
+      "learning_rate": 3.073363596427003e-05,
+      "loss": 1.4472,
+      "step": 4497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08641220909827614,
+      "learning_rate": 3.072045687509153e-05,
+      "loss": 1.4554,
+      "step": 4498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0791919818908351,
+      "learning_rate": 3.0707277785913015e-05,
+      "loss": 1.457,
+      "step": 4499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07714157243850911,
+      "learning_rate": 3.0694098696734516e-05,
+      "loss": 1.4485,
+      "step": 4500
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.4203224182128906,
+      "eval_runtime": 549.2243,
+      "eval_samples_per_second": 32.801,
+      "eval_steps_per_second": 0.049,
+      "step": 4500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08356593696369101,
+      "learning_rate": 3.0680919607556016e-05,
+      "loss": 1.4538,
+      "step": 4501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09165145234198988,
+      "learning_rate": 3.066774051837751e-05,
+      "loss": 1.4536,
+      "step": 4502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09512299976388268,
+      "learning_rate": 3.0654561429199004e-05,
+      "loss": 1.4418,
+      "step": 4503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08672512262436763,
+      "learning_rate": 3.0641382340020504e-05,
+      "loss": 1.4501,
+      "step": 4504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08383922843633079,
+      "learning_rate": 3.0628203250842e-05,
+      "loss": 1.4514,
+      "step": 4505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0843936563969388,
+      "learning_rate": 3.06150241616635e-05,
+      "loss": 1.451,
+      "step": 4506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08735560110157085,
+      "learning_rate": 3.060184507248499e-05,
+      "loss": 1.4569,
+      "step": 4507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08689527992336371,
+      "learning_rate": 3.0588665983306486e-05,
+      "loss": 1.4492,
+      "step": 4508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08327031632283971,
+      "learning_rate": 3.0575486894127987e-05,
+      "loss": 1.4538,
+      "step": 4509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09061408158057449,
+      "learning_rate": 3.056230780494949e-05,
+      "loss": 1.4603,
+      "step": 4510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08568038576804399,
+      "learning_rate": 3.0549128715770974e-05,
+      "loss": 1.4544,
+      "step": 4511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08205872879814077,
+      "learning_rate": 3.0535949626592475e-05,
+      "loss": 1.4607,
+      "step": 4512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0928029280334615,
+      "learning_rate": 3.0522770537413975e-05,
+      "loss": 1.4633,
+      "step": 4513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08082402664062466,
+      "learning_rate": 3.050959144823547e-05,
+      "loss": 1.4506,
+      "step": 4514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09661349434371591,
+      "learning_rate": 3.0496412359056963e-05,
+      "loss": 1.4575,
+      "step": 4515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08503880628491993,
+      "learning_rate": 3.048323326987846e-05,
+      "loss": 1.4565,
+      "step": 4516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09099658825972995,
+      "learning_rate": 3.0470054180699957e-05,
+      "loss": 1.4534,
+      "step": 4517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0890650816512495,
+      "learning_rate": 3.0456875091521457e-05,
+      "loss": 1.4575,
+      "step": 4518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.09684321668105236,
+      "learning_rate": 3.0443696002342948e-05,
+      "loss": 1.4476,
+      "step": 4519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.10234373144306894,
+      "learning_rate": 3.0430516913164448e-05,
+      "loss": 1.4543,
+      "step": 4520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08415782251131573,
+      "learning_rate": 3.0417337823985945e-05,
+      "loss": 1.4515,
+      "step": 4521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08191163011480833,
+      "learning_rate": 3.0404158734807442e-05,
+      "loss": 1.454,
+      "step": 4522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08453094218145892,
+      "learning_rate": 3.0390979645628936e-05,
+      "loss": 1.4481,
+      "step": 4523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08508956588756905,
+      "learning_rate": 3.0377800556450433e-05,
+      "loss": 1.4553,
+      "step": 4524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07952116631056984,
+      "learning_rate": 3.036462146727193e-05,
+      "loss": 1.4493,
+      "step": 4525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07600467448537306,
+      "learning_rate": 3.0351442378093428e-05,
+      "loss": 1.4522,
+      "step": 4526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08153011391951004,
+      "learning_rate": 3.033826328891492e-05,
+      "loss": 1.4485,
+      "step": 4527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07755369847077705,
+      "learning_rate": 3.032508419973642e-05,
+      "loss": 1.4562,
+      "step": 4528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08211718533123372,
+      "learning_rate": 3.0311905110557916e-05,
+      "loss": 1.457,
+      "step": 4529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07598057400989146,
+      "learning_rate": 3.0298726021379416e-05,
+      "loss": 1.4528,
+      "step": 4530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08580344622272958,
+      "learning_rate": 3.0285546932200906e-05,
+      "loss": 1.4502,
+      "step": 4531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0913019002792481,
+      "learning_rate": 3.0272367843022407e-05,
+      "loss": 1.4542,
+      "step": 4532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.07344513226288683,
+      "learning_rate": 3.0259188753843904e-05,
+      "loss": 1.4567,
+      "step": 4533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08915743076762138,
+      "learning_rate": 3.02460096646654e-05,
+      "loss": 1.4511,
+      "step": 4534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08413501406520063,
+      "learning_rate": 3.0232830575486895e-05,
+      "loss": 1.4576,
+      "step": 4535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0839779439384897,
+      "learning_rate": 3.0219651486308392e-05,
+      "loss": 1.4526,
+      "step": 4536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08939478009998518,
+      "learning_rate": 3.020647239712989e-05,
+      "loss": 1.4567,
+      "step": 4537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08337889025760009,
+      "learning_rate": 3.0193293307951386e-05,
+      "loss": 1.4446,
+      "step": 4538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08935205876429751,
+      "learning_rate": 3.018011421877288e-05,
+      "loss": 1.4568,
+      "step": 4539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0883498497716169,
+      "learning_rate": 3.0166935129594377e-05,
+      "loss": 1.4557,
+      "step": 4540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.08727405993584944,
+      "learning_rate": 3.0153756040415874e-05,
+      "loss": 1.4418,
+      "step": 4541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0838402301102651,
+      "learning_rate": 3.0140576951237375e-05,
+      "loss": 1.458,
+      "step": 4542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08368129263829488,
+      "learning_rate": 3.0127397862058865e-05,
+      "loss": 1.4517,
+      "step": 4543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08587762389801423,
+      "learning_rate": 3.0114218772880366e-05,
+      "loss": 1.4565,
+      "step": 4544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0869094744660967,
+      "learning_rate": 3.0101039683701863e-05,
+      "loss": 1.4538,
+      "step": 4545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09016018782195655,
+      "learning_rate": 3.008786059452336e-05,
+      "loss": 1.4562,
+      "step": 4546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08281797307362836,
+      "learning_rate": 3.0074681505344854e-05,
+      "loss": 1.4565,
+      "step": 4547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09000567844862169,
+      "learning_rate": 3.006150241616635e-05,
+      "loss": 1.448,
+      "step": 4548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07835147132400604,
+      "learning_rate": 3.0048323326987848e-05,
+      "loss": 1.4432,
+      "step": 4549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08533934790057168,
+      "learning_rate": 3.0035144237809345e-05,
+      "loss": 1.4456,
+      "step": 4550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10465996215820199,
+      "learning_rate": 3.002196514863084e-05,
+      "loss": 1.4513,
+      "step": 4551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09123573426435794,
+      "learning_rate": 3.0008786059452336e-05,
+      "loss": 1.4473,
+      "step": 4552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08720962010974859,
+      "learning_rate": 2.9995606970273833e-05,
+      "loss": 1.451,
+      "step": 4553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10737688516371625,
+      "learning_rate": 2.9982427881095334e-05,
+      "loss": 1.4504,
+      "step": 4554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10013178384777048,
+      "learning_rate": 2.9969248791916824e-05,
+      "loss": 1.4548,
+      "step": 4555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08221822554960416,
+      "learning_rate": 2.9956069702738325e-05,
+      "loss": 1.4537,
+      "step": 4556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10366062971830026,
+      "learning_rate": 2.994289061355982e-05,
+      "loss": 1.4486,
+      "step": 4557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08688296331465387,
+      "learning_rate": 2.992971152438132e-05,
+      "loss": 1.4477,
+      "step": 4558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08949213023325238,
+      "learning_rate": 2.9916532435202813e-05,
+      "loss": 1.4594,
+      "step": 4559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08350690093335296,
+      "learning_rate": 2.990335334602431e-05,
+      "loss": 1.4577,
+      "step": 4560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0845009599561705,
+      "learning_rate": 2.9890174256845807e-05,
+      "loss": 1.4525,
+      "step": 4561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08701657666017208,
+      "learning_rate": 2.9876995167667304e-05,
+      "loss": 1.4564,
+      "step": 4562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08407601853385709,
+      "learning_rate": 2.9863816078488798e-05,
+      "loss": 1.4511,
+      "step": 4563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08736592734469392,
+      "learning_rate": 2.9850636989310295e-05,
+      "loss": 1.4458,
+      "step": 4564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08120208076798967,
+      "learning_rate": 2.9837457900131792e-05,
+      "loss": 1.4511,
+      "step": 4565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09590298688511056,
+      "learning_rate": 2.9824278810953292e-05,
+      "loss": 1.4485,
+      "step": 4566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0824599650939604,
+      "learning_rate": 2.9811099721774783e-05,
+      "loss": 1.4424,
+      "step": 4567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09421478310058519,
+      "learning_rate": 2.9797920632596283e-05,
+      "loss": 1.4596,
+      "step": 4568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08280930096881925,
+      "learning_rate": 2.978474154341778e-05,
+      "loss": 1.4551,
+      "step": 4569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09271378279616198,
+      "learning_rate": 2.9771562454239278e-05,
+      "loss": 1.4554,
+      "step": 4570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08521361249484924,
+      "learning_rate": 2.975838336506077e-05,
+      "loss": 1.4498,
+      "step": 4571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07882853265908257,
+      "learning_rate": 2.974520427588227e-05,
+      "loss": 1.4558,
+      "step": 4572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07844409347219758,
+      "learning_rate": 2.9732025186703766e-05,
+      "loss": 1.4501,
+      "step": 4573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07885744445162363,
+      "learning_rate": 2.9718846097525263e-05,
+      "loss": 1.4564,
+      "step": 4574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08243306204525723,
+      "learning_rate": 2.9705667008346756e-05,
+      "loss": 1.45,
+      "step": 4575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07472430449076868,
+      "learning_rate": 2.9692487919168254e-05,
+      "loss": 1.4571,
+      "step": 4576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07994447521479492,
+      "learning_rate": 2.967930882998975e-05,
+      "loss": 1.4471,
+      "step": 4577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0835954051588335,
+      "learning_rate": 2.966612974081125e-05,
+      "loss": 1.4484,
+      "step": 4578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07692095376404672,
+      "learning_rate": 2.965295065163274e-05,
+      "loss": 1.4552,
+      "step": 4579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08549932080926789,
+      "learning_rate": 2.963977156245424e-05,
+      "loss": 1.4499,
+      "step": 4580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08220562312836388,
+      "learning_rate": 2.962659247327574e-05,
+      "loss": 1.4627,
+      "step": 4581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08380946554754604,
+      "learning_rate": 2.9613413384097236e-05,
+      "loss": 1.4494,
+      "step": 4582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08375200315743402,
+      "learning_rate": 2.960023429491873e-05,
+      "loss": 1.4522,
+      "step": 4583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08498213769886453,
+      "learning_rate": 2.9587055205740227e-05,
+      "loss": 1.4569,
+      "step": 4584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07916688221981356,
+      "learning_rate": 2.9573876116561724e-05,
+      "loss": 1.4507,
+      "step": 4585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09169516854272375,
+      "learning_rate": 2.956069702738322e-05,
+      "loss": 1.4548,
+      "step": 4586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09779969617916318,
+      "learning_rate": 2.9547517938204715e-05,
+      "loss": 1.4565,
+      "step": 4587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09470678673563097,
+      "learning_rate": 2.9534338849026212e-05,
+      "loss": 1.4534,
+      "step": 4588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08991524932941135,
+      "learning_rate": 2.952115975984771e-05,
+      "loss": 1.4578,
+      "step": 4589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09669466091594565,
+      "learning_rate": 2.950798067066921e-05,
+      "loss": 1.4502,
+      "step": 4590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10576221548040503,
+      "learning_rate": 2.94948015814907e-05,
+      "loss": 1.448,
+      "step": 4591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09040445081462692,
+      "learning_rate": 2.9481622492312197e-05,
+      "loss": 1.4514,
+      "step": 4592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.10327775793621302,
+      "learning_rate": 2.9468443403133698e-05,
+      "loss": 1.4593,
+      "step": 4593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09695722170951568,
+      "learning_rate": 2.9455264313955195e-05,
+      "loss": 1.4482,
+      "step": 4594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09159265582513396,
+      "learning_rate": 2.944208522477669e-05,
+      "loss": 1.4529,
+      "step": 4595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09456844281568226,
+      "learning_rate": 2.9428906135598186e-05,
+      "loss": 1.4568,
+      "step": 4596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09128649737999947,
+      "learning_rate": 2.9415727046419683e-05,
+      "loss": 1.4509,
+      "step": 4597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09212675273278148,
+      "learning_rate": 2.940254795724118e-05,
+      "loss": 1.4525,
+      "step": 4598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.07755024220747553,
+      "learning_rate": 2.9389368868062674e-05,
+      "loss": 1.4565,
+      "step": 4599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09292434405379912,
+      "learning_rate": 2.937618977888417e-05,
+      "loss": 1.443,
+      "step": 4600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08226631977002516,
+      "learning_rate": 2.9363010689705668e-05,
+      "loss": 1.4455,
+      "step": 4601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08453136153759702,
+      "learning_rate": 2.934983160052717e-05,
+      "loss": 1.4436,
+      "step": 4602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08398084519335428,
+      "learning_rate": 2.933665251134866e-05,
+      "loss": 1.4544,
+      "step": 4603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08693474476032585,
+      "learning_rate": 2.9323473422170156e-05,
+      "loss": 1.451,
+      "step": 4604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0855361638611918,
+      "learning_rate": 2.9310294332991657e-05,
+      "loss": 1.4514,
+      "step": 4605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08369443062727408,
+      "learning_rate": 2.9297115243813154e-05,
+      "loss": 1.4607,
+      "step": 4606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.09330924220585464,
+      "learning_rate": 2.9283936154634648e-05,
+      "loss": 1.4522,
+      "step": 4607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08729563664338486,
+      "learning_rate": 2.9270757065456145e-05,
+      "loss": 1.4451,
+      "step": 4608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08078361453915936,
+      "learning_rate": 2.9257577976277642e-05,
+      "loss": 1.4457,
+      "step": 4609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.08539984019444716,
+      "learning_rate": 2.924439888709914e-05,
+      "loss": 1.4558,
+      "step": 4610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07493613753653947,
+      "learning_rate": 2.9231219797920633e-05,
+      "loss": 1.4545,
+      "step": 4611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08093723089208055,
+      "learning_rate": 2.921804070874213e-05,
+      "loss": 1.4491,
+      "step": 4612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08391229047486859,
+      "learning_rate": 2.9204861619563627e-05,
+      "loss": 1.455,
+      "step": 4613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07317993872530625,
+      "learning_rate": 2.9191682530385128e-05,
+      "loss": 1.4456,
+      "step": 4614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08581642041319724,
+      "learning_rate": 2.9178503441206618e-05,
+      "loss": 1.4506,
+      "step": 4615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08183059419544549,
+      "learning_rate": 2.9165324352028115e-05,
+      "loss": 1.4504,
+      "step": 4616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08167582199024016,
+      "learning_rate": 2.9152145262849616e-05,
+      "loss": 1.4541,
+      "step": 4617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08397053318523638,
+      "learning_rate": 2.9138966173671113e-05,
+      "loss": 1.4554,
+      "step": 4618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08193283191743575,
+      "learning_rate": 2.9125787084492606e-05,
+      "loss": 1.4458,
+      "step": 4619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0748804089299795,
+      "learning_rate": 2.9112607995314104e-05,
+      "loss": 1.4533,
+      "step": 4620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0895393030826907,
+      "learning_rate": 2.90994289061356e-05,
+      "loss": 1.4511,
+      "step": 4621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10356685349550286,
+      "learning_rate": 2.9086249816957098e-05,
+      "loss": 1.4457,
+      "step": 4622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07769092967262105,
+      "learning_rate": 2.907307072777859e-05,
+      "loss": 1.4532,
+      "step": 4623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08543961197170315,
+      "learning_rate": 2.905989163860009e-05,
+      "loss": 1.4523,
+      "step": 4624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08791175934034158,
+      "learning_rate": 2.9046712549421586e-05,
+      "loss": 1.4517,
+      "step": 4625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08095071654183318,
+      "learning_rate": 2.9033533460243083e-05,
+      "loss": 1.4529,
+      "step": 4626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08010592629678795,
+      "learning_rate": 2.9020354371064577e-05,
+      "loss": 1.4558,
+      "step": 4627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09010538844268881,
+      "learning_rate": 2.9007175281886074e-05,
+      "loss": 1.4543,
+      "step": 4628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08162557815715987,
+      "learning_rate": 2.8993996192707574e-05,
+      "loss": 1.454,
+      "step": 4629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08937475333708497,
+      "learning_rate": 2.898081710352907e-05,
+      "loss": 1.4513,
+      "step": 4630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08374598427297755,
+      "learning_rate": 2.8967638014350565e-05,
+      "loss": 1.4514,
+      "step": 4631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08900114522540892,
+      "learning_rate": 2.8954458925172062e-05,
+      "loss": 1.4506,
+      "step": 4632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09042595074496407,
+      "learning_rate": 2.894127983599356e-05,
+      "loss": 1.4446,
+      "step": 4633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0811156196642568,
+      "learning_rate": 2.8928100746815057e-05,
+      "loss": 1.45,
+      "step": 4634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09016703252773726,
+      "learning_rate": 2.891492165763655e-05,
+      "loss": 1.4435,
+      "step": 4635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0929893375919825,
+      "learning_rate": 2.8901742568458047e-05,
+      "loss": 1.4486,
+      "step": 4636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08022936500478967,
+      "learning_rate": 2.8888563479279545e-05,
+      "loss": 1.4553,
+      "step": 4637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10197299070620164,
+      "learning_rate": 2.8875384390101042e-05,
+      "loss": 1.4571,
+      "step": 4638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07945963206231542,
+      "learning_rate": 2.8862205300922535e-05,
+      "loss": 1.4426,
+      "step": 4639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08889916168350295,
+      "learning_rate": 2.8849026211744033e-05,
+      "loss": 1.4529,
+      "step": 4640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07773330481621414,
+      "learning_rate": 2.8835847122565533e-05,
+      "loss": 1.4547,
+      "step": 4641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07825726055570628,
+      "learning_rate": 2.882266803338703e-05,
+      "loss": 1.4457,
+      "step": 4642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07514486848891966,
+      "learning_rate": 2.8809488944208524e-05,
+      "loss": 1.4478,
+      "step": 4643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07931682915224512,
+      "learning_rate": 2.879630985503002e-05,
+      "loss": 1.4491,
+      "step": 4644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08543266559780685,
+      "learning_rate": 2.8783130765851518e-05,
+      "loss": 1.445,
+      "step": 4645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07467877174886342,
+      "learning_rate": 2.8769951676673015e-05,
+      "loss": 1.4534,
+      "step": 4646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0929163276390207,
+      "learning_rate": 2.875677258749451e-05,
+      "loss": 1.4484,
+      "step": 4647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08955988763630049,
+      "learning_rate": 2.8743593498316006e-05,
+      "loss": 1.4446,
+      "step": 4648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07767224839048038,
+      "learning_rate": 2.8730414409137503e-05,
+      "loss": 1.4563,
+      "step": 4649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09875440850361732,
+      "learning_rate": 2.8717235319959e-05,
+      "loss": 1.452,
+      "step": 4650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09272512972905479,
+      "learning_rate": 2.8704056230780494e-05,
+      "loss": 1.4531,
+      "step": 4651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08346237952672055,
+      "learning_rate": 2.869087714160199e-05,
+      "loss": 1.4545,
+      "step": 4652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09479592235911957,
+      "learning_rate": 2.8677698052423492e-05,
+      "loss": 1.4573,
+      "step": 4653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09027089647479691,
+      "learning_rate": 2.866451896324499e-05,
+      "loss": 1.4599,
+      "step": 4654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09086682889068703,
+      "learning_rate": 2.8651339874066483e-05,
+      "loss": 1.4508,
+      "step": 4655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08561021754130817,
+      "learning_rate": 2.863816078488798e-05,
+      "loss": 1.4544,
+      "step": 4656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08157994819583177,
+      "learning_rate": 2.8624981695709477e-05,
+      "loss": 1.4544,
+      "step": 4657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0869789518669381,
+      "learning_rate": 2.8611802606530974e-05,
+      "loss": 1.4453,
+      "step": 4658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07643984632914762,
+      "learning_rate": 2.8598623517352468e-05,
+      "loss": 1.4412,
+      "step": 4659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08404877954246472,
+      "learning_rate": 2.8585444428173965e-05,
+      "loss": 1.4539,
+      "step": 4660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07704687687889049,
+      "learning_rate": 2.8572265338995462e-05,
+      "loss": 1.4491,
+      "step": 4661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07762891922130172,
+      "learning_rate": 2.855908624981696e-05,
+      "loss": 1.4549,
+      "step": 4662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07538130935669118,
+      "learning_rate": 2.8545907160638453e-05,
+      "loss": 1.4527,
+      "step": 4663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07536582353408797,
+      "learning_rate": 2.853272807145995e-05,
+      "loss": 1.4407,
+      "step": 4664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08112948384895526,
+      "learning_rate": 2.851954898228145e-05,
+      "loss": 1.4542,
+      "step": 4665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0885270773418604,
+      "learning_rate": 2.8506369893102948e-05,
+      "loss": 1.4441,
+      "step": 4666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07859082065928558,
+      "learning_rate": 2.8493190803924438e-05,
+      "loss": 1.4547,
+      "step": 4667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08684702211201384,
+      "learning_rate": 2.848001171474594e-05,
+      "loss": 1.4468,
+      "step": 4668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08909282522164097,
+      "learning_rate": 2.8466832625567436e-05,
+      "loss": 1.4506,
+      "step": 4669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.07914463528248371,
+      "learning_rate": 2.8453653536388933e-05,
+      "loss": 1.4493,
+      "step": 4670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10078961832019201,
+      "learning_rate": 2.8440474447210427e-05,
+      "loss": 1.4484,
+      "step": 4671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08636589955869338,
+      "learning_rate": 2.8427295358031924e-05,
+      "loss": 1.4496,
+      "step": 4672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08445518066288324,
+      "learning_rate": 2.841411626885342e-05,
+      "loss": 1.4535,
+      "step": 4673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.09236093740535598,
+      "learning_rate": 2.8400937179674918e-05,
+      "loss": 1.4468,
+      "step": 4674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08049863718877365,
+      "learning_rate": 2.8387758090496412e-05,
+      "loss": 1.4473,
+      "step": 4675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08302783079412268,
+      "learning_rate": 2.837457900131791e-05,
+      "loss": 1.4501,
+      "step": 4676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.10436984536790785,
+      "learning_rate": 2.836139991213941e-05,
+      "loss": 1.4447,
+      "step": 4677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.08375191439626568,
+      "learning_rate": 2.8348220822960907e-05,
+      "loss": 1.4477,
+      "step": 4678
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08857322327287176,
+      "learning_rate": 2.8335041733782397e-05,
+      "loss": 1.439,
+      "step": 4679
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09505243848436914,
+      "learning_rate": 2.8321862644603897e-05,
+      "loss": 1.4505,
+      "step": 4680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08235886782620656,
+      "learning_rate": 2.8308683555425395e-05,
+      "loss": 1.4488,
+      "step": 4681
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08761943965926765,
+      "learning_rate": 2.8295504466246892e-05,
+      "loss": 1.4536,
+      "step": 4682
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0950930158914687,
+      "learning_rate": 2.8282325377068385e-05,
+      "loss": 1.4582,
+      "step": 4683
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0814790314458038,
+      "learning_rate": 2.8269146287889883e-05,
+      "loss": 1.4497,
+      "step": 4684
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08558794914381099,
+      "learning_rate": 2.825596719871138e-05,
+      "loss": 1.4522,
+      "step": 4685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0935734246737233,
+      "learning_rate": 2.8242788109532877e-05,
+      "loss": 1.4492,
+      "step": 4686
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08023142428224887,
+      "learning_rate": 2.822960902035437e-05,
+      "loss": 1.4542,
+      "step": 4687
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09114683203313578,
+      "learning_rate": 2.8216429931175868e-05,
+      "loss": 1.4512,
+      "step": 4688
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08875545690675923,
+      "learning_rate": 2.8203250841997368e-05,
+      "loss": 1.4467,
+      "step": 4689
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08248003688981376,
+      "learning_rate": 2.8190071752818865e-05,
+      "loss": 1.4541,
+      "step": 4690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.10450013879449245,
+      "learning_rate": 2.8176892663640356e-05,
+      "loss": 1.454,
+      "step": 4691
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.093677138886609,
+      "learning_rate": 2.8163713574461856e-05,
+      "loss": 1.4592,
+      "step": 4692
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08607964920240532,
+      "learning_rate": 2.8150534485283353e-05,
+      "loss": 1.4508,
+      "step": 4693
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08070248133052339,
+      "learning_rate": 2.813735539610485e-05,
+      "loss": 1.452,
+      "step": 4694
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08431319340737728,
+      "learning_rate": 2.8124176306926344e-05,
+      "loss": 1.4377,
+      "step": 4695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08373095380974385,
+      "learning_rate": 2.811099721774784e-05,
+      "loss": 1.4531,
+      "step": 4696
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0803815204610658,
+      "learning_rate": 2.809781812856934e-05,
+      "loss": 1.4486,
+      "step": 4697
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08608468142853881,
+      "learning_rate": 2.8084639039390836e-05,
+      "loss": 1.4469,
+      "step": 4698
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08312299566905326,
+      "learning_rate": 2.807145995021233e-05,
+      "loss": 1.4509,
+      "step": 4699
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08483027879896984,
+      "learning_rate": 2.8058280861033827e-05,
+      "loss": 1.4568,
+      "step": 4700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07998439949456133,
+      "learning_rate": 2.8045101771855324e-05,
+      "loss": 1.4492,
+      "step": 4701
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0728622101620674,
+      "learning_rate": 2.8031922682676824e-05,
+      "loss": 1.4601,
+      "step": 4702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07640259303322634,
+      "learning_rate": 2.8018743593498315e-05,
+      "loss": 1.4504,
+      "step": 4703
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07842252489995029,
+      "learning_rate": 2.8005564504319815e-05,
+      "loss": 1.4491,
+      "step": 4704
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09102240278717807,
+      "learning_rate": 2.7992385415141312e-05,
+      "loss": 1.4471,
+      "step": 4705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0762844623279824,
+      "learning_rate": 2.797920632596281e-05,
+      "loss": 1.4533,
+      "step": 4706
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08031490631588435,
+      "learning_rate": 2.7966027236784303e-05,
+      "loss": 1.4529,
+      "step": 4707
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08120332281601293,
+      "learning_rate": 2.79528481476058e-05,
+      "loss": 1.4549,
+      "step": 4708
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08070031481961136,
+      "learning_rate": 2.7939669058427297e-05,
+      "loss": 1.4434,
+      "step": 4709
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07870560589159324,
+      "learning_rate": 2.7926489969248794e-05,
+      "loss": 1.4568,
+      "step": 4710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08343125220453432,
+      "learning_rate": 2.7913310880070288e-05,
+      "loss": 1.4493,
+      "step": 4711
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07394836768785543,
+      "learning_rate": 2.7900131790891785e-05,
+      "loss": 1.4486,
+      "step": 4712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08228956691223728,
+      "learning_rate": 2.7886952701713282e-05,
+      "loss": 1.4472,
+      "step": 4713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0763642307285805,
+      "learning_rate": 2.7873773612534783e-05,
+      "loss": 1.4532,
+      "step": 4714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08263802562965704,
+      "learning_rate": 2.7860594523356273e-05,
+      "loss": 1.4452,
+      "step": 4715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09494922326387883,
+      "learning_rate": 2.7847415434177774e-05,
+      "loss": 1.4499,
+      "step": 4716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08400446226128407,
+      "learning_rate": 2.783423634499927e-05,
+      "loss": 1.452,
+      "step": 4717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08787781230488946,
+      "learning_rate": 2.7821057255820768e-05,
+      "loss": 1.4437,
+      "step": 4718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09168527607510248,
+      "learning_rate": 2.7807878166642262e-05,
+      "loss": 1.4521,
+      "step": 4719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07544002146204742,
+      "learning_rate": 2.779469907746376e-05,
+      "loss": 1.4534,
+      "step": 4720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09606076768579346,
+      "learning_rate": 2.7781519988285256e-05,
+      "loss": 1.4497,
+      "step": 4721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.10079721000825605,
+      "learning_rate": 2.7768340899106753e-05,
+      "loss": 1.4558,
+      "step": 4722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08327336283085854,
+      "learning_rate": 2.7755161809928247e-05,
+      "loss": 1.4628,
+      "step": 4723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0992081496174711,
+      "learning_rate": 2.7741982720749744e-05,
+      "loss": 1.4453,
+      "step": 4724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08503909634983722,
+      "learning_rate": 2.772880363157124e-05,
+      "loss": 1.446,
+      "step": 4725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09492195797144382,
+      "learning_rate": 2.7715624542392742e-05,
+      "loss": 1.4482,
+      "step": 4726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08990282158309862,
+      "learning_rate": 2.7702445453214232e-05,
+      "loss": 1.4535,
+      "step": 4727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09871317442051923,
+      "learning_rate": 2.7689266364035733e-05,
+      "loss": 1.4565,
+      "step": 4728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09228694568557795,
+      "learning_rate": 2.767608727485723e-05,
+      "loss": 1.4463,
+      "step": 4729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08221693060689243,
+      "learning_rate": 2.7662908185678727e-05,
+      "loss": 1.4515,
+      "step": 4730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.09517219153441879,
+      "learning_rate": 2.764972909650022e-05,
+      "loss": 1.4471,
+      "step": 4731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07689117894941049,
+      "learning_rate": 2.7636550007321718e-05,
+      "loss": 1.4532,
+      "step": 4732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.079275314526976,
+      "learning_rate": 2.7623370918143215e-05,
+      "loss": 1.4462,
+      "step": 4733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0825516018089213,
+      "learning_rate": 2.7610191828964712e-05,
+      "loss": 1.4532,
+      "step": 4734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07803720507494528,
+      "learning_rate": 2.7597012739786206e-05,
+      "loss": 1.4488,
+      "step": 4735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07439753391650794,
+      "learning_rate": 2.7583833650607703e-05,
+      "loss": 1.4496,
+      "step": 4736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07775253959136212,
+      "learning_rate": 2.75706545614292e-05,
+      "loss": 1.455,
+      "step": 4737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0740174313900591,
+      "learning_rate": 2.75574754722507e-05,
+      "loss": 1.4482,
+      "step": 4738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07902211353305803,
+      "learning_rate": 2.754429638307219e-05,
+      "loss": 1.4592,
+      "step": 4739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0792119411813395,
+      "learning_rate": 2.753111729389369e-05,
+      "loss": 1.4545,
+      "step": 4740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08873840018099641,
+      "learning_rate": 2.751793820471519e-05,
+      "loss": 1.4551,
+      "step": 4741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08165445620461353,
+      "learning_rate": 2.7504759115536686e-05,
+      "loss": 1.4419,
+      "step": 4742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.07950959004901953,
+      "learning_rate": 2.749158002635818e-05,
+      "loss": 1.4461,
+      "step": 4743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08105777618086225,
+      "learning_rate": 2.7478400937179677e-05,
+      "loss": 1.4394,
+      "step": 4744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08113437863577391,
+      "learning_rate": 2.7465221848001174e-05,
+      "loss": 1.4508,
+      "step": 4745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.08340961735642223,
+      "learning_rate": 2.745204275882267e-05,
+      "loss": 1.4607,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09323222570933173,
+      "learning_rate": 2.7438863669644165e-05,
+      "loss": 1.4545,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0849884720646695,
+      "learning_rate": 2.742568458046566e-05,
+      "loss": 1.4391,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0907454257057709,
+      "learning_rate": 2.741250549128716e-05,
+      "loss": 1.4523,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08872248291137332,
+      "learning_rate": 2.739932640210866e-05,
+      "loss": 1.4502,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08606933965648479,
+      "learning_rate": 2.738614731293015e-05,
+      "loss": 1.4432,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08610671172227025,
+      "learning_rate": 2.737296822375165e-05,
+      "loss": 1.4549,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08543664205570785,
+      "learning_rate": 2.7359789134573147e-05,
+      "loss": 1.4497,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0738926947730651,
+      "learning_rate": 2.7346610045394644e-05,
+      "loss": 1.4491,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09589251032215845,
+      "learning_rate": 2.7333430956216138e-05,
+      "loss": 1.4554,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07867672239356599,
+      "learning_rate": 2.7320251867037635e-05,
+      "loss": 1.4585,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0935639571860841,
+      "learning_rate": 2.7307072777859132e-05,
+      "loss": 1.4407,
+      "step": 4757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08501768340481837,
+      "learning_rate": 2.729389368868063e-05,
+      "loss": 1.447,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08289231856821648,
+      "learning_rate": 2.7280714599502123e-05,
+      "loss": 1.4492,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07806771108792677,
+      "learning_rate": 2.726753551032362e-05,
+      "loss": 1.4389,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08292829822082114,
+      "learning_rate": 2.7254356421145118e-05,
+      "loss": 1.4489,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08073956032208286,
+      "learning_rate": 2.7241177331966618e-05,
+      "loss": 1.4469,
+      "step": 4762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08228163558123665,
+      "learning_rate": 2.722799824278811e-05,
+      "loss": 1.4439,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07598288194720217,
+      "learning_rate": 2.721481915360961e-05,
+      "loss": 1.4471,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08693409225391108,
+      "learning_rate": 2.7201640064431106e-05,
+      "loss": 1.4495,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07844558897393202,
+      "learning_rate": 2.7188460975252603e-05,
+      "loss": 1.44,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08210115546759134,
+      "learning_rate": 2.7175281886074097e-05,
+      "loss": 1.4492,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07586295889308334,
+      "learning_rate": 2.7162102796895594e-05,
+      "loss": 1.4489,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07644976236550086,
+      "learning_rate": 2.714892370771709e-05,
+      "loss": 1.4427,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08097450936340676,
+      "learning_rate": 2.713574461853859e-05,
+      "loss": 1.4464,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07764234092730872,
+      "learning_rate": 2.7122565529360082e-05,
+      "loss": 1.4517,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08088280522127649,
+      "learning_rate": 2.710938644018158e-05,
+      "loss": 1.4528,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08536133885555597,
+      "learning_rate": 2.7096207351003076e-05,
+      "loss": 1.4514,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0918096303232219,
+      "learning_rate": 2.7083028261824577e-05,
+      "loss": 1.4546,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0870110347929681,
+      "learning_rate": 2.7069849172646067e-05,
+      "loss": 1.4518,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08180077194319439,
+      "learning_rate": 2.7056670083467568e-05,
+      "loss": 1.4522,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08142062609740666,
+      "learning_rate": 2.7043490994289065e-05,
+      "loss": 1.4463,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07996949459502392,
+      "learning_rate": 2.7030311905110562e-05,
+      "loss": 1.4507,
+      "step": 4778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08368161988899189,
+      "learning_rate": 2.7017132815932056e-05,
+      "loss": 1.4476,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07905188820317437,
+      "learning_rate": 2.7003953726753553e-05,
+      "loss": 1.4509,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0863397888333044,
+      "learning_rate": 2.699077463757505e-05,
+      "loss": 1.4401,
+      "step": 4781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0771833000601143,
+      "learning_rate": 2.6977595548396547e-05,
+      "loss": 1.4547,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08928939871341285,
+      "learning_rate": 2.696441645921804e-05,
+      "loss": 1.4489,
+      "step": 4783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08801649013591803,
+      "learning_rate": 2.6951237370039538e-05,
+      "loss": 1.4601,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09406836474174589,
+      "learning_rate": 2.6938058280861035e-05,
+      "loss": 1.4495,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08777732543623297,
+      "learning_rate": 2.6924879191682536e-05,
+      "loss": 1.4503,
+      "step": 4786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.09329667475446472,
+      "learning_rate": 2.6911700102504026e-05,
+      "loss": 1.4427,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08784211696716634,
+      "learning_rate": 2.6898521013325523e-05,
+      "loss": 1.4525,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07998819415915448,
+      "learning_rate": 2.6885341924147024e-05,
+      "loss": 1.4503,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08592223889034817,
+      "learning_rate": 2.687216283496852e-05,
+      "loss": 1.4487,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07697323378811645,
+      "learning_rate": 2.6858983745790015e-05,
+      "loss": 1.4449,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08247874335406989,
+      "learning_rate": 2.684580465661151e-05,
+      "loss": 1.4391,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08588070192241365,
+      "learning_rate": 2.683262556743301e-05,
+      "loss": 1.4467,
+      "step": 4793
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08056038645717031,
+      "learning_rate": 2.6819446478254506e-05,
+      "loss": 1.4514,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08191060581693814,
+      "learning_rate": 2.6806267389076e-05,
+      "loss": 1.4537,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08059852377158154,
+      "learning_rate": 2.6793088299897497e-05,
+      "loss": 1.4568,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07561417558305641,
+      "learning_rate": 2.6779909210718994e-05,
+      "loss": 1.4466,
+      "step": 4797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07779336330847632,
+      "learning_rate": 2.6766730121540494e-05,
+      "loss": 1.4512,
+      "step": 4798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07819644843016625,
+      "learning_rate": 2.6753551032361985e-05,
+      "loss": 1.4483,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08052702135261901,
+      "learning_rate": 2.6740371943183482e-05,
+      "loss": 1.4537,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0704502817383074,
+      "learning_rate": 2.6727192854004982e-05,
+      "loss": 1.4463,
+      "step": 4801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07689101255942346,
+      "learning_rate": 2.671401376482648e-05,
+      "loss": 1.4484,
+      "step": 4802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06951821814966018,
+      "learning_rate": 2.6700834675647973e-05,
+      "loss": 1.4441,
+      "step": 4803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07761207082570747,
+      "learning_rate": 2.668765558646947e-05,
+      "loss": 1.439,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07089677912667228,
+      "learning_rate": 2.6674476497290968e-05,
+      "loss": 1.4477,
+      "step": 4805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07853582999857059,
+      "learning_rate": 2.6661297408112465e-05,
+      "loss": 1.4402,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07051683782134649,
+      "learning_rate": 2.664811831893396e-05,
+      "loss": 1.4483,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07872402279834952,
+      "learning_rate": 2.6634939229755456e-05,
+      "loss": 1.4536,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08261603310721752,
+      "learning_rate": 2.6621760140576953e-05,
+      "loss": 1.4556,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07944494784505189,
+      "learning_rate": 2.6608581051398453e-05,
+      "loss": 1.4577,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.08669400152961766,
+      "learning_rate": 2.6595401962219944e-05,
+      "loss": 1.4506,
+      "step": 4811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.06898685321926183,
+      "learning_rate": 2.658222287304144e-05,
+      "loss": 1.4395,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.07942730686774932,
+      "learning_rate": 2.656904378386294e-05,
+      "loss": 1.4406,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0725898004779907,
+      "learning_rate": 2.655586469468444e-05,
+      "loss": 1.4499,
+      "step": 4814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07854270453077,
+      "learning_rate": 2.6542685605505932e-05,
+      "loss": 1.447,
+      "step": 4815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07641183774007594,
+      "learning_rate": 2.652950651632743e-05,
+      "loss": 1.4486,
+      "step": 4816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08136679413545095,
+      "learning_rate": 2.6516327427148926e-05,
+      "loss": 1.4572,
+      "step": 4817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07599492355938085,
+      "learning_rate": 2.6503148337970424e-05,
+      "loss": 1.4473,
+      "step": 4818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08248019188656498,
+      "learning_rate": 2.6489969248791917e-05,
+      "loss": 1.452,
+      "step": 4819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07928970235330103,
+      "learning_rate": 2.6476790159613414e-05,
+      "loss": 1.4452,
+      "step": 4820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07019800890038186,
+      "learning_rate": 2.646361107043491e-05,
+      "loss": 1.4504,
+      "step": 4821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07710554622628167,
+      "learning_rate": 2.6450431981256412e-05,
+      "loss": 1.4492,
+      "step": 4822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07634469623040298,
+      "learning_rate": 2.6437252892077902e-05,
+      "loss": 1.4469,
+      "step": 4823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07472834378687636,
+      "learning_rate": 2.64240738028994e-05,
+      "loss": 1.4467,
+      "step": 4824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07306513572220284,
+      "learning_rate": 2.64108947137209e-05,
+      "loss": 1.449,
+      "step": 4825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07674259311071616,
+      "learning_rate": 2.6397715624542397e-05,
+      "loss": 1.4556,
+      "step": 4826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07948120816856642,
+      "learning_rate": 2.638453653536389e-05,
+      "loss": 1.4447,
+      "step": 4827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07703193349898022,
+      "learning_rate": 2.6371357446185388e-05,
+      "loss": 1.4386,
+      "step": 4828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0708216995002924,
+      "learning_rate": 2.6358178357006885e-05,
+      "loss": 1.4437,
+      "step": 4829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07258301243731959,
+      "learning_rate": 2.6344999267828382e-05,
+      "loss": 1.4454,
+      "step": 4830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0750206161566651,
+      "learning_rate": 2.6331820178649876e-05,
+      "loss": 1.4431,
+      "step": 4831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07608470349867763,
+      "learning_rate": 2.6318641089471373e-05,
+      "loss": 1.4443,
+      "step": 4832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0828362864060802,
+      "learning_rate": 2.630546200029287e-05,
+      "loss": 1.4471,
+      "step": 4833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08007633976492447,
+      "learning_rate": 2.6292282911114367e-05,
+      "loss": 1.4489,
+      "step": 4834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08728240510691865,
+      "learning_rate": 2.627910382193586e-05,
+      "loss": 1.4478,
+      "step": 4835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08178982627806854,
+      "learning_rate": 2.6265924732757358e-05,
+      "loss": 1.4517,
+      "step": 4836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08371048636841773,
+      "learning_rate": 2.625274564357886e-05,
+      "loss": 1.4469,
+      "step": 4837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08549108022696437,
+      "learning_rate": 2.6239566554400356e-05,
+      "loss": 1.4516,
+      "step": 4838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07820057972422832,
+      "learning_rate": 2.622638746522185e-05,
+      "loss": 1.4424,
+      "step": 4839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08559166202859,
+      "learning_rate": 2.6213208376043347e-05,
+      "loss": 1.4516,
+      "step": 4840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.087967095673252,
+      "learning_rate": 2.6200029286864844e-05,
+      "loss": 1.4454,
+      "step": 4841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07383751389184598,
+      "learning_rate": 2.618685019768634e-05,
+      "loss": 1.4476,
+      "step": 4842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.09059412644482066,
+      "learning_rate": 2.6173671108507835e-05,
+      "loss": 1.4488,
+      "step": 4843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07513608366258553,
+      "learning_rate": 2.6160492019329332e-05,
+      "loss": 1.4548,
+      "step": 4844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0862628852305564,
+      "learning_rate": 2.614731293015083e-05,
+      "loss": 1.4531,
+      "step": 4845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08158895245247359,
+      "learning_rate": 2.6134133840972326e-05,
+      "loss": 1.443,
+      "step": 4846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07738405753669782,
+      "learning_rate": 2.612095475179382e-05,
+      "loss": 1.4513,
+      "step": 4847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08429538733760977,
+      "learning_rate": 2.6107775662615317e-05,
+      "loss": 1.447,
+      "step": 4848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07800260021138349,
+      "learning_rate": 2.6094596573436818e-05,
+      "loss": 1.452,
+      "step": 4849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07689934254913565,
+      "learning_rate": 2.6081417484258315e-05,
+      "loss": 1.4466,
+      "step": 4850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08031477421493112,
+      "learning_rate": 2.606823839507981e-05,
+      "loss": 1.4523,
+      "step": 4851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0856928872421778,
+      "learning_rate": 2.6055059305901306e-05,
+      "loss": 1.4471,
+      "step": 4852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07841982152084892,
+      "learning_rate": 2.6041880216722803e-05,
+      "loss": 1.4457,
+      "step": 4853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08416400513216026,
+      "learning_rate": 2.60287011275443e-05,
+      "loss": 1.4456,
+      "step": 4854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07880292418435589,
+      "learning_rate": 2.6015522038365794e-05,
+      "loss": 1.4411,
+      "step": 4855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07675282299742109,
+      "learning_rate": 2.600234294918729e-05,
+      "loss": 1.4502,
+      "step": 4856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07879103972883483,
+      "learning_rate": 2.5989163860008788e-05,
+      "loss": 1.4495,
+      "step": 4857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0785515865321456,
+      "learning_rate": 2.5975984770830285e-05,
+      "loss": 1.4525,
+      "step": 4858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07821002706021246,
+      "learning_rate": 2.596280568165178e-05,
+      "loss": 1.4467,
+      "step": 4859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07371888345426184,
+      "learning_rate": 2.5949626592473276e-05,
+      "loss": 1.4511,
+      "step": 4860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07962672965039923,
+      "learning_rate": 2.5936447503294776e-05,
+      "loss": 1.4471,
+      "step": 4861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07528960466478021,
+      "learning_rate": 2.5923268414116274e-05,
+      "loss": 1.4451,
+      "step": 4862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08234948860466204,
+      "learning_rate": 2.5910089324937767e-05,
+      "loss": 1.4412,
+      "step": 4863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07399627807639253,
+      "learning_rate": 2.5896910235759264e-05,
+      "loss": 1.4474,
+      "step": 4864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0777394602079414,
+      "learning_rate": 2.588373114658076e-05,
+      "loss": 1.4418,
+      "step": 4865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08149819999451635,
+      "learning_rate": 2.587055205740226e-05,
+      "loss": 1.4561,
+      "step": 4866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07525508835037248,
+      "learning_rate": 2.5857372968223752e-05,
+      "loss": 1.4488,
+      "step": 4867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07557277710745296,
+      "learning_rate": 2.584419387904525e-05,
+      "loss": 1.4542,
+      "step": 4868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0791281207373783,
+      "learning_rate": 2.5831014789866747e-05,
+      "loss": 1.4459,
+      "step": 4869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07179586767376098,
+      "learning_rate": 2.5817835700688244e-05,
+      "loss": 1.4424,
+      "step": 4870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0704102179124005,
+      "learning_rate": 2.5804656611509738e-05,
+      "loss": 1.4429,
+      "step": 4871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07404031037395714,
+      "learning_rate": 2.5791477522331235e-05,
+      "loss": 1.4452,
+      "step": 4872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07790048907674996,
+      "learning_rate": 2.5778298433152735e-05,
+      "loss": 1.4582,
+      "step": 4873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08035833686621398,
+      "learning_rate": 2.5765119343974232e-05,
+      "loss": 1.4537,
+      "step": 4874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07470629972988314,
+      "learning_rate": 2.5751940254795723e-05,
+      "loss": 1.4423,
+      "step": 4875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0798160216545669,
+      "learning_rate": 2.5738761165617223e-05,
+      "loss": 1.4528,
+      "step": 4876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.08764712437680355,
+      "learning_rate": 2.572558207643872e-05,
+      "loss": 1.4464,
+      "step": 4877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07355864201374229,
+      "learning_rate": 2.5712402987260217e-05,
+      "loss": 1.4548,
+      "step": 4878
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07978017799164377,
+      "learning_rate": 2.569922389808171e-05,
+      "loss": 1.4515,
+      "step": 4879
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0714523162719173,
+      "learning_rate": 2.5686044808903208e-05,
+      "loss": 1.4571,
+      "step": 4880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07384919161089659,
+      "learning_rate": 2.5672865719724705e-05,
+      "loss": 1.4437,
+      "step": 4881
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.07371100544170531,
+      "learning_rate": 2.5659686630546203e-05,
+      "loss": 1.443,
+      "step": 4882
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.074301296434628,
+      "learning_rate": 2.5646507541367696e-05,
+      "loss": 1.4524,
+      "step": 4883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07657999270518648,
+      "learning_rate": 2.5633328452189193e-05,
+      "loss": 1.4498,
+      "step": 4884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07439078613339488,
+      "learning_rate": 2.5620149363010694e-05,
+      "loss": 1.4484,
+      "step": 4885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07277292992132538,
+      "learning_rate": 2.560697027383219e-05,
+      "loss": 1.4448,
+      "step": 4886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08008652573478621,
+      "learning_rate": 2.559379118465368e-05,
+      "loss": 1.448,
+      "step": 4887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08070902701424183,
+      "learning_rate": 2.5580612095475182e-05,
+      "loss": 1.4422,
+      "step": 4888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08157749739341305,
+      "learning_rate": 2.556743300629668e-05,
+      "loss": 1.445,
+      "step": 4889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0787520025216102,
+      "learning_rate": 2.5554253917118176e-05,
+      "loss": 1.4561,
+      "step": 4890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09426583202869902,
+      "learning_rate": 2.554107482793967e-05,
+      "loss": 1.4509,
+      "step": 4891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09763334484558628,
+      "learning_rate": 2.5527895738761167e-05,
+      "loss": 1.4425,
+      "step": 4892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08156582858456807,
+      "learning_rate": 2.5514716649582664e-05,
+      "loss": 1.4486,
+      "step": 4893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0836106273340306,
+      "learning_rate": 2.550153756040416e-05,
+      "loss": 1.4492,
+      "step": 4894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08394163205633925,
+      "learning_rate": 2.5488358471225655e-05,
+      "loss": 1.4489,
+      "step": 4895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0720338406298501,
+      "learning_rate": 2.5475179382047152e-05,
+      "loss": 1.4475,
+      "step": 4896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08418362485610752,
+      "learning_rate": 2.5462000292868653e-05,
+      "loss": 1.4411,
+      "step": 4897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08120011856433508,
+      "learning_rate": 2.544882120369015e-05,
+      "loss": 1.4409,
+      "step": 4898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08186752669484414,
+      "learning_rate": 2.543564211451164e-05,
+      "loss": 1.4478,
+      "step": 4899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0888636709216178,
+      "learning_rate": 2.542246302533314e-05,
+      "loss": 1.4477,
+      "step": 4900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08598840225101306,
+      "learning_rate": 2.5409283936154638e-05,
+      "loss": 1.453,
+      "step": 4901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07432147545778035,
+      "learning_rate": 2.5396104846976135e-05,
+      "loss": 1.4443,
+      "step": 4902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.10043922160922762,
+      "learning_rate": 2.538292575779763e-05,
+      "loss": 1.4463,
+      "step": 4903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08358011823347111,
+      "learning_rate": 2.5369746668619126e-05,
+      "loss": 1.4457,
+      "step": 4904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07887540980074934,
+      "learning_rate": 2.5356567579440623e-05,
+      "loss": 1.4474,
+      "step": 4905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.10166432561451402,
+      "learning_rate": 2.534338849026212e-05,
+      "loss": 1.453,
+      "step": 4906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07214655533958941,
+      "learning_rate": 2.5330209401083614e-05,
+      "loss": 1.4551,
+      "step": 4907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.09907147275427153,
+      "learning_rate": 2.531703031190511e-05,
+      "loss": 1.4413,
+      "step": 4908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08260163742003128,
+      "learning_rate": 2.530385122272661e-05,
+      "loss": 1.4402,
+      "step": 4909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07914505281439273,
+      "learning_rate": 2.529067213354811e-05,
+      "loss": 1.4441,
+      "step": 4910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08755977883910626,
+      "learning_rate": 2.52774930443696e-05,
+      "loss": 1.4427,
+      "step": 4911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08111589797893948,
+      "learning_rate": 2.52643139551911e-05,
+      "loss": 1.4506,
+      "step": 4912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07982065813739192,
+      "learning_rate": 2.5251134866012597e-05,
+      "loss": 1.4426,
+      "step": 4913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08126641634843686,
+      "learning_rate": 2.5237955776834094e-05,
+      "loss": 1.4493,
+      "step": 4914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08240255710115835,
+      "learning_rate": 2.5224776687655588e-05,
+      "loss": 1.4467,
+      "step": 4915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07858360732112787,
+      "learning_rate": 2.5211597598477085e-05,
+      "loss": 1.4493,
+      "step": 4916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07997647129185952,
+      "learning_rate": 2.5198418509298582e-05,
+      "loss": 1.4476,
+      "step": 4917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07616285266592067,
+      "learning_rate": 2.518523942012008e-05,
+      "loss": 1.4326,
+      "step": 4918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08892925755118707,
+      "learning_rate": 2.5172060330941573e-05,
+      "loss": 1.4472,
+      "step": 4919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0854970594220413,
+      "learning_rate": 2.515888124176307e-05,
+      "loss": 1.4457,
+      "step": 4920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08566748564449363,
+      "learning_rate": 2.5145702152584567e-05,
+      "loss": 1.4428,
+      "step": 4921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07615782651895694,
+      "learning_rate": 2.5132523063406067e-05,
+      "loss": 1.4437,
+      "step": 4922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08784709021016268,
+      "learning_rate": 2.5119343974227558e-05,
+      "loss": 1.457,
+      "step": 4923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07623587724181397,
+      "learning_rate": 2.510616488504906e-05,
+      "loss": 1.4478,
+      "step": 4924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07998966625466244,
+      "learning_rate": 2.5092985795870555e-05,
+      "loss": 1.4442,
+      "step": 4925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0785761404846103,
+      "learning_rate": 2.5079806706692053e-05,
+      "loss": 1.4541,
+      "step": 4926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0826558184759799,
+      "learning_rate": 2.5066627617513546e-05,
+      "loss": 1.4477,
+      "step": 4927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07769439800940718,
+      "learning_rate": 2.5053448528335043e-05,
+      "loss": 1.4437,
+      "step": 4928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07640983957478023,
+      "learning_rate": 2.504026943915654e-05,
+      "loss": 1.4517,
+      "step": 4929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0811519361419571,
+      "learning_rate": 2.5027090349978038e-05,
+      "loss": 1.4486,
+      "step": 4930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07224901420697367,
+      "learning_rate": 2.501391126079953e-05,
+      "loss": 1.4419,
+      "step": 4931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0816866163608234,
+      "learning_rate": 2.500073217162103e-05,
+      "loss": 1.4469,
+      "step": 4932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07345798178702347,
+      "learning_rate": 2.4987553082442526e-05,
+      "loss": 1.4521,
+      "step": 4933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07674085326876222,
+      "learning_rate": 2.4974373993264026e-05,
+      "loss": 1.4394,
+      "step": 4934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07754403282398863,
+      "learning_rate": 2.4961194904085517e-05,
+      "loss": 1.4494,
+      "step": 4935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07523819838975931,
+      "learning_rate": 2.4948015814907017e-05,
+      "loss": 1.4407,
+      "step": 4936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08845191483075163,
+      "learning_rate": 2.4934836725728514e-05,
+      "loss": 1.4522,
+      "step": 4937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07839642670332553,
+      "learning_rate": 2.492165763655001e-05,
+      "loss": 1.4465,
+      "step": 4938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08362184703608651,
+      "learning_rate": 2.4908478547371505e-05,
+      "loss": 1.4432,
+      "step": 4939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08704385678440557,
+      "learning_rate": 2.4895299458193002e-05,
+      "loss": 1.4484,
+      "step": 4940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07511847491399579,
+      "learning_rate": 2.48821203690145e-05,
+      "loss": 1.444,
+      "step": 4941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08585375243227607,
+      "learning_rate": 2.4868941279835996e-05,
+      "loss": 1.4534,
+      "step": 4942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07379711101174781,
+      "learning_rate": 2.485576219065749e-05,
+      "loss": 1.4411,
+      "step": 4943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08068885741182136,
+      "learning_rate": 2.4842583101478987e-05,
+      "loss": 1.4509,
+      "step": 4944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07675237949217098,
+      "learning_rate": 2.4829404012300484e-05,
+      "loss": 1.4453,
+      "step": 4945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07602437186117642,
+      "learning_rate": 2.4816224923121985e-05,
+      "loss": 1.4467,
+      "step": 4946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08303684412913793,
+      "learning_rate": 2.4803045833943475e-05,
+      "loss": 1.4411,
+      "step": 4947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07122156033579337,
+      "learning_rate": 2.4789866744764976e-05,
+      "loss": 1.4551,
+      "step": 4948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08807944203203162,
+      "learning_rate": 2.4776687655586473e-05,
+      "loss": 1.4455,
+      "step": 4949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.07805272842126433,
+      "learning_rate": 2.476350856640797e-05,
+      "loss": 1.4557,
+      "step": 4950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.08093968792422997,
+      "learning_rate": 2.4750329477229464e-05,
+      "loss": 1.4423,
+      "step": 4951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07704102065059389,
+      "learning_rate": 2.473715038805096e-05,
+      "loss": 1.4397,
+      "step": 4952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07507717346390616,
+      "learning_rate": 2.4723971298872458e-05,
+      "loss": 1.4437,
+      "step": 4953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.06643695963818229,
+      "learning_rate": 2.4710792209693955e-05,
+      "loss": 1.4482,
+      "step": 4954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07564692793507727,
+      "learning_rate": 2.469761312051545e-05,
+      "loss": 1.4453,
+      "step": 4955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07404168254825268,
+      "learning_rate": 2.4684434031336946e-05,
+      "loss": 1.4447,
+      "step": 4956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0740738073366309,
+      "learning_rate": 2.4671254942158443e-05,
+      "loss": 1.4513,
+      "step": 4957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07506121737184437,
+      "learning_rate": 2.4658075852979944e-05,
+      "loss": 1.4388,
+      "step": 4958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07815805364774761,
+      "learning_rate": 2.4644896763801434e-05,
+      "loss": 1.4413,
+      "step": 4959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07252533729620259,
+      "learning_rate": 2.4631717674622935e-05,
+      "loss": 1.4482,
+      "step": 4960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08212694629841608,
+      "learning_rate": 2.4618538585444432e-05,
+      "loss": 1.4405,
+      "step": 4961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07885484254730504,
+      "learning_rate": 2.460535949626593e-05,
+      "loss": 1.4502,
+      "step": 4962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07506323375830111,
+      "learning_rate": 2.4592180407087423e-05,
+      "loss": 1.446,
+      "step": 4963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08522129804301908,
+      "learning_rate": 2.457900131790892e-05,
+      "loss": 1.4446,
+      "step": 4964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07689972444959055,
+      "learning_rate": 2.4565822228730417e-05,
+      "loss": 1.4568,
+      "step": 4965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.077627891250216,
+      "learning_rate": 2.4552643139551914e-05,
+      "loss": 1.4474,
+      "step": 4966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0825840040535346,
+      "learning_rate": 2.4539464050373408e-05,
+      "loss": 1.4463,
+      "step": 4967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0746611730259906,
+      "learning_rate": 2.4526284961194905e-05,
+      "loss": 1.4506,
+      "step": 4968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0791958661777196,
+      "learning_rate": 2.4513105872016402e-05,
+      "loss": 1.4488,
+      "step": 4969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07847755980669983,
+      "learning_rate": 2.4499926782837896e-05,
+      "loss": 1.4505,
+      "step": 4970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07605567893550387,
+      "learning_rate": 2.4486747693659393e-05,
+      "loss": 1.4464,
+      "step": 4971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08396372961497776,
+      "learning_rate": 2.4473568604480893e-05,
+      "loss": 1.4483,
+      "step": 4972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07924852954372583,
+      "learning_rate": 2.446038951530239e-05,
+      "loss": 1.4476,
+      "step": 4973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0760521394090154,
+      "learning_rate": 2.444721042612388e-05,
+      "loss": 1.4492,
+      "step": 4974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08203652175556024,
+      "learning_rate": 2.443403133694538e-05,
+      "loss": 1.4467,
+      "step": 4975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08313122905139003,
+      "learning_rate": 2.442085224776688e-05,
+      "loss": 1.4473,
+      "step": 4976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07319403404639596,
+      "learning_rate": 2.4407673158588376e-05,
+      "loss": 1.4497,
+      "step": 4977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0810657447503926,
+      "learning_rate": 2.439449406940987e-05,
+      "loss": 1.4435,
+      "step": 4978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07321012994245853,
+      "learning_rate": 2.4381314980231367e-05,
+      "loss": 1.4437,
+      "step": 4979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.078358254842506,
+      "learning_rate": 2.4368135891052864e-05,
+      "loss": 1.4484,
+      "step": 4980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07782265703326127,
+      "learning_rate": 2.435495680187436e-05,
+      "loss": 1.4459,
+      "step": 4981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07679271897863749,
+      "learning_rate": 2.4341777712695855e-05,
+      "loss": 1.4474,
+      "step": 4982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07427477112246433,
+      "learning_rate": 2.4328598623517352e-05,
+      "loss": 1.4511,
+      "step": 4983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07821878935009148,
+      "learning_rate": 2.4315419534338852e-05,
+      "loss": 1.4468,
+      "step": 4984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0754665907918599,
+      "learning_rate": 2.430224044516035e-05,
+      "loss": 1.444,
+      "step": 4985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08667814372930882,
+      "learning_rate": 2.428906135598184e-05,
+      "loss": 1.4547,
+      "step": 4986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07084689104381346,
+      "learning_rate": 2.427588226680334e-05,
+      "loss": 1.4484,
+      "step": 4987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08207290067864108,
+      "learning_rate": 2.4262703177624837e-05,
+      "loss": 1.4472,
+      "step": 4988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0794067680187924,
+      "learning_rate": 2.4249524088446334e-05,
+      "loss": 1.4489,
+      "step": 4989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08062190440006538,
+      "learning_rate": 2.4236344999267828e-05,
+      "loss": 1.4422,
+      "step": 4990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07711542986638953,
+      "learning_rate": 2.4223165910089325e-05,
+      "loss": 1.4316,
+      "step": 4991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07424745948194397,
+      "learning_rate": 2.4209986820910822e-05,
+      "loss": 1.45,
+      "step": 4992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07547776186491341,
+      "learning_rate": 2.419680773173232e-05,
+      "loss": 1.4528,
+      "step": 4993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0811067799891161,
+      "learning_rate": 2.4183628642553813e-05,
+      "loss": 1.446,
+      "step": 4994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07235515264072861,
+      "learning_rate": 2.417044955337531e-05,
+      "loss": 1.4423,
+      "step": 4995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08155047015936336,
+      "learning_rate": 2.4157270464196808e-05,
+      "loss": 1.442,
+      "step": 4996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07199643745187521,
+      "learning_rate": 2.4144091375018308e-05,
+      "loss": 1.4467,
+      "step": 4997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07860098962181221,
+      "learning_rate": 2.41309122858398e-05,
+      "loss": 1.4449,
+      "step": 4998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07319674465357227,
+      "learning_rate": 2.41177331966613e-05,
+      "loss": 1.4407,
+      "step": 4999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.075532396314923,
+      "learning_rate": 2.4104554107482796e-05,
+      "loss": 1.4434,
+      "step": 5000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07469825107993598,
+      "learning_rate": 2.4091375018304293e-05,
+      "loss": 1.4453,
+      "step": 5001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07596465368313135,
+      "learning_rate": 2.4078195929125787e-05,
+      "loss": 1.4525,
+      "step": 5002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07941557711319822,
+      "learning_rate": 2.4065016839947284e-05,
+      "loss": 1.4431,
+      "step": 5003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0797506522334278,
+      "learning_rate": 2.405183775076878e-05,
+      "loss": 1.4458,
+      "step": 5004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08058957982381029,
+      "learning_rate": 2.403865866159028e-05,
+      "loss": 1.4458,
+      "step": 5005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0766755155340805,
+      "learning_rate": 2.4025479572411772e-05,
+      "loss": 1.4481,
+      "step": 5006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07746271092470039,
+      "learning_rate": 2.401230048323327e-05,
+      "loss": 1.4519,
+      "step": 5007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07913942707704202,
+      "learning_rate": 2.3999121394054766e-05,
+      "loss": 1.4476,
+      "step": 5008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08474168213469227,
+      "learning_rate": 2.3985942304876267e-05,
+      "loss": 1.4475,
+      "step": 5009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08801102185235708,
+      "learning_rate": 2.3972763215697757e-05,
+      "loss": 1.4437,
+      "step": 5010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0796807373456822,
+      "learning_rate": 2.3959584126519258e-05,
+      "loss": 1.4428,
+      "step": 5011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08100277040182734,
+      "learning_rate": 2.3946405037340755e-05,
+      "loss": 1.4458,
+      "step": 5012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07714671472531397,
+      "learning_rate": 2.3933225948162252e-05,
+      "loss": 1.4472,
+      "step": 5013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08238348872923645,
+      "learning_rate": 2.3920046858983746e-05,
+      "loss": 1.4453,
+      "step": 5014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0819901025607147,
+      "learning_rate": 2.3906867769805243e-05,
+      "loss": 1.4395,
+      "step": 5015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07878388986919325,
+      "learning_rate": 2.389368868062674e-05,
+      "loss": 1.4462,
+      "step": 5016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08772469322234605,
+      "learning_rate": 2.3880509591448237e-05,
+      "loss": 1.4458,
+      "step": 5017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.07166046191749598,
+      "learning_rate": 2.386733050226973e-05,
+      "loss": 1.4478,
+      "step": 5018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08399806409393122,
+      "learning_rate": 2.3854151413091228e-05,
+      "loss": 1.4509,
+      "step": 5019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07998607114895992,
+      "learning_rate": 2.3840972323912725e-05,
+      "loss": 1.4521,
+      "step": 5020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08145620706989636,
+      "learning_rate": 2.3827793234734226e-05,
+      "loss": 1.4484,
+      "step": 5021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07897278371129365,
+      "learning_rate": 2.3814614145555716e-05,
+      "loss": 1.4484,
+      "step": 5022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07828367763031566,
+      "learning_rate": 2.3801435056377217e-05,
+      "loss": 1.4455,
+      "step": 5023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08767997058806601,
+      "learning_rate": 2.3788255967198714e-05,
+      "loss": 1.4393,
+      "step": 5024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07783348464644269,
+      "learning_rate": 2.377507687802021e-05,
+      "loss": 1.4441,
+      "step": 5025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08096086852479069,
+      "learning_rate": 2.3761897788841705e-05,
+      "loss": 1.4506,
+      "step": 5026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08839652292205184,
+      "learning_rate": 2.3748718699663202e-05,
+      "loss": 1.4441,
+      "step": 5027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07937132420869311,
+      "learning_rate": 2.37355396104847e-05,
+      "loss": 1.4393,
+      "step": 5028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08056499546258605,
+      "learning_rate": 2.3722360521306196e-05,
+      "loss": 1.4365,
+      "step": 5029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08821806984554978,
+      "learning_rate": 2.370918143212769e-05,
+      "loss": 1.4374,
+      "step": 5030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08101103240087884,
+      "learning_rate": 2.3696002342949187e-05,
+      "loss": 1.45,
+      "step": 5031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08761220858552904,
+      "learning_rate": 2.3682823253770684e-05,
+      "loss": 1.4415,
+      "step": 5032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09443936862127413,
+      "learning_rate": 2.3669644164592184e-05,
+      "loss": 1.4451,
+      "step": 5033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07744169387408935,
+      "learning_rate": 2.3656465075413675e-05,
+      "loss": 1.4483,
+      "step": 5034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.10214482955057408,
+      "learning_rate": 2.3643285986235175e-05,
+      "loss": 1.4494,
+      "step": 5035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09871280704830478,
+      "learning_rate": 2.3630106897056672e-05,
+      "loss": 1.4456,
+      "step": 5036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07924122010550258,
+      "learning_rate": 2.361692780787817e-05,
+      "loss": 1.4461,
+      "step": 5037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1103027384392773,
+      "learning_rate": 2.3603748718699663e-05,
+      "loss": 1.4441,
+      "step": 5038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0817979740112491,
+      "learning_rate": 2.359056962952116e-05,
+      "loss": 1.4522,
+      "step": 5039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08861427336594135,
+      "learning_rate": 2.3577390540342658e-05,
+      "loss": 1.4423,
+      "step": 5040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09441396491317663,
+      "learning_rate": 2.3564211451164155e-05,
+      "loss": 1.4592,
+      "step": 5041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.075072776840673,
+      "learning_rate": 2.355103236198565e-05,
+      "loss": 1.4462,
+      "step": 5042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08858904686387951,
+      "learning_rate": 2.3537853272807146e-05,
+      "loss": 1.4566,
+      "step": 5043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07441121472986022,
+      "learning_rate": 2.3524674183628643e-05,
+      "loss": 1.4563,
+      "step": 5044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07553934451226477,
+      "learning_rate": 2.3511495094450143e-05,
+      "loss": 1.4427,
+      "step": 5045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07262812960272448,
+      "learning_rate": 2.3498316005271634e-05,
+      "loss": 1.4376,
+      "step": 5046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07699834682331902,
+      "learning_rate": 2.3485136916093134e-05,
+      "loss": 1.4439,
+      "step": 5047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07049501739411941,
+      "learning_rate": 2.347195782691463e-05,
+      "loss": 1.4495,
+      "step": 5048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07556126652942027,
+      "learning_rate": 2.345877873773613e-05,
+      "loss": 1.448,
+      "step": 5049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07140470821553045,
+      "learning_rate": 2.3445599648557622e-05,
+      "loss": 1.453,
+      "step": 5050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07513230330476844,
+      "learning_rate": 2.343242055937912e-05,
+      "loss": 1.4483,
+      "step": 5051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0724179063416859,
+      "learning_rate": 2.3419241470200616e-05,
+      "loss": 1.4389,
+      "step": 5052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07616506003924567,
+      "learning_rate": 2.3406062381022114e-05,
+      "loss": 1.4499,
+      "step": 5053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07739424955876131,
+      "learning_rate": 2.3392883291843607e-05,
+      "loss": 1.4431,
+      "step": 5054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07424786232061845,
+      "learning_rate": 2.3379704202665104e-05,
+      "loss": 1.4467,
+      "step": 5055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07262416140237654,
+      "learning_rate": 2.33665251134866e-05,
+      "loss": 1.4503,
+      "step": 5056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08405020148372654,
+      "learning_rate": 2.3353346024308102e-05,
+      "loss": 1.4445,
+      "step": 5057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07457243705384699,
+      "learning_rate": 2.3340166935129592e-05,
+      "loss": 1.4296,
+      "step": 5058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07904534494208784,
+      "learning_rate": 2.3326987845951093e-05,
+      "loss": 1.4449,
+      "step": 5059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07866003552843212,
+      "learning_rate": 2.331380875677259e-05,
+      "loss": 1.4476,
+      "step": 5060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07792020092509064,
+      "learning_rate": 2.3300629667594087e-05,
+      "loss": 1.4396,
+      "step": 5061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0850300171666119,
+      "learning_rate": 2.328745057841558e-05,
+      "loss": 1.4487,
+      "step": 5062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08416161247477531,
+      "learning_rate": 2.3274271489237078e-05,
+      "loss": 1.4444,
+      "step": 5063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08605608259378111,
+      "learning_rate": 2.3261092400058575e-05,
+      "loss": 1.4443,
+      "step": 5064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08132304335373357,
+      "learning_rate": 2.3247913310880072e-05,
+      "loss": 1.4389,
+      "step": 5065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.093295152530523,
+      "learning_rate": 2.3234734221701566e-05,
+      "loss": 1.4513,
+      "step": 5066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08180522810682148,
+      "learning_rate": 2.3221555132523063e-05,
+      "loss": 1.4444,
+      "step": 5067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0845456260251711,
+      "learning_rate": 2.320837604334456e-05,
+      "loss": 1.4427,
+      "step": 5068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08160844393164714,
+      "learning_rate": 2.319519695416606e-05,
+      "loss": 1.4355,
+      "step": 5069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0834715485268486,
+      "learning_rate": 2.318201786498755e-05,
+      "loss": 1.4519,
+      "step": 5070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0931783792841113,
+      "learning_rate": 2.3168838775809052e-05,
+      "loss": 1.4467,
+      "step": 5071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07495286003636802,
+      "learning_rate": 2.315565968663055e-05,
+      "loss": 1.4403,
+      "step": 5072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0906145681161765,
+      "learning_rate": 2.3142480597452046e-05,
+      "loss": 1.4503,
+      "step": 5073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08580778487620919,
+      "learning_rate": 2.312930150827354e-05,
+      "loss": 1.4448,
+      "step": 5074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08274076256057851,
+      "learning_rate": 2.3116122419095037e-05,
+      "loss": 1.4509,
+      "step": 5075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09196031508634128,
+      "learning_rate": 2.3102943329916534e-05,
+      "loss": 1.4498,
+      "step": 5076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07390337168857404,
+      "learning_rate": 2.308976424073803e-05,
+      "loss": 1.4545,
+      "step": 5077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09157321813580277,
+      "learning_rate": 2.3076585151559525e-05,
+      "loss": 1.446,
+      "step": 5078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07702021425686817,
+      "learning_rate": 2.3063406062381022e-05,
+      "loss": 1.435,
+      "step": 5079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08806830991778292,
+      "learning_rate": 2.305022697320252e-05,
+      "loss": 1.4467,
+      "step": 5080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07542779037058828,
+      "learning_rate": 2.303704788402402e-05,
+      "loss": 1.4466,
+      "step": 5081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08345056993929043,
+      "learning_rate": 2.302386879484551e-05,
+      "loss": 1.4441,
+      "step": 5082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08256615901460639,
+      "learning_rate": 2.3010689705667007e-05,
+      "loss": 1.4425,
+      "step": 5083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07920188938196653,
+      "learning_rate": 2.2997510616488508e-05,
+      "loss": 1.4414,
+      "step": 5084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08887275173422157,
+      "learning_rate": 2.2984331527310005e-05,
+      "loss": 1.4461,
+      "step": 5085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07005099101789881,
+      "learning_rate": 2.29711524381315e-05,
+      "loss": 1.439,
+      "step": 5086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.08224063751000159,
+      "learning_rate": 2.2957973348952996e-05,
+      "loss": 1.4449,
+      "step": 5087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.07003217622771513,
+      "learning_rate": 2.2944794259774493e-05,
+      "loss": 1.4429,
+      "step": 5088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0774070077545315,
+      "learning_rate": 2.293161517059599e-05,
+      "loss": 1.4394,
+      "step": 5089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07215882293499469,
+      "learning_rate": 2.2918436081417484e-05,
+      "loss": 1.4425,
+      "step": 5090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07772017699449815,
+      "learning_rate": 2.290525699223898e-05,
+      "loss": 1.4441,
+      "step": 5091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07837657963737095,
+      "learning_rate": 2.2892077903060478e-05,
+      "loss": 1.4464,
+      "step": 5092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0755403575028706,
+      "learning_rate": 2.287889881388198e-05,
+      "loss": 1.453,
+      "step": 5093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08433881844043978,
+      "learning_rate": 2.286571972470347e-05,
+      "loss": 1.4412,
+      "step": 5094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07445990609632445,
+      "learning_rate": 2.2852540635524966e-05,
+      "loss": 1.4447,
+      "step": 5095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09420243218781044,
+      "learning_rate": 2.2839361546346466e-05,
+      "loss": 1.4449,
+      "step": 5096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08567030799582746,
+      "learning_rate": 2.2826182457167964e-05,
+      "loss": 1.4441,
+      "step": 5097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08319845952784173,
+      "learning_rate": 2.2813003367989457e-05,
+      "loss": 1.4467,
+      "step": 5098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09178076767191547,
+      "learning_rate": 2.2799824278810954e-05,
+      "loss": 1.4472,
+      "step": 5099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07904016785593118,
+      "learning_rate": 2.278664518963245e-05,
+      "loss": 1.4466,
+      "step": 5100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.09250287461920997,
+      "learning_rate": 2.277346610045395e-05,
+      "loss": 1.4364,
+      "step": 5101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0778467568233248,
+      "learning_rate": 2.2760287011275442e-05,
+      "loss": 1.4435,
+      "step": 5102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07914392455449246,
+      "learning_rate": 2.274710792209694e-05,
+      "loss": 1.443,
+      "step": 5103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08172614274888121,
+      "learning_rate": 2.2733928832918437e-05,
+      "loss": 1.4431,
+      "step": 5104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07431811487258455,
+      "learning_rate": 2.2720749743739937e-05,
+      "loss": 1.4467,
+      "step": 5105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08033978173134251,
+      "learning_rate": 2.2707570654561428e-05,
+      "loss": 1.4462,
+      "step": 5106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07113035139859623,
+      "learning_rate": 2.2694391565382925e-05,
+      "loss": 1.439,
+      "step": 5107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0829437462062766,
+      "learning_rate": 2.2681212476204425e-05,
+      "loss": 1.4478,
+      "step": 5108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07255139029680703,
+      "learning_rate": 2.2668033387025922e-05,
+      "loss": 1.4535,
+      "step": 5109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08153768781069708,
+      "learning_rate": 2.2654854297847416e-05,
+      "loss": 1.4563,
+      "step": 5110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07604920895136073,
+      "learning_rate": 2.2641675208668913e-05,
+      "loss": 1.4518,
+      "step": 5111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0786715232385936,
+      "learning_rate": 2.262849611949041e-05,
+      "loss": 1.4438,
+      "step": 5112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08990418489963536,
+      "learning_rate": 2.2615317030311907e-05,
+      "loss": 1.4373,
+      "step": 5113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07119278883271916,
+      "learning_rate": 2.26021379411334e-05,
+      "loss": 1.4479,
+      "step": 5114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07902847068271979,
+      "learning_rate": 2.25889588519549e-05,
+      "loss": 1.448,
+      "step": 5115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07623222241662977,
+      "learning_rate": 2.2575779762776395e-05,
+      "loss": 1.441,
+      "step": 5116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07416342914452614,
+      "learning_rate": 2.2562600673597896e-05,
+      "loss": 1.4408,
+      "step": 5117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07854161280050583,
+      "learning_rate": 2.2549421584419386e-05,
+      "loss": 1.4464,
+      "step": 5118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.06957574060095317,
+      "learning_rate": 2.2536242495240883e-05,
+      "loss": 1.4452,
+      "step": 5119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08371029700585757,
+      "learning_rate": 2.2523063406062384e-05,
+      "loss": 1.4445,
+      "step": 5120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07654243664177138,
+      "learning_rate": 2.250988431688388e-05,
+      "loss": 1.4491,
+      "step": 5121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0801597566867057,
+      "learning_rate": 2.2496705227705375e-05,
+      "loss": 1.4434,
+      "step": 5122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07807264700486409,
+      "learning_rate": 2.2483526138526872e-05,
+      "loss": 1.4536,
+      "step": 5123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08519556693121638,
+      "learning_rate": 2.247034704934837e-05,
+      "loss": 1.4448,
+      "step": 5124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07550687973727663,
+      "learning_rate": 2.2457167960169863e-05,
+      "loss": 1.4444,
+      "step": 5125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0842184558729363,
+      "learning_rate": 2.2443988870991363e-05,
+      "loss": 1.4483,
+      "step": 5126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07452879242170955,
+      "learning_rate": 2.2430809781812857e-05,
+      "loss": 1.4468,
+      "step": 5127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08610091773340986,
+      "learning_rate": 2.2417630692634354e-05,
+      "loss": 1.4494,
+      "step": 5128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07452331469620652,
+      "learning_rate": 2.240445160345585e-05,
+      "loss": 1.4487,
+      "step": 5129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08318039890275929,
+      "learning_rate": 2.239127251427735e-05,
+      "loss": 1.4574,
+      "step": 5130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07321540336839424,
+      "learning_rate": 2.2378093425098842e-05,
+      "loss": 1.4482,
+      "step": 5131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08638422826118979,
+      "learning_rate": 2.2364914335920343e-05,
+      "loss": 1.447,
+      "step": 5132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07168227554925212,
+      "learning_rate": 2.2351735246741837e-05,
+      "loss": 1.4482,
+      "step": 5133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0828997695991488,
+      "learning_rate": 2.2338556157563334e-05,
+      "loss": 1.4382,
+      "step": 5134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07378725695061984,
+      "learning_rate": 2.232537706838483e-05,
+      "loss": 1.4518,
+      "step": 5135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.08185842241516435,
+      "learning_rate": 2.2312197979206328e-05,
+      "loss": 1.4466,
+      "step": 5136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07902830304804848,
+      "learning_rate": 2.229901889002782e-05,
+      "loss": 1.4467,
+      "step": 5137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0797132476344473,
+      "learning_rate": 2.2285839800849322e-05,
+      "loss": 1.4465,
+      "step": 5138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07848578939482552,
+      "learning_rate": 2.2272660711670816e-05,
+      "loss": 1.4445,
+      "step": 5139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07915782483937879,
+      "learning_rate": 2.2259481622492313e-05,
+      "loss": 1.4513,
+      "step": 5140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07294408269567776,
+      "learning_rate": 2.224630253331381e-05,
+      "loss": 1.4454,
+      "step": 5141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07191201784584524,
+      "learning_rate": 2.2233123444135307e-05,
+      "loss": 1.4473,
+      "step": 5142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07615730533511005,
+      "learning_rate": 2.22199443549568e-05,
+      "loss": 1.4463,
+      "step": 5143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0735759499845662,
+      "learning_rate": 2.22067652657783e-05,
+      "loss": 1.4425,
+      "step": 5144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07057903225849087,
+      "learning_rate": 2.2193586176599795e-05,
+      "loss": 1.4372,
+      "step": 5145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07205322415178919,
+      "learning_rate": 2.2180407087421292e-05,
+      "loss": 1.4373,
+      "step": 5146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07126040665870133,
+      "learning_rate": 2.216722799824279e-05,
+      "loss": 1.4472,
+      "step": 5147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07784183403224093,
+      "learning_rate": 2.2154048909064287e-05,
+      "loss": 1.4401,
+      "step": 5148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07366990043574155,
+      "learning_rate": 2.214086981988578e-05,
+      "loss": 1.4491,
+      "step": 5149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07303929155776889,
+      "learning_rate": 2.212769073070728e-05,
+      "loss": 1.4476,
+      "step": 5150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07524232317295294,
+      "learning_rate": 2.2114511641528775e-05,
+      "loss": 1.4381,
+      "step": 5151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07354980783078278,
+      "learning_rate": 2.2101332552350272e-05,
+      "loss": 1.4534,
+      "step": 5152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07538017943803083,
+      "learning_rate": 2.208815346317177e-05,
+      "loss": 1.4501,
+      "step": 5153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07231713191187358,
+      "learning_rate": 2.2074974373993266e-05,
+      "loss": 1.4381,
+      "step": 5154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07223848201360787,
+      "learning_rate": 2.206179528481476e-05,
+      "loss": 1.4428,
+      "step": 5155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0723109473467658,
+      "learning_rate": 2.204861619563626e-05,
+      "loss": 1.4409,
+      "step": 5156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07417377749266274,
+      "learning_rate": 2.2035437106457754e-05,
+      "loss": 1.4517,
+      "step": 5157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07112563846444456,
+      "learning_rate": 2.202225801727925e-05,
+      "loss": 1.4442,
+      "step": 5158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07665449622857923,
+      "learning_rate": 2.200907892810075e-05,
+      "loss": 1.4396,
+      "step": 5159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07171274966983575,
+      "learning_rate": 2.1995899838922245e-05,
+      "loss": 1.4342,
+      "step": 5160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0758205285369927,
+      "learning_rate": 2.198272074974374e-05,
+      "loss": 1.4404,
+      "step": 5161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07896806298017185,
+      "learning_rate": 2.196954166056524e-05,
+      "loss": 1.4446,
+      "step": 5162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07610675025798999,
+      "learning_rate": 2.1956362571386733e-05,
+      "loss": 1.4411,
+      "step": 5163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08886772469424734,
+      "learning_rate": 2.194318348220823e-05,
+      "loss": 1.4478,
+      "step": 5164
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09010868710045351,
+      "learning_rate": 2.1930004393029728e-05,
+      "loss": 1.44,
+      "step": 5165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08017436022814749,
+      "learning_rate": 2.1916825303851225e-05,
+      "loss": 1.4463,
+      "step": 5166
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.09526638932279191,
+      "learning_rate": 2.190364621467272e-05,
+      "loss": 1.4539,
+      "step": 5167
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07478477343336154,
+      "learning_rate": 2.189046712549422e-05,
+      "loss": 1.4329,
+      "step": 5168
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0874079223953518,
+      "learning_rate": 2.1877288036315713e-05,
+      "loss": 1.448,
+      "step": 5169
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08045463327290767,
+      "learning_rate": 2.186410894713721e-05,
+      "loss": 1.4472,
+      "step": 5170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07774766666112756,
+      "learning_rate": 2.1850929857958707e-05,
+      "loss": 1.4468,
+      "step": 5171
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08019399135241922,
+      "learning_rate": 2.1837750768780204e-05,
+      "loss": 1.4511,
+      "step": 5172
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08658145099952294,
+      "learning_rate": 2.1824571679601698e-05,
+      "loss": 1.4433,
+      "step": 5173
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08214069635828508,
+      "learning_rate": 2.18113925904232e-05,
+      "loss": 1.4405,
+      "step": 5174
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0855959234765811,
+      "learning_rate": 2.1798213501244692e-05,
+      "loss": 1.4372,
+      "step": 5175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07340944880946049,
+      "learning_rate": 2.178503441206619e-05,
+      "loss": 1.4506,
+      "step": 5176
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08783872248650436,
+      "learning_rate": 2.1771855322887687e-05,
+      "loss": 1.4336,
+      "step": 5177
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07428158659922204,
+      "learning_rate": 2.1758676233709184e-05,
+      "loss": 1.4459,
+      "step": 5178
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08648087922815073,
+      "learning_rate": 2.1745497144530677e-05,
+      "loss": 1.4442,
+      "step": 5179
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07735199547772495,
+      "learning_rate": 2.1732318055352178e-05,
+      "loss": 1.4422,
+      "step": 5180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0826368527327218,
+      "learning_rate": 2.171913896617367e-05,
+      "loss": 1.4459,
+      "step": 5181
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07559995766475787,
+      "learning_rate": 2.170595987699517e-05,
+      "loss": 1.4434,
+      "step": 5182
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07974736365812579,
+      "learning_rate": 2.1692780787816666e-05,
+      "loss": 1.4399,
+      "step": 5183
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08094204242635095,
+      "learning_rate": 2.1679601698638163e-05,
+      "loss": 1.449,
+      "step": 5184
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07207493921217599,
+      "learning_rate": 2.1666422609459657e-05,
+      "loss": 1.4457,
+      "step": 5185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08098239532586071,
+      "learning_rate": 2.1653243520281157e-05,
+      "loss": 1.44,
+      "step": 5186
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0785050392095796,
+      "learning_rate": 2.164006443110265e-05,
+      "loss": 1.4406,
+      "step": 5187
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07656608696454836,
+      "learning_rate": 2.1626885341924148e-05,
+      "loss": 1.4325,
+      "step": 5188
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07494857805594374,
+      "learning_rate": 2.1613706252745645e-05,
+      "loss": 1.46,
+      "step": 5189
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08425765587327527,
+      "learning_rate": 2.1600527163567142e-05,
+      "loss": 1.4471,
+      "step": 5190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07379200629859359,
+      "learning_rate": 2.1587348074388636e-05,
+      "loss": 1.4419,
+      "step": 5191
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0856107868116821,
+      "learning_rate": 2.1574168985210137e-05,
+      "loss": 1.4396,
+      "step": 5192
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08002414578733587,
+      "learning_rate": 2.156098989603163e-05,
+      "loss": 1.4443,
+      "step": 5193
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07889139534471498,
+      "learning_rate": 2.1547810806853128e-05,
+      "loss": 1.4476,
+      "step": 5194
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07561743050501325,
+      "learning_rate": 2.1534631717674625e-05,
+      "loss": 1.4439,
+      "step": 5195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07846755003408278,
+      "learning_rate": 2.1521452628496122e-05,
+      "loss": 1.4401,
+      "step": 5196
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0755433391593136,
+      "learning_rate": 2.1508273539317616e-05,
+      "loss": 1.4454,
+      "step": 5197
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08230177642336702,
+      "learning_rate": 2.1495094450139116e-05,
+      "loss": 1.4405,
+      "step": 5198
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07454930406321077,
+      "learning_rate": 2.148191536096061e-05,
+      "loss": 1.4342,
+      "step": 5199
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0722391845711022,
+      "learning_rate": 2.1468736271782107e-05,
+      "loss": 1.4387,
+      "step": 5200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07146399281921982,
+      "learning_rate": 2.1455557182603604e-05,
+      "loss": 1.4431,
+      "step": 5201
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0758289925242572,
+      "learning_rate": 2.14423780934251e-05,
+      "loss": 1.4509,
+      "step": 5202
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07931241184951322,
+      "learning_rate": 2.1429199004246595e-05,
+      "loss": 1.4517,
+      "step": 5203
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07084310449966312,
+      "learning_rate": 2.1416019915068095e-05,
+      "loss": 1.448,
+      "step": 5204
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07240846143898143,
+      "learning_rate": 2.140284082588959e-05,
+      "loss": 1.4424,
+      "step": 5205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07101492660098167,
+      "learning_rate": 2.1389661736711086e-05,
+      "loss": 1.4474,
+      "step": 5206
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07418651424026342,
+      "learning_rate": 2.1376482647532583e-05,
+      "loss": 1.4394,
+      "step": 5207
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07595126849424101,
+      "learning_rate": 2.136330355835408e-05,
+      "loss": 1.4339,
+      "step": 5208
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07648703206561194,
+      "learning_rate": 2.1350124469175574e-05,
+      "loss": 1.4382,
+      "step": 5209
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07516588888032873,
+      "learning_rate": 2.133694537999707e-05,
+      "loss": 1.4403,
+      "step": 5210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0731757087359834,
+      "learning_rate": 2.132376629081857e-05,
+      "loss": 1.4416,
+      "step": 5211
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.073575446949595,
+      "learning_rate": 2.1310587201640066e-05,
+      "loss": 1.4367,
+      "step": 5212
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07214264142409765,
+      "learning_rate": 2.1297408112461563e-05,
+      "loss": 1.4398,
+      "step": 5213
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0751794840797808,
+      "learning_rate": 2.128422902328306e-05,
+      "loss": 1.4468,
+      "step": 5214
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07647456231563013,
+      "learning_rate": 2.1271049934104554e-05,
+      "loss": 1.434,
+      "step": 5215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.06684308880409935,
+      "learning_rate": 2.125787084492605e-05,
+      "loss": 1.4388,
+      "step": 5216
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.08010381953084092,
+      "learning_rate": 2.1244691755747548e-05,
+      "loss": 1.4503,
+      "step": 5217
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07430893344544048,
+      "learning_rate": 2.1231512666569045e-05,
+      "loss": 1.441,
+      "step": 5218
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07451400020825438,
+      "learning_rate": 2.1218333577390542e-05,
+      "loss": 1.4542,
+      "step": 5219
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07737958996086805,
+      "learning_rate": 2.120515448821204e-05,
+      "loss": 1.4417,
+      "step": 5220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0727160023089977,
+      "learning_rate": 2.1191975399033533e-05,
+      "loss": 1.4343,
+      "step": 5221
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07215009710587389,
+      "learning_rate": 2.117879630985503e-05,
+      "loss": 1.4361,
+      "step": 5222
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0740895512362248,
+      "learning_rate": 2.1165617220676527e-05,
+      "loss": 1.4377,
+      "step": 5223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.07081129388329852,
+      "learning_rate": 2.1152438131498025e-05,
+      "loss": 1.4425,
+      "step": 5224
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07450061232625847,
+      "learning_rate": 2.113925904231952e-05,
+      "loss": 1.439,
+      "step": 5225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07429704691273743,
+      "learning_rate": 2.112607995314102e-05,
+      "loss": 1.4384,
+      "step": 5226
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06866293510448063,
+      "learning_rate": 2.1112900863962513e-05,
+      "loss": 1.4372,
+      "step": 5227
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06968981384727986,
+      "learning_rate": 2.109972177478401e-05,
+      "loss": 1.441,
+      "step": 5228
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07372037910899788,
+      "learning_rate": 2.1086542685605507e-05,
+      "loss": 1.4462,
+      "step": 5229
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0673052960980992,
+      "learning_rate": 2.1073363596427004e-05,
+      "loss": 1.4431,
+      "step": 5230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07591056973648318,
+      "learning_rate": 2.10601845072485e-05,
+      "loss": 1.4485,
+      "step": 5231
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0678496562937225,
+      "learning_rate": 2.1047005418069998e-05,
+      "loss": 1.4401,
+      "step": 5232
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07547744574887282,
+      "learning_rate": 2.1033826328891492e-05,
+      "loss": 1.4391,
+      "step": 5233
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07020894256119718,
+      "learning_rate": 2.102064723971299e-05,
+      "loss": 1.4402,
+      "step": 5234
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07030133193434782,
+      "learning_rate": 2.1007468150534486e-05,
+      "loss": 1.446,
+      "step": 5235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07241463507564218,
+      "learning_rate": 2.0994289061355983e-05,
+      "loss": 1.4372,
+      "step": 5236
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06798299550211485,
+      "learning_rate": 2.098110997217748e-05,
+      "loss": 1.4484,
+      "step": 5237
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06910799380407255,
+      "learning_rate": 2.0967930882998978e-05,
+      "loss": 1.4389,
+      "step": 5238
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07187915702351018,
+      "learning_rate": 2.095475179382047e-05,
+      "loss": 1.4435,
+      "step": 5239
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06813029568363102,
+      "learning_rate": 2.094157270464197e-05,
+      "loss": 1.4434,
+      "step": 5240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0688807431632514,
+      "learning_rate": 2.0928393615463466e-05,
+      "loss": 1.4464,
+      "step": 5241
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07192449466274732,
+      "learning_rate": 2.0915214526284963e-05,
+      "loss": 1.4432,
+      "step": 5242
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0750078037878121,
+      "learning_rate": 2.090203543710646e-05,
+      "loss": 1.4422,
+      "step": 5243
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06979678312759856,
+      "learning_rate": 2.0888856347927957e-05,
+      "loss": 1.4374,
+      "step": 5244
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07283482822063149,
+      "learning_rate": 2.087567725874945e-05,
+      "loss": 1.4295,
+      "step": 5245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07483831754020644,
+      "learning_rate": 2.0862498169570948e-05,
+      "loss": 1.4423,
+      "step": 5246
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07074822230180933,
+      "learning_rate": 2.0849319080392445e-05,
+      "loss": 1.4516,
+      "step": 5247
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07509450611021706,
+      "learning_rate": 2.0836139991213942e-05,
+      "loss": 1.439,
+      "step": 5248
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07063525395389832,
+      "learning_rate": 2.082296090203544e-05,
+      "loss": 1.448,
+      "step": 5249
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07810206570236145,
+      "learning_rate": 2.0809781812856936e-05,
+      "loss": 1.4533,
+      "step": 5250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0731262608463827,
+      "learning_rate": 2.079660272367843e-05,
+      "loss": 1.442,
+      "step": 5251
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07628085584090835,
+      "learning_rate": 2.0783423634499927e-05,
+      "loss": 1.4372,
+      "step": 5252
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07786287273999554,
+      "learning_rate": 2.0770244545321424e-05,
+      "loss": 1.4387,
+      "step": 5253
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07031334277712674,
+      "learning_rate": 2.075706545614292e-05,
+      "loss": 1.4507,
+      "step": 5254
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07547531843584546,
+      "learning_rate": 2.074388636696442e-05,
+      "loss": 1.4437,
+      "step": 5255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07408449038423273,
+      "learning_rate": 2.0730707277785916e-05,
+      "loss": 1.4349,
+      "step": 5256
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07920364971815677,
+      "learning_rate": 2.071752818860741e-05,
+      "loss": 1.4513,
+      "step": 5257
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.09422569020877054,
+      "learning_rate": 2.0704349099428907e-05,
+      "loss": 1.4412,
+      "step": 5258
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07765932620337068,
+      "learning_rate": 2.0691170010250404e-05,
+      "loss": 1.4427,
+      "step": 5259
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08837974933533015,
+      "learning_rate": 2.06779909210719e-05,
+      "loss": 1.4442,
+      "step": 5260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08284943057308215,
+      "learning_rate": 2.0664811831893398e-05,
+      "loss": 1.4397,
+      "step": 5261
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0724518879754711,
+      "learning_rate": 2.0651632742714895e-05,
+      "loss": 1.4424,
+      "step": 5262
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07681188809401879,
+      "learning_rate": 2.063845365353639e-05,
+      "loss": 1.4337,
+      "step": 5263
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07101079089787046,
+      "learning_rate": 2.0625274564357886e-05,
+      "loss": 1.4479,
+      "step": 5264
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07573295529567776,
+      "learning_rate": 2.0612095475179383e-05,
+      "loss": 1.435,
+      "step": 5265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07452276495933718,
+      "learning_rate": 2.059891638600088e-05,
+      "loss": 1.446,
+      "step": 5266
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07366321439853195,
+      "learning_rate": 2.0585737296822377e-05,
+      "loss": 1.4457,
+      "step": 5267
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07736497459478803,
+      "learning_rate": 2.0572558207643875e-05,
+      "loss": 1.4291,
+      "step": 5268
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07325589114529062,
+      "learning_rate": 2.0559379118465368e-05,
+      "loss": 1.4471,
+      "step": 5269
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0807518913271351,
+      "learning_rate": 2.0546200029286865e-05,
+      "loss": 1.4521,
+      "step": 5270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06947931855123606,
+      "learning_rate": 2.0533020940108363e-05,
+      "loss": 1.4441,
+      "step": 5271
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07582038956606409,
+      "learning_rate": 2.051984185092986e-05,
+      "loss": 1.4442,
+      "step": 5272
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07879814281970236,
+      "learning_rate": 2.0506662761751357e-05,
+      "loss": 1.4556,
+      "step": 5273
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07192130910217426,
+      "learning_rate": 2.0493483672572854e-05,
+      "loss": 1.4475,
+      "step": 5274
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07351937364231856,
+      "learning_rate": 2.0480304583394348e-05,
+      "loss": 1.4325,
+      "step": 5275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06945371113607203,
+      "learning_rate": 2.0467125494215845e-05,
+      "loss": 1.4402,
+      "step": 5276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07412863144757505,
+      "learning_rate": 2.0453946405037342e-05,
+      "loss": 1.442,
+      "step": 5277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07304775921083298,
+      "learning_rate": 2.044076731585884e-05,
+      "loss": 1.445,
+      "step": 5278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07422170374173968,
+      "learning_rate": 2.0427588226680336e-05,
+      "loss": 1.4435,
+      "step": 5279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07239880018215895,
+      "learning_rate": 2.0414409137501833e-05,
+      "loss": 1.4389,
+      "step": 5280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07649617536609282,
+      "learning_rate": 2.0401230048323327e-05,
+      "loss": 1.4427,
+      "step": 5281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.06909303496835723,
+      "learning_rate": 2.0388050959144824e-05,
+      "loss": 1.4422,
+      "step": 5282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07306667447587163,
+      "learning_rate": 2.037487186996632e-05,
+      "loss": 1.4297,
+      "step": 5283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07196183934142673,
+      "learning_rate": 2.036169278078782e-05,
+      "loss": 1.4368,
+      "step": 5284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07506692391852186,
+      "learning_rate": 2.0348513691609316e-05,
+      "loss": 1.436,
+      "step": 5285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07303442727623291,
+      "learning_rate": 2.0335334602430813e-05,
+      "loss": 1.4489,
+      "step": 5286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08316043282493489,
+      "learning_rate": 2.0322155513252306e-05,
+      "loss": 1.4414,
+      "step": 5287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07742693296811622,
+      "learning_rate": 2.0308976424073804e-05,
+      "loss": 1.4475,
+      "step": 5288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0719055880649899,
+      "learning_rate": 2.02957973348953e-05,
+      "loss": 1.4426,
+      "step": 5289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08549936733974174,
+      "learning_rate": 2.0282618245716798e-05,
+      "loss": 1.4445,
+      "step": 5290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.07255531734888275,
+      "learning_rate": 2.026943915653829e-05,
+      "loss": 1.4449,
+      "step": 5291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.08196746150527218,
+      "learning_rate": 2.0256260067359792e-05,
+      "loss": 1.4414,
+      "step": 5292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07641577685828455,
+      "learning_rate": 2.0243080978181286e-05,
+      "loss": 1.4449,
+      "step": 5293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07916154646430804,
+      "learning_rate": 2.0229901889002783e-05,
+      "loss": 1.4415,
+      "step": 5294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07843138778952255,
+      "learning_rate": 2.021672279982428e-05,
+      "loss": 1.4408,
+      "step": 5295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07037107269538123,
+      "learning_rate": 2.0203543710645777e-05,
+      "loss": 1.435,
+      "step": 5296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08037550025895233,
+      "learning_rate": 2.019036462146727e-05,
+      "loss": 1.4406,
+      "step": 5297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07088002819438141,
+      "learning_rate": 2.017718553228877e-05,
+      "loss": 1.4454,
+      "step": 5298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08576747962044898,
+      "learning_rate": 2.0164006443110265e-05,
+      "loss": 1.4324,
+      "step": 5299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07405012025406182,
+      "learning_rate": 2.0150827353931762e-05,
+      "loss": 1.4419,
+      "step": 5300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08955228353164654,
+      "learning_rate": 2.013764826475326e-05,
+      "loss": 1.4438,
+      "step": 5301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07699646091404869,
+      "learning_rate": 2.0124469175574757e-05,
+      "loss": 1.4401,
+      "step": 5302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10071097293876922,
+      "learning_rate": 2.011129008639625e-05,
+      "loss": 1.4483,
+      "step": 5303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0770760752564519,
+      "learning_rate": 2.009811099721775e-05,
+      "loss": 1.4406,
+      "step": 5304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09320028140276188,
+      "learning_rate": 2.0084931908039245e-05,
+      "loss": 1.4421,
+      "step": 5305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08426431034458194,
+      "learning_rate": 2.0071752818860742e-05,
+      "loss": 1.4408,
+      "step": 5306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08479736321377682,
+      "learning_rate": 2.005857372968224e-05,
+      "loss": 1.4451,
+      "step": 5307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0963128273375083,
+      "learning_rate": 2.0045394640503736e-05,
+      "loss": 1.4469,
+      "step": 5308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08196329387355805,
+      "learning_rate": 2.003221555132523e-05,
+      "loss": 1.4264,
+      "step": 5309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0945879007549834,
+      "learning_rate": 2.001903646214673e-05,
+      "loss": 1.439,
+      "step": 5310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08428433226337698,
+      "learning_rate": 2.0005857372968224e-05,
+      "loss": 1.4506,
+      "step": 5311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09390737294396427,
+      "learning_rate": 1.999267828378972e-05,
+      "loss": 1.4478,
+      "step": 5312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08835661951085298,
+      "learning_rate": 1.9979499194611218e-05,
+      "loss": 1.4396,
+      "step": 5313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07915409106215739,
+      "learning_rate": 1.9966320105432715e-05,
+      "loss": 1.4381,
+      "step": 5314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09665089307647438,
+      "learning_rate": 1.995314101625421e-05,
+      "loss": 1.4433,
+      "step": 5315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0704222171425156,
+      "learning_rate": 1.993996192707571e-05,
+      "loss": 1.439,
+      "step": 5316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09468340566031734,
+      "learning_rate": 1.9926782837897203e-05,
+      "loss": 1.4502,
+      "step": 5317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07484980231290964,
+      "learning_rate": 1.99136037487187e-05,
+      "loss": 1.4438,
+      "step": 5318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08157112067681922,
+      "learning_rate": 1.9900424659540198e-05,
+      "loss": 1.4449,
+      "step": 5319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07652557139322982,
+      "learning_rate": 1.9887245570361695e-05,
+      "loss": 1.4494,
+      "step": 5320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07667609037141304,
+      "learning_rate": 1.987406648118319e-05,
+      "loss": 1.4493,
+      "step": 5321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07727343135385019,
+      "learning_rate": 1.986088739200469e-05,
+      "loss": 1.4454,
+      "step": 5322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08130539827160851,
+      "learning_rate": 1.9847708302826183e-05,
+      "loss": 1.4416,
+      "step": 5323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07787781504113413,
+      "learning_rate": 1.983452921364768e-05,
+      "loss": 1.4403,
+      "step": 5324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08426556792001223,
+      "learning_rate": 1.9821350124469177e-05,
+      "loss": 1.4338,
+      "step": 5325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07029411942267813,
+      "learning_rate": 1.9808171035290674e-05,
+      "loss": 1.442,
+      "step": 5326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.09317953232098186,
+      "learning_rate": 1.9794991946112168e-05,
+      "loss": 1.4464,
+      "step": 5327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07227732861003998,
+      "learning_rate": 1.978181285693367e-05,
+      "loss": 1.4473,
+      "step": 5328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.08168887972250445,
+      "learning_rate": 1.9768633767755162e-05,
+      "loss": 1.4391,
+      "step": 5329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07694451841273235,
+      "learning_rate": 1.975545467857666e-05,
+      "loss": 1.4381,
+      "step": 5330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07442223421244215,
+      "learning_rate": 1.9742275589398156e-05,
+      "loss": 1.4366,
+      "step": 5331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0736788216963836,
+      "learning_rate": 1.9729096500219654e-05,
+      "loss": 1.4307,
+      "step": 5332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07021373413521304,
+      "learning_rate": 1.9715917411041147e-05,
+      "loss": 1.4433,
+      "step": 5333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07804391999481822,
+      "learning_rate": 1.9702738321862648e-05,
+      "loss": 1.4476,
+      "step": 5334
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07843549001391978,
+      "learning_rate": 1.968955923268414e-05,
+      "loss": 1.4469,
+      "step": 5335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0761151701777107,
+      "learning_rate": 1.967638014350564e-05,
+      "loss": 1.4441,
+      "step": 5336
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07557353636949023,
+      "learning_rate": 1.9663201054327136e-05,
+      "loss": 1.4411,
+      "step": 5337
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07810649134337913,
+      "learning_rate": 1.9650021965148633e-05,
+      "loss": 1.4421,
+      "step": 5338
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07428395750966402,
+      "learning_rate": 1.9636842875970127e-05,
+      "loss": 1.4517,
+      "step": 5339
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07586648361183178,
+      "learning_rate": 1.9623663786791627e-05,
+      "loss": 1.4416,
+      "step": 5340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0746001407192746,
+      "learning_rate": 1.961048469761312e-05,
+      "loss": 1.4399,
+      "step": 5341
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07944847015025697,
+      "learning_rate": 1.9597305608434618e-05,
+      "loss": 1.4399,
+      "step": 5342
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07612615927935437,
+      "learning_rate": 1.9584126519256115e-05,
+      "loss": 1.4415,
+      "step": 5343
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07883100370494517,
+      "learning_rate": 1.9570947430077612e-05,
+      "loss": 1.4463,
+      "step": 5344
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0708242067673502,
+      "learning_rate": 1.9557768340899106e-05,
+      "loss": 1.4424,
+      "step": 5345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07461996891316976,
+      "learning_rate": 1.9544589251720607e-05,
+      "loss": 1.4324,
+      "step": 5346
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.070797351321823,
+      "learning_rate": 1.95314101625421e-05,
+      "loss": 1.4511,
+      "step": 5347
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07942258132294246,
+      "learning_rate": 1.9518231073363597e-05,
+      "loss": 1.445,
+      "step": 5348
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07379064370987876,
+      "learning_rate": 1.9505051984185095e-05,
+      "loss": 1.4351,
+      "step": 5349
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07291071455897694,
+      "learning_rate": 1.9491872895006592e-05,
+      "loss": 1.4453,
+      "step": 5350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07062596416501056,
+      "learning_rate": 1.9478693805828085e-05,
+      "loss": 1.4401,
+      "step": 5351
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07088579665357408,
+      "learning_rate": 1.9465514716649586e-05,
+      "loss": 1.4404,
+      "step": 5352
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07312687552102666,
+      "learning_rate": 1.945233562747108e-05,
+      "loss": 1.4416,
+      "step": 5353
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06916610393656455,
+      "learning_rate": 1.9439156538292577e-05,
+      "loss": 1.4358,
+      "step": 5354
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07582367038398782,
+      "learning_rate": 1.9425977449114074e-05,
+      "loss": 1.4514,
+      "step": 5355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06650151410843223,
+      "learning_rate": 1.941279835993557e-05,
+      "loss": 1.435,
+      "step": 5356
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07497641717766314,
+      "learning_rate": 1.9399619270757065e-05,
+      "loss": 1.4441,
+      "step": 5357
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.07173818224740919,
+      "learning_rate": 1.9386440181578565e-05,
+      "loss": 1.4429,
+      "step": 5358
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06992793152427831,
+      "learning_rate": 1.937326109240006e-05,
+      "loss": 1.4444,
+      "step": 5359
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0735331680227713,
+      "learning_rate": 1.9360082003221556e-05,
+      "loss": 1.4381,
+      "step": 5360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.06981468975874108,
+      "learning_rate": 1.9346902914043053e-05,
+      "loss": 1.4346,
+      "step": 5361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07390733099727066,
+      "learning_rate": 1.933372382486455e-05,
+      "loss": 1.4394,
+      "step": 5362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07308099027506454,
+      "learning_rate": 1.9320544735686044e-05,
+      "loss": 1.436,
+      "step": 5363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0721586398600628,
+      "learning_rate": 1.9307365646507545e-05,
+      "loss": 1.4479,
+      "step": 5364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07518444419156688,
+      "learning_rate": 1.929418655732904e-05,
+      "loss": 1.4389,
+      "step": 5365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07209014673911408,
+      "learning_rate": 1.9281007468150536e-05,
+      "loss": 1.4383,
+      "step": 5366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07518134503475331,
+      "learning_rate": 1.9267828378972033e-05,
+      "loss": 1.446,
+      "step": 5367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06749646611516946,
+      "learning_rate": 1.925464928979353e-05,
+      "loss": 1.4405,
+      "step": 5368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06925509537136233,
+      "learning_rate": 1.9241470200615024e-05,
+      "loss": 1.4408,
+      "step": 5369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06832273350341453,
+      "learning_rate": 1.9228291111436524e-05,
+      "loss": 1.4402,
+      "step": 5370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07159063137027855,
+      "learning_rate": 1.9215112022258018e-05,
+      "loss": 1.4425,
+      "step": 5371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06549717250919453,
+      "learning_rate": 1.9201932933079515e-05,
+      "loss": 1.4383,
+      "step": 5372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07410591495052175,
+      "learning_rate": 1.9188753843901012e-05,
+      "loss": 1.4491,
+      "step": 5373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06595319168026842,
+      "learning_rate": 1.917557475472251e-05,
+      "loss": 1.4413,
+      "step": 5374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06919540105313828,
+      "learning_rate": 1.9162395665544003e-05,
+      "loss": 1.4486,
+      "step": 5375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06978434659211218,
+      "learning_rate": 1.9149216576365504e-05,
+      "loss": 1.4422,
+      "step": 5376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07286707119642495,
+      "learning_rate": 1.9136037487186997e-05,
+      "loss": 1.433,
+      "step": 5377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0745198605230862,
+      "learning_rate": 1.9122858398008494e-05,
+      "loss": 1.4453,
+      "step": 5378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07146593937411284,
+      "learning_rate": 1.910967930882999e-05,
+      "loss": 1.4445,
+      "step": 5379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07849383649206047,
+      "learning_rate": 1.909650021965149e-05,
+      "loss": 1.4477,
+      "step": 5380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07287605947184543,
+      "learning_rate": 1.9083321130472982e-05,
+      "loss": 1.4412,
+      "step": 5381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07688039227851792,
+      "learning_rate": 1.9070142041294483e-05,
+      "loss": 1.4435,
+      "step": 5382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07142228059967491,
+      "learning_rate": 1.9056962952115977e-05,
+      "loss": 1.4406,
+      "step": 5383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07239195318797143,
+      "learning_rate": 1.9043783862937474e-05,
+      "loss": 1.4341,
+      "step": 5384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07157373865619121,
+      "learning_rate": 1.903060477375897e-05,
+      "loss": 1.4421,
+      "step": 5385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07086167003825293,
+      "learning_rate": 1.9017425684580468e-05,
+      "loss": 1.4417,
+      "step": 5386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06929304300307879,
+      "learning_rate": 1.9004246595401962e-05,
+      "loss": 1.4319,
+      "step": 5387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0706520200630132,
+      "learning_rate": 1.8991067506223462e-05,
+      "loss": 1.4431,
+      "step": 5388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07566123542004871,
+      "learning_rate": 1.8977888417044956e-05,
+      "loss": 1.4445,
+      "step": 5389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06683551490171326,
+      "learning_rate": 1.8964709327866453e-05,
+      "loss": 1.4481,
+      "step": 5390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07485210320490993,
+      "learning_rate": 1.895153023868795e-05,
+      "loss": 1.447,
+      "step": 5391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07598189276944789,
+      "learning_rate": 1.8938351149509447e-05,
+      "loss": 1.4446,
+      "step": 5392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06858245894299334,
+      "learning_rate": 1.892517206033094e-05,
+      "loss": 1.4372,
+      "step": 5393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07215289814940697,
+      "learning_rate": 1.8911992971152442e-05,
+      "loss": 1.45,
+      "step": 5394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07343903096913885,
+      "learning_rate": 1.8898813881973935e-05,
+      "loss": 1.436,
+      "step": 5395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06821556834528009,
+      "learning_rate": 1.8885634792795433e-05,
+      "loss": 1.443,
+      "step": 5396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07372179467191475,
+      "learning_rate": 1.887245570361693e-05,
+      "loss": 1.443,
+      "step": 5397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07151391746286419,
+      "learning_rate": 1.8859276614438427e-05,
+      "loss": 1.4385,
+      "step": 5398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07332574560128335,
+      "learning_rate": 1.884609752525992e-05,
+      "loss": 1.4456,
+      "step": 5399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07095399075172829,
+      "learning_rate": 1.883291843608142e-05,
+      "loss": 1.4374,
+      "step": 5400
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.4084757566452026,
+      "eval_runtime": 445.5815,
+      "eval_samples_per_second": 40.43,
+      "eval_steps_per_second": 0.061,
+      "step": 5400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07497295198448943,
+      "learning_rate": 1.8819739346902915e-05,
+      "loss": 1.4419,
+      "step": 5401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0704524232925466,
+      "learning_rate": 1.8806560257724412e-05,
+      "loss": 1.4438,
+      "step": 5402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0769657286513065,
+      "learning_rate": 1.879338116854591e-05,
+      "loss": 1.4491,
+      "step": 5403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06991658417422238,
+      "learning_rate": 1.8780202079367406e-05,
+      "loss": 1.4406,
+      "step": 5404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07789564665991736,
+      "learning_rate": 1.87670229901889e-05,
+      "loss": 1.4489,
+      "step": 5405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07205437209499263,
+      "learning_rate": 1.87538439010104e-05,
+      "loss": 1.4417,
+      "step": 5406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07798772090251502,
+      "learning_rate": 1.8740664811831894e-05,
+      "loss": 1.4402,
+      "step": 5407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07471625831242182,
+      "learning_rate": 1.872748572265339e-05,
+      "loss": 1.4453,
+      "step": 5408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07733644972922762,
+      "learning_rate": 1.871430663347489e-05,
+      "loss": 1.4414,
+      "step": 5409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07535949278883941,
+      "learning_rate": 1.8701127544296386e-05,
+      "loss": 1.441,
+      "step": 5410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07322575401783393,
+      "learning_rate": 1.868794845511788e-05,
+      "loss": 1.4361,
+      "step": 5411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07776077736551508,
+      "learning_rate": 1.867476936593938e-05,
+      "loss": 1.4434,
+      "step": 5412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07207156713140642,
+      "learning_rate": 1.8661590276760874e-05,
+      "loss": 1.4472,
+      "step": 5413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06892685949674439,
+      "learning_rate": 1.864841118758237e-05,
+      "loss": 1.4392,
+      "step": 5414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.070262494006996,
+      "learning_rate": 1.8635232098403868e-05,
+      "loss": 1.4405,
+      "step": 5415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06998393524938062,
+      "learning_rate": 1.8622053009225365e-05,
+      "loss": 1.4362,
+      "step": 5416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0738861028640021,
+      "learning_rate": 1.860887392004686e-05,
+      "loss": 1.4454,
+      "step": 5417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07109275520464542,
+      "learning_rate": 1.8595694830868356e-05,
+      "loss": 1.4403,
+      "step": 5418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0738360180006969,
+      "learning_rate": 1.8582515741689853e-05,
+      "loss": 1.4431,
+      "step": 5419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06922074936423654,
+      "learning_rate": 1.856933665251135e-05,
+      "loss": 1.4377,
+      "step": 5420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07307027717807728,
+      "learning_rate": 1.8556157563332847e-05,
+      "loss": 1.4434,
+      "step": 5421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0734344498716073,
+      "learning_rate": 1.8542978474154344e-05,
+      "loss": 1.437,
+      "step": 5422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07178631845280475,
+      "learning_rate": 1.8529799384975838e-05,
+      "loss": 1.4391,
+      "step": 5423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07219298540259585,
+      "learning_rate": 1.8516620295797335e-05,
+      "loss": 1.4458,
+      "step": 5424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07288099700401504,
+      "learning_rate": 1.8503441206618832e-05,
+      "loss": 1.444,
+      "step": 5425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06763607529942632,
+      "learning_rate": 1.849026211744033e-05,
+      "loss": 1.4395,
+      "step": 5426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07375554584048609,
+      "learning_rate": 1.8477083028261827e-05,
+      "loss": 1.4394,
+      "step": 5427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.07603445629320019,
+      "learning_rate": 1.8463903939083324e-05,
+      "loss": 1.4475,
+      "step": 5428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.06813092551432809,
+      "learning_rate": 1.8450724849904818e-05,
+      "loss": 1.4397,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06884011050164907,
+      "learning_rate": 1.8437545760726315e-05,
+      "loss": 1.4285,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0693785863306255,
+      "learning_rate": 1.8424366671547812e-05,
+      "loss": 1.4471,
+      "step": 5431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07292501732729495,
+      "learning_rate": 1.841118758236931e-05,
+      "loss": 1.4418,
+      "step": 5432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07428410449336251,
+      "learning_rate": 1.8398008493190806e-05,
+      "loss": 1.4432,
+      "step": 5433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07418460710346911,
+      "learning_rate": 1.8384829404012303e-05,
+      "loss": 1.4465,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08233920590032036,
+      "learning_rate": 1.8371650314833797e-05,
+      "loss": 1.4347,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07178516349476441,
+      "learning_rate": 1.8358471225655294e-05,
+      "loss": 1.4396,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07866959013633094,
+      "learning_rate": 1.834529213647679e-05,
+      "loss": 1.4391,
+      "step": 5437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07256483624372856,
+      "learning_rate": 1.833211304729829e-05,
+      "loss": 1.4385,
+      "step": 5438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07748141120023748,
+      "learning_rate": 1.8318933958119785e-05,
+      "loss": 1.4465,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07592449310698503,
+      "learning_rate": 1.8305754868941283e-05,
+      "loss": 1.4421,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06909453351318431,
+      "learning_rate": 1.8292575779762776e-05,
+      "loss": 1.443,
+      "step": 5441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07834087544192946,
+      "learning_rate": 1.8279396690584273e-05,
+      "loss": 1.4421,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06521891358617028,
+      "learning_rate": 1.826621760140577e-05,
+      "loss": 1.4354,
+      "step": 5443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07694137533179915,
+      "learning_rate": 1.8253038512227268e-05,
+      "loss": 1.4452,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06929518486376159,
+      "learning_rate": 1.8239859423048765e-05,
+      "loss": 1.4367,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07373625588057962,
+      "learning_rate": 1.8226680333870262e-05,
+      "loss": 1.437,
+      "step": 5446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07030529987190413,
+      "learning_rate": 1.8213501244691756e-05,
+      "loss": 1.4432,
+      "step": 5447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0711226486257986,
+      "learning_rate": 1.8200322155513253e-05,
+      "loss": 1.4449,
+      "step": 5448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07352805261662668,
+      "learning_rate": 1.818714306633475e-05,
+      "loss": 1.4474,
+      "step": 5449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06857776263961464,
+      "learning_rate": 1.8173963977156247e-05,
+      "loss": 1.4407,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07551257526109177,
+      "learning_rate": 1.8160784887977744e-05,
+      "loss": 1.4381,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0700700404949806,
+      "learning_rate": 1.814760579879924e-05,
+      "loss": 1.4392,
+      "step": 5452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07042073120578339,
+      "learning_rate": 1.8134426709620735e-05,
+      "loss": 1.4376,
+      "step": 5453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07003665213396129,
+      "learning_rate": 1.8121247620442232e-05,
+      "loss": 1.4397,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06846093431161937,
+      "learning_rate": 1.810806853126373e-05,
+      "loss": 1.4368,
+      "step": 5455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07225380109984297,
+      "learning_rate": 1.8094889442085227e-05,
+      "loss": 1.4351,
+      "step": 5456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07001089306932014,
+      "learning_rate": 1.8081710352906724e-05,
+      "loss": 1.4345,
+      "step": 5457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07095599734262874,
+      "learning_rate": 1.806853126372822e-05,
+      "loss": 1.4461,
+      "step": 5458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06966412623871784,
+      "learning_rate": 1.8055352174549715e-05,
+      "loss": 1.4422,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06745211158704313,
+      "learning_rate": 1.804217308537121e-05,
+      "loss": 1.4414,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06906107738276043,
+      "learning_rate": 1.802899399619271e-05,
+      "loss": 1.4391,
+      "step": 5461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06808966745939936,
+      "learning_rate": 1.8015814907014206e-05,
+      "loss": 1.4341,
+      "step": 5462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07417408123090526,
+      "learning_rate": 1.8002635817835703e-05,
+      "loss": 1.4323,
+      "step": 5463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06957455444963626,
+      "learning_rate": 1.79894567286572e-05,
+      "loss": 1.44,
+      "step": 5464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07678404387485445,
+      "learning_rate": 1.7976277639478694e-05,
+      "loss": 1.4475,
+      "step": 5465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07315806367331867,
+      "learning_rate": 1.796309855030019e-05,
+      "loss": 1.4482,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06568114254391344,
+      "learning_rate": 1.7949919461121688e-05,
+      "loss": 1.4393,
+      "step": 5467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0747976975843836,
+      "learning_rate": 1.7936740371943185e-05,
+      "loss": 1.4476,
+      "step": 5468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06563948105589983,
+      "learning_rate": 1.7923561282764682e-05,
+      "loss": 1.4328,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07427138893403579,
+      "learning_rate": 1.791038219358618e-05,
+      "loss": 1.4409,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07271509129506189,
+      "learning_rate": 1.7897203104407673e-05,
+      "loss": 1.4375,
+      "step": 5471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07785130905794531,
+      "learning_rate": 1.788402401522917e-05,
+      "loss": 1.4412,
+      "step": 5472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07034803568542534,
+      "learning_rate": 1.7870844926050668e-05,
+      "loss": 1.4384,
+      "step": 5473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07363037869945846,
+      "learning_rate": 1.7857665836872165e-05,
+      "loss": 1.4407,
+      "step": 5474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07097631642558923,
+      "learning_rate": 1.7844486747693662e-05,
+      "loss": 1.44,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.08324522810584771,
+      "learning_rate": 1.783130765851516e-05,
+      "loss": 1.434,
+      "step": 5476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0784518512522241,
+      "learning_rate": 1.7818128569336653e-05,
+      "loss": 1.4381,
+      "step": 5477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06979406770048766,
+      "learning_rate": 1.780494948015815e-05,
+      "loss": 1.4417,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07334222540513458,
+      "learning_rate": 1.7791770390979647e-05,
+      "loss": 1.4323,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06906899246935655,
+      "learning_rate": 1.7778591301801144e-05,
+      "loss": 1.4329,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0819706014702987,
+      "learning_rate": 1.776541221262264e-05,
+      "loss": 1.4393,
+      "step": 5481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07170050048501976,
+      "learning_rate": 1.775223312344414e-05,
+      "loss": 1.4518,
+      "step": 5482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07406286967008459,
+      "learning_rate": 1.7739054034265632e-05,
+      "loss": 1.4401,
+      "step": 5483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07941724114743716,
+      "learning_rate": 1.772587494508713e-05,
+      "loss": 1.4487,
+      "step": 5484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06707992293363182,
+      "learning_rate": 1.7712695855908626e-05,
+      "loss": 1.4431,
+      "step": 5485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07979086788319567,
+      "learning_rate": 1.7699516766730124e-05,
+      "loss": 1.4525,
+      "step": 5486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0794078333119225,
+      "learning_rate": 1.768633767755162e-05,
+      "loss": 1.4496,
+      "step": 5487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06992632187596937,
+      "learning_rate": 1.7673158588373118e-05,
+      "loss": 1.4459,
+      "step": 5488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07824804094300161,
+      "learning_rate": 1.765997949919461e-05,
+      "loss": 1.4467,
+      "step": 5489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07511884973841888,
+      "learning_rate": 1.764680041001611e-05,
+      "loss": 1.4393,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07679002851308955,
+      "learning_rate": 1.7633621320837606e-05,
+      "loss": 1.4442,
+      "step": 5491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07106723403007098,
+      "learning_rate": 1.7620442231659103e-05,
+      "loss": 1.4361,
+      "step": 5492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07232276602034142,
+      "learning_rate": 1.76072631424806e-05,
+      "loss": 1.4422,
+      "step": 5493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07485621504477255,
+      "learning_rate": 1.7594084053302097e-05,
+      "loss": 1.4366,
+      "step": 5494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06926062346923498,
+      "learning_rate": 1.758090496412359e-05,
+      "loss": 1.4365,
+      "step": 5495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.07596706510726241,
+      "learning_rate": 1.7567725874945088e-05,
+      "loss": 1.4433,
+      "step": 5496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.06717646401622375,
+      "learning_rate": 1.7554546785766585e-05,
+      "loss": 1.4374,
+      "step": 5497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07281559731915824,
+      "learning_rate": 1.7541367696588082e-05,
+      "loss": 1.4378,
+      "step": 5498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07152377014823152,
+      "learning_rate": 1.752818860740958e-05,
+      "loss": 1.4447,
+      "step": 5499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06723721939128424,
+      "learning_rate": 1.7515009518231077e-05,
+      "loss": 1.4369,
+      "step": 5500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07444190118416973,
+      "learning_rate": 1.750183042905257e-05,
+      "loss": 1.4434,
+      "step": 5501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06950793245640345,
+      "learning_rate": 1.7488651339874067e-05,
+      "loss": 1.4446,
+      "step": 5502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06605593528329798,
+      "learning_rate": 1.7475472250695565e-05,
+      "loss": 1.4408,
+      "step": 5503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07430729484422417,
+      "learning_rate": 1.746229316151706e-05,
+      "loss": 1.4474,
+      "step": 5504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06825751568325211,
+      "learning_rate": 1.7449114072338555e-05,
+      "loss": 1.4433,
+      "step": 5505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07733248545579885,
+      "learning_rate": 1.7435934983160056e-05,
+      "loss": 1.4463,
+      "step": 5506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07024583696127931,
+      "learning_rate": 1.742275589398155e-05,
+      "loss": 1.4377,
+      "step": 5507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07765989034397697,
+      "learning_rate": 1.7409576804803047e-05,
+      "loss": 1.4347,
+      "step": 5508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07434520017631363,
+      "learning_rate": 1.7396397715624544e-05,
+      "loss": 1.4411,
+      "step": 5509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0786623596433099,
+      "learning_rate": 1.738321862644604e-05,
+      "loss": 1.4463,
+      "step": 5510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06887501221605094,
+      "learning_rate": 1.7370039537267535e-05,
+      "loss": 1.4394,
+      "step": 5511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07728063393523894,
+      "learning_rate": 1.7356860448089035e-05,
+      "loss": 1.4391,
+      "step": 5512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07699814212917322,
+      "learning_rate": 1.734368135891053e-05,
+      "loss": 1.4436,
+      "step": 5513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07280224481827438,
+      "learning_rate": 1.7330502269732026e-05,
+      "loss": 1.4347,
+      "step": 5514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07756955537421403,
+      "learning_rate": 1.7317323180553523e-05,
+      "loss": 1.4457,
+      "step": 5515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07751887751549462,
+      "learning_rate": 1.730414409137502e-05,
+      "loss": 1.437,
+      "step": 5516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07284498713795141,
+      "learning_rate": 1.7290965002196514e-05,
+      "loss": 1.4442,
+      "step": 5517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07261307757011265,
+      "learning_rate": 1.7277785913018015e-05,
+      "loss": 1.4414,
+      "step": 5518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07142791636533134,
+      "learning_rate": 1.726460682383951e-05,
+      "loss": 1.4364,
+      "step": 5519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0759152242213032,
+      "learning_rate": 1.7251427734661006e-05,
+      "loss": 1.4357,
+      "step": 5520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0714071383424332,
+      "learning_rate": 1.7238248645482503e-05,
+      "loss": 1.4404,
+      "step": 5521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07826809824206812,
+      "learning_rate": 1.7225069556304e-05,
+      "loss": 1.446,
+      "step": 5522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06951027781214732,
+      "learning_rate": 1.7211890467125494e-05,
+      "loss": 1.4482,
+      "step": 5523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07646793018949519,
+      "learning_rate": 1.7198711377946994e-05,
+      "loss": 1.4444,
+      "step": 5524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07644023292603892,
+      "learning_rate": 1.7185532288768488e-05,
+      "loss": 1.4405,
+      "step": 5525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07290590817283331,
+      "learning_rate": 1.7172353199589985e-05,
+      "loss": 1.4408,
+      "step": 5526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07828266434652154,
+      "learning_rate": 1.7159174110411482e-05,
+      "loss": 1.4341,
+      "step": 5527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06672856305131888,
+      "learning_rate": 1.714599502123298e-05,
+      "loss": 1.4405,
+      "step": 5528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07759235140134038,
+      "learning_rate": 1.7132815932054473e-05,
+      "loss": 1.4445,
+      "step": 5529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06431309830531086,
+      "learning_rate": 1.7119636842875974e-05,
+      "loss": 1.4375,
+      "step": 5530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0731667034202279,
+      "learning_rate": 1.7106457753697467e-05,
+      "loss": 1.4421,
+      "step": 5531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06870932160193705,
+      "learning_rate": 1.7093278664518964e-05,
+      "loss": 1.4398,
+      "step": 5532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0729520976402211,
+      "learning_rate": 1.708009957534046e-05,
+      "loss": 1.4402,
+      "step": 5533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06843225253391294,
+      "learning_rate": 1.706692048616196e-05,
+      "loss": 1.4417,
+      "step": 5534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06814029451940137,
+      "learning_rate": 1.7053741396983452e-05,
+      "loss": 1.4419,
+      "step": 5535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06895344476114704,
+      "learning_rate": 1.7040562307804953e-05,
+      "loss": 1.4392,
+      "step": 5536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06847890058722803,
+      "learning_rate": 1.7027383218626447e-05,
+      "loss": 1.4392,
+      "step": 5537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0714342618289602,
+      "learning_rate": 1.7014204129447944e-05,
+      "loss": 1.44,
+      "step": 5538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06904427027441162,
+      "learning_rate": 1.700102504026944e-05,
+      "loss": 1.4328,
+      "step": 5539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07361575675826765,
+      "learning_rate": 1.6987845951090938e-05,
+      "loss": 1.438,
+      "step": 5540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07384818092518805,
+      "learning_rate": 1.6974666861912432e-05,
+      "loss": 1.4408,
+      "step": 5541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06840070374777613,
+      "learning_rate": 1.6961487772733932e-05,
+      "loss": 1.4383,
+      "step": 5542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07434193121389956,
+      "learning_rate": 1.6948308683555426e-05,
+      "loss": 1.4374,
+      "step": 5543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06919892791391662,
+      "learning_rate": 1.6935129594376923e-05,
+      "loss": 1.4411,
+      "step": 5544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0738260900507658,
+      "learning_rate": 1.692195050519842e-05,
+      "loss": 1.4354,
+      "step": 5545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06578772080786978,
+      "learning_rate": 1.6908771416019917e-05,
+      "loss": 1.4412,
+      "step": 5546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.067146502726598,
+      "learning_rate": 1.689559232684141e-05,
+      "loss": 1.4317,
+      "step": 5547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07124456887423081,
+      "learning_rate": 1.688241323766291e-05,
+      "loss": 1.4268,
+      "step": 5548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07502634940309258,
+      "learning_rate": 1.6869234148484405e-05,
+      "loss": 1.4412,
+      "step": 5549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07591494753271882,
+      "learning_rate": 1.6856055059305903e-05,
+      "loss": 1.4364,
+      "step": 5550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07394448726899978,
+      "learning_rate": 1.68428759701274e-05,
+      "loss": 1.4473,
+      "step": 5551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08143951638490719,
+      "learning_rate": 1.6829696880948897e-05,
+      "loss": 1.4416,
+      "step": 5552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0708772804859353,
+      "learning_rate": 1.681651779177039e-05,
+      "loss": 1.4486,
+      "step": 5553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08406923633322634,
+      "learning_rate": 1.680333870259189e-05,
+      "loss": 1.4397,
+      "step": 5554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07189768819644575,
+      "learning_rate": 1.6790159613413385e-05,
+      "loss": 1.454,
+      "step": 5555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.08009889156477058,
+      "learning_rate": 1.6776980524234882e-05,
+      "loss": 1.4384,
+      "step": 5556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07354550452687067,
+      "learning_rate": 1.676380143505638e-05,
+      "loss": 1.4458,
+      "step": 5557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07790274264366022,
+      "learning_rate": 1.6750622345877876e-05,
+      "loss": 1.445,
+      "step": 5558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06752041318348874,
+      "learning_rate": 1.673744325669937e-05,
+      "loss": 1.429,
+      "step": 5559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07340335013993957,
+      "learning_rate": 1.672426416752087e-05,
+      "loss": 1.4433,
+      "step": 5560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07248432133838952,
+      "learning_rate": 1.6711085078342364e-05,
+      "loss": 1.4346,
+      "step": 5561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06953795324425187,
+      "learning_rate": 1.669790598916386e-05,
+      "loss": 1.443,
+      "step": 5562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07674907902852868,
+      "learning_rate": 1.668472689998536e-05,
+      "loss": 1.4381,
+      "step": 5563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.06948410127849615,
+      "learning_rate": 1.6671547810806856e-05,
+      "loss": 1.4361,
+      "step": 5564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.07437072132410909,
+      "learning_rate": 1.665836872162835e-05,
+      "loss": 1.4394,
+      "step": 5565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0710663201391156,
+      "learning_rate": 1.664518963244985e-05,
+      "loss": 1.4482,
+      "step": 5566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08404892802846024,
+      "learning_rate": 1.6632010543271344e-05,
+      "loss": 1.4466,
+      "step": 5567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07109640524798633,
+      "learning_rate": 1.661883145409284e-05,
+      "loss": 1.4379,
+      "step": 5568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07596768467050637,
+      "learning_rate": 1.6605652364914338e-05,
+      "loss": 1.4451,
+      "step": 5569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06801148518458383,
+      "learning_rate": 1.6592473275735835e-05,
+      "loss": 1.4396,
+      "step": 5570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07490409698519002,
+      "learning_rate": 1.657929418655733e-05,
+      "loss": 1.4432,
+      "step": 5571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07839177151324521,
+      "learning_rate": 1.656611509737883e-05,
+      "loss": 1.4439,
+      "step": 5572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08092035511403438,
+      "learning_rate": 1.6552936008200323e-05,
+      "loss": 1.4446,
+      "step": 5573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07432429113234777,
+      "learning_rate": 1.653975691902182e-05,
+      "loss": 1.4406,
+      "step": 5574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07964295082015344,
+      "learning_rate": 1.6526577829843317e-05,
+      "loss": 1.4453,
+      "step": 5575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07144861583365916,
+      "learning_rate": 1.6513398740664814e-05,
+      "loss": 1.4355,
+      "step": 5576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07477686771502723,
+      "learning_rate": 1.6500219651486308e-05,
+      "loss": 1.4433,
+      "step": 5577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06656791028249633,
+      "learning_rate": 1.648704056230781e-05,
+      "loss": 1.4463,
+      "step": 5578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07898746460853817,
+      "learning_rate": 1.6473861473129302e-05,
+      "loss": 1.4342,
+      "step": 5579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07027846082141008,
+      "learning_rate": 1.64606823839508e-05,
+      "loss": 1.4436,
+      "step": 5580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07231254117951577,
+      "learning_rate": 1.6447503294772297e-05,
+      "loss": 1.442,
+      "step": 5581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07010880006396623,
+      "learning_rate": 1.6434324205593794e-05,
+      "loss": 1.4388,
+      "step": 5582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07225855855314424,
+      "learning_rate": 1.6421145116415288e-05,
+      "loss": 1.4322,
+      "step": 5583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0705303195845332,
+      "learning_rate": 1.6407966027236788e-05,
+      "loss": 1.435,
+      "step": 5584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06891819746890626,
+      "learning_rate": 1.6394786938058282e-05,
+      "loss": 1.4362,
+      "step": 5585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0708577824029735,
+      "learning_rate": 1.638160784887978e-05,
+      "loss": 1.4318,
+      "step": 5586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06630403449978989,
+      "learning_rate": 1.6368428759701276e-05,
+      "loss": 1.4437,
+      "step": 5587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0669858795145188,
+      "learning_rate": 1.635524967052277e-05,
+      "loss": 1.4358,
+      "step": 5588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06734833740907027,
+      "learning_rate": 1.6342070581344267e-05,
+      "loss": 1.4468,
+      "step": 5589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07010873102253068,
+      "learning_rate": 1.6328891492165764e-05,
+      "loss": 1.4424,
+      "step": 5590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0728714132626799,
+      "learning_rate": 1.631571240298726e-05,
+      "loss": 1.4383,
+      "step": 5591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06499140775153212,
+      "learning_rate": 1.6302533313808755e-05,
+      "loss": 1.4386,
+      "step": 5592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06585588080513202,
+      "learning_rate": 1.6289354224630255e-05,
+      "loss": 1.4462,
+      "step": 5593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07068103761456729,
+      "learning_rate": 1.627617513545175e-05,
+      "loss": 1.4485,
+      "step": 5594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06463392977467264,
+      "learning_rate": 1.6262996046273246e-05,
+      "loss": 1.4421,
+      "step": 5595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0690511227055262,
+      "learning_rate": 1.6249816957094743e-05,
+      "loss": 1.4383,
+      "step": 5596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07263586755780452,
+      "learning_rate": 1.623663786791624e-05,
+      "loss": 1.4317,
+      "step": 5597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06346136450791254,
+      "learning_rate": 1.6223458778737734e-05,
+      "loss": 1.4476,
+      "step": 5598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07520105352047664,
+      "learning_rate": 1.6210279689559235e-05,
+      "loss": 1.4391,
+      "step": 5599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06924293851245693,
+      "learning_rate": 1.619710060038073e-05,
+      "loss": 1.4412,
+      "step": 5600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07334752554987714,
+      "learning_rate": 1.6183921511202226e-05,
+      "loss": 1.4391,
+      "step": 5601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06951644018985781,
+      "learning_rate": 1.6170742422023723e-05,
+      "loss": 1.4412,
+      "step": 5602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06985353483840168,
+      "learning_rate": 1.615756333284522e-05,
+      "loss": 1.4453,
+      "step": 5603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07458597827628442,
+      "learning_rate": 1.6144384243666714e-05,
+      "loss": 1.4341,
+      "step": 5604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07977094678793809,
+      "learning_rate": 1.6131205154488214e-05,
+      "loss": 1.4383,
+      "step": 5605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0774163492654342,
+      "learning_rate": 1.6118026065309708e-05,
+      "loss": 1.4437,
+      "step": 5606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0767984959661084,
+      "learning_rate": 1.6104846976131205e-05,
+      "loss": 1.4407,
+      "step": 5607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08225439624228076,
+      "learning_rate": 1.6091667886952702e-05,
+      "loss": 1.436,
+      "step": 5608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07173592680045988,
+      "learning_rate": 1.60784887977742e-05,
+      "loss": 1.434,
+      "step": 5609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07372896724035626,
+      "learning_rate": 1.6065309708595693e-05,
+      "loss": 1.4482,
+      "step": 5610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07642794837239919,
+      "learning_rate": 1.6052130619417194e-05,
+      "loss": 1.4391,
+      "step": 5611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07213145966829426,
+      "learning_rate": 1.6038951530238687e-05,
+      "loss": 1.4401,
+      "step": 5612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07417593658986094,
+      "learning_rate": 1.6025772441060184e-05,
+      "loss": 1.4423,
+      "step": 5613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07765470821602635,
+      "learning_rate": 1.601259335188168e-05,
+      "loss": 1.4403,
+      "step": 5614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0737867606581381,
+      "learning_rate": 1.599941426270318e-05,
+      "loss": 1.4452,
+      "step": 5615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07223168666992522,
+      "learning_rate": 1.5986235173524672e-05,
+      "loss": 1.4411,
+      "step": 5616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07282677836974226,
+      "learning_rate": 1.5973056084346173e-05,
+      "loss": 1.4425,
+      "step": 5617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07003134866854419,
+      "learning_rate": 1.5959876995167667e-05,
+      "loss": 1.4517,
+      "step": 5618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0704013395011192,
+      "learning_rate": 1.5946697905989164e-05,
+      "loss": 1.4389,
+      "step": 5619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06865798612656224,
+      "learning_rate": 1.593351881681066e-05,
+      "loss": 1.4375,
+      "step": 5620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07097803379678512,
+      "learning_rate": 1.5920339727632158e-05,
+      "loss": 1.4313,
+      "step": 5621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06800886887774793,
+      "learning_rate": 1.5907160638453652e-05,
+      "loss": 1.4346,
+      "step": 5622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07087822609529106,
+      "learning_rate": 1.5893981549275152e-05,
+      "loss": 1.4292,
+      "step": 5623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07506130362322896,
+      "learning_rate": 1.5880802460096646e-05,
+      "loss": 1.4501,
+      "step": 5624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0699275965500837,
+      "learning_rate": 1.5867623370918143e-05,
+      "loss": 1.4417,
+      "step": 5625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07932108637178233,
+      "learning_rate": 1.585444428173964e-05,
+      "loss": 1.4303,
+      "step": 5626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07344062360626115,
+      "learning_rate": 1.5841265192561138e-05,
+      "loss": 1.4312,
+      "step": 5627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07456560664758148,
+      "learning_rate": 1.582808610338263e-05,
+      "loss": 1.445,
+      "step": 5628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07634127226903387,
+      "learning_rate": 1.5814907014204132e-05,
+      "loss": 1.4389,
+      "step": 5629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06997577227764613,
+      "learning_rate": 1.5801727925025626e-05,
+      "loss": 1.4388,
+      "step": 5630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07077429489886254,
+      "learning_rate": 1.5788548835847123e-05,
+      "loss": 1.4466,
+      "step": 5631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.06920391280761734,
+      "learning_rate": 1.577536974666862e-05,
+      "loss": 1.4332,
+      "step": 5632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07872174493443125,
+      "learning_rate": 1.5762190657490117e-05,
+      "loss": 1.4365,
+      "step": 5633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.07028779345044214,
+      "learning_rate": 1.574901156831161e-05,
+      "loss": 1.4342,
+      "step": 5634
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.08210528020520562,
+      "learning_rate": 1.573583247913311e-05,
+      "loss": 1.4396,
+      "step": 5635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06745831470113438,
+      "learning_rate": 1.5722653389954605e-05,
+      "loss": 1.4446,
+      "step": 5636
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07093291081583777,
+      "learning_rate": 1.5709474300776102e-05,
+      "loss": 1.4332,
+      "step": 5637
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06606040469661836,
+      "learning_rate": 1.56962952115976e-05,
+      "loss": 1.4487,
+      "step": 5638
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06809985539060585,
+      "learning_rate": 1.5683116122419096e-05,
+      "loss": 1.4403,
+      "step": 5639
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06656735556934738,
+      "learning_rate": 1.566993703324059e-05,
+      "loss": 1.4382,
+      "step": 5640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0697475014143814,
+      "learning_rate": 1.565675794406209e-05,
+      "loss": 1.4371,
+      "step": 5641
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06761250999756914,
+      "learning_rate": 1.5643578854883584e-05,
+      "loss": 1.4351,
+      "step": 5642
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07340178531385398,
+      "learning_rate": 1.563039976570508e-05,
+      "loss": 1.44,
+      "step": 5643
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06319887085256984,
+      "learning_rate": 1.561722067652658e-05,
+      "loss": 1.4387,
+      "step": 5644
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07248017315998241,
+      "learning_rate": 1.5604041587348076e-05,
+      "loss": 1.435,
+      "step": 5645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0652418957742279,
+      "learning_rate": 1.559086249816957e-05,
+      "loss": 1.4312,
+      "step": 5646
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06958686557141844,
+      "learning_rate": 1.557768340899107e-05,
+      "loss": 1.4291,
+      "step": 5647
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0687934983607891,
+      "learning_rate": 1.5564504319812564e-05,
+      "loss": 1.4421,
+      "step": 5648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06856072728501372,
+      "learning_rate": 1.555132523063406e-05,
+      "loss": 1.4365,
+      "step": 5649
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07004714911427527,
+      "learning_rate": 1.5538146141455558e-05,
+      "loss": 1.435,
+      "step": 5650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06954101628113421,
+      "learning_rate": 1.5524967052277055e-05,
+      "loss": 1.4329,
+      "step": 5651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06766818697479357,
+      "learning_rate": 1.551178796309855e-05,
+      "loss": 1.435,
+      "step": 5652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06684617827296709,
+      "learning_rate": 1.549860887392005e-05,
+      "loss": 1.4399,
+      "step": 5653
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06976108449070996,
+      "learning_rate": 1.5485429784741543e-05,
+      "loss": 1.4305,
+      "step": 5654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06887961987817987,
+      "learning_rate": 1.547225069556304e-05,
+      "loss": 1.4478,
+      "step": 5655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07102910282770773,
+      "learning_rate": 1.5459071606384537e-05,
+      "loss": 1.4386,
+      "step": 5656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07161216850434166,
+      "learning_rate": 1.5445892517206034e-05,
+      "loss": 1.4299,
+      "step": 5657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07062105514511686,
+      "learning_rate": 1.5432713428027528e-05,
+      "loss": 1.4417,
+      "step": 5658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0728572903383431,
+      "learning_rate": 1.541953433884903e-05,
+      "loss": 1.4347,
+      "step": 5659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07137439282596927,
+      "learning_rate": 1.5406355249670522e-05,
+      "loss": 1.4316,
+      "step": 5660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07172805747172924,
+      "learning_rate": 1.539317616049202e-05,
+      "loss": 1.4348,
+      "step": 5661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06726222644798381,
+      "learning_rate": 1.5379997071313517e-05,
+      "loss": 1.4445,
+      "step": 5662
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06802673329845242,
+      "learning_rate": 1.5366817982135014e-05,
+      "loss": 1.4385,
+      "step": 5663
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07099172180961795,
+      "learning_rate": 1.5353638892956508e-05,
+      "loss": 1.4375,
+      "step": 5664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06898818008437907,
+      "learning_rate": 1.5340459803778008e-05,
+      "loss": 1.4391,
+      "step": 5665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06660045215396766,
+      "learning_rate": 1.5327280714599502e-05,
+      "loss": 1.4403,
+      "step": 5666
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07157034573791544,
+      "learning_rate": 1.5314101625421e-05,
+      "loss": 1.4384,
+      "step": 5667
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07263944830610182,
+      "learning_rate": 1.5300922536242496e-05,
+      "loss": 1.4367,
+      "step": 5668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07072039214246044,
+      "learning_rate": 1.5287743447063993e-05,
+      "loss": 1.4308,
+      "step": 5669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06677063422940431,
+      "learning_rate": 1.5274564357885487e-05,
+      "loss": 1.4431,
+      "step": 5670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07017444578869322,
+      "learning_rate": 1.5261385268706988e-05,
+      "loss": 1.4376,
+      "step": 5671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06899828645125414,
+      "learning_rate": 1.5248206179528481e-05,
+      "loss": 1.4352,
+      "step": 5672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06694477988259331,
+      "learning_rate": 1.5235027090349978e-05,
+      "loss": 1.4311,
+      "step": 5673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06735006641102312,
+      "learning_rate": 1.5221848001171474e-05,
+      "loss": 1.4335,
+      "step": 5674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06864713731352294,
+      "learning_rate": 1.5208668911992973e-05,
+      "loss": 1.4341,
+      "step": 5675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07027438302229246,
+      "learning_rate": 1.5195489822814468e-05,
+      "loss": 1.4357,
+      "step": 5676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06710603731710538,
+      "learning_rate": 1.5182310733635965e-05,
+      "loss": 1.43,
+      "step": 5677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07049898917009807,
+      "learning_rate": 1.516913164445746e-05,
+      "loss": 1.4483,
+      "step": 5678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06982378281151665,
+      "learning_rate": 1.5155952555278958e-05,
+      "loss": 1.4407,
+      "step": 5679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06845020300552049,
+      "learning_rate": 1.5142773466100453e-05,
+      "loss": 1.4401,
+      "step": 5680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0714692602875079,
+      "learning_rate": 1.5129594376921952e-05,
+      "loss": 1.4407,
+      "step": 5681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0715333827474906,
+      "learning_rate": 1.5116415287743447e-05,
+      "loss": 1.4386,
+      "step": 5682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07338505744633932,
+      "learning_rate": 1.5103236198564945e-05,
+      "loss": 1.4344,
+      "step": 5683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07002292253794551,
+      "learning_rate": 1.509005710938644e-05,
+      "loss": 1.4391,
+      "step": 5684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07216300340768128,
+      "learning_rate": 1.5076878020207937e-05,
+      "loss": 1.4362,
+      "step": 5685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07564451850819455,
+      "learning_rate": 1.5063698931029433e-05,
+      "loss": 1.4439,
+      "step": 5686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07282106968473069,
+      "learning_rate": 1.5050519841850931e-05,
+      "loss": 1.4364,
+      "step": 5687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07040911186294593,
+      "learning_rate": 1.5037340752672427e-05,
+      "loss": 1.4384,
+      "step": 5688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06797663827967552,
+      "learning_rate": 1.5024161663493924e-05,
+      "loss": 1.4349,
+      "step": 5689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06689068182529082,
+      "learning_rate": 1.501098257431542e-05,
+      "loss": 1.4363,
+      "step": 5690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07035219239688761,
+      "learning_rate": 1.4997803485136917e-05,
+      "loss": 1.4404,
+      "step": 5691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06911783053123281,
+      "learning_rate": 1.4984624395958412e-05,
+      "loss": 1.438,
+      "step": 5692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06771992259047283,
+      "learning_rate": 1.497144530677991e-05,
+      "loss": 1.432,
+      "step": 5693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06537726388298089,
+      "learning_rate": 1.4958266217601406e-05,
+      "loss": 1.4364,
+      "step": 5694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07034885705113526,
+      "learning_rate": 1.4945087128422903e-05,
+      "loss": 1.4239,
+      "step": 5695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07037929540901101,
+      "learning_rate": 1.4931908039244399e-05,
+      "loss": 1.4307,
+      "step": 5696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06889452422597926,
+      "learning_rate": 1.4918728950065896e-05,
+      "loss": 1.4425,
+      "step": 5697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.07109110937062302,
+      "learning_rate": 1.4905549860887391e-05,
+      "loss": 1.44,
+      "step": 5698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06865496383187907,
+      "learning_rate": 1.489237077170889e-05,
+      "loss": 1.4265,
+      "step": 5699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.06387202851927402,
+      "learning_rate": 1.4879191682530386e-05,
+      "loss": 1.4365,
+      "step": 5700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0676068962417411,
+      "learning_rate": 1.4866012593351883e-05,
+      "loss": 1.4345,
+      "step": 5701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0658460819136214,
+      "learning_rate": 1.4852833504173378e-05,
+      "loss": 1.4418,
+      "step": 5702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07414940928594166,
+      "learning_rate": 1.4839654414994875e-05,
+      "loss": 1.437,
+      "step": 5703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06770368813063847,
+      "learning_rate": 1.482647532581637e-05,
+      "loss": 1.4376,
+      "step": 5704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07497446440318321,
+      "learning_rate": 1.481329623663787e-05,
+      "loss": 1.4383,
+      "step": 5705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07309696117389146,
+      "learning_rate": 1.4800117147459365e-05,
+      "loss": 1.4375,
+      "step": 5706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07514222170774895,
+      "learning_rate": 1.4786938058280862e-05,
+      "loss": 1.4368,
+      "step": 5707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07463821194269105,
+      "learning_rate": 1.4773758969102358e-05,
+      "loss": 1.4347,
+      "step": 5708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0721250589408403,
+      "learning_rate": 1.4760579879923855e-05,
+      "loss": 1.4385,
+      "step": 5709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07717408141672805,
+      "learning_rate": 1.474740079074535e-05,
+      "loss": 1.4424,
+      "step": 5710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0735249919845923,
+      "learning_rate": 1.4734221701566849e-05,
+      "loss": 1.4416,
+      "step": 5711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07381238429040704,
+      "learning_rate": 1.4721042612388344e-05,
+      "loss": 1.4464,
+      "step": 5712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07106441077896332,
+      "learning_rate": 1.4707863523209842e-05,
+      "loss": 1.4226,
+      "step": 5713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07263863988938336,
+      "learning_rate": 1.4694684434031337e-05,
+      "loss": 1.4373,
+      "step": 5714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07046352181425125,
+      "learning_rate": 1.4681505344852834e-05,
+      "loss": 1.4408,
+      "step": 5715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.08002526486865331,
+      "learning_rate": 1.466832625567433e-05,
+      "loss": 1.4353,
+      "step": 5716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06888043226055253,
+      "learning_rate": 1.4655147166495828e-05,
+      "loss": 1.4408,
+      "step": 5717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07966066938697891,
+      "learning_rate": 1.4641968077317324e-05,
+      "loss": 1.4376,
+      "step": 5718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06723135921146074,
+      "learning_rate": 1.4628788988138821e-05,
+      "loss": 1.445,
+      "step": 5719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07601075504716964,
+      "learning_rate": 1.4615609898960316e-05,
+      "loss": 1.4466,
+      "step": 5720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0727367622444813,
+      "learning_rate": 1.4602430809781814e-05,
+      "loss": 1.4331,
+      "step": 5721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07340798519770167,
+      "learning_rate": 1.4589251720603309e-05,
+      "loss": 1.4462,
+      "step": 5722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07824803911946751,
+      "learning_rate": 1.4576072631424808e-05,
+      "loss": 1.4354,
+      "step": 5723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07065969251332352,
+      "learning_rate": 1.4562893542246303e-05,
+      "loss": 1.4367,
+      "step": 5724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0749250053481166,
+      "learning_rate": 1.45497144530678e-05,
+      "loss": 1.4359,
+      "step": 5725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07147068068010794,
+      "learning_rate": 1.4536535363889296e-05,
+      "loss": 1.4354,
+      "step": 5726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06447849842621906,
+      "learning_rate": 1.4523356274710793e-05,
+      "loss": 1.4353,
+      "step": 5727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07254895386954738,
+      "learning_rate": 1.4510177185532288e-05,
+      "loss": 1.4402,
+      "step": 5728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06947613625659632,
+      "learning_rate": 1.4496998096353787e-05,
+      "loss": 1.4291,
+      "step": 5729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07053924938128145,
+      "learning_rate": 1.4483819007175283e-05,
+      "loss": 1.4392,
+      "step": 5730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07150741535032405,
+      "learning_rate": 1.447063991799678e-05,
+      "loss": 1.4277,
+      "step": 5731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06532157299645903,
+      "learning_rate": 1.4457460828818275e-05,
+      "loss": 1.4488,
+      "step": 5732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07106752120949417,
+      "learning_rate": 1.4444281739639772e-05,
+      "loss": 1.4357,
+      "step": 5733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06418334738539747,
+      "learning_rate": 1.4431102650461268e-05,
+      "loss": 1.4396,
+      "step": 5734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07016589135453476,
+      "learning_rate": 1.4417923561282767e-05,
+      "loss": 1.436,
+      "step": 5735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06777658388526672,
+      "learning_rate": 1.4404744472104262e-05,
+      "loss": 1.4376,
+      "step": 5736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06703611509546493,
+      "learning_rate": 1.4391565382925759e-05,
+      "loss": 1.4378,
+      "step": 5737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06976986745820801,
+      "learning_rate": 1.4378386293747255e-05,
+      "loss": 1.4392,
+      "step": 5738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06576902614863372,
+      "learning_rate": 1.4365207204568752e-05,
+      "loss": 1.4399,
+      "step": 5739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06656810616399383,
+      "learning_rate": 1.4352028115390247e-05,
+      "loss": 1.4406,
+      "step": 5740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07164252769787428,
+      "learning_rate": 1.4338849026211746e-05,
+      "loss": 1.4321,
+      "step": 5741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07232444045001955,
+      "learning_rate": 1.4325669937033241e-05,
+      "loss": 1.4423,
+      "step": 5742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07170573469508447,
+      "learning_rate": 1.4312490847854739e-05,
+      "loss": 1.4416,
+      "step": 5743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06678964386859829,
+      "learning_rate": 1.4299311758676234e-05,
+      "loss": 1.4399,
+      "step": 5744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06868591481234393,
+      "learning_rate": 1.4286132669497731e-05,
+      "loss": 1.4329,
+      "step": 5745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06935737442197973,
+      "learning_rate": 1.4272953580319227e-05,
+      "loss": 1.4442,
+      "step": 5746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06959724054571566,
+      "learning_rate": 1.4259774491140725e-05,
+      "loss": 1.4385,
+      "step": 5747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06869538548845723,
+      "learning_rate": 1.4246595401962219e-05,
+      "loss": 1.436,
+      "step": 5748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0657989480311263,
+      "learning_rate": 1.4233416312783718e-05,
+      "loss": 1.4369,
+      "step": 5749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0694434968805375,
+      "learning_rate": 1.4220237223605213e-05,
+      "loss": 1.4387,
+      "step": 5750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06551858755810379,
+      "learning_rate": 1.420705813442671e-05,
+      "loss": 1.4408,
+      "step": 5751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07105537256379536,
+      "learning_rate": 1.4193879045248206e-05,
+      "loss": 1.4346,
+      "step": 5752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0660556229805991,
+      "learning_rate": 1.4180699956069705e-05,
+      "loss": 1.4373,
+      "step": 5753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06577338111462862,
+      "learning_rate": 1.4167520866891198e-05,
+      "loss": 1.4366,
+      "step": 5754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06475560350894426,
+      "learning_rate": 1.4154341777712697e-05,
+      "loss": 1.446,
+      "step": 5755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06437075049439099,
+      "learning_rate": 1.4141162688534193e-05,
+      "loss": 1.4457,
+      "step": 5756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06735355310509238,
+      "learning_rate": 1.412798359935569e-05,
+      "loss": 1.4398,
+      "step": 5757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0714177199333853,
+      "learning_rate": 1.4114804510177185e-05,
+      "loss": 1.439,
+      "step": 5758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06480566772251892,
+      "learning_rate": 1.4101625420998684e-05,
+      "loss": 1.4449,
+      "step": 5759
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06927752188701565,
+      "learning_rate": 1.4088446331820178e-05,
+      "loss": 1.443,
+      "step": 5760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06487315948247291,
+      "learning_rate": 1.4075267242641677e-05,
+      "loss": 1.4333,
+      "step": 5761
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06542287767183998,
+      "learning_rate": 1.4062088153463172e-05,
+      "loss": 1.4474,
+      "step": 5762
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06464944931450539,
+      "learning_rate": 1.404890906428467e-05,
+      "loss": 1.4351,
+      "step": 5763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06608412398267707,
+      "learning_rate": 1.4035729975106165e-05,
+      "loss": 1.4403,
+      "step": 5764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06573897589575929,
+      "learning_rate": 1.4022550885927662e-05,
+      "loss": 1.4371,
+      "step": 5765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.066638282675829,
+      "learning_rate": 1.4009371796749157e-05,
+      "loss": 1.4335,
+      "step": 5766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06618770729986596,
+      "learning_rate": 1.3996192707570656e-05,
+      "loss": 1.4385,
+      "step": 5767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06722008976881554,
+      "learning_rate": 1.3983013618392152e-05,
+      "loss": 1.4406,
+      "step": 5768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.06682507364930695,
+      "learning_rate": 1.3969834529213649e-05,
+      "loss": 1.435,
+      "step": 5769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.07082379995705217,
+      "learning_rate": 1.3956655440035144e-05,
+      "loss": 1.4334,
+      "step": 5770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0723877523372393,
+      "learning_rate": 1.3943476350856641e-05,
+      "loss": 1.4304,
+      "step": 5771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06592914144049894,
+      "learning_rate": 1.3930297261678137e-05,
+      "loss": 1.4416,
+      "step": 5772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08080409560424803,
+      "learning_rate": 1.3917118172499635e-05,
+      "loss": 1.4438,
+      "step": 5773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0707080976351095,
+      "learning_rate": 1.3903939083321131e-05,
+      "loss": 1.4432,
+      "step": 5774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08136050919219275,
+      "learning_rate": 1.3890759994142628e-05,
+      "loss": 1.4424,
+      "step": 5775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07372071516787128,
+      "learning_rate": 1.3877580904964123e-05,
+      "loss": 1.4309,
+      "step": 5776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.08119585103670353,
+      "learning_rate": 1.386440181578562e-05,
+      "loss": 1.4413,
+      "step": 5777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07497185299147686,
+      "learning_rate": 1.3851222726607116e-05,
+      "loss": 1.441,
+      "step": 5778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07205214720399425,
+      "learning_rate": 1.3838043637428615e-05,
+      "loss": 1.4489,
+      "step": 5779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07850613506513904,
+      "learning_rate": 1.382486454825011e-05,
+      "loss": 1.4306,
+      "step": 5780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07069046043649256,
+      "learning_rate": 1.3811685459071607e-05,
+      "loss": 1.4407,
+      "step": 5781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07423023642769812,
+      "learning_rate": 1.3798506369893103e-05,
+      "loss": 1.4377,
+      "step": 5782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06558533985793935,
+      "learning_rate": 1.37853272807146e-05,
+      "loss": 1.4256,
+      "step": 5783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07720056569267685,
+      "learning_rate": 1.3772148191536095e-05,
+      "loss": 1.4322,
+      "step": 5784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06643077828437759,
+      "learning_rate": 1.3758969102357594e-05,
+      "loss": 1.443,
+      "step": 5785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07038524384380188,
+      "learning_rate": 1.374579001317909e-05,
+      "loss": 1.4316,
+      "step": 5786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06769627602504971,
+      "learning_rate": 1.3732610924000587e-05,
+      "loss": 1.4365,
+      "step": 5787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06650384014833126,
+      "learning_rate": 1.3719431834822082e-05,
+      "loss": 1.4417,
+      "step": 5788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06863305177630002,
+      "learning_rate": 1.370625274564358e-05,
+      "loss": 1.4479,
+      "step": 5789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07053884467735118,
+      "learning_rate": 1.3693073656465075e-05,
+      "loss": 1.4419,
+      "step": 5790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06590957489887532,
+      "learning_rate": 1.3679894567286574e-05,
+      "loss": 1.4378,
+      "step": 5791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06574366683002669,
+      "learning_rate": 1.3666715478108069e-05,
+      "loss": 1.431,
+      "step": 5792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06536153796959611,
+      "learning_rate": 1.3653536388929566e-05,
+      "loss": 1.4392,
+      "step": 5793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06468954151068969,
+      "learning_rate": 1.3640357299751062e-05,
+      "loss": 1.4333,
+      "step": 5794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06390819664792309,
+      "learning_rate": 1.3627178210572559e-05,
+      "loss": 1.4389,
+      "step": 5795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06562575251628563,
+      "learning_rate": 1.3613999121394054e-05,
+      "loss": 1.4441,
+      "step": 5796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07340998708637635,
+      "learning_rate": 1.3600820032215553e-05,
+      "loss": 1.4287,
+      "step": 5797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0647630402151328,
+      "learning_rate": 1.3587640943037048e-05,
+      "loss": 1.4396,
+      "step": 5798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07014018535281608,
+      "learning_rate": 1.3574461853858546e-05,
+      "loss": 1.4433,
+      "step": 5799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06704009514830976,
+      "learning_rate": 1.3561282764680041e-05,
+      "loss": 1.4385,
+      "step": 5800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06602606667669766,
+      "learning_rate": 1.3548103675501538e-05,
+      "loss": 1.4443,
+      "step": 5801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06572675565789518,
+      "learning_rate": 1.3534924586323034e-05,
+      "loss": 1.4345,
+      "step": 5802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06762562454084196,
+      "learning_rate": 1.3521745497144532e-05,
+      "loss": 1.4399,
+      "step": 5803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06756999911175646,
+      "learning_rate": 1.3508566407966028e-05,
+      "loss": 1.4372,
+      "step": 5804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06953399350451961,
+      "learning_rate": 1.3495387318787525e-05,
+      "loss": 1.4345,
+      "step": 5805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0684559281504442,
+      "learning_rate": 1.348220822960902e-05,
+      "loss": 1.4355,
+      "step": 5806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07133793221303567,
+      "learning_rate": 1.3469029140430518e-05,
+      "loss": 1.4399,
+      "step": 5807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06556999864741482,
+      "learning_rate": 1.3455850051252013e-05,
+      "loss": 1.4415,
+      "step": 5808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06790942370331549,
+      "learning_rate": 1.3442670962073512e-05,
+      "loss": 1.4355,
+      "step": 5809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06557027299621976,
+      "learning_rate": 1.3429491872895007e-05,
+      "loss": 1.4414,
+      "step": 5810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06803936517871116,
+      "learning_rate": 1.3416312783716504e-05,
+      "loss": 1.4382,
+      "step": 5811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06635591301434474,
+      "learning_rate": 1.3403133694538e-05,
+      "loss": 1.4382,
+      "step": 5812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06880623871007958,
+      "learning_rate": 1.3389954605359497e-05,
+      "loss": 1.4344,
+      "step": 5813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06656997986048219,
+      "learning_rate": 1.3376775516180992e-05,
+      "loss": 1.4358,
+      "step": 5814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06743129563405764,
+      "learning_rate": 1.3363596427002491e-05,
+      "loss": 1.4323,
+      "step": 5815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06558860640736316,
+      "learning_rate": 1.3350417337823987e-05,
+      "loss": 1.4306,
+      "step": 5816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06856149524577222,
+      "learning_rate": 1.3337238248645484e-05,
+      "loss": 1.4343,
+      "step": 5817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06779986230272042,
+      "learning_rate": 1.332405915946698e-05,
+      "loss": 1.4359,
+      "step": 5818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07275292228161709,
+      "learning_rate": 1.3310880070288476e-05,
+      "loss": 1.4458,
+      "step": 5819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06835754025975596,
+      "learning_rate": 1.3297700981109972e-05,
+      "loss": 1.4376,
+      "step": 5820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07237512580980045,
+      "learning_rate": 1.328452189193147e-05,
+      "loss": 1.4353,
+      "step": 5821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0692716246671913,
+      "learning_rate": 1.3271342802752966e-05,
+      "loss": 1.4426,
+      "step": 5822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06666420933940564,
+      "learning_rate": 1.3258163713574463e-05,
+      "loss": 1.4392,
+      "step": 5823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0658349216290791,
+      "learning_rate": 1.3244984624395959e-05,
+      "loss": 1.44,
+      "step": 5824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06618397752303114,
+      "learning_rate": 1.3231805535217456e-05,
+      "loss": 1.4257,
+      "step": 5825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06724422935043456,
+      "learning_rate": 1.3218626446038951e-05,
+      "loss": 1.4423,
+      "step": 5826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06467263652553935,
+      "learning_rate": 1.320544735686045e-05,
+      "loss": 1.4359,
+      "step": 5827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06847146418156899,
+      "learning_rate": 1.3192268267681945e-05,
+      "loss": 1.4332,
+      "step": 5828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0676879939689068,
+      "learning_rate": 1.3179089178503443e-05,
+      "loss": 1.437,
+      "step": 5829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06591954428388089,
+      "learning_rate": 1.3165910089324938e-05,
+      "loss": 1.4369,
+      "step": 5830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06789456017380503,
+      "learning_rate": 1.3152731000146435e-05,
+      "loss": 1.4291,
+      "step": 5831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06619202833279934,
+      "learning_rate": 1.313955191096793e-05,
+      "loss": 1.4471,
+      "step": 5832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06875564875064381,
+      "learning_rate": 1.312637282178943e-05,
+      "loss": 1.4396,
+      "step": 5833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06555611721679766,
+      "learning_rate": 1.3113193732610925e-05,
+      "loss": 1.4282,
+      "step": 5834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.07285953774268647,
+      "learning_rate": 1.3100014643432422e-05,
+      "loss": 1.4412,
+      "step": 5835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06814014863961454,
+      "learning_rate": 1.3086835554253917e-05,
+      "loss": 1.4405,
+      "step": 5836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06851754867191612,
+      "learning_rate": 1.3073656465075415e-05,
+      "loss": 1.4401,
+      "step": 5837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06667961504285423,
+      "learning_rate": 1.306047737589691e-05,
+      "loss": 1.4379,
+      "step": 5838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.06805313206478761,
+      "learning_rate": 1.3047298286718409e-05,
+      "loss": 1.4466,
+      "step": 5839
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06673376775266611,
+      "learning_rate": 1.3034119197539904e-05,
+      "loss": 1.4373,
+      "step": 5840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07125326022272807,
+      "learning_rate": 1.3020940108361401e-05,
+      "loss": 1.4329,
+      "step": 5841
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06469644078701907,
+      "learning_rate": 1.3007761019182897e-05,
+      "loss": 1.4336,
+      "step": 5842
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0661785645360638,
+      "learning_rate": 1.2994581930004394e-05,
+      "loss": 1.4393,
+      "step": 5843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06308997361231014,
+      "learning_rate": 1.298140284082589e-05,
+      "loss": 1.4393,
+      "step": 5844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06607644497557429,
+      "learning_rate": 1.2968223751647388e-05,
+      "loss": 1.4421,
+      "step": 5845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06871900676534892,
+      "learning_rate": 1.2955044662468884e-05,
+      "loss": 1.4385,
+      "step": 5846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06441314998231182,
+      "learning_rate": 1.294186557329038e-05,
+      "loss": 1.4267,
+      "step": 5847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07857182486052236,
+      "learning_rate": 1.2928686484111876e-05,
+      "loss": 1.4429,
+      "step": 5848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06247671503692748,
+      "learning_rate": 1.2915507394933373e-05,
+      "loss": 1.4389,
+      "step": 5849
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.08205440640291385,
+      "learning_rate": 1.2902328305754869e-05,
+      "loss": 1.4259,
+      "step": 5850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06958419062261491,
+      "learning_rate": 1.2889149216576368e-05,
+      "loss": 1.4316,
+      "step": 5851
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07134658611274781,
+      "learning_rate": 1.2875970127397861e-05,
+      "loss": 1.4407,
+      "step": 5852
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07057143380491167,
+      "learning_rate": 1.286279103821936e-05,
+      "loss": 1.4468,
+      "step": 5853
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06582697883208687,
+      "learning_rate": 1.2849611949040856e-05,
+      "loss": 1.4386,
+      "step": 5854
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07020514855972793,
+      "learning_rate": 1.2836432859862353e-05,
+      "loss": 1.4417,
+      "step": 5855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06822458485176383,
+      "learning_rate": 1.2823253770683848e-05,
+      "loss": 1.4399,
+      "step": 5856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06872020869098086,
+      "learning_rate": 1.2810074681505347e-05,
+      "loss": 1.4379,
+      "step": 5857
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06430498624218695,
+      "learning_rate": 1.279689559232684e-05,
+      "loss": 1.4309,
+      "step": 5858
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06794693294635164,
+      "learning_rate": 1.278371650314834e-05,
+      "loss": 1.437,
+      "step": 5859
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06396101061172474,
+      "learning_rate": 1.2770537413969835e-05,
+      "loss": 1.4316,
+      "step": 5860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06788988227092442,
+      "learning_rate": 1.2757358324791332e-05,
+      "loss": 1.4367,
+      "step": 5861
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0646376739327382,
+      "learning_rate": 1.2744179235612828e-05,
+      "loss": 1.4437,
+      "step": 5862
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06645539591275466,
+      "learning_rate": 1.2731000146434326e-05,
+      "loss": 1.4395,
+      "step": 5863
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.05952017901954691,
+      "learning_rate": 1.271782105725582e-05,
+      "loss": 1.432,
+      "step": 5864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06470028987538516,
+      "learning_rate": 1.2704641968077319e-05,
+      "loss": 1.433,
+      "step": 5865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06060624749867768,
+      "learning_rate": 1.2691462878898814e-05,
+      "loss": 1.4303,
+      "step": 5866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06506649247707916,
+      "learning_rate": 1.2678283789720312e-05,
+      "loss": 1.431,
+      "step": 5867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.061840050308159165,
+      "learning_rate": 1.2665104700541807e-05,
+      "loss": 1.4359,
+      "step": 5868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06057217718725473,
+      "learning_rate": 1.2651925611363306e-05,
+      "loss": 1.4408,
+      "step": 5869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06643913074695947,
+      "learning_rate": 1.26387465221848e-05,
+      "loss": 1.4374,
+      "step": 5870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06615003796527581,
+      "learning_rate": 1.2625567433006298e-05,
+      "loss": 1.43,
+      "step": 5871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06619780204292701,
+      "learning_rate": 1.2612388343827794e-05,
+      "loss": 1.4409,
+      "step": 5872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07025291529844159,
+      "learning_rate": 1.2599209254649291e-05,
+      "loss": 1.4403,
+      "step": 5873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06811167754588343,
+      "learning_rate": 1.2586030165470786e-05,
+      "loss": 1.4411,
+      "step": 5874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06981502296120048,
+      "learning_rate": 1.2572851076292283e-05,
+      "loss": 1.4327,
+      "step": 5875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06472661552662565,
+      "learning_rate": 1.2559671987113779e-05,
+      "loss": 1.4376,
+      "step": 5876
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0676663998720753,
+      "learning_rate": 1.2546492897935278e-05,
+      "loss": 1.4319,
+      "step": 5877
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06768935321899246,
+      "learning_rate": 1.2533313808756773e-05,
+      "loss": 1.4402,
+      "step": 5878
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06952133241369987,
+      "learning_rate": 1.252013471957827e-05,
+      "loss": 1.4293,
+      "step": 5879
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06539731365821803,
+      "learning_rate": 1.2506955630399766e-05,
+      "loss": 1.4369,
+      "step": 5880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07121097816593011,
+      "learning_rate": 1.2493776541221263e-05,
+      "loss": 1.444,
+      "step": 5881
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06738638194409409,
+      "learning_rate": 1.2480597452042758e-05,
+      "loss": 1.4392,
+      "step": 5882
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0683679194416749,
+      "learning_rate": 1.2467418362864257e-05,
+      "loss": 1.4374,
+      "step": 5883
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07874282451516422,
+      "learning_rate": 1.2454239273685753e-05,
+      "loss": 1.4403,
+      "step": 5884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06393675502044685,
+      "learning_rate": 1.244106018450725e-05,
+      "loss": 1.4359,
+      "step": 5885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07869776469804085,
+      "learning_rate": 1.2427881095328745e-05,
+      "loss": 1.4384,
+      "step": 5886
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06689165138155057,
+      "learning_rate": 1.2414702006150242e-05,
+      "loss": 1.4298,
+      "step": 5887
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07440695067852736,
+      "learning_rate": 1.2401522916971738e-05,
+      "loss": 1.4362,
+      "step": 5888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06591246974220859,
+      "learning_rate": 1.2388343827793237e-05,
+      "loss": 1.4256,
+      "step": 5889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06948629139629062,
+      "learning_rate": 1.2375164738614732e-05,
+      "loss": 1.4337,
+      "step": 5890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06680756545440576,
+      "learning_rate": 1.2361985649436229e-05,
+      "loss": 1.4391,
+      "step": 5891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.068962952637926,
+      "learning_rate": 1.2348806560257725e-05,
+      "loss": 1.4397,
+      "step": 5892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06641884189524577,
+      "learning_rate": 1.2335627471079222e-05,
+      "loss": 1.4384,
+      "step": 5893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06782272762370473,
+      "learning_rate": 1.2322448381900717e-05,
+      "loss": 1.4345,
+      "step": 5894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06904124580875798,
+      "learning_rate": 1.2309269292722216e-05,
+      "loss": 1.449,
+      "step": 5895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0643600198157894,
+      "learning_rate": 1.2296090203543711e-05,
+      "loss": 1.4383,
+      "step": 5896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07288910014698792,
+      "learning_rate": 1.2282911114365208e-05,
+      "loss": 1.434,
+      "step": 5897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06896071139917359,
+      "learning_rate": 1.2269732025186704e-05,
+      "loss": 1.428,
+      "step": 5898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06456746645086497,
+      "learning_rate": 1.2256552936008201e-05,
+      "loss": 1.44,
+      "step": 5899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06790746562710528,
+      "learning_rate": 1.2243373846829696e-05,
+      "loss": 1.4323,
+      "step": 5900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06780343862475774,
+      "learning_rate": 1.2230194757651195e-05,
+      "loss": 1.4338,
+      "step": 5901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06531804155083032,
+      "learning_rate": 1.221701566847269e-05,
+      "loss": 1.4398,
+      "step": 5902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.07137185736649608,
+      "learning_rate": 1.2203836579294188e-05,
+      "loss": 1.4456,
+      "step": 5903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06311931102810903,
+      "learning_rate": 1.2190657490115683e-05,
+      "loss": 1.439,
+      "step": 5904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06485993005501432,
+      "learning_rate": 1.217747840093718e-05,
+      "loss": 1.4332,
+      "step": 5905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06609256541724243,
+      "learning_rate": 1.2164299311758676e-05,
+      "loss": 1.4334,
+      "step": 5906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.06810219399804832,
+      "learning_rate": 1.2151120222580175e-05,
+      "loss": 1.4413,
+      "step": 5907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06461709824125597,
+      "learning_rate": 1.213794113340167e-05,
+      "loss": 1.4305,
+      "step": 5908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06893003288757156,
+      "learning_rate": 1.2124762044223167e-05,
+      "loss": 1.4369,
+      "step": 5909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06887180432551317,
+      "learning_rate": 1.2111582955044663e-05,
+      "loss": 1.4408,
+      "step": 5910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06799252173886035,
+      "learning_rate": 1.209840386586616e-05,
+      "loss": 1.4456,
+      "step": 5911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0665605617331159,
+      "learning_rate": 1.2085224776687655e-05,
+      "loss": 1.4336,
+      "step": 5912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06758517677330712,
+      "learning_rate": 1.2072045687509154e-05,
+      "loss": 1.4402,
+      "step": 5913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06651959628646072,
+      "learning_rate": 1.205886659833065e-05,
+      "loss": 1.4363,
+      "step": 5914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06767103336260362,
+      "learning_rate": 1.2045687509152147e-05,
+      "loss": 1.4361,
+      "step": 5915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.061798143933354144,
+      "learning_rate": 1.2032508419973642e-05,
+      "loss": 1.4457,
+      "step": 5916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06336500641523708,
+      "learning_rate": 1.201932933079514e-05,
+      "loss": 1.4382,
+      "step": 5917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0686146320911596,
+      "learning_rate": 1.2006150241616635e-05,
+      "loss": 1.4336,
+      "step": 5918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06539676920054435,
+      "learning_rate": 1.1992971152438133e-05,
+      "loss": 1.4388,
+      "step": 5919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06819899273159276,
+      "learning_rate": 1.1979792063259629e-05,
+      "loss": 1.4478,
+      "step": 5920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06847934327984204,
+      "learning_rate": 1.1966612974081126e-05,
+      "loss": 1.4314,
+      "step": 5921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0688406875588286,
+      "learning_rate": 1.1953433884902621e-05,
+      "loss": 1.4412,
+      "step": 5922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06950667994159515,
+      "learning_rate": 1.1940254795724119e-05,
+      "loss": 1.451,
+      "step": 5923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0650285863158027,
+      "learning_rate": 1.1927075706545614e-05,
+      "loss": 1.4357,
+      "step": 5924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06662957426279098,
+      "learning_rate": 1.1913896617367113e-05,
+      "loss": 1.4398,
+      "step": 5925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06890887203141297,
+      "learning_rate": 1.1900717528188608e-05,
+      "loss": 1.446,
+      "step": 5926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06818039640751385,
+      "learning_rate": 1.1887538439010105e-05,
+      "loss": 1.4373,
+      "step": 5927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06781456576513409,
+      "learning_rate": 1.1874359349831601e-05,
+      "loss": 1.4251,
+      "step": 5928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.07104876283398286,
+      "learning_rate": 1.1861180260653098e-05,
+      "loss": 1.44,
+      "step": 5929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06893124934626793,
+      "learning_rate": 1.1848001171474593e-05,
+      "loss": 1.4314,
+      "step": 5930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.07468454735833618,
+      "learning_rate": 1.1834822082296092e-05,
+      "loss": 1.4314,
+      "step": 5931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06938706636953508,
+      "learning_rate": 1.1821642993117588e-05,
+      "loss": 1.437,
+      "step": 5932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06950295261386742,
+      "learning_rate": 1.1808463903939085e-05,
+      "loss": 1.4414,
+      "step": 5933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06370890227150129,
+      "learning_rate": 1.179528481476058e-05,
+      "loss": 1.4362,
+      "step": 5934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06911000319409191,
+      "learning_rate": 1.1782105725582077e-05,
+      "loss": 1.4349,
+      "step": 5935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06337337600947157,
+      "learning_rate": 1.1768926636403573e-05,
+      "loss": 1.4296,
+      "step": 5936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06830788985438378,
+      "learning_rate": 1.1755747547225072e-05,
+      "loss": 1.4412,
+      "step": 5937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0626086279539502,
+      "learning_rate": 1.1742568458046567e-05,
+      "loss": 1.4367,
+      "step": 5938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06396543242106374,
+      "learning_rate": 1.1729389368868064e-05,
+      "loss": 1.4342,
+      "step": 5939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.062244896508636904,
+      "learning_rate": 1.171621027968956e-05,
+      "loss": 1.4318,
+      "step": 5940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0687217009837408,
+      "learning_rate": 1.1703031190511057e-05,
+      "loss": 1.433,
+      "step": 5941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06367765018380928,
+      "learning_rate": 1.1689852101332552e-05,
+      "loss": 1.4312,
+      "step": 5942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06949817766937069,
+      "learning_rate": 1.1676673012154051e-05,
+      "loss": 1.4444,
+      "step": 5943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06841489964482655,
+      "learning_rate": 1.1663493922975546e-05,
+      "loss": 1.4401,
+      "step": 5944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06874931680877218,
+      "learning_rate": 1.1650314833797044e-05,
+      "loss": 1.4346,
+      "step": 5945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06862554287661533,
+      "learning_rate": 1.1637135744618539e-05,
+      "loss": 1.4309,
+      "step": 5946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06625175265751038,
+      "learning_rate": 1.1623956655440036e-05,
+      "loss": 1.4354,
+      "step": 5947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06539299840250044,
+      "learning_rate": 1.1610777566261532e-05,
+      "loss": 1.437,
+      "step": 5948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06592969020780118,
+      "learning_rate": 1.159759847708303e-05,
+      "loss": 1.4304,
+      "step": 5949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06598571575651625,
+      "learning_rate": 1.1584419387904526e-05,
+      "loss": 1.4457,
+      "step": 5950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06635936757211634,
+      "learning_rate": 1.1571240298726023e-05,
+      "loss": 1.4352,
+      "step": 5951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06718347971675434,
+      "learning_rate": 1.1558061209547518e-05,
+      "loss": 1.444,
+      "step": 5952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.07017445914125092,
+      "learning_rate": 1.1544882120369016e-05,
+      "loss": 1.4292,
+      "step": 5953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.07154600157168409,
+      "learning_rate": 1.1531703031190511e-05,
+      "loss": 1.4371,
+      "step": 5954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06273677644643572,
+      "learning_rate": 1.151852394201201e-05,
+      "loss": 1.4395,
+      "step": 5955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06882395831604127,
+      "learning_rate": 1.1505344852833504e-05,
+      "loss": 1.436,
+      "step": 5956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06520356967769371,
+      "learning_rate": 1.1492165763655002e-05,
+      "loss": 1.4402,
+      "step": 5957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06743939394203598,
+      "learning_rate": 1.1478986674476498e-05,
+      "loss": 1.4389,
+      "step": 5958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06673544609569626,
+      "learning_rate": 1.1465807585297995e-05,
+      "loss": 1.4292,
+      "step": 5959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06478964309657191,
+      "learning_rate": 1.145262849611949e-05,
+      "loss": 1.4361,
+      "step": 5960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06572930956287792,
+      "learning_rate": 1.143944940694099e-05,
+      "loss": 1.4358,
+      "step": 5961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06652919265710448,
+      "learning_rate": 1.1426270317762483e-05,
+      "loss": 1.4312,
+      "step": 5962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06729482300243306,
+      "learning_rate": 1.1413091228583982e-05,
+      "loss": 1.4426,
+      "step": 5963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06418739916537965,
+      "learning_rate": 1.1399912139405477e-05,
+      "loss": 1.4321,
+      "step": 5964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06560871267139573,
+      "learning_rate": 1.1386733050226974e-05,
+      "loss": 1.443,
+      "step": 5965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06686769490131236,
+      "learning_rate": 1.137355396104847e-05,
+      "loss": 1.4384,
+      "step": 5966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06425448366281784,
+      "learning_rate": 1.1360374871869969e-05,
+      "loss": 1.4428,
+      "step": 5967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06709038031403763,
+      "learning_rate": 1.1347195782691462e-05,
+      "loss": 1.4371,
+      "step": 5968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06104745045521677,
+      "learning_rate": 1.1334016693512961e-05,
+      "loss": 1.4362,
+      "step": 5969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0694248508660351,
+      "learning_rate": 1.1320837604334457e-05,
+      "loss": 1.4343,
+      "step": 5970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06515257097422994,
+      "learning_rate": 1.1307658515155954e-05,
+      "loss": 1.4417,
+      "step": 5971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0648651904628546,
+      "learning_rate": 1.129447942597745e-05,
+      "loss": 1.4419,
+      "step": 5972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06863138293875976,
+      "learning_rate": 1.1281300336798948e-05,
+      "loss": 1.4408,
+      "step": 5973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06209999107873858,
+      "learning_rate": 1.1268121247620442e-05,
+      "loss": 1.4358,
+      "step": 5974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.06400539269193384,
+      "learning_rate": 1.125494215844194e-05,
+      "loss": 1.4448,
+      "step": 5975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06434142030718859,
+      "learning_rate": 1.1241763069263436e-05,
+      "loss": 1.4345,
+      "step": 5976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06298945584720195,
+      "learning_rate": 1.1228583980084931e-05,
+      "loss": 1.4468,
+      "step": 5977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06472069900831948,
+      "learning_rate": 1.1215404890906429e-05,
+      "loss": 1.4303,
+      "step": 5978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06422542712359172,
+      "learning_rate": 1.1202225801727926e-05,
+      "loss": 1.437,
+      "step": 5979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06933068051705961,
+      "learning_rate": 1.1189046712549421e-05,
+      "loss": 1.4422,
+      "step": 5980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06205150040546381,
+      "learning_rate": 1.1175867623370918e-05,
+      "loss": 1.4352,
+      "step": 5981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06436086639347746,
+      "learning_rate": 1.1162688534192415e-05,
+      "loss": 1.4377,
+      "step": 5982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06553912419402443,
+      "learning_rate": 1.114950944501391e-05,
+      "loss": 1.4324,
+      "step": 5983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06986601412328841,
+      "learning_rate": 1.1136330355835408e-05,
+      "loss": 1.4282,
+      "step": 5984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.061111145232567825,
+      "learning_rate": 1.1123151266656905e-05,
+      "loss": 1.4354,
+      "step": 5985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06578236799601206,
+      "learning_rate": 1.11099721774784e-05,
+      "loss": 1.4423,
+      "step": 5986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06070886593657367,
+      "learning_rate": 1.1096793088299898e-05,
+      "loss": 1.4362,
+      "step": 5987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06490164647791355,
+      "learning_rate": 1.1083613999121395e-05,
+      "loss": 1.4391,
+      "step": 5988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06889817414896807,
+      "learning_rate": 1.107043490994289e-05,
+      "loss": 1.4315,
+      "step": 5989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06258961928752997,
+      "learning_rate": 1.1057255820764387e-05,
+      "loss": 1.4304,
+      "step": 5990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06391317427585395,
+      "learning_rate": 1.1044076731585884e-05,
+      "loss": 1.445,
+      "step": 5991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06969115416056029,
+      "learning_rate": 1.103089764240738e-05,
+      "loss": 1.4316,
+      "step": 5992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06656191615375655,
+      "learning_rate": 1.1017718553228877e-05,
+      "loss": 1.4388,
+      "step": 5993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0677271125525193,
+      "learning_rate": 1.1004539464050374e-05,
+      "loss": 1.4389,
+      "step": 5994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06615562212095258,
+      "learning_rate": 1.099136037487187e-05,
+      "loss": 1.4426,
+      "step": 5995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06994430250733294,
+      "learning_rate": 1.0978181285693367e-05,
+      "loss": 1.4348,
+      "step": 5996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.062428745527894416,
+      "learning_rate": 1.0965002196514864e-05,
+      "loss": 1.4349,
+      "step": 5997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06443541890165502,
+      "learning_rate": 1.095182310733636e-05,
+      "loss": 1.4402,
+      "step": 5998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06621499122485648,
+      "learning_rate": 1.0938644018157856e-05,
+      "loss": 1.4409,
+      "step": 5999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06822943106869231,
+      "learning_rate": 1.0925464928979354e-05,
+      "loss": 1.4417,
+      "step": 6000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06515911938307899,
+      "learning_rate": 1.0912285839800849e-05,
+      "loss": 1.4383,
+      "step": 6001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06790437699682043,
+      "learning_rate": 1.0899106750622346e-05,
+      "loss": 1.4387,
+      "step": 6002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06500726480080386,
+      "learning_rate": 1.0885927661443843e-05,
+      "loss": 1.4316,
+      "step": 6003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06631786127026709,
+      "learning_rate": 1.0872748572265339e-05,
+      "loss": 1.434,
+      "step": 6004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06409895523848612,
+      "learning_rate": 1.0859569483086836e-05,
+      "loss": 1.431,
+      "step": 6005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06532660306326486,
+      "learning_rate": 1.0846390393908333e-05,
+      "loss": 1.4281,
+      "step": 6006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06428842521444966,
+      "learning_rate": 1.0833211304729828e-05,
+      "loss": 1.4342,
+      "step": 6007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06706499556729344,
+      "learning_rate": 1.0820032215551326e-05,
+      "loss": 1.4375,
+      "step": 6008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06524698040831797,
+      "learning_rate": 1.0806853126372823e-05,
+      "loss": 1.4335,
+      "step": 6009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06568370479378922,
+      "learning_rate": 1.0793674037194318e-05,
+      "loss": 1.4324,
+      "step": 6010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06503883461368082,
+      "learning_rate": 1.0780494948015815e-05,
+      "loss": 1.431,
+      "step": 6011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06577690379209912,
+      "learning_rate": 1.0767315858837312e-05,
+      "loss": 1.4417,
+      "step": 6012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06671869360349715,
+      "learning_rate": 1.0754136769658808e-05,
+      "loss": 1.4375,
+      "step": 6013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06660588792758845,
+      "learning_rate": 1.0740957680480305e-05,
+      "loss": 1.4377,
+      "step": 6014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06478192201329919,
+      "learning_rate": 1.0727778591301802e-05,
+      "loss": 1.4325,
+      "step": 6015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06453358181699449,
+      "learning_rate": 1.0714599502123297e-05,
+      "loss": 1.4292,
+      "step": 6016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06768921012462091,
+      "learning_rate": 1.0701420412944795e-05,
+      "loss": 1.4478,
+      "step": 6017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06396025049247851,
+      "learning_rate": 1.0688241323766292e-05,
+      "loss": 1.4364,
+      "step": 6018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06433003334414848,
+      "learning_rate": 1.0675062234587787e-05,
+      "loss": 1.4393,
+      "step": 6019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06832896214062094,
+      "learning_rate": 1.0661883145409284e-05,
+      "loss": 1.438,
+      "step": 6020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06769184811658314,
+      "learning_rate": 1.0648704056230781e-05,
+      "loss": 1.4253,
+      "step": 6021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06522744693357006,
+      "learning_rate": 1.0635524967052277e-05,
+      "loss": 1.4417,
+      "step": 6022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06447730448066029,
+      "learning_rate": 1.0622345877873774e-05,
+      "loss": 1.4351,
+      "step": 6023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.07405153952827405,
+      "learning_rate": 1.0609166788695271e-05,
+      "loss": 1.4382,
+      "step": 6024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06104121535904047,
+      "learning_rate": 1.0595987699516767e-05,
+      "loss": 1.4324,
+      "step": 6025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06643540892850965,
+      "learning_rate": 1.0582808610338264e-05,
+      "loss": 1.4395,
+      "step": 6026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06760098688669318,
+      "learning_rate": 1.056962952115976e-05,
+      "loss": 1.4406,
+      "step": 6027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06604577955810581,
+      "learning_rate": 1.0556450431981256e-05,
+      "loss": 1.4407,
+      "step": 6028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06508491766023883,
+      "learning_rate": 1.0543271342802753e-05,
+      "loss": 1.4268,
+      "step": 6029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06406974259478121,
+      "learning_rate": 1.053009225362425e-05,
+      "loss": 1.437,
+      "step": 6030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06437814283048654,
+      "learning_rate": 1.0516913164445746e-05,
+      "loss": 1.438,
+      "step": 6031
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06654005113189997,
+      "learning_rate": 1.0503734075267243e-05,
+      "loss": 1.4272,
+      "step": 6032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06348674054041376,
+      "learning_rate": 1.049055498608874e-05,
+      "loss": 1.4466,
+      "step": 6033
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06540639708330451,
+      "learning_rate": 1.0477375896910236e-05,
+      "loss": 1.433,
+      "step": 6034
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06706264777051127,
+      "learning_rate": 1.0464196807731733e-05,
+      "loss": 1.4395,
+      "step": 6035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06796348842741971,
+      "learning_rate": 1.045101771855323e-05,
+      "loss": 1.4328,
+      "step": 6036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06561828127147891,
+      "learning_rate": 1.0437838629374725e-05,
+      "loss": 1.4389,
+      "step": 6037
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06485227049115559,
+      "learning_rate": 1.0424659540196222e-05,
+      "loss": 1.4357,
+      "step": 6038
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06805896258380968,
+      "learning_rate": 1.041148045101772e-05,
+      "loss": 1.4379,
+      "step": 6039
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06555098607017713,
+      "learning_rate": 1.0398301361839215e-05,
+      "loss": 1.4311,
+      "step": 6040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0645628700280191,
+      "learning_rate": 1.0385122272660712e-05,
+      "loss": 1.4436,
+      "step": 6041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06544778818963852,
+      "learning_rate": 1.037194318348221e-05,
+      "loss": 1.4357,
+      "step": 6042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06701861860015688,
+      "learning_rate": 1.0358764094303705e-05,
+      "loss": 1.433,
+      "step": 6043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.06486091737028554,
+      "learning_rate": 1.0345585005125202e-05,
+      "loss": 1.4358,
+      "step": 6044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06608347847812067,
+      "learning_rate": 1.0332405915946699e-05,
+      "loss": 1.4389,
+      "step": 6045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06414643833339287,
+      "learning_rate": 1.0319226826768194e-05,
+      "loss": 1.4432,
+      "step": 6046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.062128494453997755,
+      "learning_rate": 1.0306047737589692e-05,
+      "loss": 1.4484,
+      "step": 6047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06513308004580041,
+      "learning_rate": 1.0292868648411189e-05,
+      "loss": 1.4385,
+      "step": 6048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061323053537994436,
+      "learning_rate": 1.0279689559232684e-05,
+      "loss": 1.4353,
+      "step": 6049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06515871747152772,
+      "learning_rate": 1.0266510470054181e-05,
+      "loss": 1.4363,
+      "step": 6050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06454791428279974,
+      "learning_rate": 1.0253331380875678e-05,
+      "loss": 1.434,
+      "step": 6051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06728981149112225,
+      "learning_rate": 1.0240152291697174e-05,
+      "loss": 1.4317,
+      "step": 6052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06370011145192643,
+      "learning_rate": 1.0226973202518671e-05,
+      "loss": 1.4361,
+      "step": 6053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06715126621857026,
+      "learning_rate": 1.0213794113340168e-05,
+      "loss": 1.4407,
+      "step": 6054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0646430860979474,
+      "learning_rate": 1.0200615024161664e-05,
+      "loss": 1.4436,
+      "step": 6055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06316390799725154,
+      "learning_rate": 1.018743593498316e-05,
+      "loss": 1.4291,
+      "step": 6056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.07033774624862169,
+      "learning_rate": 1.0174256845804658e-05,
+      "loss": 1.4368,
+      "step": 6057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0621850615680952,
+      "learning_rate": 1.0161077756626153e-05,
+      "loss": 1.4348,
+      "step": 6058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.066590709757892,
+      "learning_rate": 1.014789866744765e-05,
+      "loss": 1.4244,
+      "step": 6059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06623524218564149,
+      "learning_rate": 1.0134719578269146e-05,
+      "loss": 1.4395,
+      "step": 6060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06492392962069761,
+      "learning_rate": 1.0121540489090643e-05,
+      "loss": 1.4357,
+      "step": 6061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06756573614545423,
+      "learning_rate": 1.010836139991214e-05,
+      "loss": 1.4391,
+      "step": 6062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06715544187539339,
+      "learning_rate": 1.0095182310733635e-05,
+      "loss": 1.4388,
+      "step": 6063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06801751734390689,
+      "learning_rate": 1.0082003221555133e-05,
+      "loss": 1.4333,
+      "step": 6064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0669655303412041,
+      "learning_rate": 1.006882413237663e-05,
+      "loss": 1.4409,
+      "step": 6065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06279024830981697,
+      "learning_rate": 1.0055645043198125e-05,
+      "loss": 1.442,
+      "step": 6066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06984753823640102,
+      "learning_rate": 1.0042465954019622e-05,
+      "loss": 1.4343,
+      "step": 6067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06251631509365348,
+      "learning_rate": 1.002928686484112e-05,
+      "loss": 1.4398,
+      "step": 6068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.07011090954183213,
+      "learning_rate": 1.0016107775662615e-05,
+      "loss": 1.4304,
+      "step": 6069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06387009198346798,
+      "learning_rate": 1.0002928686484112e-05,
+      "loss": 1.431,
+      "step": 6070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06969584641813628,
+      "learning_rate": 9.989749597305609e-06,
+      "loss": 1.4418,
+      "step": 6071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06645735914865665,
+      "learning_rate": 9.976570508127105e-06,
+      "loss": 1.4355,
+      "step": 6072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06474449791548188,
+      "learning_rate": 9.963391418948602e-06,
+      "loss": 1.4363,
+      "step": 6073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061232225893339765,
+      "learning_rate": 9.950212329770099e-06,
+      "loss": 1.4381,
+      "step": 6074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06575473422028738,
+      "learning_rate": 9.937033240591594e-06,
+      "loss": 1.4307,
+      "step": 6075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06545690336661897,
+      "learning_rate": 9.923854151413091e-06,
+      "loss": 1.4359,
+      "step": 6076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06597216943229298,
+      "learning_rate": 9.910675062234589e-06,
+      "loss": 1.4376,
+      "step": 6077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06499316040625325,
+      "learning_rate": 9.897495973056084e-06,
+      "loss": 1.4423,
+      "step": 6078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06813434696075031,
+      "learning_rate": 9.884316883877581e-06,
+      "loss": 1.4279,
+      "step": 6079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0634451208260268,
+      "learning_rate": 9.871137794699078e-06,
+      "loss": 1.4332,
+      "step": 6080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061622873967822574,
+      "learning_rate": 9.857958705520574e-06,
+      "loss": 1.4282,
+      "step": 6081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06790710556954899,
+      "learning_rate": 9.84477961634207e-06,
+      "loss": 1.4389,
+      "step": 6082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06204086318806439,
+      "learning_rate": 9.831600527163568e-06,
+      "loss": 1.4436,
+      "step": 6083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061821499704648966,
+      "learning_rate": 9.818421437985063e-06,
+      "loss": 1.4411,
+      "step": 6084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0660314972183723,
+      "learning_rate": 9.80524234880656e-06,
+      "loss": 1.4357,
+      "step": 6085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06427514554417893,
+      "learning_rate": 9.792063259628058e-06,
+      "loss": 1.4389,
+      "step": 6086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06575206221367454,
+      "learning_rate": 9.778884170449553e-06,
+      "loss": 1.4287,
+      "step": 6087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06506175780162182,
+      "learning_rate": 9.76570508127105e-06,
+      "loss": 1.4334,
+      "step": 6088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0669006812522229,
+      "learning_rate": 9.752525992092547e-06,
+      "loss": 1.4352,
+      "step": 6089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06376640234740956,
+      "learning_rate": 9.739346902914043e-06,
+      "loss": 1.4446,
+      "step": 6090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06436742639787939,
+      "learning_rate": 9.72616781373554e-06,
+      "loss": 1.4346,
+      "step": 6091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06583480099248624,
+      "learning_rate": 9.712988724557037e-06,
+      "loss": 1.4267,
+      "step": 6092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06309111119407648,
+      "learning_rate": 9.699809635378532e-06,
+      "loss": 1.4473,
+      "step": 6093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06586849081191586,
+      "learning_rate": 9.68663054620003e-06,
+      "loss": 1.43,
+      "step": 6094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06401069007305492,
+      "learning_rate": 9.673451457021527e-06,
+      "loss": 1.4318,
+      "step": 6095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06520101827452124,
+      "learning_rate": 9.660272367843022e-06,
+      "loss": 1.4321,
+      "step": 6096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06127194378657592,
+      "learning_rate": 9.64709327866452e-06,
+      "loss": 1.4357,
+      "step": 6097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06308046247433378,
+      "learning_rate": 9.633914189486016e-06,
+      "loss": 1.4401,
+      "step": 6098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06425488431617672,
+      "learning_rate": 9.620735100307512e-06,
+      "loss": 1.4362,
+      "step": 6099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06374726194150748,
+      "learning_rate": 9.607556011129009e-06,
+      "loss": 1.4276,
+      "step": 6100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06064823639790571,
+      "learning_rate": 9.594376921950506e-06,
+      "loss": 1.4363,
+      "step": 6101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06313811138589377,
+      "learning_rate": 9.581197832772002e-06,
+      "loss": 1.4373,
+      "step": 6102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06415312815091533,
+      "learning_rate": 9.568018743593499e-06,
+      "loss": 1.4319,
+      "step": 6103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06196266184953971,
+      "learning_rate": 9.554839654414996e-06,
+      "loss": 1.4327,
+      "step": 6104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0616041326030903,
+      "learning_rate": 9.541660565236491e-06,
+      "loss": 1.4308,
+      "step": 6105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06327506715684017,
+      "learning_rate": 9.528481476057988e-06,
+      "loss": 1.4344,
+      "step": 6106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061771970530746796,
+      "learning_rate": 9.515302386879485e-06,
+      "loss": 1.43,
+      "step": 6107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06272927602274427,
+      "learning_rate": 9.502123297700981e-06,
+      "loss": 1.4227,
+      "step": 6108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06321836144757215,
+      "learning_rate": 9.488944208522478e-06,
+      "loss": 1.4302,
+      "step": 6109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.061794422765590995,
+      "learning_rate": 9.475765119343975e-06,
+      "loss": 1.4299,
+      "step": 6110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.062435828917456186,
+      "learning_rate": 9.46258603016547e-06,
+      "loss": 1.4424,
+      "step": 6111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.06500891091987575,
+      "learning_rate": 9.449406940986968e-06,
+      "loss": 1.4345,
+      "step": 6112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.060097349895352,
+      "learning_rate": 9.436227851808465e-06,
+      "loss": 1.435,
+      "step": 6113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06575259675807373,
+      "learning_rate": 9.42304876262996e-06,
+      "loss": 1.4393,
+      "step": 6114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06250752139262115,
+      "learning_rate": 9.409869673451457e-06,
+      "loss": 1.4325,
+      "step": 6115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06278364024229915,
+      "learning_rate": 9.396690584272955e-06,
+      "loss": 1.4382,
+      "step": 6116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06265863486023963,
+      "learning_rate": 9.38351149509445e-06,
+      "loss": 1.432,
+      "step": 6117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06699204838066128,
+      "learning_rate": 9.370332405915947e-06,
+      "loss": 1.4347,
+      "step": 6118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06444266291951972,
+      "learning_rate": 9.357153316737444e-06,
+      "loss": 1.4345,
+      "step": 6119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06385543531624617,
+      "learning_rate": 9.34397422755894e-06,
+      "loss": 1.4298,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06351929817467228,
+      "learning_rate": 9.330795138380437e-06,
+      "loss": 1.4434,
+      "step": 6121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06469464229502106,
+      "learning_rate": 9.317616049201934e-06,
+      "loss": 1.4372,
+      "step": 6122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.064590123404273,
+      "learning_rate": 9.30443696002343e-06,
+      "loss": 1.4311,
+      "step": 6123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06314075910848023,
+      "learning_rate": 9.291257870844927e-06,
+      "loss": 1.4381,
+      "step": 6124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06212195632331721,
+      "learning_rate": 9.278078781666424e-06,
+      "loss": 1.4379,
+      "step": 6125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06795163365783068,
+      "learning_rate": 9.264899692487919e-06,
+      "loss": 1.4393,
+      "step": 6126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06164986976765548,
+      "learning_rate": 9.251720603309416e-06,
+      "loss": 1.4393,
+      "step": 6127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0654280442560464,
+      "learning_rate": 9.238541514130913e-06,
+      "loss": 1.4319,
+      "step": 6128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06407266079216027,
+      "learning_rate": 9.225362424952409e-06,
+      "loss": 1.4368,
+      "step": 6129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06421902887334512,
+      "learning_rate": 9.212183335773906e-06,
+      "loss": 1.4381,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06658696863040897,
+      "learning_rate": 9.199004246595403e-06,
+      "loss": 1.4297,
+      "step": 6131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06381201909488832,
+      "learning_rate": 9.185825157416898e-06,
+      "loss": 1.4322,
+      "step": 6132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05983444048502091,
+      "learning_rate": 9.172646068238396e-06,
+      "loss": 1.4346,
+      "step": 6133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06984964931795609,
+      "learning_rate": 9.159466979059893e-06,
+      "loss": 1.4418,
+      "step": 6134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06007514266066988,
+      "learning_rate": 9.146287889881388e-06,
+      "loss": 1.4341,
+      "step": 6135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06616622592737174,
+      "learning_rate": 9.133108800702885e-06,
+      "loss": 1.432,
+      "step": 6136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06815839525628971,
+      "learning_rate": 9.119929711524382e-06,
+      "loss": 1.4452,
+      "step": 6137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.062404815309564377,
+      "learning_rate": 9.106750622345878e-06,
+      "loss": 1.43,
+      "step": 6138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07450242693120039,
+      "learning_rate": 9.093571533167375e-06,
+      "loss": 1.4386,
+      "step": 6139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06048013246391036,
+      "learning_rate": 9.080392443988872e-06,
+      "loss": 1.4225,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06971113301344627,
+      "learning_rate": 9.067213354810368e-06,
+      "loss": 1.4349,
+      "step": 6141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06359416083689612,
+      "learning_rate": 9.054034265631865e-06,
+      "loss": 1.4365,
+      "step": 6142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06184559818773727,
+      "learning_rate": 9.040855176453362e-06,
+      "loss": 1.4268,
+      "step": 6143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06652310760310479,
+      "learning_rate": 9.027676087274857e-06,
+      "loss": 1.4266,
+      "step": 6144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06163372554860876,
+      "learning_rate": 9.014496998096354e-06,
+      "loss": 1.4368,
+      "step": 6145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06189088203784553,
+      "learning_rate": 9.001317908917852e-06,
+      "loss": 1.4309,
+      "step": 6146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06285246602847477,
+      "learning_rate": 8.988138819739347e-06,
+      "loss": 1.4297,
+      "step": 6147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05896603906493099,
+      "learning_rate": 8.974959730560844e-06,
+      "loss": 1.429,
+      "step": 6148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06501435439793256,
+      "learning_rate": 8.961780641382341e-06,
+      "loss": 1.4386,
+      "step": 6149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06205258582589413,
+      "learning_rate": 8.948601552203837e-06,
+      "loss": 1.4302,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06010890157854434,
+      "learning_rate": 8.935422463025334e-06,
+      "loss": 1.4346,
+      "step": 6151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.07039761346209718,
+      "learning_rate": 8.922243373846831e-06,
+      "loss": 1.4413,
+      "step": 6152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.060336854141732205,
+      "learning_rate": 8.909064284668326e-06,
+      "loss": 1.4342,
+      "step": 6153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06344690805197455,
+      "learning_rate": 8.895885195489823e-06,
+      "loss": 1.4305,
+      "step": 6154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06565102981199115,
+      "learning_rate": 8.88270610631132e-06,
+      "loss": 1.4376,
+      "step": 6155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05866961818315174,
+      "learning_rate": 8.869527017132816e-06,
+      "loss": 1.434,
+      "step": 6156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06705594873781512,
+      "learning_rate": 8.856347927954313e-06,
+      "loss": 1.4375,
+      "step": 6157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06464064400850375,
+      "learning_rate": 8.84316883877581e-06,
+      "loss": 1.4385,
+      "step": 6158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06301912018849326,
+      "learning_rate": 8.829989749597306e-06,
+      "loss": 1.44,
+      "step": 6159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06440049587463827,
+      "learning_rate": 8.816810660418803e-06,
+      "loss": 1.4364,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06275955512478563,
+      "learning_rate": 8.8036315712403e-06,
+      "loss": 1.4352,
+      "step": 6161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06377206131841018,
+      "learning_rate": 8.790452482061795e-06,
+      "loss": 1.4333,
+      "step": 6162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06321599378164398,
+      "learning_rate": 8.777273392883293e-06,
+      "loss": 1.4287,
+      "step": 6163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06214396769022291,
+      "learning_rate": 8.76409430370479e-06,
+      "loss": 1.4293,
+      "step": 6164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06354036987609832,
+      "learning_rate": 8.750915214526285e-06,
+      "loss": 1.437,
+      "step": 6165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06139108481353918,
+      "learning_rate": 8.737736125347782e-06,
+      "loss": 1.4339,
+      "step": 6166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06201912885267105,
+      "learning_rate": 8.724557036169278e-06,
+      "loss": 1.4408,
+      "step": 6167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06684376992382816,
+      "learning_rate": 8.711377946990775e-06,
+      "loss": 1.4329,
+      "step": 6168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06424242026726204,
+      "learning_rate": 8.698198857812272e-06,
+      "loss": 1.439,
+      "step": 6169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.061338409089202066,
+      "learning_rate": 8.685019768633767e-06,
+      "loss": 1.4298,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06378833076487013,
+      "learning_rate": 8.671840679455265e-06,
+      "loss": 1.4302,
+      "step": 6171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06369580510786389,
+      "learning_rate": 8.658661590276762e-06,
+      "loss": 1.4363,
+      "step": 6172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06420392543615804,
+      "learning_rate": 8.645482501098257e-06,
+      "loss": 1.4368,
+      "step": 6173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.05992040951245675,
+      "learning_rate": 8.632303411919754e-06,
+      "loss": 1.4418,
+      "step": 6174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06226991260377558,
+      "learning_rate": 8.619124322741251e-06,
+      "loss": 1.4408,
+      "step": 6175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06224416219475911,
+      "learning_rate": 8.605945233562747e-06,
+      "loss": 1.4396,
+      "step": 6176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06122093231157885,
+      "learning_rate": 8.592766144384244e-06,
+      "loss": 1.4302,
+      "step": 6177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06265290684667044,
+      "learning_rate": 8.579587055205741e-06,
+      "loss": 1.4305,
+      "step": 6178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0628627294773095,
+      "learning_rate": 8.566407966027236e-06,
+      "loss": 1.435,
+      "step": 6179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.06554841415748917,
+      "learning_rate": 8.553228876848734e-06,
+      "loss": 1.44,
+      "step": 6180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06568567346106646,
+      "learning_rate": 8.54004978767023e-06,
+      "loss": 1.4326,
+      "step": 6181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06327530734405615,
+      "learning_rate": 8.526870698491726e-06,
+      "loss": 1.4342,
+      "step": 6182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06491169423594782,
+      "learning_rate": 8.513691609313223e-06,
+      "loss": 1.4449,
+      "step": 6183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06403774799204805,
+      "learning_rate": 8.50051252013472e-06,
+      "loss": 1.4449,
+      "step": 6184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06840629531246462,
+      "learning_rate": 8.487333430956216e-06,
+      "loss": 1.4409,
+      "step": 6185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06395314017665517,
+      "learning_rate": 8.474154341777713e-06,
+      "loss": 1.433,
+      "step": 6186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06470988344035852,
+      "learning_rate": 8.46097525259921e-06,
+      "loss": 1.4304,
+      "step": 6187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06399682686588079,
+      "learning_rate": 8.447796163420706e-06,
+      "loss": 1.4345,
+      "step": 6188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.061003551024045005,
+      "learning_rate": 8.434617074242203e-06,
+      "loss": 1.4332,
+      "step": 6189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06796908721946342,
+      "learning_rate": 8.4214379850637e-06,
+      "loss": 1.4278,
+      "step": 6190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06943226910079794,
+      "learning_rate": 8.408258895885195e-06,
+      "loss": 1.4283,
+      "step": 6191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0643855955109695,
+      "learning_rate": 8.395079806706692e-06,
+      "loss": 1.4337,
+      "step": 6192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06646614295160048,
+      "learning_rate": 8.38190071752819e-06,
+      "loss": 1.4357,
+      "step": 6193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06264037505301301,
+      "learning_rate": 8.368721628349685e-06,
+      "loss": 1.4336,
+      "step": 6194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06483772978995561,
+      "learning_rate": 8.355542539171182e-06,
+      "loss": 1.4349,
+      "step": 6195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06302262339698532,
+      "learning_rate": 8.34236344999268e-06,
+      "loss": 1.4328,
+      "step": 6196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0634742706770919,
+      "learning_rate": 8.329184360814175e-06,
+      "loss": 1.4301,
+      "step": 6197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06692024158985467,
+      "learning_rate": 8.316005271635672e-06,
+      "loss": 1.4343,
+      "step": 6198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06384508254869707,
+      "learning_rate": 8.302826182457169e-06,
+      "loss": 1.4337,
+      "step": 6199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06359195099726682,
+      "learning_rate": 8.289647093278664e-06,
+      "loss": 1.4344,
+      "step": 6200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06617733815907735,
+      "learning_rate": 8.276468004100161e-06,
+      "loss": 1.4293,
+      "step": 6201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06338136971850244,
+      "learning_rate": 8.263288914921659e-06,
+      "loss": 1.4279,
+      "step": 6202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06683860469657686,
+      "learning_rate": 8.250109825743154e-06,
+      "loss": 1.4343,
+      "step": 6203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06438655025109374,
+      "learning_rate": 8.236930736564651e-06,
+      "loss": 1.4384,
+      "step": 6204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06428304955613946,
+      "learning_rate": 8.223751647386148e-06,
+      "loss": 1.4352,
+      "step": 6205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.062494926759871505,
+      "learning_rate": 8.210572558207644e-06,
+      "loss": 1.4443,
+      "step": 6206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06250327767052558,
+      "learning_rate": 8.197393469029141e-06,
+      "loss": 1.4342,
+      "step": 6207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06892165622338925,
+      "learning_rate": 8.184214379850638e-06,
+      "loss": 1.4445,
+      "step": 6208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06175083435812131,
+      "learning_rate": 8.171035290672133e-06,
+      "loss": 1.4362,
+      "step": 6209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06362811847961415,
+      "learning_rate": 8.15785620149363e-06,
+      "loss": 1.4367,
+      "step": 6210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06398930731319037,
+      "learning_rate": 8.144677112315128e-06,
+      "loss": 1.433,
+      "step": 6211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06542154159788409,
+      "learning_rate": 8.131498023136623e-06,
+      "loss": 1.4242,
+      "step": 6212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06512025018677536,
+      "learning_rate": 8.11831893395812e-06,
+      "loss": 1.4333,
+      "step": 6213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06558734331617985,
+      "learning_rate": 8.105139844779617e-06,
+      "loss": 1.426,
+      "step": 6214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06542008861791147,
+      "learning_rate": 8.091960755601113e-06,
+      "loss": 1.4338,
+      "step": 6215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06272410502856376,
+      "learning_rate": 8.07878166642261e-06,
+      "loss": 1.4466,
+      "step": 6216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06929886513108202,
+      "learning_rate": 8.065602577244107e-06,
+      "loss": 1.4346,
+      "step": 6217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.061329125714744165,
+      "learning_rate": 8.052423488065603e-06,
+      "loss": 1.4407,
+      "step": 6218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06370547117789471,
+      "learning_rate": 8.0392443988871e-06,
+      "loss": 1.4266,
+      "step": 6219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06464283335001156,
+      "learning_rate": 8.026065309708597e-06,
+      "loss": 1.439,
+      "step": 6220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06341562324877628,
+      "learning_rate": 8.012886220530092e-06,
+      "loss": 1.438,
+      "step": 6221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06447065300565007,
+      "learning_rate": 7.99970713135159e-06,
+      "loss": 1.4475,
+      "step": 6222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06563533121605698,
+      "learning_rate": 7.986528042173086e-06,
+      "loss": 1.4352,
+      "step": 6223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06287494990297589,
+      "learning_rate": 7.973348952994582e-06,
+      "loss": 1.4324,
+      "step": 6224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06223308078821612,
+      "learning_rate": 7.960169863816079e-06,
+      "loss": 1.4393,
+      "step": 6225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06686736215901734,
+      "learning_rate": 7.946990774637576e-06,
+      "loss": 1.4373,
+      "step": 6226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06203284880998706,
+      "learning_rate": 7.933811685459072e-06,
+      "loss": 1.4259,
+      "step": 6227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06021254448393435,
+      "learning_rate": 7.920632596280569e-06,
+      "loss": 1.4365,
+      "step": 6228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.07185722310782694,
+      "learning_rate": 7.907453507102066e-06,
+      "loss": 1.4383,
+      "step": 6229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06240380659732702,
+      "learning_rate": 7.894274417923561e-06,
+      "loss": 1.4351,
+      "step": 6230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.061462909395150624,
+      "learning_rate": 7.881095328745058e-06,
+      "loss": 1.4363,
+      "step": 6231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06418139338964386,
+      "learning_rate": 7.867916239566556e-06,
+      "loss": 1.4342,
+      "step": 6232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06243098522379834,
+      "learning_rate": 7.854737150388051e-06,
+      "loss": 1.439,
+      "step": 6233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06327883667849411,
+      "learning_rate": 7.841558061209548e-06,
+      "loss": 1.4366,
+      "step": 6234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06395022269300782,
+      "learning_rate": 7.828378972031045e-06,
+      "loss": 1.4407,
+      "step": 6235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06207221485603822,
+      "learning_rate": 7.81519988285254e-06,
+      "loss": 1.431,
+      "step": 6236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06667876712759523,
+      "learning_rate": 7.802020793674038e-06,
+      "loss": 1.4333,
+      "step": 6237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.05840710604396758,
+      "learning_rate": 7.788841704495535e-06,
+      "loss": 1.4325,
+      "step": 6238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06417236089242655,
+      "learning_rate": 7.77566261531703e-06,
+      "loss": 1.4299,
+      "step": 6239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06418478050709706,
+      "learning_rate": 7.762483526138528e-06,
+      "loss": 1.431,
+      "step": 6240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.059362407027923184,
+      "learning_rate": 7.749304436960025e-06,
+      "loss": 1.4337,
+      "step": 6241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0634174714614011,
+      "learning_rate": 7.73612534778152e-06,
+      "loss": 1.437,
+      "step": 6242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.061539044160764264,
+      "learning_rate": 7.722946258603017e-06,
+      "loss": 1.4285,
+      "step": 6243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0619548826275607,
+      "learning_rate": 7.709767169424514e-06,
+      "loss": 1.4427,
+      "step": 6244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06017135504640313,
+      "learning_rate": 7.69658808024601e-06,
+      "loss": 1.435,
+      "step": 6245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06864329896072131,
+      "learning_rate": 7.683408991067507e-06,
+      "loss": 1.4376,
+      "step": 6246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06044999948270044,
+      "learning_rate": 7.670229901889004e-06,
+      "loss": 1.4301,
+      "step": 6247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.06900515037272281,
+      "learning_rate": 7.6570508127105e-06,
+      "loss": 1.4505,
+      "step": 6248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.060844541807808836,
+      "learning_rate": 7.643871723531997e-06,
+      "loss": 1.4327,
+      "step": 6249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.063978307697954,
+      "learning_rate": 7.630692634353494e-06,
+      "loss": 1.4267,
+      "step": 6250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06635501457775746,
+      "learning_rate": 7.617513545174989e-06,
+      "loss": 1.4335,
+      "step": 6251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.062452575685040514,
+      "learning_rate": 7.604334455996486e-06,
+      "loss": 1.4295,
+      "step": 6252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06363869604435912,
+      "learning_rate": 7.591155366817983e-06,
+      "loss": 1.4356,
+      "step": 6253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.062147942086659094,
+      "learning_rate": 7.577976277639479e-06,
+      "loss": 1.4342,
+      "step": 6254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0626584536616682,
+      "learning_rate": 7.564797188460976e-06,
+      "loss": 1.4347,
+      "step": 6255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06084611563006369,
+      "learning_rate": 7.551618099282472e-06,
+      "loss": 1.4284,
+      "step": 6256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06033582599305746,
+      "learning_rate": 7.538439010103969e-06,
+      "loss": 1.439,
+      "step": 6257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06155499066560376,
+      "learning_rate": 7.525259920925466e-06,
+      "loss": 1.4328,
+      "step": 6258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0606303324157998,
+      "learning_rate": 7.512080831746962e-06,
+      "loss": 1.4284,
+      "step": 6259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.062117609449529085,
+      "learning_rate": 7.498901742568458e-06,
+      "loss": 1.4298,
+      "step": 6260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06246476730303114,
+      "learning_rate": 7.485722653389955e-06,
+      "loss": 1.4249,
+      "step": 6261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05946730971465832,
+      "learning_rate": 7.472543564211452e-06,
+      "loss": 1.438,
+      "step": 6262
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.060147515904830424,
+      "learning_rate": 7.459364475032948e-06,
+      "loss": 1.441,
+      "step": 6263
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05981362446521417,
+      "learning_rate": 7.446185385854445e-06,
+      "loss": 1.4381,
+      "step": 6264
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06303790248753263,
+      "learning_rate": 7.433006296675941e-06,
+      "loss": 1.4258,
+      "step": 6265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05765619618322617,
+      "learning_rate": 7.419827207497438e-06,
+      "loss": 1.4368,
+      "step": 6266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05957752741692598,
+      "learning_rate": 7.406648118318935e-06,
+      "loss": 1.4386,
+      "step": 6267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06189955311384636,
+      "learning_rate": 7.393469029140431e-06,
+      "loss": 1.4213,
+      "step": 6268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.060277267753673446,
+      "learning_rate": 7.380289939961927e-06,
+      "loss": 1.4335,
+      "step": 6269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.059570796469992304,
+      "learning_rate": 7.3671108507834245e-06,
+      "loss": 1.4346,
+      "step": 6270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0631501855745631,
+      "learning_rate": 7.353931761604921e-06,
+      "loss": 1.4253,
+      "step": 6271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06083202985001111,
+      "learning_rate": 7.340752672426417e-06,
+      "loss": 1.427,
+      "step": 6272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.059960584467936826,
+      "learning_rate": 7.327573583247914e-06,
+      "loss": 1.4363,
+      "step": 6273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06029403804552976,
+      "learning_rate": 7.3143944940694105e-06,
+      "loss": 1.438,
+      "step": 6274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.058157432452003055,
+      "learning_rate": 7.301215404890907e-06,
+      "loss": 1.4366,
+      "step": 6275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05817647214662568,
+      "learning_rate": 7.288036315712404e-06,
+      "loss": 1.4355,
+      "step": 6276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06065265880944192,
+      "learning_rate": 7.2748572265339e-06,
+      "loss": 1.4319,
+      "step": 6277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05872382268800923,
+      "learning_rate": 7.2616781373553965e-06,
+      "loss": 1.4386,
+      "step": 6278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06349127625929062,
+      "learning_rate": 7.248499048176894e-06,
+      "loss": 1.4368,
+      "step": 6279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.061355463960109054,
+      "learning_rate": 7.23531995899839e-06,
+      "loss": 1.4385,
+      "step": 6280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06199311839638587,
+      "learning_rate": 7.222140869819886e-06,
+      "loss": 1.4312,
+      "step": 6281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06117350404999558,
+      "learning_rate": 7.208961780641383e-06,
+      "loss": 1.4372,
+      "step": 6282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.059888050314714766,
+      "learning_rate": 7.1957826914628796e-06,
+      "loss": 1.435,
+      "step": 6283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06222583966946974,
+      "learning_rate": 7.182603602284376e-06,
+      "loss": 1.4277,
+      "step": 6284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06506635973689386,
+      "learning_rate": 7.169424513105873e-06,
+      "loss": 1.4388,
+      "step": 6285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.058985864607668054,
+      "learning_rate": 7.156245423927369e-06,
+      "loss": 1.4418,
+      "step": 6286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06512348844303958,
+      "learning_rate": 7.1430663347488655e-06,
+      "loss": 1.4405,
+      "step": 6287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06366189230085452,
+      "learning_rate": 7.129887245570363e-06,
+      "loss": 1.4362,
+      "step": 6288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06395713003043214,
+      "learning_rate": 7.116708156391859e-06,
+      "loss": 1.4325,
+      "step": 6289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06678359398631033,
+      "learning_rate": 7.103529067213355e-06,
+      "loss": 1.4325,
+      "step": 6290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06760479172126554,
+      "learning_rate": 7.090349978034852e-06,
+      "loss": 1.4346,
+      "step": 6291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06529042950804774,
+      "learning_rate": 7.077170888856349e-06,
+      "loss": 1.4292,
+      "step": 6292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0643211290975194,
+      "learning_rate": 7.063991799677845e-06,
+      "loss": 1.4416,
+      "step": 6293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06538492078858188,
+      "learning_rate": 7.050812710499342e-06,
+      "loss": 1.4376,
+      "step": 6294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.062126604029742315,
+      "learning_rate": 7.037633621320838e-06,
+      "loss": 1.4363,
+      "step": 6295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06487390567596735,
+      "learning_rate": 7.024454532142335e-06,
+      "loss": 1.442,
+      "step": 6296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06256975674914898,
+      "learning_rate": 7.011275442963831e-06,
+      "loss": 1.4224,
+      "step": 6297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06038715057663583,
+      "learning_rate": 6.998096353785328e-06,
+      "loss": 1.4355,
+      "step": 6298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06427425852899962,
+      "learning_rate": 6.984917264606824e-06,
+      "loss": 1.429,
+      "step": 6299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0588445158010086,
+      "learning_rate": 6.971738175428321e-06,
+      "loss": 1.4278,
+      "step": 6300
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.4013365507125854,
+      "eval_runtime": 455.239,
+      "eval_samples_per_second": 39.573,
+      "eval_steps_per_second": 0.059,
+      "step": 6300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.062960455044957,
+      "learning_rate": 6.958559086249818e-06,
+      "loss": 1.4389,
+      "step": 6301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.060372968836952805,
+      "learning_rate": 6.945379997071314e-06,
+      "loss": 1.4396,
+      "step": 6302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.061750810961367,
+      "learning_rate": 6.93220090789281e-06,
+      "loss": 1.4306,
+      "step": 6303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06083556394404097,
+      "learning_rate": 6.9190218187143074e-06,
+      "loss": 1.4361,
+      "step": 6304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.05955963404364769,
+      "learning_rate": 6.905842729535804e-06,
+      "loss": 1.4348,
+      "step": 6305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06213170147859839,
+      "learning_rate": 6.8926636403573e-06,
+      "loss": 1.4334,
+      "step": 6306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06446792502394483,
+      "learning_rate": 6.879484551178797e-06,
+      "loss": 1.4377,
+      "step": 6307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06048451613948348,
+      "learning_rate": 6.866305462000293e-06,
+      "loss": 1.4368,
+      "step": 6308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06141311664342151,
+      "learning_rate": 6.85312637282179e-06,
+      "loss": 1.4311,
+      "step": 6309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06347062577295937,
+      "learning_rate": 6.839947283643287e-06,
+      "loss": 1.4312,
+      "step": 6310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06144456406027158,
+      "learning_rate": 6.826768194464783e-06,
+      "loss": 1.4363,
+      "step": 6311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06140560484169489,
+      "learning_rate": 6.813589105286279e-06,
+      "loss": 1.4385,
+      "step": 6312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06237109930083527,
+      "learning_rate": 6.8004100161077765e-06,
+      "loss": 1.4374,
+      "step": 6313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0599735669958614,
+      "learning_rate": 6.787230926929273e-06,
+      "loss": 1.4289,
+      "step": 6314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.06321250645825423,
+      "learning_rate": 6.774051837750769e-06,
+      "loss": 1.4443,
+      "step": 6315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.060797178393291046,
+      "learning_rate": 6.760872748572266e-06,
+      "loss": 1.4352,
+      "step": 6316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.060590987774522945,
+      "learning_rate": 6.7476936593937625e-06,
+      "loss": 1.4347,
+      "step": 6317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059793806485032984,
+      "learning_rate": 6.734514570215259e-06,
+      "loss": 1.4333,
+      "step": 6318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0611311798781099,
+      "learning_rate": 6.721335481036756e-06,
+      "loss": 1.432,
+      "step": 6319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05816318390896619,
+      "learning_rate": 6.708156391858252e-06,
+      "loss": 1.4289,
+      "step": 6320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06315445527398442,
+      "learning_rate": 6.6949773026797485e-06,
+      "loss": 1.4351,
+      "step": 6321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06011350091531714,
+      "learning_rate": 6.681798213501246e-06,
+      "loss": 1.4309,
+      "step": 6322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05999321593313511,
+      "learning_rate": 6.668619124322742e-06,
+      "loss": 1.4381,
+      "step": 6323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06268174685755258,
+      "learning_rate": 6.655440035144238e-06,
+      "loss": 1.439,
+      "step": 6324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05939965794014491,
+      "learning_rate": 6.642260945965735e-06,
+      "loss": 1.4296,
+      "step": 6325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06157425962518378,
+      "learning_rate": 6.629081856787232e-06,
+      "loss": 1.4347,
+      "step": 6326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06260292553708877,
+      "learning_rate": 6.615902767608728e-06,
+      "loss": 1.4274,
+      "step": 6327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060923202543837374,
+      "learning_rate": 6.602723678430225e-06,
+      "loss": 1.4349,
+      "step": 6328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.057431647034082024,
+      "learning_rate": 6.589544589251721e-06,
+      "loss": 1.4305,
+      "step": 6329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06449015360527569,
+      "learning_rate": 6.5763655000732176e-06,
+      "loss": 1.4363,
+      "step": 6330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06351860783694457,
+      "learning_rate": 6.563186410894715e-06,
+      "loss": 1.4278,
+      "step": 6331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06078945927514134,
+      "learning_rate": 6.550007321716211e-06,
+      "loss": 1.438,
+      "step": 6332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06476866130162616,
+      "learning_rate": 6.536828232537707e-06,
+      "loss": 1.4334,
+      "step": 6333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05949053065401424,
+      "learning_rate": 6.523649143359204e-06,
+      "loss": 1.435,
+      "step": 6334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.063984126350644,
+      "learning_rate": 6.510470054180701e-06,
+      "loss": 1.4374,
+      "step": 6335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060103935634252786,
+      "learning_rate": 6.497290965002197e-06,
+      "loss": 1.4316,
+      "step": 6336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06364421895407323,
+      "learning_rate": 6.484111875823694e-06,
+      "loss": 1.4382,
+      "step": 6337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06074418125682762,
+      "learning_rate": 6.47093278664519e-06,
+      "loss": 1.4337,
+      "step": 6338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06032254481362157,
+      "learning_rate": 6.457753697466687e-06,
+      "loss": 1.4314,
+      "step": 6339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.058752888150861944,
+      "learning_rate": 6.444574608288184e-06,
+      "loss": 1.4276,
+      "step": 6340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059872624306414025,
+      "learning_rate": 6.43139551910968e-06,
+      "loss": 1.4295,
+      "step": 6341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05986572813998283,
+      "learning_rate": 6.418216429931176e-06,
+      "loss": 1.4369,
+      "step": 6342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060242789027599636,
+      "learning_rate": 6.4050373407526735e-06,
+      "loss": 1.4293,
+      "step": 6343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06062013024953972,
+      "learning_rate": 6.39185825157417e-06,
+      "loss": 1.4361,
+      "step": 6344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06017461044004338,
+      "learning_rate": 6.378679162395666e-06,
+      "loss": 1.4357,
+      "step": 6345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059572850013007946,
+      "learning_rate": 6.365500073217163e-06,
+      "loss": 1.4335,
+      "step": 6346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05984575244899501,
+      "learning_rate": 6.3523209840386595e-06,
+      "loss": 1.4318,
+      "step": 6347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06232193436860319,
+      "learning_rate": 6.339141894860156e-06,
+      "loss": 1.4352,
+      "step": 6348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05942796779665714,
+      "learning_rate": 6.325962805681653e-06,
+      "loss": 1.4253,
+      "step": 6349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06085267024984417,
+      "learning_rate": 6.312783716503149e-06,
+      "loss": 1.4281,
+      "step": 6350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06464734931797352,
+      "learning_rate": 6.2996046273246454e-06,
+      "loss": 1.4369,
+      "step": 6351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05947329761668079,
+      "learning_rate": 6.286425538146142e-06,
+      "loss": 1.4401,
+      "step": 6352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06455417145987827,
+      "learning_rate": 6.273246448967639e-06,
+      "loss": 1.4307,
+      "step": 6353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06268779244148391,
+      "learning_rate": 6.260067359789135e-06,
+      "loss": 1.4401,
+      "step": 6354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0643223568673735,
+      "learning_rate": 6.2468882706106314e-06,
+      "loss": 1.4369,
+      "step": 6355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0659897786555212,
+      "learning_rate": 6.2337091814321286e-06,
+      "loss": 1.432,
+      "step": 6356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05992601683291393,
+      "learning_rate": 6.220530092253625e-06,
+      "loss": 1.4348,
+      "step": 6357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06285528852025891,
+      "learning_rate": 6.207351003075121e-06,
+      "loss": 1.4312,
+      "step": 6358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06250083626595948,
+      "learning_rate": 6.194171913896618e-06,
+      "loss": 1.4301,
+      "step": 6359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06184270170987738,
+      "learning_rate": 6.1809928247181145e-06,
+      "loss": 1.4305,
+      "step": 6360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0595763920636672,
+      "learning_rate": 6.167813735539611e-06,
+      "loss": 1.4274,
+      "step": 6361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0638042785069529,
+      "learning_rate": 6.154634646361108e-06,
+      "loss": 1.4369,
+      "step": 6362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06029466743389012,
+      "learning_rate": 6.141455557182604e-06,
+      "loss": 1.4375,
+      "step": 6363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06026946801244364,
+      "learning_rate": 6.1282764680041005e-06,
+      "loss": 1.4302,
+      "step": 6364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.061841008445221654,
+      "learning_rate": 6.115097378825598e-06,
+      "loss": 1.4367,
+      "step": 6365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0595052250559155,
+      "learning_rate": 6.101918289647094e-06,
+      "loss": 1.4371,
+      "step": 6366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05841750142192837,
+      "learning_rate": 6.08873920046859e-06,
+      "loss": 1.4368,
+      "step": 6367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06065979876989543,
+      "learning_rate": 6.075560111290087e-06,
+      "loss": 1.4367,
+      "step": 6368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059651409025947635,
+      "learning_rate": 6.062381022111584e-06,
+      "loss": 1.4307,
+      "step": 6369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06123289058860281,
+      "learning_rate": 6.04920193293308e-06,
+      "loss": 1.4328,
+      "step": 6370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06317684055105313,
+      "learning_rate": 6.036022843754577e-06,
+      "loss": 1.4396,
+      "step": 6371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06108697906592062,
+      "learning_rate": 6.022843754576073e-06,
+      "loss": 1.4398,
+      "step": 6372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.058182395700098415,
+      "learning_rate": 6.00966466539757e-06,
+      "loss": 1.4272,
+      "step": 6373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060044478949088195,
+      "learning_rate": 5.996485576219067e-06,
+      "loss": 1.4319,
+      "step": 6374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059472950317042285,
+      "learning_rate": 5.983306487040563e-06,
+      "loss": 1.4355,
+      "step": 6375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.057531849769174914,
+      "learning_rate": 5.970127397862059e-06,
+      "loss": 1.4351,
+      "step": 6376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05871335347454856,
+      "learning_rate": 5.9569483086835564e-06,
+      "loss": 1.4262,
+      "step": 6377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.060055318675141876,
+      "learning_rate": 5.943769219505053e-06,
+      "loss": 1.439,
+      "step": 6378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059028733548699384,
+      "learning_rate": 5.930590130326549e-06,
+      "loss": 1.432,
+      "step": 6379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.057964954830235546,
+      "learning_rate": 5.917411041148046e-06,
+      "loss": 1.4321,
+      "step": 6380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06067990276807016,
+      "learning_rate": 5.904231951969542e-06,
+      "loss": 1.4306,
+      "step": 6381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.06068597941299835,
+      "learning_rate": 5.891052862791039e-06,
+      "loss": 1.43,
+      "step": 6382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.059264602027940975,
+      "learning_rate": 5.877873773612536e-06,
+      "loss": 1.4293,
+      "step": 6383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.061185122620191594,
+      "learning_rate": 5.864694684434032e-06,
+      "loss": 1.4257,
+      "step": 6384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.05754565227990208,
+      "learning_rate": 5.851515595255528e-06,
+      "loss": 1.4355,
+      "step": 6385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06942067483352427,
+      "learning_rate": 5.8383365060770255e-06,
+      "loss": 1.4334,
+      "step": 6386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06086483053857746,
+      "learning_rate": 5.825157416898522e-06,
+      "loss": 1.4398,
+      "step": 6387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.060136290510190045,
+      "learning_rate": 5.811978327720018e-06,
+      "loss": 1.4325,
+      "step": 6388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06333474049851208,
+      "learning_rate": 5.798799238541515e-06,
+      "loss": 1.4266,
+      "step": 6389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06263434232002797,
+      "learning_rate": 5.7856201493630115e-06,
+      "loss": 1.4299,
+      "step": 6390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05792035677670144,
+      "learning_rate": 5.772441060184508e-06,
+      "loss": 1.4305,
+      "step": 6391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06020440660217716,
+      "learning_rate": 5.759261971006005e-06,
+      "loss": 1.4333,
+      "step": 6392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05924875464434944,
+      "learning_rate": 5.746082881827501e-06,
+      "loss": 1.44,
+      "step": 6393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0585282964841361,
+      "learning_rate": 5.7329037926489975e-06,
+      "loss": 1.4307,
+      "step": 6394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06085569127995643,
+      "learning_rate": 5.719724703470495e-06,
+      "loss": 1.421,
+      "step": 6395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.059590270686805044,
+      "learning_rate": 5.706545614291991e-06,
+      "loss": 1.4234,
+      "step": 6396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06030367867017081,
+      "learning_rate": 5.693366525113487e-06,
+      "loss": 1.4303,
+      "step": 6397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06093875154078761,
+      "learning_rate": 5.680187435934984e-06,
+      "loss": 1.4335,
+      "step": 6398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.061680560720730616,
+      "learning_rate": 5.667008346756481e-06,
+      "loss": 1.4315,
+      "step": 6399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06065465520613908,
+      "learning_rate": 5.653829257577977e-06,
+      "loss": 1.4371,
+      "step": 6400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06015412310100039,
+      "learning_rate": 5.640650168399474e-06,
+      "loss": 1.4408,
+      "step": 6401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05864772576252199,
+      "learning_rate": 5.62747107922097e-06,
+      "loss": 1.4277,
+      "step": 6402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06049512649902905,
+      "learning_rate": 5.614291990042466e-06,
+      "loss": 1.4341,
+      "step": 6403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06042334557284515,
+      "learning_rate": 5.601112900863963e-06,
+      "loss": 1.4352,
+      "step": 6404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.060383021993253315,
+      "learning_rate": 5.587933811685459e-06,
+      "loss": 1.4393,
+      "step": 6405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05776869924949384,
+      "learning_rate": 5.574754722506955e-06,
+      "loss": 1.4371,
+      "step": 6406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06258771691916884,
+      "learning_rate": 5.5615756333284525e-06,
+      "loss": 1.4391,
+      "step": 6407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05965483881866278,
+      "learning_rate": 5.548396544149949e-06,
+      "loss": 1.4339,
+      "step": 6408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.057613830509082795,
+      "learning_rate": 5.535217454971445e-06,
+      "loss": 1.4241,
+      "step": 6409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05822323059377668,
+      "learning_rate": 5.522038365792942e-06,
+      "loss": 1.431,
+      "step": 6410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06058754677828698,
+      "learning_rate": 5.5088592766144385e-06,
+      "loss": 1.4328,
+      "step": 6411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.058518235755108484,
+      "learning_rate": 5.495680187435935e-06,
+      "loss": 1.4276,
+      "step": 6412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06079798918565879,
+      "learning_rate": 5.482501098257432e-06,
+      "loss": 1.4257,
+      "step": 6413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0573517245921447,
+      "learning_rate": 5.469322009078928e-06,
+      "loss": 1.4313,
+      "step": 6414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.061039586899693715,
+      "learning_rate": 5.4561429199004245e-06,
+      "loss": 1.4258,
+      "step": 6415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.061707301186939637,
+      "learning_rate": 5.442963830721922e-06,
+      "loss": 1.4434,
+      "step": 6416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.059496146700274026,
+      "learning_rate": 5.429784741543418e-06,
+      "loss": 1.4329,
+      "step": 6417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06164091698844467,
+      "learning_rate": 5.416605652364914e-06,
+      "loss": 1.4428,
+      "step": 6418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0594510483334285,
+      "learning_rate": 5.403426563186411e-06,
+      "loss": 1.4324,
+      "step": 6419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06161599439923712,
+      "learning_rate": 5.390247474007908e-06,
+      "loss": 1.4393,
+      "step": 6420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06260321517559571,
+      "learning_rate": 5.377068384829404e-06,
+      "loss": 1.4363,
+      "step": 6421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05860839766387973,
+      "learning_rate": 5.363889295650901e-06,
+      "loss": 1.4345,
+      "step": 6422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05847936344831541,
+      "learning_rate": 5.350710206472397e-06,
+      "loss": 1.4332,
+      "step": 6423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06014905657613253,
+      "learning_rate": 5.337531117293894e-06,
+      "loss": 1.437,
+      "step": 6424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.059542067719688514,
+      "learning_rate": 5.324352028115391e-06,
+      "loss": 1.4413,
+      "step": 6425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05766293510540477,
+      "learning_rate": 5.311172938936887e-06,
+      "loss": 1.4291,
+      "step": 6426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06304126849836976,
+      "learning_rate": 5.297993849758383e-06,
+      "loss": 1.4277,
+      "step": 6427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06030391797435429,
+      "learning_rate": 5.28481476057988e-06,
+      "loss": 1.4277,
+      "step": 6428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06054648409090382,
+      "learning_rate": 5.271635671401377e-06,
+      "loss": 1.4314,
+      "step": 6429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05914291577118935,
+      "learning_rate": 5.258456582222873e-06,
+      "loss": 1.4312,
+      "step": 6430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05925917181733758,
+      "learning_rate": 5.24527749304437e-06,
+      "loss": 1.4355,
+      "step": 6431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0598689410103994,
+      "learning_rate": 5.232098403865866e-06,
+      "loss": 1.4436,
+      "step": 6432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06393001016757424,
+      "learning_rate": 5.218919314687363e-06,
+      "loss": 1.4313,
+      "step": 6433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.060326465381481884,
+      "learning_rate": 5.20574022550886e-06,
+      "loss": 1.4393,
+      "step": 6434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.055847515778179875,
+      "learning_rate": 5.192561136330356e-06,
+      "loss": 1.4329,
+      "step": 6435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.060941589755893234,
+      "learning_rate": 5.179382047151852e-06,
+      "loss": 1.429,
+      "step": 6436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.062103887759994536,
+      "learning_rate": 5.1662029579733495e-06,
+      "loss": 1.4316,
+      "step": 6437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.057787775000608,
+      "learning_rate": 5.153023868794846e-06,
+      "loss": 1.4245,
+      "step": 6438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06113291203998644,
+      "learning_rate": 5.139844779616342e-06,
+      "loss": 1.4299,
+      "step": 6439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06113212105002214,
+      "learning_rate": 5.126665690437839e-06,
+      "loss": 1.4369,
+      "step": 6440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05818029609243109,
+      "learning_rate": 5.1134866012593355e-06,
+      "loss": 1.4302,
+      "step": 6441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.057066197121866176,
+      "learning_rate": 5.100307512080832e-06,
+      "loss": 1.4217,
+      "step": 6442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06041336928568847,
+      "learning_rate": 5.087128422902329e-06,
+      "loss": 1.4374,
+      "step": 6443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05798068246423678,
+      "learning_rate": 5.073949333723825e-06,
+      "loss": 1.4401,
+      "step": 6444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05780511020266807,
+      "learning_rate": 5.0607702445453215e-06,
+      "loss": 1.4369,
+      "step": 6445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.061839277617603806,
+      "learning_rate": 5.047591155366818e-06,
+      "loss": 1.4254,
+      "step": 6446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.058950396850539294,
+      "learning_rate": 5.034412066188315e-06,
+      "loss": 1.4294,
+      "step": 6447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05963645259475734,
+      "learning_rate": 5.021232977009811e-06,
+      "loss": 1.4341,
+      "step": 6448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0604876959224841,
+      "learning_rate": 5.0080538878313074e-06,
+      "loss": 1.4255,
+      "step": 6449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.06047012267571835,
+      "learning_rate": 4.9948747986528046e-06,
+      "loss": 1.44,
+      "step": 6450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05719921507734182,
+      "learning_rate": 4.981695709474301e-06,
+      "loss": 1.4347,
+      "step": 6451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05931913016111754,
+      "learning_rate": 4.968516620295797e-06,
+      "loss": 1.4367,
+      "step": 6452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.060921999733696934,
+      "learning_rate": 4.955337531117294e-06,
+      "loss": 1.4315,
+      "step": 6453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.05879041168600063,
+      "learning_rate": 4.9421584419387905e-06,
+      "loss": 1.4359,
+      "step": 6454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05875386466456466,
+      "learning_rate": 4.928979352760287e-06,
+      "loss": 1.4321,
+      "step": 6455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.059684768171254086,
+      "learning_rate": 4.915800263581784e-06,
+      "loss": 1.4347,
+      "step": 6456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05905030267911483,
+      "learning_rate": 4.90262117440328e-06,
+      "loss": 1.4365,
+      "step": 6457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.058017026669682795,
+      "learning_rate": 4.8894420852247765e-06,
+      "loss": 1.4329,
+      "step": 6458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06124133290958073,
+      "learning_rate": 4.876262996046274e-06,
+      "loss": 1.4343,
+      "step": 6459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05898835430106844,
+      "learning_rate": 4.86308390686777e-06,
+      "loss": 1.4362,
+      "step": 6460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.059352092219155016,
+      "learning_rate": 4.849904817689266e-06,
+      "loss": 1.4426,
+      "step": 6461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0581033131997353,
+      "learning_rate": 4.836725728510763e-06,
+      "loss": 1.4252,
+      "step": 6462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05939743935713474,
+      "learning_rate": 4.82354663933226e-06,
+      "loss": 1.4333,
+      "step": 6463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05871106638584779,
+      "learning_rate": 4.810367550153756e-06,
+      "loss": 1.4314,
+      "step": 6464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05740476293362118,
+      "learning_rate": 4.797188460975253e-06,
+      "loss": 1.4418,
+      "step": 6465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06172088016784583,
+      "learning_rate": 4.784009371796749e-06,
+      "loss": 1.4339,
+      "step": 6466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.057909915234823015,
+      "learning_rate": 4.770830282618246e-06,
+      "loss": 1.4384,
+      "step": 6467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05909712694265872,
+      "learning_rate": 4.757651193439743e-06,
+      "loss": 1.4306,
+      "step": 6468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06046728664840102,
+      "learning_rate": 4.744472104261239e-06,
+      "loss": 1.4398,
+      "step": 6469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05840145461951979,
+      "learning_rate": 4.731293015082735e-06,
+      "loss": 1.4275,
+      "step": 6470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05946848138973423,
+      "learning_rate": 4.7181139259042324e-06,
+      "loss": 1.4309,
+      "step": 6471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05953572530471102,
+      "learning_rate": 4.704934836725729e-06,
+      "loss": 1.4415,
+      "step": 6472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05920504831618106,
+      "learning_rate": 4.691755747547225e-06,
+      "loss": 1.4325,
+      "step": 6473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.058451987977440945,
+      "learning_rate": 4.678576658368722e-06,
+      "loss": 1.4383,
+      "step": 6474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06028442645181864,
+      "learning_rate": 4.665397569190218e-06,
+      "loss": 1.4454,
+      "step": 6475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05922715358066251,
+      "learning_rate": 4.652218480011715e-06,
+      "loss": 1.4358,
+      "step": 6476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05740528976382221,
+      "learning_rate": 4.639039390833212e-06,
+      "loss": 1.4332,
+      "step": 6477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05716275950065533,
+      "learning_rate": 4.625860301654708e-06,
+      "loss": 1.4312,
+      "step": 6478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06165957050457203,
+      "learning_rate": 4.612681212476204e-06,
+      "loss": 1.4239,
+      "step": 6479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05748434004763071,
+      "learning_rate": 4.5995021232977015e-06,
+      "loss": 1.427,
+      "step": 6480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06047618351463182,
+      "learning_rate": 4.586323034119198e-06,
+      "loss": 1.4319,
+      "step": 6481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05971325425320952,
+      "learning_rate": 4.573143944940694e-06,
+      "loss": 1.4366,
+      "step": 6482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0588607024340089,
+      "learning_rate": 4.559964855762191e-06,
+      "loss": 1.4314,
+      "step": 6483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0585211059828163,
+      "learning_rate": 4.5467857665836875e-06,
+      "loss": 1.4345,
+      "step": 6484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.060050806590629044,
+      "learning_rate": 4.533606677405184e-06,
+      "loss": 1.4294,
+      "step": 6485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06140283432897425,
+      "learning_rate": 4.520427588226681e-06,
+      "loss": 1.4377,
+      "step": 6486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05636775977729596,
+      "learning_rate": 4.507248499048177e-06,
+      "loss": 1.4368,
+      "step": 6487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0576809291477504,
+      "learning_rate": 4.4940694098696735e-06,
+      "loss": 1.435,
+      "step": 6488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05993007705954973,
+      "learning_rate": 4.480890320691171e-06,
+      "loss": 1.4392,
+      "step": 6489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.058928245667267676,
+      "learning_rate": 4.467711231512667e-06,
+      "loss": 1.4357,
+      "step": 6490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05692544350632601,
+      "learning_rate": 4.454532142334163e-06,
+      "loss": 1.433,
+      "step": 6491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05909830401389043,
+      "learning_rate": 4.44135305315566e-06,
+      "loss": 1.422,
+      "step": 6492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05813095903462277,
+      "learning_rate": 4.428173963977157e-06,
+      "loss": 1.4319,
+      "step": 6493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.056784720947455555,
+      "learning_rate": 4.414994874798653e-06,
+      "loss": 1.4417,
+      "step": 6494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05841264787115348,
+      "learning_rate": 4.40181578562015e-06,
+      "loss": 1.4389,
+      "step": 6495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05981292917642305,
+      "learning_rate": 4.388636696441646e-06,
+      "loss": 1.4378,
+      "step": 6496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05877016984672713,
+      "learning_rate": 4.3754576072631426e-06,
+      "loss": 1.4388,
+      "step": 6497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05820454751219376,
+      "learning_rate": 4.362278518084639e-06,
+      "loss": 1.4232,
+      "step": 6498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06087267476942221,
+      "learning_rate": 4.349099428906136e-06,
+      "loss": 1.4289,
+      "step": 6499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05955539269436869,
+      "learning_rate": 4.335920339727632e-06,
+      "loss": 1.4274,
+      "step": 6500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.057742730069847474,
+      "learning_rate": 4.3227412505491286e-06,
+      "loss": 1.4367,
+      "step": 6501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.059448277023537714,
+      "learning_rate": 4.309562161370626e-06,
+      "loss": 1.4347,
+      "step": 6502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.059258831530530645,
+      "learning_rate": 4.296383072192122e-06,
+      "loss": 1.438,
+      "step": 6503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05810695380568931,
+      "learning_rate": 4.283203983013618e-06,
+      "loss": 1.4381,
+      "step": 6504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05942124986740077,
+      "learning_rate": 4.270024893835115e-06,
+      "loss": 1.4406,
+      "step": 6505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05815761558759913,
+      "learning_rate": 4.256845804656612e-06,
+      "loss": 1.4378,
+      "step": 6506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05827606890793725,
+      "learning_rate": 4.243666715478108e-06,
+      "loss": 1.4239,
+      "step": 6507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.058549984309410504,
+      "learning_rate": 4.230487626299605e-06,
+      "loss": 1.4208,
+      "step": 6508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05785923255777377,
+      "learning_rate": 4.217308537121101e-06,
+      "loss": 1.4323,
+      "step": 6509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05944890931261336,
+      "learning_rate": 4.204129447942598e-06,
+      "loss": 1.4327,
+      "step": 6510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05709268738337199,
+      "learning_rate": 4.190950358764095e-06,
+      "loss": 1.4362,
+      "step": 6511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.057755841570702855,
+      "learning_rate": 4.177771269585591e-06,
+      "loss": 1.4313,
+      "step": 6512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.058821434692839694,
+      "learning_rate": 4.164592180407087e-06,
+      "loss": 1.4331,
+      "step": 6513
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06101792342194566,
+      "learning_rate": 4.1514130912285845e-06,
+      "loss": 1.4337,
+      "step": 6514
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05762335119903353,
+      "learning_rate": 4.138234002050081e-06,
+      "loss": 1.4339,
+      "step": 6515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05773126513647721,
+      "learning_rate": 4.125054912871577e-06,
+      "loss": 1.4301,
+      "step": 6516
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06189985721377937,
+      "learning_rate": 4.111875823693074e-06,
+      "loss": 1.4292,
+      "step": 6517
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06082408896269985,
+      "learning_rate": 4.0986967345145704e-06,
+      "loss": 1.4327,
+      "step": 6518
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.062460295752213575,
+      "learning_rate": 4.085517645336067e-06,
+      "loss": 1.4398,
+      "step": 6519
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.05592509264515622,
+      "learning_rate": 4.072338556157564e-06,
+      "loss": 1.4276,
+      "step": 6520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.06124493854334662,
+      "learning_rate": 4.05915946697906e-06,
+      "loss": 1.4379,
+      "step": 6521
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.062034789189380074,
+      "learning_rate": 4.045980377800556e-06,
+      "loss": 1.4379,
+      "step": 6522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05802652148364103,
+      "learning_rate": 4.0328012886220536e-06,
+      "loss": 1.4362,
+      "step": 6523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05962306649987125,
+      "learning_rate": 4.01962219944355e-06,
+      "loss": 1.4367,
+      "step": 6524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05944118745563274,
+      "learning_rate": 4.006443110265046e-06,
+      "loss": 1.432,
+      "step": 6525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05843067128160271,
+      "learning_rate": 3.993264021086543e-06,
+      "loss": 1.4347,
+      "step": 6526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05697179344974319,
+      "learning_rate": 3.9800849319080395e-06,
+      "loss": 1.4296,
+      "step": 6527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.057803580389223175,
+      "learning_rate": 3.966905842729536e-06,
+      "loss": 1.4348,
+      "step": 6528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05844788676243462,
+      "learning_rate": 3.953726753551033e-06,
+      "loss": 1.4346,
+      "step": 6529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05795441613405829,
+      "learning_rate": 3.940547664372529e-06,
+      "loss": 1.4284,
+      "step": 6530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05991816613542307,
+      "learning_rate": 3.9273685751940255e-06,
+      "loss": 1.4361,
+      "step": 6531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.058027308746794584,
+      "learning_rate": 3.914189486015523e-06,
+      "loss": 1.4382,
+      "step": 6532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05705332006726062,
+      "learning_rate": 3.901010396837019e-06,
+      "loss": 1.4325,
+      "step": 6533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05773928986678718,
+      "learning_rate": 3.887831307658515e-06,
+      "loss": 1.4335,
+      "step": 6534
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.059997344685739494,
+      "learning_rate": 3.874652218480012e-06,
+      "loss": 1.4313,
+      "step": 6535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.059997897231283674,
+      "learning_rate": 3.861473129301509e-06,
+      "loss": 1.4307,
+      "step": 6536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05790232166070596,
+      "learning_rate": 3.848294040123005e-06,
+      "loss": 1.437,
+      "step": 6537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0573676421321397,
+      "learning_rate": 3.835114950944502e-06,
+      "loss": 1.4293,
+      "step": 6538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05844006787838829,
+      "learning_rate": 3.821935861765998e-06,
+      "loss": 1.4343,
+      "step": 6539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05758572118453741,
+      "learning_rate": 3.8087567725874946e-06,
+      "loss": 1.4308,
+      "step": 6540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0570069728207488,
+      "learning_rate": 3.7955776834089913e-06,
+      "loss": 1.4343,
+      "step": 6541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055343688773736466,
+      "learning_rate": 3.782398594230488e-06,
+      "loss": 1.435,
+      "step": 6542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05631343306074639,
+      "learning_rate": 3.7692195050519843e-06,
+      "loss": 1.4331,
+      "step": 6543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05656610768346374,
+      "learning_rate": 3.756040415873481e-06,
+      "loss": 1.4353,
+      "step": 6544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05937066263966702,
+      "learning_rate": 3.7428613266949777e-06,
+      "loss": 1.4303,
+      "step": 6545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05578440281669389,
+      "learning_rate": 3.729682237516474e-06,
+      "loss": 1.4341,
+      "step": 6546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05732288070479703,
+      "learning_rate": 3.7165031483379707e-06,
+      "loss": 1.4329,
+      "step": 6547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05754413513390572,
+      "learning_rate": 3.7033240591594674e-06,
+      "loss": 1.428,
+      "step": 6548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05748718499786988,
+      "learning_rate": 3.6901449699809637e-06,
+      "loss": 1.4371,
+      "step": 6549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05688817306229156,
+      "learning_rate": 3.6769658808024604e-06,
+      "loss": 1.4283,
+      "step": 6550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05695152238749002,
+      "learning_rate": 3.663786791623957e-06,
+      "loss": 1.4353,
+      "step": 6551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05714074605761397,
+      "learning_rate": 3.6506077024454534e-06,
+      "loss": 1.4313,
+      "step": 6552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056921714062294924,
+      "learning_rate": 3.63742861326695e-06,
+      "loss": 1.4298,
+      "step": 6553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.058471247082576026,
+      "learning_rate": 3.624249524088447e-06,
+      "loss": 1.4276,
+      "step": 6554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.058588607618244165,
+      "learning_rate": 3.611070434909943e-06,
+      "loss": 1.4372,
+      "step": 6555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05641445490166259,
+      "learning_rate": 3.5978913457314398e-06,
+      "loss": 1.435,
+      "step": 6556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.06003281737004796,
+      "learning_rate": 3.5847122565529365e-06,
+      "loss": 1.4257,
+      "step": 6557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0574661265963837,
+      "learning_rate": 3.5715331673744328e-06,
+      "loss": 1.4319,
+      "step": 6558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056187991724706975,
+      "learning_rate": 3.5583540781959295e-06,
+      "loss": 1.4282,
+      "step": 6559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056721062763505296,
+      "learning_rate": 3.545174989017426e-06,
+      "loss": 1.432,
+      "step": 6560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055757654365913414,
+      "learning_rate": 3.5319958998389225e-06,
+      "loss": 1.4324,
+      "step": 6561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05651935627172529,
+      "learning_rate": 3.518816810660419e-06,
+      "loss": 1.4276,
+      "step": 6562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0574481345826132,
+      "learning_rate": 3.5056377214819155e-06,
+      "loss": 1.4316,
+      "step": 6563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055701409590212234,
+      "learning_rate": 3.492458632303412e-06,
+      "loss": 1.4269,
+      "step": 6564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05682272040339066,
+      "learning_rate": 3.479279543124909e-06,
+      "loss": 1.4358,
+      "step": 6565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05810451269184857,
+      "learning_rate": 3.466100453946405e-06,
+      "loss": 1.4346,
+      "step": 6566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05688465687725515,
+      "learning_rate": 3.452921364767902e-06,
+      "loss": 1.4337,
+      "step": 6567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05583254602642187,
+      "learning_rate": 3.4397422755893986e-06,
+      "loss": 1.4354,
+      "step": 6568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055608192585576754,
+      "learning_rate": 3.426563186410895e-06,
+      "loss": 1.4275,
+      "step": 6569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05776069780636299,
+      "learning_rate": 3.4133840972323916e-06,
+      "loss": 1.4411,
+      "step": 6570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055975031547654215,
+      "learning_rate": 3.4002050080538883e-06,
+      "loss": 1.4363,
+      "step": 6571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056087068589374914,
+      "learning_rate": 3.3870259188753845e-06,
+      "loss": 1.4394,
+      "step": 6572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05661335452086053,
+      "learning_rate": 3.3738468296968813e-06,
+      "loss": 1.4309,
+      "step": 6573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05759013955991416,
+      "learning_rate": 3.360667740518378e-06,
+      "loss": 1.4285,
+      "step": 6574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0569067717285498,
+      "learning_rate": 3.3474886513398742e-06,
+      "loss": 1.4333,
+      "step": 6575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056191765370738916,
+      "learning_rate": 3.334309562161371e-06,
+      "loss": 1.4328,
+      "step": 6576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.057540146610637485,
+      "learning_rate": 3.3211304729828677e-06,
+      "loss": 1.4469,
+      "step": 6577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05673449894069986,
+      "learning_rate": 3.307951383804364e-06,
+      "loss": 1.4346,
+      "step": 6578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056601332076424656,
+      "learning_rate": 3.2947722946258606e-06,
+      "loss": 1.4295,
+      "step": 6579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05658500435304479,
+      "learning_rate": 3.2815932054473574e-06,
+      "loss": 1.434,
+      "step": 6580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05769460622071621,
+      "learning_rate": 3.2684141162688536e-06,
+      "loss": 1.4208,
+      "step": 6581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05619467490896604,
+      "learning_rate": 3.2552350270903503e-06,
+      "loss": 1.4321,
+      "step": 6582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05844935411427417,
+      "learning_rate": 3.242055937911847e-06,
+      "loss": 1.4296,
+      "step": 6583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.057958324867481086,
+      "learning_rate": 3.2288768487333433e-06,
+      "loss": 1.4315,
+      "step": 6584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05782745006466921,
+      "learning_rate": 3.21569775955484e-06,
+      "loss": 1.4274,
+      "step": 6585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055952058907107305,
+      "learning_rate": 3.2025186703763367e-06,
+      "loss": 1.4383,
+      "step": 6586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.05799606190916191,
+      "learning_rate": 3.189339581197833e-06,
+      "loss": 1.423,
+      "step": 6587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.055414071516140376,
+      "learning_rate": 3.1761604920193297e-06,
+      "loss": 1.4274,
+      "step": 6588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.057403791955715384,
+      "learning_rate": 3.1629814028408264e-06,
+      "loss": 1.4272,
+      "step": 6589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.056716219937422124,
+      "learning_rate": 3.1498023136623227e-06,
+      "loss": 1.4309,
+      "step": 6590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056868207490370686,
+      "learning_rate": 3.1366232244838194e-06,
+      "loss": 1.4323,
+      "step": 6591
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05626331938803584,
+      "learning_rate": 3.1234441353053157e-06,
+      "loss": 1.4312,
+      "step": 6592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056858407174036445,
+      "learning_rate": 3.1102650461268124e-06,
+      "loss": 1.4376,
+      "step": 6593
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05750252776451947,
+      "learning_rate": 3.097085956948309e-06,
+      "loss": 1.43,
+      "step": 6594
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05776184664278512,
+      "learning_rate": 3.0839068677698054e-06,
+      "loss": 1.4298,
+      "step": 6595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05798998586914948,
+      "learning_rate": 3.070727778591302e-06,
+      "loss": 1.4312,
+      "step": 6596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.057471159601939485,
+      "learning_rate": 3.057548689412799e-06,
+      "loss": 1.4249,
+      "step": 6597
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05852696913491435,
+      "learning_rate": 3.044369600234295e-06,
+      "loss": 1.4394,
+      "step": 6598
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05776092706777341,
+      "learning_rate": 3.031190511055792e-06,
+      "loss": 1.4324,
+      "step": 6599
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05568777756645315,
+      "learning_rate": 3.0180114218772885e-06,
+      "loss": 1.4339,
+      "step": 6600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05907179731756402,
+      "learning_rate": 3.004832332698785e-06,
+      "loss": 1.4346,
+      "step": 6601
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05867273821614117,
+      "learning_rate": 2.9916532435202815e-06,
+      "loss": 1.4257,
+      "step": 6602
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.054848192445978196,
+      "learning_rate": 2.9784741543417782e-06,
+      "loss": 1.4291,
+      "step": 6603
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056321896814115015,
+      "learning_rate": 2.9652950651632745e-06,
+      "loss": 1.4314,
+      "step": 6604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05617922614116348,
+      "learning_rate": 2.952115975984771e-06,
+      "loss": 1.4364,
+      "step": 6605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05730761709968418,
+      "learning_rate": 2.938936886806268e-06,
+      "loss": 1.4378,
+      "step": 6606
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05707472488114627,
+      "learning_rate": 2.925757797627764e-06,
+      "loss": 1.4336,
+      "step": 6607
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.057570059862604585,
+      "learning_rate": 2.912578708449261e-06,
+      "loss": 1.4325,
+      "step": 6608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05689275691543308,
+      "learning_rate": 2.8993996192707576e-06,
+      "loss": 1.4284,
+      "step": 6609
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05768716443051353,
+      "learning_rate": 2.886220530092254e-06,
+      "loss": 1.4338,
+      "step": 6610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05745732922285093,
+      "learning_rate": 2.8730414409137506e-06,
+      "loss": 1.4292,
+      "step": 6611
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05756916467018062,
+      "learning_rate": 2.8598623517352473e-06,
+      "loss": 1.4353,
+      "step": 6612
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.058180097558391895,
+      "learning_rate": 2.8466832625567436e-06,
+      "loss": 1.4414,
+      "step": 6613
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05858831429193423,
+      "learning_rate": 2.8335041733782403e-06,
+      "loss": 1.4294,
+      "step": 6614
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05608763638305404,
+      "learning_rate": 2.820325084199737e-06,
+      "loss": 1.4299,
+      "step": 6615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05819839815088209,
+      "learning_rate": 2.807145995021233e-06,
+      "loss": 1.4288,
+      "step": 6616
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05929894388043529,
+      "learning_rate": 2.7939669058427296e-06,
+      "loss": 1.4351,
+      "step": 6617
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05891120740600694,
+      "learning_rate": 2.7807878166642263e-06,
+      "loss": 1.4332,
+      "step": 6618
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05574330013760471,
+      "learning_rate": 2.7676087274857226e-06,
+      "loss": 1.4319,
+      "step": 6619
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05630580571691185,
+      "learning_rate": 2.7544296383072193e-06,
+      "loss": 1.4332,
+      "step": 6620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05857122478546528,
+      "learning_rate": 2.741250549128716e-06,
+      "loss": 1.4338,
+      "step": 6621
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05811745303179609,
+      "learning_rate": 2.7280714599502122e-06,
+      "loss": 1.4357,
+      "step": 6622
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05535734278596345,
+      "learning_rate": 2.714892370771709e-06,
+      "loss": 1.4385,
+      "step": 6623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05566752504461375,
+      "learning_rate": 2.7017132815932057e-06,
+      "loss": 1.4337,
+      "step": 6624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05750418174306987,
+      "learning_rate": 2.688534192414702e-06,
+      "loss": 1.4227,
+      "step": 6625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05504003795322238,
+      "learning_rate": 2.6753551032361987e-06,
+      "loss": 1.4354,
+      "step": 6626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05499090699282595,
+      "learning_rate": 2.6621760140576954e-06,
+      "loss": 1.438,
+      "step": 6627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05642726980570827,
+      "learning_rate": 2.6489969248791916e-06,
+      "loss": 1.4305,
+      "step": 6628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05485644849768627,
+      "learning_rate": 2.6358178357006883e-06,
+      "loss": 1.4386,
+      "step": 6629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056233847315851036,
+      "learning_rate": 2.622638746522185e-06,
+      "loss": 1.4395,
+      "step": 6630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056412760043376774,
+      "learning_rate": 2.6094596573436813e-06,
+      "loss": 1.4385,
+      "step": 6631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05498318901877126,
+      "learning_rate": 2.596280568165178e-06,
+      "loss": 1.4265,
+      "step": 6632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05515702748848189,
+      "learning_rate": 2.5831014789866747e-06,
+      "loss": 1.4296,
+      "step": 6633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05689445626084397,
+      "learning_rate": 2.569922389808171e-06,
+      "loss": 1.4328,
+      "step": 6634
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05476422101962456,
+      "learning_rate": 2.5567433006296677e-06,
+      "loss": 1.432,
+      "step": 6635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.054995090957054106,
+      "learning_rate": 2.5435642114511644e-06,
+      "loss": 1.4376,
+      "step": 6636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05563823302423544,
+      "learning_rate": 2.5303851222726607e-06,
+      "loss": 1.4399,
+      "step": 6637
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05439405407120048,
+      "learning_rate": 2.5172060330941574e-06,
+      "loss": 1.426,
+      "step": 6638
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.055318392214265315,
+      "learning_rate": 2.5040269439156537e-06,
+      "loss": 1.4302,
+      "step": 6639
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05579395926930291,
+      "learning_rate": 2.4908478547371504e-06,
+      "loss": 1.4303,
+      "step": 6640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05477954071616192,
+      "learning_rate": 2.477668765558647e-06,
+      "loss": 1.4416,
+      "step": 6641
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.057934928375215064,
+      "learning_rate": 2.4644896763801434e-06,
+      "loss": 1.4326,
+      "step": 6642
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05517508141774911,
+      "learning_rate": 2.45131058720164e-06,
+      "loss": 1.4287,
+      "step": 6643
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05562580306364615,
+      "learning_rate": 2.438131498023137e-06,
+      "loss": 1.4381,
+      "step": 6644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05498062227712358,
+      "learning_rate": 2.424952408844633e-06,
+      "loss": 1.4288,
+      "step": 6645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.055264774785557046,
+      "learning_rate": 2.41177331966613e-06,
+      "loss": 1.4267,
+      "step": 6646
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.054557226891796165,
+      "learning_rate": 2.3985942304876265e-06,
+      "loss": 1.4255,
+      "step": 6647
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.054424216896889414,
+      "learning_rate": 2.385415141309123e-06,
+      "loss": 1.4362,
+      "step": 6648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05522357649648958,
+      "learning_rate": 2.3722360521306195e-06,
+      "loss": 1.4349,
+      "step": 6649
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05570973130568429,
+      "learning_rate": 2.3590569629521162e-06,
+      "loss": 1.4355,
+      "step": 6650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05890800850942938,
+      "learning_rate": 2.3458778737736125e-06,
+      "loss": 1.4349,
+      "step": 6651
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05557317474622765,
+      "learning_rate": 2.332698784595109e-06,
+      "loss": 1.4334,
+      "step": 6652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05477773657509897,
+      "learning_rate": 2.319519695416606e-06,
+      "loss": 1.4347,
+      "step": 6653
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.054808301644687134,
+      "learning_rate": 2.306340606238102e-06,
+      "loss": 1.4328,
+      "step": 6654
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056745802775319834,
+      "learning_rate": 2.293161517059599e-06,
+      "loss": 1.4405,
+      "step": 6655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.058007874697365715,
+      "learning_rate": 2.2799824278810956e-06,
+      "loss": 1.437,
+      "step": 6656
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.05612976716411118,
+      "learning_rate": 2.266803338702592e-06,
+      "loss": 1.4319,
+      "step": 6657
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.056424612003365245,
+      "learning_rate": 2.2536242495240886e-06,
+      "loss": 1.4379,
+      "step": 6658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05656484522675573,
+      "learning_rate": 2.2404451603455853e-06,
+      "loss": 1.4297,
+      "step": 6659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.057212616271806925,
+      "learning_rate": 2.2272660711670816e-06,
+      "loss": 1.4284,
+      "step": 6660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05482798129525187,
+      "learning_rate": 2.2140869819885783e-06,
+      "loss": 1.4336,
+      "step": 6661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05452685137236266,
+      "learning_rate": 2.200907892810075e-06,
+      "loss": 1.4289,
+      "step": 6662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.058396640214046794,
+      "learning_rate": 2.1877288036315713e-06,
+      "loss": 1.4352,
+      "step": 6663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05514214220817007,
+      "learning_rate": 2.174549714453068e-06,
+      "loss": 1.4353,
+      "step": 6664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.056740161695418505,
+      "learning_rate": 2.1613706252745643e-06,
+      "loss": 1.4288,
+      "step": 6665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05496884916856552,
+      "learning_rate": 2.148191536096061e-06,
+      "loss": 1.4314,
+      "step": 6666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05583234670604205,
+      "learning_rate": 2.1350124469175577e-06,
+      "loss": 1.4277,
+      "step": 6667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055533444970553486,
+      "learning_rate": 2.121833357739054e-06,
+      "loss": 1.4418,
+      "step": 6668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0544913242062098,
+      "learning_rate": 2.1086542685605507e-06,
+      "loss": 1.4368,
+      "step": 6669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0553145055635935,
+      "learning_rate": 2.0954751793820474e-06,
+      "loss": 1.4312,
+      "step": 6670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05589370636294872,
+      "learning_rate": 2.0822960902035437e-06,
+      "loss": 1.4265,
+      "step": 6671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05455202452219359,
+      "learning_rate": 2.0691170010250404e-06,
+      "loss": 1.4298,
+      "step": 6672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05613405206377541,
+      "learning_rate": 2.055937911846537e-06,
+      "loss": 1.4423,
+      "step": 6673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0605793326532818,
+      "learning_rate": 2.0427588226680334e-06,
+      "loss": 1.4308,
+      "step": 6674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05666830248948377,
+      "learning_rate": 2.02957973348953e-06,
+      "loss": 1.43,
+      "step": 6675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055280912350215795,
+      "learning_rate": 2.0164006443110268e-06,
+      "loss": 1.4359,
+      "step": 6676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05554549953239934,
+      "learning_rate": 2.003221555132523e-06,
+      "loss": 1.4288,
+      "step": 6677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05532125629595095,
+      "learning_rate": 1.9900424659540198e-06,
+      "loss": 1.4319,
+      "step": 6678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05445235538960028,
+      "learning_rate": 1.9768633767755165e-06,
+      "loss": 1.4427,
+      "step": 6679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05522269423477533,
+      "learning_rate": 1.9636842875970128e-06,
+      "loss": 1.427,
+      "step": 6680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053342733939603086,
+      "learning_rate": 1.9505051984185095e-06,
+      "loss": 1.4295,
+      "step": 6681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.056437989247005216,
+      "learning_rate": 1.937326109240006e-06,
+      "loss": 1.4351,
+      "step": 6682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055471935243057255,
+      "learning_rate": 1.9241470200615025e-06,
+      "loss": 1.4382,
+      "step": 6683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05634282164200721,
+      "learning_rate": 1.910967930882999e-06,
+      "loss": 1.4295,
+      "step": 6684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05715479171740012,
+      "learning_rate": 1.8977888417044957e-06,
+      "loss": 1.4309,
+      "step": 6685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05504295123968306,
+      "learning_rate": 1.8846097525259921e-06,
+      "loss": 1.4335,
+      "step": 6686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05442820364589842,
+      "learning_rate": 1.8714306633474889e-06,
+      "loss": 1.4274,
+      "step": 6687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05564363598278277,
+      "learning_rate": 1.8582515741689853e-06,
+      "loss": 1.4364,
+      "step": 6688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05442159268601778,
+      "learning_rate": 1.8450724849904818e-06,
+      "loss": 1.4272,
+      "step": 6689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054377742248360075,
+      "learning_rate": 1.8318933958119785e-06,
+      "loss": 1.4272,
+      "step": 6690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05393483816845512,
+      "learning_rate": 1.818714306633475e-06,
+      "loss": 1.4241,
+      "step": 6691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05481492043144851,
+      "learning_rate": 1.8055352174549715e-06,
+      "loss": 1.4392,
+      "step": 6692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05577465240564933,
+      "learning_rate": 1.7923561282764682e-06,
+      "loss": 1.4316,
+      "step": 6693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05501187724768036,
+      "learning_rate": 1.7791770390979647e-06,
+      "loss": 1.4285,
+      "step": 6694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05496562715441004,
+      "learning_rate": 1.7659979499194612e-06,
+      "loss": 1.4355,
+      "step": 6695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053976409698713895,
+      "learning_rate": 1.7528188607409577e-06,
+      "loss": 1.4265,
+      "step": 6696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054494932998870614,
+      "learning_rate": 1.7396397715624544e-06,
+      "loss": 1.426,
+      "step": 6697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05614155486179393,
+      "learning_rate": 1.726460682383951e-06,
+      "loss": 1.4344,
+      "step": 6698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05471671543544784,
+      "learning_rate": 1.7132815932054474e-06,
+      "loss": 1.4269,
+      "step": 6699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05733302643873667,
+      "learning_rate": 1.7001025040269441e-06,
+      "loss": 1.4326,
+      "step": 6700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05336659962087332,
+      "learning_rate": 1.6869234148484406e-06,
+      "loss": 1.4335,
+      "step": 6701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05521231470026388,
+      "learning_rate": 1.6737443256699371e-06,
+      "loss": 1.4239,
+      "step": 6702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05489538328186641,
+      "learning_rate": 1.6605652364914338e-06,
+      "loss": 1.4408,
+      "step": 6703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05475141830609558,
+      "learning_rate": 1.6473861473129303e-06,
+      "loss": 1.4362,
+      "step": 6704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05516975874369124,
+      "learning_rate": 1.6342070581344268e-06,
+      "loss": 1.4309,
+      "step": 6705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05357584207036973,
+      "learning_rate": 1.6210279689559235e-06,
+      "loss": 1.436,
+      "step": 6706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054078580660205884,
+      "learning_rate": 1.60784887977742e-06,
+      "loss": 1.4295,
+      "step": 6707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054557172178731066,
+      "learning_rate": 1.5946697905989165e-06,
+      "loss": 1.4374,
+      "step": 6708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05360810412038727,
+      "learning_rate": 1.5814907014204132e-06,
+      "loss": 1.4247,
+      "step": 6709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05643341946801149,
+      "learning_rate": 1.5683116122419097e-06,
+      "loss": 1.4331,
+      "step": 6710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05401928274094889,
+      "learning_rate": 1.5551325230634062e-06,
+      "loss": 1.4333,
+      "step": 6711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05364940472359365,
+      "learning_rate": 1.5419534338849027e-06,
+      "loss": 1.4381,
+      "step": 6712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05452394891045753,
+      "learning_rate": 1.5287743447063994e-06,
+      "loss": 1.4429,
+      "step": 6713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05631412394073035,
+      "learning_rate": 1.515595255527896e-06,
+      "loss": 1.4397,
+      "step": 6714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055846665699376845,
+      "learning_rate": 1.5024161663493924e-06,
+      "loss": 1.4246,
+      "step": 6715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.054223126589573054,
+      "learning_rate": 1.4892370771708891e-06,
+      "loss": 1.4243,
+      "step": 6716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053085278221861265,
+      "learning_rate": 1.4760579879923856e-06,
+      "loss": 1.4285,
+      "step": 6717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05416833024826791,
+      "learning_rate": 1.462878898813882e-06,
+      "loss": 1.4311,
+      "step": 6718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053764157561497036,
+      "learning_rate": 1.4496998096353788e-06,
+      "loss": 1.4323,
+      "step": 6719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.055360397082277496,
+      "learning_rate": 1.4365207204568753e-06,
+      "loss": 1.4331,
+      "step": 6720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0538334341836582,
+      "learning_rate": 1.4233416312783718e-06,
+      "loss": 1.4286,
+      "step": 6721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05389464935082431,
+      "learning_rate": 1.4101625420998685e-06,
+      "loss": 1.4283,
+      "step": 6722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053216975437069144,
+      "learning_rate": 1.3969834529213648e-06,
+      "loss": 1.4393,
+      "step": 6723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.053553851519952615,
+      "learning_rate": 1.3838043637428613e-06,
+      "loss": 1.4232,
+      "step": 6724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.05344645212706313,
+      "learning_rate": 1.370625274564358e-06,
+      "loss": 1.4306,
+      "step": 6725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.057170179152825934,
+      "learning_rate": 1.3574461853858545e-06,
+      "loss": 1.4257,
+      "step": 6726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.057464182499361874,
+      "learning_rate": 1.344267096207351e-06,
+      "loss": 1.4317,
+      "step": 6727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05472553194219716,
+      "learning_rate": 1.3310880070288477e-06,
+      "loss": 1.4241,
+      "step": 6728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054244642846635494,
+      "learning_rate": 1.3179089178503442e-06,
+      "loss": 1.4332,
+      "step": 6729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05510659894982475,
+      "learning_rate": 1.3047298286718407e-06,
+      "loss": 1.4342,
+      "step": 6730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.055566875367151214,
+      "learning_rate": 1.2915507394933374e-06,
+      "loss": 1.424,
+      "step": 6731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05367403123203731,
+      "learning_rate": 1.2783716503148339e-06,
+      "loss": 1.424,
+      "step": 6732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05556284113767626,
+      "learning_rate": 1.2651925611363304e-06,
+      "loss": 1.4263,
+      "step": 6733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054558265959605265,
+      "learning_rate": 1.2520134719578269e-06,
+      "loss": 1.4337,
+      "step": 6734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05481185291442955,
+      "learning_rate": 1.2388343827793236e-06,
+      "loss": 1.4388,
+      "step": 6735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05424990898551823,
+      "learning_rate": 1.22565529360082e-06,
+      "loss": 1.4271,
+      "step": 6736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.055388981215352046,
+      "learning_rate": 1.2124762044223166e-06,
+      "loss": 1.4323,
+      "step": 6737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05360492096096235,
+      "learning_rate": 1.1992971152438133e-06,
+      "loss": 1.4382,
+      "step": 6738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05282013782030302,
+      "learning_rate": 1.1861180260653098e-06,
+      "loss": 1.4427,
+      "step": 6739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05536431621455136,
+      "learning_rate": 1.1729389368868063e-06,
+      "loss": 1.4362,
+      "step": 6740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05615545404774777,
+      "learning_rate": 1.159759847708303e-06,
+      "loss": 1.4343,
+      "step": 6741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053508387058535986,
+      "learning_rate": 1.1465807585297995e-06,
+      "loss": 1.4246,
+      "step": 6742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052952841786724575,
+      "learning_rate": 1.133401669351296e-06,
+      "loss": 1.4273,
+      "step": 6743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05336287922495352,
+      "learning_rate": 1.1202225801727927e-06,
+      "loss": 1.4378,
+      "step": 6744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054163046010663976,
+      "learning_rate": 1.1070434909942891e-06,
+      "loss": 1.434,
+      "step": 6745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0545833563284415,
+      "learning_rate": 1.0938644018157856e-06,
+      "loss": 1.4329,
+      "step": 6746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.056094923535855316,
+      "learning_rate": 1.0806853126372821e-06,
+      "loss": 1.4326,
+      "step": 6747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054228054316257254,
+      "learning_rate": 1.0675062234587788e-06,
+      "loss": 1.4313,
+      "step": 6748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054255183803733655,
+      "learning_rate": 1.0543271342802753e-06,
+      "loss": 1.4324,
+      "step": 6749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05419399666218521,
+      "learning_rate": 1.0411480451017718e-06,
+      "loss": 1.4319,
+      "step": 6750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05373509566423256,
+      "learning_rate": 1.0279689559232685e-06,
+      "loss": 1.4284,
+      "step": 6751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0538816014693077,
+      "learning_rate": 1.014789866744765e-06,
+      "loss": 1.4302,
+      "step": 6752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053616543774326134,
+      "learning_rate": 1.0016107775662615e-06,
+      "loss": 1.4323,
+      "step": 6753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05307365420154263,
+      "learning_rate": 9.884316883877582e-07,
+      "loss": 1.427,
+      "step": 6754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05357123584720764,
+      "learning_rate": 9.752525992092547e-07,
+      "loss": 1.4316,
+      "step": 6755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054041773654542456,
+      "learning_rate": 9.620735100307512e-07,
+      "loss": 1.4275,
+      "step": 6756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05399128803128821,
+      "learning_rate": 9.488944208522478e-07,
+      "loss": 1.4424,
+      "step": 6757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053288317951229534,
+      "learning_rate": 9.357153316737444e-07,
+      "loss": 1.4281,
+      "step": 6758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053319550942173784,
+      "learning_rate": 9.225362424952409e-07,
+      "loss": 1.4381,
+      "step": 6759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054098204582864316,
+      "learning_rate": 9.093571533167375e-07,
+      "loss": 1.4337,
+      "step": 6760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05469662276907384,
+      "learning_rate": 8.961780641382341e-07,
+      "loss": 1.4326,
+      "step": 6761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05414494388566933,
+      "learning_rate": 8.829989749597306e-07,
+      "loss": 1.4346,
+      "step": 6762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05343067260983589,
+      "learning_rate": 8.698198857812272e-07,
+      "loss": 1.4323,
+      "step": 6763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052998945131993686,
+      "learning_rate": 8.566407966027237e-07,
+      "loss": 1.4371,
+      "step": 6764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05345502243878837,
+      "learning_rate": 8.434617074242203e-07,
+      "loss": 1.4344,
+      "step": 6765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05357470088304918,
+      "learning_rate": 8.302826182457169e-07,
+      "loss": 1.4307,
+      "step": 6766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053792911763616176,
+      "learning_rate": 8.171035290672134e-07,
+      "loss": 1.4397,
+      "step": 6767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.056067870310262086,
+      "learning_rate": 8.0392443988871e-07,
+      "loss": 1.4307,
+      "step": 6768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0530417164032033,
+      "learning_rate": 7.907453507102066e-07,
+      "loss": 1.4328,
+      "step": 6769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053229870690655656,
+      "learning_rate": 7.775662615317031e-07,
+      "loss": 1.4311,
+      "step": 6770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05308908194500047,
+      "learning_rate": 7.643871723531997e-07,
+      "loss": 1.4309,
+      "step": 6771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05382324520997576,
+      "learning_rate": 7.512080831746962e-07,
+      "loss": 1.4299,
+      "step": 6772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05391551061832538,
+      "learning_rate": 7.380289939961928e-07,
+      "loss": 1.4334,
+      "step": 6773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053117469915411025,
+      "learning_rate": 7.248499048176894e-07,
+      "loss": 1.4297,
+      "step": 6774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05344039654662785,
+      "learning_rate": 7.116708156391859e-07,
+      "loss": 1.4362,
+      "step": 6775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052993196735069886,
+      "learning_rate": 6.984917264606824e-07,
+      "loss": 1.4343,
+      "step": 6776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05342515529937354,
+      "learning_rate": 6.85312637282179e-07,
+      "loss": 1.4394,
+      "step": 6777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.054375921313173134,
+      "learning_rate": 6.721335481036755e-07,
+      "loss": 1.4327,
+      "step": 6778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05343754284704609,
+      "learning_rate": 6.589544589251721e-07,
+      "loss": 1.4284,
+      "step": 6779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.055089890351672645,
+      "learning_rate": 6.457753697466687e-07,
+      "loss": 1.4354,
+      "step": 6780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05416683250902679,
+      "learning_rate": 6.325962805681652e-07,
+      "loss": 1.4433,
+      "step": 6781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05276091959702106,
+      "learning_rate": 6.194171913896618e-07,
+      "loss": 1.4273,
+      "step": 6782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05376528996269787,
+      "learning_rate": 6.062381022111583e-07,
+      "loss": 1.4201,
+      "step": 6783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05232993743804833,
+      "learning_rate": 5.930590130326549e-07,
+      "loss": 1.4279,
+      "step": 6784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05286378845390823,
+      "learning_rate": 5.798799238541515e-07,
+      "loss": 1.432,
+      "step": 6785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053409540314138,
+      "learning_rate": 5.66700834675648e-07,
+      "loss": 1.4327,
+      "step": 6786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05299270098099772,
+      "learning_rate": 5.535217454971446e-07,
+      "loss": 1.4312,
+      "step": 6787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05314065536427632,
+      "learning_rate": 5.403426563186411e-07,
+      "loss": 1.4291,
+      "step": 6788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05337621585874362,
+      "learning_rate": 5.271635671401377e-07,
+      "loss": 1.4239,
+      "step": 6789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05482540102507366,
+      "learning_rate": 5.139844779616343e-07,
+      "loss": 1.4382,
+      "step": 6790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.05365724374545666,
+      "learning_rate": 5.008053887831308e-07,
+      "loss": 1.4349,
+      "step": 6791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052490691407228694,
+      "learning_rate": 4.876262996046274e-07,
+      "loss": 1.4285,
+      "step": 6792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0519045445655633,
+      "learning_rate": 4.744472104261239e-07,
+      "loss": 1.4288,
+      "step": 6793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.053523979173320003,
+      "learning_rate": 4.6126812124762046e-07,
+      "loss": 1.4353,
+      "step": 6794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.052630805474571093,
+      "learning_rate": 4.4808903206911706e-07,
+      "loss": 1.4316,
+      "step": 6795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05384385855600645,
+      "learning_rate": 4.349099428906136e-07,
+      "loss": 1.4307,
+      "step": 6796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05398178493595745,
+      "learning_rate": 4.2173085371211016e-07,
+      "loss": 1.4387,
+      "step": 6797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052823421370621196,
+      "learning_rate": 4.085517645336067e-07,
+      "loss": 1.4326,
+      "step": 6798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052305313266300654,
+      "learning_rate": 3.953726753551033e-07,
+      "loss": 1.4342,
+      "step": 6799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052205269439430046,
+      "learning_rate": 3.8219358617659985e-07,
+      "loss": 1.4382,
+      "step": 6800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05291254059708552,
+      "learning_rate": 3.690144969980964e-07,
+      "loss": 1.4391,
+      "step": 6801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0534410879894612,
+      "learning_rate": 3.5583540781959295e-07,
+      "loss": 1.4329,
+      "step": 6802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052772447157195056,
+      "learning_rate": 3.426563186410895e-07,
+      "loss": 1.4397,
+      "step": 6803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052585002193206436,
+      "learning_rate": 3.2947722946258604e-07,
+      "loss": 1.4321,
+      "step": 6804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05285812278265277,
+      "learning_rate": 3.162981402840826e-07,
+      "loss": 1.4302,
+      "step": 6805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.053753252066701424,
+      "learning_rate": 3.0311905110557914e-07,
+      "loss": 1.4315,
+      "step": 6806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052260107443514725,
+      "learning_rate": 2.8993996192707574e-07,
+      "loss": 1.4301,
+      "step": 6807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0518132075746276,
+      "learning_rate": 2.767608727485723e-07,
+      "loss": 1.4253,
+      "step": 6808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05315527037075237,
+      "learning_rate": 2.6358178357006883e-07,
+      "loss": 1.4394,
+      "step": 6809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05320122582883372,
+      "learning_rate": 2.504026943915654e-07,
+      "loss": 1.4283,
+      "step": 6810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.054257980503040565,
+      "learning_rate": 2.3722360521306196e-07,
+      "loss": 1.4359,
+      "step": 6811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05230609580305224,
+      "learning_rate": 2.2404451603455853e-07,
+      "loss": 1.4241,
+      "step": 6812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052778059523719986,
+      "learning_rate": 2.1086542685605508e-07,
+      "loss": 1.4384,
+      "step": 6813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05240411170814292,
+      "learning_rate": 1.9768633767755165e-07,
+      "loss": 1.424,
+      "step": 6814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.053596924198039526,
+      "learning_rate": 1.845072484990482e-07,
+      "loss": 1.4309,
+      "step": 6815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.053808119999413206,
+      "learning_rate": 1.7132815932054475e-07,
+      "loss": 1.4359,
+      "step": 6816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.053900741620548843,
+      "learning_rate": 1.581490701420413e-07,
+      "loss": 1.4356,
+      "step": 6817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052268345935564516,
+      "learning_rate": 1.4496998096353787e-07,
+      "loss": 1.4183,
+      "step": 6818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.051697443109882855,
+      "learning_rate": 1.3179089178503442e-07,
+      "loss": 1.4314,
+      "step": 6819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05248763154900252,
+      "learning_rate": 1.1861180260653098e-07,
+      "loss": 1.432,
+      "step": 6820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052891750593259484,
+      "learning_rate": 1.0543271342802754e-07,
+      "loss": 1.4303,
+      "step": 6821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05318823499553962,
+      "learning_rate": 9.22536242495241e-08,
+      "loss": 1.434,
+      "step": 6822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052853190561140476,
+      "learning_rate": 7.907453507102065e-08,
+      "loss": 1.444,
+      "step": 6823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052429086052799216,
+      "learning_rate": 6.589544589251721e-08,
+      "loss": 1.435,
+      "step": 6824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05269668898885189,
+      "learning_rate": 5.271635671401377e-08,
+      "loss": 1.4398,
+      "step": 6825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05189119875681513,
+      "learning_rate": 3.9537267535510324e-08,
+      "loss": 1.4251,
+      "step": 6826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.052591139376407636,
+      "learning_rate": 2.6358178357006885e-08,
+      "loss": 1.4363,
+      "step": 6827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05432255480179053,
+      "learning_rate": 1.3179089178503442e-08,
+      "loss": 1.436,
+      "step": 6828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.05242631273739993,
+      "learning_rate": 0.0,
+      "loss": 1.4258,
+      "step": 6829
+    },
+    {
+      "epoch": 1.0,
+      "step": 6829,
+      "total_flos": 1.6190892499009536e+17,
+      "train_loss": 0.11099688539673983,
+      "train_runtime": 96632.7363,
+      "train_samples_per_second": 142.483,
+      "train_steps_per_second": 0.071
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6829,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 900,
+  "total_flos": 1.6190892499009536e+17,
+  "train_batch_size": 42,
+  "trial_name": null,
+  "trial_params": null
+}