diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,82063 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 11719,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.930480908312965,
+      "learning_rate": 5.681818181818182e-08,
+      "loss": 2.4502,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.433437332776066,
+      "learning_rate": 1.1363636363636364e-07,
+      "loss": 2.5879,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.56981104315501,
+      "learning_rate": 1.7045454545454545e-07,
+      "loss": 2.7832,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 37.375433900955414,
+      "learning_rate": 2.2727272727272729e-07,
+      "loss": 2.6519,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.629759488080225,
+      "learning_rate": 2.840909090909091e-07,
+      "loss": 2.4839,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.939583530212456,
+      "learning_rate": 3.409090909090909e-07,
+      "loss": 2.4238,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.13407096016863,
+      "learning_rate": 3.9772727272727276e-07,
+      "loss": 1.7344,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.650321803464188,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 2.3633,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 27.478679409483913,
+      "learning_rate": 5.113636363636364e-07,
+      "loss": 2.3882,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.84593248264153,
+      "learning_rate": 5.681818181818182e-07,
+      "loss": 2.6533,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.419513485127599,
+      "learning_rate": 6.25e-07,
+      "loss": 1.8589,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.81757121069861,
+      "learning_rate": 6.818181818181818e-07,
+      "loss": 2.4019,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.138074946667803,
+      "learning_rate": 7.386363636363638e-07,
+      "loss": 2.2656,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.935344777704852,
+      "learning_rate": 7.954545454545455e-07,
+      "loss": 2.1802,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.519160169844017,
+      "learning_rate": 8.522727272727273e-07,
+      "loss": 2.0845,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.11597573676965,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 2.042,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.23882140833321,
+      "learning_rate": 9.65909090909091e-07,
+      "loss": 1.9878,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.260542929769937,
+      "learning_rate": 1.0227272727272729e-06,
+      "loss": 2.0317,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.633083940395943,
+      "learning_rate": 1.0795454545454546e-06,
+      "loss": 1.853,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.659672357751758,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 1.8652,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.56538110050701,
+      "learning_rate": 1.1931818181818183e-06,
+      "loss": 1.8477,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.863634484400828,
+      "learning_rate": 1.25e-06,
+      "loss": 1.8486,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.421492756469984,
+      "learning_rate": 1.3068181818181819e-06,
+      "loss": 1.9072,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.624234656159247,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.7666,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.128939498337541,
+      "learning_rate": 1.4204545454545458e-06,
+      "loss": 1.769,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.801952461794798,
+      "learning_rate": 1.4772727272727275e-06,
+      "loss": 1.7534,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.956581233882105,
+      "learning_rate": 1.5340909090909093e-06,
+      "loss": 1.7588,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.627098698026594,
+      "learning_rate": 1.590909090909091e-06,
+      "loss": 1.6587,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.603846870148859,
+      "learning_rate": 1.6477272727272728e-06,
+      "loss": 1.7783,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.749985969714313,
+      "learning_rate": 1.7045454545454546e-06,
+      "loss": 1.7144,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.625541396639092,
+      "learning_rate": 1.7613636363636365e-06,
+      "loss": 1.7476,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.912718735003618,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.5596,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.813309551125995,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 1.7744,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.117579379648124,
+      "learning_rate": 1.931818181818182e-06,
+      "loss": 1.6357,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.182319455591836,
+      "learning_rate": 1.9886363636363638e-06,
+      "loss": 1.6714,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.669374799353258,
+      "learning_rate": 2.0454545454545457e-06,
+      "loss": 1.5981,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.324576081273775,
+      "learning_rate": 2.1022727272727277e-06,
+      "loss": 1.6875,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.339902488375668,
+      "learning_rate": 2.1590909090909092e-06,
+      "loss": 1.6387,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.242443536114178,
+      "learning_rate": 2.2159090909090912e-06,
+      "loss": 1.6309,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.168231381845414,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.6904,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.705783559917247,
+      "learning_rate": 2.3295454545454547e-06,
+      "loss": 1.627,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.788244039230979,
+      "learning_rate": 2.3863636363636367e-06,
+      "loss": 1.5869,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.002744471998939,
+      "learning_rate": 2.4431818181818182e-06,
+      "loss": 1.5278,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.804975190698689,
+      "learning_rate": 2.5e-06,
+      "loss": 1.6909,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.203544396281607,
+      "learning_rate": 2.556818181818182e-06,
+      "loss": 1.6572,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.851012475552404,
+      "learning_rate": 2.6136363636363637e-06,
+      "loss": 1.7183,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.650823714562088,
+      "learning_rate": 2.6704545454545457e-06,
+      "loss": 1.5894,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.25169795293493,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.5591,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.104327555881191,
+      "learning_rate": 2.784090909090909e-06,
+      "loss": 1.5977,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.567597705727648,
+      "learning_rate": 2.8409090909090916e-06,
+      "loss": 1.624,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.755174325486782,
+      "learning_rate": 2.897727272727273e-06,
+      "loss": 1.5664,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.266240771203778,
+      "learning_rate": 2.954545454545455e-06,
+      "loss": 1.605,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.362675729488098,
+      "learning_rate": 3.0113636363636366e-06,
+      "loss": 1.624,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.459676000707437,
+      "learning_rate": 3.0681818181818186e-06,
+      "loss": 1.6523,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.339614599231526,
+      "learning_rate": 3.125e-06,
+      "loss": 1.5762,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.372798798489947,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 1.5332,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.673945885556424,
+      "learning_rate": 3.2386363636363637e-06,
+      "loss": 1.5156,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.620883029122156,
+      "learning_rate": 3.2954545454545456e-06,
+      "loss": 1.6016,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.635216012082727,
+      "learning_rate": 3.352272727272727e-06,
+      "loss": 1.6187,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.335549356081966,
+      "learning_rate": 3.409090909090909e-06,
+      "loss": 1.4927,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.997169237944966,
+      "learning_rate": 3.4659090909090915e-06,
+      "loss": 1.6299,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0433518353491795,
+      "learning_rate": 3.522727272727273e-06,
+      "loss": 1.4927,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.216809196109625,
+      "learning_rate": 3.579545454545455e-06,
+      "loss": 1.584,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.306424219354283,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.5923,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.094386588852722,
+      "learning_rate": 3.6931818181818186e-06,
+      "loss": 1.6011,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.243638694367268,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.623,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.507053695451891,
+      "learning_rate": 3.806818181818182e-06,
+      "loss": 1.6343,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.066226941648021,
+      "learning_rate": 3.863636363636364e-06,
+      "loss": 1.5483,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9151070256565537,
+      "learning_rate": 3.9204545454545456e-06,
+      "loss": 1.5181,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.383867909773355,
+      "learning_rate": 3.9772727272727275e-06,
+      "loss": 1.562,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.3521709415920675,
+      "learning_rate": 4.0340909090909095e-06,
+      "loss": 1.5688,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.286565910316133,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.5591,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.989326184526131,
+      "learning_rate": 4.1477272727272734e-06,
+      "loss": 1.5356,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.14946570131989,
+      "learning_rate": 4.204545454545455e-06,
+      "loss": 1.5127,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.780557886608984,
+      "learning_rate": 4.2613636363636365e-06,
+      "loss": 1.5161,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.135394113591355,
+      "learning_rate": 4.3181818181818185e-06,
+      "loss": 1.5117,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.683498124957659,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 1.5127,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.76963576679545,
+      "learning_rate": 4.4318181818181824e-06,
+      "loss": 1.5811,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.283312101669452,
+      "learning_rate": 4.4886363636363636e-06,
+      "loss": 1.5356,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.988236155474516,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.4927,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.389751972848762,
+      "learning_rate": 4.6022727272727275e-06,
+      "loss": 1.4668,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4668087488966535,
+      "learning_rate": 4.6590909090909095e-06,
+      "loss": 1.5229,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.889722409647566,
+      "learning_rate": 4.715909090909091e-06,
+      "loss": 1.5537,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.31568001154695,
+      "learning_rate": 4.772727272727273e-06,
+      "loss": 1.4998,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.486242705359419,
+      "learning_rate": 4.829545454545455e-06,
+      "loss": 1.3662,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6085286758313373,
+      "learning_rate": 4.8863636363636365e-06,
+      "loss": 1.4048,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.437330632373407,
+      "learning_rate": 4.9431818181818184e-06,
+      "loss": 1.5112,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.023300341471916,
+      "learning_rate": 5e-06,
+      "loss": 1.4995,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9128389880362535,
+      "learning_rate": 5.056818181818182e-06,
+      "loss": 1.4717,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3925049332840023,
+      "learning_rate": 5.113636363636364e-06,
+      "loss": 1.3672,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.378095418302056,
+      "learning_rate": 5.170454545454546e-06,
+      "loss": 1.5283,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7380654871667383,
+      "learning_rate": 5.2272727272727274e-06,
+      "loss": 1.4644,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.34939471011863,
+      "learning_rate": 5.28409090909091e-06,
+      "loss": 1.4651,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.032079195644996,
+      "learning_rate": 5.340909090909091e-06,
+      "loss": 1.5088,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.731554063980816,
+      "learning_rate": 5.397727272727273e-06,
+      "loss": 1.4932,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7689332549475614,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.4531,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.25023214741951,
+      "learning_rate": 5.511363636363637e-06,
+      "loss": 1.5342,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.306358276678666,
+      "learning_rate": 5.568181818181818e-06,
+      "loss": 1.3994,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.596696398854396,
+      "learning_rate": 5.625e-06,
+      "loss": 1.4956,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0254376977604895,
+      "learning_rate": 5.681818181818183e-06,
+      "loss": 1.4844,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.346819726457485,
+      "learning_rate": 5.738636363636364e-06,
+      "loss": 1.4607,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8029483646545246,
+      "learning_rate": 5.795454545454546e-06,
+      "loss": 1.4341,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.371025894032942,
+      "learning_rate": 5.852272727272727e-06,
+      "loss": 1.459,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.657619335072327,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 1.5327,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.351054808849783,
+      "learning_rate": 5.965909090909091e-06,
+      "loss": 1.4956,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.444369416350719,
+      "learning_rate": 6.022727272727273e-06,
+      "loss": 1.4678,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8945888878030432,
+      "learning_rate": 6.079545454545454e-06,
+      "loss": 1.5337,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.727888781104697,
+      "learning_rate": 6.136363636363637e-06,
+      "loss": 1.4043,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.095036826398434,
+      "learning_rate": 6.193181818181818e-06,
+      "loss": 1.5415,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9231378128219845,
+      "learning_rate": 6.25e-06,
+      "loss": 1.4951,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7692207717518467,
+      "learning_rate": 6.306818181818183e-06,
+      "loss": 1.4697,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9865573341024123,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 1.4878,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9666008919648585,
+      "learning_rate": 6.420454545454546e-06,
+      "loss": 1.4243,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.185445485337548,
+      "learning_rate": 6.477272727272727e-06,
+      "loss": 1.4316,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8437741725149923,
+      "learning_rate": 6.53409090909091e-06,
+      "loss": 1.4878,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.860668500603393,
+      "learning_rate": 6.590909090909091e-06,
+      "loss": 1.4136,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7767060736703097,
+      "learning_rate": 6.647727272727273e-06,
+      "loss": 1.3984,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1849084240058105,
+      "learning_rate": 6.704545454545454e-06,
+      "loss": 1.335,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.137080812504202,
+      "learning_rate": 6.761363636363637e-06,
+      "loss": 1.3711,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.043728110508066,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 1.4229,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7191225061154567,
+      "learning_rate": 6.875e-06,
+      "loss": 1.4419,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.99962457742828,
+      "learning_rate": 6.931818181818183e-06,
+      "loss": 1.3193,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.736738513935739,
+      "learning_rate": 6.988636363636364e-06,
+      "loss": 1.2754,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8087207259506046,
+      "learning_rate": 7.045454545454546e-06,
+      "loss": 1.479,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.886473820627131,
+      "learning_rate": 7.102272727272727e-06,
+      "loss": 1.4551,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.526362441964733,
+      "learning_rate": 7.15909090909091e-06,
+      "loss": 1.4575,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.808886898108861,
+      "learning_rate": 7.215909090909091e-06,
+      "loss": 1.4873,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7074458815655817,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 1.3906,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4875090813433083,
+      "learning_rate": 7.329545454545455e-06,
+      "loss": 1.4087,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8984388476142007,
+      "learning_rate": 7.386363636363637e-06,
+      "loss": 1.3784,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.222617633953549,
+      "learning_rate": 7.443181818181818e-06,
+      "loss": 1.4482,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.53977630808359,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.394,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.237231873543606,
+      "learning_rate": 7.556818181818183e-06,
+      "loss": 1.4292,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.116628304405642,
+      "learning_rate": 7.613636363636364e-06,
+      "loss": 1.4297,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.031079932232249,
+      "learning_rate": 7.670454545454547e-06,
+      "loss": 1.4937,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9857418506693394,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 1.4561,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.105595697423818,
+      "learning_rate": 7.784090909090911e-06,
+      "loss": 1.4736,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.059941078843444,
+      "learning_rate": 7.840909090909091e-06,
+      "loss": 1.498,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.149279586631144,
+      "learning_rate": 7.897727272727273e-06,
+      "loss": 1.3662,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9149340853148895,
+      "learning_rate": 7.954545454545455e-06,
+      "loss": 1.3906,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8526007960393174,
+      "learning_rate": 8.011363636363637e-06,
+      "loss": 1.3481,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.137232651952521,
+      "learning_rate": 8.068181818181819e-06,
+      "loss": 1.374,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8134327954358467,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.417,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9300893088510453,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 1.4399,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.105453483469677,
+      "learning_rate": 8.238636363636365e-06,
+      "loss": 1.373,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.227108526898493,
+      "learning_rate": 8.295454545454547e-06,
+      "loss": 1.4536,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.119282797231501,
+      "learning_rate": 8.352272727272727e-06,
+      "loss": 1.3994,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5716523073443405,
+      "learning_rate": 8.40909090909091e-06,
+      "loss": 1.2642,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.924829129054961,
+      "learning_rate": 8.465909090909091e-06,
+      "loss": 1.5,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9817896816366,
+      "learning_rate": 8.522727272727273e-06,
+      "loss": 1.4204,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9626957465677393,
+      "learning_rate": 8.579545454545455e-06,
+      "loss": 1.4387,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7762502363686785,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 1.4248,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.975343025646361,
+      "learning_rate": 8.693181818181819e-06,
+      "loss": 1.4058,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9816973928987185,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.4363,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.623694176114208,
+      "learning_rate": 8.806818181818183e-06,
+      "loss": 1.4663,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7762153344358325,
+      "learning_rate": 8.863636363636365e-06,
+      "loss": 1.4678,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.145140934717821,
+      "learning_rate": 8.920454545454547e-06,
+      "loss": 1.3809,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.5406422812283305,
+      "learning_rate": 8.977272727272727e-06,
+      "loss": 1.3662,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.833993545196003,
+      "learning_rate": 9.03409090909091e-06,
+      "loss": 1.4673,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8953977884775144,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.2139,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2145329143453045,
+      "learning_rate": 9.147727272727273e-06,
+      "loss": 1.4009,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.26871968349517,
+      "learning_rate": 9.204545454545455e-06,
+      "loss": 1.4082,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.232909504332312,
+      "learning_rate": 9.261363636363637e-06,
+      "loss": 1.3555,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.848105457723595,
+      "learning_rate": 9.318181818181819e-06,
+      "loss": 1.4175,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.261541372064156,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.3853,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.534185348932054,
+      "learning_rate": 9.431818181818183e-06,
+      "loss": 1.3442,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.288922324087422,
+      "learning_rate": 9.488636363636365e-06,
+      "loss": 1.4204,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.39763408591516,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 1.397,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.194534254548171,
+      "learning_rate": 9.602272727272727e-06,
+      "loss": 1.4172,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.198874493611496,
+      "learning_rate": 9.65909090909091e-06,
+      "loss": 1.3567,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7217378793225295,
+      "learning_rate": 9.715909090909091e-06,
+      "loss": 1.4062,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.164725087888128,
+      "learning_rate": 9.772727272727273e-06,
+      "loss": 1.4448,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2288124772725295,
+      "learning_rate": 9.829545454545455e-06,
+      "loss": 1.4106,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3725058998436745,
+      "learning_rate": 9.886363636363637e-06,
+      "loss": 1.4736,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.756328958831512,
+      "learning_rate": 9.943181818181819e-06,
+      "loss": 1.3108,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.301105936116914,
+      "learning_rate": 1e-05,
+      "loss": 1.4155,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.608452415473536,
+      "learning_rate": 1.0056818181818183e-05,
+      "loss": 1.4243,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0830806278524605,
+      "learning_rate": 1.0113636363636365e-05,
+      "loss": 1.2969,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.05367975664105,
+      "learning_rate": 1.0170454545454547e-05,
+      "loss": 1.4053,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.312930428794867,
+      "learning_rate": 1.0227272727272729e-05,
+      "loss": 1.4414,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.294398124509312,
+      "learning_rate": 1.0284090909090909e-05,
+      "loss": 1.4399,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.73999853248862,
+      "learning_rate": 1.0340909090909093e-05,
+      "loss": 1.4451,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0710931103097705,
+      "learning_rate": 1.0397727272727275e-05,
+      "loss": 1.3799,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.500946821700946,
+      "learning_rate": 1.0454545454545455e-05,
+      "loss": 1.3484,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.062219519630023,
+      "learning_rate": 1.0511363636363637e-05,
+      "loss": 1.3359,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.561729971608431,
+      "learning_rate": 1.056818181818182e-05,
+      "loss": 1.3533,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.078446101787263,
+      "learning_rate": 1.0625e-05,
+      "loss": 1.3535,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.769049707597702,
+      "learning_rate": 1.0681818181818183e-05,
+      "loss": 1.3525,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.430320169371985,
+      "learning_rate": 1.0738636363636365e-05,
+      "loss": 1.4277,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.995719064279274,
+      "learning_rate": 1.0795454545454547e-05,
+      "loss": 1.3472,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.323979311459056,
+      "learning_rate": 1.0852272727272729e-05,
+      "loss": 1.4473,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9689245981104166,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 1.3286,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.459305842721448,
+      "learning_rate": 1.0965909090909093e-05,
+      "loss": 1.3755,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.655311292634843,
+      "learning_rate": 1.1022727272727275e-05,
+      "loss": 1.4307,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.368194215816722,
+      "learning_rate": 1.1079545454545455e-05,
+      "loss": 1.4961,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.145125005456112,
+      "learning_rate": 1.1136363636363637e-05,
+      "loss": 1.4268,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9374691110725815,
+      "learning_rate": 1.119318181818182e-05,
+      "loss": 1.3381,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.453951644288471,
+      "learning_rate": 1.125e-05,
+      "loss": 1.4163,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.617594792039605,
+      "learning_rate": 1.1306818181818183e-05,
+      "loss": 1.3682,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.315665849450732,
+      "learning_rate": 1.1363636363636366e-05,
+      "loss": 1.4097,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.724660066295671,
+      "learning_rate": 1.1420454545454547e-05,
+      "loss": 1.3467,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.328639279420323,
+      "learning_rate": 1.1477272727272729e-05,
+      "loss": 1.4185,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.76564985357092,
+      "learning_rate": 1.1534090909090909e-05,
+      "loss": 1.4243,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.317109554481436,
+      "learning_rate": 1.1590909090909093e-05,
+      "loss": 1.28,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.172174489362356,
+      "learning_rate": 1.1647727272727274e-05,
+      "loss": 1.3721,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5050661044372635,
+      "learning_rate": 1.1704545454545455e-05,
+      "loss": 1.4766,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3769961948224765,
+      "learning_rate": 1.1761363636363637e-05,
+      "loss": 1.4316,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7419023579369446,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 1.3804,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6762796800331197,
+      "learning_rate": 1.1875e-05,
+      "loss": 1.4316,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.128065494625384,
+      "learning_rate": 1.1931818181818183e-05,
+      "loss": 1.2266,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2785461553816315,
+      "learning_rate": 1.1988636363636366e-05,
+      "loss": 1.4155,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.200874852809631,
+      "learning_rate": 1.2045454545454547e-05,
+      "loss": 1.2708,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5448013689770765,
+      "learning_rate": 1.2102272727272729e-05,
+      "loss": 1.3359,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.141799207741766,
+      "learning_rate": 1.2159090909090909e-05,
+      "loss": 1.3296,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.997850853317357,
+      "learning_rate": 1.2215909090909092e-05,
+      "loss": 1.396,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.001857201388679,
+      "learning_rate": 1.2272727272727274e-05,
+      "loss": 1.3804,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7878056487885847,
+      "learning_rate": 1.2329545454545455e-05,
+      "loss": 1.301,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.42488562784146,
+      "learning_rate": 1.2386363636363637e-05,
+      "loss": 1.4238,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8849449998535293,
+      "learning_rate": 1.244318181818182e-05,
+      "loss": 1.3921,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9632986471112774,
+      "learning_rate": 1.25e-05,
+      "loss": 1.3467,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.072219835796943,
+      "learning_rate": 1.2556818181818183e-05,
+      "loss": 1.3706,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.162543062540064,
+      "learning_rate": 1.2613636363636366e-05,
+      "loss": 1.4373,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.767926119221732,
+      "learning_rate": 1.2670454545454547e-05,
+      "loss": 1.3442,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.025680302752144,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 1.3018,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.711473640561914,
+      "learning_rate": 1.2784090909090909e-05,
+      "loss": 1.4028,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.707872185701933,
+      "learning_rate": 1.2840909090909092e-05,
+      "loss": 1.377,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8517993875674503,
+      "learning_rate": 1.2897727272727274e-05,
+      "loss": 1.3481,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.593387400174386,
+      "learning_rate": 1.2954545454545455e-05,
+      "loss": 1.3535,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9360605485601305,
+      "learning_rate": 1.3011363636363637e-05,
+      "loss": 1.4282,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5543520683887095,
+      "learning_rate": 1.306818181818182e-05,
+      "loss": 1.4106,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.373446412957313,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.4307,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.174777189911527,
+      "learning_rate": 1.3181818181818183e-05,
+      "loss": 1.3262,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5117903237745187,
+      "learning_rate": 1.3238636363636366e-05,
+      "loss": 1.3047,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6384613372412105,
+      "learning_rate": 1.3295454545454546e-05,
+      "loss": 1.3398,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9813408168911857,
+      "learning_rate": 1.3352272727272728e-05,
+      "loss": 1.3633,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.531394596193499,
+      "learning_rate": 1.3409090909090909e-05,
+      "loss": 1.2843,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.125356841073173,
+      "learning_rate": 1.3465909090909092e-05,
+      "loss": 1.3887,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3169590230343635,
+      "learning_rate": 1.3522727272727274e-05,
+      "loss": 1.3252,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.267399718704057,
+      "learning_rate": 1.3579545454545455e-05,
+      "loss": 1.3455,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.871503141530147,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.3271,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9941740610546574,
+      "learning_rate": 1.369318181818182e-05,
+      "loss": 1.3369,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.259385441575228,
+      "learning_rate": 1.375e-05,
+      "loss": 1.4399,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.774235115493736,
+      "learning_rate": 1.3806818181818182e-05,
+      "loss": 1.292,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4928884193512677,
+      "learning_rate": 1.3863636363636366e-05,
+      "loss": 1.3176,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4830061288176153,
+      "learning_rate": 1.3920454545454546e-05,
+      "loss": 1.3945,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.00765197601605,
+      "learning_rate": 1.3977272727272728e-05,
+      "loss": 1.4043,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6884096006169615,
+      "learning_rate": 1.4034090909090909e-05,
+      "loss": 1.4351,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.978765699971217,
+      "learning_rate": 1.4090909090909092e-05,
+      "loss": 1.4189,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.872188945242129,
+      "learning_rate": 1.4147727272727274e-05,
+      "loss": 1.4009,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1695083170938676,
+      "learning_rate": 1.4204545454545455e-05,
+      "loss": 1.2678,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.015243204359287,
+      "learning_rate": 1.4261363636363637e-05,
+      "loss": 1.3979,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.973365064747145,
+      "learning_rate": 1.431818181818182e-05,
+      "loss": 1.3779,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.823651942466219,
+      "learning_rate": 1.4375e-05,
+      "loss": 1.3435,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4865226938412395,
+      "learning_rate": 1.4431818181818182e-05,
+      "loss": 1.2939,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4004889332409785,
+      "learning_rate": 1.4488636363636366e-05,
+      "loss": 1.3887,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.911300832018295,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 1.3491,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6401482889565915,
+      "learning_rate": 1.4602272727272728e-05,
+      "loss": 1.2892,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0878628097841805,
+      "learning_rate": 1.465909090909091e-05,
+      "loss": 1.377,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7810669892382562,
+      "learning_rate": 1.4715909090909092e-05,
+      "loss": 1.3364,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.359874565438019,
+      "learning_rate": 1.4772727272727274e-05,
+      "loss": 1.3555,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.58783267960249,
+      "learning_rate": 1.4829545454545456e-05,
+      "loss": 1.3657,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.187443682128192,
+      "learning_rate": 1.4886363636363636e-05,
+      "loss": 1.4238,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9683986056269003,
+      "learning_rate": 1.494318181818182e-05,
+      "loss": 1.2542,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.099512968335502,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.3657,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.664041773919453,
+      "learning_rate": 1.5056818181818182e-05,
+      "loss": 1.3086,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.314306421794856,
+      "learning_rate": 1.5113636363636366e-05,
+      "loss": 1.302,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.616992289653167,
+      "learning_rate": 1.5170454545454546e-05,
+      "loss": 1.4058,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.267641754898127,
+      "learning_rate": 1.5227272727272728e-05,
+      "loss": 1.3926,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.837593294034632,
+      "learning_rate": 1.528409090909091e-05,
+      "loss": 1.4717,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.388850864697399,
+      "learning_rate": 1.5340909090909094e-05,
+      "loss": 1.2014,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3449353421934367,
+      "learning_rate": 1.5397727272727276e-05,
+      "loss": 1.2014,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.974289312593837,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 1.3564,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.100588996342797,
+      "learning_rate": 1.5511363636363636e-05,
+      "loss": 1.4092,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.734615065685456,
+      "learning_rate": 1.5568181818181822e-05,
+      "loss": 1.3999,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.423204890586678,
+      "learning_rate": 1.5625e-05,
+      "loss": 1.3794,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.957725541661371,
+      "learning_rate": 1.5681818181818182e-05,
+      "loss": 1.374,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.569330210966471,
+      "learning_rate": 1.5738636363636364e-05,
+      "loss": 1.313,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.703574091042077,
+      "learning_rate": 1.5795454545454546e-05,
+      "loss": 1.395,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.913894005867577,
+      "learning_rate": 1.5852272727272728e-05,
+      "loss": 1.4512,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.419438463481698,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 1.3633,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.132381994230466,
+      "learning_rate": 1.5965909090909092e-05,
+      "loss": 1.353,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.466283181546603,
+      "learning_rate": 1.6022727272727274e-05,
+      "loss": 1.4248,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.753967513384739,
+      "learning_rate": 1.6079545454545456e-05,
+      "loss": 1.3857,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.017004253843036,
+      "learning_rate": 1.6136363636363638e-05,
+      "loss": 1.3511,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.145469511718163,
+      "learning_rate": 1.619318181818182e-05,
+      "loss": 1.3677,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.889875242318787,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 1.3213,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8697411398751935,
+      "learning_rate": 1.6306818181818184e-05,
+      "loss": 1.4111,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.196320060240857,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 1.3396,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.542086400154194,
+      "learning_rate": 1.6420454545454548e-05,
+      "loss": 1.3662,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8586464124912134,
+      "learning_rate": 1.647727272727273e-05,
+      "loss": 1.3613,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6981394206836,
+      "learning_rate": 1.653409090909091e-05,
+      "loss": 1.2766,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.622987287592926,
+      "learning_rate": 1.6590909090909094e-05,
+      "loss": 1.396,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.185148262876516,
+      "learning_rate": 1.6647727272727276e-05,
+      "loss": 1.3616,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.60241853914143,
+      "learning_rate": 1.6704545454545454e-05,
+      "loss": 1.2737,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.336008266912438,
+      "learning_rate": 1.6761363636363636e-05,
+      "loss": 1.2693,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5837525633181415,
+      "learning_rate": 1.681818181818182e-05,
+      "loss": 1.3369,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.6559950244500055,
+      "learning_rate": 1.6875e-05,
+      "loss": 1.3672,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.3072354543535125,
+      "learning_rate": 1.6931818181818182e-05,
+      "loss": 1.354,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.21588052344784,
+      "learning_rate": 1.6988636363636368e-05,
+      "loss": 1.4058,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.3940225499621155,
+      "learning_rate": 1.7045454545454546e-05,
+      "loss": 1.3926,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.497451732653,
+      "learning_rate": 1.7102272727272728e-05,
+      "loss": 1.4014,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6825784298302064,
+      "learning_rate": 1.715909090909091e-05,
+      "loss": 1.4092,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.776270055436438,
+      "learning_rate": 1.7215909090909092e-05,
+      "loss": 1.3599,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.509181294453535,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 1.3452,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.955509480630775,
+      "learning_rate": 1.7329545454545456e-05,
+      "loss": 1.3774,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.579264937221738,
+      "learning_rate": 1.7386363636363638e-05,
+      "loss": 1.4199,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.48442339123392,
+      "learning_rate": 1.744318181818182e-05,
+      "loss": 1.3252,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9902458854776284,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.3752,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.733450167014986,
+      "learning_rate": 1.7556818181818184e-05,
+      "loss": 1.3774,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4656017789293685,
+      "learning_rate": 1.7613636363636366e-05,
+      "loss": 1.3037,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.358334536471822,
+      "learning_rate": 1.7670454545454548e-05,
+      "loss": 1.3721,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.232861732991846,
+      "learning_rate": 1.772727272727273e-05,
+      "loss": 1.3313,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.486791920239651,
+      "learning_rate": 1.7784090909090908e-05,
+      "loss": 1.3938,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.208879375033146,
+      "learning_rate": 1.7840909090909094e-05,
+      "loss": 1.4243,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.682861163055463,
+      "learning_rate": 1.7897727272727276e-05,
+      "loss": 1.3955,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8835575357516343,
+      "learning_rate": 1.7954545454545454e-05,
+      "loss": 1.3286,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8075629124291637,
+      "learning_rate": 1.8011363636363636e-05,
+      "loss": 1.314,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6788334069145945,
+      "learning_rate": 1.806818181818182e-05,
+      "loss": 1.3064,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.641391104399936,
+      "learning_rate": 1.8125e-05,
+      "loss": 1.396,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.856159147303474,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.3726,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.429450695349465,
+      "learning_rate": 1.8238636363636367e-05,
+      "loss": 1.3838,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.470489334354247,
+      "learning_rate": 1.8295454545454546e-05,
+      "loss": 1.3794,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2172740654429273,
+      "learning_rate": 1.8352272727272728e-05,
+      "loss": 1.3999,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.029429833404596,
+      "learning_rate": 1.840909090909091e-05,
+      "loss": 1.3069,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7469749268909656,
+      "learning_rate": 1.8465909090909092e-05,
+      "loss": 1.3311,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5774214186352267,
+      "learning_rate": 1.8522727272727274e-05,
+      "loss": 1.3142,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.772863017183955,
+      "learning_rate": 1.8579545454545456e-05,
+      "loss": 1.3574,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8539958903269076,
+      "learning_rate": 1.8636363636363638e-05,
+      "loss": 1.4443,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.988685534528282,
+      "learning_rate": 1.869318181818182e-05,
+      "loss": 1.3906,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7243583655379977,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.3237,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.529122384194696,
+      "learning_rate": 1.8806818181818184e-05,
+      "loss": 1.4072,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9170170005482245,
+      "learning_rate": 1.8863636363636366e-05,
+      "loss": 1.4185,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7212162588871585,
+      "learning_rate": 1.8920454545454548e-05,
+      "loss": 1.3467,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.656348484268761,
+      "learning_rate": 1.897727272727273e-05,
+      "loss": 1.3579,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7142917922965495,
+      "learning_rate": 1.9034090909090908e-05,
+      "loss": 1.2998,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5566785232052984,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 1.2634,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.868697907279285,
+      "learning_rate": 1.9147727272727276e-05,
+      "loss": 1.3984,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3708102716512967,
+      "learning_rate": 1.9204545454545454e-05,
+      "loss": 1.3203,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.842176380331417,
+      "learning_rate": 1.9261363636363636e-05,
+      "loss": 1.344,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.073870383617819,
+      "learning_rate": 1.931818181818182e-05,
+      "loss": 1.3223,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.541042225207043,
+      "learning_rate": 1.9375e-05,
+      "loss": 1.3162,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8028691637243264,
+      "learning_rate": 1.9431818181818182e-05,
+      "loss": 1.3994,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.233982217140837,
+      "learning_rate": 1.9488636363636367e-05,
+      "loss": 1.3135,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.111956833046347,
+      "learning_rate": 1.9545454545454546e-05,
+      "loss": 1.3682,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.004816296949566,
+      "learning_rate": 1.9602272727272728e-05,
+      "loss": 1.3979,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.656960818308839,
+      "learning_rate": 1.965909090909091e-05,
+      "loss": 1.2761,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0784363478209995,
+      "learning_rate": 1.9715909090909092e-05,
+      "loss": 1.3398,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.59378362593261,
+      "learning_rate": 1.9772727272727274e-05,
+      "loss": 1.416,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.009465368236238,
+      "learning_rate": 1.9829545454545456e-05,
+      "loss": 1.2988,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.739539793894413,
+      "learning_rate": 1.9886363636363638e-05,
+      "loss": 1.3579,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5615085766761796,
+      "learning_rate": 1.994318181818182e-05,
+      "loss": 1.3599,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.756545664711726,
+      "learning_rate": 2e-05,
+      "loss": 1.3691,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.111930483187007,
+      "learning_rate": 1.9999999618075055e-05,
+      "loss": 1.2876,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.144918296288623,
+      "learning_rate": 1.9999998472300245e-05,
+      "loss": 1.4282,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.05492434448453,
+      "learning_rate": 1.9999996562675653e-05,
+      "loss": 1.3408,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.781718023841436,
+      "learning_rate": 1.9999993889201435e-05,
+      "loss": 1.217,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.176071777130726,
+      "learning_rate": 1.9999990451877787e-05,
+      "loss": 1.3008,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9022543200516386,
+      "learning_rate": 1.9999986250704974e-05,
+      "loss": 1.3572,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.59614657847542,
+      "learning_rate": 1.9999981285683318e-05,
+      "loss": 1.4141,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8079354455750023,
+      "learning_rate": 1.9999975556813198e-05,
+      "loss": 1.3711,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7438201527110526,
+      "learning_rate": 1.9999969064095054e-05,
+      "loss": 1.394,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.316207599383989,
+      "learning_rate": 1.9999961807529375e-05,
+      "loss": 1.3691,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.055960125587236,
+      "learning_rate": 1.9999953787116723e-05,
+      "loss": 1.3789,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6116550087693406,
+      "learning_rate": 1.9999945002857704e-05,
+      "loss": 1.3218,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3880379556596583,
+      "learning_rate": 1.9999935454752993e-05,
+      "loss": 1.3745,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.765949016941916,
+      "learning_rate": 1.9999925142803317e-05,
+      "loss": 1.3657,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.191578249144169,
+      "learning_rate": 1.9999914067009467e-05,
+      "loss": 1.3623,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7891015751288775,
+      "learning_rate": 1.9999902227372287e-05,
+      "loss": 1.3198,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.746533283546408,
+      "learning_rate": 1.999988962389268e-05,
+      "loss": 1.3281,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.000750122687617,
+      "learning_rate": 1.9999876256571612e-05,
+      "loss": 1.2971,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3791091375683324,
+      "learning_rate": 1.99998621254101e-05,
+      "loss": 1.3481,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.43922858031181,
+      "learning_rate": 1.9999847230409228e-05,
+      "loss": 1.3481,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.050082665284306,
+      "learning_rate": 1.999983157157013e-05,
+      "loss": 1.4268,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.322103925051326,
+      "learning_rate": 1.9999815148894005e-05,
+      "loss": 1.3735,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.299479257266036,
+      "learning_rate": 1.9999797962382104e-05,
+      "loss": 1.3608,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.121422204469389,
+      "learning_rate": 1.999978001203574e-05,
+      "loss": 1.334,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.262207930998349,
+      "learning_rate": 1.999976129785629e-05,
+      "loss": 1.3196,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.450502659587759,
+      "learning_rate": 1.9999741819845178e-05,
+      "loss": 1.3525,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.92248060648359,
+      "learning_rate": 1.9999721578003894e-05,
+      "loss": 1.3472,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.464384592894435,
+      "learning_rate": 1.9999700572333983e-05,
+      "loss": 1.3005,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6060874212887084,
+      "learning_rate": 1.999967880283705e-05,
+      "loss": 1.2998,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.878224644920712,
+      "learning_rate": 1.9999656269514758e-05,
+      "loss": 1.4551,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.479052282059771,
+      "learning_rate": 1.9999632972368824e-05,
+      "loss": 1.3472,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.881261295164263,
+      "learning_rate": 1.9999608911401037e-05,
+      "loss": 1.3567,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.691201510558193,
+      "learning_rate": 1.9999584086613226e-05,
+      "loss": 1.4038,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.455864431351744,
+      "learning_rate": 1.9999558498007295e-05,
+      "loss": 1.4897,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.030773892406895,
+      "learning_rate": 1.999953214558519e-05,
+      "loss": 1.3501,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.731305951955159,
+      "learning_rate": 1.9999505029348927e-05,
+      "loss": 1.4619,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.202494178361304,
+      "learning_rate": 1.9999477149300583e-05,
+      "loss": 1.3833,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.626970940640547,
+      "learning_rate": 1.9999448505442278e-05,
+      "loss": 1.373,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.643657249279224,
+      "learning_rate": 1.999941909777621e-05,
+      "loss": 1.3687,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.744697389977843,
+      "learning_rate": 1.9999388926304617e-05,
+      "loss": 1.3594,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.65674476822078,
+      "learning_rate": 1.9999357991029808e-05,
+      "loss": 1.3569,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.15258612591877,
+      "learning_rate": 1.9999326291954145e-05,
+      "loss": 1.3628,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.957282501305781,
+      "learning_rate": 1.9999293829080048e-05,
+      "loss": 1.3809,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7254859086488037,
+      "learning_rate": 1.999926060241e-05,
+      "loss": 1.0457,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.058104422921121,
+      "learning_rate": 1.9999226611946536e-05,
+      "loss": 1.2966,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.705299622415969,
+      "learning_rate": 1.9999191857692252e-05,
+      "loss": 1.4121,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6751450920122246,
+      "learning_rate": 1.9999156339649806e-05,
+      "loss": 1.1707,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.802998329788373,
+      "learning_rate": 1.999912005782191e-05,
+      "loss": 1.3071,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.452767808577896,
+      "learning_rate": 1.999908301221133e-05,
+      "loss": 1.4336,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.232793976794093,
+      "learning_rate": 1.9999045202820907e-05,
+      "loss": 1.3926,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.254257248640603,
+      "learning_rate": 1.9999006629653518e-05,
+      "loss": 1.2954,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.552206401576211,
+      "learning_rate": 1.9998967292712112e-05,
+      "loss": 1.3279,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.811660764641815,
+      "learning_rate": 1.99989271919997e-05,
+      "loss": 1.3755,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.22070844942081,
+      "learning_rate": 1.9998886327519337e-05,
+      "loss": 1.3428,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.29943239076167,
+      "learning_rate": 1.999884469927415e-05,
+      "loss": 1.3569,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.240892095919714,
+      "learning_rate": 1.9998802307267315e-05,
+      "loss": 1.3789,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.475304318061393,
+      "learning_rate": 1.9998759151502075e-05,
+      "loss": 1.3704,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.560095059724591,
+      "learning_rate": 1.9998715231981717e-05,
+      "loss": 1.3384,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.489556809951442,
+      "learning_rate": 1.999867054870961e-05,
+      "loss": 1.3254,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.671009839887578,
+      "learning_rate": 1.9998625101689155e-05,
+      "loss": 1.3359,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.326765026919866,
+      "learning_rate": 1.9998578890923828e-05,
+      "loss": 1.3909,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.45929945464559,
+      "learning_rate": 1.999853191641716e-05,
+      "loss": 1.3262,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4743073659461263,
+      "learning_rate": 1.9998484178172735e-05,
+      "loss": 1.3237,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.370558953519804,
+      "learning_rate": 1.9998435676194202e-05,
+      "loss": 1.3569,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4432486839969325,
+      "learning_rate": 1.9998386410485268e-05,
+      "loss": 1.3643,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6698641660841678,
+      "learning_rate": 1.9998336381049694e-05,
+      "loss": 1.395,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.636378133773945,
+      "learning_rate": 1.99982855878913e-05,
+      "loss": 1.3286,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9562159596974213,
+      "learning_rate": 1.999823403101397e-05,
+      "loss": 1.3174,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.461183717625731,
+      "learning_rate": 1.999818171042164e-05,
+      "loss": 1.2737,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9792395898888366,
+      "learning_rate": 1.9998128626118302e-05,
+      "loss": 1.3413,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7534385686542087,
+      "learning_rate": 1.999807477810802e-05,
+      "loss": 1.2996,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.012880372342811,
+      "learning_rate": 1.9998020166394898e-05,
+      "loss": 1.3701,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2479247522133683,
+      "learning_rate": 1.9997964790983112e-05,
+      "loss": 1.2197,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6939377802099007,
+      "learning_rate": 1.999790865187689e-05,
+      "loss": 1.2932,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5051667757382345,
+      "learning_rate": 1.999785174908053e-05,
+      "loss": 1.3638,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7401827900767968,
+      "learning_rate": 1.999779408259836e-05,
+      "loss": 1.2915,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3409637575372795,
+      "learning_rate": 1.99977356524348e-05,
+      "loss": 1.3052,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.334751543568062,
+      "learning_rate": 1.9997676458594308e-05,
+      "loss": 1.3823,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.688065062407599,
+      "learning_rate": 1.999761650108141e-05,
+      "loss": 1.2588,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9459391774889374,
+      "learning_rate": 1.9997555779900673e-05,
+      "loss": 1.3862,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.69243497037332,
+      "learning_rate": 1.999749429505675e-05,
+      "loss": 1.3606,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.707712383597495,
+      "learning_rate": 1.9997432046554327e-05,
+      "loss": 1.312,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6587302277034084,
+      "learning_rate": 1.999736903439817e-05,
+      "loss": 1.2969,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.067493021869799,
+      "learning_rate": 1.999730525859308e-05,
+      "loss": 1.3936,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0870116178284417,
+      "learning_rate": 1.9997240719143937e-05,
+      "loss": 1.2781,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7607051717035547,
+      "learning_rate": 1.9997175416055662e-05,
+      "loss": 1.3823,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7137878247993927,
+      "learning_rate": 1.9997109349333254e-05,
+      "loss": 1.3638,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5245921410229424,
+      "learning_rate": 1.9997042518981752e-05,
+      "loss": 1.2725,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.695714817012496,
+      "learning_rate": 1.9996974925006264e-05,
+      "loss": 1.3232,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.695301263562271,
+      "learning_rate": 1.9996906567411957e-05,
+      "loss": 1.2847,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8310781463435375,
+      "learning_rate": 1.9996837446204044e-05,
+      "loss": 1.3086,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.443452161040808,
+      "learning_rate": 1.9996767561387807e-05,
+      "loss": 1.1604,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1066711004048617,
+      "learning_rate": 1.9996696912968587e-05,
+      "loss": 1.2573,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.678012023403452,
+      "learning_rate": 1.9996625500951778e-05,
+      "loss": 1.3984,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.456365977946773,
+      "learning_rate": 1.999655332534284e-05,
+      "loss": 1.3433,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.660784761811891,
+      "learning_rate": 1.999648038614728e-05,
+      "loss": 1.3735,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8311043385310697,
+      "learning_rate": 1.999640668337067e-05,
+      "loss": 1.3633,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.130638348187639,
+      "learning_rate": 1.9996332217018645e-05,
+      "loss": 1.3604,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3271142296114418,
+      "learning_rate": 1.999625698709689e-05,
+      "loss": 1.2527,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2502878582000783,
+      "learning_rate": 1.999618099361115e-05,
+      "loss": 1.2122,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.336829043458802,
+      "learning_rate": 1.999610423656723e-05,
+      "loss": 1.3589,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.139760838382213,
+      "learning_rate": 1.9996026715970995e-05,
+      "loss": 1.3408,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.504912918612004,
+      "learning_rate": 1.999594843182836e-05,
+      "loss": 1.4053,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6359608094191986,
+      "learning_rate": 1.999586938414532e-05,
+      "loss": 1.2612,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.953679905583176,
+      "learning_rate": 1.9995789572927897e-05,
+      "loss": 1.2888,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.725311797168922,
+      "learning_rate": 1.9995708998182196e-05,
+      "loss": 1.333,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4727006563162326,
+      "learning_rate": 1.9995627659914366e-05,
+      "loss": 1.2405,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.030688684885521,
+      "learning_rate": 1.9995545558130624e-05,
+      "loss": 1.2402,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7498333357491473,
+      "learning_rate": 1.999546269283724e-05,
+      "loss": 1.3242,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.066335861421708,
+      "learning_rate": 1.9995379064040547e-05,
+      "loss": 1.3098,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9211648104720065,
+      "learning_rate": 1.999529467174693e-05,
+      "loss": 1.334,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.385000153619743,
+      "learning_rate": 1.9995209515962835e-05,
+      "loss": 1.2373,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3465487096522892,
+      "learning_rate": 1.999512359669477e-05,
+      "loss": 1.3352,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6580410480746153,
+      "learning_rate": 1.9995036913949288e-05,
+      "loss": 1.3047,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8501140007538104,
+      "learning_rate": 1.999494946773302e-05,
+      "loss": 1.3938,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8849069658686974,
+      "learning_rate": 1.9994861258052647e-05,
+      "loss": 1.4114,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6003368515655314,
+      "learning_rate": 1.9994772284914897e-05,
+      "loss": 1.3477,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.650023799138505,
+      "learning_rate": 1.9994682548326576e-05,
+      "loss": 1.3501,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4489785850071564,
+      "learning_rate": 1.9994592048294534e-05,
+      "loss": 1.271,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.590690761599221,
+      "learning_rate": 1.9994500784825685e-05,
+      "loss": 1.3203,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.741195251005779,
+      "learning_rate": 1.9994408757926997e-05,
+      "loss": 1.2734,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.731639542649874,
+      "learning_rate": 1.9994315967605506e-05,
+      "loss": 1.3618,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.348058032484115,
+      "learning_rate": 1.9994222413868293e-05,
+      "loss": 1.373,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.158039226298638,
+      "learning_rate": 1.9994128096722508e-05,
+      "loss": 1.3726,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.109270258573098,
+      "learning_rate": 1.9994033016175357e-05,
+      "loss": 1.3315,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.558902732298717,
+      "learning_rate": 1.9993937172234094e-05,
+      "loss": 1.311,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1435470595034687,
+      "learning_rate": 1.999384056490605e-05,
+      "loss": 1.2915,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7780913537233456,
+      "learning_rate": 1.99937431941986e-05,
+      "loss": 1.3164,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.087201823537328,
+      "learning_rate": 1.999364506011918e-05,
+      "loss": 1.2827,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.460805864876927,
+      "learning_rate": 1.999354616267529e-05,
+      "loss": 1.3469,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.920177491059535,
+      "learning_rate": 1.9993446501874482e-05,
+      "loss": 1.3276,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.709488561248927,
+      "learning_rate": 1.9993346077724373e-05,
+      "loss": 1.3325,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6685922141563196,
+      "learning_rate": 1.9993244890232624e-05,
+      "loss": 1.3105,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5088249813723618,
+      "learning_rate": 1.9993142939406974e-05,
+      "loss": 1.3018,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3859127184379507,
+      "learning_rate": 1.9993040225255205e-05,
+      "loss": 1.2944,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.027407412229321,
+      "learning_rate": 1.9992936747785167e-05,
+      "loss": 1.3411,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.444983254430995,
+      "learning_rate": 1.9992832507004757e-05,
+      "loss": 1.2976,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4692353807634526,
+      "learning_rate": 1.9992727502921943e-05,
+      "loss": 1.3467,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3134763890791135,
+      "learning_rate": 1.9992621735544746e-05,
+      "loss": 1.3623,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.336001947353296,
+      "learning_rate": 1.9992515204881244e-05,
+      "loss": 1.2268,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.576214530186255,
+      "learning_rate": 1.9992407910939575e-05,
+      "loss": 1.3086,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5132613350416397,
+      "learning_rate": 1.999229985372793e-05,
+      "loss": 1.3218,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5714341042387416,
+      "learning_rate": 1.9992191033254568e-05,
+      "loss": 1.3613,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3980905227590954,
+      "learning_rate": 1.9992081449527804e-05,
+      "loss": 1.3623,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.128161864040858,
+      "learning_rate": 1.9991971102556e-05,
+      "loss": 1.3013,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4623016363756576,
+      "learning_rate": 1.9991859992347594e-05,
+      "loss": 1.4106,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5096535967300095,
+      "learning_rate": 1.9991748118911064e-05,
+      "loss": 1.3379,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.241628786859793,
+      "learning_rate": 1.9991635482254962e-05,
+      "loss": 1.2856,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1439222841994785,
+      "learning_rate": 1.9991522082387886e-05,
+      "loss": 1.3237,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.127344962847725,
+      "learning_rate": 1.9991407919318507e-05,
+      "loss": 1.3164,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.005619896431299,
+      "learning_rate": 1.9991292993055536e-05,
+      "loss": 1.3687,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.51869617519662,
+      "learning_rate": 1.9991177303607762e-05,
+      "loss": 1.2974,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.634109969092319,
+      "learning_rate": 1.9991060850984007e-05,
+      "loss": 1.3328,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.411498548016471,
+      "learning_rate": 1.999094363519318e-05,
+      "loss": 1.3267,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.558645920104119,
+      "learning_rate": 1.999082565624423e-05,
+      "loss": 1.3389,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.427134856005675,
+      "learning_rate": 1.9990706914146164e-05,
+      "loss": 1.2715,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.901689850041093,
+      "learning_rate": 1.999058740890806e-05,
+      "loss": 1.3384,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5559342312343114,
+      "learning_rate": 1.9990467140539043e-05,
+      "loss": 1.3833,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7657590419063713,
+      "learning_rate": 1.99903461090483e-05,
+      "loss": 1.3984,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1753253553822183,
+      "learning_rate": 1.9990224314445074e-05,
+      "loss": 1.3979,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.760072841497174,
+      "learning_rate": 1.999010175673867e-05,
+      "loss": 1.1562,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5010173305760737,
+      "learning_rate": 1.998997843593845e-05,
+      "loss": 1.356,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5277123061327122,
+      "learning_rate": 1.998985435205383e-05,
+      "loss": 1.3125,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.480429972781536,
+      "learning_rate": 1.9989729505094293e-05,
+      "loss": 1.2732,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1566956764492757,
+      "learning_rate": 1.9989603895069372e-05,
+      "loss": 1.1489,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8217696001796364,
+      "learning_rate": 1.9989477521988668e-05,
+      "loss": 1.3164,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.438268226774146,
+      "learning_rate": 1.998935038586183e-05,
+      "loss": 1.3413,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9974268229559033,
+      "learning_rate": 1.9989222486698564e-05,
+      "loss": 1.354,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3149496492603046,
+      "learning_rate": 1.9989093824508646e-05,
+      "loss": 1.3877,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.676306806055094,
+      "learning_rate": 1.9988964399301905e-05,
+      "loss": 1.3843,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5779996027743435,
+      "learning_rate": 1.998883421108822e-05,
+      "loss": 1.2954,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.768085338088118,
+      "learning_rate": 1.9988703259877544e-05,
+      "loss": 1.3623,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2025843996319505,
+      "learning_rate": 1.9988571545679873e-05,
+      "loss": 1.3076,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.62163747813662,
+      "learning_rate": 1.9988439068505267e-05,
+      "loss": 1.3848,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5567843222511284,
+      "learning_rate": 1.9988305828363853e-05,
+      "loss": 1.3354,
+      "step": 527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3675039237022286,
+      "learning_rate": 1.9988171825265807e-05,
+      "loss": 1.3333,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5284116251722897,
+      "learning_rate": 1.9988037059221356e-05,
+      "loss": 1.2751,
+      "step": 529
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6459053300804753,
+      "learning_rate": 1.9987901530240804e-05,
+      "loss": 1.3633,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4778872112522667,
+      "learning_rate": 1.99877652383345e-05,
+      "loss": 1.2886,
+      "step": 531
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.022848809562594,
+      "learning_rate": 1.9987628183512854e-05,
+      "loss": 1.3096,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6762561835853433,
+      "learning_rate": 1.9987490365786334e-05,
+      "loss": 1.2773,
+      "step": 533
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.600031159455667,
+      "learning_rate": 1.9987351785165465e-05,
+      "loss": 1.3298,
+      "step": 534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.104721492061442,
+      "learning_rate": 1.998721244166084e-05,
+      "loss": 1.2866,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.248148530905035,
+      "learning_rate": 1.99870723352831e-05,
+      "loss": 1.3789,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.975696848908391,
+      "learning_rate": 1.9986931466042943e-05,
+      "loss": 1.3677,
+      "step": 537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.825392673680417,
+      "learning_rate": 1.998678983395113e-05,
+      "loss": 1.2651,
+      "step": 538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7863758883350855,
+      "learning_rate": 1.9986647439018486e-05,
+      "loss": 1.2839,
+      "step": 539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5828963291232356,
+      "learning_rate": 1.998650428125588e-05,
+      "loss": 1.3003,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3141747721246273,
+      "learning_rate": 1.9986360360674252e-05,
+      "loss": 1.2729,
+      "step": 541
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.475259056888451,
+      "learning_rate": 1.9986215677284596e-05,
+      "loss": 1.3137,
+      "step": 542
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8036154290623445,
+      "learning_rate": 1.9986070231097957e-05,
+      "loss": 1.3364,
+      "step": 543
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5008311338558897,
+      "learning_rate": 1.9985924022125453e-05,
+      "loss": 1.2979,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.162076607028745,
+      "learning_rate": 1.9985777050378245e-05,
+      "loss": 1.3135,
+      "step": 545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.521854094013388,
+      "learning_rate": 1.9985629315867564e-05,
+      "loss": 1.3506,
+      "step": 546
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.12687240671492,
+      "learning_rate": 1.9985480818604698e-05,
+      "loss": 1.3579,
+      "step": 547
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.682371700469151,
+      "learning_rate": 1.998533155860098e-05,
+      "loss": 1.252,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.093688304974041,
+      "learning_rate": 1.998518153586782e-05,
+      "loss": 1.3491,
+      "step": 549
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4422471004816266,
+      "learning_rate": 1.998503075041667e-05,
+      "loss": 1.2773,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.163601161220485,
+      "learning_rate": 1.998487920225906e-05,
+      "loss": 1.3486,
+      "step": 551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.425143024754384,
+      "learning_rate": 1.998472689140655e-05,
+      "loss": 1.2988,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.452660517097362,
+      "learning_rate": 1.9984573817870787e-05,
+      "loss": 1.3237,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.430928501337027,
+      "learning_rate": 1.9984419981663457e-05,
+      "loss": 1.3391,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.60507760207044,
+      "learning_rate": 1.9984265382796315e-05,
+      "loss": 1.353,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.293889739627148,
+      "learning_rate": 1.9984110021281165e-05,
+      "loss": 1.3442,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0424489039536375,
+      "learning_rate": 1.9983953897129876e-05,
+      "loss": 1.3472,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.091042442836965,
+      "learning_rate": 1.9983797010354377e-05,
+      "loss": 1.3308,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7518240679577644,
+      "learning_rate": 1.998363936096665e-05,
+      "loss": 1.2969,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6271090366149656,
+      "learning_rate": 1.9983480948978737e-05,
+      "loss": 1.2983,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5996902361525844,
+      "learning_rate": 1.9983321774402735e-05,
+      "loss": 1.3196,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7449375123164748,
+      "learning_rate": 1.9983161837250806e-05,
+      "loss": 1.3867,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.185946778085656,
+      "learning_rate": 1.9983001137535167e-05,
+      "loss": 1.3657,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5220110566036853,
+      "learning_rate": 1.9982839675268087e-05,
+      "loss": 1.2598,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.637624289440149,
+      "learning_rate": 1.998267745046191e-05,
+      "loss": 1.301,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.288808522731118,
+      "learning_rate": 1.9982514463129022e-05,
+      "loss": 1.29,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9987869570052212,
+      "learning_rate": 1.998235071328187e-05,
+      "loss": 1.3779,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5789007027744804,
+      "learning_rate": 1.9982186200932964e-05,
+      "loss": 1.3386,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.622291240012273,
+      "learning_rate": 1.9982020926094875e-05,
+      "loss": 1.3298,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4543153977309995,
+      "learning_rate": 1.998185488878022e-05,
+      "loss": 1.3462,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4224800280984597,
+      "learning_rate": 1.9981688089001685e-05,
+      "loss": 1.2681,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.819061579668251,
+      "learning_rate": 1.9981520526772012e-05,
+      "loss": 1.354,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1679727644574776,
+      "learning_rate": 1.9981352202104003e-05,
+      "loss": 1.3745,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7755309444919987,
+      "learning_rate": 1.9981183115010508e-05,
+      "loss": 1.2949,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0380924051431144,
+      "learning_rate": 1.998101326550445e-05,
+      "loss": 1.2877,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9858786508506854,
+      "learning_rate": 1.9980842653598796e-05,
+      "loss": 1.3628,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.160639386822461,
+      "learning_rate": 1.998067127930658e-05,
+      "loss": 1.2476,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7942342815054575,
+      "learning_rate": 1.9980499142640897e-05,
+      "loss": 1.3081,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.540424652015008,
+      "learning_rate": 1.9980326243614895e-05,
+      "loss": 1.2979,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6529023781080765,
+      "learning_rate": 1.9980152582241775e-05,
+      "loss": 1.2346,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.794408136414203,
+      "learning_rate": 1.9979978158534808e-05,
+      "loss": 1.3643,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4644547102130216,
+      "learning_rate": 1.9979802972507316e-05,
+      "loss": 1.2783,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4424984984201554,
+      "learning_rate": 1.997962702417268e-05,
+      "loss": 1.2927,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3603282822066096,
+      "learning_rate": 1.9979450313544337e-05,
+      "loss": 1.2559,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.660489492350285,
+      "learning_rate": 1.997927284063579e-05,
+      "loss": 1.3154,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.073482230708938,
+      "learning_rate": 1.997909460546059e-05,
+      "loss": 1.283,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5723480720349547,
+      "learning_rate": 1.997891560803236e-05,
+      "loss": 1.3154,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8246988693085933,
+      "learning_rate": 1.9978735848364762e-05,
+      "loss": 1.3662,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4051420561407957,
+      "learning_rate": 1.9978555326471535e-05,
+      "loss": 1.2808,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.158052341038253,
+      "learning_rate": 1.9978374042366464e-05,
+      "loss": 1.311,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4322782701875822,
+      "learning_rate": 1.99781919960634e-05,
+      "loss": 1.2822,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9738406848605865,
+      "learning_rate": 1.997800918757624e-05,
+      "loss": 1.3896,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.484454914890377,
+      "learning_rate": 1.997782561691896e-05,
+      "loss": 1.345,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4526479141211373,
+      "learning_rate": 1.9977641284105576e-05,
+      "loss": 1.4058,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.923741543447377,
+      "learning_rate": 1.9977456189150164e-05,
+      "loss": 1.2791,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.859165138767623,
+      "learning_rate": 1.997727033206687e-05,
+      "loss": 1.3481,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5597173220884963,
+      "learning_rate": 1.9977083712869887e-05,
+      "loss": 1.2759,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.604070843201808,
+      "learning_rate": 1.997689633157347e-05,
+      "loss": 1.3896,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6417849845655295,
+      "learning_rate": 1.9976708188191934e-05,
+      "loss": 1.3296,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 59.146925816574836,
+      "learning_rate": 1.9976519282739648e-05,
+      "loss": 1.939,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.718152972303104,
+      "learning_rate": 1.997632961523104e-05,
+      "loss": 1.3528,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.755155291803678,
+      "learning_rate": 1.9976139185680606e-05,
+      "loss": 1.4502,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.733239418991268,
+      "learning_rate": 1.9975947994102883e-05,
+      "loss": 1.4434,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.518206546071597,
+      "learning_rate": 1.9975756040512477e-05,
+      "loss": 1.3594,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.24137089119195,
+      "learning_rate": 1.9975563324924052e-05,
+      "loss": 1.4751,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.534610193369163,
+      "learning_rate": 1.997536984735233e-05,
+      "loss": 1.3271,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.722770721833446,
+      "learning_rate": 1.9975175607812086e-05,
+      "loss": 1.4258,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.5167497501991445,
+      "learning_rate": 1.997498060631816e-05,
+      "loss": 1.3438,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.9229782944613465,
+      "learning_rate": 1.9974784842885447e-05,
+      "loss": 1.3682,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.493110313583215,
+      "learning_rate": 1.9974588317528902e-05,
+      "loss": 1.3267,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.35422741292535,
+      "learning_rate": 1.997439103026353e-05,
+      "loss": 1.3014,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.402144390980332,
+      "learning_rate": 1.9974192981104406e-05,
+      "loss": 1.4414,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.436909176422717,
+      "learning_rate": 1.9973994170066654e-05,
+      "loss": 1.3477,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.008724963238588,
+      "learning_rate": 1.9973794597165467e-05,
+      "loss": 1.417,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9905611120097104,
+      "learning_rate": 1.9973594262416087e-05,
+      "loss": 1.3625,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.677738973538602,
+      "learning_rate": 1.9973393165833813e-05,
+      "loss": 1.3853,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 17.82583130040632,
+      "learning_rate": 1.997319130743401e-05,
+      "loss": 1.6763,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.499509286206163,
+      "learning_rate": 1.9972988687232088e-05,
+      "loss": 1.3772,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.082800620520958,
+      "learning_rate": 1.9972785305243537e-05,
+      "loss": 1.374,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.51390265670667,
+      "learning_rate": 1.9972581161483888e-05,
+      "loss": 1.3745,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.360530599630574,
+      "learning_rate": 1.9972376255968728e-05,
+      "loss": 1.3511,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.131195208544769,
+      "learning_rate": 1.9972170588713715e-05,
+      "loss": 1.4292,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.123942100055345,
+      "learning_rate": 1.9971964159734557e-05,
+      "loss": 1.3389,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.563704226296613,
+      "learning_rate": 1.9971756969047024e-05,
+      "loss": 1.3535,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.36087862639203,
+      "learning_rate": 1.997154901666694e-05,
+      "loss": 1.3101,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.522059725537015,
+      "learning_rate": 1.9971340302610187e-05,
+      "loss": 1.3958,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.8259560466398606,
+      "learning_rate": 1.9971130826892714e-05,
+      "loss": 1.4199,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.500138503241586,
+      "learning_rate": 1.9970920589530517e-05,
+      "loss": 1.3799,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.082785031462788,
+      "learning_rate": 1.997070959053966e-05,
+      "loss": 1.3901,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4531534283317904,
+      "learning_rate": 1.9970497829936254e-05,
+      "loss": 1.3721,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.372291744944041,
+      "learning_rate": 1.997028530773648e-05,
+      "loss": 1.3682,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.943532836127151,
+      "learning_rate": 1.9970072023956562e-05,
+      "loss": 1.4351,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.165686039354895,
+      "learning_rate": 1.9969857978612806e-05,
+      "loss": 1.3804,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.691538889775421,
+      "learning_rate": 1.996964317172155e-05,
+      "loss": 1.4204,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.171821644572202,
+      "learning_rate": 1.9969427603299206e-05,
+      "loss": 1.292,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8832418406601543,
+      "learning_rate": 1.996921127336224e-05,
+      "loss": 1.3726,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5072663601278937,
+      "learning_rate": 1.9968994181927178e-05,
+      "loss": 1.3013,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.095698930481626,
+      "learning_rate": 1.99687763290106e-05,
+      "loss": 1.2935,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.601685936906042,
+      "learning_rate": 1.996855771462915e-05,
+      "loss": 1.283,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7364897275505453,
+      "learning_rate": 1.9968338338799525e-05,
+      "loss": 1.321,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.029183764656781,
+      "learning_rate": 1.9968118201538478e-05,
+      "loss": 1.313,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7117347677746757,
+      "learning_rate": 1.996789730286283e-05,
+      "loss": 1.3154,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1331627694841035,
+      "learning_rate": 1.996767564278945e-05,
+      "loss": 1.3071,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.338427644831074,
+      "learning_rate": 1.996745322133528e-05,
+      "loss": 1.3315,
+      "step": 644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8158767639395634,
+      "learning_rate": 1.9967230038517295e-05,
+      "loss": 1.2598,
+      "step": 645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.478345687705536,
+      "learning_rate": 1.996700609435255e-05,
+      "loss": 1.3081,
+      "step": 646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6900517434693922,
+      "learning_rate": 1.9966781388858147e-05,
+      "loss": 1.3975,
+      "step": 647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7555556360458278,
+      "learning_rate": 1.9966555922051257e-05,
+      "loss": 1.3296,
+      "step": 648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.351253249749705,
+      "learning_rate": 1.9966329693949098e-05,
+      "loss": 1.3127,
+      "step": 649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.217623098881597,
+      "learning_rate": 1.9966102704568952e-05,
+      "loss": 1.3066,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.538379599618774,
+      "learning_rate": 1.9965874953928156e-05,
+      "loss": 1.355,
+      "step": 651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4904759651105928,
+      "learning_rate": 1.9965646442044106e-05,
+      "loss": 1.2781,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3643162705667344,
+      "learning_rate": 1.9965417168934264e-05,
+      "loss": 1.3486,
+      "step": 653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2958131216804305,
+      "learning_rate": 1.9965187134616132e-05,
+      "loss": 1.3118,
+      "step": 654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.627095672885471,
+      "learning_rate": 1.9964956339107285e-05,
+      "loss": 1.3169,
+      "step": 655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1127925289378138,
+      "learning_rate": 1.9964724782425358e-05,
+      "loss": 1.2661,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5364074178350635,
+      "learning_rate": 1.996449246458803e-05,
+      "loss": 1.2419,
+      "step": 657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5056245236380312,
+      "learning_rate": 1.9964259385613053e-05,
+      "loss": 1.3413,
+      "step": 658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0958412542724205,
+      "learning_rate": 1.996402554551823e-05,
+      "loss": 1.3213,
+      "step": 659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.663224873721173,
+      "learning_rate": 1.996379094432142e-05,
+      "loss": 1.3071,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5138557501534624,
+      "learning_rate": 1.9963555582040546e-05,
+      "loss": 1.3662,
+      "step": 661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4093276641647066,
+      "learning_rate": 1.996331945869358e-05,
+      "loss": 1.312,
+      "step": 662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0652425256322764,
+      "learning_rate": 1.9963082574298566e-05,
+      "loss": 1.2136,
+      "step": 663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0228702493661315,
+      "learning_rate": 1.9962844928873598e-05,
+      "loss": 1.2866,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5256602303080817,
+      "learning_rate": 1.9962606522436824e-05,
+      "loss": 1.3708,
+      "step": 665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.572615502382351,
+      "learning_rate": 1.9962367355006457e-05,
+      "loss": 1.2925,
+      "step": 666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4384410290776675,
+      "learning_rate": 1.996212742660076e-05,
+      "loss": 1.3042,
+      "step": 667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4838904268503,
+      "learning_rate": 1.996188673723807e-05,
+      "loss": 1.1904,
+      "step": 668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1350793622082977,
+      "learning_rate": 1.9961645286936767e-05,
+      "loss": 1.2446,
+      "step": 669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7709784456092246,
+      "learning_rate": 1.9961403075715292e-05,
+      "loss": 1.2769,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7672617171580085,
+      "learning_rate": 1.996116010359215e-05,
+      "loss": 1.2856,
+      "step": 671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.079300148298796,
+      "learning_rate": 1.9960916370585904e-05,
+      "loss": 1.3613,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5044884894336157,
+      "learning_rate": 1.996067187671516e-05,
+      "loss": 1.2666,
+      "step": 673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4451509169393635,
+      "learning_rate": 1.9960426621998605e-05,
+      "loss": 1.3877,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.942020007890894,
+      "learning_rate": 1.9960180606454964e-05,
+      "loss": 1.3047,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4264371302814105,
+      "learning_rate": 1.9959933830103034e-05,
+      "loss": 1.3174,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.423719604881023,
+      "learning_rate": 1.995968629296167e-05,
+      "loss": 1.2932,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.272613532647432,
+      "learning_rate": 1.9959437995049768e-05,
+      "loss": 1.2773,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.394725622657974,
+      "learning_rate": 1.99591889363863e-05,
+      "loss": 1.3066,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4045564319457147,
+      "learning_rate": 1.9958939116990293e-05,
+      "loss": 1.301,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4290452294807543,
+      "learning_rate": 1.995868853688083e-05,
+      "loss": 1.3574,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5344359387206556,
+      "learning_rate": 1.995843719607704e-05,
+      "loss": 1.3525,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.455777386607647,
+      "learning_rate": 1.995818509459814e-05,
+      "loss": 1.3154,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3479186765307287,
+      "learning_rate": 1.995793223246337e-05,
+      "loss": 1.2859,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.197887373836333,
+      "learning_rate": 1.9957678609692056e-05,
+      "loss": 1.4739,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.197730690755586,
+      "learning_rate": 1.9957424226303567e-05,
+      "loss": 1.3242,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.382324202386118,
+      "learning_rate": 1.9957169082317333e-05,
+      "loss": 1.3188,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9784388101697106,
+      "learning_rate": 1.9956913177752846e-05,
+      "loss": 1.3579,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1006888357494637,
+      "learning_rate": 1.995665651262965e-05,
+      "loss": 1.2764,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.745667980579006,
+      "learning_rate": 1.9956399086967354e-05,
+      "loss": 1.2742,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.685658787133534,
+      "learning_rate": 1.9956140900785617e-05,
+      "loss": 1.2905,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.776545275259056,
+      "learning_rate": 1.995588195410416e-05,
+      "loss": 1.2837,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4161141282127963,
+      "learning_rate": 1.9955622246942772e-05,
+      "loss": 1.2925,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.248921820722914,
+      "learning_rate": 1.995536177932128e-05,
+      "loss": 1.2593,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5031778416765804,
+      "learning_rate": 1.9955100551259587e-05,
+      "loss": 1.2693,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6032894084191773,
+      "learning_rate": 1.9954838562777643e-05,
+      "loss": 1.2896,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2479189910934134,
+      "learning_rate": 1.9954575813895462e-05,
+      "loss": 1.3184,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.607883115112884,
+      "learning_rate": 1.995431230463311e-05,
+      "loss": 1.354,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4302872275171024,
+      "learning_rate": 1.995404803501072e-05,
+      "loss": 1.3062,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.734189295241357,
+      "learning_rate": 1.9953783005048478e-05,
+      "loss": 1.2578,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4796471645563716,
+      "learning_rate": 1.9953517214766625e-05,
+      "loss": 1.3716,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1255761090849767,
+      "learning_rate": 1.9953250664185467e-05,
+      "loss": 1.3135,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2639949733963167,
+      "learning_rate": 1.9952983353325358e-05,
+      "loss": 1.2085,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5781068712924653,
+      "learning_rate": 1.9952715282206727e-05,
+      "loss": 1.4067,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.747593261111076,
+      "learning_rate": 1.9952446450850042e-05,
+      "loss": 1.2878,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.883361557276629,
+      "learning_rate": 1.995217685927584e-05,
+      "loss": 1.3452,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.231300162545337,
+      "learning_rate": 1.9951906507504714e-05,
+      "loss": 1.3025,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5180181270333626,
+      "learning_rate": 1.9951635395557315e-05,
+      "loss": 1.3335,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.141459896426009,
+      "learning_rate": 1.9951363523454353e-05,
+      "loss": 1.2761,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4032706262527244,
+      "learning_rate": 1.9951090891216593e-05,
+      "loss": 1.2354,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.779990226638794,
+      "learning_rate": 1.9950817498864864e-05,
+      "loss": 1.3569,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7887309587462323,
+      "learning_rate": 1.9950543346420042e-05,
+      "loss": 1.3555,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6491427791579056,
+      "learning_rate": 1.9950268433903073e-05,
+      "loss": 1.3115,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.894864701735775,
+      "learning_rate": 1.9949992761334953e-05,
+      "loss": 1.4211,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.193737527682885,
+      "learning_rate": 1.9949716328736746e-05,
+      "loss": 1.3113,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.446080660254638,
+      "learning_rate": 1.994943913612956e-05,
+      "loss": 1.2744,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8986128868429737,
+      "learning_rate": 1.9949161183534572e-05,
+      "loss": 1.3408,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7874346731590487,
+      "learning_rate": 1.994888247097301e-05,
+      "loss": 1.3035,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.455189401609391,
+      "learning_rate": 1.994860299846617e-05,
+      "loss": 1.2388,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4676175149579964,
+      "learning_rate": 1.9948322766035395e-05,
+      "loss": 1.3069,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.377923070774838,
+      "learning_rate": 1.994804177370209e-05,
+      "loss": 1.312,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6494323854034922,
+      "learning_rate": 1.994776002148772e-05,
+      "loss": 1.3306,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.363898809431975,
+      "learning_rate": 1.9947477509413806e-05,
+      "loss": 1.24,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.997920885218377,
+      "learning_rate": 1.9947194237501928e-05,
+      "loss": 1.3042,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.632742555903756,
+      "learning_rate": 1.9946910205773723e-05,
+      "loss": 1.334,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7990841451978503,
+      "learning_rate": 1.994662541425089e-05,
+      "loss": 1.3638,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.615648015943497,
+      "learning_rate": 1.9946339862955177e-05,
+      "loss": 1.3398,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.594044187544689,
+      "learning_rate": 1.9946053551908403e-05,
+      "loss": 1.3047,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5541636458255876,
+      "learning_rate": 1.9945766481132432e-05,
+      "loss": 1.3979,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.458113952721757,
+      "learning_rate": 1.9945478650649192e-05,
+      "loss": 1.3308,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7494308862494017,
+      "learning_rate": 1.9945190060480676e-05,
+      "loss": 1.2795,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.116449900412526,
+      "learning_rate": 1.994490071064892e-05,
+      "loss": 1.291,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2348832035530863,
+      "learning_rate": 1.9944610601176023e-05,
+      "loss": 1.2842,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.859497643657222,
+      "learning_rate": 1.9944319732084156e-05,
+      "loss": 1.3059,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.589146900824583,
+      "learning_rate": 1.994402810339553e-05,
+      "loss": 1.3477,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.343194916588298,
+      "learning_rate": 1.9943735715132427e-05,
+      "loss": 1.3533,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.723903714306857,
+      "learning_rate": 1.9943442567317173e-05,
+      "loss": 1.3203,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.414508484788855,
+      "learning_rate": 1.9943148659972165e-05,
+      "loss": 1.2583,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.563986616307243,
+      "learning_rate": 1.9942853993119852e-05,
+      "loss": 1.2974,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.322710076743036,
+      "learning_rate": 1.9942558566782745e-05,
+      "loss": 1.2715,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.586257878343499,
+      "learning_rate": 1.9942262380983405e-05,
+      "loss": 1.2104,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.127072301669496,
+      "learning_rate": 1.9941965435744455e-05,
+      "loss": 1.2439,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.5947238368119105,
+      "learning_rate": 1.9941667731088586e-05,
+      "loss": 1.3665,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.484299595967397,
+      "learning_rate": 1.994136926703853e-05,
+      "loss": 1.3271,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.521065321211717,
+      "learning_rate": 1.994107004361709e-05,
+      "loss": 1.4395,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.360018157169895,
+      "learning_rate": 1.994077006084712e-05,
+      "loss": 1.3765,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.2954068036991595,
+      "learning_rate": 1.994046931875153e-05,
+      "loss": 1.3652,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.8941217378277315,
+      "learning_rate": 1.99401678173533e-05,
+      "loss": 1.3154,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.130643583627689,
+      "learning_rate": 1.993986555667546e-05,
+      "loss": 1.3342,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.512752676429674,
+      "learning_rate": 1.9939562536741093e-05,
+      "loss": 1.3608,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.78545039173749,
+      "learning_rate": 1.9939258757573345e-05,
+      "loss": 1.2642,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.7952163041381075,
+      "learning_rate": 1.9938954219195424e-05,
+      "loss": 1.3171,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.02047902586675,
+      "learning_rate": 1.9938648921630587e-05,
+      "loss": 1.212,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8901328822648846,
+      "learning_rate": 1.993834286490216e-05,
+      "loss": 1.3232,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.02779344629026,
+      "learning_rate": 1.993803604903352e-05,
+      "loss": 1.3081,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.32330131027177,
+      "learning_rate": 1.9937728474048098e-05,
+      "loss": 1.3291,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.63123670262054,
+      "learning_rate": 1.9937420139969397e-05,
+      "loss": 1.3726,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.016075660042374,
+      "learning_rate": 1.9937111046820962e-05,
+      "loss": 1.3091,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.332603124140773,
+      "learning_rate": 1.9936801194626404e-05,
+      "loss": 1.3081,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6201653989273046,
+      "learning_rate": 1.9936490583409393e-05,
+      "loss": 1.3157,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.620492907836736,
+      "learning_rate": 1.9936179213193653e-05,
+      "loss": 1.364,
+      "step": 761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.7022963838287914,
+      "learning_rate": 1.9935867084002974e-05,
+      "loss": 1.3013,
+      "step": 762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.527838010947892,
+      "learning_rate": 1.9935554195861187e-05,
+      "loss": 1.2256,
+      "step": 763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.152891369664262,
+      "learning_rate": 1.9935240548792204e-05,
+      "loss": 1.3398,
+      "step": 764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.112300410797279,
+      "learning_rate": 1.9934926142819972e-05,
+      "loss": 1.3284,
+      "step": 765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.47630474254814,
+      "learning_rate": 1.9934610977968514e-05,
+      "loss": 1.2705,
+      "step": 766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.517191731710104,
+      "learning_rate": 1.9934295054261903e-05,
+      "loss": 1.2854,
+      "step": 767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8092929305905012,
+      "learning_rate": 1.9933978371724267e-05,
+      "loss": 1.3823,
+      "step": 768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4826014495704953,
+      "learning_rate": 1.9933660930379803e-05,
+      "loss": 1.2668,
+      "step": 769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7743287157646956,
+      "learning_rate": 1.9933342730252748e-05,
+      "loss": 1.2327,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8766434870143702,
+      "learning_rate": 1.993302377136742e-05,
+      "loss": 1.292,
+      "step": 771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.575230702045649,
+      "learning_rate": 1.993270405374817e-05,
+      "loss": 1.2146,
+      "step": 772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.536698099314745,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 1.2476,
+      "step": 773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.207773035192295,
+      "learning_rate": 1.9932062342405677e-05,
+      "loss": 1.354,
+      "step": 774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9226441907238767,
+      "learning_rate": 1.9931740348731447e-05,
+      "loss": 1.2098,
+      "step": 775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.022740672702353,
+      "learning_rate": 1.9931417596421333e-05,
+      "loss": 1.3252,
+      "step": 776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7076879539954466,
+      "learning_rate": 1.9931094085499993e-05,
+      "loss": 1.2578,
+      "step": 777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3975402129799046,
+      "learning_rate": 1.9930769815992138e-05,
+      "loss": 1.192,
+      "step": 778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.939311556401114,
+      "learning_rate": 1.9930444787922535e-05,
+      "loss": 1.3101,
+      "step": 779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6705771696603504,
+      "learning_rate": 1.9930119001316014e-05,
+      "loss": 1.2695,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3241437851424305,
+      "learning_rate": 1.9929792456197456e-05,
+      "loss": 1.2559,
+      "step": 781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.51806266699757,
+      "learning_rate": 1.9929465152591807e-05,
+      "loss": 1.2998,
+      "step": 782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3626777250589184,
+      "learning_rate": 1.9929137090524068e-05,
+      "loss": 1.2666,
+      "step": 783
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9131192169506956,
+      "learning_rate": 1.9928808270019297e-05,
+      "loss": 1.353,
+      "step": 784
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.024253542211286,
+      "learning_rate": 1.9928478691102612e-05,
+      "loss": 1.376,
+      "step": 785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.360519015056924,
+      "learning_rate": 1.992814835379919e-05,
+      "loss": 1.2661,
+      "step": 786
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5353277220158477,
+      "learning_rate": 1.992781725813426e-05,
+      "loss": 1.2712,
+      "step": 787
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.263581238607639,
+      "learning_rate": 1.992748540413311e-05,
+      "loss": 1.4458,
+      "step": 788
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4440915485018824,
+      "learning_rate": 1.9927152791821097e-05,
+      "loss": 1.2654,
+      "step": 789
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.36623174396987,
+      "learning_rate": 1.9926819421223622e-05,
+      "loss": 1.3438,
+      "step": 790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9898093034814814,
+      "learning_rate": 1.992648529236615e-05,
+      "loss": 1.2537,
+      "step": 791
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2090868463367217,
+      "learning_rate": 1.992615040527421e-05,
+      "loss": 1.29,
+      "step": 792
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.820932105979316,
+      "learning_rate": 1.992581475997337e-05,
+      "loss": 1.2891,
+      "step": 793
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.57384499161664,
+      "learning_rate": 1.9925478356489277e-05,
+      "loss": 1.3062,
+      "step": 794
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7429043001457667,
+      "learning_rate": 1.9925141194847626e-05,
+      "loss": 1.2622,
+      "step": 795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.143271005178086,
+      "learning_rate": 1.9924803275074167e-05,
+      "loss": 1.2979,
+      "step": 796
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.573592195026204,
+      "learning_rate": 1.9924464597194718e-05,
+      "loss": 1.2998,
+      "step": 797
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.195932250965072,
+      "learning_rate": 1.9924125161235147e-05,
+      "loss": 1.3101,
+      "step": 798
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5235212757772554,
+      "learning_rate": 1.9923784967221376e-05,
+      "loss": 1.2515,
+      "step": 799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5532348824464073,
+      "learning_rate": 1.99234440151794e-05,
+      "loss": 1.3335,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.297248130232666,
+      "learning_rate": 1.9923102305135255e-05,
+      "loss": 1.2605,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0198705019681014,
+      "learning_rate": 1.992275983711505e-05,
+      "loss": 1.3018,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.091036398958955,
+      "learning_rate": 1.992241661114494e-05,
+      "loss": 1.3042,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2637216553781045,
+      "learning_rate": 1.992207262725114e-05,
+      "loss": 1.3394,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3521515727604987,
+      "learning_rate": 1.992172788545993e-05,
+      "loss": 1.3506,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5090267285110954,
+      "learning_rate": 1.9921382385797635e-05,
+      "loss": 1.248,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.905058548422149,
+      "learning_rate": 1.9921036128290658e-05,
+      "loss": 1.2939,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.915025510738342,
+      "learning_rate": 1.9920689112965434e-05,
+      "loss": 1.3374,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4730766373681536,
+      "learning_rate": 1.9920341339848482e-05,
+      "loss": 1.3193,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1746803807623185,
+      "learning_rate": 1.9919992808966366e-05,
+      "loss": 1.2512,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3741304492403947,
+      "learning_rate": 1.9919643520345698e-05,
+      "loss": 1.3379,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2490767483741876,
+      "learning_rate": 1.991929347401317e-05,
+      "loss": 1.3059,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.099209277994612,
+      "learning_rate": 1.9918942669995514e-05,
+      "loss": 1.2842,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2359118798248785,
+      "learning_rate": 1.9918591108319523e-05,
+      "loss": 1.2686,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2278517493706724,
+      "learning_rate": 1.991823878901206e-05,
+      "loss": 1.2744,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.860481254117778,
+      "learning_rate": 1.991788571210003e-05,
+      "loss": 1.3486,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5921104958277636,
+      "learning_rate": 1.9917531877610408e-05,
+      "loss": 1.3232,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4819112856434766,
+      "learning_rate": 1.9917177285570216e-05,
+      "loss": 1.2292,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.540702900961756,
+      "learning_rate": 1.9916821936006545e-05,
+      "loss": 1.3076,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.804189675384878,
+      "learning_rate": 1.9916465828946533e-05,
+      "loss": 1.3813,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.097872518167074,
+      "learning_rate": 1.9916108964417388e-05,
+      "loss": 1.2954,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.645543015762509,
+      "learning_rate": 1.991575134244636e-05,
+      "loss": 1.344,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.194334701600823,
+      "learning_rate": 1.9915392963060775e-05,
+      "loss": 1.2043,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.503518197224243,
+      "learning_rate": 1.9915033826288e-05,
+      "loss": 1.2908,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.817893434228164,
+      "learning_rate": 1.9914673932155473e-05,
+      "loss": 1.3274,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5100547112111053,
+      "learning_rate": 1.991431328069068e-05,
+      "loss": 1.3062,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2721760069961583,
+      "learning_rate": 1.9913951871921176e-05,
+      "loss": 1.2615,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2237115243387056,
+      "learning_rate": 1.9913589705874563e-05,
+      "loss": 1.3091,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.125534204703761,
+      "learning_rate": 1.9913226782578506e-05,
+      "loss": 1.2817,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3845710035327023,
+      "learning_rate": 1.9912863102060727e-05,
+      "loss": 1.2651,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.355768960516939,
+      "learning_rate": 1.9912498664349002e-05,
+      "loss": 1.2939,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5062359944290575,
+      "learning_rate": 1.9912133469471174e-05,
+      "loss": 1.3618,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.194912438892329,
+      "learning_rate": 1.9911767517455132e-05,
+      "loss": 1.2842,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.723732969091088,
+      "learning_rate": 1.9911400808328838e-05,
+      "loss": 1.3022,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.01242048611336,
+      "learning_rate": 1.9911033342120297e-05,
+      "loss": 1.2871,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.04898962907238,
+      "learning_rate": 1.991066511885758e-05,
+      "loss": 1.1948,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1401221676570747,
+      "learning_rate": 1.9910296138568814e-05,
+      "loss": 1.3105,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.319352519298907,
+      "learning_rate": 1.990992640128218e-05,
+      "loss": 1.2185,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.497544242140315,
+      "learning_rate": 1.9909555907025923e-05,
+      "loss": 1.23,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.873019153479384,
+      "learning_rate": 1.9909184655828346e-05,
+      "loss": 1.2537,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3578570587610774,
+      "learning_rate": 1.9908812647717805e-05,
+      "loss": 1.231,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.468922963668296,
+      "learning_rate": 1.990843988272271e-05,
+      "loss": 1.3171,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3706808485189566,
+      "learning_rate": 1.9908066360871544e-05,
+      "loss": 1.3018,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.011511543984483,
+      "learning_rate": 1.9907692082192832e-05,
+      "loss": 1.2227,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0118089753921353,
+      "learning_rate": 1.9907317046715166e-05,
+      "loss": 1.2852,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.129955461703357,
+      "learning_rate": 1.9906941254467193e-05,
+      "loss": 1.2102,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.353069725535377,
+      "learning_rate": 1.9906564705477616e-05,
+      "loss": 1.2983,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.378276176774129,
+      "learning_rate": 1.9906187399775198e-05,
+      "loss": 1.3027,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.373546237916313,
+      "learning_rate": 1.9905809337388764e-05,
+      "loss": 1.2524,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3423619122883537,
+      "learning_rate": 1.9905430518347185e-05,
+      "loss": 1.2441,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.073833961501878,
+      "learning_rate": 1.9905050942679406e-05,
+      "loss": 1.2703,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.317397626002098,
+      "learning_rate": 1.9904670610414413e-05,
+      "loss": 1.2439,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1921965422772987,
+      "learning_rate": 1.990428952158126e-05,
+      "loss": 1.2351,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0677039150051457,
+      "learning_rate": 1.990390767620906e-05,
+      "loss": 1.2756,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.963447827078108,
+      "learning_rate": 1.9903525074326974e-05,
+      "loss": 1.2639,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9175144069086287,
+      "learning_rate": 1.9903141715964233e-05,
+      "loss": 1.2578,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.567327601430836,
+      "learning_rate": 1.9902757601150112e-05,
+      "loss": 1.29,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.931445891706659,
+      "learning_rate": 1.990237272991396e-05,
+      "loss": 1.1841,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7076917617354272,
+      "learning_rate": 1.9901987102285172e-05,
+      "loss": 1.2974,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.159632628803787,
+      "learning_rate": 1.9901600718293205e-05,
+      "loss": 1.3774,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3384280243703146,
+      "learning_rate": 1.9901213577967572e-05,
+      "loss": 1.2524,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.325817860430279,
+      "learning_rate": 1.9900825681337842e-05,
+      "loss": 1.2832,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0957882794241804,
+      "learning_rate": 1.990043702843365e-05,
+      "loss": 1.2861,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1046103328700574,
+      "learning_rate": 1.9900047619284684e-05,
+      "loss": 1.3164,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2729027762028218,
+      "learning_rate": 1.989965745392068e-05,
+      "loss": 1.3174,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.023816777379861,
+      "learning_rate": 1.9899266532371447e-05,
+      "loss": 1.3262,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.325566142625954,
+      "learning_rate": 1.9898874854666848e-05,
+      "loss": 1.2871,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3875514802944764,
+      "learning_rate": 1.9898482420836797e-05,
+      "loss": 1.2871,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.335133948662597,
+      "learning_rate": 1.989808923091127e-05,
+      "loss": 1.332,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.058470252772668,
+      "learning_rate": 1.9897695284920305e-05,
+      "loss": 1.3008,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4404534282374124,
+      "learning_rate": 1.9897300582893986e-05,
+      "loss": 1.2739,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4590309015814222,
+      "learning_rate": 1.9896905124862475e-05,
+      "loss": 1.3218,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.000657187455322,
+      "learning_rate": 1.9896508910855962e-05,
+      "loss": 1.3062,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5110859357098545,
+      "learning_rate": 1.9896111940904724e-05,
+      "loss": 1.2368,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4879947554181685,
+      "learning_rate": 1.989571421503908e-05,
+      "loss": 1.3687,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4276366763306614,
+      "learning_rate": 1.9895315733289413e-05,
+      "loss": 1.2642,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7994163387526885,
+      "learning_rate": 1.989491649568616e-05,
+      "loss": 1.188,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9690421918235566,
+      "learning_rate": 1.9894516502259812e-05,
+      "loss": 1.2285,
+      "step": 878
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3530431624573103,
+      "learning_rate": 1.9894115753040927e-05,
+      "loss": 1.3198,
+      "step": 879
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0995568422434094,
+      "learning_rate": 1.9893714248060118e-05,
+      "loss": 1.272,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.825296559258712,
+      "learning_rate": 1.9893311987348046e-05,
+      "loss": 1.3196,
+      "step": 881
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.316161589814484,
+      "learning_rate": 1.9892908970935445e-05,
+      "loss": 1.3491,
+      "step": 882
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.006114983705239,
+      "learning_rate": 1.9892505198853097e-05,
+      "loss": 1.3071,
+      "step": 883
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.766859325888923,
+      "learning_rate": 1.9892100671131843e-05,
+      "loss": 1.3721,
+      "step": 884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5280234813550257,
+      "learning_rate": 1.9891695387802587e-05,
+      "loss": 1.3066,
+      "step": 885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.872719067682987,
+      "learning_rate": 1.9891289348896283e-05,
+      "loss": 1.2671,
+      "step": 886
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.102386900586813,
+      "learning_rate": 1.989088255444395e-05,
+      "loss": 1.2856,
+      "step": 887
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.320995807039149,
+      "learning_rate": 1.9890475004476654e-05,
+      "loss": 1.2678,
+      "step": 888
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4800840548121075,
+      "learning_rate": 1.989006669902553e-05,
+      "loss": 1.2949,
+      "step": 889
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3620149815572615,
+      "learning_rate": 1.9889657638121767e-05,
+      "loss": 1.3345,
+      "step": 890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3537448211534673,
+      "learning_rate": 1.9889247821796608e-05,
+      "loss": 1.2605,
+      "step": 891
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7709393932303477,
+      "learning_rate": 1.988883725008136e-05,
+      "loss": 1.2627,
+      "step": 892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4114176965403176,
+      "learning_rate": 1.9888425923007387e-05,
+      "loss": 1.3232,
+      "step": 893
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.999117753739152,
+      "learning_rate": 1.98880138406061e-05,
+      "loss": 1.2219,
+      "step": 894
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1894532953824095,
+      "learning_rate": 1.9887601002908984e-05,
+      "loss": 1.2253,
+      "step": 895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5256338693387685,
+      "learning_rate": 1.9887187409947568e-05,
+      "loss": 1.2993,
+      "step": 896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.676803503094558,
+      "learning_rate": 1.988677306175345e-05,
+      "loss": 1.2808,
+      "step": 897
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.235620428330648,
+      "learning_rate": 1.9886357958358276e-05,
+      "loss": 1.3057,
+      "step": 898
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6349415152360196,
+      "learning_rate": 1.988594209979375e-05,
+      "loss": 1.3015,
+      "step": 899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.605250983568956,
+      "learning_rate": 1.9885525486091642e-05,
+      "loss": 1.3296,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2464278946591163,
+      "learning_rate": 1.988510811728378e-05,
+      "loss": 1.2295,
+      "step": 901
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.328912316972351,
+      "learning_rate": 1.988468999340204e-05,
+      "loss": 1.2871,
+      "step": 902
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6200064678718467,
+      "learning_rate": 1.9884271114478352e-05,
+      "loss": 1.293,
+      "step": 903
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1133109144247806,
+      "learning_rate": 1.9883851480544726e-05,
+      "loss": 1.3091,
+      "step": 904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.105710936645344,
+      "learning_rate": 1.9883431091633208e-05,
+      "loss": 1.1487,
+      "step": 905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8645887148846207,
+      "learning_rate": 1.988300994777591e-05,
+      "loss": 1.3643,
+      "step": 906
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3380608339477846,
+      "learning_rate": 1.9882588049005002e-05,
+      "loss": 1.252,
+      "step": 907
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.071316308694184,
+      "learning_rate": 1.9882165395352708e-05,
+      "loss": 1.2637,
+      "step": 908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2858744275253198,
+      "learning_rate": 1.988174198685132e-05,
+      "loss": 1.2415,
+      "step": 909
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.985141204518322,
+      "learning_rate": 1.9881317823533176e-05,
+      "loss": 1.3408,
+      "step": 910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.038114134333737,
+      "learning_rate": 1.988089290543067e-05,
+      "loss": 1.2156,
+      "step": 911
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2974048725139196,
+      "learning_rate": 1.9880467232576267e-05,
+      "loss": 1.2764,
+      "step": 912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.092687914821079,
+      "learning_rate": 1.988004080500248e-05,
+      "loss": 1.2371,
+      "step": 913
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9457437420028967,
+      "learning_rate": 1.9879613622741876e-05,
+      "loss": 1.3032,
+      "step": 914
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9555520086159923,
+      "learning_rate": 1.9879185685827097e-05,
+      "loss": 1.2563,
+      "step": 915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.754853436317803,
+      "learning_rate": 1.987875699429082e-05,
+      "loss": 1.3081,
+      "step": 916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.903986345135992,
+      "learning_rate": 1.9878327548165798e-05,
+      "loss": 1.2456,
+      "step": 917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.175005792152339,
+      "learning_rate": 1.9877897347484828e-05,
+      "loss": 1.2817,
+      "step": 918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.483539808896285,
+      "learning_rate": 1.9877466392280773e-05,
+      "loss": 1.2056,
+      "step": 919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8765795243032417,
+      "learning_rate": 1.9877034682586555e-05,
+      "loss": 1.2397,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2942291454710624,
+      "learning_rate": 1.9876602218435148e-05,
+      "loss": 1.2522,
+      "step": 921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1003059355394047,
+      "learning_rate": 1.9876168999859585e-05,
+      "loss": 1.2783,
+      "step": 922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1975723540672893,
+      "learning_rate": 1.9875735026892956e-05,
+      "loss": 1.2344,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6065504891320592,
+      "learning_rate": 1.9875300299568414e-05,
+      "loss": 1.2827,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3798783404781685,
+      "learning_rate": 1.987486481791916e-05,
+      "loss": 1.2251,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4437816592726134,
+      "learning_rate": 1.9874428581978463e-05,
+      "loss": 1.3198,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.507957917849348,
+      "learning_rate": 1.9873991591779645e-05,
+      "loss": 1.1582,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5241172483220624,
+      "learning_rate": 1.9873553847356087e-05,
+      "loss": 1.2278,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.137685169597381,
+      "learning_rate": 1.9873115348741217e-05,
+      "loss": 1.3179,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.32024001432353,
+      "learning_rate": 1.9872676095968543e-05,
+      "loss": 1.24,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.402144520671536,
+      "learning_rate": 1.98722360890716e-05,
+      "loss": 1.2974,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.240092499624805,
+      "learning_rate": 1.9871795328084017e-05,
+      "loss": 1.3416,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.278847584954337,
+      "learning_rate": 1.987135381303945e-05,
+      "loss": 1.2612,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3691336837641765,
+      "learning_rate": 1.9870911543971626e-05,
+      "loss": 1.2825,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.607021902755737,
+      "learning_rate": 1.9870468520914327e-05,
+      "loss": 1.25,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.102035251120789,
+      "learning_rate": 1.9870024743901397e-05,
+      "loss": 1.3154,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.281483003810151,
+      "learning_rate": 1.986958021296673e-05,
+      "loss": 1.2944,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2579302728614534,
+      "learning_rate": 1.9869134928144286e-05,
+      "loss": 1.2078,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.840544955055379,
+      "learning_rate": 1.9868688889468074e-05,
+      "loss": 1.2278,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9832587435743547,
+      "learning_rate": 1.9868242096972166e-05,
+      "loss": 1.2932,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5095765756014816,
+      "learning_rate": 1.9867794550690688e-05,
+      "loss": 1.3042,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9470297157431586,
+      "learning_rate": 1.986734625065783e-05,
+      "loss": 1.2107,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4555974938445724,
+      "learning_rate": 1.986689719690783e-05,
+      "loss": 1.3262,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.192876869767198,
+      "learning_rate": 1.9866447389474997e-05,
+      "loss": 1.3052,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2145378460023974,
+      "learning_rate": 1.9865996828393684e-05,
+      "loss": 1.2576,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5328921839683467,
+      "learning_rate": 1.9865545513698304e-05,
+      "loss": 1.2988,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4834748502509574,
+      "learning_rate": 1.986509344542334e-05,
+      "loss": 1.2168,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.952199831830398,
+      "learning_rate": 1.9864640623603314e-05,
+      "loss": 1.1707,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8179253519322605,
+      "learning_rate": 1.986418704827282e-05,
+      "loss": 1.3066,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8568092918802934,
+      "learning_rate": 1.98637327194665e-05,
+      "loss": 1.2363,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4974870706403007,
+      "learning_rate": 1.9863277637219068e-05,
+      "loss": 1.3105,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2623102796705674,
+      "learning_rate": 1.9862821801565272e-05,
+      "loss": 1.2415,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0719324093112883,
+      "learning_rate": 1.986236521253994e-05,
+      "loss": 1.3213,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2975182465074844,
+      "learning_rate": 1.9861907870177946e-05,
+      "loss": 1.2593,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.30444150317971,
+      "learning_rate": 1.9861449774514225e-05,
+      "loss": 1.312,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7848327720430754,
+      "learning_rate": 1.9860990925583765e-05,
+      "loss": 1.2319,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4246599452604713,
+      "learning_rate": 1.986053132342162e-05,
+      "loss": 1.3569,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.424597055402796,
+      "learning_rate": 1.986007096806289e-05,
+      "loss": 1.2463,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0854210428954763,
+      "learning_rate": 1.9859609859542746e-05,
+      "loss": 1.2402,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7426621048012407,
+      "learning_rate": 1.985914799789641e-05,
+      "loss": 1.3354,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6203734394920204,
+      "learning_rate": 1.9858685383159158e-05,
+      "loss": 1.3103,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5121797076387122,
+      "learning_rate": 1.9858222015366325e-05,
+      "loss": 1.2681,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2376806346527007,
+      "learning_rate": 1.9857757894553312e-05,
+      "loss": 1.2192,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.969677963285679,
+      "learning_rate": 1.9857293020755565e-05,
+      "loss": 1.3428,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4003034189655765,
+      "learning_rate": 1.9856827394008593e-05,
+      "loss": 1.2344,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6528709806260826,
+      "learning_rate": 1.9856361014347967e-05,
+      "loss": 1.2043,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7749546914681513,
+      "learning_rate": 1.985589388180931e-05,
+      "loss": 1.1877,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.838787113442408,
+      "learning_rate": 1.9855425996428303e-05,
+      "loss": 1.2349,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.465152821577194,
+      "learning_rate": 1.9854957358240684e-05,
+      "loss": 1.2593,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.381117148615985,
+      "learning_rate": 1.9854487967282256e-05,
+      "loss": 1.3052,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9523548909324564,
+      "learning_rate": 1.9854017823588867e-05,
+      "loss": 1.3706,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.384672352248049,
+      "learning_rate": 1.985354692719643e-05,
+      "loss": 1.2766,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.344345755125963,
+      "learning_rate": 1.9853075278140913e-05,
+      "loss": 1.311,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.394186491024698,
+      "learning_rate": 1.9852602876458353e-05,
+      "loss": 1.2202,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.031203801746887,
+      "learning_rate": 1.985212972218482e-05,
+      "loss": 1.3057,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6825047987694544,
+      "learning_rate": 1.9851655815356468e-05,
+      "loss": 1.2407,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7231944840629088,
+      "learning_rate": 1.9851181156009485e-05,
+      "loss": 1.2671,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.083479510343513,
+      "learning_rate": 1.985070574418014e-05,
+      "loss": 1.2446,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3569578742796438,
+      "learning_rate": 1.9850229579904737e-05,
+      "loss": 1.2007,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8205211647995423,
+      "learning_rate": 1.9849752663219655e-05,
+      "loss": 1.2803,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4556867877139883,
+      "learning_rate": 1.9849274994161315e-05,
+      "loss": 1.2698,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.613846226287988,
+      "learning_rate": 1.9848796572766214e-05,
+      "loss": 1.3008,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6611868648807464,
+      "learning_rate": 1.984831739907089e-05,
+      "loss": 1.2896,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6671201660742554,
+      "learning_rate": 1.9847837473111945e-05,
+      "loss": 1.3047,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4147443747848176,
+      "learning_rate": 1.9847356794926044e-05,
+      "loss": 1.2791,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0090049854514636,
+      "learning_rate": 1.9846875364549895e-05,
+      "loss": 1.2839,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9457639367482025,
+      "learning_rate": 1.9846393182020277e-05,
+      "loss": 1.3484,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.714731189153642,
+      "learning_rate": 1.9845910247374017e-05,
+      "loss": 1.29,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.406158469823035,
+      "learning_rate": 1.984542656064801e-05,
+      "loss": 1.2827,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7555021435627003,
+      "learning_rate": 1.9844942121879197e-05,
+      "loss": 1.2908,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4386069737031604,
+      "learning_rate": 1.9844456931104588e-05,
+      "loss": 1.2856,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7078903508412795,
+      "learning_rate": 1.9843970988361238e-05,
+      "loss": 1.2505,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.511534267253737,
+      "learning_rate": 1.984348429368627e-05,
+      "loss": 1.3374,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.26828703253792,
+      "learning_rate": 1.9842996847116856e-05,
+      "loss": 1.2017,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4402065204719805,
+      "learning_rate": 1.984250864869024e-05,
+      "loss": 1.3228,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3463209503021347,
+      "learning_rate": 1.9842019698443696e-05,
+      "loss": 1.3511,
+      "step": 996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8013071028071717,
+      "learning_rate": 1.9841529996414586e-05,
+      "loss": 1.2354,
+      "step": 997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6400251250556175,
+      "learning_rate": 1.9841039542640313e-05,
+      "loss": 1.3613,
+      "step": 998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0161528326142824,
+      "learning_rate": 1.9840548337158336e-05,
+      "loss": 1.2529,
+      "step": 999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.080849249832697,
+      "learning_rate": 1.984005638000618e-05,
+      "loss": 1.1602,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2264078795375215,
+      "learning_rate": 1.983956367122142e-05,
+      "loss": 1.2568,
+      "step": 1001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.585336935900105,
+      "learning_rate": 1.98390702108417e-05,
+      "loss": 1.2693,
+      "step": 1002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.604046127765601,
+      "learning_rate": 1.98385759989047e-05,
+      "loss": 1.3064,
+      "step": 1003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.57480118883753,
+      "learning_rate": 1.9838081035448177e-05,
+      "loss": 1.2783,
+      "step": 1004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.390487805735371,
+      "learning_rate": 1.983758532050994e-05,
+      "loss": 1.2781,
+      "step": 1005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.638740217946106,
+      "learning_rate": 1.9837088854127852e-05,
+      "loss": 1.3052,
+      "step": 1006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6230745555483472,
+      "learning_rate": 1.983659163633984e-05,
+      "loss": 1.249,
+      "step": 1007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3554872875484265,
+      "learning_rate": 1.9836093667183874e-05,
+      "loss": 1.2319,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.324670296588059,
+      "learning_rate": 1.9835594946698e-05,
+      "loss": 1.2505,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.345968419383559,
+      "learning_rate": 1.983509547492031e-05,
+      "loss": 1.2769,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5482297049274645,
+      "learning_rate": 1.9834595251888958e-05,
+      "loss": 1.2883,
+      "step": 1011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.511845277441789,
+      "learning_rate": 1.9834094277642154e-05,
+      "loss": 1.2485,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.509680792335773,
+      "learning_rate": 1.983359255221816e-05,
+      "loss": 1.3193,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.30141974825223,
+      "learning_rate": 1.9833090075655303e-05,
+      "loss": 1.3088,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.756618101137357,
+      "learning_rate": 1.9832586847991964e-05,
+      "loss": 1.3052,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5596355068253858,
+      "learning_rate": 1.9832082869266583e-05,
+      "loss": 1.2107,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4444182583213703,
+      "learning_rate": 1.983157813951766e-05,
+      "loss": 1.2437,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8719677589017003,
+      "learning_rate": 1.9831072658783745e-05,
+      "loss": 1.2012,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7931156026892254,
+      "learning_rate": 1.9830566427103445e-05,
+      "loss": 1.2805,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2953252286843795,
+      "learning_rate": 1.9830059444515436e-05,
+      "loss": 1.3345,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4284748982414803,
+      "learning_rate": 1.982955171105844e-05,
+      "loss": 1.2646,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.709496708642948,
+      "learning_rate": 1.982904322677124e-05,
+      "loss": 1.2783,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5513429299768977,
+      "learning_rate": 1.9828533991692675e-05,
+      "loss": 1.2622,
+      "step": 1023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7495409927983956,
+      "learning_rate": 1.982802400586165e-05,
+      "loss": 1.2769,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7547128618444825,
+      "learning_rate": 1.9827513269317112e-05,
+      "loss": 1.314,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4001044589675087,
+      "learning_rate": 1.982700178209808e-05,
+      "loss": 1.3276,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.314105990585137,
+      "learning_rate": 1.9826489544243623e-05,
+      "loss": 1.2798,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4845282133110227,
+      "learning_rate": 1.9825976555792863e-05,
+      "loss": 1.3118,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.435269661487241,
+      "learning_rate": 1.9825462816784985e-05,
+      "loss": 1.2927,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.458085444246082,
+      "learning_rate": 1.9824948327259238e-05,
+      "loss": 1.269,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.603835981170523,
+      "learning_rate": 1.982443308725492e-05,
+      "loss": 1.3555,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1219774036406576,
+      "learning_rate": 1.982391709681138e-05,
+      "loss": 1.21,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7756501149094643,
+      "learning_rate": 1.982340035596804e-05,
+      "loss": 1.2886,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0181150023617103,
+      "learning_rate": 1.9822882864764366e-05,
+      "loss": 1.1824,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3194734100746475,
+      "learning_rate": 1.982236462323989e-05,
+      "loss": 1.3313,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.816043697078485,
+      "learning_rate": 1.9821845631434192e-05,
+      "loss": 1.2078,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4626433631965208,
+      "learning_rate": 1.9821325889386927e-05,
+      "loss": 1.2681,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3149376875508167,
+      "learning_rate": 1.9820805397137783e-05,
+      "loss": 1.2285,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.304791111539301,
+      "learning_rate": 1.9820284154726526e-05,
+      "loss": 1.238,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1287267077709537,
+      "learning_rate": 1.9819762162192964e-05,
+      "loss": 1.2197,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4524436771826785,
+      "learning_rate": 1.981923941957698e-05,
+      "loss": 1.2563,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.82360455471221,
+      "learning_rate": 1.981871592691849e-05,
+      "loss": 1.2754,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7182872861507894,
+      "learning_rate": 1.9818191684257492e-05,
+      "loss": 1.2705,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.726560513802765,
+      "learning_rate": 1.9817666691634025e-05,
+      "loss": 1.2983,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.836104305009529,
+      "learning_rate": 1.9817140949088193e-05,
+      "loss": 1.3281,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0666897036815373,
+      "learning_rate": 1.9816614456660152e-05,
+      "loss": 1.1826,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9500282111973135,
+      "learning_rate": 1.9816087214390123e-05,
+      "loss": 1.2991,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4558831412432856,
+      "learning_rate": 1.9815559222318374e-05,
+      "loss": 1.3235,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7836926497554475,
+      "learning_rate": 1.9815030480485238e-05,
+      "loss": 1.3091,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.105793149966776,
+      "learning_rate": 1.9814500988931103e-05,
+      "loss": 1.1909,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9823401351849244,
+      "learning_rate": 1.9813970747696415e-05,
+      "loss": 1.2717,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.360386017773243,
+      "learning_rate": 1.9813439756821674e-05,
+      "loss": 1.2954,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0506192558435643,
+      "learning_rate": 1.9812908016347444e-05,
+      "loss": 1.2471,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4556752913136344,
+      "learning_rate": 1.981237552631434e-05,
+      "loss": 1.1599,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8737552863502254,
+      "learning_rate": 1.981184228676303e-05,
+      "loss": 1.314,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8485862417863883,
+      "learning_rate": 1.9811308297734256e-05,
+      "loss": 1.249,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8883210468919076,
+      "learning_rate": 1.9810773559268797e-05,
+      "loss": 1.2998,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.342882630172088,
+      "learning_rate": 1.981023807140751e-05,
+      "loss": 1.218,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7045914953018717,
+      "learning_rate": 1.9809701834191285e-05,
+      "loss": 1.2524,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.279910344379296,
+      "learning_rate": 1.980916484766109e-05,
+      "loss": 1.292,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.314317670694452,
+      "learning_rate": 1.980862711185795e-05,
+      "loss": 1.2368,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.954697620378548,
+      "learning_rate": 1.9808088626822926e-05,
+      "loss": 1.282,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0255271777371444,
+      "learning_rate": 1.980754939259716e-05,
+      "loss": 1.2722,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3761663875256076,
+      "learning_rate": 1.9807009409221837e-05,
+      "loss": 1.2905,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.124319438663383,
+      "learning_rate": 1.98064686767382e-05,
+      "loss": 1.304,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4891437575401647,
+      "learning_rate": 1.980592719518756e-05,
+      "loss": 1.2588,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.110137534151106,
+      "learning_rate": 1.9805384964611276e-05,
+      "loss": 1.1995,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8529102075175103,
+      "learning_rate": 1.9804841985050765e-05,
+      "loss": 1.2085,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.007781956876768,
+      "learning_rate": 1.98042982565475e-05,
+      "loss": 1.2471,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2576296385380696,
+      "learning_rate": 1.9803753779143023e-05,
+      "loss": 1.3042,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9592394970609144,
+      "learning_rate": 1.9803208552878912e-05,
+      "loss": 1.2222,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4899692397528055,
+      "learning_rate": 1.9802662577796822e-05,
+      "loss": 1.3574,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.427975576695767,
+      "learning_rate": 1.9802115853938454e-05,
+      "loss": 1.291,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.227385494557931,
+      "learning_rate": 1.9801568381345572e-05,
+      "loss": 1.2439,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1201286741054233,
+      "learning_rate": 1.9801020160059995e-05,
+      "loss": 1.2336,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1622096582430492,
+      "learning_rate": 1.980047119012359e-05,
+      "loss": 1.2593,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8664700566452717,
+      "learning_rate": 1.9799921471578306e-05,
+      "loss": 1.324,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4059009293848552,
+      "learning_rate": 1.979937100446612e-05,
+      "loss": 1.3057,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.266746271460596,
+      "learning_rate": 1.979881978882908e-05,
+      "loss": 1.2651,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.588709452211583,
+      "learning_rate": 1.9798267824709302e-05,
+      "loss": 1.176,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2246618660837147,
+      "learning_rate": 1.9797715112148937e-05,
+      "loss": 1.2339,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.514894340028656,
+      "learning_rate": 1.9797161651190206e-05,
+      "loss": 1.3328,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.349014445707873,
+      "learning_rate": 1.9796607441875386e-05,
+      "loss": 1.2676,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0421606807060475,
+      "learning_rate": 1.9796052484246812e-05,
+      "loss": 1.2871,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3424320397658125,
+      "learning_rate": 1.9795496778346872e-05,
+      "loss": 1.2637,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7519242321810267,
+      "learning_rate": 1.9794940324218014e-05,
+      "loss": 1.3809,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3886035300791466,
+      "learning_rate": 1.9794383121902744e-05,
+      "loss": 1.2971,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9192534534852665,
+      "learning_rate": 1.9793825171443624e-05,
+      "loss": 1.2036,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7322630764892235,
+      "learning_rate": 1.979326647288327e-05,
+      "loss": 1.24,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.092562952456795,
+      "learning_rate": 1.9792707026264364e-05,
+      "loss": 1.2903,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.215779223441671,
+      "learning_rate": 1.9792146831629632e-05,
+      "loss": 1.3179,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.326108084888474,
+      "learning_rate": 1.9791585889021872e-05,
+      "loss": 1.2671,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5341290354096477,
+      "learning_rate": 1.9791024198483922e-05,
+      "loss": 1.2632,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.781229946815602,
+      "learning_rate": 1.9790461760058696e-05,
+      "loss": 1.2075,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3890844124212687,
+      "learning_rate": 1.9789898573789154e-05,
+      "loss": 1.2876,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3824009442620717,
+      "learning_rate": 1.978933463971831e-05,
+      "loss": 1.2175,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.203914537685938,
+      "learning_rate": 1.9788769957889246e-05,
+      "loss": 1.2014,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6507360423534148,
+      "learning_rate": 1.978820452834509e-05,
+      "loss": 1.2573,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8846479626304427,
+      "learning_rate": 1.978763835112904e-05,
+      "loss": 1.3013,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0478678106713972,
+      "learning_rate": 1.9787071426284333e-05,
+      "loss": 1.2278,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.38986629246489,
+      "learning_rate": 1.9786503753854284e-05,
+      "loss": 1.2412,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0350102730652466,
+      "learning_rate": 1.9785935333882243e-05,
+      "loss": 1.2451,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4479319460704114,
+      "learning_rate": 1.9785366166411643e-05,
+      "loss": 1.2668,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.686549088092012,
+      "learning_rate": 1.9784796251485945e-05,
+      "loss": 1.2705,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.618197471924123,
+      "learning_rate": 1.9784225589148692e-05,
+      "loss": 1.1833,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.189955134868056,
+      "learning_rate": 1.978365417944347e-05,
+      "loss": 1.281,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3754712746819586,
+      "learning_rate": 1.978308202241393e-05,
+      "loss": 1.1827,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8353475081266026,
+      "learning_rate": 1.9782509118103773e-05,
+      "loss": 1.3196,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.264858285605028,
+      "learning_rate": 1.9781935466556762e-05,
+      "loss": 1.3071,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.112521035078314,
+      "learning_rate": 1.978136106781671e-05,
+      "loss": 1.2544,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.121265733382365,
+      "learning_rate": 1.97807859219275e-05,
+      "loss": 1.2512,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.331813304393114,
+      "learning_rate": 1.9780210028933058e-05,
+      "loss": 1.2673,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5037150395278616,
+      "learning_rate": 1.977963338887738e-05,
+      "loss": 1.2271,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0901734698237697,
+      "learning_rate": 1.977905600180451e-05,
+      "loss": 1.3306,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3242437891795977,
+      "learning_rate": 1.9778477867758547e-05,
+      "loss": 1.2744,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.731537565560003,
+      "learning_rate": 1.977789898678366e-05,
+      "loss": 1.2612,
+      "step": 1116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9625503353113616,
+      "learning_rate": 1.977731935892406e-05,
+      "loss": 1.2705,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.406627849387153,
+      "learning_rate": 1.9776738984224025e-05,
+      "loss": 1.3594,
+      "step": 1118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6035660335715995,
+      "learning_rate": 1.9776157862727886e-05,
+      "loss": 1.189,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.865584146617011,
+      "learning_rate": 1.9775575994480034e-05,
+      "loss": 1.3037,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3047981021062003,
+      "learning_rate": 1.977499337952491e-05,
+      "loss": 1.3062,
+      "step": 1121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.012218611464792,
+      "learning_rate": 1.9774410017907024e-05,
+      "loss": 1.3516,
+      "step": 1122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.221688897858885,
+      "learning_rate": 1.977382590967093e-05,
+      "loss": 1.2671,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3621055762753596,
+      "learning_rate": 1.977324105486125e-05,
+      "loss": 1.2974,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6513609238018425,
+      "learning_rate": 1.9772655453522655e-05,
+      "loss": 1.1433,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.192091075330849,
+      "learning_rate": 1.9772069105699878e-05,
+      "loss": 1.2595,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.236566711692076,
+      "learning_rate": 1.9771482011437702e-05,
+      "loss": 1.2905,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.081449494476336,
+      "learning_rate": 1.977089417078098e-05,
+      "loss": 1.2681,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1114228067910363,
+      "learning_rate": 1.977030558377461e-05,
+      "loss": 1.2705,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.850701367004734,
+      "learning_rate": 1.976971625046355e-05,
+      "loss": 1.2637,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8844372587535347,
+      "learning_rate": 1.9769126170892822e-05,
+      "loss": 1.2871,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4915083766491777,
+      "learning_rate": 1.976853534510749e-05,
+      "loss": 1.2642,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3849597392080923,
+      "learning_rate": 1.976794377315269e-05,
+      "loss": 1.2256,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.905321565918617,
+      "learning_rate": 1.976735145507361e-05,
+      "loss": 1.2,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.281434540051797,
+      "learning_rate": 1.9766758390915494e-05,
+      "loss": 1.3291,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1967008746009005,
+      "learning_rate": 1.976616458072364e-05,
+      "loss": 1.2017,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.724627479765267,
+      "learning_rate": 1.9765570024543407e-05,
+      "loss": 1.2334,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9089346064832746,
+      "learning_rate": 1.9764974722420215e-05,
+      "loss": 1.2681,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.405325074251521,
+      "learning_rate": 1.9764378674399528e-05,
+      "loss": 1.3049,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.2100038145762575,
+      "learning_rate": 1.9763781880526883e-05,
+      "loss": 1.2737,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0655696312420275,
+      "learning_rate": 1.976318434084786e-05,
+      "loss": 1.3267,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.71804631142144,
+      "learning_rate": 1.9762586055408106e-05,
+      "loss": 1.2068,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.7566542716040825,
+      "learning_rate": 1.976198702425332e-05,
+      "loss": 1.312,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4130593361588764,
+      "learning_rate": 1.976138724742926e-05,
+      "loss": 1.2788,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.253671710915363,
+      "learning_rate": 1.9760786724981734e-05,
+      "loss": 1.395,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7429054749657715,
+      "learning_rate": 1.9760185456956623e-05,
+      "loss": 1.2915,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.482802326009762,
+      "learning_rate": 1.9759583443399847e-05,
+      "loss": 1.2529,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2760349272917404,
+      "learning_rate": 1.9758980684357394e-05,
+      "loss": 1.2686,
+      "step": 1148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5612834225094145,
+      "learning_rate": 1.9758377179875304e-05,
+      "loss": 1.3071,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9892746442902287,
+      "learning_rate": 1.9757772929999678e-05,
+      "loss": 1.2471,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5622515795702054,
+      "learning_rate": 1.975716793477667e-05,
+      "loss": 1.2458,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9467932699890347,
+      "learning_rate": 1.975656219425249e-05,
+      "loss": 1.2354,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.125733520577936,
+      "learning_rate": 1.9755955708473416e-05,
+      "loss": 1.25,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.236459233678715,
+      "learning_rate": 1.9755348477485766e-05,
+      "loss": 1.176,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.688030236573677,
+      "learning_rate": 1.9754740501335923e-05,
+      "loss": 1.2812,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.601668336447387,
+      "learning_rate": 1.9754131780070333e-05,
+      "loss": 1.2373,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.692096135220399,
+      "learning_rate": 1.975352231373549e-05,
+      "loss": 1.3286,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.3264986345358585,
+      "learning_rate": 1.975291210237795e-05,
+      "loss": 1.2588,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9304463595312789,
+      "learning_rate": 1.9752301146044323e-05,
+      "loss": 1.0332,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.368839456186262,
+      "learning_rate": 1.9751689444781272e-05,
+      "loss": 1.272,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.428065843259737,
+      "learning_rate": 1.975107699863553e-05,
+      "loss": 1.2866,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2914478929340465,
+      "learning_rate": 1.9750463807653873e-05,
+      "loss": 1.2222,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3629395988991275,
+      "learning_rate": 1.9749849871883145e-05,
+      "loss": 1.2051,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2104039386993706,
+      "learning_rate": 1.974923519137023e-05,
+      "loss": 1.2192,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.526533876111914,
+      "learning_rate": 1.9748619766162095e-05,
+      "loss": 1.2739,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.230917766062076,
+      "learning_rate": 1.9748003596305746e-05,
+      "loss": 1.3694,
+      "step": 1166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.597661835617468,
+      "learning_rate": 1.974738668184824e-05,
+      "loss": 1.2612,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8232779688762406,
+      "learning_rate": 1.9746769022836707e-05,
+      "loss": 1.24,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.552902927956375,
+      "learning_rate": 1.9746150619318325e-05,
+      "loss": 1.269,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3236723134420036,
+      "learning_rate": 1.974553147134033e-05,
+      "loss": 1.2563,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3878993478484816,
+      "learning_rate": 1.9744911578950023e-05,
+      "loss": 1.2627,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.204001255087207,
+      "learning_rate": 1.974429094219474e-05,
+      "loss": 1.2046,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2139424740038707,
+      "learning_rate": 1.9743669561121903e-05,
+      "loss": 1.2078,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.863715647412473,
+      "learning_rate": 1.974304743577897e-05,
+      "loss": 1.1543,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.340601089564149,
+      "learning_rate": 1.9742424566213456e-05,
+      "loss": 1.3032,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3290152953998056,
+      "learning_rate": 1.974180095247295e-05,
+      "loss": 1.2097,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1664594854505275,
+      "learning_rate": 1.9741176594605082e-05,
+      "loss": 1.2188,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9402426155289576,
+      "learning_rate": 1.974055149265754e-05,
+      "loss": 1.2227,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.293994554298049,
+      "learning_rate": 1.9739925646678077e-05,
+      "loss": 1.1963,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3255266109958797,
+      "learning_rate": 1.9739299056714497e-05,
+      "loss": 1.2725,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.309309941008927,
+      "learning_rate": 1.9738671722814663e-05,
+      "loss": 1.2549,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.081753257415044,
+      "learning_rate": 1.973804364502649e-05,
+      "loss": 1.3296,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2933777928982644,
+      "learning_rate": 1.973741482339796e-05,
+      "loss": 1.3438,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2260056793382543,
+      "learning_rate": 1.97367852579771e-05,
+      "loss": 1.2,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9892099272909474,
+      "learning_rate": 1.9736154948812003e-05,
+      "loss": 1.2451,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6995375436961386,
+      "learning_rate": 1.973552389595081e-05,
+      "loss": 1.2773,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8189064838615114,
+      "learning_rate": 1.973489209944173e-05,
+      "loss": 1.2869,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.748110455067656,
+      "learning_rate": 1.9734259559333022e-05,
+      "loss": 1.2573,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7236420899912632,
+      "learning_rate": 1.9733626275673e-05,
+      "loss": 1.2971,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7343103732917347,
+      "learning_rate": 1.9732992248510035e-05,
+      "loss": 1.2893,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.185968846475122,
+      "learning_rate": 1.9732357477892563e-05,
+      "loss": 1.26,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4808221739859033,
+      "learning_rate": 1.9731721963869067e-05,
+      "loss": 1.3096,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2947687322339925,
+      "learning_rate": 1.973108570648809e-05,
+      "loss": 1.2969,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.242460898712626,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 1.2532,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.444551910111098,
+      "learning_rate": 1.9729810961848164e-05,
+      "loss": 1.2529,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4850899698919466,
+      "learning_rate": 1.9729172474686585e-05,
+      "loss": 1.272,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7708044351646945,
+      "learning_rate": 1.9728533244362266e-05,
+      "loss": 1.2529,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4376982111572127,
+      "learning_rate": 1.9727893270924043e-05,
+      "loss": 1.2227,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7436719331419117,
+      "learning_rate": 1.9727252554420793e-05,
+      "loss": 1.3237,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.103993561678892,
+      "learning_rate": 1.972661109490146e-05,
+      "loss": 1.3193,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.501563877789951,
+      "learning_rate": 1.972596889241504e-05,
+      "loss": 1.2615,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.635529555968793,
+      "learning_rate": 1.972532594701059e-05,
+      "loss": 1.3228,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.037329954695726,
+      "learning_rate": 1.9724682258737224e-05,
+      "loss": 1.2219,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.694637490851572,
+      "learning_rate": 1.9724037827644103e-05,
+      "loss": 1.2329,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8559320956111405,
+      "learning_rate": 1.972339265378046e-05,
+      "loss": 1.3103,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.192251899126128,
+      "learning_rate": 1.972274673719557e-05,
+      "loss": 1.2798,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5685456614369473,
+      "learning_rate": 1.972210007793877e-05,
+      "loss": 1.2358,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.561327193009963,
+      "learning_rate": 1.972145267605946e-05,
+      "loss": 1.2905,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.12840266832028,
+      "learning_rate": 1.9720804531607097e-05,
+      "loss": 1.2891,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.550098689837315,
+      "learning_rate": 1.9720155644631177e-05,
+      "loss": 1.2109,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.287336155519426,
+      "learning_rate": 1.9719506015181273e-05,
+      "loss": 1.2197,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0891181359637763,
+      "learning_rate": 1.9718855643307004e-05,
+      "loss": 1.2241,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5753969900436537,
+      "learning_rate": 1.971820452905805e-05,
+      "loss": 1.3076,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.365119142823262,
+      "learning_rate": 1.9717552672484147e-05,
+      "loss": 1.3242,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5793293974101505,
+      "learning_rate": 1.9716900073635086e-05,
+      "loss": 1.261,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.964257431576929,
+      "learning_rate": 1.9716246732560715e-05,
+      "loss": 1.2798,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7687613580581614,
+      "learning_rate": 1.971559264931094e-05,
+      "loss": 1.3359,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1831293881988274,
+      "learning_rate": 1.9714937823935725e-05,
+      "loss": 1.3149,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.03781872588584,
+      "learning_rate": 1.971428225648509e-05,
+      "loss": 1.2268,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.642262485906229,
+      "learning_rate": 1.9713625947009105e-05,
+      "loss": 1.2427,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.004402781924198,
+      "learning_rate": 1.9712968895557905e-05,
+      "loss": 1.2651,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.770059225872864,
+      "learning_rate": 1.9712311102181682e-05,
+      "loss": 1.1321,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.464960368640905,
+      "learning_rate": 1.9711652566930673e-05,
+      "loss": 1.4365,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6377556456612554,
+      "learning_rate": 1.971099328985519e-05,
+      "loss": 1.1792,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.566945413752673,
+      "learning_rate": 1.9710333271005586e-05,
+      "loss": 1.2227,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.155287068184395,
+      "learning_rate": 1.9709672510432275e-05,
+      "loss": 1.2505,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8262413094302157,
+      "learning_rate": 1.9709011008185737e-05,
+      "loss": 1.2732,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.028749786891383,
+      "learning_rate": 1.9708348764316497e-05,
+      "loss": 1.2207,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.879250039157958,
+      "learning_rate": 1.9707685778875137e-05,
+      "loss": 1.2234,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.04007702144179,
+      "learning_rate": 1.97070220519123e-05,
+      "loss": 1.2439,
+      "step": 1230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6945455808295034,
+      "learning_rate": 1.970635758347869e-05,
+      "loss": 1.2251,
+      "step": 1231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9374505711189123,
+      "learning_rate": 1.9705692373625056e-05,
+      "loss": 1.3096,
+      "step": 1232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.87003450957644,
+      "learning_rate": 1.9705026422402215e-05,
+      "loss": 1.0837,
+      "step": 1233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7061748462542785,
+      "learning_rate": 1.9704359729861035e-05,
+      "loss": 1.2759,
+      "step": 1234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.352731376081371,
+      "learning_rate": 1.9703692296052435e-05,
+      "loss": 1.2798,
+      "step": 1235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8524560190635113,
+      "learning_rate": 1.9703024121027408e-05,
+      "loss": 1.2454,
+      "step": 1236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4486506304464517,
+      "learning_rate": 1.9702355204836985e-05,
+      "loss": 1.291,
+      "step": 1237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.158052238241538,
+      "learning_rate": 1.9701685547532262e-05,
+      "loss": 1.1394,
+      "step": 1238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3786692434219465,
+      "learning_rate": 1.970101514916439e-05,
+      "loss": 1.2686,
+      "step": 1239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7376432190740445,
+      "learning_rate": 1.970034400978458e-05,
+      "loss": 1.2712,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.790565531193211,
+      "learning_rate": 1.9699672129444094e-05,
+      "loss": 1.2048,
+      "step": 1241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.354099701182449,
+      "learning_rate": 1.969899950819426e-05,
+      "loss": 1.2749,
+      "step": 1242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0853752817070035,
+      "learning_rate": 1.9698326146086446e-05,
+      "loss": 1.2344,
+      "step": 1243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5417376510405485,
+      "learning_rate": 1.9697652043172092e-05,
+      "loss": 1.2817,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.238165013176494,
+      "learning_rate": 1.9696977199502695e-05,
+      "loss": 1.2644,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.339660823299074,
+      "learning_rate": 1.9696301615129793e-05,
+      "loss": 1.2559,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6878103954243966,
+      "learning_rate": 1.9695625290105e-05,
+      "loss": 1.2764,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9027852548730992,
+      "learning_rate": 1.9694948224479965e-05,
+      "loss": 1.2661,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2889184979721744,
+      "learning_rate": 1.9694270418306418e-05,
+      "loss": 1.1577,
+      "step": 1249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4795896074605928,
+      "learning_rate": 1.9693591871636126e-05,
+      "loss": 1.3613,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1964704107790434,
+      "learning_rate": 1.9692912584520922e-05,
+      "loss": 1.2661,
+      "step": 1251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1564916580760474,
+      "learning_rate": 1.9692232557012693e-05,
+      "loss": 1.2947,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.451641968598114,
+      "learning_rate": 1.9691551789163384e-05,
+      "loss": 1.2637,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.016189249821717,
+      "learning_rate": 1.9690870281024992e-05,
+      "loss": 1.2319,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3361215479240562,
+      "learning_rate": 1.969018803264958e-05,
+      "loss": 1.2432,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0265862377013275,
+      "learning_rate": 1.9689505044089253e-05,
+      "loss": 1.2646,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3431148481471173,
+      "learning_rate": 1.9688821315396192e-05,
+      "loss": 1.1602,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.627494246405701,
+      "learning_rate": 1.968813684662261e-05,
+      "loss": 1.2441,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4635978907630607,
+      "learning_rate": 1.9687451637820803e-05,
+      "loss": 1.2681,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.456356123870087,
+      "learning_rate": 1.968676568904311e-05,
+      "loss": 1.2617,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3517218866718594,
+      "learning_rate": 1.9686079000341914e-05,
+      "loss": 1.2378,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5061059187692334,
+      "learning_rate": 1.9685391571769677e-05,
+      "loss": 1.2593,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6965892334756707,
+      "learning_rate": 1.9684703403378912e-05,
+      "loss": 1.303,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.346222031184034,
+      "learning_rate": 1.9684014495222178e-05,
+      "loss": 1.314,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3779194355544684,
+      "learning_rate": 1.96833248473521e-05,
+      "loss": 1.2366,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.245417002727963,
+      "learning_rate": 1.9682634459821356e-05,
+      "loss": 1.3027,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8545843488003597,
+      "learning_rate": 1.9681943332682683e-05,
+      "loss": 1.2395,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9513175610334463,
+      "learning_rate": 1.968125146598887e-05,
+      "loss": 1.23,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6109486482900843,
+      "learning_rate": 1.9680558859792768e-05,
+      "loss": 1.3486,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.470710776524007,
+      "learning_rate": 1.967986551414728e-05,
+      "loss": 1.2732,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1772914912616876,
+      "learning_rate": 1.9679171429105363e-05,
+      "loss": 1.1689,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.177518228197765,
+      "learning_rate": 1.9678476604720046e-05,
+      "loss": 1.2,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4851169142451845,
+      "learning_rate": 1.967778104104439e-05,
+      "loss": 1.2183,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.6883991800599425,
+      "learning_rate": 1.9677084738131535e-05,
+      "loss": 1.292,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.06853177121528,
+      "learning_rate": 1.9676387696034665e-05,
+      "loss": 1.2312,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8653119849499125,
+      "learning_rate": 1.9675689914807023e-05,
+      "loss": 1.2627,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.07661787875394,
+      "learning_rate": 1.967499139450191e-05,
+      "loss": 1.2808,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.357019074340022,
+      "learning_rate": 1.9674292135172683e-05,
+      "loss": 1.2319,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9097759760852906,
+      "learning_rate": 1.967359213687275e-05,
+      "loss": 1.3008,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8932646928530388,
+      "learning_rate": 1.9672891399655587e-05,
+      "loss": 1.2039,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0758289507050143,
+      "learning_rate": 1.9672189923574718e-05,
+      "loss": 1.2576,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.519082612306214,
+      "learning_rate": 1.9671487708683723e-05,
+      "loss": 1.2886,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9560879442506915,
+      "learning_rate": 1.967078475503624e-05,
+      "loss": 1.2588,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1641448057325126,
+      "learning_rate": 1.9670081062685972e-05,
+      "loss": 1.2683,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7599341773361497,
+      "learning_rate": 1.9669376631686658e-05,
+      "loss": 1.3035,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.513735077884851,
+      "learning_rate": 1.9668671462092113e-05,
+      "loss": 1.123,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5456725192768968,
+      "learning_rate": 1.9667965553956206e-05,
+      "loss": 1.2607,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.425700331899081,
+      "learning_rate": 1.966725890733285e-05,
+      "loss": 1.2935,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.087446486805911,
+      "learning_rate": 1.9666551522276024e-05,
+      "loss": 1.2405,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9903399447586825,
+      "learning_rate": 1.9665843398839763e-05,
+      "loss": 1.2324,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7868443652599777,
+      "learning_rate": 1.966513453707816e-05,
+      "loss": 1.1904,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.358136661515128,
+      "learning_rate": 1.9664424937045353e-05,
+      "loss": 1.3022,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.373848937776347,
+      "learning_rate": 1.966371459879555e-05,
+      "loss": 1.2332,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2597002060853026,
+      "learning_rate": 1.9663003522383013e-05,
+      "loss": 1.2234,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4869584028067626,
+      "learning_rate": 1.9662291707862055e-05,
+      "loss": 1.2473,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.068096861837584,
+      "learning_rate": 1.9661579155287046e-05,
+      "loss": 1.3247,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.426213062174945,
+      "learning_rate": 1.9660865864712413e-05,
+      "loss": 1.2527,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9616163380380565,
+      "learning_rate": 1.9660151836192648e-05,
+      "loss": 1.2964,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.501182911589572,
+      "learning_rate": 1.9659437069782282e-05,
+      "loss": 1.3052,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6491695344587454,
+      "learning_rate": 1.9658721565535923e-05,
+      "loss": 1.3198,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.798604999074978,
+      "learning_rate": 1.965800532350822e-05,
+      "loss": 1.2986,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.401007395040289,
+      "learning_rate": 1.965728834375388e-05,
+      "loss": 1.3213,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1313263600783854,
+      "learning_rate": 1.965657062632767e-05,
+      "loss": 1.2493,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1708578737740667,
+      "learning_rate": 1.965585217128442e-05,
+      "loss": 1.2563,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.581583513395744,
+      "learning_rate": 1.9655132978679003e-05,
+      "loss": 1.2034,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7536351457695005,
+      "learning_rate": 1.9654413048566353e-05,
+      "loss": 1.2087,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1356262081186728,
+      "learning_rate": 1.9653692381001467e-05,
+      "loss": 1.2627,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5851953614791316,
+      "learning_rate": 1.9652970976039393e-05,
+      "loss": 1.2542,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.066667651162878,
+      "learning_rate": 1.965224883373523e-05,
+      "loss": 1.2275,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.622314095231972,
+      "learning_rate": 1.9651525954144148e-05,
+      "loss": 1.2554,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.287332112738072,
+      "learning_rate": 1.965080233732135e-05,
+      "loss": 1.3179,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.372741825963426,
+      "learning_rate": 1.9650077983322124e-05,
+      "loss": 1.1863,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9488289829719823,
+      "learning_rate": 1.9649352892201793e-05,
+      "loss": 1.2551,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4175228928909855,
+      "learning_rate": 1.964862706401574e-05,
+      "loss": 1.1743,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9377247826857666,
+      "learning_rate": 1.9647900498819412e-05,
+      "loss": 1.2837,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4240674415595196,
+      "learning_rate": 1.964717319666831e-05,
+      "loss": 1.3252,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.281124931460957,
+      "learning_rate": 1.9646445157617983e-05,
+      "loss": 1.2837,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.939053480570753,
+      "learning_rate": 1.964571638172405e-05,
+      "loss": 1.217,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.08357416146249,
+      "learning_rate": 1.9644986869042167e-05,
+      "loss": 1.2068,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.551452747776152,
+      "learning_rate": 1.9644256619628064e-05,
+      "loss": 1.2832,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1561765365678998,
+      "learning_rate": 1.9643525633537525e-05,
+      "loss": 1.2588,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2760491428557796,
+      "learning_rate": 1.9642793910826382e-05,
+      "loss": 1.2283,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4259324831513953,
+      "learning_rate": 1.964206145155053e-05,
+      "loss": 1.2312,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8762719264179304,
+      "learning_rate": 1.9641328255765916e-05,
+      "loss": 1.2373,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.473280600466387,
+      "learning_rate": 1.9640594323528544e-05,
+      "loss": 1.3203,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6970301402991432,
+      "learning_rate": 1.9639859654894476e-05,
+      "loss": 1.3032,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.407274230643812,
+      "learning_rate": 1.963912424991983e-05,
+      "loss": 1.2192,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.356703784775248,
+      "learning_rate": 1.9638388108660783e-05,
+      "loss": 1.1919,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.652841022114779,
+      "learning_rate": 1.9637651231173566e-05,
+      "loss": 1.2781,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.079587292250542,
+      "learning_rate": 1.9636913617514456e-05,
+      "loss": 1.2827,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.839602383892733,
+      "learning_rate": 1.9636175267739805e-05,
+      "loss": 1.1709,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2488598025592648,
+      "learning_rate": 1.9635436181906004e-05,
+      "loss": 1.2891,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.591364956561953,
+      "learning_rate": 1.9634696360069517e-05,
+      "loss": 1.2002,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5634551062007223,
+      "learning_rate": 1.963395580228685e-05,
+      "loss": 1.2373,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0669557883842873,
+      "learning_rate": 1.963321450861457e-05,
+      "loss": 1.2849,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.598627284160907,
+      "learning_rate": 1.96324724791093e-05,
+      "loss": 1.2493,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5460743122456844,
+      "learning_rate": 1.9631729713827722e-05,
+      "loss": 1.3208,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6878232254847036,
+      "learning_rate": 1.9630986212826576e-05,
+      "loss": 1.3379,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.127772203188451,
+      "learning_rate": 1.963024197616265e-05,
+      "loss": 1.2627,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.15619394723531,
+      "learning_rate": 1.962949700389279e-05,
+      "loss": 1.1785,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7064150586116202,
+      "learning_rate": 1.9628751296073906e-05,
+      "loss": 1.2627,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1864409545641923,
+      "learning_rate": 1.9628004852762952e-05,
+      "loss": 1.2888,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.59849124109788,
+      "learning_rate": 1.9627257674016954e-05,
+      "loss": 1.2803,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.305824522268333,
+      "learning_rate": 1.962650975989298e-05,
+      "loss": 1.2271,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5197841366086835,
+      "learning_rate": 1.962576111044816e-05,
+      "loss": 1.2432,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1412879805428595,
+      "learning_rate": 1.962501172573968e-05,
+      "loss": 1.2754,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.309629069600568,
+      "learning_rate": 1.962426160582478e-05,
+      "loss": 1.2236,
+      "step": 1347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.376573918746721,
+      "learning_rate": 1.962351075076076e-05,
+      "loss": 1.1929,
+      "step": 1348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4498546732080353,
+      "learning_rate": 1.9622759160604974e-05,
+      "loss": 1.2368,
+      "step": 1349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4256756012109766,
+      "learning_rate": 1.962200683541483e-05,
+      "loss": 1.2314,
+      "step": 1350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.995931220967792,
+      "learning_rate": 1.96212537752478e-05,
+      "loss": 1.2864,
+      "step": 1351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2343481662807334,
+      "learning_rate": 1.9620499980161397e-05,
+      "loss": 1.23,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.279064843907596,
+      "learning_rate": 1.9619745450213208e-05,
+      "loss": 1.2017,
+      "step": 1353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.221008337718894,
+      "learning_rate": 1.9618990185460865e-05,
+      "loss": 1.1284,
+      "step": 1354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1944676019023857,
+      "learning_rate": 1.9618234185962056e-05,
+      "loss": 1.2368,
+      "step": 1355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.534678117943284,
+      "learning_rate": 1.9617477451774537e-05,
+      "loss": 1.2559,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.826789132775429,
+      "learning_rate": 1.9616719982956097e-05,
+      "loss": 1.2842,
+      "step": 1357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.675478031642721,
+      "learning_rate": 1.961596177956461e-05,
+      "loss": 1.2446,
+      "step": 1358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7020987625834545,
+      "learning_rate": 1.9615202841657985e-05,
+      "loss": 1.3052,
+      "step": 1359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.998100779635901,
+      "learning_rate": 1.9614443169294186e-05,
+      "loss": 1.2549,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3227216468859515,
+      "learning_rate": 1.9613682762531252e-05,
+      "loss": 1.3164,
+      "step": 1361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8821633132504934,
+      "learning_rate": 1.961292162142726e-05,
+      "loss": 1.2915,
+      "step": 1362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3830738298561993,
+      "learning_rate": 1.9612159746040355e-05,
+      "loss": 1.2151,
+      "step": 1363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5269658944003033,
+      "learning_rate": 1.9611397136428725e-05,
+      "loss": 1.2817,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2136054284490116,
+      "learning_rate": 1.9610633792650632e-05,
+      "loss": 1.1204,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4044685983871927,
+      "learning_rate": 1.9609869714764377e-05,
+      "loss": 1.3364,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.208810792147363,
+      "learning_rate": 1.9609104902828325e-05,
+      "loss": 1.269,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.18310674853465,
+      "learning_rate": 1.9608339356900897e-05,
+      "loss": 1.3057,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.267761582447746,
+      "learning_rate": 1.960757307704057e-05,
+      "loss": 1.2959,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.084626954238022,
+      "learning_rate": 1.9606806063305873e-05,
+      "loss": 1.2495,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5489995359458075,
+      "learning_rate": 1.96060383157554e-05,
+      "loss": 1.2732,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.128477408501015,
+      "learning_rate": 1.960526983444779e-05,
+      "loss": 1.1853,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9845751728469314,
+      "learning_rate": 1.9604500619441745e-05,
+      "loss": 1.2505,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.343627155703983,
+      "learning_rate": 1.9603730670796025e-05,
+      "loss": 1.1741,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7307335440043587,
+      "learning_rate": 1.960295998856944e-05,
+      "loss": 1.3003,
+      "step": 1375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.017749965506203,
+      "learning_rate": 1.9602188572820854e-05,
+      "loss": 1.2317,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5160052303160136,
+      "learning_rate": 1.96014164236092e-05,
+      "loss": 1.2715,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2736699120243666,
+      "learning_rate": 1.9600643540993453e-05,
+      "loss": 1.2371,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3974725918242368,
+      "learning_rate": 1.9599869925032647e-05,
+      "loss": 1.2478,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0968592911049035,
+      "learning_rate": 1.9599095575785885e-05,
+      "loss": 1.2627,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.012606311174533,
+      "learning_rate": 1.9598320493312307e-05,
+      "loss": 1.2065,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3338937840671385,
+      "learning_rate": 1.959754467767112e-05,
+      "loss": 1.2856,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.243192508777218,
+      "learning_rate": 1.959676812892158e-05,
+      "loss": 1.2407,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.733654399601925,
+      "learning_rate": 1.9595990847123013e-05,
+      "loss": 1.2576,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.366433772576363,
+      "learning_rate": 1.9595212832334786e-05,
+      "loss": 1.2371,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.048787988591674,
+      "learning_rate": 1.9594434084616328e-05,
+      "loss": 1.2534,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.913478986041895,
+      "learning_rate": 1.959365460402713e-05,
+      "loss": 1.3838,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3765998185156216,
+      "learning_rate": 1.959287439062672e-05,
+      "loss": 1.2197,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6357297915058573,
+      "learning_rate": 1.9592093444474704e-05,
+      "loss": 1.1953,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2861548815924486,
+      "learning_rate": 1.959131176563073e-05,
+      "loss": 1.1882,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.377958436288329,
+      "learning_rate": 1.9590529354154512e-05,
+      "loss": 1.2725,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.730754107401189,
+      "learning_rate": 1.958974621010581e-05,
+      "loss": 1.3027,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.228654909868341,
+      "learning_rate": 1.9588962333544442e-05,
+      "loss": 1.3086,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.63625565273374,
+      "learning_rate": 1.958817772453029e-05,
+      "loss": 1.2974,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5305479752830795,
+      "learning_rate": 1.9587392383123288e-05,
+      "loss": 1.2944,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4734658445661664,
+      "learning_rate": 1.9586606309383415e-05,
+      "loss": 1.2087,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.175640712334617,
+      "learning_rate": 1.9585819503370725e-05,
+      "loss": 1.2358,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7792120771982898,
+      "learning_rate": 1.958503196514531e-05,
+      "loss": 1.157,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.410220811255674,
+      "learning_rate": 1.9584243694767332e-05,
+      "loss": 1.2366,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.822565869180531,
+      "learning_rate": 1.9583454692297003e-05,
+      "loss": 1.222,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.699661099658657,
+      "learning_rate": 1.9582664957794587e-05,
+      "loss": 1.313,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5563064466119365,
+      "learning_rate": 1.9581874491320406e-05,
+      "loss": 1.2122,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9343834815953924,
+      "learning_rate": 1.9581083292934847e-05,
+      "loss": 1.2104,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.584735974881004,
+      "learning_rate": 1.9580291362698344e-05,
+      "loss": 1.2898,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.460254623617134,
+      "learning_rate": 1.9579498700671386e-05,
+      "loss": 1.2251,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.105441033866446,
+      "learning_rate": 1.9578705306914524e-05,
+      "loss": 1.209,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2186201639255705,
+      "learning_rate": 1.9577911181488357e-05,
+      "loss": 1.2637,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0596707267002445,
+      "learning_rate": 1.957711632445354e-05,
+      "loss": 1.2495,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5674484440608105,
+      "learning_rate": 1.9576320735870802e-05,
+      "loss": 1.1377,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2401341770915515,
+      "learning_rate": 1.9575524415800903e-05,
+      "loss": 1.1602,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.156326496345581,
+      "learning_rate": 1.9574727364304674e-05,
+      "loss": 1.2498,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6875623460576614,
+      "learning_rate": 1.9573929581442997e-05,
+      "loss": 1.2358,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4684323861366004,
+      "learning_rate": 1.957313106727681e-05,
+      "loss": 1.2861,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3724295956131796,
+      "learning_rate": 1.9572331821867108e-05,
+      "loss": 1.2749,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3361187328405317,
+      "learning_rate": 1.957153184527494e-05,
+      "loss": 1.3081,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.163967474446027,
+      "learning_rate": 1.9570731137561417e-05,
+      "loss": 1.1909,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.606457045799689,
+      "learning_rate": 1.9569929698787694e-05,
+      "loss": 1.3135,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2496389312548692,
+      "learning_rate": 1.9569127529014996e-05,
+      "loss": 1.2227,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.108496830084396,
+      "learning_rate": 1.956832462830459e-05,
+      "loss": 1.3066,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.166615866076609,
+      "learning_rate": 1.956752099671781e-05,
+      "loss": 1.2271,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2022549517554006,
+      "learning_rate": 1.956671663431604e-05,
+      "loss": 1.2227,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9588438124020553,
+      "learning_rate": 1.956591154116072e-05,
+      "loss": 1.2002,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1031207872693005,
+      "learning_rate": 1.9565105717313352e-05,
+      "loss": 1.2039,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2666846894833492,
+      "learning_rate": 1.9564299162835483e-05,
+      "loss": 1.239,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.10415432736392,
+      "learning_rate": 1.9563491877788726e-05,
+      "loss": 1.2959,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4700023505015,
+      "learning_rate": 1.956268386223474e-05,
+      "loss": 1.3677,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9107775353010177,
+      "learning_rate": 1.956187511623525e-05,
+      "loss": 1.231,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6558312426916455,
+      "learning_rate": 1.956106563985203e-05,
+      "loss": 1.2437,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.288317830938594,
+      "learning_rate": 1.9560255433146916e-05,
+      "loss": 1.2112,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2661719985005244,
+      "learning_rate": 1.955944449618179e-05,
+      "loss": 1.252,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.104176056867543,
+      "learning_rate": 1.9558632829018596e-05,
+      "loss": 1.2847,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.845123754897243,
+      "learning_rate": 1.9557820431719333e-05,
+      "loss": 1.1721,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5410755767050004,
+      "learning_rate": 1.955700730434606e-05,
+      "loss": 1.2461,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.316181198954885,
+      "learning_rate": 1.955619344696089e-05,
+      "loss": 1.2212,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4522604450686023,
+      "learning_rate": 1.9555378859625978e-05,
+      "loss": 1.2915,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4791590310028893,
+      "learning_rate": 1.9554563542403558e-05,
+      "loss": 1.3052,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.071614166816688,
+      "learning_rate": 1.95537474953559e-05,
+      "loss": 1.1755,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1911457569303665,
+      "learning_rate": 1.9552930718545346e-05,
+      "loss": 1.3032,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4144628707935185,
+      "learning_rate": 1.9552113212034276e-05,
+      "loss": 1.2778,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2047443856856743,
+      "learning_rate": 1.955129497588514e-05,
+      "loss": 1.3018,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.736703832933073,
+      "learning_rate": 1.955047601016044e-05,
+      "loss": 1.2683,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.940641145536436,
+      "learning_rate": 1.954965631492273e-05,
+      "loss": 1.2637,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3135827598676433,
+      "learning_rate": 1.9548835890234623e-05,
+      "loss": 1.3018,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4647267013001075,
+      "learning_rate": 1.954801473615879e-05,
+      "loss": 1.2739,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.099553672930975,
+      "learning_rate": 1.954719285275795e-05,
+      "loss": 1.29,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1425804017558367,
+      "learning_rate": 1.9546370240094888e-05,
+      "loss": 1.2412,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.020201128740013,
+      "learning_rate": 1.9545546898232437e-05,
+      "loss": 1.2217,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8425306725476163,
+      "learning_rate": 1.9544722827233487e-05,
+      "loss": 1.1248,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5114090771551907,
+      "learning_rate": 1.9543898027160984e-05,
+      "loss": 1.2395,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9966679601608286,
+      "learning_rate": 1.954307249807793e-05,
+      "loss": 1.1948,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3558263831653234,
+      "learning_rate": 1.9542246240047385e-05,
+      "loss": 1.2524,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6693830942218324,
+      "learning_rate": 1.9541419253132467e-05,
+      "loss": 1.2261,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5554061122668794,
+      "learning_rate": 1.9540591537396336e-05,
+      "loss": 1.2432,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.665513752415085,
+      "learning_rate": 1.9539763092902222e-05,
+      "loss": 1.1482,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2370756363137607,
+      "learning_rate": 1.9538933919713406e-05,
+      "loss": 1.2271,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2514154134276416,
+      "learning_rate": 1.953810401789322e-05,
+      "loss": 1.335,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.194337915258399,
+      "learning_rate": 1.9537273387505067e-05,
+      "loss": 1.1062,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.294372430812874,
+      "learning_rate": 1.9536442028612383e-05,
+      "loss": 1.2483,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.899425767551795,
+      "learning_rate": 1.9535609941278676e-05,
+      "loss": 1.2324,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.704102773889957,
+      "learning_rate": 1.9534777125567507e-05,
+      "loss": 1.251,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0634710967282976,
+      "learning_rate": 1.9533943581542488e-05,
+      "loss": 1.293,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0214301987854064,
+      "learning_rate": 1.953310930926729e-05,
+      "loss": 1.1873,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9001148240453185,
+      "learning_rate": 1.9532274308805636e-05,
+      "loss": 1.2378,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.73459584642823,
+      "learning_rate": 1.953143858022131e-05,
+      "loss": 1.2368,
+      "step": 1464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.485652443512886,
+      "learning_rate": 1.953060212357815e-05,
+      "loss": 1.2563,
+      "step": 1465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3791557317757284,
+      "learning_rate": 1.952976493894005e-05,
+      "loss": 1.2351,
+      "step": 1466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.14568063516712,
+      "learning_rate": 1.9528927026370954e-05,
+      "loss": 1.2629,
+      "step": 1467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.693358212409802,
+      "learning_rate": 1.952808838593487e-05,
+      "loss": 1.1536,
+      "step": 1468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.929279876802863,
+      "learning_rate": 1.9527249017695852e-05,
+      "loss": 1.3086,
+      "step": 1469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0228168279561536,
+      "learning_rate": 1.9526408921718023e-05,
+      "loss": 1.1992,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4019766670025264,
+      "learning_rate": 1.952556809806555e-05,
+      "loss": 1.3374,
+      "step": 1471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0940579486276363,
+      "learning_rate": 1.9524726546802655e-05,
+      "loss": 1.2388,
+      "step": 1472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4030188540538524,
+      "learning_rate": 1.9523884267993626e-05,
+      "loss": 1.2852,
+      "step": 1473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.359151103077032,
+      "learning_rate": 1.95230412617028e-05,
+      "loss": 1.2439,
+      "step": 1474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4797973236709714,
+      "learning_rate": 1.9522197527994565e-05,
+      "loss": 1.3408,
+      "step": 1475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2494431239797743,
+      "learning_rate": 1.9521353066933374e-05,
+      "loss": 1.2051,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.250293374288638,
+      "learning_rate": 1.9520507878583728e-05,
+      "loss": 1.2578,
+      "step": 1477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0913168270318985,
+      "learning_rate": 1.951966196301019e-05,
+      "loss": 1.2549,
+      "step": 1478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1650665022188167,
+      "learning_rate": 1.951881532027738e-05,
+      "loss": 1.2705,
+      "step": 1479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.155069573067177,
+      "learning_rate": 1.9517967950449956e-05,
+      "loss": 1.2126,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2486197602369846,
+      "learning_rate": 1.9517119853592653e-05,
+      "loss": 1.2188,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.047133613272701,
+      "learning_rate": 1.951627102977025e-05,
+      "loss": 1.2935,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.118268754741922,
+      "learning_rate": 1.9515421479047588e-05,
+      "loss": 1.2671,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4695460653450434,
+      "learning_rate": 1.9514571201489553e-05,
+      "loss": 1.3069,
+      "step": 1484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.562298027465081,
+      "learning_rate": 1.95137201971611e-05,
+      "loss": 1.2739,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.384810357147669,
+      "learning_rate": 1.9512868466127232e-05,
+      "loss": 1.2925,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3517674778566144,
+      "learning_rate": 1.9512016008453005e-05,
+      "loss": 1.3047,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2183702298939845,
+      "learning_rate": 1.951116282420354e-05,
+      "loss": 1.218,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.17556600087554,
+      "learning_rate": 1.9510308913443997e-05,
+      "loss": 1.2549,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.169134511428046,
+      "learning_rate": 1.9509454276239614e-05,
+      "loss": 1.1907,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.415420023232102,
+      "learning_rate": 1.9508598912655662e-05,
+      "loss": 1.3096,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.505125603752304,
+      "learning_rate": 1.9507742822757486e-05,
+      "loss": 1.2827,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.37781130794748,
+      "learning_rate": 1.950688600661048e-05,
+      "loss": 1.2705,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8554475727417974,
+      "learning_rate": 1.9506028464280082e-05,
+      "loss": 1.2905,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4264209737493543,
+      "learning_rate": 1.95051701958318e-05,
+      "loss": 1.2783,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.017717675806171,
+      "learning_rate": 1.9504311201331196e-05,
+      "loss": 1.1943,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.142763477161724,
+      "learning_rate": 1.950345148084388e-05,
+      "loss": 1.2417,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2755619043669877,
+      "learning_rate": 1.9502591034435525e-05,
+      "loss": 1.2783,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.234940965115558,
+      "learning_rate": 1.950172986217185e-05,
+      "loss": 1.3203,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4704430012531127,
+      "learning_rate": 1.950086796411865e-05,
+      "loss": 1.2456,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.642947582960144,
+      "learning_rate": 1.950000534034174e-05,
+      "loss": 1.333,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9204293125970704,
+      "learning_rate": 1.949914199090703e-05,
+      "loss": 1.2754,
+      "step": 1502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1200328651340787,
+      "learning_rate": 1.949827791588046e-05,
+      "loss": 1.229,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2443776212049964,
+      "learning_rate": 1.9497413115328028e-05,
+      "loss": 1.2241,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4033226274512582,
+      "learning_rate": 1.9496547589315798e-05,
+      "loss": 1.2993,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.276894509408774,
+      "learning_rate": 1.949568133790988e-05,
+      "loss": 1.2625,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4399276691836613,
+      "learning_rate": 1.9494814361176448e-05,
+      "loss": 1.2996,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8059894963557266,
+      "learning_rate": 1.9493946659181716e-05,
+      "loss": 1.1973,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0208119858372506,
+      "learning_rate": 1.9493078231991975e-05,
+      "loss": 1.2185,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2006445031939705,
+      "learning_rate": 1.9492209079673553e-05,
+      "loss": 1.2278,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7519770001679946,
+      "learning_rate": 1.9491339202292838e-05,
+      "loss": 1.2944,
+      "step": 1511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.604661343418237,
+      "learning_rate": 1.949046859991628e-05,
+      "loss": 1.2561,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.344854760184383,
+      "learning_rate": 1.9489597272610377e-05,
+      "loss": 1.1562,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2330611312809907,
+      "learning_rate": 1.948872522044169e-05,
+      "loss": 1.2202,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.255731105280065,
+      "learning_rate": 1.9487852443476828e-05,
+      "loss": 1.3076,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.386324968648969,
+      "learning_rate": 1.9486978941782454e-05,
+      "loss": 1.2788,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.859741234765031,
+      "learning_rate": 1.9486104715425295e-05,
+      "loss": 1.3423,
+      "step": 1517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.249471063309411,
+      "learning_rate": 1.948522976447213e-05,
+      "loss": 1.2886,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3206528879042314,
+      "learning_rate": 1.9484354088989793e-05,
+      "loss": 1.248,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7201537675371403,
+      "learning_rate": 1.9483477689045163e-05,
+      "loss": 1.2793,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1312043023201235,
+      "learning_rate": 1.9482600564705193e-05,
+      "loss": 1.2029,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.218580376770059,
+      "learning_rate": 1.9481722716036883e-05,
+      "loss": 1.2778,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7626191737186367,
+      "learning_rate": 1.948084414310728e-05,
+      "loss": 1.2529,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.399839666543512,
+      "learning_rate": 1.9479964845983495e-05,
+      "loss": 1.1311,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.175231215427761,
+      "learning_rate": 1.9479084824732702e-05,
+      "loss": 1.2703,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2945623057477733,
+      "learning_rate": 1.9478204079422112e-05,
+      "loss": 1.3022,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.294007421049925,
+      "learning_rate": 1.9477322610119005e-05,
+      "loss": 1.2979,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3473084466894867,
+      "learning_rate": 1.947644041689071e-05,
+      "loss": 1.3203,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1315246712233966,
+      "learning_rate": 1.9475557499804614e-05,
+      "loss": 1.2164,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.646430289337839,
+      "learning_rate": 1.947467385892816e-05,
+      "loss": 1.2034,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3166887684757893,
+      "learning_rate": 1.9473789494328844e-05,
+      "loss": 1.3276,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5366265649633393,
+      "learning_rate": 1.9472904406074217e-05,
+      "loss": 1.3154,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.133298488834274,
+      "learning_rate": 1.947201859423189e-05,
+      "loss": 1.2417,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.602202438069618,
+      "learning_rate": 1.947113205886952e-05,
+      "loss": 1.1675,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.121612140839086,
+      "learning_rate": 1.947024480005483e-05,
+      "loss": 1.2646,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8752921103925524,
+      "learning_rate": 1.9469356817855593e-05,
+      "loss": 1.2729,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5617590842752684,
+      "learning_rate": 1.9468468112339633e-05,
+      "loss": 1.2446,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8537047884512403,
+      "learning_rate": 1.9467578683574836e-05,
+      "loss": 1.1929,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.095150060994252,
+      "learning_rate": 1.946668853162914e-05,
+      "loss": 1.2979,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.140272348891808,
+      "learning_rate": 1.9465797656570546e-05,
+      "loss": 1.2517,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7607203825687727,
+      "learning_rate": 1.9464906058467097e-05,
+      "loss": 1.2407,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.072425402444351,
+      "learning_rate": 1.9464013737386898e-05,
+      "loss": 1.2625,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.43822390715365,
+      "learning_rate": 1.9463120693398113e-05,
+      "loss": 1.2542,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.173370646737976,
+      "learning_rate": 1.9462226926568954e-05,
+      "loss": 1.2397,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.303218554884006,
+      "learning_rate": 1.946133243696769e-05,
+      "loss": 1.2024,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.809980781906535,
+      "learning_rate": 1.9460437224662644e-05,
+      "loss": 1.2456,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.030396525349968,
+      "learning_rate": 1.9459541289722206e-05,
+      "loss": 1.2781,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6318098467592046,
+      "learning_rate": 1.9458644632214808e-05,
+      "loss": 1.3232,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4312254999145733,
+      "learning_rate": 1.9457747252208936e-05,
+      "loss": 1.3467,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5212352999681915,
+      "learning_rate": 1.9456849149773142e-05,
+      "loss": 1.2065,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.207538775698287,
+      "learning_rate": 1.9455950324976026e-05,
+      "loss": 1.2209,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.646286105214462,
+      "learning_rate": 1.9455050777886245e-05,
+      "loss": 1.2524,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.601193447660826,
+      "learning_rate": 1.945415050857251e-05,
+      "loss": 1.2778,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.259719015905912,
+      "learning_rate": 1.945324951710359e-05,
+      "loss": 1.2363,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.124363208538889,
+      "learning_rate": 1.9452347803548307e-05,
+      "loss": 1.2388,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.133113890732492,
+      "learning_rate": 1.9451445367975535e-05,
+      "loss": 1.239,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1957268473944573,
+      "learning_rate": 1.945054221045421e-05,
+      "loss": 1.2112,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.725988183736037,
+      "learning_rate": 1.944963833105332e-05,
+      "loss": 1.2344,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9963148093331142,
+      "learning_rate": 1.9448733729841903e-05,
+      "loss": 1.2297,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0059444333039247,
+      "learning_rate": 1.944782840688906e-05,
+      "loss": 1.2393,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6116665800950907,
+      "learning_rate": 1.944692236226395e-05,
+      "loss": 1.3276,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.345386568649925,
+      "learning_rate": 1.944601559603577e-05,
+      "loss": 1.2578,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.126849326648636,
+      "learning_rate": 1.9445108108273792e-05,
+      "loss": 1.3296,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.688579721247993,
+      "learning_rate": 1.944419989904733e-05,
+      "loss": 1.25,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.131807348104793,
+      "learning_rate": 1.944329096842576e-05,
+      "loss": 1.1904,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.626041540708689,
+      "learning_rate": 1.944238131647851e-05,
+      "loss": 1.2476,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.37592225862834,
+      "learning_rate": 1.944147094327506e-05,
+      "loss": 1.3137,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0409858485634054,
+      "learning_rate": 1.9440559848884954e-05,
+      "loss": 1.2378,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4634643376353593,
+      "learning_rate": 1.9439648033377785e-05,
+      "loss": 1.2034,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.921503567897164,
+      "learning_rate": 1.94387354968232e-05,
+      "loss": 1.2341,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9046991506743542,
+      "learning_rate": 1.9437822239290903e-05,
+      "loss": 1.2109,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.257737531359712,
+      "learning_rate": 1.943690826085066e-05,
+      "loss": 1.2727,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.205768305151787,
+      "learning_rate": 1.9435993561572272e-05,
+      "loss": 1.2375,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.248085060027069,
+      "learning_rate": 1.9435078141525622e-05,
+      "loss": 1.2622,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.139700964564949,
+      "learning_rate": 1.9434162000780626e-05,
+      "loss": 1.2537,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4460785205053113,
+      "learning_rate": 1.9433245139407266e-05,
+      "loss": 1.2202,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6172603451539165,
+      "learning_rate": 1.9432327557475575e-05,
+      "loss": 1.2898,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5617608283348483,
+      "learning_rate": 1.9431409255055644e-05,
+      "loss": 1.2891,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.113156574031139,
+      "learning_rate": 1.943049023221762e-05,
+      "loss": 1.2043,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2570497388308093,
+      "learning_rate": 1.9429570489031696e-05,
+      "loss": 1.2612,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7481740544316478,
+      "learning_rate": 1.942865002556813e-05,
+      "loss": 1.1987,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5052099170924493,
+      "learning_rate": 1.9427728841897234e-05,
+      "loss": 1.2422,
+      "step": 1582
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.689047741226769,
+      "learning_rate": 1.942680693808937e-05,
+      "loss": 1.2881,
+      "step": 1583
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9875765329485255,
+      "learning_rate": 1.942588431421496e-05,
+      "loss": 1.1934,
+      "step": 1584
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1378595871539416,
+      "learning_rate": 1.942496097034447e-05,
+      "loss": 1.2419,
+      "step": 1585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1015280673916705,
+      "learning_rate": 1.942403690654844e-05,
+      "loss": 1.3035,
+      "step": 1586
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9305641480790263,
+      "learning_rate": 1.942311212289745e-05,
+      "loss": 1.1648,
+      "step": 1587
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3193069443647305,
+      "learning_rate": 1.9422186619462144e-05,
+      "loss": 1.1162,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5278974380501715,
+      "learning_rate": 1.942126039631321e-05,
+      "loss": 1.2988,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4578198518116827,
+      "learning_rate": 1.9420333453521404e-05,
+      "loss": 1.2563,
+      "step": 1590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.434675795611355,
+      "learning_rate": 1.9419405791157527e-05,
+      "loss": 1.2642,
+      "step": 1591
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3084120352891415,
+      "learning_rate": 1.9418477409292437e-05,
+      "loss": 1.2437,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2409129368499503,
+      "learning_rate": 1.9417548307997046e-05,
+      "loss": 1.1448,
+      "step": 1593
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0759674979239895,
+      "learning_rate": 1.9416618487342333e-05,
+      "loss": 1.2305,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.60118727190382,
+      "learning_rate": 1.9415687947399313e-05,
+      "loss": 1.2207,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.215675327856688,
+      "learning_rate": 1.9414756688239073e-05,
+      "loss": 1.2036,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.234562397333244,
+      "learning_rate": 1.941382470993274e-05,
+      "loss": 1.2783,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2069744220942216,
+      "learning_rate": 1.9412892012551507e-05,
+      "loss": 1.2439,
+      "step": 1598
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.546247510424465,
+      "learning_rate": 1.9411958596166615e-05,
+      "loss": 1.3032,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.390234997151838,
+      "learning_rate": 1.941102446084937e-05,
+      "loss": 1.2937,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.971764137561138,
+      "learning_rate": 1.9410089606671118e-05,
+      "loss": 1.2109,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.122956368191748,
+      "learning_rate": 1.9409154033703276e-05,
+      "loss": 1.2671,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.506271007009762,
+      "learning_rate": 1.94082177420173e-05,
+      "loss": 1.2876,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.238102563537781,
+      "learning_rate": 1.940728073168471e-05,
+      "loss": 1.2085,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2921501591105504,
+      "learning_rate": 1.9406343002777084e-05,
+      "loss": 1.2163,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.023354146221781,
+      "learning_rate": 1.9405404555366046e-05,
+      "loss": 1.208,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.420140473069377,
+      "learning_rate": 1.940446538952328e-05,
+      "loss": 1.2656,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0109871264491983,
+      "learning_rate": 1.9403525505320527e-05,
+      "loss": 1.2061,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1733043760826014,
+      "learning_rate": 1.9402584902829575e-05,
+      "loss": 1.292,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4442752195197484,
+      "learning_rate": 1.9401643582122277e-05,
+      "loss": 1.2786,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.384631292312878,
+      "learning_rate": 1.9400701543270535e-05,
+      "loss": 1.2549,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9375864778538507,
+      "learning_rate": 1.9399758786346305e-05,
+      "loss": 1.1929,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1529648863455737,
+      "learning_rate": 1.93988153114216e-05,
+      "loss": 1.2271,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.015811446366891,
+      "learning_rate": 1.9397871118568485e-05,
+      "loss": 1.2529,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.80468279089705,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 1.1873,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.12298150617407,
+      "learning_rate": 1.9395980579365576e-05,
+      "loss": 1.1609,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.048915430617538,
+      "learning_rate": 1.9395034233160194e-05,
+      "loss": 1.2324,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3493988290683125,
+      "learning_rate": 1.9394087169315217e-05,
+      "loss": 1.2185,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3170338101221937,
+      "learning_rate": 1.9393139387902993e-05,
+      "loss": 1.3076,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.043182487810276,
+      "learning_rate": 1.9392190888995918e-05,
+      "loss": 1.166,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4144863448247014,
+      "learning_rate": 1.9391241672666438e-05,
+      "loss": 1.2705,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7630894595843603,
+      "learning_rate": 1.9390291738987066e-05,
+      "loss": 1.2271,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.688110444111173,
+      "learning_rate": 1.938934108803036e-05,
+      "loss": 1.2944,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.360363982830912,
+      "learning_rate": 1.938838971986893e-05,
+      "loss": 1.2397,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9974822831433134,
+      "learning_rate": 1.9387437634575456e-05,
+      "loss": 1.1172,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4153814464599512,
+      "learning_rate": 1.9386484832222652e-05,
+      "loss": 1.3008,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0909408641495233,
+      "learning_rate": 1.938553131288331e-05,
+      "loss": 1.2019,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1432952767899347,
+      "learning_rate": 1.9384577076630252e-05,
+      "loss": 1.2605,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1756564394535616,
+      "learning_rate": 1.9383622123536376e-05,
+      "loss": 1.4072,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.159540455602768,
+      "learning_rate": 1.9382666453674625e-05,
+      "loss": 1.2534,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7421199660020736,
+      "learning_rate": 1.9381710067117995e-05,
+      "loss": 1.1641,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.292491322897357,
+      "learning_rate": 1.9380752963939545e-05,
+      "loss": 1.3296,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.911190262732764,
+      "learning_rate": 1.9379795144212374e-05,
+      "loss": 1.1902,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8158906089792377,
+      "learning_rate": 1.9378836608009655e-05,
+      "loss": 1.1086,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8310253602616053,
+      "learning_rate": 1.9377877355404596e-05,
+      "loss": 1.2466,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9241287706605137,
+      "learning_rate": 1.937691738647048e-05,
+      "loss": 1.2043,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.105509885278813,
+      "learning_rate": 1.9375956701280625e-05,
+      "loss": 1.2112,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4769204571565737,
+      "learning_rate": 1.937499529990842e-05,
+      "loss": 1.1482,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6712058284420324,
+      "learning_rate": 1.9374033182427297e-05,
+      "loss": 1.1543,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.945004730802144,
+      "learning_rate": 1.9373070348910747e-05,
+      "loss": 1.1775,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8685917895911572,
+      "learning_rate": 1.9372106799432318e-05,
+      "loss": 1.2944,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.857748001895151,
+      "learning_rate": 1.9371142534065613e-05,
+      "loss": 1.2449,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.982455412932375,
+      "learning_rate": 1.9370177552884287e-05,
+      "loss": 1.2954,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.094471267815526,
+      "learning_rate": 1.9369211855962044e-05,
+      "loss": 1.2139,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.477593354111286,
+      "learning_rate": 1.9368245443372652e-05,
+      "loss": 1.2078,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.851974946410804,
+      "learning_rate": 1.9367278315189938e-05,
+      "loss": 1.314,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0435390077331603,
+      "learning_rate": 1.9366310471487763e-05,
+      "loss": 1.2571,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.648507796109113,
+      "learning_rate": 1.936534191234006e-05,
+      "loss": 1.2529,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7760947005614467,
+      "learning_rate": 1.936437263782082e-05,
+      "loss": 1.2034,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6360737141998154,
+      "learning_rate": 1.9363402648004074e-05,
+      "loss": 1.26,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1751810789155885,
+      "learning_rate": 1.9362431942963915e-05,
+      "loss": 1.187,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7710613280653655,
+      "learning_rate": 1.9361460522774493e-05,
+      "loss": 1.3486,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4956273031478657,
+      "learning_rate": 1.9360488387510007e-05,
+      "loss": 1.1836,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.803511235181797,
+      "learning_rate": 1.9359515537244715e-05,
+      "loss": 1.2502,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.404919275804002,
+      "learning_rate": 1.9358541972052926e-05,
+      "loss": 1.2061,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.160722015270287,
+      "learning_rate": 1.935756769200901e-05,
+      "loss": 1.2871,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5911200500464737,
+      "learning_rate": 1.9356592697187383e-05,
+      "loss": 1.2622,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2158915876430303,
+      "learning_rate": 1.9355616987662526e-05,
+      "loss": 1.2339,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1287120632034107,
+      "learning_rate": 1.9354640563508962e-05,
+      "loss": 1.2551,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.92185866591863,
+      "learning_rate": 1.935366342480128e-05,
+      "loss": 1.2998,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1371253402998627,
+      "learning_rate": 1.9352685571614117e-05,
+      "loss": 1.137,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.098933445711985,
+      "learning_rate": 1.9351707004022162e-05,
+      "loss": 1.3005,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.068423659374869,
+      "learning_rate": 1.935072772210017e-05,
+      "loss": 1.2544,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8379272869935765,
+      "learning_rate": 1.9349747725922937e-05,
+      "loss": 1.22,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.177319002434898,
+      "learning_rate": 1.9348767015565325e-05,
+      "loss": 1.2593,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.905046649463948,
+      "learning_rate": 1.9347785591102244e-05,
+      "loss": 1.248,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4672364865545435,
+      "learning_rate": 1.934680345260866e-05,
+      "loss": 1.2285,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2083017293710188,
+      "learning_rate": 1.9345820600159593e-05,
+      "loss": 1.2661,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.409177722646234,
+      "learning_rate": 1.9344837033830116e-05,
+      "loss": 1.2893,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.180479281782153,
+      "learning_rate": 1.9343852753695366e-05,
+      "loss": 1.2085,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.859683485926647,
+      "learning_rate": 1.934286775983052e-05,
+      "loss": 1.2231,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.251560508808991,
+      "learning_rate": 1.934188205231082e-05,
+      "loss": 1.1914,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.641779685529702,
+      "learning_rate": 1.9340895631211557e-05,
+      "loss": 1.3296,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.216966737311048,
+      "learning_rate": 1.933990849660808e-05,
+      "loss": 1.2151,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9030279099414398,
+      "learning_rate": 1.9338920648575798e-05,
+      "loss": 1.2432,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4333524952846317,
+      "learning_rate": 1.9337932087190154e-05,
+      "loss": 1.312,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.503955186001922,
+      "learning_rate": 1.9336942812526667e-05,
+      "loss": 1.2708,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.686116038920826,
+      "learning_rate": 1.9335952824660908e-05,
+      "loss": 1.2605,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.4158915696958765,
+      "learning_rate": 1.9334962123668487e-05,
+      "loss": 1.26,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.642315012094044,
+      "learning_rate": 1.9333970709625083e-05,
+      "loss": 1.136,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7022162358655075,
+      "learning_rate": 1.9332978582606428e-05,
+      "loss": 1.1875,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3838514803899775,
+      "learning_rate": 1.93319857426883e-05,
+      "loss": 1.2246,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.233147705247238,
+      "learning_rate": 1.9330992189946542e-05,
+      "loss": 1.1965,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2157770405037884,
+      "learning_rate": 1.9329997924457046e-05,
+      "loss": 1.2666,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.943787391058571,
+      "learning_rate": 1.9329002946295755e-05,
+      "loss": 1.3135,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.258870732244476,
+      "learning_rate": 1.932800725553867e-05,
+      "loss": 1.3438,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9195567682621975,
+      "learning_rate": 1.9327010852261855e-05,
+      "loss": 1.2438,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.137841586155768,
+      "learning_rate": 1.9326013736541416e-05,
+      "loss": 1.2192,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.966673709172504,
+      "learning_rate": 1.9325015908453512e-05,
+      "loss": 1.2773,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.135467478505036,
+      "learning_rate": 1.932401736807437e-05,
+      "loss": 1.2068,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.300976810721223,
+      "learning_rate": 1.932301811548026e-05,
+      "loss": 1.2585,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.235914571191927,
+      "learning_rate": 1.932201815074751e-05,
+      "loss": 1.209,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.447798735134417,
+      "learning_rate": 1.93210174739525e-05,
+      "loss": 1.2786,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0028931955059415,
+      "learning_rate": 1.9320016085171672e-05,
+      "loss": 1.2686,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.431125984893807,
+      "learning_rate": 1.9319013984481514e-05,
+      "loss": 1.2942,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.236056058253058,
+      "learning_rate": 1.931801117195857e-05,
+      "loss": 1.2515,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.306421356020028,
+      "learning_rate": 1.9317007647679444e-05,
+      "loss": 1.2566,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.689101119078642,
+      "learning_rate": 1.9316003411720785e-05,
+      "loss": 1.2046,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6571511285028215,
+      "learning_rate": 1.9314998464159305e-05,
+      "loss": 1.3477,
+      "step": 1699
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7781661401688735,
+      "learning_rate": 1.9313992805071765e-05,
+      "loss": 1.1843,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6696732108432344,
+      "learning_rate": 1.9312986434534984e-05,
+      "loss": 1.1875,
+      "step": 1701
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8309179530075643,
+      "learning_rate": 1.9311979352625837e-05,
+      "loss": 1.2498,
+      "step": 1702
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9817863338893105,
+      "learning_rate": 1.9310971559421242e-05,
+      "loss": 1.249,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.475401566785018,
+      "learning_rate": 1.9309963054998183e-05,
+      "loss": 1.1545,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.23582032063807,
+      "learning_rate": 1.93089538394337e-05,
+      "loss": 1.2048,
+      "step": 1705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5296098700147525,
+      "learning_rate": 1.930794391280487e-05,
+      "loss": 1.2944,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8923526419421557,
+      "learning_rate": 1.9306933275188848e-05,
+      "loss": 1.1455,
+      "step": 1707
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2161033264303547,
+      "learning_rate": 1.930592192666282e-05,
+      "loss": 1.2319,
+      "step": 1708
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6383612158312726,
+      "learning_rate": 1.9304909867304052e-05,
+      "loss": 1.2646,
+      "step": 1709
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1245414506237146,
+      "learning_rate": 1.930389709718984e-05,
+      "loss": 1.312,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3751276174784817,
+      "learning_rate": 1.9302883616397546e-05,
+      "loss": 1.2715,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1691363764915073,
+      "learning_rate": 1.9301869425004586e-05,
+      "loss": 1.2026,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.331107845262484,
+      "learning_rate": 1.9300854523088426e-05,
+      "loss": 1.2634,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.264540886769116,
+      "learning_rate": 1.9299838910726594e-05,
+      "loss": 1.2742,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.26400921486155,
+      "learning_rate": 1.9298822587996668e-05,
+      "loss": 1.2373,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.491680541833413,
+      "learning_rate": 1.9297805554976276e-05,
+      "loss": 1.2153,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3497728838925243,
+      "learning_rate": 1.9296787811743104e-05,
+      "loss": 1.2798,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8478297994557926,
+      "learning_rate": 1.9295769358374894e-05,
+      "loss": 1.1806,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5777410556360087,
+      "learning_rate": 1.929475019494944e-05,
+      "loss": 1.2208,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.215885332419256,
+      "learning_rate": 1.929373032154459e-05,
+      "loss": 1.2344,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.109875711933387,
+      "learning_rate": 1.929270973823825e-05,
+      "loss": 1.2764,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.921661588074297,
+      "learning_rate": 1.9291688445108378e-05,
+      "loss": 1.3135,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.059162709206627,
+      "learning_rate": 1.9290666442232976e-05,
+      "loss": 1.2344,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6345491537651506,
+      "learning_rate": 1.928964372969012e-05,
+      "loss": 1.1802,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.694836160517738,
+      "learning_rate": 1.9288620307557926e-05,
+      "loss": 1.2527,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.439550557752121,
+      "learning_rate": 1.9287596175914568e-05,
+      "loss": 1.283,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8868311869350136,
+      "learning_rate": 1.9286571334838277e-05,
+      "loss": 1.3091,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.177599448174772,
+      "learning_rate": 1.928554578440733e-05,
+      "loss": 1.2363,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3953666151606208,
+      "learning_rate": 1.928451952470007e-05,
+      "loss": 1.2468,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.151159853006159,
+      "learning_rate": 1.928349255579488e-05,
+      "loss": 1.1873,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3052275368020614,
+      "learning_rate": 1.9282464877770213e-05,
+      "loss": 1.3076,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1500697876416233,
+      "learning_rate": 1.928143649070456e-05,
+      "loss": 1.1724,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6534944868541803,
+      "learning_rate": 1.9280407394676484e-05,
+      "loss": 1.2268,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4839543542531928,
+      "learning_rate": 1.9279377589764586e-05,
+      "loss": 1.2759,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3084497364142553,
+      "learning_rate": 1.9278347076047533e-05,
+      "loss": 1.2014,
+      "step": 1735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.463784261473901,
+      "learning_rate": 1.9277315853604033e-05,
+      "loss": 1.2837,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.179886022668871,
+      "learning_rate": 1.927628392251286e-05,
+      "loss": 1.2153,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5022485685714053,
+      "learning_rate": 1.927525128285284e-05,
+      "loss": 1.3223,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7093376148185797,
+      "learning_rate": 1.927421793470285e-05,
+      "loss": 1.2,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1006956258592107,
+      "learning_rate": 1.927318387814182e-05,
+      "loss": 1.1763,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3278912744253475,
+      "learning_rate": 1.9272149113248737e-05,
+      "loss": 1.1528,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3474761200563936,
+      "learning_rate": 1.927111364010264e-05,
+      "loss": 1.2402,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4059354160820727,
+      "learning_rate": 1.927007745878263e-05,
+      "loss": 1.2563,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.563667215294903,
+      "learning_rate": 1.926904056936785e-05,
+      "loss": 1.224,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.021695446977837,
+      "learning_rate": 1.9268002971937508e-05,
+      "loss": 1.2173,
+      "step": 1745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.104947427607811,
+      "learning_rate": 1.9266964666570852e-05,
+      "loss": 1.1606,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.638043247711128,
+      "learning_rate": 1.9265925653347198e-05,
+      "loss": 1.1934,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4581296959086223,
+      "learning_rate": 1.9264885932345917e-05,
+      "loss": 1.2905,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.29091386751439,
+      "learning_rate": 1.926384550364642e-05,
+      "loss": 1.2705,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0617287866574396,
+      "learning_rate": 1.926280436732818e-05,
+      "loss": 1.2798,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.387638885658214,
+      "learning_rate": 1.9261762523470727e-05,
+      "loss": 1.2815,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.4517165573026745,
+      "learning_rate": 1.9260719972153644e-05,
+      "loss": 1.3232,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.116269490341339,
+      "learning_rate": 1.9259676713456565e-05,
+      "loss": 1.2664,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2676574162733076,
+      "learning_rate": 1.9258632747459176e-05,
+      "loss": 1.1707,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.008675984577834,
+      "learning_rate": 1.9257588074241225e-05,
+      "loss": 1.2612,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.017568257449081,
+      "learning_rate": 1.9256542693882505e-05,
+      "loss": 1.1936,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4260240390481065,
+      "learning_rate": 1.9255496606462872e-05,
+      "loss": 1.2139,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.185541982020421,
+      "learning_rate": 1.9254449812062227e-05,
+      "loss": 1.1902,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9014347198284707,
+      "learning_rate": 1.9253402310760532e-05,
+      "loss": 1.199,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.975648753242241,
+      "learning_rate": 1.9252354102637797e-05,
+      "loss": 1.2153,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.029851412554894,
+      "learning_rate": 1.9251305187774096e-05,
+      "loss": 1.2505,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5576272404644125,
+      "learning_rate": 1.9250255566249546e-05,
+      "loss": 1.2861,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.254434378065944,
+      "learning_rate": 1.9249205238144322e-05,
+      "loss": 1.2007,
+      "step": 1763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1514824571915496,
+      "learning_rate": 1.9248154203538653e-05,
+      "loss": 1.1575,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.118863558810055,
+      "learning_rate": 1.9247102462512823e-05,
+      "loss": 1.2217,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1286525770824305,
+      "learning_rate": 1.924605001514717e-05,
+      "loss": 1.2646,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8286400473849573,
+      "learning_rate": 1.9244996861522085e-05,
+      "loss": 1.2637,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.186706682519259,
+      "learning_rate": 1.924394300171801e-05,
+      "loss": 1.2959,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.752828766507837,
+      "learning_rate": 1.9242888435815455e-05,
+      "loss": 1.2307,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.200781156419106,
+      "learning_rate": 1.924183316389496e-05,
+      "loss": 1.2202,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.289184770922107,
+      "learning_rate": 1.9240777186037137e-05,
+      "loss": 1.2092,
+      "step": 1771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.724643962394464,
+      "learning_rate": 1.9239720502322647e-05,
+      "loss": 1.189,
+      "step": 1772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9544115019249406,
+      "learning_rate": 1.92386631128322e-05,
+      "loss": 1.3184,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9454714435914453,
+      "learning_rate": 1.9237605017646573e-05,
+      "loss": 1.2483,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.049659542774044,
+      "learning_rate": 1.9236546216846584e-05,
+      "loss": 1.2866,
+      "step": 1775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.989740300675218,
+      "learning_rate": 1.9235486710513114e-05,
+      "loss": 1.2578,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.450671903494075,
+      "learning_rate": 1.9234426498727085e-05,
+      "loss": 1.3003,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.655215124169079,
+      "learning_rate": 1.9233365581569486e-05,
+      "loss": 1.2866,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.304139800796684,
+      "learning_rate": 1.923230395912136e-05,
+      "loss": 1.1694,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2745674824347204,
+      "learning_rate": 1.923124163146379e-05,
+      "loss": 1.2605,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.916223100567053,
+      "learning_rate": 1.9230178598677924e-05,
+      "loss": 1.2966,
+      "step": 1781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7651040748731917,
+      "learning_rate": 1.9229114860844968e-05,
+      "loss": 1.2383,
+      "step": 1782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.591053288848492,
+      "learning_rate": 1.922805041804617e-05,
+      "loss": 1.2717,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.016168148751521,
+      "learning_rate": 1.9226985270362837e-05,
+      "loss": 1.1763,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1072673916554527,
+      "learning_rate": 1.9225919417876332e-05,
+      "loss": 1.2185,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9724003990929635,
+      "learning_rate": 1.9224852860668072e-05,
+      "loss": 1.2087,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.215779923719141,
+      "learning_rate": 1.9223785598819523e-05,
+      "loss": 1.2715,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.99095769900591,
+      "learning_rate": 1.9222717632412208e-05,
+      "loss": 1.2852,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2934594276651383,
+      "learning_rate": 1.922164896152771e-05,
+      "loss": 1.1895,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2465296222709643,
+      "learning_rate": 1.9220579586247647e-05,
+      "loss": 1.2612,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.335360787215781,
+      "learning_rate": 1.9219509506653712e-05,
+      "loss": 1.2822,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8199294030194317,
+      "learning_rate": 1.9218438722827644e-05,
+      "loss": 1.2246,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.068417951047812,
+      "learning_rate": 1.921736723485123e-05,
+      "loss": 1.2546,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.0105318193551565,
+      "learning_rate": 1.921629504280632e-05,
+      "loss": 1.2698,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0438380145993236,
+      "learning_rate": 1.9215222146774806e-05,
+      "loss": 1.2292,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3092143947139117,
+      "learning_rate": 1.9214148546838647e-05,
+      "loss": 1.3628,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1649700751373464,
+      "learning_rate": 1.921307424307985e-05,
+      "loss": 1.2515,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7432733059414645,
+      "learning_rate": 1.9211999235580475e-05,
+      "loss": 1.2312,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6862727507244792,
+      "learning_rate": 1.921092352442264e-05,
+      "loss": 1.2126,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.240641775626373,
+      "learning_rate": 1.9209847109688506e-05,
+      "loss": 1.269,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8537290859311093,
+      "learning_rate": 1.9208769991460298e-05,
+      "loss": 1.2129,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.353167579688052,
+      "learning_rate": 1.920769216982029e-05,
+      "loss": 1.2549,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.223218550610136,
+      "learning_rate": 1.9206613644850817e-05,
+      "loss": 1.2295,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6917735139744203,
+      "learning_rate": 1.9205534416634254e-05,
+      "loss": 1.2734,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.475931664089166,
+      "learning_rate": 1.9204454485253046e-05,
+      "loss": 1.2832,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6812233490451782,
+      "learning_rate": 1.920337385078968e-05,
+      "loss": 1.269,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8925135879171866,
+      "learning_rate": 1.9202292513326698e-05,
+      "loss": 1.2671,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7886779440077927,
+      "learning_rate": 1.92012104729467e-05,
+      "loss": 1.3789,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2879552386099666,
+      "learning_rate": 1.920012772973234e-05,
+      "loss": 1.229,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9359024540354413,
+      "learning_rate": 1.919904428376632e-05,
+      "loss": 1.3062,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5315021789068997,
+      "learning_rate": 1.91979601351314e-05,
+      "loss": 1.24,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9316693021616267,
+      "learning_rate": 1.919687528391039e-05,
+      "loss": 1.3213,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.765680094198171,
+      "learning_rate": 1.919578973018616e-05,
+      "loss": 1.2246,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0224665553914307,
+      "learning_rate": 1.9194703474041632e-05,
+      "loss": 1.2715,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.391322113173683,
+      "learning_rate": 1.9193616515559775e-05,
+      "loss": 1.0984,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.509738400759504,
+      "learning_rate": 1.919252885482362e-05,
+      "loss": 1.3682,
+      "step": 1816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8089278042840617,
+      "learning_rate": 1.9191440491916244e-05,
+      "loss": 1.2507,
+      "step": 1817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.576752486615811,
+      "learning_rate": 1.9190351426920785e-05,
+      "loss": 1.2306,
+      "step": 1818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.778742117110999,
+      "learning_rate": 1.9189261659920428e-05,
+      "loss": 1.2358,
+      "step": 1819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.280824753791218,
+      "learning_rate": 1.9188171190998416e-05,
+      "loss": 1.1631,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.754867384770997,
+      "learning_rate": 1.9187080020238046e-05,
+      "loss": 1.2314,
+      "step": 1821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9706601803981503,
+      "learning_rate": 1.9185988147722664e-05,
+      "loss": 1.3281,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.628907761282862,
+      "learning_rate": 1.9184895573535678e-05,
+      "loss": 1.188,
+      "step": 1823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.93254004225883,
+      "learning_rate": 1.9183802297760538e-05,
+      "loss": 1.21,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.664251414783912,
+      "learning_rate": 1.918270832048076e-05,
+      "loss": 1.2778,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6100631035764637,
+      "learning_rate": 1.9181613641779905e-05,
+      "loss": 1.2117,
+      "step": 1826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.652106904571406,
+      "learning_rate": 1.9180518261741586e-05,
+      "loss": 1.2239,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.557109573772524,
+      "learning_rate": 1.9179422180449478e-05,
+      "loss": 1.1931,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.352082156501959,
+      "learning_rate": 1.9178325397987305e-05,
+      "loss": 1.2458,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.160712804862022,
+      "learning_rate": 1.9177227914438844e-05,
+      "loss": 1.2051,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.143256121414249,
+      "learning_rate": 1.9176129729887925e-05,
+      "loss": 1.174,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4289919950698633,
+      "learning_rate": 1.9175030844418435e-05,
+      "loss": 1.2188,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9185794283565887,
+      "learning_rate": 1.917393125811431e-05,
+      "loss": 1.2432,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.095875496232119,
+      "learning_rate": 1.9172830971059546e-05,
+      "loss": 1.208,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3535581333607785,
+      "learning_rate": 1.9171729983338184e-05,
+      "loss": 1.2533,
+      "step": 1835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5726243942266134,
+      "learning_rate": 1.9170628295034324e-05,
+      "loss": 1.2751,
+      "step": 1836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.013317319381145,
+      "learning_rate": 1.916952590623212e-05,
+      "loss": 1.2039,
+      "step": 1837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.080768220175902,
+      "learning_rate": 1.9168422817015778e-05,
+      "loss": 1.1968,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3960846039943076,
+      "learning_rate": 1.9167319027469554e-05,
+      "loss": 1.2546,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9059626111287464,
+      "learning_rate": 1.916621453767776e-05,
+      "loss": 1.2297,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.604338447839148,
+      "learning_rate": 1.9165109347724774e-05,
+      "loss": 1.2671,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.12115994148702,
+      "learning_rate": 1.9164003457695e-05,
+      "loss": 1.209,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.461665658418847,
+      "learning_rate": 1.9162896867672926e-05,
+      "loss": 1.2656,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.818603904630935,
+      "learning_rate": 1.9161789577743067e-05,
+      "loss": 1.2363,
+      "step": 1844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0925025801763026,
+      "learning_rate": 1.916068158799001e-05,
+      "loss": 1.2048,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2441843965472352,
+      "learning_rate": 1.9159572898498387e-05,
+      "loss": 1.2087,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4804731781339466,
+      "learning_rate": 1.9158463509352887e-05,
+      "loss": 1.2715,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2906693635088122,
+      "learning_rate": 1.9157353420638248e-05,
+      "loss": 1.1826,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.426933775569643,
+      "learning_rate": 1.9156242632439264e-05,
+      "loss": 1.2163,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9789221406465787,
+      "learning_rate": 1.9155131144840784e-05,
+      "loss": 1.2451,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.110348591511144,
+      "learning_rate": 1.915401895792771e-05,
+      "loss": 1.2124,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9147372707778256,
+      "learning_rate": 1.9152906071784993e-05,
+      "loss": 1.2012,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8037926456994304,
+      "learning_rate": 1.9151792486497642e-05,
+      "loss": 1.2483,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.140611063602471,
+      "learning_rate": 1.915067820215072e-05,
+      "loss": 1.2078,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9234378119543196,
+      "learning_rate": 1.9149563218829345e-05,
+      "loss": 1.175,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4634912164199885,
+      "learning_rate": 1.9148447536618675e-05,
+      "loss": 1.3152,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1842082957524758,
+      "learning_rate": 1.9147331155603937e-05,
+      "loss": 1.1777,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.841180072984733,
+      "learning_rate": 1.914621407587041e-05,
+      "loss": 1.2944,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3460091055966177,
+      "learning_rate": 1.9145096297503416e-05,
+      "loss": 1.2283,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.736870841055135,
+      "learning_rate": 1.9143977820588337e-05,
+      "loss": 1.2402,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3413507644102687,
+      "learning_rate": 1.914285864521061e-05,
+      "loss": 1.2769,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4534807345609737,
+      "learning_rate": 1.9141738771455722e-05,
+      "loss": 1.3264,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0077160158721408,
+      "learning_rate": 1.914061819940922e-05,
+      "loss": 1.2537,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.598304120105751,
+      "learning_rate": 1.9139496929156685e-05,
+      "loss": 1.1658,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.878190676133189,
+      "learning_rate": 1.9138374960783782e-05,
+      "loss": 1.2148,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2427545496847707,
+      "learning_rate": 1.91372522943762e-05,
+      "loss": 1.2485,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.812534795486264,
+      "learning_rate": 1.91361289300197e-05,
+      "loss": 1.2539,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.171953872113459,
+      "learning_rate": 1.9135004867800088e-05,
+      "loss": 1.1721,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.570873328094927,
+      "learning_rate": 1.9133880107803223e-05,
+      "loss": 1.2397,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.849039655174978,
+      "learning_rate": 1.9132754650115028e-05,
+      "loss": 1.2847,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2550091876307827,
+      "learning_rate": 1.9131628494821464e-05,
+      "loss": 1.2188,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9070550303683413,
+      "learning_rate": 1.9130501642008553e-05,
+      "loss": 1.2749,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.918127537161049,
+      "learning_rate": 1.9129374091762375e-05,
+      "loss": 1.1809,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.273216876570527,
+      "learning_rate": 1.9128245844169048e-05,
+      "loss": 1.2646,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.952121525968949,
+      "learning_rate": 1.9127116899314764e-05,
+      "loss": 1.1899,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.656440236902288,
+      "learning_rate": 1.912598725728575e-05,
+      "loss": 1.3174,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1732621501850677,
+      "learning_rate": 1.91248569181683e-05,
+      "loss": 1.1797,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5266115118616255,
+      "learning_rate": 1.9123725882048743e-05,
+      "loss": 1.2827,
+      "step": 1878
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.488692900508122,
+      "learning_rate": 1.9122594149013488e-05,
+      "loss": 1.3047,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2672996103028766,
+      "learning_rate": 1.9121461719148977e-05,
+      "loss": 1.2593,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.188792249874168,
+      "learning_rate": 1.9120328592541702e-05,
+      "loss": 1.2437,
+      "step": 1881
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.267347818004438,
+      "learning_rate": 1.911919476927823e-05,
+      "loss": 1.2646,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.240582078991587,
+      "learning_rate": 1.911806024944516e-05,
+      "loss": 1.2144,
+      "step": 1883
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0067680553481213,
+      "learning_rate": 1.9116925033129158e-05,
+      "loss": 1.1323,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.258358780510333,
+      "learning_rate": 1.911578912041693e-05,
+      "loss": 1.2883,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.176005094576097,
+      "learning_rate": 1.9114652511395248e-05,
+      "loss": 1.238,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4298439410419097,
+      "learning_rate": 1.9113515206150933e-05,
+      "loss": 1.2502,
+      "step": 1887
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.896207820601942,
+      "learning_rate": 1.9112377204770853e-05,
+      "loss": 1.2305,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.723340552232769,
+      "learning_rate": 1.911123850734194e-05,
+      "loss": 1.1533,
+      "step": 1889
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0373259795628456,
+      "learning_rate": 1.9110099113951165e-05,
+      "loss": 1.2107,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9921856102364752,
+      "learning_rate": 1.910895902468557e-05,
+      "loss": 1.2346,
+      "step": 1891
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6553431444912863,
+      "learning_rate": 1.9107818239632234e-05,
+      "loss": 1.2749,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.924622249008107,
+      "learning_rate": 1.91066767588783e-05,
+      "loss": 1.1482,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.587827816664194,
+      "learning_rate": 1.9105534582510956e-05,
+      "loss": 1.2617,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.91322396651129,
+      "learning_rate": 1.9104391710617452e-05,
+      "loss": 1.2383,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.179213804013856,
+      "learning_rate": 1.9103248143285084e-05,
+      "loss": 1.3018,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4481236810990237,
+      "learning_rate": 1.9102103880601205e-05,
+      "loss": 1.207,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2718612209028106,
+      "learning_rate": 1.9100958922653214e-05,
+      "loss": 1.261,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6086815104169947,
+      "learning_rate": 1.9099813269528573e-05,
+      "loss": 1.176,
+      "step": 1899
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.323983103404713,
+      "learning_rate": 1.9098666921314793e-05,
+      "loss": 1.2119,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8898230719387303,
+      "learning_rate": 1.9097519878099436e-05,
+      "loss": 1.1909,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.541736093286506,
+      "learning_rate": 1.909637213997012e-05,
+      "loss": 1.2744,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1743290872618353,
+      "learning_rate": 1.9095223707014514e-05,
+      "loss": 1.2644,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4030176405768753,
+      "learning_rate": 1.909407457932034e-05,
+      "loss": 1.1765,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3944560326318802,
+      "learning_rate": 1.9092924756975378e-05,
+      "loss": 1.1738,
+      "step": 1905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.090155701640692,
+      "learning_rate": 1.9091774240067455e-05,
+      "loss": 1.2593,
+      "step": 1906
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.686583719457745,
+      "learning_rate": 1.9090623028684455e-05,
+      "loss": 1.23,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8385026121082966,
+      "learning_rate": 1.9089471122914308e-05,
+      "loss": 1.2373,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0608028134138943,
+      "learning_rate": 1.908831852284501e-05,
+      "loss": 1.3223,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8419147319540445,
+      "learning_rate": 1.9087165228564594e-05,
+      "loss": 1.2798,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4954363405961044,
+      "learning_rate": 1.908601124016116e-05,
+      "loss": 1.2573,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6444167002106953,
+      "learning_rate": 1.9084856557722854e-05,
+      "loss": 1.2163,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.126392130041255,
+      "learning_rate": 1.9083701181337876e-05,
+      "loss": 1.2175,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.995918664050125,
+      "learning_rate": 1.908254511109448e-05,
+      "loss": 1.2949,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.845767073803205,
+      "learning_rate": 1.908138834708097e-05,
+      "loss": 1.2397,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.212642805780223,
+      "learning_rate": 1.9080230889385705e-05,
+      "loss": 1.2402,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.078100935665212,
+      "learning_rate": 1.9079072738097106e-05,
+      "loss": 1.252,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6530611570983176,
+      "learning_rate": 1.907791389330363e-05,
+      "loss": 1.2456,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.351509599973998,
+      "learning_rate": 1.9076754355093797e-05,
+      "loss": 1.2549,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4875770760641625,
+      "learning_rate": 1.907559412355618e-05,
+      "loss": 1.2856,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.989275068131367,
+      "learning_rate": 1.90744331987794e-05,
+      "loss": 1.2588,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6506235577501767,
+      "learning_rate": 1.907327158085214e-05,
+      "loss": 1.3276,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0972476606477097,
+      "learning_rate": 1.9072109269863123e-05,
+      "loss": 1.2739,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.367511972433545,
+      "learning_rate": 1.9070946265901138e-05,
+      "loss": 1.2249,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2694460100453036,
+      "learning_rate": 1.906978256905502e-05,
+      "loss": 1.2629,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4865475079354806,
+      "learning_rate": 1.9068618179413652e-05,
+      "loss": 1.2786,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9190499361794857,
+      "learning_rate": 1.9067453097065986e-05,
+      "loss": 1.2056,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.628932171870218,
+      "learning_rate": 1.9066287322101007e-05,
+      "loss": 1.2559,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1675613140728447,
+      "learning_rate": 1.906512085460777e-05,
+      "loss": 1.2239,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1333813247505833,
+      "learning_rate": 1.906395369467537e-05,
+      "loss": 1.2251,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2565087754611373,
+      "learning_rate": 1.9062785842392967e-05,
+      "loss": 1.2319,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3083357040409194,
+      "learning_rate": 1.9061617297849758e-05,
+      "loss": 1.248,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3476185792352475,
+      "learning_rate": 1.9060448061135012e-05,
+      "loss": 1.2461,
+      "step": 1933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5997731370253714,
+      "learning_rate": 1.9059278132338038e-05,
+      "loss": 1.2515,
+      "step": 1934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1036797893299006,
+      "learning_rate": 1.9058107511548197e-05,
+      "loss": 1.197,
+      "step": 1935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.559809743420567,
+      "learning_rate": 1.905693619885491e-05,
+      "loss": 1.2146,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9624631752702415,
+      "learning_rate": 1.9055764194347647e-05,
+      "loss": 1.114,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0250828123407305,
+      "learning_rate": 1.9054591498115937e-05,
+      "loss": 1.2055,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5646544238953584,
+      "learning_rate": 1.9053418110249343e-05,
+      "loss": 1.29,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3585574645426908,
+      "learning_rate": 1.905224403083751e-05,
+      "loss": 1.2407,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.805703404115303,
+      "learning_rate": 1.9051069259970112e-05,
+      "loss": 1.2739,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9600365742968795,
+      "learning_rate": 1.9049893797736883e-05,
+      "loss": 1.186,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1515917476359334,
+      "learning_rate": 1.904871764422761e-05,
+      "loss": 1.2219,
+      "step": 1943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6387457649409187,
+      "learning_rate": 1.904754079953214e-05,
+      "loss": 1.2568,
+      "step": 1944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4759401035518795,
+      "learning_rate": 1.904636326374036e-05,
+      "loss": 1.1836,
+      "step": 1945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.774246629601335,
+      "learning_rate": 1.904518503694222e-05,
+      "loss": 1.2559,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1346015270753282,
+      "learning_rate": 1.9044006119227713e-05,
+      "loss": 1.1545,
+      "step": 1947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0956976830347407,
+      "learning_rate": 1.9042826510686895e-05,
+      "loss": 1.3403,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.27055744940041,
+      "learning_rate": 1.9041646211409876e-05,
+      "loss": 1.2559,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2485829440411638,
+      "learning_rate": 1.9040465221486803e-05,
+      "loss": 1.2065,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3620273284982978,
+      "learning_rate": 1.9039283541007888e-05,
+      "loss": 1.1912,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.471394290004182,
+      "learning_rate": 1.9038101170063395e-05,
+      "loss": 1.2065,
+      "step": 1952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8522691512471274,
+      "learning_rate": 1.9036918108743645e-05,
+      "loss": 1.2659,
+      "step": 1953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1297104603569843,
+      "learning_rate": 1.9035734357139e-05,
+      "loss": 1.1704,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.514469005681972,
+      "learning_rate": 1.903454991533988e-05,
+      "loss": 1.196,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8559878331046216,
+      "learning_rate": 1.903336478343676e-05,
+      "loss": 1.2698,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4755121975156005,
+      "learning_rate": 1.903217896152017e-05,
+      "loss": 1.2644,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5140427650905246,
+      "learning_rate": 1.9030992449680684e-05,
+      "loss": 1.314,
+      "step": 1958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.054371232434427,
+      "learning_rate": 1.9029805248008935e-05,
+      "loss": 1.243,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3795732965686236,
+      "learning_rate": 1.9028617356595608e-05,
+      "loss": 1.2021,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.091679983699959,
+      "learning_rate": 1.902742877553144e-05,
+      "loss": 1.2505,
+      "step": 1961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2388293394343237,
+      "learning_rate": 1.902623950490722e-05,
+      "loss": 1.2058,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2695971171690132,
+      "learning_rate": 1.9025049544813793e-05,
+      "loss": 1.1917,
+      "step": 1963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7272302304995075,
+      "learning_rate": 1.902385889534205e-05,
+      "loss": 1.2251,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1357047605345127,
+      "learning_rate": 1.9022667556582944e-05,
+      "loss": 1.2815,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.026085256998653,
+      "learning_rate": 1.902147552862747e-05,
+      "loss": 1.218,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7552055351656497,
+      "learning_rate": 1.9020282811566687e-05,
+      "loss": 1.1973,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.501878886677903,
+      "learning_rate": 1.9019089405491693e-05,
+      "loss": 1.2539,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.791601208711811,
+      "learning_rate": 1.901789531049365e-05,
+      "loss": 1.2383,
+      "step": 1969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9449642565875847,
+      "learning_rate": 1.9016700526663772e-05,
+      "loss": 1.2561,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.409471798186117,
+      "learning_rate": 1.901550505409332e-05,
+      "loss": 1.2461,
+      "step": 1971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.316370827650348,
+      "learning_rate": 1.9014308892873612e-05,
+      "loss": 1.1938,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.064459740842506,
+      "learning_rate": 1.9013112043096012e-05,
+      "loss": 1.2427,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.846363863796114,
+      "learning_rate": 1.9011914504851943e-05,
+      "loss": 1.2036,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.161231922723163,
+      "learning_rate": 1.9010716278232882e-05,
+      "loss": 1.239,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.034371414026534,
+      "learning_rate": 1.9009517363330356e-05,
+      "loss": 1.1699,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4876176702469723,
+      "learning_rate": 1.9008317760235938e-05,
+      "loss": 1.1245,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6136801053618273,
+      "learning_rate": 1.9007117469041265e-05,
+      "loss": 1.1733,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.079586774057286,
+      "learning_rate": 1.900591648983802e-05,
+      "loss": 1.1509,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.4916285556858435,
+      "learning_rate": 1.9004714822717938e-05,
+      "loss": 1.2061,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8802210522102922,
+      "learning_rate": 1.900351246777281e-05,
+      "loss": 1.2998,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.075204411460181,
+      "learning_rate": 1.900230942509448e-05,
+      "loss": 1.1738,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.639289918390531,
+      "learning_rate": 1.900110569477484e-05,
+      "loss": 1.2549,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6256629583234052,
+      "learning_rate": 1.8999901276905835e-05,
+      "loss": 1.2219,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.2320014874553635,
+      "learning_rate": 1.8998696171579467e-05,
+      "loss": 1.2725,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.414281793175941,
+      "learning_rate": 1.899749037888779e-05,
+      "loss": 1.2212,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2359229053136414,
+      "learning_rate": 1.89962838989229e-05,
+      "loss": 1.2544,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.000427507883837,
+      "learning_rate": 1.8995076731776966e-05,
+      "loss": 1.3318,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6679076568250575,
+      "learning_rate": 1.8993868877542187e-05,
+      "loss": 1.2676,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.162904869048458,
+      "learning_rate": 1.899266033631083e-05,
+      "loss": 1.1489,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.267308064913518,
+      "learning_rate": 1.899145110817521e-05,
+      "loss": 1.2136,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.922961316604385,
+      "learning_rate": 1.899024119322769e-05,
+      "loss": 1.207,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5047152191646322,
+      "learning_rate": 1.8989030591560695e-05,
+      "loss": 1.2329,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9923493038521953,
+      "learning_rate": 1.8987819303266693e-05,
+      "loss": 1.2061,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.893788854233747,
+      "learning_rate": 1.898660732843821e-05,
+      "loss": 1.1565,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.808312620399921,
+      "learning_rate": 1.8985394667167818e-05,
+      "loss": 1.2852,
+      "step": 1996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.609681681509499,
+      "learning_rate": 1.898418131954815e-05,
+      "loss": 1.2168,
+      "step": 1997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.171400086607448,
+      "learning_rate": 1.898296728567189e-05,
+      "loss": 1.2495,
+      "step": 1998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.36430885618102,
+      "learning_rate": 1.8981752565631767e-05,
+      "loss": 1.2046,
+      "step": 1999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.053311293447201,
+      "learning_rate": 1.898053715952057e-05,
+      "loss": 1.2207,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.639085183279195,
+      "learning_rate": 1.897932106743114e-05,
+      "loss": 1.2932,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.148316667751832,
+      "learning_rate": 1.8978104289456364e-05,
+      "loss": 1.1936,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.13996024533807,
+      "learning_rate": 1.8976886825689184e-05,
+      "loss": 1.27,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8779757142957387,
+      "learning_rate": 1.89756686762226e-05,
+      "loss": 1.2222,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.029599577840123,
+      "learning_rate": 1.8974449841149663e-05,
+      "loss": 1.2896,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.359562331620903,
+      "learning_rate": 1.897323032056347e-05,
+      "loss": 1.2581,
+      "step": 2006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2798025875483203,
+      "learning_rate": 1.8972010114557172e-05,
+      "loss": 1.3157,
+      "step": 2007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6690561106738038,
+      "learning_rate": 1.8970789223223978e-05,
+      "loss": 1.2852,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.960387618660177,
+      "learning_rate": 1.896956764665714e-05,
+      "loss": 1.2505,
+      "step": 2009
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7576966880634943,
+      "learning_rate": 1.896834538494998e-05,
+      "loss": 1.3025,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1269356161517297,
+      "learning_rate": 1.8967122438195846e-05,
+      "loss": 1.2502,
+      "step": 2011
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.110971346087393,
+      "learning_rate": 1.8965898806488162e-05,
+      "loss": 1.2153,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5132112164634193,
+      "learning_rate": 1.8964674489920394e-05,
+      "loss": 1.269,
+      "step": 2013
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.969946376451613,
+      "learning_rate": 1.896344948858606e-05,
+      "loss": 1.2297,
+      "step": 2014
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2632370578667698,
+      "learning_rate": 1.8962223802578733e-05,
+      "loss": 1.1851,
+      "step": 2015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.08333423330125,
+      "learning_rate": 1.8960997431992033e-05,
+      "loss": 1.1826,
+      "step": 2016
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.498348221439925,
+      "learning_rate": 1.895977037691964e-05,
+      "loss": 1.2764,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.441082181404877,
+      "learning_rate": 1.895854263745528e-05,
+      "loss": 1.2034,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.800451919713885,
+      "learning_rate": 1.895731421369274e-05,
+      "loss": 1.144,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4546841934044643,
+      "learning_rate": 1.8956085105725845e-05,
+      "loss": 1.208,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1905947872549554,
+      "learning_rate": 1.8954855313648488e-05,
+      "loss": 1.2253,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.952863097561249,
+      "learning_rate": 1.89536248375546e-05,
+      "loss": 1.1667,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4284628915822224,
+      "learning_rate": 1.8952393677538174e-05,
+      "loss": 1.2351,
+      "step": 2023
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.095118280725797,
+      "learning_rate": 1.8951161833693255e-05,
+      "loss": 1.2183,
+      "step": 2024
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0274064105224796,
+      "learning_rate": 1.894992930611393e-05,
+      "loss": 1.2048,
+      "step": 2025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.926767142298273,
+      "learning_rate": 1.8948696094894354e-05,
+      "loss": 1.238,
+      "step": 2026
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.99074026624205,
+      "learning_rate": 1.894746220012872e-05,
+      "loss": 1.2466,
+      "step": 2027
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.581209818759619,
+      "learning_rate": 1.894622762191128e-05,
+      "loss": 1.2888,
+      "step": 2028
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3985816781109475,
+      "learning_rate": 1.8944992360336337e-05,
+      "loss": 1.2117,
+      "step": 2029
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.184181508632514,
+      "learning_rate": 1.8943756415498252e-05,
+      "loss": 1.2473,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4290135064584257,
+      "learning_rate": 1.8942519787491422e-05,
+      "loss": 1.2671,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1033271184794176,
+      "learning_rate": 1.8941282476410315e-05,
+      "loss": 1.2126,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3234773044607993,
+      "learning_rate": 1.8940044482349443e-05,
+      "loss": 1.2485,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0038088249770243,
+      "learning_rate": 1.8938805805403368e-05,
+      "loss": 1.1729,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.135708479936234,
+      "learning_rate": 1.8937566445666707e-05,
+      "loss": 1.1729,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0930685068800896,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 1.2192,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5891040204851277,
+      "learning_rate": 1.8935085678200345e-05,
+      "loss": 1.2529,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.277992582715025,
+      "learning_rate": 1.8933844270660144e-05,
+      "loss": 1.2034,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9678526267155942,
+      "learning_rate": 1.893260218070834e-05,
+      "loss": 1.2319,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0770429107648654,
+      "learning_rate": 1.8931359408439817e-05,
+      "loss": 1.2163,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.196331467071842,
+      "learning_rate": 1.8930115953949496e-05,
+      "loss": 1.1631,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2420352721958063,
+      "learning_rate": 1.8928871817332365e-05,
+      "loss": 1.2183,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9530638656583075,
+      "learning_rate": 1.8927626998683452e-05,
+      "loss": 1.1541,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.62743098526671,
+      "learning_rate": 1.892638149809785e-05,
+      "loss": 1.2131,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5096718402664893,
+      "learning_rate": 1.892513531567069e-05,
+      "loss": 1.2107,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9597983186617056,
+      "learning_rate": 1.892388845149716e-05,
+      "loss": 1.272,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7562815370223275,
+      "learning_rate": 1.892264090567251e-05,
+      "loss": 1.1626,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2899749608219557,
+      "learning_rate": 1.8921392678292027e-05,
+      "loss": 1.2219,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3510039829506995,
+      "learning_rate": 1.892014376945106e-05,
+      "loss": 1.2903,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.149821192426779,
+      "learning_rate": 1.8918894179245006e-05,
+      "loss": 1.2329,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.254432897205669,
+      "learning_rate": 1.8917643907769317e-05,
+      "loss": 1.2317,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9733746651469053,
+      "learning_rate": 1.891639295511949e-05,
+      "loss": 1.1904,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.322033422278612,
+      "learning_rate": 1.8915141321391083e-05,
+      "loss": 1.1636,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.278838899860085,
+      "learning_rate": 1.8913889006679697e-05,
+      "loss": 1.1738,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.199707690366517,
+      "learning_rate": 1.8912636011080994e-05,
+      "loss": 1.2798,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.37170566129146,
+      "learning_rate": 1.8911382334690687e-05,
+      "loss": 1.2817,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.841828786109459,
+      "learning_rate": 1.8910127977604534e-05,
+      "loss": 1.2476,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.03078346158601,
+      "learning_rate": 1.890887293991835e-05,
+      "loss": 1.2485,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.920134673210716,
+      "learning_rate": 1.8907617221728003e-05,
+      "loss": 1.2478,
+      "step": 2059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.782642606693253,
+      "learning_rate": 1.890636082312941e-05,
+      "loss": 1.2144,
+      "step": 2060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.96562981566826,
+      "learning_rate": 1.8905103744218534e-05,
+      "loss": 1.1931,
+      "step": 2061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2950826738790897,
+      "learning_rate": 1.8903845985091406e-05,
+      "loss": 1.2065,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.362254595681159,
+      "learning_rate": 1.8902587545844096e-05,
+      "loss": 1.2559,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.436923152981799,
+      "learning_rate": 1.8901328426572733e-05,
+      "loss": 1.2964,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.056673668874732,
+      "learning_rate": 1.890006862737349e-05,
+      "loss": 1.21,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2955570528047855,
+      "learning_rate": 1.8898808148342603e-05,
+      "loss": 1.2441,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3865673189863954,
+      "learning_rate": 1.8897546989576347e-05,
+      "loss": 1.2319,
+      "step": 2067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3380428360976575,
+      "learning_rate": 1.889628515117106e-05,
+      "loss": 1.2808,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1286416590033785,
+      "learning_rate": 1.8895022633223127e-05,
+      "loss": 1.2407,
+      "step": 2069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.951971696136723,
+      "learning_rate": 1.889375943582898e-05,
+      "loss": 1.1575,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0508189005183053,
+      "learning_rate": 1.889249555908512e-05,
+      "loss": 1.2832,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2695524510968923,
+      "learning_rate": 1.8891231003088073e-05,
+      "loss": 1.2468,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.996598417558353,
+      "learning_rate": 1.8889965767934444e-05,
+      "loss": 1.2068,
+      "step": 2073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3482242570285816,
+      "learning_rate": 1.8888699853720877e-05,
+      "loss": 1.2742,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6949511714282335,
+      "learning_rate": 1.888743326054406e-05,
+      "loss": 1.2268,
+      "step": 2075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8972441814308474,
+      "learning_rate": 1.8886165988500754e-05,
+      "loss": 1.0815,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8619264635357444,
+      "learning_rate": 1.888489803768775e-05,
+      "loss": 1.2361,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.887966838097008,
+      "learning_rate": 1.8883629408201904e-05,
+      "loss": 1.2302,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.540633686637702,
+      "learning_rate": 1.888236010014012e-05,
+      "loss": 1.2527,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.682272703643187,
+      "learning_rate": 1.8881090113599353e-05,
+      "loss": 1.2349,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0570637489620776,
+      "learning_rate": 1.8879819448676612e-05,
+      "loss": 1.208,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2284779692125234,
+      "learning_rate": 1.8878548105468958e-05,
+      "loss": 1.3027,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.999174335351595,
+      "learning_rate": 1.88772760840735e-05,
+      "loss": 1.1045,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5498201259712836,
+      "learning_rate": 1.8876003384587404e-05,
+      "loss": 1.2185,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.62570642784746,
+      "learning_rate": 1.8874730007107883e-05,
+      "loss": 1.2632,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.38673786055641,
+      "learning_rate": 1.8873455951732205e-05,
+      "loss": 1.2805,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2717445203203073,
+      "learning_rate": 1.8872181218557685e-05,
+      "loss": 1.218,
+      "step": 2087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.49443220282121,
+      "learning_rate": 1.88709058076817e-05,
+      "loss": 1.2175,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2529101269539566,
+      "learning_rate": 1.8869629719201668e-05,
+      "loss": 1.175,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1608897869378914,
+      "learning_rate": 1.8868352953215065e-05,
+      "loss": 1.2156,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0960591084469624,
+      "learning_rate": 1.8867075509819413e-05,
+      "loss": 1.2312,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1002999022101663,
+      "learning_rate": 1.8865797389112296e-05,
+      "loss": 1.2332,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.94174417815892,
+      "learning_rate": 1.886451859119134e-05,
+      "loss": 1.168,
+      "step": 2093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1922938182611413,
+      "learning_rate": 1.886323911615422e-05,
+      "loss": 1.2271,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.231931672352866,
+      "learning_rate": 1.886195896409868e-05,
+      "loss": 1.2549,
+      "step": 2095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.22652174719551,
+      "learning_rate": 1.8860678135122494e-05,
+      "loss": 1.2329,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.664073307488085,
+      "learning_rate": 1.8859396629323505e-05,
+      "loss": 1.2715,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.820219356297866,
+      "learning_rate": 1.88581144467996e-05,
+      "loss": 1.269,
+      "step": 2098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.401629126394126,
+      "learning_rate": 1.8856831587648715e-05,
+      "loss": 1.3232,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.971577037244324,
+      "learning_rate": 1.8855548051968843e-05,
+      "loss": 1.1448,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2510614429896734,
+      "learning_rate": 1.8854263839858028e-05,
+      "loss": 1.2734,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.836572376855537,
+      "learning_rate": 1.8852978951414363e-05,
+      "loss": 1.1953,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.705696986774715,
+      "learning_rate": 1.8851693386735994e-05,
+      "loss": 1.2234,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.297709367073946,
+      "learning_rate": 1.8850407145921118e-05,
+      "loss": 1.2271,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.005209498395589,
+      "learning_rate": 1.8849120229067992e-05,
+      "loss": 1.1709,
+      "step": 2105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7148105664230235,
+      "learning_rate": 1.8847832636274908e-05,
+      "loss": 1.3145,
+      "step": 2106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4567673121063986,
+      "learning_rate": 1.8846544367640218e-05,
+      "loss": 1.1846,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.513573064814225,
+      "learning_rate": 1.8845255423262334e-05,
+      "loss": 1.1802,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.388876053958744,
+      "learning_rate": 1.884396580323971e-05,
+      "loss": 1.2593,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3308841496151635,
+      "learning_rate": 1.884267550767085e-05,
+      "loss": 1.22,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.144096647166621,
+      "learning_rate": 1.8841384536654314e-05,
+      "loss": 1.1538,
+      "step": 2111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1183902702183257,
+      "learning_rate": 1.8840092890288715e-05,
+      "loss": 1.2388,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3590936626823282,
+      "learning_rate": 1.8838800568672716e-05,
+      "loss": 1.3284,
+      "step": 2113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7843001368642133,
+      "learning_rate": 1.8837507571905028e-05,
+      "loss": 1.2554,
+      "step": 2114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3654506811440816,
+      "learning_rate": 1.883621390008442e-05,
+      "loss": 1.241,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.98490630387092,
+      "learning_rate": 1.8834919553309704e-05,
+      "loss": 1.1948,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.210312316934537,
+      "learning_rate": 1.8833624531679753e-05,
+      "loss": 1.197,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.599462515214929,
+      "learning_rate": 1.8832328835293483e-05,
+      "loss": 1.1802,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1610099345381304,
+      "learning_rate": 1.8831032464249872e-05,
+      "loss": 1.2871,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4856867785337373,
+      "learning_rate": 1.8829735418647938e-05,
+      "loss": 1.2803,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8159779221162577,
+      "learning_rate": 1.882843769858676e-05,
+      "loss": 1.2429,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8546395867353445,
+      "learning_rate": 1.882713930416546e-05,
+      "loss": 1.1514,
+      "step": 2122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4253849355025254,
+      "learning_rate": 1.8825840235483218e-05,
+      "loss": 1.2688,
+      "step": 2123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.618180666304984,
+      "learning_rate": 1.8824540492639267e-05,
+      "loss": 1.1687,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4484351037871197,
+      "learning_rate": 1.882324007573288e-05,
+      "loss": 1.2981,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4322965553595712,
+      "learning_rate": 1.8821938984863393e-05,
+      "loss": 1.24,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2387723620784956,
+      "learning_rate": 1.8820637220130194e-05,
+      "loss": 1.1926,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2680253157090675,
+      "learning_rate": 1.881933478163271e-05,
+      "loss": 1.2415,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.512019877258164,
+      "learning_rate": 1.8818031669470437e-05,
+      "loss": 1.3076,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3405479298090373,
+      "learning_rate": 1.8816727883742904e-05,
+      "loss": 1.2903,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.263187814644437,
+      "learning_rate": 1.8815423424549707e-05,
+      "loss": 1.2466,
+      "step": 2131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7913651510371587,
+      "learning_rate": 1.8814118291990488e-05,
+      "loss": 1.2373,
+      "step": 2132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8586803303143906,
+      "learning_rate": 1.8812812486164935e-05,
+      "loss": 1.1755,
+      "step": 2133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8533081836520187,
+      "learning_rate": 1.881150600717279e-05,
+      "loss": 1.199,
+      "step": 2134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8296442363175003,
+      "learning_rate": 1.8810198855113854e-05,
+      "loss": 1.2725,
+      "step": 2135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0880936266954353,
+      "learning_rate": 1.8808891030087973e-05,
+      "loss": 1.1736,
+      "step": 2136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.724025109388185,
+      "learning_rate": 1.8807582532195046e-05,
+      "loss": 1.2852,
+      "step": 2137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.087969792415505,
+      "learning_rate": 1.8806273361535017e-05,
+      "loss": 1.2117,
+      "step": 2138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4020524746107443,
+      "learning_rate": 1.8804963518207895e-05,
+      "loss": 1.2324,
+      "step": 2139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.657179151400887,
+      "learning_rate": 1.8803653002313724e-05,
+      "loss": 1.2354,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6091557817799176,
+      "learning_rate": 1.8802341813952614e-05,
+      "loss": 1.2131,
+      "step": 2141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.387681790861238,
+      "learning_rate": 1.8801029953224717e-05,
+      "loss": 1.2417,
+      "step": 2142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.752009601108392,
+      "learning_rate": 1.8799717420230242e-05,
+      "loss": 1.2622,
+      "step": 2143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3106955757897163,
+      "learning_rate": 1.8798404215069445e-05,
+      "loss": 1.2236,
+      "step": 2144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.056932386209127,
+      "learning_rate": 1.879709033784263e-05,
+      "loss": 1.2422,
+      "step": 2145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.90634705589837,
+      "learning_rate": 1.8795775788650174e-05,
+      "loss": 1.2319,
+      "step": 2146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.131485145021431,
+      "learning_rate": 1.879446056759247e-05,
+      "loss": 1.2534,
+      "step": 2147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0845074130067625,
+      "learning_rate": 1.8793144674769988e-05,
+      "loss": 1.2222,
+      "step": 2148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.226139367371076,
+      "learning_rate": 1.8791828110283244e-05,
+      "loss": 1.2268,
+      "step": 2149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5017510668049017,
+      "learning_rate": 1.8790510874232805e-05,
+      "loss": 1.2146,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.338987725336701,
+      "learning_rate": 1.8789192966719287e-05,
+      "loss": 1.1279,
+      "step": 2151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.357963297482927,
+      "learning_rate": 1.8787874387843358e-05,
+      "loss": 1.3008,
+      "step": 2152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.300758451532598,
+      "learning_rate": 1.8786555137705732e-05,
+      "loss": 1.291,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9865755824732743,
+      "learning_rate": 1.878523521640719e-05,
+      "loss": 1.1946,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.344393401573172,
+      "learning_rate": 1.8783914624048546e-05,
+      "loss": 1.1809,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2172533379048223,
+      "learning_rate": 1.8782593360730678e-05,
+      "loss": 1.1672,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.158967373669426,
+      "learning_rate": 1.878127142655451e-05,
+      "loss": 1.2246,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.289454674190958,
+      "learning_rate": 1.8779948821621018e-05,
+      "loss": 1.2261,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4751210158992603,
+      "learning_rate": 1.8778625546031226e-05,
+      "loss": 1.2168,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6369401370638443,
+      "learning_rate": 1.8777301599886216e-05,
+      "loss": 1.2241,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1607254502091906,
+      "learning_rate": 1.8775976983287117e-05,
+      "loss": 1.2627,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8959722564208032,
+      "learning_rate": 1.8774651696335108e-05,
+      "loss": 1.2334,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.919528223696603,
+      "learning_rate": 1.8773325739131423e-05,
+      "loss": 1.1997,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2059179537441094,
+      "learning_rate": 1.8771999111777344e-05,
+      "loss": 1.252,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5719795079880865,
+      "learning_rate": 1.8770671814374207e-05,
+      "loss": 1.3291,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1513910258691564,
+      "learning_rate": 1.8769343847023397e-05,
+      "loss": 1.2161,
+      "step": 2166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1200576218365192,
+      "learning_rate": 1.8768015209826346e-05,
+      "loss": 1.1953,
+      "step": 2167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.856120132130044,
+      "learning_rate": 1.876668590288455e-05,
+      "loss": 1.1936,
+      "step": 2168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.373088109703693,
+      "learning_rate": 1.8765355926299546e-05,
+      "loss": 1.229,
+      "step": 2169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3843805035681105,
+      "learning_rate": 1.876402528017292e-05,
+      "loss": 1.2402,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8959188465472865,
+      "learning_rate": 1.8762693964606314e-05,
+      "loss": 1.2495,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2320237340227695,
+      "learning_rate": 1.8761361979701424e-05,
+      "loss": 1.2458,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3222455518049965,
+      "learning_rate": 1.8760029325559992e-05,
+      "loss": 1.2852,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7294704343476517,
+      "learning_rate": 1.8758696002283817e-05,
+      "loss": 1.271,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1443577754108687,
+      "learning_rate": 1.875736200997474e-05,
+      "loss": 1.2737,
+      "step": 2175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6612100661728495,
+      "learning_rate": 1.8756027348734654e-05,
+      "loss": 1.26,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9094775433330127,
+      "learning_rate": 1.8754692018665516e-05,
+      "loss": 1.2147,
+      "step": 2177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4598274660785684,
+      "learning_rate": 1.8753356019869318e-05,
+      "loss": 1.2998,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7147275072155925,
+      "learning_rate": 1.8752019352448114e-05,
+      "loss": 1.3069,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3781755069277195,
+      "learning_rate": 1.8750682016504004e-05,
+      "loss": 1.2827,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8597636646749405,
+      "learning_rate": 1.8749344012139145e-05,
+      "loss": 1.2651,
+      "step": 2181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.391684164092527,
+      "learning_rate": 1.8748005339455732e-05,
+      "loss": 1.2542,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9768048624086414,
+      "learning_rate": 1.8746665998556027e-05,
+      "loss": 1.165,
+      "step": 2183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.645762088084242,
+      "learning_rate": 1.8745325989542327e-05,
+      "loss": 1.2053,
+      "step": 2184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.101277704489686,
+      "learning_rate": 1.8743985312517e-05,
+      "loss": 1.2183,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2825072840095366,
+      "learning_rate": 1.874264396758245e-05,
+      "loss": 1.2043,
+      "step": 2186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3012514651305924,
+      "learning_rate": 1.8741301954841128e-05,
+      "loss": 1.2334,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.994206848216242,
+      "learning_rate": 1.873995927439555e-05,
+      "loss": 1.2461,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9309213103382996,
+      "learning_rate": 1.8738615926348274e-05,
+      "loss": 1.2441,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1625134605730962,
+      "learning_rate": 1.8737271910801916e-05,
+      "loss": 1.2627,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0100486297103397,
+      "learning_rate": 1.8735927227859136e-05,
+      "loss": 1.179,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1467913721160974,
+      "learning_rate": 1.8734581877622644e-05,
+      "loss": 1.2485,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.432671530023458,
+      "learning_rate": 1.873323586019521e-05,
+      "loss": 1.2703,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4251292728616636,
+      "learning_rate": 1.8731889175679647e-05,
+      "loss": 1.2527,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.548917615866466,
+      "learning_rate": 1.8730541824178824e-05,
+      "loss": 1.2935,
+      "step": 2195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9790810611139222,
+      "learning_rate": 1.8729193805795654e-05,
+      "loss": 1.2427,
+      "step": 2196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2392949424768522,
+      "learning_rate": 1.8727845120633108e-05,
+      "loss": 1.2344,
+      "step": 2197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0190432521684825,
+      "learning_rate": 1.8726495768794205e-05,
+      "loss": 1.156,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.508683858542327,
+      "learning_rate": 1.872514575038202e-05,
+      "loss": 1.1973,
+      "step": 2199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2268759066390476,
+      "learning_rate": 1.8723795065499663e-05,
+      "loss": 1.2244,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1173981516811837,
+      "learning_rate": 1.8722443714250317e-05,
+      "loss": 1.1846,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5004544879720445,
+      "learning_rate": 1.8721091696737198e-05,
+      "loss": 1.2285,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.121031763415124,
+      "learning_rate": 1.871973901306358e-05,
+      "loss": 1.2324,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.294111524861851,
+      "learning_rate": 1.8718385663332795e-05,
+      "loss": 1.2363,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8292964405591516,
+      "learning_rate": 1.871703164764821e-05,
+      "loss": 1.2754,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.15329881473513,
+      "learning_rate": 1.8715676966113256e-05,
+      "loss": 1.1785,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.518364329962991,
+      "learning_rate": 1.871432161883141e-05,
+      "loss": 1.2393,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4810790794236914,
+      "learning_rate": 1.87129656059062e-05,
+      "loss": 1.2393,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3024097614575365,
+      "learning_rate": 1.8711608927441203e-05,
+      "loss": 1.2378,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.853613133432272,
+      "learning_rate": 1.8710251583540054e-05,
+      "loss": 1.2122,
+      "step": 2210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.119402233207247,
+      "learning_rate": 1.8708893574306425e-05,
+      "loss": 1.27,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.093043642854872,
+      "learning_rate": 1.870753489984406e-05,
+      "loss": 1.2607,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.188116425684695,
+      "learning_rate": 1.8706175560256728e-05,
+      "loss": 1.2495,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9895593613532694,
+      "learning_rate": 1.870481555564827e-05,
+      "loss": 1.2114,
+      "step": 2214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8311199248647556,
+      "learning_rate": 1.8703454886122568e-05,
+      "loss": 1.2334,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9188183517258737,
+      "learning_rate": 1.8702093551783556e-05,
+      "loss": 1.2007,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.290480058980993,
+      "learning_rate": 1.870073155273522e-05,
+      "loss": 1.175,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.493110268135696,
+      "learning_rate": 1.86993688890816e-05,
+      "loss": 1.2332,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.118133977484434,
+      "learning_rate": 1.8698005560926776e-05,
+      "loss": 1.27,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.220186324712695,
+      "learning_rate": 1.869664156837489e-05,
+      "loss": 1.2493,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9923971303627535,
+      "learning_rate": 1.8695276911530134e-05,
+      "loss": 1.252,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.193325481468241,
+      "learning_rate": 1.8693911590496738e-05,
+      "loss": 1.2273,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9473651208919533,
+      "learning_rate": 1.8692545605379e-05,
+      "loss": 1.157,
+      "step": 2223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6233657758258917,
+      "learning_rate": 1.869117895628126e-05,
+      "loss": 1.2603,
+      "step": 2224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2966110354188167,
+      "learning_rate": 1.868981164330791e-05,
+      "loss": 1.2705,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2866121342148236,
+      "learning_rate": 1.8688443666563382e-05,
+      "loss": 1.2998,
+      "step": 2226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.103872739880784,
+      "learning_rate": 1.868707502615218e-05,
+      "loss": 1.2383,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1002677833847976,
+      "learning_rate": 1.8685705722178844e-05,
+      "loss": 1.1438,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2239881338168357,
+      "learning_rate": 1.8684335754747968e-05,
+      "loss": 1.2551,
+      "step": 2229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.570638495460667,
+      "learning_rate": 1.86829651239642e-05,
+      "loss": 1.2627,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2270666815547417,
+      "learning_rate": 1.868159382993223e-05,
+      "loss": 1.2563,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.518699934685398,
+      "learning_rate": 1.868022187275681e-05,
+      "loss": 1.2646,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.156075563367055,
+      "learning_rate": 1.8678849252542733e-05,
+      "loss": 1.2078,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5154519310048347,
+      "learning_rate": 1.8677475969394853e-05,
+      "loss": 1.1572,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0386477612755893,
+      "learning_rate": 1.8676102023418057e-05,
+      "loss": 1.2732,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.80808259076934,
+      "learning_rate": 1.8674727414717302e-05,
+      "loss": 1.2522,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9189637028180875,
+      "learning_rate": 1.8673352143397584e-05,
+      "loss": 1.1843,
+      "step": 2237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1397776930085253,
+      "learning_rate": 1.867197620956396e-05,
+      "loss": 1.2583,
+      "step": 2238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4463441070934246,
+      "learning_rate": 1.8670599613321518e-05,
+      "loss": 1.2012,
+      "step": 2239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9457022400530812,
+      "learning_rate": 1.8669222354775417e-05,
+      "loss": 1.1584,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.456545173085503,
+      "learning_rate": 1.8667844434030864e-05,
+      "loss": 1.229,
+      "step": 2241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1604774471597437,
+      "learning_rate": 1.86664658511931e-05,
+      "loss": 1.2715,
+      "step": 2242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.921828458282739,
+      "learning_rate": 1.8665086606367438e-05,
+      "loss": 1.189,
+      "step": 2243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.411764932362569,
+      "learning_rate": 1.866370669965923e-05,
+      "loss": 1.2407,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4772747383102383,
+      "learning_rate": 1.866232613117387e-05,
+      "loss": 1.167,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5317059063289893,
+      "learning_rate": 1.8660944901016827e-05,
+      "loss": 1.2429,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.782760797487225,
+      "learning_rate": 1.8659563009293597e-05,
+      "loss": 1.2053,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.548986789083154,
+      "learning_rate": 1.8658180456109738e-05,
+      "loss": 1.3086,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3513962618575253,
+      "learning_rate": 1.8656797241570858e-05,
+      "loss": 1.1926,
+      "step": 2249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.503850678897599,
+      "learning_rate": 1.865541336578261e-05,
+      "loss": 1.2495,
+      "step": 2250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.056943725755842,
+      "learning_rate": 1.8654028828850705e-05,
+      "loss": 1.1841,
+      "step": 2251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.040877738580352,
+      "learning_rate": 1.8652643630880902e-05,
+      "loss": 1.2224,
+      "step": 2252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.896004622665576,
+      "learning_rate": 1.8651257771979006e-05,
+      "loss": 1.2769,
+      "step": 2253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0924694737302083,
+      "learning_rate": 1.8649871252250874e-05,
+      "loss": 1.1465,
+      "step": 2254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9191532733359153,
+      "learning_rate": 1.8648484071802423e-05,
+      "loss": 1.1782,
+      "step": 2255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.891303104527232,
+      "learning_rate": 1.8647096230739607e-05,
+      "loss": 1.1812,
+      "step": 2256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.12918481006489,
+      "learning_rate": 1.864570772916843e-05,
+      "loss": 1.2134,
+      "step": 2257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4767490452775403,
+      "learning_rate": 1.864431856719497e-05,
+      "loss": 1.2263,
+      "step": 2258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5321852648912415,
+      "learning_rate": 1.864292874492532e-05,
+      "loss": 1.2268,
+      "step": 2259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9792960922511553,
+      "learning_rate": 1.8641538262465656e-05,
+      "loss": 1.2395,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4004242864992382,
+      "learning_rate": 1.864014711992218e-05,
+      "loss": 1.127,
+      "step": 2261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2605202550927945,
+      "learning_rate": 1.8638755317401157e-05,
+      "loss": 1.2537,
+      "step": 2262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1375566857669948,
+      "learning_rate": 1.8637362855008902e-05,
+      "loss": 1.2244,
+      "step": 2263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.439789681510037,
+      "learning_rate": 1.863596973285178e-05,
+      "loss": 1.2417,
+      "step": 2264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0872501447840324,
+      "learning_rate": 1.8634575951036197e-05,
+      "loss": 1.2229,
+      "step": 2265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.167041582621611,
+      "learning_rate": 1.8633181509668623e-05,
+      "loss": 1.2354,
+      "step": 2266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.06586724588836,
+      "learning_rate": 1.8631786408855572e-05,
+      "loss": 1.1807,
+      "step": 2267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1012262341788435,
+      "learning_rate": 1.8630390648703604e-05,
+      "loss": 1.2139,
+      "step": 2268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4905237144387526,
+      "learning_rate": 1.862899422931934e-05,
+      "loss": 1.187,
+      "step": 2269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9818328080198677,
+      "learning_rate": 1.8627597150809445e-05,
+      "loss": 1.271,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.310281456932891,
+      "learning_rate": 1.8626199413280635e-05,
+      "loss": 1.2771,
+      "step": 2271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.685157917639287,
+      "learning_rate": 1.862480101683967e-05,
+      "loss": 1.1598,
+      "step": 2272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.452396363850125,
+      "learning_rate": 1.8623401961593375e-05,
+      "loss": 1.2988,
+      "step": 2273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6470512908229487,
+      "learning_rate": 1.862200224764861e-05,
+      "loss": 1.196,
+      "step": 2274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.131105243169297,
+      "learning_rate": 1.8620601875112295e-05,
+      "loss": 1.2407,
+      "step": 2275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.745310654810979,
+      "learning_rate": 1.8619200844091398e-05,
+      "loss": 1.1709,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4523236506779758,
+      "learning_rate": 1.8617799154692933e-05,
+      "loss": 1.2004,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9236274228098633,
+      "learning_rate": 1.8616396807023975e-05,
+      "loss": 1.2578,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1702518445640577,
+      "learning_rate": 1.8614993801191637e-05,
+      "loss": 1.2517,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.195816613515828,
+      "learning_rate": 1.861359013730309e-05,
+      "loss": 1.2275,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.962507899327652,
+      "learning_rate": 1.8612185815465547e-05,
+      "loss": 1.167,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0207471908387507,
+      "learning_rate": 1.8610780835786288e-05,
+      "loss": 1.1704,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.298384567313304,
+      "learning_rate": 1.8609375198372623e-05,
+      "loss": 1.2637,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.544522088471803,
+      "learning_rate": 1.860796890333192e-05,
+      "loss": 1.2668,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.192280605216877,
+      "learning_rate": 1.8606561950771612e-05,
+      "loss": 1.1782,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.12493255927666,
+      "learning_rate": 1.8605154340799156e-05,
+      "loss": 1.251,
+      "step": 2286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5214440841453887,
+      "learning_rate": 1.860374607352208e-05,
+      "loss": 1.218,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2014474415949055,
+      "learning_rate": 1.860233714904795e-05,
+      "loss": 1.2393,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3885319814196433,
+      "learning_rate": 1.8600927567484387e-05,
+      "loss": 1.2422,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.774950728687984,
+      "learning_rate": 1.859951732893907e-05,
+      "loss": 1.2334,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1285379458826807,
+      "learning_rate": 1.8598106433519706e-05,
+      "loss": 1.1584,
+      "step": 2291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0291144019704657,
+      "learning_rate": 1.8596694881334073e-05,
+      "loss": 1.2388,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.51875277553736,
+      "learning_rate": 1.8595282672489993e-05,
+      "loss": 1.2097,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.596310527716529,
+      "learning_rate": 1.859386980709534e-05,
+      "loss": 1.1602,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4504282098310433,
+      "learning_rate": 1.859245628525803e-05,
+      "loss": 1.2312,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4730709556971417,
+      "learning_rate": 1.859104210708604e-05,
+      "loss": 1.26,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9784138646487284,
+      "learning_rate": 1.8589627272687386e-05,
+      "loss": 1.1648,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.847168675296482,
+      "learning_rate": 1.858821178217015e-05,
+      "loss": 1.1555,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.789439107795046,
+      "learning_rate": 1.8586795635642444e-05,
+      "loss": 1.2646,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0525636562696907,
+      "learning_rate": 1.8585378833212446e-05,
+      "loss": 1.1746,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.726205217630372,
+      "learning_rate": 1.8583961374988377e-05,
+      "loss": 1.2458,
+      "step": 2301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3970531831326025,
+      "learning_rate": 1.858254326107851e-05,
+      "loss": 1.2964,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7481195152838316,
+      "learning_rate": 1.8581124491591166e-05,
+      "loss": 1.1438,
+      "step": 2303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1833612259100574,
+      "learning_rate": 1.8579705066634718e-05,
+      "loss": 1.1855,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1343728503485377,
+      "learning_rate": 1.8578284986317592e-05,
+      "loss": 1.1506,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0335369726229735,
+      "learning_rate": 1.8576864250748254e-05,
+      "loss": 1.1509,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.229440530280783,
+      "learning_rate": 1.8575442860035235e-05,
+      "loss": 1.2546,
+      "step": 2307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4029723740734474,
+      "learning_rate": 1.8574020814287104e-05,
+      "loss": 1.2573,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.889662985750571,
+      "learning_rate": 1.8572598113612482e-05,
+      "loss": 1.1489,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.59511829440651,
+      "learning_rate": 1.8571174758120046e-05,
+      "loss": 1.1675,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7453438340609657,
+      "learning_rate": 1.8569750747918518e-05,
+      "loss": 1.2446,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.038963661633695,
+      "learning_rate": 1.8568326083116667e-05,
+      "loss": 1.2896,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.235867265404068,
+      "learning_rate": 1.856690076382332e-05,
+      "loss": 1.2224,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1835861985579257,
+      "learning_rate": 1.8565474790147352e-05,
+      "loss": 1.2456,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.500036459226255,
+      "learning_rate": 1.856404816219768e-05,
+      "loss": 1.1934,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.535595315728455,
+      "learning_rate": 1.856262088008328e-05,
+      "loss": 1.3486,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0799743488060707,
+      "learning_rate": 1.8561192943913173e-05,
+      "loss": 1.2202,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1203664481602518,
+      "learning_rate": 1.8559764353796436e-05,
+      "loss": 1.23,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.065959520541091,
+      "learning_rate": 1.8558335109842193e-05,
+      "loss": 1.3013,
+      "step": 2319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.053555469532359,
+      "learning_rate": 1.8556905212159607e-05,
+      "loss": 1.1667,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6290087017522326,
+      "learning_rate": 1.8555474660857912e-05,
+      "loss": 1.2383,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9884294848669675,
+      "learning_rate": 1.8554043456046374e-05,
+      "loss": 1.2764,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9458902494962564,
+      "learning_rate": 1.855261159783432e-05,
+      "loss": 1.2061,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.237416625209302,
+      "learning_rate": 1.8551179086331115e-05,
+      "loss": 1.2446,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2215204717002153,
+      "learning_rate": 1.854974592164619e-05,
+      "loss": 1.2407,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.160279954992886,
+      "learning_rate": 1.8548312103889014e-05,
+      "loss": 1.2097,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.341062790100554,
+      "learning_rate": 1.8546877633169106e-05,
+      "loss": 1.2153,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.167791608936117,
+      "learning_rate": 1.854544250959604e-05,
+      "loss": 1.2175,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.952464808942658,
+      "learning_rate": 1.854400673327944e-05,
+      "loss": 1.2441,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8083171644461524,
+      "learning_rate": 1.8542570304328976e-05,
+      "loss": 1.2812,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5363248952919113,
+      "learning_rate": 1.854113322285437e-05,
+      "loss": 1.2334,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6525974042166944,
+      "learning_rate": 1.8539695488965396e-05,
+      "loss": 1.231,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0018230593794635,
+      "learning_rate": 1.853825710277187e-05,
+      "loss": 1.1328,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.219246066205682,
+      "learning_rate": 1.8536818064383667e-05,
+      "loss": 1.1575,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.422048954659991,
+      "learning_rate": 1.8535378373910706e-05,
+      "loss": 1.2173,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.156406550446461,
+      "learning_rate": 1.853393803146296e-05,
+      "loss": 1.2383,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7455143788759804,
+      "learning_rate": 1.8532497037150446e-05,
+      "loss": 1.1672,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2404173075208575,
+      "learning_rate": 1.8531055391083236e-05,
+      "loss": 1.2168,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2155182472418002,
+      "learning_rate": 1.852961309337145e-05,
+      "loss": 1.2153,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.312512744966766,
+      "learning_rate": 1.852817014412526e-05,
+      "loss": 1.2478,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5844000079015084,
+      "learning_rate": 1.8526726543454883e-05,
+      "loss": 1.2205,
+      "step": 2341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0886291107782786,
+      "learning_rate": 1.852528229147059e-05,
+      "loss": 1.1611,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7964081134377277,
+      "learning_rate": 1.8523837388282703e-05,
+      "loss": 1.2334,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.917716810158402,
+      "learning_rate": 1.8522391834001582e-05,
+      "loss": 1.2256,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1678823087831183,
+      "learning_rate": 1.8520945628737653e-05,
+      "loss": 1.2095,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8284967068075453,
+      "learning_rate": 1.8519498772601386e-05,
+      "loss": 1.2329,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3755012577071777,
+      "learning_rate": 1.8518051265703293e-05,
+      "loss": 1.1799,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.115464096431045,
+      "learning_rate": 1.8516603108153944e-05,
+      "loss": 1.2925,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0679694326468967,
+      "learning_rate": 1.8515154300063955e-05,
+      "loss": 1.2529,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3207626778792423,
+      "learning_rate": 1.8513704841543997e-05,
+      "loss": 1.2717,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.963763247588883,
+      "learning_rate": 1.851225473270479e-05,
+      "loss": 1.1538,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.055586095429696,
+      "learning_rate": 1.851080397365709e-05,
+      "loss": 1.2183,
+      "step": 2352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3995947076151682,
+      "learning_rate": 1.850935256451172e-05,
+      "loss": 1.3057,
+      "step": 2353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.49401869169301,
+      "learning_rate": 1.8507900505379544e-05,
+      "loss": 1.2471,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1380886958594414,
+      "learning_rate": 1.8506447796371474e-05,
+      "loss": 1.2446,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5378967591780386,
+      "learning_rate": 1.8504994437598488e-05,
+      "loss": 1.1306,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.416047067481149,
+      "learning_rate": 1.8503540429171587e-05,
+      "loss": 1.2568,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.965844787318307,
+      "learning_rate": 1.850208577120184e-05,
+      "loss": 1.2437,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0914882932452414,
+      "learning_rate": 1.8500630463800364e-05,
+      "loss": 1.2263,
+      "step": 2359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.456367799504969,
+      "learning_rate": 1.849917450707832e-05,
+      "loss": 1.2937,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2923800191515302,
+      "learning_rate": 1.849771790114692e-05,
+      "loss": 1.1028,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3751615529363956,
+      "learning_rate": 1.849626064611743e-05,
+      "loss": 1.2104,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.170527870493354,
+      "learning_rate": 1.8494802742101157e-05,
+      "loss": 1.1963,
+      "step": 2363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9550257134186158,
+      "learning_rate": 1.8493344189209473e-05,
+      "loss": 1.1604,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.752323493891925,
+      "learning_rate": 1.849188498755378e-05,
+      "loss": 1.27,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9800116868487994,
+      "learning_rate": 1.849042513724554e-05,
+      "loss": 1.196,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.504046260373853,
+      "learning_rate": 1.8488964638396268e-05,
+      "loss": 1.2361,
+      "step": 2367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.071052261496611,
+      "learning_rate": 1.8487503491117523e-05,
+      "loss": 1.197,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.081343577575112,
+      "learning_rate": 1.8486041695520912e-05,
+      "loss": 1.2568,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.310042834326037,
+      "learning_rate": 1.84845792517181e-05,
+      "loss": 1.2292,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.991961872350162,
+      "learning_rate": 1.8483116159820785e-05,
+      "loss": 1.2354,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2529231271531605,
+      "learning_rate": 1.848165241994074e-05,
+      "loss": 1.2234,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7995340988587856,
+      "learning_rate": 1.848018803218976e-05,
+      "loss": 1.2246,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.516141098737766,
+      "learning_rate": 1.8478722996679708e-05,
+      "loss": 1.2803,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.063245585394988,
+      "learning_rate": 1.847725731352249e-05,
+      "loss": 1.2295,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9481779917892004,
+      "learning_rate": 1.847579098283006e-05,
+      "loss": 1.2358,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1003000577410074,
+      "learning_rate": 1.847432400471443e-05,
+      "loss": 1.2021,
+      "step": 2377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5750891618409684,
+      "learning_rate": 1.847285637928765e-05,
+      "loss": 1.2905,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2587315868256717,
+      "learning_rate": 1.8471388106661826e-05,
+      "loss": 1.2456,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4495341391950243,
+      "learning_rate": 1.8469919186949105e-05,
+      "loss": 1.2451,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8903425955879802,
+      "learning_rate": 1.8468449620261704e-05,
+      "loss": 1.1685,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.095117978312302,
+      "learning_rate": 1.8466979406711866e-05,
+      "loss": 1.248,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.026105858146917,
+      "learning_rate": 1.8465508546411894e-05,
+      "loss": 1.2495,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7171659222492086,
+      "learning_rate": 1.8464037039474144e-05,
+      "loss": 1.269,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.080466765067844,
+      "learning_rate": 1.8462564886011015e-05,
+      "loss": 1.2483,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.077990954404369,
+      "learning_rate": 1.8461092086134953e-05,
+      "loss": 1.1804,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.057786462824308,
+      "learning_rate": 1.8459618639958465e-05,
+      "loss": 1.2305,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1896574900410317,
+      "learning_rate": 1.8458144547594097e-05,
+      "loss": 1.1809,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7488616787357536,
+      "learning_rate": 1.8456669809154444e-05,
+      "loss": 1.3081,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.178003583480788,
+      "learning_rate": 1.8455194424752162e-05,
+      "loss": 1.2627,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3562952130984423,
+      "learning_rate": 1.8453718394499942e-05,
+      "loss": 1.2764,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4160757930586643,
+      "learning_rate": 1.8452241718510532e-05,
+      "loss": 1.2656,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1310456737953474,
+      "learning_rate": 1.845076439689673e-05,
+      "loss": 1.176,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.390622330450612,
+      "learning_rate": 1.844928642977138e-05,
+      "loss": 1.1438,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.0894948986253565,
+      "learning_rate": 1.8447807817247368e-05,
+      "loss": 1.2021,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5618561051166333,
+      "learning_rate": 1.844632855943765e-05,
+      "loss": 1.27,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.99184232912413,
+      "learning_rate": 1.8444848656455218e-05,
+      "loss": 1.0815,
+      "step": 2397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9637093131173327,
+      "learning_rate": 1.844336810841311e-05,
+      "loss": 1.1997,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0641257911751603,
+      "learning_rate": 1.8441886915424416e-05,
+      "loss": 1.1704,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.861366543658384,
+      "learning_rate": 1.844040507760228e-05,
+      "loss": 1.1299,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.558631778830547,
+      "learning_rate": 1.843892259505989e-05,
+      "loss": 1.2378,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1250099344393654,
+      "learning_rate": 1.843743946791049e-05,
+      "loss": 1.2373,
+      "step": 2402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.39880547860233,
+      "learning_rate": 1.8435955696267366e-05,
+      "loss": 1.2278,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.917551336582935,
+      "learning_rate": 1.8434471280243854e-05,
+      "loss": 1.1997,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6309207968915667,
+      "learning_rate": 1.843298621995334e-05,
+      "loss": 1.2065,
+      "step": 2405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.770815830867967,
+      "learning_rate": 1.8431500515509267e-05,
+      "loss": 1.259,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5769960157218934,
+      "learning_rate": 1.843001416702512e-05,
+      "loss": 1.2332,
+      "step": 2407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6642386443941803,
+      "learning_rate": 1.8428527174614425e-05,
+      "loss": 1.2363,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7215332703377864,
+      "learning_rate": 1.842703953839077e-05,
+      "loss": 1.2314,
+      "step": 2409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.443220449761991,
+      "learning_rate": 1.8425551258467792e-05,
+      "loss": 1.2566,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9993511637000667,
+      "learning_rate": 1.842406233495917e-05,
+      "loss": 1.1816,
+      "step": 2411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.027992126505478,
+      "learning_rate": 1.8422572767978636e-05,
+      "loss": 1.2148,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.814596759180634,
+      "learning_rate": 1.8421082557639973e-05,
+      "loss": 1.1858,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4444216091462,
+      "learning_rate": 1.8419591704057006e-05,
+      "loss": 1.238,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.686973488096324,
+      "learning_rate": 1.8418100207343615e-05,
+      "loss": 1.2175,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0309724782088034,
+      "learning_rate": 1.8416608067613732e-05,
+      "loss": 1.2051,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5939824841547865,
+      "learning_rate": 1.8415115284981325e-05,
+      "loss": 1.2146,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7552202710527607,
+      "learning_rate": 1.8413621859560433e-05,
+      "loss": 1.2344,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.187491352639904,
+      "learning_rate": 1.841212779146512e-05,
+      "loss": 1.2188,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6610703978124053,
+      "learning_rate": 1.8410633080809518e-05,
+      "loss": 1.186,
+      "step": 2420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.105607510936415,
+      "learning_rate": 1.8409137727707795e-05,
+      "loss": 1.209,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3439566721616583,
+      "learning_rate": 1.8407641732274174e-05,
+      "loss": 1.178,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6509537670031174,
+      "learning_rate": 1.8406145094622934e-05,
+      "loss": 1.2808,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.532878007528203,
+      "learning_rate": 1.8404647814868387e-05,
+      "loss": 1.2205,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6569702259142782,
+      "learning_rate": 1.8403149893124907e-05,
+      "loss": 1.1936,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.375827900104694,
+      "learning_rate": 1.8401651329506912e-05,
+      "loss": 1.2949,
+      "step": 2426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.666207690839817,
+      "learning_rate": 1.8400152124128866e-05,
+      "loss": 1.197,
+      "step": 2427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8484121468448826,
+      "learning_rate": 1.8398652277105293e-05,
+      "loss": 1.2495,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4116567195169485,
+      "learning_rate": 1.8397151788550753e-05,
+      "loss": 1.2583,
+      "step": 2429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8817752022460374,
+      "learning_rate": 1.8395650658579864e-05,
+      "loss": 1.219,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3199541953241996,
+      "learning_rate": 1.8394148887307286e-05,
+      "loss": 1.2319,
+      "step": 2431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0033958637662583,
+      "learning_rate": 1.8392646474847734e-05,
+      "loss": 1.1536,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9544965624920483,
+      "learning_rate": 1.8391143421315972e-05,
+      "loss": 1.2104,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5710328399804334,
+      "learning_rate": 1.838963972682681e-05,
+      "loss": 1.1511,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.102684636928821,
+      "learning_rate": 1.8388135391495105e-05,
+      "loss": 1.2058,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.910683450053897,
+      "learning_rate": 1.8386630415435766e-05,
+      "loss": 1.2183,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.195943421662295,
+      "learning_rate": 1.8385124798763755e-05,
+      "loss": 1.3411,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.057279748033002,
+      "learning_rate": 1.8383618541594073e-05,
+      "loss": 1.2957,
+      "step": 2438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.565365384952661,
+      "learning_rate": 1.8382111644041777e-05,
+      "loss": 1.2251,
+      "step": 2439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6297850413586086,
+      "learning_rate": 1.8380604106221973e-05,
+      "loss": 1.1245,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5997469136315408,
+      "learning_rate": 1.8379095928249815e-05,
+      "loss": 1.2363,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1407627465752666,
+      "learning_rate": 1.83775871102405e-05,
+      "loss": 1.1987,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4955796774268695,
+      "learning_rate": 1.8376077652309284e-05,
+      "loss": 1.1333,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.710597770976169,
+      "learning_rate": 1.8374567554571466e-05,
+      "loss": 1.2454,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.439611057097656,
+      "learning_rate": 1.8373056817142393e-05,
+      "loss": 1.2744,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.659544537545047,
+      "learning_rate": 1.8371545440137467e-05,
+      "loss": 1.1484,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9242831301678875,
+      "learning_rate": 1.837003342367213e-05,
+      "loss": 1.1965,
+      "step": 2447
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.058148899528281,
+      "learning_rate": 1.8368520767861876e-05,
+      "loss": 1.1948,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9248946430784675,
+      "learning_rate": 1.8367007472822253e-05,
+      "loss": 1.1948,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2922685875477278,
+      "learning_rate": 1.8365493538668854e-05,
+      "loss": 1.2871,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5341483804128275,
+      "learning_rate": 1.836397896551732e-05,
+      "loss": 1.2563,
+      "step": 2451
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.455425390018489,
+      "learning_rate": 1.836246375348334e-05,
+      "loss": 1.1748,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.579967664475489,
+      "learning_rate": 1.8360947902682657e-05,
+      "loss": 1.3125,
+      "step": 2453
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5892324786572147,
+      "learning_rate": 1.8359431413231058e-05,
+      "loss": 1.2239,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9198739232875153,
+      "learning_rate": 1.8357914285244373e-05,
+      "loss": 1.1736,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9858877460445674,
+      "learning_rate": 1.83563965188385e-05,
+      "loss": 1.2068,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.523967836358194,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 1.2371,
+      "step": 2457
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.149480432801675,
+      "learning_rate": 1.8353359071232954e-05,
+      "loss": 1.2729,
+      "step": 2458
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.602673932529742,
+      "learning_rate": 1.8351839390265304e-05,
+      "loss": 1.1233,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.143681945662679,
+      "learning_rate": 1.8350319071342483e-05,
+      "loss": 1.1941,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.138682306871128,
+      "learning_rate": 1.834879811458063e-05,
+      "loss": 1.2114,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6497531242870194,
+      "learning_rate": 1.8347276520095927e-05,
+      "loss": 1.2441,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.46057628224989,
+      "learning_rate": 1.8345754288004593e-05,
+      "loss": 1.2695,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.099988581567643,
+      "learning_rate": 1.8344231418422903e-05,
+      "loss": 1.1855,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9549882232034532,
+      "learning_rate": 1.8342707911467186e-05,
+      "loss": 1.2249,
+      "step": 2465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.358850713296485,
+      "learning_rate": 1.8341183767253818e-05,
+      "loss": 1.1504,
+      "step": 2466
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.291768053109023,
+      "learning_rate": 1.833965898589921e-05,
+      "loss": 1.2964,
+      "step": 2467
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.480462406012986,
+      "learning_rate": 1.8338133567519844e-05,
+      "loss": 1.137,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9262501613969487,
+      "learning_rate": 1.833660751223223e-05,
+      "loss": 1.2432,
+      "step": 2469
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.692386319603474,
+      "learning_rate": 1.833508082015294e-05,
+      "loss": 1.1775,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.668068858030806,
+      "learning_rate": 1.8333553491398593e-05,
+      "loss": 1.2239,
+      "step": 2471
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.572245499364934,
+      "learning_rate": 1.833202552608585e-05,
+      "loss": 1.2061,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7536608560451494,
+      "learning_rate": 1.8330496924331425e-05,
+      "loss": 1.1411,
+      "step": 2473
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1434296799620105,
+      "learning_rate": 1.832896768625208e-05,
+      "loss": 1.187,
+      "step": 2474
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.40176051872296,
+      "learning_rate": 1.832743781196463e-05,
+      "loss": 1.293,
+      "step": 2475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1107876208485523,
+      "learning_rate": 1.8325907301585924e-05,
+      "loss": 1.2297,
+      "step": 2476
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1812229248935324,
+      "learning_rate": 1.8324376155232887e-05,
+      "loss": 1.1592,
+      "step": 2477
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7308319884854644,
+      "learning_rate": 1.8322844373022458e-05,
+      "loss": 1.2598,
+      "step": 2478
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.746045639481944,
+      "learning_rate": 1.832131195507165e-05,
+      "loss": 1.187,
+      "step": 2479
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.221225531987836,
+      "learning_rate": 1.831977890149752e-05,
+      "loss": 1.2139,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9658745255789336,
+      "learning_rate": 1.8318245212417167e-05,
+      "loss": 1.2095,
+      "step": 2481
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3995972839184505,
+      "learning_rate": 1.831671088794774e-05,
+      "loss": 1.239,
+      "step": 2482
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8656859110051545,
+      "learning_rate": 1.831517592820644e-05,
+      "loss": 1.2422,
+      "step": 2483
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.646463232944538,
+      "learning_rate": 1.8313640333310517e-05,
+      "loss": 1.2642,
+      "step": 2484
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3633636517877883,
+      "learning_rate": 1.8312104103377266e-05,
+      "loss": 1.1265,
+      "step": 2485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.159787429995474,
+      "learning_rate": 1.8310567238524028e-05,
+      "loss": 1.2686,
+      "step": 2486
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.580871620523482,
+      "learning_rate": 1.83090297388682e-05,
+      "loss": 1.291,
+      "step": 2487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.813466073473839,
+      "learning_rate": 1.8307491604527224e-05,
+      "loss": 1.2793,
+      "step": 2488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.148268861076046,
+      "learning_rate": 1.830595283561859e-05,
+      "loss": 1.2095,
+      "step": 2489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3602387825915088,
+      "learning_rate": 1.830441343225984e-05,
+      "loss": 1.1895,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7684803621382033,
+      "learning_rate": 1.8302873394568553e-05,
+      "loss": 1.2864,
+      "step": 2491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7818876586678805,
+      "learning_rate": 1.8301332722662372e-05,
+      "loss": 1.1763,
+      "step": 2492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.051351761923045,
+      "learning_rate": 1.8299791416658982e-05,
+      "loss": 1.2048,
+      "step": 2493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.786672435844319,
+      "learning_rate": 1.8298249476676107e-05,
+      "loss": 1.2573,
+      "step": 2494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.084450340235559,
+      "learning_rate": 1.8296706902831534e-05,
+      "loss": 1.2017,
+      "step": 2495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4463559096342795,
+      "learning_rate": 1.8295163695243097e-05,
+      "loss": 1.2615,
+      "step": 2496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.655161518169077,
+      "learning_rate": 1.8293619854028666e-05,
+      "loss": 1.2117,
+      "step": 2497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.892734692223735,
+      "learning_rate": 1.8292075379306174e-05,
+      "loss": 1.1575,
+      "step": 2498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4814582667825436,
+      "learning_rate": 1.829053027119359e-05,
+      "loss": 1.2278,
+      "step": 2499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.359647257824698,
+      "learning_rate": 1.8288984529808937e-05,
+      "loss": 1.1987,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5949135134281685,
+      "learning_rate": 1.8287438155270288e-05,
+      "loss": 1.27,
+      "step": 2501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.310469086751799,
+      "learning_rate": 1.828589114769577e-05,
+      "loss": 1.3142,
+      "step": 2502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4748333879663935,
+      "learning_rate": 1.828434350720354e-05,
+      "loss": 1.3164,
+      "step": 2503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.715121183209195,
+      "learning_rate": 1.8282795233911818e-05,
+      "loss": 1.2607,
+      "step": 2504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8961178031142194,
+      "learning_rate": 1.8281246327938875e-05,
+      "loss": 1.2554,
+      "step": 2505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5181970524404598,
+      "learning_rate": 1.8279696789403014e-05,
+      "loss": 1.1936,
+      "step": 2506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.820138343481122,
+      "learning_rate": 1.8278146618422603e-05,
+      "loss": 1.2029,
+      "step": 2507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5626116501023644,
+      "learning_rate": 1.8276595815116048e-05,
+      "loss": 1.2083,
+      "step": 2508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.087124347859866,
+      "learning_rate": 1.8275044379601814e-05,
+      "loss": 1.2275,
+      "step": 2509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.511961197775739,
+      "learning_rate": 1.82734923119984e-05,
+      "loss": 1.189,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.244873539103223,
+      "learning_rate": 1.8271939612424364e-05,
+      "loss": 1.2244,
+      "step": 2511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.962547284200098,
+      "learning_rate": 1.827038628099831e-05,
+      "loss": 1.1948,
+      "step": 2512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8115016290097943,
+      "learning_rate": 1.8268832317838886e-05,
+      "loss": 1.2373,
+      "step": 2513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8715321347254545,
+      "learning_rate": 1.8267277723064797e-05,
+      "loss": 1.2122,
+      "step": 2514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.086651004346875,
+      "learning_rate": 1.8265722496794782e-05,
+      "loss": 1.2512,
+      "step": 2515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.268512649339881,
+      "learning_rate": 1.8264166639147644e-05,
+      "loss": 1.2471,
+      "step": 2516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.820617030786134,
+      "learning_rate": 1.8262610150242223e-05,
+      "loss": 1.2354,
+      "step": 2517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2063678670700537,
+      "learning_rate": 1.8261053030197414e-05,
+      "loss": 1.2048,
+      "step": 2518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4625531291392986,
+      "learning_rate": 1.825949527913216e-05,
+      "loss": 1.251,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7956973119005957,
+      "learning_rate": 1.8257936897165445e-05,
+      "loss": 1.2373,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.050145375906522,
+      "learning_rate": 1.8256377884416307e-05,
+      "loss": 1.2,
+      "step": 2521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.301051956142088,
+      "learning_rate": 1.825481824100383e-05,
+      "loss": 1.1782,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.126473340421061,
+      "learning_rate": 1.825325796704715e-05,
+      "loss": 1.1877,
+      "step": 2523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0335999060897936,
+      "learning_rate": 1.825169706266545e-05,
+      "loss": 1.23,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6860307889774275,
+      "learning_rate": 1.825013552797796e-05,
+      "loss": 1.2559,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5044576558953078,
+      "learning_rate": 1.824857336310395e-05,
+      "loss": 1.2559,
+      "step": 2526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2476504944057956,
+      "learning_rate": 1.8247010568162752e-05,
+      "loss": 1.1716,
+      "step": 2527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.097761977966208,
+      "learning_rate": 1.824544714327374e-05,
+      "loss": 1.2456,
+      "step": 2528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1540951587556014,
+      "learning_rate": 1.8243883088556336e-05,
+      "loss": 1.2112,
+      "step": 2529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.014567515896996,
+      "learning_rate": 1.824231840413001e-05,
+      "loss": 1.1892,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3058277395288242,
+      "learning_rate": 1.8240753090114278e-05,
+      "loss": 1.2832,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8683503252489544,
+      "learning_rate": 1.8239187146628712e-05,
+      "loss": 1.1831,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.384838377241696,
+      "learning_rate": 1.823762057379292e-05,
+      "loss": 1.2566,
+      "step": 2533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0792250068593434,
+      "learning_rate": 1.8236053371726573e-05,
+      "loss": 1.2368,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.105173826758403,
+      "learning_rate": 1.823448554054937e-05,
+      "loss": 1.1941,
+      "step": 2535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2320868721897584,
+      "learning_rate": 1.823291708038108e-05,
+      "loss": 1.25,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2415588749015134,
+      "learning_rate": 1.8231347991341505e-05,
+      "loss": 1.2595,
+      "step": 2537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8470158192033908,
+      "learning_rate": 1.8229778273550503e-05,
+      "loss": 1.2305,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.802921314449711,
+      "learning_rate": 1.822820792712797e-05,
+      "loss": 1.1965,
+      "step": 2539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.263842441415264,
+      "learning_rate": 1.8226636952193868e-05,
+      "loss": 1.1978,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.755499103404432,
+      "learning_rate": 1.822506534886819e-05,
+      "loss": 1.2944,
+      "step": 2541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2548145478166317,
+      "learning_rate": 1.8223493117270976e-05,
+      "loss": 1.2871,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.637434446978856,
+      "learning_rate": 1.8221920257522333e-05,
+      "loss": 1.2283,
+      "step": 2543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7914292839699852,
+      "learning_rate": 1.8220346769742395e-05,
+      "loss": 1.1885,
+      "step": 2544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.488976504604572,
+      "learning_rate": 1.8218772654051357e-05,
+      "loss": 1.207,
+      "step": 2545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8060483077770932,
+      "learning_rate": 1.821719791056946e-05,
+      "loss": 1.2302,
+      "step": 2546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3871119437605564,
+      "learning_rate": 1.821562253941698e-05,
+      "loss": 1.2056,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2467146726039537,
+      "learning_rate": 1.8214046540714268e-05,
+      "loss": 1.2532,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.765012014221982,
+      "learning_rate": 1.8212469914581694e-05,
+      "loss": 1.0963,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5767395349954674,
+      "learning_rate": 1.8210892661139695e-05,
+      "loss": 1.1749,
+      "step": 2550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8332222567543157,
+      "learning_rate": 1.8209314780508745e-05,
+      "loss": 1.2798,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1296278761946748,
+      "learning_rate": 1.8207736272809377e-05,
+      "loss": 1.2207,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2603601815009164,
+      "learning_rate": 1.8206157138162154e-05,
+      "loss": 1.2578,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.252242277581454,
+      "learning_rate": 1.8204577376687707e-05,
+      "loss": 1.1714,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9361098233366345,
+      "learning_rate": 1.820299698850671e-05,
+      "loss": 1.2205,
+      "step": 2555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4295315549541754,
+      "learning_rate": 1.820141597373987e-05,
+      "loss": 1.2141,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.3488170384298535,
+      "learning_rate": 1.8199834332507955e-05,
+      "loss": 1.3071,
+      "step": 2557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.39369222993467,
+      "learning_rate": 1.8198252064931784e-05,
+      "loss": 1.2441,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.326510252381525,
+      "learning_rate": 1.819666917113222e-05,
+      "loss": 1.1255,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.144913112794513,
+      "learning_rate": 1.8195085651230165e-05,
+      "loss": 1.2837,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.027311794204639,
+      "learning_rate": 1.819350150534658e-05,
+      "loss": 1.2432,
+      "step": 2561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.242286394805594,
+      "learning_rate": 1.8191916733602467e-05,
+      "loss": 1.1841,
+      "step": 2562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.719893951986188,
+      "learning_rate": 1.8190331336118885e-05,
+      "loss": 1.2034,
+      "step": 2563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.402791380382867,
+      "learning_rate": 1.818874531301693e-05,
+      "loss": 1.2092,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.067704890817412,
+      "learning_rate": 1.818715866441775e-05,
+      "loss": 1.1438,
+      "step": 2565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5564529075082665,
+      "learning_rate": 1.8185571390442542e-05,
+      "loss": 1.2598,
+      "step": 2566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.310341449991961,
+      "learning_rate": 1.8183983491212554e-05,
+      "loss": 1.1665,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5332273462149595,
+      "learning_rate": 1.8182394966849072e-05,
+      "loss": 1.2925,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7769847164491415,
+      "learning_rate": 1.8180805817473436e-05,
+      "loss": 1.1289,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.688431036376976,
+      "learning_rate": 1.8179216043207033e-05,
+      "loss": 1.1169,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0362523004946245,
+      "learning_rate": 1.8177625644171303e-05,
+      "loss": 1.2466,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.76029411911623,
+      "learning_rate": 1.817603462048772e-05,
+      "loss": 1.1985,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7522014456810133,
+      "learning_rate": 1.8174442972277822e-05,
+      "loss": 1.2358,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0879049160570538,
+      "learning_rate": 1.8172850699663184e-05,
+      "loss": 1.1785,
+      "step": 2574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8491634159454393,
+      "learning_rate": 1.8171257802765436e-05,
+      "loss": 1.1401,
+      "step": 2575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6013422580469356,
+      "learning_rate": 1.816966428170624e-05,
+      "loss": 1.2236,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.34428379260684,
+      "learning_rate": 1.816807013660733e-05,
+      "loss": 1.2297,
+      "step": 2577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6074646428056107,
+      "learning_rate": 1.8166475367590467e-05,
+      "loss": 1.2095,
+      "step": 2578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2502343979887707,
+      "learning_rate": 1.816487997477747e-05,
+      "loss": 1.2002,
+      "step": 2579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3462048503450355,
+      "learning_rate": 1.8163283958290203e-05,
+      "loss": 1.2556,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7147867956017335,
+      "learning_rate": 1.8161687318250577e-05,
+      "loss": 1.252,
+      "step": 2581
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.351229209202817,
+      "learning_rate": 1.8160090054780553e-05,
+      "loss": 1.2439,
+      "step": 2582
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.104770644612388,
+      "learning_rate": 1.8158492168002138e-05,
+      "loss": 1.1658,
+      "step": 2583
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8503849645445762,
+      "learning_rate": 1.8156893658037386e-05,
+      "loss": 1.2676,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2468825774309766,
+      "learning_rate": 1.8155294525008394e-05,
+      "loss": 1.2026,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6460992073854417,
+      "learning_rate": 1.815369476903732e-05,
+      "loss": 1.271,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.400146036400721,
+      "learning_rate": 1.8152094390246353e-05,
+      "loss": 1.2612,
+      "step": 2587
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.030345300141349,
+      "learning_rate": 1.8150493388757747e-05,
+      "loss": 1.2244,
+      "step": 2588
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.805881202884494,
+      "learning_rate": 1.8148891764693786e-05,
+      "loss": 1.2363,
+      "step": 2589
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7927484598137333,
+      "learning_rate": 1.814728951817682e-05,
+      "loss": 1.1689,
+      "step": 2590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7246566951613285,
+      "learning_rate": 1.8145686649329223e-05,
+      "loss": 1.2102,
+      "step": 2591
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0708776226459173,
+      "learning_rate": 1.814408315827344e-05,
+      "loss": 1.1936,
+      "step": 2592
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.52937887693488,
+      "learning_rate": 1.8142479045131956e-05,
+      "loss": 1.1311,
+      "step": 2593
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2854341879545825,
+      "learning_rate": 1.8140874310027293e-05,
+      "loss": 1.3091,
+      "step": 2594
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.071079761636288,
+      "learning_rate": 1.813926895308203e-05,
+      "loss": 1.2539,
+      "step": 2595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4172615841510483,
+      "learning_rate": 1.8137662974418798e-05,
+      "loss": 1.2346,
+      "step": 2596
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9780130336753534,
+      "learning_rate": 1.8136056374160267e-05,
+      "loss": 1.2131,
+      "step": 2597
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.490066406203299,
+      "learning_rate": 1.8134449152429153e-05,
+      "loss": 1.2329,
+      "step": 2598
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.359932013702432,
+      "learning_rate": 1.813284130934823e-05,
+      "loss": 1.1296,
+      "step": 2599
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4433579455701055,
+      "learning_rate": 1.8131232845040307e-05,
+      "loss": 1.1777,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.756896090957611,
+      "learning_rate": 1.812962375962825e-05,
+      "loss": 1.2205,
+      "step": 2601
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0098972486718836,
+      "learning_rate": 1.812801405323497e-05,
+      "loss": 1.2039,
+      "step": 2602
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.078179143006481,
+      "learning_rate": 1.8126403725983423e-05,
+      "loss": 1.1792,
+      "step": 2603
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.209135634020277,
+      "learning_rate": 1.8124792777996612e-05,
+      "loss": 1.2383,
+      "step": 2604
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.232019602315093,
+      "learning_rate": 1.812318120939759e-05,
+      "loss": 1.2751,
+      "step": 2605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.970676864159023,
+      "learning_rate": 1.812156902030946e-05,
+      "loss": 1.2053,
+      "step": 2606
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8363432676029037,
+      "learning_rate": 1.8119956210855366e-05,
+      "loss": 1.2363,
+      "step": 2607
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6658376373993575,
+      "learning_rate": 1.81183427811585e-05,
+      "loss": 1.1917,
+      "step": 2608
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.872237213788784,
+      "learning_rate": 1.811672873134211e-05,
+      "loss": 1.2808,
+      "step": 2609
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.108383550197798,
+      "learning_rate": 1.8115114061529483e-05,
+      "loss": 1.2129,
+      "step": 2610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4995769199521054,
+      "learning_rate": 1.811349877184395e-05,
+      "loss": 1.2534,
+      "step": 2611
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7959911864078393,
+      "learning_rate": 1.8111882862408902e-05,
+      "loss": 1.1262,
+      "step": 2612
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0719931844538433,
+      "learning_rate": 1.811026633334777e-05,
+      "loss": 1.2278,
+      "step": 2613
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.176319458639742,
+      "learning_rate": 1.8108649184784024e-05,
+      "loss": 1.24,
+      "step": 2614
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1500243135395714,
+      "learning_rate": 1.81070314168412e-05,
+      "loss": 1.2681,
+      "step": 2615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.412219319335497,
+      "learning_rate": 1.810541302964287e-05,
+      "loss": 1.2822,
+      "step": 2616
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1093832796432253,
+      "learning_rate": 1.8103794023312645e-05,
+      "loss": 1.2266,
+      "step": 2617
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2410634627357164,
+      "learning_rate": 1.8102174397974205e-05,
+      "loss": 1.2456,
+      "step": 2618
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.899853131955289,
+      "learning_rate": 1.8100554153751257e-05,
+      "loss": 1.2397,
+      "step": 2619
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3045570993735924,
+      "learning_rate": 1.809893329076757e-05,
+      "loss": 1.1882,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.24350286876111,
+      "learning_rate": 1.8097311809146946e-05,
+      "loss": 1.2134,
+      "step": 2621
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3648789871847185,
+      "learning_rate": 1.8095689709013245e-05,
+      "loss": 1.2778,
+      "step": 2622
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.984918096270865,
+      "learning_rate": 1.809406699049037e-05,
+      "loss": 1.1497,
+      "step": 2623
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7971609075357455,
+      "learning_rate": 1.8092443653702275e-05,
+      "loss": 1.1824,
+      "step": 2624
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5671207006397445,
+      "learning_rate": 1.809081969877296e-05,
+      "loss": 1.2212,
+      "step": 2625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.193591698965225,
+      "learning_rate": 1.8089195125826466e-05,
+      "loss": 1.2451,
+      "step": 2626
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5937623317263325,
+      "learning_rate": 1.808756993498689e-05,
+      "loss": 1.1545,
+      "step": 2627
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4709733600605546,
+      "learning_rate": 1.808594412637837e-05,
+      "loss": 1.1853,
+      "step": 2628
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8587972123062455,
+      "learning_rate": 1.808431770012509e-05,
+      "loss": 1.1895,
+      "step": 2629
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.011975073544319,
+      "learning_rate": 1.808269065635129e-05,
+      "loss": 1.155,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9904642558670735,
+      "learning_rate": 1.8081062995181254e-05,
+      "loss": 1.1399,
+      "step": 2631
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.617310209914352,
+      "learning_rate": 1.80794347167393e-05,
+      "loss": 1.2827,
+      "step": 2632
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3267198197106786,
+      "learning_rate": 1.8077805821149818e-05,
+      "loss": 1.2419,
+      "step": 2633
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.202786151914347,
+      "learning_rate": 1.807617630853722e-05,
+      "loss": 1.1887,
+      "step": 2634
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.136469452833615,
+      "learning_rate": 1.807454617902598e-05,
+      "loss": 1.2102,
+      "step": 2635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2292039606769776,
+      "learning_rate": 1.8072915432740617e-05,
+      "loss": 1.208,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9563932661276247,
+      "learning_rate": 1.8071284069805693e-05,
+      "loss": 1.1912,
+      "step": 2637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4262743885027054,
+      "learning_rate": 1.8069652090345825e-05,
+      "loss": 1.1924,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1423754737870735,
+      "learning_rate": 1.8068019494485664e-05,
+      "loss": 1.1602,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7832547313772396,
+      "learning_rate": 1.806638628234992e-05,
+      "loss": 1.1682,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9095035095359973,
+      "learning_rate": 1.8064752454063347e-05,
+      "loss": 1.0909,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8788514861549035,
+      "learning_rate": 1.806311800975074e-05,
+      "loss": 1.2273,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1510789879514443,
+      "learning_rate": 1.806148294953695e-05,
+      "loss": 1.2339,
+      "step": 2643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.612835759201951,
+      "learning_rate": 1.8059847273546876e-05,
+      "loss": 1.2468,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0891665188667736,
+      "learning_rate": 1.805821098190545e-05,
+      "loss": 1.2424,
+      "step": 2645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6911294075176877,
+      "learning_rate": 1.8056574074737662e-05,
+      "loss": 1.1367,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5540851771771536,
+      "learning_rate": 1.8054936552168548e-05,
+      "loss": 1.2729,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.611206738374854,
+      "learning_rate": 1.8053298414323195e-05,
+      "loss": 1.2202,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.57255087303339,
+      "learning_rate": 1.8051659661326726e-05,
+      "loss": 1.2937,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.096621974736331,
+      "learning_rate": 1.805002029330432e-05,
+      "loss": 1.2017,
+      "step": 2650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.362954701627282,
+      "learning_rate": 1.80483803103812e-05,
+      "loss": 1.2617,
+      "step": 2651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0712676854140155,
+      "learning_rate": 1.804673971268263e-05,
+      "loss": 1.1987,
+      "step": 2652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9076991899967415,
+      "learning_rate": 1.8045098500333938e-05,
+      "loss": 1.1907,
+      "step": 2653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1980602097608952,
+      "learning_rate": 1.804345667346048e-05,
+      "loss": 1.2229,
+      "step": 2654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.964359974473896,
+      "learning_rate": 1.804181423218767e-05,
+      "loss": 1.2246,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.889105420107583,
+      "learning_rate": 1.8040171176640967e-05,
+      "loss": 1.2456,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.942460575494186,
+      "learning_rate": 1.8038527506945872e-05,
+      "loss": 1.2856,
+      "step": 2657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3733279665924587,
+      "learning_rate": 1.8036883223227937e-05,
+      "loss": 1.2026,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.472820374919624,
+      "learning_rate": 1.8035238325612764e-05,
+      "loss": 1.2656,
+      "step": 2659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.040445447339224,
+      "learning_rate": 1.8033592814226e-05,
+      "loss": 1.1807,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0022541590450347,
+      "learning_rate": 1.803194668919333e-05,
+      "loss": 1.23,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0572138392042874,
+      "learning_rate": 1.8030299950640496e-05,
+      "loss": 1.1777,
+      "step": 2662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9608456812382404,
+      "learning_rate": 1.802865259869329e-05,
+      "loss": 1.2095,
+      "step": 2663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.283907759108376,
+      "learning_rate": 1.802700463347754e-05,
+      "loss": 1.1892,
+      "step": 2664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4224678208087567,
+      "learning_rate": 1.8025356055119122e-05,
+      "loss": 1.2505,
+      "step": 2665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3878690876271222,
+      "learning_rate": 1.802370686374397e-05,
+      "loss": 1.1599,
+      "step": 2666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.315583257952846,
+      "learning_rate": 1.8022057059478054e-05,
+      "loss": 1.1987,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1826807678838693,
+      "learning_rate": 1.8020406642447393e-05,
+      "loss": 1.2158,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7654795904249645,
+      "learning_rate": 1.8018755612778062e-05,
+      "loss": 1.2568,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.980715713449648,
+      "learning_rate": 1.801710397059616e-05,
+      "loss": 1.2273,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.273788374450234,
+      "learning_rate": 1.8015451716027863e-05,
+      "loss": 1.2876,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9550093256265217,
+      "learning_rate": 1.8013798849199368e-05,
+      "loss": 1.1892,
+      "step": 2672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0930513454939326,
+      "learning_rate": 1.8012145370236932e-05,
+      "loss": 1.2068,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.157954434456808,
+      "learning_rate": 1.801049127926686e-05,
+      "loss": 1.2358,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.907630732382347,
+      "learning_rate": 1.8008836576415493e-05,
+      "loss": 1.1985,
+      "step": 2675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9290443366904233,
+      "learning_rate": 1.8007181261809235e-05,
+      "loss": 1.219,
+      "step": 2676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.217212077248875,
+      "learning_rate": 1.8005525335574515e-05,
+      "loss": 1.1978,
+      "step": 2677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.825641645281977,
+      "learning_rate": 1.800386879783783e-05,
+      "loss": 1.2053,
+      "step": 2678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1597423913143716,
+      "learning_rate": 1.8002211648725714e-05,
+      "loss": 1.2051,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3111377171939296,
+      "learning_rate": 1.8000553888364743e-05,
+      "loss": 1.2485,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.470243675806716,
+      "learning_rate": 1.799889551688155e-05,
+      "loss": 1.2056,
+      "step": 2681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1969410660786317,
+      "learning_rate": 1.7997236534402803e-05,
+      "loss": 1.3003,
+      "step": 2682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6757713009656285,
+      "learning_rate": 1.7995576941055236e-05,
+      "loss": 1.2107,
+      "step": 2683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0783157092889515,
+      "learning_rate": 1.7993916736965604e-05,
+      "loss": 1.2476,
+      "step": 2684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3064769644097964,
+      "learning_rate": 1.7992255922260726e-05,
+      "loss": 1.1587,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.826388603642371,
+      "learning_rate": 1.799059449706747e-05,
+      "loss": 1.2927,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4109489422486976,
+      "learning_rate": 1.7988932461512733e-05,
+      "loss": 1.2256,
+      "step": 2687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.846877188399067,
+      "learning_rate": 1.7987269815723474e-05,
+      "loss": 1.178,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.457522527895155,
+      "learning_rate": 1.79856065598267e-05,
+      "loss": 1.1611,
+      "step": 2689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.001187140242127,
+      "learning_rate": 1.798394269394945e-05,
+      "loss": 1.2224,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.642137119735271,
+      "learning_rate": 1.798227821821882e-05,
+      "loss": 1.1748,
+      "step": 2691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9738328562384644,
+      "learning_rate": 1.7980613132761956e-05,
+      "loss": 1.1541,
+      "step": 2692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.943318688200656,
+      "learning_rate": 1.797894743770604e-05,
+      "loss": 1.2056,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.26838584674101,
+      "learning_rate": 1.7977281133178312e-05,
+      "loss": 1.2097,
+      "step": 2694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.180088803586519,
+      "learning_rate": 1.797561421930605e-05,
+      "loss": 1.2573,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1099034501600684,
+      "learning_rate": 1.7973946696216576e-05,
+      "loss": 1.2412,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.630023513809248,
+      "learning_rate": 1.797227856403727e-05,
+      "loss": 1.2422,
+      "step": 2697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.12711884326332,
+      "learning_rate": 1.7970609822895552e-05,
+      "loss": 1.2236,
+      "step": 2698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6417479724765616,
+      "learning_rate": 1.7968940472918887e-05,
+      "loss": 1.2559,
+      "step": 2699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3847138337048728,
+      "learning_rate": 1.7967270514234788e-05,
+      "loss": 1.27,
+      "step": 2700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4555608549365875,
+      "learning_rate": 1.7965599946970814e-05,
+      "loss": 1.1641,
+      "step": 2701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.630234069450089,
+      "learning_rate": 1.7963928771254574e-05,
+      "loss": 1.1777,
+      "step": 2702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.810516723752082,
+      "learning_rate": 1.796225698721372e-05,
+      "loss": 1.1555,
+      "step": 2703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3380334783378953,
+      "learning_rate": 1.796058459497595e-05,
+      "loss": 1.2163,
+      "step": 2704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.526251603011823,
+      "learning_rate": 1.795891159466901e-05,
+      "loss": 1.2759,
+      "step": 2705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.989006625597699,
+      "learning_rate": 1.7957237986420692e-05,
+      "loss": 1.2546,
+      "step": 2706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.820371992651381,
+      "learning_rate": 1.7955563770358834e-05,
+      "loss": 1.2192,
+      "step": 2707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.504909253545486,
+      "learning_rate": 1.7953888946611325e-05,
+      "loss": 1.2397,
+      "step": 2708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7788594914454197,
+      "learning_rate": 1.795221351530609e-05,
+      "loss": 1.2832,
+      "step": 2709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4936532928924016,
+      "learning_rate": 1.795053747657111e-05,
+      "loss": 1.2148,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.122617887048826,
+      "learning_rate": 1.794886083053441e-05,
+      "loss": 1.2024,
+      "step": 2711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5240538108574713,
+      "learning_rate": 1.794718357732406e-05,
+      "loss": 1.2393,
+      "step": 2712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.332423219791186,
+      "learning_rate": 1.7945505717068176e-05,
+      "loss": 1.2131,
+      "step": 2713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.354035220331127,
+      "learning_rate": 1.7943827249894926e-05,
+      "loss": 1.2627,
+      "step": 2714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1383108428113475,
+      "learning_rate": 1.794214817593251e-05,
+      "loss": 1.2368,
+      "step": 2715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1018333024069222,
+      "learning_rate": 1.7940468495309192e-05,
+      "loss": 1.2385,
+      "step": 2716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4776113639065365,
+      "learning_rate": 1.7938788208153274e-05,
+      "loss": 1.1536,
+      "step": 2717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.997199744502663,
+      "learning_rate": 1.7937107314593102e-05,
+      "loss": 1.2559,
+      "step": 2718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.148673968924332,
+      "learning_rate": 1.793542581475707e-05,
+      "loss": 1.2019,
+      "step": 2719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.272270895725558,
+      "learning_rate": 1.7933743708773624e-05,
+      "loss": 1.2366,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1258490258453335,
+      "learning_rate": 1.793206099677125e-05,
+      "loss": 1.1738,
+      "step": 2721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.030256261909126,
+      "learning_rate": 1.793037767887848e-05,
+      "loss": 1.1787,
+      "step": 2722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.951741906699162,
+      "learning_rate": 1.7928693755223894e-05,
+      "loss": 1.1824,
+      "step": 2723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1380613154669708,
+      "learning_rate": 1.792700922593612e-05,
+      "loss": 1.2339,
+      "step": 2724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.995896561200463,
+      "learning_rate": 1.792532409114383e-05,
+      "loss": 1.2822,
+      "step": 2725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5570732785431907,
+      "learning_rate": 1.792363835097574e-05,
+      "loss": 1.1841,
+      "step": 2726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0859208630161645,
+      "learning_rate": 1.7921952005560627e-05,
+      "loss": 1.1221,
+      "step": 2727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.707314161745712,
+      "learning_rate": 1.7920265055027285e-05,
+      "loss": 1.354,
+      "step": 2728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.368990740551882,
+      "learning_rate": 1.7918577499504587e-05,
+      "loss": 1.2466,
+      "step": 2729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.897020380622683,
+      "learning_rate": 1.791688933912143e-05,
+      "loss": 1.2693,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.523242044869145,
+      "learning_rate": 1.791520057400676e-05,
+      "loss": 1.2983,
+      "step": 2731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.129737045491186,
+      "learning_rate": 1.7913511204289585e-05,
+      "loss": 1.1924,
+      "step": 2732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9006497175764934,
+      "learning_rate": 1.7911821230098937e-05,
+      "loss": 1.1375,
+      "step": 2733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2097821191651987,
+      "learning_rate": 1.791013065156391e-05,
+      "loss": 1.2117,
+      "step": 2734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.020080572777109,
+      "learning_rate": 1.7908439468813634e-05,
+      "loss": 1.2,
+      "step": 2735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1048847636053356,
+      "learning_rate": 1.7906747681977296e-05,
+      "loss": 1.2427,
+      "step": 2736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5567160712419814,
+      "learning_rate": 1.790505529118412e-05,
+      "loss": 1.249,
+      "step": 2737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.068640308002033,
+      "learning_rate": 1.7903362296563377e-05,
+      "loss": 1.2817,
+      "step": 2738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.338933514903848,
+      "learning_rate": 1.790166869824439e-05,
+      "loss": 1.1917,
+      "step": 2739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.544218968226937,
+      "learning_rate": 1.7899974496356523e-05,
+      "loss": 1.228,
+      "step": 2740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.541979637602309,
+      "learning_rate": 1.789827969102919e-05,
+      "loss": 1.1851,
+      "step": 2741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.182100527398148,
+      "learning_rate": 1.7896584282391844e-05,
+      "loss": 1.1768,
+      "step": 2742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.313041822426718,
+      "learning_rate": 1.7894888270573997e-05,
+      "loss": 1.1724,
+      "step": 2743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6733933193920048,
+      "learning_rate": 1.7893191655705188e-05,
+      "loss": 1.2563,
+      "step": 2744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4724259740580656,
+      "learning_rate": 1.7891494437915017e-05,
+      "loss": 1.1648,
+      "step": 2745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8094262700917487,
+      "learning_rate": 1.788979661733313e-05,
+      "loss": 1.25,
+      "step": 2746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1192133677204756,
+      "learning_rate": 1.788809819408921e-05,
+      "loss": 1.1973,
+      "step": 2747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9345886717839296,
+      "learning_rate": 1.7886399168312995e-05,
+      "loss": 1.1909,
+      "step": 2748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8464128446487535,
+      "learning_rate": 1.788469954013426e-05,
+      "loss": 1.1526,
+      "step": 2749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.783231394272841,
+      "learning_rate": 1.788299930968284e-05,
+      "loss": 1.2285,
+      "step": 2750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3645901990555704,
+      "learning_rate": 1.78812984770886e-05,
+      "loss": 1.2021,
+      "step": 2751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.146390671107296,
+      "learning_rate": 1.7879597042481456e-05,
+      "loss": 1.2024,
+      "step": 2752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5914897871695497,
+      "learning_rate": 1.7877895005991376e-05,
+      "loss": 1.2651,
+      "step": 2753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.084212719377022,
+      "learning_rate": 1.787619236774837e-05,
+      "loss": 1.1313,
+      "step": 2754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1404394828480635,
+      "learning_rate": 1.7874489127882496e-05,
+      "loss": 1.1938,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.388968437901819,
+      "learning_rate": 1.787278528652385e-05,
+      "loss": 1.2441,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6383955463247273,
+      "learning_rate": 1.7871080843802588e-05,
+      "loss": 1.1658,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.745098965102097,
+      "learning_rate": 1.7869375799848896e-05,
+      "loss": 1.2065,
+      "step": 2758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4945996837452165,
+      "learning_rate": 1.7867670154793018e-05,
+      "loss": 1.2515,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5243368115233813,
+      "learning_rate": 1.7865963908765237e-05,
+      "loss": 1.2146,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.11848861253568,
+      "learning_rate": 1.786425706189589e-05,
+      "loss": 1.2808,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.733458169262,
+      "learning_rate": 1.786254961431535e-05,
+      "loss": 1.1943,
+      "step": 2762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.433437350220125,
+      "learning_rate": 1.7860841566154036e-05,
+      "loss": 1.2075,
+      "step": 2763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.613014242127071,
+      "learning_rate": 1.7859132917542427e-05,
+      "loss": 1.3062,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.579882472301165,
+      "learning_rate": 1.7857423668611033e-05,
+      "loss": 1.1882,
+      "step": 2765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6878039717279667,
+      "learning_rate": 1.7855713819490415e-05,
+      "loss": 1.1809,
+      "step": 2766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7618730441355104,
+      "learning_rate": 1.785400337031118e-05,
+      "loss": 1.2048,
+      "step": 2767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.14367021365456,
+      "learning_rate": 1.785229232120398e-05,
+      "loss": 1.2402,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7532488425174098,
+      "learning_rate": 1.7850580672299517e-05,
+      "loss": 1.1738,
+      "step": 2769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1648636475510568,
+      "learning_rate": 1.7848868423728527e-05,
+      "loss": 1.2141,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.651713751705468,
+      "learning_rate": 1.7847155575621813e-05,
+      "loss": 1.1982,
+      "step": 2771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.23814312598025,
+      "learning_rate": 1.78454421281102e-05,
+      "loss": 1.1636,
+      "step": 2772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.196248762258087,
+      "learning_rate": 1.7843728081324575e-05,
+      "loss": 1.1948,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.472735528770448,
+      "learning_rate": 1.7842013435395862e-05,
+      "loss": 1.2146,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9577896438195146,
+      "learning_rate": 1.7840298190455033e-05,
+      "loss": 1.2104,
+      "step": 2775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8937923610263874,
+      "learning_rate": 1.7838582346633116e-05,
+      "loss": 1.2019,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.922128456584316,
+      "learning_rate": 1.7836865904061165e-05,
+      "loss": 1.1553,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9632936584981984,
+      "learning_rate": 1.7835148862870298e-05,
+      "loss": 1.1077,
+      "step": 2778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.027971963270297,
+      "learning_rate": 1.7833431223191666e-05,
+      "loss": 1.2266,
+      "step": 2779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.635394087609985,
+      "learning_rate": 1.7831712985156472e-05,
+      "loss": 1.2004,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.430929867811702,
+      "learning_rate": 1.7829994148895972e-05,
+      "loss": 1.25,
+      "step": 2781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1844358101427845,
+      "learning_rate": 1.7828274714541445e-05,
+      "loss": 1.2817,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8797632746725563,
+      "learning_rate": 1.7826554682224236e-05,
+      "loss": 1.1416,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.184673872360237,
+      "learning_rate": 1.7824834052075737e-05,
+      "loss": 1.2395,
+      "step": 2784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.004805424141257,
+      "learning_rate": 1.782311282422737e-05,
+      "loss": 1.2019,
+      "step": 2785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.976053795889173,
+      "learning_rate": 1.782139099881061e-05,
+      "loss": 1.1497,
+      "step": 2786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.553263066675215,
+      "learning_rate": 1.7819668575956984e-05,
+      "loss": 1.178,
+      "step": 2787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.10749005846315,
+      "learning_rate": 1.7817945555798054e-05,
+      "loss": 1.2515,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2442946548170926,
+      "learning_rate": 1.781622193846544e-05,
+      "loss": 1.228,
+      "step": 2789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.264489418847435,
+      "learning_rate": 1.7814497724090793e-05,
+      "loss": 1.2234,
+      "step": 2790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0628568926537167,
+      "learning_rate": 1.781277291280582e-05,
+      "loss": 1.1909,
+      "step": 2791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4903981823310755,
+      "learning_rate": 1.7811047504742273e-05,
+      "loss": 1.2324,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.185627427886519,
+      "learning_rate": 1.7809321500031945e-05,
+      "loss": 1.2222,
+      "step": 2793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2942378236287833,
+      "learning_rate": 1.7807594898806676e-05,
+      "loss": 1.1782,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.514652139312499,
+      "learning_rate": 1.780586770119835e-05,
+      "loss": 1.136,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1432436100026924,
+      "learning_rate": 1.780413990733891e-05,
+      "loss": 1.1951,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1444656334126164,
+      "learning_rate": 1.7802411517360322e-05,
+      "loss": 1.229,
+      "step": 2797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.1018073863765645,
+      "learning_rate": 1.780068253139461e-05,
+      "loss": 1.2046,
+      "step": 2798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.463604578872547,
+      "learning_rate": 1.779895294957385e-05,
+      "loss": 1.2539,
+      "step": 2799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4053182886143945,
+      "learning_rate": 1.7797222772030155e-05,
+      "loss": 1.2065,
+      "step": 2800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3570881231633627,
+      "learning_rate": 1.7795491998895673e-05,
+      "loss": 1.2329,
+      "step": 2801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9310877144847125,
+      "learning_rate": 1.779376063030262e-05,
+      "loss": 1.1855,
+      "step": 2802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.075262733264323,
+      "learning_rate": 1.7792028666383247e-05,
+      "loss": 1.2373,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.049912533130151,
+      "learning_rate": 1.7790296107269845e-05,
+      "loss": 1.2078,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3588022001546025,
+      "learning_rate": 1.7788562953094758e-05,
+      "loss": 1.2849,
+      "step": 2805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1952175095511866,
+      "learning_rate": 1.7786829203990375e-05,
+      "loss": 1.1968,
+      "step": 2806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8695173120147173,
+      "learning_rate": 1.778509486008912e-05,
+      "loss": 1.1819,
+      "step": 2807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3614855888174864,
+      "learning_rate": 1.778335992152348e-05,
+      "loss": 1.2598,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0445324882468583,
+      "learning_rate": 1.7781624388425974e-05,
+      "loss": 1.0889,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.660376899744107,
+      "learning_rate": 1.7779888260929173e-05,
+      "loss": 1.2427,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.271254599174062,
+      "learning_rate": 1.7778151539165688e-05,
+      "loss": 1.207,
+      "step": 2811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.395691694552848,
+      "learning_rate": 1.7776414223268184e-05,
+      "loss": 1.2107,
+      "step": 2812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.307557844041287,
+      "learning_rate": 1.7774676313369358e-05,
+      "loss": 1.1736,
+      "step": 2813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5025838217770753,
+      "learning_rate": 1.7772937809601967e-05,
+      "loss": 1.2778,
+      "step": 2814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.444555823524407,
+      "learning_rate": 1.77711987120988e-05,
+      "loss": 1.2803,
+      "step": 2815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.283929575114286,
+      "learning_rate": 1.7769459020992706e-05,
+      "loss": 1.269,
+      "step": 2816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9199998102025773,
+      "learning_rate": 1.7767718736416565e-05,
+      "loss": 1.157,
+      "step": 2817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.185286903064244,
+      "learning_rate": 1.7765977858503314e-05,
+      "loss": 1.2295,
+      "step": 2818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8373734456791353,
+      "learning_rate": 1.776423638738592e-05,
+      "loss": 1.2202,
+      "step": 2819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.868778019806126,
+      "learning_rate": 1.776249432319742e-05,
+      "loss": 1.2041,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.451962438697562,
+      "learning_rate": 1.7760751666070867e-05,
+      "loss": 1.1484,
+      "step": 2821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2074318569818048,
+      "learning_rate": 1.775900841613938e-05,
+      "loss": 1.2632,
+      "step": 2822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0114311714580593,
+      "learning_rate": 1.7757264573536124e-05,
+      "loss": 1.2249,
+      "step": 2823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.562156902280209,
+      "learning_rate": 1.7755520138394292e-05,
+      "loss": 1.2002,
+      "step": 2824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8864721501315804,
+      "learning_rate": 1.7753775110847137e-05,
+      "loss": 1.228,
+      "step": 2825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.054000593567928,
+      "learning_rate": 1.7752029491027952e-05,
+      "loss": 1.2224,
+      "step": 2826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.656557698068986,
+      "learning_rate": 1.7750283279070077e-05,
+      "loss": 1.2632,
+      "step": 2827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.983901776809663,
+      "learning_rate": 1.7748536475106895e-05,
+      "loss": 1.2122,
+      "step": 2828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3088911077289413,
+      "learning_rate": 1.7746789079271836e-05,
+      "loss": 1.2039,
+      "step": 2829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.172740267136887,
+      "learning_rate": 1.774504109169838e-05,
+      "loss": 1.1611,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9192230732809765,
+      "learning_rate": 1.7743292512520038e-05,
+      "loss": 1.2412,
+      "step": 2831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0092626328660406,
+      "learning_rate": 1.774154334187038e-05,
+      "loss": 1.2024,
+      "step": 2832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1755421382488542,
+      "learning_rate": 1.7739793579883015e-05,
+      "loss": 1.1689,
+      "step": 2833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0586434466245263,
+      "learning_rate": 1.77380432266916e-05,
+      "loss": 1.1729,
+      "step": 2834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.79865821786929,
+      "learning_rate": 1.7736292282429838e-05,
+      "loss": 1.251,
+      "step": 2835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2109760203220494,
+      "learning_rate": 1.773454074723147e-05,
+      "loss": 1.2239,
+      "step": 2836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.932556927398796,
+      "learning_rate": 1.773278862123029e-05,
+      "loss": 1.0732,
+      "step": 2837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1667036084721616,
+      "learning_rate": 1.773103590456013e-05,
+      "loss": 1.2749,
+      "step": 2838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.72069796142718,
+      "learning_rate": 1.772928259735488e-05,
+      "loss": 1.1709,
+      "step": 2839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.352684751490275,
+      "learning_rate": 1.772752869974846e-05,
+      "loss": 1.1692,
+      "step": 2840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1931847035323453,
+      "learning_rate": 1.7725774211874837e-05,
+      "loss": 1.199,
+      "step": 2841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.124761717026569,
+      "learning_rate": 1.7724019133868035e-05,
+      "loss": 1.2073,
+      "step": 2842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.724200173105194,
+      "learning_rate": 1.7722263465862114e-05,
+      "loss": 1.2424,
+      "step": 2843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.986416556361647,
+      "learning_rate": 1.7720507207991183e-05,
+      "loss": 1.2012,
+      "step": 2844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.51638375770153,
+      "learning_rate": 1.771875036038939e-05,
+      "loss": 1.1733,
+      "step": 2845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.008448759403535,
+      "learning_rate": 1.771699292319093e-05,
+      "loss": 1.1914,
+      "step": 2846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2002659031192384,
+      "learning_rate": 1.771523489653005e-05,
+      "loss": 1.1848,
+      "step": 2847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.330644101977467,
+      "learning_rate": 1.7713476280541032e-05,
+      "loss": 1.2419,
+      "step": 2848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.949321561426879,
+      "learning_rate": 1.771171707535821e-05,
+      "loss": 1.2268,
+      "step": 2849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.813957917059783,
+      "learning_rate": 1.770995728111596e-05,
+      "loss": 1.2056,
+      "step": 2850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4185757276548694,
+      "learning_rate": 1.7708196897948708e-05,
+      "loss": 1.1819,
+      "step": 2851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5041051423303458,
+      "learning_rate": 1.7706435925990915e-05,
+      "loss": 1.2451,
+      "step": 2852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.429285264405626,
+      "learning_rate": 1.7704674365377097e-05,
+      "loss": 1.1853,
+      "step": 2853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.969595220358995,
+      "learning_rate": 1.770291221624181e-05,
+      "loss": 1.2319,
+      "step": 2854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1335396679312955,
+      "learning_rate": 1.7701149478719654e-05,
+      "loss": 1.21,
+      "step": 2855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.041860196646866,
+      "learning_rate": 1.7699386152945277e-05,
+      "loss": 1.2109,
+      "step": 2856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0424422753155573,
+      "learning_rate": 1.769762223905337e-05,
+      "loss": 1.229,
+      "step": 2857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.067530054640901,
+      "learning_rate": 1.7695857737178673e-05,
+      "loss": 1.2776,
+      "step": 2858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9443773069053334,
+      "learning_rate": 1.769409264745596e-05,
+      "loss": 1.2122,
+      "step": 2859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0578748410320378,
+      "learning_rate": 1.7692326970020066e-05,
+      "loss": 1.2266,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7432160778281425,
+      "learning_rate": 1.7690560705005854e-05,
+      "loss": 1.1251,
+      "step": 2861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.156834458746672,
+      "learning_rate": 1.7688793852548246e-05,
+      "loss": 1.252,
+      "step": 2862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.918831602410599,
+      "learning_rate": 1.76870264127822e-05,
+      "loss": 1.2266,
+      "step": 2863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1602325565806026,
+      "learning_rate": 1.768525838584273e-05,
+      "loss": 1.2297,
+      "step": 2864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.43818884539346,
+      "learning_rate": 1.768348977186487e-05,
+      "loss": 1.1702,
+      "step": 2865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.53734132484359,
+      "learning_rate": 1.7681720570983733e-05,
+      "loss": 1.2227,
+      "step": 2866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3212240455867863,
+      "learning_rate": 1.7679950783334447e-05,
+      "loss": 1.188,
+      "step": 2867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5411209351428568,
+      "learning_rate": 1.7678180409052202e-05,
+      "loss": 1.217,
+      "step": 2868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.422664070368076,
+      "learning_rate": 1.7676409448272228e-05,
+      "loss": 1.1492,
+      "step": 2869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8083063369503938,
+      "learning_rate": 1.76746379011298e-05,
+      "loss": 1.217,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.185169398383234,
+      "learning_rate": 1.7672865767760237e-05,
+      "loss": 1.2212,
+      "step": 2871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1160758164631988,
+      "learning_rate": 1.7671093048298905e-05,
+      "loss": 1.239,
+      "step": 2872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0781962696135468,
+      "learning_rate": 1.7669319742881216e-05,
+      "loss": 1.167,
+      "step": 2873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.946436395705582,
+      "learning_rate": 1.7667545851642612e-05,
+      "loss": 1.2427,
+      "step": 2874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.044078193651779,
+      "learning_rate": 1.7665771374718604e-05,
+      "loss": 1.2588,
+      "step": 2875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0154057608969094,
+      "learning_rate": 1.7663996312244728e-05,
+      "loss": 1.1772,
+      "step": 2876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2221855879034704,
+      "learning_rate": 1.7662220664356577e-05,
+      "loss": 1.2542,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7119147808464126,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 1.2231,
+      "step": 2878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9927401254109838,
+      "learning_rate": 1.765866761288002e-05,
+      "loss": 1.1541,
+      "step": 2879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3135217202030036,
+      "learning_rate": 1.7656890209563015e-05,
+      "loss": 1.251,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1353173593333086,
+      "learning_rate": 1.765511222137453e-05,
+      "loss": 1.2144,
+      "step": 2881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8535155767845852,
+      "learning_rate": 1.765333364845038e-05,
+      "loss": 1.1973,
+      "step": 2882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8979740509730405,
+      "learning_rate": 1.765155449092642e-05,
+      "loss": 1.2056,
+      "step": 2883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5364325639818066,
+      "learning_rate": 1.7649774748938553e-05,
+      "loss": 1.2263,
+      "step": 2884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.045489744584048,
+      "learning_rate": 1.7647994422622723e-05,
+      "loss": 1.1628,
+      "step": 2885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.084238044502409,
+      "learning_rate": 1.764621351211492e-05,
+      "loss": 1.293,
+      "step": 2886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.223312797858227,
+      "learning_rate": 1.7644432017551177e-05,
+      "loss": 1.2175,
+      "step": 2887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5948202610110664,
+      "learning_rate": 1.7642649939067577e-05,
+      "loss": 1.2124,
+      "step": 2888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.322400324491537,
+      "learning_rate": 1.7640867276800242e-05,
+      "loss": 1.2102,
+      "step": 2889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1520649626125334,
+      "learning_rate": 1.763908403088534e-05,
+      "loss": 1.1897,
+      "step": 2890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3703627052014173,
+      "learning_rate": 1.763730020145909e-05,
+      "loss": 1.2637,
+      "step": 2891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.973758573527168,
+      "learning_rate": 1.7635515788657742e-05,
+      "loss": 1.1855,
+      "step": 2892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3622396950613016,
+      "learning_rate": 1.76337307926176e-05,
+      "loss": 1.1406,
+      "step": 2893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2473032648375484,
+      "learning_rate": 1.7631945213475016e-05,
+      "loss": 1.2148,
+      "step": 2894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1497711711808645,
+      "learning_rate": 1.7630159051366376e-05,
+      "loss": 1.2329,
+      "step": 2895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.539801535234168,
+      "learning_rate": 1.7628372306428113e-05,
+      "loss": 1.3188,
+      "step": 2896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5317495637410468,
+      "learning_rate": 1.762658497879672e-05,
+      "loss": 1.2847,
+      "step": 2897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0696436298138488,
+      "learning_rate": 1.7624797068608713e-05,
+      "loss": 1.1958,
+      "step": 2898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.077909709117118,
+      "learning_rate": 1.762300857600066e-05,
+      "loss": 1.2278,
+      "step": 2899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2305357839169053,
+      "learning_rate": 1.762121950110918e-05,
+      "loss": 1.21,
+      "step": 2900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.989333578097943,
+      "learning_rate": 1.7619429844070926e-05,
+      "loss": 1.2285,
+      "step": 2901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1729534240837554,
+      "learning_rate": 1.7617639605022606e-05,
+      "loss": 1.2446,
+      "step": 2902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9360334826010575,
+      "learning_rate": 1.7615848784100967e-05,
+      "loss": 1.207,
+      "step": 2903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9253629732741606,
+      "learning_rate": 1.76140573814428e-05,
+      "loss": 1.1604,
+      "step": 2904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1854433107761326,
+      "learning_rate": 1.7612265397184938e-05,
+      "loss": 1.2705,
+      "step": 2905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.524092908413508,
+      "learning_rate": 1.7610472831464265e-05,
+      "loss": 1.2502,
+      "step": 2906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.70862231187327,
+      "learning_rate": 1.7608679684417708e-05,
+      "loss": 1.2104,
+      "step": 2907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.568916554868253,
+      "learning_rate": 1.7606885956182228e-05,
+      "loss": 1.2222,
+      "step": 2908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.764572534402072,
+      "learning_rate": 1.760509164689485e-05,
+      "loss": 1.2432,
+      "step": 2909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.898385152078298,
+      "learning_rate": 1.760329675669263e-05,
+      "loss": 1.2109,
+      "step": 2910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.771424033750521,
+      "learning_rate": 1.760150128571266e-05,
+      "loss": 1.2002,
+      "step": 2911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9785524569250907,
+      "learning_rate": 1.75997052340921e-05,
+      "loss": 1.2134,
+      "step": 2912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7656037453194036,
+      "learning_rate": 1.7597908601968136e-05,
+      "loss": 1.1938,
+      "step": 2913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.073207283943072,
+      "learning_rate": 1.7596111389478003e-05,
+      "loss": 1.1831,
+      "step": 2914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9610180620882813,
+      "learning_rate": 1.759431359675898e-05,
+      "loss": 1.179,
+      "step": 2915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1437862947295176,
+      "learning_rate": 1.75925152239484e-05,
+      "loss": 1.1038,
+      "step": 2916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5343854739769993,
+      "learning_rate": 1.759071627118362e-05,
+      "loss": 1.259,
+      "step": 2917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2566695203685447,
+      "learning_rate": 1.758891673860206e-05,
+      "loss": 1.1238,
+      "step": 2918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2896028598498064,
+      "learning_rate": 1.7587116626341173e-05,
+      "loss": 1.2742,
+      "step": 2919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9055375853823286,
+      "learning_rate": 1.7585315934538468e-05,
+      "loss": 1.2461,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.892826098589966,
+      "learning_rate": 1.758351466333148e-05,
+      "loss": 1.1978,
+      "step": 2921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.758475700574752,
+      "learning_rate": 1.7581712812857808e-05,
+      "loss": 1.1863,
+      "step": 2922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2201301507173508,
+      "learning_rate": 1.7579910383255085e-05,
+      "loss": 1.1614,
+      "step": 2923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3485547038130226,
+      "learning_rate": 1.7578107374660985e-05,
+      "loss": 1.1453,
+      "step": 2924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.190662997565487,
+      "learning_rate": 1.7576303787213232e-05,
+      "loss": 1.1777,
+      "step": 2925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.419930471265698,
+      "learning_rate": 1.75744996210496e-05,
+      "loss": 1.2144,
+      "step": 2926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.690480534476322,
+      "learning_rate": 1.7572694876307892e-05,
+      "loss": 1.1719,
+      "step": 2927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4001824937267977,
+      "learning_rate": 1.7570889553125963e-05,
+      "loss": 1.2417,
+      "step": 2928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3538678543313107,
+      "learning_rate": 1.756908365164172e-05,
+      "loss": 1.2183,
+      "step": 2929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1140068604422093,
+      "learning_rate": 1.7567277171993104e-05,
+      "loss": 1.2559,
+      "step": 2930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4388038164114807,
+      "learning_rate": 1.75654701143181e-05,
+      "loss": 1.0486,
+      "step": 2931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.912930090223677,
+      "learning_rate": 1.7563662478754737e-05,
+      "loss": 1.2466,
+      "step": 2932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.085965455809853,
+      "learning_rate": 1.7561854265441097e-05,
+      "loss": 1.2642,
+      "step": 2933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9997963131486505,
+      "learning_rate": 1.7560045474515305e-05,
+      "loss": 1.1471,
+      "step": 2934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.445055359751054,
+      "learning_rate": 1.7558236106115514e-05,
+      "loss": 1.1765,
+      "step": 2935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4901656824224765,
+      "learning_rate": 1.755642616037994e-05,
+      "loss": 1.2529,
+      "step": 2936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.191464819379458,
+      "learning_rate": 1.7554615637446836e-05,
+      "loss": 1.135,
+      "step": 2937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4693023746116163,
+      "learning_rate": 1.7552804537454494e-05,
+      "loss": 1.1121,
+      "step": 2938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.691634865021856,
+      "learning_rate": 1.755099286054126e-05,
+      "loss": 1.1997,
+      "step": 2939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4690464362305065,
+      "learning_rate": 1.7549180606845513e-05,
+      "loss": 1.2844,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.34612100470915,
+      "learning_rate": 1.7547367776505688e-05,
+      "loss": 1.2661,
+      "step": 2941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.355816988615716,
+      "learning_rate": 1.7545554369660256e-05,
+      "loss": 1.1924,
+      "step": 2942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0593037308032813,
+      "learning_rate": 1.7543740386447732e-05,
+      "loss": 1.2039,
+      "step": 2943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5480662748696155,
+      "learning_rate": 1.754192582700668e-05,
+      "loss": 1.1975,
+      "step": 2944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3099764422486126,
+      "learning_rate": 1.75401106914757e-05,
+      "loss": 1.2361,
+      "step": 2945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.051483424783243,
+      "learning_rate": 1.753829497999345e-05,
+      "loss": 1.1812,
+      "step": 2946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.210430379892056,
+      "learning_rate": 1.7536478692698615e-05,
+      "loss": 1.1865,
+      "step": 2947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8107622743699547,
+      "learning_rate": 1.7534661829729934e-05,
+      "loss": 1.167,
+      "step": 2948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.008286091673133,
+      "learning_rate": 1.753284439122619e-05,
+      "loss": 1.2202,
+      "step": 2949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.81171667589951,
+      "learning_rate": 1.753102637732621e-05,
+      "loss": 1.177,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9155228593530187,
+      "learning_rate": 1.7529207788168857e-05,
+      "loss": 1.1987,
+      "step": 2951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.1789322718917115,
+      "learning_rate": 1.752738862389305e-05,
+      "loss": 1.2607,
+      "step": 2952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7205774359862405,
+      "learning_rate": 1.752556888463774e-05,
+      "loss": 1.1152,
+      "step": 2953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.993142013950879,
+      "learning_rate": 1.7523748570541933e-05,
+      "loss": 1.1165,
+      "step": 2954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.280982776913553,
+      "learning_rate": 1.752192768174467e-05,
+      "loss": 1.2324,
+      "step": 2955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.472416332413689,
+      "learning_rate": 1.7520106218385038e-05,
+      "loss": 1.178,
+      "step": 2956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3368535071201673,
+      "learning_rate": 1.7518284180602174e-05,
+      "loss": 1.2258,
+      "step": 2957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.505111769971719,
+      "learning_rate": 1.7516461568535254e-05,
+      "loss": 1.1499,
+      "step": 2958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2419860468462196,
+      "learning_rate": 1.7514638382323497e-05,
+      "loss": 1.2126,
+      "step": 2959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.557299051937295,
+      "learning_rate": 1.7512814622106165e-05,
+      "loss": 1.2053,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.486088568740995,
+      "learning_rate": 1.7510990288022568e-05,
+      "loss": 1.1694,
+      "step": 2961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2103733922612223,
+      "learning_rate": 1.750916538021206e-05,
+      "loss": 1.2146,
+      "step": 2962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.381691316155499,
+      "learning_rate": 1.750733989881403e-05,
+      "loss": 1.2312,
+      "step": 2963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2585009492396733,
+      "learning_rate": 1.7505513843967924e-05,
+      "loss": 1.165,
+      "step": 2964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2848282318860544,
+      "learning_rate": 1.750368721581322e-05,
+      "loss": 1.2249,
+      "step": 2965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.328779354935635,
+      "learning_rate": 1.7501860014489454e-05,
+      "loss": 1.175,
+      "step": 2966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.251080876582183,
+      "learning_rate": 1.7500032240136188e-05,
+      "loss": 1.1416,
+      "step": 2967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8881903699084366,
+      "learning_rate": 1.749820389289304e-05,
+      "loss": 1.23,
+      "step": 2968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6208682712755365,
+      "learning_rate": 1.749637497289966e-05,
+      "loss": 1.1914,
+      "step": 2969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3243481654841496,
+      "learning_rate": 1.7494545480295765e-05,
+      "loss": 1.2637,
+      "step": 2970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2741687459775033,
+      "learning_rate": 1.7492715415221087e-05,
+      "loss": 1.1711,
+      "step": 2971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.925232712738419,
+      "learning_rate": 1.7490884777815426e-05,
+      "loss": 1.2092,
+      "step": 2972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.631001241927846,
+      "learning_rate": 1.748905356821861e-05,
+      "loss": 1.2046,
+      "step": 2973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8284483826476583,
+      "learning_rate": 1.7487221786570517e-05,
+      "loss": 1.1885,
+      "step": 2974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.890199954769756,
+      "learning_rate": 1.7485389433011066e-05,
+      "loss": 1.1226,
+      "step": 2975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.287151227181428,
+      "learning_rate": 1.748355650768022e-05,
+      "loss": 1.2019,
+      "step": 2976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.21570075309808,
+      "learning_rate": 1.7481723010717994e-05,
+      "loss": 1.2161,
+      "step": 2977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0137261429373536,
+      "learning_rate": 1.7479888942264434e-05,
+      "loss": 1.2163,
+      "step": 2978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.387667685247897,
+      "learning_rate": 1.7478054302459635e-05,
+      "loss": 1.2529,
+      "step": 2979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6099888898269548,
+      "learning_rate": 1.7476219091443738e-05,
+      "loss": 1.2607,
+      "step": 2980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1329957344807373,
+      "learning_rate": 1.7474383309356927e-05,
+      "loss": 1.2014,
+      "step": 2981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4228784817903652,
+      "learning_rate": 1.7472546956339424e-05,
+      "loss": 1.2107,
+      "step": 2982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3197558953264705,
+      "learning_rate": 1.7470710032531502e-05,
+      "loss": 1.2012,
+      "step": 2983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1816328325521583,
+      "learning_rate": 1.7468872538073473e-05,
+      "loss": 1.1743,
+      "step": 2984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5908204700624515,
+      "learning_rate": 1.7467034473105696e-05,
+      "loss": 1.176,
+      "step": 2985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4201663337433432,
+      "learning_rate": 1.7465195837768568e-05,
+      "loss": 1.1929,
+      "step": 2986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7103624349736015,
+      "learning_rate": 1.7463356632202538e-05,
+      "loss": 1.2017,
+      "step": 2987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.686141872490881,
+      "learning_rate": 1.7461516856548088e-05,
+      "loss": 1.1931,
+      "step": 2988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.32742536199632,
+      "learning_rate": 1.745967651094575e-05,
+      "loss": 1.1689,
+      "step": 2989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7210171405565533,
+      "learning_rate": 1.7457835595536105e-05,
+      "loss": 1.187,
+      "step": 2990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.299645188318906,
+      "learning_rate": 1.7455994110459768e-05,
+      "loss": 1.2236,
+      "step": 2991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8447013503035974,
+      "learning_rate": 1.7454152055857396e-05,
+      "loss": 1.1545,
+      "step": 2992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.971642284537138,
+      "learning_rate": 1.74523094318697e-05,
+      "loss": 1.2554,
+      "step": 2993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3978350628801266,
+      "learning_rate": 1.745046623863743e-05,
+      "loss": 1.186,
+      "step": 2994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.138509268500947,
+      "learning_rate": 1.744862247630137e-05,
+      "loss": 1.2048,
+      "step": 2995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0132662984785172,
+      "learning_rate": 1.7446778145002364e-05,
+      "loss": 1.1538,
+      "step": 2996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3850324394052724,
+      "learning_rate": 1.7444933244881287e-05,
+      "loss": 1.1904,
+      "step": 2997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.052929352078856,
+      "learning_rate": 1.7443087776079068e-05,
+      "loss": 1.166,
+      "step": 2998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.009944695852237,
+      "learning_rate": 1.744124173873666e-05,
+      "loss": 1.1497,
+      "step": 2999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.250804366870931,
+      "learning_rate": 1.743939513299509e-05,
+      "loss": 1.2144,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.218001283618318,
+      "learning_rate": 1.7437547958995394e-05,
+      "loss": 1.1223,
+      "step": 3001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8298541864670077,
+      "learning_rate": 1.743570021687868e-05,
+      "loss": 1.2002,
+      "step": 3002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.395601502372016,
+      "learning_rate": 1.7433851906786082e-05,
+      "loss": 1.1956,
+      "step": 3003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2288348084294087,
+      "learning_rate": 1.7432003028858785e-05,
+      "loss": 1.2061,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0829716697836327,
+      "learning_rate": 1.743015358323802e-05,
+      "loss": 1.2844,
+      "step": 3005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1742200200143817,
+      "learning_rate": 1.7428303570065048e-05,
+      "loss": 1.2314,
+      "step": 3006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.002654276212863,
+      "learning_rate": 1.7426452989481187e-05,
+      "loss": 1.0852,
+      "step": 3007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.430318954934107,
+      "learning_rate": 1.742460184162779e-05,
+      "loss": 1.1646,
+      "step": 3008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.11128188207832,
+      "learning_rate": 1.7422750126646268e-05,
+      "loss": 1.1792,
+      "step": 3009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0849161443691635,
+      "learning_rate": 1.742089784467805e-05,
+      "loss": 1.1521,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3463337971773033,
+      "learning_rate": 1.7419044995864633e-05,
+      "loss": 1.2729,
+      "step": 3011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7578619893093963,
+      "learning_rate": 1.741719158034754e-05,
+      "loss": 1.2051,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.789143897932312,
+      "learning_rate": 1.7415337598268346e-05,
+      "loss": 1.2544,
+      "step": 3013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.335573112717642,
+      "learning_rate": 1.741348304976867e-05,
+      "loss": 1.1641,
+      "step": 3014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.96632400759864,
+      "learning_rate": 1.7411627934990168e-05,
+      "loss": 1.25,
+      "step": 3015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.489657808571903,
+      "learning_rate": 1.740977225407455e-05,
+      "loss": 1.2188,
+      "step": 3016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.390063545536771,
+      "learning_rate": 1.740791600716355e-05,
+      "loss": 1.1892,
+      "step": 3017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.44943668510229,
+      "learning_rate": 1.740605919439897e-05,
+      "loss": 1.1875,
+      "step": 3018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.261560012472501,
+      "learning_rate": 1.7404201815922635e-05,
+      "loss": 1.21,
+      "step": 3019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3557861436693717,
+      "learning_rate": 1.7402343871876422e-05,
+      "loss": 1.3049,
+      "step": 3020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9881525890851606,
+      "learning_rate": 1.740048536240225e-05,
+      "loss": 1.126,
+      "step": 3021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3248451769775227,
+      "learning_rate": 1.7398626287642083e-05,
+      "loss": 1.1843,
+      "step": 3022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1436366775357922,
+      "learning_rate": 1.739676664773793e-05,
+      "loss": 1.2046,
+      "step": 3023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.161869074585195,
+      "learning_rate": 1.7394906442831828e-05,
+      "loss": 1.1499,
+      "step": 3024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.017270158917334,
+      "learning_rate": 1.739304567306588e-05,
+      "loss": 1.209,
+      "step": 3025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6052938033463384,
+      "learning_rate": 1.739118433858222e-05,
+      "loss": 1.2378,
+      "step": 3026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5399049327038106,
+      "learning_rate": 1.7389322439523014e-05,
+      "loss": 1.1855,
+      "step": 3027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5185972950176936,
+      "learning_rate": 1.73874599760305e-05,
+      "loss": 1.1782,
+      "step": 3028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5172034351452024,
+      "learning_rate": 1.738559694824693e-05,
+      "loss": 1.2354,
+      "step": 3029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8965317554748204,
+      "learning_rate": 1.7383733356314612e-05,
+      "loss": 1.2051,
+      "step": 3030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.012569508638537,
+      "learning_rate": 1.738186920037591e-05,
+      "loss": 1.1665,
+      "step": 3031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.190982025688499,
+      "learning_rate": 1.73800044805732e-05,
+      "loss": 1.1211,
+      "step": 3032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.101396843312381,
+      "learning_rate": 1.7378139197048926e-05,
+      "loss": 1.2327,
+      "step": 3033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5408814094317544,
+      "learning_rate": 1.7376273349945568e-05,
+      "loss": 1.1633,
+      "step": 3034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.168810802131502,
+      "learning_rate": 1.737440693940565e-05,
+      "loss": 1.1489,
+      "step": 3035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6544246082328833,
+      "learning_rate": 1.737253996557174e-05,
+      "loss": 1.146,
+      "step": 3036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9008206450090923,
+      "learning_rate": 1.7370672428586436e-05,
+      "loss": 1.2344,
+      "step": 3037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.948686417645651,
+      "learning_rate": 1.7368804328592398e-05,
+      "loss": 1.1931,
+      "step": 3038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2489333017009887,
+      "learning_rate": 1.736693566573232e-05,
+      "loss": 1.2268,
+      "step": 3039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8988630903646553,
+      "learning_rate": 1.7365066440148942e-05,
+      "loss": 1.1738,
+      "step": 3040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6254246365133787,
+      "learning_rate": 1.736319665198504e-05,
+      "loss": 1.2246,
+      "step": 3041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9018479894824876,
+      "learning_rate": 1.736132630138344e-05,
+      "loss": 1.2229,
+      "step": 3042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2155409619772306,
+      "learning_rate": 1.7359455388487008e-05,
+      "loss": 1.1046,
+      "step": 3043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.879426434608461,
+      "learning_rate": 1.7357583913438655e-05,
+      "loss": 1.2197,
+      "step": 3044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4275122578702217,
+      "learning_rate": 1.7355711876381335e-05,
+      "loss": 1.1589,
+      "step": 3045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4097397878092734,
+      "learning_rate": 1.735383927745804e-05,
+      "loss": 1.2458,
+      "step": 3046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1368057143464787,
+      "learning_rate": 1.735196611681181e-05,
+      "loss": 1.1692,
+      "step": 3047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.176708642410201,
+      "learning_rate": 1.7350092394585724e-05,
+      "loss": 1.2202,
+      "step": 3048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.626110977359878,
+      "learning_rate": 1.734821811092291e-05,
+      "loss": 1.2593,
+      "step": 3049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1306038986229296,
+      "learning_rate": 1.734634326596653e-05,
+      "loss": 1.2607,
+      "step": 3050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1459847274946764,
+      "learning_rate": 1.7344467859859807e-05,
+      "loss": 1.2214,
+      "step": 3051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.480677539945333,
+      "learning_rate": 1.7342591892745978e-05,
+      "loss": 1.3013,
+      "step": 3052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7794539151054014,
+      "learning_rate": 1.7340715364768345e-05,
+      "loss": 1.1609,
+      "step": 3053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4239042572717375,
+      "learning_rate": 1.7338838276070248e-05,
+      "loss": 1.1929,
+      "step": 3054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2687051430429306,
+      "learning_rate": 1.7336960626795064e-05,
+      "loss": 1.2322,
+      "step": 3055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.362289757948876,
+      "learning_rate": 1.7335082417086223e-05,
+      "loss": 1.2285,
+      "step": 3056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.929614725480608,
+      "learning_rate": 1.7333203647087188e-05,
+      "loss": 1.2205,
+      "step": 3057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.630389245455423,
+      "learning_rate": 1.7331324316941474e-05,
+      "loss": 1.1838,
+      "step": 3058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.805226253207587,
+      "learning_rate": 1.732944442679263e-05,
+      "loss": 1.2646,
+      "step": 3059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.612706462854552,
+      "learning_rate": 1.7327563976784245e-05,
+      "loss": 1.2393,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6352606690691673,
+      "learning_rate": 1.732568296705997e-05,
+      "loss": 1.1714,
+      "step": 3061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1958216936783788,
+      "learning_rate": 1.7323801397763475e-05,
+      "loss": 1.146,
+      "step": 3062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.306928194379734,
+      "learning_rate": 1.7321919269038487e-05,
+      "loss": 1.1772,
+      "step": 3063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8301373744406866,
+      "learning_rate": 1.7320036581028778e-05,
+      "loss": 1.1946,
+      "step": 3064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3870326415624117,
+      "learning_rate": 1.731815333387815e-05,
+      "loss": 1.2378,
+      "step": 3065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.54554609143972,
+      "learning_rate": 1.7316269527730455e-05,
+      "loss": 1.1851,
+      "step": 3066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2232695527687136,
+      "learning_rate": 1.7314385162729595e-05,
+      "loss": 1.1897,
+      "step": 3067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.245073058453666,
+      "learning_rate": 1.7312500239019495e-05,
+      "loss": 1.2241,
+      "step": 3068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0647274065385455,
+      "learning_rate": 1.7310614756744146e-05,
+      "loss": 1.2393,
+      "step": 3069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.26639695303647,
+      "learning_rate": 1.7308728716047568e-05,
+      "loss": 1.1702,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7059345130384047,
+      "learning_rate": 1.730684211707382e-05,
+      "loss": 1.2808,
+      "step": 3071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1334335707855234,
+      "learning_rate": 1.730495495996702e-05,
+      "loss": 1.2593,
+      "step": 3072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8996836921118248,
+      "learning_rate": 1.730306724487131e-05,
+      "loss": 1.2842,
+      "step": 3073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2090709045408565,
+      "learning_rate": 1.7301178971930886e-05,
+      "loss": 1.2461,
+      "step": 3074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.816298209103785,
+      "learning_rate": 1.7299290141289983e-05,
+      "loss": 1.2217,
+      "step": 3075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2302343176510675,
+      "learning_rate": 1.729740075309288e-05,
+      "loss": 1.2268,
+      "step": 3076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.182603072406039,
+      "learning_rate": 1.72955108074839e-05,
+      "loss": 1.252,
+      "step": 3077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4056030542748537,
+      "learning_rate": 1.7293620304607405e-05,
+      "loss": 1.0996,
+      "step": 3078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0782563065343673,
+      "learning_rate": 1.72917292446078e-05,
+      "loss": 1.2393,
+      "step": 3079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1462734450364054,
+      "learning_rate": 1.728983762762953e-05,
+      "loss": 1.1838,
+      "step": 3080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9701767282017317,
+      "learning_rate": 1.728794545381709e-05,
+      "loss": 1.1003,
+      "step": 3081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.261995596543632,
+      "learning_rate": 1.728605272331502e-05,
+      "loss": 1.1398,
+      "step": 3082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.251047432061986,
+      "learning_rate": 1.7284159436267885e-05,
+      "loss": 1.2881,
+      "step": 3083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.096935145891595,
+      "learning_rate": 1.7282265592820308e-05,
+      "loss": 1.1943,
+      "step": 3084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8346025194031768,
+      "learning_rate": 1.7280371193116955e-05,
+      "loss": 1.2427,
+      "step": 3085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.156602578664964,
+      "learning_rate": 1.7278476237302523e-05,
+      "loss": 1.1772,
+      "step": 3086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7932385065947862,
+      "learning_rate": 1.7276580725521763e-05,
+      "loss": 1.1028,
+      "step": 3087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.504871025873424,
+      "learning_rate": 1.727468465791946e-05,
+      "loss": 1.228,
+      "step": 3088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1993509301394076,
+      "learning_rate": 1.7272788034640448e-05,
+      "loss": 1.1106,
+      "step": 3089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.264555050966,
+      "learning_rate": 1.7270890855829595e-05,
+      "loss": 1.2017,
+      "step": 3090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.084668573422382,
+      "learning_rate": 1.7268993121631825e-05,
+      "loss": 1.2305,
+      "step": 3091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9820395978725696,
+      "learning_rate": 1.7267094832192093e-05,
+      "loss": 1.2244,
+      "step": 3092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.980693736844743,
+      "learning_rate": 1.7265195987655397e-05,
+      "loss": 1.155,
+      "step": 3093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1169270013391226,
+      "learning_rate": 1.7263296588166784e-05,
+      "loss": 1.1169,
+      "step": 3094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.277848560118828,
+      "learning_rate": 1.7261396633871336e-05,
+      "loss": 1.1995,
+      "step": 3095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.026637234538083,
+      "learning_rate": 1.7259496124914187e-05,
+      "loss": 1.2539,
+      "step": 3096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1277992483054695,
+      "learning_rate": 1.7257595061440502e-05,
+      "loss": 1.1697,
+      "step": 3097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.598410599523532,
+      "learning_rate": 1.7255693443595494e-05,
+      "loss": 1.248,
+      "step": 3098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0678551675145425,
+      "learning_rate": 1.725379127152442e-05,
+      "loss": 1.2378,
+      "step": 3099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1134544722081685,
+      "learning_rate": 1.725188854537258e-05,
+      "loss": 1.1545,
+      "step": 3100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2491647156744095,
+      "learning_rate": 1.7249985265285306e-05,
+      "loss": 1.2405,
+      "step": 3101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9000590064179876,
+      "learning_rate": 1.7248081431407985e-05,
+      "loss": 1.2427,
+      "step": 3102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.655255522023911,
+      "learning_rate": 1.724617704388604e-05,
+      "loss": 1.2461,
+      "step": 3103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.727300582448234,
+      "learning_rate": 1.7244272102864942e-05,
+      "loss": 1.1853,
+      "step": 3104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.171489068010337,
+      "learning_rate": 1.724236660849019e-05,
+      "loss": 1.2578,
+      "step": 3105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6291865361297475,
+      "learning_rate": 1.7240460560907345e-05,
+      "loss": 1.2039,
+      "step": 3106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2059898831792353,
+      "learning_rate": 1.7238553960262e-05,
+      "loss": 1.2466,
+      "step": 3107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6697608056594366,
+      "learning_rate": 1.7236646806699782e-05,
+      "loss": 1.1719,
+      "step": 3108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.681455316916694,
+      "learning_rate": 1.723473910036638e-05,
+      "loss": 1.145,
+      "step": 3109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2267284548351998,
+      "learning_rate": 1.7232830841407504e-05,
+      "loss": 1.2092,
+      "step": 3110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4214676670756834,
+      "learning_rate": 1.7230922029968925e-05,
+      "loss": 1.1812,
+      "step": 3111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.513546131135956,
+      "learning_rate": 1.722901266619644e-05,
+      "loss": 1.2676,
+      "step": 3112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5467442221100565,
+      "learning_rate": 1.72271027502359e-05,
+      "loss": 1.2463,
+      "step": 3113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0751655314600383,
+      "learning_rate": 1.7225192282233195e-05,
+      "loss": 1.2886,
+      "step": 3114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3404329639046932,
+      "learning_rate": 1.7223281262334252e-05,
+      "loss": 1.2502,
+      "step": 3115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9578852192034946,
+      "learning_rate": 1.7221369690685047e-05,
+      "loss": 1.1313,
+      "step": 3116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.856736443917743,
+      "learning_rate": 1.7219457567431595e-05,
+      "loss": 1.166,
+      "step": 3117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4579895614793075,
+      "learning_rate": 1.7217544892719952e-05,
+      "loss": 1.1521,
+      "step": 3118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.219820478236479,
+      "learning_rate": 1.721563166669622e-05,
+      "loss": 1.1802,
+      "step": 3119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3953287751751984,
+      "learning_rate": 1.721371788950654e-05,
+      "loss": 1.2612,
+      "step": 3120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0628564443786406,
+      "learning_rate": 1.7211803561297097e-05,
+      "loss": 1.1758,
+      "step": 3121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0023009074576725,
+      "learning_rate": 1.7209888682214115e-05,
+      "loss": 1.1423,
+      "step": 3122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.420854971672432,
+      "learning_rate": 1.720797325240386e-05,
+      "loss": 1.2771,
+      "step": 3123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.032196305994755,
+      "learning_rate": 1.7206057272012647e-05,
+      "loss": 1.2422,
+      "step": 3124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.993490064690277,
+      "learning_rate": 1.7204140741186825e-05,
+      "loss": 1.2256,
+      "step": 3125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.898952160336636,
+      "learning_rate": 1.720222366007279e-05,
+      "loss": 1.1677,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0687465821026687,
+      "learning_rate": 1.7200306028816974e-05,
+      "loss": 1.2637,
+      "step": 3127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.268916189001657,
+      "learning_rate": 1.7198387847565862e-05,
+      "loss": 1.2729,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.868216300193663,
+      "learning_rate": 1.719646911646597e-05,
+      "loss": 1.292,
+      "step": 3129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.16306544064944,
+      "learning_rate": 1.7194549835663858e-05,
+      "loss": 1.2009,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.277664410854016,
+      "learning_rate": 1.7192630005306134e-05,
+      "loss": 1.2878,
+      "step": 3131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7601541688067948,
+      "learning_rate": 1.7190709625539445e-05,
+      "loss": 1.2107,
+      "step": 3132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1472055212891155,
+      "learning_rate": 1.7188788696510477e-05,
+      "loss": 1.1292,
+      "step": 3133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5663815903698963,
+      "learning_rate": 1.718686721836596e-05,
+      "loss": 1.219,
+      "step": 3134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8679174514586365,
+      "learning_rate": 1.7184945191252665e-05,
+      "loss": 1.2515,
+      "step": 3135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9590246481090445,
+      "learning_rate": 1.7183022615317413e-05,
+      "loss": 1.1489,
+      "step": 3136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.812063738843136,
+      "learning_rate": 1.7181099490707055e-05,
+      "loss": 1.2559,
+      "step": 3137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.103113503839388,
+      "learning_rate": 1.7179175817568484e-05,
+      "loss": 1.2053,
+      "step": 3138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6379831308705692,
+      "learning_rate": 1.7177251596048647e-05,
+      "loss": 1.1877,
+      "step": 3139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0137499590861454,
+      "learning_rate": 1.717532682629452e-05,
+      "loss": 1.207,
+      "step": 3140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.064217269271703,
+      "learning_rate": 1.7173401508453133e-05,
+      "loss": 1.147,
+      "step": 3141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3392334065256906,
+      "learning_rate": 1.7171475642671546e-05,
+      "loss": 1.1816,
+      "step": 3142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4841169279235773,
+      "learning_rate": 1.7169549229096865e-05,
+      "loss": 1.1094,
+      "step": 3143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9079983905805626,
+      "learning_rate": 1.7167622267876248e-05,
+      "loss": 1.1594,
+      "step": 3144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8100175788265807,
+      "learning_rate": 1.7165694759156878e-05,
+      "loss": 1.2112,
+      "step": 3145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.088949687126392,
+      "learning_rate": 1.716376670308599e-05,
+      "loss": 1.2163,
+      "step": 3146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0431501262724683,
+      "learning_rate": 1.7161838099810855e-05,
+      "loss": 1.1174,
+      "step": 3147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.336819841105745,
+      "learning_rate": 1.7159908949478797e-05,
+      "loss": 1.2798,
+      "step": 3148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.097736647979037,
+      "learning_rate": 1.715797925223717e-05,
+      "loss": 1.187,
+      "step": 3149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4300808275790593,
+      "learning_rate": 1.7156049008233374e-05,
+      "loss": 1.2229,
+      "step": 3150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7137940320852825,
+      "learning_rate": 1.7154118217614844e-05,
+      "loss": 1.2847,
+      "step": 3151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2481563734457612,
+      "learning_rate": 1.715218688052908e-05,
+      "loss": 1.2271,
+      "step": 3152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.416224743827951,
+      "learning_rate": 1.715025499712359e-05,
+      "loss": 1.2527,
+      "step": 3153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4077607445687135,
+      "learning_rate": 1.714832256754595e-05,
+      "loss": 1.2151,
+      "step": 3154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7831427641224957,
+      "learning_rate": 1.7146389591943763e-05,
+      "loss": 1.1926,
+      "step": 3155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4665624417034007,
+      "learning_rate": 1.714445607046469e-05,
+      "loss": 1.2612,
+      "step": 3156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.045654493037582,
+      "learning_rate": 1.714252200325641e-05,
+      "loss": 1.2554,
+      "step": 3157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.470154332472987,
+      "learning_rate": 1.7140587390466662e-05,
+      "loss": 1.2087,
+      "step": 3158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9099069043681105,
+      "learning_rate": 1.713865223224323e-05,
+      "loss": 1.158,
+      "step": 3159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1380102493442683,
+      "learning_rate": 1.7136716528733916e-05,
+      "loss": 1.2146,
+      "step": 3160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2942352173001965,
+      "learning_rate": 1.7134780280086585e-05,
+      "loss": 1.2119,
+      "step": 3161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9993099260892087,
+      "learning_rate": 1.713284348644914e-05,
+      "loss": 1.2263,
+      "step": 3162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.078742651524062,
+      "learning_rate": 1.713090614796952e-05,
+      "loss": 1.2488,
+      "step": 3163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.178492604432267,
+      "learning_rate": 1.712896826479571e-05,
+      "loss": 1.2817,
+      "step": 3164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7344784573158694,
+      "learning_rate": 1.7127029837075736e-05,
+      "loss": 1.1821,
+      "step": 3165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0733468910954187,
+      "learning_rate": 1.7125090864957664e-05,
+      "loss": 1.1467,
+      "step": 3166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0803265315704547,
+      "learning_rate": 1.71231513485896e-05,
+      "loss": 1.147,
+      "step": 3167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1521781941580054,
+      "learning_rate": 1.71212112881197e-05,
+      "loss": 1.2869,
+      "step": 3168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.632363146231039,
+      "learning_rate": 1.7119270683696144e-05,
+      "loss": 1.2063,
+      "step": 3169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.108113376853558,
+      "learning_rate": 1.7117329535467177e-05,
+      "loss": 1.2139,
+      "step": 3170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.49012111812984,
+      "learning_rate": 1.7115387843581068e-05,
+      "loss": 1.2036,
+      "step": 3171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7583178289605925,
+      "learning_rate": 1.7113445608186134e-05,
+      "loss": 1.2158,
+      "step": 3172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3324555526432826,
+      "learning_rate": 1.7111502829430734e-05,
+      "loss": 1.2056,
+      "step": 3173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1111924150232073,
+      "learning_rate": 1.7109559507463264e-05,
+      "loss": 1.1765,
+      "step": 3174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.497763938821389,
+      "learning_rate": 1.7107615642432166e-05,
+      "loss": 1.2488,
+      "step": 3175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.993024385703013,
+      "learning_rate": 1.7105671234485925e-05,
+      "loss": 1.2261,
+      "step": 3176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.88080158128215,
+      "learning_rate": 1.710372628377306e-05,
+      "loss": 1.1611,
+      "step": 3177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9657134485728305,
+      "learning_rate": 1.710178079044214e-05,
+      "loss": 1.1963,
+      "step": 3178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1783558047979477,
+      "learning_rate": 1.709983475464177e-05,
+      "loss": 1.1433,
+      "step": 3179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.119029624563908,
+      "learning_rate": 1.7097888176520597e-05,
+      "loss": 1.2544,
+      "step": 3180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0592970951346383,
+      "learning_rate": 1.7095941056227306e-05,
+      "loss": 1.1436,
+      "step": 3181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.84610867686768,
+      "learning_rate": 1.709399339391064e-05,
+      "loss": 1.2397,
+      "step": 3182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.366099844989877,
+      "learning_rate": 1.709204518971936e-05,
+      "loss": 1.1829,
+      "step": 3183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7792114778599446,
+      "learning_rate": 1.7090096443802286e-05,
+      "loss": 1.1833,
+      "step": 3184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.231671731791031,
+      "learning_rate": 1.7088147156308267e-05,
+      "loss": 1.2358,
+      "step": 3185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6128245458139356,
+      "learning_rate": 1.7086197327386204e-05,
+      "loss": 1.2686,
+      "step": 3186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.838486577493911,
+      "learning_rate": 1.7084246957185036e-05,
+      "loss": 1.2407,
+      "step": 3187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.114181361661098,
+      "learning_rate": 1.7082296045853735e-05,
+      "loss": 1.1738,
+      "step": 3188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.516128160312502,
+      "learning_rate": 1.7080344593541326e-05,
+      "loss": 1.2385,
+      "step": 3189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.556655109193774,
+      "learning_rate": 1.7078392600396876e-05,
+      "loss": 1.1682,
+      "step": 3190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.514569077863747,
+      "learning_rate": 1.707644006656948e-05,
+      "loss": 1.2537,
+      "step": 3191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.564254153935329,
+      "learning_rate": 1.707448699220828e-05,
+      "loss": 1.1196,
+      "step": 3192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.40289979281124,
+      "learning_rate": 1.707253337746247e-05,
+      "loss": 1.1814,
+      "step": 3193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.276631641573876,
+      "learning_rate": 1.7070579222481272e-05,
+      "loss": 1.2485,
+      "step": 3194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.810186714330227,
+      "learning_rate": 1.7068624527413955e-05,
+      "loss": 1.1599,
+      "step": 3195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.61486411065063,
+      "learning_rate": 1.706666929240983e-05,
+      "loss": 1.2537,
+      "step": 3196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1251350416867383,
+      "learning_rate": 1.7064713517618244e-05,
+      "loss": 1.1409,
+      "step": 3197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.0088596062415025,
+      "learning_rate": 1.7062757203188592e-05,
+      "loss": 1.1831,
+      "step": 3198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.40207433123572,
+      "learning_rate": 1.7060800349270307e-05,
+      "loss": 1.2444,
+      "step": 3199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.462401868862926,
+      "learning_rate": 1.705884295601286e-05,
+      "loss": 1.1414,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.775473080350794,
+      "learning_rate": 1.705688502356577e-05,
+      "loss": 1.2646,
+      "step": 3201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6100478018753344,
+      "learning_rate": 1.705492655207859e-05,
+      "loss": 1.2175,
+      "step": 3202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5311414403390224,
+      "learning_rate": 1.7052967541700925e-05,
+      "loss": 1.1709,
+      "step": 3203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.051032917363089,
+      "learning_rate": 1.7051007992582406e-05,
+      "loss": 1.1962,
+      "step": 3204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8001139319382102,
+      "learning_rate": 1.7049047904872715e-05,
+      "loss": 1.2476,
+      "step": 3205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.533186436079632,
+      "learning_rate": 1.7047087278721578e-05,
+      "loss": 1.1973,
+      "step": 3206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.010642857037919,
+      "learning_rate": 1.7045126114278753e-05,
+      "loss": 1.0891,
+      "step": 3207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4372945856903656,
+      "learning_rate": 1.7043164411694043e-05,
+      "loss": 1.1765,
+      "step": 3208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9470122986643745,
+      "learning_rate": 1.7041202171117295e-05,
+      "loss": 1.2307,
+      "step": 3209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.703664739533799,
+      "learning_rate": 1.7039239392698394e-05,
+      "loss": 1.2441,
+      "step": 3210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.156217754017528,
+      "learning_rate": 1.7037276076587267e-05,
+      "loss": 1.2351,
+      "step": 3211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7799712806570476,
+      "learning_rate": 1.7035312222933883e-05,
+      "loss": 1.2183,
+      "step": 3212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.278508116331607,
+      "learning_rate": 1.7033347831888246e-05,
+      "loss": 1.2793,
+      "step": 3213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8042145122866273,
+      "learning_rate": 1.703138290360041e-05,
+      "loss": 1.2031,
+      "step": 3214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.046464709478787,
+      "learning_rate": 1.7029417438220467e-05,
+      "loss": 1.1157,
+      "step": 3215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1810978718658833,
+      "learning_rate": 1.702745143589855e-05,
+      "loss": 1.176,
+      "step": 3216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.967233850497897,
+      "learning_rate": 1.7025484896784824e-05,
+      "loss": 1.1912,
+      "step": 3217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5680916855563867,
+      "learning_rate": 1.702351782102951e-05,
+      "loss": 1.175,
+      "step": 3218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.316927743883325,
+      "learning_rate": 1.7021550208782867e-05,
+      "loss": 1.2244,
+      "step": 3219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1668838962681147,
+      "learning_rate": 1.701958206019518e-05,
+      "loss": 1.1697,
+      "step": 3220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.636088089443268,
+      "learning_rate": 1.7017613375416796e-05,
+      "loss": 1.1616,
+      "step": 3221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0261965796956822,
+      "learning_rate": 1.7015644154598086e-05,
+      "loss": 1.1504,
+      "step": 3222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2872005271229643,
+      "learning_rate": 1.701367439788947e-05,
+      "loss": 1.23,
+      "step": 3223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2762129477513167,
+      "learning_rate": 1.7011704105441412e-05,
+      "loss": 1.2009,
+      "step": 3224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8058508787243355,
+      "learning_rate": 1.700973327740441e-05,
+      "loss": 1.2437,
+      "step": 3225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.098938305339955,
+      "learning_rate": 1.7007761913929007e-05,
+      "loss": 1.2563,
+      "step": 3226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.558388908784276,
+      "learning_rate": 1.7005790015165784e-05,
+      "loss": 1.2107,
+      "step": 3227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.356853396272706,
+      "learning_rate": 1.7003817581265363e-05,
+      "loss": 1.1816,
+      "step": 3228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8634607971422987,
+      "learning_rate": 1.7001844612378415e-05,
+      "loss": 1.1628,
+      "step": 3229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8710914564772287,
+      "learning_rate": 1.6999871108655635e-05,
+      "loss": 1.1711,
+      "step": 3230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8722976363889874,
+      "learning_rate": 1.699789707024778e-05,
+      "loss": 1.1838,
+      "step": 3231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8935758612131957,
+      "learning_rate": 1.699592249730563e-05,
+      "loss": 1.1953,
+      "step": 3232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4789155914072856,
+      "learning_rate": 1.699394738998001e-05,
+      "loss": 1.2803,
+      "step": 3233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1781660419373363,
+      "learning_rate": 1.69919717484218e-05,
+      "loss": 1.2112,
+      "step": 3234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.541836184822763,
+      "learning_rate": 1.6989995572781898e-05,
+      "loss": 1.2761,
+      "step": 3235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1767939626487918,
+      "learning_rate": 1.6988018863211256e-05,
+      "loss": 1.2041,
+      "step": 3236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.1822144951109825,
+      "learning_rate": 1.6986041619860868e-05,
+      "loss": 1.2249,
+      "step": 3237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.269434627805233,
+      "learning_rate": 1.698406384288177e-05,
+      "loss": 1.1963,
+      "step": 3238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.2264735543909975,
+      "learning_rate": 1.698208553242503e-05,
+      "loss": 1.2046,
+      "step": 3239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8638106691961904,
+      "learning_rate": 1.6980106688641754e-05,
+      "loss": 1.2366,
+      "step": 3240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.712161926874804,
+      "learning_rate": 1.6978127311683103e-05,
+      "loss": 1.2634,
+      "step": 3241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9019921650844385,
+      "learning_rate": 1.6976147401700276e-05,
+      "loss": 1.2195,
+      "step": 3242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.481969641743757,
+      "learning_rate": 1.6974166958844503e-05,
+      "loss": 1.2532,
+      "step": 3243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.757972609535146,
+      "learning_rate": 1.697218598326706e-05,
+      "loss": 1.1279,
+      "step": 3244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.599814313155729,
+      "learning_rate": 1.6970204475119262e-05,
+      "loss": 1.252,
+      "step": 3245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5822760853584414,
+      "learning_rate": 1.6968222434552474e-05,
+      "loss": 1.1653,
+      "step": 3246
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.94948957732167,
+      "learning_rate": 1.6966239861718083e-05,
+      "loss": 1.2385,
+      "step": 3247
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.933538106036317,
+      "learning_rate": 1.696425675676754e-05,
+      "loss": 1.1689,
+      "step": 3248
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6686841220907587,
+      "learning_rate": 1.6962273119852314e-05,
+      "loss": 1.2085,
+      "step": 3249
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1580617042294596,
+      "learning_rate": 1.6960288951123933e-05,
+      "loss": 1.1411,
+      "step": 3250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5891670854307653,
+      "learning_rate": 1.6958304250733954e-05,
+      "loss": 1.2261,
+      "step": 3251
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.853200395193363,
+      "learning_rate": 1.6956319018833975e-05,
+      "loss": 1.2773,
+      "step": 3252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9896678990428036,
+      "learning_rate": 1.6954333255575646e-05,
+      "loss": 1.2649,
+      "step": 3253
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0788922702274224,
+      "learning_rate": 1.6952346961110643e-05,
+      "loss": 1.2061,
+      "step": 3254
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7773761299149853,
+      "learning_rate": 1.695036013559069e-05,
+      "loss": 1.2134,
+      "step": 3255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4409026802895846,
+      "learning_rate": 1.6948372779167558e-05,
+      "loss": 1.3149,
+      "step": 3256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3193224712770926,
+      "learning_rate": 1.6946384891993036e-05,
+      "loss": 1.1741,
+      "step": 3257
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.849814783153425,
+      "learning_rate": 1.6944396474218984e-05,
+      "loss": 1.2808,
+      "step": 3258
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.063882653399251,
+      "learning_rate": 1.6942407525997277e-05,
+      "loss": 1.2048,
+      "step": 3259
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.863269445491256,
+      "learning_rate": 1.694041804747985e-05,
+      "loss": 1.1587,
+      "step": 3260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9597241562768586,
+      "learning_rate": 1.693842803881866e-05,
+      "loss": 1.2441,
+      "step": 3261
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.556969356951435,
+      "learning_rate": 1.693643750016572e-05,
+      "loss": 1.3188,
+      "step": 3262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2651469227782703,
+      "learning_rate": 1.6934446431673073e-05,
+      "loss": 1.2446,
+      "step": 3263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3387798799015584,
+      "learning_rate": 1.693245483349281e-05,
+      "loss": 1.1318,
+      "step": 3264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.292493844984531,
+      "learning_rate": 1.6930462705777057e-05,
+      "loss": 1.2427,
+      "step": 3265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.607655922604369,
+      "learning_rate": 1.6928470048677984e-05,
+      "loss": 1.2454,
+      "step": 3266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.119632276606223,
+      "learning_rate": 1.6926476862347803e-05,
+      "loss": 1.2273,
+      "step": 3267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.64882122422409,
+      "learning_rate": 1.6924483146938756e-05,
+      "loss": 1.2068,
+      "step": 3268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2804356851737446,
+      "learning_rate": 1.6922488902603143e-05,
+      "loss": 1.166,
+      "step": 3269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8010147512824703,
+      "learning_rate": 1.692049412949328e-05,
+      "loss": 1.1892,
+      "step": 3270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1024531789917305,
+      "learning_rate": 1.6918498827761554e-05,
+      "loss": 1.1821,
+      "step": 3271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5911466630725397,
+      "learning_rate": 1.6916502997560367e-05,
+      "loss": 1.1956,
+      "step": 3272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6133950078300368,
+      "learning_rate": 1.691450663904217e-05,
+      "loss": 1.21,
+      "step": 3273
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2571306514223495,
+      "learning_rate": 1.6912509752359457e-05,
+      "loss": 1.2168,
+      "step": 3274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9498140646079443,
+      "learning_rate": 1.691051233766476e-05,
+      "loss": 1.1189,
+      "step": 3275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4426362239279165,
+      "learning_rate": 1.6908514395110652e-05,
+      "loss": 1.2061,
+      "step": 3276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.409283790894793,
+      "learning_rate": 1.6906515924849745e-05,
+      "loss": 1.1741,
+      "step": 3277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7731695221931965,
+      "learning_rate": 1.6904516927034692e-05,
+      "loss": 1.2,
+      "step": 3278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.213857931542782,
+      "learning_rate": 1.6902517401818186e-05,
+      "loss": 1.2163,
+      "step": 3279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.572312213958063,
+      "learning_rate": 1.6900517349352964e-05,
+      "loss": 1.1938,
+      "step": 3280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0450784983674968,
+      "learning_rate": 1.6898516769791794e-05,
+      "loss": 1.1675,
+      "step": 3281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1852735839840816,
+      "learning_rate": 1.6896515663287497e-05,
+      "loss": 1.2319,
+      "step": 3282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9193623456413236,
+      "learning_rate": 1.6894514029992922e-05,
+      "loss": 1.1792,
+      "step": 3283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2502754037712127,
+      "learning_rate": 1.689251187006097e-05,
+      "loss": 1.2568,
+      "step": 3284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.763680008751771,
+      "learning_rate": 1.6890509183644568e-05,
+      "loss": 1.2229,
+      "step": 3285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.016025728446679,
+      "learning_rate": 1.6888505970896696e-05,
+      "loss": 1.2424,
+      "step": 3286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2385604772381757,
+      "learning_rate": 1.6886502231970368e-05,
+      "loss": 1.1907,
+      "step": 3287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0510118915667963,
+      "learning_rate": 1.688449796701864e-05,
+      "loss": 1.1726,
+      "step": 3288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2219502950258954,
+      "learning_rate": 1.688249317619461e-05,
+      "loss": 1.2192,
+      "step": 3289
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2114393221787334,
+      "learning_rate": 1.688048785965141e-05,
+      "loss": 1.2075,
+      "step": 3290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.839354802817925,
+      "learning_rate": 1.6878482017542217e-05,
+      "loss": 1.2339,
+      "step": 3291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.877117384114171,
+      "learning_rate": 1.687647565002025e-05,
+      "loss": 1.2363,
+      "step": 3292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.827355161744899,
+      "learning_rate": 1.6874468757238763e-05,
+      "loss": 1.2314,
+      "step": 3293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2689091657900007,
+      "learning_rate": 1.6872461339351053e-05,
+      "loss": 1.2,
+      "step": 3294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2331752972294217,
+      "learning_rate": 1.6870453396510456e-05,
+      "loss": 1.2229,
+      "step": 3295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9615157438421162,
+      "learning_rate": 1.6868444928870346e-05,
+      "loss": 1.1921,
+      "step": 3296
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9163338937196244,
+      "learning_rate": 1.6866435936584148e-05,
+      "loss": 1.1047,
+      "step": 3297
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1748818306305555,
+      "learning_rate": 1.6864426419805312e-05,
+      "loss": 1.1841,
+      "step": 3298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9012038297416285,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 1.219,
+      "step": 3299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.651450853289605,
+      "learning_rate": 1.686040581338376e-05,
+      "loss": 1.1919,
+      "step": 3300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7443946083645034,
+      "learning_rate": 1.6858394724048156e-05,
+      "loss": 1.1609,
+      "step": 3301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0246409533405454,
+      "learning_rate": 1.6856383110834146e-05,
+      "loss": 1.1958,
+      "step": 3302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.358914124401566,
+      "learning_rate": 1.6854370973895383e-05,
+      "loss": 1.2219,
+      "step": 3303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9968079015170996,
+      "learning_rate": 1.6852358313385568e-05,
+      "loss": 1.2024,
+      "step": 3304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.540256056441182,
+      "learning_rate": 1.6850345129458435e-05,
+      "loss": 1.1343,
+      "step": 3305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1924257870684634,
+      "learning_rate": 1.684833142226776e-05,
+      "loss": 1.2046,
+      "step": 3306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.14765104553529,
+      "learning_rate": 1.6846317191967364e-05,
+      "loss": 1.1382,
+      "step": 3307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.736911642287484,
+      "learning_rate": 1.6844302438711106e-05,
+      "loss": 1.147,
+      "step": 3308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.440323249113541,
+      "learning_rate": 1.684228716265287e-05,
+      "loss": 1.2346,
+      "step": 3309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.508434543435511,
+      "learning_rate": 1.6840271363946608e-05,
+      "loss": 1.2239,
+      "step": 3310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.735815718304622,
+      "learning_rate": 1.683825504274629e-05,
+      "loss": 1.1516,
+      "step": 3311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0702468393975484,
+      "learning_rate": 1.683623819920593e-05,
+      "loss": 1.2102,
+      "step": 3312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.920605675263037,
+      "learning_rate": 1.683422083347959e-05,
+      "loss": 1.1855,
+      "step": 3313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3942785494118257,
+      "learning_rate": 1.6832202945721362e-05,
+      "loss": 1.1526,
+      "step": 3314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.168455792604799,
+      "learning_rate": 1.6830184536085387e-05,
+      "loss": 1.2227,
+      "step": 3315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.692907026908719,
+      "learning_rate": 1.682816560472584e-05,
+      "loss": 1.2446,
+      "step": 3316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0335270784804296,
+      "learning_rate": 1.682614615179693e-05,
+      "loss": 1.2144,
+      "step": 3317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4616042554048008,
+      "learning_rate": 1.6824126177452923e-05,
+      "loss": 1.2,
+      "step": 3318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.265974169583133,
+      "learning_rate": 1.6822105681848112e-05,
+      "loss": 1.1948,
+      "step": 3319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.313729287658243,
+      "learning_rate": 1.6820084665136828e-05,
+      "loss": 1.1858,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7068448795564786,
+      "learning_rate": 1.6818063127473447e-05,
+      "loss": 1.1685,
+      "step": 3321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2742034950779333,
+      "learning_rate": 1.681604106901239e-05,
+      "loss": 1.2058,
+      "step": 3322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0485617711747284,
+      "learning_rate": 1.681401848990811e-05,
+      "loss": 1.2393,
+      "step": 3323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.99552406054297,
+      "learning_rate": 1.6811995390315093e-05,
+      "loss": 1.1792,
+      "step": 3324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7657576076422843,
+      "learning_rate": 1.6809971770387888e-05,
+      "loss": 1.1326,
+      "step": 3325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.697305053152081,
+      "learning_rate": 1.6807947630281056e-05,
+      "loss": 1.2129,
+      "step": 3326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9757956221082713,
+      "learning_rate": 1.680592297014922e-05,
+      "loss": 1.1897,
+      "step": 3327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1372939406502223,
+      "learning_rate": 1.6803897790147026e-05,
+      "loss": 1.2632,
+      "step": 3328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7452018001477767,
+      "learning_rate": 1.6801872090429176e-05,
+      "loss": 1.2056,
+      "step": 3329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5265266078175457,
+      "learning_rate": 1.6799845871150395e-05,
+      "loss": 1.1865,
+      "step": 3330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.416678208240114,
+      "learning_rate": 1.6797819132465462e-05,
+      "loss": 1.2344,
+      "step": 3331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1770246207093225,
+      "learning_rate": 1.6795791874529187e-05,
+      "loss": 1.1328,
+      "step": 3332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.044242591342121,
+      "learning_rate": 1.679376409749642e-05,
+      "loss": 1.2437,
+      "step": 3333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.193607135808856,
+      "learning_rate": 1.6791735801522056e-05,
+      "loss": 1.2007,
+      "step": 3334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3128770690108063,
+      "learning_rate": 1.6789706986761026e-05,
+      "loss": 1.2273,
+      "step": 3335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.910365963050929,
+      "learning_rate": 1.6787677653368295e-05,
+      "loss": 1.2178,
+      "step": 3336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.541095476964984,
+      "learning_rate": 1.678564780149888e-05,
+      "loss": 1.2061,
+      "step": 3337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.618748755387375,
+      "learning_rate": 1.678361743130783e-05,
+      "loss": 1.209,
+      "step": 3338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3888448258823516,
+      "learning_rate": 1.678158654295024e-05,
+      "loss": 1.1692,
+      "step": 3339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2134486797891917,
+      "learning_rate": 1.6779555136581228e-05,
+      "loss": 1.2607,
+      "step": 3340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2863745881025284,
+      "learning_rate": 1.6777523212355967e-05,
+      "loss": 1.2058,
+      "step": 3341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2195328510142054,
+      "learning_rate": 1.677549077042967e-05,
+      "loss": 1.1797,
+      "step": 3342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0182607719626517,
+      "learning_rate": 1.6773457810957584e-05,
+      "loss": 1.1577,
+      "step": 3343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2916704406507904,
+      "learning_rate": 1.6771424334094994e-05,
+      "loss": 1.1472,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5063743575683346,
+      "learning_rate": 1.6769390339997228e-05,
+      "loss": 1.2271,
+      "step": 3345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.697602778037925,
+      "learning_rate": 1.6767355828819653e-05,
+      "loss": 1.1577,
+      "step": 3346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.185201058512337,
+      "learning_rate": 1.6765320800717675e-05,
+      "loss": 1.2383,
+      "step": 3347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9755630988373505,
+      "learning_rate": 1.676328525584674e-05,
+      "loss": 1.2085,
+      "step": 3348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1070187678540786,
+      "learning_rate": 1.676124919436233e-05,
+      "loss": 1.259,
+      "step": 3349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8792372006080655,
+      "learning_rate": 1.6759212616419974e-05,
+      "loss": 1.2515,
+      "step": 3350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.153086388689916,
+      "learning_rate": 1.675717552217523e-05,
+      "loss": 1.1792,
+      "step": 3351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0774748001705423,
+      "learning_rate": 1.6755137911783713e-05,
+      "loss": 1.1665,
+      "step": 3352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.152237484944801,
+      "learning_rate": 1.6753099785401052e-05,
+      "loss": 1.1929,
+      "step": 3353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.973001620134058,
+      "learning_rate": 1.6751061143182937e-05,
+      "loss": 1.1423,
+      "step": 3354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1223125982730187,
+      "learning_rate": 1.674902198528509e-05,
+      "loss": 1.1494,
+      "step": 3355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1442849880246597,
+      "learning_rate": 1.674698231186327e-05,
+      "loss": 1.1809,
+      "step": 3356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.300301505457457,
+      "learning_rate": 1.6744942123073275e-05,
+      "loss": 1.2468,
+      "step": 3357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3055852822871485,
+      "learning_rate": 1.674290141907095e-05,
+      "loss": 1.2065,
+      "step": 3358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.907212777436634,
+      "learning_rate": 1.674086020001217e-05,
+      "loss": 1.1548,
+      "step": 3359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3289502378868603,
+      "learning_rate": 1.6738818466052858e-05,
+      "loss": 1.2573,
+      "step": 3360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4076570209976493,
+      "learning_rate": 1.6736776217348966e-05,
+      "loss": 1.1389,
+      "step": 3361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8181334557980553,
+      "learning_rate": 1.6734733454056495e-05,
+      "loss": 1.1309,
+      "step": 3362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5131099298878103,
+      "learning_rate": 1.6732690176331483e-05,
+      "loss": 1.1733,
+      "step": 3363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3867005820155534,
+      "learning_rate": 1.673064638433e-05,
+      "loss": 1.2559,
+      "step": 3364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5066636772068396,
+      "learning_rate": 1.6728602078208168e-05,
+      "loss": 1.2427,
+      "step": 3365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0895386594520136,
+      "learning_rate": 1.6726557258122133e-05,
+      "loss": 1.0881,
+      "step": 3366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6231696317324698,
+      "learning_rate": 1.6724511924228093e-05,
+      "loss": 1.2021,
+      "step": 3367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6487415820571414,
+      "learning_rate": 1.6722466076682287e-05,
+      "loss": 1.1875,
+      "step": 3368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1994412391247278,
+      "learning_rate": 1.6720419715640975e-05,
+      "loss": 1.2241,
+      "step": 3369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.335728861305684,
+      "learning_rate": 1.6718372841260478e-05,
+      "loss": 1.1919,
+      "step": 3370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.253422533873965,
+      "learning_rate": 1.671632545369714e-05,
+      "loss": 1.1877,
+      "step": 3371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8871927298225377,
+      "learning_rate": 1.6714277553107355e-05,
+      "loss": 1.259,
+      "step": 3372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.336603047867273,
+      "learning_rate": 1.6712229139647552e-05,
+      "loss": 1.1653,
+      "step": 3373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.868104085272409,
+      "learning_rate": 1.6710180213474197e-05,
+      "loss": 1.1382,
+      "step": 3374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.161943961299747,
+      "learning_rate": 1.6708130774743793e-05,
+      "loss": 1.2686,
+      "step": 3375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3017565742181207,
+      "learning_rate": 1.6706080823612897e-05,
+      "loss": 1.187,
+      "step": 3376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8451033936349677,
+      "learning_rate": 1.6704030360238085e-05,
+      "loss": 1.2593,
+      "step": 3377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.818763058511513,
+      "learning_rate": 1.6701979384775985e-05,
+      "loss": 1.2871,
+      "step": 3378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.110998411677365,
+      "learning_rate": 1.6699927897383262e-05,
+      "loss": 1.2227,
+      "step": 3379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9879615439699774,
+      "learning_rate": 1.6697875898216617e-05,
+      "loss": 1.1553,
+      "step": 3380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6633460190988587,
+      "learning_rate": 1.6695823387432795e-05,
+      "loss": 1.1821,
+      "step": 3381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.226243152889305,
+      "learning_rate": 1.6693770365188573e-05,
+      "loss": 1.1731,
+      "step": 3382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.525334647439582,
+      "learning_rate": 1.6691716831640768e-05,
+      "loss": 1.2637,
+      "step": 3383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2159101193906143,
+      "learning_rate": 1.668966278694625e-05,
+      "loss": 1.207,
+      "step": 3384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.157583087943574,
+      "learning_rate": 1.668760823126191e-05,
+      "loss": 1.2708,
+      "step": 3385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.781174283116833,
+      "learning_rate": 1.6685553164744686e-05,
+      "loss": 1.1526,
+      "step": 3386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2274562501887742,
+      "learning_rate": 1.6683497587551552e-05,
+      "loss": 1.2285,
+      "step": 3387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7548269884054766,
+      "learning_rate": 1.6681441499839527e-05,
+      "loss": 1.2212,
+      "step": 3388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1688976795631185,
+      "learning_rate": 1.6679384901765666e-05,
+      "loss": 1.1838,
+      "step": 3389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.900497755005197,
+      "learning_rate": 1.6677327793487056e-05,
+      "loss": 1.1997,
+      "step": 3390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5286656150550777,
+      "learning_rate": 1.6675270175160837e-05,
+      "loss": 1.3032,
+      "step": 3391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.229275232426884,
+      "learning_rate": 1.6673212046944173e-05,
+      "loss": 1.2666,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.237078648726684,
+      "learning_rate": 1.6671153408994282e-05,
+      "loss": 1.2134,
+      "step": 3393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.185733036171713,
+      "learning_rate": 1.6669094261468405e-05,
+      "loss": 1.2444,
+      "step": 3394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8432570151483887,
+      "learning_rate": 1.6667034604523838e-05,
+      "loss": 1.177,
+      "step": 3395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.089471386916887,
+      "learning_rate": 1.66649744383179e-05,
+      "loss": 1.1643,
+      "step": 3396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1575226394191582,
+      "learning_rate": 1.666291376300796e-05,
+      "loss": 1.1804,
+      "step": 3397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4116321063024326,
+      "learning_rate": 1.6660852578751426e-05,
+      "loss": 1.1897,
+      "step": 3398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.835428708847911,
+      "learning_rate": 1.6658790885705737e-05,
+      "loss": 1.1477,
+      "step": 3399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5310427028071336,
+      "learning_rate": 1.665672868402837e-05,
+      "loss": 1.2085,
+      "step": 3400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.626934509558248,
+      "learning_rate": 1.6654665973876866e-05,
+      "loss": 1.2048,
+      "step": 3401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.708581616129391,
+      "learning_rate": 1.6652602755408762e-05,
+      "loss": 1.2297,
+      "step": 3402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.379666427038687,
+      "learning_rate": 1.665053902878167e-05,
+      "loss": 1.1807,
+      "step": 3403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.375010276467311,
+      "learning_rate": 1.6648474794153226e-05,
+      "loss": 1.2246,
+      "step": 3404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.280415707773958,
+      "learning_rate": 1.6646410051681107e-05,
+      "loss": 1.1584,
+      "step": 3405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.997401300208074,
+      "learning_rate": 1.6644344801523022e-05,
+      "loss": 1.1831,
+      "step": 3406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.302522459711999,
+      "learning_rate": 1.6642279043836734e-05,
+      "loss": 1.1965,
+      "step": 3407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.569452944310659,
+      "learning_rate": 1.664021277878003e-05,
+      "loss": 1.2227,
+      "step": 3408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6321064829547653,
+      "learning_rate": 1.6638146006510744e-05,
+      "loss": 1.158,
+      "step": 3409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.147306317441206,
+      "learning_rate": 1.663607872718674e-05,
+      "loss": 1.291,
+      "step": 3410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2271475164881616,
+      "learning_rate": 1.663401094096594e-05,
+      "loss": 1.241,
+      "step": 3411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.050746113685472,
+      "learning_rate": 1.6631942648006284e-05,
+      "loss": 1.2295,
+      "step": 3412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9694307117026018,
+      "learning_rate": 1.6629873848465755e-05,
+      "loss": 1.1565,
+      "step": 3413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7562793050845085,
+      "learning_rate": 1.6627804542502385e-05,
+      "loss": 1.1411,
+      "step": 3414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.251333607266329,
+      "learning_rate": 1.6625734730274238e-05,
+      "loss": 1.1846,
+      "step": 3415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.782522609735633,
+      "learning_rate": 1.662366441193941e-05,
+      "loss": 1.2236,
+      "step": 3416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.050875895670775,
+      "learning_rate": 1.6621593587656047e-05,
+      "loss": 1.1692,
+      "step": 3417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1686071071393074,
+      "learning_rate": 1.661952225758233e-05,
+      "loss": 1.249,
+      "step": 3418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8996572622978736,
+      "learning_rate": 1.6617450421876477e-05,
+      "loss": 1.2122,
+      "step": 3419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.399494292179018,
+      "learning_rate": 1.6615378080696737e-05,
+      "loss": 1.1743,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.924450300237119,
+      "learning_rate": 1.661330523420142e-05,
+      "loss": 1.2322,
+      "step": 3421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9096245471974727,
+      "learning_rate": 1.6611231882548848e-05,
+      "loss": 1.209,
+      "step": 3422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.61052944249715,
+      "learning_rate": 1.66091580258974e-05,
+      "loss": 1.2246,
+      "step": 3423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8862489127942954,
+      "learning_rate": 1.6607083664405487e-05,
+      "loss": 1.1621,
+      "step": 3424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.305334382683531,
+      "learning_rate": 1.6605008798231557e-05,
+      "loss": 1.2424,
+      "step": 3425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.009948024925653,
+      "learning_rate": 1.66029334275341e-05,
+      "loss": 1.1555,
+      "step": 3426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2775869465405685,
+      "learning_rate": 1.6600857552471644e-05,
+      "loss": 1.1858,
+      "step": 3427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6271155255010603,
+      "learning_rate": 1.659878117320276e-05,
+      "loss": 1.2842,
+      "step": 3428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.785804802766517,
+      "learning_rate": 1.6596704289886036e-05,
+      "loss": 1.1875,
+      "step": 3429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2369365301556847,
+      "learning_rate": 1.659462690268013e-05,
+      "loss": 1.1851,
+      "step": 3430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0359919062273195,
+      "learning_rate": 1.6592549011743717e-05,
+      "loss": 1.2231,
+      "step": 3431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.014626734534667,
+      "learning_rate": 1.6590470617235515e-05,
+      "loss": 1.1492,
+      "step": 3432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.89980133488995,
+      "learning_rate": 1.6588391719314288e-05,
+      "loss": 1.1814,
+      "step": 3433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2107235932683387,
+      "learning_rate": 1.658631231813883e-05,
+      "loss": 1.2327,
+      "step": 3434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5139061353782175,
+      "learning_rate": 1.6584232413867976e-05,
+      "loss": 1.2063,
+      "step": 3435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.970030381762959,
+      "learning_rate": 1.6582152006660597e-05,
+      "loss": 1.1477,
+      "step": 3436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3540236531204624,
+      "learning_rate": 1.6580071096675605e-05,
+      "loss": 1.1411,
+      "step": 3437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.750097466359509,
+      "learning_rate": 1.6577989684071956e-05,
+      "loss": 1.1821,
+      "step": 3438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9541151150479252,
+      "learning_rate": 1.6575907769008633e-05,
+      "loss": 1.1975,
+      "step": 3439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2668574405462443,
+      "learning_rate": 1.6573825351644663e-05,
+      "loss": 1.1765,
+      "step": 3440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1179549772552906,
+      "learning_rate": 1.6571742432139117e-05,
+      "loss": 1.2175,
+      "step": 3441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3547295642554453,
+      "learning_rate": 1.6569659010651093e-05,
+      "loss": 1.2144,
+      "step": 3442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.175248276555693,
+      "learning_rate": 1.6567575087339734e-05,
+      "loss": 1.1716,
+      "step": 3443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9586595218470366,
+      "learning_rate": 1.6565490662364222e-05,
+      "loss": 1.1997,
+      "step": 3444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0021673641373847,
+      "learning_rate": 1.656340573588378e-05,
+      "loss": 1.2217,
+      "step": 3445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8586267115271298,
+      "learning_rate": 1.6561320308057655e-05,
+      "loss": 1.1375,
+      "step": 3446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5112152463985895,
+      "learning_rate": 1.6559234379045152e-05,
+      "loss": 1.1687,
+      "step": 3447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.326661257461596,
+      "learning_rate": 1.6557147949005597e-05,
+      "loss": 1.1973,
+      "step": 3448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9099456885348136,
+      "learning_rate": 1.6555061018098367e-05,
+      "loss": 1.1543,
+      "step": 3449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.006443004124638,
+      "learning_rate": 1.6552973586482867e-05,
+      "loss": 1.2168,
+      "step": 3450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9422913347831634,
+      "learning_rate": 1.6550885654318556e-05,
+      "loss": 1.1016,
+      "step": 3451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.857860137199325,
+      "learning_rate": 1.654879722176491e-05,
+      "loss": 1.1863,
+      "step": 3452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.865964724202572,
+      "learning_rate": 1.654670828898146e-05,
+      "loss": 1.1665,
+      "step": 3453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.951679385779684,
+      "learning_rate": 1.6544618856127764e-05,
+      "loss": 1.1782,
+      "step": 3454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.142371697518511,
+      "learning_rate": 1.654252892336343e-05,
+      "loss": 1.2405,
+      "step": 3455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.065464450083701,
+      "learning_rate": 1.6540438490848087e-05,
+      "loss": 1.2124,
+      "step": 3456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.893141039034558,
+      "learning_rate": 1.6538347558741424e-05,
+      "loss": 1.2202,
+      "step": 3457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1432930467729534,
+      "learning_rate": 1.653625612720315e-05,
+      "loss": 1.1882,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.521000663210536,
+      "learning_rate": 1.6534164196393022e-05,
+      "loss": 1.2048,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4135003123510588,
+      "learning_rate": 1.653207176647083e-05,
+      "loss": 1.2114,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1025140212437994,
+      "learning_rate": 1.652997883759641e-05,
+      "loss": 1.1729,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1629545235136063,
+      "learning_rate": 1.6527885409929622e-05,
+      "loss": 1.2153,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.049387451562867,
+      "learning_rate": 1.6525791483630377e-05,
+      "loss": 1.2004,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9990242265869815,
+      "learning_rate": 1.652369705885862e-05,
+      "loss": 1.2017,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.77217797186607,
+      "learning_rate": 1.652160213577433e-05,
+      "loss": 1.1616,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.552401357121892,
+      "learning_rate": 1.651950671453753e-05,
+      "loss": 1.2112,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.852411266952452,
+      "learning_rate": 1.651741079530828e-05,
+      "loss": 1.1453,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8620442761404554,
+      "learning_rate": 1.6515314378246675e-05,
+      "loss": 1.1492,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.944137356990262,
+      "learning_rate": 1.651321746351285e-05,
+      "loss": 1.1833,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9041049334264226,
+      "learning_rate": 1.6511120051266977e-05,
+      "loss": 1.218,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8387954977421246,
+      "learning_rate": 1.6509022141669273e-05,
+      "loss": 1.168,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.208518706099658,
+      "learning_rate": 1.6506923734879977e-05,
+      "loss": 1.2515,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3373252240237443,
+      "learning_rate": 1.6504824831059384e-05,
+      "loss": 1.1799,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9522601178112406,
+      "learning_rate": 1.6502725430367814e-05,
+      "loss": 1.1401,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3292740143254362,
+      "learning_rate": 1.650062553296563e-05,
+      "loss": 1.2324,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.465333604402415,
+      "learning_rate": 1.6498525139013234e-05,
+      "loss": 1.1858,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5589221666221245,
+      "learning_rate": 1.6496424248671064e-05,
+      "loss": 1.1951,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8688007634503263,
+      "learning_rate": 1.64943228620996e-05,
+      "loss": 1.1655,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.800456211299716,
+      "learning_rate": 1.6492220979459345e-05,
+      "loss": 1.2068,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.535455708175995,
+      "learning_rate": 1.6490118600910867e-05,
+      "loss": 1.187,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.939537692119775,
+      "learning_rate": 1.6488015726614745e-05,
+      "loss": 1.1873,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9470020981996408,
+      "learning_rate": 1.6485912356731612e-05,
+      "loss": 1.2021,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3958319866368143,
+      "learning_rate": 1.648380849142213e-05,
+      "loss": 1.2253,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7218134857141094,
+      "learning_rate": 1.6481704130847013e-05,
+      "loss": 1.1586,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.094276854328976,
+      "learning_rate": 1.6479599275166986e-05,
+      "loss": 1.1848,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.746638888165415,
+      "learning_rate": 1.647749392454284e-05,
+      "loss": 1.2788,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0455177612126287,
+      "learning_rate": 1.6475388079135387e-05,
+      "loss": 1.2578,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4518336742727658,
+      "learning_rate": 1.6473281739105488e-05,
+      "loss": 1.1948,
+      "step": 3488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.044461535896855,
+      "learning_rate": 1.647117490461403e-05,
+      "loss": 1.1365,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.204105982650968,
+      "learning_rate": 1.6469067575821947e-05,
+      "loss": 1.1604,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.705607564358709,
+      "learning_rate": 1.6466959752890207e-05,
+      "loss": 1.1885,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1798227426989114,
+      "learning_rate": 1.6464851435979813e-05,
+      "loss": 1.1179,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2688779538579946,
+      "learning_rate": 1.646274262525181e-05,
+      "loss": 1.2275,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.173717382699788,
+      "learning_rate": 1.646063332086728e-05,
+      "loss": 1.1301,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.218178879110543,
+      "learning_rate": 1.6458523522987345e-05,
+      "loss": 1.2393,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2685338429395236,
+      "learning_rate": 1.6456413231773157e-05,
+      "loss": 1.2266,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.795118311510679,
+      "learning_rate": 1.645430244738591e-05,
+      "loss": 1.2104,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0991056445481813,
+      "learning_rate": 1.6452191169986848e-05,
+      "loss": 1.1548,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4225121728459964,
+      "learning_rate": 1.6450079399737223e-05,
+      "loss": 1.1685,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.444928102474689,
+      "learning_rate": 1.6447967136798354e-05,
+      "loss": 1.1138,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.276225516330536,
+      "learning_rate": 1.6445854381331585e-05,
+      "loss": 1.2371,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.850398434771539,
+      "learning_rate": 1.6443741133498297e-05,
+      "loss": 1.0986,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.975288605626785,
+      "learning_rate": 1.644162739345991e-05,
+      "loss": 1.1379,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.620926634266525,
+      "learning_rate": 1.6439513161377886e-05,
+      "loss": 1.2161,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.467264588657385,
+      "learning_rate": 1.6437398437413713e-05,
+      "loss": 1.1416,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.892634724957758,
+      "learning_rate": 1.643528322172893e-05,
+      "loss": 1.2275,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8561772129900076,
+      "learning_rate": 1.6433167514485105e-05,
+      "loss": 1.1599,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.007777381904848,
+      "learning_rate": 1.6431051315843853e-05,
+      "loss": 1.1853,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.8529624913891025,
+      "learning_rate": 1.6428934625966812e-05,
+      "loss": 1.1987,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.373589584401813,
+      "learning_rate": 1.6426817445015667e-05,
+      "loss": 1.2085,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9030355285658738,
+      "learning_rate": 1.642469977315214e-05,
+      "loss": 1.1912,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.262336208130834,
+      "learning_rate": 1.642258161053799e-05,
+      "loss": 1.2573,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.047607593638764,
+      "learning_rate": 1.642046295733501e-05,
+      "loss": 1.2329,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1606443808985922,
+      "learning_rate": 1.6418343813705036e-05,
+      "loss": 1.2327,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.651735094679895,
+      "learning_rate": 1.6416224179809936e-05,
+      "loss": 1.2568,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6614932475272592,
+      "learning_rate": 1.641410405581162e-05,
+      "loss": 1.2073,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8355369962391954,
+      "learning_rate": 1.6411983441872035e-05,
+      "loss": 1.1804,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4746052153371068,
+      "learning_rate": 1.6409862338153164e-05,
+      "loss": 1.2026,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9768062360192777,
+      "learning_rate": 1.6407740744817028e-05,
+      "loss": 1.1936,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4290746310887337,
+      "learning_rate": 1.6405618662025678e-05,
+      "loss": 1.1599,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.244209380759028,
+      "learning_rate": 1.640349608994122e-05,
+      "loss": 1.2007,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8723445521932613,
+      "learning_rate": 1.640137302872578e-05,
+      "loss": 1.2378,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.778801748434566,
+      "learning_rate": 1.6399249478541527e-05,
+      "loss": 1.1494,
+      "step": 3523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.524536228731531,
+      "learning_rate": 1.639712543955067e-05,
+      "loss": 1.1968,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.40515253826312,
+      "learning_rate": 1.6395000911915456e-05,
+      "loss": 1.175,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6179596595448382,
+      "learning_rate": 1.639287589579816e-05,
+      "loss": 1.1624,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.470435323778797,
+      "learning_rate": 1.6390750391361113e-05,
+      "loss": 1.157,
+      "step": 3527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.52286215851052,
+      "learning_rate": 1.6388624398766662e-05,
+      "loss": 1.2168,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.19537389918938,
+      "learning_rate": 1.6386497918177208e-05,
+      "loss": 1.2141,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2925540369077964,
+      "learning_rate": 1.6384370949755174e-05,
+      "loss": 1.1785,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9868334627373847,
+      "learning_rate": 1.6382243493663038e-05,
+      "loss": 1.1892,
+      "step": 3531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.498727792899524,
+      "learning_rate": 1.6380115550063294e-05,
+      "loss": 1.2812,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.12766333260887,
+      "learning_rate": 1.6377987119118493e-05,
+      "loss": 1.1479,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.18564395388246,
+      "learning_rate": 1.6375858200991217e-05,
+      "loss": 1.1738,
+      "step": 3534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4022879556732324,
+      "learning_rate": 1.637372879584408e-05,
+      "loss": 1.2041,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.168307878014383,
+      "learning_rate": 1.637159890383973e-05,
+      "loss": 1.095,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0810884872099993,
+      "learning_rate": 1.636946852514087e-05,
+      "loss": 1.2317,
+      "step": 3537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.211502920932121,
+      "learning_rate": 1.6367337659910223e-05,
+      "loss": 1.1802,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.914884773561023,
+      "learning_rate": 1.6365206308310557e-05,
+      "loss": 1.2114,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3887222216216286,
+      "learning_rate": 1.6363074470504672e-05,
+      "loss": 1.2327,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6852691090361676,
+      "learning_rate": 1.636094214665542e-05,
+      "loss": 1.2383,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6575682231863724,
+      "learning_rate": 1.6358809336925662e-05,
+      "loss": 1.2534,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.018995514326609,
+      "learning_rate": 1.6356676041478322e-05,
+      "loss": 1.1794,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.592441334980316,
+      "learning_rate": 1.635454226047635e-05,
+      "loss": 1.134,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.932332501025126,
+      "learning_rate": 1.6352407994082734e-05,
+      "loss": 1.1626,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.302131314802121,
+      "learning_rate": 1.6350273242460502e-05,
+      "loss": 1.228,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2490825128874543,
+      "learning_rate": 1.6348138005772717e-05,
+      "loss": 1.2322,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3131434189941196,
+      "learning_rate": 1.6346002284182476e-05,
+      "loss": 1.2522,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.053350573431078,
+      "learning_rate": 1.6343866077852917e-05,
+      "loss": 1.2166,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4860472438432883,
+      "learning_rate": 1.6341729386947214e-05,
+      "loss": 1.1655,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6998911543161106,
+      "learning_rate": 1.6339592211628583e-05,
+      "loss": 1.1484,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4075872702385883,
+      "learning_rate": 1.6337454552060264e-05,
+      "loss": 1.2612,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.496493031568085,
+      "learning_rate": 1.6335316408405548e-05,
+      "loss": 1.1616,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5867824745811436,
+      "learning_rate": 1.633317778082776e-05,
+      "loss": 1.2114,
+      "step": 3554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.208251048001512,
+      "learning_rate": 1.6331038669490248e-05,
+      "loss": 1.1724,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.805912727951814,
+      "learning_rate": 1.6328899074556415e-05,
+      "loss": 1.1709,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1850979298313637,
+      "learning_rate": 1.6326758996189696e-05,
+      "loss": 1.1753,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.033799702763256,
+      "learning_rate": 1.6324618434553556e-05,
+      "loss": 1.2417,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2709649456504,
+      "learning_rate": 1.6322477389811507e-05,
+      "loss": 1.2292,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3547775254781724,
+      "learning_rate": 1.6320335862127087e-05,
+      "loss": 1.2163,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.365437779268242,
+      "learning_rate": 1.6318193851663883e-05,
+      "loss": 1.2351,
+      "step": 3561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3102037086673612,
+      "learning_rate": 1.6316051358585503e-05,
+      "loss": 1.1705,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9127757978040325,
+      "learning_rate": 1.631390838305561e-05,
+      "loss": 1.1335,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7878077645944335,
+      "learning_rate": 1.6311764925237893e-05,
+      "loss": 1.2134,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8151555069006657,
+      "learning_rate": 1.6309620985296075e-05,
+      "loss": 1.1799,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7175776689608706,
+      "learning_rate": 1.6307476563393933e-05,
+      "loss": 1.229,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8096600384818036,
+      "learning_rate": 1.6305331659695252e-05,
+      "loss": 1.2026,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3925323184697787,
+      "learning_rate": 1.6303186274363884e-05,
+      "loss": 1.0982,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.966284689595831,
+      "learning_rate": 1.6301040407563697e-05,
+      "loss": 1.1892,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.433656529933796,
+      "learning_rate": 1.6298894059458607e-05,
+      "loss": 1.1746,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.768134901895349,
+      "learning_rate": 1.629674723021256e-05,
+      "loss": 1.1479,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.756276874862138,
+      "learning_rate": 1.6294599919989544e-05,
+      "loss": 1.2664,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.026063112293989,
+      "learning_rate": 1.629245212895358e-05,
+      "loss": 1.1375,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3587058236684455,
+      "learning_rate": 1.6290303857268724e-05,
+      "loss": 1.2305,
+      "step": 3574
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.248920812195674,
+      "learning_rate": 1.6288155105099077e-05,
+      "loss": 1.2214,
+      "step": 3575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.258806213517249,
+      "learning_rate": 1.6286005872608768e-05,
+      "loss": 1.2148,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.629650356772155,
+      "learning_rate": 1.6283856159961967e-05,
+      "loss": 1.1233,
+      "step": 3577
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9566201402340355,
+      "learning_rate": 1.6281705967322882e-05,
+      "loss": 1.1482,
+      "step": 3578
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5687690274220714,
+      "learning_rate": 1.627955529485575e-05,
+      "loss": 1.2578,
+      "step": 3579
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.016788908456323,
+      "learning_rate": 1.6277404142724854e-05,
+      "loss": 1.1877,
+      "step": 3580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.879184585545489,
+      "learning_rate": 1.6275252511094515e-05,
+      "loss": 1.1621,
+      "step": 3581
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.073393826763461,
+      "learning_rate": 1.6273100400129072e-05,
+      "loss": 1.1138,
+      "step": 3582
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7250382364462964,
+      "learning_rate": 1.6270947809992924e-05,
+      "loss": 1.1123,
+      "step": 3583
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.002508328889765,
+      "learning_rate": 1.62687947408505e-05,
+      "loss": 1.1667,
+      "step": 3584
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.555462119044535,
+      "learning_rate": 1.6266641192866248e-05,
+      "loss": 1.2378,
+      "step": 3585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9733384946671655,
+      "learning_rate": 1.6264487166204677e-05,
+      "loss": 1.187,
+      "step": 3586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1551465328036503,
+      "learning_rate": 1.6262332661030322e-05,
+      "loss": 1.1492,
+      "step": 3587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8767589011174697,
+      "learning_rate": 1.626017767750775e-05,
+      "loss": 1.1858,
+      "step": 3588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0373777240309394,
+      "learning_rate": 1.6258022215801575e-05,
+      "loss": 1.199,
+      "step": 3589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.490343161954476,
+      "learning_rate": 1.625586627607644e-05,
+      "loss": 1.2307,
+      "step": 3590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1821546128419844,
+      "learning_rate": 1.6253709858497028e-05,
+      "loss": 1.147,
+      "step": 3591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.37559221428573,
+      "learning_rate": 1.625155296322805e-05,
+      "loss": 1.1624,
+      "step": 3592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.506932540642277,
+      "learning_rate": 1.6249395590434267e-05,
+      "loss": 1.1377,
+      "step": 3593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0718647761431868,
+      "learning_rate": 1.624723774028047e-05,
+      "loss": 1.1899,
+      "step": 3594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1725424800624107,
+      "learning_rate": 1.6245079412931483e-05,
+      "loss": 1.238,
+      "step": 3595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.528185016512208,
+      "learning_rate": 1.624292060855217e-05,
+      "loss": 1.1926,
+      "step": 3596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.396903227310016,
+      "learning_rate": 1.6240761327307434e-05,
+      "loss": 1.2402,
+      "step": 3597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.156535612461306,
+      "learning_rate": 1.623860156936221e-05,
+      "loss": 1.1084,
+      "step": 3598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.789543969639989,
+      "learning_rate": 1.623644133488147e-05,
+      "loss": 1.1382,
+      "step": 3599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9550777228557465,
+      "learning_rate": 1.6234280624030226e-05,
+      "loss": 1.1887,
+      "step": 3600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.874006704728273,
+      "learning_rate": 1.6232119436973525e-05,
+      "loss": 1.0977,
+      "step": 3601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0084492802789256,
+      "learning_rate": 1.622995777387644e-05,
+      "loss": 1.1362,
+      "step": 3602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4647363101220385,
+      "learning_rate": 1.6227795634904103e-05,
+      "loss": 1.1365,
+      "step": 3603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.375728389224155,
+      "learning_rate": 1.622563302022166e-05,
+      "loss": 1.1987,
+      "step": 3604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.190905807664575,
+      "learning_rate": 1.6223469929994305e-05,
+      "loss": 1.1492,
+      "step": 3605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7250984655316506,
+      "learning_rate": 1.6221306364387265e-05,
+      "loss": 1.1328,
+      "step": 3606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2503011433873628,
+      "learning_rate": 1.621914232356581e-05,
+      "loss": 1.2102,
+      "step": 3607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9876501124343764,
+      "learning_rate": 1.6216977807695227e-05,
+      "loss": 1.1689,
+      "step": 3608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.95035160661506,
+      "learning_rate": 1.6214812816940864e-05,
+      "loss": 1.1438,
+      "step": 3609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4667621531387383,
+      "learning_rate": 1.6212647351468088e-05,
+      "loss": 1.1347,
+      "step": 3610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1579651276978797,
+      "learning_rate": 1.6210481411442312e-05,
+      "loss": 1.165,
+      "step": 3611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.446780974978896,
+      "learning_rate": 1.6208314997028977e-05,
+      "loss": 1.158,
+      "step": 3612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.812171013783556,
+      "learning_rate": 1.6206148108393568e-05,
+      "loss": 1.2041,
+      "step": 3613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.739533882543173,
+      "learning_rate": 1.62039807457016e-05,
+      "loss": 1.155,
+      "step": 3614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.75583931753404,
+      "learning_rate": 1.6201812909118633e-05,
+      "loss": 1.2542,
+      "step": 3615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4415394085302355,
+      "learning_rate": 1.619964459881025e-05,
+      "loss": 1.1653,
+      "step": 3616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.57161817952975,
+      "learning_rate": 1.619747581494208e-05,
+      "loss": 1.197,
+      "step": 3617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9506762157450837,
+      "learning_rate": 1.6195306557679788e-05,
+      "loss": 1.147,
+      "step": 3618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4064942338531052,
+      "learning_rate": 1.6193136827189067e-05,
+      "loss": 1.1782,
+      "step": 3619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.479677795247296,
+      "learning_rate": 1.6190966623635656e-05,
+      "loss": 1.2534,
+      "step": 3620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.666161620855055,
+      "learning_rate": 1.6188795947185326e-05,
+      "loss": 1.1858,
+      "step": 3621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.763230920545753,
+      "learning_rate": 1.618662479800388e-05,
+      "loss": 1.2214,
+      "step": 3622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.473561497869118,
+      "learning_rate": 1.618445317625717e-05,
+      "loss": 1.2195,
+      "step": 3623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.347303058894846,
+      "learning_rate": 1.6182281082111066e-05,
+      "loss": 1.2524,
+      "step": 3624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8799721865499173,
+      "learning_rate": 1.6180108515731487e-05,
+      "loss": 1.1521,
+      "step": 3625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.512074093057532,
+      "learning_rate": 1.6177935477284386e-05,
+      "loss": 1.1379,
+      "step": 3626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.167671955678002,
+      "learning_rate": 1.617576196693575e-05,
+      "loss": 1.106,
+      "step": 3627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6178392740697043,
+      "learning_rate": 1.6173587984851596e-05,
+      "loss": 1.239,
+      "step": 3628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4457245133532624,
+      "learning_rate": 1.6171413531197995e-05,
+      "loss": 1.2754,
+      "step": 3629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.63203817371458,
+      "learning_rate": 1.6169238606141034e-05,
+      "loss": 1.105,
+      "step": 3630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8858644439405468,
+      "learning_rate": 1.616706320984685e-05,
+      "loss": 1.1414,
+      "step": 3631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4660993074385806,
+      "learning_rate": 1.616488734248161e-05,
+      "loss": 1.2051,
+      "step": 3632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2637946950687113,
+      "learning_rate": 1.6162711004211513e-05,
+      "loss": 1.1719,
+      "step": 3633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.293476308210555,
+      "learning_rate": 1.61605341952028e-05,
+      "loss": 1.2058,
+      "step": 3634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1337321443580666,
+      "learning_rate": 1.615835691562175e-05,
+      "loss": 1.1892,
+      "step": 3635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6414565735829347,
+      "learning_rate": 1.615617916563467e-05,
+      "loss": 1.1514,
+      "step": 3636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8638834637143815,
+      "learning_rate": 1.6154000945407914e-05,
+      "loss": 1.0781,
+      "step": 3637
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7867102355562268,
+      "learning_rate": 1.6151822255107855e-05,
+      "loss": 1.1309,
+      "step": 3638
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.470009323935292,
+      "learning_rate": 1.614964309490092e-05,
+      "loss": 1.1411,
+      "step": 3639
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.977130229300281,
+      "learning_rate": 1.614746346495357e-05,
+      "loss": 1.2034,
+      "step": 3640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.54371553471329,
+      "learning_rate": 1.614528336543228e-05,
+      "loss": 1.2139,
+      "step": 3641
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.135980912556915,
+      "learning_rate": 1.6143102796503587e-05,
+      "loss": 1.1675,
+      "step": 3642
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0157398232282424,
+      "learning_rate": 1.6140921758334053e-05,
+      "loss": 1.208,
+      "step": 3643
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.94867746193277,
+      "learning_rate": 1.6138740251090274e-05,
+      "loss": 1.2458,
+      "step": 3644
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0888253738143248,
+      "learning_rate": 1.613655827493889e-05,
+      "loss": 1.2166,
+      "step": 3645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.677295613566454,
+      "learning_rate": 1.6134375830046566e-05,
+      "loss": 1.1748,
+      "step": 3646
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1281748778885454,
+      "learning_rate": 1.6132192916580006e-05,
+      "loss": 1.1492,
+      "step": 3647
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1014172830204765,
+      "learning_rate": 1.6130009534705958e-05,
+      "loss": 1.1467,
+      "step": 3648
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.164073170979484,
+      "learning_rate": 1.6127825684591195e-05,
+      "loss": 1.1497,
+      "step": 3649
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0806379455800887,
+      "learning_rate": 1.6125641366402535e-05,
+      "loss": 1.198,
+      "step": 3650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2063777056714926,
+      "learning_rate": 1.612345658030682e-05,
+      "loss": 1.2244,
+      "step": 3651
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.10873699969432,
+      "learning_rate": 1.6121271326470945e-05,
+      "loss": 1.1724,
+      "step": 3652
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.869033345261752,
+      "learning_rate": 1.611908560506182e-05,
+      "loss": 1.217,
+      "step": 3653
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5199672682130405,
+      "learning_rate": 1.611689941624641e-05,
+      "loss": 1.1924,
+      "step": 3654
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9558834792563733,
+      "learning_rate": 1.6114712760191697e-05,
+      "loss": 1.228,
+      "step": 3655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3762236817266995,
+      "learning_rate": 1.611252563706472e-05,
+      "loss": 1.2556,
+      "step": 3656
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8036512439442633,
+      "learning_rate": 1.6110338047032536e-05,
+      "loss": 1.1633,
+      "step": 3657
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7745460464090304,
+      "learning_rate": 1.610814999026224e-05,
+      "loss": 1.2385,
+      "step": 3658
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.181469475812881,
+      "learning_rate": 1.6105961466920977e-05,
+      "loss": 1.2764,
+      "step": 3659
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7839099674700423,
+      "learning_rate": 1.6103772477175913e-05,
+      "loss": 1.2317,
+      "step": 3660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9648887864249422,
+      "learning_rate": 1.610158302119425e-05,
+      "loss": 1.1975,
+      "step": 3661
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2015640448129714,
+      "learning_rate": 1.6099393099143236e-05,
+      "loss": 1.1797,
+      "step": 3662
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.129923179294741,
+      "learning_rate": 1.609720271119014e-05,
+      "loss": 1.1677,
+      "step": 3663
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.910036031508888,
+      "learning_rate": 1.6095011857502282e-05,
+      "loss": 1.1848,
+      "step": 3664
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1840914064499515,
+      "learning_rate": 1.6092820538247006e-05,
+      "loss": 1.1692,
+      "step": 3665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.572744002787421,
+      "learning_rate": 1.6090628753591703e-05,
+      "loss": 1.1372,
+      "step": 3666
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.350594202018647,
+      "learning_rate": 1.6088436503703784e-05,
+      "loss": 1.1519,
+      "step": 3667
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.199938214407453,
+      "learning_rate": 1.6086243788750707e-05,
+      "loss": 1.1931,
+      "step": 3668
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8164277751307716,
+      "learning_rate": 1.6084050608899965e-05,
+      "loss": 1.1326,
+      "step": 3669
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.434980523876801,
+      "learning_rate": 1.608185696431908e-05,
+      "loss": 1.1682,
+      "step": 3670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4596365142115113,
+      "learning_rate": 1.6079662855175615e-05,
+      "loss": 1.2085,
+      "step": 3671
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.810213199574778,
+      "learning_rate": 1.6077468281637166e-05,
+      "loss": 1.1614,
+      "step": 3672
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.129906476025858,
+      "learning_rate": 1.607527324387137e-05,
+      "loss": 1.1956,
+      "step": 3673
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0962737219431724,
+      "learning_rate": 1.607307774204589e-05,
+      "loss": 1.1724,
+      "step": 3674
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6640131555368107,
+      "learning_rate": 1.6070881776328434e-05,
+      "loss": 1.1262,
+      "step": 3675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8689260415528444,
+      "learning_rate": 1.606868534688673e-05,
+      "loss": 1.147,
+      "step": 3676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6107035039860103,
+      "learning_rate": 1.6066488453888568e-05,
+      "loss": 1.2222,
+      "step": 3677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.536920375108444,
+      "learning_rate": 1.6064291097501746e-05,
+      "loss": 1.1709,
+      "step": 3678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0480345289441564,
+      "learning_rate": 1.6062093277894116e-05,
+      "loss": 1.1753,
+      "step": 3679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.532477769093615,
+      "learning_rate": 1.6059894995233554e-05,
+      "loss": 1.1814,
+      "step": 3680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4052354322587575,
+      "learning_rate": 1.605769624968798e-05,
+      "loss": 1.1741,
+      "step": 3681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2293171868378217,
+      "learning_rate": 1.6055497041425333e-05,
+      "loss": 1.208,
+      "step": 3682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3743647757985613,
+      "learning_rate": 1.6053297370613617e-05,
+      "loss": 1.0393,
+      "step": 3683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.095832980688345,
+      "learning_rate": 1.6051097237420843e-05,
+      "loss": 1.2173,
+      "step": 3684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1885300786608135,
+      "learning_rate": 1.604889664201507e-05,
+      "loss": 1.1208,
+      "step": 3685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.611598661764117,
+      "learning_rate": 1.6046695584564393e-05,
+      "loss": 1.2385,
+      "step": 3686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.56556714253888,
+      "learning_rate": 1.6044494065236938e-05,
+      "loss": 1.2236,
+      "step": 3687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.168441139577564,
+      "learning_rate": 1.6042292084200864e-05,
+      "loss": 1.1765,
+      "step": 3688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7284083221066155,
+      "learning_rate": 1.6040089641624377e-05,
+      "loss": 1.2092,
+      "step": 3689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.874938235276145,
+      "learning_rate": 1.6037886737675708e-05,
+      "loss": 1.1816,
+      "step": 3690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8760928636457987,
+      "learning_rate": 1.603568337252312e-05,
+      "loss": 1.1956,
+      "step": 3691
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.588710426763052,
+      "learning_rate": 1.6033479546334924e-05,
+      "loss": 1.1899,
+      "step": 3692
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.176930522826574,
+      "learning_rate": 1.603127525927946e-05,
+      "loss": 1.1921,
+      "step": 3693
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.352681724197597,
+      "learning_rate": 1.6029070511525094e-05,
+      "loss": 1.2676,
+      "step": 3694
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7592759144194896,
+      "learning_rate": 1.6026865303240243e-05,
+      "loss": 1.1221,
+      "step": 3695
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.982809000600181,
+      "learning_rate": 1.602465963459335e-05,
+      "loss": 1.1841,
+      "step": 3696
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9138707300473117,
+      "learning_rate": 1.6022453505752896e-05,
+      "loss": 1.2227,
+      "step": 3697
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.900580620563348,
+      "learning_rate": 1.6020246916887394e-05,
+      "loss": 1.1755,
+      "step": 3698
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1410275487171773,
+      "learning_rate": 1.6018039868165398e-05,
+      "loss": 1.1885,
+      "step": 3699
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.570423141722566,
+      "learning_rate": 1.6015832359755483e-05,
+      "loss": 1.1794,
+      "step": 3700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.310943344451072,
+      "learning_rate": 1.6013624391826282e-05,
+      "loss": 1.2351,
+      "step": 3701
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.380627996933434,
+      "learning_rate": 1.6011415964546448e-05,
+      "loss": 1.1995,
+      "step": 3702
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6507228412614965,
+      "learning_rate": 1.6009207078084666e-05,
+      "loss": 1.1555,
+      "step": 3703
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.174242673357025,
+      "learning_rate": 1.6006997732609665e-05,
+      "loss": 1.2551,
+      "step": 3704
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.838974546776511,
+      "learning_rate": 1.6004787928290208e-05,
+      "loss": 1.1826,
+      "step": 3705
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.804149584403174,
+      "learning_rate": 1.6002577665295085e-05,
+      "loss": 1.1931,
+      "step": 3706
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.656106221027793,
+      "learning_rate": 1.6000366943793133e-05,
+      "loss": 1.2102,
+      "step": 3707
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5776023875463965,
+      "learning_rate": 1.5998155763953217e-05,
+      "loss": 1.1626,
+      "step": 3708
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.041785567631126,
+      "learning_rate": 1.5995944125944235e-05,
+      "loss": 1.1802,
+      "step": 3709
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.054932562890947,
+      "learning_rate": 1.5993732029935123e-05,
+      "loss": 1.1487,
+      "step": 3710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.355154870166263,
+      "learning_rate": 1.5991519476094858e-05,
+      "loss": 1.1738,
+      "step": 3711
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8631579751829643,
+      "learning_rate": 1.5989306464592434e-05,
+      "loss": 1.1917,
+      "step": 3712
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.376604316186384,
+      "learning_rate": 1.5987092995596908e-05,
+      "loss": 1.1479,
+      "step": 3713
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2627858694132374,
+      "learning_rate": 1.598487906927734e-05,
+      "loss": 1.1948,
+      "step": 3714
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.321481098241624,
+      "learning_rate": 1.5982664685802852e-05,
+      "loss": 1.2107,
+      "step": 3715
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8851978017613766,
+      "learning_rate": 1.5980449845342586e-05,
+      "loss": 1.1643,
+      "step": 3716
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1346177072178736,
+      "learning_rate": 1.5978234548065718e-05,
+      "loss": 1.125,
+      "step": 3717
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.860741737916856,
+      "learning_rate": 1.597601879414147e-05,
+      "loss": 1.2505,
+      "step": 3718
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9990944011624894,
+      "learning_rate": 1.597380258373909e-05,
+      "loss": 1.1848,
+      "step": 3719
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2319011355215483,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 1.2354,
+      "step": 3720
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2225657348492422,
+      "learning_rate": 1.5969368794177108e-05,
+      "loss": 1.1428,
+      "step": 3721
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1228698628266525,
+      "learning_rate": 1.5967151215356183e-05,
+      "loss": 1.1836,
+      "step": 3722
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2047721314565867,
+      "learning_rate": 1.5964933180734473e-05,
+      "loss": 1.176,
+      "step": 3723
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.69224206428673,
+      "learning_rate": 1.5962714690481408e-05,
+      "loss": 1.1833,
+      "step": 3724
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4456447720330114,
+      "learning_rate": 1.5960495744766445e-05,
+      "loss": 1.2073,
+      "step": 3725
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.486871415609615,
+      "learning_rate": 1.5958276343759072e-05,
+      "loss": 1.2231,
+      "step": 3726
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.068679685798008,
+      "learning_rate": 1.5956056487628832e-05,
+      "loss": 1.1738,
+      "step": 3727
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.554290829392492,
+      "learning_rate": 1.5953836176545273e-05,
+      "loss": 1.1702,
+      "step": 3728
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.57883832261921,
+      "learning_rate": 1.5951615410678005e-05,
+      "loss": 1.2058,
+      "step": 3729
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.426844378268119,
+      "learning_rate": 1.5949394190196658e-05,
+      "loss": 1.2253,
+      "step": 3730
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1728483928301783,
+      "learning_rate": 1.5947172515270896e-05,
+      "loss": 1.1763,
+      "step": 3731
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3897532012359184,
+      "learning_rate": 1.5944950386070425e-05,
+      "loss": 1.2244,
+      "step": 3732
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.999262739883265,
+      "learning_rate": 1.5942727802764982e-05,
+      "loss": 1.1272,
+      "step": 3733
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9506782271768643,
+      "learning_rate": 1.594050476552434e-05,
+      "loss": 1.1633,
+      "step": 3734
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.310132703923591,
+      "learning_rate": 1.5938281274518304e-05,
+      "loss": 1.2268,
+      "step": 3735
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0806324695953267,
+      "learning_rate": 1.5936057329916718e-05,
+      "loss": 1.2427,
+      "step": 3736
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.869478057848143,
+      "learning_rate": 1.593383293188945e-05,
+      "loss": 1.1147,
+      "step": 3737
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9494999181522426,
+      "learning_rate": 1.5931608080606424e-05,
+      "loss": 1.2131,
+      "step": 3738
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1072655810223133,
+      "learning_rate": 1.592938277623757e-05,
+      "loss": 1.1428,
+      "step": 3739
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.893512370752905,
+      "learning_rate": 1.592715701895288e-05,
+      "loss": 1.2095,
+      "step": 3740
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7481712723951346,
+      "learning_rate": 1.5924930808922363e-05,
+      "loss": 1.1807,
+      "step": 3741
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.761472913472176,
+      "learning_rate": 1.5922704146316068e-05,
+      "loss": 1.1553,
+      "step": 3742
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.113173136515112,
+      "learning_rate": 1.5920477031304083e-05,
+      "loss": 1.178,
+      "step": 3743
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.029741238191986,
+      "learning_rate": 1.5918249464056518e-05,
+      "loss": 1.144,
+      "step": 3744
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1361540066492353,
+      "learning_rate": 1.5916021444743535e-05,
+      "loss": 1.1587,
+      "step": 3745
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.205646049776848,
+      "learning_rate": 1.5913792973535317e-05,
+      "loss": 1.2041,
+      "step": 3746
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.572986321907368,
+      "learning_rate": 1.591156405060208e-05,
+      "loss": 1.2168,
+      "step": 3747
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.237857883175418,
+      "learning_rate": 1.590933467611409e-05,
+      "loss": 1.1877,
+      "step": 3748
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.022727844466495,
+      "learning_rate": 1.590710485024163e-05,
+      "loss": 1.1692,
+      "step": 3749
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9379277861822404,
+      "learning_rate": 1.5904874573155035e-05,
+      "loss": 1.1919,
+      "step": 3750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.047430408366143,
+      "learning_rate": 1.5902643845024653e-05,
+      "loss": 1.2239,
+      "step": 3751
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.410461604783988,
+      "learning_rate": 1.5900412666020882e-05,
+      "loss": 1.2241,
+      "step": 3752
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.975456043694523,
+      "learning_rate": 1.5898181036314157e-05,
+      "loss": 1.1873,
+      "step": 3753
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0982818086119512,
+      "learning_rate": 1.5895948956074937e-05,
+      "loss": 1.2349,
+      "step": 3754
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2366465782801455,
+      "learning_rate": 1.5893716425473713e-05,
+      "loss": 1.2488,
+      "step": 3755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4080940968308924,
+      "learning_rate": 1.5891483444681027e-05,
+      "loss": 1.1467,
+      "step": 3756
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.093648154455995,
+      "learning_rate": 1.5889250013867437e-05,
+      "loss": 1.1755,
+      "step": 3757
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.916721810455733,
+      "learning_rate": 1.588701613320355e-05,
+      "loss": 1.2268,
+      "step": 3758
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4907915634416815,
+      "learning_rate": 1.5884781802859995e-05,
+      "loss": 1.2156,
+      "step": 3759
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3343092080512533,
+      "learning_rate": 1.588254702300745e-05,
+      "loss": 1.1846,
+      "step": 3760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2031228337338726,
+      "learning_rate": 1.5880311793816606e-05,
+      "loss": 1.2532,
+      "step": 3761
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.046189678417247,
+      "learning_rate": 1.5878076115458215e-05,
+      "loss": 1.1995,
+      "step": 3762
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.93213754473199,
+      "learning_rate": 1.587583998810304e-05,
+      "loss": 1.1587,
+      "step": 3763
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9340015926625327,
+      "learning_rate": 1.587360341192189e-05,
+      "loss": 1.1111,
+      "step": 3764
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.600962184327039,
+      "learning_rate": 1.58713663870856e-05,
+      "loss": 1.2285,
+      "step": 3765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4280197244860564,
+      "learning_rate": 1.5869128913765056e-05,
+      "loss": 1.1362,
+      "step": 3766
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4777714914892597,
+      "learning_rate": 1.5866890992131163e-05,
+      "loss": 1.271,
+      "step": 3767
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.560259048852048,
+      "learning_rate": 1.586465262235486e-05,
+      "loss": 1.1665,
+      "step": 3768
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.995460821137302,
+      "learning_rate": 1.586241380460713e-05,
+      "loss": 1.2661,
+      "step": 3769
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0224229965848313,
+      "learning_rate": 1.5860174539058988e-05,
+      "loss": 1.1904,
+      "step": 3770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.507708795876757,
+      "learning_rate": 1.585793482588147e-05,
+      "loss": 1.187,
+      "step": 3771
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.904041774355462,
+      "learning_rate": 1.5855694665245665e-05,
+      "loss": 1.1401,
+      "step": 3772
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.193493303791883,
+      "learning_rate": 1.5853454057322684e-05,
+      "loss": 1.1143,
+      "step": 3773
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4852925229062266,
+      "learning_rate": 1.585121300228368e-05,
+      "loss": 1.189,
+      "step": 3774
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4041387092552635,
+      "learning_rate": 1.584897150029983e-05,
+      "loss": 1.157,
+      "step": 3775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.537451619700538,
+      "learning_rate": 1.5846729551542357e-05,
+      "loss": 1.1819,
+      "step": 3776
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2745101358851794,
+      "learning_rate": 1.5844487156182506e-05,
+      "loss": 1.2241,
+      "step": 3777
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.833220676668162,
+      "learning_rate": 1.5842244314391568e-05,
+      "loss": 1.2915,
+      "step": 3778
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.470708966216569,
+      "learning_rate": 1.5840001026340858e-05,
+      "loss": 1.2092,
+      "step": 3779
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.526541022876302,
+      "learning_rate": 1.5837757292201733e-05,
+      "loss": 1.1675,
+      "step": 3780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.066890744032493,
+      "learning_rate": 1.5835513112145583e-05,
+      "loss": 1.2629,
+      "step": 3781
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.972579815515733,
+      "learning_rate": 1.5833268486343817e-05,
+      "loss": 1.1353,
+      "step": 3782
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.175421902612299,
+      "learning_rate": 1.5831023414967906e-05,
+      "loss": 1.2441,
+      "step": 3783
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.348740085236354,
+      "learning_rate": 1.5828777898189333e-05,
+      "loss": 1.1694,
+      "step": 3784
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.123903708988493,
+      "learning_rate": 1.5826531936179617e-05,
+      "loss": 1.2625,
+      "step": 3785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6848668294586293,
+      "learning_rate": 1.5824285529110327e-05,
+      "loss": 1.2104,
+      "step": 3786
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.802037545035946,
+      "learning_rate": 1.5822038677153045e-05,
+      "loss": 1.2004,
+      "step": 3787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.630942293881338,
+      "learning_rate": 1.5819791380479403e-05,
+      "loss": 1.1982,
+      "step": 3788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.4826629133479425,
+      "learning_rate": 1.5817543639261057e-05,
+      "loss": 1.1946,
+      "step": 3789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.682801138662739,
+      "learning_rate": 1.5815295453669702e-05,
+      "loss": 1.1821,
+      "step": 3790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.606967163452617,
+      "learning_rate": 1.5813046823877064e-05,
+      "loss": 1.1589,
+      "step": 3791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.160030604395221,
+      "learning_rate": 1.5810797750054908e-05,
+      "loss": 1.1746,
+      "step": 3792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.2753391890082035,
+      "learning_rate": 1.5808548232375026e-05,
+      "loss": 1.1917,
+      "step": 3793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.238556399492076,
+      "learning_rate": 1.5806298271009248e-05,
+      "loss": 1.1968,
+      "step": 3794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.162286061056603,
+      "learning_rate": 1.5804047866129442e-05,
+      "loss": 1.1863,
+      "step": 3795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.224638380611868,
+      "learning_rate": 1.58017970179075e-05,
+      "loss": 1.197,
+      "step": 3796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.290634380141474,
+      "learning_rate": 1.5799545726515354e-05,
+      "loss": 1.1914,
+      "step": 3797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3033505603192808,
+      "learning_rate": 1.5797293992124964e-05,
+      "loss": 1.1272,
+      "step": 3798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.963658764197357,
+      "learning_rate": 1.579504181490834e-05,
+      "loss": 1.1704,
+      "step": 3799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.9347778054359335,
+      "learning_rate": 1.579278919503751e-05,
+      "loss": 1.1685,
+      "step": 3800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0774124240197493,
+      "learning_rate": 1.5790536132684536e-05,
+      "loss": 1.1965,
+      "step": 3801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.6055867751887165,
+      "learning_rate": 1.578828262802152e-05,
+      "loss": 1.1829,
+      "step": 3802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.599717376437212,
+      "learning_rate": 1.5786028681220597e-05,
+      "loss": 1.2329,
+      "step": 3803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8328695884017514,
+      "learning_rate": 1.5783774292453936e-05,
+      "loss": 1.1689,
+      "step": 3804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.710532652206074,
+      "learning_rate": 1.5781519461893734e-05,
+      "loss": 1.1309,
+      "step": 3805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.511763431897546,
+      "learning_rate": 1.5779264189712234e-05,
+      "loss": 1.1609,
+      "step": 3806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7188319100210028,
+      "learning_rate": 1.5777008476081695e-05,
+      "loss": 1.1638,
+      "step": 3807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.734871898471119,
+      "learning_rate": 1.5774752321174428e-05,
+      "loss": 1.1667,
+      "step": 3808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.342577427442388,
+      "learning_rate": 1.577249572516277e-05,
+      "loss": 1.166,
+      "step": 3809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.749368925896805,
+      "learning_rate": 1.5770238688219078e-05,
+      "loss": 1.2222,
+      "step": 3810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5705016054003815,
+      "learning_rate": 1.5767981210515772e-05,
+      "loss": 1.1953,
+      "step": 3811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3213583543060996,
+      "learning_rate": 1.576572329222528e-05,
+      "loss": 1.1646,
+      "step": 3812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2908939081543664,
+      "learning_rate": 1.5763464933520077e-05,
+      "loss": 1.1667,
+      "step": 3813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2126892486179695,
+      "learning_rate": 1.5761206134572665e-05,
+      "loss": 1.1765,
+      "step": 3814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4309777872560194,
+      "learning_rate": 1.5758946895555584e-05,
+      "loss": 1.1934,
+      "step": 3815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1861821271533453,
+      "learning_rate": 1.5756687216641404e-05,
+      "loss": 1.165,
+      "step": 3816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.745185613815299,
+      "learning_rate": 1.5754427098002735e-05,
+      "loss": 1.2627,
+      "step": 3817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.628437797854973,
+      "learning_rate": 1.5752166539812212e-05,
+      "loss": 1.2024,
+      "step": 3818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1513479301353646,
+      "learning_rate": 1.5749905542242507e-05,
+      "loss": 1.1758,
+      "step": 3819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.682950568959227,
+      "learning_rate": 1.5747644105466327e-05,
+      "loss": 1.1201,
+      "step": 3820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.910156489378644,
+      "learning_rate": 1.5745382229656417e-05,
+      "loss": 1.1794,
+      "step": 3821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3456952336558436,
+      "learning_rate": 1.5743119914985546e-05,
+      "loss": 1.0986,
+      "step": 3822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6305764859145255,
+      "learning_rate": 1.574085716162652e-05,
+      "loss": 1.1816,
+      "step": 3823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0600651503109666,
+      "learning_rate": 1.5738593969752178e-05,
+      "loss": 1.1829,
+      "step": 3824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.053480236813547,
+      "learning_rate": 1.5736330339535402e-05,
+      "loss": 1.1685,
+      "step": 3825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5638389908812487,
+      "learning_rate": 1.573406627114909e-05,
+      "loss": 1.2229,
+      "step": 3826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1050829308837713,
+      "learning_rate": 1.573180176476619e-05,
+      "loss": 1.2131,
+      "step": 3827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7755965321565257,
+      "learning_rate": 1.572953682055967e-05,
+      "loss": 1.1487,
+      "step": 3828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5232214697723934,
+      "learning_rate": 1.572727143870254e-05,
+      "loss": 1.1086,
+      "step": 3829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.053507831027811,
+      "learning_rate": 1.5725005619367845e-05,
+      "loss": 1.166,
+      "step": 3830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.088744166942843,
+      "learning_rate": 1.5722739362728654e-05,
+      "loss": 1.147,
+      "step": 3831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.149639962359315,
+      "learning_rate": 1.5720472668958076e-05,
+      "loss": 1.2378,
+      "step": 3832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8400658081378025,
+      "learning_rate": 1.571820553822926e-05,
+      "loss": 1.0923,
+      "step": 3833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.678701843593263,
+      "learning_rate": 1.571593797071537e-05,
+      "loss": 1.2078,
+      "step": 3834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.7921345201555265,
+      "learning_rate": 1.571366996658962e-05,
+      "loss": 1.1643,
+      "step": 3835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.321390683978188,
+      "learning_rate": 1.571140152602525e-05,
+      "loss": 1.1511,
+      "step": 3836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3604603957356463,
+      "learning_rate": 1.5709132649195537e-05,
+      "loss": 1.1069,
+      "step": 3837
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.221638490016033,
+      "learning_rate": 1.5706863336273783e-05,
+      "loss": 1.207,
+      "step": 3838
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3130955544027545,
+      "learning_rate": 1.570459358743334e-05,
+      "loss": 1.2136,
+      "step": 3839
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.674098644583947,
+      "learning_rate": 1.570232340284757e-05,
+      "loss": 1.2004,
+      "step": 3840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.397091162852552,
+      "learning_rate": 1.570005278268989e-05,
+      "loss": 1.1069,
+      "step": 3841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8858953311064233,
+      "learning_rate": 1.5697781727133735e-05,
+      "loss": 1.1326,
+      "step": 3842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8802370042483045,
+      "learning_rate": 1.569551023635259e-05,
+      "loss": 1.137,
+      "step": 3843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.451393014125637,
+      "learning_rate": 1.5693238310519952e-05,
+      "loss": 1.2495,
+      "step": 3844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1094929146213817,
+      "learning_rate": 1.5690965949809364e-05,
+      "loss": 1.1169,
+      "step": 3845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.379162140576505,
+      "learning_rate": 1.5688693154394406e-05,
+      "loss": 1.1973,
+      "step": 3846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3632578613606663,
+      "learning_rate": 1.568641992444868e-05,
+      "loss": 1.1677,
+      "step": 3847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.248192710277728,
+      "learning_rate": 1.568414626014583e-05,
+      "loss": 1.2158,
+      "step": 3848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.598239984513208,
+      "learning_rate": 1.5681872161659528e-05,
+      "loss": 1.189,
+      "step": 3849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.695545344384765,
+      "learning_rate": 1.567959762916348e-05,
+      "loss": 1.1831,
+      "step": 3850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.056332674601339,
+      "learning_rate": 1.5677322662831426e-05,
+      "loss": 1.1868,
+      "step": 3851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.061248231797099,
+      "learning_rate": 1.5675047262837146e-05,
+      "loss": 1.176,
+      "step": 3852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.645862075532294,
+      "learning_rate": 1.5672771429354436e-05,
+      "loss": 1.1677,
+      "step": 3853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2944542436954705,
+      "learning_rate": 1.5670495162557144e-05,
+      "loss": 1.1599,
+      "step": 3854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.161470940901642,
+      "learning_rate": 1.5668218462619136e-05,
+      "loss": 1.1631,
+      "step": 3855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0410960080554847,
+      "learning_rate": 1.566594132971432e-05,
+      "loss": 1.2102,
+      "step": 3856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.019237451738,
+      "learning_rate": 1.566366376401664e-05,
+      "loss": 1.1445,
+      "step": 3857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.371883030575601,
+      "learning_rate": 1.5661385765700062e-05,
+      "loss": 1.1682,
+      "step": 3858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5977662294375112,
+      "learning_rate": 1.5659107334938592e-05,
+      "loss": 1.1826,
+      "step": 3859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.936188070685322,
+      "learning_rate": 1.5656828471906265e-05,
+      "loss": 1.2148,
+      "step": 3860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.490311231257937,
+      "learning_rate": 1.565454917677716e-05,
+      "loss": 1.179,
+      "step": 3861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5458119847267944,
+      "learning_rate": 1.5652269449725375e-05,
+      "loss": 1.2827,
+      "step": 3862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.57067336376215,
+      "learning_rate": 1.5649989290925047e-05,
+      "loss": 1.0845,
+      "step": 3863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.965824683713408,
+      "learning_rate": 1.5647708700550347e-05,
+      "loss": 1.1997,
+      "step": 3864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3762611544935495,
+      "learning_rate": 1.5645427678775478e-05,
+      "loss": 1.228,
+      "step": 3865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0730597629321723,
+      "learning_rate": 1.5643146225774676e-05,
+      "loss": 1.1484,
+      "step": 3866
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0582304508264038,
+      "learning_rate": 1.564086434172221e-05,
+      "loss": 1.1936,
+      "step": 3867
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.58421800771794,
+      "learning_rate": 1.5638582026792378e-05,
+      "loss": 1.1716,
+      "step": 3868
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.722693915684678,
+      "learning_rate": 1.563629928115952e-05,
+      "loss": 1.1414,
+      "step": 3869
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4337525077426783,
+      "learning_rate": 1.5634016104998004e-05,
+      "loss": 1.1978,
+      "step": 3870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.920219322906947,
+      "learning_rate": 1.5631732498482222e-05,
+      "loss": 1.1995,
+      "step": 3871
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0475060258832967,
+      "learning_rate": 1.562944846178662e-05,
+      "loss": 1.2351,
+      "step": 3872
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.714707453550901,
+      "learning_rate": 1.562716399508565e-05,
+      "loss": 1.2183,
+      "step": 3873
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5932777819512247,
+      "learning_rate": 1.562487909855382e-05,
+      "loss": 1.2466,
+      "step": 3874
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0392560416106695,
+      "learning_rate": 1.562259377236566e-05,
+      "loss": 1.1455,
+      "step": 3875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.278622822387891,
+      "learning_rate": 1.5620308016695736e-05,
+      "loss": 1.1448,
+      "step": 3876
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2913944581711543,
+      "learning_rate": 1.5618021831718642e-05,
+      "loss": 1.2227,
+      "step": 3877
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.286607280107666,
+      "learning_rate": 1.5615735217609008e-05,
+      "loss": 1.2568,
+      "step": 3878
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1907287416046044,
+      "learning_rate": 1.5613448174541503e-05,
+      "loss": 1.229,
+      "step": 3879
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.17635180552228,
+      "learning_rate": 1.561116070269082e-05,
+      "loss": 1.1904,
+      "step": 3880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6081924960430114,
+      "learning_rate": 1.560887280223168e-05,
+      "loss": 1.1465,
+      "step": 3881
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.927228056921284,
+      "learning_rate": 1.5606584473338857e-05,
+      "loss": 1.1943,
+      "step": 3882
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1036175638481804,
+      "learning_rate": 1.5604295716187135e-05,
+      "loss": 1.135,
+      "step": 3883
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4706252525531247,
+      "learning_rate": 1.5602006530951345e-05,
+      "loss": 1.1211,
+      "step": 3884
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9036568574275594,
+      "learning_rate": 1.5599716917806344e-05,
+      "loss": 1.1692,
+      "step": 3885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8574815482820313,
+      "learning_rate": 1.559742687692703e-05,
+      "loss": 1.231,
+      "step": 3886
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9376507702013397,
+      "learning_rate": 1.5595136408488318e-05,
+      "loss": 1.2129,
+      "step": 3887
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.798462000589983,
+      "learning_rate": 1.5592845512665174e-05,
+      "loss": 1.1699,
+      "step": 3888
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.413147978157393,
+      "learning_rate": 1.5590554189632585e-05,
+      "loss": 1.1638,
+      "step": 3889
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.22202599228991,
+      "learning_rate": 1.558826243956557e-05,
+      "loss": 1.1978,
+      "step": 3890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.521111207804071,
+      "learning_rate": 1.5585970262639188e-05,
+      "loss": 1.1931,
+      "step": 3891
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7953059969117424,
+      "learning_rate": 1.5583677659028535e-05,
+      "loss": 1.1809,
+      "step": 3892
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5449698585488965,
+      "learning_rate": 1.5581384628908715e-05,
+      "loss": 1.1768,
+      "step": 3893
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.39987892215128,
+      "learning_rate": 1.5579091172454888e-05,
+      "loss": 1.209,
+      "step": 3894
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.166893800924013,
+      "learning_rate": 1.557679728984225e-05,
+      "loss": 1.23,
+      "step": 3895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2975801019263415,
+      "learning_rate": 1.5574502981246006e-05,
+      "loss": 1.1863,
+      "step": 3896
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4351571884321,
+      "learning_rate": 1.5572208246841408e-05,
+      "loss": 1.2358,
+      "step": 3897
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2181034387794876,
+      "learning_rate": 1.5569913086803744e-05,
+      "loss": 1.1941,
+      "step": 3898
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4321740535925773,
+      "learning_rate": 1.5567617501308328e-05,
+      "loss": 1.1853,
+      "step": 3899
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2994375608882502,
+      "learning_rate": 1.5565321490530507e-05,
+      "loss": 1.1777,
+      "step": 3900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.424372860094893,
+      "learning_rate": 1.5563025054645665e-05,
+      "loss": 1.1704,
+      "step": 3901
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.478139774609487,
+      "learning_rate": 1.5560728193829214e-05,
+      "loss": 1.0525,
+      "step": 3902
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.17258061217502,
+      "learning_rate": 1.5558430908256596e-05,
+      "loss": 1.1702,
+      "step": 3903
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.999373012516928,
+      "learning_rate": 1.5556133198103292e-05,
+      "loss": 1.175,
+      "step": 3904
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.390338267972673,
+      "learning_rate": 1.5553835063544814e-05,
+      "loss": 1.1675,
+      "step": 3905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.312752462159177,
+      "learning_rate": 1.5551536504756705e-05,
+      "loss": 1.1985,
+      "step": 3906
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.421215852575987,
+      "learning_rate": 1.5549237521914537e-05,
+      "loss": 1.1841,
+      "step": 3907
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.714914990595008,
+      "learning_rate": 1.5546938115193922e-05,
+      "loss": 1.1851,
+      "step": 3908
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9741584685924995,
+      "learning_rate": 1.5544638284770496e-05,
+      "loss": 1.2261,
+      "step": 3909
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0919695240886798,
+      "learning_rate": 1.5542338030819933e-05,
+      "loss": 1.2058,
+      "step": 3910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3150120948315367,
+      "learning_rate": 1.554003735351794e-05,
+      "loss": 1.0955,
+      "step": 3911
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.198555411223223,
+      "learning_rate": 1.553773625304025e-05,
+      "loss": 1.2227,
+      "step": 3912
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2204475691819114,
+      "learning_rate": 1.5535434729562637e-05,
+      "loss": 1.2246,
+      "step": 3913
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1244628896175084,
+      "learning_rate": 1.55331327832609e-05,
+      "loss": 1.1814,
+      "step": 3914
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5180430026999963,
+      "learning_rate": 1.5530830414310875e-05,
+      "loss": 1.2703,
+      "step": 3915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2604067417958307,
+      "learning_rate": 1.552852762288843e-05,
+      "loss": 1.2698,
+      "step": 3916
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.673151485517769,
+      "learning_rate": 1.5526224409169456e-05,
+      "loss": 1.2456,
+      "step": 3917
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9534211041840295,
+      "learning_rate": 1.5523920773329894e-05,
+      "loss": 1.0727,
+      "step": 3918
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1485792617194592,
+      "learning_rate": 1.5521616715545696e-05,
+      "loss": 1.1785,
+      "step": 3919
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.014357560446517,
+      "learning_rate": 1.551931223599287e-05,
+      "loss": 1.1611,
+      "step": 3920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7417361492207197,
+      "learning_rate": 1.5517007334847435e-05,
+      "loss": 1.1624,
+      "step": 3921
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7101232884722335,
+      "learning_rate": 1.5514702012285455e-05,
+      "loss": 1.2349,
+      "step": 3922
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.843441412712804,
+      "learning_rate": 1.5512396268483017e-05,
+      "loss": 1.1692,
+      "step": 3923
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0992594784651275,
+      "learning_rate": 1.551009010361625e-05,
+      "loss": 1.2302,
+      "step": 3924
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.88753219717233,
+      "learning_rate": 1.5507783517861308e-05,
+      "loss": 1.1665,
+      "step": 3925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.149142630213589,
+      "learning_rate": 1.5505476511394384e-05,
+      "loss": 1.1909,
+      "step": 3926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2744924839235616,
+      "learning_rate": 1.5503169084391688e-05,
+      "loss": 1.2253,
+      "step": 3927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.999361347491756,
+      "learning_rate": 1.5500861237029488e-05,
+      "loss": 1.178,
+      "step": 3928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.537446518217187,
+      "learning_rate": 1.5498552969484058e-05,
+      "loss": 1.1982,
+      "step": 3929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3517976443014175,
+      "learning_rate": 1.5496244281931718e-05,
+      "loss": 1.2141,
+      "step": 3930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1637177347948193,
+      "learning_rate": 1.5493935174548815e-05,
+      "loss": 1.113,
+      "step": 3931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8587289848108246,
+      "learning_rate": 1.5491625647511732e-05,
+      "loss": 1.1716,
+      "step": 3932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.325948264036177,
+      "learning_rate": 1.5489315700996883e-05,
+      "loss": 1.2029,
+      "step": 3933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.978641462950956,
+      "learning_rate": 1.548700533518071e-05,
+      "loss": 1.2427,
+      "step": 3934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8080375243209246,
+      "learning_rate": 1.54846945502397e-05,
+      "loss": 1.1877,
+      "step": 3935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5765432974697555,
+      "learning_rate": 1.548238334635035e-05,
+      "loss": 1.1921,
+      "step": 3936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0691308474315173,
+      "learning_rate": 1.5480071723689204e-05,
+      "loss": 1.13,
+      "step": 3937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.00448431137849,
+      "learning_rate": 1.547775968243284e-05,
+      "loss": 1.2251,
+      "step": 3938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6868122808718136,
+      "learning_rate": 1.547544722275786e-05,
+      "loss": 1.1646,
+      "step": 3939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.846916064359686,
+      "learning_rate": 1.5473134344840905e-05,
+      "loss": 1.1682,
+      "step": 3940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.528180971807159,
+      "learning_rate": 1.5470821048858638e-05,
+      "loss": 1.1323,
+      "step": 3941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.973727278391229,
+      "learning_rate": 1.5468507334987763e-05,
+      "loss": 1.166,
+      "step": 3942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.261920652450822,
+      "learning_rate": 1.5466193203405017e-05,
+      "loss": 1.175,
+      "step": 3943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.796104457334337,
+      "learning_rate": 1.5463878654287157e-05,
+      "loss": 1.1577,
+      "step": 3944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1673233455759386,
+      "learning_rate": 1.5461563687810987e-05,
+      "loss": 1.1646,
+      "step": 3945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.673805610092405,
+      "learning_rate": 1.545924830415333e-05,
+      "loss": 1.2131,
+      "step": 3946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1979306954562556,
+      "learning_rate": 1.545693250349105e-05,
+      "loss": 1.2124,
+      "step": 3947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9202169593766634,
+      "learning_rate": 1.5454616286001037e-05,
+      "loss": 1.1011,
+      "step": 3948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9863213846756183,
+      "learning_rate": 1.5452299651860217e-05,
+      "loss": 1.2588,
+      "step": 3949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.056913745237371,
+      "learning_rate": 1.544998260124555e-05,
+      "loss": 1.2373,
+      "step": 3950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5635704998808064,
+      "learning_rate": 1.5447665134334017e-05,
+      "loss": 1.1086,
+      "step": 3951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.652626186685396,
+      "learning_rate": 1.5445347251302642e-05,
+      "loss": 1.1382,
+      "step": 3952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3383525545009873,
+      "learning_rate": 1.5443028952328474e-05,
+      "loss": 1.1399,
+      "step": 3953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4686738451688104,
+      "learning_rate": 1.5440710237588598e-05,
+      "loss": 1.1331,
+      "step": 3954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.971138867616636,
+      "learning_rate": 1.5438391107260128e-05,
+      "loss": 1.1111,
+      "step": 3955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.104096600449298,
+      "learning_rate": 1.5436071561520208e-05,
+      "loss": 1.2515,
+      "step": 3956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0990899506502623,
+      "learning_rate": 1.5433751600546025e-05,
+      "loss": 1.2175,
+      "step": 3957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.882372105953668,
+      "learning_rate": 1.5431431224514777e-05,
+      "loss": 1.1062,
+      "step": 3958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.707846448267828,
+      "learning_rate": 1.542911043360372e-05,
+      "loss": 1.1721,
+      "step": 3959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1125782621049525,
+      "learning_rate": 1.5426789227990115e-05,
+      "loss": 1.2256,
+      "step": 3960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7736283972013127,
+      "learning_rate": 1.5424467607851274e-05,
+      "loss": 1.1516,
+      "step": 3961
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8355334828115097,
+      "learning_rate": 1.542214557336453e-05,
+      "loss": 1.156,
+      "step": 3962
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0864232817200525,
+      "learning_rate": 1.541982312470726e-05,
+      "loss": 1.2175,
+      "step": 3963
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6187725976663745,
+      "learning_rate": 1.541750026205685e-05,
+      "loss": 1.1619,
+      "step": 3964
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3807658167950603,
+      "learning_rate": 1.5415176985590745e-05,
+      "loss": 1.1799,
+      "step": 3965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3315572479061686,
+      "learning_rate": 1.5412853295486404e-05,
+      "loss": 1.175,
+      "step": 3966
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.275170474225825,
+      "learning_rate": 1.541052919192132e-05,
+      "loss": 1.187,
+      "step": 3967
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9038622809322603,
+      "learning_rate": 1.540820467507302e-05,
+      "loss": 1.2112,
+      "step": 3968
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.333174500640893,
+      "learning_rate": 1.5405879745119065e-05,
+      "loss": 1.2102,
+      "step": 3969
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3034560901349086,
+      "learning_rate": 1.540355440223704e-05,
+      "loss": 1.2251,
+      "step": 3970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9387307811404844,
+      "learning_rate": 1.5401228646604573e-05,
+      "loss": 1.2622,
+      "step": 3971
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2290748944303544,
+      "learning_rate": 1.5398902478399314e-05,
+      "loss": 1.2322,
+      "step": 3972
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.627503508992444,
+      "learning_rate": 1.539657589779894e-05,
+      "loss": 1.1836,
+      "step": 3973
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.107408333814214,
+      "learning_rate": 1.539424890498118e-05,
+      "loss": 1.176,
+      "step": 3974
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3385234685410277,
+      "learning_rate": 1.539192150012377e-05,
+      "loss": 1.1401,
+      "step": 3975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4086797307208903,
+      "learning_rate": 1.53895936834045e-05,
+      "loss": 1.1233,
+      "step": 3976
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2338302382165365,
+      "learning_rate": 1.5387265455001168e-05,
+      "loss": 1.1287,
+      "step": 3977
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.357586640247916,
+      "learning_rate": 1.5384936815091624e-05,
+      "loss": 1.2185,
+      "step": 3978
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.253141389581297,
+      "learning_rate": 1.5382607763853736e-05,
+      "loss": 1.1228,
+      "step": 3979
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1035770153014015,
+      "learning_rate": 1.538027830146541e-05,
+      "loss": 1.1189,
+      "step": 3980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6743109555378277,
+      "learning_rate": 1.537794842810459e-05,
+      "loss": 1.2124,
+      "step": 3981
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.730862577494044,
+      "learning_rate": 1.537561814394923e-05,
+      "loss": 1.1448,
+      "step": 3982
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3686864581345706,
+      "learning_rate": 1.5373287449177335e-05,
+      "loss": 1.1179,
+      "step": 3983
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1343220613816385,
+      "learning_rate": 1.5370956343966936e-05,
+      "loss": 1.1538,
+      "step": 3984
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.662469967763827,
+      "learning_rate": 1.5368624828496098e-05,
+      "loss": 1.2378,
+      "step": 3985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.070839555319355,
+      "learning_rate": 1.5366292902942906e-05,
+      "loss": 1.1204,
+      "step": 3986
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3939651678731355,
+      "learning_rate": 1.5363960567485486e-05,
+      "loss": 1.2268,
+      "step": 3987
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4541818448611434,
+      "learning_rate": 1.5361627822301997e-05,
+      "loss": 1.187,
+      "step": 3988
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6245048111225864,
+      "learning_rate": 1.5359294667570624e-05,
+      "loss": 1.0972,
+      "step": 3989
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0584185637347723,
+      "learning_rate": 1.5356961103469583e-05,
+      "loss": 1.1904,
+      "step": 3990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.131955067665045,
+      "learning_rate": 1.5354627130177123e-05,
+      "loss": 1.2837,
+      "step": 3991
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2868057348446067,
+      "learning_rate": 1.5352292747871532e-05,
+      "loss": 1.1724,
+      "step": 3992
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0489470924807316,
+      "learning_rate": 1.5349957956731113e-05,
+      "loss": 1.2458,
+      "step": 3993
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.571400746786345,
+      "learning_rate": 1.5347622756934214e-05,
+      "loss": 1.1909,
+      "step": 3994
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.635728207166066,
+      "learning_rate": 1.5345287148659207e-05,
+      "loss": 1.219,
+      "step": 3995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1751758477433953,
+      "learning_rate": 1.5342951132084496e-05,
+      "loss": 1.1345,
+      "step": 3996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.107909166840833,
+      "learning_rate": 1.534061470738852e-05,
+      "loss": 1.1672,
+      "step": 3997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.659699159811194,
+      "learning_rate": 1.5338277874749748e-05,
+      "loss": 1.2227,
+      "step": 3998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1338762415037706,
+      "learning_rate": 1.5335940634346677e-05,
+      "loss": 1.1851,
+      "step": 3999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.085933278856529,
+      "learning_rate": 1.5333602986357833e-05,
+      "loss": 1.1743,
+      "step": 4000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.426396860079655,
+      "learning_rate": 1.5331264930961788e-05,
+      "loss": 1.179,
+      "step": 4001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3186569492307116,
+      "learning_rate": 1.5328926468337123e-05,
+      "loss": 1.1621,
+      "step": 4002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2672043718849215,
+      "learning_rate": 1.5326587598662468e-05,
+      "loss": 1.1855,
+      "step": 4003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9448916943485277,
+      "learning_rate": 1.532424832211648e-05,
+      "loss": 1.2168,
+      "step": 4004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3591282972970236,
+      "learning_rate": 1.5321908638877834e-05,
+      "loss": 1.1924,
+      "step": 4005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9222826999973166,
+      "learning_rate": 1.5319568549125253e-05,
+      "loss": 1.2034,
+      "step": 4006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5093912151540376,
+      "learning_rate": 1.5317228053037486e-05,
+      "loss": 1.2004,
+      "step": 4007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.910448155708376,
+      "learning_rate": 1.5314887150793314e-05,
+      "loss": 1.1924,
+      "step": 4008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.828466573674339,
+      "learning_rate": 1.531254584257154e-05,
+      "loss": 1.16,
+      "step": 4009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.71957535725753,
+      "learning_rate": 1.531020412855101e-05,
+      "loss": 1.1926,
+      "step": 4010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.740188291704029,
+      "learning_rate": 1.5307862008910592e-05,
+      "loss": 1.1646,
+      "step": 4011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7943359556576066,
+      "learning_rate": 1.5305519483829193e-05,
+      "loss": 1.1868,
+      "step": 4012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9132063491020537,
+      "learning_rate": 1.5303176553485745e-05,
+      "loss": 1.2314,
+      "step": 4013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.793509079028712,
+      "learning_rate": 1.530083321805921e-05,
+      "loss": 1.1399,
+      "step": 4014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.54461456233078,
+      "learning_rate": 1.5298489477728587e-05,
+      "loss": 1.189,
+      "step": 4015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0385562996425604,
+      "learning_rate": 1.52961453326729e-05,
+      "loss": 1.2317,
+      "step": 4016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.483818636297825,
+      "learning_rate": 1.5293800783071212e-05,
+      "loss": 1.2263,
+      "step": 4017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0641711461193846,
+      "learning_rate": 1.5291455829102603e-05,
+      "loss": 1.1809,
+      "step": 4018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8590885700273927,
+      "learning_rate": 1.52891104709462e-05,
+      "loss": 1.2222,
+      "step": 4019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3856555555189947,
+      "learning_rate": 1.5286764708781143e-05,
+      "loss": 1.2217,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.450749137746395,
+      "learning_rate": 1.5284418542786625e-05,
+      "loss": 1.1619,
+      "step": 4021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0478549726247732,
+      "learning_rate": 1.5282071973141853e-05,
+      "loss": 1.21,
+      "step": 4022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0388740841836133,
+      "learning_rate": 1.5279725000026072e-05,
+      "loss": 1.1377,
+      "step": 4023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.057553511486213,
+      "learning_rate": 1.527737762361855e-05,
+      "loss": 1.1914,
+      "step": 4024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.511836461687568,
+      "learning_rate": 1.5275029844098593e-05,
+      "loss": 1.1384,
+      "step": 4025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0535350650711677,
+      "learning_rate": 1.5272681661645538e-05,
+      "loss": 1.1975,
+      "step": 4026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2060516542158326,
+      "learning_rate": 1.527033307643875e-05,
+      "loss": 1.2268,
+      "step": 4027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.171617125950602,
+      "learning_rate": 1.5267984088657627e-05,
+      "loss": 1.1475,
+      "step": 4028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.182018967228267,
+      "learning_rate": 1.526563469848159e-05,
+      "loss": 1.1799,
+      "step": 4029
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0288164359161382,
+      "learning_rate": 1.5263284906090107e-05,
+      "loss": 1.1545,
+      "step": 4030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0096103823428675,
+      "learning_rate": 1.5260934711662664e-05,
+      "loss": 1.2053,
+      "step": 4031
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.185670255321692,
+      "learning_rate": 1.5258584115378779e-05,
+      "loss": 1.1418,
+      "step": 4032
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4977516405691644,
+      "learning_rate": 1.5256233117418e-05,
+      "loss": 1.2148,
+      "step": 4033
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.817356485181005,
+      "learning_rate": 1.5253881717959908e-05,
+      "loss": 1.1694,
+      "step": 4034
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3009241039354964,
+      "learning_rate": 1.525152991718412e-05,
+      "loss": 1.1277,
+      "step": 4035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0196680655456563,
+      "learning_rate": 1.5249177715270276e-05,
+      "loss": 1.1897,
+      "step": 4036
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.557640446684515,
+      "learning_rate": 1.5246825112398043e-05,
+      "loss": 1.1732,
+      "step": 4037
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1466904296866853,
+      "learning_rate": 1.5244472108747134e-05,
+      "loss": 1.2173,
+      "step": 4038
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.975134296730037,
+      "learning_rate": 1.5242118704497277e-05,
+      "loss": 1.1594,
+      "step": 4039
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8656725695334466,
+      "learning_rate": 1.5239764899828237e-05,
+      "loss": 1.1936,
+      "step": 4040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.434521598592507,
+      "learning_rate": 1.5237410694919814e-05,
+      "loss": 1.2612,
+      "step": 4041
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.784995276191725,
+      "learning_rate": 1.523505608995183e-05,
+      "loss": 1.1641,
+      "step": 4042
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.661083446622254,
+      "learning_rate": 1.5232701085104138e-05,
+      "loss": 1.2319,
+      "step": 4043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.103602266512657,
+      "learning_rate": 1.5230345680556631e-05,
+      "loss": 1.2109,
+      "step": 4044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.720437518286369,
+      "learning_rate": 1.5227989876489227e-05,
+      "loss": 1.2896,
+      "step": 4045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2054719283452875,
+      "learning_rate": 1.5225633673081873e-05,
+      "loss": 1.1787,
+      "step": 4046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5095579943694912,
+      "learning_rate": 1.5223277070514542e-05,
+      "loss": 1.1592,
+      "step": 4047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.15038181842914,
+      "learning_rate": 1.522092006896725e-05,
+      "loss": 1.1895,
+      "step": 4048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.570601404610452,
+      "learning_rate": 1.5218562668620034e-05,
+      "loss": 1.1477,
+      "step": 4049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.401600389566359,
+      "learning_rate": 1.5216204869652961e-05,
+      "loss": 1.1997,
+      "step": 4050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.197727339269136,
+      "learning_rate": 1.5213846672246139e-05,
+      "loss": 1.074,
+      "step": 4051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.292300092640462,
+      "learning_rate": 1.5211488076579692e-05,
+      "loss": 1.1787,
+      "step": 4052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.978776986740103,
+      "learning_rate": 1.520912908283378e-05,
+      "loss": 1.1558,
+      "step": 4053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.791175335765857,
+      "learning_rate": 1.5206769691188604e-05,
+      "loss": 1.1931,
+      "step": 4054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.133945892679697,
+      "learning_rate": 1.5204409901824377e-05,
+      "loss": 1.1062,
+      "step": 4055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8201283615782073,
+      "learning_rate": 1.5202049714921356e-05,
+      "loss": 1.188,
+      "step": 4056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.146701602157832,
+      "learning_rate": 1.5199689130659822e-05,
+      "loss": 1.1562,
+      "step": 4057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.740048628566107,
+      "learning_rate": 1.519732814922009e-05,
+      "loss": 1.1887,
+      "step": 4058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3561737097994873,
+      "learning_rate": 1.5194966770782502e-05,
+      "loss": 1.1667,
+      "step": 4059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.743203748564278,
+      "learning_rate": 1.5192604995527433e-05,
+      "loss": 1.1533,
+      "step": 4060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.387850378140622,
+      "learning_rate": 1.5190242823635286e-05,
+      "loss": 1.178,
+      "step": 4061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.437207026783555,
+      "learning_rate": 1.5187880255286495e-05,
+      "loss": 1.23,
+      "step": 4062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.100172789073744,
+      "learning_rate": 1.5185517290661526e-05,
+      "loss": 1.1331,
+      "step": 4063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.615193721505569,
+      "learning_rate": 1.5183153929940875e-05,
+      "loss": 1.2285,
+      "step": 4064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4397102695240673,
+      "learning_rate": 1.5180790173305067e-05,
+      "loss": 1.1655,
+      "step": 4065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.636065431559556,
+      "learning_rate": 1.5178426020934653e-05,
+      "loss": 1.165,
+      "step": 4066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9947608052266865,
+      "learning_rate": 1.5176061473010223e-05,
+      "loss": 1.1541,
+      "step": 4067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.231634932402109,
+      "learning_rate": 1.5173696529712397e-05,
+      "loss": 1.2109,
+      "step": 4068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8053889221843655,
+      "learning_rate": 1.5171331191221814e-05,
+      "loss": 1.1147,
+      "step": 4069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.5417778068707175,
+      "learning_rate": 1.516896545771915e-05,
+      "loss": 1.1934,
+      "step": 4070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3871878951578203,
+      "learning_rate": 1.5166599329385116e-05,
+      "loss": 1.144,
+      "step": 4071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.943780750704524,
+      "learning_rate": 1.5164232806400447e-05,
+      "loss": 1.1614,
+      "step": 4072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.678261330365763,
+      "learning_rate": 1.516186588894591e-05,
+      "loss": 1.2822,
+      "step": 4073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.372723534572633,
+      "learning_rate": 1.5159498577202303e-05,
+      "loss": 1.1367,
+      "step": 4074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5620673894943935,
+      "learning_rate": 1.5157130871350451e-05,
+      "loss": 1.135,
+      "step": 4075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.957994873109021,
+      "learning_rate": 1.5154762771571213e-05,
+      "loss": 1.1958,
+      "step": 4076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.673856437883605,
+      "learning_rate": 1.5152394278045473e-05,
+      "loss": 1.1418,
+      "step": 4077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.492022944732827,
+      "learning_rate": 1.5150025390954153e-05,
+      "loss": 1.1116,
+      "step": 4078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.103258453049473,
+      "learning_rate": 1.5147656110478198e-05,
+      "loss": 1.1914,
+      "step": 4079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4750202936016885,
+      "learning_rate": 1.5145286436798583e-05,
+      "loss": 1.2205,
+      "step": 4080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0782522755235573,
+      "learning_rate": 1.5142916370096324e-05,
+      "loss": 1.1875,
+      "step": 4081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.247467126118866,
+      "learning_rate": 1.5140545910552446e-05,
+      "loss": 1.1409,
+      "step": 4082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.531494463812783,
+      "learning_rate": 1.5138175058348026e-05,
+      "loss": 1.2241,
+      "step": 4083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.025571312104703,
+      "learning_rate": 1.5135803813664159e-05,
+      "loss": 1.2422,
+      "step": 4084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.275783641488406,
+      "learning_rate": 1.5133432176681973e-05,
+      "loss": 1.1975,
+      "step": 4085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9118744419912153,
+      "learning_rate": 1.5131060147582622e-05,
+      "loss": 1.127,
+      "step": 4086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.14810906867214,
+      "learning_rate": 1.5128687726547297e-05,
+      "loss": 1.1995,
+      "step": 4087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4680216617687747,
+      "learning_rate": 1.5126314913757218e-05,
+      "loss": 1.1814,
+      "step": 4088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.61652501107598,
+      "learning_rate": 1.5123941709393624e-05,
+      "loss": 1.1821,
+      "step": 4089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3768552861952355,
+      "learning_rate": 1.5121568113637796e-05,
+      "loss": 1.0913,
+      "step": 4090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7836670647701025,
+      "learning_rate": 1.5119194126671046e-05,
+      "loss": 1.1707,
+      "step": 4091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9358656635429994,
+      "learning_rate": 1.5116819748674705e-05,
+      "loss": 1.1536,
+      "step": 4092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8698147867212422,
+      "learning_rate": 1.5114444979830145e-05,
+      "loss": 1.166,
+      "step": 4093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8294735754200357,
+      "learning_rate": 1.5112069820318755e-05,
+      "loss": 1.0974,
+      "step": 4094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8872022883914936,
+      "learning_rate": 1.5109694270321968e-05,
+      "loss": 1.1826,
+      "step": 4095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3418833727263095,
+      "learning_rate": 1.5107318330021237e-05,
+      "loss": 1.1995,
+      "step": 4096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9504352209702405,
+      "learning_rate": 1.5104941999598052e-05,
+      "loss": 1.1619,
+      "step": 4097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.66688807992128,
+      "learning_rate": 1.5102565279233928e-05,
+      "loss": 1.1775,
+      "step": 4098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8191602304871712,
+      "learning_rate": 1.5100188169110403e-05,
+      "loss": 1.1777,
+      "step": 4099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0391301377248303,
+      "learning_rate": 1.5097810669409061e-05,
+      "loss": 1.1814,
+      "step": 4100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4589301674659283,
+      "learning_rate": 1.509543278031151e-05,
+      "loss": 1.1709,
+      "step": 4101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.77341810604991,
+      "learning_rate": 1.5093054501999376e-05,
+      "loss": 1.186,
+      "step": 4102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1220308308197544,
+      "learning_rate": 1.5090675834654329e-05,
+      "loss": 1.1719,
+      "step": 4103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8781553603231904,
+      "learning_rate": 1.5088296778458063e-05,
+      "loss": 1.1901,
+      "step": 4104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.879919766091659,
+      "learning_rate": 1.5085917333592299e-05,
+      "loss": 1.2031,
+      "step": 4105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2204636508762383,
+      "learning_rate": 1.5083537500238797e-05,
+      "loss": 1.2012,
+      "step": 4106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3555397900742836,
+      "learning_rate": 1.5081157278579332e-05,
+      "loss": 1.2422,
+      "step": 4107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7789541792683967,
+      "learning_rate": 1.5078776668795727e-05,
+      "loss": 1.0918,
+      "step": 4108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.042984510013252,
+      "learning_rate": 1.5076395671069816e-05,
+      "loss": 1.2695,
+      "step": 4109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7164408955062784,
+      "learning_rate": 1.5074014285583479e-05,
+      "loss": 1.2114,
+      "step": 4110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2055755244675215,
+      "learning_rate": 1.5071632512518612e-05,
+      "loss": 1.1943,
+      "step": 4111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.03227876676532,
+      "learning_rate": 1.5069250352057152e-05,
+      "loss": 1.2534,
+      "step": 4112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7866997542273926,
+      "learning_rate": 1.5066867804381055e-05,
+      "loss": 1.1819,
+      "step": 4113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.836673127791523,
+      "learning_rate": 1.5064484869672318e-05,
+      "loss": 1.1414,
+      "step": 4114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.093033470096722,
+      "learning_rate": 1.5062101548112952e-05,
+      "loss": 1.2581,
+      "step": 4115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9172074297106447,
+      "learning_rate": 1.505971783988502e-05,
+      "loss": 1.2241,
+      "step": 4116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.207985356802419,
+      "learning_rate": 1.505733374517059e-05,
+      "loss": 1.1802,
+      "step": 4117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.085052220131492,
+      "learning_rate": 1.5054949264151778e-05,
+      "loss": 1.2803,
+      "step": 4118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2178881461228346,
+      "learning_rate": 1.5052564397010718e-05,
+      "loss": 1.1238,
+      "step": 4119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0794851735310322,
+      "learning_rate": 1.505017914392958e-05,
+      "loss": 1.1626,
+      "step": 4120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3666392906089078,
+      "learning_rate": 1.5047793505090564e-05,
+      "loss": 1.2031,
+      "step": 4121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1186443720408543,
+      "learning_rate": 1.5045407480675892e-05,
+      "loss": 1.1807,
+      "step": 4122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.599129537852733,
+      "learning_rate": 1.5043021070867825e-05,
+      "loss": 1.2102,
+      "step": 4123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2512883288660723,
+      "learning_rate": 1.5040634275848647e-05,
+      "loss": 1.2178,
+      "step": 4124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6915130811703256,
+      "learning_rate": 1.5038247095800673e-05,
+      "loss": 1.2612,
+      "step": 4125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9255256720280856,
+      "learning_rate": 1.5035859530906247e-05,
+      "loss": 1.2373,
+      "step": 4126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0575850554209167,
+      "learning_rate": 1.5033471581347745e-05,
+      "loss": 1.1926,
+      "step": 4127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9831023683530926,
+      "learning_rate": 1.5031083247307572e-05,
+      "loss": 1.166,
+      "step": 4128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3754818509433924,
+      "learning_rate": 1.5028694528968153e-05,
+      "loss": 1.2351,
+      "step": 4129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.136982296935519,
+      "learning_rate": 1.502630542651196e-05,
+      "loss": 1.1858,
+      "step": 4130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0082268102121117,
+      "learning_rate": 1.5023915940121482e-05,
+      "loss": 1.2114,
+      "step": 4131
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6280196045248823,
+      "learning_rate": 1.5021526069979232e-05,
+      "loss": 1.1462,
+      "step": 4132
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.647603908606728,
+      "learning_rate": 1.5019135816267771e-05,
+      "loss": 1.2512,
+      "step": 4133
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8237759306801955,
+      "learning_rate": 1.5016745179169672e-05,
+      "loss": 1.1414,
+      "step": 4134
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9371015812419987,
+      "learning_rate": 1.5014354158867549e-05,
+      "loss": 1.145,
+      "step": 4135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.063589342976226,
+      "learning_rate": 1.5011962755544032e-05,
+      "loss": 1.1951,
+      "step": 4136
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0784314449012795,
+      "learning_rate": 1.5009570969381795e-05,
+      "loss": 1.1924,
+      "step": 4137
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0260767613480075,
+      "learning_rate": 1.5007178800563535e-05,
+      "loss": 1.1599,
+      "step": 4138
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2287808463376533,
+      "learning_rate": 1.5004786249271975e-05,
+      "loss": 1.1606,
+      "step": 4139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1781095836871742,
+      "learning_rate": 1.5002393315689867e-05,
+      "loss": 1.2207,
+      "step": 4140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.864631160441583,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1113,
+      "step": 4141
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.220267578832691,
+      "learning_rate": 1.4997606302385189e-05,
+      "loss": 1.187,
+      "step": 4142
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.985965857104157,
+      "learning_rate": 1.499521222302827e-05,
+      "loss": 1.1606,
+      "step": 4143
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.949451083227622,
+      "learning_rate": 1.4992817762112122e-05,
+      "loss": 1.1499,
+      "step": 4144
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8822048059284597,
+      "learning_rate": 1.4990422919819639e-05,
+      "loss": 1.2043,
+      "step": 4145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.631917335985898,
+      "learning_rate": 1.4988027696333753e-05,
+      "loss": 1.2102,
+      "step": 4146
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7604995287092238,
+      "learning_rate": 1.4985632091837428e-05,
+      "loss": 1.198,
+      "step": 4147
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7866361557309647,
+      "learning_rate": 1.4983236106513647e-05,
+      "loss": 1.1631,
+      "step": 4148
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.735412708854103,
+      "learning_rate": 1.498083974054543e-05,
+      "loss": 1.1887,
+      "step": 4149
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1670193063761154,
+      "learning_rate": 1.4978442994115821e-05,
+      "loss": 1.2554,
+      "step": 4150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.519146913877871,
+      "learning_rate": 1.4976045867407898e-05,
+      "loss": 1.1533,
+      "step": 4151
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6731143279663594,
+      "learning_rate": 1.4973648360604763e-05,
+      "loss": 1.1519,
+      "step": 4152
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.922631844789844,
+      "learning_rate": 1.4971250473889552e-05,
+      "loss": 1.2007,
+      "step": 4153
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3155671695834035,
+      "learning_rate": 1.4968852207445425e-05,
+      "loss": 1.1685,
+      "step": 4154
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0081105795148964,
+      "learning_rate": 1.4966453561455573e-05,
+      "loss": 1.1685,
+      "step": 4155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9919683583000007,
+      "learning_rate": 1.4964054536103221e-05,
+      "loss": 1.1877,
+      "step": 4156
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.464800454044453,
+      "learning_rate": 1.4961655131571614e-05,
+      "loss": 1.041,
+      "step": 4157
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7613261727675544,
+      "learning_rate": 1.4959255348044037e-05,
+      "loss": 1.1475,
+      "step": 4158
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1107882281505894,
+      "learning_rate": 1.4956855185703787e-05,
+      "loss": 1.1726,
+      "step": 4159
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.268755253985661,
+      "learning_rate": 1.4954454644734208e-05,
+      "loss": 1.1709,
+      "step": 4160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.265038462718268,
+      "learning_rate": 1.4952053725318667e-05,
+      "loss": 1.1611,
+      "step": 4161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.611194168929749,
+      "learning_rate": 1.494965242764055e-05,
+      "loss": 1.1274,
+      "step": 4162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4710628626846747,
+      "learning_rate": 1.4947250751883288e-05,
+      "loss": 1.1421,
+      "step": 4163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8161048971508156,
+      "learning_rate": 1.494484869823033e-05,
+      "loss": 1.126,
+      "step": 4164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4921672344286465,
+      "learning_rate": 1.4942446266865154e-05,
+      "loss": 1.1101,
+      "step": 4165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.03311871978335,
+      "learning_rate": 1.4940043457971274e-05,
+      "loss": 1.1055,
+      "step": 4166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1048693586800566,
+      "learning_rate": 1.4937640271732227e-05,
+      "loss": 1.2073,
+      "step": 4167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5031607939535965,
+      "learning_rate": 1.4935236708331582e-05,
+      "loss": 1.1846,
+      "step": 4168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.646078286651607,
+      "learning_rate": 1.493283276795293e-05,
+      "loss": 1.1675,
+      "step": 4169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1637013566284593,
+      "learning_rate": 1.49304284507799e-05,
+      "loss": 1.1387,
+      "step": 4170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8678461794222887,
+      "learning_rate": 1.4928023756996147e-05,
+      "loss": 1.188,
+      "step": 4171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.192795167741564,
+      "learning_rate": 1.492561868678535e-05,
+      "loss": 1.2051,
+      "step": 4172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.005383797315825,
+      "learning_rate": 1.4923213240331223e-05,
+      "loss": 1.1436,
+      "step": 4173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.102291365051272,
+      "learning_rate": 1.4920807417817504e-05,
+      "loss": 1.1936,
+      "step": 4174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.473469572892469,
+      "learning_rate": 1.4918401219427963e-05,
+      "loss": 1.2051,
+      "step": 4175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5783675613394625,
+      "learning_rate": 1.4915994645346397e-05,
+      "loss": 1.2004,
+      "step": 4176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.425513307604131,
+      "learning_rate": 1.4913587695756632e-05,
+      "loss": 1.1897,
+      "step": 4177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6588112201084906,
+      "learning_rate": 1.4911180370842524e-05,
+      "loss": 1.1575,
+      "step": 4178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9666634665247695,
+      "learning_rate": 1.4908772670787953e-05,
+      "loss": 1.1611,
+      "step": 4179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7994247844403373,
+      "learning_rate": 1.4906364595776834e-05,
+      "loss": 1.2441,
+      "step": 4180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.061814167401879,
+      "learning_rate": 1.4903956145993111e-05,
+      "loss": 1.1963,
+      "step": 4181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9831893142885257,
+      "learning_rate": 1.4901547321620746e-05,
+      "loss": 1.1191,
+      "step": 4182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7120277409457256,
+      "learning_rate": 1.4899138122843742e-05,
+      "loss": 1.1028,
+      "step": 4183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.839022104654737,
+      "learning_rate": 1.4896728549846126e-05,
+      "loss": 1.1736,
+      "step": 4184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4253556325323395,
+      "learning_rate": 1.489431860281195e-05,
+      "loss": 1.1492,
+      "step": 4185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.210681171561162,
+      "learning_rate": 1.48919082819253e-05,
+      "loss": 1.3213,
+      "step": 4186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.197543076115012,
+      "learning_rate": 1.4889497587370287e-05,
+      "loss": 1.1836,
+      "step": 4187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.738551497282625,
+      "learning_rate": 1.4887086519331054e-05,
+      "loss": 1.2378,
+      "step": 4188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.619909717354002,
+      "learning_rate": 1.4884675077991766e-05,
+      "loss": 1.1389,
+      "step": 4189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.284492756936851,
+      "learning_rate": 1.488226326353663e-05,
+      "loss": 1.2017,
+      "step": 4190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.392086671443215,
+      "learning_rate": 1.4879851076149863e-05,
+      "loss": 1.208,
+      "step": 4191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.579194830893108,
+      "learning_rate": 1.487743851601572e-05,
+      "loss": 1.2402,
+      "step": 4192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.449738790550821,
+      "learning_rate": 1.487502558331849e-05,
+      "loss": 1.1765,
+      "step": 4193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6848823156561,
+      "learning_rate": 1.4872612278242486e-05,
+      "loss": 1.1685,
+      "step": 4194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.818353380823403,
+      "learning_rate": 1.4870198600972043e-05,
+      "loss": 1.1367,
+      "step": 4195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8058672484347564,
+      "learning_rate": 1.4867784551691528e-05,
+      "loss": 1.1699,
+      "step": 4196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.180724668609296,
+      "learning_rate": 1.4865370130585346e-05,
+      "loss": 1.1875,
+      "step": 4197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.618559822903475,
+      "learning_rate": 1.4862955337837917e-05,
+      "loss": 1.1638,
+      "step": 4198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.11654994758939,
+      "learning_rate": 1.4860540173633693e-05,
+      "loss": 1.2324,
+      "step": 4199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8842106353410126,
+      "learning_rate": 1.4858124638157163e-05,
+      "loss": 1.1458,
+      "step": 4200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.020394066658569,
+      "learning_rate": 1.4855708731592835e-05,
+      "loss": 1.1587,
+      "step": 4201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0383299555152017,
+      "learning_rate": 1.4853292454125243e-05,
+      "loss": 1.2451,
+      "step": 4202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2896069435103215,
+      "learning_rate": 1.4850875805938963e-05,
+      "loss": 1.1719,
+      "step": 4203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8693809059718864,
+      "learning_rate": 1.4848458787218583e-05,
+      "loss": 1.1875,
+      "step": 4204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3974060937940083,
+      "learning_rate": 1.4846041398148732e-05,
+      "loss": 1.2578,
+      "step": 4205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9380433320081343,
+      "learning_rate": 1.4843623638914058e-05,
+      "loss": 1.1467,
+      "step": 4206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.238062618534505,
+      "learning_rate": 1.4841205509699243e-05,
+      "loss": 1.1765,
+      "step": 4207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7103542167948564,
+      "learning_rate": 1.4838787010688997e-05,
+      "loss": 1.1658,
+      "step": 4208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.770572323513643,
+      "learning_rate": 1.483636814206806e-05,
+      "loss": 1.2109,
+      "step": 4209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.891416229072611,
+      "learning_rate": 1.4833948904021192e-05,
+      "loss": 1.1648,
+      "step": 4210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9948729291719802,
+      "learning_rate": 1.4831529296733186e-05,
+      "loss": 1.1985,
+      "step": 4211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9793059058894658,
+      "learning_rate": 1.4829109320388869e-05,
+      "loss": 1.2378,
+      "step": 4212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1330064485510216,
+      "learning_rate": 1.4826688975173085e-05,
+      "loss": 1.2114,
+      "step": 4213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0444562469104612,
+      "learning_rate": 1.482426826127072e-05,
+      "loss": 1.1992,
+      "step": 4214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8375644576393344,
+      "learning_rate": 1.4821847178866668e-05,
+      "loss": 1.2234,
+      "step": 4215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.889629309643462,
+      "learning_rate": 1.4819425728145875e-05,
+      "loss": 1.1199,
+      "step": 4216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0238851340931525,
+      "learning_rate": 1.4817003909293296e-05,
+      "loss": 1.0959,
+      "step": 4217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9424203263520696,
+      "learning_rate": 1.4814581722493925e-05,
+      "loss": 1.178,
+      "step": 4218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4973669384385024,
+      "learning_rate": 1.4812159167932784e-05,
+      "loss": 1.1606,
+      "step": 4219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.243912235932914,
+      "learning_rate": 1.4809736245794912e-05,
+      "loss": 1.1775,
+      "step": 4220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.418699810387672,
+      "learning_rate": 1.4807312956265389e-05,
+      "loss": 1.1357,
+      "step": 4221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0936754585266457,
+      "learning_rate": 1.480488929952932e-05,
+      "loss": 1.2383,
+      "step": 4222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3297751288852195,
+      "learning_rate": 1.4802465275771829e-05,
+      "loss": 1.1589,
+      "step": 4223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5829278726983755,
+      "learning_rate": 1.480004088517808e-05,
+      "loss": 1.2292,
+      "step": 4224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5577700977234894,
+      "learning_rate": 1.4797616127933259e-05,
+      "loss": 1.1716,
+      "step": 4225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8918722732227553,
+      "learning_rate": 1.479519100422258e-05,
+      "loss": 1.1335,
+      "step": 4226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6795313733938886,
+      "learning_rate": 1.4792765514231286e-05,
+      "loss": 1.1775,
+      "step": 4227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1485596384928365,
+      "learning_rate": 1.4790339658144652e-05,
+      "loss": 1.2808,
+      "step": 4228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.748745804838613,
+      "learning_rate": 1.4787913436147974e-05,
+      "loss": 1.1699,
+      "step": 4229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.833954740286534,
+      "learning_rate": 1.4785486848426578e-05,
+      "loss": 1.1467,
+      "step": 4230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1854405514031323,
+      "learning_rate": 1.4783059895165818e-05,
+      "loss": 1.1638,
+      "step": 4231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.458442861335028,
+      "learning_rate": 1.4780632576551081e-05,
+      "loss": 1.2283,
+      "step": 4232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0129046762294074,
+      "learning_rate": 1.4778204892767774e-05,
+      "loss": 1.2695,
+      "step": 4233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1968987184617217,
+      "learning_rate": 1.477577684400134e-05,
+      "loss": 1.1711,
+      "step": 4234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.816922981232932,
+      "learning_rate": 1.4773348430437237e-05,
+      "loss": 1.1411,
+      "step": 4235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.318564270007492,
+      "learning_rate": 1.4770919652260968e-05,
+      "loss": 1.2415,
+      "step": 4236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0296438909668253,
+      "learning_rate": 1.4768490509658053e-05,
+      "loss": 1.1577,
+      "step": 4237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.951115726496014,
+      "learning_rate": 1.476606100281404e-05,
+      "loss": 1.1599,
+      "step": 4238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9379670016510477,
+      "learning_rate": 1.476363113191451e-05,
+      "loss": 1.2256,
+      "step": 4239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4547071848105344,
+      "learning_rate": 1.4761200897145063e-05,
+      "loss": 1.1995,
+      "step": 4240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.095632422629599,
+      "learning_rate": 1.4758770298691337e-05,
+      "loss": 1.1353,
+      "step": 4241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.088635169920522,
+      "learning_rate": 1.4756339336738995e-05,
+      "loss": 1.2302,
+      "step": 4242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.241355385626048,
+      "learning_rate": 1.4753908011473717e-05,
+      "loss": 1.228,
+      "step": 4243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.295207684576989,
+      "learning_rate": 1.4751476323081231e-05,
+      "loss": 1.1416,
+      "step": 4244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.823592373150775,
+      "learning_rate": 1.4749044271747274e-05,
+      "loss": 1.1821,
+      "step": 4245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.509028258718628,
+      "learning_rate": 1.4746611857657624e-05,
+      "loss": 1.1562,
+      "step": 4246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9069416334501303,
+      "learning_rate": 1.4744179080998076e-05,
+      "loss": 1.187,
+      "step": 4247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.264657496680249,
+      "learning_rate": 1.4741745941954457e-05,
+      "loss": 1.2117,
+      "step": 4248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9497329377361385,
+      "learning_rate": 1.4739312440712627e-05,
+      "loss": 1.1191,
+      "step": 4249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3818938434588213,
+      "learning_rate": 1.4736878577458467e-05,
+      "loss": 1.2898,
+      "step": 4250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0010298024559092,
+      "learning_rate": 1.4734444352377886e-05,
+      "loss": 1.1187,
+      "step": 4251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7719951816614627,
+      "learning_rate": 1.4732009765656824e-05,
+      "loss": 1.2275,
+      "step": 4252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.264656717239707,
+      "learning_rate": 1.4729574817481243e-05,
+      "loss": 1.2163,
+      "step": 4253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.039772115714831,
+      "learning_rate": 1.4727139508037143e-05,
+      "loss": 1.1912,
+      "step": 4254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3485734444959943,
+      "learning_rate": 1.4724703837510542e-05,
+      "loss": 1.1162,
+      "step": 4255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8034548858453334,
+      "learning_rate": 1.4722267806087487e-05,
+      "loss": 1.1799,
+      "step": 4256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0145262354224935,
+      "learning_rate": 1.4719831413954059e-05,
+      "loss": 1.2495,
+      "step": 4257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.737462969596805,
+      "learning_rate": 1.4717394661296353e-05,
+      "loss": 1.1682,
+      "step": 4258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1408736849743453,
+      "learning_rate": 1.4714957548300508e-05,
+      "loss": 1.2012,
+      "step": 4259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0312357241930994,
+      "learning_rate": 1.4712520075152683e-05,
+      "loss": 1.1477,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.98883355609159,
+      "learning_rate": 1.471008224203906e-05,
+      "loss": 1.1897,
+      "step": 4261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.106761028984988,
+      "learning_rate": 1.4707644049145854e-05,
+      "loss": 1.1243,
+      "step": 4262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.099555938280515,
+      "learning_rate": 1.470520549665931e-05,
+      "loss": 1.1367,
+      "step": 4263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1862381200341643,
+      "learning_rate": 1.4702766584765693e-05,
+      "loss": 1.1602,
+      "step": 4264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0635561975264816,
+      "learning_rate": 1.4700327313651302e-05,
+      "loss": 1.1204,
+      "step": 4265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.876257929857712,
+      "learning_rate": 1.4697887683502457e-05,
+      "loss": 1.1533,
+      "step": 4266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0172492442179384,
+      "learning_rate": 1.4695447694505512e-05,
+      "loss": 1.1344,
+      "step": 4267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5984532987024855,
+      "learning_rate": 1.4693007346846841e-05,
+      "loss": 1.1058,
+      "step": 4268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.064522208473884,
+      "learning_rate": 1.4690566640712855e-05,
+      "loss": 1.1962,
+      "step": 4269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4733921480629784,
+      "learning_rate": 1.4688125576289992e-05,
+      "loss": 1.1653,
+      "step": 4270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0880338250620425,
+      "learning_rate": 1.4685684153764702e-05,
+      "loss": 1.2134,
+      "step": 4271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.402571305533366,
+      "learning_rate": 1.4683242373323477e-05,
+      "loss": 1.1821,
+      "step": 4272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7923087543714744,
+      "learning_rate": 1.4680800235152833e-05,
+      "loss": 1.1938,
+      "step": 4273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3759371649964995,
+      "learning_rate": 1.4678357739439313e-05,
+      "loss": 1.1826,
+      "step": 4274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.165136535871447,
+      "learning_rate": 1.4675914886369489e-05,
+      "loss": 1.1887,
+      "step": 4275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.086153190410274,
+      "learning_rate": 1.467347167612995e-05,
+      "loss": 1.2373,
+      "step": 4276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.518273525375465,
+      "learning_rate": 1.4671028108907335e-05,
+      "loss": 1.2344,
+      "step": 4277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.981938179336237,
+      "learning_rate": 1.4668584184888282e-05,
+      "loss": 1.093,
+      "step": 4278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.861121666428433,
+      "learning_rate": 1.4666139904259478e-05,
+      "loss": 1.1321,
+      "step": 4279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0984342247245413,
+      "learning_rate": 1.4663695267207625e-05,
+      "loss": 1.187,
+      "step": 4280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.123871298743614,
+      "learning_rate": 1.4661250273919459e-05,
+      "loss": 1.1589,
+      "step": 4281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5590226289075795,
+      "learning_rate": 1.4658804924581738e-05,
+      "loss": 1.2095,
+      "step": 4282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5690888284735727,
+      "learning_rate": 1.4656359219381256e-05,
+      "loss": 1.1609,
+      "step": 4283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.972655888612552,
+      "learning_rate": 1.4653913158504824e-05,
+      "loss": 1.1157,
+      "step": 4284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5589030664744343,
+      "learning_rate": 1.4651466742139284e-05,
+      "loss": 1.1895,
+      "step": 4285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.760498299611378,
+      "learning_rate": 1.4649019970471506e-05,
+      "loss": 1.2419,
+      "step": 4286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1705054084208486,
+      "learning_rate": 1.4646572843688387e-05,
+      "loss": 1.2021,
+      "step": 4287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.498660236819328,
+      "learning_rate": 1.464412536197685e-05,
+      "loss": 1.1978,
+      "step": 4288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7072980781441274,
+      "learning_rate": 1.4641677525523848e-05,
+      "loss": 1.2314,
+      "step": 4289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.950264434768645,
+      "learning_rate": 1.4639229334516358e-05,
+      "loss": 1.1545,
+      "step": 4290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3391607997967188,
+      "learning_rate": 1.4636780789141383e-05,
+      "loss": 1.1367,
+      "step": 4291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.102781299077512,
+      "learning_rate": 1.4634331889585956e-05,
+      "loss": 1.1165,
+      "step": 4292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.889801576191917,
+      "learning_rate": 1.4631882636037139e-05,
+      "loss": 1.1697,
+      "step": 4293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0699350837006407,
+      "learning_rate": 1.4629433028682014e-05,
+      "loss": 1.1628,
+      "step": 4294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9051196631439127,
+      "learning_rate": 1.4626983067707696e-05,
+      "loss": 1.2148,
+      "step": 4295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.099923205998726,
+      "learning_rate": 1.4624532753301326e-05,
+      "loss": 1.113,
+      "step": 4296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9303327922648816,
+      "learning_rate": 1.4622082085650072e-05,
+      "loss": 1.0803,
+      "step": 4297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9952884658681027,
+      "learning_rate": 1.4619631064941126e-05,
+      "loss": 1.1929,
+      "step": 4298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7025785164652487,
+      "learning_rate": 1.4617179691361709e-05,
+      "loss": 1.1648,
+      "step": 4299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1731015140978904,
+      "learning_rate": 1.4614727965099074e-05,
+      "loss": 1.2268,
+      "step": 4300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0140826485520544,
+      "learning_rate": 1.4612275886340488e-05,
+      "loss": 1.1384,
+      "step": 4301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8220236068131492,
+      "learning_rate": 1.460982345527326e-05,
+      "loss": 1.2031,
+      "step": 4302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7487859781169073,
+      "learning_rate": 1.4607370672084717e-05,
+      "loss": 1.1743,
+      "step": 4303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7405879364598493,
+      "learning_rate": 1.4604917536962211e-05,
+      "loss": 1.1467,
+      "step": 4304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.180652767963664,
+      "learning_rate": 1.4602464050093128e-05,
+      "loss": 1.24,
+      "step": 4305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1162463952440835,
+      "learning_rate": 1.4600010211664877e-05,
+      "loss": 1.1929,
+      "step": 4306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.416721236425974,
+      "learning_rate": 1.4597556021864899e-05,
+      "loss": 1.1992,
+      "step": 4307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4661458050162754,
+      "learning_rate": 1.459510148088065e-05,
+      "loss": 1.1873,
+      "step": 4308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1248092101387632,
+      "learning_rate": 1.4592646588899623e-05,
+      "loss": 1.1677,
+      "step": 4309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.096857929785219,
+      "learning_rate": 1.4590191346109338e-05,
+      "loss": 1.2026,
+      "step": 4310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8202686010851385,
+      "learning_rate": 1.4587735752697331e-05,
+      "loss": 1.1621,
+      "step": 4311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1578990878989326,
+      "learning_rate": 1.458527980885118e-05,
+      "loss": 1.1353,
+      "step": 4312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3081868800037935,
+      "learning_rate": 1.4582823514758481e-05,
+      "loss": 1.1016,
+      "step": 4313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.034504983350621,
+      "learning_rate": 1.4580366870606852e-05,
+      "loss": 1.165,
+      "step": 4314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.280116517554485,
+      "learning_rate": 1.4577909876583951e-05,
+      "loss": 1.1318,
+      "step": 4315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.403937294734511,
+      "learning_rate": 1.4575452532877451e-05,
+      "loss": 1.1584,
+      "step": 4316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.869158961748407,
+      "learning_rate": 1.4572994839675063e-05,
+      "loss": 1.187,
+      "step": 4317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.600742928825021,
+      "learning_rate": 1.4570536797164507e-05,
+      "loss": 1.2212,
+      "step": 4318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.092578026935936,
+      "learning_rate": 1.4568078405533548e-05,
+      "loss": 1.1819,
+      "step": 4319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.014556046297513,
+      "learning_rate": 1.456561966496997e-05,
+      "loss": 1.1829,
+      "step": 4320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.109960898689991,
+      "learning_rate": 1.456316057566158e-05,
+      "loss": 1.1819,
+      "step": 4321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.715222822283294,
+      "learning_rate": 1.4560701137796218e-05,
+      "loss": 1.0703,
+      "step": 4322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2529026443075373,
+      "learning_rate": 1.4558241351561751e-05,
+      "loss": 1.1875,
+      "step": 4323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2080418584198567,
+      "learning_rate": 1.4555781217146063e-05,
+      "loss": 1.1848,
+      "step": 4324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3037544043813702,
+      "learning_rate": 1.4553320734737075e-05,
+      "loss": 1.2002,
+      "step": 4325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1810509492080783,
+      "learning_rate": 1.4550859904522735e-05,
+      "loss": 1.1458,
+      "step": 4326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2300752115519056,
+      "learning_rate": 1.4548398726691006e-05,
+      "loss": 1.1541,
+      "step": 4327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2904003149641627,
+      "learning_rate": 1.454593720142989e-05,
+      "loss": 1.1982,
+      "step": 4328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1203675363662593,
+      "learning_rate": 1.4543475328927407e-05,
+      "loss": 1.2183,
+      "step": 4329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6352522317794107,
+      "learning_rate": 1.454101310937161e-05,
+      "loss": 1.2275,
+      "step": 4330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7238034081742883,
+      "learning_rate": 1.4538550542950577e-05,
+      "loss": 1.1655,
+      "step": 4331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.152216684812925,
+      "learning_rate": 1.453608762985241e-05,
+      "loss": 1.272,
+      "step": 4332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5654286773746446,
+      "learning_rate": 1.4533624370265233e-05,
+      "loss": 1.2036,
+      "step": 4333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.268548329557398,
+      "learning_rate": 1.4531160764377209e-05,
+      "loss": 1.1777,
+      "step": 4334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5841938989525435,
+      "learning_rate": 1.4528696812376519e-05,
+      "loss": 1.1152,
+      "step": 4335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9251586882128624,
+      "learning_rate": 1.4526232514451369e-05,
+      "loss": 1.2458,
+      "step": 4336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3900696661151795,
+      "learning_rate": 1.4523767870789998e-05,
+      "loss": 1.196,
+      "step": 4337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.948939821177208,
+      "learning_rate": 1.4521302881580663e-05,
+      "loss": 1.1355,
+      "step": 4338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0386738744727424,
+      "learning_rate": 1.4518837547011657e-05,
+      "loss": 1.1343,
+      "step": 4339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.33648537922145,
+      "learning_rate": 1.4516371867271298e-05,
+      "loss": 1.1733,
+      "step": 4340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8496554363128337,
+      "learning_rate": 1.4513905842547917e-05,
+      "loss": 1.1912,
+      "step": 4341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4148060224432704,
+      "learning_rate": 1.4511439473029884e-05,
+      "loss": 1.1514,
+      "step": 4342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.831842779925716,
+      "learning_rate": 1.45089727589056e-05,
+      "loss": 1.0815,
+      "step": 4343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4595118737134722,
+      "learning_rate": 1.4506505700363476e-05,
+      "loss": 1.1924,
+      "step": 4344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.999756851779801,
+      "learning_rate": 1.4504038297591962e-05,
+      "loss": 1.0859,
+      "step": 4345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.848831375644063,
+      "learning_rate": 1.4501570550779532e-05,
+      "loss": 1.1326,
+      "step": 4346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5218026061834258,
+      "learning_rate": 1.4499102460114682e-05,
+      "loss": 1.2061,
+      "step": 4347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.953883783809417,
+      "learning_rate": 1.4496634025785938e-05,
+      "loss": 1.1548,
+      "step": 4348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.227275011487752,
+      "learning_rate": 1.4494165247981855e-05,
+      "loss": 1.1096,
+      "step": 4349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3919567495526546,
+      "learning_rate": 1.4491696126891007e-05,
+      "loss": 1.1208,
+      "step": 4350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.20117325561638,
+      "learning_rate": 1.4489226662701994e-05,
+      "loss": 1.1707,
+      "step": 4351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.23145505689216,
+      "learning_rate": 1.448675685560345e-05,
+      "loss": 1.2,
+      "step": 4352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3161507309810614,
+      "learning_rate": 1.4484286705784035e-05,
+      "loss": 1.2112,
+      "step": 4353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.343909446359335,
+      "learning_rate": 1.4481816213432428e-05,
+      "loss": 1.1475,
+      "step": 4354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.4690381572840785,
+      "learning_rate": 1.4479345378737333e-05,
+      "loss": 1.231,
+      "step": 4355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4571032992088,
+      "learning_rate": 1.447687420188749e-05,
+      "loss": 1.0769,
+      "step": 4356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.418798423032855,
+      "learning_rate": 1.4474402683071659e-05,
+      "loss": 1.2185,
+      "step": 4357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.895474626775535,
+      "learning_rate": 1.4471930822478624e-05,
+      "loss": 1.1228,
+      "step": 4358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.155258241101492,
+      "learning_rate": 1.4469458620297206e-05,
+      "loss": 1.1711,
+      "step": 4359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.802456078781614,
+      "learning_rate": 1.4466986076716235e-05,
+      "loss": 1.1326,
+      "step": 4360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.359815067828288,
+      "learning_rate": 1.446451319192458e-05,
+      "loss": 1.1904,
+      "step": 4361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.724368263751023,
+      "learning_rate": 1.4462039966111133e-05,
+      "loss": 1.0933,
+      "step": 4362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.1098464026903665,
+      "learning_rate": 1.445956639946481e-05,
+      "loss": 1.1921,
+      "step": 4363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4825052314231053,
+      "learning_rate": 1.4457092492174558e-05,
+      "loss": 1.1177,
+      "step": 4364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5002774063193085,
+      "learning_rate": 1.4454618244429339e-05,
+      "loss": 1.2227,
+      "step": 4365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.157355953951271,
+      "learning_rate": 1.4452143656418157e-05,
+      "loss": 1.1692,
+      "step": 4366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.182201934573233,
+      "learning_rate": 1.4449668728330023e-05,
+      "loss": 1.1748,
+      "step": 4367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.497195008852893,
+      "learning_rate": 1.4447193460353996e-05,
+      "loss": 1.1858,
+      "step": 4368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.230945130470602,
+      "learning_rate": 1.4444717852679139e-05,
+      "loss": 1.1787,
+      "step": 4369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.871881308999783,
+      "learning_rate": 1.444224190549456e-05,
+      "loss": 1.2495,
+      "step": 4370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7834509830883674,
+      "learning_rate": 1.4439765618989378e-05,
+      "loss": 1.2278,
+      "step": 4371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.121564797513516,
+      "learning_rate": 1.443728899335275e-05,
+      "loss": 1.1633,
+      "step": 4372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0765874934610076,
+      "learning_rate": 1.4434812028773845e-05,
+      "loss": 1.1899,
+      "step": 4373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0842702232885943,
+      "learning_rate": 1.4432334725441873e-05,
+      "loss": 1.1931,
+      "step": 4374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.03433000972149,
+      "learning_rate": 1.4429857083546054e-05,
+      "loss": 1.1455,
+      "step": 4375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.443419153037571,
+      "learning_rate": 1.4427379103275655e-05,
+      "loss": 1.2278,
+      "step": 4376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6529187865915502,
+      "learning_rate": 1.4424900784819949e-05,
+      "loss": 1.1084,
+      "step": 4377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7523926978803286,
+      "learning_rate": 1.4422422128368242e-05,
+      "loss": 1.1838,
+      "step": 4378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9139397567499588,
+      "learning_rate": 1.4419943134109868e-05,
+      "loss": 1.1282,
+      "step": 4379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.217309750937532,
+      "learning_rate": 1.4417463802234189e-05,
+      "loss": 1.1682,
+      "step": 4380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4343823020056425,
+      "learning_rate": 1.4414984132930579e-05,
+      "loss": 1.1841,
+      "step": 4381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4185246082356304,
+      "learning_rate": 1.4412504126388453e-05,
+      "loss": 1.1965,
+      "step": 4382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.640729892216189,
+      "learning_rate": 1.441002378279725e-05,
+      "loss": 1.1453,
+      "step": 4383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7853533910700934,
+      "learning_rate": 1.4407543102346423e-05,
+      "loss": 1.2649,
+      "step": 4384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7864191297271854,
+      "learning_rate": 1.4405062085225465e-05,
+      "loss": 1.1089,
+      "step": 4385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.085359924776206,
+      "learning_rate": 1.4402580731623885e-05,
+      "loss": 1.1565,
+      "step": 4386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4667171545995203,
+      "learning_rate": 1.4400099041731226e-05,
+      "loss": 1.2026,
+      "step": 4387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.824826550572187,
+      "learning_rate": 1.4397617015737044e-05,
+      "loss": 1.1013,
+      "step": 4388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.160996982436239,
+      "learning_rate": 1.4395134653830934e-05,
+      "loss": 1.166,
+      "step": 4389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.895204782630927,
+      "learning_rate": 1.4392651956202512e-05,
+      "loss": 1.176,
+      "step": 4390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.339056489693155,
+      "learning_rate": 1.4390168923041417e-05,
+      "loss": 1.1799,
+      "step": 4391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.927592745689643,
+      "learning_rate": 1.4387685554537312e-05,
+      "loss": 1.1477,
+      "step": 4392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0656995399334366,
+      "learning_rate": 1.4385201850879895e-05,
+      "loss": 1.1641,
+      "step": 4393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7237532286595574,
+      "learning_rate": 1.4382717812258882e-05,
+      "loss": 1.1177,
+      "step": 4394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8719926567111496,
+      "learning_rate": 1.4380233438864013e-05,
+      "loss": 1.1521,
+      "step": 4395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.114098968444248,
+      "learning_rate": 1.4377748730885061e-05,
+      "loss": 1.1646,
+      "step": 4396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.190132446388238,
+      "learning_rate": 1.4375263688511819e-05,
+      "loss": 1.2485,
+      "step": 4397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.105015603778684,
+      "learning_rate": 1.4372778311934105e-05,
+      "loss": 1.2163,
+      "step": 4398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.506727898159621,
+      "learning_rate": 1.4370292601341764e-05,
+      "loss": 1.1479,
+      "step": 4399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.143510226776033,
+      "learning_rate": 1.4367806556924676e-05,
+      "loss": 1.1665,
+      "step": 4400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8849421951078207,
+      "learning_rate": 1.4365320178872726e-05,
+      "loss": 1.136,
+      "step": 4401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.608734647723684,
+      "learning_rate": 1.4362833467375839e-05,
+      "loss": 1.2627,
+      "step": 4402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.589913964882031,
+      "learning_rate": 1.4360346422623968e-05,
+      "loss": 1.1692,
+      "step": 4403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.751247350899208,
+      "learning_rate": 1.4357859044807077e-05,
+      "loss": 1.1631,
+      "step": 4404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.909282514538182,
+      "learning_rate": 1.435537133411517e-05,
+      "loss": 1.2036,
+      "step": 4405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4157791933756987,
+      "learning_rate": 1.4352883290738275e-05,
+      "loss": 1.1299,
+      "step": 4406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.981785946324169,
+      "learning_rate": 1.4350394914866432e-05,
+      "loss": 1.186,
+      "step": 4407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8801371892812013,
+      "learning_rate": 1.4347906206689719e-05,
+      "loss": 1.1445,
+      "step": 4408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2870417941836383,
+      "learning_rate": 1.4345417166398235e-05,
+      "loss": 1.1738,
+      "step": 4409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5891501723688957,
+      "learning_rate": 1.4342927794182113e-05,
+      "loss": 1.3167,
+      "step": 4410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6978706633337883,
+      "learning_rate": 1.4340438090231492e-05,
+      "loss": 1.1111,
+      "step": 4411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.892544137998452,
+      "learning_rate": 1.4337948054736555e-05,
+      "loss": 1.1599,
+      "step": 4412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1171898309291404,
+      "learning_rate": 1.4335457687887504e-05,
+      "loss": 1.1223,
+      "step": 4413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.211609409564004,
+      "learning_rate": 1.433296698987456e-05,
+      "loss": 1.1816,
+      "step": 4414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2442335486381118,
+      "learning_rate": 1.4330475960887981e-05,
+      "loss": 1.1731,
+      "step": 4415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.123299933617996,
+      "learning_rate": 1.4327984601118043e-05,
+      "loss": 1.1582,
+      "step": 4416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.279509688626537,
+      "learning_rate": 1.4325492910755043e-05,
+      "loss": 1.2021,
+      "step": 4417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9333039640954266,
+      "learning_rate": 1.4323000889989313e-05,
+      "loss": 1.1748,
+      "step": 4418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4699888224200133,
+      "learning_rate": 1.4320508539011211e-05,
+      "loss": 1.2539,
+      "step": 4419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3759596454283676,
+      "learning_rate": 1.4318015858011108e-05,
+      "loss": 1.1868,
+      "step": 4420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.722500513127505,
+      "learning_rate": 1.4315522847179407e-05,
+      "loss": 1.2593,
+      "step": 4421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.019022469386963,
+      "learning_rate": 1.431302950670654e-05,
+      "loss": 1.2168,
+      "step": 4422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.668552183651791,
+      "learning_rate": 1.431053583678296e-05,
+      "loss": 1.1541,
+      "step": 4423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9362686258846735,
+      "learning_rate": 1.4308041837599146e-05,
+      "loss": 1.0752,
+      "step": 4424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.328422130581779,
+      "learning_rate": 1.4305547509345603e-05,
+      "loss": 1.2319,
+      "step": 4425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.463606442536508,
+      "learning_rate": 1.430305285221286e-05,
+      "loss": 1.1914,
+      "step": 4426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.191568358406004,
+      "learning_rate": 1.4300557866391467e-05,
+      "loss": 1.1848,
+      "step": 4427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.372430918270472,
+      "learning_rate": 1.429806255207201e-05,
+      "loss": 1.1306,
+      "step": 4428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3049299428630694,
+      "learning_rate": 1.429556690944509e-05,
+      "loss": 1.1658,
+      "step": 4429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0867871150071293,
+      "learning_rate": 1.4293070938701336e-05,
+      "loss": 1.2061,
+      "step": 4430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.288690061306546,
+      "learning_rate": 1.4290574640031404e-05,
+      "loss": 1.1006,
+      "step": 4431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.290204620187158,
+      "learning_rate": 1.4288078013625975e-05,
+      "loss": 1.2004,
+      "step": 4432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5493068153093277,
+      "learning_rate": 1.428558105967575e-05,
+      "loss": 1.1807,
+      "step": 4433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.134532736285572,
+      "learning_rate": 1.4283083778371465e-05,
+      "loss": 1.1226,
+      "step": 4434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0550002076136527,
+      "learning_rate": 1.4280586169903869e-05,
+      "loss": 1.1414,
+      "step": 4435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.086713176355734,
+      "learning_rate": 1.4278088234463745e-05,
+      "loss": 1.1648,
+      "step": 4436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.04563710263278,
+      "learning_rate": 1.4275589972241894e-05,
+      "loss": 1.1787,
+      "step": 4437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6070522371720126,
+      "learning_rate": 1.427309138342915e-05,
+      "loss": 1.208,
+      "step": 4438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5397312919738084,
+      "learning_rate": 1.4270592468216367e-05,
+      "loss": 1.1572,
+      "step": 4439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.123817231938139,
+      "learning_rate": 1.4268093226794422e-05,
+      "loss": 1.1914,
+      "step": 4440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3816800506352904,
+      "learning_rate": 1.426559365935422e-05,
+      "loss": 1.1816,
+      "step": 4441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9938547552529604,
+      "learning_rate": 1.4263093766086696e-05,
+      "loss": 1.1558,
+      "step": 4442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.378788989918236,
+      "learning_rate": 1.4260593547182797e-05,
+      "loss": 1.1819,
+      "step": 4443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.54981853903258,
+      "learning_rate": 1.4258093002833504e-05,
+      "loss": 1.2209,
+      "step": 4444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.343836409533524,
+      "learning_rate": 1.4255592133229823e-05,
+      "loss": 1.1538,
+      "step": 4445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2888779750543757,
+      "learning_rate": 1.4253090938562784e-05,
+      "loss": 1.168,
+      "step": 4446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1011757864797085,
+      "learning_rate": 1.4250589419023435e-05,
+      "loss": 1.2004,
+      "step": 4447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.174022548295844,
+      "learning_rate": 1.424808757480286e-05,
+      "loss": 1.1497,
+      "step": 4448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.1495596562073045,
+      "learning_rate": 1.4245585406092164e-05,
+      "loss": 1.1675,
+      "step": 4449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6840122673856386,
+      "learning_rate": 1.4243082913082467e-05,
+      "loss": 1.1748,
+      "step": 4450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.041017802571133,
+      "learning_rate": 1.4240580095964927e-05,
+      "loss": 1.2302,
+      "step": 4451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.493228096597684,
+      "learning_rate": 1.4238076954930726e-05,
+      "loss": 1.2244,
+      "step": 4452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.442511014309646,
+      "learning_rate": 1.4235573490171057e-05,
+      "loss": 1.1995,
+      "step": 4453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.645450908509031,
+      "learning_rate": 1.4233069701877153e-05,
+      "loss": 1.2612,
+      "step": 4454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1159838037341308,
+      "learning_rate": 1.4230565590240266e-05,
+      "loss": 1.134,
+      "step": 4455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8949291176847556,
+      "learning_rate": 1.4228061155451671e-05,
+      "loss": 1.2117,
+      "step": 4456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3160332840293023,
+      "learning_rate": 1.4225556397702671e-05,
+      "loss": 1.1741,
+      "step": 4457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2410903904224404,
+      "learning_rate": 1.4223051317184586e-05,
+      "loss": 1.1882,
+      "step": 4458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4713175595484826,
+      "learning_rate": 1.4220545914088777e-05,
+      "loss": 1.2302,
+      "step": 4459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8864570235767144,
+      "learning_rate": 1.4218040188606608e-05,
+      "loss": 1.1353,
+      "step": 4460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3005906838646415,
+      "learning_rate": 1.4215534140929488e-05,
+      "loss": 1.1069,
+      "step": 4461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.769801225705663,
+      "learning_rate": 1.4213027771248833e-05,
+      "loss": 1.179,
+      "step": 4462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0354139371729367,
+      "learning_rate": 1.4210521079756099e-05,
+      "loss": 1.1785,
+      "step": 4463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.24546646397549,
+      "learning_rate": 1.4208014066642757e-05,
+      "loss": 1.1821,
+      "step": 4464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.081421987703159,
+      "learning_rate": 1.4205506732100303e-05,
+      "loss": 1.2107,
+      "step": 4465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.273916179989953,
+      "learning_rate": 1.4202999076320262e-05,
+      "loss": 1.1423,
+      "step": 4466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.4495880308201095,
+      "learning_rate": 1.4200491099494185e-05,
+      "loss": 1.1589,
+      "step": 4467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.020506885342189,
+      "learning_rate": 1.4197982801813637e-05,
+      "loss": 1.1655,
+      "step": 4468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.984694201349859,
+      "learning_rate": 1.4195474183470218e-05,
+      "loss": 1.2202,
+      "step": 4469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.254436485065893,
+      "learning_rate": 1.4192965244655548e-05,
+      "loss": 1.1199,
+      "step": 4470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.599811980459655,
+      "learning_rate": 1.4190455985561273e-05,
+      "loss": 1.1836,
+      "step": 4471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.568539412544165,
+      "learning_rate": 1.418794640637906e-05,
+      "loss": 1.2517,
+      "step": 4472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3087293194296827,
+      "learning_rate": 1.4185436507300607e-05,
+      "loss": 1.2168,
+      "step": 4473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.120554141224323,
+      "learning_rate": 1.4182926288517628e-05,
+      "loss": 1.1851,
+      "step": 4474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9948330637327,
+      "learning_rate": 1.4180415750221874e-05,
+      "loss": 1.0764,
+      "step": 4475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2016381702508863,
+      "learning_rate": 1.4177904892605105e-05,
+      "loss": 1.1685,
+      "step": 4476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3239074989545894,
+      "learning_rate": 1.4175393715859112e-05,
+      "loss": 1.0933,
+      "step": 4477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5124531923753457,
+      "learning_rate": 1.4172882220175717e-05,
+      "loss": 1.1863,
+      "step": 4478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.4097004483635835,
+      "learning_rate": 1.4170370405746758e-05,
+      "loss": 1.2026,
+      "step": 4479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3964747008678913,
+      "learning_rate": 1.4167858272764101e-05,
+      "loss": 1.189,
+      "step": 4480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.777909668480233,
+      "learning_rate": 1.4165345821419631e-05,
+      "loss": 1.1873,
+      "step": 4481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.117329187092835,
+      "learning_rate": 1.4162833051905267e-05,
+      "loss": 1.1943,
+      "step": 4482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.803300245161781,
+      "learning_rate": 1.4160319964412943e-05,
+      "loss": 1.1887,
+      "step": 4483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.397181679641176,
+      "learning_rate": 1.415780655913462e-05,
+      "loss": 1.2292,
+      "step": 4484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.054720460708348,
+      "learning_rate": 1.4155292836262293e-05,
+      "loss": 1.1318,
+      "step": 4485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0852295792393067,
+      "learning_rate": 1.4152778795987965e-05,
+      "loss": 1.1897,
+      "step": 4486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.05028235456336,
+      "learning_rate": 1.4150264438503668e-05,
+      "loss": 1.2173,
+      "step": 4487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.646829067789223,
+      "learning_rate": 1.4147749764001471e-05,
+      "loss": 1.1782,
+      "step": 4488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0582596110790465,
+      "learning_rate": 1.4145234772673452e-05,
+      "loss": 1.1274,
+      "step": 4489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.942345640346481,
+      "learning_rate": 1.4142719464711719e-05,
+      "loss": 1.1602,
+      "step": 4490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.390702560824281,
+      "learning_rate": 1.4140203840308402e-05,
+      "loss": 1.1963,
+      "step": 4491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.895489833194949,
+      "learning_rate": 1.4137687899655662e-05,
+      "loss": 1.1414,
+      "step": 4492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.955553372252125,
+      "learning_rate": 1.4135171642945672e-05,
+      "loss": 1.1162,
+      "step": 4493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.095163055109287,
+      "learning_rate": 1.413265507037064e-05,
+      "loss": 1.2141,
+      "step": 4494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6874153214113154,
+      "learning_rate": 1.4130138182122793e-05,
+      "loss": 1.2109,
+      "step": 4495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1679002021530103,
+      "learning_rate": 1.4127620978394388e-05,
+      "loss": 1.1589,
+      "step": 4496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.092664897979507,
+      "learning_rate": 1.4125103459377697e-05,
+      "loss": 1.1528,
+      "step": 4497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.74137310977528,
+      "learning_rate": 1.4122585625265019e-05,
+      "loss": 1.1853,
+      "step": 4498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.615779339450063,
+      "learning_rate": 1.412006747624869e-05,
+      "loss": 1.2966,
+      "step": 4499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.034613290157512,
+      "learning_rate": 1.4117549012521043e-05,
+      "loss": 1.1431,
+      "step": 4500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.740909496317511,
+      "learning_rate": 1.4115030234274461e-05,
+      "loss": 1.2302,
+      "step": 4501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.484266652640631,
+      "learning_rate": 1.4112511141701339e-05,
+      "loss": 1.1919,
+      "step": 4502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.397864885487632,
+      "learning_rate": 1.4109991734994096e-05,
+      "loss": 1.1177,
+      "step": 4503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7154444717713204,
+      "learning_rate": 1.4107472014345179e-05,
+      "loss": 1.1265,
+      "step": 4504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.961434208134734,
+      "learning_rate": 1.4104951979947052e-05,
+      "loss": 1.2437,
+      "step": 4505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.106833241998149,
+      "learning_rate": 1.4102431631992218e-05,
+      "loss": 1.1821,
+      "step": 4506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.644352257392169,
+      "learning_rate": 1.409991097067318e-05,
+      "loss": 1.1721,
+      "step": 4507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.972070582177001,
+      "learning_rate": 1.4097389996182491e-05,
+      "loss": 1.0864,
+      "step": 4508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.3874659468293995,
+      "learning_rate": 1.409486870871271e-05,
+      "loss": 1.1587,
+      "step": 4509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.874205244378875,
+      "learning_rate": 1.4092347108456425e-05,
+      "loss": 1.1921,
+      "step": 4510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.427090007318814,
+      "learning_rate": 1.408982519560625e-05,
+      "loss": 1.1118,
+      "step": 4511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.128915396362005,
+      "learning_rate": 1.408730297035482e-05,
+      "loss": 1.2058,
+      "step": 4512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.554834498913453,
+      "learning_rate": 1.40847804328948e-05,
+      "loss": 1.1282,
+      "step": 4513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.659846781601851,
+      "learning_rate": 1.4082257583418865e-05,
+      "loss": 1.1943,
+      "step": 4514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.769306889194547,
+      "learning_rate": 1.4079734422119728e-05,
+      "loss": 1.149,
+      "step": 4515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.083705437561821,
+      "learning_rate": 1.4077210949190124e-05,
+      "loss": 1.1401,
+      "step": 4516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.162350241875845,
+      "learning_rate": 1.4074687164822802e-05,
+      "loss": 1.1885,
+      "step": 4517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.709204043771188,
+      "learning_rate": 1.4072163069210546e-05,
+      "loss": 1.1997,
+      "step": 4518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.603384732413761,
+      "learning_rate": 1.4069638662546158e-05,
+      "loss": 1.1047,
+      "step": 4519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.502976173764031,
+      "learning_rate": 1.4067113945022462e-05,
+      "loss": 1.1416,
+      "step": 4520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1998335232495116,
+      "learning_rate": 1.4064588916832309e-05,
+      "loss": 1.1973,
+      "step": 4521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4176087736213305,
+      "learning_rate": 1.4062063578168577e-05,
+      "loss": 1.1765,
+      "step": 4522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.826704632640973,
+      "learning_rate": 1.4059537929224164e-05,
+      "loss": 1.1418,
+      "step": 4523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5284604568823577,
+      "learning_rate": 1.4057011970191985e-05,
+      "loss": 1.2075,
+      "step": 4524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.586823394526795,
+      "learning_rate": 1.4054485701264989e-05,
+      "loss": 1.1218,
+      "step": 4525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1114674662394637,
+      "learning_rate": 1.4051959122636149e-05,
+      "loss": 1.1472,
+      "step": 4526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5771831436046138,
+      "learning_rate": 1.4049432234498455e-05,
+      "loss": 1.1816,
+      "step": 4527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7895390433357137,
+      "learning_rate": 1.404690503704492e-05,
+      "loss": 1.1206,
+      "step": 4528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.620392397490571,
+      "learning_rate": 1.4044377530468589e-05,
+      "loss": 1.1855,
+      "step": 4529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5305305302923586,
+      "learning_rate": 1.404184971496252e-05,
+      "loss": 1.1389,
+      "step": 4530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0033896182400843,
+      "learning_rate": 1.4039321590719807e-05,
+      "loss": 1.0908,
+      "step": 4531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5902298002656963,
+      "learning_rate": 1.4036793157933557e-05,
+      "loss": 1.136,
+      "step": 4532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.38620028327629,
+      "learning_rate": 1.40342644167969e-05,
+      "loss": 1.2141,
+      "step": 4533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.056890646395158,
+      "learning_rate": 1.4031735367503e-05,
+      "loss": 1.1787,
+      "step": 4534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8919480376938633,
+      "learning_rate": 1.402920601024504e-05,
+      "loss": 1.2053,
+      "step": 4535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.238144449345819,
+      "learning_rate": 1.4026676345216221e-05,
+      "loss": 1.1482,
+      "step": 4536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.073915929857602,
+      "learning_rate": 1.402414637260977e-05,
+      "loss": 1.2175,
+      "step": 4537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1353472728577523,
+      "learning_rate": 1.4021616092618941e-05,
+      "loss": 1.1575,
+      "step": 4538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.033588347963145,
+      "learning_rate": 1.4019085505437011e-05,
+      "loss": 1.1965,
+      "step": 4539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9671696246042116,
+      "learning_rate": 1.4016554611257276e-05,
+      "loss": 1.1389,
+      "step": 4540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.913698431103166,
+      "learning_rate": 1.401402341027306e-05,
+      "loss": 1.1455,
+      "step": 4541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.771998463124675,
+      "learning_rate": 1.4011491902677711e-05,
+      "loss": 1.1914,
+      "step": 4542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.570866106397938,
+      "learning_rate": 1.4008960088664591e-05,
+      "loss": 1.4709,
+      "step": 4543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.09704111942884,
+      "learning_rate": 1.4006427968427101e-05,
+      "loss": 1.2061,
+      "step": 4544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.208923003849487,
+      "learning_rate": 1.4003895542158652e-05,
+      "loss": 1.1594,
+      "step": 4545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1168451521562845,
+      "learning_rate": 1.4001362810052684e-05,
+      "loss": 1.1626,
+      "step": 4546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.823141981584068,
+      "learning_rate": 1.399882977230266e-05,
+      "loss": 1.1392,
+      "step": 4547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.124578584986398,
+      "learning_rate": 1.3996296429102066e-05,
+      "loss": 1.1558,
+      "step": 4548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.334414703308503,
+      "learning_rate": 1.3993762780644413e-05,
+      "loss": 1.1157,
+      "step": 4549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.16178277985506,
+      "learning_rate": 1.399122882712323e-05,
+      "loss": 1.2295,
+      "step": 4550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.319086624950159,
+      "learning_rate": 1.3988694568732077e-05,
+      "loss": 1.1892,
+      "step": 4551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3449401642471885,
+      "learning_rate": 1.3986160005664533e-05,
+      "loss": 1.1113,
+      "step": 4552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.791694078136502,
+      "learning_rate": 1.3983625138114198e-05,
+      "loss": 1.272,
+      "step": 4553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.349307293797065,
+      "learning_rate": 1.3981089966274697e-05,
+      "loss": 1.1165,
+      "step": 4554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7727127089312478,
+      "learning_rate": 1.3978554490339687e-05,
+      "loss": 1.2273,
+      "step": 4555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.446807200287246,
+      "learning_rate": 1.397601871050283e-05,
+      "loss": 1.1284,
+      "step": 4556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.570586987244859,
+      "learning_rate": 1.3973482626957826e-05,
+      "loss": 1.2104,
+      "step": 4557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1801864246651252,
+      "learning_rate": 1.3970946239898394e-05,
+      "loss": 1.1465,
+      "step": 4558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.230898414047277,
+      "learning_rate": 1.3968409549518276e-05,
+      "loss": 1.1187,
+      "step": 4559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.58611783042231,
+      "learning_rate": 1.3965872556011237e-05,
+      "loss": 1.125,
+      "step": 4560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6851053048285207,
+      "learning_rate": 1.3963335259571063e-05,
+      "loss": 1.1016,
+      "step": 4561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.539489117755016,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 1.2449,
+      "step": 4562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3432878492127265,
+      "learning_rate": 1.3958259758666588e-05,
+      "loss": 1.127,
+      "step": 4563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.88444152927855,
+      "learning_rate": 1.3955721554589979e-05,
+      "loss": 1.1028,
+      "step": 4564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.682577893556587,
+      "learning_rate": 1.395318304835562e-05,
+      "loss": 1.1362,
+      "step": 4565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3349060235591916,
+      "learning_rate": 1.3950644240157412e-05,
+      "loss": 1.2205,
+      "step": 4566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.378227032353228,
+      "learning_rate": 1.3948105130189287e-05,
+      "loss": 1.2068,
+      "step": 4567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.502178769945333,
+      "learning_rate": 1.3945565718645195e-05,
+      "loss": 1.1614,
+      "step": 4568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8046018337008207,
+      "learning_rate": 1.3943026005719108e-05,
+      "loss": 1.1733,
+      "step": 4569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.604562921675613,
+      "learning_rate": 1.3940485991605022e-05,
+      "loss": 1.1775,
+      "step": 4570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.949678646475824,
+      "learning_rate": 1.3937945676496952e-05,
+      "loss": 1.1389,
+      "step": 4571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.337768341302734,
+      "learning_rate": 1.3935405060588947e-05,
+      "loss": 1.2087,
+      "step": 4572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6222393057403957,
+      "learning_rate": 1.3932864144075064e-05,
+      "loss": 1.1667,
+      "step": 4573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3979322523084603,
+      "learning_rate": 1.39303229271494e-05,
+      "loss": 1.1614,
+      "step": 4574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.914499033853423,
+      "learning_rate": 1.3927781410006058e-05,
+      "loss": 1.1304,
+      "step": 4575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.68408888531168,
+      "learning_rate": 1.3925239592839175e-05,
+      "loss": 1.1492,
+      "step": 4576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.202776219769488,
+      "learning_rate": 1.3922697475842908e-05,
+      "loss": 1.1953,
+      "step": 4577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5044240876543697,
+      "learning_rate": 1.3920155059211436e-05,
+      "loss": 1.1987,
+      "step": 4578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5414055336344106,
+      "learning_rate": 1.3917612343138962e-05,
+      "loss": 1.1331,
+      "step": 4579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.037252115870318,
+      "learning_rate": 1.3915069327819708e-05,
+      "loss": 1.3127,
+      "step": 4580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.791670300360647,
+      "learning_rate": 1.3912526013447924e-05,
+      "loss": 1.1812,
+      "step": 4581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6973260861345834,
+      "learning_rate": 1.3909982400217885e-05,
+      "loss": 1.1455,
+      "step": 4582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.7829307236268574,
+      "learning_rate": 1.3907438488323883e-05,
+      "loss": 1.2627,
+      "step": 4583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4242921698770674,
+      "learning_rate": 1.3904894277960233e-05,
+      "loss": 1.1643,
+      "step": 4584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3160707715620403,
+      "learning_rate": 1.3902349769321272e-05,
+      "loss": 1.2034,
+      "step": 4585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.962734455929159,
+      "learning_rate": 1.3899804962601367e-05,
+      "loss": 1.1409,
+      "step": 4586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1620285930705796,
+      "learning_rate": 1.38972598579949e-05,
+      "loss": 1.2195,
+      "step": 4587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0070846806821034,
+      "learning_rate": 1.3894714455696283e-05,
+      "loss": 1.1189,
+      "step": 4588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1846292639273592,
+      "learning_rate": 1.3892168755899944e-05,
+      "loss": 1.1577,
+      "step": 4589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6363342059569472,
+      "learning_rate": 1.3889622758800334e-05,
+      "loss": 1.1646,
+      "step": 4590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.422948328065412,
+      "learning_rate": 1.388707646459193e-05,
+      "loss": 1.2402,
+      "step": 4591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.423933898164423,
+      "learning_rate": 1.3884529873469232e-05,
+      "loss": 1.1282,
+      "step": 4592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.378149456798954,
+      "learning_rate": 1.3881982985626763e-05,
+      "loss": 1.166,
+      "step": 4593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1995337945969253,
+      "learning_rate": 1.387943580125906e-05,
+      "loss": 1.2024,
+      "step": 4594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.8464987376313795,
+      "learning_rate": 1.38768883205607e-05,
+      "loss": 1.2483,
+      "step": 4595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.828464863849577,
+      "learning_rate": 1.3874340543726263e-05,
+      "loss": 1.1226,
+      "step": 4596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7356277055829885,
+      "learning_rate": 1.3871792470950367e-05,
+      "loss": 1.1973,
+      "step": 4597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.5062149000189695,
+      "learning_rate": 1.3869244102427643e-05,
+      "loss": 1.1892,
+      "step": 4598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.30552350496484,
+      "learning_rate": 1.386669543835275e-05,
+      "loss": 1.2266,
+      "step": 4599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6651628417641784,
+      "learning_rate": 1.3864146478920366e-05,
+      "loss": 1.1799,
+      "step": 4600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3459939664624567,
+      "learning_rate": 1.3861597224325193e-05,
+      "loss": 1.1833,
+      "step": 4601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9917354131801686,
+      "learning_rate": 1.3859047674761961e-05,
+      "loss": 1.1458,
+      "step": 4602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.576870806483024,
+      "learning_rate": 1.385649783042541e-05,
+      "loss": 1.198,
+      "step": 4603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0393248803480595,
+      "learning_rate": 1.3853947691510314e-05,
+      "loss": 1.1033,
+      "step": 4604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.827644117605675,
+      "learning_rate": 1.3851397258211466e-05,
+      "loss": 1.1406,
+      "step": 4605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7352042893442396,
+      "learning_rate": 1.3848846530723678e-05,
+      "loss": 1.1013,
+      "step": 4606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.752109624606229,
+      "learning_rate": 1.3846295509241788e-05,
+      "loss": 1.1736,
+      "step": 4607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8441311963432887,
+      "learning_rate": 1.3843744193960658e-05,
+      "loss": 1.0889,
+      "step": 4608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.5771197430448,
+      "learning_rate": 1.3841192585075164e-05,
+      "loss": 1.1602,
+      "step": 4609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.146497251352854,
+      "learning_rate": 1.383864068278022e-05,
+      "loss": 1.1833,
+      "step": 4610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5548420394326867,
+      "learning_rate": 1.3836088487270746e-05,
+      "loss": 1.2041,
+      "step": 4611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.139866573231897,
+      "learning_rate": 1.3833535998741696e-05,
+      "loss": 1.1379,
+      "step": 4612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.382479988576492,
+      "learning_rate": 1.3830983217388037e-05,
+      "loss": 1.1826,
+      "step": 4613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.45225386297066,
+      "learning_rate": 1.3828430143404765e-05,
+      "loss": 1.2231,
+      "step": 4614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5841990226008726,
+      "learning_rate": 1.3825876776986899e-05,
+      "loss": 1.085,
+      "step": 4615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1278724825961026,
+      "learning_rate": 1.3823323118329478e-05,
+      "loss": 1.1228,
+      "step": 4616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6275317818442234,
+      "learning_rate": 1.3820769167627556e-05,
+      "loss": 1.1665,
+      "step": 4617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8638772161270993,
+      "learning_rate": 1.3818214925076226e-05,
+      "loss": 1.2126,
+      "step": 4618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1393934239010033,
+      "learning_rate": 1.3815660390870587e-05,
+      "loss": 1.1199,
+      "step": 4619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.111241496707044,
+      "learning_rate": 1.381310556520577e-05,
+      "loss": 1.3501,
+      "step": 4620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.419922720153591,
+      "learning_rate": 1.3810550448276925e-05,
+      "loss": 1.1541,
+      "step": 4621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.379503274553397,
+      "learning_rate": 1.3807995040279226e-05,
+      "loss": 1.2183,
+      "step": 4622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.946388875255518,
+      "learning_rate": 1.3805439341407864e-05,
+      "loss": 1.231,
+      "step": 4623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8303138221437747,
+      "learning_rate": 1.3802883351858057e-05,
+      "loss": 1.1401,
+      "step": 4624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8910729500963446,
+      "learning_rate": 1.3800327071825051e-05,
+      "loss": 1.1558,
+      "step": 4625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.034316743549051,
+      "learning_rate": 1.3797770501504098e-05,
+      "loss": 1.1023,
+      "step": 4626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1557764504236534,
+      "learning_rate": 1.3795213641090483e-05,
+      "loss": 1.1816,
+      "step": 4627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8475915263058758,
+      "learning_rate": 1.379265649077952e-05,
+      "loss": 1.1318,
+      "step": 4628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8431264872671385,
+      "learning_rate": 1.3790099050766528e-05,
+      "loss": 1.1609,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0525725740373435,
+      "learning_rate": 1.3787541321246863e-05,
+      "loss": 1.2288,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.256092597369357,
+      "learning_rate": 1.3784983302415888e-05,
+      "loss": 1.1638,
+      "step": 4631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.639781468797714,
+      "learning_rate": 1.378242499446901e-05,
+      "loss": 1.2122,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3840302887786984,
+      "learning_rate": 1.3779866397601635e-05,
+      "loss": 1.2061,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.830644199802011,
+      "learning_rate": 1.3777307512009208e-05,
+      "loss": 1.1543,
+      "step": 4634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.164546257731544,
+      "learning_rate": 1.3774748337887187e-05,
+      "loss": 1.1604,
+      "step": 4635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.546681540891158,
+      "learning_rate": 1.3772188875431054e-05,
+      "loss": 1.1868,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.034599976917278,
+      "learning_rate": 1.376962912483631e-05,
+      "loss": 1.1292,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8202430569270396,
+      "learning_rate": 1.376706908629849e-05,
+      "loss": 1.1333,
+      "step": 4638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.319604633973003,
+      "learning_rate": 1.3764508760013136e-05,
+      "loss": 1.1543,
+      "step": 4639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.989060761111098,
+      "learning_rate": 1.3761948146175819e-05,
+      "loss": 1.1501,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8821753888848742,
+      "learning_rate": 1.3759387244982134e-05,
+      "loss": 1.1306,
+      "step": 4641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5507395604700434,
+      "learning_rate": 1.3756826056627694e-05,
+      "loss": 1.1953,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.577756950177968,
+      "learning_rate": 1.3754264581308134e-05,
+      "loss": 1.1868,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7659828731876357,
+      "learning_rate": 1.3751702819219119e-05,
+      "loss": 1.1155,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5346577790829192,
+      "learning_rate": 1.3749140770556322e-05,
+      "loss": 1.189,
+      "step": 4645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2730516552562983,
+      "learning_rate": 1.3746578435515444e-05,
+      "loss": 1.2568,
+      "step": 4646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.316989614571838,
+      "learning_rate": 1.374401581429221e-05,
+      "loss": 1.1895,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.090077488798766,
+      "learning_rate": 1.3741452907082373e-05,
+      "loss": 1.1438,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.41408131182876,
+      "learning_rate": 1.3738889714081695e-05,
+      "loss": 1.1804,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1426700407601476,
+      "learning_rate": 1.3736326235485964e-05,
+      "loss": 1.1396,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.047822385227846,
+      "learning_rate": 1.3733762471490992e-05,
+      "loss": 1.1442,
+      "step": 4651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4036464371742396,
+      "learning_rate": 1.3731198422292615e-05,
+      "loss": 1.2327,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.853594900116216,
+      "learning_rate": 1.3728634088086687e-05,
+      "loss": 1.1624,
+      "step": 4653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9418775293372685,
+      "learning_rate": 1.3726069469069083e-05,
+      "loss": 1.1313,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0849550704652713,
+      "learning_rate": 1.3723504565435703e-05,
+      "loss": 1.2134,
+      "step": 4655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.922617196243661,
+      "learning_rate": 1.3720939377382463e-05,
+      "loss": 1.1926,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.342192228417242,
+      "learning_rate": 1.3718373905105313e-05,
+      "loss": 1.1101,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1402182390506908,
+      "learning_rate": 1.371580814880021e-05,
+      "loss": 1.1423,
+      "step": 4658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0140457310038298,
+      "learning_rate": 1.3713242108663139e-05,
+      "loss": 1.1533,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.016831468526674,
+      "learning_rate": 1.3710675784890108e-05,
+      "loss": 1.2612,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2096393650145423,
+      "learning_rate": 1.3708109177677148e-05,
+      "loss": 1.1514,
+      "step": 4661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9607606221011475,
+      "learning_rate": 1.3705542287220308e-05,
+      "loss": 1.125,
+      "step": 4662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.617086091188055,
+      "learning_rate": 1.3702975113715661e-05,
+      "loss": 1.1221,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.391829764720592,
+      "learning_rate": 1.3700407657359298e-05,
+      "loss": 1.1777,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.882507671545433,
+      "learning_rate": 1.3697839918347336e-05,
+      "loss": 1.2075,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.412577161860596,
+      "learning_rate": 1.369527189687591e-05,
+      "loss": 1.1421,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7081803903186183,
+      "learning_rate": 1.3692703593141181e-05,
+      "loss": 1.1062,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.030105726172417,
+      "learning_rate": 1.3690135007339327e-05,
+      "loss": 1.1638,
+      "step": 4668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.272675745986358,
+      "learning_rate": 1.3687566139666549e-05,
+      "loss": 1.1765,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2436629668002532,
+      "learning_rate": 1.3684996990319068e-05,
+      "loss": 1.1477,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2653567717902643,
+      "learning_rate": 1.3682427559493134e-05,
+      "loss": 1.1553,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.838931134153981,
+      "learning_rate": 1.367985784738501e-05,
+      "loss": 1.1812,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.846144767774393,
+      "learning_rate": 1.3677287854190984e-05,
+      "loss": 1.1667,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.448034530751306,
+      "learning_rate": 1.3674717580107361e-05,
+      "loss": 1.1604,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.332980790826464,
+      "learning_rate": 1.3672147025330477e-05,
+      "loss": 1.147,
+      "step": 4675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.986032853395581,
+      "learning_rate": 1.3669576190056682e-05,
+      "loss": 1.155,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4925985218432096,
+      "learning_rate": 1.3667005074482346e-05,
+      "loss": 1.1543,
+      "step": 4677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4221925661600614,
+      "learning_rate": 1.366443367880387e-05,
+      "loss": 1.1543,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0061702717761154,
+      "learning_rate": 1.3661862003217663e-05,
+      "loss": 1.1648,
+      "step": 4679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4046067469157864,
+      "learning_rate": 1.3659290047920166e-05,
+      "loss": 1.2002,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8093573746788687,
+      "learning_rate": 1.365671781310784e-05,
+      "loss": 1.1252,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.978808268100671,
+      "learning_rate": 1.3654145298977162e-05,
+      "loss": 1.1541,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2260256583622438,
+      "learning_rate": 1.3651572505724633e-05,
+      "loss": 1.2563,
+      "step": 4683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.358568225089442,
+      "learning_rate": 1.3648999433546775e-05,
+      "loss": 1.1875,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8515712870175047,
+      "learning_rate": 1.3646426082640139e-05,
+      "loss": 1.1936,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0443060802730897,
+      "learning_rate": 1.3643852453201282e-05,
+      "loss": 1.2144,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2779178098151887,
+      "learning_rate": 1.3641278545426793e-05,
+      "loss": 1.1548,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8315763331758337,
+      "learning_rate": 1.3638704359513285e-05,
+      "loss": 1.1729,
+      "step": 4688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.114959191183264,
+      "learning_rate": 1.3636129895657378e-05,
+      "loss": 1.1592,
+      "step": 4689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.763731988515936,
+      "learning_rate": 1.3633555154055733e-05,
+      "loss": 1.1421,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.227601011636839,
+      "learning_rate": 1.3630980134905012e-05,
+      "loss": 1.1602,
+      "step": 4691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.614244026069146,
+      "learning_rate": 1.3628404838401914e-05,
+      "loss": 1.166,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3079208078962656,
+      "learning_rate": 1.362582926474315e-05,
+      "loss": 1.1912,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.828196453259665,
+      "learning_rate": 1.3623253414125452e-05,
+      "loss": 1.125,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.49937437936879,
+      "learning_rate": 1.362067728674559e-05,
+      "loss": 1.1658,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.957995197776779,
+      "learning_rate": 1.3618100882800324e-05,
+      "loss": 1.166,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3070056424692162,
+      "learning_rate": 1.3615524202486463e-05,
+      "loss": 1.1543,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.528298661083473,
+      "learning_rate": 1.3612947246000824e-05,
+      "loss": 1.2483,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4420184319997906,
+      "learning_rate": 1.361037001354025e-05,
+      "loss": 1.1804,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.009911483778631,
+      "learning_rate": 1.3607792505301598e-05,
+      "loss": 1.1824,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2024347977655414,
+      "learning_rate": 1.3605214721481756e-05,
+      "loss": 1.2129,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.601174464155021,
+      "learning_rate": 1.3602636662277622e-05,
+      "loss": 1.1001,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6875745229660057,
+      "learning_rate": 1.3600058327886126e-05,
+      "loss": 1.1953,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.71018504556397,
+      "learning_rate": 1.3597479718504216e-05,
+      "loss": 1.1599,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.572361106584603,
+      "learning_rate": 1.3594900834328855e-05,
+      "loss": 1.1943,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6615793291516288,
+      "learning_rate": 1.359232167555703e-05,
+      "loss": 1.1255,
+      "step": 4706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.897543500508879,
+      "learning_rate": 1.358974224238575e-05,
+      "loss": 1.1606,
+      "step": 4707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5189669847522382,
+      "learning_rate": 1.3587162535012051e-05,
+      "loss": 1.1787,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8561238015831996,
+      "learning_rate": 1.3584582553632978e-05,
+      "loss": 1.124,
+      "step": 4709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7722923505356762,
+      "learning_rate": 1.3582002298445605e-05,
+      "loss": 1.1414,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.683471505608353,
+      "learning_rate": 1.3579421769647026e-05,
+      "loss": 1.2305,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.385036142372349,
+      "learning_rate": 1.357684096743435e-05,
+      "loss": 1.1804,
+      "step": 4712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.000058965954094,
+      "learning_rate": 1.3574259892004716e-05,
+      "loss": 1.1299,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1355801898081808,
+      "learning_rate": 1.357167854355528e-05,
+      "loss": 1.1555,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9417979306429496,
+      "learning_rate": 1.3569096922283216e-05,
+      "loss": 1.1465,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0089431767798493,
+      "learning_rate": 1.3566515028385718e-05,
+      "loss": 1.2373,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.435487334492873,
+      "learning_rate": 1.356393286206001e-05,
+      "loss": 1.1987,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.949918528421406,
+      "learning_rate": 1.3561350423503328e-05,
+      "loss": 1.1936,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1817012623604817,
+      "learning_rate": 1.3558767712912933e-05,
+      "loss": 1.1514,
+      "step": 4719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.252696603000407,
+      "learning_rate": 1.35561847304861e-05,
+      "loss": 1.2505,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.901957936612862,
+      "learning_rate": 1.355360147642014e-05,
+      "loss": 1.1792,
+      "step": 4721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.618635174532998,
+      "learning_rate": 1.3551017950912364e-05,
+      "loss": 1.145,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9626239692890217,
+      "learning_rate": 1.3548434154160122e-05,
+      "loss": 1.197,
+      "step": 4723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5221510569201095,
+      "learning_rate": 1.354585008636077e-05,
+      "loss": 1.1396,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.205339313025624,
+      "learning_rate": 1.3543265747711702e-05,
+      "loss": 1.156,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.512498925923348,
+      "learning_rate": 1.3540681138410317e-05,
+      "loss": 1.1858,
+      "step": 4726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1191055854830947,
+      "learning_rate": 1.353809625865404e-05,
+      "loss": 1.207,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.68670142217793,
+      "learning_rate": 1.3535511108640318e-05,
+      "loss": 1.1418,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.441556366605363,
+      "learning_rate": 1.3532925688566616e-05,
+      "loss": 1.0878,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6236741394625227,
+      "learning_rate": 1.3530339998630422e-05,
+      "loss": 1.1548,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8008986487383942,
+      "learning_rate": 1.3527754039029249e-05,
+      "loss": 1.1187,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.921860457745791,
+      "learning_rate": 1.3525167809960618e-05,
+      "loss": 1.1902,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.347516414026389,
+      "learning_rate": 1.3522581311622083e-05,
+      "loss": 1.1619,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3272264577531323,
+      "learning_rate": 1.3519994544211211e-05,
+      "loss": 1.1968,
+      "step": 4734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.2987282081476215,
+      "learning_rate": 1.3517407507925593e-05,
+      "loss": 1.1519,
+      "step": 4735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8732012763795236,
+      "learning_rate": 1.3514820202962838e-05,
+      "loss": 1.1101,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9305458806114952,
+      "learning_rate": 1.351223262952058e-05,
+      "loss": 1.1292,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.536768149431602,
+      "learning_rate": 1.3509644787796475e-05,
+      "loss": 1.1589,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1296160081345876,
+      "learning_rate": 1.3507056677988181e-05,
+      "loss": 1.1614,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3789935383243908,
+      "learning_rate": 1.3504468300293405e-05,
+      "loss": 1.1304,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0460685891596726,
+      "learning_rate": 1.3501879654909855e-05,
+      "loss": 1.1064,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9065701988615236,
+      "learning_rate": 1.3499290742035267e-05,
+      "loss": 1.1973,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7307640928209613,
+      "learning_rate": 1.349670156186739e-05,
+      "loss": 1.1333,
+      "step": 4743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9679339319142826,
+      "learning_rate": 1.3494112114604001e-05,
+      "loss": 1.2236,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7396737954731423,
+      "learning_rate": 1.3491522400442896e-05,
+      "loss": 1.2505,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1493374654045745,
+      "learning_rate": 1.348893241958189e-05,
+      "loss": 1.1799,
+      "step": 4746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9310570577260933,
+      "learning_rate": 1.3486342172218817e-05,
+      "loss": 1.1365,
+      "step": 4747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3132298233205306,
+      "learning_rate": 1.3483751658551535e-05,
+      "loss": 1.1375,
+      "step": 4748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1939352219181414,
+      "learning_rate": 1.3481160878777919e-05,
+      "loss": 1.1431,
+      "step": 4749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5845407875519655,
+      "learning_rate": 1.3478569833095864e-05,
+      "loss": 1.1819,
+      "step": 4750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9988085546497194,
+      "learning_rate": 1.3475978521703292e-05,
+      "loss": 1.1765,
+      "step": 4751
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.391798290977613,
+      "learning_rate": 1.347338694479814e-05,
+      "loss": 1.175,
+      "step": 4752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.294418645751069,
+      "learning_rate": 1.3470795102578358e-05,
+      "loss": 1.1719,
+      "step": 4753
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.151609644770071,
+      "learning_rate": 1.3468202995241932e-05,
+      "loss": 1.1309,
+      "step": 4754
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8032854649919945,
+      "learning_rate": 1.3465610622986854e-05,
+      "loss": 1.1418,
+      "step": 4755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.644404213997303,
+      "learning_rate": 1.3463017986011148e-05,
+      "loss": 1.2266,
+      "step": 4756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6014542406287533,
+      "learning_rate": 1.3460425084512849e-05,
+      "loss": 1.198,
+      "step": 4757
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4708283185919653,
+      "learning_rate": 1.3457831918690019e-05,
+      "loss": 1.2283,
+      "step": 4758
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.199014958867604,
+      "learning_rate": 1.345523848874073e-05,
+      "loss": 1.1528,
+      "step": 4759
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.397162251063706,
+      "learning_rate": 1.345264479486309e-05,
+      "loss": 1.1809,
+      "step": 4760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.929515776210842,
+      "learning_rate": 1.3450050837255212e-05,
+      "loss": 1.1875,
+      "step": 4761
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5369889580078473,
+      "learning_rate": 1.3447456616115238e-05,
+      "loss": 1.1455,
+      "step": 4762
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1512449719290268,
+      "learning_rate": 1.3444862131641326e-05,
+      "loss": 1.147,
+      "step": 4763
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4023908485662986,
+      "learning_rate": 1.3442267384031658e-05,
+      "loss": 1.1763,
+      "step": 4764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.346287138863818,
+      "learning_rate": 1.343967237348443e-05,
+      "loss": 1.2017,
+      "step": 4765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9635752388014067,
+      "learning_rate": 1.3437077100197866e-05,
+      "loss": 1.0947,
+      "step": 4766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4815605765786075,
+      "learning_rate": 1.3434481564370204e-05,
+      "loss": 1.1699,
+      "step": 4767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4776814611568487,
+      "learning_rate": 1.3431885766199704e-05,
+      "loss": 1.1602,
+      "step": 4768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.284547262722785,
+      "learning_rate": 1.3429289705884644e-05,
+      "loss": 1.1682,
+      "step": 4769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.164343801329804,
+      "learning_rate": 1.342669338362333e-05,
+      "loss": 1.1343,
+      "step": 4770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.505711064017492,
+      "learning_rate": 1.342409679961408e-05,
+      "loss": 1.2529,
+      "step": 4771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1877675731575126,
+      "learning_rate": 1.3421499954055225e-05,
+      "loss": 1.1587,
+      "step": 4772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1868581832762617,
+      "learning_rate": 1.3418902847145137e-05,
+      "loss": 1.146,
+      "step": 4773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.184213013405272,
+      "learning_rate": 1.3416305479082192e-05,
+      "loss": 1.1038,
+      "step": 4774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0164302712153153,
+      "learning_rate": 1.3413707850064787e-05,
+      "loss": 1.0762,
+      "step": 4775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.953722098230496,
+      "learning_rate": 1.3411109960291345e-05,
+      "loss": 1.1228,
+      "step": 4776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6365771478626097,
+      "learning_rate": 1.3408511809960305e-05,
+      "loss": 1.2239,
+      "step": 4777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6621617390239773,
+      "learning_rate": 1.3405913399270127e-05,
+      "loss": 1.1062,
+      "step": 4778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0549157772598923,
+      "learning_rate": 1.340331472841929e-05,
+      "loss": 1.1838,
+      "step": 4779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1938368188843,
+      "learning_rate": 1.3400715797606293e-05,
+      "loss": 1.1768,
+      "step": 4780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2559934577324148,
+      "learning_rate": 1.3398116607029657e-05,
+      "loss": 1.2209,
+      "step": 4781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9059214537259885,
+      "learning_rate": 1.3395517156887916e-05,
+      "loss": 1.1621,
+      "step": 4782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.462679307748834,
+      "learning_rate": 1.3392917447379636e-05,
+      "loss": 1.0964,
+      "step": 4783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3159860377317942,
+      "learning_rate": 1.3390317478703397e-05,
+      "loss": 1.1924,
+      "step": 4784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8530740135093176,
+      "learning_rate": 1.3387717251057788e-05,
+      "loss": 1.1267,
+      "step": 4785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5645162376930366,
+      "learning_rate": 1.3385116764641433e-05,
+      "loss": 1.1587,
+      "step": 4786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9314473134171948,
+      "learning_rate": 1.3382516019652968e-05,
+      "loss": 1.1326,
+      "step": 4787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.906952335041824,
+      "learning_rate": 1.3379915016291056e-05,
+      "loss": 1.1777,
+      "step": 4788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.160916403220615,
+      "learning_rate": 1.3377313754754373e-05,
+      "loss": 1.228,
+      "step": 4789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4199533783095086,
+      "learning_rate": 1.337471223524161e-05,
+      "loss": 1.2021,
+      "step": 4790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.999579138404074,
+      "learning_rate": 1.3372110457951495e-05,
+      "loss": 1.2461,
+      "step": 4791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0842293605170674,
+      "learning_rate": 1.336950842308275e-05,
+      "loss": 1.1714,
+      "step": 4792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1983915291200042,
+      "learning_rate": 1.3366906130834148e-05,
+      "loss": 1.1389,
+      "step": 4793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8879618086349894,
+      "learning_rate": 1.3364303581404453e-05,
+      "loss": 1.1597,
+      "step": 4794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9525781453440256,
+      "learning_rate": 1.3361700774992466e-05,
+      "loss": 1.1084,
+      "step": 4795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6879007205682073,
+      "learning_rate": 1.3359097711796997e-05,
+      "loss": 1.1565,
+      "step": 4796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.744476899995556,
+      "learning_rate": 1.335649439201689e-05,
+      "loss": 1.1436,
+      "step": 4797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.766531901992352,
+      "learning_rate": 1.3353890815850993e-05,
+      "loss": 1.1375,
+      "step": 4798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8186396552305384,
+      "learning_rate": 1.3351286983498182e-05,
+      "loss": 1.123,
+      "step": 4799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1175288374855654,
+      "learning_rate": 1.3348682895157348e-05,
+      "loss": 1.1733,
+      "step": 4800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.705324150940612,
+      "learning_rate": 1.3346078551027408e-05,
+      "loss": 1.1736,
+      "step": 4801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2369726282591036,
+      "learning_rate": 1.3343473951307294e-05,
+      "loss": 1.2148,
+      "step": 4802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.869463385513747,
+      "learning_rate": 1.3340869096195956e-05,
+      "loss": 1.1321,
+      "step": 4803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.746360232743619,
+      "learning_rate": 1.333826398589237e-05,
+      "loss": 1.0642,
+      "step": 4804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.280220379958206,
+      "learning_rate": 1.3335658620595521e-05,
+      "loss": 1.2822,
+      "step": 4805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2420626051815167,
+      "learning_rate": 1.3333053000504426e-05,
+      "loss": 1.1714,
+      "step": 4806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8138221876332943,
+      "learning_rate": 1.3330447125818115e-05,
+      "loss": 1.0999,
+      "step": 4807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0178838142165443,
+      "learning_rate": 1.3327840996735635e-05,
+      "loss": 1.1697,
+      "step": 4808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.757910139459983,
+      "learning_rate": 1.3325234613456054e-05,
+      "loss": 1.12,
+      "step": 4809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.155152796720883,
+      "learning_rate": 1.3322627976178462e-05,
+      "loss": 1.2029,
+      "step": 4810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.449618580974035,
+      "learning_rate": 1.3320021085101968e-05,
+      "loss": 1.1436,
+      "step": 4811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.912673758785974,
+      "learning_rate": 1.3317413940425698e-05,
+      "loss": 1.1707,
+      "step": 4812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0787520780442117,
+      "learning_rate": 1.33148065423488e-05,
+      "loss": 1.1384,
+      "step": 4813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2479421932919026,
+      "learning_rate": 1.331219889107044e-05,
+      "loss": 1.1318,
+      "step": 4814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5085437753553523,
+      "learning_rate": 1.3309590986789802e-05,
+      "loss": 1.1465,
+      "step": 4815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0703850713614944,
+      "learning_rate": 1.3306982829706093e-05,
+      "loss": 1.1152,
+      "step": 4816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7565020736743167,
+      "learning_rate": 1.3304374420018536e-05,
+      "loss": 1.1311,
+      "step": 4817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2050843009215093,
+      "learning_rate": 1.330176575792637e-05,
+      "loss": 1.1592,
+      "step": 4818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9743512365510756,
+      "learning_rate": 1.3299156843628865e-05,
+      "loss": 1.207,
+      "step": 4819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5016690378065545,
+      "learning_rate": 1.3296547677325298e-05,
+      "loss": 1.1523,
+      "step": 4820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6605262201685336,
+      "learning_rate": 1.3293938259214974e-05,
+      "loss": 1.2178,
+      "step": 4821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.385797332023569,
+      "learning_rate": 1.329132858949721e-05,
+      "loss": 1.1709,
+      "step": 4822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1334626517290283,
+      "learning_rate": 1.3288718668371345e-05,
+      "loss": 1.1282,
+      "step": 4823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.021528006749052,
+      "learning_rate": 1.3286108496036741e-05,
+      "loss": 1.1458,
+      "step": 4824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9791623290090943,
+      "learning_rate": 1.3283498072692771e-05,
+      "loss": 1.1621,
+      "step": 4825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9952167680067587,
+      "learning_rate": 1.3280887398538842e-05,
+      "loss": 1.1218,
+      "step": 4826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2884293413144854,
+      "learning_rate": 1.327827647377436e-05,
+      "loss": 1.2114,
+      "step": 4827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.087056232130776,
+      "learning_rate": 1.3275665298598763e-05,
+      "loss": 1.1187,
+      "step": 4828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5584050476833173,
+      "learning_rate": 1.3273053873211508e-05,
+      "loss": 1.2065,
+      "step": 4829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.387271548208437,
+      "learning_rate": 1.327044219781207e-05,
+      "loss": 1.1772,
+      "step": 4830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.382346669052774,
+      "learning_rate": 1.3267830272599938e-05,
+      "loss": 1.1296,
+      "step": 4831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.110380972808534,
+      "learning_rate": 1.3265218097774623e-05,
+      "loss": 1.1501,
+      "step": 4832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.300407707363388,
+      "learning_rate": 1.326260567353566e-05,
+      "loss": 1.1443,
+      "step": 4833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.601786205187492,
+      "learning_rate": 1.3259993000082599e-05,
+      "loss": 1.176,
+      "step": 4834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.480635314973814,
+      "learning_rate": 1.3257380077615006e-05,
+      "loss": 1.1638,
+      "step": 4835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7076565563349058,
+      "learning_rate": 1.3254766906332468e-05,
+      "loss": 1.1978,
+      "step": 4836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.066333797838064,
+      "learning_rate": 1.3252153486434597e-05,
+      "loss": 1.1797,
+      "step": 4837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2406201183299337,
+      "learning_rate": 1.3249539818121015e-05,
+      "loss": 1.106,
+      "step": 4838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.135638858853358,
+      "learning_rate": 1.324692590159137e-05,
+      "loss": 1.0959,
+      "step": 4839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.993226688377042,
+      "learning_rate": 1.3244311737045324e-05,
+      "loss": 1.1467,
+      "step": 4840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4123007171262336,
+      "learning_rate": 1.3241697324682561e-05,
+      "loss": 1.2395,
+      "step": 4841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8152800237225266,
+      "learning_rate": 1.323908266470278e-05,
+      "loss": 1.1689,
+      "step": 4842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1865264224436807,
+      "learning_rate": 1.3236467757305704e-05,
+      "loss": 1.1616,
+      "step": 4843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.879313536283189,
+      "learning_rate": 1.3233852602691075e-05,
+      "loss": 1.1326,
+      "step": 4844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.094945427680874,
+      "learning_rate": 1.3231237201058651e-05,
+      "loss": 1.1262,
+      "step": 4845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0889898222911647,
+      "learning_rate": 1.3228621552608203e-05,
+      "loss": 1.1875,
+      "step": 4846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1830832854678666,
+      "learning_rate": 1.3226005657539536e-05,
+      "loss": 1.1611,
+      "step": 4847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3159500026995197,
+      "learning_rate": 1.3223389516052456e-05,
+      "loss": 1.2131,
+      "step": 4848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0628028760609376,
+      "learning_rate": 1.3220773128346806e-05,
+      "loss": 1.1787,
+      "step": 4849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9016771298635735,
+      "learning_rate": 1.3218156494622434e-05,
+      "loss": 1.1067,
+      "step": 4850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8997619505008907,
+      "learning_rate": 1.3215539615079213e-05,
+      "loss": 1.1899,
+      "step": 4851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.077402530519401,
+      "learning_rate": 1.321292248991703e-05,
+      "loss": 1.179,
+      "step": 4852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2799763620929947,
+      "learning_rate": 1.3210305119335799e-05,
+      "loss": 1.1968,
+      "step": 4853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1573702186305272,
+      "learning_rate": 1.3207687503535448e-05,
+      "loss": 1.2151,
+      "step": 4854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.616287597113604,
+      "learning_rate": 1.3205069642715915e-05,
+      "loss": 1.1458,
+      "step": 4855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.582951353703879,
+      "learning_rate": 1.3202451537077175e-05,
+      "loss": 1.1455,
+      "step": 4856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8407212158470143,
+      "learning_rate": 1.3199833186819209e-05,
+      "loss": 1.1736,
+      "step": 4857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.636387176525424,
+      "learning_rate": 1.3197214592142015e-05,
+      "loss": 1.1162,
+      "step": 4858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4446464391317377,
+      "learning_rate": 1.3194595753245623e-05,
+      "loss": 1.145,
+      "step": 4859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.001297785055292,
+      "learning_rate": 1.3191976670330066e-05,
+      "loss": 1.1736,
+      "step": 4860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.606495904980802,
+      "learning_rate": 1.3189357343595405e-05,
+      "loss": 1.188,
+      "step": 4861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9340037742559155,
+      "learning_rate": 1.3186737773241719e-05,
+      "loss": 1.1841,
+      "step": 4862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9170999894976664,
+      "learning_rate": 1.31841179594691e-05,
+      "loss": 1.1587,
+      "step": 4863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.630534447558745,
+      "learning_rate": 1.3181497902477665e-05,
+      "loss": 1.095,
+      "step": 4864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0895876184514917,
+      "learning_rate": 1.3178877602467546e-05,
+      "loss": 1.1372,
+      "step": 4865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7241624743369433,
+      "learning_rate": 1.3176257059638892e-05,
+      "loss": 1.1018,
+      "step": 4866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4825226973879224,
+      "learning_rate": 1.3173636274191884e-05,
+      "loss": 1.1519,
+      "step": 4867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3005134918306966,
+      "learning_rate": 1.3171015246326699e-05,
+      "loss": 1.1855,
+      "step": 4868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0760129719518976,
+      "learning_rate": 1.3168393976243546e-05,
+      "loss": 1.1384,
+      "step": 4869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5937786732182837,
+      "learning_rate": 1.3165772464142654e-05,
+      "loss": 1.1899,
+      "step": 4870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.912298675437781,
+      "learning_rate": 1.3163150710224266e-05,
+      "loss": 1.1948,
+      "step": 4871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.847657878144399,
+      "learning_rate": 1.3160528714688642e-05,
+      "loss": 1.1526,
+      "step": 4872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4042552685151133,
+      "learning_rate": 1.315790647773607e-05,
+      "loss": 1.1436,
+      "step": 4873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9202933711453105,
+      "learning_rate": 1.3155283999566846e-05,
+      "loss": 1.1167,
+      "step": 4874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1471889835815716,
+      "learning_rate": 1.3152661280381282e-05,
+      "loss": 1.1658,
+      "step": 4875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0095679667226842,
+      "learning_rate": 1.3150038320379723e-05,
+      "loss": 1.1763,
+      "step": 4876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.003737439011941,
+      "learning_rate": 1.314741511976252e-05,
+      "loss": 1.0857,
+      "step": 4877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2040911788853563,
+      "learning_rate": 1.3144791678730047e-05,
+      "loss": 1.1733,
+      "step": 4878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3978843227788764,
+      "learning_rate": 1.3142167997482693e-05,
+      "loss": 1.1963,
+      "step": 4879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.706653190128881,
+      "learning_rate": 1.3139544076220872e-05,
+      "loss": 1.1631,
+      "step": 4880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8529798946859883,
+      "learning_rate": 1.3136919915145008e-05,
+      "loss": 1.2156,
+      "step": 4881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.044000137953337,
+      "learning_rate": 1.3134295514455552e-05,
+      "loss": 1.199,
+      "step": 4882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7771741186454717,
+      "learning_rate": 1.3131670874352965e-05,
+      "loss": 1.1602,
+      "step": 4883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4544113048527776,
+      "learning_rate": 1.312904599503773e-05,
+      "loss": 1.1318,
+      "step": 4884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2907150292579144,
+      "learning_rate": 1.3126420876710352e-05,
+      "loss": 1.1292,
+      "step": 4885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1060232519134865,
+      "learning_rate": 1.3123795519571347e-05,
+      "loss": 1.1218,
+      "step": 4886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4161697650635694,
+      "learning_rate": 1.3121169923821257e-05,
+      "loss": 1.1309,
+      "step": 4887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.821792016170263,
+      "learning_rate": 1.3118544089660635e-05,
+      "loss": 1.1075,
+      "step": 4888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3574451908413767,
+      "learning_rate": 1.3115918017290052e-05,
+      "loss": 1.167,
+      "step": 4889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2393703243489638,
+      "learning_rate": 1.3113291706910107e-05,
+      "loss": 1.2236,
+      "step": 4890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1535638356029154,
+      "learning_rate": 1.3110665158721407e-05,
+      "loss": 1.1611,
+      "step": 4891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.584961175853625,
+      "learning_rate": 1.3108038372924586e-05,
+      "loss": 1.1858,
+      "step": 4892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7966767152900602,
+      "learning_rate": 1.310541134972028e-05,
+      "loss": 1.1155,
+      "step": 4893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.874717073442126,
+      "learning_rate": 1.3102784089309166e-05,
+      "loss": 1.1379,
+      "step": 4894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4730261474527593,
+      "learning_rate": 1.3100156591891919e-05,
+      "loss": 1.125,
+      "step": 4895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.00611734328989,
+      "learning_rate": 1.3097528857669248e-05,
+      "loss": 1.2002,
+      "step": 4896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.593782436766089,
+      "learning_rate": 1.3094900886841867e-05,
+      "loss": 1.1145,
+      "step": 4897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.00232498531349,
+      "learning_rate": 1.3092272679610512e-05,
+      "loss": 1.197,
+      "step": 4898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1371611381766047,
+      "learning_rate": 1.308964423617594e-05,
+      "loss": 1.1555,
+      "step": 4899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6583003849943667,
+      "learning_rate": 1.3087015556738928e-05,
+      "loss": 1.2148,
+      "step": 4900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1497190556961807,
+      "learning_rate": 1.3084386641500268e-05,
+      "loss": 1.157,
+      "step": 4901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.166763100571026,
+      "learning_rate": 1.3081757490660763e-05,
+      "loss": 1.2039,
+      "step": 4902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2229402250826684,
+      "learning_rate": 1.3079128104421246e-05,
+      "loss": 1.1653,
+      "step": 4903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.344266740349523,
+      "learning_rate": 1.3076498482982563e-05,
+      "loss": 1.1533,
+      "step": 4904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0414430140941784,
+      "learning_rate": 1.3073868626545576e-05,
+      "loss": 1.1809,
+      "step": 4905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9945399725835733,
+      "learning_rate": 1.3071238535311164e-05,
+      "loss": 1.1743,
+      "step": 4906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9790542621261302,
+      "learning_rate": 1.306860820948023e-05,
+      "loss": 1.1616,
+      "step": 4907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3527873720534673,
+      "learning_rate": 1.3065977649253692e-05,
+      "loss": 1.1438,
+      "step": 4908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.267602774438157,
+      "learning_rate": 1.306334685483248e-05,
+      "loss": 1.1658,
+      "step": 4909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4982065849783126,
+      "learning_rate": 1.3060715826417557e-05,
+      "loss": 1.1663,
+      "step": 4910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.985347750607978,
+      "learning_rate": 1.3058084564209885e-05,
+      "loss": 1.1409,
+      "step": 4911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0979086953742603,
+      "learning_rate": 1.3055453068410455e-05,
+      "loss": 1.1597,
+      "step": 4912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.090900029303959,
+      "learning_rate": 1.3052821339220274e-05,
+      "loss": 1.0825,
+      "step": 4913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6795088869301393,
+      "learning_rate": 1.305018937684037e-05,
+      "loss": 1.1353,
+      "step": 4914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0024352561377747,
+      "learning_rate": 1.3047557181471784e-05,
+      "loss": 1.2173,
+      "step": 4915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6148940178092794,
+      "learning_rate": 1.304492475331557e-05,
+      "loss": 1.2407,
+      "step": 4916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5010200436604033,
+      "learning_rate": 1.3042292092572817e-05,
+      "loss": 1.1304,
+      "step": 4917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9721582946663774,
+      "learning_rate": 1.303965919944461e-05,
+      "loss": 1.21,
+      "step": 4918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1451784380120142,
+      "learning_rate": 1.3037026074132068e-05,
+      "loss": 1.1257,
+      "step": 4919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.725626448939737,
+      "learning_rate": 1.3034392716836326e-05,
+      "loss": 1.1511,
+      "step": 4920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.323606572280457,
+      "learning_rate": 1.3031759127758524e-05,
+      "loss": 1.1523,
+      "step": 4921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.168301483283134,
+      "learning_rate": 1.3029125307099832e-05,
+      "loss": 1.1321,
+      "step": 4922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6606235020158016,
+      "learning_rate": 1.3026491255061438e-05,
+      "loss": 1.1978,
+      "step": 4923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9844909503254353,
+      "learning_rate": 1.3023856971844543e-05,
+      "loss": 1.2065,
+      "step": 4924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1706642589264256,
+      "learning_rate": 1.3021222457650365e-05,
+      "loss": 1.1904,
+      "step": 4925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8878131844978276,
+      "learning_rate": 1.3018587712680137e-05,
+      "loss": 1.156,
+      "step": 4926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1678093549384543,
+      "learning_rate": 1.3015952737135124e-05,
+      "loss": 1.1558,
+      "step": 4927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0916274281702507,
+      "learning_rate": 1.3013317531216592e-05,
+      "loss": 1.1116,
+      "step": 4928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.410029264480328,
+      "learning_rate": 1.301068209512583e-05,
+      "loss": 1.1384,
+      "step": 4929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3424691190729474,
+      "learning_rate": 1.3008046429064152e-05,
+      "loss": 1.168,
+      "step": 4930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.986945412509384,
+      "learning_rate": 1.3005410533232876e-05,
+      "loss": 1.1426,
+      "step": 4931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6452846354633963,
+      "learning_rate": 1.3002774407833348e-05,
+      "loss": 1.1094,
+      "step": 4932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8172839311010427,
+      "learning_rate": 1.3000138053066931e-05,
+      "loss": 1.1316,
+      "step": 4933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9080641059524104,
+      "learning_rate": 1.2997501469135002e-05,
+      "loss": 1.1458,
+      "step": 4934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.648093720910266,
+      "learning_rate": 1.299486465623895e-05,
+      "loss": 1.0784,
+      "step": 4935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0001848668021593,
+      "learning_rate": 1.2992227614580196e-05,
+      "loss": 1.1195,
+      "step": 4936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6867405586220077,
+      "learning_rate": 1.2989590344360168e-05,
+      "loss": 1.1282,
+      "step": 4937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.891624777500039,
+      "learning_rate": 1.2986952845780314e-05,
+      "loss": 1.1489,
+      "step": 4938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7850515423209576,
+      "learning_rate": 1.2984315119042095e-05,
+      "loss": 1.1252,
+      "step": 4939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0645469291947016,
+      "learning_rate": 1.2981677164347e-05,
+      "loss": 1.1619,
+      "step": 4940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2404443873695765,
+      "learning_rate": 1.2979038981896526e-05,
+      "loss": 1.2041,
+      "step": 4941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7810283926545796,
+      "learning_rate": 1.2976400571892189e-05,
+      "loss": 1.1298,
+      "step": 4942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.672981135283822,
+      "learning_rate": 1.2973761934535531e-05,
+      "loss": 1.1299,
+      "step": 4943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0560172450033796,
+      "learning_rate": 1.2971123070028093e-05,
+      "loss": 1.2,
+      "step": 4944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.249981197545792,
+      "learning_rate": 1.2968483978571453e-05,
+      "loss": 1.1619,
+      "step": 4945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.313675409995508,
+      "learning_rate": 1.2965844660367197e-05,
+      "loss": 1.1648,
+      "step": 4946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3640642766553595,
+      "learning_rate": 1.2963205115616928e-05,
+      "loss": 1.1694,
+      "step": 4947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3927480145726676,
+      "learning_rate": 1.2960565344522268e-05,
+      "loss": 1.1489,
+      "step": 4948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.643316072754348,
+      "learning_rate": 1.2957925347284854e-05,
+      "loss": 1.1362,
+      "step": 4949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.21336563470184,
+      "learning_rate": 1.2955285124106344e-05,
+      "loss": 1.165,
+      "step": 4950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8056748209931675,
+      "learning_rate": 1.2952644675188413e-05,
+      "loss": 1.166,
+      "step": 4951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.177731439387461,
+      "learning_rate": 1.295000400073275e-05,
+      "loss": 1.1899,
+      "step": 4952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6287843918717684,
+      "learning_rate": 1.2947363100941062e-05,
+      "loss": 1.1196,
+      "step": 4953
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.725883355753921,
+      "learning_rate": 1.2944721976015071e-05,
+      "loss": 1.2036,
+      "step": 4954
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0127992224123754,
+      "learning_rate": 1.2942080626156526e-05,
+      "loss": 1.1375,
+      "step": 4955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.538129888504257,
+      "learning_rate": 1.2939439051567185e-05,
+      "loss": 1.1824,
+      "step": 4956
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.919807385190352,
+      "learning_rate": 1.2936797252448823e-05,
+      "loss": 1.147,
+      "step": 4957
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.28813363432385,
+      "learning_rate": 1.293415522900323e-05,
+      "loss": 1.082,
+      "step": 4958
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.4339102016644025,
+      "learning_rate": 1.2931512981432223e-05,
+      "loss": 1.2124,
+      "step": 4959
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.917954261230994,
+      "learning_rate": 1.2928870509937633e-05,
+      "loss": 1.2014,
+      "step": 4960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.404007246957366,
+      "learning_rate": 1.2926227814721295e-05,
+      "loss": 1.1973,
+      "step": 4961
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.462242988918543,
+      "learning_rate": 1.2923584895985077e-05,
+      "loss": 1.2007,
+      "step": 4962
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.722302734890782,
+      "learning_rate": 1.292094175393086e-05,
+      "loss": 1.1863,
+      "step": 4963
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5898019089818427,
+      "learning_rate": 1.2918298388760536e-05,
+      "loss": 1.2043,
+      "step": 4964
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.005515967254259,
+      "learning_rate": 1.2915654800676018e-05,
+      "loss": 1.1616,
+      "step": 4965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.424902006936322,
+      "learning_rate": 1.2913010989879244e-05,
+      "loss": 1.0835,
+      "step": 4966
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3992005173597155,
+      "learning_rate": 1.2910366956572156e-05,
+      "loss": 1.2053,
+      "step": 4967
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9442367295673444,
+      "learning_rate": 1.2907722700956715e-05,
+      "loss": 1.1528,
+      "step": 4968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.569392363738854,
+      "learning_rate": 1.2905078223234907e-05,
+      "loss": 1.0317,
+      "step": 4969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2929278919291685,
+      "learning_rate": 1.2902433523608733e-05,
+      "loss": 1.1902,
+      "step": 4970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.478564917118249,
+      "learning_rate": 1.2899788602280206e-05,
+      "loss": 1.2166,
+      "step": 4971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.754979569403951,
+      "learning_rate": 1.2897143459451353e-05,
+      "loss": 1.1892,
+      "step": 4972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6176465677926877,
+      "learning_rate": 1.289449809532423e-05,
+      "loss": 1.175,
+      "step": 4973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.04988905949247,
+      "learning_rate": 1.28918525101009e-05,
+      "loss": 1.1121,
+      "step": 4974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8070269041743097,
+      "learning_rate": 1.2889206703983448e-05,
+      "loss": 1.1809,
+      "step": 4975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2731447379130985,
+      "learning_rate": 1.288656067717397e-05,
+      "loss": 1.1384,
+      "step": 4976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.13987167279301,
+      "learning_rate": 1.2883914429874586e-05,
+      "loss": 1.1567,
+      "step": 4977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3118991705442737,
+      "learning_rate": 1.288126796228743e-05,
+      "loss": 1.1501,
+      "step": 4978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.328730177067416,
+      "learning_rate": 1.2878621274614651e-05,
+      "loss": 1.075,
+      "step": 4979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.56473570297061,
+      "learning_rate": 1.287597436705842e-05,
+      "loss": 1.2136,
+      "step": 4980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.034918417688908,
+      "learning_rate": 1.2873327239820912e-05,
+      "loss": 1.2449,
+      "step": 4981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.204547660256311,
+      "learning_rate": 1.2870679893104334e-05,
+      "loss": 1.249,
+      "step": 4982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.3417602905176995,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 1.1865,
+      "step": 4983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.048860074286175,
+      "learning_rate": 1.2865384542042856e-05,
+      "loss": 1.2444,
+      "step": 4984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4675068266023095,
+      "learning_rate": 1.2862736538102438e-05,
+      "loss": 1.137,
+      "step": 4985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.277990584749237,
+      "learning_rate": 1.286008831549192e-05,
+      "loss": 1.1108,
+      "step": 4986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0793425924348496,
+      "learning_rate": 1.2857439874413589e-05,
+      "loss": 1.1599,
+      "step": 4987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2264853381947516,
+      "learning_rate": 1.2854791215069742e-05,
+      "loss": 1.1753,
+      "step": 4988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.171691643600301,
+      "learning_rate": 1.2852142337662698e-05,
+      "loss": 1.207,
+      "step": 4989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.906735550949478,
+      "learning_rate": 1.2849493242394796e-05,
+      "loss": 1.1836,
+      "step": 4990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0221380508090325,
+      "learning_rate": 1.2846843929468375e-05,
+      "loss": 1.1033,
+      "step": 4991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.857129659207635,
+      "learning_rate": 1.2844194399085814e-05,
+      "loss": 1.1863,
+      "step": 4992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8629467371621753,
+      "learning_rate": 1.2841544651449498e-05,
+      "loss": 1.0205,
+      "step": 4993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5080425698354354,
+      "learning_rate": 1.283889468676182e-05,
+      "loss": 1.1772,
+      "step": 4994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.694494189715069,
+      "learning_rate": 1.2836244505225201e-05,
+      "loss": 1.1973,
+      "step": 4995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8932905437868253,
+      "learning_rate": 1.2833594107042078e-05,
+      "loss": 1.1089,
+      "step": 4996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9618556398207776,
+      "learning_rate": 1.2830943492414898e-05,
+      "loss": 1.1244,
+      "step": 4997
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.230185341194501,
+      "learning_rate": 1.2828292661546127e-05,
+      "loss": 1.1333,
+      "step": 4998
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.792490811152882,
+      "learning_rate": 1.2825641614638254e-05,
+      "loss": 1.1077,
+      "step": 4999
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.927810715626313,
+      "learning_rate": 1.2822990351893778e-05,
+      "loss": 1.2324,
+      "step": 5000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.152595036363865,
+      "learning_rate": 1.2820338873515211e-05,
+      "loss": 1.1362,
+      "step": 5001
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2083982892645113,
+      "learning_rate": 1.2817687179705087e-05,
+      "loss": 1.2217,
+      "step": 5002
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5629157838820826,
+      "learning_rate": 1.281503527066596e-05,
+      "loss": 1.1846,
+      "step": 5003
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3053183237453276,
+      "learning_rate": 1.2812383146600396e-05,
+      "loss": 1.1672,
+      "step": 5004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.823561850616771,
+      "learning_rate": 1.2809730807710973e-05,
+      "loss": 1.1665,
+      "step": 5005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.70088276784508,
+      "learning_rate": 1.2807078254200294e-05,
+      "loss": 1.1611,
+      "step": 5006
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.090272478469186,
+      "learning_rate": 1.2804425486270968e-05,
+      "loss": 1.1663,
+      "step": 5007
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.073079281775239,
+      "learning_rate": 1.2801772504125636e-05,
+      "loss": 1.1523,
+      "step": 5008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6057126829591666,
+      "learning_rate": 1.2799119307966937e-05,
+      "loss": 1.1633,
+      "step": 5009
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.933527956812958,
+      "learning_rate": 1.279646589799754e-05,
+      "loss": 1.1553,
+      "step": 5010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9619845791631367,
+      "learning_rate": 1.2793812274420127e-05,
+      "loss": 1.1409,
+      "step": 5011
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.044068221886909,
+      "learning_rate": 1.2791158437437391e-05,
+      "loss": 1.158,
+      "step": 5012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1329509207029287,
+      "learning_rate": 1.2788504387252053e-05,
+      "loss": 1.1902,
+      "step": 5013
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0564203859393047,
+      "learning_rate": 1.2785850124066829e-05,
+      "loss": 1.1423,
+      "step": 5014
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4341531106423475,
+      "learning_rate": 1.2783195648084477e-05,
+      "loss": 1.1128,
+      "step": 5015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9726444719108813,
+      "learning_rate": 1.2780540959507754e-05,
+      "loss": 1.0823,
+      "step": 5016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.622896775249351,
+      "learning_rate": 1.277788605853944e-05,
+      "loss": 1.1787,
+      "step": 5017
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.479956962947106,
+      "learning_rate": 1.2775230945382329e-05,
+      "loss": 1.2083,
+      "step": 5018
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9679984508411748,
+      "learning_rate": 1.2772575620239229e-05,
+      "loss": 1.1675,
+      "step": 5019
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1932634686291674,
+      "learning_rate": 1.2769920083312972e-05,
+      "loss": 1.1526,
+      "step": 5020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6993715849847915,
+      "learning_rate": 1.27672643348064e-05,
+      "loss": 1.1553,
+      "step": 5021
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6663413027900433,
+      "learning_rate": 1.276460837492237e-05,
+      "loss": 1.1396,
+      "step": 5022
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.527928080239887,
+      "learning_rate": 1.2761952203863759e-05,
+      "loss": 1.1919,
+      "step": 5023
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0528262439383846,
+      "learning_rate": 1.2759295821833457e-05,
+      "loss": 1.1848,
+      "step": 5024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8955833840921463,
+      "learning_rate": 1.275663922903437e-05,
+      "loss": 1.1185,
+      "step": 5025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.256250524884072,
+      "learning_rate": 1.275398242566943e-05,
+      "loss": 1.2253,
+      "step": 5026
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4697046191978864,
+      "learning_rate": 1.2751325411941571e-05,
+      "loss": 1.1335,
+      "step": 5027
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.982279719647766,
+      "learning_rate": 1.2748668188053748e-05,
+      "loss": 1.1821,
+      "step": 5028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2481952015172078,
+      "learning_rate": 1.2746010754208933e-05,
+      "loss": 1.1738,
+      "step": 5029
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.566382281801547,
+      "learning_rate": 1.274335311061012e-05,
+      "loss": 1.1826,
+      "step": 5030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.120494700636622,
+      "learning_rate": 1.2740695257460307e-05,
+      "loss": 1.1484,
+      "step": 5031
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.7043077021012305,
+      "learning_rate": 1.2738037194962516e-05,
+      "loss": 1.2,
+      "step": 5032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8105224880374706,
+      "learning_rate": 1.273537892331978e-05,
+      "loss": 1.1509,
+      "step": 5033
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.238039924963022,
+      "learning_rate": 1.2732720442735155e-05,
+      "loss": 1.1875,
+      "step": 5034
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.054189285616242,
+      "learning_rate": 1.273006175341171e-05,
+      "loss": 1.1136,
+      "step": 5035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.27024327314849,
+      "learning_rate": 1.2727402855552527e-05,
+      "loss": 1.1255,
+      "step": 5036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2538153219512673,
+      "learning_rate": 1.2724743749360707e-05,
+      "loss": 1.1113,
+      "step": 5037
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4145675676768636,
+      "learning_rate": 1.2722084435039362e-05,
+      "loss": 1.2393,
+      "step": 5038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8492918546835897,
+      "learning_rate": 1.2719424912791625e-05,
+      "loss": 1.1328,
+      "step": 5039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.747439090607054,
+      "learning_rate": 1.2716765182820648e-05,
+      "loss": 1.1211,
+      "step": 5040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6903284778354264,
+      "learning_rate": 1.2714105245329591e-05,
+      "loss": 1.1616,
+      "step": 5041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.247939089376175,
+      "learning_rate": 1.2711445100521631e-05,
+      "loss": 1.2827,
+      "step": 5042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.126802780973675,
+      "learning_rate": 1.2708784748599969e-05,
+      "loss": 1.147,
+      "step": 5043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8157206160748616,
+      "learning_rate": 1.270612418976781e-05,
+      "loss": 1.0371,
+      "step": 5044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5178267977095827,
+      "learning_rate": 1.2703463424228386e-05,
+      "loss": 1.1084,
+      "step": 5045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4319253937964374,
+      "learning_rate": 1.2700802452184937e-05,
+      "loss": 1.2112,
+      "step": 5046
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5508074452397302,
+      "learning_rate": 1.2698141273840719e-05,
+      "loss": 1.2153,
+      "step": 5047
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1534143321923094,
+      "learning_rate": 1.269547988939901e-05,
+      "loss": 1.1968,
+      "step": 5048
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.261958973141832,
+      "learning_rate": 1.2692818299063094e-05,
+      "loss": 1.1951,
+      "step": 5049
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.175419520115131,
+      "learning_rate": 1.2690156503036288e-05,
+      "loss": 1.1082,
+      "step": 5050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7472477762179994,
+      "learning_rate": 1.2687494501521901e-05,
+      "loss": 1.1526,
+      "step": 5051
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.447573230408593,
+      "learning_rate": 1.2684832294723278e-05,
+      "loss": 1.092,
+      "step": 5052
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.986025773807248,
+      "learning_rate": 1.2682169882843766e-05,
+      "loss": 1.1089,
+      "step": 5053
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2379942729146807,
+      "learning_rate": 1.2679507266086737e-05,
+      "loss": 1.2144,
+      "step": 5054
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.495232977205506,
+      "learning_rate": 1.2676844444655576e-05,
+      "loss": 1.187,
+      "step": 5055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.922521607771725,
+      "learning_rate": 1.267418141875368e-05,
+      "loss": 1.1338,
+      "step": 5056
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8811264076774856,
+      "learning_rate": 1.2671518188584462e-05,
+      "loss": 1.2075,
+      "step": 5057
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.223468221618806,
+      "learning_rate": 1.2668854754351357e-05,
+      "loss": 1.1169,
+      "step": 5058
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6576448405401214,
+      "learning_rate": 1.266619111625781e-05,
+      "loss": 1.1614,
+      "step": 5059
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.143596622522705,
+      "learning_rate": 1.2663527274507283e-05,
+      "loss": 1.1794,
+      "step": 5060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.168853678502583,
+      "learning_rate": 1.2660863229303254e-05,
+      "loss": 1.1865,
+      "step": 5061
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.292482530895042,
+      "learning_rate": 1.2658198980849213e-05,
+      "loss": 1.1372,
+      "step": 5062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.521169955900445,
+      "learning_rate": 1.2655534529348672e-05,
+      "loss": 1.1428,
+      "step": 5063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.342037617013365,
+      "learning_rate": 1.2652869875005156e-05,
+      "loss": 1.209,
+      "step": 5064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.042951600494098,
+      "learning_rate": 1.2650205018022201e-05,
+      "loss": 1.1726,
+      "step": 5065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.366519274131682,
+      "learning_rate": 1.2647539958603365e-05,
+      "loss": 1.1128,
+      "step": 5066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.504314494843474,
+      "learning_rate": 1.2644874696952214e-05,
+      "loss": 1.1072,
+      "step": 5067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1641941874674235,
+      "learning_rate": 1.264220923327234e-05,
+      "loss": 1.1968,
+      "step": 5068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8179864901933964,
+      "learning_rate": 1.2639543567767346e-05,
+      "loss": 1.093,
+      "step": 5069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7054464233070847,
+      "learning_rate": 1.2636877700640837e-05,
+      "loss": 1.1516,
+      "step": 5070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8531137491212752,
+      "learning_rate": 1.2634211632096457e-05,
+      "loss": 1.1497,
+      "step": 5071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5701385354323554,
+      "learning_rate": 1.2631545362337848e-05,
+      "loss": 1.1248,
+      "step": 5072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7785208659459584,
+      "learning_rate": 1.2628878891568674e-05,
+      "loss": 1.1257,
+      "step": 5073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2595328958028875,
+      "learning_rate": 1.2626212219992617e-05,
+      "loss": 1.2029,
+      "step": 5074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.307037973991311,
+      "learning_rate": 1.2623545347813365e-05,
+      "loss": 1.1821,
+      "step": 5075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9044093580500183,
+      "learning_rate": 1.262087827523463e-05,
+      "loss": 1.1539,
+      "step": 5076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8174122295628923,
+      "learning_rate": 1.2618211002460135e-05,
+      "loss": 1.177,
+      "step": 5077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.875413923084302,
+      "learning_rate": 1.2615543529693621e-05,
+      "loss": 1.158,
+      "step": 5078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.21451817707271,
+      "learning_rate": 1.2612875857138842e-05,
+      "loss": 1.1875,
+      "step": 5079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5452368292912158,
+      "learning_rate": 1.2610207984999567e-05,
+      "loss": 1.2139,
+      "step": 5080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.455471455664398,
+      "learning_rate": 1.2607539913479583e-05,
+      "loss": 1.1882,
+      "step": 5081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.167850657327508,
+      "learning_rate": 1.2604871642782694e-05,
+      "loss": 1.1448,
+      "step": 5082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.690132877934288,
+      "learning_rate": 1.260220317311271e-05,
+      "loss": 1.2148,
+      "step": 5083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3697591765279276,
+      "learning_rate": 1.2599534504673463e-05,
+      "loss": 1.1848,
+      "step": 5084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.115571412989097,
+      "learning_rate": 1.25968656376688e-05,
+      "loss": 1.1433,
+      "step": 5085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.148444980881298,
+      "learning_rate": 1.2594196572302586e-05,
+      "loss": 1.2397,
+      "step": 5086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.474298392998776,
+      "learning_rate": 1.2591527308778694e-05,
+      "loss": 1.1401,
+      "step": 5087
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4244192354925462,
+      "learning_rate": 1.2588857847301013e-05,
+      "loss": 1.1514,
+      "step": 5088
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3423973332173653,
+      "learning_rate": 1.2586188188073456e-05,
+      "loss": 1.144,
+      "step": 5089
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.105355413486014,
+      "learning_rate": 1.2583518331299937e-05,
+      "loss": 1.1157,
+      "step": 5090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7473828919009486,
+      "learning_rate": 1.2580848277184401e-05,
+      "loss": 1.1389,
+      "step": 5091
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2920635988540963,
+      "learning_rate": 1.2578178025930798e-05,
+      "loss": 1.1794,
+      "step": 5092
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.414325060230442,
+      "learning_rate": 1.2575507577743094e-05,
+      "loss": 1.1399,
+      "step": 5093
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2507915711807795,
+      "learning_rate": 1.2572836932825268e-05,
+      "loss": 1.2073,
+      "step": 5094
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.064738016944762,
+      "learning_rate": 1.257016609138132e-05,
+      "loss": 1.1934,
+      "step": 5095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.059080046693759,
+      "learning_rate": 1.2567495053615267e-05,
+      "loss": 1.145,
+      "step": 5096
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4127362249484343,
+      "learning_rate": 1.2564823819731127e-05,
+      "loss": 1.1267,
+      "step": 5097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.254095536538888,
+      "learning_rate": 1.2562152389932947e-05,
+      "loss": 1.1958,
+      "step": 5098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.566423066434875,
+      "learning_rate": 1.2559480764424788e-05,
+      "loss": 1.2002,
+      "step": 5099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1398466682444246,
+      "learning_rate": 1.2556808943410712e-05,
+      "loss": 1.1995,
+      "step": 5100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4238224884401083,
+      "learning_rate": 1.2554136927094817e-05,
+      "loss": 1.167,
+      "step": 5101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.11836182865862,
+      "learning_rate": 1.2551464715681197e-05,
+      "loss": 1.1892,
+      "step": 5102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8368946367770693,
+      "learning_rate": 1.2548792309373973e-05,
+      "loss": 1.1726,
+      "step": 5103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.810564287289462,
+      "learning_rate": 1.2546119708377273e-05,
+      "loss": 1.0791,
+      "step": 5104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.717281800850692,
+      "learning_rate": 1.254344691289525e-05,
+      "loss": 1.1958,
+      "step": 5105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7922826656201445,
+      "learning_rate": 1.254077392313206e-05,
+      "loss": 1.1167,
+      "step": 5106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.804628865351659,
+      "learning_rate": 1.2538100739291882e-05,
+      "loss": 1.1147,
+      "step": 5107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.135725695319322,
+      "learning_rate": 1.2535427361578903e-05,
+      "loss": 1.1409,
+      "step": 5108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.249753473218079,
+      "learning_rate": 1.2532753790197334e-05,
+      "loss": 1.1199,
+      "step": 5109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.198871013178228,
+      "learning_rate": 1.2530080025351392e-05,
+      "loss": 1.145,
+      "step": 5110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.787130983973683,
+      "learning_rate": 1.2527406067245318e-05,
+      "loss": 1.1785,
+      "step": 5111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3831783449522725,
+      "learning_rate": 1.2524731916083355e-05,
+      "loss": 1.1509,
+      "step": 5112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.400768713935673,
+      "learning_rate": 1.2522057572069772e-05,
+      "loss": 1.1626,
+      "step": 5113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4498606776518854,
+      "learning_rate": 1.2519383035408848e-05,
+      "loss": 1.2004,
+      "step": 5114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.322274868227364,
+      "learning_rate": 1.251670830630488e-05,
+      "loss": 1.1113,
+      "step": 5115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5209245475453312,
+      "learning_rate": 1.2514033384962173e-05,
+      "loss": 1.2129,
+      "step": 5116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1849855180375375,
+      "learning_rate": 1.251135827158505e-05,
+      "loss": 1.1345,
+      "step": 5117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1791831869917235,
+      "learning_rate": 1.2508682966377852e-05,
+      "loss": 1.1252,
+      "step": 5118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.077256060987841,
+      "learning_rate": 1.2506007469544936e-05,
+      "loss": 1.1389,
+      "step": 5119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.119934556582927,
+      "learning_rate": 1.2503331781290665e-05,
+      "loss": 1.0938,
+      "step": 5120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0714834464075538,
+      "learning_rate": 1.250065590181942e-05,
+      "loss": 1.1558,
+      "step": 5121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6175511120035813,
+      "learning_rate": 1.2497979831335604e-05,
+      "loss": 1.1973,
+      "step": 5122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9525801740114432,
+      "learning_rate": 1.2495303570043623e-05,
+      "loss": 1.0535,
+      "step": 5123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7983089722649996,
+      "learning_rate": 1.2492627118147905e-05,
+      "loss": 1.1895,
+      "step": 5124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9865011301692537,
+      "learning_rate": 1.2489950475852892e-05,
+      "loss": 1.1162,
+      "step": 5125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.948572965601328,
+      "learning_rate": 1.2487273643363041e-05,
+      "loss": 1.1489,
+      "step": 5126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.848035126289602,
+      "learning_rate": 1.2484596620882814e-05,
+      "loss": 1.1367,
+      "step": 5127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0514954687653413,
+      "learning_rate": 1.24819194086167e-05,
+      "loss": 1.1606,
+      "step": 5128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2862385110779,
+      "learning_rate": 1.2479242006769203e-05,
+      "loss": 1.2366,
+      "step": 5129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9333605194206402,
+      "learning_rate": 1.247656441554483e-05,
+      "loss": 1.1895,
+      "step": 5130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.386014159980253,
+      "learning_rate": 1.2473886635148109e-05,
+      "loss": 1.033,
+      "step": 5131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.718627235383231,
+      "learning_rate": 1.2471208665783586e-05,
+      "loss": 1.1863,
+      "step": 5132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.072944772106552,
+      "learning_rate": 1.2468530507655814e-05,
+      "loss": 1.1799,
+      "step": 5133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.02993796137684,
+      "learning_rate": 1.2465852160969364e-05,
+      "loss": 1.1375,
+      "step": 5134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5888211665803427,
+      "learning_rate": 1.2463173625928824e-05,
+      "loss": 1.1619,
+      "step": 5135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9207819747568053,
+      "learning_rate": 1.2460494902738793e-05,
+      "loss": 1.1804,
+      "step": 5136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.698531430305798,
+      "learning_rate": 1.2457815991603886e-05,
+      "loss": 1.2012,
+      "step": 5137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.792311575001899,
+      "learning_rate": 1.2455136892728727e-05,
+      "loss": 1.0745,
+      "step": 5138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3641274925260065,
+      "learning_rate": 1.2452457606317966e-05,
+      "loss": 1.248,
+      "step": 5139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8552480170531167,
+      "learning_rate": 1.2449778132576256e-05,
+      "loss": 1.1631,
+      "step": 5140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6997185279255045,
+      "learning_rate": 1.2447098471708268e-05,
+      "loss": 1.1177,
+      "step": 5141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1167737440031895,
+      "learning_rate": 1.244441862391869e-05,
+      "loss": 1.1909,
+      "step": 5142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.245576023391258,
+      "learning_rate": 1.2441738589412222e-05,
+      "loss": 1.2117,
+      "step": 5143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.012214818163098,
+      "learning_rate": 1.2439058368393578e-05,
+      "loss": 1.2078,
+      "step": 5144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.07045120267892,
+      "learning_rate": 1.2436377961067486e-05,
+      "loss": 1.1533,
+      "step": 5145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.253191660247082,
+      "learning_rate": 1.2433697367638689e-05,
+      "loss": 1.1274,
+      "step": 5146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.102167840724755,
+      "learning_rate": 1.2431016588311943e-05,
+      "loss": 1.0984,
+      "step": 5147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6437303755830435,
+      "learning_rate": 1.2428335623292023e-05,
+      "loss": 1.1313,
+      "step": 5148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6903714530669065,
+      "learning_rate": 1.2425654472783712e-05,
+      "loss": 1.1746,
+      "step": 5149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.408116378404366,
+      "learning_rate": 1.2422973136991806e-05,
+      "loss": 1.1338,
+      "step": 5150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3957170488358845,
+      "learning_rate": 1.2420291616121128e-05,
+      "loss": 1.2097,
+      "step": 5151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.710290224312835,
+      "learning_rate": 1.2417609910376498e-05,
+      "loss": 1.2302,
+      "step": 5152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2123071568295716,
+      "learning_rate": 1.2414928019962763e-05,
+      "loss": 1.1436,
+      "step": 5153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.358696449720102,
+      "learning_rate": 1.2412245945084772e-05,
+      "loss": 1.0874,
+      "step": 5154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.122979384766229,
+      "learning_rate": 1.2409563685947404e-05,
+      "loss": 1.198,
+      "step": 5155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.368011335519806,
+      "learning_rate": 1.2406881242755537e-05,
+      "loss": 1.1089,
+      "step": 5156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2845935419694476,
+      "learning_rate": 1.2404198615714074e-05,
+      "loss": 1.1951,
+      "step": 5157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.775409956836691,
+      "learning_rate": 1.2401515805027924e-05,
+      "loss": 1.1772,
+      "step": 5158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.278327903250144,
+      "learning_rate": 1.2398832810902015e-05,
+      "loss": 1.1545,
+      "step": 5159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.938919130642556,
+      "learning_rate": 1.2396149633541289e-05,
+      "loss": 1.1304,
+      "step": 5160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.123111908547493,
+      "learning_rate": 1.2393466273150693e-05,
+      "loss": 1.1328,
+      "step": 5161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.612664442512582,
+      "learning_rate": 1.2390782729935207e-05,
+      "loss": 1.1462,
+      "step": 5162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.357378685314493,
+      "learning_rate": 1.2388099004099807e-05,
+      "loss": 1.1946,
+      "step": 5163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.806142563834848,
+      "learning_rate": 1.2385415095849487e-05,
+      "loss": 1.1987,
+      "step": 5164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.523941809139562,
+      "learning_rate": 1.2382731005389261e-05,
+      "loss": 1.1101,
+      "step": 5165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4811622139374543,
+      "learning_rate": 1.2380046732924154e-05,
+      "loss": 1.2068,
+      "step": 5166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7649993181151,
+      "learning_rate": 1.2377362278659202e-05,
+      "loss": 1.0706,
+      "step": 5167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8191668070158777,
+      "learning_rate": 1.2374677642799457e-05,
+      "loss": 1.1646,
+      "step": 5168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7697506412098587,
+      "learning_rate": 1.2371992825549985e-05,
+      "loss": 1.1377,
+      "step": 5169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.052388199967575,
+      "learning_rate": 1.2369307827115866e-05,
+      "loss": 1.2112,
+      "step": 5170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.41191339933079,
+      "learning_rate": 1.2366622647702196e-05,
+      "loss": 1.2166,
+      "step": 5171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.809405837778152,
+      "learning_rate": 1.236393728751408e-05,
+      "loss": 1.1814,
+      "step": 5172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1081463278300556,
+      "learning_rate": 1.2361251746756637e-05,
+      "loss": 1.1658,
+      "step": 5173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9319098493408955,
+      "learning_rate": 1.2358566025635006e-05,
+      "loss": 1.1707,
+      "step": 5174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3107371255902214,
+      "learning_rate": 1.2355880124354332e-05,
+      "loss": 1.2085,
+      "step": 5175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.159662403655335,
+      "learning_rate": 1.2353194043119783e-05,
+      "loss": 1.1431,
+      "step": 5176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.445584093413747,
+      "learning_rate": 1.235050778213653e-05,
+      "loss": 1.2444,
+      "step": 5177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3208001681836135,
+      "learning_rate": 1.2347821341609765e-05,
+      "loss": 1.2324,
+      "step": 5178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9067026583002455,
+      "learning_rate": 1.2345134721744693e-05,
+      "loss": 1.1775,
+      "step": 5179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7118244842156964,
+      "learning_rate": 1.2342447922746528e-05,
+      "loss": 1.1169,
+      "step": 5180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.355178018005107,
+      "learning_rate": 1.2339760944820504e-05,
+      "loss": 1.1592,
+      "step": 5181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4718577934805745,
+      "learning_rate": 1.2337073788171869e-05,
+      "loss": 1.1262,
+      "step": 5182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.739728418857073,
+      "learning_rate": 1.2334386453005873e-05,
+      "loss": 1.0715,
+      "step": 5183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.774354296797165,
+      "learning_rate": 1.2331698939527792e-05,
+      "loss": 1.0333,
+      "step": 5184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.570091956328736,
+      "learning_rate": 1.2329011247942915e-05,
+      "loss": 1.1707,
+      "step": 5185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.516282036896329,
+      "learning_rate": 1.2326323378456539e-05,
+      "loss": 1.2129,
+      "step": 5186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6097558241633596,
+      "learning_rate": 1.2323635331273973e-05,
+      "loss": 1.1038,
+      "step": 5187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.221269405080428,
+      "learning_rate": 1.232094710660055e-05,
+      "loss": 1.1235,
+      "step": 5188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7217320566165797,
+      "learning_rate": 1.2318258704641605e-05,
+      "loss": 1.186,
+      "step": 5189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.187786673032531,
+      "learning_rate": 1.2315570125602495e-05,
+      "loss": 1.1377,
+      "step": 5190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9214612859486073,
+      "learning_rate": 1.2312881369688584e-05,
+      "loss": 1.1309,
+      "step": 5191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.016152898229673,
+      "learning_rate": 1.2310192437105255e-05,
+      "loss": 1.1284,
+      "step": 5192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9284806839143385,
+      "learning_rate": 1.2307503328057899e-05,
+      "loss": 1.1206,
+      "step": 5193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1645712890808078,
+      "learning_rate": 1.2304814042751927e-05,
+      "loss": 1.1714,
+      "step": 5194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.011793650459086,
+      "learning_rate": 1.230212458139276e-05,
+      "loss": 1.1375,
+      "step": 5195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1891002614595276,
+      "learning_rate": 1.2299434944185833e-05,
+      "loss": 1.1794,
+      "step": 5196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.170277774663737,
+      "learning_rate": 1.2296745131336589e-05,
+      "loss": 1.157,
+      "step": 5197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9845314435810013,
+      "learning_rate": 1.2294055143050492e-05,
+      "loss": 1.1443,
+      "step": 5198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.358796580899544,
+      "learning_rate": 1.229136497953302e-05,
+      "loss": 1.1409,
+      "step": 5199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.33619479041748,
+      "learning_rate": 1.2288674640989657e-05,
+      "loss": 1.2007,
+      "step": 5200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.950732090128692,
+      "learning_rate": 1.2285984127625907e-05,
+      "loss": 1.1377,
+      "step": 5201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.388517183846088,
+      "learning_rate": 1.2283293439647283e-05,
+      "loss": 1.146,
+      "step": 5202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7808358042805277,
+      "learning_rate": 1.2280602577259312e-05,
+      "loss": 1.1326,
+      "step": 5203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.424155021932179,
+      "learning_rate": 1.227791154066754e-05,
+      "loss": 1.0891,
+      "step": 5204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.061560644337407,
+      "learning_rate": 1.227522033007752e-05,
+      "loss": 1.1155,
+      "step": 5205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0447010963619756,
+      "learning_rate": 1.2272528945694814e-05,
+      "loss": 1.1082,
+      "step": 5206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9544404037286545,
+      "learning_rate": 1.2269837387725012e-05,
+      "loss": 1.1165,
+      "step": 5207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9328572465262797,
+      "learning_rate": 1.2267145656373703e-05,
+      "loss": 1.1423,
+      "step": 5208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7598413629458496,
+      "learning_rate": 1.2264453751846504e-05,
+      "loss": 1.1377,
+      "step": 5209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1281215485021203,
+      "learning_rate": 1.2261761674349021e-05,
+      "loss": 1.1438,
+      "step": 5210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1483939186456786,
+      "learning_rate": 1.22590694240869e-05,
+      "loss": 1.1711,
+      "step": 5211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.413752823048975,
+      "learning_rate": 1.2256377001265785e-05,
+      "loss": 1.1084,
+      "step": 5212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0840518956406004,
+      "learning_rate": 1.2253684406091334e-05,
+      "loss": 1.1648,
+      "step": 5213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4206470299908416,
+      "learning_rate": 1.225099163876923e-05,
+      "loss": 1.1519,
+      "step": 5214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7846830073926374,
+      "learning_rate": 1.224829869950515e-05,
+      "loss": 1.1353,
+      "step": 5215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.335706854854864,
+      "learning_rate": 1.2245605588504796e-05,
+      "loss": 1.2388,
+      "step": 5216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8722287936555335,
+      "learning_rate": 1.2242912305973881e-05,
+      "loss": 1.1233,
+      "step": 5217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.506306223564103,
+      "learning_rate": 1.224021885211814e-05,
+      "loss": 1.1077,
+      "step": 5218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1421009188014475,
+      "learning_rate": 1.2237525227143304e-05,
+      "loss": 1.1902,
+      "step": 5219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.020707048011303,
+      "learning_rate": 1.2234831431255122e-05,
+      "loss": 1.1292,
+      "step": 5220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.764721814202205,
+      "learning_rate": 1.2232137464659368e-05,
+      "loss": 1.083,
+      "step": 5221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3632910539459817,
+      "learning_rate": 1.2229443327561818e-05,
+      "loss": 1.0936,
+      "step": 5222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.026200329900513,
+      "learning_rate": 1.2226749020168264e-05,
+      "loss": 1.2073,
+      "step": 5223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2563691890695026,
+      "learning_rate": 1.2224054542684507e-05,
+      "loss": 1.0776,
+      "step": 5224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1087720020602814,
+      "learning_rate": 1.2221359895316372e-05,
+      "loss": 1.1741,
+      "step": 5225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1133266693054455,
+      "learning_rate": 1.2218665078269681e-05,
+      "loss": 1.0356,
+      "step": 5226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.507773809138233,
+      "learning_rate": 1.221597009175028e-05,
+      "loss": 1.1582,
+      "step": 5227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9764748302887813,
+      "learning_rate": 1.221327493596403e-05,
+      "loss": 1.1409,
+      "step": 5228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.879280884616344,
+      "learning_rate": 1.22105796111168e-05,
+      "loss": 1.1235,
+      "step": 5229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.116576890100112,
+      "learning_rate": 1.2207884117414465e-05,
+      "loss": 1.157,
+      "step": 5230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4472695927421495,
+      "learning_rate": 1.2205188455062927e-05,
+      "loss": 1.2065,
+      "step": 5231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.802433023966294,
+      "learning_rate": 1.2202492624268092e-05,
+      "loss": 1.1174,
+      "step": 5232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.16206694401317,
+      "learning_rate": 1.2199796625235883e-05,
+      "loss": 1.2246,
+      "step": 5233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0812672851534195,
+      "learning_rate": 1.219710045817223e-05,
+      "loss": 1.1448,
+      "step": 5234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9649063378288028,
+      "learning_rate": 1.2194404123283083e-05,
+      "loss": 1.1287,
+      "step": 5235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1529256949440523,
+      "learning_rate": 1.2191707620774397e-05,
+      "loss": 1.1938,
+      "step": 5236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.917583283883183,
+      "learning_rate": 1.2189010950852153e-05,
+      "loss": 1.1465,
+      "step": 5237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6619317041113812,
+      "learning_rate": 1.2186314113722326e-05,
+      "loss": 1.0784,
+      "step": 5238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6120195294288338,
+      "learning_rate": 1.2183617109590925e-05,
+      "loss": 1.1509,
+      "step": 5239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.960429639655591,
+      "learning_rate": 1.2180919938663946e-05,
+      "loss": 1.1565,
+      "step": 5240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9507508352295075,
+      "learning_rate": 1.2178222601147428e-05,
+      "loss": 1.1277,
+      "step": 5241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.006909074390112,
+      "learning_rate": 1.2175525097247396e-05,
+      "loss": 1.2034,
+      "step": 5242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.126942352661172,
+      "learning_rate": 1.2172827427169901e-05,
+      "loss": 1.2512,
+      "step": 5243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3784107487840243,
+      "learning_rate": 1.2170129591121007e-05,
+      "loss": 1.1418,
+      "step": 5244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7203701955305575,
+      "learning_rate": 1.2167431589306791e-05,
+      "loss": 1.2712,
+      "step": 5245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2228080245604374,
+      "learning_rate": 1.2164733421933331e-05,
+      "loss": 1.075,
+      "step": 5246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.232521420643039,
+      "learning_rate": 1.2162035089206734e-05,
+      "loss": 1.1235,
+      "step": 5247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1319925797188324,
+      "learning_rate": 1.215933659133311e-05,
+      "loss": 1.1692,
+      "step": 5248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.997779595392708,
+      "learning_rate": 1.2156637928518582e-05,
+      "loss": 1.0913,
+      "step": 5249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.015419156402172,
+      "learning_rate": 1.215393910096929e-05,
+      "loss": 1.1394,
+      "step": 5250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5874234460848933,
+      "learning_rate": 1.2151240108891382e-05,
+      "loss": 1.2207,
+      "step": 5251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.99668557126633,
+      "learning_rate": 1.2148540952491024e-05,
+      "loss": 1.2026,
+      "step": 5252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7207457177264316,
+      "learning_rate": 1.2145841631974382e-05,
+      "loss": 1.1086,
+      "step": 5253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4582262734383087,
+      "learning_rate": 1.2143142147547653e-05,
+      "loss": 1.1411,
+      "step": 5254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.73867438654745,
+      "learning_rate": 1.2140442499417036e-05,
+      "loss": 1.1289,
+      "step": 5255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0619587108949315,
+      "learning_rate": 1.2137742687788738e-05,
+      "loss": 1.1274,
+      "step": 5256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8779900701134293,
+      "learning_rate": 1.213504271286899e-05,
+      "loss": 1.1172,
+      "step": 5257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.062792888500424,
+      "learning_rate": 1.2132342574864026e-05,
+      "loss": 1.1338,
+      "step": 5258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2597303754296028,
+      "learning_rate": 1.2129642273980096e-05,
+      "loss": 1.1506,
+      "step": 5259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2599259249115593,
+      "learning_rate": 1.2126941810423465e-05,
+      "loss": 1.1489,
+      "step": 5260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.366253870968863,
+      "learning_rate": 1.2124241184400405e-05,
+      "loss": 1.1667,
+      "step": 5261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1983102267067727,
+      "learning_rate": 1.2121540396117206e-05,
+      "loss": 1.1431,
+      "step": 5262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0975724044470763,
+      "learning_rate": 1.2118839445780164e-05,
+      "loss": 1.1877,
+      "step": 5263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.035049572227339,
+      "learning_rate": 1.2116138333595595e-05,
+      "loss": 1.1484,
+      "step": 5264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5467345623091537,
+      "learning_rate": 1.2113437059769826e-05,
+      "loss": 1.127,
+      "step": 5265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9370123208807173,
+      "learning_rate": 1.2110735624509184e-05,
+      "loss": 1.1646,
+      "step": 5266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9320619145092675,
+      "learning_rate": 1.2108034028020024e-05,
+      "loss": 1.1731,
+      "step": 5267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6209606939136996,
+      "learning_rate": 1.210533227050871e-05,
+      "loss": 1.1667,
+      "step": 5268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.127126118135561,
+      "learning_rate": 1.210263035218161e-05,
+      "loss": 1.0769,
+      "step": 5269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.335443933818094,
+      "learning_rate": 1.2099928273245115e-05,
+      "loss": 1.1887,
+      "step": 5270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.804767866995484,
+      "learning_rate": 1.2097226033905619e-05,
+      "loss": 1.1707,
+      "step": 5271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.734160994114763,
+      "learning_rate": 1.2094523634369538e-05,
+      "loss": 1.1143,
+      "step": 5272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.245983919303725,
+      "learning_rate": 1.2091821074843288e-05,
+      "loss": 1.0881,
+      "step": 5273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9856906338930576,
+      "learning_rate": 1.208911835553331e-05,
+      "loss": 1.0522,
+      "step": 5274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.388489162667722,
+      "learning_rate": 1.2086415476646047e-05,
+      "loss": 1.1462,
+      "step": 5275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.65299464076596,
+      "learning_rate": 1.208371243838796e-05,
+      "loss": 1.1375,
+      "step": 5276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.294346487941302,
+      "learning_rate": 1.208100924096552e-05,
+      "loss": 1.1409,
+      "step": 5277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.207619251730938,
+      "learning_rate": 1.2078305884585212e-05,
+      "loss": 1.1631,
+      "step": 5278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4951545994982074,
+      "learning_rate": 1.2075602369453532e-05,
+      "loss": 1.1338,
+      "step": 5279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9912677569039894,
+      "learning_rate": 1.2072898695776987e-05,
+      "loss": 1.1541,
+      "step": 5280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.278607519806441,
+      "learning_rate": 1.2070194863762094e-05,
+      "loss": 1.16,
+      "step": 5281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2734182866210904,
+      "learning_rate": 1.2067490873615393e-05,
+      "loss": 1.0771,
+      "step": 5282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.449020441921677,
+      "learning_rate": 1.2064786725543418e-05,
+      "loss": 1.1785,
+      "step": 5283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3003671434410915,
+      "learning_rate": 1.2062082419752737e-05,
+      "loss": 1.1377,
+      "step": 5284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7043575988620767,
+      "learning_rate": 1.205937795644991e-05,
+      "loss": 1.0513,
+      "step": 5285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.890885971501123,
+      "learning_rate": 1.2056673335841515e-05,
+      "loss": 1.1926,
+      "step": 5286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.15557147018779,
+      "learning_rate": 1.2053968558134152e-05,
+      "loss": 1.1514,
+      "step": 5287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.462730327490272,
+      "learning_rate": 1.2051263623534425e-05,
+      "loss": 1.1987,
+      "step": 5288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3449308092185275,
+      "learning_rate": 1.2048558532248948e-05,
+      "loss": 1.1489,
+      "step": 5289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.011129450723474,
+      "learning_rate": 1.2045853284484346e-05,
+      "loss": 1.123,
+      "step": 5290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.068161436522651,
+      "learning_rate": 1.2043147880447264e-05,
+      "loss": 1.1394,
+      "step": 5291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.157038129087228,
+      "learning_rate": 1.2040442320344355e-05,
+      "loss": 1.1182,
+      "step": 5292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2660302413806814,
+      "learning_rate": 1.2037736604382279e-05,
+      "loss": 1.1318,
+      "step": 5293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.056663668106899,
+      "learning_rate": 1.2035030732767713e-05,
+      "loss": 1.1345,
+      "step": 5294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3736322054677195,
+      "learning_rate": 1.2032324705707347e-05,
+      "loss": 1.1169,
+      "step": 5295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.963395519606012,
+      "learning_rate": 1.202961852340788e-05,
+      "loss": 1.1841,
+      "step": 5296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1810323257817346,
+      "learning_rate": 1.2026912186076023e-05,
+      "loss": 1.1501,
+      "step": 5297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0282935544516816,
+      "learning_rate": 1.2024205693918502e-05,
+      "loss": 1.0815,
+      "step": 5298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.082768362224041,
+      "learning_rate": 1.202149904714205e-05,
+      "loss": 1.1506,
+      "step": 5299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.015168064278823,
+      "learning_rate": 1.2018792245953412e-05,
+      "loss": 1.1533,
+      "step": 5300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.34222773659071,
+      "learning_rate": 1.2016085290559352e-05,
+      "loss": 1.1289,
+      "step": 5301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9554559922403043,
+      "learning_rate": 1.201337818116664e-05,
+      "loss": 1.1252,
+      "step": 5302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.389911477051906,
+      "learning_rate": 1.2010670917982055e-05,
+      "loss": 1.0464,
+      "step": 5303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.24661322166969,
+      "learning_rate": 1.2007963501212392e-05,
+      "loss": 1.155,
+      "step": 5304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3727042506526503,
+      "learning_rate": 1.200525593106446e-05,
+      "loss": 1.1843,
+      "step": 5305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.115937602537365,
+      "learning_rate": 1.2002548207745075e-05,
+      "loss": 1.1821,
+      "step": 5306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3868400690597276,
+      "learning_rate": 1.1999840331461068e-05,
+      "loss": 1.0464,
+      "step": 5307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.305189395560348,
+      "learning_rate": 1.1997132302419277e-05,
+      "loss": 1.158,
+      "step": 5308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8016513990359027,
+      "learning_rate": 1.1994424120826555e-05,
+      "loss": 1.1968,
+      "step": 5309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9124128660367474,
+      "learning_rate": 1.1991715786889768e-05,
+      "loss": 1.1758,
+      "step": 5310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8904236280368876,
+      "learning_rate": 1.1989007300815793e-05,
+      "loss": 1.1443,
+      "step": 5311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.004711573103746,
+      "learning_rate": 1.1986298662811518e-05,
+      "loss": 1.1938,
+      "step": 5312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3116085422067987,
+      "learning_rate": 1.1983589873083837e-05,
+      "loss": 1.1462,
+      "step": 5313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.075233575226126,
+      "learning_rate": 1.1980880931839664e-05,
+      "loss": 1.1221,
+      "step": 5314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3195291658203265,
+      "learning_rate": 1.1978171839285925e-05,
+      "loss": 1.0825,
+      "step": 5315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.935938260991656,
+      "learning_rate": 1.1975462595629552e-05,
+      "loss": 1.1418,
+      "step": 5316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1574620792663737,
+      "learning_rate": 1.1972753201077486e-05,
+      "loss": 1.1331,
+      "step": 5317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5258250609458353,
+      "learning_rate": 1.1970043655836691e-05,
+      "loss": 1.1182,
+      "step": 5318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7042594732567378,
+      "learning_rate": 1.1967333960114131e-05,
+      "loss": 1.1594,
+      "step": 5319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1987858608657325,
+      "learning_rate": 1.1964624114116784e-05,
+      "loss": 1.0867,
+      "step": 5320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9760381921742174,
+      "learning_rate": 1.1961914118051649e-05,
+      "loss": 1.1799,
+      "step": 5321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1188554952955156,
+      "learning_rate": 1.1959203972125726e-05,
+      "loss": 1.1812,
+      "step": 5322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.002808994100759,
+      "learning_rate": 1.1956493676546026e-05,
+      "loss": 1.1057,
+      "step": 5323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.214543904302566,
+      "learning_rate": 1.1953783231519576e-05,
+      "loss": 1.1833,
+      "step": 5324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6820323469405314,
+      "learning_rate": 1.1951072637253418e-05,
+      "loss": 1.0847,
+      "step": 5325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.416314601764783,
+      "learning_rate": 1.1948361893954597e-05,
+      "loss": 1.0901,
+      "step": 5326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2959420366582433,
+      "learning_rate": 1.194565100183017e-05,
+      "loss": 1.1438,
+      "step": 5327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.208725738463622,
+      "learning_rate": 1.1942939961087215e-05,
+      "loss": 1.1384,
+      "step": 5328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.067769066964204,
+      "learning_rate": 1.1940228771932808e-05,
+      "loss": 1.1479,
+      "step": 5329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.780200170597381,
+      "learning_rate": 1.193751743457405e-05,
+      "loss": 1.1362,
+      "step": 5330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.463792906944405,
+      "learning_rate": 1.1934805949218045e-05,
+      "loss": 1.0405,
+      "step": 5331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1669517498187227,
+      "learning_rate": 1.1932094316071905e-05,
+      "loss": 1.1174,
+      "step": 5332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3063359213113577,
+      "learning_rate": 1.192938253534276e-05,
+      "loss": 1.1294,
+      "step": 5333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.621058710851037,
+      "learning_rate": 1.192667060723775e-05,
+      "loss": 1.1143,
+      "step": 5334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7628371309171187,
+      "learning_rate": 1.1923958531964033e-05,
+      "loss": 1.1155,
+      "step": 5335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.461901064123331,
+      "learning_rate": 1.1921246309728757e-05,
+      "loss": 1.1558,
+      "step": 5336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1089188435684507,
+      "learning_rate": 1.1918533940739106e-05,
+      "loss": 1.168,
+      "step": 5337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1069947089395793,
+      "learning_rate": 1.1915821425202258e-05,
+      "loss": 1.2114,
+      "step": 5338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.668955278168632,
+      "learning_rate": 1.1913108763325412e-05,
+      "loss": 1.1675,
+      "step": 5339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.607437941175166,
+      "learning_rate": 1.1910395955315772e-05,
+      "loss": 1.1099,
+      "step": 5340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0409602046552027,
+      "learning_rate": 1.1907683001380562e-05,
+      "loss": 1.1006,
+      "step": 5341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0932777538350615,
+      "learning_rate": 1.1904969901726999e-05,
+      "loss": 1.2361,
+      "step": 5342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.017841603680298,
+      "learning_rate": 1.1902256656562331e-05,
+      "loss": 1.1316,
+      "step": 5343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.775941251022773,
+      "learning_rate": 1.1899543266093814e-05,
+      "loss": 1.1675,
+      "step": 5344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5218775415210386,
+      "learning_rate": 1.1896829730528702e-05,
+      "loss": 1.1602,
+      "step": 5345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.284711822731918,
+      "learning_rate": 1.1894116050074271e-05,
+      "loss": 1.1602,
+      "step": 5346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.427653445473075,
+      "learning_rate": 1.1891402224937805e-05,
+      "loss": 1.1736,
+      "step": 5347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.581478352019055,
+      "learning_rate": 1.1888688255326602e-05,
+      "loss": 1.1809,
+      "step": 5348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.658570041448324,
+      "learning_rate": 1.1885974141447965e-05,
+      "loss": 1.1233,
+      "step": 5349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.317962171905869,
+      "learning_rate": 1.1883259883509213e-05,
+      "loss": 1.1509,
+      "step": 5350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.977717113277513,
+      "learning_rate": 1.1880545481717675e-05,
+      "loss": 1.144,
+      "step": 5351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.883059597587248,
+      "learning_rate": 1.187783093628069e-05,
+      "loss": 1.1421,
+      "step": 5352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0813037531058924,
+      "learning_rate": 1.187511624740561e-05,
+      "loss": 1.1519,
+      "step": 5353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3708941156324372,
+      "learning_rate": 1.1872401415299794e-05,
+      "loss": 1.1301,
+      "step": 5354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.813439617391082,
+      "learning_rate": 1.1869686440170618e-05,
+      "loss": 1.1035,
+      "step": 5355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.024284062533949,
+      "learning_rate": 1.1866971322225462e-05,
+      "loss": 1.1333,
+      "step": 5356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.442901085698631,
+      "learning_rate": 1.1864256061671719e-05,
+      "loss": 1.1257,
+      "step": 5357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3944004873545346,
+      "learning_rate": 1.1861540658716802e-05,
+      "loss": 1.1853,
+      "step": 5358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.959541256123227,
+      "learning_rate": 1.1858825113568118e-05,
+      "loss": 1.0735,
+      "step": 5359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3033675770994604,
+      "learning_rate": 1.1856109426433096e-05,
+      "loss": 1.1973,
+      "step": 5360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9304630086004058,
+      "learning_rate": 1.1853393597519179e-05,
+      "loss": 1.1147,
+      "step": 5361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7729831092110167,
+      "learning_rate": 1.1850677627033807e-05,
+      "loss": 1.1724,
+      "step": 5362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.974149879669484,
+      "learning_rate": 1.1847961515184447e-05,
+      "loss": 1.0532,
+      "step": 5363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1672474176478396,
+      "learning_rate": 1.1845245262178567e-05,
+      "loss": 1.1311,
+      "step": 5364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.892324494278108,
+      "learning_rate": 1.1842528868223645e-05,
+      "loss": 1.136,
+      "step": 5365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6691777830846624,
+      "learning_rate": 1.1839812333527176e-05,
+      "loss": 1.0664,
+      "step": 5366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4260983632966915,
+      "learning_rate": 1.183709565829666e-05,
+      "loss": 1.166,
+      "step": 5367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.606215926785786,
+      "learning_rate": 1.1834378842739616e-05,
+      "loss": 1.1597,
+      "step": 5368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.454699707816616,
+      "learning_rate": 1.1831661887063559e-05,
+      "loss": 1.0798,
+      "step": 5369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.32610630611089,
+      "learning_rate": 1.1828944791476029e-05,
+      "loss": 1.1829,
+      "step": 5370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6085393343867596,
+      "learning_rate": 1.1826227556184574e-05,
+      "loss": 1.167,
+      "step": 5371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8638933339815726,
+      "learning_rate": 1.1823510181396743e-05,
+      "loss": 1.1831,
+      "step": 5372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0459284216036977,
+      "learning_rate": 1.1820792667320107e-05,
+      "loss": 1.1365,
+      "step": 5373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6000936943538404,
+      "learning_rate": 1.1818075014162243e-05,
+      "loss": 1.2241,
+      "step": 5374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3891194056292506,
+      "learning_rate": 1.1815357222130737e-05,
+      "loss": 1.1292,
+      "step": 5375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1698735858355787,
+      "learning_rate": 1.1812639291433188e-05,
+      "loss": 1.1501,
+      "step": 5376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.09982195445519,
+      "learning_rate": 1.180992122227721e-05,
+      "loss": 1.1326,
+      "step": 5377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.34366109528735,
+      "learning_rate": 1.1807203014870416e-05,
+      "loss": 1.1736,
+      "step": 5378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.593241390746682,
+      "learning_rate": 1.180448466942044e-05,
+      "loss": 1.134,
+      "step": 5379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.109229777005313,
+      "learning_rate": 1.1801766186134917e-05,
+      "loss": 1.1555,
+      "step": 5380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6872859973764895,
+      "learning_rate": 1.1799047565221508e-05,
+      "loss": 1.1565,
+      "step": 5381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.091959576714599,
+      "learning_rate": 1.179632880688787e-05,
+      "loss": 1.1199,
+      "step": 5382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.018019073523692,
+      "learning_rate": 1.1793609911341672e-05,
+      "loss": 1.2551,
+      "step": 5383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3451140954711684,
+      "learning_rate": 1.1790890878790605e-05,
+      "loss": 1.1172,
+      "step": 5384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1641387040989946,
+      "learning_rate": 1.1788171709442351e-05,
+      "loss": 1.1128,
+      "step": 5385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4477316802048175,
+      "learning_rate": 1.1785452403504625e-05,
+      "loss": 1.1943,
+      "step": 5386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.445078708256561,
+      "learning_rate": 1.1782732961185132e-05,
+      "loss": 1.1174,
+      "step": 5387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2922641629918967,
+      "learning_rate": 1.1780013382691604e-05,
+      "loss": 1.1411,
+      "step": 5388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.401777125152252,
+      "learning_rate": 1.177729366823177e-05,
+      "loss": 1.166,
+      "step": 5389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.234476028079092,
+      "learning_rate": 1.1774573818013377e-05,
+      "loss": 1.1104,
+      "step": 5390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3706628807535384,
+      "learning_rate": 1.1771853832244187e-05,
+      "loss": 1.2085,
+      "step": 5391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.454907884818961,
+      "learning_rate": 1.176913371113196e-05,
+      "loss": 1.1172,
+      "step": 5392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.118917409300363,
+      "learning_rate": 1.1766413454884469e-05,
+      "loss": 1.2261,
+      "step": 5393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.472941501756241,
+      "learning_rate": 1.1763693063709507e-05,
+      "loss": 1.1909,
+      "step": 5394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8253755069439217,
+      "learning_rate": 1.176097253781487e-05,
+      "loss": 1.156,
+      "step": 5395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.960327299412169,
+      "learning_rate": 1.1758251877408363e-05,
+      "loss": 1.1436,
+      "step": 5396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.914835824063582,
+      "learning_rate": 1.1755531082697805e-05,
+      "loss": 1.1133,
+      "step": 5397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5748091671247666,
+      "learning_rate": 1.1752810153891026e-05,
+      "loss": 1.1375,
+      "step": 5398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4613874622047627,
+      "learning_rate": 1.1750089091195858e-05,
+      "loss": 1.147,
+      "step": 5399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8761235662414113,
+      "learning_rate": 1.1747367894820156e-05,
+      "loss": 1.1533,
+      "step": 5400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.659757208992614,
+      "learning_rate": 1.1744646564971777e-05,
+      "loss": 1.1677,
+      "step": 5401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7749954748008143,
+      "learning_rate": 1.1741925101858587e-05,
+      "loss": 1.1758,
+      "step": 5402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.383617206499086,
+      "learning_rate": 1.1739203505688466e-05,
+      "loss": 1.155,
+      "step": 5403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.960668823127624,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 1.0598,
+      "step": 5404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4289356862623572,
+      "learning_rate": 1.1733759915009002e-05,
+      "loss": 1.1484,
+      "step": 5405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6915747056211736,
+      "learning_rate": 1.1731037920915465e-05,
+      "loss": 1.1416,
+      "step": 5406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.334668342131614,
+      "learning_rate": 1.1728315794596614e-05,
+      "loss": 1.1721,
+      "step": 5407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7248902461255007,
+      "learning_rate": 1.1725593536260382e-05,
+      "loss": 1.126,
+      "step": 5408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.330533511873623,
+      "learning_rate": 1.1722871146114705e-05,
+      "loss": 1.1699,
+      "step": 5409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.506197677430578,
+      "learning_rate": 1.1720148624367533e-05,
+      "loss": 1.1709,
+      "step": 5410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.124269061883,
+      "learning_rate": 1.171742597122683e-05,
+      "loss": 1.1418,
+      "step": 5411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.139589447369072,
+      "learning_rate": 1.1714703186900557e-05,
+      "loss": 1.1589,
+      "step": 5412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5866404047187586,
+      "learning_rate": 1.17119802715967e-05,
+      "loss": 1.2261,
+      "step": 5413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4840619015319527,
+      "learning_rate": 1.1709257225523255e-05,
+      "loss": 1.251,
+      "step": 5414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2642710584837777,
+      "learning_rate": 1.170653404888821e-05,
+      "loss": 1.1077,
+      "step": 5415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.870063103344379,
+      "learning_rate": 1.1703810741899577e-05,
+      "loss": 1.1636,
+      "step": 5416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.385769280176057,
+      "learning_rate": 1.1701087304765382e-05,
+      "loss": 1.155,
+      "step": 5417
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1801398629752344,
+      "learning_rate": 1.169836373769365e-05,
+      "loss": 1.1428,
+      "step": 5418
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9029307311773995,
+      "learning_rate": 1.1695640040892424e-05,
+      "loss": 1.1658,
+      "step": 5419
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.901508762717505,
+      "learning_rate": 1.1692916214569749e-05,
+      "loss": 1.1045,
+      "step": 5420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4110867016392588,
+      "learning_rate": 1.169019225893369e-05,
+      "loss": 1.1619,
+      "step": 5421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.234558039543584,
+      "learning_rate": 1.1687468174192309e-05,
+      "loss": 1.1694,
+      "step": 5422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.00331218929614,
+      "learning_rate": 1.1684743960553691e-05,
+      "loss": 1.1572,
+      "step": 5423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6432204443884757,
+      "learning_rate": 1.1682019618225926e-05,
+      "loss": 1.0473,
+      "step": 5424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.623620275255163,
+      "learning_rate": 1.1679295147417109e-05,
+      "loss": 1.1277,
+      "step": 5425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.121288118818343,
+      "learning_rate": 1.167657054833535e-05,
+      "loss": 1.1235,
+      "step": 5426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6730966218723453,
+      "learning_rate": 1.1673845821188762e-05,
+      "loss": 1.1655,
+      "step": 5427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.840680240044714,
+      "learning_rate": 1.1671120966185486e-05,
+      "loss": 1.1172,
+      "step": 5428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.92776813886831,
+      "learning_rate": 1.1668395983533651e-05,
+      "loss": 1.0801,
+      "step": 5429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.5250341460631605,
+      "learning_rate": 1.1665670873441405e-05,
+      "loss": 1.1965,
+      "step": 5430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.398693508406704,
+      "learning_rate": 1.166294563611691e-05,
+      "loss": 1.1504,
+      "step": 5431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.060444389734932,
+      "learning_rate": 1.1660220271768326e-05,
+      "loss": 1.144,
+      "step": 5432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.707765430309061,
+      "learning_rate": 1.1657494780603838e-05,
+      "loss": 1.1509,
+      "step": 5433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.181189620120912,
+      "learning_rate": 1.1654769162831627e-05,
+      "loss": 1.2488,
+      "step": 5434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3585066086036393,
+      "learning_rate": 1.165204341865989e-05,
+      "loss": 1.2166,
+      "step": 5435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9930066442445464,
+      "learning_rate": 1.1649317548296834e-05,
+      "loss": 1.1848,
+      "step": 5436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8175065816350306,
+      "learning_rate": 1.1646591551950677e-05,
+      "loss": 1.012,
+      "step": 5437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.144346122607787,
+      "learning_rate": 1.1643865429829639e-05,
+      "loss": 1.1497,
+      "step": 5438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4658662821724513,
+      "learning_rate": 1.1641139182141958e-05,
+      "loss": 1.1523,
+      "step": 5439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.335309866584107,
+      "learning_rate": 1.1638412809095878e-05,
+      "loss": 1.115,
+      "step": 5440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.000235230064621,
+      "learning_rate": 1.1635686310899656e-05,
+      "loss": 1.1929,
+      "step": 5441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.307260146438276,
+      "learning_rate": 1.1632959687761549e-05,
+      "loss": 1.1365,
+      "step": 5442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4991675207356243,
+      "learning_rate": 1.1630232939889835e-05,
+      "loss": 1.1492,
+      "step": 5443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.287838391274048,
+      "learning_rate": 1.1627506067492798e-05,
+      "loss": 1.1826,
+      "step": 5444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1432714407256688,
+      "learning_rate": 1.1624779070778722e-05,
+      "loss": 1.1121,
+      "step": 5445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3022272834117996,
+      "learning_rate": 1.1622051949955915e-05,
+      "loss": 1.2524,
+      "step": 5446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.890182573252443,
+      "learning_rate": 1.1619324705232691e-05,
+      "loss": 1.1208,
+      "step": 5447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7067027225271696,
+      "learning_rate": 1.1616597336817364e-05,
+      "loss": 1.1887,
+      "step": 5448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.44197296922318,
+      "learning_rate": 1.1613869844918266e-05,
+      "loss": 1.1455,
+      "step": 5449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.523046276334388,
+      "learning_rate": 1.1611142229743734e-05,
+      "loss": 1.114,
+      "step": 5450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6773001545042456,
+      "learning_rate": 1.1608414491502127e-05,
+      "loss": 1.1477,
+      "step": 5451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8858841063987426,
+      "learning_rate": 1.1605686630401794e-05,
+      "loss": 1.1394,
+      "step": 5452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.070510795704371,
+      "learning_rate": 1.16029586466511e-05,
+      "loss": 1.1943,
+      "step": 5453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2887998253622417,
+      "learning_rate": 1.1600230540458432e-05,
+      "loss": 1.2112,
+      "step": 5454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.900249478447524,
+      "learning_rate": 1.159750231203217e-05,
+      "loss": 1.1113,
+      "step": 5455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.103597023087166,
+      "learning_rate": 1.159477396158071e-05,
+      "loss": 1.1262,
+      "step": 5456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.472700774239807,
+      "learning_rate": 1.1592045489312458e-05,
+      "loss": 1.1421,
+      "step": 5457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.483595300514435,
+      "learning_rate": 1.1589316895435827e-05,
+      "loss": 1.1313,
+      "step": 5458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8186126276822963,
+      "learning_rate": 1.1586588180159241e-05,
+      "loss": 1.1155,
+      "step": 5459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0646809183440054,
+      "learning_rate": 1.1583859343691135e-05,
+      "loss": 1.0659,
+      "step": 5460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.955441148016663,
+      "learning_rate": 1.1581130386239952e-05,
+      "loss": 1.1851,
+      "step": 5461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.049912772949109,
+      "learning_rate": 1.157840130801414e-05,
+      "loss": 1.1482,
+      "step": 5462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8983800057694364,
+      "learning_rate": 1.1575672109222161e-05,
+      "loss": 1.0522,
+      "step": 5463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.822271314210498,
+      "learning_rate": 1.1572942790072486e-05,
+      "loss": 1.0898,
+      "step": 5464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4982371771707714,
+      "learning_rate": 1.157021335077359e-05,
+      "loss": 1.2314,
+      "step": 5465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.817046776109075,
+      "learning_rate": 1.1567483791533964e-05,
+      "loss": 1.0852,
+      "step": 5466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.557018605903915,
+      "learning_rate": 1.1564754112562107e-05,
+      "loss": 1.1628,
+      "step": 5467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2891812460240697,
+      "learning_rate": 1.1562024314066523e-05,
+      "loss": 1.2346,
+      "step": 5468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.591032801069932,
+      "learning_rate": 1.1559294396255724e-05,
+      "loss": 1.1411,
+      "step": 5469
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5410677712173078,
+      "learning_rate": 1.1556564359338244e-05,
+      "loss": 1.1165,
+      "step": 5470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9907600841126505,
+      "learning_rate": 1.1553834203522614e-05,
+      "loss": 1.1567,
+      "step": 5471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8785729891760776,
+      "learning_rate": 1.1551103929017371e-05,
+      "loss": 1.2046,
+      "step": 5472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2811205232426586,
+      "learning_rate": 1.154837353603107e-05,
+      "loss": 1.1343,
+      "step": 5473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9745118209085906,
+      "learning_rate": 1.1545643024772276e-05,
+      "loss": 1.0649,
+      "step": 5474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6611997592449423,
+      "learning_rate": 1.1542912395449554e-05,
+      "loss": 1.0574,
+      "step": 5475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8515229304879974,
+      "learning_rate": 1.1540181648271487e-05,
+      "loss": 1.1111,
+      "step": 5476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.557307437064218,
+      "learning_rate": 1.1537450783446659e-05,
+      "loss": 1.1182,
+      "step": 5477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6961412473917417,
+      "learning_rate": 1.1534719801183669e-05,
+      "loss": 1.0803,
+      "step": 5478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.669284097338104,
+      "learning_rate": 1.1531988701691124e-05,
+      "loss": 1.1411,
+      "step": 5479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0331807734445175,
+      "learning_rate": 1.1529257485177639e-05,
+      "loss": 1.1565,
+      "step": 5480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9504430203461616,
+      "learning_rate": 1.1526526151851838e-05,
+      "loss": 1.1072,
+      "step": 5481
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1468122231286197,
+      "learning_rate": 1.1523794701922351e-05,
+      "loss": 1.1765,
+      "step": 5482
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.146860505037417,
+      "learning_rate": 1.1521063135597821e-05,
+      "loss": 1.1421,
+      "step": 5483
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.045875126280745,
+      "learning_rate": 1.1518331453086901e-05,
+      "loss": 1.0969,
+      "step": 5484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0044388908826276,
+      "learning_rate": 1.151559965459825e-05,
+      "loss": 1.1716,
+      "step": 5485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.910282675025939,
+      "learning_rate": 1.1512867740340534e-05,
+      "loss": 1.1174,
+      "step": 5486
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.85754308464666,
+      "learning_rate": 1.1510135710522435e-05,
+      "loss": 1.1545,
+      "step": 5487
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.45678935901917,
+      "learning_rate": 1.1507403565352631e-05,
+      "loss": 1.114,
+      "step": 5488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.092092183951373,
+      "learning_rate": 1.1504671305039825e-05,
+      "loss": 1.1101,
+      "step": 5489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2367219393231283,
+      "learning_rate": 1.1501938929792716e-05,
+      "loss": 1.166,
+      "step": 5490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8120449128454643,
+      "learning_rate": 1.149920643982002e-05,
+      "loss": 1.1672,
+      "step": 5491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.841466319262014,
+      "learning_rate": 1.149647383533045e-05,
+      "loss": 1.072,
+      "step": 5492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.950661542533252,
+      "learning_rate": 1.1493741116532747e-05,
+      "loss": 1.1292,
+      "step": 5493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0498746916743205,
+      "learning_rate": 1.1491008283635648e-05,
+      "loss": 1.1531,
+      "step": 5494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2966250340675045,
+      "learning_rate": 1.148827533684789e-05,
+      "loss": 1.1282,
+      "step": 5495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2378275405897385,
+      "learning_rate": 1.1485542276378237e-05,
+      "loss": 1.0642,
+      "step": 5496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.415394526169915,
+      "learning_rate": 1.1482809102435458e-05,
+      "loss": 1.1479,
+      "step": 5497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.093497169493864,
+      "learning_rate": 1.1480075815228316e-05,
+      "loss": 1.2129,
+      "step": 5498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.939716822338418,
+      "learning_rate": 1.1477342414965602e-05,
+      "loss": 1.127,
+      "step": 5499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.017120677647212,
+      "learning_rate": 1.1474608901856103e-05,
+      "loss": 1.1616,
+      "step": 5500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4204106929895715,
+      "learning_rate": 1.1471875276108617e-05,
+      "loss": 1.1831,
+      "step": 5501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0060547080624596,
+      "learning_rate": 1.1469141537931953e-05,
+      "loss": 1.1493,
+      "step": 5502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9779550237422865,
+      "learning_rate": 1.146640768753493e-05,
+      "loss": 1.1235,
+      "step": 5503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9615534044174936,
+      "learning_rate": 1.1463673725126373e-05,
+      "loss": 1.1011,
+      "step": 5504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1810232480277247,
+      "learning_rate": 1.146093965091511e-05,
+      "loss": 1.1475,
+      "step": 5505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0267151301825574,
+      "learning_rate": 1.1458205465109988e-05,
+      "loss": 1.1341,
+      "step": 5506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.423981351682442,
+      "learning_rate": 1.145547116791986e-05,
+      "loss": 1.1191,
+      "step": 5507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6893667696382884,
+      "learning_rate": 1.1452736759553581e-05,
+      "loss": 1.1621,
+      "step": 5508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1110390593521053,
+      "learning_rate": 1.145000224022002e-05,
+      "loss": 1.1853,
+      "step": 5509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7507449653126224,
+      "learning_rate": 1.1447267610128052e-05,
+      "loss": 1.1423,
+      "step": 5510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.735330679280366,
+      "learning_rate": 1.1444532869486564e-05,
+      "loss": 1.2083,
+      "step": 5511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7053274065624144,
+      "learning_rate": 1.1441798018504447e-05,
+      "loss": 1.1069,
+      "step": 5512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8951145036564094,
+      "learning_rate": 1.1439063057390606e-05,
+      "loss": 1.0466,
+      "step": 5513
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.760877491071353,
+      "learning_rate": 1.1436327986353948e-05,
+      "loss": 1.0598,
+      "step": 5514
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8967882827205833,
+      "learning_rate": 1.143359280560339e-05,
+      "loss": 1.1809,
+      "step": 5515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7171771757224206,
+      "learning_rate": 1.1430857515347862e-05,
+      "loss": 1.1013,
+      "step": 5516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.387980006959978,
+      "learning_rate": 1.1428122115796299e-05,
+      "loss": 1.1321,
+      "step": 5517
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3893906230962565,
+      "learning_rate": 1.1425386607157642e-05,
+      "loss": 1.1294,
+      "step": 5518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4171173431739046,
+      "learning_rate": 1.1422650989640844e-05,
+      "loss": 1.1731,
+      "step": 5519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.232516888017283,
+      "learning_rate": 1.1419915263454867e-05,
+      "loss": 1.1162,
+      "step": 5520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.026016287249588,
+      "learning_rate": 1.1417179428808675e-05,
+      "loss": 1.0801,
+      "step": 5521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.955446557067034,
+      "learning_rate": 1.1414443485911248e-05,
+      "loss": 1.1611,
+      "step": 5522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5661155009558567,
+      "learning_rate": 1.1411707434971571e-05,
+      "loss": 1.061,
+      "step": 5523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.264484218240728,
+      "learning_rate": 1.1408971276198638e-05,
+      "loss": 1.1323,
+      "step": 5524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.123639381855931,
+      "learning_rate": 1.1406235009801447e-05,
+      "loss": 1.1362,
+      "step": 5525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.1925361569224595,
+      "learning_rate": 1.140349863598901e-05,
+      "loss": 1.1987,
+      "step": 5526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.428541505099189,
+      "learning_rate": 1.1400762154970346e-05,
+      "loss": 1.1919,
+      "step": 5527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.232532624207078,
+      "learning_rate": 1.1398025566954478e-05,
+      "loss": 1.1863,
+      "step": 5528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2244797207744775,
+      "learning_rate": 1.139528887215044e-05,
+      "loss": 1.168,
+      "step": 5529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.968783151761469,
+      "learning_rate": 1.1392552070767282e-05,
+      "loss": 1.1438,
+      "step": 5530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5161826582199835,
+      "learning_rate": 1.1389815163014044e-05,
+      "loss": 1.1226,
+      "step": 5531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.031521284069423,
+      "learning_rate": 1.1387078149099792e-05,
+      "loss": 1.1057,
+      "step": 5532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6974405378555404,
+      "learning_rate": 1.1384341029233588e-05,
+      "loss": 1.0581,
+      "step": 5533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4892628990262247,
+      "learning_rate": 1.1381603803624512e-05,
+      "loss": 1.1724,
+      "step": 5534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.804183503166095,
+      "learning_rate": 1.1378866472481642e-05,
+      "loss": 1.1606,
+      "step": 5535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.841506406762177,
+      "learning_rate": 1.1376129036014073e-05,
+      "loss": 1.1045,
+      "step": 5536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.967244688619949,
+      "learning_rate": 1.1373391494430901e-05,
+      "loss": 1.0874,
+      "step": 5537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7637912754688636,
+      "learning_rate": 1.1370653847941233e-05,
+      "loss": 1.1724,
+      "step": 5538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8786993744167133,
+      "learning_rate": 1.1367916096754185e-05,
+      "loss": 1.1199,
+      "step": 5539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.159319615411806,
+      "learning_rate": 1.1365178241078884e-05,
+      "loss": 1.1914,
+      "step": 5540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0300317832967925,
+      "learning_rate": 1.1362440281124456e-05,
+      "loss": 1.072,
+      "step": 5541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.928571196313523,
+      "learning_rate": 1.1359702217100041e-05,
+      "loss": 1.1016,
+      "step": 5542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.745479500377566,
+      "learning_rate": 1.1356964049214784e-05,
+      "loss": 1.1094,
+      "step": 5543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4664228769505,
+      "learning_rate": 1.1354225777677848e-05,
+      "loss": 1.1602,
+      "step": 5544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.443029588446821,
+      "learning_rate": 1.1351487402698388e-05,
+      "loss": 1.1248,
+      "step": 5545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.598191915765113,
+      "learning_rate": 1.1348748924485577e-05,
+      "loss": 1.1013,
+      "step": 5546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.192881710916119,
+      "learning_rate": 1.1346010343248594e-05,
+      "loss": 1.2009,
+      "step": 5547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.301502429458105,
+      "learning_rate": 1.1343271659196624e-05,
+      "loss": 1.1831,
+      "step": 5548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.905598399396722,
+      "learning_rate": 1.1340532872538861e-05,
+      "loss": 1.0742,
+      "step": 5549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.228408559166158,
+      "learning_rate": 1.1337793983484513e-05,
+      "loss": 1.1895,
+      "step": 5550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.594183285264849,
+      "learning_rate": 1.1335054992242784e-05,
+      "loss": 1.127,
+      "step": 5551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3119575971848954,
+      "learning_rate": 1.1332315899022892e-05,
+      "loss": 1.1018,
+      "step": 5552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2452246961397866,
+      "learning_rate": 1.1329576704034064e-05,
+      "loss": 1.1379,
+      "step": 5553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5403530788582533,
+      "learning_rate": 1.1326837407485537e-05,
+      "loss": 1.1631,
+      "step": 5554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.262359962353096,
+      "learning_rate": 1.1324098009586548e-05,
+      "loss": 1.106,
+      "step": 5555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8872401211237664,
+      "learning_rate": 1.1321358510546346e-05,
+      "loss": 1.1604,
+      "step": 5556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3028124562321777,
+      "learning_rate": 1.1318618910574186e-05,
+      "loss": 1.0913,
+      "step": 5557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.306478312745292,
+      "learning_rate": 1.1315879209879337e-05,
+      "loss": 1.0999,
+      "step": 5558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1942933958638378,
+      "learning_rate": 1.1313139408671068e-05,
+      "loss": 1.063,
+      "step": 5559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2519286071007736,
+      "learning_rate": 1.131039950715866e-05,
+      "loss": 1.0769,
+      "step": 5560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.067717305244434,
+      "learning_rate": 1.1307659505551396e-05,
+      "loss": 1.1545,
+      "step": 5561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0248980565229306,
+      "learning_rate": 1.1304919404058576e-05,
+      "loss": 1.1062,
+      "step": 5562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.91363427124416,
+      "learning_rate": 1.1302179202889505e-05,
+      "loss": 1.0474,
+      "step": 5563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.353074125544352,
+      "learning_rate": 1.1299438902253486e-05,
+      "loss": 1.1311,
+      "step": 5564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.533663980917696,
+      "learning_rate": 1.129669850235984e-05,
+      "loss": 1.1487,
+      "step": 5565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0721887383658824,
+      "learning_rate": 1.1293958003417894e-05,
+      "loss": 1.116,
+      "step": 5566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.980351898000634,
+      "learning_rate": 1.1291217405636978e-05,
+      "loss": 1.125,
+      "step": 5567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.657829495477688,
+      "learning_rate": 1.1288476709226434e-05,
+      "loss": 1.1768,
+      "step": 5568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.71661616774112,
+      "learning_rate": 1.1285735914395613e-05,
+      "loss": 1.1372,
+      "step": 5569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.414832856464059,
+      "learning_rate": 1.1282995021353866e-05,
+      "loss": 1.1924,
+      "step": 5570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0356188839446716,
+      "learning_rate": 1.1280254030310558e-05,
+      "loss": 1.1282,
+      "step": 5571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.651602738351137,
+      "learning_rate": 1.1277512941475059e-05,
+      "loss": 1.0078,
+      "step": 5572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1201057457751413,
+      "learning_rate": 1.1274771755056748e-05,
+      "loss": 1.1167,
+      "step": 5573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9325991460732377,
+      "learning_rate": 1.1272030471265012e-05,
+      "loss": 1.1465,
+      "step": 5574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6368296029623357,
+      "learning_rate": 1.1269289090309239e-05,
+      "loss": 1.1643,
+      "step": 5575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.220490089869533,
+      "learning_rate": 1.1266547612398832e-05,
+      "loss": 1.1787,
+      "step": 5576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.071558169678212,
+      "learning_rate": 1.12638060377432e-05,
+      "loss": 1.1711,
+      "step": 5577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4105148401362633,
+      "learning_rate": 1.126106436655176e-05,
+      "loss": 1.1548,
+      "step": 5578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7950073236789508,
+      "learning_rate": 1.1258322599033929e-05,
+      "loss": 1.1821,
+      "step": 5579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.66695989132309,
+      "learning_rate": 1.1255580735399141e-05,
+      "loss": 1.2146,
+      "step": 5580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.683812877841381,
+      "learning_rate": 1.1252838775856829e-05,
+      "loss": 1.1509,
+      "step": 5581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.091744940731837,
+      "learning_rate": 1.1250096720616441e-05,
+      "loss": 1.1899,
+      "step": 5582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0091159269092467,
+      "learning_rate": 1.1247354569887434e-05,
+      "loss": 1.1313,
+      "step": 5583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0283446240241148,
+      "learning_rate": 1.1244612323879256e-05,
+      "loss": 1.1487,
+      "step": 5584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.991853956694547,
+      "learning_rate": 1.1241869982801378e-05,
+      "loss": 1.1138,
+      "step": 5585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2489624582409498,
+      "learning_rate": 1.1239127546863276e-05,
+      "loss": 1.198,
+      "step": 5586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.197761535344188,
+      "learning_rate": 1.1236385016274432e-05,
+      "loss": 1.1033,
+      "step": 5587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.964628517275707,
+      "learning_rate": 1.1233642391244329e-05,
+      "loss": 1.1365,
+      "step": 5588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3666194109783953,
+      "learning_rate": 1.1230899671982466e-05,
+      "loss": 1.1057,
+      "step": 5589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.499194335996054,
+      "learning_rate": 1.1228156858698344e-05,
+      "loss": 1.1284,
+      "step": 5590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9691443189267512,
+      "learning_rate": 1.1225413951601472e-05,
+      "loss": 1.0991,
+      "step": 5591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3314800443805206,
+      "learning_rate": 1.122267095090137e-05,
+      "loss": 1.1836,
+      "step": 5592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5360668956340664,
+      "learning_rate": 1.1219927856807561e-05,
+      "loss": 1.1218,
+      "step": 5593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.138373230617035,
+      "learning_rate": 1.1217184669529574e-05,
+      "loss": 1.144,
+      "step": 5594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6275717108018184,
+      "learning_rate": 1.1214441389276952e-05,
+      "loss": 1.1328,
+      "step": 5595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.130482566192294,
+      "learning_rate": 1.1211698016259234e-05,
+      "loss": 1.1682,
+      "step": 5596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5580463734317274,
+      "learning_rate": 1.1208954550685978e-05,
+      "loss": 1.1433,
+      "step": 5597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.440861603924928,
+      "learning_rate": 1.120621099276674e-05,
+      "loss": 1.0479,
+      "step": 5598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3348803116325225,
+      "learning_rate": 1.1203467342711087e-05,
+      "loss": 1.0896,
+      "step": 5599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.907029753114099,
+      "learning_rate": 1.1200723600728596e-05,
+      "loss": 1.1165,
+      "step": 5600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9013969532368384,
+      "learning_rate": 1.1197979767028843e-05,
+      "loss": 1.1084,
+      "step": 5601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4366177931182182,
+      "learning_rate": 1.1195235841821423e-05,
+      "loss": 1.1101,
+      "step": 5602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.399243719766636,
+      "learning_rate": 1.1192491825315922e-05,
+      "loss": 1.1399,
+      "step": 5603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.922736100306989,
+      "learning_rate": 1.1189747717721949e-05,
+      "loss": 1.1509,
+      "step": 5604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.361566392027786,
+      "learning_rate": 1.1187003519249104e-05,
+      "loss": 1.1482,
+      "step": 5605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7882256168911526,
+      "learning_rate": 1.118425923010701e-05,
+      "loss": 1.1104,
+      "step": 5606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.404633873787557,
+      "learning_rate": 1.118151485050529e-05,
+      "loss": 1.1821,
+      "step": 5607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.788422159408508,
+      "learning_rate": 1.1178770380653567e-05,
+      "loss": 1.1472,
+      "step": 5608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9579679743373464,
+      "learning_rate": 1.1176025820761483e-05,
+      "loss": 1.1345,
+      "step": 5609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2348971637894706,
+      "learning_rate": 1.117328117103868e-05,
+      "loss": 1.1858,
+      "step": 5610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.042326806218967,
+      "learning_rate": 1.1170536431694805e-05,
+      "loss": 1.134,
+      "step": 5611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6018675780430844,
+      "learning_rate": 1.1167791602939518e-05,
+      "loss": 1.1321,
+      "step": 5612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.023301787035073,
+      "learning_rate": 1.1165046684982483e-05,
+      "loss": 1.1243,
+      "step": 5613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6797540354559857,
+      "learning_rate": 1.1162301678033367e-05,
+      "loss": 1.1538,
+      "step": 5614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1886790839250096,
+      "learning_rate": 1.1159556582301851e-05,
+      "loss": 1.0562,
+      "step": 5615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4210156228186297,
+      "learning_rate": 1.1156811397997615e-05,
+      "loss": 1.167,
+      "step": 5616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9695811587309335,
+      "learning_rate": 1.1154066125330358e-05,
+      "loss": 1.124,
+      "step": 5617
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0116026179674256,
+      "learning_rate": 1.1151320764509767e-05,
+      "loss": 1.1736,
+      "step": 5618
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.219958041697891,
+      "learning_rate": 1.1148575315745553e-05,
+      "loss": 1.105,
+      "step": 5619
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8897292377884907,
+      "learning_rate": 1.114582977924743e-05,
+      "loss": 1.1548,
+      "step": 5620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9830569533848275,
+      "learning_rate": 1.1143084155225108e-05,
+      "loss": 1.1423,
+      "step": 5621
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1794680383757434,
+      "learning_rate": 1.1140338443888314e-05,
+      "loss": 1.1514,
+      "step": 5622
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.36494109131632,
+      "learning_rate": 1.1137592645446781e-05,
+      "loss": 1.1731,
+      "step": 5623
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8021914569349375,
+      "learning_rate": 1.1134846760110248e-05,
+      "loss": 1.114,
+      "step": 5624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.662773060040859,
+      "learning_rate": 1.1132100788088456e-05,
+      "loss": 1.1702,
+      "step": 5625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6237644806150158,
+      "learning_rate": 1.1129354729591158e-05,
+      "loss": 1.1404,
+      "step": 5626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8302538631205145,
+      "learning_rate": 1.1126608584828106e-05,
+      "loss": 1.1748,
+      "step": 5627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3379314667316327,
+      "learning_rate": 1.1123862354009072e-05,
+      "loss": 1.1392,
+      "step": 5628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8495835163532863,
+      "learning_rate": 1.1121116037343826e-05,
+      "loss": 1.1702,
+      "step": 5629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6972628849976896,
+      "learning_rate": 1.1118369635042143e-05,
+      "loss": 1.1733,
+      "step": 5630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.235954040065353,
+      "learning_rate": 1.1115623147313806e-05,
+      "loss": 1.1077,
+      "step": 5631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1304453207440637,
+      "learning_rate": 1.1112876574368606e-05,
+      "loss": 1.1211,
+      "step": 5632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5980169543093927,
+      "learning_rate": 1.1110129916416343e-05,
+      "loss": 1.0386,
+      "step": 5633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.55198162761763,
+      "learning_rate": 1.110738317366682e-05,
+      "loss": 1.0143,
+      "step": 5634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3052163375899988,
+      "learning_rate": 1.1104636346329838e-05,
+      "loss": 1.1311,
+      "step": 5635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0521438147671605,
+      "learning_rate": 1.1101889434615227e-05,
+      "loss": 1.1675,
+      "step": 5636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.874110241204265,
+      "learning_rate": 1.1099142438732799e-05,
+      "loss": 1.0642,
+      "step": 5637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.921247464794527,
+      "learning_rate": 1.1096395358892389e-05,
+      "loss": 1.1785,
+      "step": 5638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.198665036356354,
+      "learning_rate": 1.1093648195303832e-05,
+      "loss": 1.1633,
+      "step": 5639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.611527798456971,
+      "learning_rate": 1.1090900948176968e-05,
+      "loss": 1.1011,
+      "step": 5640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3346121503705213,
+      "learning_rate": 1.1088153617721646e-05,
+      "loss": 1.1333,
+      "step": 5641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.204649635941704,
+      "learning_rate": 1.1085406204147722e-05,
+      "loss": 1.1411,
+      "step": 5642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6960339362464634,
+      "learning_rate": 1.108265870766506e-05,
+      "loss": 1.1904,
+      "step": 5643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.507747634413272,
+      "learning_rate": 1.107991112848352e-05,
+      "loss": 1.1357,
+      "step": 5644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8662142956663663,
+      "learning_rate": 1.1077163466812981e-05,
+      "loss": 1.1362,
+      "step": 5645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1652827410361075,
+      "learning_rate": 1.1074415722863323e-05,
+      "loss": 1.1797,
+      "step": 5646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6433739291432405,
+      "learning_rate": 1.1071667896844432e-05,
+      "loss": 1.0984,
+      "step": 5647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0677425258768487,
+      "learning_rate": 1.1068919988966198e-05,
+      "loss": 1.1648,
+      "step": 5648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.316218564054901,
+      "learning_rate": 1.1066171999438521e-05,
+      "loss": 1.1479,
+      "step": 5649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9004576144342002,
+      "learning_rate": 1.106342392847131e-05,
+      "loss": 1.0791,
+      "step": 5650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.66557531781484,
+      "learning_rate": 1.1060675776274472e-05,
+      "loss": 1.1379,
+      "step": 5651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.244304260706456,
+      "learning_rate": 1.1057927543057925e-05,
+      "loss": 1.1213,
+      "step": 5652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6071064640240054,
+      "learning_rate": 1.1055179229031598e-05,
+      "loss": 1.1128,
+      "step": 5653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8739939414139752,
+      "learning_rate": 1.1052430834405414e-05,
+      "loss": 1.1807,
+      "step": 5654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.969137666512847,
+      "learning_rate": 1.1049682359389311e-05,
+      "loss": 1.1909,
+      "step": 5655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7961378329813074,
+      "learning_rate": 1.1046933804193234e-05,
+      "loss": 1.1553,
+      "step": 5656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.328775418722491,
+      "learning_rate": 1.1044185169027126e-05,
+      "loss": 1.1191,
+      "step": 5657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.739119727008282,
+      "learning_rate": 1.1041436454100949e-05,
+      "loss": 1.1748,
+      "step": 5658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.220980608054877,
+      "learning_rate": 1.1038687659624659e-05,
+      "loss": 1.1299,
+      "step": 5659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6839962951209246,
+      "learning_rate": 1.1035938785808223e-05,
+      "loss": 1.0795,
+      "step": 5660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1698894764965626,
+      "learning_rate": 1.1033189832861611e-05,
+      "loss": 1.1555,
+      "step": 5661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.655263143830236,
+      "learning_rate": 1.1030440800994806e-05,
+      "loss": 1.1082,
+      "step": 5662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3989877254698113,
+      "learning_rate": 1.1027691690417795e-05,
+      "loss": 1.175,
+      "step": 5663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0832543312309455,
+      "learning_rate": 1.1024942501340561e-05,
+      "loss": 1.1289,
+      "step": 5664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3842811544548512,
+      "learning_rate": 1.1022193233973104e-05,
+      "loss": 1.1885,
+      "step": 5665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.879085121136306,
+      "learning_rate": 1.101944388852543e-05,
+      "loss": 1.137,
+      "step": 5666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9489357967303276,
+      "learning_rate": 1.1016694465207546e-05,
+      "loss": 1.1633,
+      "step": 5667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.379223329765459,
+      "learning_rate": 1.1013944964229464e-05,
+      "loss": 1.136,
+      "step": 5668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7827692828724166,
+      "learning_rate": 1.1011195385801207e-05,
+      "loss": 1.1704,
+      "step": 5669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.299156274701956,
+      "learning_rate": 1.1008445730132805e-05,
+      "loss": 1.1265,
+      "step": 5670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6206612383327883,
+      "learning_rate": 1.100569599743428e-05,
+      "loss": 1.0901,
+      "step": 5671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0947375550891896,
+      "learning_rate": 1.1002946187915683e-05,
+      "loss": 1.1008,
+      "step": 5672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.652795973361289,
+      "learning_rate": 1.1000196301787049e-05,
+      "loss": 1.134,
+      "step": 5673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0987844424052264,
+      "learning_rate": 1.099744633925843e-05,
+      "loss": 1.1255,
+      "step": 5674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1211064166700027,
+      "learning_rate": 1.0994696300539884e-05,
+      "loss": 1.1318,
+      "step": 5675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.267713523640053,
+      "learning_rate": 1.0991946185841473e-05,
+      "loss": 1.1458,
+      "step": 5676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.568499737034574,
+      "learning_rate": 1.0989195995373266e-05,
+      "loss": 1.1345,
+      "step": 5677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.454138352778144,
+      "learning_rate": 1.0986445729345327e-05,
+      "loss": 1.1802,
+      "step": 5678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0479788478623524,
+      "learning_rate": 1.0983695387967743e-05,
+      "loss": 1.1248,
+      "step": 5679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5478666216638084,
+      "learning_rate": 1.09809449714506e-05,
+      "loss": 1.1077,
+      "step": 5680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.428008816266386,
+      "learning_rate": 1.0978194480003986e-05,
+      "loss": 1.1611,
+      "step": 5681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3598539729683075,
+      "learning_rate": 1.0975443913837993e-05,
+      "loss": 1.1379,
+      "step": 5682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.037936603524794,
+      "learning_rate": 1.0972693273162728e-05,
+      "loss": 1.1687,
+      "step": 5683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1672507503854237,
+      "learning_rate": 1.09699425581883e-05,
+      "loss": 1.1101,
+      "step": 5684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.776032042484241,
+      "learning_rate": 1.0967191769124818e-05,
+      "loss": 1.1743,
+      "step": 5685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5207119283351287,
+      "learning_rate": 1.0964440906182406e-05,
+      "loss": 1.1045,
+      "step": 5686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.717521353317495,
+      "learning_rate": 1.0961689969571183e-05,
+      "loss": 1.1909,
+      "step": 5687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4750708368601435,
+      "learning_rate": 1.0958938959501279e-05,
+      "loss": 1.0522,
+      "step": 5688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.231744448188298,
+      "learning_rate": 1.0956187876182837e-05,
+      "loss": 1.1926,
+      "step": 5689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.617892190111447,
+      "learning_rate": 1.0953436719825996e-05,
+      "loss": 1.1763,
+      "step": 5690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.36716128076891,
+      "learning_rate": 1.0950685490640898e-05,
+      "loss": 1.1335,
+      "step": 5691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.258192562061826,
+      "learning_rate": 1.0947934188837697e-05,
+      "loss": 1.1504,
+      "step": 5692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.855419644972023,
+      "learning_rate": 1.0945182814626557e-05,
+      "loss": 1.1245,
+      "step": 5693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1582299221988634,
+      "learning_rate": 1.0942431368217635e-05,
+      "loss": 1.1343,
+      "step": 5694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6734972453187438,
+      "learning_rate": 1.0939679849821106e-05,
+      "loss": 1.1294,
+      "step": 5695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.392860504887957,
+      "learning_rate": 1.0936928259647139e-05,
+      "loss": 1.1738,
+      "step": 5696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.340196845702525,
+      "learning_rate": 1.0934176597905916e-05,
+      "loss": 1.1426,
+      "step": 5697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.476203300907578,
+      "learning_rate": 1.0931424864807624e-05,
+      "loss": 1.092,
+      "step": 5698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.727579413414981,
+      "learning_rate": 1.0928673060562458e-05,
+      "loss": 1.1133,
+      "step": 5699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.424536041633867,
+      "learning_rate": 1.0925921185380608e-05,
+      "loss": 1.1628,
+      "step": 5700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.626902216819549,
+      "learning_rate": 1.0923169239472275e-05,
+      "loss": 1.084,
+      "step": 5701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1017522320334767,
+      "learning_rate": 1.0920417223047671e-05,
+      "loss": 1.0601,
+      "step": 5702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.83223954076038,
+      "learning_rate": 1.0917665136317008e-05,
+      "loss": 1.146,
+      "step": 5703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8906357061020307,
+      "learning_rate": 1.0914912979490503e-05,
+      "loss": 1.1768,
+      "step": 5704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6853468437258794,
+      "learning_rate": 1.0912160752778378e-05,
+      "loss": 1.1033,
+      "step": 5705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1818183040826904,
+      "learning_rate": 1.0909408456390867e-05,
+      "loss": 1.1675,
+      "step": 5706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7275344146568723,
+      "learning_rate": 1.0906656090538196e-05,
+      "loss": 1.0801,
+      "step": 5707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8806444039392707,
+      "learning_rate": 1.0903903655430612e-05,
+      "loss": 1.0425,
+      "step": 5708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.872470946290299,
+      "learning_rate": 1.0901151151278359e-05,
+      "loss": 1.1099,
+      "step": 5709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.205996611025854,
+      "learning_rate": 1.0898398578291682e-05,
+      "loss": 1.1145,
+      "step": 5710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1506885025159512,
+      "learning_rate": 1.0895645936680838e-05,
+      "loss": 1.1697,
+      "step": 5711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1743864281452026,
+      "learning_rate": 1.0892893226656092e-05,
+      "loss": 1.1602,
+      "step": 5712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9434154445987977,
+      "learning_rate": 1.0890140448427705e-05,
+      "loss": 1.2036,
+      "step": 5713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3888023737479025,
+      "learning_rate": 1.0887387602205949e-05,
+      "loss": 1.1445,
+      "step": 5714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0567719709341623,
+      "learning_rate": 1.08846346882011e-05,
+      "loss": 1.1763,
+      "step": 5715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3132516971994685,
+      "learning_rate": 1.0881881706623443e-05,
+      "loss": 1.1328,
+      "step": 5716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1141184680401754,
+      "learning_rate": 1.0879128657683256e-05,
+      "loss": 1.1436,
+      "step": 5717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5849638450234598,
+      "learning_rate": 1.087637554159084e-05,
+      "loss": 1.1921,
+      "step": 5718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2034814847020763,
+      "learning_rate": 1.0873622358556486e-05,
+      "loss": 1.1377,
+      "step": 5719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.051398411683029,
+      "learning_rate": 1.0870869108790496e-05,
+      "loss": 1.1348,
+      "step": 5720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2869434749535658,
+      "learning_rate": 1.086811579250318e-05,
+      "loss": 1.093,
+      "step": 5721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8671532233889283,
+      "learning_rate": 1.0865362409904848e-05,
+      "loss": 1.0718,
+      "step": 5722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3815572515315977,
+      "learning_rate": 1.0862608961205818e-05,
+      "loss": 1.2024,
+      "step": 5723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.860312447015098,
+      "learning_rate": 1.0859855446616412e-05,
+      "loss": 1.0754,
+      "step": 5724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.75589709827226,
+      "learning_rate": 1.0857101866346953e-05,
+      "loss": 1.0977,
+      "step": 5725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0992319244405735,
+      "learning_rate": 1.0854348220607783e-05,
+      "loss": 1.1851,
+      "step": 5726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3323507429702324,
+      "learning_rate": 1.085159450960923e-05,
+      "loss": 1.2117,
+      "step": 5727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4260840426224974,
+      "learning_rate": 1.0848840733561638e-05,
+      "loss": 1.0662,
+      "step": 5728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.824067906731295,
+      "learning_rate": 1.0846086892675358e-05,
+      "loss": 1.0818,
+      "step": 5729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.88068040219108,
+      "learning_rate": 1.084333298716074e-05,
+      "loss": 1.1753,
+      "step": 5730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.154502950423468,
+      "learning_rate": 1.0840579017228137e-05,
+      "loss": 1.1487,
+      "step": 5731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6748380426337155,
+      "learning_rate": 1.083782498308792e-05,
+      "loss": 1.1912,
+      "step": 5732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.862754017973029,
+      "learning_rate": 1.0835070884950447e-05,
+      "loss": 1.1228,
+      "step": 5733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6897016823704565,
+      "learning_rate": 1.0832316723026092e-05,
+      "loss": 1.1079,
+      "step": 5734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1618293368861616,
+      "learning_rate": 1.0829562497525237e-05,
+      "loss": 1.1497,
+      "step": 5735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7786638364287604,
+      "learning_rate": 1.0826808208658257e-05,
+      "loss": 1.2334,
+      "step": 5736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.038985574573579,
+      "learning_rate": 1.0824053856635543e-05,
+      "loss": 1.0889,
+      "step": 5737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.996799249469848,
+      "learning_rate": 1.0821299441667481e-05,
+      "loss": 1.0688,
+      "step": 5738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8845903166219853,
+      "learning_rate": 1.0818544963964472e-05,
+      "loss": 1.1975,
+      "step": 5739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.292623481715673,
+      "learning_rate": 1.0815790423736914e-05,
+      "loss": 1.1257,
+      "step": 5740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6346815732453894,
+      "learning_rate": 1.0813035821195212e-05,
+      "loss": 1.1553,
+      "step": 5741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.577392105941309,
+      "learning_rate": 1.0810281156549778e-05,
+      "loss": 1.1123,
+      "step": 5742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.176451937153492,
+      "learning_rate": 1.0807526430011028e-05,
+      "loss": 1.1042,
+      "step": 5743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0838093034506544,
+      "learning_rate": 1.0804771641789377e-05,
+      "loss": 1.1072,
+      "step": 5744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.166637550782509,
+      "learning_rate": 1.0802016792095254e-05,
+      "loss": 1.0986,
+      "step": 5745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.987350457047538,
+      "learning_rate": 1.079926188113909e-05,
+      "loss": 1.0903,
+      "step": 5746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6611852392826947,
+      "learning_rate": 1.0796506909131312e-05,
+      "loss": 1.072,
+      "step": 5747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.216370414530271,
+      "learning_rate": 1.0793751876282363e-05,
+      "loss": 1.1943,
+      "step": 5748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2195651482804153,
+      "learning_rate": 1.0790996782802687e-05,
+      "loss": 1.2241,
+      "step": 5749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.917784970565413,
+      "learning_rate": 1.0788241628902727e-05,
+      "loss": 1.0833,
+      "step": 5750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8924643061583284,
+      "learning_rate": 1.0785486414792943e-05,
+      "loss": 1.0754,
+      "step": 5751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.416045435643972,
+      "learning_rate": 1.0782731140683786e-05,
+      "loss": 1.1443,
+      "step": 5752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1939220100511063,
+      "learning_rate": 1.0779975806785716e-05,
+      "loss": 1.1855,
+      "step": 5753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7207068736140076,
+      "learning_rate": 1.0777220413309204e-05,
+      "loss": 1.1797,
+      "step": 5754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2534845066408673,
+      "learning_rate": 1.0774464960464721e-05,
+      "loss": 1.1445,
+      "step": 5755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9450908513140392,
+      "learning_rate": 1.077170944846274e-05,
+      "loss": 1.1133,
+      "step": 5756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7360906581573516,
+      "learning_rate": 1.0768953877513737e-05,
+      "loss": 1.1709,
+      "step": 5757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2763815073646936,
+      "learning_rate": 1.0766198247828201e-05,
+      "loss": 1.1792,
+      "step": 5758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.041846435833268,
+      "learning_rate": 1.0763442559616624e-05,
+      "loss": 1.0669,
+      "step": 5759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.022933363314506,
+      "learning_rate": 1.0760686813089493e-05,
+      "loss": 1.1047,
+      "step": 5760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.544203541751258,
+      "learning_rate": 1.0757931008457308e-05,
+      "loss": 1.0781,
+      "step": 5761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7918926857648567,
+      "learning_rate": 1.075517514593057e-05,
+      "loss": 1.1875,
+      "step": 5762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.618933110964401,
+      "learning_rate": 1.0752419225719785e-05,
+      "loss": 1.0953,
+      "step": 5763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.023149435716902,
+      "learning_rate": 1.0749663248035469e-05,
+      "loss": 1.1497,
+      "step": 5764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0194369701728236,
+      "learning_rate": 1.0746907213088134e-05,
+      "loss": 1.1357,
+      "step": 5765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3114261277524943,
+      "learning_rate": 1.0744151121088297e-05,
+      "loss": 1.1404,
+      "step": 5766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.049868606401472,
+      "learning_rate": 1.0741394972246484e-05,
+      "loss": 1.1511,
+      "step": 5767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.51149905287346,
+      "learning_rate": 1.0738638766773224e-05,
+      "loss": 1.062,
+      "step": 5768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.74265906877015,
+      "learning_rate": 1.0735882504879054e-05,
+      "loss": 1.1951,
+      "step": 5769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.913317122197895,
+      "learning_rate": 1.0733126186774504e-05,
+      "loss": 1.0972,
+      "step": 5770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6779971684176553,
+      "learning_rate": 1.0730369812670117e-05,
+      "loss": 1.0435,
+      "step": 5771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2604063451761527,
+      "learning_rate": 1.0727613382776441e-05,
+      "loss": 1.199,
+      "step": 5772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0137582845606437,
+      "learning_rate": 1.0724856897304022e-05,
+      "loss": 1.1111,
+      "step": 5773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1401142498595176,
+      "learning_rate": 1.0722100356463418e-05,
+      "loss": 1.145,
+      "step": 5774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.843723874951318,
+      "learning_rate": 1.0719343760465186e-05,
+      "loss": 1.1272,
+      "step": 5775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3675772052533985,
+      "learning_rate": 1.0716587109519889e-05,
+      "loss": 1.1006,
+      "step": 5776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2636806877720375,
+      "learning_rate": 1.0713830403838092e-05,
+      "loss": 1.1133,
+      "step": 5777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.208527347439763,
+      "learning_rate": 1.0711073643630368e-05,
+      "loss": 1.1232,
+      "step": 5778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.324451878214199,
+      "learning_rate": 1.0708316829107295e-05,
+      "loss": 1.1338,
+      "step": 5779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.34053229545913,
+      "learning_rate": 1.0705559960479444e-05,
+      "loss": 1.1272,
+      "step": 5780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3160855756891303,
+      "learning_rate": 1.0702803037957401e-05,
+      "loss": 1.0476,
+      "step": 5781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9187143497413017,
+      "learning_rate": 1.0700046061751758e-05,
+      "loss": 1.1326,
+      "step": 5782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8092411269064197,
+      "learning_rate": 1.0697289032073104e-05,
+      "loss": 1.158,
+      "step": 5783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1698378600564454,
+      "learning_rate": 1.0694531949132036e-05,
+      "loss": 1.1033,
+      "step": 5784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.926367527323234,
+      "learning_rate": 1.0691774813139151e-05,
+      "loss": 1.1152,
+      "step": 5785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.749081702402576,
+      "learning_rate": 1.0689017624305056e-05,
+      "loss": 1.1345,
+      "step": 5786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.308779156699152,
+      "learning_rate": 1.0686260382840354e-05,
+      "loss": 1.0938,
+      "step": 5787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.008506507430171,
+      "learning_rate": 1.0683503088955663e-05,
+      "loss": 1.1121,
+      "step": 5788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.05050578406173,
+      "learning_rate": 1.0680745742861595e-05,
+      "loss": 1.2,
+      "step": 5789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1643070802672844,
+      "learning_rate": 1.0677988344768768e-05,
+      "loss": 1.0977,
+      "step": 5790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.333308675762957,
+      "learning_rate": 1.0675230894887811e-05,
+      "loss": 1.156,
+      "step": 5791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.035819759974917,
+      "learning_rate": 1.067247339342935e-05,
+      "loss": 1.1499,
+      "step": 5792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0411330177762066,
+      "learning_rate": 1.0669715840604016e-05,
+      "loss": 1.114,
+      "step": 5793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.187477856651462,
+      "learning_rate": 1.0666958236622442e-05,
+      "loss": 1.0781,
+      "step": 5794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2975884922292282,
+      "learning_rate": 1.066420058169527e-05,
+      "loss": 1.1421,
+      "step": 5795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.205616088374686,
+      "learning_rate": 1.0661442876033145e-05,
+      "loss": 1.157,
+      "step": 5796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4048075076876314,
+      "learning_rate": 1.0658685119846713e-05,
+      "loss": 1.2085,
+      "step": 5797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.03788392426251,
+      "learning_rate": 1.0655927313346626e-05,
+      "loss": 1.1262,
+      "step": 5798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0159338127612862,
+      "learning_rate": 1.065316945674354e-05,
+      "loss": 1.095,
+      "step": 5799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4092420450868497,
+      "learning_rate": 1.0650411550248107e-05,
+      "loss": 1.0867,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1335817632823955,
+      "learning_rate": 1.0647653594070994e-05,
+      "loss": 1.1167,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.535647198402506,
+      "learning_rate": 1.0644895588422874e-05,
+      "loss": 1.1191,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4721187731702123,
+      "learning_rate": 1.0642137533514409e-05,
+      "loss": 1.1304,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7805892162806205,
+      "learning_rate": 1.0639379429556274e-05,
+      "loss": 1.1653,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.063761803028125,
+      "learning_rate": 1.0636621276759147e-05,
+      "loss": 1.1084,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0550944381647067,
+      "learning_rate": 1.0633863075333713e-05,
+      "loss": 1.2217,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1983663140299026,
+      "learning_rate": 1.0631104825490655e-05,
+      "loss": 1.1919,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1819380191785904,
+      "learning_rate": 1.062834652744066e-05,
+      "loss": 1.1443,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.851153532270454,
+      "learning_rate": 1.0625588181394424e-05,
+      "loss": 1.1982,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.192502629262485,
+      "learning_rate": 1.0622829787562639e-05,
+      "loss": 1.158,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.782965750526672,
+      "learning_rate": 1.0620071346156008e-05,
+      "loss": 1.1233,
+      "step": 5811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4399081411980106,
+      "learning_rate": 1.0617312857385235e-05,
+      "loss": 1.0837,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3594523786210773,
+      "learning_rate": 1.0614554321461025e-05,
+      "loss": 1.1721,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.618684247110089,
+      "learning_rate": 1.0611795738594089e-05,
+      "loss": 1.1482,
+      "step": 5814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8893261931563967,
+      "learning_rate": 1.0609037108995143e-05,
+      "loss": 1.1619,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.269732527194488,
+      "learning_rate": 1.0606278432874904e-05,
+      "loss": 1.2568,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9297380911114694,
+      "learning_rate": 1.0603519710444095e-05,
+      "loss": 1.0151,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.03749054865706,
+      "learning_rate": 1.0600760941913435e-05,
+      "loss": 1.1118,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3159190649627055,
+      "learning_rate": 1.059800212749366e-05,
+      "loss": 1.165,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.307375578929675,
+      "learning_rate": 1.0595243267395497e-05,
+      "loss": 1.1624,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.203034606063701,
+      "learning_rate": 1.0592484361829687e-05,
+      "loss": 1.1238,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.379990511503441,
+      "learning_rate": 1.0589725411006964e-05,
+      "loss": 1.1094,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.731150127142352,
+      "learning_rate": 1.0586966415138067e-05,
+      "loss": 1.1316,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.954701091759934,
+      "learning_rate": 1.058420737443375e-05,
+      "loss": 1.1418,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.509614008869726,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 1.1091,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1380355283990213,
+      "learning_rate": 1.0578689159361848e-05,
+      "loss": 1.2583,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.991026607015371,
+      "learning_rate": 1.057592998541577e-05,
+      "loss": 1.1897,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.133755685950092,
+      "learning_rate": 1.0573170767477285e-05,
+      "loss": 1.155,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.993092206089322,
+      "learning_rate": 1.0570411505757159e-05,
+      "loss": 1.1133,
+      "step": 5829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1159195170454126,
+      "learning_rate": 1.0567652200466157e-05,
+      "loss": 1.1399,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.206910436732445,
+      "learning_rate": 1.0564892851815045e-05,
+      "loss": 1.1694,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.312486835781966,
+      "learning_rate": 1.05621334600146e-05,
+      "loss": 1.2026,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.416471852447543,
+      "learning_rate": 1.0559374025275597e-05,
+      "loss": 1.1233,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.639467455387,
+      "learning_rate": 1.0556614547808812e-05,
+      "loss": 1.1411,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.679087901249185,
+      "learning_rate": 1.0553855027825035e-05,
+      "loss": 1.1064,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9562088690183836,
+      "learning_rate": 1.0551095465535047e-05,
+      "loss": 1.0742,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1629355522343827,
+      "learning_rate": 1.0548335861149635e-05,
+      "loss": 1.1097,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7755256307759537,
+      "learning_rate": 1.0545576214879594e-05,
+      "loss": 1.0767,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.899740895473844,
+      "learning_rate": 1.0542816526935722e-05,
+      "loss": 1.1497,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.388960094817284,
+      "learning_rate": 1.0540056797528817e-05,
+      "loss": 1.1458,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.882956121511356,
+      "learning_rate": 1.0537297026869676e-05,
+      "loss": 1.1375,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9147128483337825,
+      "learning_rate": 1.0534537215169106e-05,
+      "loss": 1.1614,
+      "step": 5842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.215619111578358,
+      "learning_rate": 1.0531777362637919e-05,
+      "loss": 1.1443,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.414370089211791,
+      "learning_rate": 1.0529017469486924e-05,
+      "loss": 1.1111,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.897745675049683,
+      "learning_rate": 1.0526257535926936e-05,
+      "loss": 1.1423,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.072384298094528,
+      "learning_rate": 1.052349756216877e-05,
+      "loss": 1.1021,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.959406034874345,
+      "learning_rate": 1.0520737548423246e-05,
+      "loss": 1.1255,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9767194064215534,
+      "learning_rate": 1.0517977494901192e-05,
+      "loss": 1.0801,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2949641722619516,
+      "learning_rate": 1.0515217401813435e-05,
+      "loss": 1.1772,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3120412731018987,
+      "learning_rate": 1.05124572693708e-05,
+      "loss": 1.0957,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3070141500492323,
+      "learning_rate": 1.0509697097784122e-05,
+      "loss": 1.105,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.228110842475238,
+      "learning_rate": 1.0506936887264237e-05,
+      "loss": 1.1086,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2487675355685073,
+      "learning_rate": 1.0504176638021982e-05,
+      "loss": 1.0635,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.571674697258889,
+      "learning_rate": 1.0501416350268201e-05,
+      "loss": 1.1802,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0818330061301635,
+      "learning_rate": 1.0498656024213737e-05,
+      "loss": 1.1443,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2863144166421554,
+      "learning_rate": 1.0495895660069437e-05,
+      "loss": 1.1394,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0062606282204936,
+      "learning_rate": 1.0493135258046155e-05,
+      "loss": 1.1272,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0816437067542792,
+      "learning_rate": 1.0490374818354739e-05,
+      "loss": 1.2305,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6269083754696396,
+      "learning_rate": 1.048761434120605e-05,
+      "loss": 1.1599,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.382320211718468,
+      "learning_rate": 1.0484853826810942e-05,
+      "loss": 1.199,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.362794784704931,
+      "learning_rate": 1.048209327538028e-05,
+      "loss": 1.2051,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.965641068125107,
+      "learning_rate": 1.047933268712493e-05,
+      "loss": 1.137,
+      "step": 5862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0824567906849887,
+      "learning_rate": 1.0476572062255758e-05,
+      "loss": 1.1211,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.78804015649223,
+      "learning_rate": 1.0473811400983632e-05,
+      "loss": 1.0999,
+      "step": 5864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.049852990736878,
+      "learning_rate": 1.047105070351943e-05,
+      "loss": 1.082,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6292927176150247,
+      "learning_rate": 1.0468289970074021e-05,
+      "loss": 1.2019,
+      "step": 5866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9865736819865525,
+      "learning_rate": 1.046552920085829e-05,
+      "loss": 1.1125,
+      "step": 5867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7018932668663354,
+      "learning_rate": 1.0462768396083116e-05,
+      "loss": 1.115,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2547213532648582,
+      "learning_rate": 1.0460007555959384e-05,
+      "loss": 1.1572,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.502387621424196,
+      "learning_rate": 1.0457246680697977e-05,
+      "loss": 1.1287,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6462655534302604,
+      "learning_rate": 1.0454485770509785e-05,
+      "loss": 1.1777,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9540732327198373,
+      "learning_rate": 1.045172482560571e-05,
+      "loss": 1.1594,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8930546825509404,
+      "learning_rate": 1.0448963846196634e-05,
+      "loss": 1.1116,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.051951389734587,
+      "learning_rate": 1.044620283249346e-05,
+      "loss": 1.1711,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.940217576823577,
+      "learning_rate": 1.0443441784707088e-05,
+      "loss": 1.2188,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7265562769949834,
+      "learning_rate": 1.0440680703048419e-05,
+      "loss": 1.2332,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.902437009093209,
+      "learning_rate": 1.043791958772836e-05,
+      "loss": 1.0918,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.418539482354335,
+      "learning_rate": 1.043515843895782e-05,
+      "loss": 1.1311,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4282790611868923,
+      "learning_rate": 1.0432397256947702e-05,
+      "loss": 1.1101,
+      "step": 5879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.157516131353638,
+      "learning_rate": 1.0429636041908926e-05,
+      "loss": 1.1343,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4845417074735865,
+      "learning_rate": 1.042687479405241e-05,
+      "loss": 1.083,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3012620299250823,
+      "learning_rate": 1.0424113513589062e-05,
+      "loss": 1.095,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.555168626591358,
+      "learning_rate": 1.0421352200729809e-05,
+      "loss": 1.208,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7661369507959668,
+      "learning_rate": 1.0418590855685573e-05,
+      "loss": 1.1382,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.133936406723885,
+      "learning_rate": 1.041582947866728e-05,
+      "loss": 1.0701,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2843306873980516,
+      "learning_rate": 1.0413068069885854e-05,
+      "loss": 1.0696,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9348711706135493,
+      "learning_rate": 1.0410306629552231e-05,
+      "loss": 1.1238,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.585445990818737,
+      "learning_rate": 1.040754515787734e-05,
+      "loss": 1.1843,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6349719762519213,
+      "learning_rate": 1.0404783655072114e-05,
+      "loss": 1.1182,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7131348441311633,
+      "learning_rate": 1.0402022121347495e-05,
+      "loss": 1.1509,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.563367852180038,
+      "learning_rate": 1.0399260556914421e-05,
+      "loss": 1.1367,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2730640608626236,
+      "learning_rate": 1.0396498961983835e-05,
+      "loss": 1.2183,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8495085749984326,
+      "learning_rate": 1.0393737336766674e-05,
+      "loss": 1.1047,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2203273257550986,
+      "learning_rate": 1.0390975681473894e-05,
+      "loss": 1.1772,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3924731346491876,
+      "learning_rate": 1.0388213996316442e-05,
+      "loss": 1.1157,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.414595934823958,
+      "learning_rate": 1.0385452281505269e-05,
+      "loss": 1.0896,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.064918656847411,
+      "learning_rate": 1.0382690537251324e-05,
+      "loss": 1.0911,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0404299727629844,
+      "learning_rate": 1.0379928763765568e-05,
+      "loss": 1.0774,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7424538557144933,
+      "learning_rate": 1.0377166961258955e-05,
+      "loss": 1.103,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2692295898660166,
+      "learning_rate": 1.0374405129942452e-05,
+      "loss": 1.0962,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.014259222736374,
+      "learning_rate": 1.0371643270027014e-05,
+      "loss": 1.167,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.506495438789405,
+      "learning_rate": 1.0368881381723612e-05,
+      "loss": 1.1296,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.768611361252175,
+      "learning_rate": 1.0366119465243205e-05,
+      "loss": 1.1135,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6878469856263383,
+      "learning_rate": 1.0363357520796768e-05,
+      "loss": 1.1282,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.432256798797433,
+      "learning_rate": 1.0360595548595274e-05,
+      "loss": 1.1238,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.631390649947006,
+      "learning_rate": 1.0357833548849692e-05,
+      "loss": 1.1108,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1743891823111094,
+      "learning_rate": 1.0355071521770994e-05,
+      "loss": 1.0959,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.010858301604957,
+      "learning_rate": 1.0352309467570165e-05,
+      "loss": 1.1699,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1841445178759633,
+      "learning_rate": 1.034954738645818e-05,
+      "loss": 1.1458,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.413485296703893,
+      "learning_rate": 1.0346785278646023e-05,
+      "loss": 1.126,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.091794199830933,
+      "learning_rate": 1.0344023144344673e-05,
+      "loss": 1.1216,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5098829296242755,
+      "learning_rate": 1.0341260983765121e-05,
+      "loss": 1.1331,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.244602956158591,
+      "learning_rate": 1.0338498797118353e-05,
+      "loss": 1.1121,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7899506246391805,
+      "learning_rate": 1.0335736584615357e-05,
+      "loss": 1.1006,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.941863672566997,
+      "learning_rate": 1.0332974346467128e-05,
+      "loss": 1.1421,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8704854204777854,
+      "learning_rate": 1.0330212082884655e-05,
+      "loss": 1.1418,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7025557957136437,
+      "learning_rate": 1.0327449794078932e-05,
+      "loss": 1.0967,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.69673013140533,
+      "learning_rate": 1.0324687480260965e-05,
+      "loss": 1.0579,
+      "step": 5918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.511272066536896,
+      "learning_rate": 1.0321925141641746e-05,
+      "loss": 1.1316,
+      "step": 5919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.894879514471321,
+      "learning_rate": 1.031916277843228e-05,
+      "loss": 1.1633,
+      "step": 5920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4043630868408785,
+      "learning_rate": 1.0316400390843567e-05,
+      "loss": 1.1467,
+      "step": 5921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.09493439205804,
+      "learning_rate": 1.0313637979086616e-05,
+      "loss": 1.1118,
+      "step": 5922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.638897767489556,
+      "learning_rate": 1.0310875543372425e-05,
+      "loss": 1.095,
+      "step": 5923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6525535652001646,
+      "learning_rate": 1.0308113083912015e-05,
+      "loss": 1.1621,
+      "step": 5924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.26232160826091,
+      "learning_rate": 1.030535060091639e-05,
+      "loss": 1.1978,
+      "step": 5925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.724771780106253,
+      "learning_rate": 1.0302588094596559e-05,
+      "loss": 1.1418,
+      "step": 5926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5345877198143048,
+      "learning_rate": 1.0299825565163541e-05,
+      "loss": 1.1506,
+      "step": 5927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.36697740947663,
+      "learning_rate": 1.0297063012828352e-05,
+      "loss": 1.1135,
+      "step": 5928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.016646833950704,
+      "learning_rate": 1.029430043780201e-05,
+      "loss": 1.1118,
+      "step": 5929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5080805219625253,
+      "learning_rate": 1.0291537840295526e-05,
+      "loss": 1.1199,
+      "step": 5930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.403530273364385,
+      "learning_rate": 1.028877522051993e-05,
+      "loss": 1.1062,
+      "step": 5931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.790667021356135,
+      "learning_rate": 1.0286012578686244e-05,
+      "loss": 1.146,
+      "step": 5932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.579944414099122,
+      "learning_rate": 1.028324991500549e-05,
+      "loss": 1.1646,
+      "step": 5933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2622979744834875,
+      "learning_rate": 1.0280487229688691e-05,
+      "loss": 1.1335,
+      "step": 5934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.896568113636642,
+      "learning_rate": 1.0277724522946882e-05,
+      "loss": 1.1394,
+      "step": 5935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5626050836507734,
+      "learning_rate": 1.0274961794991084e-05,
+      "loss": 1.1091,
+      "step": 5936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1633858755663367,
+      "learning_rate": 1.0272199046032334e-05,
+      "loss": 1.1602,
+      "step": 5937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.733860351626042,
+      "learning_rate": 1.0269436276281665e-05,
+      "loss": 1.1575,
+      "step": 5938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0766099060875773,
+      "learning_rate": 1.0266673485950107e-05,
+      "loss": 1.1646,
+      "step": 5939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.447819846370475,
+      "learning_rate": 1.0263910675248696e-05,
+      "loss": 1.1255,
+      "step": 5940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5851136100772476,
+      "learning_rate": 1.0261147844388472e-05,
+      "loss": 1.1133,
+      "step": 5941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.569546946165833,
+      "learning_rate": 1.0258384993580472e-05,
+      "loss": 1.1631,
+      "step": 5942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7123259923861167,
+      "learning_rate": 1.025562212303574e-05,
+      "loss": 1.0742,
+      "step": 5943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9492301990432064,
+      "learning_rate": 1.025285923296531e-05,
+      "loss": 1.0945,
+      "step": 5944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.472367874525227,
+      "learning_rate": 1.0250096323580236e-05,
+      "loss": 1.1978,
+      "step": 5945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.146771447768366,
+      "learning_rate": 1.0247333395091552e-05,
+      "loss": 1.1433,
+      "step": 5946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.393011521567856,
+      "learning_rate": 1.0244570447710314e-05,
+      "loss": 1.0618,
+      "step": 5947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.48696629324417,
+      "learning_rate": 1.0241807481647564e-05,
+      "loss": 1.1511,
+      "step": 5948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3075556854232513,
+      "learning_rate": 1.0239044497114348e-05,
+      "loss": 1.1335,
+      "step": 5949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7524163216774373,
+      "learning_rate": 1.023628149432172e-05,
+      "loss": 1.1514,
+      "step": 5950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1834565090659193,
+      "learning_rate": 1.0233518473480737e-05,
+      "loss": 1.1611,
+      "step": 5951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.682394178385891,
+      "learning_rate": 1.0230755434802448e-05,
+      "loss": 1.0996,
+      "step": 5952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2777367880900656,
+      "learning_rate": 1.0227992378497905e-05,
+      "loss": 1.1387,
+      "step": 5953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.240541861769982,
+      "learning_rate": 1.0225229304778166e-05,
+      "loss": 1.1665,
+      "step": 5954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7125385374339226,
+      "learning_rate": 1.0222466213854292e-05,
+      "loss": 1.1228,
+      "step": 5955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3921074065773342,
+      "learning_rate": 1.0219703105937335e-05,
+      "loss": 1.1565,
+      "step": 5956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.945091331651594,
+      "learning_rate": 1.0216939981238362e-05,
+      "loss": 1.1353,
+      "step": 5957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0045169219207994,
+      "learning_rate": 1.021417683996843e-05,
+      "loss": 1.1211,
+      "step": 5958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2067294674307965,
+      "learning_rate": 1.0211413682338603e-05,
+      "loss": 1.1514,
+      "step": 5959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5740397375055863,
+      "learning_rate": 1.0208650508559941e-05,
+      "loss": 1.2361,
+      "step": 5960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3779169084054126,
+      "learning_rate": 1.0205887318843514e-05,
+      "loss": 1.1194,
+      "step": 5961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.629254991181197,
+      "learning_rate": 1.020312411340039e-05,
+      "loss": 1.1016,
+      "step": 5962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2856119767661194,
+      "learning_rate": 1.0200360892441625e-05,
+      "loss": 1.1304,
+      "step": 5963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.489515551614844,
+      "learning_rate": 1.01975976561783e-05,
+      "loss": 1.1377,
+      "step": 5964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1071886892838223,
+      "learning_rate": 1.0194834404821481e-05,
+      "loss": 1.0925,
+      "step": 5965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0747179587298596,
+      "learning_rate": 1.0192071138582238e-05,
+      "loss": 1.1736,
+      "step": 5966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9973332484588853,
+      "learning_rate": 1.018930785767164e-05,
+      "loss": 1.0916,
+      "step": 5967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5872040023066054,
+      "learning_rate": 1.0186544562300766e-05,
+      "loss": 1.1646,
+      "step": 5968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7819491589889744,
+      "learning_rate": 1.0183781252680684e-05,
+      "loss": 1.116,
+      "step": 5969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.657330376911377,
+      "learning_rate": 1.0181017929022478e-05,
+      "loss": 1.0698,
+      "step": 5970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5541983960928047,
+      "learning_rate": 1.0178254591537219e-05,
+      "loss": 1.2397,
+      "step": 5971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6975944736889517,
+      "learning_rate": 1.0175491240435982e-05,
+      "loss": 1.1196,
+      "step": 5972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.338192060920903,
+      "learning_rate": 1.0172727875929846e-05,
+      "loss": 1.071,
+      "step": 5973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.602726920002898,
+      "learning_rate": 1.0169964498229896e-05,
+      "loss": 1.1467,
+      "step": 5974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.278919372405656,
+      "learning_rate": 1.0167201107547212e-05,
+      "loss": 1.1174,
+      "step": 5975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4628526281402516,
+      "learning_rate": 1.0164437704092873e-05,
+      "loss": 1.106,
+      "step": 5976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3052799401740294,
+      "learning_rate": 1.0161674288077959e-05,
+      "loss": 1.1292,
+      "step": 5977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9148323155094253,
+      "learning_rate": 1.0158910859713558e-05,
+      "loss": 1.0881,
+      "step": 5978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.934945674936394,
+      "learning_rate": 1.015614741921075e-05,
+      "loss": 1.1338,
+      "step": 5979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9531421496679897,
+      "learning_rate": 1.0153383966780626e-05,
+      "loss": 1.1392,
+      "step": 5980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2505299332138913,
+      "learning_rate": 1.015062050263427e-05,
+      "loss": 1.127,
+      "step": 5981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1746114500711857,
+      "learning_rate": 1.0147857026982765e-05,
+      "loss": 1.1436,
+      "step": 5982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.526229711510645,
+      "learning_rate": 1.0145093540037204e-05,
+      "loss": 1.1885,
+      "step": 5983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8408789990901306,
+      "learning_rate": 1.0142330042008675e-05,
+      "loss": 1.2097,
+      "step": 5984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0455769747319965,
+      "learning_rate": 1.0139566533108267e-05,
+      "loss": 1.0664,
+      "step": 5985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8809065190179353,
+      "learning_rate": 1.0136803013547071e-05,
+      "loss": 1.1262,
+      "step": 5986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9662421814819098,
+      "learning_rate": 1.0134039483536178e-05,
+      "loss": 1.0969,
+      "step": 5987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.015255213196737,
+      "learning_rate": 1.013127594328668e-05,
+      "loss": 1.1099,
+      "step": 5988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5359895046885748,
+      "learning_rate": 1.0128512393009673e-05,
+      "loss": 1.0977,
+      "step": 5989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4699987904074767,
+      "learning_rate": 1.0125748832916245e-05,
+      "loss": 1.1428,
+      "step": 5990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0523679283613405,
+      "learning_rate": 1.0122985263217496e-05,
+      "loss": 1.1062,
+      "step": 5991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.735089220857523,
+      "learning_rate": 1.0120221684124516e-05,
+      "loss": 1.2048,
+      "step": 5992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.588833771738409,
+      "learning_rate": 1.0117458095848405e-05,
+      "loss": 1.0837,
+      "step": 5993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7952947141306463,
+      "learning_rate": 1.0114694498600261e-05,
+      "loss": 1.1101,
+      "step": 5994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.164157935783559,
+      "learning_rate": 1.011193089259118e-05,
+      "loss": 1.1467,
+      "step": 5995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2685224300977653,
+      "learning_rate": 1.0109167278032254e-05,
+      "loss": 1.1621,
+      "step": 5996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0801798702245624,
+      "learning_rate": 1.0106403655134589e-05,
+      "loss": 1.1411,
+      "step": 5997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5954215879574654,
+      "learning_rate": 1.0103640024109282e-05,
+      "loss": 1.0972,
+      "step": 5998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.18806363785506,
+      "learning_rate": 1.0100876385167433e-05,
+      "loss": 1.1382,
+      "step": 5999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1250787301874863,
+      "learning_rate": 1.0098112738520143e-05,
+      "loss": 1.1572,
+      "step": 6000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7045977831713075,
+      "learning_rate": 1.0095349084378512e-05,
+      "loss": 1.1448,
+      "step": 6001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4532150734929674,
+      "learning_rate": 1.009258542295364e-05,
+      "loss": 1.0479,
+      "step": 6002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1542456126697873,
+      "learning_rate": 1.0089821754456634e-05,
+      "loss": 1.2048,
+      "step": 6003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2085620501514853,
+      "learning_rate": 1.0087058079098595e-05,
+      "loss": 1.1365,
+      "step": 6004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.116160867851255,
+      "learning_rate": 1.0084294397090624e-05,
+      "loss": 1.1536,
+      "step": 6005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3984810922967403,
+      "learning_rate": 1.0081530708643825e-05,
+      "loss": 1.1602,
+      "step": 6006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.852616557688329,
+      "learning_rate": 1.0078767013969303e-05,
+      "loss": 1.1013,
+      "step": 6007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.733201679362873,
+      "learning_rate": 1.0076003313278171e-05,
+      "loss": 1.1377,
+      "step": 6008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6522177645174834,
+      "learning_rate": 1.007323960678152e-05,
+      "loss": 1.136,
+      "step": 6009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.48349587659605,
+      "learning_rate": 1.0070475894690463e-05,
+      "loss": 1.0593,
+      "step": 6010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.056162270235615,
+      "learning_rate": 1.0067712177216108e-05,
+      "loss": 1.137,
+      "step": 6011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4754522810519064,
+      "learning_rate": 1.0064948454569556e-05,
+      "loss": 1.1604,
+      "step": 6012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.27238423426345,
+      "learning_rate": 1.0062184726961918e-05,
+      "loss": 1.1958,
+      "step": 6013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8586251701051606,
+      "learning_rate": 1.0059420994604303e-05,
+      "loss": 1.145,
+      "step": 6014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.839235669280263,
+      "learning_rate": 1.0056657257707811e-05,
+      "loss": 1.1272,
+      "step": 6015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.87631074990938,
+      "learning_rate": 1.0053893516483557e-05,
+      "loss": 1.084,
+      "step": 6016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.512638909772281,
+      "learning_rate": 1.005112977114265e-05,
+      "loss": 1.0793,
+      "step": 6017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.583750801115026,
+      "learning_rate": 1.0048366021896194e-05,
+      "loss": 1.175,
+      "step": 6018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0743034116655097,
+      "learning_rate": 1.0045602268955298e-05,
+      "loss": 1.1392,
+      "step": 6019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5739425319240827,
+      "learning_rate": 1.0042838512531075e-05,
+      "loss": 1.1152,
+      "step": 6020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2704363932888856,
+      "learning_rate": 1.0040074752834635e-05,
+      "loss": 1.1091,
+      "step": 6021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1412230516487236,
+      "learning_rate": 1.0037310990077083e-05,
+      "loss": 1.1086,
+      "step": 6022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9740897868857017,
+      "learning_rate": 1.003454722446953e-05,
+      "loss": 1.0515,
+      "step": 6023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.1621534504295346,
+      "learning_rate": 1.0031783456223092e-05,
+      "loss": 1.1738,
+      "step": 6024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.89739593508361,
+      "learning_rate": 1.002901968554887e-05,
+      "loss": 1.1448,
+      "step": 6025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5448203150464117,
+      "learning_rate": 1.0026255912657981e-05,
+      "loss": 1.1726,
+      "step": 6026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.378088601496991,
+      "learning_rate": 1.0023492137761536e-05,
+      "loss": 1.1826,
+      "step": 6027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.465346863964706,
+      "learning_rate": 1.0020728361070644e-05,
+      "loss": 1.0583,
+      "step": 6028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9367804301443936,
+      "learning_rate": 1.0017964582796416e-05,
+      "loss": 1.1736,
+      "step": 6029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.717800770514878,
+      "learning_rate": 1.0015200803149963e-05,
+      "loss": 1.1519,
+      "step": 6030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.245097111258938,
+      "learning_rate": 1.0012437022342399e-05,
+      "loss": 1.1194,
+      "step": 6031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.256271902523738,
+      "learning_rate": 1.0009673240584833e-05,
+      "loss": 1.1475,
+      "step": 6032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5400198977379143,
+      "learning_rate": 1.0006909458088373e-05,
+      "loss": 1.1785,
+      "step": 6033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.152787509387165,
+      "learning_rate": 1.0004145675064137e-05,
+      "loss": 1.144,
+      "step": 6034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2482765664055595,
+      "learning_rate": 1.0001381891723232e-05,
+      "loss": 1.1987,
+      "step": 6035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.069878950134302,
+      "learning_rate": 9.998618108276773e-06,
+      "loss": 1.1848,
+      "step": 6036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.181007276321822,
+      "learning_rate": 9.995854324935867e-06,
+      "loss": 1.1394,
+      "step": 6037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.951360937244067,
+      "learning_rate": 9.993090541911629e-06,
+      "loss": 1.1184,
+      "step": 6038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1969336634193857,
+      "learning_rate": 9.990326759415174e-06,
+      "loss": 1.0925,
+      "step": 6039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.652892994891432,
+      "learning_rate": 9.987562977657605e-06,
+      "loss": 1.0984,
+      "step": 6040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.143796159583604,
+      "learning_rate": 9.98479919685004e-06,
+      "loss": 1.1191,
+      "step": 6041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.156580810315466,
+      "learning_rate": 9.982035417203587e-06,
+      "loss": 1.1511,
+      "step": 6042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2626949489847923,
+      "learning_rate": 9.979271638929358e-06,
+      "loss": 1.0806,
+      "step": 6043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0427827571652957,
+      "learning_rate": 9.976507862238464e-06,
+      "loss": 1.1211,
+      "step": 6044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.37026620431507,
+      "learning_rate": 9.97374408734202e-06,
+      "loss": 1.115,
+      "step": 6045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.248813216672396,
+      "learning_rate": 9.970980314451136e-06,
+      "loss": 1.2241,
+      "step": 6046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7576668692760005,
+      "learning_rate": 9.968216543776915e-06,
+      "loss": 1.2109,
+      "step": 6047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2775661198551855,
+      "learning_rate": 9.965452775530471e-06,
+      "loss": 1.1438,
+      "step": 6048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.186576217806973,
+      "learning_rate": 9.962689009922918e-06,
+      "loss": 1.0891,
+      "step": 6049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1254157764289094,
+      "learning_rate": 9.959925247165369e-06,
+      "loss": 1.113,
+      "step": 6050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1690204771617383,
+      "learning_rate": 9.957161487468925e-06,
+      "loss": 1.0742,
+      "step": 6051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1477255789863547,
+      "learning_rate": 9.954397731044702e-06,
+      "loss": 1.1282,
+      "step": 6052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1682539253255415,
+      "learning_rate": 9.951633978103811e-06,
+      "loss": 1.1699,
+      "step": 6053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.547893040232794,
+      "learning_rate": 9.948870228857354e-06,
+      "loss": 1.0476,
+      "step": 6054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1044619852164668,
+      "learning_rate": 9.946106483516446e-06,
+      "loss": 1.145,
+      "step": 6055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5735704540189555,
+      "learning_rate": 9.94334274229219e-06,
+      "loss": 1.1177,
+      "step": 6056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.871014412364404,
+      "learning_rate": 9.9405790053957e-06,
+      "loss": 1.2134,
+      "step": 6057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8876775729434154,
+      "learning_rate": 9.937815273038082e-06,
+      "loss": 1.0806,
+      "step": 6058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.446536996787289,
+      "learning_rate": 9.935051545430448e-06,
+      "loss": 1.0723,
+      "step": 6059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7275190945081165,
+      "learning_rate": 9.932287822783897e-06,
+      "loss": 1.1948,
+      "step": 6060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2463782994269272,
+      "learning_rate": 9.92952410530954e-06,
+      "loss": 1.1189,
+      "step": 6061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.181262500082379,
+      "learning_rate": 9.926760393218483e-06,
+      "loss": 1.1152,
+      "step": 6062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0161599595180375,
+      "learning_rate": 9.923996686721832e-06,
+      "loss": 1.116,
+      "step": 6063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.425611945304736,
+      "learning_rate": 9.921232986030695e-06,
+      "loss": 1.0789,
+      "step": 6064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.344292304813642,
+      "learning_rate": 9.918469291356174e-06,
+      "loss": 1.1418,
+      "step": 6065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.001805078613504,
+      "learning_rate": 9.915705602909381e-06,
+      "loss": 1.1409,
+      "step": 6066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0260044043866747,
+      "learning_rate": 9.912941920901408e-06,
+      "loss": 1.0457,
+      "step": 6067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.165394007916628,
+      "learning_rate": 9.910178245543368e-06,
+      "loss": 1.0823,
+      "step": 6068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2512061339393736,
+      "learning_rate": 9.907414577046363e-06,
+      "loss": 1.1672,
+      "step": 6069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6749988364472057,
+      "learning_rate": 9.904650915621491e-06,
+      "loss": 1.1731,
+      "step": 6070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1676700098554518,
+      "learning_rate": 9.901887261479859e-06,
+      "loss": 1.1284,
+      "step": 6071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2612904812847954,
+      "learning_rate": 9.89912361483257e-06,
+      "loss": 1.1003,
+      "step": 6072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0883677071530387,
+      "learning_rate": 9.89635997589072e-06,
+      "loss": 1.1101,
+      "step": 6073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8618287634733273,
+      "learning_rate": 9.893596344865414e-06,
+      "loss": 1.1008,
+      "step": 6074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.23132117249993,
+      "learning_rate": 9.89083272196775e-06,
+      "loss": 1.2119,
+      "step": 6075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.163308961348384,
+      "learning_rate": 9.888069107408824e-06,
+      "loss": 1.0781,
+      "step": 6076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3200477984044414,
+      "learning_rate": 9.885305501399739e-06,
+      "loss": 1.1616,
+      "step": 6077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3200952886609523,
+      "learning_rate": 9.882541904151595e-06,
+      "loss": 1.1812,
+      "step": 6078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2996609440022215,
+      "learning_rate": 9.879778315875487e-06,
+      "loss": 1.1011,
+      "step": 6079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.997953894505998,
+      "learning_rate": 9.877014736782509e-06,
+      "loss": 1.0808,
+      "step": 6080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.501318595849835,
+      "learning_rate": 9.874251167083758e-06,
+      "loss": 1.0662,
+      "step": 6081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.023579368355538,
+      "learning_rate": 9.87148760699033e-06,
+      "loss": 1.1885,
+      "step": 6082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.299831617894032,
+      "learning_rate": 9.868724056713323e-06,
+      "loss": 1.1152,
+      "step": 6083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.964585827598527,
+      "learning_rate": 9.865960516463824e-06,
+      "loss": 1.1406,
+      "step": 6084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9826047677474774,
+      "learning_rate": 9.86319698645293e-06,
+      "loss": 1.0811,
+      "step": 6085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4018400228047594,
+      "learning_rate": 9.860433466891737e-06,
+      "loss": 1.1216,
+      "step": 6086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9502931517540696,
+      "learning_rate": 9.857669957991328e-06,
+      "loss": 1.1274,
+      "step": 6087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6854421385368603,
+      "learning_rate": 9.8549064599628e-06,
+      "loss": 1.1079,
+      "step": 6088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2960388833957794,
+      "learning_rate": 9.852142973017238e-06,
+      "loss": 1.188,
+      "step": 6089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2769563710038563,
+      "learning_rate": 9.849379497365732e-06,
+      "loss": 1.1211,
+      "step": 6090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.858794453220923,
+      "learning_rate": 9.846616033219377e-06,
+      "loss": 1.1067,
+      "step": 6091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8724531744917368,
+      "learning_rate": 9.843852580789253e-06,
+      "loss": 1.145,
+      "step": 6092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8135374304110194,
+      "learning_rate": 9.841089140286447e-06,
+      "loss": 1.0565,
+      "step": 6093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.8458891517298355,
+      "learning_rate": 9.838325711922045e-06,
+      "loss": 1.093,
+      "step": 6094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4446068765461795,
+      "learning_rate": 9.83556229590713e-06,
+      "loss": 1.0671,
+      "step": 6095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.2637631486934175,
+      "learning_rate": 9.83279889245279e-06,
+      "loss": 1.1851,
+      "step": 6096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.004462674194554,
+      "learning_rate": 9.830035501770104e-06,
+      "loss": 1.0923,
+      "step": 6097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3112725572321295,
+      "learning_rate": 9.827272124070152e-06,
+      "loss": 1.116,
+      "step": 6098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.784424094902694,
+      "learning_rate": 9.824508759564025e-06,
+      "loss": 1.2227,
+      "step": 6099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5528173831482444,
+      "learning_rate": 9.821745408462787e-06,
+      "loss": 1.064,
+      "step": 6100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.835265539531603,
+      "learning_rate": 9.818982070977525e-06,
+      "loss": 1.1624,
+      "step": 6101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.992908690459855,
+      "learning_rate": 9.816218747319318e-06,
+      "loss": 1.1099,
+      "step": 6102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.017793771606015,
+      "learning_rate": 9.813455437699238e-06,
+      "loss": 1.0286,
+      "step": 6103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.426266610018064,
+      "learning_rate": 9.810692142328361e-06,
+      "loss": 1.0916,
+      "step": 6104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.448167529981849,
+      "learning_rate": 9.807928861417765e-06,
+      "loss": 1.1475,
+      "step": 6105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2126953302614525,
+      "learning_rate": 9.805165595178524e-06,
+      "loss": 1.2202,
+      "step": 6106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.8063348035777915,
+      "learning_rate": 9.802402343821703e-06,
+      "loss": 1.1145,
+      "step": 6107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.648697269236485,
+      "learning_rate": 9.799639107558377e-06,
+      "loss": 1.1475,
+      "step": 6108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.14789733685685,
+      "learning_rate": 9.796875886599616e-06,
+      "loss": 1.1172,
+      "step": 6109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.558010086324301,
+      "learning_rate": 9.794112681156487e-06,
+      "loss": 1.126,
+      "step": 6110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8575616262631107,
+      "learning_rate": 9.79134949144006e-06,
+      "loss": 1.1775,
+      "step": 6111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6740816022657166,
+      "learning_rate": 9.788586317661404e-06,
+      "loss": 1.0637,
+      "step": 6112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.27910770335819,
+      "learning_rate": 9.785823160031575e-06,
+      "loss": 1.0903,
+      "step": 6113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3449988670602906,
+      "learning_rate": 9.783060018761641e-06,
+      "loss": 1.0723,
+      "step": 6114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.387095154345381,
+      "learning_rate": 9.780296894062667e-06,
+      "loss": 1.1052,
+      "step": 6115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9854747803253754,
+      "learning_rate": 9.777533786145711e-06,
+      "loss": 1.0881,
+      "step": 6116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7584211173610016,
+      "learning_rate": 9.774770695221834e-06,
+      "loss": 1.1201,
+      "step": 6117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8414963616591113,
+      "learning_rate": 9.772007621502095e-06,
+      "loss": 1.1431,
+      "step": 6118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.080786210210445,
+      "learning_rate": 9.769244565197557e-06,
+      "loss": 1.1318,
+      "step": 6119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.262917217702631,
+      "learning_rate": 9.766481526519265e-06,
+      "loss": 1.1721,
+      "step": 6120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0980255800022034,
+      "learning_rate": 9.763718505678282e-06,
+      "loss": 1.1189,
+      "step": 6121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.079654083259358,
+      "learning_rate": 9.760955502885655e-06,
+      "loss": 1.1196,
+      "step": 6122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.867636190664336,
+      "learning_rate": 9.758192518352441e-06,
+      "loss": 1.114,
+      "step": 6123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.810608852543464,
+      "learning_rate": 9.75542955228969e-06,
+      "loss": 1.1013,
+      "step": 6124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5114610225894953,
+      "learning_rate": 9.752666604908448e-06,
+      "loss": 1.1602,
+      "step": 6125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.655942114173675,
+      "learning_rate": 9.749903676419769e-06,
+      "loss": 1.0969,
+      "step": 6126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4359177430038286,
+      "learning_rate": 9.747140767034692e-06,
+      "loss": 1.0669,
+      "step": 6127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.996449965037114,
+      "learning_rate": 9.744377876964262e-06,
+      "loss": 1.1279,
+      "step": 6128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.79186472138132,
+      "learning_rate": 9.74161500641953e-06,
+      "loss": 1.1221,
+      "step": 6129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2830778686088142,
+      "learning_rate": 9.738852155611531e-06,
+      "loss": 1.124,
+      "step": 6130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7224036440750234,
+      "learning_rate": 9.736089324751305e-06,
+      "loss": 1.1133,
+      "step": 6131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3526510255509505,
+      "learning_rate": 9.7333265140499e-06,
+      "loss": 1.1653,
+      "step": 6132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.044609830721745,
+      "learning_rate": 9.730563723718338e-06,
+      "loss": 1.0781,
+      "step": 6133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0901678659222984,
+      "learning_rate": 9.72780095396767e-06,
+      "loss": 1.1416,
+      "step": 6134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0982205242579646,
+      "learning_rate": 9.725038205008919e-06,
+      "loss": 1.1592,
+      "step": 6135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.77684132475292,
+      "learning_rate": 9.722275477053122e-06,
+      "loss": 1.1321,
+      "step": 6136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.256168457645561,
+      "learning_rate": 9.719512770311309e-06,
+      "loss": 1.1311,
+      "step": 6137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.818094729682956,
+      "learning_rate": 9.716750084994513e-06,
+      "loss": 1.0964,
+      "step": 6138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.945683332486009,
+      "learning_rate": 9.71398742131376e-06,
+      "loss": 1.1711,
+      "step": 6139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.189153017674745,
+      "learning_rate": 9.711224779480073e-06,
+      "loss": 1.1423,
+      "step": 6140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8990967349000116,
+      "learning_rate": 9.708462159704476e-06,
+      "loss": 1.0979,
+      "step": 6141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7937418513685786,
+      "learning_rate": 9.705699562197993e-06,
+      "loss": 1.2007,
+      "step": 6142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.018487216139243,
+      "learning_rate": 9.70293698717165e-06,
+      "loss": 1.188,
+      "step": 6143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3880918774613478,
+      "learning_rate": 9.700174434836459e-06,
+      "loss": 1.1533,
+      "step": 6144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9225185173794253,
+      "learning_rate": 9.697411905403444e-06,
+      "loss": 1.0327,
+      "step": 6145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7428880162541627,
+      "learning_rate": 9.694649399083616e-06,
+      "loss": 1.1306,
+      "step": 6146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0725946857961293,
+      "learning_rate": 9.691886916087987e-06,
+      "loss": 1.0833,
+      "step": 6147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4757292165428346,
+      "learning_rate": 9.689124456627577e-06,
+      "loss": 1.0718,
+      "step": 6148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.179153823616328,
+      "learning_rate": 9.68636202091339e-06,
+      "loss": 1.1078,
+      "step": 6149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3407451311186374,
+      "learning_rate": 9.683599609156435e-06,
+      "loss": 1.1978,
+      "step": 6150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0871118276573775,
+      "learning_rate": 9.680837221567722e-06,
+      "loss": 1.0916,
+      "step": 6151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.198651840386062,
+      "learning_rate": 9.678074858358257e-06,
+      "loss": 1.0767,
+      "step": 6152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8105295270539483,
+      "learning_rate": 9.67531251973904e-06,
+      "loss": 1.0503,
+      "step": 6153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.700435650323353,
+      "learning_rate": 9.672550205921071e-06,
+      "loss": 1.1284,
+      "step": 6154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9946699722760926,
+      "learning_rate": 9.66978791711535e-06,
+      "loss": 1.1194,
+      "step": 6155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6754913034565266,
+      "learning_rate": 9.667025653532876e-06,
+      "loss": 1.106,
+      "step": 6156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3120737189510563,
+      "learning_rate": 9.664263415384644e-06,
+      "loss": 1.104,
+      "step": 6157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1824647602697893,
+      "learning_rate": 9.661501202881648e-06,
+      "loss": 1.1865,
+      "step": 6158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.930668006444812,
+      "learning_rate": 9.658739016234882e-06,
+      "loss": 1.1809,
+      "step": 6159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0126882997397195,
+      "learning_rate": 9.655976855655329e-06,
+      "loss": 1.0867,
+      "step": 6160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.225469626202778,
+      "learning_rate": 9.65321472135398e-06,
+      "loss": 1.0781,
+      "step": 6161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7341118624086436,
+      "learning_rate": 9.650452613541823e-06,
+      "loss": 1.1189,
+      "step": 6162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.246793569014898,
+      "learning_rate": 9.647690532429837e-06,
+      "loss": 1.0999,
+      "step": 6163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4235286376974416,
+      "learning_rate": 9.644928478229006e-06,
+      "loss": 1.1108,
+      "step": 6164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9398139307616846,
+      "learning_rate": 9.642166451150315e-06,
+      "loss": 1.0774,
+      "step": 6165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8770046219900816,
+      "learning_rate": 9.639404451404729e-06,
+      "loss": 1.0664,
+      "step": 6166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0725754368234943,
+      "learning_rate": 9.636642479203234e-06,
+      "loss": 1.1235,
+      "step": 6167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2393076970885626,
+      "learning_rate": 9.633880534756797e-06,
+      "loss": 1.1272,
+      "step": 6168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4079706749441256,
+      "learning_rate": 9.631118618276391e-06,
+      "loss": 1.1108,
+      "step": 6169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.233776586576419,
+      "learning_rate": 9.628356729972986e-06,
+      "loss": 1.0664,
+      "step": 6170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1378988557043015,
+      "learning_rate": 9.62559487005755e-06,
+      "loss": 1.1172,
+      "step": 6171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.004845674608147,
+      "learning_rate": 9.622833038741048e-06,
+      "loss": 1.1382,
+      "step": 6172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2293372448704694,
+      "learning_rate": 9.620071236234437e-06,
+      "loss": 1.1091,
+      "step": 6173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4147599592523132,
+      "learning_rate": 9.61730946274868e-06,
+      "loss": 1.0891,
+      "step": 6174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6625652573707064,
+      "learning_rate": 9.614547718494736e-06,
+      "loss": 1.1541,
+      "step": 6175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1027240204995516,
+      "learning_rate": 9.61178600368356e-06,
+      "loss": 1.1277,
+      "step": 6176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.200923605444803,
+      "learning_rate": 9.609024318526108e-06,
+      "loss": 1.114,
+      "step": 6177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.409838809699512,
+      "learning_rate": 9.606262663233326e-06,
+      "loss": 1.1306,
+      "step": 6178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0208297937745154,
+      "learning_rate": 9.603501038016172e-06,
+      "loss": 1.1008,
+      "step": 6179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2097173111786708,
+      "learning_rate": 9.600739443085582e-06,
+      "loss": 1.1497,
+      "step": 6180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.903921406242991,
+      "learning_rate": 9.597977878652508e-06,
+      "loss": 1.1265,
+      "step": 6181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4580449176651618,
+      "learning_rate": 9.595216344927887e-06,
+      "loss": 1.1787,
+      "step": 6182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.737072396541516,
+      "learning_rate": 9.592454842122663e-06,
+      "loss": 1.208,
+      "step": 6183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8475260413715824,
+      "learning_rate": 9.589693370447769e-06,
+      "loss": 1.0649,
+      "step": 6184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.355762135937736,
+      "learning_rate": 9.586931930114148e-06,
+      "loss": 1.157,
+      "step": 6185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.242689282584922,
+      "learning_rate": 9.584170521332724e-06,
+      "loss": 1.1138,
+      "step": 6186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2864620471760566,
+      "learning_rate": 9.58140914431443e-06,
+      "loss": 1.1262,
+      "step": 6187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1724426683225024,
+      "learning_rate": 9.578647799270193e-06,
+      "loss": 1.1099,
+      "step": 6188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.223399456722003,
+      "learning_rate": 9.575886486410941e-06,
+      "loss": 1.1609,
+      "step": 6189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9495215749953245,
+      "learning_rate": 9.573125205947594e-06,
+      "loss": 1.0872,
+      "step": 6190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.5336975565464455,
+      "learning_rate": 9.570363958091075e-06,
+      "loss": 1.1604,
+      "step": 6191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5210168356754656,
+      "learning_rate": 9.567602743052303e-06,
+      "loss": 1.1716,
+      "step": 6192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.390038778911411,
+      "learning_rate": 9.564841561042187e-06,
+      "loss": 1.1292,
+      "step": 6193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.365862123767257,
+      "learning_rate": 9.562080412271643e-06,
+      "loss": 1.123,
+      "step": 6194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.112324120311557,
+      "learning_rate": 9.559319296951585e-06,
+      "loss": 1.0615,
+      "step": 6195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5199680674512535,
+      "learning_rate": 9.556558215292915e-06,
+      "loss": 1.115,
+      "step": 6196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6521724540625384,
+      "learning_rate": 9.55379716750654e-06,
+      "loss": 1.0652,
+      "step": 6197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0044019971947273,
+      "learning_rate": 9.55103615380337e-06,
+      "loss": 1.0762,
+      "step": 6198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1497751986279363,
+      "learning_rate": 9.548275174394295e-06,
+      "loss": 1.1021,
+      "step": 6199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9083281977538937,
+      "learning_rate": 9.545514229490216e-06,
+      "loss": 1.104,
+      "step": 6200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3911506117983,
+      "learning_rate": 9.542753319302027e-06,
+      "loss": 1.0806,
+      "step": 6201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6332410332102016,
+      "learning_rate": 9.53999244404062e-06,
+      "loss": 1.1946,
+      "step": 6202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3807341135357416,
+      "learning_rate": 9.537231603916886e-06,
+      "loss": 1.0903,
+      "step": 6203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9258447186477956,
+      "learning_rate": 9.534470799141712e-06,
+      "loss": 1.1409,
+      "step": 6204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.771832924387317,
+      "learning_rate": 9.531710029925985e-06,
+      "loss": 1.1775,
+      "step": 6205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.562505016263245,
+      "learning_rate": 9.528949296480576e-06,
+      "loss": 1.158,
+      "step": 6206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.506655437176435,
+      "learning_rate": 9.526188599016372e-06,
+      "loss": 1.1753,
+      "step": 6207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8601714567862815,
+      "learning_rate": 9.523427937744245e-06,
+      "loss": 1.0918,
+      "step": 6208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.235278714964999,
+      "learning_rate": 9.520667312875073e-06,
+      "loss": 1.0317,
+      "step": 6209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.41009670314274,
+      "learning_rate": 9.517906724619721e-06,
+      "loss": 1.1133,
+      "step": 6210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3445045196312213,
+      "learning_rate": 9.515146173189058e-06,
+      "loss": 1.1091,
+      "step": 6211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.477733969348291,
+      "learning_rate": 9.512385658793955e-06,
+      "loss": 1.1177,
+      "step": 6212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3437450170201313,
+      "learning_rate": 9.509625181645263e-06,
+      "loss": 1.1626,
+      "step": 6213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.802008987882726,
+      "learning_rate": 9.50686474195385e-06,
+      "loss": 1.1448,
+      "step": 6214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.509489899335622,
+      "learning_rate": 9.504104339930565e-06,
+      "loss": 1.0688,
+      "step": 6215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4082546915427057,
+      "learning_rate": 9.501343975786264e-06,
+      "loss": 1.1436,
+      "step": 6216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.231866143009184,
+      "learning_rate": 9.498583649731802e-06,
+      "loss": 1.1387,
+      "step": 6217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.310103550881002,
+      "learning_rate": 9.495823361978022e-06,
+      "loss": 1.1538,
+      "step": 6218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.279208821139861,
+      "learning_rate": 9.493063112735768e-06,
+      "loss": 1.048,
+      "step": 6219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2503812241369605,
+      "learning_rate": 9.490302902215882e-06,
+      "loss": 1.0952,
+      "step": 6220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.070770859781055,
+      "learning_rate": 9.487542730629204e-06,
+      "loss": 1.1396,
+      "step": 6221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.272431874281332,
+      "learning_rate": 9.484782598186567e-06,
+      "loss": 1.1592,
+      "step": 6222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.129681937248002,
+      "learning_rate": 9.48202250509881e-06,
+      "loss": 1.2065,
+      "step": 6223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.190405431765575,
+      "learning_rate": 9.479262451576754e-06,
+      "loss": 1.1985,
+      "step": 6224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.895750897739168,
+      "learning_rate": 9.476502437831236e-06,
+      "loss": 1.0867,
+      "step": 6225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4047619564327847,
+      "learning_rate": 9.473742464073069e-06,
+      "loss": 1.1392,
+      "step": 6226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9131972943517974,
+      "learning_rate": 9.470982530513078e-06,
+      "loss": 1.1543,
+      "step": 6227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8968121998398355,
+      "learning_rate": 9.468222637362083e-06,
+      "loss": 1.134,
+      "step": 6228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.890044293518447,
+      "learning_rate": 9.465462784830895e-06,
+      "loss": 1.1423,
+      "step": 6229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.199799305731958,
+      "learning_rate": 9.462702973130326e-06,
+      "loss": 1.0706,
+      "step": 6230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.13830109648261,
+      "learning_rate": 9.459943202471186e-06,
+      "loss": 1.0608,
+      "step": 6231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.818147299312386,
+      "learning_rate": 9.457183473064279e-06,
+      "loss": 1.1377,
+      "step": 6232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8103581504173247,
+      "learning_rate": 9.454423785120407e-06,
+      "loss": 1.1233,
+      "step": 6233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2739278978231416,
+      "learning_rate": 9.451664138850368e-06,
+      "loss": 1.0776,
+      "step": 6234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.111136796439127,
+      "learning_rate": 9.448904534464957e-06,
+      "loss": 1.1616,
+      "step": 6235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0547524727926287,
+      "learning_rate": 9.446144972174968e-06,
+      "loss": 1.1118,
+      "step": 6236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8446273790917633,
+      "learning_rate": 9.443385452191189e-06,
+      "loss": 1.1228,
+      "step": 6237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.975990586031927,
+      "learning_rate": 9.440625974724408e-06,
+      "loss": 1.1587,
+      "step": 6238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.902213222198518,
+      "learning_rate": 9.437866539985405e-06,
+      "loss": 1.1284,
+      "step": 6239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.633690932162932,
+      "learning_rate": 9.435107148184958e-06,
+      "loss": 1.157,
+      "step": 6240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2321847158771964,
+      "learning_rate": 9.432347799533845e-06,
+      "loss": 1.0752,
+      "step": 6241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0741801970650413,
+      "learning_rate": 9.429588494242845e-06,
+      "loss": 1.0376,
+      "step": 6242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.895347126945661,
+      "learning_rate": 9.426829232522716e-06,
+      "loss": 1.1702,
+      "step": 6243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2612849568102873,
+      "learning_rate": 9.424070014584232e-06,
+      "loss": 1.1091,
+      "step": 6244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4560876017725395,
+      "learning_rate": 9.421310840638157e-06,
+      "loss": 1.1567,
+      "step": 6245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.193356619623728,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 1.1128,
+      "step": 6246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.954706822968515,
+      "learning_rate": 9.415792625566253e-06,
+      "loss": 1.1101,
+      "step": 6247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.283452565607767,
+      "learning_rate": 9.413033584861935e-06,
+      "loss": 1.1948,
+      "step": 6248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8855863776849415,
+      "learning_rate": 9.41027458899304e-06,
+      "loss": 1.1125,
+      "step": 6249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.611798036917192,
+      "learning_rate": 9.407515638170317e-06,
+      "loss": 1.103,
+      "step": 6250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9551847770188173,
+      "learning_rate": 9.404756732604503e-06,
+      "loss": 1.187,
+      "step": 6251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9537100225860997,
+      "learning_rate": 9.401997872506344e-06,
+      "loss": 1.071,
+      "step": 6252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.615290604251468,
+      "learning_rate": 9.399239058086568e-06,
+      "loss": 1.0664,
+      "step": 6253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.68681988812584,
+      "learning_rate": 9.39648028955591e-06,
+      "loss": 1.1377,
+      "step": 6254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4260571446452843,
+      "learning_rate": 9.3937215671251e-06,
+      "loss": 1.1719,
+      "step": 6255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.211770363352381,
+      "learning_rate": 9.390962891004858e-06,
+      "loss": 1.1277,
+      "step": 6256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2690689607232217,
+      "learning_rate": 9.388204261405911e-06,
+      "loss": 1.1689,
+      "step": 6257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9302952042360904,
+      "learning_rate": 9.38544567853898e-06,
+      "loss": 1.1099,
+      "step": 6258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.362340583050143,
+      "learning_rate": 9.382687142614768e-06,
+      "loss": 1.1304,
+      "step": 6259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.982055878150255,
+      "learning_rate": 9.379928653843993e-06,
+      "loss": 1.1382,
+      "step": 6260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0220168317425014,
+      "learning_rate": 9.377170212437364e-06,
+      "loss": 1.1079,
+      "step": 6261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2200116707288777,
+      "learning_rate": 9.37441181860558e-06,
+      "loss": 1.1094,
+      "step": 6262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.612890256656294,
+      "learning_rate": 9.371653472559341e-06,
+      "loss": 1.1309,
+      "step": 6263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8758574080597157,
+      "learning_rate": 9.368895174509347e-06,
+      "loss": 1.085,
+      "step": 6264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7195462958145145,
+      "learning_rate": 9.366136924666289e-06,
+      "loss": 1.1433,
+      "step": 6265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8996909161294835,
+      "learning_rate": 9.363378723240856e-06,
+      "loss": 1.1235,
+      "step": 6266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4886350770645955,
+      "learning_rate": 9.36062057044373e-06,
+      "loss": 1.1357,
+      "step": 6267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.926265897960244,
+      "learning_rate": 9.357862466485594e-06,
+      "loss": 1.0776,
+      "step": 6268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.746202979916825,
+      "learning_rate": 9.35510441157713e-06,
+      "loss": 1.1699,
+      "step": 6269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.896530822367125,
+      "learning_rate": 9.352346405929004e-06,
+      "loss": 1.144,
+      "step": 6270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7538690515467863,
+      "learning_rate": 9.349588449751898e-06,
+      "loss": 1.1355,
+      "step": 6271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.415893672810472,
+      "learning_rate": 9.346830543256467e-06,
+      "loss": 1.145,
+      "step": 6272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.575369861116968,
+      "learning_rate": 9.344072686653377e-06,
+      "loss": 1.1553,
+      "step": 6273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9056055771881018,
+      "learning_rate": 9.34131488015329e-06,
+      "loss": 1.1021,
+      "step": 6274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5208452490913786,
+      "learning_rate": 9.338557123966858e-06,
+      "loss": 1.1157,
+      "step": 6275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.218073029456759,
+      "learning_rate": 9.335799418304731e-06,
+      "loss": 1.1599,
+      "step": 6276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6857190676885447,
+      "learning_rate": 9.33304176337756e-06,
+      "loss": 1.1218,
+      "step": 6277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5179053221388807,
+      "learning_rate": 9.33028415939599e-06,
+      "loss": 1.207,
+      "step": 6278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.931289524404089,
+      "learning_rate": 9.327526606570654e-06,
+      "loss": 1.1951,
+      "step": 6279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.163235575957316,
+      "learning_rate": 9.324769105112192e-06,
+      "loss": 1.1995,
+      "step": 6280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1865375321497686,
+      "learning_rate": 9.322011655231234e-06,
+      "loss": 1.1343,
+      "step": 6281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3235174688849973,
+      "learning_rate": 9.319254257138408e-06,
+      "loss": 1.1079,
+      "step": 6282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.926861722413885,
+      "learning_rate": 9.31649691104434e-06,
+      "loss": 1.125,
+      "step": 6283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1565647819557716,
+      "learning_rate": 9.313739617159646e-06,
+      "loss": 1.135,
+      "step": 6284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.117353045338949,
+      "learning_rate": 9.310982375694949e-06,
+      "loss": 1.1113,
+      "step": 6285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.251907709595011,
+      "learning_rate": 9.308225186860852e-06,
+      "loss": 1.1611,
+      "step": 6286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3211988484794657,
+      "learning_rate": 9.305468050867966e-06,
+      "loss": 1.1504,
+      "step": 6287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.252089009295548,
+      "learning_rate": 9.302710967926898e-06,
+      "loss": 1.1848,
+      "step": 6288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3252006357769357,
+      "learning_rate": 9.299953938248244e-06,
+      "loss": 1.1963,
+      "step": 6289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.888686585501716,
+      "learning_rate": 9.2971969620426e-06,
+      "loss": 1.1357,
+      "step": 6290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0817923243968917,
+      "learning_rate": 9.294440039520563e-06,
+      "loss": 1.114,
+      "step": 6291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1153142067566826,
+      "learning_rate": 9.291683170892712e-06,
+      "loss": 1.1494,
+      "step": 6292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.07507989961522,
+      "learning_rate": 9.288926356369635e-06,
+      "loss": 1.1409,
+      "step": 6293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.104066057079548,
+      "learning_rate": 9.28616959616191e-06,
+      "loss": 1.0798,
+      "step": 6294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5435287469117194,
+      "learning_rate": 9.283412890480114e-06,
+      "loss": 1.0762,
+      "step": 6295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5975205301739783,
+      "learning_rate": 9.280656239534814e-06,
+      "loss": 1.1211,
+      "step": 6296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0471292916883685,
+      "learning_rate": 9.277899643536584e-06,
+      "loss": 1.0535,
+      "step": 6297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.949386801100136,
+      "learning_rate": 9.275143102695983e-06,
+      "loss": 1.0581,
+      "step": 6298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.071068919286927,
+      "learning_rate": 9.272386617223565e-06,
+      "loss": 1.0972,
+      "step": 6299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0840012132850556,
+      "learning_rate": 9.269630187329887e-06,
+      "loss": 1.1636,
+      "step": 6300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.148777081147181,
+      "learning_rate": 9.266873813225498e-06,
+      "loss": 1.1526,
+      "step": 6301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6547936998390855,
+      "learning_rate": 9.26411749512095e-06,
+      "loss": 1.1548,
+      "step": 6302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9433556140908537,
+      "learning_rate": 9.261361233226776e-06,
+      "loss": 1.1003,
+      "step": 6303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0218137296194603,
+      "learning_rate": 9.258605027753516e-06,
+      "loss": 1.1465,
+      "step": 6304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.827848245387025,
+      "learning_rate": 9.255848878911708e-06,
+      "loss": 1.1472,
+      "step": 6305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7072518472611637,
+      "learning_rate": 9.253092786911871e-06,
+      "loss": 1.1194,
+      "step": 6306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6620200206028013,
+      "learning_rate": 9.250336751964534e-06,
+      "loss": 1.0984,
+      "step": 6307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.684084712091359,
+      "learning_rate": 9.247580774280216e-06,
+      "loss": 1.1355,
+      "step": 6308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.911023043748364,
+      "learning_rate": 9.244824854069432e-06,
+      "loss": 1.1199,
+      "step": 6309
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.131732109580236,
+      "learning_rate": 9.242068991542694e-06,
+      "loss": 1.187,
+      "step": 6310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.529284323631477,
+      "learning_rate": 9.239313186910512e-06,
+      "loss": 1.1438,
+      "step": 6311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4580015733474596,
+      "learning_rate": 9.236557440383381e-06,
+      "loss": 1.1179,
+      "step": 6312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.126020491540739,
+      "learning_rate": 9.233801752171802e-06,
+      "loss": 1.0454,
+      "step": 6313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6404459681179366,
+      "learning_rate": 9.231046122486266e-06,
+      "loss": 1.1072,
+      "step": 6314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.085314196660884,
+      "learning_rate": 9.228290551537264e-06,
+      "loss": 1.1375,
+      "step": 6315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.621390594257437,
+      "learning_rate": 9.225535039535282e-06,
+      "loss": 1.1331,
+      "step": 6316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.449849198685986,
+      "learning_rate": 9.222779586690798e-06,
+      "loss": 1.1389,
+      "step": 6317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8096527267185287,
+      "learning_rate": 9.220024193214289e-06,
+      "loss": 1.1382,
+      "step": 6318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.013177368665764,
+      "learning_rate": 9.217268859316219e-06,
+      "loss": 1.0664,
+      "step": 6319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9806355917768004,
+      "learning_rate": 9.21451358520706e-06,
+      "loss": 1.0708,
+      "step": 6320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.276164512292269,
+      "learning_rate": 9.211758371097274e-06,
+      "loss": 1.1431,
+      "step": 6321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.972210699303499,
+      "learning_rate": 9.209003217197316e-06,
+      "loss": 1.1174,
+      "step": 6322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.296290977838214,
+      "learning_rate": 9.206248123717637e-06,
+      "loss": 1.1296,
+      "step": 6323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.365342726405006,
+      "learning_rate": 9.203493090868692e-06,
+      "loss": 1.1057,
+      "step": 6324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0432949469008594,
+      "learning_rate": 9.200738118860914e-06,
+      "loss": 1.1128,
+      "step": 6325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.855115691003881,
+      "learning_rate": 9.197983207904748e-06,
+      "loss": 1.1326,
+      "step": 6326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.259293072123206,
+      "learning_rate": 9.195228358210624e-06,
+      "loss": 1.1052,
+      "step": 6327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.680721681285045,
+      "learning_rate": 9.192473569988975e-06,
+      "loss": 1.1077,
+      "step": 6328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.391505933402522,
+      "learning_rate": 9.189718843450222e-06,
+      "loss": 1.0833,
+      "step": 6329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4437120104780394,
+      "learning_rate": 9.18696417880479e-06,
+      "loss": 1.1228,
+      "step": 6330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.634676453307781,
+      "learning_rate": 9.184209576263093e-06,
+      "loss": 1.0793,
+      "step": 6331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.2477323412372465,
+      "learning_rate": 9.181455036035533e-06,
+      "loss": 1.1064,
+      "step": 6332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.409144717706073,
+      "learning_rate": 9.178700558332522e-06,
+      "loss": 1.1045,
+      "step": 6333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.076698103949818,
+      "learning_rate": 9.17594614336446e-06,
+      "loss": 1.1448,
+      "step": 6334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.659130932580088,
+      "learning_rate": 9.173191791341745e-06,
+      "loss": 1.1636,
+      "step": 6335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3385550613734067,
+      "learning_rate": 9.170437502474765e-06,
+      "loss": 1.1787,
+      "step": 6336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.012698492534915,
+      "learning_rate": 9.167683276973906e-06,
+      "loss": 1.1338,
+      "step": 6337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.21405370988789,
+      "learning_rate": 9.164929115049556e-06,
+      "loss": 1.0945,
+      "step": 6338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1267368079060254,
+      "learning_rate": 9.162175016912084e-06,
+      "loss": 1.0833,
+      "step": 6339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9631825374115617,
+      "learning_rate": 9.159420982771865e-06,
+      "loss": 1.1389,
+      "step": 6340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5648178458761777,
+      "learning_rate": 9.156667012839264e-06,
+      "loss": 1.1248,
+      "step": 6341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.413249062731597,
+      "learning_rate": 9.153913107324644e-06,
+      "loss": 1.1777,
+      "step": 6342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.479317594855192,
+      "learning_rate": 9.151159266438362e-06,
+      "loss": 1.1274,
+      "step": 6343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7647149295352658,
+      "learning_rate": 9.148405490390774e-06,
+      "loss": 1.1294,
+      "step": 6344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.438965813402253,
+      "learning_rate": 9.145651779392222e-06,
+      "loss": 1.1643,
+      "step": 6345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.671021964026244,
+      "learning_rate": 9.142898133653049e-06,
+      "loss": 1.1394,
+      "step": 6346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2551808465432943,
+      "learning_rate": 9.140144553383592e-06,
+      "loss": 1.1213,
+      "step": 6347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.755456077867906,
+      "learning_rate": 9.137391038794184e-06,
+      "loss": 1.1372,
+      "step": 6348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.874438538906881,
+      "learning_rate": 9.134637590095154e-06,
+      "loss": 1.1499,
+      "step": 6349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7372288984326003,
+      "learning_rate": 9.131884207496822e-06,
+      "loss": 1.116,
+      "step": 6350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.235389368499589,
+      "learning_rate": 9.129130891209508e-06,
+      "loss": 1.1106,
+      "step": 6351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2526804436432672,
+      "learning_rate": 9.126377641443518e-06,
+      "loss": 1.1985,
+      "step": 6352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6324668736209125,
+      "learning_rate": 9.123624458409163e-06,
+      "loss": 1.0696,
+      "step": 6353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.364418288252834,
+      "learning_rate": 9.120871342316746e-06,
+      "loss": 1.1787,
+      "step": 6354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.731890769247617,
+      "learning_rate": 9.11811829337656e-06,
+      "loss": 1.1458,
+      "step": 6355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.165736802491367,
+      "learning_rate": 9.115365311798901e-06,
+      "loss": 1.1135,
+      "step": 6356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.476980481711392,
+      "learning_rate": 9.112612397794052e-06,
+      "loss": 1.1282,
+      "step": 6357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.118968039287161,
+      "learning_rate": 9.109859551572297e-06,
+      "loss": 1.1169,
+      "step": 6358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.406361300875345,
+      "learning_rate": 9.107106773343913e-06,
+      "loss": 1.0906,
+      "step": 6359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7025846843390933,
+      "learning_rate": 9.104354063319163e-06,
+      "loss": 1.209,
+      "step": 6360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4028108196963145,
+      "learning_rate": 9.10160142170832e-06,
+      "loss": 1.1001,
+      "step": 6361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3275286179054127,
+      "learning_rate": 9.098848848721643e-06,
+      "loss": 1.1338,
+      "step": 6362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4939076333037673,
+      "learning_rate": 9.096096344569388e-06,
+      "loss": 1.0813,
+      "step": 6363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.358128927031698,
+      "learning_rate": 9.093343909461807e-06,
+      "loss": 1.104,
+      "step": 6364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.156494864012961,
+      "learning_rate": 9.090591543609138e-06,
+      "loss": 1.0994,
+      "step": 6365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3326779453647064,
+      "learning_rate": 9.087839247221624e-06,
+      "loss": 1.0845,
+      "step": 6366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6710124932925465,
+      "learning_rate": 9.0850870205095e-06,
+      "loss": 1.0652,
+      "step": 6367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.032004232133889,
+      "learning_rate": 9.082334863682995e-06,
+      "loss": 1.0708,
+      "step": 6368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3994129643086426,
+      "learning_rate": 9.07958277695233e-06,
+      "loss": 1.0659,
+      "step": 6369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.124559754787657,
+      "learning_rate": 9.076830760527727e-06,
+      "loss": 1.1675,
+      "step": 6370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.701404638759162,
+      "learning_rate": 9.074078814619398e-06,
+      "loss": 1.1487,
+      "step": 6371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4245365679363653,
+      "learning_rate": 9.071326939437545e-06,
+      "loss": 1.022,
+      "step": 6372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4807433862628545,
+      "learning_rate": 9.068575135192377e-06,
+      "loss": 1.1479,
+      "step": 6373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2267145809142774,
+      "learning_rate": 9.065823402094086e-06,
+      "loss": 1.0945,
+      "step": 6374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2034623991813826,
+      "learning_rate": 9.063071740352863e-06,
+      "loss": 1.1218,
+      "step": 6375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.054995092347151,
+      "learning_rate": 9.060320150178898e-06,
+      "loss": 1.1382,
+      "step": 6376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.208677424768336,
+      "learning_rate": 9.057568631782365e-06,
+      "loss": 1.0957,
+      "step": 6377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.40428056830202,
+      "learning_rate": 9.054817185373448e-06,
+      "loss": 1.1042,
+      "step": 6378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8864832110640997,
+      "learning_rate": 9.052065811162305e-06,
+      "loss": 1.0813,
+      "step": 6379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7786100389867037,
+      "learning_rate": 9.049314509359105e-06,
+      "loss": 1.1216,
+      "step": 6380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9037772368269024,
+      "learning_rate": 9.046563280174007e-06,
+      "loss": 1.0894,
+      "step": 6381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.836522342209137,
+      "learning_rate": 9.043812123817165e-06,
+      "loss": 1.1111,
+      "step": 6382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.007103835183398,
+      "learning_rate": 9.04106104049872e-06,
+      "loss": 1.0674,
+      "step": 6383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7041328099735793,
+      "learning_rate": 9.038310030428824e-06,
+      "loss": 1.123,
+      "step": 6384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.461972382354196,
+      "learning_rate": 9.035559093817598e-06,
+      "loss": 1.1199,
+      "step": 6385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2642517005677396,
+      "learning_rate": 9.032808230875184e-06,
+      "loss": 1.0671,
+      "step": 6386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6951501343887125,
+      "learning_rate": 9.030057441811703e-06,
+      "loss": 1.1606,
+      "step": 6387
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0638700852046337,
+      "learning_rate": 9.027306726837273e-06,
+      "loss": 1.1914,
+      "step": 6388
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.306275974517674,
+      "learning_rate": 9.024556086162009e-06,
+      "loss": 1.1772,
+      "step": 6389
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4615510109583916,
+      "learning_rate": 9.02180551999602e-06,
+      "loss": 1.2986,
+      "step": 6390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0718727297109942,
+      "learning_rate": 9.019055028549403e-06,
+      "loss": 1.1072,
+      "step": 6391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0564773752109313,
+      "learning_rate": 9.01630461203226e-06,
+      "loss": 1.0791,
+      "step": 6392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0500036011230254,
+      "learning_rate": 9.013554270654676e-06,
+      "loss": 1.1121,
+      "step": 6393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1263233359372595,
+      "learning_rate": 9.010804004626739e-06,
+      "loss": 1.1062,
+      "step": 6394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2401326547537006,
+      "learning_rate": 9.008053814158529e-06,
+      "loss": 1.0315,
+      "step": 6395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.857080926031684,
+      "learning_rate": 9.005303699460116e-06,
+      "loss": 0.9848,
+      "step": 6396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6360361324693504,
+      "learning_rate": 9.002553660741575e-06,
+      "loss": 1.1519,
+      "step": 6397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2889441818732967,
+      "learning_rate": 8.999803698212956e-06,
+      "loss": 1.1099,
+      "step": 6398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0367806774229376,
+      "learning_rate": 8.99705381208432e-06,
+      "loss": 1.0786,
+      "step": 6399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6785051175141295,
+      "learning_rate": 8.994304002565723e-06,
+      "loss": 1.113,
+      "step": 6400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0331488748832,
+      "learning_rate": 8.9915542698672e-06,
+      "loss": 1.1184,
+      "step": 6401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1058666837997526,
+      "learning_rate": 8.988804614198795e-06,
+      "loss": 1.155,
+      "step": 6402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7041016619405975,
+      "learning_rate": 8.986055035770536e-06,
+      "loss": 1.0496,
+      "step": 6403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7017217632876096,
+      "learning_rate": 8.983305534792459e-06,
+      "loss": 1.1074,
+      "step": 6404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.955225149177019,
+      "learning_rate": 8.980556111474572e-06,
+      "loss": 1.1423,
+      "step": 6405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1507089615738715,
+      "learning_rate": 8.9778067660269e-06,
+      "loss": 1.1262,
+      "step": 6406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0546799640428417,
+      "learning_rate": 8.975057498659444e-06,
+      "loss": 1.0544,
+      "step": 6407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3002469188978356,
+      "learning_rate": 8.97230830958221e-06,
+      "loss": 1.1736,
+      "step": 6408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2973129750050334,
+      "learning_rate": 8.969559199005196e-06,
+      "loss": 1.1514,
+      "step": 6409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.406085378724695,
+      "learning_rate": 8.96681016713839e-06,
+      "loss": 1.1763,
+      "step": 6410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.943621670405987,
+      "learning_rate": 8.964061214191784e-06,
+      "loss": 1.1248,
+      "step": 6411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9374588918574163,
+      "learning_rate": 8.961312340375346e-06,
+      "loss": 1.1013,
+      "step": 6412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.944519072540962,
+      "learning_rate": 8.958563545899055e-06,
+      "loss": 1.1492,
+      "step": 6413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.13968831018592,
+      "learning_rate": 8.955814830972875e-06,
+      "loss": 1.1184,
+      "step": 6414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.322580247078618,
+      "learning_rate": 8.95306619580677e-06,
+      "loss": 1.0957,
+      "step": 6415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.304906603074494,
+      "learning_rate": 8.95031764061069e-06,
+      "loss": 1.137,
+      "step": 6416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.001548676430134,
+      "learning_rate": 8.947569165594591e-06,
+      "loss": 1.1338,
+      "step": 6417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.119115804409349,
+      "learning_rate": 8.944820770968407e-06,
+      "loss": 1.1548,
+      "step": 6418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9072747361084454,
+      "learning_rate": 8.942072456942078e-06,
+      "loss": 1.0684,
+      "step": 6419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.541146458179454,
+      "learning_rate": 8.939324223725532e-06,
+      "loss": 1.1863,
+      "step": 6420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6239333571644416,
+      "learning_rate": 8.936576071528693e-06,
+      "loss": 1.0994,
+      "step": 6421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.389214887626825,
+      "learning_rate": 8.93382800056148e-06,
+      "loss": 1.0764,
+      "step": 6422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.069295368296649,
+      "learning_rate": 8.931080011033805e-06,
+      "loss": 1.1802,
+      "step": 6423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.080301709620625,
+      "learning_rate": 8.928332103155572e-06,
+      "loss": 1.0784,
+      "step": 6424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.556606865039183,
+      "learning_rate": 8.925584277136682e-06,
+      "loss": 1.1147,
+      "step": 6425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0266916234063777,
+      "learning_rate": 8.922836533187022e-06,
+      "loss": 1.1082,
+      "step": 6426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.880495657912687,
+      "learning_rate": 8.920088871516482e-06,
+      "loss": 1.1516,
+      "step": 6427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.171626979780581,
+      "learning_rate": 8.917341292334945e-06,
+      "loss": 1.0898,
+      "step": 6428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9189726959697615,
+      "learning_rate": 8.914593795852278e-06,
+      "loss": 1.0586,
+      "step": 6429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.814098077880715,
+      "learning_rate": 8.911846382278354e-06,
+      "loss": 1.0779,
+      "step": 6430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.935095387608779,
+      "learning_rate": 8.909099051823036e-06,
+      "loss": 1.1506,
+      "step": 6431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.002233569941723,
+      "learning_rate": 8.906351804696171e-06,
+      "loss": 1.0505,
+      "step": 6432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.20048806803655,
+      "learning_rate": 8.903604641107615e-06,
+      "loss": 1.1304,
+      "step": 6433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.832347917125725,
+      "learning_rate": 8.900857561267204e-06,
+      "loss": 1.0996,
+      "step": 6434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.281570529751053,
+      "learning_rate": 8.898110565384776e-06,
+      "loss": 1.1089,
+      "step": 6435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.639433626872034,
+      "learning_rate": 8.895363653670162e-06,
+      "loss": 1.1294,
+      "step": 6436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.375295927372276,
+      "learning_rate": 8.892616826333187e-06,
+      "loss": 1.1223,
+      "step": 6437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.165023503681654,
+      "learning_rate": 8.889870083583662e-06,
+      "loss": 1.0459,
+      "step": 6438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.442278974308606,
+      "learning_rate": 8.887123425631396e-06,
+      "loss": 1.0735,
+      "step": 6439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.1977707620796805,
+      "learning_rate": 8.884376852686196e-06,
+      "loss": 1.1672,
+      "step": 6440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1856312304986485,
+      "learning_rate": 8.881630364957859e-06,
+      "loss": 1.1277,
+      "step": 6441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0804795026226675,
+      "learning_rate": 8.878883962656176e-06,
+      "loss": 1.1404,
+      "step": 6442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.727688906965597,
+      "learning_rate": 8.876137645990928e-06,
+      "loss": 1.1321,
+      "step": 6443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.406421967841384,
+      "learning_rate": 8.873391415171899e-06,
+      "loss": 1.1235,
+      "step": 6444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8275448253595523,
+      "learning_rate": 8.870645270408849e-06,
+      "loss": 1.1238,
+      "step": 6445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.933455736841649,
+      "learning_rate": 8.867899211911549e-06,
+      "loss": 1.1333,
+      "step": 6446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.220382595681767,
+      "learning_rate": 8.865153239889757e-06,
+      "loss": 1.0959,
+      "step": 6447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3283049796134994,
+      "learning_rate": 8.86240735455322e-06,
+      "loss": 1.1179,
+      "step": 6448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.634134192532286,
+      "learning_rate": 8.859661556111686e-06,
+      "loss": 1.1145,
+      "step": 6449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1010242701997557,
+      "learning_rate": 8.856915844774897e-06,
+      "loss": 1.1145,
+      "step": 6450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3357932564924244,
+      "learning_rate": 8.854170220752573e-06,
+      "loss": 1.1458,
+      "step": 6451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5503765064107355,
+      "learning_rate": 8.851424684254449e-06,
+      "loss": 1.0977,
+      "step": 6452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1387868182458534,
+      "learning_rate": 8.848679235490235e-06,
+      "loss": 1.1394,
+      "step": 6453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.973002267615614,
+      "learning_rate": 8.845933874669645e-06,
+      "loss": 1.104,
+      "step": 6454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.314348799329983,
+      "learning_rate": 8.843188602002385e-06,
+      "loss": 1.1631,
+      "step": 6455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.2293929265867956,
+      "learning_rate": 8.840443417698152e-06,
+      "loss": 1.114,
+      "step": 6456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5933817111747737,
+      "learning_rate": 8.837698321966638e-06,
+      "loss": 1.0564,
+      "step": 6457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0250033919765134,
+      "learning_rate": 8.834953315017522e-06,
+      "loss": 1.1018,
+      "step": 6458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1338956576571215,
+      "learning_rate": 8.832208397060486e-06,
+      "loss": 1.0037,
+      "step": 6459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.855307541610434,
+      "learning_rate": 8.829463568305197e-06,
+      "loss": 1.0747,
+      "step": 6460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.602703334390688,
+      "learning_rate": 8.826718828961324e-06,
+      "loss": 1.1174,
+      "step": 6461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8472137653003706,
+      "learning_rate": 8.823974179238519e-06,
+      "loss": 1.1206,
+      "step": 6462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7487471212958137,
+      "learning_rate": 8.821229619346433e-06,
+      "loss": 1.1116,
+      "step": 6463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.171789452527832,
+      "learning_rate": 8.818485149494714e-06,
+      "loss": 1.0935,
+      "step": 6464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.150909631615675,
+      "learning_rate": 8.815740769892991e-06,
+      "loss": 1.1528,
+      "step": 6465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7650582756092525,
+      "learning_rate": 8.812996480750899e-06,
+      "loss": 1.105,
+      "step": 6466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.933799681661119,
+      "learning_rate": 8.810252282278056e-06,
+      "loss": 1.1431,
+      "step": 6467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.293552537767633,
+      "learning_rate": 8.80750817468408e-06,
+      "loss": 1.0786,
+      "step": 6468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1503748254799606,
+      "learning_rate": 8.804764158178579e-06,
+      "loss": 1.1433,
+      "step": 6469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.96569039515878,
+      "learning_rate": 8.802020232971158e-06,
+      "loss": 1.1265,
+      "step": 6470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.735537566372152,
+      "learning_rate": 8.799276399271407e-06,
+      "loss": 1.0935,
+      "step": 6471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4634600903036468,
+      "learning_rate": 8.796532657288916e-06,
+      "loss": 1.0691,
+      "step": 6472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.269853412828356,
+      "learning_rate": 8.793789007233264e-06,
+      "loss": 1.0972,
+      "step": 6473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6016965275358954,
+      "learning_rate": 8.791045449314023e-06,
+      "loss": 1.1531,
+      "step": 6474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.126957006931576,
+      "learning_rate": 8.78830198374077e-06,
+      "loss": 1.1343,
+      "step": 6475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9213304536991638,
+      "learning_rate": 8.78555861072305e-06,
+      "loss": 1.1807,
+      "step": 6476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2970408011927423,
+      "learning_rate": 8.78281533047043e-06,
+      "loss": 1.1099,
+      "step": 6477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0403702221912745,
+      "learning_rate": 8.780072143192444e-06,
+      "loss": 1.1465,
+      "step": 6478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5174822102543466,
+      "learning_rate": 8.777329049098632e-06,
+      "loss": 1.1216,
+      "step": 6479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.232404592643348,
+      "learning_rate": 8.77458604839853e-06,
+      "loss": 1.1365,
+      "step": 6480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.805277696561689,
+      "learning_rate": 8.771843141301659e-06,
+      "loss": 1.1277,
+      "step": 6481
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7148008149340246,
+      "learning_rate": 8.769100328017536e-06,
+      "loss": 1.0879,
+      "step": 6482
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.42465536433943,
+      "learning_rate": 8.766357608755671e-06,
+      "loss": 1.1521,
+      "step": 6483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1409819771357155,
+      "learning_rate": 8.763614983725573e-06,
+      "loss": 1.1096,
+      "step": 6484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.024679470960605,
+      "learning_rate": 8.760872453136727e-06,
+      "loss": 1.1169,
+      "step": 6485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.356251248174103,
+      "learning_rate": 8.758130017198625e-06,
+      "loss": 1.0959,
+      "step": 6486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4043166742170374,
+      "learning_rate": 8.755387676120747e-06,
+      "loss": 1.1062,
+      "step": 6487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.981081273132136,
+      "learning_rate": 8.752645430112568e-06,
+      "loss": 1.073,
+      "step": 6488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.293376179340244,
+      "learning_rate": 8.749903279383557e-06,
+      "loss": 1.1624,
+      "step": 6489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4945753782873124,
+      "learning_rate": 8.747161224143175e-06,
+      "loss": 1.0818,
+      "step": 6490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.41522875313,
+      "learning_rate": 8.744419264600864e-06,
+      "loss": 1.1138,
+      "step": 6491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.050678329576552,
+      "learning_rate": 8.741677400966075e-06,
+      "loss": 1.1096,
+      "step": 6492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.242425223446269,
+      "learning_rate": 8.738935633448242e-06,
+      "loss": 1.1128,
+      "step": 6493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.423982967089449,
+      "learning_rate": 8.736193962256801e-06,
+      "loss": 1.1677,
+      "step": 6494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.004643777488972,
+      "learning_rate": 8.73345238760117e-06,
+      "loss": 1.1226,
+      "step": 6495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7435465152530067,
+      "learning_rate": 8.730710909690761e-06,
+      "loss": 1.1694,
+      "step": 6496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.389399757181628,
+      "learning_rate": 8.727969528734993e-06,
+      "loss": 1.0933,
+      "step": 6497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.223089664386939,
+      "learning_rate": 8.725228244943254e-06,
+      "loss": 1.0994,
+      "step": 6498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7853884176109824,
+      "learning_rate": 8.722487058524945e-06,
+      "loss": 1.0613,
+      "step": 6499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4100377993806004,
+      "learning_rate": 8.719745969689444e-06,
+      "loss": 1.0659,
+      "step": 6500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6645635158614898,
+      "learning_rate": 8.717004978646135e-06,
+      "loss": 1.1794,
+      "step": 6501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.387907118920395,
+      "learning_rate": 8.714264085604387e-06,
+      "loss": 1.1929,
+      "step": 6502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.31085599378136,
+      "learning_rate": 8.711523290773567e-06,
+      "loss": 1.1528,
+      "step": 6503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.955424697903043,
+      "learning_rate": 8.708782594363027e-06,
+      "loss": 1.0764,
+      "step": 6504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.326896558156191,
+      "learning_rate": 8.706041996582111e-06,
+      "loss": 1.146,
+      "step": 6505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7849725555517613,
+      "learning_rate": 8.703301497640164e-06,
+      "loss": 1.0967,
+      "step": 6506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4787888316548448,
+      "learning_rate": 8.700561097746516e-06,
+      "loss": 1.1558,
+      "step": 6507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8935344318545195,
+      "learning_rate": 8.697820797110499e-06,
+      "loss": 1.1169,
+      "step": 6508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5568068217133773,
+      "learning_rate": 8.695080595941424e-06,
+      "loss": 1.1423,
+      "step": 6509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.304035161895299,
+      "learning_rate": 8.692340494448608e-06,
+      "loss": 1.1921,
+      "step": 6510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1034589990955923,
+      "learning_rate": 8.689600492841345e-06,
+      "loss": 1.1123,
+      "step": 6511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3061108091486573,
+      "learning_rate": 8.686860591328936e-06,
+      "loss": 1.0784,
+      "step": 6512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6067807193705543,
+      "learning_rate": 8.684120790120666e-06,
+      "loss": 1.1284,
+      "step": 6513
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3639106713956117,
+      "learning_rate": 8.681381089425816e-06,
+      "loss": 1.1594,
+      "step": 6514
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.174122126033209,
+      "learning_rate": 8.678641489453656e-06,
+      "loss": 1.1204,
+      "step": 6515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.292009378882687,
+      "learning_rate": 8.675901990413454e-06,
+      "loss": 1.1301,
+      "step": 6516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.936563547762443,
+      "learning_rate": 8.673162592514466e-06,
+      "loss": 1.0728,
+      "step": 6517
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.182259175910038,
+      "learning_rate": 8.670423295965938e-06,
+      "loss": 1.04,
+      "step": 6518
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.375791632757263,
+      "learning_rate": 8.66768410097711e-06,
+      "loss": 1.1501,
+      "step": 6519
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1975912267382087,
+      "learning_rate": 8.664945007757218e-06,
+      "loss": 1.1167,
+      "step": 6520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3128244237922018,
+      "learning_rate": 8.66220601651549e-06,
+      "loss": 1.1182,
+      "step": 6521
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.162294083184564,
+      "learning_rate": 8.659467127461139e-06,
+      "loss": 1.0967,
+      "step": 6522
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.266176131190247,
+      "learning_rate": 8.656728340803382e-06,
+      "loss": 1.1904,
+      "step": 6523
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.055665994304304,
+      "learning_rate": 8.65398965675141e-06,
+      "loss": 1.1824,
+      "step": 6524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3324167668913027,
+      "learning_rate": 8.651251075514426e-06,
+      "loss": 1.094,
+      "step": 6525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.290735288843046,
+      "learning_rate": 8.648512597301613e-06,
+      "loss": 1.1069,
+      "step": 6526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5186502428867197,
+      "learning_rate": 8.645774222322154e-06,
+      "loss": 1.1572,
+      "step": 6527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7353533196128694,
+      "learning_rate": 8.643035950785214e-06,
+      "loss": 1.0965,
+      "step": 6528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5086765257988524,
+      "learning_rate": 8.64029778289996e-06,
+      "loss": 1.1404,
+      "step": 6529
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5522320695318395,
+      "learning_rate": 8.637559718875547e-06,
+      "loss": 1.1484,
+      "step": 6530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.426409920595193,
+      "learning_rate": 8.634821758921119e-06,
+      "loss": 1.1328,
+      "step": 6531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.300332932466828,
+      "learning_rate": 8.632083903245816e-06,
+      "loss": 1.1475,
+      "step": 6532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8535724752704414,
+      "learning_rate": 8.62934615205877e-06,
+      "loss": 1.0437,
+      "step": 6533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9537382936447574,
+      "learning_rate": 8.626608505569102e-06,
+      "loss": 1.207,
+      "step": 6534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.39164799709596,
+      "learning_rate": 8.62387096398593e-06,
+      "loss": 1.0583,
+      "step": 6535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.927157994552082,
+      "learning_rate": 8.62113352751836e-06,
+      "loss": 1.1072,
+      "step": 6536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.129326292847424,
+      "learning_rate": 8.618396196375494e-06,
+      "loss": 1.1238,
+      "step": 6537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.049570404314874,
+      "learning_rate": 8.615658970766414e-06,
+      "loss": 1.1062,
+      "step": 6538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3101367004290814,
+      "learning_rate": 8.612921850900211e-06,
+      "loss": 1.1504,
+      "step": 6539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5756763873352586,
+      "learning_rate": 8.61018483698596e-06,
+      "loss": 1.1709,
+      "step": 6540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.262793982358237,
+      "learning_rate": 8.607447929232721e-06,
+      "loss": 1.1665,
+      "step": 6541
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3864767822880135,
+      "learning_rate": 8.604711127849558e-06,
+      "loss": 1.1335,
+      "step": 6542
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.473343503303754,
+      "learning_rate": 8.601974433045527e-06,
+      "loss": 1.1365,
+      "step": 6543
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.232763099789186,
+      "learning_rate": 8.599237845029657e-06,
+      "loss": 1.1541,
+      "step": 6544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9383471692781526,
+      "learning_rate": 8.596501364010992e-06,
+      "loss": 1.0586,
+      "step": 6545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7440065443604924,
+      "learning_rate": 8.593764990198556e-06,
+      "loss": 1.1191,
+      "step": 6546
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.38287031187426,
+      "learning_rate": 8.591028723801364e-06,
+      "loss": 1.0791,
+      "step": 6547
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.018738818314528,
+      "learning_rate": 8.588292565028429e-06,
+      "loss": 1.2144,
+      "step": 6548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.396830275717514,
+      "learning_rate": 8.585556514088754e-06,
+      "loss": 1.1677,
+      "step": 6549
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.214481589957566,
+      "learning_rate": 8.582820571191328e-06,
+      "loss": 1.1533,
+      "step": 6550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2279533648377003,
+      "learning_rate": 8.580084736545138e-06,
+      "loss": 1.1096,
+      "step": 6551
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.356926081898132,
+      "learning_rate": 8.57734901035916e-06,
+      "loss": 1.1709,
+      "step": 6552
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.181950048685042,
+      "learning_rate": 8.574613392842361e-06,
+      "loss": 1.167,
+      "step": 6553
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3628275055400003,
+      "learning_rate": 8.571877884203705e-06,
+      "loss": 1.1204,
+      "step": 6554
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2318960002632493,
+      "learning_rate": 8.56914248465214e-06,
+      "loss": 1.0801,
+      "step": 6555
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.168839680840706,
+      "learning_rate": 8.566407194396612e-06,
+      "loss": 1.1382,
+      "step": 6556
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1394766928179876,
+      "learning_rate": 8.563672013646057e-06,
+      "loss": 1.1699,
+      "step": 6557
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2401399250158947,
+      "learning_rate": 8.560936942609397e-06,
+      "loss": 1.1167,
+      "step": 6558
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.862813508617217,
+      "learning_rate": 8.558201981495555e-06,
+      "loss": 1.0276,
+      "step": 6559
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9412405182969423,
+      "learning_rate": 8.555467130513438e-06,
+      "loss": 1.1013,
+      "step": 6560
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.33626206304915,
+      "learning_rate": 8.55273238987195e-06,
+      "loss": 1.1104,
+      "step": 6561
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.1722740585113165,
+      "learning_rate": 8.549997759779981e-06,
+      "loss": 1.1284,
+      "step": 6562
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.554515966793548,
+      "learning_rate": 8.547263240446424e-06,
+      "loss": 1.1045,
+      "step": 6563
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2725835355010995,
+      "learning_rate": 8.544528832080142e-06,
+      "loss": 1.0896,
+      "step": 6564
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4154370379831036,
+      "learning_rate": 8.541794534890014e-06,
+      "loss": 1.0762,
+      "step": 6565
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1515221143201377,
+      "learning_rate": 8.539060349084892e-06,
+      "loss": 1.1089,
+      "step": 6566
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.15782810967762,
+      "learning_rate": 8.536326274873629e-06,
+      "loss": 1.1465,
+      "step": 6567
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.92510289954993,
+      "learning_rate": 8.533592312465073e-06,
+      "loss": 1.1362,
+      "step": 6568
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.744047275748094,
+      "learning_rate": 8.530858462068048e-06,
+      "loss": 1.1582,
+      "step": 6569
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.486891590926971,
+      "learning_rate": 8.528124723891388e-06,
+      "loss": 1.0873,
+      "step": 6570
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.971491633456709,
+      "learning_rate": 8.525391098143902e-06,
+      "loss": 1.0696,
+      "step": 6571
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.163187940288993,
+      "learning_rate": 8.522657585034403e-06,
+      "loss": 1.0337,
+      "step": 6572
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.564911122096137,
+      "learning_rate": 8.519924184771688e-06,
+      "loss": 1.1384,
+      "step": 6573
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9674218481137196,
+      "learning_rate": 8.517190897564547e-06,
+      "loss": 1.1172,
+      "step": 6574
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.814499494798027,
+      "learning_rate": 8.514457723621763e-06,
+      "loss": 1.1016,
+      "step": 6575
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3249890854473074,
+      "learning_rate": 8.511724663152116e-06,
+      "loss": 1.0898,
+      "step": 6576
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3672871709054055,
+      "learning_rate": 8.508991716364359e-06,
+      "loss": 1.0706,
+      "step": 6577
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6255071697321055,
+      "learning_rate": 8.506258883467257e-06,
+      "loss": 1.1003,
+      "step": 6578
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7431803206637273,
+      "learning_rate": 8.503526164669551e-06,
+      "loss": 1.0942,
+      "step": 6579
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.307186324182799,
+      "learning_rate": 8.500793560179984e-06,
+      "loss": 1.1165,
+      "step": 6580
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.637652600568882,
+      "learning_rate": 8.498061070207286e-06,
+      "loss": 1.2097,
+      "step": 6581
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.0351558934107254,
+      "learning_rate": 8.495328694960177e-06,
+      "loss": 1.146,
+      "step": 6582
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5998455150356237,
+      "learning_rate": 8.49259643464737e-06,
+      "loss": 1.1655,
+      "step": 6583
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4475112808599166,
+      "learning_rate": 8.48986428947757e-06,
+      "loss": 1.1069,
+      "step": 6584
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1010930095758598,
+      "learning_rate": 8.487132259659467e-06,
+      "loss": 1.1052,
+      "step": 6585
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3455347903588826,
+      "learning_rate": 8.484400345401751e-06,
+      "loss": 1.1013,
+      "step": 6586
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4670749233122398,
+      "learning_rate": 8.4816685469131e-06,
+      "loss": 1.1763,
+      "step": 6587
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.487841327232657,
+      "learning_rate": 8.478936864402179e-06,
+      "loss": 1.1106,
+      "step": 6588
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4513750533936682,
+      "learning_rate": 8.47620529807765e-06,
+      "loss": 1.0728,
+      "step": 6589
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2838190336135376,
+      "learning_rate": 8.473473848148167e-06,
+      "loss": 1.0691,
+      "step": 6590
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.644638475453722,
+      "learning_rate": 8.470742514822363e-06,
+      "loss": 1.1121,
+      "step": 6591
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8584128784820875,
+      "learning_rate": 8.46801129830888e-06,
+      "loss": 1.1165,
+      "step": 6592
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.993466091475192,
+      "learning_rate": 8.465280198816334e-06,
+      "loss": 1.1458,
+      "step": 6593
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.562422170384916,
+      "learning_rate": 8.462549216553343e-06,
+      "loss": 1.1226,
+      "step": 6594
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.757290063545395,
+      "learning_rate": 8.459818351728514e-06,
+      "loss": 1.1138,
+      "step": 6595
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9374217421013125,
+      "learning_rate": 8.457087604550449e-06,
+      "loss": 1.1353,
+      "step": 6596
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4326322047853335,
+      "learning_rate": 8.45435697522773e-06,
+      "loss": 1.0532,
+      "step": 6597
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.713066535099856,
+      "learning_rate": 8.451626463968932e-06,
+      "loss": 1.1907,
+      "step": 6598
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7011705894266678,
+      "learning_rate": 8.448896070982634e-06,
+      "loss": 1.0691,
+      "step": 6599
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.46303776638282,
+      "learning_rate": 8.44616579647739e-06,
+      "loss": 1.0649,
+      "step": 6600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8204742094638773,
+      "learning_rate": 8.443435640661757e-06,
+      "loss": 1.1245,
+      "step": 6601
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8853663018003193,
+      "learning_rate": 8.440705603744274e-06,
+      "loss": 1.1003,
+      "step": 6602
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.577832732434957,
+      "learning_rate": 8.437975685933484e-06,
+      "loss": 1.0815,
+      "step": 6603
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.990355699329003,
+      "learning_rate": 8.435245887437898e-06,
+      "loss": 1.1218,
+      "step": 6604
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2639296908965605,
+      "learning_rate": 8.432516208466038e-06,
+      "loss": 1.1162,
+      "step": 6605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7773823228841996,
+      "learning_rate": 8.429786649226414e-06,
+      "loss": 1.1069,
+      "step": 6606
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.224871973780093,
+      "learning_rate": 8.427057209927519e-06,
+      "loss": 1.1238,
+      "step": 6607
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2584883475685316,
+      "learning_rate": 8.42432789077784e-06,
+      "loss": 1.1091,
+      "step": 6608
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.419735058932921,
+      "learning_rate": 8.421598691985859e-06,
+      "loss": 1.1118,
+      "step": 6609
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.233018105022174,
+      "learning_rate": 8.41886961376005e-06,
+      "loss": 1.1392,
+      "step": 6610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.417069744340332,
+      "learning_rate": 8.416140656308866e-06,
+      "loss": 1.1243,
+      "step": 6611
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.97294342728508,
+      "learning_rate": 8.41341181984076e-06,
+      "loss": 1.1707,
+      "step": 6612
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.027975347486993,
+      "learning_rate": 8.410683104564175e-06,
+      "loss": 1.0779,
+      "step": 6613
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2325279737041734,
+      "learning_rate": 8.407954510687544e-06,
+      "loss": 1.1775,
+      "step": 6614
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0038675687392793,
+      "learning_rate": 8.405226038419293e-06,
+      "loss": 1.1021,
+      "step": 6615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.575061726585904,
+      "learning_rate": 8.402497687967837e-06,
+      "loss": 1.1616,
+      "step": 6616
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.536538851270299,
+      "learning_rate": 8.399769459541573e-06,
+      "loss": 1.0659,
+      "step": 6617
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.518537295019257,
+      "learning_rate": 8.397041353348901e-06,
+      "loss": 1.1548,
+      "step": 6618
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.153336825518472,
+      "learning_rate": 8.394313369598211e-06,
+      "loss": 1.0754,
+      "step": 6619
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.956899036580155,
+      "learning_rate": 8.391585508497876e-06,
+      "loss": 1.0803,
+      "step": 6620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.983554262240291,
+      "learning_rate": 8.388857770256264e-06,
+      "loss": 1.064,
+      "step": 6621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4615888438114952,
+      "learning_rate": 8.386130155081736e-06,
+      "loss": 1.1399,
+      "step": 6622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.018756800527519,
+      "learning_rate": 8.38340266318264e-06,
+      "loss": 1.075,
+      "step": 6623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.452908168917921,
+      "learning_rate": 8.380675294767312e-06,
+      "loss": 1.1133,
+      "step": 6624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7003474187281595,
+      "learning_rate": 8.377948050044087e-06,
+      "loss": 1.1528,
+      "step": 6625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0956005237375823,
+      "learning_rate": 8.375220929221281e-06,
+      "loss": 1.1113,
+      "step": 6626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.565301254498745,
+      "learning_rate": 8.372493932507207e-06,
+      "loss": 1.1475,
+      "step": 6627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7097980634092136,
+      "learning_rate": 8.369767060110165e-06,
+      "loss": 1.136,
+      "step": 6628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2254601674790813,
+      "learning_rate": 8.367040312238451e-06,
+      "loss": 1.0962,
+      "step": 6629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5503487512421468,
+      "learning_rate": 8.36431368910035e-06,
+      "loss": 1.1243,
+      "step": 6630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.533359435524553,
+      "learning_rate": 8.361587190904124e-06,
+      "loss": 1.0984,
+      "step": 6631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4546129834508816,
+      "learning_rate": 8.358860817858044e-06,
+      "loss": 1.0784,
+      "step": 6632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2720515326047135,
+      "learning_rate": 8.356134570170363e-06,
+      "loss": 1.0652,
+      "step": 6633
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3285034811257015,
+      "learning_rate": 8.353408448049327e-06,
+      "loss": 1.0544,
+      "step": 6634
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8359530582374677,
+      "learning_rate": 8.350682451703167e-06,
+      "loss": 1.0942,
+      "step": 6635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4822512252948847,
+      "learning_rate": 8.347956581340115e-06,
+      "loss": 1.1675,
+      "step": 6636
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5152041659108773,
+      "learning_rate": 8.34523083716838e-06,
+      "loss": 1.1743,
+      "step": 6637
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.036380732749185,
+      "learning_rate": 8.342505219396167e-06,
+      "loss": 1.1763,
+      "step": 6638
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.867918634288548,
+      "learning_rate": 8.339779728231677e-06,
+      "loss": 1.0879,
+      "step": 6639
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.809581394591368,
+      "learning_rate": 8.337054363883095e-06,
+      "loss": 1.0639,
+      "step": 6640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3416300277497,
+      "learning_rate": 8.334329126558596e-06,
+      "loss": 1.1133,
+      "step": 6641
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.900646920252974,
+      "learning_rate": 8.331604016466352e-06,
+      "loss": 1.0698,
+      "step": 6642
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.955866758458451,
+      "learning_rate": 8.328879033814516e-06,
+      "loss": 0.9919,
+      "step": 6643
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7540266700551324,
+      "learning_rate": 8.32615417881124e-06,
+      "loss": 1.0524,
+      "step": 6644
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3244774971407387,
+      "learning_rate": 8.323429451664656e-06,
+      "loss": 1.116,
+      "step": 6645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2843095922172836,
+      "learning_rate": 8.320704852582894e-06,
+      "loss": 1.146,
+      "step": 6646
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.398406092792468,
+      "learning_rate": 8.317980381774076e-06,
+      "loss": 1.1924,
+      "step": 6647
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0734740506326266,
+      "learning_rate": 8.31525603944631e-06,
+      "loss": 1.1074,
+      "step": 6648
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.521991481252074,
+      "learning_rate": 8.312531825807696e-06,
+      "loss": 1.1165,
+      "step": 6649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1545019887661274,
+      "learning_rate": 8.309807741066316e-06,
+      "loss": 1.1372,
+      "step": 6650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7726677978325664,
+      "learning_rate": 8.307083785430252e-06,
+      "loss": 1.1182,
+      "step": 6651
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6758896884907717,
+      "learning_rate": 8.304359959107579e-06,
+      "loss": 1.0908,
+      "step": 6652
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.454762242669271,
+      "learning_rate": 8.301636262306351e-06,
+      "loss": 1.1292,
+      "step": 6653
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.876582016547146,
+      "learning_rate": 8.29891269523462e-06,
+      "loss": 1.1233,
+      "step": 6654
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3150966763567853,
+      "learning_rate": 8.296189258100423e-06,
+      "loss": 1.1514,
+      "step": 6655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.453016410092246,
+      "learning_rate": 8.293465951111796e-06,
+      "loss": 1.0969,
+      "step": 6656
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.820024456659637,
+      "learning_rate": 8.29074277447675e-06,
+      "loss": 1.0513,
+      "step": 6657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.839513894521536,
+      "learning_rate": 8.288019728403301e-06,
+      "loss": 1.1357,
+      "step": 6658
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.665521684583546,
+      "learning_rate": 8.285296813099445e-06,
+      "loss": 1.1282,
+      "step": 6659
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.157556042911744,
+      "learning_rate": 8.282574028773174e-06,
+      "loss": 1.1323,
+      "step": 6660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.212736454195327,
+      "learning_rate": 8.279851375632469e-06,
+      "loss": 1.1235,
+      "step": 6661
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.571374635033684,
+      "learning_rate": 8.277128853885297e-06,
+      "loss": 1.1265,
+      "step": 6662
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8623125077163243,
+      "learning_rate": 8.274406463739623e-06,
+      "loss": 1.0168,
+      "step": 6663
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.223296076808315,
+      "learning_rate": 8.271684205403388e-06,
+      "loss": 1.0522,
+      "step": 6664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1278709126368525,
+      "learning_rate": 8.268962079084537e-06,
+      "loss": 1.1729,
+      "step": 6665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.630435466952332,
+      "learning_rate": 8.266240084991e-06,
+      "loss": 1.0879,
+      "step": 6666
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0087653206826737,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 1.1777,
+      "step": 6667
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9877324264160317,
+      "learning_rate": 8.260796494311534e-06,
+      "loss": 1.0974,
+      "step": 6668
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8524355882658914,
+      "learning_rate": 8.258074898141418e-06,
+      "loss": 1.1707,
+      "step": 6669
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7560752432477464,
+      "learning_rate": 8.255353435028228e-06,
+      "loss": 1.1348,
+      "step": 6670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.984930760472495,
+      "learning_rate": 8.252632105179846e-06,
+      "loss": 1.1138,
+      "step": 6671
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9656255073680082,
+      "learning_rate": 8.249910908804144e-06,
+      "loss": 1.0791,
+      "step": 6672
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.002153670053462,
+      "learning_rate": 8.247189846108978e-06,
+      "loss": 1.0334,
+      "step": 6673
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.063123816001967,
+      "learning_rate": 8.244468917302197e-06,
+      "loss": 1.095,
+      "step": 6674
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2917877478646727,
+      "learning_rate": 8.24174812259164e-06,
+      "loss": 1.0786,
+      "step": 6675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2329166617109824,
+      "learning_rate": 8.239027462185134e-06,
+      "loss": 1.1479,
+      "step": 6676
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5877418140674995,
+      "learning_rate": 8.236306936290497e-06,
+      "loss": 1.1812,
+      "step": 6677
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.754400699616365,
+      "learning_rate": 8.233586545115534e-06,
+      "loss": 1.123,
+      "step": 6678
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0471371916908754,
+      "learning_rate": 8.230866288868045e-06,
+      "loss": 1.075,
+      "step": 6679
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.985225567047839,
+      "learning_rate": 8.228146167755816e-06,
+      "loss": 1.1357,
+      "step": 6680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9498380288423656,
+      "learning_rate": 8.225426181986621e-06,
+      "loss": 1.0852,
+      "step": 6681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.174840856734486,
+      "learning_rate": 8.222706331768231e-06,
+      "loss": 1.186,
+      "step": 6682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7133277927168886,
+      "learning_rate": 8.2199866173084e-06,
+      "loss": 1.0327,
+      "step": 6683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.205666307634374,
+      "learning_rate": 8.21726703881487e-06,
+      "loss": 1.0894,
+      "step": 6684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.036010626002657,
+      "learning_rate": 8.214547596495378e-06,
+      "loss": 1.1418,
+      "step": 6685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3187764718437665,
+      "learning_rate": 8.21182829055765e-06,
+      "loss": 1.1533,
+      "step": 6686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.428228838920585,
+      "learning_rate": 8.209109121209399e-06,
+      "loss": 1.0715,
+      "step": 6687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.00230733842308,
+      "learning_rate": 8.206390088658326e-06,
+      "loss": 1.0964,
+      "step": 6688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.111307413869796,
+      "learning_rate": 8.203671193112134e-06,
+      "loss": 1.0522,
+      "step": 6689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2069479081200813,
+      "learning_rate": 8.200952434778494e-06,
+      "loss": 1.1475,
+      "step": 6690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9850638143752297,
+      "learning_rate": 8.198233813865084e-06,
+      "loss": 1.0901,
+      "step": 6691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8164116613690564,
+      "learning_rate": 8.195515330579564e-06,
+      "loss": 1.1165,
+      "step": 6692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.882705325625878,
+      "learning_rate": 8.192796985129587e-06,
+      "loss": 1.0647,
+      "step": 6693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1954643327401193,
+      "learning_rate": 8.190078777722794e-06,
+      "loss": 1.0293,
+      "step": 6694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.666527749173423,
+      "learning_rate": 8.187360708566812e-06,
+      "loss": 1.1523,
+      "step": 6695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.80380442758091,
+      "learning_rate": 8.184642777869268e-06,
+      "loss": 1.0439,
+      "step": 6696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.267579812170848,
+      "learning_rate": 8.181924985837762e-06,
+      "loss": 1.1033,
+      "step": 6697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.635373658231087,
+      "learning_rate": 8.179207332679897e-06,
+      "loss": 1.1729,
+      "step": 6698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8708178018611106,
+      "learning_rate": 8.17648981860326e-06,
+      "loss": 1.0676,
+      "step": 6699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3558248129338044,
+      "learning_rate": 8.17377244381543e-06,
+      "loss": 1.1025,
+      "step": 6700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.826760805351081,
+      "learning_rate": 8.171055208523971e-06,
+      "loss": 1.1023,
+      "step": 6701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8003354769426205,
+      "learning_rate": 8.168338112936441e-06,
+      "loss": 1.0334,
+      "step": 6702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5845901238509184,
+      "learning_rate": 8.165621157260389e-06,
+      "loss": 1.1287,
+      "step": 6703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.740871418524929,
+      "learning_rate": 8.162904341703342e-06,
+      "loss": 1.1064,
+      "step": 6704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.176787691711668,
+      "learning_rate": 8.160187666472828e-06,
+      "loss": 1.166,
+      "step": 6705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.050598896922779,
+      "learning_rate": 8.157471131776356e-06,
+      "loss": 1.1685,
+      "step": 6706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4723302262532547,
+      "learning_rate": 8.154754737821435e-06,
+      "loss": 1.1814,
+      "step": 6707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5302652890492547,
+      "learning_rate": 8.152038484815555e-06,
+      "loss": 1.146,
+      "step": 6708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.777638652760471,
+      "learning_rate": 8.149322372966195e-06,
+      "loss": 1.1201,
+      "step": 6709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7473674900580125,
+      "learning_rate": 8.146606402480826e-06,
+      "loss": 1.0835,
+      "step": 6710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1780322533300454,
+      "learning_rate": 8.143890573566906e-06,
+      "loss": 1.1128,
+      "step": 6711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9478537092552997,
+      "learning_rate": 8.141174886431885e-06,
+      "loss": 1.0715,
+      "step": 6712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1494090886039787,
+      "learning_rate": 8.138459341283203e-06,
+      "loss": 1.1746,
+      "step": 6713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5020834634719114,
+      "learning_rate": 8.135743938328281e-06,
+      "loss": 1.1101,
+      "step": 6714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3562398845483306,
+      "learning_rate": 8.13302867777454e-06,
+      "loss": 1.0935,
+      "step": 6715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.08036733054516,
+      "learning_rate": 8.130313559829386e-06,
+      "loss": 1.0562,
+      "step": 6716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.452430499705982,
+      "learning_rate": 8.127598584700207e-06,
+      "loss": 1.1243,
+      "step": 6717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2547609937059914,
+      "learning_rate": 8.124883752594394e-06,
+      "loss": 1.1311,
+      "step": 6718
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9540790400354715,
+      "learning_rate": 8.122169063719312e-06,
+      "loss": 1.1636,
+      "step": 6719
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1683267957405428,
+      "learning_rate": 8.119454518282327e-06,
+      "loss": 1.1367,
+      "step": 6720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1189338722152966,
+      "learning_rate": 8.116740116490787e-06,
+      "loss": 1.1475,
+      "step": 6721
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.142408638159916,
+      "learning_rate": 8.114025858552039e-06,
+      "loss": 1.1096,
+      "step": 6722
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0105762440597825,
+      "learning_rate": 8.111311744673405e-06,
+      "loss": 1.1044,
+      "step": 6723
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.093496495575093,
+      "learning_rate": 8.108597775062198e-06,
+      "loss": 1.1013,
+      "step": 6724
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1818313951976793,
+      "learning_rate": 8.105883949925732e-06,
+      "loss": 1.1821,
+      "step": 6725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.343086329584558,
+      "learning_rate": 8.1031702694713e-06,
+      "loss": 1.1091,
+      "step": 6726
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.190616011104849,
+      "learning_rate": 8.100456733906189e-06,
+      "loss": 1.0815,
+      "step": 6727
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2312091763883277,
+      "learning_rate": 8.097743343437667e-06,
+      "loss": 1.1018,
+      "step": 6728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.645213222531406,
+      "learning_rate": 8.095030098273006e-06,
+      "loss": 1.1199,
+      "step": 6729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.21040706693877,
+      "learning_rate": 8.092316998619445e-06,
+      "loss": 1.1284,
+      "step": 6730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9869977379334918,
+      "learning_rate": 8.08960404468423e-06,
+      "loss": 1.0977,
+      "step": 6731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9245082251223806,
+      "learning_rate": 8.086891236674593e-06,
+      "loss": 1.1973,
+      "step": 6732
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4827653436839796,
+      "learning_rate": 8.084178574797745e-06,
+      "loss": 1.1013,
+      "step": 6733
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4478371621876365,
+      "learning_rate": 8.081466059260896e-06,
+      "loss": 1.0439,
+      "step": 6734
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.394394881024619,
+      "learning_rate": 8.078753690271243e-06,
+      "loss": 1.2031,
+      "step": 6735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7405687617807506,
+      "learning_rate": 8.076041468035972e-06,
+      "loss": 1.186,
+      "step": 6736
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.569431927508297,
+      "learning_rate": 8.073329392762251e-06,
+      "loss": 1.1216,
+      "step": 6737
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5101721089201394,
+      "learning_rate": 8.070617464657242e-06,
+      "loss": 1.1289,
+      "step": 6738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3504316658009468,
+      "learning_rate": 8.067905683928098e-06,
+      "loss": 1.1204,
+      "step": 6739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.104424498984993,
+      "learning_rate": 8.065194050781958e-06,
+      "loss": 1.0435,
+      "step": 6740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.041483026408871,
+      "learning_rate": 8.062482565425951e-06,
+      "loss": 1.0808,
+      "step": 6741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.252073493081616,
+      "learning_rate": 8.059771228067197e-06,
+      "loss": 1.0806,
+      "step": 6742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.396661319615005,
+      "learning_rate": 8.05706003891279e-06,
+      "loss": 1.1567,
+      "step": 6743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4298358985739457,
+      "learning_rate": 8.054348998169834e-06,
+      "loss": 1.1638,
+      "step": 6744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.286463845089549,
+      "learning_rate": 8.051638106045408e-06,
+      "loss": 1.127,
+      "step": 6745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2053150605314493,
+      "learning_rate": 8.048927362746585e-06,
+      "loss": 1.0107,
+      "step": 6746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.522261372470402,
+      "learning_rate": 8.046216768480424e-06,
+      "loss": 1.0938,
+      "step": 6747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2694874527548876,
+      "learning_rate": 8.043506323453976e-06,
+      "loss": 1.1826,
+      "step": 6748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1199130551926486,
+      "learning_rate": 8.040796027874279e-06,
+      "loss": 1.124,
+      "step": 6749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8856576479552016,
+      "learning_rate": 8.038085881948353e-06,
+      "loss": 1.1501,
+      "step": 6750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4107585418257527,
+      "learning_rate": 8.035375885883217e-06,
+      "loss": 1.1902,
+      "step": 6751
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.21208490998576,
+      "learning_rate": 8.032666039885872e-06,
+      "loss": 1.1101,
+      "step": 6752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.11757084573561,
+      "learning_rate": 8.02995634416331e-06,
+      "loss": 1.1306,
+      "step": 6753
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0411504064519455,
+      "learning_rate": 8.027246798922513e-06,
+      "loss": 1.1489,
+      "step": 6754
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.935967230296472,
+      "learning_rate": 8.024537404370451e-06,
+      "loss": 1.2109,
+      "step": 6755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2819329535931794,
+      "learning_rate": 8.02182816071408e-06,
+      "loss": 1.104,
+      "step": 6756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.210616564867456,
+      "learning_rate": 8.019119068160338e-06,
+      "loss": 1.1328,
+      "step": 6757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2786450541962213,
+      "learning_rate": 8.016410126916165e-06,
+      "loss": 1.1504,
+      "step": 6758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6942589437040936,
+      "learning_rate": 8.013701337188485e-06,
+      "loss": 1.0425,
+      "step": 6759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.759392363176913,
+      "learning_rate": 8.010992699184208e-06,
+      "loss": 1.0376,
+      "step": 6760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5561899279879996,
+      "learning_rate": 8.008284213110234e-06,
+      "loss": 1.0793,
+      "step": 6761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1286990932640713,
+      "learning_rate": 8.00557587917345e-06,
+      "loss": 1.1021,
+      "step": 6762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9636345717069874,
+      "learning_rate": 8.002867697580727e-06,
+      "loss": 1.0171,
+      "step": 6763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8396863257521296,
+      "learning_rate": 8.000159668538936e-06,
+      "loss": 1.0857,
+      "step": 6764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1777389049497744,
+      "learning_rate": 7.997451792254927e-06,
+      "loss": 1.0842,
+      "step": 6765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5574023838730278,
+      "learning_rate": 7.994744068935541e-06,
+      "loss": 1.1375,
+      "step": 6766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.386670717802947,
+      "learning_rate": 7.992036498787607e-06,
+      "loss": 1.0979,
+      "step": 6767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3078096353212603,
+      "learning_rate": 7.989329082017945e-06,
+      "loss": 1.1514,
+      "step": 6768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.167162819734683,
+      "learning_rate": 7.986621818833364e-06,
+      "loss": 1.0894,
+      "step": 6769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.02886472241367,
+      "learning_rate": 7.983914709440651e-06,
+      "loss": 1.0386,
+      "step": 6770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2246274120154457,
+      "learning_rate": 7.98120775404659e-06,
+      "loss": 1.1372,
+      "step": 6771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0468032469309043,
+      "learning_rate": 7.978500952857952e-06,
+      "loss": 1.0906,
+      "step": 6772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.46905701273463,
+      "learning_rate": 7.9757943060815e-06,
+      "loss": 1.0184,
+      "step": 6773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5101528128625814,
+      "learning_rate": 7.973087813923979e-06,
+      "loss": 1.134,
+      "step": 6774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.106857919378414,
+      "learning_rate": 7.970381476592124e-06,
+      "loss": 1.1157,
+      "step": 6775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3521707582778038,
+      "learning_rate": 7.967675294292658e-06,
+      "loss": 1.0735,
+      "step": 6776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.399681313674148,
+      "learning_rate": 7.964969267232289e-06,
+      "loss": 1.0659,
+      "step": 6777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.186468967266796,
+      "learning_rate": 7.962263395617724e-06,
+      "loss": 1.1477,
+      "step": 6778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1734616295412734,
+      "learning_rate": 7.959557679655649e-06,
+      "loss": 1.0935,
+      "step": 6779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8945074188302966,
+      "learning_rate": 7.956852119552738e-06,
+      "loss": 1.1274,
+      "step": 6780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.257987611453133,
+      "learning_rate": 7.954146715515654e-06,
+      "loss": 1.1208,
+      "step": 6781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1741228591810287,
+      "learning_rate": 7.951441467751057e-06,
+      "loss": 1.0698,
+      "step": 6782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5779190351674455,
+      "learning_rate": 7.948736376465576e-06,
+      "loss": 1.1362,
+      "step": 6783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4145685293242876,
+      "learning_rate": 7.94603144186585e-06,
+      "loss": 1.1318,
+      "step": 6784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.479740093172692,
+      "learning_rate": 7.943326664158487e-06,
+      "loss": 1.1831,
+      "step": 6785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0929433049058397,
+      "learning_rate": 7.940622043550094e-06,
+      "loss": 1.0808,
+      "step": 6786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4756356368822607,
+      "learning_rate": 7.937917580247265e-06,
+      "loss": 1.0762,
+      "step": 6787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2401398691820678,
+      "learning_rate": 7.935213274456582e-06,
+      "loss": 1.0708,
+      "step": 6788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.678527123638944,
+      "learning_rate": 7.932509126384612e-06,
+      "loss": 1.1172,
+      "step": 6789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5541210219107295,
+      "learning_rate": 7.92980513623791e-06,
+      "loss": 1.075,
+      "step": 6790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1930015682858444,
+      "learning_rate": 7.927101304223017e-06,
+      "loss": 1.052,
+      "step": 6791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8752100045178812,
+      "learning_rate": 7.92439763054647e-06,
+      "loss": 1.0989,
+      "step": 6792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1920181113793196,
+      "learning_rate": 7.92169411541479e-06,
+      "loss": 1.1096,
+      "step": 6793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1503587057195874,
+      "learning_rate": 7.918990759034482e-06,
+      "loss": 1.1047,
+      "step": 6794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5750049305327267,
+      "learning_rate": 7.916287561612045e-06,
+      "loss": 1.0121,
+      "step": 6795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.267322101156674,
+      "learning_rate": 7.913584523353958e-06,
+      "loss": 1.0647,
+      "step": 6796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9458596206130476,
+      "learning_rate": 7.910881644466694e-06,
+      "loss": 1.0916,
+      "step": 6797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.241686540516685,
+      "learning_rate": 7.908178925156715e-06,
+      "loss": 1.0706,
+      "step": 6798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.997881767481815,
+      "learning_rate": 7.905476365630466e-06,
+      "loss": 1.116,
+      "step": 6799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8306309508564653,
+      "learning_rate": 7.902773966094381e-06,
+      "loss": 1.1328,
+      "step": 6800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.27612518135932,
+      "learning_rate": 7.900071726754887e-06,
+      "loss": 1.0986,
+      "step": 6801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5019540691002136,
+      "learning_rate": 7.897369647818394e-06,
+      "loss": 1.1108,
+      "step": 6802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6668855759555683,
+      "learning_rate": 7.894667729491295e-06,
+      "loss": 1.1167,
+      "step": 6803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5551374059009295,
+      "learning_rate": 7.89196597197998e-06,
+      "loss": 1.1477,
+      "step": 6804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5685660935800523,
+      "learning_rate": 7.88926437549082e-06,
+      "loss": 1.1094,
+      "step": 6805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7707492320132827,
+      "learning_rate": 7.886562940230178e-06,
+      "loss": 1.0269,
+      "step": 6806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.934969133103018,
+      "learning_rate": 7.883861666404405e-06,
+      "loss": 1.0703,
+      "step": 6807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.162477621242089,
+      "learning_rate": 7.881160554219836e-06,
+      "loss": 1.166,
+      "step": 6808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.08652448383196,
+      "learning_rate": 7.878459603882799e-06,
+      "loss": 1.0117,
+      "step": 6809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.062130171903132,
+      "learning_rate": 7.875758815599597e-06,
+      "loss": 1.1504,
+      "step": 6810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9398805842929248,
+      "learning_rate": 7.873058189576537e-06,
+      "loss": 1.1389,
+      "step": 6811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3072659865237144,
+      "learning_rate": 7.870357726019907e-06,
+      "loss": 1.1106,
+      "step": 6812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3332757480978756,
+      "learning_rate": 7.867657425135977e-06,
+      "loss": 1.0713,
+      "step": 6813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.798031505736623,
+      "learning_rate": 7.864957287131012e-06,
+      "loss": 1.1074,
+      "step": 6814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6007163561146025,
+      "learning_rate": 7.862257312211266e-06,
+      "loss": 1.0381,
+      "step": 6815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6574735980770092,
+      "learning_rate": 7.859557500582968e-06,
+      "loss": 1.1106,
+      "step": 6816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.279815588144471,
+      "learning_rate": 7.85685785245235e-06,
+      "loss": 1.0588,
+      "step": 6817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8753482062263123,
+      "learning_rate": 7.85415836802562e-06,
+      "loss": 1.0879,
+      "step": 6818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.737345326339504,
+      "learning_rate": 7.85145904750898e-06,
+      "loss": 1.0542,
+      "step": 6819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0099930627801745,
+      "learning_rate": 7.84875989110862e-06,
+      "loss": 1.094,
+      "step": 6820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.028094016981147,
+      "learning_rate": 7.846060899030712e-06,
+      "loss": 1.0669,
+      "step": 6821
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.302633388665793,
+      "learning_rate": 7.843362071481422e-06,
+      "loss": 1.1716,
+      "step": 6822
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.043852701115334,
+      "learning_rate": 7.840663408666893e-06,
+      "loss": 1.0757,
+      "step": 6823
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3253005167973564,
+      "learning_rate": 7.837964910793268e-06,
+      "loss": 1.1631,
+      "step": 6824
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0162211789925166,
+      "learning_rate": 7.83526657806667e-06,
+      "loss": 1.0432,
+      "step": 6825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.079482471278982,
+      "learning_rate": 7.832568410693212e-06,
+      "loss": 1.1157,
+      "step": 6826
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4889273082595293,
+      "learning_rate": 7.829870408878993e-06,
+      "loss": 1.1338,
+      "step": 6827
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2232148577873265,
+      "learning_rate": 7.827172572830098e-06,
+      "loss": 1.0559,
+      "step": 6828
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8598401326132814,
+      "learning_rate": 7.824474902752609e-06,
+      "loss": 1.0955,
+      "step": 6829
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.992244244250966,
+      "learning_rate": 7.821777398852577e-06,
+      "loss": 1.1191,
+      "step": 6830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7099138172451998,
+      "learning_rate": 7.819080061336056e-06,
+      "loss": 1.1685,
+      "step": 6831
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4028140350984772,
+      "learning_rate": 7.81638289040908e-06,
+      "loss": 1.1467,
+      "step": 6832
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5585021527736367,
+      "learning_rate": 7.813685886277673e-06,
+      "loss": 1.0974,
+      "step": 6833
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.333907664606578,
+      "learning_rate": 7.81098904914785e-06,
+      "loss": 1.1287,
+      "step": 6834
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2181028125155144,
+      "learning_rate": 7.808292379225605e-06,
+      "loss": 1.1658,
+      "step": 6835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7353825906167004,
+      "learning_rate": 7.805595876716922e-06,
+      "loss": 1.1147,
+      "step": 6836
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2294747211109707,
+      "learning_rate": 7.802899541827772e-06,
+      "loss": 1.1682,
+      "step": 6837
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9531012942072885,
+      "learning_rate": 7.80020337476412e-06,
+      "loss": 1.1191,
+      "step": 6838
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.21789000883219,
+      "learning_rate": 7.797507375731911e-06,
+      "loss": 1.1145,
+      "step": 6839
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0954626762409636,
+      "learning_rate": 7.794811544937075e-06,
+      "loss": 1.1182,
+      "step": 6840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.909164398186331,
+      "learning_rate": 7.792115882585536e-06,
+      "loss": 1.1072,
+      "step": 6841
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.359675738407697,
+      "learning_rate": 7.789420388883206e-06,
+      "loss": 1.0847,
+      "step": 6842
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.828382412250276,
+      "learning_rate": 7.786725064035972e-06,
+      "loss": 1.0667,
+      "step": 6843
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.048139730533513,
+      "learning_rate": 7.784029908249722e-06,
+      "loss": 1.1255,
+      "step": 6844
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.966237788199103,
+      "learning_rate": 7.781334921730322e-06,
+      "loss": 1.092,
+      "step": 6845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2463147992453707,
+      "learning_rate": 7.778640104683631e-06,
+      "loss": 1.0676,
+      "step": 6846
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.996093296464089,
+      "learning_rate": 7.775945457315493e-06,
+      "loss": 1.0923,
+      "step": 6847
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.951727831784339,
+      "learning_rate": 7.77325097983174e-06,
+      "loss": 1.1807,
+      "step": 6848
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4475854087750397,
+      "learning_rate": 7.770556672438184e-06,
+      "loss": 1.1042,
+      "step": 6849
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.839542657416714,
+      "learning_rate": 7.767862535340635e-06,
+      "loss": 1.1172,
+      "step": 6850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5588364250136646,
+      "learning_rate": 7.765168568744881e-06,
+      "loss": 1.1055,
+      "step": 6851
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8747627482251756,
+      "learning_rate": 7.762474772856701e-06,
+      "loss": 1.1062,
+      "step": 6852
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.393943669561789,
+      "learning_rate": 7.759781147881864e-06,
+      "loss": 1.0994,
+      "step": 6853
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7720956978590348,
+      "learning_rate": 7.757087694026119e-06,
+      "loss": 1.1123,
+      "step": 6854
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9983990131612144,
+      "learning_rate": 7.75439441149521e-06,
+      "loss": 1.1025,
+      "step": 6855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.065751699500075,
+      "learning_rate": 7.751701300494857e-06,
+      "loss": 1.1077,
+      "step": 6856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.661162186695302,
+      "learning_rate": 7.749008361230775e-06,
+      "loss": 1.1091,
+      "step": 6857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.400713844976495,
+      "learning_rate": 7.746315593908667e-06,
+      "loss": 1.1089,
+      "step": 6858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.239740635381409,
+      "learning_rate": 7.743622998734217e-06,
+      "loss": 1.1541,
+      "step": 6859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.944585233934385,
+      "learning_rate": 7.740930575913101e-06,
+      "loss": 1.0979,
+      "step": 6860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.008638102840455,
+      "learning_rate": 7.73823832565098e-06,
+      "loss": 1.1038,
+      "step": 6861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.085789878775509,
+      "learning_rate": 7.735546248153503e-06,
+      "loss": 1.1433,
+      "step": 6862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.065342450219618,
+      "learning_rate": 7.732854343626299e-06,
+      "loss": 1.1208,
+      "step": 6863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9714393770843412,
+      "learning_rate": 7.730162612274991e-06,
+      "loss": 1.1309,
+      "step": 6864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.497503267571057,
+      "learning_rate": 7.727471054305187e-06,
+      "loss": 1.1733,
+      "step": 6865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5406509528956547,
+      "learning_rate": 7.724779669922485e-06,
+      "loss": 1.0485,
+      "step": 6866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2202344462887686,
+      "learning_rate": 7.722088459332463e-06,
+      "loss": 1.1743,
+      "step": 6867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.241958617258418,
+      "learning_rate": 7.71939742274069e-06,
+      "loss": 1.0952,
+      "step": 6868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.786413609394251,
+      "learning_rate": 7.716706560352722e-06,
+      "loss": 1.1365,
+      "step": 6869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2894296863061965,
+      "learning_rate": 7.714015872374096e-06,
+      "loss": 1.127,
+      "step": 6870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3682662266392676,
+      "learning_rate": 7.711325359010344e-06,
+      "loss": 1.0454,
+      "step": 6871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.954688543782294,
+      "learning_rate": 7.708635020466983e-06,
+      "loss": 1.1279,
+      "step": 6872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.920435811981246,
+      "learning_rate": 7.705944856949508e-06,
+      "loss": 1.1135,
+      "step": 6873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6590248428416796,
+      "learning_rate": 7.703254868663413e-06,
+      "loss": 1.1399,
+      "step": 6874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.849439708266346,
+      "learning_rate": 7.700565055814172e-06,
+      "loss": 1.1489,
+      "step": 6875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.931953627285617,
+      "learning_rate": 7.697875418607241e-06,
+      "loss": 1.2292,
+      "step": 6876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0929927862572253,
+      "learning_rate": 7.695185957248074e-06,
+      "loss": 1.095,
+      "step": 6877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.773581326962724,
+      "learning_rate": 7.692496671942103e-06,
+      "loss": 1.1328,
+      "step": 6878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.612331889853947,
+      "learning_rate": 7.689807562894747e-06,
+      "loss": 1.1772,
+      "step": 6879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4240278366932735,
+      "learning_rate": 7.687118630311418e-06,
+      "loss": 1.1575,
+      "step": 6880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5181960522114233,
+      "learning_rate": 7.684429874397507e-06,
+      "loss": 1.1079,
+      "step": 6881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.132811950448774,
+      "learning_rate": 7.6817412953584e-06,
+      "loss": 1.0938,
+      "step": 6882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.063820278337731,
+      "learning_rate": 7.679052893399454e-06,
+      "loss": 1.1836,
+      "step": 6883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8557937834335716,
+      "learning_rate": 7.67636466872603e-06,
+      "loss": 1.0847,
+      "step": 6884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.75665206148515,
+      "learning_rate": 7.673676621543464e-06,
+      "loss": 1.0969,
+      "step": 6885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.738979686192477,
+      "learning_rate": 7.670988752057088e-06,
+      "loss": 1.1042,
+      "step": 6886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7966572634159417,
+      "learning_rate": 7.668301060472208e-06,
+      "loss": 1.095,
+      "step": 6887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.245989686691802,
+      "learning_rate": 7.665613546994133e-06,
+      "loss": 1.1438,
+      "step": 6888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.929937210121423,
+      "learning_rate": 7.662926211828136e-06,
+      "loss": 1.0739,
+      "step": 6889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8035037090768724,
+      "learning_rate": 7.660239055179497e-06,
+      "loss": 1.1414,
+      "step": 6890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9450145558181555,
+      "learning_rate": 7.657552077253474e-06,
+      "loss": 1.1235,
+      "step": 6891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.221220794729875,
+      "learning_rate": 7.65486527825531e-06,
+      "loss": 1.1277,
+      "step": 6892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7066385755353632,
+      "learning_rate": 7.652178658390235e-06,
+      "loss": 1.0459,
+      "step": 6893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.071876807748315,
+      "learning_rate": 7.64949221786347e-06,
+      "loss": 1.073,
+      "step": 6894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.356999850792698,
+      "learning_rate": 7.646805956880222e-06,
+      "loss": 1.1355,
+      "step": 6895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.152225910965835,
+      "learning_rate": 7.644119875645672e-06,
+      "loss": 1.1106,
+      "step": 6896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7149367598683263,
+      "learning_rate": 7.641433974364998e-06,
+      "loss": 1.0994,
+      "step": 6897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.253125815404035,
+      "learning_rate": 7.638748253243366e-06,
+      "loss": 1.1377,
+      "step": 6898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.229977598354826,
+      "learning_rate": 7.636062712485922e-06,
+      "loss": 1.115,
+      "step": 6899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.947674262547604,
+      "learning_rate": 7.633377352297805e-06,
+      "loss": 1.0864,
+      "step": 6900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2304146632691104,
+      "learning_rate": 7.630692172884137e-06,
+      "loss": 1.1758,
+      "step": 6901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.014716156574906,
+      "learning_rate": 7.628007174450018e-06,
+      "loss": 1.1438,
+      "step": 6902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4393069744874922,
+      "learning_rate": 7.625322357200546e-06,
+      "loss": 1.1375,
+      "step": 6903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2496560201554026,
+      "learning_rate": 7.622637721340801e-06,
+      "loss": 1.144,
+      "step": 6904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3909992709675665,
+      "learning_rate": 7.619953267075849e-06,
+      "loss": 1.179,
+      "step": 6905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0173851894376953,
+      "learning_rate": 7.61726899461074e-06,
+      "loss": 1.1155,
+      "step": 6906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2211573750783495,
+      "learning_rate": 7.614584904150514e-06,
+      "loss": 1.1064,
+      "step": 6907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2996076815474256,
+      "learning_rate": 7.611900995900199e-06,
+      "loss": 1.1475,
+      "step": 6908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1970539344081677,
+      "learning_rate": 7.609217270064796e-06,
+      "loss": 1.1719,
+      "step": 6909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9167296955913598,
+      "learning_rate": 7.606533726849309e-06,
+      "loss": 1.0637,
+      "step": 6910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.386169701342458,
+      "learning_rate": 7.603850366458716e-06,
+      "loss": 1.064,
+      "step": 6911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5065486666672157,
+      "learning_rate": 7.601167189097987e-06,
+      "loss": 1.0911,
+      "step": 6912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5307496111143504,
+      "learning_rate": 7.598484194972076e-06,
+      "loss": 1.1567,
+      "step": 6913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2568813639651197,
+      "learning_rate": 7.595801384285928e-06,
+      "loss": 1.1052,
+      "step": 6914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0224312518916,
+      "learning_rate": 7.593118757244468e-06,
+      "loss": 1.1113,
+      "step": 6915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.920869293931577,
+      "learning_rate": 7.5904363140526e-06,
+      "loss": 1.1477,
+      "step": 6916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4744930057541183,
+      "learning_rate": 7.587754054915231e-06,
+      "loss": 1.1416,
+      "step": 6917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1588327677241206,
+      "learning_rate": 7.585071980037241e-06,
+      "loss": 1.0547,
+      "step": 6918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5191081534594053,
+      "learning_rate": 7.582390089623505e-06,
+      "loss": 1.0298,
+      "step": 6919
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.283490702795385,
+      "learning_rate": 7.579708383878875e-06,
+      "loss": 1.0952,
+      "step": 6920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2327023570312186,
+      "learning_rate": 7.577026863008197e-06,
+      "loss": 1.1213,
+      "step": 6921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0865534566806128,
+      "learning_rate": 7.574345527216293e-06,
+      "loss": 1.1238,
+      "step": 6922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.418653202735112,
+      "learning_rate": 7.57166437670798e-06,
+      "loss": 1.147,
+      "step": 6923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3467756079654545,
+      "learning_rate": 7.56898341168806e-06,
+      "loss": 1.1169,
+      "step": 6924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.128831028331747,
+      "learning_rate": 7.566302632361314e-06,
+      "loss": 1.1294,
+      "step": 6925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7811267250680283,
+      "learning_rate": 7.563622038932517e-06,
+      "loss": 1.061,
+      "step": 6926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3011123513734835,
+      "learning_rate": 7.560941631606425e-06,
+      "loss": 1.0786,
+      "step": 6927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.833318920157639,
+      "learning_rate": 7.558261410587781e-06,
+      "loss": 1.1128,
+      "step": 6928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.418191962310542,
+      "learning_rate": 7.555581376081314e-06,
+      "loss": 1.1096,
+      "step": 6929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3948080402163567,
+      "learning_rate": 7.552901528291735e-06,
+      "loss": 1.1155,
+      "step": 6930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.592722697857527,
+      "learning_rate": 7.550221867423746e-06,
+      "loss": 1.135,
+      "step": 6931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8037527559397755,
+      "learning_rate": 7.547542393682034e-06,
+      "loss": 1.0522,
+      "step": 6932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0965748351585325,
+      "learning_rate": 7.544863107271274e-06,
+      "loss": 1.155,
+      "step": 6933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.068598366280147,
+      "learning_rate": 7.542184008396116e-06,
+      "loss": 1.126,
+      "step": 6934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.485390204132317,
+      "learning_rate": 7.539505097261212e-06,
+      "loss": 1.0901,
+      "step": 6935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.388335694723752,
+      "learning_rate": 7.536826374071179e-06,
+      "loss": 1.1648,
+      "step": 6936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.120778029339471,
+      "learning_rate": 7.534147839030638e-06,
+      "loss": 1.0837,
+      "step": 6937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3602662402997177,
+      "learning_rate": 7.53146949234419e-06,
+      "loss": 1.1196,
+      "step": 6938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.041026263702271,
+      "learning_rate": 7.528791334216417e-06,
+      "loss": 1.062,
+      "step": 6939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1043871696519525,
+      "learning_rate": 7.526113364851891e-06,
+      "loss": 1.0679,
+      "step": 6940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.006329322475275,
+      "learning_rate": 7.5234355844551745e-06,
+      "loss": 1.1338,
+      "step": 6941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1713753126671986,
+      "learning_rate": 7.520757993230801e-06,
+      "loss": 1.0933,
+      "step": 6942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3221819927634004,
+      "learning_rate": 7.518080591383302e-06,
+      "loss": 1.1709,
+      "step": 6943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.970818621038058,
+      "learning_rate": 7.51540337911719e-06,
+      "loss": 1.1523,
+      "step": 6944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.283151786029509,
+      "learning_rate": 7.512726356636963e-06,
+      "loss": 1.1038,
+      "step": 6945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.212164099562416,
+      "learning_rate": 7.510049524147111e-06,
+      "loss": 1.1082,
+      "step": 6946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5417111845117013,
+      "learning_rate": 7.507372881852096e-06,
+      "loss": 1.1421,
+      "step": 6947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.016341787133324,
+      "learning_rate": 7.5046964299563815e-06,
+      "loss": 1.0723,
+      "step": 6948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.754868785321434,
+      "learning_rate": 7.5020201686644e-06,
+      "loss": 1.0549,
+      "step": 6949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.545481595600339,
+      "learning_rate": 7.499344098180581e-06,
+      "loss": 1.0605,
+      "step": 6950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0351652331814574,
+      "learning_rate": 7.496668218709338e-06,
+      "loss": 1.0447,
+      "step": 6951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.950816099583731,
+      "learning_rate": 7.493992530455068e-06,
+      "loss": 1.1323,
+      "step": 6952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.33858192836958,
+      "learning_rate": 7.491317033622147e-06,
+      "loss": 1.0532,
+      "step": 6953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.831999333382674,
+      "learning_rate": 7.488641728414952e-06,
+      "loss": 1.1067,
+      "step": 6954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8431212771983687,
+      "learning_rate": 7.485966615037834e-06,
+      "loss": 1.0374,
+      "step": 6955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.362643474381588,
+      "learning_rate": 7.483291693695125e-06,
+      "loss": 1.1555,
+      "step": 6956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.49250550530305,
+      "learning_rate": 7.4806169645911554e-06,
+      "loss": 1.0779,
+      "step": 6957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3682378256508283,
+      "learning_rate": 7.47794242793023e-06,
+      "loss": 1.0867,
+      "step": 6958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2722085275127863,
+      "learning_rate": 7.475268083916647e-06,
+      "loss": 1.0769,
+      "step": 6959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3360531883760767,
+      "learning_rate": 7.472593932754686e-06,
+      "loss": 1.0918,
+      "step": 6960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.921881170189814,
+      "learning_rate": 7.46991997464861e-06,
+      "loss": 1.1001,
+      "step": 6961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6114462222683064,
+      "learning_rate": 7.46724620980267e-06,
+      "loss": 1.073,
+      "step": 6962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5236092275017636,
+      "learning_rate": 7.4645726384211e-06,
+      "loss": 1.1355,
+      "step": 6963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2464030865232334,
+      "learning_rate": 7.461899260708121e-06,
+      "loss": 1.1206,
+      "step": 6964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.694680963694332,
+      "learning_rate": 7.459226076867943e-06,
+      "loss": 1.0298,
+      "step": 6965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.887524662737833,
+      "learning_rate": 7.456553087104752e-06,
+      "loss": 1.1594,
+      "step": 6966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9280127436891537,
+      "learning_rate": 7.453880291622726e-06,
+      "loss": 1.092,
+      "step": 6967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.232368973734168,
+      "learning_rate": 7.4512076906260315e-06,
+      "loss": 1.0974,
+      "step": 6968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8253415808712856,
+      "learning_rate": 7.448535284318805e-06,
+      "loss": 1.115,
+      "step": 6969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8467542208805847,
+      "learning_rate": 7.445863072905186e-06,
+      "loss": 1.0898,
+      "step": 6970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.845763895888738,
+      "learning_rate": 7.443191056589289e-06,
+      "loss": 1.0798,
+      "step": 6971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5181081854191283,
+      "learning_rate": 7.4405192355752164e-06,
+      "loss": 1.0615,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.130717381840733,
+      "learning_rate": 7.437847610067053e-06,
+      "loss": 1.0613,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.915437990236704,
+      "learning_rate": 7.435176180268877e-06,
+      "loss": 1.1248,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.034878818432788,
+      "learning_rate": 7.432504946384738e-06,
+      "loss": 1.105,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0780697729268436,
+      "learning_rate": 7.429833908618682e-06,
+      "loss": 1.0918,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.449487894229364,
+      "learning_rate": 7.427163067174735e-06,
+      "loss": 1.1079,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.311637760574991,
+      "learning_rate": 7.424492422256911e-06,
+      "loss": 1.1387,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.522633075011297,
+      "learning_rate": 7.421821974069205e-06,
+      "loss": 1.1301,
+      "step": 6979
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2455652501882875,
+      "learning_rate": 7.4191517228156e-06,
+      "loss": 1.0112,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.137223650817785,
+      "learning_rate": 7.416481668700066e-06,
+      "loss": 1.125,
+      "step": 6981
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.286378293979055,
+      "learning_rate": 7.41381181192655e-06,
+      "loss": 1.1072,
+      "step": 6982
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2917324160005497,
+      "learning_rate": 7.411142152698989e-06,
+      "loss": 1.1113,
+      "step": 6983
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.656489451831697,
+      "learning_rate": 7.4084726912213115e-06,
+      "loss": 1.0583,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0174198338277476,
+      "learning_rate": 7.405803427697417e-06,
+      "loss": 1.0786,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9899763739438967,
+      "learning_rate": 7.4031343623312e-06,
+      "loss": 1.0811,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3564777413628244,
+      "learning_rate": 7.400465495326537e-06,
+      "loss": 1.1252,
+      "step": 6987
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.575512144704941,
+      "learning_rate": 7.397796826887295e-06,
+      "loss": 1.1833,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4967598524291907,
+      "learning_rate": 7.39512835721731e-06,
+      "loss": 1.1199,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2794520428448295,
+      "learning_rate": 7.392460086520419e-06,
+      "loss": 1.1331,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5463982223046853,
+      "learning_rate": 7.389792015000434e-06,
+      "loss": 1.0676,
+      "step": 6991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.333309648629573,
+      "learning_rate": 7.3871241428611594e-06,
+      "loss": 1.1006,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.417505778428664,
+      "learning_rate": 7.384456470306383e-06,
+      "loss": 1.1589,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.906932464513785,
+      "learning_rate": 7.381788997539869e-06,
+      "loss": 1.1633,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.004033247528878,
+      "learning_rate": 7.3791217247653746e-06,
+      "loss": 1.0496,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8100233102035164,
+      "learning_rate": 7.3764546521866385e-06,
+      "loss": 1.1392,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9479967868399446,
+      "learning_rate": 7.373787780007387e-06,
+      "loss": 1.0793,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.431589633101973,
+      "learning_rate": 7.371121108431327e-06,
+      "loss": 0.9775,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7200717884985526,
+      "learning_rate": 7.368454637662153e-06,
+      "loss": 1.1038,
+      "step": 6999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.003579139881314,
+      "learning_rate": 7.365788367903543e-06,
+      "loss": 1.1738,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1948131514883302,
+      "learning_rate": 7.363122299359167e-06,
+      "loss": 1.1606,
+      "step": 7001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2264506546360163,
+      "learning_rate": 7.36045643223266e-06,
+      "loss": 1.0449,
+      "step": 7002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9498595953039466,
+      "learning_rate": 7.357790766727663e-06,
+      "loss": 1.1138,
+      "step": 7003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9857339543915122,
+      "learning_rate": 7.355125303047788e-06,
+      "loss": 1.0823,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3956719969901714,
+      "learning_rate": 7.352460041396638e-06,
+      "loss": 1.1257,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2250112189174702,
+      "learning_rate": 7.3497949819778e-06,
+      "loss": 1.1692,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8618169868180967,
+      "learning_rate": 7.347130124994847e-06,
+      "loss": 1.085,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2642086476252787,
+      "learning_rate": 7.344465470651333e-06,
+      "loss": 1.1487,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1391609494637307,
+      "learning_rate": 7.3418010191507915e-06,
+      "loss": 1.1492,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.105651744063364,
+      "learning_rate": 7.339136770696751e-06,
+      "loss": 1.0811,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7413799767478784,
+      "learning_rate": 7.336472725492721e-06,
+      "loss": 1.1553,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.069883961116062,
+      "learning_rate": 7.333808883742192e-06,
+      "loss": 1.1155,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.194715019133972,
+      "learning_rate": 7.331145245648646e-06,
+      "loss": 1.0933,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.191028591593832,
+      "learning_rate": 7.328481811415543e-06,
+      "loss": 1.0834,
+      "step": 7014
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.796885464757916,
+      "learning_rate": 7.325818581246325e-06,
+      "loss": 1.1094,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.206122838237207,
+      "learning_rate": 7.323155555344428e-06,
+      "loss": 1.094,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6673385682577186,
+      "learning_rate": 7.320492733913264e-06,
+      "loss": 1.0701,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.042897899987932,
+      "learning_rate": 7.317830117156235e-06,
+      "loss": 1.126,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0651963944268044,
+      "learning_rate": 7.315167705276725e-06,
+      "loss": 1.1008,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5418770867107736,
+      "learning_rate": 7.3125054984780985e-06,
+      "loss": 1.1294,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.010880976059654,
+      "learning_rate": 7.309843496963715e-06,
+      "loss": 1.1089,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8575413239596714,
+      "learning_rate": 7.307181700936908e-06,
+      "loss": 1.1924,
+      "step": 7022
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5501144049094107,
+      "learning_rate": 7.304520110600994e-06,
+      "loss": 1.1111,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2059115875522357,
+      "learning_rate": 7.301858726159284e-06,
+      "loss": 1.0894,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7314156824590854,
+      "learning_rate": 7.299197547815067e-06,
+      "loss": 1.05,
+      "step": 7025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6103721050830595,
+      "learning_rate": 7.296536575771616e-06,
+      "loss": 1.1255,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.883123533121836,
+      "learning_rate": 7.293875810232195e-06,
+      "loss": 1.0566,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9331262320889664,
+      "learning_rate": 7.291215251400037e-06,
+      "loss": 1.0378,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5394766100283372,
+      "learning_rate": 7.28855489947837e-06,
+      "loss": 1.1338,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.831239327745213,
+      "learning_rate": 7.285894754670413e-06,
+      "loss": 1.052,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.247264720575213,
+      "learning_rate": 7.283234817179355e-06,
+      "loss": 1.085,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.079287440127535,
+      "learning_rate": 7.280575087208376e-06,
+      "loss": 1.1162,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.308887072214338,
+      "learning_rate": 7.27791556496064e-06,
+      "loss": 1.094,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.243989305050347,
+      "learning_rate": 7.275256250639299e-06,
+      "loss": 1.1675,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1937070022376637,
+      "learning_rate": 7.272597144447475e-06,
+      "loss": 1.0791,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6210004744441058,
+      "learning_rate": 7.269938246588293e-06,
+      "loss": 1.0288,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.862142615543731,
+      "learning_rate": 7.267279557264846e-06,
+      "loss": 1.0649,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.241889662879294,
+      "learning_rate": 7.2646210766802205e-06,
+      "loss": 1.1128,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1081065173570885,
+      "learning_rate": 7.261962805037486e-06,
+      "loss": 1.1245,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4031367484207493,
+      "learning_rate": 7.259304742539695e-06,
+      "loss": 1.0867,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3569360377216113,
+      "learning_rate": 7.2566468893898845e-06,
+      "loss": 1.1091,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.322716751984024,
+      "learning_rate": 7.253989245791069e-06,
+      "loss": 1.1606,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.409297949532627,
+      "learning_rate": 7.251331811946255e-06,
+      "loss": 1.0691,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7768421114295143,
+      "learning_rate": 7.248674588058431e-06,
+      "loss": 1.1687,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2162202524267003,
+      "learning_rate": 7.2460175743305726e-06,
+      "loss": 1.0771,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1298156404502957,
+      "learning_rate": 7.243360770965629e-06,
+      "loss": 1.1191,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6683385350802533,
+      "learning_rate": 7.240704178166549e-06,
+      "loss": 1.1125,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.011272137240566,
+      "learning_rate": 7.238047796136247e-06,
+      "loss": 1.0874,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.028721349512191,
+      "learning_rate": 7.235391625077635e-06,
+      "loss": 1.134,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.632476453081893,
+      "learning_rate": 7.232735665193604e-06,
+      "loss": 1.1382,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.173096925503809,
+      "learning_rate": 7.230079916687029e-06,
+      "loss": 1.0801,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.10818891301489,
+      "learning_rate": 7.227424379760772e-06,
+      "loss": 1.1602,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2762642480395496,
+      "learning_rate": 7.224769054617673e-06,
+      "loss": 1.1179,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.139781788448502,
+      "learning_rate": 7.222113941460564e-06,
+      "loss": 1.105,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.887566296906344,
+      "learning_rate": 7.219459040492251e-06,
+      "loss": 1.1819,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.178555896892318,
+      "learning_rate": 7.216804351915528e-06,
+      "loss": 1.1338,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.058115520305459,
+      "learning_rate": 7.214149875933173e-06,
+      "loss": 1.0725,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8190835742829,
+      "learning_rate": 7.211495612747952e-06,
+      "loss": 1.1558,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.653487766037637,
+      "learning_rate": 7.2088415625626096e-06,
+      "loss": 1.1157,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7756923992422093,
+      "learning_rate": 7.206187725579874e-06,
+      "loss": 1.1809,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.794517211019732,
+      "learning_rate": 7.203534102002463e-06,
+      "loss": 1.1182,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1948091951011457,
+      "learning_rate": 7.200880692033065e-06,
+      "loss": 1.1326,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1141214047783357,
+      "learning_rate": 7.198227495874368e-06,
+      "loss": 1.0869,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1096411556959986,
+      "learning_rate": 7.195574513729033e-06,
+      "loss": 1.1167,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.85081481970346,
+      "learning_rate": 7.19292174579971e-06,
+      "loss": 1.1729,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7719052545148726,
+      "learning_rate": 7.190269192289028e-06,
+      "loss": 1.0615,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.242870832584467,
+      "learning_rate": 7.187616853399609e-06,
+      "loss": 1.0879,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.202799387917057,
+      "learning_rate": 7.184964729334041e-06,
+      "loss": 1.1379,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.826552902509043,
+      "learning_rate": 7.182312820294916e-06,
+      "loss": 1.0728,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1616525259627943,
+      "learning_rate": 7.179661126484794e-06,
+      "loss": 1.0056,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1762086041447053,
+      "learning_rate": 7.177009648106226e-06,
+      "loss": 1.0146,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.193738767050442,
+      "learning_rate": 7.174358385361746e-06,
+      "loss": 1.073,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6301491970981896,
+      "learning_rate": 7.171707338453873e-06,
+      "loss": 1.0437,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8190864564657963,
+      "learning_rate": 7.169056507585108e-06,
+      "loss": 1.1108,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8962305884452664,
+      "learning_rate": 7.166405892957926e-06,
+      "loss": 1.1343,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7872906730248883,
+      "learning_rate": 7.163755494774802e-06,
+      "loss": 1.0522,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.433219578102207,
+      "learning_rate": 7.161105313238182e-06,
+      "loss": 1.0959,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1157335283953964,
+      "learning_rate": 7.158455348550506e-06,
+      "loss": 1.0947,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.751845228240507,
+      "learning_rate": 7.155805600914185e-06,
+      "loss": 1.0703,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9647507524166326,
+      "learning_rate": 7.153156070531624e-06,
+      "loss": 1.1035,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5724221266184304,
+      "learning_rate": 7.150506757605211e-06,
+      "loss": 1.0503,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.335782185385116,
+      "learning_rate": 7.147857662337304e-06,
+      "loss": 1.0801,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0876854073435998,
+      "learning_rate": 7.145208784930262e-06,
+      "loss": 1.1406,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6570536312833526,
+      "learning_rate": 7.142560125586413e-06,
+      "loss": 1.1145,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.548332527077511,
+      "learning_rate": 7.13991168450808e-06,
+      "loss": 1.1704,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3603127031648614,
+      "learning_rate": 7.137263461897563e-06,
+      "loss": 1.1292,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0759351602182043,
+      "learning_rate": 7.134615457957148e-06,
+      "loss": 1.0537,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1881650782760356,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 1.0627,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1437384268629898,
+      "learning_rate": 7.12932010689567e-06,
+      "loss": 1.0605,
+      "step": 7089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9828252507082627,
+      "learning_rate": 7.126672760179092e-06,
+      "loss": 1.0938,
+      "step": 7090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.449547879785724,
+      "learning_rate": 7.124025632941584e-06,
+      "loss": 1.0881,
+      "step": 7091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1555404122889614,
+      "learning_rate": 7.121378725385351e-06,
+      "loss": 1.051,
+      "step": 7092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.426892051964605,
+      "learning_rate": 7.11873203771257e-06,
+      "loss": 1.106,
+      "step": 7093
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1703101454775884,
+      "learning_rate": 7.116085570125417e-06,
+      "loss": 1.1562,
+      "step": 7094
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3664621099528214,
+      "learning_rate": 7.113439322826032e-06,
+      "loss": 1.0808,
+      "step": 7095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0845303431182005,
+      "learning_rate": 7.110793296016556e-06,
+      "loss": 1.1499,
+      "step": 7096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9175038822581767,
+      "learning_rate": 7.108147489899103e-06,
+      "loss": 1.0901,
+      "step": 7097
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.060800265118469,
+      "learning_rate": 7.105501904675772e-06,
+      "loss": 1.1306,
+      "step": 7098
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5850275064067385,
+      "learning_rate": 7.102856540548648e-06,
+      "loss": 1.0715,
+      "step": 7099
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3789753212236056,
+      "learning_rate": 7.100211397719799e-06,
+      "loss": 1.1321,
+      "step": 7100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3929788699230836,
+      "learning_rate": 7.097566476391269e-06,
+      "loss": 1.1021,
+      "step": 7101
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7231279258983085,
+      "learning_rate": 7.094921776765095e-06,
+      "loss": 1.0906,
+      "step": 7102
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.19424083199904,
+      "learning_rate": 7.0922772990432865e-06,
+      "loss": 1.0674,
+      "step": 7103
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0970606094425346,
+      "learning_rate": 7.089633043427849e-06,
+      "loss": 1.0693,
+      "step": 7104
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1231641240443353,
+      "learning_rate": 7.08698901012076e-06,
+      "loss": 1.1482,
+      "step": 7105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.796140953155372,
+      "learning_rate": 7.084345199323982e-06,
+      "loss": 1.0457,
+      "step": 7106
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0234125196979904,
+      "learning_rate": 7.08170161123947e-06,
+      "loss": 1.1121,
+      "step": 7107
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3595529400335473,
+      "learning_rate": 7.079058246069146e-06,
+      "loss": 1.0964,
+      "step": 7108
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.381894884138284,
+      "learning_rate": 7.076415104014925e-06,
+      "loss": 1.0742,
+      "step": 7109
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7839998870743683,
+      "learning_rate": 7.073772185278707e-06,
+      "loss": 1.1624,
+      "step": 7110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8822182245360746,
+      "learning_rate": 7.071129490062372e-06,
+      "loss": 1.0767,
+      "step": 7111
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.065250039741713,
+      "learning_rate": 7.068487018567776e-06,
+      "loss": 1.113,
+      "step": 7112
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9950676743849223,
+      "learning_rate": 7.0658447709967695e-06,
+      "loss": 1.1333,
+      "step": 7113
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8023578669744684,
+      "learning_rate": 7.0632027475511824e-06,
+      "loss": 1.1506,
+      "step": 7114
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3754518201512305,
+      "learning_rate": 7.060560948432818e-06,
+      "loss": 1.1208,
+      "step": 7115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7770156325820365,
+      "learning_rate": 7.057919373843477e-06,
+      "loss": 1.1379,
+      "step": 7116
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5577576635741033,
+      "learning_rate": 7.05527802398493e-06,
+      "loss": 1.124,
+      "step": 7117
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2037442911248277,
+      "learning_rate": 7.052636899058943e-06,
+      "loss": 1.0864,
+      "step": 7118
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.766840193332354,
+      "learning_rate": 7.049995999267254e-06,
+      "loss": 1.1646,
+      "step": 7119
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1683188531885396,
+      "learning_rate": 7.047355324811591e-06,
+      "loss": 1.1497,
+      "step": 7120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0397889371276525,
+      "learning_rate": 7.044714875893659e-06,
+      "loss": 1.0601,
+      "step": 7121
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7918417335501364,
+      "learning_rate": 7.04207465271515e-06,
+      "loss": 1.1262,
+      "step": 7122
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.3546380379380425,
+      "learning_rate": 7.039434655477735e-06,
+      "loss": 1.1299,
+      "step": 7123
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.106662179643568,
+      "learning_rate": 7.0367948843830755e-06,
+      "loss": 1.1274,
+      "step": 7124
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.157059459396818,
+      "learning_rate": 7.034155339632804e-06,
+      "loss": 1.1272,
+      "step": 7125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5838167291665077,
+      "learning_rate": 7.031516021428546e-06,
+      "loss": 1.1084,
+      "step": 7126
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.165752850376669,
+      "learning_rate": 7.02887692997191e-06,
+      "loss": 1.0957,
+      "step": 7127
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9334668874881586,
+      "learning_rate": 7.026238065464475e-06,
+      "loss": 1.0918,
+      "step": 7128
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2712865131952893,
+      "learning_rate": 7.023599428107815e-06,
+      "loss": 1.042,
+      "step": 7129
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.964901484341568,
+      "learning_rate": 7.020961018103478e-06,
+      "loss": 1.1577,
+      "step": 7130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.544525063980999,
+      "learning_rate": 7.018322835653002e-06,
+      "loss": 1.1187,
+      "step": 7131
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9419616553872423,
+      "learning_rate": 7.015684880957906e-06,
+      "loss": 1.1328,
+      "step": 7132
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.710590150832356,
+      "learning_rate": 7.01304715421969e-06,
+      "loss": 1.1907,
+      "step": 7133
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4771641609789046,
+      "learning_rate": 7.010409655639836e-06,
+      "loss": 1.0583,
+      "step": 7134
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.34362142592246,
+      "learning_rate": 7.0077723854198075e-06,
+      "loss": 1.1543,
+      "step": 7135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4581239103668113,
+      "learning_rate": 7.005135343761052e-06,
+      "loss": 1.1079,
+      "step": 7136
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.428172925529442,
+      "learning_rate": 7.002498530865003e-06,
+      "loss": 1.085,
+      "step": 7137
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0422574603448336,
+      "learning_rate": 6.999861946933072e-06,
+      "loss": 1.0632,
+      "step": 7138
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0117234298500235,
+      "learning_rate": 6.997225592166653e-06,
+      "loss": 1.0576,
+      "step": 7139
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9954859973164423,
+      "learning_rate": 6.99458946676713e-06,
+      "loss": 1.1526,
+      "step": 7140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.41223437746827,
+      "learning_rate": 6.991953570935853e-06,
+      "loss": 1.0818,
+      "step": 7141
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4823931124810366,
+      "learning_rate": 6.989317904874172e-06,
+      "loss": 1.1416,
+      "step": 7142
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2465296617869095,
+      "learning_rate": 6.986682468783412e-06,
+      "loss": 1.0549,
+      "step": 7143
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.068832439758159,
+      "learning_rate": 6.984047262864879e-06,
+      "loss": 1.0858,
+      "step": 7144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.844363035767304,
+      "learning_rate": 6.981412287319862e-06,
+      "loss": 1.1177,
+      "step": 7145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.674702368806056,
+      "learning_rate": 6.9787775423496375e-06,
+      "loss": 1.1079,
+      "step": 7146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.30899022488149,
+      "learning_rate": 6.97614302815546e-06,
+      "loss": 1.0996,
+      "step": 7147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.372642251917464,
+      "learning_rate": 6.973508744938566e-06,
+      "loss": 1.0955,
+      "step": 7148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.259382515685577,
+      "learning_rate": 6.97087469290017e-06,
+      "loss": 1.0945,
+      "step": 7149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4660936937296465,
+      "learning_rate": 6.968240872241479e-06,
+      "loss": 1.0203,
+      "step": 7150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2004025218030825,
+      "learning_rate": 6.965607283163678e-06,
+      "loss": 1.1165,
+      "step": 7151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8934835303869004,
+      "learning_rate": 6.962973925867933e-06,
+      "loss": 1.0417,
+      "step": 7152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0137075336637227,
+      "learning_rate": 6.960340800555393e-06,
+      "loss": 1.1199,
+      "step": 7153
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.631688097534476,
+      "learning_rate": 6.9577079074271895e-06,
+      "loss": 1.1472,
+      "step": 7154
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2135156576431685,
+      "learning_rate": 6.9550752466844305e-06,
+      "loss": 1.1331,
+      "step": 7155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2078515765206532,
+      "learning_rate": 6.95244281852822e-06,
+      "loss": 1.0771,
+      "step": 7156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.4545645775810705,
+      "learning_rate": 6.949810623159632e-06,
+      "loss": 1.0745,
+      "step": 7157
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.176757886304951,
+      "learning_rate": 6.9471786607797254e-06,
+      "loss": 1.032,
+      "step": 7158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.715733753203716,
+      "learning_rate": 6.944546931589546e-06,
+      "loss": 1.0918,
+      "step": 7159
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.009152595132267,
+      "learning_rate": 6.94191543579012e-06,
+      "loss": 1.1025,
+      "step": 7160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.001335155359322,
+      "learning_rate": 6.939284173582447e-06,
+      "loss": 1.0657,
+      "step": 7161
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.061300311559097,
+      "learning_rate": 6.936653145167521e-06,
+      "loss": 1.0208,
+      "step": 7162
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.334442814447506,
+      "learning_rate": 6.934022350746311e-06,
+      "loss": 1.0723,
+      "step": 7163
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.635132973723719,
+      "learning_rate": 6.93139179051977e-06,
+      "loss": 1.0962,
+      "step": 7164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.060475139559956,
+      "learning_rate": 6.928761464688836e-06,
+      "loss": 1.1191,
+      "step": 7165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8178103361992455,
+      "learning_rate": 6.926131373454428e-06,
+      "loss": 1.1443,
+      "step": 7166
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.087789157327156,
+      "learning_rate": 6.9235015170174414e-06,
+      "loss": 1.1321,
+      "step": 7167
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7012021028924926,
+      "learning_rate": 6.920871895578757e-06,
+      "loss": 1.1316,
+      "step": 7168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.290607694394007,
+      "learning_rate": 6.918242509339239e-06,
+      "loss": 1.1401,
+      "step": 7169
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.5898810658053995,
+      "learning_rate": 6.915613358499736e-06,
+      "loss": 1.1299,
+      "step": 7170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3580116105224445,
+      "learning_rate": 6.912984443261073e-06,
+      "loss": 1.0918,
+      "step": 7171
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5359530049949464,
+      "learning_rate": 6.91035576382406e-06,
+      "loss": 1.1199,
+      "step": 7172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.26737333194801,
+      "learning_rate": 6.9077273203894945e-06,
+      "loss": 1.1238,
+      "step": 7173
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.9774961769078185,
+      "learning_rate": 6.905099113158138e-06,
+      "loss": 1.0796,
+      "step": 7174
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6733114601614205,
+      "learning_rate": 6.9024711423307555e-06,
+      "loss": 1.0906,
+      "step": 7175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.257385113356422,
+      "learning_rate": 6.899843408108083e-06,
+      "loss": 1.1692,
+      "step": 7176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.79920199437303,
+      "learning_rate": 6.897215910690837e-06,
+      "loss": 1.1055,
+      "step": 7177
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.233498538270011,
+      "learning_rate": 6.89458865027972e-06,
+      "loss": 1.1262,
+      "step": 7178
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.204367323738909,
+      "learning_rate": 6.8919616270754165e-06,
+      "loss": 1.103,
+      "step": 7179
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.809879061572515,
+      "learning_rate": 6.889334841278596e-06,
+      "loss": 1.187,
+      "step": 7180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.222018881538373,
+      "learning_rate": 6.8867082930898964e-06,
+      "loss": 1.1543,
+      "step": 7181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.198391972011853,
+      "learning_rate": 6.884081982709952e-06,
+      "loss": 1.1086,
+      "step": 7182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.499200585495818,
+      "learning_rate": 6.881455910339369e-06,
+      "loss": 1.0786,
+      "step": 7183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9762171418976493,
+      "learning_rate": 6.878830076178744e-06,
+      "loss": 1.072,
+      "step": 7184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1374832160885955,
+      "learning_rate": 6.876204480428654e-06,
+      "loss": 1.0366,
+      "step": 7185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9538855268757365,
+      "learning_rate": 6.873579123289648e-06,
+      "loss": 1.1333,
+      "step": 7186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.415535320793473,
+      "learning_rate": 6.870954004962274e-06,
+      "loss": 1.1438,
+      "step": 7187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.045242440197404,
+      "learning_rate": 6.868329125647039e-06,
+      "loss": 1.2056,
+      "step": 7188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4063004806588255,
+      "learning_rate": 6.865704485544451e-06,
+      "loss": 1.1008,
+      "step": 7189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2180241470785496,
+      "learning_rate": 6.863080084854994e-06,
+      "loss": 1.1348,
+      "step": 7190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9966220687091147,
+      "learning_rate": 6.86045592377913e-06,
+      "loss": 1.1492,
+      "step": 7191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6221908738643496,
+      "learning_rate": 6.857832002517307e-06,
+      "loss": 1.1257,
+      "step": 7192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.932803702703247,
+      "learning_rate": 6.855208321269957e-06,
+      "loss": 1.0906,
+      "step": 7193
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5576104248615064,
+      "learning_rate": 6.852584880237483e-06,
+      "loss": 1.0925,
+      "step": 7194
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.082927492410721,
+      "learning_rate": 6.849961679620281e-06,
+      "loss": 1.123,
+      "step": 7195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.567474989623045,
+      "learning_rate": 6.8473387196187195e-06,
+      "loss": 1.1499,
+      "step": 7196
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.752971504058924,
+      "learning_rate": 6.844716000433157e-06,
+      "loss": 1.0559,
+      "step": 7197
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.330291714139076,
+      "learning_rate": 6.84209352226393e-06,
+      "loss": 1.0457,
+      "step": 7198
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.446419887889665,
+      "learning_rate": 6.839471285311357e-06,
+      "loss": 1.0715,
+      "step": 7199
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.391353924220635,
+      "learning_rate": 6.836849289775739e-06,
+      "loss": 1.0969,
+      "step": 7200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1916337898060183,
+      "learning_rate": 6.834227535857349e-06,
+      "loss": 1.0801,
+      "step": 7201
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5218550677510794,
+      "learning_rate": 6.8316060237564575e-06,
+      "loss": 1.1355,
+      "step": 7202
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.437900381247247,
+      "learning_rate": 6.828984753673304e-06,
+      "loss": 1.1021,
+      "step": 7203
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.663754173986515,
+      "learning_rate": 6.82636372580812e-06,
+      "loss": 1.1025,
+      "step": 7204
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2932216820354534,
+      "learning_rate": 6.823742940361107e-06,
+      "loss": 1.1077,
+      "step": 7205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9481873917833967,
+      "learning_rate": 6.8211223975324555e-06,
+      "loss": 1.1431,
+      "step": 7206
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9536257273121214,
+      "learning_rate": 6.818502097522339e-06,
+      "loss": 1.1055,
+      "step": 7207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.858392943848445,
+      "learning_rate": 6.815882040530904e-06,
+      "loss": 1.1509,
+      "step": 7208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1749057002734915,
+      "learning_rate": 6.813262226758286e-06,
+      "loss": 1.1208,
+      "step": 7209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.000737100678656,
+      "learning_rate": 6.8106426564045965e-06,
+      "loss": 1.1448,
+      "step": 7210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.115431487191311,
+      "learning_rate": 6.8080233296699346e-06,
+      "loss": 1.126,
+      "step": 7211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0282793639667367,
+      "learning_rate": 6.8054042467543765e-06,
+      "loss": 1.0962,
+      "step": 7212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.10804464449274,
+      "learning_rate": 6.802785407857987e-06,
+      "loss": 1.1243,
+      "step": 7213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.767335356117585,
+      "learning_rate": 6.800166813180796e-06,
+      "loss": 1.1399,
+      "step": 7214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9993634764819737,
+      "learning_rate": 6.797548462922828e-06,
+      "loss": 1.0476,
+      "step": 7215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.268390664966252,
+      "learning_rate": 6.794930357284086e-06,
+      "loss": 1.1187,
+      "step": 7216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.211684416821958,
+      "learning_rate": 6.792312496464555e-06,
+      "loss": 1.1257,
+      "step": 7217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.321656296206818,
+      "learning_rate": 6.789694880664203e-06,
+      "loss": 1.0918,
+      "step": 7218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.828876980692256,
+      "learning_rate": 6.78707751008297e-06,
+      "loss": 1.1438,
+      "step": 7219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3121746461963535,
+      "learning_rate": 6.784460384920792e-06,
+      "loss": 1.1313,
+      "step": 7220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.599242090875505,
+      "learning_rate": 6.781843505377569e-06,
+      "loss": 1.04,
+      "step": 7221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7629180809957963,
+      "learning_rate": 6.779226871653195e-06,
+      "loss": 1.1523,
+      "step": 7222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.394797023291386,
+      "learning_rate": 6.776610483947546e-06,
+      "loss": 1.0925,
+      "step": 7223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4987752205155025,
+      "learning_rate": 6.773994342460468e-06,
+      "loss": 1.1611,
+      "step": 7224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.195774938538627,
+      "learning_rate": 6.771378447391799e-06,
+      "loss": 1.073,
+      "step": 7225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1250978283088617,
+      "learning_rate": 6.768762798941355e-06,
+      "loss": 1.0435,
+      "step": 7226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.889527509130784,
+      "learning_rate": 6.7661473973089275e-06,
+      "loss": 1.1199,
+      "step": 7227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8153796768774813,
+      "learning_rate": 6.763532242694298e-06,
+      "loss": 1.1101,
+      "step": 7228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5885778205689083,
+      "learning_rate": 6.760917335297224e-06,
+      "loss": 1.0752,
+      "step": 7229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8191308780570954,
+      "learning_rate": 6.758302675317443e-06,
+      "loss": 1.0498,
+      "step": 7230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.987788146604689,
+      "learning_rate": 6.755688262954677e-06,
+      "loss": 1.1116,
+      "step": 7231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4785178209669496,
+      "learning_rate": 6.753074098408632e-06,
+      "loss": 1.0466,
+      "step": 7232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.159767205471031,
+      "learning_rate": 6.750460181878989e-06,
+      "loss": 1.0603,
+      "step": 7233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.92186238275318,
+      "learning_rate": 6.747846513565407e-06,
+      "loss": 1.1221,
+      "step": 7234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7503685529855586,
+      "learning_rate": 6.745233093667534e-06,
+      "loss": 1.1172,
+      "step": 7235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.757232058190548,
+      "learning_rate": 6.742619922384997e-06,
+      "loss": 1.0808,
+      "step": 7236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.521111497824778,
+      "learning_rate": 6.740006999917406e-06,
+      "loss": 1.1235,
+      "step": 7237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2044578032730664,
+      "learning_rate": 6.737394326464342e-06,
+      "loss": 1.054,
+      "step": 7238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.38366378315568,
+      "learning_rate": 6.734781902225378e-06,
+      "loss": 1.1567,
+      "step": 7239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1397528274321305,
+      "learning_rate": 6.732169727400068e-06,
+      "loss": 1.0742,
+      "step": 7240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7534100786902562,
+      "learning_rate": 6.729557802187934e-06,
+      "loss": 1.1045,
+      "step": 7241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9277873114518873,
+      "learning_rate": 6.726946126788495e-06,
+      "loss": 1.1606,
+      "step": 7242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0316517630624267,
+      "learning_rate": 6.724334701401239e-06,
+      "loss": 1.1355,
+      "step": 7243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2712870338332753,
+      "learning_rate": 6.721723526225643e-06,
+      "loss": 1.0414,
+      "step": 7244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.226932513078008,
+      "learning_rate": 6.719112601461162e-06,
+      "loss": 1.0798,
+      "step": 7245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4953491019948846,
+      "learning_rate": 6.716501927307231e-06,
+      "loss": 1.1721,
+      "step": 7246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.998634441492601,
+      "learning_rate": 6.713891503963265e-06,
+      "loss": 1.1418,
+      "step": 7247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.115692375598212,
+      "learning_rate": 6.711281331628658e-06,
+      "loss": 1.0742,
+      "step": 7248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3746883952979636,
+      "learning_rate": 6.708671410502794e-06,
+      "loss": 1.1562,
+      "step": 7249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0274563341586997,
+      "learning_rate": 6.706061740785031e-06,
+      "loss": 1.0576,
+      "step": 7250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.255692811119623,
+      "learning_rate": 6.703452322674704e-06,
+      "loss": 1.0791,
+      "step": 7251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.102278719311958,
+      "learning_rate": 6.7008431563711365e-06,
+      "loss": 1.1631,
+      "step": 7252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.09009130673168,
+      "learning_rate": 6.698234242073633e-06,
+      "loss": 1.1519,
+      "step": 7253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5509871508656214,
+      "learning_rate": 6.6956255799814686e-06,
+      "loss": 1.115,
+      "step": 7254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.925157192048314,
+      "learning_rate": 6.69301717029391e-06,
+      "loss": 1.0315,
+      "step": 7255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.977384088971047,
+      "learning_rate": 6.6904090132102004e-06,
+      "loss": 1.1328,
+      "step": 7256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9181319595497706,
+      "learning_rate": 6.687801108929561e-06,
+      "loss": 1.1287,
+      "step": 7257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8787481136731192,
+      "learning_rate": 6.6851934576512e-06,
+      "loss": 1.1816,
+      "step": 7258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.634140090972876,
+      "learning_rate": 6.682586059574303e-06,
+      "loss": 1.0813,
+      "step": 7259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8786705991749053,
+      "learning_rate": 6.679978914898035e-06,
+      "loss": 1.0933,
+      "step": 7260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0041094106384802,
+      "learning_rate": 6.677372023821542e-06,
+      "loss": 1.1211,
+      "step": 7261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.133218363057953,
+      "learning_rate": 6.67476538654395e-06,
+      "loss": 1.1536,
+      "step": 7262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7301915852492855,
+      "learning_rate": 6.672159003264369e-06,
+      "loss": 1.0955,
+      "step": 7263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.716086622769176,
+      "learning_rate": 6.669552874181888e-06,
+      "loss": 1.062,
+      "step": 7264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6057543516180655,
+      "learning_rate": 6.666946999495574e-06,
+      "loss": 1.0679,
+      "step": 7265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.274317516428819,
+      "learning_rate": 6.664341379404482e-06,
+      "loss": 1.1006,
+      "step": 7266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.076019977573477,
+      "learning_rate": 6.661736014107635e-06,
+      "loss": 1.0281,
+      "step": 7267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.914239595907686,
+      "learning_rate": 6.659130903804046e-06,
+      "loss": 1.1267,
+      "step": 7268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.485884725599944,
+      "learning_rate": 6.6565260486927105e-06,
+      "loss": 1.1208,
+      "step": 7269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.959098307458824,
+      "learning_rate": 6.653921448972594e-06,
+      "loss": 1.0706,
+      "step": 7270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.21670069128025,
+      "learning_rate": 6.651317104842654e-06,
+      "loss": 1.1067,
+      "step": 7271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3495283598041445,
+      "learning_rate": 6.64871301650182e-06,
+      "loss": 1.104,
+      "step": 7272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2958984825699416,
+      "learning_rate": 6.646109184149011e-06,
+      "loss": 1.1089,
+      "step": 7273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.951016036216469,
+      "learning_rate": 6.643505607983112e-06,
+      "loss": 1.0796,
+      "step": 7274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5010720394262127,
+      "learning_rate": 6.640902288203004e-06,
+      "loss": 1.0881,
+      "step": 7275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.518064876002787,
+      "learning_rate": 6.638299225007537e-06,
+      "loss": 1.1394,
+      "step": 7276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.243496624721557,
+      "learning_rate": 6.635696418595548e-06,
+      "loss": 1.0652,
+      "step": 7277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0784058138695696,
+      "learning_rate": 6.633093869165854e-06,
+      "loss": 1.0537,
+      "step": 7278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8180370103267562,
+      "learning_rate": 6.6304915769172505e-06,
+      "loss": 1.135,
+      "step": 7279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.667975276671058,
+      "learning_rate": 6.627889542048511e-06,
+      "loss": 1.0955,
+      "step": 7280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0425513898554235,
+      "learning_rate": 6.625287764758391e-06,
+      "loss": 1.1074,
+      "step": 7281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3288751034464634,
+      "learning_rate": 6.6226862452456295e-06,
+      "loss": 1.1296,
+      "step": 7282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4530174588701406,
+      "learning_rate": 6.620084983708945e-06,
+      "loss": 1.1394,
+      "step": 7283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.995779525655814,
+      "learning_rate": 6.6174839803470324e-06,
+      "loss": 1.0176,
+      "step": 7284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.115862151124302,
+      "learning_rate": 6.614883235358569e-06,
+      "loss": 1.0535,
+      "step": 7285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.823062448724529,
+      "learning_rate": 6.612282748942218e-06,
+      "loss": 1.0881,
+      "step": 7286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0894342051903543,
+      "learning_rate": 6.609682521296609e-06,
+      "loss": 1.1182,
+      "step": 7287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3721262238515757,
+      "learning_rate": 6.607082552620368e-06,
+      "loss": 1.0901,
+      "step": 7288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2512834152179826,
+      "learning_rate": 6.604482843112086e-06,
+      "loss": 1.0667,
+      "step": 7289
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4777910595982866,
+      "learning_rate": 6.601883392970347e-06,
+      "loss": 1.1038,
+      "step": 7290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.587810746615595,
+      "learning_rate": 6.599284202393709e-06,
+      "loss": 1.1145,
+      "step": 7291
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2716742955777636,
+      "learning_rate": 6.5966852715807126e-06,
+      "loss": 1.0964,
+      "step": 7292
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2952322150925846,
+      "learning_rate": 6.594086600729876e-06,
+      "loss": 1.1064,
+      "step": 7293
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1922701736701224,
+      "learning_rate": 6.591488190039699e-06,
+      "loss": 1.0793,
+      "step": 7294
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2522036139856203,
+      "learning_rate": 6.588890039708657e-06,
+      "loss": 1.1155,
+      "step": 7295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1099273655871436,
+      "learning_rate": 6.586292149935214e-06,
+      "loss": 1.0828,
+      "step": 7296
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1625271097725065,
+      "learning_rate": 6.583694520917812e-06,
+      "loss": 1.1047,
+      "step": 7297
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8631301753674534,
+      "learning_rate": 6.581097152854864e-06,
+      "loss": 1.0896,
+      "step": 7298
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.524170341663448,
+      "learning_rate": 6.578500045944778e-06,
+      "loss": 1.1436,
+      "step": 7299
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.488476560409616,
+      "learning_rate": 6.575903200385927e-06,
+      "loss": 1.0977,
+      "step": 7300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.258338431945847,
+      "learning_rate": 6.573306616376671e-06,
+      "loss": 1.0793,
+      "step": 7301
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.389260865358332,
+      "learning_rate": 6.5707102941153565e-06,
+      "loss": 1.0984,
+      "step": 7302
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.658112032693564,
+      "learning_rate": 6.568114233800298e-06,
+      "loss": 1.1189,
+      "step": 7303
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8690203633434317,
+      "learning_rate": 6.565518435629797e-06,
+      "loss": 1.1345,
+      "step": 7304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1247066047521646,
+      "learning_rate": 6.562922899802135e-06,
+      "loss": 1.1223,
+      "step": 7305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6549986750167394,
+      "learning_rate": 6.5603276265155734e-06,
+      "loss": 1.1995,
+      "step": 7306
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.637601753575572,
+      "learning_rate": 6.5577326159683486e-06,
+      "loss": 1.1272,
+      "step": 7307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.050897930080824,
+      "learning_rate": 6.5551378683586785e-06,
+      "loss": 1.0222,
+      "step": 7308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7084944588178477,
+      "learning_rate": 6.552543383884766e-06,
+      "loss": 1.0476,
+      "step": 7309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.869383685407717,
+      "learning_rate": 6.54994916274479e-06,
+      "loss": 1.1282,
+      "step": 7310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.937914499749311,
+      "learning_rate": 6.547355205136914e-06,
+      "loss": 1.0989,
+      "step": 7311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.963872290360207,
+      "learning_rate": 6.54476151125927e-06,
+      "loss": 1.0759,
+      "step": 7312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5857218499492323,
+      "learning_rate": 6.542168081309987e-06,
+      "loss": 1.1108,
+      "step": 7313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.300512389205829,
+      "learning_rate": 6.539574915487154e-06,
+      "loss": 1.1003,
+      "step": 7314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.108647629681246,
+      "learning_rate": 6.536982013988853e-06,
+      "loss": 1.1938,
+      "step": 7315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5348094612022996,
+      "learning_rate": 6.534389377013148e-06,
+      "loss": 1.0994,
+      "step": 7316
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.8113251306111495,
+      "learning_rate": 6.531797004758071e-06,
+      "loss": 1.157,
+      "step": 7317
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2969233193800163,
+      "learning_rate": 6.529204897421644e-06,
+      "loss": 1.0925,
+      "step": 7318
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4780780928311463,
+      "learning_rate": 6.526613055201867e-06,
+      "loss": 1.0815,
+      "step": 7319
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.694853198556631,
+      "learning_rate": 6.5240214782967105e-06,
+      "loss": 1.1541,
+      "step": 7320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.031900370054858,
+      "learning_rate": 6.521430166904139e-06,
+      "loss": 1.1299,
+      "step": 7321
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9644656660389224,
+      "learning_rate": 6.5188391212220845e-06,
+      "loss": 1.0247,
+      "step": 7322
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7303964061113457,
+      "learning_rate": 6.516248341448468e-06,
+      "loss": 1.0718,
+      "step": 7323
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.376664353710888,
+      "learning_rate": 6.513657827781184e-06,
+      "loss": 1.0918,
+      "step": 7324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2098541540843426,
+      "learning_rate": 6.511067580418113e-06,
+      "loss": 1.0225,
+      "step": 7325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.231429478052309,
+      "learning_rate": 6.508477599557109e-06,
+      "loss": 1.0562,
+      "step": 7326
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.627196227998781,
+      "learning_rate": 6.505887885396003e-06,
+      "loss": 1.1431,
+      "step": 7327
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9413218527000695,
+      "learning_rate": 6.503298438132614e-06,
+      "loss": 1.125,
+      "step": 7328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.914029476602505,
+      "learning_rate": 6.500709257964736e-06,
+      "loss": 1.0601,
+      "step": 7329
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.708640027390409,
+      "learning_rate": 6.4981203450901464e-06,
+      "loss": 1.1309,
+      "step": 7330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.487383639468143,
+      "learning_rate": 6.495531699706594e-06,
+      "loss": 1.092,
+      "step": 7331
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0648969441480185,
+      "learning_rate": 6.492943322011817e-06,
+      "loss": 1.1211,
+      "step": 7332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.476970808563244,
+      "learning_rate": 6.490355212203532e-06,
+      "loss": 1.1052,
+      "step": 7333
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.049473769229305,
+      "learning_rate": 6.487767370479421e-06,
+      "loss": 1.1572,
+      "step": 7334
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6276856468699235,
+      "learning_rate": 6.485179797037164e-06,
+      "loss": 1.1313,
+      "step": 7335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6934614915060093,
+      "learning_rate": 6.482592492074411e-06,
+      "loss": 1.1399,
+      "step": 7336
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.983979119236332,
+      "learning_rate": 6.480005455788792e-06,
+      "loss": 1.0754,
+      "step": 7337
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.769755525229688,
+      "learning_rate": 6.477418688377919e-06,
+      "loss": 1.0867,
+      "step": 7338
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1245247080086265,
+      "learning_rate": 6.4748321900393865e-06,
+      "loss": 1.0781,
+      "step": 7339
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.228573099895442,
+      "learning_rate": 6.472245960970755e-06,
+      "loss": 1.0769,
+      "step": 7340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4016755012988646,
+      "learning_rate": 6.46966000136958e-06,
+      "loss": 1.0708,
+      "step": 7341
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.622009829013356,
+      "learning_rate": 6.467074311433385e-06,
+      "loss": 1.0583,
+      "step": 7342
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7295939598171657,
+      "learning_rate": 6.464488891359683e-06,
+      "loss": 1.103,
+      "step": 7343
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.667992466605134,
+      "learning_rate": 6.461903741345963e-06,
+      "loss": 1.0762,
+      "step": 7344
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.400713479509097,
+      "learning_rate": 6.4593188615896855e-06,
+      "loss": 1.1404,
+      "step": 7345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.517010447778908,
+      "learning_rate": 6.456734252288302e-06,
+      "loss": 1.1177,
+      "step": 7346
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2635400877894063,
+      "learning_rate": 6.454149913639232e-06,
+      "loss": 1.0403,
+      "step": 7347
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3697868552540866,
+      "learning_rate": 6.451565845839882e-06,
+      "loss": 1.0706,
+      "step": 7348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1957583454482377,
+      "learning_rate": 6.44898204908764e-06,
+      "loss": 1.1106,
+      "step": 7349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.338101032948457,
+      "learning_rate": 6.446398523579864e-06,
+      "loss": 1.103,
+      "step": 7350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0628671973370807,
+      "learning_rate": 6.4438152695139e-06,
+      "loss": 1.1055,
+      "step": 7351
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.09808398609206,
+      "learning_rate": 6.441232287087073e-06,
+      "loss": 1.1216,
+      "step": 7352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9641130720734865,
+      "learning_rate": 6.438649576496675e-06,
+      "loss": 1.0796,
+      "step": 7353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.572755828586847,
+      "learning_rate": 6.436067137939994e-06,
+      "loss": 1.0928,
+      "step": 7354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7094989789234956,
+      "learning_rate": 6.4334849716142845e-06,
+      "loss": 1.1494,
+      "step": 7355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.287240179128836,
+      "learning_rate": 6.430903077716788e-06,
+      "loss": 1.0901,
+      "step": 7356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5866321769204714,
+      "learning_rate": 6.428321456444722e-06,
+      "loss": 1.0962,
+      "step": 7357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6355474614506265,
+      "learning_rate": 6.425740107995284e-06,
+      "loss": 1.0321,
+      "step": 7358
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5144791562776128,
+      "learning_rate": 6.423159032565653e-06,
+      "loss": 1.0693,
+      "step": 7359
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2892003545338495,
+      "learning_rate": 6.420578230352978e-06,
+      "loss": 1.1008,
+      "step": 7360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7094673777084886,
+      "learning_rate": 6.417997701554397e-06,
+      "loss": 1.0732,
+      "step": 7361
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0851194988370745,
+      "learning_rate": 6.415417446367022e-06,
+      "loss": 1.1113,
+      "step": 7362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8844839198185612,
+      "learning_rate": 6.412837464987951e-06,
+      "loss": 1.0443,
+      "step": 7363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0416018941334433,
+      "learning_rate": 6.41025775761425e-06,
+      "loss": 1.2312,
+      "step": 7364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7522759727898927,
+      "learning_rate": 6.407678324442972e-06,
+      "loss": 1.0815,
+      "step": 7365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3089639155154247,
+      "learning_rate": 6.40509916567115e-06,
+      "loss": 1.0791,
+      "step": 7366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.375406871135637,
+      "learning_rate": 6.402520281495787e-06,
+      "loss": 1.1057,
+      "step": 7367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0188968472083597,
+      "learning_rate": 6.399941672113877e-06,
+      "loss": 1.0457,
+      "step": 7368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.254348173333569,
+      "learning_rate": 6.39736333772238e-06,
+      "loss": 1.0615,
+      "step": 7369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.941051248596801,
+      "learning_rate": 6.394785278518247e-06,
+      "loss": 1.1079,
+      "step": 7370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0435050431250876,
+      "learning_rate": 6.392207494698406e-06,
+      "loss": 1.0913,
+      "step": 7371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.020428153667108,
+      "learning_rate": 6.389629986459756e-06,
+      "loss": 1.1362,
+      "step": 7372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2256188025721757,
+      "learning_rate": 6.38705275399918e-06,
+      "loss": 1.0872,
+      "step": 7373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9012262776948825,
+      "learning_rate": 6.384475797513541e-06,
+      "loss": 0.9766,
+      "step": 7374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.148336484184237,
+      "learning_rate": 6.381899117199678e-06,
+      "loss": 1.1284,
+      "step": 7375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.485455940711233,
+      "learning_rate": 6.379322713254414e-06,
+      "loss": 1.019,
+      "step": 7376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.389579536509229,
+      "learning_rate": 6.376746585874547e-06,
+      "loss": 1.0891,
+      "step": 7377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.453638459481537,
+      "learning_rate": 6.374170735256853e-06,
+      "loss": 1.0942,
+      "step": 7378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.16543839587399,
+      "learning_rate": 6.371595161598092e-06,
+      "loss": 1.1174,
+      "step": 7379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.038624776094414,
+      "learning_rate": 6.369019865094992e-06,
+      "loss": 1.0359,
+      "step": 7380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.629983053871514,
+      "learning_rate": 6.36644484594427e-06,
+      "loss": 1.0625,
+      "step": 7381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5438868135978283,
+      "learning_rate": 6.363870104342624e-06,
+      "loss": 1.137,
+      "step": 7382
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.890552030543584,
+      "learning_rate": 6.3612956404867185e-06,
+      "loss": 1.104,
+      "step": 7383
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3126280187439607,
+      "learning_rate": 6.3587214545732066e-06,
+      "loss": 1.0835,
+      "step": 7384
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.889522805939758,
+      "learning_rate": 6.356147546798721e-06,
+      "loss": 1.1021,
+      "step": 7385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.722576898986105,
+      "learning_rate": 6.353573917359865e-06,
+      "loss": 1.1118,
+      "step": 7386
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.38087631507279,
+      "learning_rate": 6.351000566453227e-06,
+      "loss": 1.1062,
+      "step": 7387
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.381555623949729,
+      "learning_rate": 6.348427494275371e-06,
+      "loss": 1.0388,
+      "step": 7388
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.993240304679187,
+      "learning_rate": 6.345854701022841e-06,
+      "loss": 1.0723,
+      "step": 7389
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4069651482642054,
+      "learning_rate": 6.343282186892163e-06,
+      "loss": 1.0757,
+      "step": 7390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.837190971227316,
+      "learning_rate": 6.340709952079835e-06,
+      "loss": 1.0427,
+      "step": 7391
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.673993660667247,
+      "learning_rate": 6.338137996782342e-06,
+      "loss": 1.1733,
+      "step": 7392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8766266682967205,
+      "learning_rate": 6.335566321196135e-06,
+      "loss": 1.0181,
+      "step": 7393
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6050285208899115,
+      "learning_rate": 6.332994925517656e-06,
+      "loss": 1.0627,
+      "step": 7394
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.565492503657127,
+      "learning_rate": 6.33042380994332e-06,
+      "loss": 1.135,
+      "step": 7395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.85546588136688,
+      "learning_rate": 6.327852974669526e-06,
+      "loss": 1.1296,
+      "step": 7396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7375196746060366,
+      "learning_rate": 6.32528241989264e-06,
+      "loss": 1.0715,
+      "step": 7397
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.059149184863713,
+      "learning_rate": 6.322712145809018e-06,
+      "loss": 1.0828,
+      "step": 7398
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.611281054044548,
+      "learning_rate": 6.3201421526149945e-06,
+      "loss": 1.0532,
+      "step": 7399
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.221136585341286,
+      "learning_rate": 6.317572440506868e-06,
+      "loss": 1.1272,
+      "step": 7400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.627057677265922,
+      "learning_rate": 6.315003009680935e-06,
+      "loss": 1.0886,
+      "step": 7401
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5597369896100024,
+      "learning_rate": 6.312433860333456e-06,
+      "loss": 1.1653,
+      "step": 7402
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.759209121538923,
+      "learning_rate": 6.309864992660676e-06,
+      "loss": 1.0991,
+      "step": 7403
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4002549552748236,
+      "learning_rate": 6.307296406858822e-06,
+      "loss": 1.075,
+      "step": 7404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.141795804567687,
+      "learning_rate": 6.304728103124094e-06,
+      "loss": 1.0923,
+      "step": 7405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5809375206338494,
+      "learning_rate": 6.3021600816526684e-06,
+      "loss": 1.0291,
+      "step": 7406
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.74498009780437,
+      "learning_rate": 6.299592342640706e-06,
+      "loss": 1.1077,
+      "step": 7407
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.243040446033627,
+      "learning_rate": 6.297024886284342e-06,
+      "loss": 1.1008,
+      "step": 7408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.571703987368198,
+      "learning_rate": 6.294457712779693e-06,
+      "loss": 1.0562,
+      "step": 7409
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.374831504501082,
+      "learning_rate": 6.291890822322854e-06,
+      "loss": 1.1343,
+      "step": 7410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1552833251774586,
+      "learning_rate": 6.2893242151098936e-06,
+      "loss": 1.0562,
+      "step": 7411
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5151090926875663,
+      "learning_rate": 6.286757891336866e-06,
+      "loss": 1.0852,
+      "step": 7412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2779174417782775,
+      "learning_rate": 6.284191851199796e-06,
+      "loss": 1.1113,
+      "step": 7413
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.223446487437574,
+      "learning_rate": 6.2816260948946905e-06,
+      "loss": 1.0979,
+      "step": 7414
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.0779091370364995,
+      "learning_rate": 6.279060622617539e-06,
+      "loss": 1.1365,
+      "step": 7415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2615141982908216,
+      "learning_rate": 6.2764954345643e-06,
+      "loss": 1.04,
+      "step": 7416
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.450829779331414,
+      "learning_rate": 6.273930530930919e-06,
+      "loss": 1.1201,
+      "step": 7417
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5009127188092544,
+      "learning_rate": 6.2713659119133145e-06,
+      "loss": 1.1445,
+      "step": 7418
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.560392457236552,
+      "learning_rate": 6.268801577707386e-06,
+      "loss": 1.0767,
+      "step": 7419
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.545422592411839,
+      "learning_rate": 6.26623752850901e-06,
+      "loss": 1.0991,
+      "step": 7420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.132146355281973,
+      "learning_rate": 6.263673764514039e-06,
+      "loss": 1.1245,
+      "step": 7421
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.052948403314819,
+      "learning_rate": 6.261110285918308e-06,
+      "loss": 1.0447,
+      "step": 7422
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2667946045689877,
+      "learning_rate": 6.2585470929176285e-06,
+      "loss": 1.1233,
+      "step": 7423
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.498477707449075,
+      "learning_rate": 6.255984185707789e-06,
+      "loss": 1.0588,
+      "step": 7424
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.436616649160318,
+      "learning_rate": 6.253421564484562e-06,
+      "loss": 1.1143,
+      "step": 7425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.161811129628668,
+      "learning_rate": 6.250859229443684e-06,
+      "loss": 1.0415,
+      "step": 7426
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.322085771788783,
+      "learning_rate": 6.248297180780884e-06,
+      "loss": 1.1062,
+      "step": 7427
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.909662654759509,
+      "learning_rate": 6.245735418691867e-06,
+      "loss": 1.1345,
+      "step": 7428
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4385844501913976,
+      "learning_rate": 6.243173943372307e-06,
+      "loss": 1.1353,
+      "step": 7429
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6879941605894286,
+      "learning_rate": 6.240612755017867e-06,
+      "loss": 1.0559,
+      "step": 7430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1067126348563656,
+      "learning_rate": 6.238051853824181e-06,
+      "loss": 1.1128,
+      "step": 7431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1201539077000917,
+      "learning_rate": 6.235491239986869e-06,
+      "loss": 1.1499,
+      "step": 7432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.062431680530957,
+      "learning_rate": 6.232930913701516e-06,
+      "loss": 1.0483,
+      "step": 7433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.169094033781995,
+      "learning_rate": 6.230370875163692e-06,
+      "loss": 1.1396,
+      "step": 7434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0459648076476897,
+      "learning_rate": 6.2278111245689495e-06,
+      "loss": 1.0381,
+      "step": 7435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.01239157015728,
+      "learning_rate": 6.225251662112814e-06,
+      "loss": 1.0928,
+      "step": 7436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.152526371533131,
+      "learning_rate": 6.222692487990793e-06,
+      "loss": 1.0625,
+      "step": 7437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.384421067577056,
+      "learning_rate": 6.220133602398365e-06,
+      "loss": 1.1086,
+      "step": 7438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.043420376614114,
+      "learning_rate": 6.217575005530995e-06,
+      "loss": 1.0366,
+      "step": 7439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0903705676492774,
+      "learning_rate": 6.215016697584113e-06,
+      "loss": 1.0518,
+      "step": 7440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4488473667433683,
+      "learning_rate": 6.2124586787531415e-06,
+      "loss": 1.1631,
+      "step": 7441
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0203395926779466,
+      "learning_rate": 6.209900949233476e-06,
+      "loss": 1.1245,
+      "step": 7442
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.929648691238215,
+      "learning_rate": 6.207343509220483e-06,
+      "loss": 1.1279,
+      "step": 7443
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.281444636104108,
+      "learning_rate": 6.204786358909516e-06,
+      "loss": 1.0813,
+      "step": 7444
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.914677064689894,
+      "learning_rate": 6.202229498495908e-06,
+      "loss": 1.0525,
+      "step": 7445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3476864992553335,
+      "learning_rate": 6.199672928174953e-06,
+      "loss": 1.0933,
+      "step": 7446
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0646845353037695,
+      "learning_rate": 6.1971166481419455e-06,
+      "loss": 1.0522,
+      "step": 7447
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1036621601412064,
+      "learning_rate": 6.194560658592139e-06,
+      "loss": 1.1592,
+      "step": 7448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2198211403221975,
+      "learning_rate": 6.192004959720777e-06,
+      "loss": 1.0767,
+      "step": 7449
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6742558980904465,
+      "learning_rate": 6.189449551723076e-06,
+      "loss": 1.0303,
+      "step": 7450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0864503102413385,
+      "learning_rate": 6.186894434794231e-06,
+      "loss": 1.0879,
+      "step": 7451
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4030640226487026,
+      "learning_rate": 6.184339609129418e-06,
+      "loss": 1.1248,
+      "step": 7452
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.321630385497231,
+      "learning_rate": 6.181785074923778e-06,
+      "loss": 1.1335,
+      "step": 7453
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.284196339467756,
+      "learning_rate": 6.1792308323724456e-06,
+      "loss": 1.0806,
+      "step": 7454
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8784053529257805,
+      "learning_rate": 6.176676881670525e-06,
+      "loss": 1.1123,
+      "step": 7455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8085826301885137,
+      "learning_rate": 6.174123223013102e-06,
+      "loss": 1.1614,
+      "step": 7456
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.127192711899416,
+      "learning_rate": 6.171569856595236e-06,
+      "loss": 1.0791,
+      "step": 7457
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.192980121773023,
+      "learning_rate": 6.1690167826119625e-06,
+      "loss": 1.041,
+      "step": 7458
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.733664645583339,
+      "learning_rate": 6.166464001258307e-06,
+      "loss": 1.1492,
+      "step": 7459
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4178686822820437,
+      "learning_rate": 6.1639115127292545e-06,
+      "loss": 1.1379,
+      "step": 7460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6601056168717068,
+      "learning_rate": 6.1613593172197836e-06,
+      "loss": 1.146,
+      "step": 7461
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7180267063093693,
+      "learning_rate": 6.1588074149248365e-06,
+      "loss": 1.1042,
+      "step": 7462
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0010641974660786,
+      "learning_rate": 6.156255806039345e-06,
+      "loss": 1.0331,
+      "step": 7463
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1211524129517967,
+      "learning_rate": 6.153704490758212e-06,
+      "loss": 1.1401,
+      "step": 7464
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.179061515288344,
+      "learning_rate": 6.151153469276326e-06,
+      "loss": 1.093,
+      "step": 7465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7908882382494955,
+      "learning_rate": 6.148602741788539e-06,
+      "loss": 1.0099,
+      "step": 7466
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.205934378486907,
+      "learning_rate": 6.14605230848969e-06,
+      "loss": 1.1621,
+      "step": 7467
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.395381176753646,
+      "learning_rate": 6.143502169574591e-06,
+      "loss": 1.0891,
+      "step": 7468
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.242870311763962,
+      "learning_rate": 6.140952325238041e-06,
+      "loss": 1.1313,
+      "step": 7469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7947076792715193,
+      "learning_rate": 6.138402775674807e-06,
+      "loss": 1.083,
+      "step": 7470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.98093266301423,
+      "learning_rate": 6.135853521079636e-06,
+      "loss": 1.1682,
+      "step": 7471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.820970556004441,
+      "learning_rate": 6.133304561647255e-06,
+      "loss": 1.0762,
+      "step": 7472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.838746241811571,
+      "learning_rate": 6.13075589757236e-06,
+      "loss": 1.1331,
+      "step": 7473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6887952659917294,
+      "learning_rate": 6.128207529049634e-06,
+      "loss": 1.0808,
+      "step": 7474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0807346551469847,
+      "learning_rate": 6.125659456273739e-06,
+      "loss": 1.0935,
+      "step": 7475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8394991431678913,
+      "learning_rate": 6.123111679439302e-06,
+      "loss": 1.0859,
+      "step": 7476
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3632271902987685,
+      "learning_rate": 6.120564198740939e-06,
+      "loss": 1.1304,
+      "step": 7477
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1089605622518537,
+      "learning_rate": 6.118017014373243e-06,
+      "loss": 1.1067,
+      "step": 7478
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.520536309327753,
+      "learning_rate": 6.115470126530771e-06,
+      "loss": 1.1824,
+      "step": 7479
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6117888007734167,
+      "learning_rate": 6.112923535408074e-06,
+      "loss": 1.1526,
+      "step": 7480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2279240635402635,
+      "learning_rate": 6.11037724119967e-06,
+      "loss": 1.022,
+      "step": 7481
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5649499667108904,
+      "learning_rate": 6.107831244100059e-06,
+      "loss": 1.0837,
+      "step": 7482
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9221934586057534,
+      "learning_rate": 6.105285544303717e-06,
+      "loss": 1.1494,
+      "step": 7483
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2059605319117224,
+      "learning_rate": 6.102740142005099e-06,
+      "loss": 1.1484,
+      "step": 7484
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0382851369345665,
+      "learning_rate": 6.100195037398638e-06,
+      "loss": 1.0947,
+      "step": 7485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3185779076203294,
+      "learning_rate": 6.097650230678731e-06,
+      "loss": 1.0552,
+      "step": 7486
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.614547206926469,
+      "learning_rate": 6.095105722039771e-06,
+      "loss": 1.1123,
+      "step": 7487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8103375023036734,
+      "learning_rate": 6.092561511676118e-06,
+      "loss": 0.9907,
+      "step": 7488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.708726113597617,
+      "learning_rate": 6.0900175997821165e-06,
+      "loss": 1.0996,
+      "step": 7489
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2446460639465515,
+      "learning_rate": 6.087473986552076e-06,
+      "loss": 1.1025,
+      "step": 7490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.519100986141261,
+      "learning_rate": 6.084930672180294e-06,
+      "loss": 1.1531,
+      "step": 7491
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1652323023995144,
+      "learning_rate": 6.082387656861044e-06,
+      "loss": 1.0776,
+      "step": 7492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.291909210564807,
+      "learning_rate": 6.079844940788569e-06,
+      "loss": 1.1218,
+      "step": 7493
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.412347668300711,
+      "learning_rate": 6.077302524157096e-06,
+      "loss": 1.1062,
+      "step": 7494
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2777432731295786,
+      "learning_rate": 6.074760407160828e-06,
+      "loss": 1.0942,
+      "step": 7495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3079335792337705,
+      "learning_rate": 6.072218589993944e-06,
+      "loss": 1.1226,
+      "step": 7496
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4135613880481093,
+      "learning_rate": 6.069677072850602e-06,
+      "loss": 1.1211,
+      "step": 7497
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8058844663547955,
+      "learning_rate": 6.067135855924937e-06,
+      "loss": 1.052,
+      "step": 7498
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6021676049117843,
+      "learning_rate": 6.064594939411058e-06,
+      "loss": 1.1699,
+      "step": 7499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8751783973217524,
+      "learning_rate": 6.062054323503051e-06,
+      "loss": 1.0674,
+      "step": 7500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.052651650032587,
+      "learning_rate": 6.059514008394982e-06,
+      "loss": 1.1494,
+      "step": 7501
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.938168502770379,
+      "learning_rate": 6.056973994280892e-06,
+      "loss": 1.0461,
+      "step": 7502
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5943705386670466,
+      "learning_rate": 6.054434281354806e-06,
+      "loss": 1.085,
+      "step": 7503
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9250412738620675,
+      "learning_rate": 6.0518948698107125e-06,
+      "loss": 1.0706,
+      "step": 7504
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.645730520086009,
+      "learning_rate": 6.049355759842592e-06,
+      "loss": 1.1438,
+      "step": 7505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.123462386982665,
+      "learning_rate": 6.046816951644387e-06,
+      "loss": 1.0555,
+      "step": 7506
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3638474633553836,
+      "learning_rate": 6.044278445410025e-06,
+      "loss": 1.033,
+      "step": 7507
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.148872936930826,
+      "learning_rate": 6.041740241333415e-06,
+      "loss": 1.0801,
+      "step": 7508
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.503175341349068,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 1.1396,
+      "step": 7509
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2327347511092475,
+      "learning_rate": 6.036664740428937e-06,
+      "loss": 1.1143,
+      "step": 7510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.235253678145717,
+      "learning_rate": 6.034127443988766e-06,
+      "loss": 1.0808,
+      "step": 7511
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7736809565732363,
+      "learning_rate": 6.031590450481728e-06,
+      "loss": 1.0825,
+      "step": 7512
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0896073023633615,
+      "learning_rate": 6.0290537601016105e-06,
+      "loss": 1.0559,
+      "step": 7513
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.118205920696212,
+      "learning_rate": 6.026517373042178e-06,
+      "loss": 1.0884,
+      "step": 7514
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.36864342277693,
+      "learning_rate": 6.023981289497174e-06,
+      "loss": 1.1553,
+      "step": 7515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7477990856728165,
+      "learning_rate": 6.021445509660315e-06,
+      "loss": 1.1008,
+      "step": 7516
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.077998191380527,
+      "learning_rate": 6.018910033725303e-06,
+      "loss": 1.0725,
+      "step": 7517
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5575454890816682,
+      "learning_rate": 6.0163748618858074e-06,
+      "loss": 1.1025,
+      "step": 7518
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9317366132445817,
+      "learning_rate": 6.013839994335471e-06,
+      "loss": 1.1165,
+      "step": 7519
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3582670934017385,
+      "learning_rate": 6.011305431267924e-06,
+      "loss": 1.0634,
+      "step": 7520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.476438570492577,
+      "learning_rate": 6.008771172876771e-06,
+      "loss": 1.0728,
+      "step": 7521
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5648963409676924,
+      "learning_rate": 6.006237219355591e-06,
+      "loss": 1.1533,
+      "step": 7522
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9972384221711708,
+      "learning_rate": 6.003703570897935e-06,
+      "loss": 1.0303,
+      "step": 7523
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.263824389271094,
+      "learning_rate": 6.001170227697342e-06,
+      "loss": 1.1609,
+      "step": 7524
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.205451427817812,
+      "learning_rate": 5.9986371899473205e-06,
+      "loss": 1.0792,
+      "step": 7525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.981954590205206,
+      "learning_rate": 5.9961044578413515e-06,
+      "loss": 1.0894,
+      "step": 7526
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.219409579266361,
+      "learning_rate": 5.993572031572903e-06,
+      "loss": 1.1287,
+      "step": 7527
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7992492951016934,
+      "learning_rate": 5.99103991133541e-06,
+      "loss": 1.085,
+      "step": 7528
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8445265225243093,
+      "learning_rate": 5.988508097322292e-06,
+      "loss": 1.0122,
+      "step": 7529
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.555276149634404,
+      "learning_rate": 5.98597658972694e-06,
+      "loss": 1.1055,
+      "step": 7530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.223479845909631,
+      "learning_rate": 5.983445388742726e-06,
+      "loss": 1.0706,
+      "step": 7531
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.574812011456455,
+      "learning_rate": 5.980914494562992e-06,
+      "loss": 1.1631,
+      "step": 7532
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3841455697290925,
+      "learning_rate": 5.978383907381061e-06,
+      "loss": 1.0769,
+      "step": 7533
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.651968048274125,
+      "learning_rate": 5.975853627390233e-06,
+      "loss": 1.1401,
+      "step": 7534
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2354770558916086,
+      "learning_rate": 5.973323654783781e-06,
+      "loss": 1.1018,
+      "step": 7535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.693479677467113,
+      "learning_rate": 5.970793989754962e-06,
+      "loss": 1.1558,
+      "step": 7536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8360484554880965,
+      "learning_rate": 5.968264632496998e-06,
+      "loss": 1.1125,
+      "step": 7537
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1709687329739844,
+      "learning_rate": 5.965735583203103e-06,
+      "loss": 1.0483,
+      "step": 7538
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.381610322727603,
+      "learning_rate": 5.963206842066449e-06,
+      "loss": 1.0901,
+      "step": 7539
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7987765303787895,
+      "learning_rate": 5.960678409280196e-06,
+      "loss": 1.1199,
+      "step": 7540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3382790723387803,
+      "learning_rate": 5.958150285037483e-06,
+      "loss": 1.0867,
+      "step": 7541
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.803939186884847,
+      "learning_rate": 5.955622469531414e-06,
+      "loss": 1.0168,
+      "step": 7542
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.236574666391475,
+      "learning_rate": 5.953094962955081e-06,
+      "loss": 1.1042,
+      "step": 7543
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.170073839110056,
+      "learning_rate": 5.950567765501548e-06,
+      "loss": 1.1274,
+      "step": 7544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8053978831668127,
+      "learning_rate": 5.9480408773638544e-06,
+      "loss": 1.0303,
+      "step": 7545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7382134727384715,
+      "learning_rate": 5.945514298735014e-06,
+      "loss": 1.1301,
+      "step": 7546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.006490752147927,
+      "learning_rate": 5.942988029808019e-06,
+      "loss": 1.1311,
+      "step": 7547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6142994141180886,
+      "learning_rate": 5.940462070775841e-06,
+      "loss": 1.1438,
+      "step": 7548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9226608259538183,
+      "learning_rate": 5.937936421831425e-06,
+      "loss": 1.0381,
+      "step": 7549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.841097775393604,
+      "learning_rate": 5.935411083167693e-06,
+      "loss": 1.0857,
+      "step": 7550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.539853347247746,
+      "learning_rate": 5.932886054977544e-06,
+      "loss": 1.1887,
+      "step": 7551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7435390356855316,
+      "learning_rate": 5.930361337453847e-06,
+      "loss": 1.0645,
+      "step": 7552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6037615009875075,
+      "learning_rate": 5.927836930789457e-06,
+      "loss": 1.1353,
+      "step": 7553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3443258907185784,
+      "learning_rate": 5.9253128351772e-06,
+      "loss": 1.0964,
+      "step": 7554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4600115204951174,
+      "learning_rate": 5.922789050809878e-06,
+      "loss": 1.0769,
+      "step": 7555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2894212862346395,
+      "learning_rate": 5.920265577880271e-06,
+      "loss": 1.0437,
+      "step": 7556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2123834683648966,
+      "learning_rate": 5.9177424165811356e-06,
+      "loss": 1.1587,
+      "step": 7557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.484841740063131,
+      "learning_rate": 5.915219567105205e-06,
+      "loss": 1.1345,
+      "step": 7558
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8645288813538086,
+      "learning_rate": 5.91269702964518e-06,
+      "loss": 1.1301,
+      "step": 7559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3294567054124005,
+      "learning_rate": 5.910174804393754e-06,
+      "loss": 1.0642,
+      "step": 7560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4490269240935185,
+      "learning_rate": 5.907652891543576e-06,
+      "loss": 1.1333,
+      "step": 7561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6513600781686013,
+      "learning_rate": 5.905131291287291e-06,
+      "loss": 1.1191,
+      "step": 7562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.069897457990145,
+      "learning_rate": 5.9026100038175105e-06,
+      "loss": 1.0972,
+      "step": 7563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.444728992816773,
+      "learning_rate": 5.900089029326819e-06,
+      "loss": 1.157,
+      "step": 7564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.916781800061131,
+      "learning_rate": 5.897568368007789e-06,
+      "loss": 1.0608,
+      "step": 7565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1030870789781613,
+      "learning_rate": 5.895048020052949e-06,
+      "loss": 1.0671,
+      "step": 7566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9092524483851845,
+      "learning_rate": 5.892527985654824e-06,
+      "loss": 1.1157,
+      "step": 7567
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.464193319218345,
+      "learning_rate": 5.890008265005907e-06,
+      "loss": 1.0366,
+      "step": 7568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.432502229489494,
+      "learning_rate": 5.887488858298665e-06,
+      "loss": 1.1443,
+      "step": 7569
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.955576363663901,
+      "learning_rate": 5.88496976572554e-06,
+      "loss": 1.0933,
+      "step": 7570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5420738481101166,
+      "learning_rate": 5.88245098747896e-06,
+      "loss": 1.1108,
+      "step": 7571
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5391054531752895,
+      "learning_rate": 5.8799325237513146e-06,
+      "loss": 1.0823,
+      "step": 7572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.992119928231836,
+      "learning_rate": 5.877414374734982e-06,
+      "loss": 1.1057,
+      "step": 7573
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.616841119279728,
+      "learning_rate": 5.874896540622306e-06,
+      "loss": 1.0317,
+      "step": 7574
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.15429285007846,
+      "learning_rate": 5.872379021605614e-06,
+      "loss": 1.1736,
+      "step": 7575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.8638371068539525,
+      "learning_rate": 5.869861817877207e-06,
+      "loss": 1.1038,
+      "step": 7576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9682756233370706,
+      "learning_rate": 5.867344929629364e-06,
+      "loss": 1.1021,
+      "step": 7577
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4518489324634247,
+      "learning_rate": 5.864828357054333e-06,
+      "loss": 1.1318,
+      "step": 7578
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.99681563265301,
+      "learning_rate": 5.862312100344345e-06,
+      "loss": 1.0266,
+      "step": 7579
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.952124503860786,
+      "learning_rate": 5.859796159691601e-06,
+      "loss": 1.1179,
+      "step": 7580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.180669916816339,
+      "learning_rate": 5.8572805352882835e-06,
+      "loss": 1.1206,
+      "step": 7581
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6478007008334474,
+      "learning_rate": 5.854765227326551e-06,
+      "loss": 1.1179,
+      "step": 7582
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1743933938699076,
+      "learning_rate": 5.852250235998529e-06,
+      "loss": 1.1033,
+      "step": 7583
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7446640848409256,
+      "learning_rate": 5.849735561496331e-06,
+      "loss": 1.0642,
+      "step": 7584
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.709725525040195,
+      "learning_rate": 5.8472212040120405e-06,
+      "loss": 1.0544,
+      "step": 7585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.155207841503339,
+      "learning_rate": 5.844707163737711e-06,
+      "loss": 1.0249,
+      "step": 7586
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.431346039959337,
+      "learning_rate": 5.842193440865381e-06,
+      "loss": 1.1304,
+      "step": 7587
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.498021444527481,
+      "learning_rate": 5.839680035587061e-06,
+      "loss": 1.0972,
+      "step": 7588
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.441613292733119,
+      "learning_rate": 5.837166948094737e-06,
+      "loss": 1.1216,
+      "step": 7589
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0611062127569335,
+      "learning_rate": 5.834654178580372e-06,
+      "loss": 1.0977,
+      "step": 7590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6725991678883734,
+      "learning_rate": 5.8321417272359045e-06,
+      "loss": 1.1279,
+      "step": 7591
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7810838047276585,
+      "learning_rate": 5.829629594253248e-06,
+      "loss": 1.1123,
+      "step": 7592
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.26069367690705,
+      "learning_rate": 5.827117779824285e-06,
+      "loss": 1.1016,
+      "step": 7593
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.741058932019161,
+      "learning_rate": 5.82460628414089e-06,
+      "loss": 1.0823,
+      "step": 7594
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1260248689634222,
+      "learning_rate": 5.8220951073949e-06,
+      "loss": 1.0737,
+      "step": 7595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.086316858337301,
+      "learning_rate": 5.819584249778128e-06,
+      "loss": 1.1536,
+      "step": 7596
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.78830329473388,
+      "learning_rate": 5.8170737114823715e-06,
+      "loss": 1.1216,
+      "step": 7597
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.21452345243263,
+      "learning_rate": 5.8145634926993954e-06,
+      "loss": 1.0898,
+      "step": 7598
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9823489881129808,
+      "learning_rate": 5.812053593620943e-06,
+      "loss": 1.1072,
+      "step": 7599
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.790413610356472,
+      "learning_rate": 5.809544014438732e-06,
+      "loss": 1.0691,
+      "step": 7600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.29352184398721,
+      "learning_rate": 5.807034755344454e-06,
+      "loss": 1.0203,
+      "step": 7601
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.745202576663903,
+      "learning_rate": 5.804525816529785e-06,
+      "loss": 1.1072,
+      "step": 7602
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6055683968046885,
+      "learning_rate": 5.802017198186366e-06,
+      "loss": 1.0351,
+      "step": 7603
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.0176585564267855,
+      "learning_rate": 5.799508900505819e-06,
+      "loss": 1.1697,
+      "step": 7604
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.104609307236542,
+      "learning_rate": 5.797000923679741e-06,
+      "loss": 1.116,
+      "step": 7605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.527257656409668,
+      "learning_rate": 5.794493267899699e-06,
+      "loss": 1.0872,
+      "step": 7606
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.789857795725576,
+      "learning_rate": 5.791985933357247e-06,
+      "loss": 1.0525,
+      "step": 7607
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.2373065657258735,
+      "learning_rate": 5.789478920243905e-06,
+      "loss": 1.1133,
+      "step": 7608
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3471788849294386,
+      "learning_rate": 5.786972228751169e-06,
+      "loss": 1.1228,
+      "step": 7609
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1372153904725564,
+      "learning_rate": 5.784465859070516e-06,
+      "loss": 1.0281,
+      "step": 7610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.411597753286334,
+      "learning_rate": 5.781959811393395e-06,
+      "loss": 1.0483,
+      "step": 7611
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7069208005081773,
+      "learning_rate": 5.779454085911229e-06,
+      "loss": 1.0625,
+      "step": 7612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1416880604887094,
+      "learning_rate": 5.776948682815417e-06,
+      "loss": 1.0818,
+      "step": 7613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.117358708221244,
+      "learning_rate": 5.774443602297332e-06,
+      "loss": 1.1052,
+      "step": 7614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9778462013224685,
+      "learning_rate": 5.771938844548331e-06,
+      "loss": 1.1025,
+      "step": 7615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.630902230133993,
+      "learning_rate": 5.7694344097597335e-06,
+      "loss": 1.1052,
+      "step": 7616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.861178830307493,
+      "learning_rate": 5.766930298122846e-06,
+      "loss": 1.0637,
+      "step": 7617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.877198746669341,
+      "learning_rate": 5.7644265098289485e-06,
+      "loss": 1.1277,
+      "step": 7618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.513432520157351,
+      "learning_rate": 5.7619230450692775e-06,
+      "loss": 1.1196,
+      "step": 7619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.739784368405567,
+      "learning_rate": 5.759419904035075e-06,
+      "loss": 1.1741,
+      "step": 7620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2621153285585813,
+      "learning_rate": 5.756917086917533e-06,
+      "loss": 1.0222,
+      "step": 7621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.447358896574106,
+      "learning_rate": 5.75441459390784e-06,
+      "loss": 1.1167,
+      "step": 7622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4199445769756354,
+      "learning_rate": 5.751912425197141e-06,
+      "loss": 1.0803,
+      "step": 7623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.925741604286824,
+      "learning_rate": 5.749410580976569e-06,
+      "loss": 1.075,
+      "step": 7624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.38921258982918,
+      "learning_rate": 5.746909061437223e-06,
+      "loss": 1.0889,
+      "step": 7625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9590315670123384,
+      "learning_rate": 5.744407866770178e-06,
+      "loss": 0.9832,
+      "step": 7626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.833302139205904,
+      "learning_rate": 5.7419069971664995e-06,
+      "loss": 1.0862,
+      "step": 7627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.058118356356756,
+      "learning_rate": 5.739406452817209e-06,
+      "loss": 1.0581,
+      "step": 7628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0270893431363004,
+      "learning_rate": 5.736906233913307e-06,
+      "loss": 1.0806,
+      "step": 7629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9164774954699215,
+      "learning_rate": 5.734406340645781e-06,
+      "loss": 1.0955,
+      "step": 7630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9446101438965595,
+      "learning_rate": 5.731906773205581e-06,
+      "loss": 1.1487,
+      "step": 7631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.62936457260052,
+      "learning_rate": 5.729407531783638e-06,
+      "loss": 1.0803,
+      "step": 7632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.209218167316966,
+      "learning_rate": 5.7269086165708546e-06,
+      "loss": 1.042,
+      "step": 7633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.256283216887048,
+      "learning_rate": 5.724410027758107e-06,
+      "loss": 1.0806,
+      "step": 7634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1531876332028235,
+      "learning_rate": 5.721911765536259e-06,
+      "loss": 1.0696,
+      "step": 7635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1830416037485176,
+      "learning_rate": 5.719413830096134e-06,
+      "loss": 1.0798,
+      "step": 7636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5572013388114203,
+      "learning_rate": 5.716916221628535e-06,
+      "loss": 1.1321,
+      "step": 7637
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.910800571117296,
+      "learning_rate": 5.714418940324253e-06,
+      "loss": 1.0481,
+      "step": 7638
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9000433741233653,
+      "learning_rate": 5.711921986374027e-06,
+      "loss": 1.0635,
+      "step": 7639
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.5799092093879965,
+      "learning_rate": 5.709425359968599e-06,
+      "loss": 1.1533,
+      "step": 7640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0778063258026505,
+      "learning_rate": 5.706929061298667e-06,
+      "loss": 1.0752,
+      "step": 7641
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1360448886753702,
+      "learning_rate": 5.704433090554912e-06,
+      "loss": 1.0938,
+      "step": 7642
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.780237614307446,
+      "learning_rate": 5.701937447927992e-06,
+      "loss": 1.1345,
+      "step": 7643
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4635994468676015,
+      "learning_rate": 5.699442133608535e-06,
+      "loss": 1.0681,
+      "step": 7644
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5654901621111366,
+      "learning_rate": 5.696947147787145e-06,
+      "loss": 1.1633,
+      "step": 7645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.175520154646992,
+      "learning_rate": 5.694452490654402e-06,
+      "loss": 1.0374,
+      "step": 7646
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.386317352522449,
+      "learning_rate": 5.6919581624008545e-06,
+      "loss": 1.1445,
+      "step": 7647
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.59967751536813,
+      "learning_rate": 5.689464163217042e-06,
+      "loss": 1.0854,
+      "step": 7648
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.568397834632691,
+      "learning_rate": 5.68697049329346e-06,
+      "loss": 1.093,
+      "step": 7649
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3304509123443737,
+      "learning_rate": 5.684477152820594e-06,
+      "loss": 1.0928,
+      "step": 7650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.378076737865881,
+      "learning_rate": 5.6819841419889e-06,
+      "loss": 1.0789,
+      "step": 7651
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.490375425090823,
+      "learning_rate": 5.679491460988793e-06,
+      "loss": 1.0415,
+      "step": 7652
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2270987778827482,
+      "learning_rate": 5.676999110010688e-06,
+      "loss": 1.0679,
+      "step": 7653
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.64461724375336,
+      "learning_rate": 5.674507089244957e-06,
+      "loss": 1.1685,
+      "step": 7654
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.990040497123934,
+      "learning_rate": 5.672015398881961e-06,
+      "loss": 1.0605,
+      "step": 7655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.020976131417031,
+      "learning_rate": 5.669524039112022e-06,
+      "loss": 1.0923,
+      "step": 7656
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3776232946330227,
+      "learning_rate": 5.667033010125442e-06,
+      "loss": 1.1128,
+      "step": 7657
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5597435562583075,
+      "learning_rate": 5.664542312112502e-06,
+      "loss": 1.0586,
+      "step": 7658
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.597659094209205,
+      "learning_rate": 5.662051945263445e-06,
+      "loss": 1.1296,
+      "step": 7659
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9861660240545267,
+      "learning_rate": 5.659561909768509e-06,
+      "loss": 1.1084,
+      "step": 7660
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0914347699242577,
+      "learning_rate": 5.657072205817892e-06,
+      "loss": 1.1047,
+      "step": 7661
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.215360358212452,
+      "learning_rate": 5.654582833601763e-06,
+      "loss": 1.019,
+      "step": 7662
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.69843952583214,
+      "learning_rate": 5.652093793310283e-06,
+      "loss": 1.0688,
+      "step": 7663
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3965464158421614,
+      "learning_rate": 5.649605085133572e-06,
+      "loss": 1.1133,
+      "step": 7664
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0859580667027675,
+      "learning_rate": 5.64711670926173e-06,
+      "loss": 1.104,
+      "step": 7665
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.431821696962516,
+      "learning_rate": 5.644628665884833e-06,
+      "loss": 1.0474,
+      "step": 7666
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3737594658566716,
+      "learning_rate": 5.642140955192923e-06,
+      "loss": 1.0603,
+      "step": 7667
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8621543177426907,
+      "learning_rate": 5.639653577376036e-06,
+      "loss": 1.1116,
+      "step": 7668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.915460844052964,
+      "learning_rate": 5.6371665326241635e-06,
+      "loss": 1.0923,
+      "step": 7669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.758097638360257,
+      "learning_rate": 5.634679821127276e-06,
+      "loss": 1.0754,
+      "step": 7670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.007626209602573,
+      "learning_rate": 5.632193443075331e-06,
+      "loss": 1.1619,
+      "step": 7671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0087691132757937,
+      "learning_rate": 5.629707398658235e-06,
+      "loss": 1.0481,
+      "step": 7672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7209905120416895,
+      "learning_rate": 5.627221688065898e-06,
+      "loss": 1.199,
+      "step": 7673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.326913495035979,
+      "learning_rate": 5.6247363114881856e-06,
+      "loss": 1.0581,
+      "step": 7674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2864184154856093,
+      "learning_rate": 5.62225126911494e-06,
+      "loss": 1.0251,
+      "step": 7675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7989686801424276,
+      "learning_rate": 5.619766561135989e-06,
+      "loss": 1.1101,
+      "step": 7676
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.309360879353406,
+      "learning_rate": 5.617282187741122e-06,
+      "loss": 1.1196,
+      "step": 7677
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6617169607996205,
+      "learning_rate": 5.614798149120108e-06,
+      "loss": 1.095,
+      "step": 7678
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9458303959055163,
+      "learning_rate": 5.612314445462692e-06,
+      "loss": 1.1052,
+      "step": 7679
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.208454519898562,
+      "learning_rate": 5.609831076958586e-06,
+      "loss": 1.1436,
+      "step": 7680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.491061138593973,
+      "learning_rate": 5.607348043797491e-06,
+      "loss": 0.9929,
+      "step": 7681
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.808725692537082,
+      "learning_rate": 5.604865346169068e-06,
+      "loss": 1.1543,
+      "step": 7682
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4738617989932545,
+      "learning_rate": 5.602382984262957e-06,
+      "loss": 1.0891,
+      "step": 7683
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.145025473257985,
+      "learning_rate": 5.599900958268781e-06,
+      "loss": 1.1326,
+      "step": 7684
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.076036870500118,
+      "learning_rate": 5.5974192683761165e-06,
+      "loss": 1.0674,
+      "step": 7685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3026063420233607,
+      "learning_rate": 5.5949379147745385e-06,
+      "loss": 1.1145,
+      "step": 7686
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.913616051220946,
+      "learning_rate": 5.5924568976535775e-06,
+      "loss": 1.0891,
+      "step": 7687
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.690170344524257,
+      "learning_rate": 5.589976217202753e-06,
+      "loss": 1.103,
+      "step": 7688
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6817700016122132,
+      "learning_rate": 5.5874958736115485e-06,
+      "loss": 1.1296,
+      "step": 7689
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5257084928403155,
+      "learning_rate": 5.585015867069421e-06,
+      "loss": 1.0967,
+      "step": 7690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.692975350814661,
+      "learning_rate": 5.582536197765818e-06,
+      "loss": 1.114,
+      "step": 7691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.189036391008614,
+      "learning_rate": 5.580056865890131e-06,
+      "loss": 1.0742,
+      "step": 7692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.505412082334767,
+      "learning_rate": 5.577577871631758e-06,
+      "loss": 1.075,
+      "step": 7693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5976334024084227,
+      "learning_rate": 5.575099215180054e-06,
+      "loss": 1.0488,
+      "step": 7694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9059353198694664,
+      "learning_rate": 5.572620896724345e-06,
+      "loss": 1.1238,
+      "step": 7695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.232529034091648,
+      "learning_rate": 5.570142916453944e-06,
+      "loss": 1.1213,
+      "step": 7696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3296510623932325,
+      "learning_rate": 5.567665274558131e-06,
+      "loss": 1.0654,
+      "step": 7697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.758066820447952,
+      "learning_rate": 5.565187971226159e-06,
+      "loss": 1.1218,
+      "step": 7698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.083897798378551,
+      "learning_rate": 5.5627110066472564e-06,
+      "loss": 1.0867,
+      "step": 7699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5389451578246898,
+      "learning_rate": 5.560234381010623e-06,
+      "loss": 1.0842,
+      "step": 7700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.986639398381846,
+      "learning_rate": 5.557758094505442e-06,
+      "loss": 1.0571,
+      "step": 7701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.783402017855869,
+      "learning_rate": 5.555282147320862e-06,
+      "loss": 1.1494,
+      "step": 7702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.325397170167489,
+      "learning_rate": 5.5528065396460055e-06,
+      "loss": 1.0642,
+      "step": 7703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.644216072180857,
+      "learning_rate": 5.550331271669981e-06,
+      "loss": 1.0999,
+      "step": 7704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8042938753484026,
+      "learning_rate": 5.547856343581848e-06,
+      "loss": 1.0898,
+      "step": 7705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.277878221265666,
+      "learning_rate": 5.545381755570663e-06,
+      "loss": 1.0852,
+      "step": 7706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.665640982185907,
+      "learning_rate": 5.542907507825448e-06,
+      "loss": 1.0771,
+      "step": 7707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3979119076755477,
+      "learning_rate": 5.5404336005351896e-06,
+      "loss": 1.1418,
+      "step": 7708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3051387265191288,
+      "learning_rate": 5.537960033888868e-06,
+      "loss": 1.0703,
+      "step": 7709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.380814351741242,
+      "learning_rate": 5.535486808075422e-06,
+      "loss": 1.1006,
+      "step": 7710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.210467230597616,
+      "learning_rate": 5.53301392328377e-06,
+      "loss": 1.1094,
+      "step": 7711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9194928078867752,
+      "learning_rate": 5.5305413797028e-06,
+      "loss": 1.0254,
+      "step": 7712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2736192217649935,
+      "learning_rate": 5.528069177521376e-06,
+      "loss": 1.1304,
+      "step": 7713
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6715150219151313,
+      "learning_rate": 5.525597316928344e-06,
+      "loss": 1.1028,
+      "step": 7714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7782825345888438,
+      "learning_rate": 5.523125798112514e-06,
+      "loss": 1.1221,
+      "step": 7715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2751511603260535,
+      "learning_rate": 5.520654621262669e-06,
+      "loss": 1.1489,
+      "step": 7716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1506251678261488,
+      "learning_rate": 5.518183786567582e-06,
+      "loss": 1.0891,
+      "step": 7717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3469106636509163,
+      "learning_rate": 5.5157132942159685e-06,
+      "loss": 1.0278,
+      "step": 7718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3190725659188245,
+      "learning_rate": 5.513243144396552e-06,
+      "loss": 1.1221,
+      "step": 7719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2665704058542637,
+      "learning_rate": 5.510773337298011e-06,
+      "loss": 1.0969,
+      "step": 7720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3670494522596885,
+      "learning_rate": 5.508303873108998e-06,
+      "loss": 1.0723,
+      "step": 7721
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.274347684971314,
+      "learning_rate": 5.505834752018147e-06,
+      "loss": 1.1287,
+      "step": 7722
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.879569958962196,
+      "learning_rate": 5.503365974214059e-06,
+      "loss": 1.0417,
+      "step": 7723
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.835957801094059,
+      "learning_rate": 5.500897539885321e-06,
+      "loss": 1.2046,
+      "step": 7724
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.474183205540569,
+      "learning_rate": 5.498429449220473e-06,
+      "loss": 1.0586,
+      "step": 7725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.553042051142641,
+      "learning_rate": 5.495961702408039e-06,
+      "loss": 1.0874,
+      "step": 7726
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.124281436992016,
+      "learning_rate": 5.493494299636528e-06,
+      "loss": 1.0713,
+      "step": 7727
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8265796948377986,
+      "learning_rate": 5.491027241094402e-06,
+      "loss": 1.1052,
+      "step": 7728
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.112513207824566,
+      "learning_rate": 5.488560526970116e-06,
+      "loss": 1.1553,
+      "step": 7729
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.116564196279954,
+      "learning_rate": 5.486094157452087e-06,
+      "loss": 1.0359,
+      "step": 7730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6221005559858375,
+      "learning_rate": 5.483628132728707e-06,
+      "loss": 1.0835,
+      "step": 7731
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.0542531766564265,
+      "learning_rate": 5.481162452988344e-06,
+      "loss": 1.0615,
+      "step": 7732
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8865150147799197,
+      "learning_rate": 5.478697118419336e-06,
+      "loss": 1.092,
+      "step": 7733
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.936792818475211,
+      "learning_rate": 5.476232129210005e-06,
+      "loss": 1.0547,
+      "step": 7734
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7609232184375077,
+      "learning_rate": 5.473767485548635e-06,
+      "loss": 1.0513,
+      "step": 7735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.775835131413483,
+      "learning_rate": 5.471303187623483e-06,
+      "loss": 1.1238,
+      "step": 7736
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7447751886265452,
+      "learning_rate": 5.4688392356227956e-06,
+      "loss": 1.1477,
+      "step": 7737
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6168720281456945,
+      "learning_rate": 5.466375629734768e-06,
+      "loss": 1.0764,
+      "step": 7738
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.3672051218271015,
+      "learning_rate": 5.463912370147595e-06,
+      "loss": 1.1165,
+      "step": 7739
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.726512536290249,
+      "learning_rate": 5.461449457049426e-06,
+      "loss": 1.1421,
+      "step": 7740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9715891757535577,
+      "learning_rate": 5.458986890628389e-06,
+      "loss": 1.1299,
+      "step": 7741
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.106517029543114,
+      "learning_rate": 5.456524671072594e-06,
+      "loss": 1.064,
+      "step": 7742
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1519055141989085,
+      "learning_rate": 5.454062798570113e-06,
+      "loss": 1.0793,
+      "step": 7743
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.231275072971238,
+      "learning_rate": 5.451601273308997e-06,
+      "loss": 1.0254,
+      "step": 7744
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.354935499671891,
+      "learning_rate": 5.449140095477271e-06,
+      "loss": 1.0964,
+      "step": 7745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.849116822812139,
+      "learning_rate": 5.446679265262925e-06,
+      "loss": 1.0789,
+      "step": 7746
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.59655533009042,
+      "learning_rate": 5.4442187828539406e-06,
+      "loss": 1.165,
+      "step": 7747
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9152346352925056,
+      "learning_rate": 5.441758648438255e-06,
+      "loss": 1.0518,
+      "step": 7748
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8831869324762027,
+      "learning_rate": 5.439298862203783e-06,
+      "loss": 1.0835,
+      "step": 7749
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.8944503804328106,
+      "learning_rate": 5.436839424338426e-06,
+      "loss": 1.1882,
+      "step": 7750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7911764202430267,
+      "learning_rate": 5.434380335030034e-06,
+      "loss": 1.1858,
+      "step": 7751
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3232542377402536,
+      "learning_rate": 5.431921594466456e-06,
+      "loss": 1.0186,
+      "step": 7752
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.070113393994238,
+      "learning_rate": 5.429463202835498e-06,
+      "loss": 1.0977,
+      "step": 7753
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.935419107700412,
+      "learning_rate": 5.4270051603249404e-06,
+      "loss": 1.0913,
+      "step": 7754
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9323067439550226,
+      "learning_rate": 5.424547467122549e-06,
+      "loss": 1.0571,
+      "step": 7755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9508413889858125,
+      "learning_rate": 5.422090123416048e-06,
+      "loss": 1.1531,
+      "step": 7756
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.838832850037018,
+      "learning_rate": 5.419633129393151e-06,
+      "loss": 1.094,
+      "step": 7757
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.877051177946895,
+      "learning_rate": 5.417176485241527e-06,
+      "loss": 1.0227,
+      "step": 7758
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.928102114710387,
+      "learning_rate": 5.414720191148822e-06,
+      "loss": 1.0862,
+      "step": 7759
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.349404483970881,
+      "learning_rate": 5.412264247302672e-06,
+      "loss": 1.0957,
+      "step": 7760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.195779585713389,
+      "learning_rate": 5.409808653890665e-06,
+      "loss": 1.082,
+      "step": 7761
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3795880046106124,
+      "learning_rate": 5.407353411100377e-06,
+      "loss": 1.0486,
+      "step": 7762
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.028081044429447,
+      "learning_rate": 5.404898519119353e-06,
+      "loss": 1.0691,
+      "step": 7763
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9411924892112697,
+      "learning_rate": 5.4024439781351045e-06,
+      "loss": 1.051,
+      "step": 7764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8234883435807414,
+      "learning_rate": 5.399989788335125e-06,
+      "loss": 1.1216,
+      "step": 7765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0723811477699883,
+      "learning_rate": 5.3975359499068736e-06,
+      "loss": 1.0305,
+      "step": 7766
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3791703379381084,
+      "learning_rate": 5.395082463037793e-06,
+      "loss": 1.1353,
+      "step": 7767
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.239798632082932,
+      "learning_rate": 5.392629327915289e-06,
+      "loss": 1.1028,
+      "step": 7768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.885873367987745,
+      "learning_rate": 5.390176544726742e-06,
+      "loss": 1.176,
+      "step": 7769
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.7083219813271375,
+      "learning_rate": 5.3877241136595174e-06,
+      "loss": 1.1711,
+      "step": 7770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6225923810947074,
+      "learning_rate": 5.38527203490093e-06,
+      "loss": 1.0759,
+      "step": 7771
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5844546172742913,
+      "learning_rate": 5.382820308638293e-06,
+      "loss": 1.0386,
+      "step": 7772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.37740451510782,
+      "learning_rate": 5.380368935058878e-06,
+      "loss": 1.1069,
+      "step": 7773
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5978691805694623,
+      "learning_rate": 5.377917914349929e-06,
+      "loss": 1.1709,
+      "step": 7774
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2691332026870388,
+      "learning_rate": 5.3754672466986756e-06,
+      "loss": 1.061,
+      "step": 7775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3725745234700737,
+      "learning_rate": 5.373016932292306e-06,
+      "loss": 1.1362,
+      "step": 7776
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.974864492916599,
+      "learning_rate": 5.37056697131799e-06,
+      "loss": 1.0571,
+      "step": 7777
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5733866848729052,
+      "learning_rate": 5.368117363962867e-06,
+      "loss": 1.0371,
+      "step": 7778
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.941773739467859,
+      "learning_rate": 5.3656681104140465e-06,
+      "loss": 1.1152,
+      "step": 7779
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.570580536069547,
+      "learning_rate": 5.363219210858621e-06,
+      "loss": 1.0876,
+      "step": 7780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2391847115468924,
+      "learning_rate": 5.360770665483647e-06,
+      "loss": 0.9985,
+      "step": 7781
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.031717629346211,
+      "learning_rate": 5.358322474476152e-06,
+      "loss": 1.011,
+      "step": 7782
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6247386615130504,
+      "learning_rate": 5.355874638023155e-06,
+      "loss": 1.0592,
+      "step": 7783
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1175124392217315,
+      "learning_rate": 5.353427156311615e-06,
+      "loss": 1.0486,
+      "step": 7784
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0939646005298114,
+      "learning_rate": 5.350980029528497e-06,
+      "loss": 1.073,
+      "step": 7785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.863368523368356,
+      "learning_rate": 5.3485332578607205e-06,
+      "loss": 1.1545,
+      "step": 7786
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7327578433846056,
+      "learning_rate": 5.3460868414951775e-06,
+      "loss": 1.006,
+      "step": 7787
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6238300484671404,
+      "learning_rate": 5.343640780618747e-06,
+      "loss": 1.0974,
+      "step": 7788
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6084041074055273,
+      "learning_rate": 5.34119507541826e-06,
+      "loss": 1.0918,
+      "step": 7789
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4251023607246465,
+      "learning_rate": 5.3387497260805465e-06,
+      "loss": 1.0188,
+      "step": 7790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6608102630430404,
+      "learning_rate": 5.336304732792381e-06,
+      "loss": 1.0513,
+      "step": 7791
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.2146670036626475,
+      "learning_rate": 5.333860095740526e-06,
+      "loss": 1.0371,
+      "step": 7792
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.20569473708455,
+      "learning_rate": 5.3314158151117226e-06,
+      "loss": 1.0789,
+      "step": 7793
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7306500232303157,
+      "learning_rate": 5.328971891092667e-06,
+      "loss": 1.0493,
+      "step": 7794
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8376596123447593,
+      "learning_rate": 5.326528323870049e-06,
+      "loss": 1.0684,
+      "step": 7795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0293341048084685,
+      "learning_rate": 5.324085113630515e-06,
+      "loss": 1.0283,
+      "step": 7796
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.53549059530251,
+      "learning_rate": 5.321642260560691e-06,
+      "loss": 1.0889,
+      "step": 7797
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2530407850819145,
+      "learning_rate": 5.319199764847171e-06,
+      "loss": 0.9695,
+      "step": 7798
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.564751944361956,
+      "learning_rate": 5.316757626676525e-06,
+      "loss": 1.0532,
+      "step": 7799
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.135324085116203,
+      "learning_rate": 5.3143158462353026e-06,
+      "loss": 1.0146,
+      "step": 7800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7930619584382885,
+      "learning_rate": 5.311874423710013e-06,
+      "loss": 1.0842,
+      "step": 7801
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.628389895889555,
+      "learning_rate": 5.309433359287142e-06,
+      "loss": 1.1235,
+      "step": 7802
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.8159640973873685,
+      "learning_rate": 5.306992653153163e-06,
+      "loss": 1.0527,
+      "step": 7803
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.346667015184154,
+      "learning_rate": 5.304552305494493e-06,
+      "loss": 1.054,
+      "step": 7804
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9391054307692204,
+      "learning_rate": 5.302112316497548e-06,
+      "loss": 1.0315,
+      "step": 7805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.349589313213296,
+      "learning_rate": 5.299672686348704e-06,
+      "loss": 1.0708,
+      "step": 7806
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.553163359544213,
+      "learning_rate": 5.297233415234308e-06,
+      "loss": 1.1448,
+      "step": 7807
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5485212876903187,
+      "learning_rate": 5.2947945033406925e-06,
+      "loss": 1.0679,
+      "step": 7808
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.235389031925269,
+      "learning_rate": 5.292355950854146e-06,
+      "loss": 1.1187,
+      "step": 7809
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.671072386719866,
+      "learning_rate": 5.289917757960943e-06,
+      "loss": 1.1211,
+      "step": 7810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7320853653796044,
+      "learning_rate": 5.287479924847322e-06,
+      "loss": 1.0994,
+      "step": 7811
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.961620357910944,
+      "learning_rate": 5.285042451699493e-06,
+      "loss": 1.0796,
+      "step": 7812
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5620736835102225,
+      "learning_rate": 5.2826053387036495e-06,
+      "loss": 1.0955,
+      "step": 7813
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.353667779396749,
+      "learning_rate": 5.280168586045947e-06,
+      "loss": 1.0945,
+      "step": 7814
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.882412356425423,
+      "learning_rate": 5.277732193912513e-06,
+      "loss": 1.1121,
+      "step": 7815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8933159777679087,
+      "learning_rate": 5.275296162489461e-06,
+      "loss": 1.1191,
+      "step": 7816
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.206990594510365,
+      "learning_rate": 5.272860491962859e-06,
+      "loss": 1.1353,
+      "step": 7817
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.675309554868618,
+      "learning_rate": 5.270425182518759e-06,
+      "loss": 1.0671,
+      "step": 7818
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8490215686908344,
+      "learning_rate": 5.267990234343182e-06,
+      "loss": 1.1052,
+      "step": 7819
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8352131488277528,
+      "learning_rate": 5.265555647622116e-06,
+      "loss": 1.075,
+      "step": 7820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.166166568454873,
+      "learning_rate": 5.2631214225415374e-06,
+      "loss": 1.1074,
+      "step": 7821
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6135750581885158,
+      "learning_rate": 5.260687559287376e-06,
+      "loss": 1.0515,
+      "step": 7822
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8394956955924333,
+      "learning_rate": 5.258254058045547e-06,
+      "loss": 1.0972,
+      "step": 7823
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0277196889151288,
+      "learning_rate": 5.25582091900193e-06,
+      "loss": 1.0732,
+      "step": 7824
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.98471965222699,
+      "learning_rate": 5.253388142342379e-06,
+      "loss": 1.1143,
+      "step": 7825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9514531275144917,
+      "learning_rate": 5.250955728252728e-06,
+      "loss": 1.0828,
+      "step": 7826
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9872430660667413,
+      "learning_rate": 5.248523676918773e-06,
+      "loss": 1.0781,
+      "step": 7827
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.074463476627633,
+      "learning_rate": 5.2460919885262825e-06,
+      "loss": 1.0747,
+      "step": 7828
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.885966364064832,
+      "learning_rate": 5.24366066326101e-06,
+      "loss": 1.0825,
+      "step": 7829
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.003404206256967,
+      "learning_rate": 5.2412297013086655e-06,
+      "loss": 1.0408,
+      "step": 7830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9745807613870596,
+      "learning_rate": 5.238799102854941e-06,
+      "loss": 1.158,
+      "step": 7831
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.066683254820525,
+      "learning_rate": 5.236368868085493e-06,
+      "loss": 1.1216,
+      "step": 7832
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2435432235127903,
+      "learning_rate": 5.233938997185961e-06,
+      "loss": 1.0701,
+      "step": 7833
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1701445054773933,
+      "learning_rate": 5.231509490341949e-06,
+      "loss": 1.0825,
+      "step": 7834
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1726768362033306,
+      "learning_rate": 5.229080347739031e-06,
+      "loss": 1.02,
+      "step": 7835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3815274630089247,
+      "learning_rate": 5.226651569562762e-06,
+      "loss": 1.1475,
+      "step": 7836
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.797932403990986,
+      "learning_rate": 5.224223155998663e-06,
+      "loss": 1.054,
+      "step": 7837
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234867390671982,
+      "learning_rate": 5.221795107232228e-06,
+      "loss": 1.1057,
+      "step": 7838
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.451567844697679,
+      "learning_rate": 5.219367423448922e-06,
+      "loss": 1.0063,
+      "step": 7839
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3501754918360276,
+      "learning_rate": 5.2169401048341815e-06,
+      "loss": 1.0504,
+      "step": 7840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2134228909706066,
+      "learning_rate": 5.214513151573424e-06,
+      "loss": 1.1023,
+      "step": 7841
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2638728103043686,
+      "learning_rate": 5.2120865638520305e-06,
+      "loss": 1.1455,
+      "step": 7842
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.097935569425755,
+      "learning_rate": 5.209660341855351e-06,
+      "loss": 1.1067,
+      "step": 7843
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.018238416024706,
+      "learning_rate": 5.2072344857687175e-06,
+      "loss": 1.1509,
+      "step": 7844
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3339602045480325,
+      "learning_rate": 5.2048089957774215e-06,
+      "loss": 1.1177,
+      "step": 7845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.789260102098737,
+      "learning_rate": 5.202383872066744e-06,
+      "loss": 1.135,
+      "step": 7846
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1080174921264154,
+      "learning_rate": 5.199959114821924e-06,
+      "loss": 1.0718,
+      "step": 7847
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.298636451503254,
+      "learning_rate": 5.197534724228173e-06,
+      "loss": 1.0759,
+      "step": 7848
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3654364991773584,
+      "learning_rate": 5.1951107004706825e-06,
+      "loss": 1.0815,
+      "step": 7849
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8086868351616996,
+      "learning_rate": 5.192687043734612e-06,
+      "loss": 1.0872,
+      "step": 7850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3200145772901544,
+      "learning_rate": 5.19026375420509e-06,
+      "loss": 1.1597,
+      "step": 7851
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.383821907819832,
+      "learning_rate": 5.187840832067221e-06,
+      "loss": 1.1196,
+      "step": 7852
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9408915501268824,
+      "learning_rate": 5.185418277506074e-06,
+      "loss": 1.0244,
+      "step": 7853
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.691741054158455,
+      "learning_rate": 5.182996090706704e-06,
+      "loss": 1.147,
+      "step": 7854
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9550848988848073,
+      "learning_rate": 5.180574271854129e-06,
+      "loss": 1.189,
+      "step": 7855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5531443894848245,
+      "learning_rate": 5.178152821133335e-06,
+      "loss": 1.0867,
+      "step": 7856
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8655592022075367,
+      "learning_rate": 5.175731738729289e-06,
+      "loss": 1.1101,
+      "step": 7857
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8973734121310413,
+      "learning_rate": 5.173311024826916e-06,
+      "loss": 1.0933,
+      "step": 7858
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2268514184874704,
+      "learning_rate": 5.170890679611135e-06,
+      "loss": 1.0852,
+      "step": 7859
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.102766416340537,
+      "learning_rate": 5.168470703266817e-06,
+      "loss": 1.0299,
+      "step": 7860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9055718974195357,
+      "learning_rate": 5.16605109597881e-06,
+      "loss": 1.0645,
+      "step": 7861
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.130378224357751,
+      "learning_rate": 5.163631857931942e-06,
+      "loss": 1.0254,
+      "step": 7862
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3496258756103257,
+      "learning_rate": 5.1612129893110045e-06,
+      "loss": 1.031,
+      "step": 7863
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.253809275257404,
+      "learning_rate": 5.15879449030076e-06,
+      "loss": 1.0391,
+      "step": 7864
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0529729567778023,
+      "learning_rate": 5.156376361085947e-06,
+      "loss": 1.1355,
+      "step": 7865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.854135758198435,
+      "learning_rate": 5.1539586018512725e-06,
+      "loss": 1.1423,
+      "step": 7866
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7992028362325057,
+      "learning_rate": 5.15154121278142e-06,
+      "loss": 1.072,
+      "step": 7867
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.217667880617609,
+      "learning_rate": 5.149124194061037e-06,
+      "loss": 1.0781,
+      "step": 7868
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.847207912475431,
+      "learning_rate": 5.146707545874756e-06,
+      "loss": 1.0737,
+      "step": 7869
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.081823584500817,
+      "learning_rate": 5.144291268407168e-06,
+      "loss": 1.0974,
+      "step": 7870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8625508425286403,
+      "learning_rate": 5.141875361842838e-06,
+      "loss": 1.0088,
+      "step": 7871
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8249583609762077,
+      "learning_rate": 5.139459826366309e-06,
+      "loss": 1.0901,
+      "step": 7872
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.809805764503435,
+      "learning_rate": 5.137044662162085e-06,
+      "loss": 1.1743,
+      "step": 7873
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2200568150982902,
+      "learning_rate": 5.134629869414655e-06,
+      "loss": 1.0498,
+      "step": 7874
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.601878292509014,
+      "learning_rate": 5.132215448308473e-06,
+      "loss": 1.1206,
+      "step": 7875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.143169954655654,
+      "learning_rate": 5.129801399027962e-06,
+      "loss": 1.0808,
+      "step": 7876
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4291526352522337,
+      "learning_rate": 5.127387721757519e-06,
+      "loss": 1.0605,
+      "step": 7877
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1921407473468055,
+      "learning_rate": 5.12497441668151e-06,
+      "loss": 1.0481,
+      "step": 7878
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.021790843304487,
+      "learning_rate": 5.122561483984282e-06,
+      "loss": 1.0435,
+      "step": 7879
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.571990698897547,
+      "learning_rate": 5.120148923850143e-06,
+      "loss": 1.1355,
+      "step": 7880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6020390811713083,
+      "learning_rate": 5.117736736463372e-06,
+      "loss": 1.1101,
+      "step": 7881
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.014108792579095,
+      "learning_rate": 5.1153249220082335e-06,
+      "loss": 1.0718,
+      "step": 7882
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.73817534945968,
+      "learning_rate": 5.112913480668948e-06,
+      "loss": 1.0999,
+      "step": 7883
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.108977469265138,
+      "learning_rate": 5.110502412629717e-06,
+      "loss": 1.0159,
+      "step": 7884
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7450000794986886,
+      "learning_rate": 5.108091718074706e-06,
+      "loss": 1.0239,
+      "step": 7885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2250970226087965,
+      "learning_rate": 5.105681397188052e-06,
+      "loss": 1.0913,
+      "step": 7886
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.04499517261428,
+      "learning_rate": 5.103271450153877e-06,
+      "loss": 1.0217,
+      "step": 7887
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.038145601899639,
+      "learning_rate": 5.10086187715626e-06,
+      "loss": 1.0862,
+      "step": 7888
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.233401401699132,
+      "learning_rate": 5.098452678379254e-06,
+      "loss": 1.1235,
+      "step": 7889
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.286088131633436,
+      "learning_rate": 5.096043854006896e-06,
+      "loss": 1.1096,
+      "step": 7890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.555595732722266,
+      "learning_rate": 5.093635404223168e-06,
+      "loss": 1.0676,
+      "step": 7891
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.609857865374256,
+      "learning_rate": 5.09122732921205e-06,
+      "loss": 1.0874,
+      "step": 7892
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.971347504201341,
+      "learning_rate": 5.088819629157482e-06,
+      "loss": 1.0898,
+      "step": 7893
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.485729607729159,
+      "learning_rate": 5.08641230424337e-06,
+      "loss": 1.1243,
+      "step": 7894
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5575710843910815,
+      "learning_rate": 5.084005354653605e-06,
+      "loss": 1.1169,
+      "step": 7895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.707861093258658,
+      "learning_rate": 5.081598780572041e-06,
+      "loss": 1.0325,
+      "step": 7896
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4420990463065855,
+      "learning_rate": 5.079192582182501e-06,
+      "loss": 1.1223,
+      "step": 7897
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.258047048698247,
+      "learning_rate": 5.076786759668783e-06,
+      "loss": 1.0693,
+      "step": 7898
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.754466220979059,
+      "learning_rate": 5.074381313214651e-06,
+      "loss": 1.0298,
+      "step": 7899
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8890137106053344,
+      "learning_rate": 5.071976243003856e-06,
+      "loss": 1.032,
+      "step": 7900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.793407006534772,
+      "learning_rate": 5.0695715492200996e-06,
+      "loss": 1.084,
+      "step": 7901
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9524376501499145,
+      "learning_rate": 5.067167232047071e-06,
+      "loss": 1.1067,
+      "step": 7902
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4022955564634274,
+      "learning_rate": 5.064763291668425e-06,
+      "loss": 1.0234,
+      "step": 7903
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1701420339118664,
+      "learning_rate": 5.062359728267775e-06,
+      "loss": 1.0969,
+      "step": 7904
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3647023853999927,
+      "learning_rate": 5.059956542028729e-06,
+      "loss": 1.0349,
+      "step": 7905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.244437184923285,
+      "learning_rate": 5.057553733134845e-06,
+      "loss": 1.1174,
+      "step": 7906
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.44008487499701,
+      "learning_rate": 5.0551513017696715e-06,
+      "loss": 1.0876,
+      "step": 7907
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2531550923965606,
+      "learning_rate": 5.052749248116714e-06,
+      "loss": 1.1365,
+      "step": 7908
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.162823365105146,
+      "learning_rate": 5.050347572359452e-06,
+      "loss": 1.2417,
+      "step": 7909
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2413003295140728,
+      "learning_rate": 5.047946274681339e-06,
+      "loss": 1.1052,
+      "step": 7910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0171312902529754,
+      "learning_rate": 5.045545355265792e-06,
+      "loss": 1.1309,
+      "step": 7911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5110093812243366,
+      "learning_rate": 5.043144814296214e-06,
+      "loss": 1.0862,
+      "step": 7912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0332015389710336,
+      "learning_rate": 5.040744651955969e-06,
+      "loss": 1.1162,
+      "step": 7913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0254112396713837,
+      "learning_rate": 5.0383448684283855e-06,
+      "loss": 1.1396,
+      "step": 7914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.977203778932076,
+      "learning_rate": 5.035945463896781e-06,
+      "loss": 1.1533,
+      "step": 7915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3495854228139255,
+      "learning_rate": 5.03354643854443e-06,
+      "loss": 1.1001,
+      "step": 7916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.804685306132309,
+      "learning_rate": 5.03114779255458e-06,
+      "loss": 1.1252,
+      "step": 7917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6285463005505973,
+      "learning_rate": 5.028749526110455e-06,
+      "loss": 1.0325,
+      "step": 7918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.177732801612835,
+      "learning_rate": 5.026351639395239e-06,
+      "loss": 1.1133,
+      "step": 7919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9539939942755287,
+      "learning_rate": 5.023954132592106e-06,
+      "loss": 1.0703,
+      "step": 7920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5543889385380307,
+      "learning_rate": 5.021557005884183e-06,
+      "loss": 1.0903,
+      "step": 7921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.610277953286171,
+      "learning_rate": 5.0191602594545705e-06,
+      "loss": 1.1438,
+      "step": 7922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5096137672564796,
+      "learning_rate": 5.016763893486358e-06,
+      "loss": 1.0776,
+      "step": 7923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.319604055895026,
+      "learning_rate": 5.014367908162574e-06,
+      "loss": 1.113,
+      "step": 7924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.155745321057377,
+      "learning_rate": 5.011972303666248e-06,
+      "loss": 1.0308,
+      "step": 7925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5125520636270546,
+      "learning_rate": 5.009577080180366e-06,
+      "loss": 1.0752,
+      "step": 7926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1075645416424766,
+      "learning_rate": 5.00718223788788e-06,
+      "loss": 1.0898,
+      "step": 7927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9389350805314005,
+      "learning_rate": 5.004787776971731e-06,
+      "loss": 1.0815,
+      "step": 7928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2046158817284907,
+      "learning_rate": 5.002393697614815e-06,
+      "loss": 1.1401,
+      "step": 7929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5559376298634278,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 1.0901,
+      "step": 7930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4123895691954393,
+      "learning_rate": 4.997606684310136e-06,
+      "loss": 1.124,
+      "step": 7931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.112017282963466,
+      "learning_rate": 4.995213750728028e-06,
+      "loss": 1.0767,
+      "step": 7932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6415061862748854,
+      "learning_rate": 4.992821199436469e-06,
+      "loss": 1.0583,
+      "step": 7933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2263063214396293,
+      "learning_rate": 4.990429030618204e-06,
+      "loss": 1.1318,
+      "step": 7934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.074432417349663,
+      "learning_rate": 4.9880372444559675e-06,
+      "loss": 1.0281,
+      "step": 7935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8858225033486846,
+      "learning_rate": 4.985645841132458e-06,
+      "loss": 1.0203,
+      "step": 7936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0002475215145648,
+      "learning_rate": 4.98325482083033e-06,
+      "loss": 1.0356,
+      "step": 7937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.879811222535483,
+      "learning_rate": 4.980864183732232e-06,
+      "loss": 1.1348,
+      "step": 7938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9401865155950855,
+      "learning_rate": 4.9784739300207675e-06,
+      "loss": 1.1121,
+      "step": 7939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.641518903074468,
+      "learning_rate": 4.976084059878522e-06,
+      "loss": 1.0959,
+      "step": 7940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9752792681431393,
+      "learning_rate": 4.973694573488043e-06,
+      "loss": 1.1113,
+      "step": 7941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0419461843829905,
+      "learning_rate": 4.971305471031845e-06,
+      "loss": 1.0181,
+      "step": 7942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1684090699898353,
+      "learning_rate": 4.968916752692434e-06,
+      "loss": 1.085,
+      "step": 7943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.013663284115837,
+      "learning_rate": 4.966528418652257e-06,
+      "loss": 1.0142,
+      "step": 7944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3102658899449726,
+      "learning_rate": 4.964140469093755e-06,
+      "loss": 1.1162,
+      "step": 7945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.57650293154375,
+      "learning_rate": 4.9617529041993305e-06,
+      "loss": 1.0867,
+      "step": 7946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9052699783721967,
+      "learning_rate": 4.959365724151354e-06,
+      "loss": 1.085,
+      "step": 7947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5303086540897963,
+      "learning_rate": 4.956978929132177e-06,
+      "loss": 1.1147,
+      "step": 7948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.686783295119144,
+      "learning_rate": 4.954592519324109e-06,
+      "loss": 1.0143,
+      "step": 7949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1964777630862637,
+      "learning_rate": 4.95220649490944e-06,
+      "loss": 1.095,
+      "step": 7950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.268663697608949,
+      "learning_rate": 4.949820856070424e-06,
+      "loss": 1.103,
+      "step": 7951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.883821619652889,
+      "learning_rate": 4.947435602989283e-06,
+      "loss": 1.0618,
+      "step": 7952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1495068519574105,
+      "learning_rate": 4.9450507358482256e-06,
+      "loss": 1.0454,
+      "step": 7953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5283887867336983,
+      "learning_rate": 4.942666254829414e-06,
+      "loss": 1.1589,
+      "step": 7954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.197851971415577,
+      "learning_rate": 4.940282160114982e-06,
+      "loss": 1.1384,
+      "step": 7955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5243080964334403,
+      "learning_rate": 4.937898451887051e-06,
+      "loss": 1.0872,
+      "step": 7956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9840019561508324,
+      "learning_rate": 4.935515130327686e-06,
+      "loss": 1.1228,
+      "step": 7957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.499937230305638,
+      "learning_rate": 4.933132195618948e-06,
+      "loss": 1.056,
+      "step": 7958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8183959046036664,
+      "learning_rate": 4.930749647942853e-06,
+      "loss": 1.0894,
+      "step": 7959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4310350326107146,
+      "learning_rate": 4.928367487481389e-06,
+      "loss": 1.125,
+      "step": 7960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3553511170266916,
+      "learning_rate": 4.925985714416525e-06,
+      "loss": 1.0549,
+      "step": 7961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.328027938437061,
+      "learning_rate": 4.923604328930187e-06,
+      "loss": 1.0898,
+      "step": 7962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4097903848475206,
+      "learning_rate": 4.921223331204279e-06,
+      "loss": 1.073,
+      "step": 7963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.64648049493273,
+      "learning_rate": 4.918842721420673e-06,
+      "loss": 1.0942,
+      "step": 7964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4786641607533775,
+      "learning_rate": 4.916462499761207e-06,
+      "loss": 1.0647,
+      "step": 7965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.347722090646202,
+      "learning_rate": 4.914082666407705e-06,
+      "loss": 1.1099,
+      "step": 7966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8955541359346433,
+      "learning_rate": 4.911703221541943e-06,
+      "loss": 1.0574,
+      "step": 7967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.371584695114752,
+      "learning_rate": 4.909324165345673e-06,
+      "loss": 1.0737,
+      "step": 7968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.232467211955694,
+      "learning_rate": 4.90694549800063e-06,
+      "loss": 1.0635,
+      "step": 7969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.124455897230069,
+      "learning_rate": 4.904567219688494e-06,
+      "loss": 1.1062,
+      "step": 7970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1695759466100055,
+      "learning_rate": 4.9021893305909405e-06,
+      "loss": 1.0706,
+      "step": 7971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.919775622727679,
+      "learning_rate": 4.899811830889597e-06,
+      "loss": 1.0906,
+      "step": 7972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2073142448685794,
+      "learning_rate": 4.8974347207660766e-06,
+      "loss": 1.1387,
+      "step": 7973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4714541768417138,
+      "learning_rate": 4.895058000401951e-06,
+      "loss": 1.0623,
+      "step": 7974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0070214806667086,
+      "learning_rate": 4.892681669978762e-06,
+      "loss": 1.0413,
+      "step": 7975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4665478221999018,
+      "learning_rate": 4.890305729678037e-06,
+      "loss": 1.1399,
+      "step": 7976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4536513604613583,
+      "learning_rate": 4.887930179681246e-06,
+      "loss": 1.1401,
+      "step": 7977
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2953877068864323,
+      "learning_rate": 4.885555020169859e-06,
+      "loss": 1.0166,
+      "step": 7978
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.313600351498647,
+      "learning_rate": 4.883180251325298e-06,
+      "loss": 1.0605,
+      "step": 7979
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.882038296757231,
+      "learning_rate": 4.880805873328954e-06,
+      "loss": 1.0,
+      "step": 7980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.881355801764668,
+      "learning_rate": 4.8784318863622036e-06,
+      "loss": 1.0676,
+      "step": 7981
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.962357929419293,
+      "learning_rate": 4.876058290606379e-06,
+      "loss": 1.0601,
+      "step": 7982
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.408644747609404,
+      "learning_rate": 4.8736850862427875e-06,
+      "loss": 1.1235,
+      "step": 7983
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2958213270138637,
+      "learning_rate": 4.871312273452706e-06,
+      "loss": 1.0474,
+      "step": 7984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.260263388704065,
+      "learning_rate": 4.868939852417379e-06,
+      "loss": 1.0469,
+      "step": 7985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3315557410845025,
+      "learning_rate": 4.86656782331803e-06,
+      "loss": 1.0596,
+      "step": 7986
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2453233946518414,
+      "learning_rate": 4.864196186335845e-06,
+      "loss": 1.0605,
+      "step": 7987
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.918515453867572,
+      "learning_rate": 4.8618249416519735e-06,
+      "loss": 1.0571,
+      "step": 7988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9879822961701366,
+      "learning_rate": 4.859454089447558e-06,
+      "loss": 1.1133,
+      "step": 7989
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.202628815558747,
+      "learning_rate": 4.85708362990368e-06,
+      "loss": 1.0042,
+      "step": 7990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3861809474566624,
+      "learning_rate": 4.854713563201419e-06,
+      "loss": 1.0674,
+      "step": 7991
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.519783060229944,
+      "learning_rate": 4.8523438895218066e-06,
+      "loss": 1.1245,
+      "step": 7992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2608411509156023,
+      "learning_rate": 4.849974609045849e-06,
+      "loss": 1.1069,
+      "step": 7993
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2724650836754385,
+      "learning_rate": 4.847605721954528e-06,
+      "loss": 1.1062,
+      "step": 7994
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.253947505884016,
+      "learning_rate": 4.845237228428791e-06,
+      "loss": 1.0552,
+      "step": 7995
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9260600870563542,
+      "learning_rate": 4.842869128649554e-06,
+      "loss": 1.0754,
+      "step": 7996
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.498968712854927,
+      "learning_rate": 4.840501422797702e-06,
+      "loss": 1.0815,
+      "step": 7997
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9040578664932344,
+      "learning_rate": 4.838134111054092e-06,
+      "loss": 1.0525,
+      "step": 7998
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3485664075740207,
+      "learning_rate": 4.835767193599556e-06,
+      "loss": 1.0723,
+      "step": 7999
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1651042935482825,
+      "learning_rate": 4.833400670614887e-06,
+      "loss": 1.0874,
+      "step": 8000
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1608670672144394,
+      "learning_rate": 4.831034542280852e-06,
+      "loss": 1.0649,
+      "step": 8001
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4203643519393045,
+      "learning_rate": 4.828668808778194e-06,
+      "loss": 1.1045,
+      "step": 8002
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.548080150449681,
+      "learning_rate": 4.826303470287607e-06,
+      "loss": 1.093,
+      "step": 8003
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2721201257586228,
+      "learning_rate": 4.823938526989779e-06,
+      "loss": 1.1184,
+      "step": 8004
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7879276336597045,
+      "learning_rate": 4.821573979065351e-06,
+      "loss": 1.0579,
+      "step": 8005
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1046626102574604,
+      "learning_rate": 4.819209826694936e-06,
+      "loss": 1.0911,
+      "step": 8006
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.063087399139474,
+      "learning_rate": 4.816846070059127e-06,
+      "loss": 1.0386,
+      "step": 8007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4695474949911462,
+      "learning_rate": 4.814482709338473e-06,
+      "loss": 1.1245,
+      "step": 8008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.409903551376479,
+      "learning_rate": 4.812119744713509e-06,
+      "loss": 1.1377,
+      "step": 8009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6727007412022865,
+      "learning_rate": 4.80975717636472e-06,
+      "loss": 1.0876,
+      "step": 8010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.229542837850368,
+      "learning_rate": 4.807395004472569e-06,
+      "loss": 1.0688,
+      "step": 8011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.860338424305135,
+      "learning_rate": 4.805033229217501e-06,
+      "loss": 1.1516,
+      "step": 8012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.124399715600373,
+      "learning_rate": 4.80267185077991e-06,
+      "loss": 1.0203,
+      "step": 8013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.561793080209247,
+      "learning_rate": 4.800310869340178e-06,
+      "loss": 1.0371,
+      "step": 8014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.072822105860809,
+      "learning_rate": 4.797950285078647e-06,
+      "loss": 1.0347,
+      "step": 8015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4763395783126207,
+      "learning_rate": 4.795590098175625e-06,
+      "loss": 1.0864,
+      "step": 8016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.600607718688167,
+      "learning_rate": 4.7932303088114e-06,
+      "loss": 1.1064,
+      "step": 8017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4281315591072596,
+      "learning_rate": 4.79087091716622e-06,
+      "loss": 1.1516,
+      "step": 8018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3374839534143828,
+      "learning_rate": 4.7885119234203114e-06,
+      "loss": 1.1179,
+      "step": 8019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.182994146356331,
+      "learning_rate": 4.786153327753865e-06,
+      "loss": 1.115,
+      "step": 8020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9101009870888888,
+      "learning_rate": 4.783795130347038e-06,
+      "loss": 1.1045,
+      "step": 8021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.098222807758857,
+      "learning_rate": 4.781437331379972e-06,
+      "loss": 1.0601,
+      "step": 8022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3968256143633675,
+      "learning_rate": 4.779079931032752e-06,
+      "loss": 1.0881,
+      "step": 8023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2354123841414286,
+      "learning_rate": 4.77672292948546e-06,
+      "loss": 1.0728,
+      "step": 8024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2263864170131615,
+      "learning_rate": 4.774366326918132e-06,
+      "loss": 1.0457,
+      "step": 8025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8882196901493353,
+      "learning_rate": 4.772010123510773e-06,
+      "loss": 1.1155,
+      "step": 8026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.367503502775311,
+      "learning_rate": 4.769654319443369e-06,
+      "loss": 1.0786,
+      "step": 8027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.000264656215905,
+      "learning_rate": 4.767298914895864e-06,
+      "loss": 1.0242,
+      "step": 8028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5019809104276702,
+      "learning_rate": 4.764943910048175e-06,
+      "loss": 1.0703,
+      "step": 8029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.277671126830718,
+      "learning_rate": 4.762589305080191e-06,
+      "loss": 1.156,
+      "step": 8030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.591316260673902,
+      "learning_rate": 4.760235100171764e-06,
+      "loss": 1.0901,
+      "step": 8031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6960984525001166,
+      "learning_rate": 4.757881295502725e-06,
+      "loss": 1.0408,
+      "step": 8032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.93116024257111,
+      "learning_rate": 4.75552789125287e-06,
+      "loss": 1.0911,
+      "step": 8033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.614214122240674,
+      "learning_rate": 4.753174887601956e-06,
+      "loss": 1.1001,
+      "step": 8034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2758526929658602,
+      "learning_rate": 4.750822284729727e-06,
+      "loss": 1.0803,
+      "step": 8035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.44721532560724,
+      "learning_rate": 4.7484700828158825e-06,
+      "loss": 1.1438,
+      "step": 8036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1619757278119116,
+      "learning_rate": 4.7461182820400945e-06,
+      "loss": 1.0364,
+      "step": 8037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.501395999118339,
+      "learning_rate": 4.743766882582006e-06,
+      "loss": 1.1262,
+      "step": 8038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6594637942983694,
+      "learning_rate": 4.741415884621225e-06,
+      "loss": 1.0461,
+      "step": 8039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0357107478023146,
+      "learning_rate": 4.739065288337338e-06,
+      "loss": 1.0752,
+      "step": 8040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.923945812769377,
+      "learning_rate": 4.736715093909891e-06,
+      "loss": 1.0549,
+      "step": 8041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3883953058573915,
+      "learning_rate": 4.734365301518412e-06,
+      "loss": 1.0479,
+      "step": 8042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1333685631752477,
+      "learning_rate": 4.7320159113423804e-06,
+      "loss": 1.1013,
+      "step": 8043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0239313422148393,
+      "learning_rate": 4.729666923561254e-06,
+      "loss": 1.0686,
+      "step": 8044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.932512767989363,
+      "learning_rate": 4.727318338354467e-06,
+      "loss": 1.0642,
+      "step": 8045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.455638089269625,
+      "learning_rate": 4.724970155901409e-06,
+      "loss": 1.1079,
+      "step": 8046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.593092722754008,
+      "learning_rate": 4.722622376381455e-06,
+      "loss": 1.0637,
+      "step": 8047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5061308149357813,
+      "learning_rate": 4.720274999973933e-06,
+      "loss": 1.1038,
+      "step": 8048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.131192630890197,
+      "learning_rate": 4.717928026858149e-06,
+      "loss": 1.1211,
+      "step": 8049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0224115883799594,
+      "learning_rate": 4.715581457213377e-06,
+      "loss": 1.1438,
+      "step": 8050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.358643921878671,
+      "learning_rate": 4.713235291218856e-06,
+      "loss": 1.0793,
+      "step": 8051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.734115497587853,
+      "learning_rate": 4.710889529053806e-06,
+      "loss": 1.1033,
+      "step": 8052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.335913961665844,
+      "learning_rate": 4.708544170897401e-06,
+      "loss": 1.1238,
+      "step": 8053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1386110831630143,
+      "learning_rate": 4.706199216928791e-06,
+      "loss": 1.1409,
+      "step": 8054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3277070301825207,
+      "learning_rate": 4.703854667327105e-06,
+      "loss": 1.0596,
+      "step": 8055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.193651664214522,
+      "learning_rate": 4.701510522271416e-06,
+      "loss": 1.0566,
+      "step": 8056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.557300973473248,
+      "learning_rate": 4.699166781940794e-06,
+      "loss": 1.0894,
+      "step": 8057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3213456795949754,
+      "learning_rate": 4.69682344651426e-06,
+      "loss": 1.0862,
+      "step": 8058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1461545013862042,
+      "learning_rate": 4.694480516170807e-06,
+      "loss": 1.0974,
+      "step": 8059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.191215945444836,
+      "learning_rate": 4.692137991089408e-06,
+      "loss": 1.0887,
+      "step": 8060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.413500065571735,
+      "learning_rate": 4.689795871448992e-06,
+      "loss": 1.0823,
+      "step": 8061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.658996669716662,
+      "learning_rate": 4.687454157428462e-06,
+      "loss": 1.0754,
+      "step": 8062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5095745939797274,
+      "learning_rate": 4.685112849206691e-06,
+      "loss": 1.0671,
+      "step": 8063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.602178838611705,
+      "learning_rate": 4.682771946962513e-06,
+      "loss": 1.1467,
+      "step": 8064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.895329069439104,
+      "learning_rate": 4.680431450874748e-06,
+      "loss": 1.0605,
+      "step": 8065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7145291608510216,
+      "learning_rate": 4.6780913611221714e-06,
+      "loss": 1.0759,
+      "step": 8066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.545425635791185,
+      "learning_rate": 4.675751677883524e-06,
+      "loss": 1.0779,
+      "step": 8067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2265143439945807,
+      "learning_rate": 4.673412401337533e-06,
+      "loss": 1.1016,
+      "step": 8068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3182234969829407,
+      "learning_rate": 4.6710735316628785e-06,
+      "loss": 1.0759,
+      "step": 8069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7008415604451077,
+      "learning_rate": 4.668735069038216e-06,
+      "loss": 1.0281,
+      "step": 8070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.413674381321534,
+      "learning_rate": 4.666397013642169e-06,
+      "loss": 1.0205,
+      "step": 8071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2615057735094997,
+      "learning_rate": 4.664059365653325e-06,
+      "loss": 1.064,
+      "step": 8072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8979225061010943,
+      "learning_rate": 4.6617221252502545e-06,
+      "loss": 1.105,
+      "step": 8073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.413609710716933,
+      "learning_rate": 4.659385292611479e-06,
+      "loss": 1.095,
+      "step": 8074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4911008268057926,
+      "learning_rate": 4.657048867915508e-06,
+      "loss": 1.072,
+      "step": 8075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5235602332094227,
+      "learning_rate": 4.654712851340799e-06,
+      "loss": 1.1047,
+      "step": 8076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5360185311831147,
+      "learning_rate": 4.6523772430657885e-06,
+      "loss": 1.0762,
+      "step": 8077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9800493214276615,
+      "learning_rate": 4.6500420432688894e-06,
+      "loss": 1.0742,
+      "step": 8078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8737895565386347,
+      "learning_rate": 4.647707252128469e-06,
+      "loss": 1.1428,
+      "step": 8079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2790710988914022,
+      "learning_rate": 4.645372869822876e-06,
+      "loss": 1.1318,
+      "step": 8080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3429575332255954,
+      "learning_rate": 4.6430388965304196e-06,
+      "loss": 1.124,
+      "step": 8081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.997344243326429,
+      "learning_rate": 4.64070533242938e-06,
+      "loss": 1.0571,
+      "step": 8082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3654061366044177,
+      "learning_rate": 4.638372177698007e-06,
+      "loss": 1.115,
+      "step": 8083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1363797230983184,
+      "learning_rate": 4.636039432514515e-06,
+      "loss": 0.9946,
+      "step": 8084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8326186634755057,
+      "learning_rate": 4.6337070970570985e-06,
+      "loss": 1.0862,
+      "step": 8085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.180954219648412,
+      "learning_rate": 4.631375171503907e-06,
+      "loss": 1.0562,
+      "step": 8086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.243060939404797,
+      "learning_rate": 4.629043656033063e-06,
+      "loss": 1.1016,
+      "step": 8087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.197848336087629,
+      "learning_rate": 4.6267125508226665e-06,
+      "loss": 1.0542,
+      "step": 8088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.778209327482213,
+      "learning_rate": 4.624381856050773e-06,
+      "loss": 1.0693,
+      "step": 8089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1495309623195897,
+      "learning_rate": 4.622051571895416e-06,
+      "loss": 1.0762,
+      "step": 8090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.983266320089322,
+      "learning_rate": 4.619721698534593e-06,
+      "loss": 1.0271,
+      "step": 8091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.054063959884939,
+      "learning_rate": 4.617392236146267e-06,
+      "loss": 1.1357,
+      "step": 8092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9395761899755155,
+      "learning_rate": 4.6150631849083805e-06,
+      "loss": 1.0586,
+      "step": 8093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0640860879321816,
+      "learning_rate": 4.6127345449988355e-06,
+      "loss": 1.1091,
+      "step": 8094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.809008740410849,
+      "learning_rate": 4.610406316595506e-06,
+      "loss": 1.1033,
+      "step": 8095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0821925651072983,
+      "learning_rate": 4.608078499876233e-06,
+      "loss": 1.1072,
+      "step": 8096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7206935439108832,
+      "learning_rate": 4.605751095018822e-06,
+      "loss": 1.0659,
+      "step": 8097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5761879477095455,
+      "learning_rate": 4.6034241022010604e-06,
+      "loss": 1.1179,
+      "step": 8098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.834163067131266,
+      "learning_rate": 4.601097521600692e-06,
+      "loss": 1.1184,
+      "step": 8099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.315053884601422,
+      "learning_rate": 4.598771353395427e-06,
+      "loss": 1.0723,
+      "step": 8100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.478153518143093,
+      "learning_rate": 4.59644559776296e-06,
+      "loss": 1.0759,
+      "step": 8101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1995287640633383,
+      "learning_rate": 4.594120254880939e-06,
+      "loss": 1.0647,
+      "step": 8102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1188803273359174,
+      "learning_rate": 4.591795324926983e-06,
+      "loss": 1.1206,
+      "step": 8103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.040860630879984,
+      "learning_rate": 4.589470808078686e-06,
+      "loss": 1.0908,
+      "step": 8104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7898763632997903,
+      "learning_rate": 4.587146704513599e-06,
+      "loss": 1.103,
+      "step": 8105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8662451838700815,
+      "learning_rate": 4.584823014409256e-06,
+      "loss": 1.0811,
+      "step": 8106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3420252883566715,
+      "learning_rate": 4.582499737943152e-06,
+      "loss": 1.0659,
+      "step": 8107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4104620548167635,
+      "learning_rate": 4.580176875292746e-06,
+      "loss": 1.0342,
+      "step": 8108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.929559627047017,
+      "learning_rate": 4.577854426635474e-06,
+      "loss": 1.0481,
+      "step": 8109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.956884128612631,
+      "learning_rate": 4.575532392148729e-06,
+      "loss": 1.043,
+      "step": 8110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3144332155136595,
+      "learning_rate": 4.5732107720098885e-06,
+      "loss": 1.1157,
+      "step": 8111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5968409557009444,
+      "learning_rate": 4.570889566396286e-06,
+      "loss": 1.0872,
+      "step": 8112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0918722464326254,
+      "learning_rate": 4.5685687754852225e-06,
+      "loss": 1.1199,
+      "step": 8113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.729107779878946,
+      "learning_rate": 4.566248399453979e-06,
+      "loss": 1.0598,
+      "step": 8114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2723207201098585,
+      "learning_rate": 4.563928438479793e-06,
+      "loss": 1.0784,
+      "step": 8115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2702046840768606,
+      "learning_rate": 4.5616088927398775e-06,
+      "loss": 1.0532,
+      "step": 8116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4539410742850842,
+      "learning_rate": 4.559289762411404e-06,
+      "loss": 1.1084,
+      "step": 8117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.83623296893545,
+      "learning_rate": 4.556971047671528e-06,
+      "loss": 1.1233,
+      "step": 8118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.585073295381896,
+      "learning_rate": 4.554652748697361e-06,
+      "loss": 1.0784,
+      "step": 8119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.543717299019611,
+      "learning_rate": 4.552334865665982e-06,
+      "loss": 1.1624,
+      "step": 8120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.186646666469046,
+      "learning_rate": 4.550017398754451e-06,
+      "loss": 1.0029,
+      "step": 8121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.293458426625809,
+      "learning_rate": 4.547700348139782e-06,
+      "loss": 1.0452,
+      "step": 8122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.001166513061953,
+      "learning_rate": 4.5453837139989655e-06,
+      "loss": 1.1025,
+      "step": 8123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6679199779688587,
+      "learning_rate": 4.5430674965089546e-06,
+      "loss": 1.0645,
+      "step": 8124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.732791564310053,
+      "learning_rate": 4.5407516958466725e-06,
+      "loss": 1.0803,
+      "step": 8125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.346809018349083,
+      "learning_rate": 4.538436312189017e-06,
+      "loss": 1.1655,
+      "step": 8126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.929299777123997,
+      "learning_rate": 4.536121345712847e-06,
+      "loss": 1.1025,
+      "step": 8127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.904297199197942,
+      "learning_rate": 4.533806796594989e-06,
+      "loss": 1.0396,
+      "step": 8128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9670317780040056,
+      "learning_rate": 4.531492665012242e-06,
+      "loss": 1.0405,
+      "step": 8129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3132383017487452,
+      "learning_rate": 4.529178951141364e-06,
+      "loss": 1.0737,
+      "step": 8130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.302131524637499,
+      "learning_rate": 4.526865655159099e-06,
+      "loss": 1.0847,
+      "step": 8131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3469761685678057,
+      "learning_rate": 4.524552777242143e-06,
+      "loss": 1.1033,
+      "step": 8132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.782950992708288,
+      "learning_rate": 4.5222403175671615e-06,
+      "loss": 1.052,
+      "step": 8133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.253831053998363,
+      "learning_rate": 4.519928276310798e-06,
+      "loss": 1.0969,
+      "step": 8134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.579125139905598,
+      "learning_rate": 4.517616653649655e-06,
+      "loss": 1.1191,
+      "step": 8135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.703421663948821,
+      "learning_rate": 4.5153054497603065e-06,
+      "loss": 1.134,
+      "step": 8136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.325361972379117,
+      "learning_rate": 4.5129946648192926e-06,
+      "loss": 1.0825,
+      "step": 8137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6104167342479307,
+      "learning_rate": 4.5106842990031185e-06,
+      "loss": 1.009,
+      "step": 8138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2715861064124523,
+      "learning_rate": 4.5083743524882705e-06,
+      "loss": 1.0718,
+      "step": 8139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.405035274240804,
+      "learning_rate": 4.50606482545119e-06,
+      "loss": 1.1101,
+      "step": 8140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2350365132425756,
+      "learning_rate": 4.503755718068285e-06,
+      "loss": 1.0825,
+      "step": 8141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3030460823609595,
+      "learning_rate": 4.5014470305159485e-06,
+      "loss": 1.116,
+      "step": 8142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6653603021151766,
+      "learning_rate": 4.4991387629705156e-06,
+      "loss": 1.0771,
+      "step": 8143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.524488296371772,
+      "learning_rate": 4.496830915608312e-06,
+      "loss": 1.0984,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6110356520789164,
+      "learning_rate": 4.494523488605622e-06,
+      "loss": 1.0581,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4513444799330637,
+      "learning_rate": 4.4922164821386926e-06,
+      "loss": 1.156,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5750252834283627,
+      "learning_rate": 4.489909896383752e-06,
+      "loss": 1.0486,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7367513184391252,
+      "learning_rate": 4.487603731516986e-06,
+      "loss": 1.1824,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.192620557513318,
+      "learning_rate": 4.485297987714551e-06,
+      "loss": 1.0093,
+      "step": 8149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8127159144221396,
+      "learning_rate": 4.48299266515257e-06,
+      "loss": 1.0459,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.881474143721595,
+      "learning_rate": 4.480687764007132e-06,
+      "loss": 1.0659,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9340214948428978,
+      "learning_rate": 4.478383284454305e-06,
+      "loss": 1.0171,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8725643161954655,
+      "learning_rate": 4.476079226670108e-06,
+      "loss": 1.0803,
+      "step": 8153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4728985697602393,
+      "learning_rate": 4.473775590830545e-06,
+      "loss": 1.0945,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.571369607302428,
+      "learning_rate": 4.471472377111574e-06,
+      "loss": 1.0771,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8214480672628537,
+      "learning_rate": 4.469169585689127e-06,
+      "loss": 1.1365,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0549694832666887,
+      "learning_rate": 4.466867216739102e-06,
+      "loss": 1.0649,
+      "step": 8157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4698241317004777,
+      "learning_rate": 4.464565270437363e-06,
+      "loss": 1.0325,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1306807063364324,
+      "learning_rate": 4.462263746959751e-06,
+      "loss": 1.1306,
+      "step": 8159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8472930125622695,
+      "learning_rate": 4.459962646482063e-06,
+      "loss": 1.0935,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3334447815699235,
+      "learning_rate": 4.4576619691800675e-06,
+      "loss": 1.0486,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.944400501242264,
+      "learning_rate": 4.455361715229509e-06,
+      "loss": 1.0854,
+      "step": 8162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.833084300273322,
+      "learning_rate": 4.4530618848060814e-06,
+      "loss": 1.0325,
+      "step": 8163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9697110076266116,
+      "learning_rate": 4.450762478085465e-06,
+      "loss": 1.0767,
+      "step": 8164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1082079352926146,
+      "learning_rate": 4.448463495243299e-06,
+      "loss": 1.0908,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.714001796500367,
+      "learning_rate": 4.446164936455186e-06,
+      "loss": 1.073,
+      "step": 8166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0473834105726705,
+      "learning_rate": 4.443866801896709e-06,
+      "loss": 1.1079,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5087972917468226,
+      "learning_rate": 4.4415690917434075e-06,
+      "loss": 1.0964,
+      "step": 8168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9644455077857126,
+      "learning_rate": 4.439271806170792e-06,
+      "loss": 1.0671,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.915603801840867,
+      "learning_rate": 4.436974945354339e-06,
+      "loss": 0.988,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.668067837434782,
+      "learning_rate": 4.434678509469494e-06,
+      "loss": 1.0952,
+      "step": 8171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2527940665878408,
+      "learning_rate": 4.432382498691675e-06,
+      "loss": 1.1279,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.222515968029398,
+      "learning_rate": 4.43008691319626e-06,
+      "loss": 1.0742,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.243861487022169,
+      "learning_rate": 4.427791753158594e-06,
+      "loss": 1.0342,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5015700996983306,
+      "learning_rate": 4.425497018754003e-06,
+      "loss": 1.1113,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9070132190568816,
+      "learning_rate": 4.423202710157755e-06,
+      "loss": 1.0391,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.040598024746966,
+      "learning_rate": 4.420908827545112e-06,
+      "loss": 1.0706,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1768519108811484,
+      "learning_rate": 4.418615371091289e-06,
+      "loss": 1.0835,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.965380944697059,
+      "learning_rate": 4.416322340971469e-06,
+      "loss": 1.0728,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8779549164761615,
+      "learning_rate": 4.414029737360811e-06,
+      "loss": 1.0286,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.396910140589233,
+      "learning_rate": 4.4117375604344325e-06,
+      "loss": 1.1042,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6081280585076483,
+      "learning_rate": 4.409445810367421e-06,
+      "loss": 1.1025,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7916683002569522,
+      "learning_rate": 4.407154487334832e-06,
+      "loss": 0.9438,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4443541965515143,
+      "learning_rate": 4.404863591511685e-06,
+      "loss": 1.0767,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1125041782827942,
+      "learning_rate": 4.402573123072975e-06,
+      "loss": 1.0359,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5249584050601093,
+      "learning_rate": 4.400283082193656e-06,
+      "loss": 1.1467,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4313505379997773,
+      "learning_rate": 4.397993469048657e-06,
+      "loss": 1.1194,
+      "step": 8187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.399671226969621,
+      "learning_rate": 4.395704283812872e-06,
+      "loss": 1.0671,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.184629424990344,
+      "learning_rate": 4.393415526661146e-06,
+      "loss": 1.0869,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3604895295831287,
+      "learning_rate": 4.391127197768322e-06,
+      "loss": 1.092,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.841682170489833,
+      "learning_rate": 4.388839297309182e-06,
+      "loss": 1.0583,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.811430480609307,
+      "learning_rate": 4.386551825458497e-06,
+      "loss": 1.0603,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3864933098928542,
+      "learning_rate": 4.384264782390992e-06,
+      "loss": 1.0786,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0323577225923852,
+      "learning_rate": 4.381978168281357e-06,
+      "loss": 1.0923,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7591426536684267,
+      "learning_rate": 4.379691983304269e-06,
+      "loss": 1.0161,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.563157493127595,
+      "learning_rate": 4.377406227634341e-06,
+      "loss": 1.0801,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.904615478702619,
+      "learning_rate": 4.375120901446182e-06,
+      "loss": 1.0774,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.819447782888603,
+      "learning_rate": 4.372836004914354e-06,
+      "loss": 0.9783,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.211583854713889,
+      "learning_rate": 4.370551538213383e-06,
+      "loss": 1.1382,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6869396863975417,
+      "learning_rate": 4.368267501517778e-06,
+      "loss": 1.0601,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.917404643633627,
+      "learning_rate": 4.365983895002e-06,
+      "loss": 1.0979,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.824342356954902,
+      "learning_rate": 4.3637007188404825e-06,
+      "loss": 1.0789,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.523778956746517,
+      "learning_rate": 4.361417973207625e-06,
+      "loss": 1.0332,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.188679678582715,
+      "learning_rate": 4.359135658277792e-06,
+      "loss": 1.0969,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.830992924324766,
+      "learning_rate": 4.356853774225327e-06,
+      "loss": 1.0698,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0575134126627024,
+      "learning_rate": 4.354572321224525e-06,
+      "loss": 1.061,
+      "step": 8206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2739184586445207,
+      "learning_rate": 4.352291299449654e-06,
+      "loss": 1.1187,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.220713634051637,
+      "learning_rate": 4.350010709074959e-06,
+      "loss": 1.1016,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.356738144681438,
+      "learning_rate": 4.347730550274628e-06,
+      "loss": 1.0327,
+      "step": 8209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.464226208722416,
+      "learning_rate": 4.3454508232228435e-06,
+      "loss": 1.0562,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1755300099300907,
+      "learning_rate": 4.343171528093737e-06,
+      "loss": 1.0884,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5154534361621166,
+      "learning_rate": 4.34089266506141e-06,
+      "loss": 1.0369,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.821202559711361,
+      "learning_rate": 4.338614234299941e-06,
+      "loss": 1.0229,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3611450281208826,
+      "learning_rate": 4.336336235983359e-06,
+      "loss": 1.1467,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.571276905749683,
+      "learning_rate": 4.334058670285682e-06,
+      "loss": 1.0479,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6290453437873733,
+      "learning_rate": 4.331781537380869e-06,
+      "loss": 1.1936,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.064078652293366,
+      "learning_rate": 4.32950483744286e-06,
+      "loss": 0.9758,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.140811664190318,
+      "learning_rate": 4.327228570645566e-06,
+      "loss": 1.1206,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2615257151811794,
+      "learning_rate": 4.324952737162855e-06,
+      "loss": 1.0652,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.175996376038195,
+      "learning_rate": 4.322677337168573e-06,
+      "loss": 1.0837,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1414874421933114,
+      "learning_rate": 4.320402370836525e-06,
+      "loss": 1.1033,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9365666197268068,
+      "learning_rate": 4.318127838340476e-06,
+      "loss": 1.0823,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.945006909472874,
+      "learning_rate": 4.315853739854173e-06,
+      "loss": 1.1091,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9575000385676815,
+      "learning_rate": 4.3135800755513205e-06,
+      "loss": 1.0103,
+      "step": 8224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3913774706710837,
+      "learning_rate": 4.311306845605595e-06,
+      "loss": 1.1521,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3332577765893157,
+      "learning_rate": 4.309034050190637e-06,
+      "loss": 1.134,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0976731473821446,
+      "learning_rate": 4.306761689480049e-06,
+      "loss": 1.0889,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.339874095249073,
+      "learning_rate": 4.3044897636474155e-06,
+      "loss": 1.0833,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6120471086474484,
+      "learning_rate": 4.302218272866264e-06,
+      "loss": 1.198,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0119700456088867,
+      "learning_rate": 4.299947217310115e-06,
+      "loss": 1.094,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.120386116097113,
+      "learning_rate": 4.297676597152435e-06,
+      "loss": 1.0903,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.282272080483493,
+      "learning_rate": 4.295406412566664e-06,
+      "loss": 1.0596,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9702055195817736,
+      "learning_rate": 4.293136663726217e-06,
+      "loss": 1.0474,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0816617131191655,
+      "learning_rate": 4.290867350804466e-06,
+      "loss": 1.0542,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9956920684322412,
+      "learning_rate": 4.288598473974753e-06,
+      "loss": 1.1023,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.16304552532949,
+      "learning_rate": 4.286330033410384e-06,
+      "loss": 1.0129,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9502464891143516,
+      "learning_rate": 4.2840620292846314e-06,
+      "loss": 1.0762,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5321131175983815,
+      "learning_rate": 4.281794461770743e-06,
+      "loss": 1.0452,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8300073154415677,
+      "learning_rate": 4.2795273310419235e-06,
+      "loss": 1.1335,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.168980686038118,
+      "learning_rate": 4.277260637271345e-06,
+      "loss": 1.0784,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.613135804393563,
+      "learning_rate": 4.27499438063216e-06,
+      "loss": 1.0107,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5657568046607926,
+      "learning_rate": 4.272728561297461e-06,
+      "loss": 1.1289,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3844141698572363,
+      "learning_rate": 4.270463179440334e-06,
+      "loss": 1.0381,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.043877675908545,
+      "learning_rate": 4.2681982352338155e-06,
+      "loss": 1.1331,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.014632031414868,
+      "learning_rate": 4.26593372885091e-06,
+      "loss": 1.0503,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7888945573034216,
+      "learning_rate": 4.263669660464601e-06,
+      "loss": 1.0781,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.951336216248136,
+      "learning_rate": 4.261406030247822e-06,
+      "loss": 0.9971,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.316587918296565,
+      "learning_rate": 4.259142838373484e-06,
+      "loss": 1.1189,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1437923045919067,
+      "learning_rate": 4.25688008501446e-06,
+      "loss": 1.019,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.1120944521984635,
+      "learning_rate": 4.254617770343584e-06,
+      "loss": 1.1206,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2705316527048174,
+      "learning_rate": 4.252355894533674e-06,
+      "loss": 1.0642,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.380401871681746,
+      "learning_rate": 4.250094457757497e-06,
+      "loss": 1.0381,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.29078947543885,
+      "learning_rate": 4.2478334601877915e-06,
+      "loss": 1.1128,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7147220213789645,
+      "learning_rate": 4.2455729019972715e-06,
+      "loss": 1.1023,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.162894435725744,
+      "learning_rate": 4.243312783358598e-06,
+      "loss": 1.0757,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3164578603130077,
+      "learning_rate": 4.2410531044444205e-06,
+      "loss": 1.0747,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6685591396330106,
+      "learning_rate": 4.238793865427337e-06,
+      "loss": 1.093,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4803384995899402,
+      "learning_rate": 4.2365350664799254e-06,
+      "loss": 1.1921,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9930057493718119,
+      "learning_rate": 4.234276707774722e-06,
+      "loss": 0.9756,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3267338480164916,
+      "learning_rate": 4.232018789484228e-06,
+      "loss": 1.04,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2584368864398927,
+      "learning_rate": 4.229761311780924e-06,
+      "loss": 1.1145,
+      "step": 8261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.794979139357156,
+      "learning_rate": 4.227504274837235e-06,
+      "loss": 1.0696,
+      "step": 8262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.965163989168291,
+      "learning_rate": 4.2252476788255735e-06,
+      "loss": 1.1758,
+      "step": 8263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2078025088814273,
+      "learning_rate": 4.222991523918307e-06,
+      "loss": 1.0706,
+      "step": 8264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3593153691821254,
+      "learning_rate": 4.220735810287769e-06,
+      "loss": 1.1235,
+      "step": 8265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9058459578521125,
+      "learning_rate": 4.218480538106267e-06,
+      "loss": 1.0669,
+      "step": 8266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.199953365172379,
+      "learning_rate": 4.2162257075460675e-06,
+      "loss": 1.0376,
+      "step": 8267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.574614932298014,
+      "learning_rate": 4.213971318779408e-06,
+      "loss": 1.1365,
+      "step": 8268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.807831407446443,
+      "learning_rate": 4.211717371978485e-06,
+      "loss": 1.1213,
+      "step": 8269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.381694055159166,
+      "learning_rate": 4.209463867315468e-06,
+      "loss": 1.1575,
+      "step": 8270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8196246635538005,
+      "learning_rate": 4.207210804962494e-06,
+      "loss": 1.1199,
+      "step": 8271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.739574955959469,
+      "learning_rate": 4.204958185091662e-06,
+      "loss": 1.0178,
+      "step": 8272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0130224027158374,
+      "learning_rate": 4.202706007875034e-06,
+      "loss": 1.0715,
+      "step": 8273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6848929793427163,
+      "learning_rate": 4.200454273484653e-06,
+      "loss": 1.0605,
+      "step": 8274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.231783790001101,
+      "learning_rate": 4.198202982092504e-06,
+      "loss": 1.0991,
+      "step": 8275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.127015151785281,
+      "learning_rate": 4.195952133870561e-06,
+      "loss": 1.1343,
+      "step": 8276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.398423220615423,
+      "learning_rate": 4.193701728990755e-06,
+      "loss": 1.0881,
+      "step": 8277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0353038942482007,
+      "learning_rate": 4.1914517676249744e-06,
+      "loss": 1.1155,
+      "step": 8278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1526965252045005,
+      "learning_rate": 4.1892022499450944e-06,
+      "loss": 1.1389,
+      "step": 8279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2615759381421014,
+      "learning_rate": 4.186953176122939e-06,
+      "loss": 1.1426,
+      "step": 8280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.987150585702287,
+      "learning_rate": 4.184704546330302e-06,
+      "loss": 1.063,
+      "step": 8281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5291354052357478,
+      "learning_rate": 4.182456360738948e-06,
+      "loss": 1.0535,
+      "step": 8282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8480173696830757,
+      "learning_rate": 4.180208619520599e-06,
+      "loss": 1.0857,
+      "step": 8283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.547477899368128,
+      "learning_rate": 4.1779613228469564e-06,
+      "loss": 1.1665,
+      "step": 8284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2447644496015307,
+      "learning_rate": 4.175714470889677e-06,
+      "loss": 1.0881,
+      "step": 8285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.81865951801306,
+      "learning_rate": 4.173468063820382e-06,
+      "loss": 1.0862,
+      "step": 8286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.251281698894674,
+      "learning_rate": 4.17122210181067e-06,
+      "loss": 1.1038,
+      "step": 8287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.792812482423666,
+      "learning_rate": 4.168976585032096e-06,
+      "loss": 1.1067,
+      "step": 8288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9816446189286387,
+      "learning_rate": 4.166731513656185e-06,
+      "loss": 1.0352,
+      "step": 8289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2716336292409807,
+      "learning_rate": 4.164486887854424e-06,
+      "loss": 1.1013,
+      "step": 8290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.468223868696146,
+      "learning_rate": 4.162242707798267e-06,
+      "loss": 1.1135,
+      "step": 8291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2942323664227433,
+      "learning_rate": 4.159998973659143e-06,
+      "loss": 1.0503,
+      "step": 8292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.05034177245571,
+      "learning_rate": 4.157755685608431e-06,
+      "loss": 1.1062,
+      "step": 8293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7915215448319692,
+      "learning_rate": 4.1555128438174964e-06,
+      "loss": 1.0991,
+      "step": 8294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9639601532089626,
+      "learning_rate": 4.153270448457644e-06,
+      "loss": 1.0701,
+      "step": 8295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.149942738790385,
+      "learning_rate": 4.15102849970017e-06,
+      "loss": 1.113,
+      "step": 8296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.499021212445275,
+      "learning_rate": 4.148786997716322e-06,
+      "loss": 0.9714,
+      "step": 8297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0218965641315463,
+      "learning_rate": 4.146545942677314e-06,
+      "loss": 0.9952,
+      "step": 8298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8950537507561434,
+      "learning_rate": 4.144305334754335e-06,
+      "loss": 1.093,
+      "step": 8299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2667775107441224,
+      "learning_rate": 4.142065174118531e-06,
+      "loss": 1.0728,
+      "step": 8300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6403509727811216,
+      "learning_rate": 4.139825460941017e-06,
+      "loss": 1.0613,
+      "step": 8301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2508520011237687,
+      "learning_rate": 4.137586195392872e-06,
+      "loss": 1.0789,
+      "step": 8302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.355304499171004,
+      "learning_rate": 4.13534737764514e-06,
+      "loss": 1.1421,
+      "step": 8303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.792707101261526,
+      "learning_rate": 4.133109007868841e-06,
+      "loss": 1.0857,
+      "step": 8304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3669283302266737,
+      "learning_rate": 4.1308710862349475e-06,
+      "loss": 1.0312,
+      "step": 8305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3772276781365056,
+      "learning_rate": 4.1286336129144e-06,
+      "loss": 1.0784,
+      "step": 8306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1763872598326883,
+      "learning_rate": 4.126396588078118e-06,
+      "loss": 1.064,
+      "step": 8307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6620852764225558,
+      "learning_rate": 4.124160011896965e-06,
+      "loss": 1.022,
+      "step": 8308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.234726217670418,
+      "learning_rate": 4.1219238845417895e-06,
+      "loss": 1.0481,
+      "step": 8309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2585330606943357,
+      "learning_rate": 4.1196882061833965e-06,
+      "loss": 1.1292,
+      "step": 8310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.459024346809227,
+      "learning_rate": 4.117452976992553e-06,
+      "loss": 1.187,
+      "step": 8311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8513170473069764,
+      "learning_rate": 4.115218197140005e-06,
+      "loss": 1.0627,
+      "step": 8312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9201189322304066,
+      "learning_rate": 4.112983866796454e-06,
+      "loss": 1.0334,
+      "step": 8313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.666141071800054,
+      "learning_rate": 4.110749986132566e-06,
+      "loss": 1.1187,
+      "step": 8314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.497096428524633,
+      "learning_rate": 4.108516555318979e-06,
+      "loss": 1.0786,
+      "step": 8315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1923441384882656,
+      "learning_rate": 4.1062835745262886e-06,
+      "loss": 1.1365,
+      "step": 8316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4718823403283867,
+      "learning_rate": 4.104051043925068e-06,
+      "loss": 1.0801,
+      "step": 8317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0789580916660046,
+      "learning_rate": 4.1018189636858465e-06,
+      "loss": 1.0752,
+      "step": 8318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3597468376216506,
+      "learning_rate": 4.099587333979117e-06,
+      "loss": 1.1125,
+      "step": 8319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6829111019383456,
+      "learning_rate": 4.097356154975349e-06,
+      "loss": 1.0515,
+      "step": 8320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3147292961994395,
+      "learning_rate": 4.09512542684497e-06,
+      "loss": 1.0923,
+      "step": 8321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.340047696645989,
+      "learning_rate": 4.092895149758373e-06,
+      "loss": 1.0696,
+      "step": 8322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5359101161479036,
+      "learning_rate": 4.090665323885915e-06,
+      "loss": 1.0854,
+      "step": 8323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.674643232061462,
+      "learning_rate": 4.088435949397922e-06,
+      "loss": 1.0579,
+      "step": 8324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.221229332398051,
+      "learning_rate": 4.086207026464689e-06,
+      "loss": 1.0891,
+      "step": 8325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0372782542968717,
+      "learning_rate": 4.083978555256466e-06,
+      "loss": 1.0581,
+      "step": 8326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.065168342861082,
+      "learning_rate": 4.081750535943485e-06,
+      "loss": 1.0605,
+      "step": 8327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.774951565100493,
+      "learning_rate": 4.079522968695924e-06,
+      "loss": 1.074,
+      "step": 8328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.770631161129205,
+      "learning_rate": 4.077295853683933e-06,
+      "loss": 1.1294,
+      "step": 8329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.044807392167953,
+      "learning_rate": 4.075069191077641e-06,
+      "loss": 1.1357,
+      "step": 8330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1131395088725795,
+      "learning_rate": 4.0728429810471214e-06,
+      "loss": 1.085,
+      "step": 8331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2687169940946714,
+      "learning_rate": 4.070617223762431e-06,
+      "loss": 1.1108,
+      "step": 8332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8487945386659543,
+      "learning_rate": 4.068391919393581e-06,
+      "loss": 1.0649,
+      "step": 8333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.041385116131365,
+      "learning_rate": 4.066167068110553e-06,
+      "loss": 1.0457,
+      "step": 8334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7359604844076717,
+      "learning_rate": 4.063942670083288e-06,
+      "loss": 1.0967,
+      "step": 8335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.745906040264092,
+      "learning_rate": 4.061718725481698e-06,
+      "loss": 1.0701,
+      "step": 8336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8461657867459778,
+      "learning_rate": 4.0594952344756625e-06,
+      "loss": 1.1448,
+      "step": 8337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0653971919159964,
+      "learning_rate": 4.05727219723502e-06,
+      "loss": 1.0437,
+      "step": 8338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2909888153229323,
+      "learning_rate": 4.055049613929575e-06,
+      "loss": 1.0115,
+      "step": 8339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2104737148148126,
+      "learning_rate": 4.052827484729106e-06,
+      "loss": 1.1218,
+      "step": 8340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.81334568927403,
+      "learning_rate": 4.0506058098033456e-06,
+      "loss": 1.0969,
+      "step": 8341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.861439753389167,
+      "learning_rate": 4.048384589321998e-06,
+      "loss": 1.084,
+      "step": 8342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.206119483466687,
+      "learning_rate": 4.04616382345473e-06,
+      "loss": 1.1045,
+      "step": 8343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.915214311874115,
+      "learning_rate": 4.043943512371171e-06,
+      "loss": 1.0876,
+      "step": 8344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3473310206186633,
+      "learning_rate": 4.0417236562409275e-06,
+      "loss": 1.0637,
+      "step": 8345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8838651009162044,
+      "learning_rate": 4.039504255233559e-06,
+      "loss": 1.0798,
+      "step": 8346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6237717286393076,
+      "learning_rate": 4.037285309518596e-06,
+      "loss": 1.0369,
+      "step": 8347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.404989646298062,
+      "learning_rate": 4.035066819265531e-06,
+      "loss": 1.1035,
+      "step": 8348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.340834141613759,
+      "learning_rate": 4.032848784643819e-06,
+      "loss": 1.0786,
+      "step": 8349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.101434271488298,
+      "learning_rate": 4.030631205822894e-06,
+      "loss": 1.085,
+      "step": 8350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8759766748684146,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.9944,
+      "step": 8351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0421538985822045,
+      "learning_rate": 4.026197416260911e-06,
+      "loss": 1.0066,
+      "step": 8352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.521238885093864,
+      "learning_rate": 4.02398120585853e-06,
+      "loss": 1.073,
+      "step": 8353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6237030617525137,
+      "learning_rate": 4.021765451934283e-06,
+      "loss": 1.0945,
+      "step": 8354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.339443970059831,
+      "learning_rate": 4.0195501546574185e-06,
+      "loss": 1.1345,
+      "step": 8355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.327244664982045,
+      "learning_rate": 4.017335314197152e-06,
+      "loss": 1.0664,
+      "step": 8356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.385288132847073,
+      "learning_rate": 4.01512093072266e-06,
+      "loss": 1.0938,
+      "step": 8357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.92712766779232,
+      "learning_rate": 4.012907004403096e-06,
+      "loss": 1.074,
+      "step": 8358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.325508354732916,
+      "learning_rate": 4.010693535407564e-06,
+      "loss": 1.0127,
+      "step": 8359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3900907540285563,
+      "learning_rate": 4.008480523905148e-06,
+      "loss": 1.0725,
+      "step": 8360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3050665976148963,
+      "learning_rate": 4.0062679700648824e-06,
+      "loss": 1.087,
+      "step": 8361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.223462063750219,
+      "learning_rate": 4.004055874055769e-06,
+      "loss": 1.1113,
+      "step": 8362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.202192948536041,
+      "learning_rate": 4.001844236046788e-06,
+      "loss": 1.1133,
+      "step": 8363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.842712424621746,
+      "learning_rate": 3.999633056206868e-06,
+      "loss": 1.0632,
+      "step": 8364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3933210034493717,
+      "learning_rate": 3.997422334704917e-06,
+      "loss": 1.1001,
+      "step": 8365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.962533858602223,
+      "learning_rate": 3.995212071709796e-06,
+      "loss": 1.0459,
+      "step": 8366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.078199337289186,
+      "learning_rate": 3.993002267390339e-06,
+      "loss": 1.125,
+      "step": 8367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7697512185691853,
+      "learning_rate": 3.990792921915339e-06,
+      "loss": 1.1216,
+      "step": 8368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2650498025939663,
+      "learning_rate": 3.988584035453556e-06,
+      "loss": 1.1113,
+      "step": 8369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.023744123627186,
+      "learning_rate": 3.986375608173719e-06,
+      "loss": 1.0723,
+      "step": 8370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1571867620541756,
+      "learning_rate": 3.984167640244518e-06,
+      "loss": 1.071,
+      "step": 8371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.213727991865065,
+      "learning_rate": 3.981960131834606e-06,
+      "loss": 1.0125,
+      "step": 8372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.783399332057206,
+      "learning_rate": 3.979753083112608e-06,
+      "loss": 1.0486,
+      "step": 8373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.387019390258054,
+      "learning_rate": 3.977546494247106e-06,
+      "loss": 1.0454,
+      "step": 8374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8796837056492537,
+      "learning_rate": 3.975340365406652e-06,
+      "loss": 0.9856,
+      "step": 8375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3705036408230065,
+      "learning_rate": 3.9731346967597595e-06,
+      "loss": 1.1167,
+      "step": 8376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6808697869976297,
+      "learning_rate": 3.970929488474906e-06,
+      "loss": 1.1023,
+      "step": 8377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.845059639864677,
+      "learning_rate": 3.968724740720543e-06,
+      "loss": 1.0916,
+      "step": 8378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1224696131026346,
+      "learning_rate": 3.966520453665078e-06,
+      "loss": 1.0437,
+      "step": 8379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6146841976843165,
+      "learning_rate": 3.964316627476883e-06,
+      "loss": 1.0544,
+      "step": 8380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2598887110174495,
+      "learning_rate": 3.962113262324298e-06,
+      "loss": 1.0654,
+      "step": 8381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.917002544442159,
+      "learning_rate": 3.959910358375625e-06,
+      "loss": 1.113,
+      "step": 8382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3861304438201456,
+      "learning_rate": 3.957707915799137e-06,
+      "loss": 1.1289,
+      "step": 8383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.239728992463018,
+      "learning_rate": 3.955505934763068e-06,
+      "loss": 1.1128,
+      "step": 8384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.169256454263801,
+      "learning_rate": 3.953304415435608e-06,
+      "loss": 1.0796,
+      "step": 8385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.959617888740946,
+      "learning_rate": 3.951103357984931e-06,
+      "loss": 1.0376,
+      "step": 8386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0368732099921325,
+      "learning_rate": 3.94890276257916e-06,
+      "loss": 1.0168,
+      "step": 8387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9914235983789106,
+      "learning_rate": 3.946702629386388e-06,
+      "loss": 1.1211,
+      "step": 8388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3746267362463294,
+      "learning_rate": 3.9445029585746695e-06,
+      "loss": 1.0898,
+      "step": 8389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.224664347054124,
+      "learning_rate": 3.9423037503120255e-06,
+      "loss": 1.032,
+      "step": 8390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1493709277370643,
+      "learning_rate": 3.940105004766449e-06,
+      "loss": 1.0564,
+      "step": 8391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.746781392226653,
+      "learning_rate": 3.937906722105887e-06,
+      "loss": 1.0498,
+      "step": 8392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8851627959899697,
+      "learning_rate": 3.935708902498253e-06,
+      "loss": 1.0334,
+      "step": 8393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.092551233339254,
+      "learning_rate": 3.933511546111437e-06,
+      "loss": 1.0471,
+      "step": 8394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4499125827744592,
+      "learning_rate": 3.93131465311327e-06,
+      "loss": 1.0579,
+      "step": 8395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5270700741287304,
+      "learning_rate": 3.929118223671572e-06,
+      "loss": 1.0381,
+      "step": 8396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.327818527768553,
+      "learning_rate": 3.926922257954111e-06,
+      "loss": 1.0408,
+      "step": 8397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.212981574610744,
+      "learning_rate": 3.924726756128632e-06,
+      "loss": 1.085,
+      "step": 8398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.13623540615992,
+      "learning_rate": 3.922531718362835e-06,
+      "loss": 1.0227,
+      "step": 8399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9284819727194327,
+      "learning_rate": 3.920337144824389e-06,
+      "loss": 1.0737,
+      "step": 8400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.655006525194112,
+      "learning_rate": 3.918143035680925e-06,
+      "loss": 1.0571,
+      "step": 8401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0693655758281504,
+      "learning_rate": 3.915949391100038e-06,
+      "loss": 1.0842,
+      "step": 8402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3650330154914374,
+      "learning_rate": 3.913756211249295e-06,
+      "loss": 1.03,
+      "step": 8403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.262632638977415,
+      "learning_rate": 3.911563496296219e-06,
+      "loss": 1.114,
+      "step": 8404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7496441704591787,
+      "learning_rate": 3.909371246408298e-06,
+      "loss": 1.1047,
+      "step": 8405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1555928127978485,
+      "learning_rate": 3.907179461752992e-06,
+      "loss": 1.074,
+      "step": 8406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.890650847743628,
+      "learning_rate": 3.90498814249772e-06,
+      "loss": 1.0437,
+      "step": 8407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.1752703543193315,
+      "learning_rate": 3.902797288809863e-06,
+      "loss": 1.0552,
+      "step": 8408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.725996859404378,
+      "learning_rate": 3.90060690085677e-06,
+      "loss": 1.0984,
+      "step": 8409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4771932086077495,
+      "learning_rate": 3.898416978805751e-06,
+      "loss": 1.0764,
+      "step": 8410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5205200544295243,
+      "learning_rate": 3.896227522824091e-06,
+      "loss": 1.0342,
+      "step": 8411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.758854873303152,
+      "learning_rate": 3.894038533079025e-06,
+      "loss": 0.988,
+      "step": 8412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.2048547131049,
+      "learning_rate": 3.891850009737759e-06,
+      "loss": 1.1003,
+      "step": 8413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.295854185866941,
+      "learning_rate": 3.88966195296747e-06,
+      "loss": 1.0237,
+      "step": 8414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.87279423768917,
+      "learning_rate": 3.887474362935284e-06,
+      "loss": 0.9971,
+      "step": 8415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7044406003920507,
+      "learning_rate": 3.885287239808306e-06,
+      "loss": 1.0701,
+      "step": 8416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.091767098687201,
+      "learning_rate": 3.883100583753596e-06,
+      "loss": 1.0925,
+      "step": 8417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.819417721329449,
+      "learning_rate": 3.880914394938181e-06,
+      "loss": 1.1389,
+      "step": 8418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0197867680792587,
+      "learning_rate": 3.878728673529058e-06,
+      "loss": 1.0662,
+      "step": 8419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0622873590536117,
+      "learning_rate": 3.8765434196931806e-06,
+      "loss": 1.085,
+      "step": 8420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.313228150049387,
+      "learning_rate": 3.874358633597469e-06,
+      "loss": 1.0979,
+      "step": 8421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1558270575281275,
+      "learning_rate": 3.872174315408808e-06,
+      "loss": 1.0481,
+      "step": 8422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1704768866057234,
+      "learning_rate": 3.869990465294043e-06,
+      "loss": 1.1006,
+      "step": 8423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7100999115908473,
+      "learning_rate": 3.867807083419995e-06,
+      "loss": 1.1194,
+      "step": 8424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2330794776198784,
+      "learning_rate": 3.8656241699534396e-06,
+      "loss": 1.156,
+      "step": 8425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.162879732312702,
+      "learning_rate": 3.8634417250611115e-06,
+      "loss": 1.1646,
+      "step": 8426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9746035874580032,
+      "learning_rate": 3.861259748909729e-06,
+      "loss": 0.9756,
+      "step": 8427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.672055349904854,
+      "learning_rate": 3.8590782416659485e-06,
+      "loss": 1.0762,
+      "step": 8428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.079331286602348,
+      "learning_rate": 3.856897203496416e-06,
+      "loss": 1.0757,
+      "step": 8429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.111335745099905,
+      "learning_rate": 3.854716634567724e-06,
+      "loss": 1.0945,
+      "step": 8430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.018464904665178,
+      "learning_rate": 3.852536535046435e-06,
+      "loss": 1.0835,
+      "step": 8431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.48040988225769,
+      "learning_rate": 3.850356905099079e-06,
+      "loss": 1.1531,
+      "step": 8432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.126720016357566,
+      "learning_rate": 3.848177744892146e-06,
+      "loss": 1.0537,
+      "step": 8433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.557715840638076,
+      "learning_rate": 3.845999054592092e-06,
+      "loss": 1.0161,
+      "step": 8434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.034988724566439,
+      "learning_rate": 3.843820834365334e-06,
+      "loss": 1.0344,
+      "step": 8435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.499126461402841,
+      "learning_rate": 3.8416430843782525e-06,
+      "loss": 1.0237,
+      "step": 8436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5491226145745762,
+      "learning_rate": 3.839465804797203e-06,
+      "loss": 1.0471,
+      "step": 8437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.972102417762742,
+      "learning_rate": 3.837288995788488e-06,
+      "loss": 1.0906,
+      "step": 8438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.591666269757138,
+      "learning_rate": 3.835112657518393e-06,
+      "loss": 1.1416,
+      "step": 8439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.129130168710327,
+      "learning_rate": 3.83293679015315e-06,
+      "loss": 1.0696,
+      "step": 8440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.897634771652616,
+      "learning_rate": 3.830761393858966e-06,
+      "loss": 1.1516,
+      "step": 8441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6092729111524524,
+      "learning_rate": 3.828586468802008e-06,
+      "loss": 1.1191,
+      "step": 8442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.208696665319671,
+      "learning_rate": 3.826412015148403e-06,
+      "loss": 1.0239,
+      "step": 8443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.351619705104364,
+      "learning_rate": 3.824238033064253e-06,
+      "loss": 1.0366,
+      "step": 8444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.153765028221366,
+      "learning_rate": 3.8220645227156175e-06,
+      "loss": 1.0745,
+      "step": 8445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8419075978722197,
+      "learning_rate": 3.819891484268513e-06,
+      "loss": 1.0125,
+      "step": 8446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.736536219739852,
+      "learning_rate": 3.81771891788894e-06,
+      "loss": 1.1094,
+      "step": 8447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3674160178367014,
+      "learning_rate": 3.815546823742834e-06,
+      "loss": 1.0635,
+      "step": 8448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0608381512826512,
+      "learning_rate": 3.8133752019961213e-06,
+      "loss": 1.0828,
+      "step": 8449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.074468955572667,
+      "learning_rate": 3.8112040528146787e-06,
+      "loss": 1.1575,
+      "step": 8450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.212334364925423,
+      "learning_rate": 3.809033376364346e-06,
+      "loss": 1.0298,
+      "step": 8451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7676535658693324,
+      "learning_rate": 3.8068631728109364e-06,
+      "loss": 1.0815,
+      "step": 8452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0634740267649776,
+      "learning_rate": 3.8046934423202165e-06,
+      "loss": 1.0842,
+      "step": 8453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3427494544412317,
+      "learning_rate": 3.802524185057923e-06,
+      "loss": 1.0718,
+      "step": 8454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9419583260674864,
+      "learning_rate": 3.8003554011897536e-06,
+      "loss": 1.1282,
+      "step": 8455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0631197558797454,
+      "learning_rate": 3.7981870908813677e-06,
+      "loss": 1.1167,
+      "step": 8456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.657444805310502,
+      "learning_rate": 3.7960192542984e-06,
+      "loss": 1.0771,
+      "step": 8457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.912267956694907,
+      "learning_rate": 3.7938518916064335e-06,
+      "loss": 1.0024,
+      "step": 8458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7891961692307365,
+      "learning_rate": 3.7916850029710217e-06,
+      "loss": 1.1003,
+      "step": 8459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1336234332162185,
+      "learning_rate": 3.7895185885576924e-06,
+      "loss": 1.0354,
+      "step": 8460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2973331016199934,
+      "learning_rate": 3.7873526485319135e-06,
+      "loss": 1.0618,
+      "step": 8461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3760432829041847,
+      "learning_rate": 3.785187183059139e-06,
+      "loss": 1.0959,
+      "step": 8462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.78197835304298,
+      "learning_rate": 3.783022192304776e-06,
+      "loss": 1.0725,
+      "step": 8463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9099435972597965,
+      "learning_rate": 3.780857676434193e-06,
+      "loss": 1.1204,
+      "step": 8464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9556744097442387,
+      "learning_rate": 3.7786936356127345e-06,
+      "loss": 1.0181,
+      "step": 8465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.286841353966018,
+      "learning_rate": 3.7765300700056927e-06,
+      "loss": 1.0574,
+      "step": 8466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.616473675064539,
+      "learning_rate": 3.7743669797783424e-06,
+      "loss": 1.0964,
+      "step": 8467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.360014814131588,
+      "learning_rate": 3.7722043650959008e-06,
+      "loss": 1.0791,
+      "step": 8468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.66416338118822,
+      "learning_rate": 3.7700422261235592e-06,
+      "loss": 1.1504,
+      "step": 8469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.279894060771963,
+      "learning_rate": 3.7678805630264802e-06,
+      "loss": 1.0469,
+      "step": 8470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.670342393873556,
+      "learning_rate": 3.7657193759697743e-06,
+      "loss": 1.0913,
+      "step": 8471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2703445583334845,
+      "learning_rate": 3.7635586651185305e-06,
+      "loss": 1.114,
+      "step": 8472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6555211165206005,
+      "learning_rate": 3.761398430637796e-06,
+      "loss": 1.019,
+      "step": 8473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.63772760473509,
+      "learning_rate": 3.75923867269257e-06,
+      "loss": 1.0647,
+      "step": 8474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.087464873715203,
+      "learning_rate": 3.7570793914478342e-06,
+      "loss": 1.0122,
+      "step": 8475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.948117802006053,
+      "learning_rate": 3.75492058706852e-06,
+      "loss": 1.0681,
+      "step": 8476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.467839131502466,
+      "learning_rate": 3.7527622597195336e-06,
+      "loss": 1.0327,
+      "step": 8477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.707457404995845,
+      "learning_rate": 3.7506044095657358e-06,
+      "loss": 1.1145,
+      "step": 8478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.332290302818705,
+      "learning_rate": 3.7484470367719493e-06,
+      "loss": 1.083,
+      "step": 8479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4573101166949725,
+      "learning_rate": 3.7462901415029785e-06,
+      "loss": 1.0398,
+      "step": 8480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1273297132846034,
+      "learning_rate": 3.744133723923561e-06,
+      "loss": 1.0991,
+      "step": 8481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.808450476841252,
+      "learning_rate": 3.741977784198426e-06,
+      "loss": 1.0049,
+      "step": 8482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9562596419239675,
+      "learning_rate": 3.7398223224922515e-06,
+      "loss": 1.0042,
+      "step": 8483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1225462054289475,
+      "learning_rate": 3.737667338969679e-06,
+      "loss": 1.0417,
+      "step": 8484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4408759257602326,
+      "learning_rate": 3.735512833795324e-06,
+      "loss": 1.095,
+      "step": 8485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2593914534920643,
+      "learning_rate": 3.7333588071337544e-06,
+      "loss": 1.0833,
+      "step": 8486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.351937653007181,
+      "learning_rate": 3.7312052591495065e-06,
+      "loss": 1.0862,
+      "step": 8487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3984574873654383,
+      "learning_rate": 3.7290521900070774e-06,
+      "loss": 1.1072,
+      "step": 8488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.014338480502073,
+      "learning_rate": 3.726899599870928e-06,
+      "loss": 1.095,
+      "step": 8489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.388013921047858,
+      "learning_rate": 3.7247474889054892e-06,
+      "loss": 1.0281,
+      "step": 8490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4848340642451796,
+      "learning_rate": 3.7225958572751462e-06,
+      "loss": 1.1167,
+      "step": 8491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.137196877014114,
+      "learning_rate": 3.7204447051442483e-06,
+      "loss": 1.0862,
+      "step": 8492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.297546114316013,
+      "learning_rate": 3.7182940326771232e-06,
+      "loss": 1.0771,
+      "step": 8493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2978353564214657,
+      "learning_rate": 3.7161438400380336e-06,
+      "loss": 1.1287,
+      "step": 8494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5748608662498436,
+      "learning_rate": 3.713994127391234e-06,
+      "loss": 1.1504,
+      "step": 8495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0547680199357305,
+      "learning_rate": 3.711844894900927e-06,
+      "loss": 1.0344,
+      "step": 8496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2856051148845005,
+      "learning_rate": 3.709696142731277e-06,
+      "loss": 1.1409,
+      "step": 8497
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9334796698887136,
+      "learning_rate": 3.707547871046423e-06,
+      "loss": 1.1064,
+      "step": 8498
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.45360347873694,
+      "learning_rate": 3.705400080010456e-06,
+      "loss": 1.1313,
+      "step": 8499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.031060296060287,
+      "learning_rate": 3.703252769787443e-06,
+      "loss": 1.0513,
+      "step": 8500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7790452196923963,
+      "learning_rate": 3.701105940541397e-06,
+      "loss": 1.0708,
+      "step": 8501
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.969201820784218,
+      "learning_rate": 3.698959592436304e-06,
+      "loss": 1.1162,
+      "step": 8502
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.155070889508824,
+      "learning_rate": 3.6968137256361192e-06,
+      "loss": 1.1177,
+      "step": 8503
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.854438536942737,
+      "learning_rate": 3.694668340304748e-06,
+      "loss": 1.0686,
+      "step": 8504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.081691792901406,
+      "learning_rate": 3.6925234366060712e-06,
+      "loss": 1.0938,
+      "step": 8505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7210365613623804,
+      "learning_rate": 3.6903790147039286e-06,
+      "loss": 1.0278,
+      "step": 8506
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.72851281227371,
+      "learning_rate": 3.6882350747621108e-06,
+      "loss": 1.0635,
+      "step": 8507
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.567268297820344,
+      "learning_rate": 3.6860916169443937e-06,
+      "loss": 1.1023,
+      "step": 8508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4630789491756016,
+      "learning_rate": 3.6839486414144974e-06,
+      "loss": 1.1013,
+      "step": 8509
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7184301473450843,
+      "learning_rate": 3.6818061483361212e-06,
+      "loss": 1.1003,
+      "step": 8510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2596607018516073,
+      "learning_rate": 3.679664137872915e-06,
+      "loss": 1.051,
+      "step": 8511
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.582935372418553,
+      "learning_rate": 3.6775226101884932e-06,
+      "loss": 1.1106,
+      "step": 8512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4988554163065717,
+      "learning_rate": 3.6753815654464477e-06,
+      "loss": 1.0361,
+      "step": 8513
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.258933315765297,
+      "learning_rate": 3.673241003810306e-06,
+      "loss": 1.0864,
+      "step": 8514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.21949700574753,
+      "learning_rate": 3.671100925443586e-06,
+      "loss": 1.167,
+      "step": 8515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3118721830507827,
+      "learning_rate": 3.668961330509756e-06,
+      "loss": 1.0305,
+      "step": 8516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.682405020389182,
+      "learning_rate": 3.6668222191722437e-06,
+      "loss": 1.0784,
+      "step": 8517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5254412220787343,
+      "learning_rate": 3.6646835915944525e-06,
+      "loss": 1.0898,
+      "step": 8518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8528032363364493,
+      "learning_rate": 3.662545447939737e-06,
+      "loss": 1.0876,
+      "step": 8519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.978728699140063,
+      "learning_rate": 3.660407788371422e-06,
+      "loss": 1.0806,
+      "step": 8520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.024008844025409,
+      "learning_rate": 3.658270613052789e-06,
+      "loss": 1.0862,
+      "step": 8521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.437063617533948,
+      "learning_rate": 3.656133922147086e-06,
+      "loss": 1.0818,
+      "step": 8522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.128261686323989,
+      "learning_rate": 3.6539977158175286e-06,
+      "loss": 1.0959,
+      "step": 8523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.404701148856388,
+      "learning_rate": 3.6518619942272883e-06,
+      "loss": 1.0642,
+      "step": 8524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.112793007132475,
+      "learning_rate": 3.649726757539499e-06,
+      "loss": 1.1301,
+      "step": 8525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3619640750768514,
+      "learning_rate": 3.6475920059172708e-06,
+      "loss": 1.1602,
+      "step": 8526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3570225642118334,
+      "learning_rate": 3.6454577395236533e-06,
+      "loss": 1.063,
+      "step": 8527
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.521676104335751,
+      "learning_rate": 3.643323958521682e-06,
+      "loss": 1.0857,
+      "step": 8528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.117119648784731,
+      "learning_rate": 3.641190663074342e-06,
+      "loss": 1.0178,
+      "step": 8529
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9611956221541114,
+      "learning_rate": 3.639057853344583e-06,
+      "loss": 1.041,
+      "step": 8530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.167830104798484,
+      "learning_rate": 3.6369255294953267e-06,
+      "loss": 1.053,
+      "step": 8531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9550995708545877,
+      "learning_rate": 3.6347936916894454e-06,
+      "loss": 1.0758,
+      "step": 8532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7872210296069584,
+      "learning_rate": 3.6326623400897797e-06,
+      "loss": 1.0417,
+      "step": 8533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.442617678915525,
+      "learning_rate": 3.6305314748591348e-06,
+      "loss": 1.0562,
+      "step": 8534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.363501246110174,
+      "learning_rate": 3.628401096160272e-06,
+      "loss": 1.1213,
+      "step": 8535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9111841215085152,
+      "learning_rate": 3.6262712041559265e-06,
+      "loss": 1.1086,
+      "step": 8536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.388529665158878,
+      "learning_rate": 3.624141799008788e-06,
+      "loss": 1.0967,
+      "step": 8537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7682652737956643,
+      "learning_rate": 3.622012880881508e-06,
+      "loss": 1.0586,
+      "step": 8538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1250278945512346,
+      "learning_rate": 3.619884449936708e-06,
+      "loss": 1.1157,
+      "step": 8539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5079155873733714,
+      "learning_rate": 3.617756506336968e-06,
+      "loss": 1.1653,
+      "step": 8540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.01450759843174,
+      "learning_rate": 3.615629050244829e-06,
+      "loss": 1.1033,
+      "step": 8541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8547570722523,
+      "learning_rate": 3.613502081822794e-06,
+      "loss": 1.1365,
+      "step": 8542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1812529472609565,
+      "learning_rate": 3.611375601233338e-06,
+      "loss": 1.0422,
+      "step": 8543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.229129368876844,
+      "learning_rate": 3.6092496086388882e-06,
+      "loss": 1.0698,
+      "step": 8544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6725838043015444,
+      "learning_rate": 3.607124104201837e-06,
+      "loss": 1.1335,
+      "step": 8545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7835648861726754,
+      "learning_rate": 3.60499908808455e-06,
+      "loss": 1.0867,
+      "step": 8546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0846400388369286,
+      "learning_rate": 3.6028745604493322e-06,
+      "loss": 1.0559,
+      "step": 8547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.910779941613858,
+      "learning_rate": 3.6007505214584772e-06,
+      "loss": 1.0813,
+      "step": 8548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6518238645569254,
+      "learning_rate": 3.598626971274226e-06,
+      "loss": 1.0774,
+      "step": 8549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3523689201010587,
+      "learning_rate": 3.5965039100587818e-06,
+      "loss": 1.0706,
+      "step": 8550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8641569100539623,
+      "learning_rate": 3.594381337974322e-06,
+      "loss": 1.1042,
+      "step": 8551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2467548169763467,
+      "learning_rate": 3.592259255182975e-06,
+      "loss": 1.113,
+      "step": 8552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1544774262155717,
+      "learning_rate": 3.5901376618468376e-06,
+      "loss": 1.1279,
+      "step": 8553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3102437232197635,
+      "learning_rate": 3.5880165581279667e-06,
+      "loss": 1.0586,
+      "step": 8554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.560585765396052,
+      "learning_rate": 3.58589594418838e-06,
+      "loss": 1.0833,
+      "step": 8555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1234620189944775,
+      "learning_rate": 3.5837758201900664e-06,
+      "loss": 1.0964,
+      "step": 8556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3535849871485626,
+      "learning_rate": 3.5816561862949695e-06,
+      "loss": 1.0608,
+      "step": 8557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6686741760716797,
+      "learning_rate": 3.579537042664992e-06,
+      "loss": 1.1621,
+      "step": 8558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.020179630883658,
+      "learning_rate": 3.5774183894620175e-06,
+      "loss": 1.1091,
+      "step": 8559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6403367537013405,
+      "learning_rate": 3.575300226847863e-06,
+      "loss": 1.1096,
+      "step": 8560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.36349593120417,
+      "learning_rate": 3.573182554984337e-06,
+      "loss": 1.0745,
+      "step": 8561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0523257409168134,
+      "learning_rate": 3.5710653740331922e-06,
+      "loss": 1.0989,
+      "step": 8562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9123961729912877,
+      "learning_rate": 3.5689486841561482e-06,
+      "loss": 1.0488,
+      "step": 8563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1739221470917856,
+      "learning_rate": 3.566832485514894e-06,
+      "loss": 1.0242,
+      "step": 8564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.291859859487774,
+      "learning_rate": 3.564716778271072e-06,
+      "loss": 1.1423,
+      "step": 8565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8136818066166507,
+      "learning_rate": 3.5626015625862907e-06,
+      "loss": 1.1086,
+      "step": 8566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1487800310356957,
+      "learning_rate": 3.5604868386221215e-06,
+      "loss": 1.0745,
+      "step": 8567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3063920512617195,
+      "learning_rate": 3.5583726065400915e-06,
+      "loss": 1.1077,
+      "step": 8568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0626585356606895,
+      "learning_rate": 3.5562588665017063e-06,
+      "loss": 1.0576,
+      "step": 8569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.115768393442189,
+      "learning_rate": 3.5541456186684188e-06,
+      "loss": 1.1064,
+      "step": 8570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3984116391590136,
+      "learning_rate": 3.5520328632016467e-06,
+      "loss": 1.0833,
+      "step": 8571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0665016487492056,
+      "learning_rate": 3.5499206002627796e-06,
+      "loss": 1.1196,
+      "step": 8572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9254999732221827,
+      "learning_rate": 3.5478088300131584e-06,
+      "loss": 1.0898,
+      "step": 8573
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.184504910855848,
+      "learning_rate": 3.5456975526140915e-06,
+      "loss": 1.1272,
+      "step": 8574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8014483832398036,
+      "learning_rate": 3.5435867682268487e-06,
+      "loss": 1.0544,
+      "step": 8575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2257564792943083,
+      "learning_rate": 3.5414764770126585e-06,
+      "loss": 1.127,
+      "step": 8576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.183998725151154,
+      "learning_rate": 3.539366679132722e-06,
+      "loss": 1.0911,
+      "step": 8577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.207723459069674,
+      "learning_rate": 3.5372573747481907e-06,
+      "loss": 1.061,
+      "step": 8578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.33434646776571,
+      "learning_rate": 3.5351485640201922e-06,
+      "loss": 1.0535,
+      "step": 8579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4552526148256577,
+      "learning_rate": 3.5330402471097957e-06,
+      "loss": 1.0962,
+      "step": 8580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.757306031409813,
+      "learning_rate": 3.530932424178055e-06,
+      "loss": 1.0698,
+      "step": 8581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0828013352473733,
+      "learning_rate": 3.5288250953859727e-06,
+      "loss": 1.0288,
+      "step": 8582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1132419938492677,
+      "learning_rate": 3.5267182608945127e-06,
+      "loss": 1.0369,
+      "step": 8583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.478368825535072,
+      "learning_rate": 3.5246119208646124e-06,
+      "loss": 1.0957,
+      "step": 8584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4209792534334005,
+      "learning_rate": 3.5225060754571626e-06,
+      "loss": 1.0679,
+      "step": 8585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9414763743848837,
+      "learning_rate": 3.5204007248330174e-06,
+      "loss": 1.0884,
+      "step": 8586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1391107978387787,
+      "learning_rate": 3.5182958691529945e-06,
+      "loss": 1.0486,
+      "step": 8587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.121945237114468,
+      "learning_rate": 3.5161915085778696e-06,
+      "loss": 1.0991,
+      "step": 8588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1449493201369187,
+      "learning_rate": 3.51408764326839e-06,
+      "loss": 1.0725,
+      "step": 8589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3579928808962505,
+      "learning_rate": 3.5119842733852573e-06,
+      "loss": 1.1338,
+      "step": 8590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1484977361514446,
+      "learning_rate": 3.5098813990891346e-06,
+      "loss": 1.0742,
+      "step": 8591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7014824867258485,
+      "learning_rate": 3.5077790205406546e-06,
+      "loss": 0.942,
+      "step": 8592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.909709382376801,
+      "learning_rate": 3.505677137900405e-06,
+      "loss": 1.1013,
+      "step": 8593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0580263544045434,
+      "learning_rate": 3.5035757513289393e-06,
+      "loss": 1.0789,
+      "step": 8594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.702561623736968,
+      "learning_rate": 3.50147486098677e-06,
+      "loss": 1.0039,
+      "step": 8595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1824043987911526,
+      "learning_rate": 3.4993744670343723e-06,
+      "loss": 1.1028,
+      "step": 8596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9521615824894365,
+      "learning_rate": 3.4972745696321897e-06,
+      "loss": 1.0854,
+      "step": 8597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3923859251441053,
+      "learning_rate": 3.49517516894062e-06,
+      "loss": 1.0204,
+      "step": 8598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.898767461709442,
+      "learning_rate": 3.4930762651200257e-06,
+      "loss": 1.1646,
+      "step": 8599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4038033758984514,
+      "learning_rate": 3.4909778583307317e-06,
+      "loss": 1.1204,
+      "step": 8600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1872728337099248,
+      "learning_rate": 3.4888799487330226e-06,
+      "loss": 1.0559,
+      "step": 8601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8956150674076526,
+      "learning_rate": 3.4867825364871523e-06,
+      "loss": 1.0503,
+      "step": 8602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0138495327062413,
+      "learning_rate": 3.4846856217533288e-06,
+      "loss": 1.0281,
+      "step": 8603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9553301915500123,
+      "learning_rate": 3.4825892046917207e-06,
+      "loss": 1.0916,
+      "step": 8604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.312742618709723,
+      "learning_rate": 3.4804932854624716e-06,
+      "loss": 1.0591,
+      "step": 8605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5823577454643765,
+      "learning_rate": 3.4783978642256745e-06,
+      "loss": 1.062,
+      "step": 8606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.142594207487853,
+      "learning_rate": 3.476302941141385e-06,
+      "loss": 1.0491,
+      "step": 8607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8298571378899235,
+      "learning_rate": 3.4742085163696283e-06,
+      "loss": 1.05,
+      "step": 8608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2150196770948636,
+      "learning_rate": 3.47211459007038e-06,
+      "loss": 1.0913,
+      "step": 8609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.22628666706663,
+      "learning_rate": 3.470021162403594e-06,
+      "loss": 1.0774,
+      "step": 8610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.399651278038437,
+      "learning_rate": 3.4679282335291674e-06,
+      "loss": 1.0781,
+      "step": 8611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9053154629785265,
+      "learning_rate": 3.4658358036069816e-06,
+      "loss": 1.1038,
+      "step": 8612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.659603182070972,
+      "learning_rate": 3.463743872796854e-06,
+      "loss": 1.1028,
+      "step": 8613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3807358236874996,
+      "learning_rate": 3.4616524412585797e-06,
+      "loss": 1.0576,
+      "step": 8614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5007353004551742,
+      "learning_rate": 3.4595615091519165e-06,
+      "loss": 1.0676,
+      "step": 8615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.139065860445055,
+      "learning_rate": 3.4574710766365737e-06,
+      "loss": 1.1128,
+      "step": 8616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3585003525821002,
+      "learning_rate": 3.4553811438722384e-06,
+      "loss": 1.1074,
+      "step": 8617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3301807084322155,
+      "learning_rate": 3.453291711018544e-06,
+      "loss": 1.0667,
+      "step": 8618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2426635089705798,
+      "learning_rate": 3.4512027782350934e-06,
+      "loss": 1.0825,
+      "step": 8619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.590150894966337,
+      "learning_rate": 3.449114345681448e-06,
+      "loss": 1.1167,
+      "step": 8620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4929595056339817,
+      "learning_rate": 3.4470264135171316e-06,
+      "loss": 1.0984,
+      "step": 8621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8974774689326526,
+      "learning_rate": 3.4449389819016353e-06,
+      "loss": 1.0789,
+      "step": 8622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6887748770333855,
+      "learning_rate": 3.4428520509944065e-06,
+      "loss": 1.0759,
+      "step": 8623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8687012252701125,
+      "learning_rate": 3.44076562095485e-06,
+      "loss": 1.0264,
+      "step": 8624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.872430969557531,
+      "learning_rate": 3.438679691942346e-06,
+      "loss": 1.0923,
+      "step": 8625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8736788584931783,
+      "learning_rate": 3.436594264116224e-06,
+      "loss": 1.0457,
+      "step": 8626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1806098777745686,
+      "learning_rate": 3.4345093376357785e-06,
+      "loss": 1.0002,
+      "step": 8627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5319975806957986,
+      "learning_rate": 3.4324249126602695e-06,
+      "loss": 1.1082,
+      "step": 8628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5679329344319823,
+      "learning_rate": 3.4303409893489093e-06,
+      "loss": 1.083,
+      "step": 8629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1126400113391903,
+      "learning_rate": 3.4282575678608855e-06,
+      "loss": 1.0288,
+      "step": 8630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1127735792565923,
+      "learning_rate": 3.4261746483553382e-06,
+      "loss": 1.0908,
+      "step": 8631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.334186351288776,
+      "learning_rate": 3.4240922309913715e-06,
+      "loss": 1.0771,
+      "step": 8632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1295183519302783,
+      "learning_rate": 3.4220103159280494e-06,
+      "loss": 1.0496,
+      "step": 8633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.74181595734321,
+      "learning_rate": 3.4199289033243954e-06,
+      "loss": 1.0432,
+      "step": 8634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3486744728456097,
+      "learning_rate": 3.417847993339407e-06,
+      "loss": 1.0996,
+      "step": 8635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.295976617239099,
+      "learning_rate": 3.4157675861320295e-06,
+      "loss": 1.1194,
+      "step": 8636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0396096262923264,
+      "learning_rate": 3.413687681861171e-06,
+      "loss": 1.0793,
+      "step": 8637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6047446464290163,
+      "learning_rate": 3.411608280685713e-06,
+      "loss": 1.1355,
+      "step": 8638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7297471716677535,
+      "learning_rate": 3.409529382764486e-06,
+      "loss": 1.0034,
+      "step": 8639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8391882691307297,
+      "learning_rate": 3.4074509882562877e-06,
+      "loss": 1.1184,
+      "step": 8640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1210269565739903,
+      "learning_rate": 3.405373097319875e-06,
+      "loss": 1.0574,
+      "step": 8641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.580217168327413,
+      "learning_rate": 3.403295710113966e-06,
+      "loss": 1.0435,
+      "step": 8642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.646795443640686,
+      "learning_rate": 3.401218826797247e-06,
+      "loss": 1.1316,
+      "step": 8643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.314914456237012,
+      "learning_rate": 3.3991424475283543e-06,
+      "loss": 1.0562,
+      "step": 8644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.041788278591201,
+      "learning_rate": 3.397066572465899e-06,
+      "loss": 1.0281,
+      "step": 8645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.035597138773511,
+      "learning_rate": 3.394991201768447e-06,
+      "loss": 1.095,
+      "step": 8646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1025458713429166,
+      "learning_rate": 3.392916335594516e-06,
+      "loss": 1.041,
+      "step": 8647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2444196076696343,
+      "learning_rate": 3.3908419741026034e-06,
+      "loss": 1.103,
+      "step": 8648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9043649462504537,
+      "learning_rate": 3.3887681174511534e-06,
+      "loss": 1.1421,
+      "step": 8649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5245882349586273,
+      "learning_rate": 3.3866947657985827e-06,
+      "loss": 1.0654,
+      "step": 8650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9384138942728466,
+      "learning_rate": 3.384621919303264e-06,
+      "loss": 1.0557,
+      "step": 8651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9662764587545505,
+      "learning_rate": 3.3825495781235286e-06,
+      "loss": 1.0549,
+      "step": 8652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.427124764161272,
+      "learning_rate": 3.3804777424176737e-06,
+      "loss": 1.1135,
+      "step": 8653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2084281313625085,
+      "learning_rate": 3.378406412343953e-06,
+      "loss": 1.0452,
+      "step": 8654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.744162906271187,
+      "learning_rate": 3.376335588060591e-06,
+      "loss": 0.9792,
+      "step": 8655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7910528608451695,
+      "learning_rate": 3.374265269725766e-06,
+      "loss": 1.0396,
+      "step": 8656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.798365527357813,
+      "learning_rate": 3.3721954574976137e-06,
+      "loss": 1.0845,
+      "step": 8657
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.704307012488549,
+      "learning_rate": 3.3701261515342443e-06,
+      "loss": 1.0085,
+      "step": 8658
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8658282021148245,
+      "learning_rate": 3.3680573519937187e-06,
+      "loss": 1.0981,
+      "step": 8659
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2366409546249915,
+      "learning_rate": 3.365989059034063e-06,
+      "loss": 1.1089,
+      "step": 8660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1489041946269842,
+      "learning_rate": 3.3639212728132607e-06,
+      "loss": 1.0479,
+      "step": 8661
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.695056908019172,
+      "learning_rate": 3.36185399348926e-06,
+      "loss": 1.0867,
+      "step": 8662
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3096234991139717,
+      "learning_rate": 3.3597872212199746e-06,
+      "loss": 1.084,
+      "step": 8663
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.922315381549084,
+      "learning_rate": 3.3577209561632705e-06,
+      "loss": 1.0369,
+      "step": 8664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0471098144537683,
+      "learning_rate": 3.355655198476978e-06,
+      "loss": 1.0786,
+      "step": 8665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7100953240692682,
+      "learning_rate": 3.3535899483189006e-06,
+      "loss": 1.0701,
+      "step": 8666
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.370830987981228,
+      "learning_rate": 3.351525205846776e-06,
+      "loss": 1.0962,
+      "step": 8667
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1737461760284926,
+      "learning_rate": 3.3494609712183323e-06,
+      "loss": 1.1306,
+      "step": 8668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9718239285716765,
+      "learning_rate": 3.3473972445912418e-06,
+      "loss": 1.0405,
+      "step": 8669
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4078743312861395,
+      "learning_rate": 3.345334026123138e-06,
+      "loss": 1.1006,
+      "step": 8670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.325278239031107,
+      "learning_rate": 3.343271315971629e-06,
+      "loss": 1.0232,
+      "step": 8671
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4073114734433263,
+      "learning_rate": 3.3412091142942683e-06,
+      "loss": 1.0449,
+      "step": 8672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.252594038164209,
+      "learning_rate": 3.3391474212485795e-06,
+      "loss": 1.1121,
+      "step": 8673
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.578603674617682,
+      "learning_rate": 3.3370862369920444e-06,
+      "loss": 1.0676,
+      "step": 8674
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6446827891028764,
+      "learning_rate": 3.3350255616821026e-06,
+      "loss": 1.0181,
+      "step": 8675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4993935440387243,
+      "learning_rate": 3.3329653954761655e-06,
+      "loss": 1.1074,
+      "step": 8676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3631751707507607,
+      "learning_rate": 3.3309057385315966e-06,
+      "loss": 1.0596,
+      "step": 8677
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1048221913726155,
+      "learning_rate": 3.3288465910057187e-06,
+      "loss": 1.0864,
+      "step": 8678
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.09195830094838,
+      "learning_rate": 3.32678795305583e-06,
+      "loss": 1.0779,
+      "step": 8679
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1415657077577697,
+      "learning_rate": 3.3247298248391656e-06,
+      "loss": 1.0547,
+      "step": 8680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.166478390910717,
+      "learning_rate": 3.322672206512946e-06,
+      "loss": 1.0012,
+      "step": 8681
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.153335682523452,
+      "learning_rate": 3.3206150982343354e-06,
+      "loss": 1.0916,
+      "step": 8682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4219830654610086,
+      "learning_rate": 3.3185585001604735e-06,
+      "loss": 1.0647,
+      "step": 8683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.315778469636325,
+      "learning_rate": 3.3165024124484504e-06,
+      "loss": 1.1118,
+      "step": 8684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.219516324178421,
+      "learning_rate": 3.3144468352553183e-06,
+      "loss": 1.1301,
+      "step": 8685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5553058352401226,
+      "learning_rate": 3.312391768738095e-06,
+      "loss": 0.9927,
+      "step": 8686
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2185334688623013,
+      "learning_rate": 3.3103372130537515e-06,
+      "loss": 1.0645,
+      "step": 8687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.004665023376335,
+      "learning_rate": 3.308283168359232e-06,
+      "loss": 1.0728,
+      "step": 8688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.890844497691609,
+      "learning_rate": 3.306229634811432e-06,
+      "loss": 1.113,
+      "step": 8689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.682518903520515,
+      "learning_rate": 3.304176612567207e-06,
+      "loss": 1.1062,
+      "step": 8690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0052511615823816,
+      "learning_rate": 3.3021241017833837e-06,
+      "loss": 1.1282,
+      "step": 8691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.075646073499647,
+      "learning_rate": 3.3000721026167393e-06,
+      "loss": 1.0686,
+      "step": 8692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8737168906862642,
+      "learning_rate": 3.298020615224017e-06,
+      "loss": 0.999,
+      "step": 8693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.710871295777596,
+      "learning_rate": 3.295969639761919e-06,
+      "loss": 1.1304,
+      "step": 8694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1766093619137785,
+      "learning_rate": 3.2939191763871047e-06,
+      "loss": 1.1062,
+      "step": 8695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.103679714462324,
+      "learning_rate": 3.291869225256207e-06,
+      "loss": 1.0498,
+      "step": 8696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.194212040529098,
+      "learning_rate": 3.2898197865258074e-06,
+      "loss": 1.0979,
+      "step": 8697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8489043592180394,
+      "learning_rate": 3.287770860352448e-06,
+      "loss": 1.0789,
+      "step": 8698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9318593877246264,
+      "learning_rate": 3.2857224468926487e-06,
+      "loss": 1.0535,
+      "step": 8699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1194957878500813,
+      "learning_rate": 3.2836745463028606e-06,
+      "loss": 1.1006,
+      "step": 8700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1607548317859346,
+      "learning_rate": 3.2816271587395255e-06,
+      "loss": 1.0913,
+      "step": 8701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.137401030512852,
+      "learning_rate": 3.2795802843590276e-06,
+      "loss": 1.0923,
+      "step": 8702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.727921730349188,
+      "learning_rate": 3.277533923317715e-06,
+      "loss": 1.1162,
+      "step": 8703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.980585650461302,
+      "learning_rate": 3.2754880757719065e-06,
+      "loss": 1.0354,
+      "step": 8704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.379574383776822,
+      "learning_rate": 3.2734427418778702e-06,
+      "loss": 1.0549,
+      "step": 8705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8155891270434807,
+      "learning_rate": 3.271397921791838e-06,
+      "loss": 1.0034,
+      "step": 8706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0517532363857294,
+      "learning_rate": 3.2693536156700047e-06,
+      "loss": 1.0383,
+      "step": 8707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1770241908883476,
+      "learning_rate": 3.2673098236685197e-06,
+      "loss": 1.1033,
+      "step": 8708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6549901625752645,
+      "learning_rate": 3.265266545943506e-06,
+      "loss": 1.0708,
+      "step": 8709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3284845500751663,
+      "learning_rate": 3.263223782651036e-06,
+      "loss": 1.0479,
+      "step": 8710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6041770126514927,
+      "learning_rate": 3.261181533947143e-06,
+      "loss": 1.1111,
+      "step": 8711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.133451216083836,
+      "learning_rate": 3.259139799987834e-06,
+      "loss": 1.0735,
+      "step": 8712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.791062792647217,
+      "learning_rate": 3.2570985809290525e-06,
+      "loss": 1.0657,
+      "step": 8713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.117972422705173,
+      "learning_rate": 3.255057876926727e-06,
+      "loss": 1.0181,
+      "step": 8714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.296991453695988,
+      "learning_rate": 3.2530176881367346e-06,
+      "loss": 1.105,
+      "step": 8715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4517716438203268,
+      "learning_rate": 3.250978014714912e-06,
+      "loss": 1.0898,
+      "step": 8716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.075142205798267,
+      "learning_rate": 3.2489388568170643e-06,
+      "loss": 1.0967,
+      "step": 8717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2011595270725266,
+      "learning_rate": 3.246900214598947e-06,
+      "loss": 1.0732,
+      "step": 8718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8266372773071966,
+      "learning_rate": 3.244862088216293e-06,
+      "loss": 1.0581,
+      "step": 8719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.152597037119203,
+      "learning_rate": 3.242824477824772e-06,
+      "loss": 1.0566,
+      "step": 8720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9896678657334097,
+      "learning_rate": 3.2407873835800287e-06,
+      "loss": 1.0684,
+      "step": 8721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4903750434630543,
+      "learning_rate": 3.2387508056376726e-06,
+      "loss": 0.9829,
+      "step": 8722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1491915038670757,
+      "learning_rate": 3.236714744153262e-06,
+      "loss": 1.0779,
+      "step": 8723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.08684457118733,
+      "learning_rate": 3.234679199282327e-06,
+      "loss": 1.0764,
+      "step": 8724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.826706313528844,
+      "learning_rate": 3.232644171180349e-06,
+      "loss": 1.1077,
+      "step": 8725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0967463943601232,
+      "learning_rate": 3.230609660002775e-06,
+      "loss": 1.0374,
+      "step": 8726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8751475241086677,
+      "learning_rate": 3.2285756659050094e-06,
+      "loss": 1.0503,
+      "step": 8727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8311879819773247,
+      "learning_rate": 3.2265421890424164e-06,
+      "loss": 0.9756,
+      "step": 8728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9317481301913135,
+      "learning_rate": 3.2245092295703294e-06,
+      "loss": 1.0476,
+      "step": 8729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7163339824028303,
+      "learning_rate": 3.2224767876440345e-06,
+      "loss": 1.0408,
+      "step": 8730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.804536156070923,
+      "learning_rate": 3.2204448634187734e-06,
+      "loss": 1.0706,
+      "step": 8731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8297848771585397,
+      "learning_rate": 3.218413457049767e-06,
+      "loss": 1.022,
+      "step": 8732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.465628834049187,
+      "learning_rate": 3.21638256869217e-06,
+      "loss": 1.1226,
+      "step": 8733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2713849115197324,
+      "learning_rate": 3.2143521985011218e-06,
+      "loss": 1.0513,
+      "step": 8734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6633331991587244,
+      "learning_rate": 3.212322346631709e-06,
+      "loss": 1.1482,
+      "step": 8735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3814328599516648,
+      "learning_rate": 3.210293013238979e-06,
+      "loss": 1.0281,
+      "step": 8736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8367838463662354,
+      "learning_rate": 3.2082641984779474e-06,
+      "loss": 1.0955,
+      "step": 8737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4440370802584144,
+      "learning_rate": 3.2062359025035826e-06,
+      "loss": 1.0466,
+      "step": 8738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6203300998994234,
+      "learning_rate": 3.204208125470817e-06,
+      "loss": 1.0007,
+      "step": 8739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.740233990777302,
+      "learning_rate": 3.202180867534541e-06,
+      "loss": 1.0273,
+      "step": 8740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.387960557595562,
+      "learning_rate": 3.2001541288496053e-06,
+      "loss": 1.0999,
+      "step": 8741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5454476055988122,
+      "learning_rate": 3.1981279095708264e-06,
+      "loss": 1.0769,
+      "step": 8742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6418584694369773,
+      "learning_rate": 3.196102209852976e-06,
+      "loss": 1.1221,
+      "step": 8743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.58661765644795,
+      "learning_rate": 3.1940770298507815e-06,
+      "loss": 1.0681,
+      "step": 8744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.072700106475027,
+      "learning_rate": 3.1920523697189487e-06,
+      "loss": 1.0981,
+      "step": 8745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.105900778123855,
+      "learning_rate": 3.1900282296121166e-06,
+      "loss": 1.0461,
+      "step": 8746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2035287408272715,
+      "learning_rate": 3.188004609684908e-06,
+      "loss": 1.0884,
+      "step": 8747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6821272016879725,
+      "learning_rate": 3.185981510091897e-06,
+      "loss": 1.0366,
+      "step": 8748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.913264939526826,
+      "learning_rate": 3.183958930987612e-06,
+      "loss": 1.0393,
+      "step": 8749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9816187718341016,
+      "learning_rate": 3.181936872526554e-06,
+      "loss": 1.1282,
+      "step": 8750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2079841787086054,
+      "learning_rate": 3.179915334863173e-06,
+      "loss": 1.123,
+      "step": 8751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2853747109107387,
+      "learning_rate": 3.177894318151894e-06,
+      "loss": 1.1016,
+      "step": 8752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.647470459612835,
+      "learning_rate": 3.175873822547081e-06,
+      "loss": 1.1128,
+      "step": 8753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.411157672375403,
+      "learning_rate": 3.17385384820307e-06,
+      "loss": 1.0486,
+      "step": 8754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.771600833454915,
+      "learning_rate": 3.171834395274165e-06,
+      "loss": 1.0962,
+      "step": 8755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1139578850540666,
+      "learning_rate": 3.1698154639146126e-06,
+      "loss": 1.135,
+      "step": 8756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0738806440921955,
+      "learning_rate": 3.1677970542786375e-06,
+      "loss": 1.0898,
+      "step": 8757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.086835860064632,
+      "learning_rate": 3.165779166520415e-06,
+      "loss": 1.0847,
+      "step": 8758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.69565990616466,
+      "learning_rate": 3.163761800794073e-06,
+      "loss": 1.1404,
+      "step": 8759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.952245407573835,
+      "learning_rate": 3.1617449572537152e-06,
+      "loss": 1.1191,
+      "step": 8760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8367854862326483,
+      "learning_rate": 3.159728636053393e-06,
+      "loss": 1.0984,
+      "step": 8761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8484779264794757,
+      "learning_rate": 3.15771283734713e-06,
+      "loss": 1.0659,
+      "step": 8762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.846162617868798,
+      "learning_rate": 3.1556975612888997e-06,
+      "loss": 1.0349,
+      "step": 8763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4005620366538776,
+      "learning_rate": 3.1536828080326344e-06,
+      "loss": 1.0933,
+      "step": 8764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3253621079151836,
+      "learning_rate": 3.151668577732242e-06,
+      "loss": 1.0984,
+      "step": 8765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1137221726675475,
+      "learning_rate": 3.1496548705415663e-06,
+      "loss": 1.0151,
+      "step": 8766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.228292539924857,
+      "learning_rate": 3.1476416866144343e-06,
+      "loss": 1.1682,
+      "step": 8767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.307744951768023,
+      "learning_rate": 3.1456290261046184e-06,
+      "loss": 1.0439,
+      "step": 8768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1229397477808587,
+      "learning_rate": 3.1436168891658546e-06,
+      "loss": 1.0178,
+      "step": 8769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0329976882743486,
+      "learning_rate": 3.1416052759518447e-06,
+      "loss": 1.0645,
+      "step": 8770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.869494269819767,
+      "learning_rate": 3.1395941866162426e-06,
+      "loss": 1.0134,
+      "step": 8771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.310728868261693,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 1.0308,
+      "step": 8772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3762456833827446,
+      "learning_rate": 3.1355735801946918e-06,
+      "loss": 1.0315,
+      "step": 8773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.241740375159284,
+      "learning_rate": 3.133564063415854e-06,
+      "loss": 1.0383,
+      "step": 8774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8684009248655955,
+      "learning_rate": 3.131555071129654e-06,
+      "loss": 1.0437,
+      "step": 8775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.007243567542101,
+      "learning_rate": 3.1295466034895482e-06,
+      "loss": 1.0486,
+      "step": 8776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3321418038114112,
+      "learning_rate": 3.127538660648949e-06,
+      "loss": 1.0732,
+      "step": 8777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9278154206080695,
+      "learning_rate": 3.125531242761243e-06,
+      "loss": 1.033,
+      "step": 8778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.003591051620025,
+      "learning_rate": 3.123524349979753e-06,
+      "loss": 1.0566,
+      "step": 8779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9156124495154905,
+      "learning_rate": 3.1215179824577856e-06,
+      "loss": 1.0947,
+      "step": 8780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.228765280095443,
+      "learning_rate": 3.119512140348595e-06,
+      "loss": 1.1204,
+      "step": 8781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.430122256421375,
+      "learning_rate": 3.117506823805393e-06,
+      "loss": 1.0369,
+      "step": 8782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5952402915970976,
+      "learning_rate": 3.1155020329813625e-06,
+      "loss": 1.0415,
+      "step": 8783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.508888162141925,
+      "learning_rate": 3.113497768029632e-06,
+      "loss": 1.0869,
+      "step": 8784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.192392189359853,
+      "learning_rate": 3.1114940291033092e-06,
+      "loss": 1.0317,
+      "step": 8785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6443249500579658,
+      "learning_rate": 3.1094908163554372e-06,
+      "loss": 1.1047,
+      "step": 8786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7170674195001694,
+      "learning_rate": 3.1074881299390338e-06,
+      "loss": 1.1089,
+      "step": 8787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.092089558207915,
+      "learning_rate": 3.1054859700070794e-06,
+      "loss": 1.093,
+      "step": 8788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.185981641940469,
+      "learning_rate": 3.1034843367125033e-06,
+      "loss": 1.0261,
+      "step": 8789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.140650117771331,
+      "learning_rate": 3.1014832302082064e-06,
+      "loss": 0.9929,
+      "step": 8790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.30183063231869,
+      "learning_rate": 3.099482650647039e-06,
+      "loss": 1.0747,
+      "step": 8791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.136007973895834,
+      "learning_rate": 3.097482598181816e-06,
+      "loss": 1.0361,
+      "step": 8792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6120019076435823,
+      "learning_rate": 3.0954830729653116e-06,
+      "loss": 1.0637,
+      "step": 8793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0435737090693546,
+      "learning_rate": 3.093484075150257e-06,
+      "loss": 1.0576,
+      "step": 8794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.810513536335758,
+      "learning_rate": 3.0914856048893505e-06,
+      "loss": 1.0352,
+      "step": 8795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.982348499569619,
+      "learning_rate": 3.089487662335243e-06,
+      "loss": 1.094,
+      "step": 8796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1188122506907163,
+      "learning_rate": 3.0874902476405445e-06,
+      "loss": 1.0603,
+      "step": 8797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.641413322682071,
+      "learning_rate": 3.0854933609578354e-06,
+      "loss": 1.0886,
+      "step": 8798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.206031330092114,
+      "learning_rate": 3.083497002439637e-06,
+      "loss": 1.0107,
+      "step": 8799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1284851236996314,
+      "learning_rate": 3.081501172238449e-06,
+      "loss": 1.0266,
+      "step": 8800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4714421545124488,
+      "learning_rate": 3.0795058705067216e-06,
+      "loss": 1.0962,
+      "step": 8801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4177086098312337,
+      "learning_rate": 3.0775110973968604e-06,
+      "loss": 1.0964,
+      "step": 8802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.237585107387061,
+      "learning_rate": 3.0755168530612444e-06,
+      "loss": 1.0095,
+      "step": 8803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.966789265529324,
+      "learning_rate": 3.0735231376522e-06,
+      "loss": 1.0178,
+      "step": 8804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.416300608103811,
+      "learning_rate": 3.0715299513220174e-06,
+      "loss": 1.1182,
+      "step": 8805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.752546163274374,
+      "learning_rate": 3.0695372942229462e-06,
+      "loss": 1.061,
+      "step": 8806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.453604223320027,
+      "learning_rate": 3.067545166507192e-06,
+      "loss": 0.9915,
+      "step": 8807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1936623738696555,
+      "learning_rate": 3.0655535683269298e-06,
+      "loss": 1.093,
+      "step": 8808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5869603463034316,
+      "learning_rate": 3.063562499834284e-06,
+      "loss": 1.0312,
+      "step": 8809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.364724157508487,
+      "learning_rate": 3.061571961181341e-06,
+      "loss": 1.0928,
+      "step": 8810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.970284651325236,
+      "learning_rate": 3.0595819525201554e-06,
+      "loss": 1.0789,
+      "step": 8811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2410505494009065,
+      "learning_rate": 3.057592474002723e-06,
+      "loss": 1.0129,
+      "step": 8812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.964425528975042,
+      "learning_rate": 3.055603525781019e-06,
+      "loss": 1.0332,
+      "step": 8813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7154110795992517,
+      "learning_rate": 3.0536151080069665e-06,
+      "loss": 1.0562,
+      "step": 8814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8665417663862387,
+      "learning_rate": 3.051627220832446e-06,
+      "loss": 1.0454,
+      "step": 8815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1781623814543765,
+      "learning_rate": 3.0496398644093106e-06,
+      "loss": 1.1155,
+      "step": 8816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.218507841089798,
+      "learning_rate": 3.0476530388893598e-06,
+      "loss": 1.0984,
+      "step": 8817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1582912795922167,
+      "learning_rate": 3.045666744424358e-06,
+      "loss": 1.1174,
+      "step": 8818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2211096731412705,
+      "learning_rate": 3.0436809811660283e-06,
+      "loss": 1.02,
+      "step": 8819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.026354324255629,
+      "learning_rate": 3.0416957492660494e-06,
+      "loss": 1.0312,
+      "step": 8820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4227059470750176,
+      "learning_rate": 3.0397110488760707e-06,
+      "loss": 1.1433,
+      "step": 8821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.02414586381065,
+      "learning_rate": 3.0377268801476855e-06,
+      "loss": 1.0518,
+      "step": 8822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3770971630160322,
+      "learning_rate": 3.035743243232463e-06,
+      "loss": 1.125,
+      "step": 8823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0058040324105377,
+      "learning_rate": 3.033760138281918e-06,
+      "loss": 1.0583,
+      "step": 8824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2458861165685384,
+      "learning_rate": 3.0317775654475313e-06,
+      "loss": 1.0583,
+      "step": 8825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.971230505699355,
+      "learning_rate": 3.029795524880742e-06,
+      "loss": 1.0684,
+      "step": 8826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8743671608006856,
+      "learning_rate": 3.027814016732944e-06,
+      "loss": 1.0989,
+      "step": 8827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1876006932984935,
+      "learning_rate": 3.0258330411555015e-06,
+      "loss": 1.0513,
+      "step": 8828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7275761233189617,
+      "learning_rate": 3.023852598299728e-06,
+      "loss": 1.0803,
+      "step": 8829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.058011076072028,
+      "learning_rate": 3.021872688316896e-06,
+      "loss": 1.0049,
+      "step": 8830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.763327896001516,
+      "learning_rate": 3.019893311358252e-06,
+      "loss": 1.103,
+      "step": 8831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6270190735215166,
+      "learning_rate": 3.017914467574976e-06,
+      "loss": 1.0813,
+      "step": 8832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.311655340602296,
+      "learning_rate": 3.0159361571182335e-06,
+      "loss": 1.0342,
+      "step": 8833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3613468462784533,
+      "learning_rate": 3.0139583801391324e-06,
+      "loss": 1.1016,
+      "step": 8834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.059388614205715,
+      "learning_rate": 3.0119811367887443e-06,
+      "loss": 1.1355,
+      "step": 8835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.871879091839734,
+      "learning_rate": 3.010004427218106e-06,
+      "loss": 1.0308,
+      "step": 8836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0108325174770916,
+      "learning_rate": 3.0080282515782045e-06,
+      "loss": 1.0234,
+      "step": 8837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3728641130272528,
+      "learning_rate": 3.0060526100199914e-06,
+      "loss": 1.0266,
+      "step": 8838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.085347531373802,
+      "learning_rate": 3.0040775026943757e-06,
+      "loss": 1.0791,
+      "step": 8839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.185526096994264,
+      "learning_rate": 3.002102929752222e-06,
+      "loss": 1.0774,
+      "step": 8840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.74439218757542,
+      "learning_rate": 3.000128891344365e-06,
+      "loss": 1.0,
+      "step": 8841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.630068519466281,
+      "learning_rate": 2.9981553876215883e-06,
+      "loss": 0.9883,
+      "step": 8842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.490779580364355,
+      "learning_rate": 2.9961824187346357e-06,
+      "loss": 1.1072,
+      "step": 8843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.412847563806054,
+      "learning_rate": 2.9942099848342176e-06,
+      "loss": 1.1399,
+      "step": 8844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9024680184513505,
+      "learning_rate": 2.9922380860709943e-06,
+      "loss": 1.1025,
+      "step": 8845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1261319647233505,
+      "learning_rate": 2.990266722595592e-06,
+      "loss": 1.1216,
+      "step": 8846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.982153455906342,
+      "learning_rate": 2.9882958945585907e-06,
+      "loss": 1.0542,
+      "step": 8847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.403868076322946,
+      "learning_rate": 2.98632560211053e-06,
+      "loss": 1.095,
+      "step": 8848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4495203546768494,
+      "learning_rate": 2.984355845401917e-06,
+      "loss": 1.0945,
+      "step": 8849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3635990277082497,
+      "learning_rate": 2.982386624583209e-06,
+      "loss": 1.1194,
+      "step": 8850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5495251933839724,
+      "learning_rate": 2.980417939804824e-06,
+      "loss": 1.0359,
+      "step": 8851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.139233012547337,
+      "learning_rate": 2.9784497912171395e-06,
+      "loss": 1.0161,
+      "step": 8852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0099043116405,
+      "learning_rate": 2.9764821789704903e-06,
+      "loss": 1.1589,
+      "step": 8853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1204337277380345,
+      "learning_rate": 2.9745151032151785e-06,
+      "loss": 1.0938,
+      "step": 8854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.414891150600916,
+      "learning_rate": 2.9725485641014562e-06,
+      "loss": 1.0664,
+      "step": 8855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6000994191177713,
+      "learning_rate": 2.9705825617795335e-06,
+      "loss": 1.033,
+      "step": 8856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.917508372089592,
+      "learning_rate": 2.968617096399592e-06,
+      "loss": 1.0654,
+      "step": 8857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1161067859915867,
+      "learning_rate": 2.9666521681117576e-06,
+      "loss": 1.0522,
+      "step": 8858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4785457994779074,
+      "learning_rate": 2.964687777066123e-06,
+      "loss": 1.1267,
+      "step": 8859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.480695406128728,
+      "learning_rate": 2.962723923412738e-06,
+      "loss": 1.0686,
+      "step": 8860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.53533853843338,
+      "learning_rate": 2.960760607301608e-06,
+      "loss": 1.1013,
+      "step": 8861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3749660326321282,
+      "learning_rate": 2.958797828882708e-06,
+      "loss": 1.0952,
+      "step": 8862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4599244896144996,
+      "learning_rate": 2.9568355883059573e-06,
+      "loss": 1.0581,
+      "step": 8863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.750533486828739,
+      "learning_rate": 2.9548738857212523e-06,
+      "loss": 1.0764,
+      "step": 8864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.875639643792999,
+      "learning_rate": 2.9529127212784236e-06,
+      "loss": 1.0508,
+      "step": 8865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8627471333770313,
+      "learning_rate": 2.9509520951272864e-06,
+      "loss": 1.0842,
+      "step": 8866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.668402217098528,
+      "learning_rate": 2.948992007417597e-06,
+      "loss": 1.1196,
+      "step": 8867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1211151104710555,
+      "learning_rate": 2.947032458299076e-06,
+      "loss": 1.0664,
+      "step": 8868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.527591646825976,
+      "learning_rate": 2.9450734479214095e-06,
+      "loss": 1.04,
+      "step": 8869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.056790357426596,
+      "learning_rate": 2.9431149764342327e-06,
+      "loss": 1.0625,
+      "step": 8870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.038369319136677,
+      "learning_rate": 2.941157043987143e-06,
+      "loss": 1.116,
+      "step": 8871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5540993173737534,
+      "learning_rate": 2.939199650729698e-06,
+      "loss": 1.0271,
+      "step": 8872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.531906591724541,
+      "learning_rate": 2.937242796811409e-06,
+      "loss": 1.063,
+      "step": 8873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1598884773505422,
+      "learning_rate": 2.9352864823817586e-06,
+      "loss": 1.0596,
+      "step": 8874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2389196956731445,
+      "learning_rate": 2.933330707590174e-06,
+      "loss": 1.0354,
+      "step": 8875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.929507739751081,
+      "learning_rate": 2.9313754725860455e-06,
+      "loss": 1.0261,
+      "step": 8876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.13391051736134,
+      "learning_rate": 2.929420777518729e-06,
+      "loss": 1.0234,
+      "step": 8877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2940209349233798,
+      "learning_rate": 2.9274666225375324e-06,
+      "loss": 1.0671,
+      "step": 8878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.952030721183352,
+      "learning_rate": 2.9255130077917226e-06,
+      "loss": 1.1514,
+      "step": 8879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.075360163529387,
+      "learning_rate": 2.9235599334305266e-06,
+      "loss": 1.1428,
+      "step": 8880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.049317286317941,
+      "learning_rate": 2.921607399603126e-06,
+      "loss": 1.0215,
+      "step": 8881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.866180401887978,
+      "learning_rate": 2.9196554064586734e-06,
+      "loss": 1.0527,
+      "step": 8882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9437554856406667,
+      "learning_rate": 2.917703954146267e-06,
+      "loss": 0.9736,
+      "step": 8883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3919372284024227,
+      "learning_rate": 2.9157530428149683e-06,
+      "loss": 1.02,
+      "step": 8884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1550701511197334,
+      "learning_rate": 2.913802672613799e-06,
+      "loss": 1.0308,
+      "step": 8885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.951915217060666,
+      "learning_rate": 2.911852843691734e-06,
+      "loss": 0.978,
+      "step": 8886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1238951634350696,
+      "learning_rate": 2.9099035561977175e-06,
+      "loss": 1.0701,
+      "step": 8887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.864449508116668,
+      "learning_rate": 2.9079548102806433e-06,
+      "loss": 1.0354,
+      "step": 8888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.329887486619615,
+      "learning_rate": 2.906006606089361e-06,
+      "loss": 1.0671,
+      "step": 8889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.24479975928667,
+      "learning_rate": 2.9040589437726928e-06,
+      "loss": 1.094,
+      "step": 8890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9478662019867934,
+      "learning_rate": 2.902111823479408e-06,
+      "loss": 1.0996,
+      "step": 8891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2573832591441105,
+      "learning_rate": 2.900165245358235e-06,
+      "loss": 1.0554,
+      "step": 8892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5243648763840962,
+      "learning_rate": 2.898219209557864e-06,
+      "loss": 1.0637,
+      "step": 8893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.24594304214635,
+      "learning_rate": 2.8962737162269405e-06,
+      "loss": 1.1143,
+      "step": 8894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.206510188442137,
+      "learning_rate": 2.8943287655140774e-06,
+      "loss": 1.116,
+      "step": 8895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.804332385835979,
+      "learning_rate": 2.8923843575678335e-06,
+      "loss": 1.0533,
+      "step": 8896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.033630720765998,
+      "learning_rate": 2.890440492536737e-06,
+      "loss": 1.0981,
+      "step": 8897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65909194749005,
+      "learning_rate": 2.8884971705692723e-06,
+      "loss": 1.0083,
+      "step": 8898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1830591936312898,
+      "learning_rate": 2.8865543918138685e-06,
+      "loss": 1.0869,
+      "step": 8899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.155861088218322,
+      "learning_rate": 2.8846121564189356e-06,
+      "loss": 1.0889,
+      "step": 8900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1339966504481698,
+      "learning_rate": 2.882670464532824e-06,
+      "loss": 1.0938,
+      "step": 8901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9829470821379878,
+      "learning_rate": 2.8807293163038572e-06,
+      "loss": 1.0708,
+      "step": 8902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.325051804105147,
+      "learning_rate": 2.8787887118803058e-06,
+      "loss": 1.0715,
+      "step": 8903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.765711697637907,
+      "learning_rate": 2.8768486514104033e-06,
+      "loss": 1.0894,
+      "step": 8904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5081486645759163,
+      "learning_rate": 2.8749091350423406e-06,
+      "loss": 1.1543,
+      "step": 8905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7233307518094083,
+      "learning_rate": 2.8729701629242644e-06,
+      "loss": 1.0413,
+      "step": 8906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1674721066308815,
+      "learning_rate": 2.8710317352042914e-06,
+      "loss": 1.0657,
+      "step": 8907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4049166100511226,
+      "learning_rate": 2.869093852030482e-06,
+      "loss": 1.0713,
+      "step": 8908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.072461117998515,
+      "learning_rate": 2.8671565135508606e-06,
+      "loss": 1.0591,
+      "step": 8909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2162245116442834,
+      "learning_rate": 2.865219719913417e-06,
+      "loss": 1.0889,
+      "step": 8910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.977055698682901,
+      "learning_rate": 2.8632834712660882e-06,
+      "loss": 1.0437,
+      "step": 8911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1432086067335847,
+      "learning_rate": 2.861347767756777e-06,
+      "loss": 1.1204,
+      "step": 8912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9999887927536446,
+      "learning_rate": 2.85941260953334e-06,
+      "loss": 1.1128,
+      "step": 8913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.152950833022915,
+      "learning_rate": 2.8574779967435916e-06,
+      "loss": 1.0588,
+      "step": 8914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.626803864451643,
+      "learning_rate": 2.855543929535315e-06,
+      "loss": 1.1013,
+      "step": 8915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3075695901649738,
+      "learning_rate": 2.853610408056239e-06,
+      "loss": 1.0801,
+      "step": 8916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0189254939325925,
+      "learning_rate": 2.851677432454052e-06,
+      "loss": 1.0417,
+      "step": 8917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8083408879433884,
+      "learning_rate": 2.8497450028764152e-06,
+      "loss": 1.1057,
+      "step": 8918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.380772197796072,
+      "learning_rate": 2.847813119470925e-06,
+      "loss": 1.0278,
+      "step": 8919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6949944714662446,
+      "learning_rate": 2.845881782385156e-06,
+      "loss": 1.0299,
+      "step": 8920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1197751377855436,
+      "learning_rate": 2.8439509917666317e-06,
+      "loss": 1.0745,
+      "step": 8921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0239740629944527,
+      "learning_rate": 2.8420207477628324e-06,
+      "loss": 1.0891,
+      "step": 8922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4820195120456776,
+      "learning_rate": 2.8400910505212044e-06,
+      "loss": 1.1096,
+      "step": 8923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.263335576899768,
+      "learning_rate": 2.838161900189146e-06,
+      "loss": 1.0654,
+      "step": 8924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1197633756100194,
+      "learning_rate": 2.8362332969140147e-06,
+      "loss": 1.0718,
+      "step": 8925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.881072776438979,
+      "learning_rate": 2.834305240843127e-06,
+      "loss": 1.0974,
+      "step": 8926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2015524951978396,
+      "learning_rate": 2.8323777321237544e-06,
+      "loss": 1.0444,
+      "step": 8927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.290969649228619,
+      "learning_rate": 2.830450770903136e-06,
+      "loss": 1.0645,
+      "step": 8928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.824464496011942,
+      "learning_rate": 2.828524357328456e-06,
+      "loss": 1.1218,
+      "step": 8929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5594367315530833,
+      "learning_rate": 2.82659849154687e-06,
+      "loss": 1.0579,
+      "step": 8930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3301007753172414,
+      "learning_rate": 2.8246731737054855e-06,
+      "loss": 1.1575,
+      "step": 8931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8629071152553465,
+      "learning_rate": 2.8227484039513574e-06,
+      "loss": 1.0933,
+      "step": 8932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.027657230666876,
+      "learning_rate": 2.820824182431521e-06,
+      "loss": 1.1077,
+      "step": 8933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.661577609554698,
+      "learning_rate": 2.8189005092929487e-06,
+      "loss": 1.0654,
+      "step": 8934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8900146013882244,
+      "learning_rate": 2.8169773846825887e-06,
+      "loss": 1.0547,
+      "step": 8935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3697616317185344,
+      "learning_rate": 2.815054808747334e-06,
+      "loss": 1.0325,
+      "step": 8936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.729144594471382,
+      "learning_rate": 2.8131327816340427e-06,
+      "loss": 1.0942,
+      "step": 8937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2246012869536496,
+      "learning_rate": 2.8112113034895273e-06,
+      "loss": 1.0831,
+      "step": 8938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.69062357986481,
+      "learning_rate": 2.809290374460556e-06,
+      "loss": 1.1077,
+      "step": 8939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.003296897657613,
+      "learning_rate": 2.807369994693868e-06,
+      "loss": 1.0579,
+      "step": 8940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.092741525155346,
+      "learning_rate": 2.805450164336145e-06,
+      "loss": 1.0225,
+      "step": 8941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8503311351437857,
+      "learning_rate": 2.8035308835340325e-06,
+      "loss": 1.0166,
+      "step": 8942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7745451493397653,
+      "learning_rate": 2.8016121524341413e-06,
+      "loss": 1.0828,
+      "step": 8943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7171159910102394,
+      "learning_rate": 2.799693971183027e-06,
+      "loss": 1.0483,
+      "step": 8944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.954708430479849,
+      "learning_rate": 2.797776339927214e-06,
+      "loss": 1.0496,
+      "step": 8945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5574842597542586,
+      "learning_rate": 2.7958592588131796e-06,
+      "loss": 1.1062,
+      "step": 8946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.028615817026909,
+      "learning_rate": 2.7939427279873543e-06,
+      "loss": 1.0067,
+      "step": 8947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.935798272740521,
+      "learning_rate": 2.7920267475961416e-06,
+      "loss": 1.0576,
+      "step": 8948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6042156261951566,
+      "learning_rate": 2.7901113177858887e-06,
+      "loss": 1.0664,
+      "step": 8949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.94470606750144,
+      "learning_rate": 2.7881964387029038e-06,
+      "loss": 1.083,
+      "step": 8950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.870968876121259,
+      "learning_rate": 2.7862821104934624e-06,
+      "loss": 0.9731,
+      "step": 8951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8263455544290967,
+      "learning_rate": 2.784368333303781e-06,
+      "loss": 1.0974,
+      "step": 8952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8156491194752014,
+      "learning_rate": 2.7824551072800498e-06,
+      "loss": 1.0569,
+      "step": 8953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3361832786553247,
+      "learning_rate": 2.7805424325684083e-06,
+      "loss": 1.0774,
+      "step": 8954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.012424115054057,
+      "learning_rate": 2.7786303093149536e-06,
+      "loss": 1.1353,
+      "step": 8955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3114023371297305,
+      "learning_rate": 2.7767187376657502e-06,
+      "loss": 1.1641,
+      "step": 8956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.818877677998795,
+      "learning_rate": 2.774807717766809e-06,
+      "loss": 1.0823,
+      "step": 8957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9581190118576934,
+      "learning_rate": 2.7728972497641027e-06,
+      "loss": 1.0591,
+      "step": 8958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7525466716438958,
+      "learning_rate": 2.7709873338035644e-06,
+      "loss": 1.0984,
+      "step": 8959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.390647448621494,
+      "learning_rate": 2.769077970031078e-06,
+      "loss": 1.0977,
+      "step": 8960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.584713239771307,
+      "learning_rate": 2.7671691585924975e-06,
+      "loss": 1.064,
+      "step": 8961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1595508386898805,
+      "learning_rate": 2.765260899633624e-06,
+      "loss": 1.1077,
+      "step": 8962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1063962314629414,
+      "learning_rate": 2.7633531933002177e-06,
+      "loss": 1.0859,
+      "step": 8963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.19076693322607,
+      "learning_rate": 2.761446039738006e-06,
+      "loss": 1.166,
+      "step": 8964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.311396991117804,
+      "learning_rate": 2.7595394390926557e-06,
+      "loss": 1.0376,
+      "step": 8965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1249617224822623,
+      "learning_rate": 2.757633391509812e-06,
+      "loss": 1.0107,
+      "step": 8966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.925651047958283,
+      "learning_rate": 2.755727897135062e-06,
+      "loss": 1.053,
+      "step": 8967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3953088997513707,
+      "learning_rate": 2.753822956113962e-06,
+      "loss": 1.0898,
+      "step": 8968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8136121212577505,
+      "learning_rate": 2.751918568592019e-06,
+      "loss": 1.0448,
+      "step": 8969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3249676613175065,
+      "learning_rate": 2.7500147347146966e-06,
+      "loss": 1.0847,
+      "step": 8970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7675573060330168,
+      "learning_rate": 2.748111454627428e-06,
+      "loss": 1.0881,
+      "step": 8971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.133877727770478,
+      "learning_rate": 2.7462087284755823e-06,
+      "loss": 1.127,
+      "step": 8972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.956009777037951,
+      "learning_rate": 2.744306556404508e-06,
+      "loss": 1.0544,
+      "step": 8973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.522109812305558,
+      "learning_rate": 2.742404938559502e-06,
+      "loss": 1.0779,
+      "step": 8974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2116236838155365,
+      "learning_rate": 2.740503875085814e-06,
+      "loss": 1.1038,
+      "step": 8975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8252096256548875,
+      "learning_rate": 2.738603366128664e-06,
+      "loss": 0.9829,
+      "step": 8976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.102131259705293,
+      "learning_rate": 2.736703411833218e-06,
+      "loss": 1.0872,
+      "step": 8977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.189158552615439,
+      "learning_rate": 2.734804012344605e-06,
+      "loss": 1.0891,
+      "step": 8978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0039230938068395,
+      "learning_rate": 2.732905167807911e-06,
+      "loss": 1.0691,
+      "step": 8979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3148252646924043,
+      "learning_rate": 2.7310068783681755e-06,
+      "loss": 1.0779,
+      "step": 8980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.365048343727127,
+      "learning_rate": 2.7291091441704052e-06,
+      "loss": 1.0862,
+      "step": 8981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.86036856627693,
+      "learning_rate": 2.727211965359556e-06,
+      "loss": 1.0231,
+      "step": 8982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.957374478899297,
+      "learning_rate": 2.72531534208054e-06,
+      "loss": 1.0833,
+      "step": 8983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5963289124185436,
+      "learning_rate": 2.723419274478242e-06,
+      "loss": 1.0183,
+      "step": 8984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4784372561578234,
+      "learning_rate": 2.7215237626974775e-06,
+      "loss": 1.0811,
+      "step": 8985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.396150472492864,
+      "learning_rate": 2.719628806883048e-06,
+      "loss": 1.0422,
+      "step": 8986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1594147222576505,
+      "learning_rate": 2.717734407179694e-06,
+      "loss": 1.0935,
+      "step": 8987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.931401614529926,
+      "learning_rate": 2.715840563732116e-06,
+      "loss": 1.0671,
+      "step": 8988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8109076391336214,
+      "learning_rate": 2.713947276684984e-06,
+      "loss": 1.0405,
+      "step": 8989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.029067158366838,
+      "learning_rate": 2.7120545461829107e-06,
+      "loss": 1.03,
+      "step": 8990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7011830421571177,
+      "learning_rate": 2.7101623723704738e-06,
+      "loss": 1.0896,
+      "step": 8991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3962416425713964,
+      "learning_rate": 2.708270755392207e-06,
+      "loss": 1.0439,
+      "step": 8992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7023442304799086,
+      "learning_rate": 2.706379695392599e-06,
+      "loss": 1.0566,
+      "step": 8993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9900903999067427,
+      "learning_rate": 2.704489192516102e-06,
+      "loss": 1.0784,
+      "step": 8994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.05223537004412,
+      "learning_rate": 2.7025992469071217e-06,
+      "loss": 1.0549,
+      "step": 8995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.261661761322821,
+      "learning_rate": 2.700709858710018e-06,
+      "loss": 1.073,
+      "step": 8996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.172317627488315,
+      "learning_rate": 2.6988210280691187e-06,
+      "loss": 1.0881,
+      "step": 8997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8001908888151203,
+      "learning_rate": 2.696932755128693e-06,
+      "loss": 1.0762,
+      "step": 8998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7402192728644073,
+      "learning_rate": 2.695045040032983e-06,
+      "loss": 1.0732,
+      "step": 8999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8678513709072613,
+      "learning_rate": 2.6931578829261816e-06,
+      "loss": 1.1431,
+      "step": 9000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0632888893327945,
+      "learning_rate": 2.6912712839524336e-06,
+      "loss": 1.0791,
+      "step": 9001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.248032356380006,
+      "learning_rate": 2.6893852432558543e-06,
+      "loss": 1.1243,
+      "step": 9002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3346216125499226,
+      "learning_rate": 2.687499760980503e-06,
+      "loss": 1.0879,
+      "step": 9003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.060657096343565,
+      "learning_rate": 2.6856148372704107e-06,
+      "loss": 1.0112,
+      "step": 9004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9834202567393673,
+      "learning_rate": 2.6837304722695456e-06,
+      "loss": 1.0935,
+      "step": 9005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8846598161386985,
+      "learning_rate": 2.6818466661218533e-06,
+      "loss": 1.0491,
+      "step": 9006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2423353065987937,
+      "learning_rate": 2.6799634189712264e-06,
+      "loss": 1.0625,
+      "step": 9007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6605613999690676,
+      "learning_rate": 2.6780807309615133e-06,
+      "loss": 0.9807,
+      "step": 9008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.811185299080327,
+      "learning_rate": 2.6761986022365273e-06,
+      "loss": 1.0713,
+      "step": 9009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0156921173258575,
+      "learning_rate": 2.674317032940035e-06,
+      "loss": 1.0132,
+      "step": 9010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7236979359060016,
+      "learning_rate": 2.6724360232157574e-06,
+      "loss": 1.1162,
+      "step": 9011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.598981917359104,
+      "learning_rate": 2.670555573207376e-06,
+      "loss": 1.1223,
+      "step": 9012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.799829421551187,
+      "learning_rate": 2.6686756830585282e-06,
+      "loss": 1.0276,
+      "step": 9013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0307178259059278,
+      "learning_rate": 2.6667963529128118e-06,
+      "loss": 1.0361,
+      "step": 9014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.414772805995757,
+      "learning_rate": 2.6649175829137797e-06,
+      "loss": 1.0957,
+      "step": 9015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7772303954279303,
+      "learning_rate": 2.663039373204935e-06,
+      "loss": 1.0994,
+      "step": 9016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1746790851368405,
+      "learning_rate": 2.661161723929758e-06,
+      "loss": 1.0674,
+      "step": 9017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.02128748993571,
+      "learning_rate": 2.659284635231658e-06,
+      "loss": 1.0872,
+      "step": 9018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3125246556520604,
+      "learning_rate": 2.657408107254027e-06,
+      "loss": 1.0681,
+      "step": 9019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.118142503501991,
+      "learning_rate": 2.655532140140199e-06,
+      "loss": 1.0664,
+      "step": 9020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8841404386306482,
+      "learning_rate": 2.653656734033467e-06,
+      "loss": 0.9778,
+      "step": 9021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8819027448699335,
+      "learning_rate": 2.651781889077092e-06,
+      "loss": 1.1152,
+      "step": 9022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.188951828573311,
+      "learning_rate": 2.649907605414278e-06,
+      "loss": 1.075,
+      "step": 9023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3196159252002384,
+      "learning_rate": 2.6480338831881947e-06,
+      "loss": 1.0723,
+      "step": 9024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.920776276152595,
+      "learning_rate": 2.646160722541964e-06,
+      "loss": 1.0913,
+      "step": 9025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.667694423103628,
+      "learning_rate": 2.644288123618667e-06,
+      "loss": 1.0137,
+      "step": 9026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.805042842687805,
+      "learning_rate": 2.6424160865613457e-06,
+      "loss": 1.0852,
+      "step": 9027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.954566359564533,
+      "learning_rate": 2.640544611512993e-06,
+      "loss": 1.03,
+      "step": 9028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1496926802476586,
+      "learning_rate": 2.6386736986165594e-06,
+      "loss": 1.1101,
+      "step": 9029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6001505133223994,
+      "learning_rate": 2.6368033480149647e-06,
+      "loss": 1.1394,
+      "step": 9030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.441988865380437,
+      "learning_rate": 2.6349335598510604e-06,
+      "loss": 1.074,
+      "step": 9031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1024477769343854,
+      "learning_rate": 2.633064334267682e-06,
+      "loss": 1.0918,
+      "step": 9032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3144244067506836,
+      "learning_rate": 2.631195671407605e-06,
+      "loss": 1.053,
+      "step": 9033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9413927250180087,
+      "learning_rate": 2.629327571413567e-06,
+      "loss": 1.134,
+      "step": 9034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.040120205070497,
+      "learning_rate": 2.6274600344282665e-06,
+      "loss": 1.0974,
+      "step": 9035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7882206326431795,
+      "learning_rate": 2.625593060594349e-06,
+      "loss": 1.0446,
+      "step": 9036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.770753811137741,
+      "learning_rate": 2.6237266500544344e-06,
+      "loss": 1.0337,
+      "step": 9037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9856513284356243,
+      "learning_rate": 2.6218608029510784e-06,
+      "loss": 1.1021,
+      "step": 9038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5607764482118456,
+      "learning_rate": 2.6199955194268045e-06,
+      "loss": 1.1208,
+      "step": 9039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9946215027861154,
+      "learning_rate": 2.618130799624097e-06,
+      "loss": 1.0522,
+      "step": 9040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7959688501277626,
+      "learning_rate": 2.6162666436853857e-06,
+      "loss": 0.9919,
+      "step": 9041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.105157173243868,
+      "learning_rate": 2.614403051753073e-06,
+      "loss": 1.0645,
+      "step": 9042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.977803551339383,
+      "learning_rate": 2.612540023969504e-06,
+      "loss": 1.082,
+      "step": 9043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.319376099790883,
+      "learning_rate": 2.6106775604769873e-06,
+      "loss": 1.0476,
+      "step": 9044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1513819844891797,
+      "learning_rate": 2.608815661417786e-06,
+      "loss": 1.0637,
+      "step": 9045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8785582375017693,
+      "learning_rate": 2.60695432693412e-06,
+      "loss": 0.9822,
+      "step": 9046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3502119010379388,
+      "learning_rate": 2.6050935571681725e-06,
+      "loss": 1.0818,
+      "step": 9047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3200767816290466,
+      "learning_rate": 2.6032333522620747e-06,
+      "loss": 1.1096,
+      "step": 9048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8115995553684963,
+      "learning_rate": 2.601373712357915e-06,
+      "loss": 1.0972,
+      "step": 9049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.502540504545888,
+      "learning_rate": 2.599514637597752e-06,
+      "loss": 1.0999,
+      "step": 9050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.192465549092931,
+      "learning_rate": 2.5976561281235802e-06,
+      "loss": 1.0815,
+      "step": 9051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.723794927607417,
+      "learning_rate": 2.5957981840773684e-06,
+      "loss": 1.0781,
+      "step": 9052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.844039099681203,
+      "learning_rate": 2.593940805601034e-06,
+      "loss": 1.0601,
+      "step": 9053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1024045370220774,
+      "learning_rate": 2.5920839928364494e-06,
+      "loss": 1.0566,
+      "step": 9054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8889718648740548,
+      "learning_rate": 2.5902277459254533e-06,
+      "loss": 1.0767,
+      "step": 9055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5957199388449697,
+      "learning_rate": 2.588372065009832e-06,
+      "loss": 1.0813,
+      "step": 9056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0303740916498993,
+      "learning_rate": 2.5865169502313323e-06,
+      "loss": 1.0012,
+      "step": 9057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.356943038510994,
+      "learning_rate": 2.5846624017316568e-06,
+      "loss": 1.0928,
+      "step": 9058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3462003953802673,
+      "learning_rate": 2.582808419652463e-06,
+      "loss": 1.0806,
+      "step": 9059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.730420433787575,
+      "learning_rate": 2.5809550041353705e-06,
+      "loss": 1.0583,
+      "step": 9060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.589639162869082,
+      "learning_rate": 2.579102155321952e-06,
+      "loss": 1.0356,
+      "step": 9061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2012889250360477,
+      "learning_rate": 2.577249873353734e-06,
+      "loss": 1.0364,
+      "step": 9062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0187429567157626,
+      "learning_rate": 2.575398158372212e-06,
+      "loss": 1.0796,
+      "step": 9063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2552012610517607,
+      "learning_rate": 2.5735470105188166e-06,
+      "loss": 1.041,
+      "step": 9064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9952994458454354,
+      "learning_rate": 2.5716964299349558e-06,
+      "loss": 1.0376,
+      "step": 9065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.273299150285956,
+      "learning_rate": 2.569846416761985e-06,
+      "loss": 1.0808,
+      "step": 9066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7109753674939165,
+      "learning_rate": 2.5679969711412143e-06,
+      "loss": 1.0864,
+      "step": 9067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.990456416755124,
+      "learning_rate": 2.566148093213919e-06,
+      "loss": 1.0891,
+      "step": 9068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4776145647803416,
+      "learning_rate": 2.56429978312132e-06,
+      "loss": 1.1306,
+      "step": 9069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.24607385357202,
+      "learning_rate": 2.5624520410046083e-06,
+      "loss": 1.1631,
+      "step": 9070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.252301329508585,
+      "learning_rate": 2.5606048670049167e-06,
+      "loss": 1.0073,
+      "step": 9071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2376572770789087,
+      "learning_rate": 2.55875826126334e-06,
+      "loss": 1.0562,
+      "step": 9072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6974899243602954,
+      "learning_rate": 2.5569122239209366e-06,
+      "loss": 1.0364,
+      "step": 9073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.302079422648683,
+      "learning_rate": 2.5550667551187114e-06,
+      "loss": 1.0837,
+      "step": 9074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.497128957386113,
+      "learning_rate": 2.5532218549976374e-06,
+      "loss": 1.048,
+      "step": 9075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2288087768296196,
+      "learning_rate": 2.5513775236986316e-06,
+      "loss": 1.1436,
+      "step": 9076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0565067540733986,
+      "learning_rate": 2.549533761362575e-06,
+      "loss": 0.9937,
+      "step": 9077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.341290387064024,
+      "learning_rate": 2.547690568130303e-06,
+      "loss": 1.0762,
+      "step": 9078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.146170781168245,
+      "learning_rate": 2.5458479441426045e-06,
+      "loss": 1.053,
+      "step": 9079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.396171151207823,
+      "learning_rate": 2.5440058895402353e-06,
+      "loss": 1.0332,
+      "step": 9080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2033487737517103,
+      "learning_rate": 2.542164404463897e-06,
+      "loss": 1.0857,
+      "step": 9081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.583677009265645,
+      "learning_rate": 2.5403234890542483e-06,
+      "loss": 1.0024,
+      "step": 9082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9486235978862565,
+      "learning_rate": 2.5384831434519165e-06,
+      "loss": 0.9695,
+      "step": 9083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0540145316570526,
+      "learning_rate": 2.5366433677974657e-06,
+      "loss": 1.0813,
+      "step": 9084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.141487833000511,
+      "learning_rate": 2.5348041622314335e-06,
+      "loss": 1.0537,
+      "step": 9085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.43145998207653,
+      "learning_rate": 2.5329655268943078e-06,
+      "loss": 1.0518,
+      "step": 9086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3326385069488627,
+      "learning_rate": 2.5311274619265268e-06,
+      "loss": 1.0781,
+      "step": 9087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.534887266858863,
+      "learning_rate": 2.529289967468499e-06,
+      "loss": 1.0393,
+      "step": 9088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.203501012657976,
+      "learning_rate": 2.5274530436605772e-06,
+      "loss": 1.0056,
+      "step": 9089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5854175725457473,
+      "learning_rate": 2.5256166906430757e-06,
+      "loss": 1.1157,
+      "step": 9090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9721786131589094,
+      "learning_rate": 2.523780908556265e-06,
+      "loss": 1.1418,
+      "step": 9091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0185696702694536,
+      "learning_rate": 2.5219456975403655e-06,
+      "loss": 1.0796,
+      "step": 9092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.558343715498941,
+      "learning_rate": 2.520111057735568e-06,
+      "loss": 0.98,
+      "step": 9093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.791504127361642,
+      "learning_rate": 2.518276989282009e-06,
+      "loss": 1.0459,
+      "step": 9094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2022403812616336,
+      "learning_rate": 2.516443492319779e-06,
+      "loss": 1.0964,
+      "step": 9095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.434445972214771,
+      "learning_rate": 2.5146105669889364e-06,
+      "loss": 1.0828,
+      "step": 9096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3262080116683355,
+      "learning_rate": 2.5127782134294863e-06,
+      "loss": 1.0713,
+      "step": 9097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2306507641239715,
+      "learning_rate": 2.5109464317813938e-06,
+      "loss": 1.0769,
+      "step": 9098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3128951282027006,
+      "learning_rate": 2.509115222184577e-06,
+      "loss": 1.0784,
+      "step": 9099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.578579156512927,
+      "learning_rate": 2.507284584778913e-06,
+      "loss": 0.9512,
+      "step": 9100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.128685638437492,
+      "learning_rate": 2.505454519704239e-06,
+      "loss": 1.0366,
+      "step": 9101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.104547222180025,
+      "learning_rate": 2.5036250271003415e-06,
+      "loss": 1.0757,
+      "step": 9102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7774652594798455,
+      "learning_rate": 2.501796107106966e-06,
+      "loss": 1.0237,
+      "step": 9103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0967599131472636,
+      "learning_rate": 2.499967759863817e-06,
+      "loss": 1.0972,
+      "step": 9104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.383443475941103,
+      "learning_rate": 2.498139985510548e-06,
+      "loss": 1.1045,
+      "step": 9105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.257543075435956,
+      "learning_rate": 2.4963127841867783e-06,
+      "loss": 1.0535,
+      "step": 9106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5163408402317686,
+      "learning_rate": 2.4944861560320756e-06,
+      "loss": 1.0662,
+      "step": 9107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1352915015266487,
+      "learning_rate": 2.4926601011859698e-06,
+      "loss": 1.0649,
+      "step": 9108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3966535652367638,
+      "learning_rate": 2.490834619787943e-06,
+      "loss": 1.0295,
+      "step": 9109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4018012036052308,
+      "learning_rate": 2.489009711977435e-06,
+      "loss": 1.0579,
+      "step": 9110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7418281584301605,
+      "learning_rate": 2.4871853778938394e-06,
+      "loss": 1.02,
+      "step": 9111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0036871059561743,
+      "learning_rate": 2.485361617676506e-06,
+      "loss": 1.0325,
+      "step": 9112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.449598847654908,
+      "learning_rate": 2.483538431464748e-06,
+      "loss": 1.0959,
+      "step": 9113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.265013977545894,
+      "learning_rate": 2.4817158193978287e-06,
+      "loss": 1.0398,
+      "step": 9114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.299348561868526,
+      "learning_rate": 2.479893781614963e-06,
+      "loss": 1.0828,
+      "step": 9115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.250229074289684,
+      "learning_rate": 2.4780723182553366e-06,
+      "loss": 1.1267,
+      "step": 9116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2122081671668616,
+      "learning_rate": 2.476251429458072e-06,
+      "loss": 1.0684,
+      "step": 9117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4333990100068754,
+      "learning_rate": 2.4744311153622636e-06,
+      "loss": 1.1665,
+      "step": 9118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3043072299375633,
+      "learning_rate": 2.472611376106955e-06,
+      "loss": 1.0442,
+      "step": 9119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2192113749517928,
+      "learning_rate": 2.4707922118311434e-06,
+      "loss": 1.0068,
+      "step": 9120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.222145559132112,
+      "learning_rate": 2.4689736226737916e-06,
+      "loss": 1.0649,
+      "step": 9121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2412390266642603,
+      "learning_rate": 2.46715560877381e-06,
+      "loss": 1.0693,
+      "step": 9122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6098999672749863,
+      "learning_rate": 2.4653381702700672e-06,
+      "loss": 1.1628,
+      "step": 9123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.297441452243098,
+      "learning_rate": 2.4635213073013884e-06,
+      "loss": 1.0762,
+      "step": 9124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.402210799071134,
+      "learning_rate": 2.4617050200065527e-06,
+      "loss": 1.0898,
+      "step": 9125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5121629756087938,
+      "learning_rate": 2.4598893085243004e-06,
+      "loss": 1.1028,
+      "step": 9126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6100420924971592,
+      "learning_rate": 2.4580741729933246e-06,
+      "loss": 1.1021,
+      "step": 9127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7806940296565474,
+      "learning_rate": 2.456259613552269e-06,
+      "loss": 1.1086,
+      "step": 9128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4603895063403627,
+      "learning_rate": 2.4544456303397456e-06,
+      "loss": 1.1074,
+      "step": 9129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9315809693166157,
+      "learning_rate": 2.4526322234943133e-06,
+      "loss": 1.0632,
+      "step": 9130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.603938949538638,
+      "learning_rate": 2.450819393154489e-06,
+      "loss": 1.0344,
+      "step": 9131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.449235766780346,
+      "learning_rate": 2.4490071394587445e-06,
+      "loss": 1.0908,
+      "step": 9132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6925511802276203,
+      "learning_rate": 2.4471954625455065e-06,
+      "loss": 1.1355,
+      "step": 9133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.96523206194628,
+      "learning_rate": 2.445384362553166e-06,
+      "loss": 1.0911,
+      "step": 9134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1783767728907444,
+      "learning_rate": 2.4435738396200614e-06,
+      "loss": 1.0154,
+      "step": 9135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0575338848478535,
+      "learning_rate": 2.4417638938844877e-06,
+      "loss": 1.0344,
+      "step": 9136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2947042853524744,
+      "learning_rate": 2.4399545254847002e-06,
+      "loss": 1.043,
+      "step": 9137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.011343428557653,
+      "learning_rate": 2.438145734558902e-06,
+      "loss": 1.1152,
+      "step": 9138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.280903420746888,
+      "learning_rate": 2.4363375212452655e-06,
+      "loss": 1.0984,
+      "step": 9139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7200310632032596,
+      "learning_rate": 2.4345298856819065e-06,
+      "loss": 1.0786,
+      "step": 9140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.011574815143168,
+      "learning_rate": 2.4327228280068993e-06,
+      "loss": 1.0557,
+      "step": 9141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.18430904020278,
+      "learning_rate": 2.430916348358281e-06,
+      "loss": 1.0508,
+      "step": 9142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9106916727127485,
+      "learning_rate": 2.4291104468740378e-06,
+      "loss": 1.0181,
+      "step": 9143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2032454293569814,
+      "learning_rate": 2.427305123692112e-06,
+      "loss": 1.062,
+      "step": 9144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.687802012070587,
+      "learning_rate": 2.425500378950405e-06,
+      "loss": 1.0947,
+      "step": 9145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.181842629449233,
+      "learning_rate": 2.423696212786768e-06,
+      "loss": 1.0271,
+      "step": 9146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.853406418062478,
+      "learning_rate": 2.421892625339017e-06,
+      "loss": 1.0562,
+      "step": 9147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3528010940713298,
+      "learning_rate": 2.4200896167449163e-06,
+      "loss": 1.1162,
+      "step": 9148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.033623778582385,
+      "learning_rate": 2.418287187142192e-06,
+      "loss": 1.0347,
+      "step": 9149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8462671719135275,
+      "learning_rate": 2.4164853366685206e-06,
+      "loss": 0.9712,
+      "step": 9150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0191155978953548,
+      "learning_rate": 2.4146840654615356e-06,
+      "loss": 1.0471,
+      "step": 9151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.749717527124209,
+      "learning_rate": 2.4128833736588285e-06,
+      "loss": 1.0867,
+      "step": 9152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3691653481209256,
+      "learning_rate": 2.4110832613979406e-06,
+      "loss": 1.1274,
+      "step": 9153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.678525005233725,
+      "learning_rate": 2.4092837288163807e-06,
+      "loss": 1.1128,
+      "step": 9154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6801031285048875,
+      "learning_rate": 2.407484776051603e-06,
+      "loss": 1.0508,
+      "step": 9155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4201420872951562,
+      "learning_rate": 2.40568640324102e-06,
+      "loss": 1.0374,
+      "step": 9156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.940780273354187,
+      "learning_rate": 2.4038886105220005e-06,
+      "loss": 1.1125,
+      "step": 9157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4312912599490097,
+      "learning_rate": 2.402091398031866e-06,
+      "loss": 1.1184,
+      "step": 9158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6659299793969873,
+      "learning_rate": 2.4002947659079025e-06,
+      "loss": 1.0515,
+      "step": 9159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.644613474606261,
+      "learning_rate": 2.3984987142873417e-06,
+      "loss": 1.0793,
+      "step": 9160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0182307435751023,
+      "learning_rate": 2.3967032433073733e-06,
+      "loss": 1.0925,
+      "step": 9161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.826113256704669,
+      "learning_rate": 2.394908353105151e-06,
+      "loss": 1.0415,
+      "step": 9162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.375804742816557,
+      "learning_rate": 2.3931140438177724e-06,
+      "loss": 1.1123,
+      "step": 9163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6162886040130364,
+      "learning_rate": 2.3913203155822974e-06,
+      "loss": 1.0547,
+      "step": 9164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.212201356652945,
+      "learning_rate": 2.389527168535739e-06,
+      "loss": 1.083,
+      "step": 9165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2522286816716117,
+      "learning_rate": 2.3877346028150638e-06,
+      "loss": 1.0632,
+      "step": 9166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8421695133092806,
+      "learning_rate": 2.3859426185572042e-06,
+      "loss": 1.0747,
+      "step": 9167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4190132542418565,
+      "learning_rate": 2.3841512158990367e-06,
+      "loss": 1.0308,
+      "step": 9168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.302884972729658,
+      "learning_rate": 2.3823603949773942e-06,
+      "loss": 1.0691,
+      "step": 9169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.859643407311754,
+      "learning_rate": 2.3805701559290782e-06,
+      "loss": 1.0642,
+      "step": 9170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1087947551461634,
+      "learning_rate": 2.3787804988908235e-06,
+      "loss": 1.0554,
+      "step": 9171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.28601935801554,
+      "learning_rate": 2.3769914239993432e-06,
+      "loss": 1.0862,
+      "step": 9172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8628278971230836,
+      "learning_rate": 2.375202931391292e-06,
+      "loss": 1.063,
+      "step": 9173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.211520664135244,
+      "learning_rate": 2.3734150212032813e-06,
+      "loss": 1.0718,
+      "step": 9174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.239982701266157,
+      "learning_rate": 2.3716276935718864e-06,
+      "loss": 1.0803,
+      "step": 9175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.424216707484341,
+      "learning_rate": 2.369840948633628e-06,
+      "loss": 1.0605,
+      "step": 9176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.271089915151355,
+      "learning_rate": 2.3680547865249892e-06,
+      "loss": 1.0918,
+      "step": 9177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.540719245481102,
+      "learning_rate": 2.3662692073824035e-06,
+      "loss": 1.0442,
+      "step": 9178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1723450903409867,
+      "learning_rate": 2.3644842113422606e-06,
+      "loss": 1.0908,
+      "step": 9179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.764449276894458,
+      "learning_rate": 2.3626997985409137e-06,
+      "loss": 1.0537,
+      "step": 9180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0004006291523027,
+      "learning_rate": 2.3609159691146577e-06,
+      "loss": 1.0518,
+      "step": 9181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0197508104156503,
+      "learning_rate": 2.3591327231997585e-06,
+      "loss": 1.0388,
+      "step": 9182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.029132896839047,
+      "learning_rate": 2.357350060932427e-06,
+      "loss": 1.0874,
+      "step": 9183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.626602100015791,
+      "learning_rate": 2.3555679824488254e-06,
+      "loss": 1.147,
+      "step": 9184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6166849230703546,
+      "learning_rate": 2.3537864878850847e-06,
+      "loss": 1.106,
+      "step": 9185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4333866557230475,
+      "learning_rate": 2.352005577377279e-06,
+      "loss": 1.062,
+      "step": 9186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1331077677929926,
+      "learning_rate": 2.3502252510614486e-06,
+      "loss": 1.0735,
+      "step": 9187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.119349772739284,
+      "learning_rate": 2.3484455090735814e-06,
+      "loss": 1.0557,
+      "step": 9188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.654611759366252,
+      "learning_rate": 2.3466663515496222e-06,
+      "loss": 1.0576,
+      "step": 9189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.435549142483282,
+      "learning_rate": 2.3448877786254742e-06,
+      "loss": 1.1191,
+      "step": 9190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.159718289665059,
+      "learning_rate": 2.3431097904369883e-06,
+      "loss": 1.0232,
+      "step": 9191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.223673686184712,
+      "learning_rate": 2.3413323871199824e-06,
+      "loss": 1.0947,
+      "step": 9192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0828538846560276,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 1.0796,
+      "step": 9193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9359054406721468,
+      "learning_rate": 2.337779335643423e-06,
+      "loss": 1.1116,
+      "step": 9194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9402203403794,
+      "learning_rate": 2.3360036877552727e-06,
+      "loss": 1.0225,
+      "step": 9195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5443006754125013,
+      "learning_rate": 2.3342286252813996e-06,
+      "loss": 1.0498,
+      "step": 9196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.42639851928728,
+      "learning_rate": 2.3324541483573905e-06,
+      "loss": 1.1067,
+      "step": 9197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7404983102436415,
+      "learning_rate": 2.3306802571187903e-06,
+      "loss": 1.0618,
+      "step": 9198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.576800183689574,
+      "learning_rate": 2.3289069517010953e-06,
+      "loss": 1.0933,
+      "step": 9199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.524229241768106,
+      "learning_rate": 2.327134232239764e-06,
+      "loss": 1.113,
+      "step": 9200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.685153311587139,
+      "learning_rate": 2.325362098870202e-06,
+      "loss": 1.0559,
+      "step": 9201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9886812222371906,
+      "learning_rate": 2.323590551727772e-06,
+      "loss": 1.0254,
+      "step": 9202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.961836162503681,
+      "learning_rate": 2.3218195909478024e-06,
+      "loss": 1.1301,
+      "step": 9203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6316924309044394,
+      "learning_rate": 2.320049216665556e-06,
+      "loss": 1.0327,
+      "step": 9204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.268209841868526,
+      "learning_rate": 2.3182794290162725e-06,
+      "loss": 1.155,
+      "step": 9205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.596452547369439,
+      "learning_rate": 2.3165102281351314e-06,
+      "loss": 1.1233,
+      "step": 9206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9683092321843607,
+      "learning_rate": 2.314741614157274e-06,
+      "loss": 1.0518,
+      "step": 9207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7895971445770917,
+      "learning_rate": 2.312973587217798e-06,
+      "loss": 1.0066,
+      "step": 9208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0264353632902803,
+      "learning_rate": 2.3112061474517544e-06,
+      "loss": 1.0042,
+      "step": 9209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.292959157527016,
+      "learning_rate": 2.3094392949941468e-06,
+      "loss": 1.0366,
+      "step": 9210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1208549006477746,
+      "learning_rate": 2.3076730299799378e-06,
+      "loss": 1.0681,
+      "step": 9211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1595343274283154,
+      "learning_rate": 2.30590735254404e-06,
+      "loss": 0.9653,
+      "step": 9212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.019945771457971,
+      "learning_rate": 2.3041422628213296e-06,
+      "loss": 1.0906,
+      "step": 9213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.383031555669699,
+      "learning_rate": 2.302377760946628e-06,
+      "loss": 1.0083,
+      "step": 9214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9132723750227076,
+      "learning_rate": 2.3006138470547225e-06,
+      "loss": 1.0693,
+      "step": 9215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3006694455933974,
+      "learning_rate": 2.298850521280349e-06,
+      "loss": 1.0916,
+      "step": 9216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9069286705781487,
+      "learning_rate": 2.2970877837581917e-06,
+      "loss": 1.1279,
+      "step": 9217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.848233232534447,
+      "learning_rate": 2.2953256346229045e-06,
+      "loss": 1.0715,
+      "step": 9218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.125134582601767,
+      "learning_rate": 2.2935640740090847e-06,
+      "loss": 1.0459,
+      "step": 9219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.351839430356399,
+      "learning_rate": 2.2918031020512945e-06,
+      "loss": 1.1016,
+      "step": 9220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6146760319535445,
+      "learning_rate": 2.2900427188840413e-06,
+      "loss": 1.1226,
+      "step": 9221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.516384164249896,
+      "learning_rate": 2.2882829246417914e-06,
+      "loss": 1.094,
+      "step": 9222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9703182054790442,
+      "learning_rate": 2.286523719458974e-06,
+      "loss": 1.1211,
+      "step": 9223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.278314162899973,
+      "learning_rate": 2.2847651034699536e-06,
+      "loss": 1.0579,
+      "step": 9224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0575022827392226,
+      "learning_rate": 2.283007076809073e-06,
+      "loss": 1.0586,
+      "step": 9225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.252641552864739,
+      "learning_rate": 2.2812496396106155e-06,
+      "loss": 1.074,
+      "step": 9226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2044628499733525,
+      "learning_rate": 2.2794927920088185e-06,
+      "loss": 1.0408,
+      "step": 9227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.311891627844527,
+      "learning_rate": 2.2777365341378855e-06,
+      "loss": 1.0549,
+      "step": 9228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.286943944750655,
+      "learning_rate": 2.275980866131966e-06,
+      "loss": 1.0872,
+      "step": 9229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1176023092999983,
+      "learning_rate": 2.274225788125165e-06,
+      "loss": 1.0305,
+      "step": 9230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9045907440603544,
+      "learning_rate": 2.2724713002515463e-06,
+      "loss": 1.0288,
+      "step": 9231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8767815658204565,
+      "learning_rate": 2.270717402645122e-06,
+      "loss": 1.0935,
+      "step": 9232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1744168548346385,
+      "learning_rate": 2.2689640954398694e-06,
+      "loss": 1.0356,
+      "step": 9233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.128792667976448,
+      "learning_rate": 2.2672113787697125e-06,
+      "loss": 1.0347,
+      "step": 9234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8262736096060594,
+      "learning_rate": 2.2654592527685305e-06,
+      "loss": 1.0417,
+      "step": 9235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.298533979744113,
+      "learning_rate": 2.2637077175701662e-06,
+      "loss": 1.0159,
+      "step": 9236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4749338977880777,
+      "learning_rate": 2.2619567733084004e-06,
+      "loss": 1.0684,
+      "step": 9237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5350888834641245,
+      "learning_rate": 2.260206420116987e-06,
+      "loss": 1.043,
+      "step": 9238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.246429172978768,
+      "learning_rate": 2.258456658129624e-06,
+      "loss": 1.041,
+      "step": 9239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.177671463598015,
+      "learning_rate": 2.2567074874799645e-06,
+      "loss": 1.0264,
+      "step": 9240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.291093073752909,
+      "learning_rate": 2.2549589083016242e-06,
+      "loss": 1.1025,
+      "step": 9241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9012786912084936,
+      "learning_rate": 2.253210920728166e-06,
+      "loss": 1.0508,
+      "step": 9242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.303890733480606,
+      "learning_rate": 2.2514635248931084e-06,
+      "loss": 1.0713,
+      "step": 9243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.932087958952815,
+      "learning_rate": 2.2497167209299277e-06,
+      "loss": 1.0103,
+      "step": 9244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.539053409424742,
+      "learning_rate": 2.2479705089720504e-06,
+      "loss": 1.0613,
+      "step": 9245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.853634959326173,
+      "learning_rate": 2.246224889152866e-06,
+      "loss": 1.075,
+      "step": 9246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4109028537491732,
+      "learning_rate": 2.2444798616057115e-06,
+      "loss": 1.031,
+      "step": 9247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.178818038228994,
+      "learning_rate": 2.242735426463877e-06,
+      "loss": 1.02,
+      "step": 9248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.20333307698169,
+      "learning_rate": 2.240991583860621e-06,
+      "loss": 1.0266,
+      "step": 9249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.021969698455936,
+      "learning_rate": 2.239248333929135e-06,
+      "loss": 1.0464,
+      "step": 9250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.71730034849008,
+      "learning_rate": 2.237505676802585e-06,
+      "loss": 1.1514,
+      "step": 9251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2743056396250014,
+      "learning_rate": 2.235763612614078e-06,
+      "loss": 1.0681,
+      "step": 9252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1791099272298777,
+      "learning_rate": 2.234022141496689e-06,
+      "loss": 1.0522,
+      "step": 9253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3365356450113515,
+      "learning_rate": 2.232281263583437e-06,
+      "loss": 1.04,
+      "step": 9254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6780589785227122,
+      "learning_rate": 2.230540979007294e-06,
+      "loss": 1.1257,
+      "step": 9255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6849584282797623,
+      "learning_rate": 2.2288012879012022e-06,
+      "loss": 1.0676,
+      "step": 9256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.67369949307278,
+      "learning_rate": 2.227062190398035e-06,
+      "loss": 1.0999,
+      "step": 9257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.384278285866164,
+      "learning_rate": 2.225323686630644e-06,
+      "loss": 1.0098,
+      "step": 9258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.378840475737665,
+      "learning_rate": 2.2235857767318202e-06,
+      "loss": 1.0771,
+      "step": 9259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1516236497740255,
+      "learning_rate": 2.2218484608343106e-06,
+      "loss": 1.0442,
+      "step": 9260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1967198611583405,
+      "learning_rate": 2.220111739070828e-06,
+      "loss": 1.0742,
+      "step": 9261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.017434376468563,
+      "learning_rate": 2.2183756115740274e-06,
+      "loss": 1.1252,
+      "step": 9262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4287867484878576,
+      "learning_rate": 2.216640078476523e-06,
+      "loss": 1.0986,
+      "step": 9263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.267583212860356,
+      "learning_rate": 2.2149051399108834e-06,
+      "loss": 1.0627,
+      "step": 9264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0613209467335585,
+      "learning_rate": 2.2131707960096293e-06,
+      "loss": 1.0881,
+      "step": 9265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.978508453204167,
+      "learning_rate": 2.2114370469052436e-06,
+      "loss": 1.0339,
+      "step": 9266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3845623812983465,
+      "learning_rate": 2.209703892730157e-06,
+      "loss": 1.0557,
+      "step": 9267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2215934396711927,
+      "learning_rate": 2.207971333616753e-06,
+      "loss": 1.0774,
+      "step": 9268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.031133258061052,
+      "learning_rate": 2.2062393696973814e-06,
+      "loss": 1.124,
+      "step": 9269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5939742702496895,
+      "learning_rate": 2.2045080011043286e-06,
+      "loss": 1.0999,
+      "step": 9270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.941981790760134,
+      "learning_rate": 2.202777227969851e-06,
+      "loss": 1.1111,
+      "step": 9271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1124948108966883,
+      "learning_rate": 2.2010470504261516e-06,
+      "loss": 1.1257,
+      "step": 9272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4628721493262025,
+      "learning_rate": 2.1993174686053885e-06,
+      "loss": 1.073,
+      "step": 9273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1945782220854744,
+      "learning_rate": 2.197588482639681e-06,
+      "loss": 1.0764,
+      "step": 9274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0419703674146144,
+      "learning_rate": 2.195860092661094e-06,
+      "loss": 1.0906,
+      "step": 9275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.045417893263624,
+      "learning_rate": 2.19413229880165e-06,
+      "loss": 1.0999,
+      "step": 9276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.447642971653353,
+      "learning_rate": 2.192405101193329e-06,
+      "loss": 1.106,
+      "step": 9277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.314776019349104,
+      "learning_rate": 2.190678499968059e-06,
+      "loss": 1.1008,
+      "step": 9278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4796933965243335,
+      "learning_rate": 2.18895249525773e-06,
+      "loss": 1.0906,
+      "step": 9279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5096861378399447,
+      "learning_rate": 2.1872270871941824e-06,
+      "loss": 1.0491,
+      "step": 9280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9958794665144746,
+      "learning_rate": 2.1855022759092094e-06,
+      "loss": 0.9619,
+      "step": 9281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1344858569085825,
+      "learning_rate": 2.183778061534566e-06,
+      "loss": 1.0178,
+      "step": 9282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.131850668269153,
+      "learning_rate": 2.1820544442019476e-06,
+      "loss": 1.0945,
+      "step": 9283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2437292454814473,
+      "learning_rate": 2.1803314240430197e-06,
+      "loss": 1.0776,
+      "step": 9284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0870181528937715,
+      "learning_rate": 2.1786090011893933e-06,
+      "loss": 1.0571,
+      "step": 9285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6931829943719654,
+      "learning_rate": 2.176887175772633e-06,
+      "loss": 1.1011,
+      "step": 9286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.871391835762942,
+      "learning_rate": 2.175165947924266e-06,
+      "loss": 1.0818,
+      "step": 9287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.232717327087811,
+      "learning_rate": 2.173445317775761e-06,
+      "loss": 1.1062,
+      "step": 9288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.083210827278831,
+      "learning_rate": 2.171725285458559e-06,
+      "loss": 1.0593,
+      "step": 9289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.246951904719318,
+      "learning_rate": 2.170005851104031e-06,
+      "loss": 1.0496,
+      "step": 9290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6668539070693815,
+      "learning_rate": 2.1682870148435277e-06,
+      "loss": 1.1079,
+      "step": 9291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.939739462222764,
+      "learning_rate": 2.1665687768083366e-06,
+      "loss": 1.0186,
+      "step": 9292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5721338863337775,
+      "learning_rate": 2.164851137129703e-06,
+      "loss": 1.0925,
+      "step": 9293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3456315417350933,
+      "learning_rate": 2.163134095938836e-06,
+      "loss": 1.1223,
+      "step": 9294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2673198565961727,
+      "learning_rate": 2.1614176533668874e-06,
+      "loss": 1.1577,
+      "step": 9295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9687813896790973,
+      "learning_rate": 2.1597018095449682e-06,
+      "loss": 1.0532,
+      "step": 9296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3637163644903403,
+      "learning_rate": 2.157986564604143e-06,
+      "loss": 1.0886,
+      "step": 9297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.101241773455336,
+      "learning_rate": 2.1562719186754287e-06,
+      "loss": 1.0762,
+      "step": 9298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9678302190205903,
+      "learning_rate": 2.154557871889803e-06,
+      "loss": 1.0662,
+      "step": 9299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1162379988998916,
+      "learning_rate": 2.152844424378191e-06,
+      "loss": 1.0605,
+      "step": 9300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1433314615683425,
+      "learning_rate": 2.1511315762714713e-06,
+      "loss": 0.958,
+      "step": 9301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8180751910054394,
+      "learning_rate": 2.1494193277004895e-06,
+      "loss": 1.0044,
+      "step": 9302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4775881351381894,
+      "learning_rate": 2.147707678796023e-06,
+      "loss": 1.0781,
+      "step": 9303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0602441213782283,
+      "learning_rate": 2.1459966296888235e-06,
+      "loss": 1.0671,
+      "step": 9304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.890770133827649,
+      "learning_rate": 2.1442861805095892e-06,
+      "loss": 1.0889,
+      "step": 9305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2443830935907143,
+      "learning_rate": 2.1425763313889692e-06,
+      "loss": 1.0383,
+      "step": 9306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5161033595003035,
+      "learning_rate": 2.140867082457575e-06,
+      "loss": 1.145,
+      "step": 9307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5153217319973806,
+      "learning_rate": 2.1391584338459658e-06,
+      "loss": 1.0977,
+      "step": 9308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.946273770796833,
+      "learning_rate": 2.137450385684656e-06,
+      "loss": 1.0469,
+      "step": 9309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2454985076854475,
+      "learning_rate": 2.135742938104115e-06,
+      "loss": 1.1331,
+      "step": 9310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2199202695629907,
+      "learning_rate": 2.1340360912347636e-06,
+      "loss": 1.0869,
+      "step": 9311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.459662173403049,
+      "learning_rate": 2.1323298452069853e-06,
+      "loss": 1.0923,
+      "step": 9312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.264600595961836,
+      "learning_rate": 2.130624200151107e-06,
+      "loss": 1.0977,
+      "step": 9313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.319358373180635,
+      "learning_rate": 2.128919156197413e-06,
+      "loss": 1.0801,
+      "step": 9314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5830990971259093,
+      "learning_rate": 2.1272147134761524e-06,
+      "loss": 1.0537,
+      "step": 9315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2618874940223592,
+      "learning_rate": 2.1255108721175066e-06,
+      "loss": 1.0879,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0339927733173586,
+      "learning_rate": 2.123807632251631e-06,
+      "loss": 1.0725,
+      "step": 9317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.300341528461758,
+      "learning_rate": 2.1221049940086267e-06,
+      "loss": 1.0493,
+      "step": 9318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9931819467181393,
+      "learning_rate": 2.120402957518546e-06,
+      "loss": 1.0686,
+      "step": 9319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0772293900979544,
+      "learning_rate": 2.118701522911404e-06,
+      "loss": 1.1135,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7306505364097418,
+      "learning_rate": 2.1170006903171605e-06,
+      "loss": 1.0696,
+      "step": 9321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1887243529479155,
+      "learning_rate": 2.1153004598657413e-06,
+      "loss": 1.1135,
+      "step": 9322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4669454165322118,
+      "learning_rate": 2.1136008316870095e-06,
+      "loss": 1.05,
+      "step": 9323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3436468616236654,
+      "learning_rate": 2.111901805910791e-06,
+      "loss": 1.0759,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.076494416175664,
+      "learning_rate": 2.110203382666873e-06,
+      "loss": 1.0942,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2952222065382233,
+      "learning_rate": 2.1085055620849835e-06,
+      "loss": 1.0671,
+      "step": 9326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.996963400754199,
+      "learning_rate": 2.106808344294815e-06,
+      "loss": 1.01,
+      "step": 9327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3936760334576563,
+      "learning_rate": 2.1051117294260072e-06,
+      "loss": 1.1045,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9036920628980147,
+      "learning_rate": 2.103415717608157e-06,
+      "loss": 1.0283,
+      "step": 9329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.220797311034302,
+      "learning_rate": 2.1017203089708125e-06,
+      "loss": 1.0779,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3940613190077085,
+      "learning_rate": 2.100025503643477e-06,
+      "loss": 1.1584,
+      "step": 9331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.350093825982621,
+      "learning_rate": 2.098331301755611e-06,
+      "loss": 1.0078,
+      "step": 9332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.117615900009773,
+      "learning_rate": 2.0966377034366248e-06,
+      "loss": 1.0757,
+      "step": 9333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.708561194263133,
+      "learning_rate": 2.0949447088158812e-06,
+      "loss": 1.0469,
+      "step": 9334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.144371570803708,
+      "learning_rate": 2.093252318022708e-06,
+      "loss": 1.0557,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7868669616820854,
+      "learning_rate": 2.0915605311863672e-06,
+      "loss": 1.042,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2863778059836237,
+      "learning_rate": 2.089869348436093e-06,
+      "loss": 1.0916,
+      "step": 9337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2778297449453726,
+      "learning_rate": 2.0881787699010657e-06,
+      "loss": 1.041,
+      "step": 9338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4127165425297994,
+      "learning_rate": 2.0864887957104162e-06,
+      "loss": 1.1125,
+      "step": 9339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.201482862602129,
+      "learning_rate": 2.084799425993239e-06,
+      "loss": 1.0591,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0093877123131434,
+      "learning_rate": 2.083110660878573e-06,
+      "loss": 1.0544,
+      "step": 9341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4047492354448594,
+      "learning_rate": 2.081422500495416e-06,
+      "loss": 1.0242,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1511658510224083,
+      "learning_rate": 2.079734944972717e-06,
+      "loss": 1.0278,
+      "step": 9343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1298681520663214,
+      "learning_rate": 2.0780479944393773e-06,
+      "loss": 1.0032,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0916142232010158,
+      "learning_rate": 2.07636164902426e-06,
+      "loss": 1.0198,
+      "step": 9345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.335985005819934,
+      "learning_rate": 2.074675908856174e-06,
+      "loss": 1.041,
+      "step": 9346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0348134611304682,
+      "learning_rate": 2.072990774063882e-06,
+      "loss": 0.9976,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.213162863638186,
+      "learning_rate": 2.071306244776109e-06,
+      "loss": 1.0688,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.765483737425171,
+      "learning_rate": 2.069622321121524e-06,
+      "loss": 1.0294,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.428373550122731,
+      "learning_rate": 2.067939003228754e-06,
+      "loss": 1.0227,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5175080951003044,
+      "learning_rate": 2.0662562912263784e-06,
+      "loss": 1.134,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.189088648745416,
+      "learning_rate": 2.064574185242929e-06,
+      "loss": 1.0684,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.421107383121777,
+      "learning_rate": 2.0628926854068996e-06,
+      "loss": 1.0874,
+      "step": 9353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.297691293117308,
+      "learning_rate": 2.0612117918467254e-06,
+      "loss": 1.0503,
+      "step": 9354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0616760570260912,
+      "learning_rate": 2.0595315046908094e-06,
+      "loss": 1.1113,
+      "step": 9355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4355322868423466,
+      "learning_rate": 2.057851824067493e-06,
+      "loss": 1.0317,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4921639076818503,
+      "learning_rate": 2.0561727501050778e-06,
+      "loss": 1.0322,
+      "step": 9357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.040477264877638,
+      "learning_rate": 2.054494282931825e-06,
+      "loss": 1.1023,
+      "step": 9358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1073768149105367,
+      "learning_rate": 2.0528164226759396e-06,
+      "loss": 1.0901,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5604693934154854,
+      "learning_rate": 2.05113916946559e-06,
+      "loss": 1.0261,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.19391152379712,
+      "learning_rate": 2.049462523428891e-06,
+      "loss": 1.147,
+      "step": 9361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6730587780263737,
+      "learning_rate": 2.0477864846939133e-06,
+      "loss": 1.0378,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2197073175724644,
+      "learning_rate": 2.0461110533886797e-06,
+      "loss": 1.0898,
+      "step": 9363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3938924799695496,
+      "learning_rate": 2.044436229641167e-06,
+      "loss": 1.0166,
+      "step": 9364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4584505768764005,
+      "learning_rate": 2.0427620135793103e-06,
+      "loss": 1.0183,
+      "step": 9365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.230815521734061,
+      "learning_rate": 2.0410884053309934e-06,
+      "loss": 1.0911,
+      "step": 9366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.755638558437399,
+      "learning_rate": 2.0394154050240513e-06,
+      "loss": 1.0208,
+      "step": 9367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9564789121061423,
+      "learning_rate": 2.0377430127862805e-06,
+      "loss": 1.0464,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7493759018405224,
+      "learning_rate": 2.0360712287454264e-06,
+      "loss": 1.0586,
+      "step": 9369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6794652390509985,
+      "learning_rate": 2.0344000530291875e-06,
+      "loss": 1.0398,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8018917531719456,
+      "learning_rate": 2.0327294857652157e-06,
+      "loss": 1.0494,
+      "step": 9371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6171514243985072,
+      "learning_rate": 2.031059527081114e-06,
+      "loss": 1.1785,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.373676106893943,
+      "learning_rate": 2.0293901771044487e-06,
+      "loss": 1.0833,
+      "step": 9373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.421959644556993,
+      "learning_rate": 2.0277214359627305e-06,
+      "loss": 1.0894,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0249597416220473,
+      "learning_rate": 2.0260533037834264e-06,
+      "loss": 1.0288,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.047803441004364,
+      "learning_rate": 2.024385780693956e-06,
+      "loss": 1.093,
+      "step": 9376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1095794432008668,
+      "learning_rate": 2.0227188668216903e-06,
+      "loss": 1.0959,
+      "step": 9377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1743076549081386,
+      "learning_rate": 2.0210525622939616e-06,
+      "loss": 1.1062,
+      "step": 9378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6402030560124223,
+      "learning_rate": 2.019386867238048e-06,
+      "loss": 1.0974,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9593700504595177,
+      "learning_rate": 2.017721781781181e-06,
+      "loss": 0.9622,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.05703513155924,
+      "learning_rate": 2.016057306050554e-06,
+      "loss": 1.0779,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.330497126843734,
+      "learning_rate": 2.0143934401733055e-06,
+      "loss": 1.073,
+      "step": 9382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.223319518874412,
+      "learning_rate": 2.012730184276529e-06,
+      "loss": 1.0254,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.263640704225842,
+      "learning_rate": 2.0110675384872725e-06,
+      "loss": 1.1558,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1289070718147447,
+      "learning_rate": 2.0094055029325355e-06,
+      "loss": 1.0239,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4382738824517096,
+      "learning_rate": 2.0077440777392755e-06,
+      "loss": 1.0186,
+      "step": 9386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3838313478522273,
+      "learning_rate": 2.0060832630344007e-06,
+      "loss": 1.1365,
+      "step": 9387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8032217484288338,
+      "learning_rate": 2.0044230589447698e-06,
+      "loss": 1.1433,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.477705945676218,
+      "learning_rate": 2.0027634655971994e-06,
+      "loss": 1.1479,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4878620706682395,
+      "learning_rate": 2.0011044831184546e-06,
+      "loss": 1.0081,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.943921626654671,
+      "learning_rate": 1.9994461116352603e-06,
+      "loss": 1.0837,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1644223784130356,
+      "learning_rate": 1.997788351274289e-06,
+      "loss": 1.0828,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.872353402168333,
+      "learning_rate": 1.9961312021621714e-06,
+      "loss": 1.0708,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.246587031689478,
+      "learning_rate": 1.9944746644254864e-06,
+      "loss": 1.1011,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.401933197999473,
+      "learning_rate": 1.9928187381907694e-06,
+      "loss": 1.1255,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2424908326626194,
+      "learning_rate": 1.991163423584509e-06,
+      "loss": 1.1194,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.067889682829422,
+      "learning_rate": 1.9895087207331422e-06,
+      "loss": 1.0503,
+      "step": 9397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.927378813572028,
+      "learning_rate": 1.9878546297630697e-06,
+      "loss": 1.1357,
+      "step": 9398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.877756460631163,
+      "learning_rate": 1.986201150800636e-06,
+      "loss": 1.0112,
+      "step": 9399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.248160439324982,
+      "learning_rate": 1.98454828397214e-06,
+      "loss": 1.0715,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3735011404522526,
+      "learning_rate": 1.9828960294038403e-06,
+      "loss": 1.0076,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.144036189938563,
+      "learning_rate": 1.9812443872219433e-06,
+      "loss": 1.0938,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.567603883660506,
+      "learning_rate": 1.9795933575526073e-06,
+      "loss": 1.0363,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8344743583633645,
+      "learning_rate": 1.9779429405219485e-06,
+      "loss": 1.1248,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.800182369408841,
+      "learning_rate": 1.976293136256031e-06,
+      "loss": 1.0781,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9055181990833603,
+      "learning_rate": 1.9746439448808785e-06,
+      "loss": 1.0454,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7948258348974337,
+      "learning_rate": 1.9729953665224642e-06,
+      "loss": 1.0356,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.551575656165616,
+      "learning_rate": 1.9713474013067136e-06,
+      "loss": 1.0278,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.471094333886675,
+      "learning_rate": 1.9697000493595054e-06,
+      "loss": 1.1077,
+      "step": 9409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1791261605404837,
+      "learning_rate": 1.9680533108066723e-06,
+      "loss": 1.1094,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4651859540763184,
+      "learning_rate": 1.966407185774004e-06,
+      "loss": 1.0161,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2702636979328426,
+      "learning_rate": 1.964761674387238e-06,
+      "loss": 1.0801,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9699488520259396,
+      "learning_rate": 1.9631167767720627e-06,
+      "loss": 1.0032,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2920148351360465,
+      "learning_rate": 1.9614724930541307e-06,
+      "loss": 1.0535,
+      "step": 9414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1995156693862787,
+      "learning_rate": 1.959828823359037e-06,
+      "loss": 1.0647,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.040416886746306,
+      "learning_rate": 1.9581857678123316e-06,
+      "loss": 1.0625,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2760042744964735,
+      "learning_rate": 1.9565433265395228e-06,
+      "loss": 1.0784,
+      "step": 9417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4151985679631194,
+      "learning_rate": 1.954901499666064e-06,
+      "loss": 1.104,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7852098384907515,
+      "learning_rate": 1.953260287317371e-06,
+      "loss": 1.0715,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.731480385417831,
+      "learning_rate": 1.9516196896188057e-06,
+      "loss": 1.0813,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0056676370475652,
+      "learning_rate": 1.9499797066956815e-06,
+      "loss": 1.0286,
+      "step": 9421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.023710403875393,
+      "learning_rate": 1.948340338673278e-06,
+      "loss": 1.0732,
+      "step": 9422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.797894449387975,
+      "learning_rate": 1.946701585676807e-06,
+      "loss": 1.0281,
+      "step": 9423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.598219882961907,
+      "learning_rate": 1.945063447831452e-06,
+      "loss": 1.1172,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.122813479688581,
+      "learning_rate": 1.943425925262341e-06,
+      "loss": 1.0547,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9568040769491932,
+      "learning_rate": 1.941789018094553e-06,
+      "loss": 1.0608,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9393710709522156,
+      "learning_rate": 1.9401527264531274e-06,
+      "loss": 1.0422,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4443073839345764,
+      "learning_rate": 1.938517050463049e-06,
+      "loss": 1.114,
+      "step": 9428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.014466645382364,
+      "learning_rate": 1.936881990249262e-06,
+      "loss": 1.1201,
+      "step": 9429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7036326341871493,
+      "learning_rate": 1.935247545936655e-06,
+      "loss": 1.0352,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1616003632326097,
+      "learning_rate": 1.9336137176500814e-06,
+      "loss": 1.0476,
+      "step": 9431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.114142287472014,
+      "learning_rate": 1.931980505514338e-06,
+      "loss": 1.1047,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6620358954089225,
+      "learning_rate": 1.9303479096541767e-06,
+      "loss": 1.0115,
+      "step": 9433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4992838599485543,
+      "learning_rate": 1.928715930194306e-06,
+      "loss": 1.0798,
+      "step": 9434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2170066644213025,
+      "learning_rate": 1.927084567259384e-06,
+      "loss": 1.0786,
+      "step": 9435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8749343117726345,
+      "learning_rate": 1.9254538209740216e-06,
+      "loss": 1.0332,
+      "step": 9436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1116802852313836,
+      "learning_rate": 1.923823691462783e-06,
+      "loss": 1.033,
+      "step": 9437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2135198491512287,
+      "learning_rate": 1.922194178850184e-06,
+      "loss": 1.075,
+      "step": 9438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7860463866542737,
+      "learning_rate": 1.920565283260699e-06,
+      "loss": 1.0925,
+      "step": 9439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7506954943621964,
+      "learning_rate": 1.9189370048187496e-06,
+      "loss": 1.0242,
+      "step": 9440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.529910841532782,
+      "learning_rate": 1.917309343648711e-06,
+      "loss": 1.1279,
+      "step": 9441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5219559727935628,
+      "learning_rate": 1.9156822998749126e-06,
+      "loss": 1.0261,
+      "step": 9442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.263607576899424,
+      "learning_rate": 1.914055873621633e-06,
+      "loss": 1.0752,
+      "step": 9443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5730300806744957,
+      "learning_rate": 1.9124300650131136e-06,
+      "loss": 1.0774,
+      "step": 9444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1931832151207815,
+      "learning_rate": 1.9108048741735364e-06,
+      "loss": 1.0728,
+      "step": 9445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8948644254736493,
+      "learning_rate": 1.9091803012270403e-06,
+      "loss": 1.0549,
+      "step": 9446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1917511268212313,
+      "learning_rate": 1.9075563462977253e-06,
+      "loss": 1.0999,
+      "step": 9447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.925674302510934,
+      "learning_rate": 1.9059330095096318e-06,
+      "loss": 1.0569,
+      "step": 9448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.464597031073088,
+      "learning_rate": 1.9043102909867595e-06,
+      "loss": 1.0098,
+      "step": 9449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.293864167684476,
+      "learning_rate": 1.9026881908530593e-06,
+      "loss": 1.0627,
+      "step": 9450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9621367781130026,
+      "learning_rate": 1.9010667092324342e-06,
+      "loss": 1.0713,
+      "step": 9451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.101872923922329,
+      "learning_rate": 1.8994458462487442e-06,
+      "loss": 1.0422,
+      "step": 9452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.173049289583755,
+      "learning_rate": 1.8978256020257978e-06,
+      "loss": 1.0669,
+      "step": 9453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.99522532582536,
+      "learning_rate": 1.8962059766873541e-06,
+      "loss": 1.0396,
+      "step": 9454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8807464600110473,
+      "learning_rate": 1.8945869703571362e-06,
+      "loss": 1.1187,
+      "step": 9455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4007170555833115,
+      "learning_rate": 1.8929685831588007e-06,
+      "loss": 0.9983,
+      "step": 9456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.137844289335219,
+      "learning_rate": 1.8913508152159777e-06,
+      "loss": 1.0408,
+      "step": 9457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1671352350492663,
+      "learning_rate": 1.889733666652236e-06,
+      "loss": 1.0156,
+      "step": 9458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7258139844221496,
+      "learning_rate": 1.888117137591099e-06,
+      "loss": 1.0522,
+      "step": 9459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.088387696476063,
+      "learning_rate": 1.8865012281560523e-06,
+      "loss": 1.0298,
+      "step": 9460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2796912470201423,
+      "learning_rate": 1.8848859384705221e-06,
+      "loss": 1.0842,
+      "step": 9461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.962012915654887,
+      "learning_rate": 1.8832712686578937e-06,
+      "loss": 1.0637,
+      "step": 9462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.495859429562566,
+      "learning_rate": 1.8816572188415028e-06,
+      "loss": 1.0994,
+      "step": 9463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8503612331405694,
+      "learning_rate": 1.8800437891446378e-06,
+      "loss": 1.0061,
+      "step": 9464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4443771411485113,
+      "learning_rate": 1.878430979690543e-06,
+      "loss": 1.0134,
+      "step": 9465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2038324537222165,
+      "learning_rate": 1.8768187906024104e-06,
+      "loss": 1.0627,
+      "step": 9466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4912336228707352,
+      "learning_rate": 1.87520722200339e-06,
+      "loss": 1.1128,
+      "step": 9467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.310859971938431,
+      "learning_rate": 1.8735962740165826e-06,
+      "loss": 1.0806,
+      "step": 9468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8119328643942336,
+      "learning_rate": 1.8719859467650314e-06,
+      "loss": 1.0569,
+      "step": 9469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1444074437546727,
+      "learning_rate": 1.8703762403717519e-06,
+      "loss": 1.0808,
+      "step": 9470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0271749740130063,
+      "learning_rate": 1.868767154959693e-06,
+      "loss": 1.1042,
+      "step": 9471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7862583537252474,
+      "learning_rate": 1.8671586906517713e-06,
+      "loss": 1.0566,
+      "step": 9472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5657922581428196,
+      "learning_rate": 1.8655508475708473e-06,
+      "loss": 1.0269,
+      "step": 9473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2829911386517048,
+      "learning_rate": 1.8639436258397326e-06,
+      "loss": 1.074,
+      "step": 9474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2556275949289986,
+      "learning_rate": 1.8623370255812034e-06,
+      "loss": 1.0066,
+      "step": 9475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1001156741682294,
+      "learning_rate": 1.8607310469179686e-06,
+      "loss": 1.0669,
+      "step": 9476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.282971076236691,
+      "learning_rate": 1.8591256899727094e-06,
+      "loss": 1.1343,
+      "step": 9477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.051512811604521,
+      "learning_rate": 1.8575209548680472e-06,
+      "loss": 1.1294,
+      "step": 9478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5396680715121938,
+      "learning_rate": 1.8559168417265583e-06,
+      "loss": 1.0925,
+      "step": 9479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.679815837951577,
+      "learning_rate": 1.8543133506707778e-06,
+      "loss": 1.0854,
+      "step": 9480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9929617425237356,
+      "learning_rate": 1.8527104818231857e-06,
+      "loss": 1.0415,
+      "step": 9481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1919880092215136,
+      "learning_rate": 1.851108235306216e-06,
+      "loss": 1.0806,
+      "step": 9482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.027081499892401,
+      "learning_rate": 1.8495066112422578e-06,
+      "loss": 1.0024,
+      "step": 9483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.370545119884127,
+      "learning_rate": 1.8479056097536485e-06,
+      "loss": 1.0994,
+      "step": 9484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7059666953616373,
+      "learning_rate": 1.846305230962685e-06,
+      "loss": 1.0991,
+      "step": 9485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.159698429441286,
+      "learning_rate": 1.8447054749916092e-06,
+      "loss": 1.0171,
+      "step": 9486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.761858787351725,
+      "learning_rate": 1.8431063419626172e-06,
+      "loss": 1.1167,
+      "step": 9487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.405398731645176,
+      "learning_rate": 1.8415078319978662e-06,
+      "loss": 1.0928,
+      "step": 9488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2065384545141247,
+      "learning_rate": 1.839909945219448e-06,
+      "loss": 1.0745,
+      "step": 9489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3537984459656873,
+      "learning_rate": 1.8383126817494234e-06,
+      "loss": 1.0645,
+      "step": 9490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.752404650631733,
+      "learning_rate": 1.8367160417097996e-06,
+      "loss": 1.0647,
+      "step": 9491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8761838423767383,
+      "learning_rate": 1.8351200252225299e-06,
+      "loss": 1.1003,
+      "step": 9492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2070615544263856,
+      "learning_rate": 1.8335246324095346e-06,
+      "loss": 0.9919,
+      "step": 9493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.912035456866756,
+      "learning_rate": 1.8319298633926697e-06,
+      "loss": 1.0627,
+      "step": 9494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0073409663020665,
+      "learning_rate": 1.830335718293762e-06,
+      "loss": 1.0803,
+      "step": 9495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3966229420387273,
+      "learning_rate": 1.8287421972345697e-06,
+      "loss": 1.0801,
+      "step": 9496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.699455081127649,
+      "learning_rate": 1.8271493003368169e-06,
+      "loss": 1.0967,
+      "step": 9497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.032847168049681,
+      "learning_rate": 1.82555702772218e-06,
+      "loss": 0.9917,
+      "step": 9498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.216779395845194,
+      "learning_rate": 1.8239653795122804e-06,
+      "loss": 1.0427,
+      "step": 9499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.427063404610364,
+      "learning_rate": 1.822374355828701e-06,
+      "loss": 1.0437,
+      "step": 9500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4847408839481613,
+      "learning_rate": 1.820783956792972e-06,
+      "loss": 1.0276,
+      "step": 9501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.014120029436965,
+      "learning_rate": 1.819194182526568e-06,
+      "loss": 1.0579,
+      "step": 9502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1735426304131074,
+      "learning_rate": 1.8176050331509331e-06,
+      "loss": 1.085,
+      "step": 9503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.57371920315306,
+      "learning_rate": 1.8160165087874482e-06,
+      "loss": 1.1121,
+      "step": 9504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3330811100737194,
+      "learning_rate": 1.814428609557458e-06,
+      "loss": 1.0833,
+      "step": 9505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.141651423084998,
+      "learning_rate": 1.8128413355822517e-06,
+      "loss": 1.0574,
+      "step": 9506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.422284236669467,
+      "learning_rate": 1.8112546869830705e-06,
+      "loss": 1.0867,
+      "step": 9507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.651408178318201,
+      "learning_rate": 1.809668663881118e-06,
+      "loss": 1.0305,
+      "step": 9508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.168008583291142,
+      "learning_rate": 1.8080832663975322e-06,
+      "loss": 1.0171,
+      "step": 9509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.404541959656441,
+      "learning_rate": 1.8064984946534226e-06,
+      "loss": 1.0444,
+      "step": 9510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2544938114530564,
+      "learning_rate": 1.8049143487698374e-06,
+      "loss": 1.0918,
+      "step": 9511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.892326169499697,
+      "learning_rate": 1.8033308288677808e-06,
+      "loss": 1.0239,
+      "step": 9512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6470600707088523,
+      "learning_rate": 1.8017479350682153e-06,
+      "loss": 1.0278,
+      "step": 9513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.032145557995487,
+      "learning_rate": 1.8001656674920453e-06,
+      "loss": 1.073,
+      "step": 9514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.294505702725458,
+      "learning_rate": 1.7985840262601351e-06,
+      "loss": 1.0261,
+      "step": 9515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.632883486131107,
+      "learning_rate": 1.7970030114932969e-06,
+      "loss": 1.1245,
+      "step": 9516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7297795495890806,
+      "learning_rate": 1.7954226233122928e-06,
+      "loss": 1.0098,
+      "step": 9517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0359621845114035,
+      "learning_rate": 1.7938428618378478e-06,
+      "loss": 0.97,
+      "step": 9518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9690033663372737,
+      "learning_rate": 1.7922637271906296e-06,
+      "loss": 1.031,
+      "step": 9519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.076162141524183,
+      "learning_rate": 1.7906852194912562e-06,
+      "loss": 1.1155,
+      "step": 9520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0856502590161896,
+      "learning_rate": 1.7891073388603096e-06,
+      "loss": 1.0457,
+      "step": 9521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1718051677654318,
+      "learning_rate": 1.7875300854183074e-06,
+      "loss": 1.0791,
+      "step": 9522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1160966883008405,
+      "learning_rate": 1.7859534592857353e-06,
+      "loss": 1.0737,
+      "step": 9523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8961274363581953,
+      "learning_rate": 1.784377460583021e-06,
+      "loss": 1.0105,
+      "step": 9524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.074636303907694,
+      "learning_rate": 1.7828020894305443e-06,
+      "loss": 1.083,
+      "step": 9525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2141714925672993,
+      "learning_rate": 1.7812273459486451e-06,
+      "loss": 1.0762,
+      "step": 9526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1980525114874805,
+      "learning_rate": 1.7796532302576086e-06,
+      "loss": 1.0762,
+      "step": 9527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1230297366978776,
+      "learning_rate": 1.778079742477673e-06,
+      "loss": 1.0183,
+      "step": 9528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9606591994399767,
+      "learning_rate": 1.7765068827290277e-06,
+      "loss": 1.0095,
+      "step": 9529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.250362066156753,
+      "learning_rate": 1.7749346511318154e-06,
+      "loss": 1.0762,
+      "step": 9530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.420177445265623,
+      "learning_rate": 1.7733630478061347e-06,
+      "loss": 1.0984,
+      "step": 9531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0046219683860365,
+      "learning_rate": 1.7717920728720284e-06,
+      "loss": 1.0869,
+      "step": 9532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.369380954462256,
+      "learning_rate": 1.7702217264494993e-06,
+      "loss": 1.0347,
+      "step": 9533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0393534374135283,
+      "learning_rate": 1.7686520086584992e-06,
+      "loss": 1.075,
+      "step": 9534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.007580773604869,
+      "learning_rate": 1.767082919618922e-06,
+      "loss": 1.1289,
+      "step": 9535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.396592988966946,
+      "learning_rate": 1.7655144594506324e-06,
+      "loss": 1.0261,
+      "step": 9536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.23813200092276,
+      "learning_rate": 1.7639466282734308e-06,
+      "loss": 1.0835,
+      "step": 9537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.469902308039672,
+      "learning_rate": 1.7623794262070805e-06,
+      "loss": 1.0234,
+      "step": 9538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7518931159537963,
+      "learning_rate": 1.7608128533712908e-06,
+      "loss": 1.1272,
+      "step": 9539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.041667795718595,
+      "learning_rate": 1.7592469098857212e-06,
+      "loss": 1.0203,
+      "step": 9540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.79653259447714,
+      "learning_rate": 1.7576815958699932e-06,
+      "loss": 1.0493,
+      "step": 9541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1854166360892187,
+      "learning_rate": 1.756116911443665e-06,
+      "loss": 1.0642,
+      "step": 9542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.898341402465166,
+      "learning_rate": 1.7545528567262615e-06,
+      "loss": 1.0645,
+      "step": 9543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3036210711494274,
+      "learning_rate": 1.7529894318372498e-06,
+      "loss": 1.0652,
+      "step": 9544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.314397760919604,
+      "learning_rate": 1.7514266368960509e-06,
+      "loss": 1.0605,
+      "step": 9545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5311827607051303,
+      "learning_rate": 1.7498644720220437e-06,
+      "loss": 1.1667,
+      "step": 9546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0015026831425353,
+      "learning_rate": 1.7483029373345505e-06,
+      "loss": 1.0427,
+      "step": 9547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1902753581285688,
+      "learning_rate": 1.7467420329528506e-06,
+      "loss": 1.0588,
+      "step": 9548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.309242440251064,
+      "learning_rate": 1.7451817589961727e-06,
+      "loss": 1.0505,
+      "step": 9549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9139584810072496,
+      "learning_rate": 1.7436221155836953e-06,
+      "loss": 1.1008,
+      "step": 9550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.443536859331709,
+      "learning_rate": 1.742063102834558e-06,
+      "loss": 1.0535,
+      "step": 9551
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8902550407417906,
+      "learning_rate": 1.7405047208678438e-06,
+      "loss": 1.0522,
+      "step": 9552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8850893591684774,
+      "learning_rate": 1.7389469698025851e-06,
+      "loss": 1.0396,
+      "step": 9553
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1344191116074795,
+      "learning_rate": 1.73738984975778e-06,
+      "loss": 1.0496,
+      "step": 9554
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9699662137946605,
+      "learning_rate": 1.7358333608523592e-06,
+      "loss": 1.071,
+      "step": 9555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.472560433641759,
+      "learning_rate": 1.7342775032052205e-06,
+      "loss": 1.0703,
+      "step": 9556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.236777792640004,
+      "learning_rate": 1.732722276935208e-06,
+      "loss": 1.0364,
+      "step": 9557
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1230598543969768,
+      "learning_rate": 1.7311676821611145e-06,
+      "loss": 1.0642,
+      "step": 9558
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5969517332017453,
+      "learning_rate": 1.7296137190016916e-06,
+      "loss": 1.0886,
+      "step": 9559
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1271668460285778,
+      "learning_rate": 1.7280603875756375e-06,
+      "loss": 1.0781,
+      "step": 9560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9275297312250905,
+      "learning_rate": 1.7265076880016018e-06,
+      "loss": 1.0281,
+      "step": 9561
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.653287103273093,
+      "learning_rate": 1.72495562039819e-06,
+      "loss": 1.1082,
+      "step": 9562
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4625503091904584,
+      "learning_rate": 1.7234041848839522e-06,
+      "loss": 1.0479,
+      "step": 9563
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3375790485743244,
+      "learning_rate": 1.7218533815774007e-06,
+      "loss": 1.0322,
+      "step": 9564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.134766515392997,
+      "learning_rate": 1.7203032105969897e-06,
+      "loss": 1.0322,
+      "step": 9565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5291116056410723,
+      "learning_rate": 1.7187536720611287e-06,
+      "loss": 1.0708,
+      "step": 9566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2803518296425884,
+      "learning_rate": 1.7172047660881852e-06,
+      "loss": 1.1348,
+      "step": 9567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1524239975849864,
+      "learning_rate": 1.7156564927964637e-06,
+      "loss": 1.0491,
+      "step": 9568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.976145553829211,
+      "learning_rate": 1.714108852304235e-06,
+      "loss": 1.0161,
+      "step": 9569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.444867196953834,
+      "learning_rate": 1.7125618447297132e-06,
+      "loss": 1.0737,
+      "step": 9570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2116423248648047,
+      "learning_rate": 1.7110154701910642e-06,
+      "loss": 1.0479,
+      "step": 9571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3466486623298275,
+      "learning_rate": 1.7094697288064143e-06,
+      "loss": 1.0889,
+      "step": 9572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0346875544565854,
+      "learning_rate": 1.707924620693827e-06,
+      "loss": 1.0288,
+      "step": 9573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.883282174931376,
+      "learning_rate": 1.7063801459713358e-06,
+      "loss": 1.0625,
+      "step": 9574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.813623099130926,
+      "learning_rate": 1.704836304756904e-06,
+      "loss": 1.095,
+      "step": 9575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9444804664980317,
+      "learning_rate": 1.7032930971684659e-06,
+      "loss": 0.9966,
+      "step": 9576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6427883867322444,
+      "learning_rate": 1.7017505233238962e-06,
+      "loss": 1.0684,
+      "step": 9577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9780432895671938,
+      "learning_rate": 1.700208583341022e-06,
+      "loss": 1.0576,
+      "step": 9578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.199054142189576,
+      "learning_rate": 1.69866727733763e-06,
+      "loss": 1.053,
+      "step": 9579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8572971224848653,
+      "learning_rate": 1.6971266054314494e-06,
+      "loss": 1.0811,
+      "step": 9580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.215880441494476,
+      "learning_rate": 1.6955865677401651e-06,
+      "loss": 1.0481,
+      "step": 9581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9530801347069437,
+      "learning_rate": 1.6940471643814128e-06,
+      "loss": 1.0847,
+      "step": 9582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.344037541262882,
+      "learning_rate": 1.6925083954727773e-06,
+      "loss": 1.095,
+      "step": 9583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0858912995658523,
+      "learning_rate": 1.6909702611318024e-06,
+      "loss": 1.0059,
+      "step": 9584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.659146262551472,
+      "learning_rate": 1.6894327614759754e-06,
+      "loss": 1.1042,
+      "step": 9585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4664470095792637,
+      "learning_rate": 1.6878958966227366e-06,
+      "loss": 1.0969,
+      "step": 9586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8534505653601836,
+      "learning_rate": 1.6863596666894866e-06,
+      "loss": 1.0376,
+      "step": 9587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3519824888256076,
+      "learning_rate": 1.6848240717935605e-06,
+      "loss": 1.0212,
+      "step": 9588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1692421386197354,
+      "learning_rate": 1.683289112052262e-06,
+      "loss": 1.022,
+      "step": 9589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1850278944412964,
+      "learning_rate": 1.6817547875828365e-06,
+      "loss": 1.0225,
+      "step": 9590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9050039680916226,
+      "learning_rate": 1.6802210985024802e-06,
+      "loss": 1.0579,
+      "step": 9591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0141640759987482,
+      "learning_rate": 1.6786880449283493e-06,
+      "loss": 1.0813,
+      "step": 9592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8415423214925895,
+      "learning_rate": 1.6771556269775446e-06,
+      "loss": 1.0715,
+      "step": 9593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.990870883984162,
+      "learning_rate": 1.6756238447671181e-06,
+      "loss": 1.0505,
+      "step": 9594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9775799495589688,
+      "learning_rate": 1.6740926984140771e-06,
+      "loss": 1.0278,
+      "step": 9595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5688945883565446,
+      "learning_rate": 1.6725621880353737e-06,
+      "loss": 1.0566,
+      "step": 9596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2587078336995376,
+      "learning_rate": 1.6710323137479222e-06,
+      "loss": 1.0469,
+      "step": 9597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.317622700502591,
+      "learning_rate": 1.6695030756685782e-06,
+      "loss": 1.0811,
+      "step": 9598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5438743194870637,
+      "learning_rate": 1.6679744739141512e-06,
+      "loss": 1.0845,
+      "step": 9599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.395337319667384,
+      "learning_rate": 1.6664465086014082e-06,
+      "loss": 1.1545,
+      "step": 9600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0575395487897934,
+      "learning_rate": 1.664919179847061e-06,
+      "loss": 1.0637,
+      "step": 9601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2336225817196365,
+      "learning_rate": 1.663392487767772e-06,
+      "loss": 1.1011,
+      "step": 9602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.46513037533579,
+      "learning_rate": 1.6618664324801615e-06,
+      "loss": 1.0659,
+      "step": 9603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.19191942260698,
+      "learning_rate": 1.6603410141007913e-06,
+      "loss": 1.0391,
+      "step": 9604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2305706938118997,
+      "learning_rate": 1.6588162327461866e-06,
+      "loss": 1.1216,
+      "step": 9605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3544148749947036,
+      "learning_rate": 1.6572920885328137e-06,
+      "loss": 1.0723,
+      "step": 9606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5461447096274923,
+      "learning_rate": 1.6557685815771006e-06,
+      "loss": 1.0925,
+      "step": 9607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3014950615272944,
+      "learning_rate": 1.6542457119954135e-06,
+      "loss": 1.0323,
+      "step": 9608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7665938590360253,
+      "learning_rate": 1.652723479904076e-06,
+      "loss": 1.075,
+      "step": 9609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8979180696672513,
+      "learning_rate": 1.6512018854193702e-06,
+      "loss": 1.1108,
+      "step": 9610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.907535324272242,
+      "learning_rate": 1.6496809286575177e-06,
+      "loss": 1.0479,
+      "step": 9611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7997760066874235,
+      "learning_rate": 1.6481606097347002e-06,
+      "loss": 1.01,
+      "step": 9612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0542971243896657,
+      "learning_rate": 1.646640928767047e-06,
+      "loss": 1.1528,
+      "step": 9613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8653395795327805,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 1.0283,
+      "step": 9614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0180064863333698,
+      "learning_rate": 1.6436034811615032e-06,
+      "loss": 1.0938,
+      "step": 9615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2296978057679366,
+      "learning_rate": 1.6420857147556268e-06,
+      "loss": 1.1008,
+      "step": 9616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.308853571152445,
+      "learning_rate": 1.6405685867689469e-06,
+      "loss": 1.0747,
+      "step": 9617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.653334815540628,
+      "learning_rate": 1.6390520973173463e-06,
+      "loss": 1.0508,
+      "step": 9618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7274931821630446,
+      "learning_rate": 1.63753624651666e-06,
+      "loss": 1.0947,
+      "step": 9619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1335285800081216,
+      "learning_rate": 1.636021034482682e-06,
+      "loss": 1.0569,
+      "step": 9620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.187879649774313,
+      "learning_rate": 1.634506461331148e-06,
+      "loss": 1.0945,
+      "step": 9621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1276774079155354,
+      "learning_rate": 1.6329925271777492e-06,
+      "loss": 1.0151,
+      "step": 9622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.437195940414636,
+      "learning_rate": 1.6314792321381267e-06,
+      "loss": 1.0576,
+      "step": 9623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5202589462929006,
+      "learning_rate": 1.6299665763278726e-06,
+      "loss": 1.0815,
+      "step": 9624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1070385547542,
+      "learning_rate": 1.6284545598625357e-06,
+      "loss": 1.1072,
+      "step": 9625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.085393106341605,
+      "learning_rate": 1.6269431828576077e-06,
+      "loss": 1.0486,
+      "step": 9626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.177485450538789,
+      "learning_rate": 1.6254324454285364e-06,
+      "loss": 1.0688,
+      "step": 9627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7529769634938597,
+      "learning_rate": 1.6239223476907184e-06,
+      "loss": 1.0198,
+      "step": 9628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.486424153071306,
+      "learning_rate": 1.6224128897595003e-06,
+      "loss": 1.0422,
+      "step": 9629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.053390727990851,
+      "learning_rate": 1.620904071750188e-06,
+      "loss": 1.0095,
+      "step": 9630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.032584604449541,
+      "learning_rate": 1.6193958937780285e-06,
+      "loss": 1.0171,
+      "step": 9631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.044302732749486,
+      "learning_rate": 1.6178883559582225e-06,
+      "loss": 1.0684,
+      "step": 9632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.579181623604262,
+      "learning_rate": 1.6163814584059278e-06,
+      "loss": 1.0542,
+      "step": 9633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.716780488515343,
+      "learning_rate": 1.6148752012362467e-06,
+      "loss": 1.0256,
+      "step": 9634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.99366779888048,
+      "learning_rate": 1.613369584564235e-06,
+      "loss": 1.0869,
+      "step": 9635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.371217442079883,
+      "learning_rate": 1.6118646085048972e-06,
+      "loss": 1.0591,
+      "step": 9636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0348074549522077,
+      "learning_rate": 1.6103602731731916e-06,
+      "loss": 1.0696,
+      "step": 9637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.064409975188153,
+      "learning_rate": 1.6088565786840282e-06,
+      "loss": 1.0308,
+      "step": 9638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7742583044144102,
+      "learning_rate": 1.6073535251522643e-06,
+      "loss": 1.0291,
+      "step": 9639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.157682135073243,
+      "learning_rate": 1.6058511126927178e-06,
+      "loss": 1.0562,
+      "step": 9640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.13127641479335,
+      "learning_rate": 1.6043493414201417e-06,
+      "loss": 1.0444,
+      "step": 9641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.403117145346675,
+      "learning_rate": 1.6028482114492493e-06,
+      "loss": 1.0793,
+      "step": 9642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3880515030905314,
+      "learning_rate": 1.6013477228947104e-06,
+      "loss": 1.0281,
+      "step": 9643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.147236609061804,
+      "learning_rate": 1.5998478758711345e-06,
+      "loss": 1.0587,
+      "step": 9644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.643593098696903,
+      "learning_rate": 1.5983486704930907e-06,
+      "loss": 1.0742,
+      "step": 9645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8418577736624937,
+      "learning_rate": 1.5968501068750953e-06,
+      "loss": 1.0044,
+      "step": 9646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.271089250844144,
+      "learning_rate": 1.5953521851316157e-06,
+      "loss": 1.0718,
+      "step": 9647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.636109418737928,
+      "learning_rate": 1.593854905377069e-06,
+      "loss": 1.0789,
+      "step": 9648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5344139986190717,
+      "learning_rate": 1.5923582677258254e-06,
+      "loss": 1.0964,
+      "step": 9649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.005009108000887,
+      "learning_rate": 1.5908622722922073e-06,
+      "loss": 1.0635,
+      "step": 9650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.326112113328893,
+      "learning_rate": 1.589366919190486e-06,
+      "loss": 1.0869,
+      "step": 9651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7975095939359864,
+      "learning_rate": 1.5878722085348808e-06,
+      "loss": 1.0408,
+      "step": 9652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6936850264274037,
+      "learning_rate": 1.5863781404395695e-06,
+      "loss": 1.0847,
+      "step": 9653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0350779074668046,
+      "learning_rate": 1.5848847150186752e-06,
+      "loss": 1.0593,
+      "step": 9654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8510026382160123,
+      "learning_rate": 1.5833919323862734e-06,
+      "loss": 1.0437,
+      "step": 9655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.315539078459816,
+      "learning_rate": 1.5818997926563884e-06,
+      "loss": 1.0784,
+      "step": 9656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.948827669337234,
+      "learning_rate": 1.5804082959429968e-06,
+      "loss": 1.1089,
+      "step": 9657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9543328984374515,
+      "learning_rate": 1.5789174423600307e-06,
+      "loss": 1.0454,
+      "step": 9658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9461052197606366,
+      "learning_rate": 1.5774272320213658e-06,
+      "loss": 1.0505,
+      "step": 9659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1303221824155796,
+      "learning_rate": 1.575937665040832e-06,
+      "loss": 1.1174,
+      "step": 9660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2683988944485756,
+      "learning_rate": 1.5744487415322107e-06,
+      "loss": 1.0913,
+      "step": 9661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9864450558747415,
+      "learning_rate": 1.5729604616092308e-06,
+      "loss": 1.1072,
+      "step": 9662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9756159204218764,
+      "learning_rate": 1.5714728253855793e-06,
+      "loss": 1.0232,
+      "step": 9663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.98957474589949,
+      "learning_rate": 1.5699858329748852e-06,
+      "loss": 1.043,
+      "step": 9664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.934678393387296,
+      "learning_rate": 1.5684994844907331e-06,
+      "loss": 1.0674,
+      "step": 9665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8802253856363964,
+      "learning_rate": 1.56701378004666e-06,
+      "loss": 1.0918,
+      "step": 9666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4164912444519704,
+      "learning_rate": 1.5655287197561497e-06,
+      "loss": 1.1018,
+      "step": 9667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.856226645275451,
+      "learning_rate": 1.564044303732638e-06,
+      "loss": 1.0295,
+      "step": 9668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8754625539524277,
+      "learning_rate": 1.5625605320895143e-06,
+      "loss": 1.0493,
+      "step": 9669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3153291339149,
+      "learning_rate": 1.5610774049401112e-06,
+      "loss": 1.0481,
+      "step": 9670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8183771813253773,
+      "learning_rate": 1.5595949223977235e-06,
+      "loss": 1.0593,
+      "step": 9671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.28398865655475,
+      "learning_rate": 1.5581130845755887e-06,
+      "loss": 1.022,
+      "step": 9672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5264803488081187,
+      "learning_rate": 1.556631891586894e-06,
+      "loss": 1.0774,
+      "step": 9673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.808484195521006,
+      "learning_rate": 1.5551513435447863e-06,
+      "loss": 1.0029,
+      "step": 9674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.441831827505756,
+      "learning_rate": 1.5536714405623498e-06,
+      "loss": 1.0193,
+      "step": 9675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.447389003461243,
+      "learning_rate": 1.5521921827526332e-06,
+      "loss": 1.0764,
+      "step": 9676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9211233721052694,
+      "learning_rate": 1.5507135702286247e-06,
+      "loss": 1.0657,
+      "step": 9677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1297464020382426,
+      "learning_rate": 1.5492356031032717e-06,
+      "loss": 1.0012,
+      "step": 9678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3184315133069973,
+      "learning_rate": 1.5477582814894688e-06,
+      "loss": 1.0212,
+      "step": 9679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.04766266374028,
+      "learning_rate": 1.5462816055000597e-06,
+      "loss": 1.0417,
+      "step": 9680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5321124671172317,
+      "learning_rate": 1.5448055752478397e-06,
+      "loss": 1.019,
+      "step": 9681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8732910663514946,
+      "learning_rate": 1.5433301908455544e-06,
+      "loss": 1.0344,
+      "step": 9682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1084132217387612,
+      "learning_rate": 1.5418554524059047e-06,
+      "loss": 1.0413,
+      "step": 9683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.455341518566513,
+      "learning_rate": 1.5403813600415375e-06,
+      "loss": 1.103,
+      "step": 9684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2284138068340194,
+      "learning_rate": 1.5389079138650464e-06,
+      "loss": 1.021,
+      "step": 9685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1829302652958735,
+      "learning_rate": 1.5374351139889875e-06,
+      "loss": 1.0193,
+      "step": 9686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0803410621052043,
+      "learning_rate": 1.5359629605258585e-06,
+      "loss": 1.0981,
+      "step": 9687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.825574280981172,
+      "learning_rate": 1.5344914535881084e-06,
+      "loss": 1.011,
+      "step": 9688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9132392127161846,
+      "learning_rate": 1.5330205932881382e-06,
+      "loss": 1.0466,
+      "step": 9689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7074109856229875,
+      "learning_rate": 1.531550379738298e-06,
+      "loss": 1.1758,
+      "step": 9690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1828554127345527,
+      "learning_rate": 1.5300808130508948e-06,
+      "loss": 1.0173,
+      "step": 9691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.07933677956947,
+      "learning_rate": 1.52861189333818e-06,
+      "loss": 1.0471,
+      "step": 9692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.666346831585965,
+      "learning_rate": 1.5271436207123537e-06,
+      "loss": 1.1064,
+      "step": 9693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.964626768274126,
+      "learning_rate": 1.5256759952855737e-06,
+      "loss": 1.0493,
+      "step": 9694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9269574160603593,
+      "learning_rate": 1.5242090171699397e-06,
+      "loss": 1.0129,
+      "step": 9695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2241039619874714,
+      "learning_rate": 1.522742686477512e-06,
+      "loss": 1.0449,
+      "step": 9696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9289758049868353,
+      "learning_rate": 1.521277003320295e-06,
+      "loss": 1.105,
+      "step": 9697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6053389653833987,
+      "learning_rate": 1.5198119678102418e-06,
+      "loss": 0.9562,
+      "step": 9698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.311496127597306,
+      "learning_rate": 1.5183475800592628e-06,
+      "loss": 1.0642,
+      "step": 9699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.271305023088146,
+      "learning_rate": 1.516883840179215e-06,
+      "loss": 1.0518,
+      "step": 9700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8391316462603355,
+      "learning_rate": 1.515420748281905e-06,
+      "loss": 1.1172,
+      "step": 9701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.154972743473802,
+      "learning_rate": 1.5139583044790907e-06,
+      "loss": 1.1284,
+      "step": 9702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.789113249971505,
+      "learning_rate": 1.5124965088824795e-06,
+      "loss": 1.0657,
+      "step": 9703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.212325874220939,
+      "learning_rate": 1.5110353616037332e-06,
+      "loss": 1.0574,
+      "step": 9704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0710156688241597,
+      "learning_rate": 1.5095748627544626e-06,
+      "loss": 1.1072,
+      "step": 9705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.155201645462084,
+      "learning_rate": 1.508115012446223e-06,
+      "loss": 1.0759,
+      "step": 9706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6245095459685093,
+      "learning_rate": 1.5066558107905328e-06,
+      "loss": 1.0342,
+      "step": 9707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2665162408916264,
+      "learning_rate": 1.5051972578988427e-06,
+      "loss": 1.0801,
+      "step": 9708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2296642059248333,
+      "learning_rate": 1.5037393538825728e-06,
+      "loss": 1.0605,
+      "step": 9709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2977796546161056,
+      "learning_rate": 1.502282098853083e-06,
+      "loss": 1.0422,
+      "step": 9710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.028348278924013,
+      "learning_rate": 1.5008254929216826e-06,
+      "loss": 1.03,
+      "step": 9711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.847975948588817,
+      "learning_rate": 1.4993695361996374e-06,
+      "loss": 1.0591,
+      "step": 9712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.498559826712663,
+      "learning_rate": 1.497914228798162e-06,
+      "loss": 1.0747,
+      "step": 9713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.078054496858627,
+      "learning_rate": 1.496459570828417e-06,
+      "loss": 1.0635,
+      "step": 9714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1872051094716314,
+      "learning_rate": 1.4950055624015147e-06,
+      "loss": 1.1111,
+      "step": 9715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.871159965536575,
+      "learning_rate": 1.4935522036285255e-06,
+      "loss": 1.0103,
+      "step": 9716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.338177505765822,
+      "learning_rate": 1.4920994946204603e-06,
+      "loss": 1.0564,
+      "step": 9717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7446514266497504,
+      "learning_rate": 1.4906474354882827e-06,
+      "loss": 1.1533,
+      "step": 9718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8272508372076337,
+      "learning_rate": 1.489196026342914e-06,
+      "loss": 1.0122,
+      "step": 9719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6221917252162172,
+      "learning_rate": 1.4877452672952153e-06,
+      "loss": 1.1089,
+      "step": 9720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9958452854680195,
+      "learning_rate": 1.4862951584560037e-06,
+      "loss": 1.0413,
+      "step": 9721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6060061523098117,
+      "learning_rate": 1.4848456999360472e-06,
+      "loss": 1.1135,
+      "step": 9722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.294945650277798,
+      "learning_rate": 1.4833968918460583e-06,
+      "loss": 1.0562,
+      "step": 9723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.374098953514376,
+      "learning_rate": 1.4819487342967098e-06,
+      "loss": 1.0657,
+      "step": 9724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.326391121307453,
+      "learning_rate": 1.4805012273986175e-06,
+      "loss": 1.0342,
+      "step": 9725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1285192836558373,
+      "learning_rate": 1.4790543712623461e-06,
+      "loss": 1.0281,
+      "step": 9726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3638769454628474,
+      "learning_rate": 1.477608165998421e-06,
+      "loss": 1.1167,
+      "step": 9727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.088946901882967,
+      "learning_rate": 1.4761626117173e-06,
+      "loss": 0.9551,
+      "step": 9728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3130228542584486,
+      "learning_rate": 1.4747177085294107e-06,
+      "loss": 1.1069,
+      "step": 9729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.719470404045221,
+      "learning_rate": 1.473273456545119e-06,
+      "loss": 1.0139,
+      "step": 9730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.167267335436925,
+      "learning_rate": 1.47182985587474e-06,
+      "loss": 1.0576,
+      "step": 9731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1340564005707097,
+      "learning_rate": 1.4703869066285515e-06,
+      "loss": 1.0828,
+      "step": 9732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1374469086488426,
+      "learning_rate": 1.468944608916767e-06,
+      "loss": 1.0435,
+      "step": 9733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5858344609830946,
+      "learning_rate": 1.4675029628495585e-06,
+      "loss": 1.1299,
+      "step": 9734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9059712733925327,
+      "learning_rate": 1.466061968537046e-06,
+      "loss": 1.0376,
+      "step": 9735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2102773173619887,
+      "learning_rate": 1.4646216260892966e-06,
+      "loss": 1.0461,
+      "step": 9736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2527062449705433,
+      "learning_rate": 1.4631819356163357e-06,
+      "loss": 1.073,
+      "step": 9737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8902254015281468,
+      "learning_rate": 1.4617428972281333e-06,
+      "loss": 1.0308,
+      "step": 9738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.318290266431005,
+      "learning_rate": 1.460304511034606e-06,
+      "loss": 1.0725,
+      "step": 9739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7397316066492854,
+      "learning_rate": 1.4588667771456323e-06,
+      "loss": 1.0449,
+      "step": 9740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0121354653280137,
+      "learning_rate": 1.4574296956710255e-06,
+      "loss": 1.0847,
+      "step": 9741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2675213925714615,
+      "learning_rate": 1.4559932667205623e-06,
+      "loss": 1.0959,
+      "step": 9742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2305900374814644,
+      "learning_rate": 1.4545574904039627e-06,
+      "loss": 1.1179,
+      "step": 9743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1907378855136246,
+      "learning_rate": 1.453122366830897e-06,
+      "loss": 1.0579,
+      "step": 9744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2621312957688824,
+      "learning_rate": 1.4516878961109905e-06,
+      "loss": 1.0706,
+      "step": 9745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.991487434073372,
+      "learning_rate": 1.4502540783538111e-06,
+      "loss": 1.0101,
+      "step": 9746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.822087061984978,
+      "learning_rate": 1.4488209136688879e-06,
+      "loss": 1.0454,
+      "step": 9747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.481368706998631,
+      "learning_rate": 1.447388402165686e-06,
+      "loss": 1.0559,
+      "step": 9748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.988231760905494,
+      "learning_rate": 1.4459565439536272e-06,
+      "loss": 1.0632,
+      "step": 9749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.825866057791903,
+      "learning_rate": 1.444525339142091e-06,
+      "loss": 1.0166,
+      "step": 9750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8725379553725716,
+      "learning_rate": 1.4430947878403922e-06,
+      "loss": 1.0959,
+      "step": 9751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.404290030068209,
+      "learning_rate": 1.44166489015781e-06,
+      "loss": 1.0303,
+      "step": 9752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3092255760333926,
+      "learning_rate": 1.4402356462035671e-06,
+      "loss": 1.0005,
+      "step": 9753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.386132175611401,
+      "learning_rate": 1.4388070560868284e-06,
+      "loss": 1.0959,
+      "step": 9754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2211737695153033,
+      "learning_rate": 1.4373791199167242e-06,
+      "loss": 1.052,
+      "step": 9755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8251437667241284,
+      "learning_rate": 1.435951837802323e-06,
+      "loss": 0.9952,
+      "step": 9756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.996420094622724,
+      "learning_rate": 1.4345252098526518e-06,
+      "loss": 1.0613,
+      "step": 9757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3668906463796255,
+      "learning_rate": 1.4330992361766827e-06,
+      "loss": 0.9783,
+      "step": 9758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2225011348962824,
+      "learning_rate": 1.4316739168833338e-06,
+      "loss": 1.1057,
+      "step": 9759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.95818720141691,
+      "learning_rate": 1.430249252081487e-06,
+      "loss": 1.0391,
+      "step": 9760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2331799962901133,
+      "learning_rate": 1.4288252418799554e-06,
+      "loss": 1.115,
+      "step": 9761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2283231498428386,
+      "learning_rate": 1.4274018863875194e-06,
+      "loss": 1.1003,
+      "step": 9762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8323510934643528,
+      "learning_rate": 1.4259791857128991e-06,
+      "loss": 1.0688,
+      "step": 9763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.123713621825803,
+      "learning_rate": 1.4245571399647662e-06,
+      "loss": 1.0996,
+      "step": 9764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6106315355394867,
+      "learning_rate": 1.4231357492517462e-06,
+      "loss": 1.1135,
+      "step": 9765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9805300674718,
+      "learning_rate": 1.4217150136824121e-06,
+      "loss": 1.0186,
+      "step": 9766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0430078073952758,
+      "learning_rate": 1.420294933365285e-06,
+      "loss": 1.0625,
+      "step": 9767
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.169819075189091,
+      "learning_rate": 1.418875508408838e-06,
+      "loss": 1.0662,
+      "step": 9768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3469670648611323,
+      "learning_rate": 1.4174567389214922e-06,
+      "loss": 1.1028,
+      "step": 9769
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.190578408492531,
+      "learning_rate": 1.4160386250116254e-06,
+      "loss": 1.0259,
+      "step": 9770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.459588129639718,
+      "learning_rate": 1.4146211667875565e-06,
+      "loss": 1.0093,
+      "step": 9771
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.251704716287609,
+      "learning_rate": 1.4132043643575566e-06,
+      "loss": 1.0825,
+      "step": 9772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8414903899877983,
+      "learning_rate": 1.4117882178298548e-06,
+      "loss": 1.0082,
+      "step": 9773
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.372982065801896,
+      "learning_rate": 1.4103727273126133e-06,
+      "loss": 1.0374,
+      "step": 9774
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.119609831735114,
+      "learning_rate": 1.4089578929139635e-06,
+      "loss": 1.0798,
+      "step": 9775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.863134380980078,
+      "learning_rate": 1.4075437147419736e-06,
+      "loss": 1.0938,
+      "step": 9776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2859273215685865,
+      "learning_rate": 1.4061301929046634e-06,
+      "loss": 1.0352,
+      "step": 9777
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.069592069230612,
+      "learning_rate": 1.404717327510009e-06,
+      "loss": 0.9846,
+      "step": 9778
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.48854808731913,
+      "learning_rate": 1.4033051186659285e-06,
+      "loss": 1.0989,
+      "step": 9779
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2229828254166346,
+      "learning_rate": 1.4018935664803001e-06,
+      "loss": 1.0886,
+      "step": 9780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2877293942344425,
+      "learning_rate": 1.4004826710609376e-06,
+      "loss": 1.0388,
+      "step": 9781
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6013701830648284,
+      "learning_rate": 1.3990724325156135e-06,
+      "loss": 1.135,
+      "step": 9782
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.994522663116925,
+      "learning_rate": 1.3976628509520518e-06,
+      "loss": 1.0527,
+      "step": 9783
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.941149325345776,
+      "learning_rate": 1.3962539264779206e-06,
+      "loss": 1.001,
+      "step": 9784
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.014681707636681,
+      "learning_rate": 1.3948456592008441e-06,
+      "loss": 1.0139,
+      "step": 9785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3161998932826013,
+      "learning_rate": 1.3934380492283916e-06,
+      "loss": 1.0715,
+      "step": 9786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.071001164144084,
+      "learning_rate": 1.3920310966680794e-06,
+      "loss": 1.082,
+      "step": 9787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.292575405322022,
+      "learning_rate": 1.3906248016273816e-06,
+      "loss": 1.0784,
+      "step": 9788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.808722900743983,
+      "learning_rate": 1.3892191642137153e-06,
+      "loss": 1.103,
+      "step": 9789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2366655038049035,
+      "learning_rate": 1.387814184534454e-06,
+      "loss": 1.0801,
+      "step": 9790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8582762606399,
+      "learning_rate": 1.3864098626969147e-06,
+      "loss": 1.0674,
+      "step": 9791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.114101712377869,
+      "learning_rate": 1.3850061988083652e-06,
+      "loss": 1.115,
+      "step": 9792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0391633002785663,
+      "learning_rate": 1.3836031929760296e-06,
+      "loss": 1.0464,
+      "step": 9793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1194145268094373,
+      "learning_rate": 1.382200845307068e-06,
+      "loss": 1.0334,
+      "step": 9794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2272781060257807,
+      "learning_rate": 1.3807991559086054e-06,
+      "loss": 1.0657,
+      "step": 9795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4224865125571227,
+      "learning_rate": 1.3793981248877086e-06,
+      "loss": 1.0603,
+      "step": 9796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.085787670204248,
+      "learning_rate": 1.3779977523513921e-06,
+      "loss": 1.0151,
+      "step": 9797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0252109618000453,
+      "learning_rate": 1.376598038406628e-06,
+      "loss": 1.0881,
+      "step": 9798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2824980369698054,
+      "learning_rate": 1.3751989831603308e-06,
+      "loss": 1.0083,
+      "step": 9799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0419373926000626,
+      "learning_rate": 1.3738005867193682e-06,
+      "loss": 1.1245,
+      "step": 9800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0174524810398244,
+      "learning_rate": 1.3724028491905562e-06,
+      "loss": 1.0513,
+      "step": 9801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5060570935995976,
+      "learning_rate": 1.371005770680659e-06,
+      "loss": 0.9629,
+      "step": 9802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.23372813907437,
+      "learning_rate": 1.369609351296396e-06,
+      "loss": 1.1047,
+      "step": 9803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.822271767942878,
+      "learning_rate": 1.3682135911444316e-06,
+      "loss": 1.0503,
+      "step": 9804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6590572092471243,
+      "learning_rate": 1.3668184903313785e-06,
+      "loss": 1.0466,
+      "step": 9805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.544917125869491,
+      "learning_rate": 1.3654240489638071e-06,
+      "loss": 1.0723,
+      "step": 9806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.753034175919093,
+      "learning_rate": 1.3640302671482242e-06,
+      "loss": 0.9902,
+      "step": 9807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2759423308061457,
+      "learning_rate": 1.3626371449911003e-06,
+      "loss": 1.0676,
+      "step": 9808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4132617315772005,
+      "learning_rate": 1.3612446825988456e-06,
+      "loss": 1.1138,
+      "step": 9809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.735270784913333,
+      "learning_rate": 1.359852880077822e-06,
+      "loss": 1.0925,
+      "step": 9810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.476721752462556,
+      "learning_rate": 1.3584617375343478e-06,
+      "loss": 0.9861,
+      "step": 9811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.193241538489721,
+      "learning_rate": 1.3570712550746811e-06,
+      "loss": 1.0664,
+      "step": 9812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9069122045027815,
+      "learning_rate": 1.355681432805035e-06,
+      "loss": 1.0669,
+      "step": 9813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.436240589337153,
+      "learning_rate": 1.3542922708315708e-06,
+      "loss": 1.0859,
+      "step": 9814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.605664778073155,
+      "learning_rate": 1.3529037692603974e-06,
+      "loss": 1.0957,
+      "step": 9815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4286589312975657,
+      "learning_rate": 1.3515159281975809e-06,
+      "loss": 1.0645,
+      "step": 9816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6979488882312537,
+      "learning_rate": 1.3501287477491255e-06,
+      "loss": 0.9849,
+      "step": 9817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8026719829047875,
+      "learning_rate": 1.348742228020996e-06,
+      "loss": 0.97,
+      "step": 9818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.694967765288828,
+      "learning_rate": 1.3473563691191028e-06,
+      "loss": 1.0115,
+      "step": 9819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7449017285559805,
+      "learning_rate": 1.3459711711492962e-06,
+      "loss": 1.0872,
+      "step": 9820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1982638011813562,
+      "learning_rate": 1.344586634217393e-06,
+      "loss": 1.1494,
+      "step": 9821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.516268272490789,
+      "learning_rate": 1.3432027584291441e-06,
+      "loss": 1.0371,
+      "step": 9822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.076167131677634,
+      "learning_rate": 1.3418195438902637e-06,
+      "loss": 1.0527,
+      "step": 9823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9970861176912735,
+      "learning_rate": 1.3404369907064063e-06,
+      "loss": 1.0222,
+      "step": 9824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.139050654386711,
+      "learning_rate": 1.339055098983174e-06,
+      "loss": 1.0928,
+      "step": 9825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1631637368169714,
+      "learning_rate": 1.3376738688261316e-06,
+      "loss": 1.0669,
+      "step": 9826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4186446441213674,
+      "learning_rate": 1.3362933003407741e-06,
+      "loss": 1.1169,
+      "step": 9827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2213073949992426,
+      "learning_rate": 1.334913393632563e-06,
+      "loss": 1.0269,
+      "step": 9828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.272656392351085,
+      "learning_rate": 1.3335341488068997e-06,
+      "loss": 1.1047,
+      "step": 9829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3134829612184458,
+      "learning_rate": 1.3321555659691376e-06,
+      "loss": 1.0793,
+      "step": 9830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4453574420145134,
+      "learning_rate": 1.3307776452245825e-06,
+      "loss": 1.0583,
+      "step": 9831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0228227508183885,
+      "learning_rate": 1.3294003866784845e-06,
+      "loss": 1.0986,
+      "step": 9832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1861314472540947,
+      "learning_rate": 1.328023790436046e-06,
+      "loss": 1.0515,
+      "step": 9833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.028005428682723,
+      "learning_rate": 1.3266478566024178e-06,
+      "loss": 1.0791,
+      "step": 9834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4396665150582857,
+      "learning_rate": 1.3252725852826998e-06,
+      "loss": 1.0361,
+      "step": 9835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0179592995864457,
+      "learning_rate": 1.3238979765819449e-06,
+      "loss": 1.1018,
+      "step": 9836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1022415407961366,
+      "learning_rate": 1.3225240306051512e-06,
+      "loss": 1.0764,
+      "step": 9837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.348848962246209,
+      "learning_rate": 1.321150747457265e-06,
+      "loss": 1.1504,
+      "step": 9838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2597149990143515,
+      "learning_rate": 1.319778127243192e-06,
+      "loss": 1.0303,
+      "step": 9839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1666325764333436,
+      "learning_rate": 1.3184061700677697e-06,
+      "loss": 0.9932,
+      "step": 9840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.913556845747275,
+      "learning_rate": 1.3170348760358021e-06,
+      "loss": 1.0566,
+      "step": 9841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8453433807435724,
+      "learning_rate": 1.315664245252033e-06,
+      "loss": 1.0774,
+      "step": 9842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.487712784552647,
+      "learning_rate": 1.3142942778211565e-06,
+      "loss": 1.0374,
+      "step": 9843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.889659452916918,
+      "learning_rate": 1.3129249738478222e-06,
+      "loss": 1.0684,
+      "step": 9844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8731672803426074,
+      "learning_rate": 1.311556333436621e-06,
+      "loss": 1.0376,
+      "step": 9845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.193208565781888,
+      "learning_rate": 1.310188356692097e-06,
+      "loss": 1.0837,
+      "step": 9846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.446235073139081,
+      "learning_rate": 1.308821043718743e-06,
+      "loss": 1.0833,
+      "step": 9847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1984552575982566,
+      "learning_rate": 1.3074543946210006e-06,
+      "loss": 1.0413,
+      "step": 9848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8905600540933585,
+      "learning_rate": 1.3060884095032622e-06,
+      "loss": 1.1067,
+      "step": 9849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6919853545252126,
+      "learning_rate": 1.3047230884698702e-06,
+      "loss": 1.0056,
+      "step": 9850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.160023277860985,
+      "learning_rate": 1.303358431625109e-06,
+      "loss": 1.1145,
+      "step": 9851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.248986808044665,
+      "learning_rate": 1.3019944390732254e-06,
+      "loss": 1.1133,
+      "step": 9852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.244908443705892,
+      "learning_rate": 1.3006311109184034e-06,
+      "loss": 1.0601,
+      "step": 9853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.402620431038664,
+      "learning_rate": 1.299268447264782e-06,
+      "loss": 1.0676,
+      "step": 9854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2814230223291947,
+      "learning_rate": 1.2979064482164449e-06,
+      "loss": 1.0803,
+      "step": 9855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4910686351878537,
+      "learning_rate": 1.2965451138774343e-06,
+      "loss": 1.0276,
+      "step": 9856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1138140765846183,
+      "learning_rate": 1.2951844443517336e-06,
+      "loss": 1.0376,
+      "step": 9857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.878648747550722,
+      "learning_rate": 1.2938244397432731e-06,
+      "loss": 1.0649,
+      "step": 9858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.957496412390602,
+      "learning_rate": 1.2924651001559463e-06,
+      "loss": 1.0413,
+      "step": 9859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.842041541739242,
+      "learning_rate": 1.2911064256935746e-06,
+      "loss": 1.0649,
+      "step": 9860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4296609291240028,
+      "learning_rate": 1.2897484164599495e-06,
+      "loss": 1.0542,
+      "step": 9861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0386968058004618,
+      "learning_rate": 1.2883910725587989e-06,
+      "loss": 1.0281,
+      "step": 9862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9866341104483545,
+      "learning_rate": 1.2870343940938012e-06,
+      "loss": 1.0862,
+      "step": 9863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0417067977125485,
+      "learning_rate": 1.2856783811685914e-06,
+      "loss": 1.04,
+      "step": 9864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9093538599034994,
+      "learning_rate": 1.2843230338867452e-06,
+      "loss": 1.0088,
+      "step": 9865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8721068528129545,
+      "learning_rate": 1.2829683523517923e-06,
+      "loss": 1.0349,
+      "step": 9866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2210292059743844,
+      "learning_rate": 1.2816143366672096e-06,
+      "loss": 1.0515,
+      "step": 9867
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9957554136894213,
+      "learning_rate": 1.28026098693642e-06,
+      "loss": 1.0107,
+      "step": 9868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.993787582166295,
+      "learning_rate": 1.2789083032628058e-06,
+      "loss": 1.0972,
+      "step": 9869
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5481833684206485,
+      "learning_rate": 1.277556285749687e-06,
+      "loss": 1.0256,
+      "step": 9870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.376314136468227,
+      "learning_rate": 1.276204934500338e-06,
+      "loss": 1.0679,
+      "step": 9871
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.199944991285287,
+      "learning_rate": 1.2748542496179838e-06,
+      "loss": 1.0596,
+      "step": 9872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2443684339201964,
+      "learning_rate": 1.2735042312057955e-06,
+      "loss": 1.0901,
+      "step": 9873
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.114299970211698,
+      "learning_rate": 1.2721548793668936e-06,
+      "loss": 1.0459,
+      "step": 9874
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0252469406091578,
+      "learning_rate": 1.2708061942043492e-06,
+      "loss": 1.0095,
+      "step": 9875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.034341184970191,
+      "learning_rate": 1.2694581758211788e-06,
+      "loss": 1.0613,
+      "step": 9876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.060667258191309,
+      "learning_rate": 1.2681108243203543e-06,
+      "loss": 1.0105,
+      "step": 9877
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.319767119411335,
+      "learning_rate": 1.2667641398047926e-06,
+      "loss": 1.0046,
+      "step": 9878
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.992169596142902,
+      "learning_rate": 1.2654181223773598e-06,
+      "loss": 1.0188,
+      "step": 9879
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6857618344714416,
+      "learning_rate": 1.2640727721408696e-06,
+      "loss": 1.0293,
+      "step": 9880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.947448194731672,
+      "learning_rate": 1.2627280891980875e-06,
+      "loss": 1.0706,
+      "step": 9881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9240523916758567,
+      "learning_rate": 1.261384073651729e-06,
+      "loss": 1.0144,
+      "step": 9882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6452881741940266,
+      "learning_rate": 1.2600407256044544e-06,
+      "loss": 0.9954,
+      "step": 9883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0548270962238124,
+      "learning_rate": 1.258698045158876e-06,
+      "loss": 1.0381,
+      "step": 9884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.105731685171631,
+      "learning_rate": 1.2573560324175548e-06,
+      "loss": 1.0552,
+      "step": 9885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4976759341968178,
+      "learning_rate": 1.2560146874830014e-06,
+      "loss": 1.0022,
+      "step": 9886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.397134195270029,
+      "learning_rate": 1.2546740104576737e-06,
+      "loss": 1.064,
+      "step": 9887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9888489453553553,
+      "learning_rate": 1.2533340014439788e-06,
+      "loss": 1.0032,
+      "step": 9888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.482088420021789,
+      "learning_rate": 1.25199466054427e-06,
+      "loss": 1.1057,
+      "step": 9889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.753657164959443,
+      "learning_rate": 1.2506559878608593e-06,
+      "loss": 1.0664,
+      "step": 9890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0580527997433022,
+      "learning_rate": 1.2493179834959956e-06,
+      "loss": 1.0376,
+      "step": 9891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1999215723998398,
+      "learning_rate": 1.2479806475518896e-06,
+      "loss": 1.0623,
+      "step": 9892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.069293065701701,
+      "learning_rate": 1.2466439801306874e-06,
+      "loss": 1.0627,
+      "step": 9893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.527085146000009,
+      "learning_rate": 1.2453079813344893e-06,
+      "loss": 1.073,
+      "step": 9894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.119503778505777,
+      "learning_rate": 1.2439726512653495e-06,
+      "loss": 1.0081,
+      "step": 9895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.624110853379059,
+      "learning_rate": 1.242637990025264e-06,
+      "loss": 1.0867,
+      "step": 9896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.96575996804167,
+      "learning_rate": 1.2413039977161855e-06,
+      "loss": 1.0771,
+      "step": 9897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9138212014589167,
+      "learning_rate": 1.2399706744400075e-06,
+      "loss": 0.9973,
+      "step": 9898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.173294348067038,
+      "learning_rate": 1.2386380202985772e-06,
+      "loss": 1.0139,
+      "step": 9899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2083473782114544,
+      "learning_rate": 1.2373060353936884e-06,
+      "loss": 1.0664,
+      "step": 9900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0145686922656867,
+      "learning_rate": 1.2359747198270832e-06,
+      "loss": 0.9626,
+      "step": 9901
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.507292574888857,
+      "learning_rate": 1.2346440737004584e-06,
+      "loss": 1.0825,
+      "step": 9902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.520583597576482,
+      "learning_rate": 1.2333140971154522e-06,
+      "loss": 1.0352,
+      "step": 9903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0741089768090495,
+      "learning_rate": 1.2319847901736537e-06,
+      "loss": 1.0198,
+      "step": 9904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2876529038117255,
+      "learning_rate": 1.230656152976607e-06,
+      "loss": 0.9963,
+      "step": 9905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1399027892161246,
+      "learning_rate": 1.2293281856257954e-06,
+      "loss": 0.9736,
+      "step": 9906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5289969661875693,
+      "learning_rate": 1.228000888222659e-06,
+      "loss": 1.0745,
+      "step": 9907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8137415054776986,
+      "learning_rate": 1.22667426086858e-06,
+      "loss": 1.1106,
+      "step": 9908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.895875008696644,
+      "learning_rate": 1.225348303664894e-06,
+      "loss": 1.0696,
+      "step": 9909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.952923405874174,
+      "learning_rate": 1.2240230167128863e-06,
+      "loss": 1.0796,
+      "step": 9910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.300809608639554,
+      "learning_rate": 1.2226984001137865e-06,
+      "loss": 1.0532,
+      "step": 9911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1404857878944283,
+      "learning_rate": 1.2213744539687777e-06,
+      "loss": 1.0627,
+      "step": 9912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.290677450772201,
+      "learning_rate": 1.220051178378987e-06,
+      "loss": 1.082,
+      "step": 9913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3715363066415818,
+      "learning_rate": 1.2187285734454924e-06,
+      "loss": 1.094,
+      "step": 9914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0632812893320724,
+      "learning_rate": 1.2174066392693241e-06,
+      "loss": 1.0144,
+      "step": 9915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8521167582895033,
+      "learning_rate": 1.2160853759514567e-06,
+      "loss": 1.041,
+      "step": 9916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.329898462874623,
+      "learning_rate": 1.2147647835928121e-06,
+      "loss": 1.0989,
+      "step": 9917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7987435523430495,
+      "learning_rate": 1.2134448622942696e-06,
+      "loss": 1.0029,
+      "step": 9918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5857848199065447,
+      "learning_rate": 1.212125612156647e-06,
+      "loss": 1.072,
+      "step": 9919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.473006392002625,
+      "learning_rate": 1.2108070332807164e-06,
+      "loss": 1.0837,
+      "step": 9920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.909365381154904,
+      "learning_rate": 1.209489125767197e-06,
+      "loss": 1.0488,
+      "step": 9921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.176429023667369,
+      "learning_rate": 1.208171889716756e-06,
+      "loss": 1.0684,
+      "step": 9922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2639453994902947,
+      "learning_rate": 1.2068553252300142e-06,
+      "loss": 1.1255,
+      "step": 9923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3324771991598174,
+      "learning_rate": 1.205539432407532e-06,
+      "loss": 1.0928,
+      "step": 9924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0865479435165417,
+      "learning_rate": 1.2042242113498304e-06,
+      "loss": 1.0256,
+      "step": 9925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5125596643578185,
+      "learning_rate": 1.2029096621573699e-06,
+      "loss": 1.0435,
+      "step": 9926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3252585579407263,
+      "learning_rate": 1.2015957849305582e-06,
+      "loss": 1.0757,
+      "step": 9927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6962174765656215,
+      "learning_rate": 1.2002825797697604e-06,
+      "loss": 1.0283,
+      "step": 9928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.270808762918833,
+      "learning_rate": 1.198970046775283e-06,
+      "loss": 1.0986,
+      "step": 9929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.322006049070919,
+      "learning_rate": 1.1976581860473868e-06,
+      "loss": 1.103,
+      "step": 9930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6493866389486724,
+      "learning_rate": 1.1963469976862773e-06,
+      "loss": 1.0034,
+      "step": 9931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9721418998926166,
+      "learning_rate": 1.1950364817921089e-06,
+      "loss": 1.0261,
+      "step": 9932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5564414024331534,
+      "learning_rate": 1.1937266384649848e-06,
+      "loss": 1.1052,
+      "step": 9933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5332354629558402,
+      "learning_rate": 1.1924174678049561e-06,
+      "loss": 1.0474,
+      "step": 9934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.360682674672364,
+      "learning_rate": 1.1911089699120283e-06,
+      "loss": 1.0398,
+      "step": 9935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4443046897204055,
+      "learning_rate": 1.189801144886148e-06,
+      "loss": 1.0901,
+      "step": 9936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2574113477634468,
+      "learning_rate": 1.188493992827211e-06,
+      "loss": 1.1177,
+      "step": 9937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2193412119169516,
+      "learning_rate": 1.1871875138350696e-06,
+      "loss": 1.0339,
+      "step": 9938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5050819939046014,
+      "learning_rate": 1.1858817080095165e-06,
+      "loss": 1.0295,
+      "step": 9939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.168901897244699,
+      "learning_rate": 1.1845765754502957e-06,
+      "loss": 1.0642,
+      "step": 9940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6702501909455347,
+      "learning_rate": 1.1832721162570993e-06,
+      "loss": 1.0469,
+      "step": 9941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.924227170335885,
+      "learning_rate": 1.181968330529567e-06,
+      "loss": 1.0447,
+      "step": 9942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8545142175200846,
+      "learning_rate": 1.1806652183672917e-06,
+      "loss": 1.0312,
+      "step": 9943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2820777456408448,
+      "learning_rate": 1.1793627798698104e-06,
+      "loss": 0.9968,
+      "step": 9944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9466184229533567,
+      "learning_rate": 1.17806101513661e-06,
+      "loss": 1.0194,
+      "step": 9945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.34787541153916,
+      "learning_rate": 1.1767599242671245e-06,
+      "loss": 1.0625,
+      "step": 9946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2801998011769684,
+      "learning_rate": 1.1754595073607367e-06,
+      "loss": 1.0232,
+      "step": 9947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.608694404445133,
+      "learning_rate": 1.1741597645167836e-06,
+      "loss": 1.0833,
+      "step": 9948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9446233874773955,
+      "learning_rate": 1.1728606958345422e-06,
+      "loss": 1.001,
+      "step": 9949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5693864629838505,
+      "learning_rate": 1.1715623014132416e-06,
+      "loss": 0.9933,
+      "step": 9950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2853158500687054,
+      "learning_rate": 1.170264581352063e-06,
+      "loss": 1.0549,
+      "step": 9951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.950559672035917,
+      "learning_rate": 1.1689675357501307e-06,
+      "loss": 1.0259,
+      "step": 9952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9049151922588745,
+      "learning_rate": 1.1676711647065198e-06,
+      "loss": 0.9988,
+      "step": 9953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.680996130437701,
+      "learning_rate": 1.1663754683202523e-06,
+      "loss": 1.05,
+      "step": 9954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.221397648819374,
+      "learning_rate": 1.1650804466902987e-06,
+      "loss": 1.137,
+      "step": 9955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.088521942475199,
+      "learning_rate": 1.1637860999155849e-06,
+      "loss": 1.0396,
+      "step": 9956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.252763848728393,
+      "learning_rate": 1.1624924280949735e-06,
+      "loss": 1.0791,
+      "step": 9957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.115580237481001,
+      "learning_rate": 1.1611994313272856e-06,
+      "loss": 1.0999,
+      "step": 9958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.014771577138511,
+      "learning_rate": 1.159907109711288e-06,
+      "loss": 0.9773,
+      "step": 9959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3028235154865366,
+      "learning_rate": 1.158615463345687e-06,
+      "loss": 1.1082,
+      "step": 9960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.617888629733071,
+      "learning_rate": 1.1573244923291526e-06,
+      "loss": 1.0198,
+      "step": 9961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2338783375503146,
+      "learning_rate": 1.1560341967602917e-06,
+      "loss": 1.0681,
+      "step": 9962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8460738349457455,
+      "learning_rate": 1.1547445767376664e-06,
+      "loss": 1.0984,
+      "step": 9963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.279665896874216,
+      "learning_rate": 1.1534556323597824e-06,
+      "loss": 1.0474,
+      "step": 9964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8339520678660333,
+      "learning_rate": 1.1521673637250962e-06,
+      "loss": 1.1001,
+      "step": 9965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1350628240322878,
+      "learning_rate": 1.150879770932013e-06,
+      "loss": 1.0264,
+      "step": 9966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4363710451497425,
+      "learning_rate": 1.1495928540788815e-06,
+      "loss": 1.0137,
+      "step": 9967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9435921816146133,
+      "learning_rate": 1.1483066132640075e-06,
+      "loss": 1.0901,
+      "step": 9968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.409335829722202,
+      "learning_rate": 1.1470210485856404e-06,
+      "loss": 1.0979,
+      "step": 9969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6139343247596614,
+      "learning_rate": 1.1457361601419737e-06,
+      "loss": 1.0715,
+      "step": 9970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4722110719866004,
+      "learning_rate": 1.144451948031159e-06,
+      "loss": 1.0874,
+      "step": 9971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3634878240784114,
+      "learning_rate": 1.1431684123512875e-06,
+      "loss": 1.0757,
+      "step": 9972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.440384529520141,
+      "learning_rate": 1.1418855532004025e-06,
+      "loss": 1.0542,
+      "step": 9973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9810184747360537,
+      "learning_rate": 1.1406033706764963e-06,
+      "loss": 1.0259,
+      "step": 9974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.430001632369471,
+      "learning_rate": 1.1393218648775062e-06,
+      "loss": 1.0903,
+      "step": 9975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2421679243393027,
+      "learning_rate": 1.1380410359013227e-06,
+      "loss": 1.0024,
+      "step": 9976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6206033243647555,
+      "learning_rate": 1.13676088384578e-06,
+      "loss": 1.0537,
+      "step": 9977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3078288659449737,
+      "learning_rate": 1.1354814088086618e-06,
+      "loss": 1.0171,
+      "step": 9978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0730915801646757,
+      "learning_rate": 1.1342026108877069e-06,
+      "loss": 1.0112,
+      "step": 9979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.736888086399461,
+      "learning_rate": 1.1329244901805869e-06,
+      "loss": 1.0786,
+      "step": 9980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.122744604510162,
+      "learning_rate": 1.1316470467849373e-06,
+      "loss": 1.0364,
+      "step": 9981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.612631132303673,
+      "learning_rate": 1.130370280798334e-06,
+      "loss": 1.1052,
+      "step": 9982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0796669345083894,
+      "learning_rate": 1.1290941923183008e-06,
+      "loss": 1.0447,
+      "step": 9983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.99841688766459,
+      "learning_rate": 1.1278187814423158e-06,
+      "loss": 1.0222,
+      "step": 9984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5927977469104793,
+      "learning_rate": 1.1265440482677993e-06,
+      "loss": 1.0264,
+      "step": 9985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.933572456345793,
+      "learning_rate": 1.1252699928921208e-06,
+      "loss": 1.0466,
+      "step": 9986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4229457026009706,
+      "learning_rate": 1.1239966154125991e-06,
+      "loss": 1.0789,
+      "step": 9987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9429107291911614,
+      "learning_rate": 1.1227239159265013e-06,
+      "loss": 0.9253,
+      "step": 9988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.836817466368341,
+      "learning_rate": 1.121451894531045e-06,
+      "loss": 1.0537,
+      "step": 9989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5596993827181955,
+      "learning_rate": 1.12018055132339e-06,
+      "loss": 1.1013,
+      "step": 9990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4503512162378964,
+      "learning_rate": 1.1189098864006488e-06,
+      "loss": 1.0518,
+      "step": 9991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5108104876903012,
+      "learning_rate": 1.1176398998598847e-06,
+      "loss": 1.1204,
+      "step": 9992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1571307271010363,
+      "learning_rate": 1.1163705917980994e-06,
+      "loss": 1.083,
+      "step": 9993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.009351847182989,
+      "learning_rate": 1.1151019623122539e-06,
+      "loss": 1.0164,
+      "step": 9994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.903025299388763,
+      "learning_rate": 1.1138340114992508e-06,
+      "loss": 1.0366,
+      "step": 9995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4509995480363007,
+      "learning_rate": 1.1125667394559404e-06,
+      "loss": 1.05,
+      "step": 9996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7819878425655338,
+      "learning_rate": 1.1113001462791261e-06,
+      "loss": 1.1028,
+      "step": 9997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3927474547905354,
+      "learning_rate": 1.1100342320655555e-06,
+      "loss": 1.0959,
+      "step": 9998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3419607613897178,
+      "learning_rate": 1.1087689969119286e-06,
+      "loss": 1.094,
+      "step": 9999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6414293093463757,
+      "learning_rate": 1.107504440914885e-06,
+      "loss": 1.0479,
+      "step": 10000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6043682394236956,
+      "learning_rate": 1.1062405641710205e-06,
+      "loss": 1.0537,
+      "step": 10001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8625127124744854,
+      "learning_rate": 1.104977366776877e-06,
+      "loss": 1.0513,
+      "step": 10002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.281251159310136,
+      "learning_rate": 1.1037148488289418e-06,
+      "loss": 1.0334,
+      "step": 10003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4688325033922514,
+      "learning_rate": 1.1024530104236542e-06,
+      "loss": 1.0896,
+      "step": 10004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4423091260991088,
+      "learning_rate": 1.1011918516573993e-06,
+      "loss": 1.0872,
+      "step": 10005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3681841935672985,
+      "learning_rate": 1.0999313726265104e-06,
+      "loss": 1.0581,
+      "step": 10006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2993668465839656,
+      "learning_rate": 1.0986715734272702e-06,
+      "loss": 1.124,
+      "step": 10007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3016301932630117,
+      "learning_rate": 1.097412454155904e-06,
+      "loss": 1.0693,
+      "step": 10008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3239591940677524,
+      "learning_rate": 1.0961540149085958e-06,
+      "loss": 1.0361,
+      "step": 10009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.503867444702761,
+      "learning_rate": 1.094896255781468e-06,
+      "loss": 1.0601,
+      "step": 10010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4285383755549605,
+      "learning_rate": 1.0936391768705945e-06,
+      "loss": 1.0735,
+      "step": 10011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5211098009454203,
+      "learning_rate": 1.0923827782720008e-06,
+      "loss": 1.1094,
+      "step": 10012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1628250625149983,
+      "learning_rate": 1.0911270600816503e-06,
+      "loss": 1.0383,
+      "step": 10013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9124078484119194,
+      "learning_rate": 1.0898720223954673e-06,
+      "loss": 1.0283,
+      "step": 10014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.439716684881061,
+      "learning_rate": 1.088617665309315e-06,
+      "loss": 1.0935,
+      "step": 10015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7900259210122322,
+      "learning_rate": 1.0873639889190057e-06,
+      "loss": 1.1023,
+      "step": 10016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9940963631324977,
+      "learning_rate": 1.0861109933203051e-06,
+      "loss": 1.0635,
+      "step": 10017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9166044118417447,
+      "learning_rate": 1.084858678608922e-06,
+      "loss": 1.1001,
+      "step": 10018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2483673415689402,
+      "learning_rate": 1.0836070448805147e-06,
+      "loss": 1.0139,
+      "step": 10019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.796455854204381,
+      "learning_rate": 1.0823560922306875e-06,
+      "loss": 0.9736,
+      "step": 10020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.991226334256665,
+      "learning_rate": 1.0811058207549952e-06,
+      "loss": 1.0588,
+      "step": 10021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.877385449476312,
+      "learning_rate": 1.0798562305489414e-06,
+      "loss": 1.0493,
+      "step": 10022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.129939537010409,
+      "learning_rate": 1.078607321707974e-06,
+      "loss": 1.0479,
+      "step": 10023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9511237528706755,
+      "learning_rate": 1.0773590943274904e-06,
+      "loss": 0.9805,
+      "step": 10024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.853044915263501,
+      "learning_rate": 1.0761115485028417e-06,
+      "loss": 1.075,
+      "step": 10025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.160947956169501,
+      "learning_rate": 1.0748646843293132e-06,
+      "loss": 1.1235,
+      "step": 10026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9551667841805904,
+      "learning_rate": 1.073618501902154e-06,
+      "loss": 0.978,
+      "step": 10027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2387122154178942,
+      "learning_rate": 1.07237300131655e-06,
+      "loss": 1.1123,
+      "step": 10028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1438415754320634,
+      "learning_rate": 1.0711281826676379e-06,
+      "loss": 0.9468,
+      "step": 10029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7212103441401845,
+      "learning_rate": 1.0698840460505067e-06,
+      "loss": 1.0513,
+      "step": 10030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.25905902245639,
+      "learning_rate": 1.068640591560186e-06,
+      "loss": 0.9941,
+      "step": 10031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0411586416997163,
+      "learning_rate": 1.0673978192916635e-06,
+      "loss": 1.0701,
+      "step": 10032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.367776570409654,
+      "learning_rate": 1.0661557293398606e-06,
+      "loss": 1.0903,
+      "step": 10033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0367913558024413,
+      "learning_rate": 1.064914321799656e-06,
+      "loss": 1.0908,
+      "step": 10034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9360116227788233,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 1.0144,
+      "step": 10035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0845898924047743,
+      "learning_rate": 1.0624335543332964e-06,
+      "loss": 1.03,
+      "step": 10036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.494492642095353,
+      "learning_rate": 1.0611941945966341e-06,
+      "loss": 1.0229,
+      "step": 10037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5097314275645646,
+      "learning_rate": 1.0599555176505583e-06,
+      "loss": 1.0645,
+      "step": 10038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3715579797879918,
+      "learning_rate": 1.0587175235896862e-06,
+      "loss": 1.0769,
+      "step": 10039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.802946185621721,
+      "learning_rate": 1.0574802125085804e-06,
+      "loss": 1.0574,
+      "step": 10040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.083578651996278,
+      "learning_rate": 1.0562435845017515e-06,
+      "loss": 1.0708,
+      "step": 10041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2743712309229775,
+      "learning_rate": 1.0550076396636644e-06,
+      "loss": 1.0708,
+      "step": 10042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9760817403283646,
+      "learning_rate": 1.0537723780887232e-06,
+      "loss": 1.0232,
+      "step": 10043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9455677783462586,
+      "learning_rate": 1.0525377998712827e-06,
+      "loss": 1.0061,
+      "step": 10044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3321582717151594,
+      "learning_rate": 1.0513039051056507e-06,
+      "loss": 0.9834,
+      "step": 10045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7246156540174122,
+      "learning_rate": 1.0500706938860706e-06,
+      "loss": 1.0339,
+      "step": 10046
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1553527734393514,
+      "learning_rate": 1.048838166306748e-06,
+      "loss": 1.0176,
+      "step": 10047
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1018065699157344,
+      "learning_rate": 1.0476063224618271e-06,
+      "loss": 1.0564,
+      "step": 10048
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1293895503846856,
+      "learning_rate": 1.046375162445401e-06,
+      "loss": 1.053,
+      "step": 10049
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.041525056633169,
+      "learning_rate": 1.0451446863515147e-06,
+      "loss": 1.0591,
+      "step": 10050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.288352406164199,
+      "learning_rate": 1.043914894274156e-06,
+      "loss": 1.0427,
+      "step": 10051
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.469219740187879,
+      "learning_rate": 1.0426857863072626e-06,
+      "loss": 1.1035,
+      "step": 10052
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3885108201783507,
+      "learning_rate": 1.041457362544721e-06,
+      "loss": 1.1038,
+      "step": 10053
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9939326139719924,
+      "learning_rate": 1.040229623080362e-06,
+      "loss": 1.0637,
+      "step": 10054
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.056306255199241,
+      "learning_rate": 1.0390025680079707e-06,
+      "loss": 1.0176,
+      "step": 10055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.244871599608786,
+      "learning_rate": 1.0377761974212718e-06,
+      "loss": 1.0637,
+      "step": 10056
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.074590857151003,
+      "learning_rate": 1.0365505114139417e-06,
+      "loss": 1.0583,
+      "step": 10057
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2889179574008764,
+      "learning_rate": 1.03532551007961e-06,
+      "loss": 1.0854,
+      "step": 10058
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.923754740050819,
+      "learning_rate": 1.0341011935118395e-06,
+      "loss": 1.0215,
+      "step": 10059
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8465481781880797,
+      "learning_rate": 1.0328775618041565e-06,
+      "loss": 1.1309,
+      "step": 10060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.147210558887408,
+      "learning_rate": 1.0316546150500262e-06,
+      "loss": 1.0693,
+      "step": 10061
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9701726386191396,
+      "learning_rate": 1.0304323533428595e-06,
+      "loss": 1.0009,
+      "step": 10062
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7733783446974485,
+      "learning_rate": 1.0292107767760261e-06,
+      "loss": 1.0349,
+      "step": 10063
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3603460025824505,
+      "learning_rate": 1.0279898854428294e-06,
+      "loss": 1.0664,
+      "step": 10064
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.19433341810488,
+      "learning_rate": 1.0267696794365333e-06,
+      "loss": 1.0793,
+      "step": 10065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.734025441544105,
+      "learning_rate": 1.0255501588503391e-06,
+      "loss": 1.0344,
+      "step": 10066
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1690966553102213,
+      "learning_rate": 1.024331323777399e-06,
+      "loss": 1.0649,
+      "step": 10067
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.302695658331081,
+      "learning_rate": 1.023113174310817e-06,
+      "loss": 1.0808,
+      "step": 10068
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8821888201507457,
+      "learning_rate": 1.0218957105436378e-06,
+      "loss": 1.0405,
+      "step": 10069
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.116931038150381,
+      "learning_rate": 1.0206789325688627e-06,
+      "loss": 1.0962,
+      "step": 10070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.29451561776637,
+      "learning_rate": 1.0194628404794327e-06,
+      "loss": 1.0178,
+      "step": 10071
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.425335124294075,
+      "learning_rate": 1.0182474343682346e-06,
+      "loss": 1.0784,
+      "step": 10072
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7268700159826467,
+      "learning_rate": 1.0170327143281133e-06,
+      "loss": 1.0908,
+      "step": 10073
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2107715118934346,
+      "learning_rate": 1.01581868045185e-06,
+      "loss": 1.0862,
+      "step": 10074
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.888163250829748,
+      "learning_rate": 1.0146053328321847e-06,
+      "loss": 1.0288,
+      "step": 10075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2993425944704446,
+      "learning_rate": 1.0133926715617947e-06,
+      "loss": 1.1069,
+      "step": 10076
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2059242296821995,
+      "learning_rate": 1.012180696733308e-06,
+      "loss": 1.1084,
+      "step": 10077
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9282615957045297,
+      "learning_rate": 1.0109694084393085e-06,
+      "loss": 1.0229,
+      "step": 10078
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0839419322750365,
+      "learning_rate": 1.0097588067723108e-06,
+      "loss": 1.1001,
+      "step": 10079
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.938250427761463,
+      "learning_rate": 1.0085488918247932e-06,
+      "loss": 1.0679,
+      "step": 10080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9276386203815044,
+      "learning_rate": 1.007339663689173e-06,
+      "loss": 1.0457,
+      "step": 10081
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1782008462047067,
+      "learning_rate": 1.0061311224578152e-06,
+      "loss": 1.0188,
+      "step": 10082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1289050020972633,
+      "learning_rate": 1.0049232682230382e-06,
+      "loss": 1.0933,
+      "step": 10083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.249560575454707,
+      "learning_rate": 1.0037161010771013e-06,
+      "loss": 1.0884,
+      "step": 10084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2696997086818076,
+      "learning_rate": 1.0025096211122153e-06,
+      "loss": 0.9919,
+      "step": 10085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2144120344092704,
+      "learning_rate": 1.0013038284205368e-06,
+      "loss": 1.0554,
+      "step": 10086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8241527324264126,
+      "learning_rate": 1.0000987230941673e-06,
+      "loss": 1.0369,
+      "step": 10087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0296753886531866,
+      "learning_rate": 9.988943052251632e-07,
+      "loss": 1.0212,
+      "step": 10088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.143327358460143,
+      "learning_rate": 9.976905749055222e-07,
+      "loss": 1.0408,
+      "step": 10089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1150081642825396,
+      "learning_rate": 9.964875322271895e-07,
+      "loss": 1.0171,
+      "step": 10090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1531870494255627,
+      "learning_rate": 9.952851772820649e-07,
+      "loss": 1.0383,
+      "step": 10091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.196575694995201,
+      "learning_rate": 9.940835101619818e-07,
+      "loss": 1.0901,
+      "step": 10092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1050056056675683,
+      "learning_rate": 9.928825309587375e-07,
+      "loss": 1.0771,
+      "step": 10093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.544806979553177,
+      "learning_rate": 9.916822397640646e-07,
+      "loss": 1.0603,
+      "step": 10094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4148171486802448,
+      "learning_rate": 9.904826366696462e-07,
+      "loss": 1.0908,
+      "step": 10095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8612088960214606,
+      "learning_rate": 9.89283721767118e-07,
+      "loss": 1.0864,
+      "step": 10096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0564889796333023,
+      "learning_rate": 9.880854951480579e-07,
+      "loss": 0.998,
+      "step": 10097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3344486854083755,
+      "learning_rate": 9.868879569039912e-07,
+      "loss": 1.0303,
+      "step": 10098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.937946256752596,
+      "learning_rate": 9.85691107126392e-07,
+      "loss": 1.0535,
+      "step": 10099
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5222711322269027,
+      "learning_rate": 9.8449494590668e-07,
+      "loss": 1.0938,
+      "step": 10100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8711613897705535,
+      "learning_rate": 9.832994733362277e-07,
+      "loss": 1.0317,
+      "step": 10101
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1085162691738697,
+      "learning_rate": 9.821046895063479e-07,
+      "loss": 1.0173,
+      "step": 10102
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7796854478371213,
+      "learning_rate": 9.809105945083086e-07,
+      "loss": 1.0662,
+      "step": 10103
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9402961344306053,
+      "learning_rate": 9.797171884333169e-07,
+      "loss": 1.0515,
+      "step": 10104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9268756768897672,
+      "learning_rate": 9.785244713725316e-07,
+      "loss": 1.0332,
+      "step": 10105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.870503018807386,
+      "learning_rate": 9.773324434170583e-07,
+      "loss": 1.095,
+      "step": 10106
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1186735915622914,
+      "learning_rate": 9.761411046579505e-07,
+      "loss": 1.0405,
+      "step": 10107
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1695611240039487,
+      "learning_rate": 9.749504551862087e-07,
+      "loss": 1.0449,
+      "step": 10108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.097636994708515,
+      "learning_rate": 9.737604950927826e-07,
+      "loss": 1.0615,
+      "step": 10109
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.527655057909108,
+      "learning_rate": 9.725712244685615e-07,
+      "loss": 1.0444,
+      "step": 10110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4800248106433953,
+      "learning_rate": 9.713826434043972e-07,
+      "loss": 1.0029,
+      "step": 10111
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4732448026302642,
+      "learning_rate": 9.70194751991068e-07,
+      "loss": 1.0872,
+      "step": 10112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9853037579152724,
+      "learning_rate": 9.690075503193209e-07,
+      "loss": 1.0701,
+      "step": 10113
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2102680129777332,
+      "learning_rate": 9.678210384798348e-07,
+      "loss": 1.0203,
+      "step": 10114
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1062166943679523,
+      "learning_rate": 9.66635216563241e-07,
+      "loss": 1.0581,
+      "step": 10115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2940808405776,
+      "learning_rate": 9.654500846601234e-07,
+      "loss": 1.0405,
+      "step": 10116
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3462798745508957,
+      "learning_rate": 9.642656428610042e-07,
+      "loss": 1.0122,
+      "step": 10117
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.914636545453934,
+      "learning_rate": 9.630818912563577e-07,
+      "loss": 1.0234,
+      "step": 10118
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9898673767887374,
+      "learning_rate": 9.618988299366062e-07,
+      "loss": 1.0925,
+      "step": 10119
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3435256665643434,
+      "learning_rate": 9.607164589921135e-07,
+      "loss": 1.0527,
+      "step": 10120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4264747692844715,
+      "learning_rate": 9.595347785132014e-07,
+      "loss": 1.0425,
+      "step": 10121
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.994878314585728,
+      "learning_rate": 9.583537885901284e-07,
+      "loss": 1.0735,
+      "step": 10122
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.176634122080482,
+      "learning_rate": 9.571734893131036e-07,
+      "loss": 1.0681,
+      "step": 10123
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4626738603611535,
+      "learning_rate": 9.559938807722879e-07,
+      "loss": 1.0664,
+      "step": 10124
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.114066804427346,
+      "learning_rate": 9.54814963057784e-07,
+      "loss": 1.0518,
+      "step": 10125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.849896439741137,
+      "learning_rate": 9.536367362596422e-07,
+      "loss": 1.0959,
+      "step": 10126
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.663231002830529,
+      "learning_rate": 9.524592004678634e-07,
+      "loss": 1.0649,
+      "step": 10127
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.041882842182067,
+      "learning_rate": 9.512823557723894e-07,
+      "loss": 0.9961,
+      "step": 10128
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4954534507042743,
+      "learning_rate": 9.501062022631203e-07,
+      "loss": 1.0447,
+      "step": 10129
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.942555570452678,
+      "learning_rate": 9.489307400298908e-07,
+      "loss": 1.0205,
+      "step": 10130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1048701169674144,
+      "learning_rate": 9.477559691624927e-07,
+      "loss": 1.0764,
+      "step": 10131
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3503229387688407,
+      "learning_rate": 9.465818897506573e-07,
+      "loss": 1.0864,
+      "step": 10132
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.264398917180268,
+      "learning_rate": 9.454085018840676e-07,
+      "loss": 1.1174,
+      "step": 10133
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.263102693752179,
+      "learning_rate": 9.442358056523537e-07,
+      "loss": 1.0779,
+      "step": 10134
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.10249229063305,
+      "learning_rate": 9.430638011450932e-07,
+      "loss": 1.0591,
+      "step": 10135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9578845806654024,
+      "learning_rate": 9.418924884518055e-07,
+      "loss": 1.061,
+      "step": 10136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.950602253074278,
+      "learning_rate": 9.407218676619656e-07,
+      "loss": 1.0334,
+      "step": 10137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.442165126140164,
+      "learning_rate": 9.39551938864991e-07,
+      "loss": 1.0945,
+      "step": 10138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.849882254957656,
+      "learning_rate": 9.383827021502445e-07,
+      "loss": 1.0388,
+      "step": 10139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.289100611074016,
+      "learning_rate": 9.372141576070371e-07,
+      "loss": 1.042,
+      "step": 10140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.118578582358438,
+      "learning_rate": 9.360463053246316e-07,
+      "loss": 1.0754,
+      "step": 10141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.094090382970434,
+      "learning_rate": 9.348791453922334e-07,
+      "loss": 1.0747,
+      "step": 10142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0839645451994744,
+      "learning_rate": 9.337126778989947e-07,
+      "loss": 1.0513,
+      "step": 10143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.319807489641661,
+      "learning_rate": 9.325469029340195e-07,
+      "loss": 1.1414,
+      "step": 10144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3257174597273096,
+      "learning_rate": 9.313818205863489e-07,
+      "loss": 1.0271,
+      "step": 10145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8974296357116507,
+      "learning_rate": 9.302174309449841e-07,
+      "loss": 1.02,
+      "step": 10146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.031250843067642,
+      "learning_rate": 9.290537340988647e-07,
+      "loss": 1.0183,
+      "step": 10147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3648121846270436,
+      "learning_rate": 9.278907301368778e-07,
+      "loss": 1.0686,
+      "step": 10148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2711520265322025,
+      "learning_rate": 9.26728419147862e-07,
+      "loss": 1.0513,
+      "step": 10149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8953640401563883,
+      "learning_rate": 9.25566801220601e-07,
+      "loss": 1.1609,
+      "step": 10150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.080022957591389,
+      "learning_rate": 9.244058764438224e-07,
+      "loss": 1.0679,
+      "step": 10151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9372951430270873,
+      "learning_rate": 9.232456449062055e-07,
+      "loss": 1.0137,
+      "step": 10152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.401887322308605,
+      "learning_rate": 9.220861066963715e-07,
+      "loss": 1.0,
+      "step": 10153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9895030800882303,
+      "learning_rate": 9.209272619028964e-07,
+      "loss": 1.0891,
+      "step": 10154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.770814287158864,
+      "learning_rate": 9.197691106142947e-07,
+      "loss": 1.0835,
+      "step": 10155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1796118300511473,
+      "learning_rate": 9.186116529190325e-07,
+      "loss": 1.1028,
+      "step": 10156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9180774662882762,
+      "learning_rate": 9.174548889055245e-07,
+      "loss": 1.0239,
+      "step": 10157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.154446012223281,
+      "learning_rate": 9.16298818662128e-07,
+      "loss": 1.0684,
+      "step": 10158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2133679436863436,
+      "learning_rate": 9.151434422771499e-07,
+      "loss": 1.0341,
+      "step": 10159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9623202033827165,
+      "learning_rate": 9.139887598388442e-07,
+      "loss": 1.0671,
+      "step": 10160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3576405652610024,
+      "learning_rate": 9.12834771435408e-07,
+      "loss": 1.0208,
+      "step": 10161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7285178534449637,
+      "learning_rate": 9.116814771549943e-07,
+      "loss": 1.0642,
+      "step": 10162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3094926075327082,
+      "learning_rate": 9.105288770856935e-07,
+      "loss": 1.0725,
+      "step": 10163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.407246375511491,
+      "learning_rate": 9.093769713155475e-07,
+      "loss": 1.0632,
+      "step": 10164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3308922635216587,
+      "learning_rate": 9.082257599325461e-07,
+      "loss": 1.073,
+      "step": 10165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.507882049036748,
+      "learning_rate": 9.070752430246221e-07,
+      "loss": 1.0322,
+      "step": 10166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.827471985692932,
+      "learning_rate": 9.059254206796597e-07,
+      "loss": 1.0312,
+      "step": 10167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8805674109169583,
+      "learning_rate": 9.047762929854886e-07,
+      "loss": 1.0938,
+      "step": 10168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4543424233303663,
+      "learning_rate": 9.03627860029882e-07,
+      "loss": 1.0017,
+      "step": 10169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3023605391714312,
+      "learning_rate": 9.024801219005663e-07,
+      "loss": 1.0696,
+      "step": 10170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2862503438850013,
+      "learning_rate": 9.013330786852093e-07,
+      "loss": 1.0149,
+      "step": 10171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.43424018526597,
+      "learning_rate": 9.001867304714295e-07,
+      "loss": 1.0466,
+      "step": 10172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5828659430338865,
+      "learning_rate": 8.990410773467895e-07,
+      "loss": 1.0583,
+      "step": 10173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1418406323214074,
+      "learning_rate": 8.978961193987989e-07,
+      "loss": 1.0757,
+      "step": 10174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7060863927692935,
+      "learning_rate": 8.96751856714918e-07,
+      "loss": 1.095,
+      "step": 10175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4798031849167024,
+      "learning_rate": 8.95608289382548e-07,
+      "loss": 1.0437,
+      "step": 10176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5525765349210054,
+      "learning_rate": 8.944654174890432e-07,
+      "loss": 1.0796,
+      "step": 10177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7589212218698536,
+      "learning_rate": 8.933232411217041e-07,
+      "loss": 1.0374,
+      "step": 10178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.381049047154346,
+      "learning_rate": 8.921817603677685e-07,
+      "loss": 1.0347,
+      "step": 10179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8270086071043425,
+      "learning_rate": 8.910409753144344e-07,
+      "loss": 1.0229,
+      "step": 10180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4907647638720056,
+      "learning_rate": 8.899008860488367e-07,
+      "loss": 1.0312,
+      "step": 10181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.136606885897512,
+      "learning_rate": 8.887614926580646e-07,
+      "loss": 1.0945,
+      "step": 10182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.173970528048881,
+      "learning_rate": 8.876227952291494e-07,
+      "loss": 1.0979,
+      "step": 10183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0651033404897983,
+      "learning_rate": 8.864847938490706e-07,
+      "loss": 1.1096,
+      "step": 10184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.423734704240279,
+      "learning_rate": 8.853474886047541e-07,
+      "loss": 1.0061,
+      "step": 10185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.014814919739556,
+      "learning_rate": 8.842108795830718e-07,
+      "loss": 1.0032,
+      "step": 10186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9735651013224165,
+      "learning_rate": 8.830749668708449e-07,
+      "loss": 1.1018,
+      "step": 10187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2099724034765242,
+      "learning_rate": 8.819397505548411e-07,
+      "loss": 1.0369,
+      "step": 10188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.188516789046765,
+      "learning_rate": 8.808052307217707e-07,
+      "loss": 1.042,
+      "step": 10189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.008011675407942,
+      "learning_rate": 8.796714074582979e-07,
+      "loss": 1.0303,
+      "step": 10190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0148515044384703,
+      "learning_rate": 8.785382808510279e-07,
+      "loss": 1.0271,
+      "step": 10191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1859988111470035,
+      "learning_rate": 8.774058509865146e-07,
+      "loss": 1.1248,
+      "step": 10192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.127881235378037,
+      "learning_rate": 8.762741179512579e-07,
+      "loss": 1.0439,
+      "step": 10193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.864699095117307,
+      "learning_rate": 8.751430818317042e-07,
+      "loss": 0.9915,
+      "step": 10194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.28078319444879,
+      "learning_rate": 8.740127427142519e-07,
+      "loss": 1.0393,
+      "step": 10195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3969455588782824,
+      "learning_rate": 8.72883100685239e-07,
+      "loss": 0.9851,
+      "step": 10196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8106764251742,
+      "learning_rate": 8.717541558309539e-07,
+      "loss": 1.0156,
+      "step": 10197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.608657964373342,
+      "learning_rate": 8.7062590823763e-07,
+      "loss": 1.025,
+      "step": 10198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0573608648283956,
+      "learning_rate": 8.694983579914484e-07,
+      "loss": 1.0854,
+      "step": 10199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2642147372781802,
+      "learning_rate": 8.683715051785391e-07,
+      "loss": 1.0498,
+      "step": 10200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.926631009095057,
+      "learning_rate": 8.672453498849754e-07,
+      "loss": 1.0337,
+      "step": 10201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0789038806108575,
+      "learning_rate": 8.661198921967773e-07,
+      "loss": 1.0427,
+      "step": 10202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.230475666726029,
+      "learning_rate": 8.64995132199915e-07,
+      "loss": 1.0376,
+      "step": 10203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.138835091492357,
+      "learning_rate": 8.638710699803043e-07,
+      "loss": 1.0483,
+      "step": 10204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5918609320818056,
+      "learning_rate": 8.627477056238042e-07,
+      "loss": 1.0085,
+      "step": 10205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.740996957209362,
+      "learning_rate": 8.616250392162229e-07,
+      "loss": 1.0669,
+      "step": 10206
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3378120228818884,
+      "learning_rate": 8.605030708433149e-07,
+      "loss": 1.0601,
+      "step": 10207
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.310771895669109,
+      "learning_rate": 8.59381800590785e-07,
+      "loss": 1.0422,
+      "step": 10208
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.846676030164826,
+      "learning_rate": 8.582612285442771e-07,
+      "loss": 0.9774,
+      "step": 10209
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.243823739892235,
+      "learning_rate": 8.571413547893903e-07,
+      "loss": 1.0537,
+      "step": 10210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9930574931764666,
+      "learning_rate": 8.560221794116652e-07,
+      "loss": 0.9741,
+      "step": 10211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.088159400768697,
+      "learning_rate": 8.549037024965868e-07,
+      "loss": 1.0493,
+      "step": 10212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6621440313606732,
+      "learning_rate": 8.537859241295921e-07,
+      "loss": 1.1599,
+      "step": 10213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.053965077241652,
+      "learning_rate": 8.526688443960618e-07,
+      "loss": 1.0432,
+      "step": 10214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.700228764790682,
+      "learning_rate": 8.515524633813266e-07,
+      "loss": 1.0732,
+      "step": 10215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.80221472518388,
+      "learning_rate": 8.504367811706582e-07,
+      "loss": 0.9976,
+      "step": 10216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.716742263714474,
+      "learning_rate": 8.493217978492807e-07,
+      "loss": 1.134,
+      "step": 10217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1988234147607684,
+      "learning_rate": 8.482075135023593e-07,
+      "loss": 1.092,
+      "step": 10218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4529406848432203,
+      "learning_rate": 8.470939282150093e-07,
+      "loss": 1.0549,
+      "step": 10219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7892191021303927,
+      "learning_rate": 8.459810420722936e-07,
+      "loss": 1.073,
+      "step": 10220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3468434298892524,
+      "learning_rate": 8.448688551592188e-07,
+      "loss": 1.0547,
+      "step": 10221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2786457522665966,
+      "learning_rate": 8.43757367560738e-07,
+      "loss": 1.0605,
+      "step": 10222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.130704579063487,
+      "learning_rate": 8.426465793617544e-07,
+      "loss": 1.0771,
+      "step": 10223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1817556920094328,
+      "learning_rate": 8.415364906471157e-07,
+      "loss": 1.0505,
+      "step": 10224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2443526202040407,
+      "learning_rate": 8.404271015016152e-07,
+      "loss": 1.0869,
+      "step": 10225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6914828529896453,
+      "learning_rate": 8.393184120099929e-07,
+      "loss": 1.0576,
+      "step": 10226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.365179168399998,
+      "learning_rate": 8.382104222569343e-07,
+      "loss": 1.0925,
+      "step": 10227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9229116890876066,
+      "learning_rate": 8.371031323270773e-07,
+      "loss": 1.0703,
+      "step": 10228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2256951010453947,
+      "learning_rate": 8.359965423049998e-07,
+      "loss": 1.0896,
+      "step": 10229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9276538968079335,
+      "learning_rate": 8.348906522752287e-07,
+      "loss": 0.9971,
+      "step": 10230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7251076353799824,
+      "learning_rate": 8.337854623222408e-07,
+      "loss": 1.0144,
+      "step": 10231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.021538336487136,
+      "learning_rate": 8.326809725304496e-07,
+      "loss": 1.0667,
+      "step": 10232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1358087898506573,
+      "learning_rate": 8.315771829842267e-07,
+      "loss": 1.0581,
+      "step": 10233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9838560928660733,
+      "learning_rate": 8.304740937678835e-07,
+      "loss": 1.0352,
+      "step": 10234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5109233459294305,
+      "learning_rate": 8.293717049656769e-07,
+      "loss": 1.0576,
+      "step": 10235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1527374979184954,
+      "learning_rate": 8.282700166618185e-07,
+      "loss": 0.9846,
+      "step": 10236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8665438730543777,
+      "learning_rate": 8.271690289404565e-07,
+      "loss": 1.0452,
+      "step": 10237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1996718645892046,
+      "learning_rate": 8.260687418856905e-07,
+      "loss": 1.0356,
+      "step": 10238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9675134497548523,
+      "learning_rate": 8.249691555815675e-07,
+      "loss": 0.9937,
+      "step": 10239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.281176962635655,
+      "learning_rate": 8.23870270112076e-07,
+      "loss": 1.0798,
+      "step": 10240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1899516238182226,
+      "learning_rate": 8.227720855611576e-07,
+      "loss": 1.0796,
+      "step": 10241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.172922229596292,
+      "learning_rate": 8.216746020126953e-07,
+      "loss": 1.073,
+      "step": 10242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3285494148156314,
+      "learning_rate": 8.205778195505221e-07,
+      "loss": 1.1257,
+      "step": 10243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0290779228943365,
+      "learning_rate": 8.194817382584175e-07,
+      "loss": 0.9956,
+      "step": 10244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4937881917819444,
+      "learning_rate": 8.18386358220099e-07,
+      "loss": 1.0532,
+      "step": 10245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9810723669842134,
+      "learning_rate": 8.172916795192421e-07,
+      "loss": 1.0073,
+      "step": 10246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.297442923885662,
+      "learning_rate": 8.161977022394619e-07,
+      "loss": 1.0276,
+      "step": 10247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.866877362858387,
+      "learning_rate": 8.151044264643238e-07,
+      "loss": 1.0488,
+      "step": 10248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.110775213568637,
+      "learning_rate": 8.140118522773377e-07,
+      "loss": 1.0044,
+      "step": 10249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.891646061805773,
+      "learning_rate": 8.129199797619558e-07,
+      "loss": 1.1038,
+      "step": 10250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.213869802641406,
+      "learning_rate": 8.11828809001588e-07,
+      "loss": 1.0913,
+      "step": 10251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.049771109306673,
+      "learning_rate": 8.107383400795754e-07,
+      "loss": 1.0608,
+      "step": 10252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.071739336841942,
+      "learning_rate": 8.096485730792191e-07,
+      "loss": 1.041,
+      "step": 10253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.266292444779367,
+      "learning_rate": 8.085595080837594e-07,
+      "loss": 1.0999,
+      "step": 10254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1676672212551082,
+      "learning_rate": 8.074711451763817e-07,
+      "loss": 1.0244,
+      "step": 10255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0560831449737558,
+      "learning_rate": 8.063834844402252e-07,
+      "loss": 1.0872,
+      "step": 10256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.498855664102378,
+      "learning_rate": 8.052965259583689e-07,
+      "loss": 1.1252,
+      "step": 10257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2862476226793325,
+      "learning_rate": 8.042102698138399e-07,
+      "loss": 1.0635,
+      "step": 10258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.559904257109107,
+      "learning_rate": 8.031247160896116e-07,
+      "loss": 1.0278,
+      "step": 10259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7434908282506925,
+      "learning_rate": 8.020398648686023e-07,
+      "loss": 0.9355,
+      "step": 10260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7992068528277327,
+      "learning_rate": 8.009557162336823e-07,
+      "loss": 1.0029,
+      "step": 10261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9835898710976294,
+      "learning_rate": 7.99872270267662e-07,
+      "loss": 1.0208,
+      "step": 10262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7079830644868186,
+      "learning_rate": 7.987895270532986e-07,
+      "loss": 1.0078,
+      "step": 10263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8694413428472925,
+      "learning_rate": 7.977074866733037e-07,
+      "loss": 1.0798,
+      "step": 10264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6602301400572474,
+      "learning_rate": 7.966261492103212e-07,
+      "loss": 1.1047,
+      "step": 10265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3281201209646825,
+      "learning_rate": 7.955455147469548e-07,
+      "loss": 1.1064,
+      "step": 10266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7280890483579285,
+      "learning_rate": 7.944655833657467e-07,
+      "loss": 0.947,
+      "step": 10267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6557083071909853,
+      "learning_rate": 7.933863551491849e-07,
+      "loss": 1.0429,
+      "step": 10268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8501229519527813,
+      "learning_rate": 7.923078301797105e-07,
+      "loss": 1.0276,
+      "step": 10269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.429990616198488,
+      "learning_rate": 7.912300085397051e-07,
+      "loss": 1.0999,
+      "step": 10270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.022468352963498,
+      "learning_rate": 7.901528903114975e-07,
+      "loss": 1.0469,
+      "step": 10271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5155426217070405,
+      "learning_rate": 7.890764755773639e-07,
+      "loss": 1.0991,
+      "step": 10272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.016227200799043,
+      "learning_rate": 7.880007644195242e-07,
+      "loss": 1.0093,
+      "step": 10273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.781095204800549,
+      "learning_rate": 7.869257569201493e-07,
+      "loss": 1.1023,
+      "step": 10274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.801052311159698,
+      "learning_rate": 7.858514531613537e-07,
+      "loss": 1.002,
+      "step": 10275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2554736525334116,
+      "learning_rate": 7.847778532251948e-07,
+      "loss": 1.0505,
+      "step": 10276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5425973273522513,
+      "learning_rate": 7.837049571936861e-07,
+      "loss": 0.9897,
+      "step": 10277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3132700842420957,
+      "learning_rate": 7.82632765148772e-07,
+      "loss": 1.0415,
+      "step": 10278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7595497594966116,
+      "learning_rate": 7.815612771723591e-07,
+      "loss": 1.0156,
+      "step": 10279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.468784542330595,
+      "learning_rate": 7.804904933462887e-07,
+      "loss": 1.0862,
+      "step": 10280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.713980288892442,
+      "learning_rate": 7.794204137523531e-07,
+      "loss": 1.0085,
+      "step": 10281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.443582688139818,
+      "learning_rate": 7.783510384722937e-07,
+      "loss": 1.0608,
+      "step": 10282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3927924162493235,
+      "learning_rate": 7.772823675877905e-07,
+      "loss": 1.0422,
+      "step": 10283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.251183606937502,
+      "learning_rate": 7.762144011804784e-07,
+      "loss": 1.0518,
+      "step": 10284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4519272351651153,
+      "learning_rate": 7.751471393319287e-07,
+      "loss": 1.0898,
+      "step": 10285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.034959071813823,
+      "learning_rate": 7.740805821236685e-07,
+      "loss": 1.0356,
+      "step": 10286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0115395278025807,
+      "learning_rate": 7.730147296371649e-07,
+      "loss": 1.0693,
+      "step": 10287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.972554240657018,
+      "learning_rate": 7.719495819538325e-07,
+      "loss": 1.0254,
+      "step": 10288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2412310381568057,
+      "learning_rate": 7.708851391550342e-07,
+      "loss": 1.0496,
+      "step": 10289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6789005447899457,
+      "learning_rate": 7.698214013220762e-07,
+      "loss": 0.9661,
+      "step": 10290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.888370460630493,
+      "learning_rate": 7.687583685362132e-07,
+      "loss": 1.0769,
+      "step": 10291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.827830692515817,
+      "learning_rate": 7.676960408786438e-07,
+      "loss": 1.0029,
+      "step": 10292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8375177099026034,
+      "learning_rate": 7.66634418430513e-07,
+      "loss": 1.0703,
+      "step": 10293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3691448199204577,
+      "learning_rate": 7.65573501272916e-07,
+      "loss": 1.0762,
+      "step": 10294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.306637447680114,
+      "learning_rate": 7.64513289486889e-07,
+      "loss": 1.0769,
+      "step": 10295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.812423310625064,
+      "learning_rate": 7.63453783153415e-07,
+      "loss": 1.0393,
+      "step": 10296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.07659082118159,
+      "learning_rate": 7.623949823534282e-07,
+      "loss": 1.0349,
+      "step": 10297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.173548533082205,
+      "learning_rate": 7.613368871678006e-07,
+      "loss": 1.0269,
+      "step": 10298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3450075309969605,
+      "learning_rate": 7.602794976773575e-07,
+      "loss": 1.1189,
+      "step": 10299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0326909039504013,
+      "learning_rate": 7.592228139628677e-07,
+      "loss": 0.9836,
+      "step": 10300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.770871673597105,
+      "learning_rate": 7.581668361050432e-07,
+      "loss": 1.0088,
+      "step": 10301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5507662571379637,
+      "learning_rate": 7.571115641845483e-07,
+      "loss": 1.115,
+      "step": 10302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5319831972997426,
+      "learning_rate": 7.560569982819877e-07,
+      "loss": 1.0974,
+      "step": 10303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2126460805545256,
+      "learning_rate": 7.550031384779167e-07,
+      "loss": 1.0488,
+      "step": 10304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1468808535850052,
+      "learning_rate": 7.53949984852832e-07,
+      "loss": 1.0737,
+      "step": 10305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1974116270487496,
+      "learning_rate": 7.528975374871783e-07,
+      "loss": 1.0691,
+      "step": 10306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7674448368871056,
+      "learning_rate": 7.518457964613501e-07,
+      "loss": 0.9819,
+      "step": 10307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9593516573506493,
+      "learning_rate": 7.507947618556821e-07,
+      "loss": 1.0366,
+      "step": 10308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0850880679085018,
+      "learning_rate": 7.497444337504556e-07,
+      "loss": 0.9878,
+      "step": 10309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.162387368268324,
+      "learning_rate": 7.486948122259074e-07,
+      "loss": 1.051,
+      "step": 10310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6279722018578986,
+      "learning_rate": 7.476458973622036e-07,
+      "loss": 0.9927,
+      "step": 10311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2893209486107757,
+      "learning_rate": 7.465976892394711e-07,
+      "loss": 1.0762,
+      "step": 10312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.477687877032793,
+      "learning_rate": 7.455501879377769e-07,
+      "loss": 1.0107,
+      "step": 10313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9403422569749074,
+      "learning_rate": 7.445033935371304e-07,
+      "loss": 0.991,
+      "step": 10314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9941533599932306,
+      "learning_rate": 7.434573061174966e-07,
+      "loss": 1.0659,
+      "step": 10315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3305580461746964,
+      "learning_rate": 7.424119257587759e-07,
+      "loss": 0.9919,
+      "step": 10316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.925785052996686,
+      "learning_rate": 7.413672525408266e-07,
+      "loss": 1.0437,
+      "step": 10317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2231794975775436,
+      "learning_rate": 7.403232865434385e-07,
+      "loss": 1.0166,
+      "step": 10318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1556784178953943,
+      "learning_rate": 7.392800278463575e-07,
+      "loss": 1.0356,
+      "step": 10319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4109807338226683,
+      "learning_rate": 7.382374765292733e-07,
+      "loss": 1.0571,
+      "step": 10320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3093465605318424,
+      "learning_rate": 7.371956326718211e-07,
+      "loss": 1.0674,
+      "step": 10321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2400206388441455,
+      "learning_rate": 7.361544963535827e-07,
+      "loss": 1.0491,
+      "step": 10322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.364722984540735,
+      "learning_rate": 7.351140676540858e-07,
+      "loss": 0.9707,
+      "step": 10323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.094720358946178,
+      "learning_rate": 7.340743466528021e-07,
+      "loss": 1.1045,
+      "step": 10324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.454887878583111,
+      "learning_rate": 7.330353334291507e-07,
+      "loss": 1.0957,
+      "step": 10325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1786668638903643,
+      "learning_rate": 7.319970280624955e-07,
+      "loss": 1.0222,
+      "step": 10326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0569794384039737,
+      "learning_rate": 7.30959430632151e-07,
+      "loss": 1.002,
+      "step": 10327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0848748008498212,
+      "learning_rate": 7.299225412173727e-07,
+      "loss": 1.0583,
+      "step": 10328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3034129743453433,
+      "learning_rate": 7.288863598973595e-07,
+      "loss": 1.05,
+      "step": 10329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.40008854318492,
+      "learning_rate": 7.278508867512679e-07,
+      "loss": 1.0776,
+      "step": 10330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8579066275876754,
+      "learning_rate": 7.268161218581837e-07,
+      "loss": 1.0378,
+      "step": 10331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.937012957828577,
+      "learning_rate": 7.257820652971537e-07,
+      "loss": 1.1536,
+      "step": 10332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.432441736176116,
+      "learning_rate": 7.247487171471624e-07,
+      "loss": 1.019,
+      "step": 10333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.345477527279449,
+      "learning_rate": 7.237160774871399e-07,
+      "loss": 1.0242,
+      "step": 10334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.977009655954677,
+      "learning_rate": 7.226841463959688e-07,
+      "loss": 1.1067,
+      "step": 10335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5835330910286034,
+      "learning_rate": 7.216529239524706e-07,
+      "loss": 0.9871,
+      "step": 10336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.638244733218627,
+      "learning_rate": 7.206224102354143e-07,
+      "loss": 1.0769,
+      "step": 10337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.966041891408001,
+      "learning_rate": 7.19592605323518e-07,
+      "loss": 1.0439,
+      "step": 10338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6828652951804277,
+      "learning_rate": 7.18563509295439e-07,
+      "loss": 1.063,
+      "step": 10339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.377497125077489,
+      "learning_rate": 7.175351222297911e-07,
+      "loss": 1.0691,
+      "step": 10340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3227986822537887,
+      "learning_rate": 7.165074442051234e-07,
+      "loss": 1.0195,
+      "step": 10341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.748886520196357,
+      "learning_rate": 7.154804752999344e-07,
+      "loss": 1.0027,
+      "step": 10342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.977689133087361,
+      "learning_rate": 7.144542155926748e-07,
+      "loss": 1.0647,
+      "step": 10343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.186467532638707,
+      "learning_rate": 7.134286651617273e-07,
+      "loss": 0.9944,
+      "step": 10344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4019549865697853,
+      "learning_rate": 7.124038240854348e-07,
+      "loss": 1.03,
+      "step": 10345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3696240445251058,
+      "learning_rate": 7.113796924420769e-07,
+      "loss": 1.0352,
+      "step": 10346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.744801286334288,
+      "learning_rate": 7.103562703098821e-07,
+      "loss": 1.0251,
+      "step": 10347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7308739817155976,
+      "learning_rate": 7.093335577670257e-07,
+      "loss": 1.073,
+      "step": 10348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9831783628691046,
+      "learning_rate": 7.083115548916264e-07,
+      "loss": 1.0659,
+      "step": 10349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1118127490497565,
+      "learning_rate": 7.072902617617527e-07,
+      "loss": 1.0876,
+      "step": 10350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.525070978758378,
+      "learning_rate": 7.062696784554124e-07,
+      "loss": 1.0276,
+      "step": 10351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2545504971195647,
+      "learning_rate": 7.05249805050563e-07,
+      "loss": 1.0791,
+      "step": 10352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.120840670204207,
+      "learning_rate": 7.042306416251088e-07,
+      "loss": 1.0759,
+      "step": 10353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0893557846819215,
+      "learning_rate": 7.032121882568976e-07,
+      "loss": 1.0449,
+      "step": 10354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6143017434927347,
+      "learning_rate": 7.021944450237273e-07,
+      "loss": 0.9375,
+      "step": 10355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9644455816703505,
+      "learning_rate": 7.011774120033343e-07,
+      "loss": 1.0725,
+      "step": 10356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.338780477741788,
+      "learning_rate": 7.001610892734067e-07,
+      "loss": 1.0574,
+      "step": 10357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.273480311360739,
+      "learning_rate": 6.991454769115757e-07,
+      "loss": 1.0764,
+      "step": 10358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.80121272513069,
+      "learning_rate": 6.98130574995417e-07,
+      "loss": 1.0125,
+      "step": 10359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.159711534756563,
+      "learning_rate": 6.971163836024575e-07,
+      "loss": 1.0865,
+      "step": 10360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3344735375851613,
+      "learning_rate": 6.96102902810164e-07,
+      "loss": 1.0542,
+      "step": 10361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3416999533765126,
+      "learning_rate": 6.950901326959492e-07,
+      "loss": 1.0598,
+      "step": 10362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0967588046158565,
+      "learning_rate": 6.940780733371799e-07,
+      "loss": 1.051,
+      "step": 10363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.914047692084331,
+      "learning_rate": 6.930667248111556e-07,
+      "loss": 1.075,
+      "step": 10364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.83650939717686,
+      "learning_rate": 6.920560871951309e-07,
+      "loss": 1.0581,
+      "step": 10365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2741715016843727,
+      "learning_rate": 6.910461605663044e-07,
+      "loss": 1.0552,
+      "step": 10366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.156755964372735,
+      "learning_rate": 6.900369450018163e-07,
+      "loss": 1.0718,
+      "step": 10367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5147891347021454,
+      "learning_rate": 6.890284405787584e-07,
+      "loss": 1.043,
+      "step": 10368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.577441360514869,
+      "learning_rate": 6.880206473741646e-07,
+      "loss": 1.0913,
+      "step": 10369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0667139472992817,
+      "learning_rate": 6.870135654650156e-07,
+      "loss": 1.0796,
+      "step": 10370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0342861039623097,
+      "learning_rate": 6.860071949282354e-07,
+      "loss": 1.0769,
+      "step": 10371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1166448374741207,
+      "learning_rate": 6.850015358406958e-07,
+      "loss": 1.0195,
+      "step": 10372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6621724241551292,
+      "learning_rate": 6.839965882792165e-07,
+      "loss": 1.0876,
+      "step": 10373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1745542672721596,
+      "learning_rate": 6.829923523205595e-07,
+      "loss": 1.0574,
+      "step": 10374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2042334642644206,
+      "learning_rate": 6.81988828041431e-07,
+      "loss": 1.0298,
+      "step": 10375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2211836597510812,
+      "learning_rate": 6.809860155184877e-07,
+      "loss": 1.106,
+      "step": 10376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8777851993824353,
+      "learning_rate": 6.799839148283293e-07,
+      "loss": 1.0504,
+      "step": 10377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.861445180513362,
+      "learning_rate": 6.789825260475014e-07,
+      "loss": 1.0557,
+      "step": 10378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1825534083181526,
+      "learning_rate": 6.779818492524937e-07,
+      "loss": 1.0574,
+      "step": 10379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0992494021357477,
+      "learning_rate": 6.769818845197418e-07,
+      "loss": 1.085,
+      "step": 10380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2782085243110726,
+      "learning_rate": 6.759826319256313e-07,
+      "loss": 1.0798,
+      "step": 10381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6065131119605245,
+      "learning_rate": 6.749840915464867e-07,
+      "loss": 1.0586,
+      "step": 10382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.916708696188836,
+      "learning_rate": 6.73986263458587e-07,
+      "loss": 1.0181,
+      "step": 10383
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2225048333055764,
+      "learning_rate": 6.729891477381467e-07,
+      "loss": 1.0725,
+      "step": 10384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2259609879653977,
+      "learning_rate": 6.719927444613284e-07,
+      "loss": 0.9907,
+      "step": 10385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.241082433372561,
+      "learning_rate": 6.709970537042476e-07,
+      "loss": 1.0854,
+      "step": 10386
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2776265626185177,
+      "learning_rate": 6.700020755429559e-07,
+      "loss": 1.0674,
+      "step": 10387
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4352092148673483,
+      "learning_rate": 6.690078100534591e-07,
+      "loss": 1.0039,
+      "step": 10388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.221428523863682,
+      "learning_rate": 6.68014257311701e-07,
+      "loss": 1.0461,
+      "step": 10389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6361114337247784,
+      "learning_rate": 6.670214173935751e-07,
+      "loss": 1.1174,
+      "step": 10390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.973278663636785,
+      "learning_rate": 6.660292903749188e-07,
+      "loss": 1.0559,
+      "step": 10391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7670967347977626,
+      "learning_rate": 6.650378763315147e-07,
+      "loss": 1.1021,
+      "step": 10392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3646975139271342,
+      "learning_rate": 6.640471753390954e-07,
+      "loss": 1.1853,
+      "step": 10393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2212279546675666,
+      "learning_rate": 6.630571874733338e-07,
+      "loss": 1.0842,
+      "step": 10394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1237293912429034,
+      "learning_rate": 6.620679128098473e-07,
+      "loss": 1.0979,
+      "step": 10395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.97983421155535,
+      "learning_rate": 6.610793514242075e-07,
+      "loss": 1.0662,
+      "step": 10396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.224684860725528,
+      "learning_rate": 6.600915033919197e-07,
+      "loss": 1.0706,
+      "step": 10397
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7342720606705275,
+      "learning_rate": 6.591043687884447e-07,
+      "loss": 0.9673,
+      "step": 10398
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9648323498633626,
+      "learning_rate": 6.58117947689183e-07,
+      "loss": 1.0527,
+      "step": 10399
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.285583647318208,
+      "learning_rate": 6.571322401694813e-07,
+      "loss": 1.0198,
+      "step": 10400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.672562667717287,
+      "learning_rate": 6.56147246304637e-07,
+      "loss": 0.9636,
+      "step": 10401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3069490348341195,
+      "learning_rate": 6.551629661698844e-07,
+      "loss": 1.0618,
+      "step": 10402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0158164322494834,
+      "learning_rate": 6.541793998404111e-07,
+      "loss": 1.0889,
+      "step": 10403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.033361532508008,
+      "learning_rate": 6.531965473913437e-07,
+      "loss": 1.0649,
+      "step": 10404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1724783417237767,
+      "learning_rate": 6.522144088977578e-07,
+      "loss": 1.053,
+      "step": 10405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8287304540025553,
+      "learning_rate": 6.512329844346776e-07,
+      "loss": 1.0713,
+      "step": 10406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.083310899713054,
+      "learning_rate": 6.502522740770656e-07,
+      "loss": 1.0005,
+      "step": 10407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.645207903036037,
+      "learning_rate": 6.492722778998328e-07,
+      "loss": 1.0237,
+      "step": 10408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0200090838338722,
+      "learning_rate": 6.482929959778395e-07,
+      "loss": 1.0195,
+      "step": 10409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2131110442037523,
+      "learning_rate": 6.473144283858868e-07,
+      "loss": 1.0266,
+      "step": 10410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3487358525182205,
+      "learning_rate": 6.463365751987227e-07,
+      "loss": 1.0857,
+      "step": 10411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3839276351463026,
+      "learning_rate": 6.453594364910398e-07,
+      "loss": 1.0874,
+      "step": 10412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2260503190663545,
+      "learning_rate": 6.44383012337475e-07,
+      "loss": 1.0913,
+      "step": 10413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.100807427732881,
+      "learning_rate": 6.434073028126164e-07,
+      "loss": 1.0327,
+      "step": 10414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1881396176214207,
+      "learning_rate": 6.424323079909911e-07,
+      "loss": 1.1333,
+      "step": 10415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8395178955865386,
+      "learning_rate": 6.414580279470761e-07,
+      "loss": 1.0,
+      "step": 10416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8767177227476437,
+      "learning_rate": 6.404844627552886e-07,
+      "loss": 1.0454,
+      "step": 10417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2249256004393856,
+      "learning_rate": 6.395116124899958e-07,
+      "loss": 1.0703,
+      "step": 10418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2026198797165666,
+      "learning_rate": 6.385394772255105e-07,
+      "loss": 1.0142,
+      "step": 10419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.212241664091265,
+      "learning_rate": 6.375680570360876e-07,
+      "loss": 1.0977,
+      "step": 10420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8788534537110593,
+      "learning_rate": 6.365973519959279e-07,
+      "loss": 0.9836,
+      "step": 10421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.82229187187609,
+      "learning_rate": 6.356273621791819e-07,
+      "loss": 0.9915,
+      "step": 10422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.13968977137124,
+      "learning_rate": 6.346580876599395e-07,
+      "loss": 1.1089,
+      "step": 10423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.088556688039263,
+      "learning_rate": 6.336895285122413e-07,
+      "loss": 0.9783,
+      "step": 10424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0151182966403116,
+      "learning_rate": 6.327216848100659e-07,
+      "loss": 1.0847,
+      "step": 10425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.030815608014272,
+      "learning_rate": 6.317545566273475e-07,
+      "loss": 1.0322,
+      "step": 10426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3207967050294855,
+      "learning_rate": 6.307881440379582e-07,
+      "loss": 1.0725,
+      "step": 10427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1069680147873684,
+      "learning_rate": 6.298224471157155e-07,
+      "loss": 1.0156,
+      "step": 10428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0259365270513685,
+      "learning_rate": 6.288574659343871e-07,
+      "loss": 1.093,
+      "step": 10429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.046112430341079,
+      "learning_rate": 6.278932005676819e-07,
+      "loss": 1.0188,
+      "step": 10430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9527336128836694,
+      "learning_rate": 6.269296510892542e-07,
+      "loss": 1.0679,
+      "step": 10431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2343633553895095,
+      "learning_rate": 6.259668175727063e-07,
+      "loss": 1.0898,
+      "step": 10432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2381555050535598,
+      "learning_rate": 6.250047000915826e-07,
+      "loss": 0.968,
+      "step": 10433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8327372317571857,
+      "learning_rate": 6.240432987193757e-07,
+      "loss": 1.0986,
+      "step": 10434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.044235184176321,
+      "learning_rate": 6.230826135295231e-07,
+      "loss": 1.1074,
+      "step": 10435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2554980584116775,
+      "learning_rate": 6.221226445954054e-07,
+      "loss": 1.1155,
+      "step": 10436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.934004163161105,
+      "learning_rate": 6.211633919903493e-07,
+      "loss": 0.9729,
+      "step": 10437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1092404553283455,
+      "learning_rate": 6.202048557876273e-07,
+      "loss": 1.0657,
+      "step": 10438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1327614650767,
+      "learning_rate": 6.192470360604585e-07,
+      "loss": 1.1155,
+      "step": 10439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.99283143073031,
+      "learning_rate": 6.182899328820058e-07,
+      "loss": 1.085,
+      "step": 10440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1338491493496328,
+      "learning_rate": 6.17333546325376e-07,
+      "loss": 1.0808,
+      "step": 10441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.005995344875364,
+      "learning_rate": 6.163778764636242e-07,
+      "loss": 1.082,
+      "step": 10442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1471011407102543,
+      "learning_rate": 6.154229233697495e-07,
+      "loss": 1.1025,
+      "step": 10443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2413508503407384,
+      "learning_rate": 6.144686871166949e-07,
+      "loss": 1.1172,
+      "step": 10444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.133709885134011,
+      "learning_rate": 6.135151677773499e-07,
+      "loss": 1.0095,
+      "step": 10445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1270040109438533,
+      "learning_rate": 6.125623654245483e-07,
+      "loss": 0.9978,
+      "step": 10446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5283197671661255,
+      "learning_rate": 6.116102801310719e-07,
+      "loss": 1.0032,
+      "step": 10447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.063078215537699,
+      "learning_rate": 6.106589119696448e-07,
+      "loss": 1.1277,
+      "step": 10448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.043682246425885,
+      "learning_rate": 6.097082610129368e-07,
+      "loss": 1.0305,
+      "step": 10449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1346449865913337,
+      "learning_rate": 6.08758327333564e-07,
+      "loss": 1.054,
+      "step": 10450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.747837595789613,
+      "learning_rate": 6.078091110040852e-07,
+      "loss": 1.0879,
+      "step": 10451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.192900462889835,
+      "learning_rate": 6.068606120970089e-07,
+      "loss": 1.1616,
+      "step": 10452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6723744883326344,
+      "learning_rate": 6.05912830684785e-07,
+      "loss": 1.1089,
+      "step": 10453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.213751517706151,
+      "learning_rate": 6.04965766839809e-07,
+      "loss": 1.042,
+      "step": 10454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.554692363347914,
+      "learning_rate": 6.040194206344241e-07,
+      "loss": 1.1233,
+      "step": 10455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.149549519339035,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 1.0801,
+      "step": 10456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2200722411931415,
+      "learning_rate": 6.021288814315174e-07,
+      "loss": 1.0166,
+      "step": 10457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7947285413209237,
+      "learning_rate": 6.011846885784045e-07,
+      "loss": 1.0801,
+      "step": 10458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2892771962853358,
+      "learning_rate": 6.002412136536972e-07,
+      "loss": 0.9932,
+      "step": 10459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2240249675642643,
+      "learning_rate": 5.992984567294668e-07,
+      "loss": 1.1248,
+      "step": 10460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.885801687772892,
+      "learning_rate": 5.983564178777224e-07,
+      "loss": 1.0183,
+      "step": 10461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2353765715263743,
+      "learning_rate": 5.974150971704252e-07,
+      "loss": 0.9963,
+      "step": 10462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4722301988278264,
+      "learning_rate": 5.964744946794754e-07,
+      "loss": 1.0378,
+      "step": 10463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.414458285453186,
+      "learning_rate": 5.955346104767224e-07,
+      "loss": 1.0381,
+      "step": 10464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5369227130775047,
+      "learning_rate": 5.945954446339574e-07,
+      "loss": 1.072,
+      "step": 10465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4663871731242852,
+      "learning_rate": 5.93656997222919e-07,
+      "loss": 1.0981,
+      "step": 10466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.283282575647311,
+      "learning_rate": 5.927192683152916e-07,
+      "loss": 1.0305,
+      "step": 10467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9589218454405337,
+      "learning_rate": 5.917822579827037e-07,
+      "loss": 1.0801,
+      "step": 10468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3254785442302164,
+      "learning_rate": 5.90845966296728e-07,
+      "loss": 1.0371,
+      "step": 10469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0385448438912426,
+      "learning_rate": 5.89910393328883e-07,
+      "loss": 1.0911,
+      "step": 10470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9229000637920226,
+      "learning_rate": 5.889755391506314e-07,
+      "loss": 1.0891,
+      "step": 10471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1441489066446056,
+      "learning_rate": 5.880414038333859e-07,
+      "loss": 1.0347,
+      "step": 10472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9393661368130934,
+      "learning_rate": 5.871079874484963e-07,
+      "loss": 1.0461,
+      "step": 10473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.225402836362442,
+      "learning_rate": 5.861752900672624e-07,
+      "loss": 1.1011,
+      "step": 10474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4375695083409386,
+      "learning_rate": 5.852433117609313e-07,
+      "loss": 1.0645,
+      "step": 10475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.905440721186627,
+      "learning_rate": 5.843120526006896e-07,
+      "loss": 1.0193,
+      "step": 10476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.062367071306666,
+      "learning_rate": 5.833815126576714e-07,
+      "loss": 1.0977,
+      "step": 10477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1257454127648305,
+      "learning_rate": 5.824516920029566e-07,
+      "loss": 0.9944,
+      "step": 10478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6326553903801093,
+      "learning_rate": 5.815225907075672e-07,
+      "loss": 1.0508,
+      "step": 10479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7307451074035844,
+      "learning_rate": 5.805942088424777e-07,
+      "loss": 0.9731,
+      "step": 10480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0401056229081007,
+      "learning_rate": 5.796665464785988e-07,
+      "loss": 1.0156,
+      "step": 10481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.116667737925835,
+      "learning_rate": 5.787396036867888e-07,
+      "loss": 1.0715,
+      "step": 10482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2200131562115804,
+      "learning_rate": 5.778133805378583e-07,
+      "loss": 1.1094,
+      "step": 10483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.868473305898088,
+      "learning_rate": 5.768878771025488e-07,
+      "loss": 1.0691,
+      "step": 10484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.394132532555057,
+      "learning_rate": 5.759630934515614e-07,
+      "loss": 1.0999,
+      "step": 10485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9546641102631113,
+      "learning_rate": 5.750390296555319e-07,
+      "loss": 1.0588,
+      "step": 10486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.313800869537411,
+      "learning_rate": 5.741156857850449e-07,
+      "loss": 1.1069,
+      "step": 10487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.277235929320542,
+      "learning_rate": 5.731930619106329e-07,
+      "loss": 1.0535,
+      "step": 10488
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.013777902497024,
+      "learning_rate": 5.722711581027695e-07,
+      "loss": 1.0356,
+      "step": 10489
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.382314132255757,
+      "learning_rate": 5.713499744318729e-07,
+      "loss": 1.1038,
+      "step": 10490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.204516413764428,
+      "learning_rate": 5.704295109683089e-07,
+      "loss": 1.011,
+      "step": 10491
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.398178191815411,
+      "learning_rate": 5.695097677823846e-07,
+      "loss": 1.051,
+      "step": 10492
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.405023639636675,
+      "learning_rate": 5.685907449443584e-07,
+      "loss": 1.0,
+      "step": 10493
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.976176823731952,
+      "learning_rate": 5.676724425244262e-07,
+      "loss": 1.0071,
+      "step": 10494
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.803985783608689,
+      "learning_rate": 5.667548605927365e-07,
+      "loss": 1.0515,
+      "step": 10495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3578844389791143,
+      "learning_rate": 5.658379992193785e-07,
+      "loss": 1.0339,
+      "step": 10496
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4033064545030465,
+      "learning_rate": 5.64921858474381e-07,
+      "loss": 1.0647,
+      "step": 10497
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.147580819342136,
+      "learning_rate": 5.640064384277288e-07,
+      "loss": 0.9868,
+      "step": 10498
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9354738853299214,
+      "learning_rate": 5.63091739149344e-07,
+      "loss": 1.0925,
+      "step": 10499
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4158897288096917,
+      "learning_rate": 5.621777607090972e-07,
+      "loss": 1.009,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2079893738408245,
+      "learning_rate": 5.612645031768027e-07,
+      "loss": 1.1038,
+      "step": 10501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6788370494628557,
+      "learning_rate": 5.603519666222168e-07,
+      "loss": 1.0242,
+      "step": 10502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.91919584816748,
+      "learning_rate": 5.594401511150482e-07,
+      "loss": 0.9753,
+      "step": 10503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8544653474614083,
+      "learning_rate": 5.585290567249424e-07,
+      "loss": 1.026,
+      "step": 10504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.125850882546563,
+      "learning_rate": 5.576186835214948e-07,
+      "loss": 1.075,
+      "step": 10505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.620207655511779,
+      "learning_rate": 5.56709031574243e-07,
+      "loss": 1.1072,
+      "step": 10506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6163444571297267,
+      "learning_rate": 5.558001009526715e-07,
+      "loss": 1.0424,
+      "step": 10507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1937531092176257,
+      "learning_rate": 5.548918917262103e-07,
+      "loss": 1.0762,
+      "step": 10508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8923921673414985,
+      "learning_rate": 5.539844039642317e-07,
+      "loss": 1.0178,
+      "step": 10509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.091036573862343,
+      "learning_rate": 5.530776377360536e-07,
+      "loss": 0.9722,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5445117302015117,
+      "learning_rate": 5.521715931109406e-07,
+      "loss": 1.1138,
+      "step": 10511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7318402987044554,
+      "learning_rate": 5.512662701580984e-07,
+      "loss": 1.1006,
+      "step": 10512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.262345915305877,
+      "learning_rate": 5.503616689466829e-07,
+      "loss": 1.0791,
+      "step": 10513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6737525841928407,
+      "learning_rate": 5.49457789545792e-07,
+      "loss": 1.0593,
+      "step": 10514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3058417688045383,
+      "learning_rate": 5.485546320244651e-07,
+      "loss": 1.0637,
+      "step": 10515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2147949643224316,
+      "learning_rate": 5.476521964516957e-07,
+      "loss": 1.0967,
+      "step": 10516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.907679452128225,
+      "learning_rate": 5.46750482896411e-07,
+      "loss": 1.0264,
+      "step": 10517
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0085666991268076,
+      "learning_rate": 5.458494914274903e-07,
+      "loss": 1.01,
+      "step": 10518
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1216952629920405,
+      "learning_rate": 5.449492221137564e-07,
+      "loss": 0.9697,
+      "step": 10519
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.242640918313938,
+      "learning_rate": 5.440496750239755e-07,
+      "loss": 1.0471,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9773071177765766,
+      "learning_rate": 5.4315085022686e-07,
+      "loss": 1.0549,
+      "step": 10521
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6728572297071547,
+      "learning_rate": 5.422527477910666e-07,
+      "loss": 1.0688,
+      "step": 10522
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.549165074104001,
+      "learning_rate": 5.413553677851968e-07,
+      "loss": 1.0674,
+      "step": 10523
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.696416544637072,
+      "learning_rate": 5.404587102777969e-07,
+      "loss": 1.0347,
+      "step": 10524
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.268799325155314,
+      "learning_rate": 5.395627753373567e-07,
+      "loss": 1.0674,
+      "step": 10525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.897607008965362,
+      "learning_rate": 5.386675630323146e-07,
+      "loss": 1.0967,
+      "step": 10526
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.227826956000943,
+      "learning_rate": 5.377730734310494e-07,
+      "loss": 1.0388,
+      "step": 10527
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8769322104512836,
+      "learning_rate": 5.368793066018895e-07,
+      "loss": 1.0649,
+      "step": 10528
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.649539217029089,
+      "learning_rate": 5.35986262613104e-07,
+      "loss": 1.0547,
+      "step": 10529
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7879794488864027,
+      "learning_rate": 5.350939415329049e-07,
+      "loss": 1.0676,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.917128076699001,
+      "learning_rate": 5.342023434294552e-07,
+      "loss": 0.998,
+      "step": 10531
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.57738579051913,
+      "learning_rate": 5.333114683708585e-07,
+      "loss": 1.092,
+      "step": 10532
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.086007634705893,
+      "learning_rate": 5.324213164251657e-07,
+      "loss": 1.0808,
+      "step": 10533
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.296634932140249,
+      "learning_rate": 5.315318876603714e-07,
+      "loss": 1.0959,
+      "step": 10534
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5597049878654237,
+      "learning_rate": 5.306431821444102e-07,
+      "loss": 1.0818,
+      "step": 10535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2753738531980994,
+      "learning_rate": 5.297551999451734e-07,
+      "loss": 1.1304,
+      "step": 10536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3665301541209343,
+      "learning_rate": 5.28867941130482e-07,
+      "loss": 1.0073,
+      "step": 10537
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.29567708465997,
+      "learning_rate": 5.279814057681131e-07,
+      "loss": 1.0317,
+      "step": 10538
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5748417819402163,
+      "learning_rate": 5.270955939257849e-07,
+      "loss": 0.9799,
+      "step": 10539
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4250201081253215,
+      "learning_rate": 5.262105056711575e-07,
+      "loss": 1.0327,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3560149805135087,
+      "learning_rate": 5.253261410718413e-07,
+      "loss": 1.1318,
+      "step": 10541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2132730946250687,
+      "learning_rate": 5.244425001953868e-07,
+      "loss": 1.0566,
+      "step": 10542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3120506785701096,
+      "learning_rate": 5.235595831092921e-07,
+      "loss": 1.0073,
+      "step": 10543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.910430005817825,
+      "learning_rate": 5.226773898809978e-07,
+      "loss": 1.0542,
+      "step": 10544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.968931627364358,
+      "learning_rate": 5.2179592057789e-07,
+      "loss": 1.0496,
+      "step": 10545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1865172159600084,
+      "learning_rate": 5.209151752673003e-07,
+      "loss": 1.0869,
+      "step": 10546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.675035148168541,
+      "learning_rate": 5.200351540165049e-07,
+      "loss": 1.0356,
+      "step": 10547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8903967726989785,
+      "learning_rate": 5.191558568927224e-07,
+      "loss": 1.0933,
+      "step": 10548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.662209334598701,
+      "learning_rate": 5.182772839631223e-07,
+      "loss": 1.0405,
+      "step": 10549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6787142543458784,
+      "learning_rate": 5.173994352948086e-07,
+      "loss": 1.0352,
+      "step": 10550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.579544764370047,
+      "learning_rate": 5.165223109548389e-07,
+      "loss": 1.0271,
+      "step": 10551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0092685209730186,
+      "learning_rate": 5.156459110102119e-07,
+      "loss": 1.0732,
+      "step": 10552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.308345172967217,
+      "learning_rate": 5.147702355278706e-07,
+      "loss": 1.1208,
+      "step": 10553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6526663600758296,
+      "learning_rate": 5.138952845747058e-07,
+      "loss": 0.9868,
+      "step": 10554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1835366684968385,
+      "learning_rate": 5.130210582175477e-07,
+      "loss": 1.0696,
+      "step": 10555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1041468972464927,
+      "learning_rate": 5.121475565231771e-07,
+      "loss": 1.0742,
+      "step": 10556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1037653200809032,
+      "learning_rate": 5.112747795583128e-07,
+      "loss": 1.0955,
+      "step": 10557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.322871689036761,
+      "learning_rate": 5.104027273896239e-07,
+      "loss": 1.0518,
+      "step": 10558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3714800990508156,
+      "learning_rate": 5.095314000837226e-07,
+      "loss": 1.1072,
+      "step": 10559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.189783486548575,
+      "learning_rate": 5.086607977071656e-07,
+      "loss": 1.1113,
+      "step": 10560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.104572639790843,
+      "learning_rate": 5.077909203264509e-07,
+      "loss": 1.0483,
+      "step": 10561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.555207436783078,
+      "learning_rate": 5.069217680080285e-07,
+      "loss": 1.0613,
+      "step": 10562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.64030131403485,
+      "learning_rate": 5.060533408182833e-07,
+      "loss": 1.062,
+      "step": 10563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.078922849696067,
+      "learning_rate": 5.051856388235543e-07,
+      "loss": 1.0535,
+      "step": 10564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7278890630170665,
+      "learning_rate": 5.043186620901186e-07,
+      "loss": 1.0361,
+      "step": 10565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.171374965408077,
+      "learning_rate": 5.034524106842031e-07,
+      "loss": 1.0815,
+      "step": 10566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1239580446138397,
+      "learning_rate": 5.025868846719728e-07,
+      "loss": 1.1001,
+      "step": 10567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2862858303034543,
+      "learning_rate": 5.017220841195425e-07,
+      "loss": 1.0754,
+      "step": 10568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6460353532590406,
+      "learning_rate": 5.008580090929726e-07,
+      "loss": 1.0061,
+      "step": 10569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6939441931767276,
+      "learning_rate": 4.999946596582595e-07,
+      "loss": 1.1062,
+      "step": 10570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.311109106601939,
+      "learning_rate": 4.991320358813556e-07,
+      "loss": 1.1191,
+      "step": 10571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9138099456141453,
+      "learning_rate": 4.982701378281496e-07,
+      "loss": 1.0281,
+      "step": 10572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7141957456162316,
+      "learning_rate": 4.974089655644776e-07,
+      "loss": 1.0767,
+      "step": 10573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9979798315672177,
+      "learning_rate": 4.965485191561215e-07,
+      "loss": 1.1047,
+      "step": 10574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.199394838088958,
+      "learning_rate": 4.956887986688064e-07,
+      "loss": 1.0681,
+      "step": 10575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.896940365683211,
+      "learning_rate": 4.94829804168202e-07,
+      "loss": 1.0132,
+      "step": 10576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3831629164160337,
+      "learning_rate": 4.939715357199215e-07,
+      "loss": 1.1021,
+      "step": 10577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2402342012786964,
+      "learning_rate": 4.931139933895235e-07,
+      "loss": 1.0217,
+      "step": 10578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.593303044427476,
+      "learning_rate": 4.922571772425133e-07,
+      "loss": 0.9934,
+      "step": 10579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8802720746578854,
+      "learning_rate": 4.914010873443375e-07,
+      "loss": 1.0803,
+      "step": 10580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0946464579366557,
+      "learning_rate": 4.905457237603883e-07,
+      "loss": 1.1006,
+      "step": 10581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6778580584774643,
+      "learning_rate": 4.896910865560045e-07,
+      "loss": 1.0903,
+      "step": 10582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2051281628721413,
+      "learning_rate": 4.888371757964649e-07,
+      "loss": 1.0708,
+      "step": 10583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2078338116439795,
+      "learning_rate": 4.879839915469974e-07,
+      "loss": 1.0383,
+      "step": 10584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7315158226273852,
+      "learning_rate": 4.871315338727711e-07,
+      "loss": 1.0066,
+      "step": 10585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.763610454495154,
+      "learning_rate": 4.862798028389015e-07,
+      "loss": 1.0896,
+      "step": 10586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3997374047003066,
+      "learning_rate": 4.854287985104488e-07,
+      "loss": 1.0461,
+      "step": 10587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8066489244264017,
+      "learning_rate": 4.845785209524156e-07,
+      "loss": 1.0264,
+      "step": 10588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9134224965953748,
+      "learning_rate": 4.83728970229752e-07,
+      "loss": 1.0293,
+      "step": 10589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.082717995695327,
+      "learning_rate": 4.828801464073507e-07,
+      "loss": 1.0347,
+      "step": 10590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3740623452555845,
+      "learning_rate": 4.820320495500464e-07,
+      "loss": 1.0747,
+      "step": 10591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.84538940302063,
+      "learning_rate": 4.811846797226239e-07,
+      "loss": 0.9846,
+      "step": 10592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7042637816306243,
+      "learning_rate": 4.803380369898092e-07,
+      "loss": 0.9644,
+      "step": 10593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9259432370935183,
+      "learning_rate": 4.794921214162717e-07,
+      "loss": 1.0266,
+      "step": 10594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2843928596864407,
+      "learning_rate": 4.786469330666288e-07,
+      "loss": 1.0005,
+      "step": 10595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6110466699845367,
+      "learning_rate": 4.778024720054375e-07,
+      "loss": 1.0286,
+      "step": 10596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6235750256185137,
+      "learning_rate": 4.76958738297204e-07,
+      "loss": 1.0464,
+      "step": 10597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4752928873418485,
+      "learning_rate": 4.76115732006377e-07,
+      "loss": 1.0083,
+      "step": 10598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3229360133936248,
+      "learning_rate": 4.7527345319734573e-07,
+      "loss": 1.0483,
+      "step": 10599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2968852237344026,
+      "learning_rate": 4.744319019344534e-07,
+      "loss": 1.0884,
+      "step": 10600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5836310218964735,
+      "learning_rate": 4.7359107828197747e-07,
+      "loss": 1.1572,
+      "step": 10601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.191107105145602,
+      "learning_rate": 4.7275098230414874e-07,
+      "loss": 1.0691,
+      "step": 10602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.765315228930597,
+      "learning_rate": 4.7191161406513363e-07,
+      "loss": 1.092,
+      "step": 10603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8816016325588443,
+      "learning_rate": 4.7107297362904757e-07,
+      "loss": 1.0505,
+      "step": 10604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2783223400871044,
+      "learning_rate": 4.7023506105995266e-07,
+      "loss": 1.0488,
+      "step": 10605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.396087152174575,
+      "learning_rate": 4.6939787642184986e-07,
+      "loss": 1.1299,
+      "step": 10606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1372733975229705,
+      "learning_rate": 4.685614197786914e-07,
+      "loss": 1.0654,
+      "step": 10607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4767375009938952,
+      "learning_rate": 4.677256911943673e-07,
+      "loss": 1.0479,
+      "step": 10608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.262530705616455,
+      "learning_rate": 4.668906907327142e-07,
+      "loss": 1.0391,
+      "step": 10609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3666547891286864,
+      "learning_rate": 4.660564184575156e-07,
+      "loss": 1.0688,
+      "step": 10610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6160228923345663,
+      "learning_rate": 4.6522287443249495e-07,
+      "loss": 1.1013,
+      "step": 10611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.244779679876198,
+      "learning_rate": 4.6439005872132457e-07,
+      "loss": 1.1089,
+      "step": 10612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9560556515710052,
+      "learning_rate": 4.6355797138761816e-07,
+      "loss": 1.0366,
+      "step": 10613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8893032028891823,
+      "learning_rate": 4.627266124949348e-07,
+      "loss": 1.0361,
+      "step": 10614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.843903066041396,
+      "learning_rate": 4.618959821067792e-07,
+      "loss": 1.064,
+      "step": 10615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.455780583652293,
+      "learning_rate": 4.6106608028659625e-07,
+      "loss": 1.0637,
+      "step": 10616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5743664900544307,
+      "learning_rate": 4.602369070977808e-07,
+      "loss": 1.0947,
+      "step": 10617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4726628712718646,
+      "learning_rate": 4.5940846260366766e-07,
+      "loss": 1.0059,
+      "step": 10618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.780160999695416,
+      "learning_rate": 4.585807468675363e-07,
+      "loss": 1.1187,
+      "step": 10619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.089344829542815,
+      "learning_rate": 4.577537599526149e-07,
+      "loss": 1.0579,
+      "step": 10620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3491003881499326,
+      "learning_rate": 4.569275019220709e-07,
+      "loss": 1.0046,
+      "step": 10621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7471517579779228,
+      "learning_rate": 4.5610197283901924e-07,
+      "loss": 1.0405,
+      "step": 10622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2204879170652925,
+      "learning_rate": 4.5527717276651616e-07,
+      "loss": 1.0613,
+      "step": 10623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.490568171999506,
+      "learning_rate": 4.5445310176756466e-07,
+      "loss": 1.0911,
+      "step": 10624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.352569145192122,
+      "learning_rate": 4.5362975990511203e-07,
+      "loss": 1.0142,
+      "step": 10625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.950664066166123,
+      "learning_rate": 4.528071472420503e-07,
+      "loss": 0.9949,
+      "step": 10626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.132415615541251,
+      "learning_rate": 4.519852638412103e-07,
+      "loss": 1.0698,
+      "step": 10627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.837517395142137,
+      "learning_rate": 4.511641097653774e-07,
+      "loss": 1.041,
+      "step": 10628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.460079593626482,
+      "learning_rate": 4.5034368507727134e-07,
+      "loss": 1.0251,
+      "step": 10629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3853119454937572,
+      "learning_rate": 4.4952398983956205e-07,
+      "loss": 1.0886,
+      "step": 10630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2639290644373222,
+      "learning_rate": 4.487050241148616e-07,
+      "loss": 1.1287,
+      "step": 10631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9147600904678166,
+      "learning_rate": 4.478867879657256e-07,
+      "loss": 1.0601,
+      "step": 10632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0446149954022426,
+      "learning_rate": 4.4706928145465735e-07,
+      "loss": 1.084,
+      "step": 10633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1833837480848004,
+      "learning_rate": 4.462525046440991e-07,
+      "loss": 1.1301,
+      "step": 10634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0176381571208606,
+      "learning_rate": 4.454364575964443e-07,
+      "loss": 1.0769,
+      "step": 10635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.040269019665798,
+      "learning_rate": 4.4462114037402304e-07,
+      "loss": 1.05,
+      "step": 10636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0691363507651674,
+      "learning_rate": 4.4380655303911227e-07,
+      "loss": 0.9939,
+      "step": 10637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.162541270876106,
+      "learning_rate": 4.429926956539399e-07,
+      "loss": 1.0508,
+      "step": 10638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8840960635680655,
+      "learning_rate": 4.421795682806662e-07,
+      "loss": 0.9961,
+      "step": 10639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0166328151623913,
+      "learning_rate": 4.4136717098140715e-07,
+      "loss": 1.0271,
+      "step": 10640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.101890749280775,
+      "learning_rate": 4.405555038182141e-07,
+      "loss": 1.0801,
+      "step": 10641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.171989811736114,
+      "learning_rate": 4.3974456685308864e-07,
+      "loss": 1.0688,
+      "step": 10642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.294696771129267,
+      "learning_rate": 4.3893436014797233e-07,
+      "loss": 1.0342,
+      "step": 10643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.310805201649505,
+      "learning_rate": 4.3812488376475226e-07,
+      "loss": 1.0168,
+      "step": 10644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.231604226500016,
+      "learning_rate": 4.3731613776526236e-07,
+      "loss": 1.1777,
+      "step": 10645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7701232550486736,
+      "learning_rate": 4.365081222112788e-07,
+      "loss": 1.0425,
+      "step": 10646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2542893420631223,
+      "learning_rate": 4.357008371645188e-07,
+      "loss": 1.082,
+      "step": 10647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2309058152039944,
+      "learning_rate": 4.34894282686652e-07,
+      "loss": 1.042,
+      "step": 10648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.886707306208143,
+      "learning_rate": 4.3408845883928017e-07,
+      "loss": 1.0288,
+      "step": 10649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.231393902085211,
+      "learning_rate": 4.33283365683963e-07,
+      "loss": 1.0747,
+      "step": 10650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4125708361865477,
+      "learning_rate": 4.324790032821935e-07,
+      "loss": 1.0488,
+      "step": 10651
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2916690518784932,
+      "learning_rate": 4.316753716954125e-07,
+      "loss": 0.9871,
+      "step": 10652
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.447291286037995,
+      "learning_rate": 4.3087247098500763e-07,
+      "loss": 0.9951,
+      "step": 10653
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3574507003620835,
+      "learning_rate": 4.300703012123075e-07,
+      "loss": 1.0862,
+      "step": 10654
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0131427403278197,
+      "learning_rate": 4.2926886243858655e-07,
+      "loss": 1.0957,
+      "step": 10655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.961667440009467,
+      "learning_rate": 4.284681547250624e-07,
+      "loss": 1.0349,
+      "step": 10656
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.978315208993422,
+      "learning_rate": 4.27668178132894e-07,
+      "loss": 1.0022,
+      "step": 10657
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0821286765958127,
+      "learning_rate": 4.268689327231923e-07,
+      "loss": 1.043,
+      "step": 10658
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7458331749752034,
+      "learning_rate": 4.2607041855700525e-07,
+      "loss": 1.1079,
+      "step": 10659
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2074765905025315,
+      "learning_rate": 4.252726356953274e-07,
+      "loss": 1.0867,
+      "step": 10660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6714744365209975,
+      "learning_rate": 4.244755841990977e-07,
+      "loss": 0.9907,
+      "step": 10661
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.756809896238053,
+      "learning_rate": 4.236792641291998e-07,
+      "loss": 1.1318,
+      "step": 10662
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2341608848223324,
+      "learning_rate": 4.2288367554645936e-07,
+      "loss": 0.9819,
+      "step": 10663
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.16505825289376,
+      "learning_rate": 4.2208881851164786e-07,
+      "loss": 1.0688,
+      "step": 10664
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1775000948318795,
+      "learning_rate": 4.2129469308547887e-07,
+      "loss": 1.0681,
+      "step": 10665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.951825317453308,
+      "learning_rate": 4.2050129932861394e-07,
+      "loss": 1.0776,
+      "step": 10666
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.376258173166463,
+      "learning_rate": 4.1970863730165567e-07,
+      "loss": 1.0303,
+      "step": 10667
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.608078544888041,
+      "learning_rate": 4.189167070651534e-07,
+      "loss": 1.0168,
+      "step": 10668
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5449634310490827,
+      "learning_rate": 4.1812550867959545e-07,
+      "loss": 1.0098,
+      "step": 10669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5738223538741476,
+      "learning_rate": 4.173350422054168e-07,
+      "loss": 1.0728,
+      "step": 10670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3314761495172216,
+      "learning_rate": 4.1654530770300127e-07,
+      "loss": 1.0166,
+      "step": 10671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9007295389636427,
+      "learning_rate": 4.1575630523266963e-07,
+      "loss": 1.0764,
+      "step": 10672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0115127789908525,
+      "learning_rate": 4.149680348546914e-07,
+      "loss": 1.0571,
+      "step": 10673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.339892605792305,
+      "learning_rate": 4.141804966292795e-07,
+      "loss": 0.981,
+      "step": 10674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9431910206432654,
+      "learning_rate": 4.1339369061658696e-07,
+      "loss": 1.0793,
+      "step": 10675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.140406472280272,
+      "learning_rate": 4.126076168767168e-07,
+      "loss": 1.053,
+      "step": 10676
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5046435014298822,
+      "learning_rate": 4.11822275469711e-07,
+      "loss": 1.0508,
+      "step": 10677
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.111748878851284,
+      "learning_rate": 4.110376664555593e-07,
+      "loss": 1.0496,
+      "step": 10678
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.864844850899366,
+      "learning_rate": 4.1025378989419383e-07,
+      "loss": 1.053,
+      "step": 10679
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7536440279752807,
+      "learning_rate": 4.0947064584549e-07,
+      "loss": 1.0408,
+      "step": 10680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.45277320247863,
+      "learning_rate": 4.086882343692711e-07,
+      "loss": 1.1223,
+      "step": 10681
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.933965414751228,
+      "learning_rate": 4.079065555252981e-07,
+      "loss": 1.0737,
+      "step": 10682
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.04422682826273,
+      "learning_rate": 4.0712560937328226e-07,
+      "loss": 1.0181,
+      "step": 10683
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1688587932208163,
+      "learning_rate": 4.063453959728747e-07,
+      "loss": 0.9824,
+      "step": 10684
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9449571569370696,
+      "learning_rate": 4.055659153836711e-07,
+      "loss": 1.0747,
+      "step": 10685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1778989825729753,
+      "learning_rate": 4.0478716766521373e-07,
+      "loss": 1.1042,
+      "step": 10686
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9585115561716324,
+      "learning_rate": 4.040091528769874e-07,
+      "loss": 1.0237,
+      "step": 10687
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.16547471587732,
+      "learning_rate": 4.0323187107842e-07,
+      "loss": 1.063,
+      "step": 10688
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0244216800294454,
+      "learning_rate": 4.024553223288841e-07,
+      "loss": 1.0403,
+      "step": 10689
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.347746395004066,
+      "learning_rate": 4.016795066876955e-07,
+      "loss": 0.9998,
+      "step": 10690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9098157967567913,
+      "learning_rate": 4.00904424214118e-07,
+      "loss": 1.0476,
+      "step": 10691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7496505904451762,
+      "learning_rate": 4.001300749673531e-07,
+      "loss": 1.0532,
+      "step": 10692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.370720200186483,
+      "learning_rate": 3.9935645900654906e-07,
+      "loss": 1.1079,
+      "step": 10693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.056852904008165,
+      "learning_rate": 3.985835763908019e-07,
+      "loss": 1.1018,
+      "step": 10694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.099068044767697,
+      "learning_rate": 3.9781142717914665e-07,
+      "loss": 1.0488,
+      "step": 10695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8912896490846975,
+      "learning_rate": 3.9704001143056284e-07,
+      "loss": 1.0732,
+      "step": 10696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.112563215715143,
+      "learning_rate": 3.9626932920397653e-07,
+      "loss": 1.0879,
+      "step": 10697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.185009821553892,
+      "learning_rate": 3.954993805582541e-07,
+      "loss": 0.981,
+      "step": 10698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4854777104333152,
+      "learning_rate": 3.9473016555221064e-07,
+      "loss": 1.0513,
+      "step": 10699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.981939389029706,
+      "learning_rate": 3.9396168424460256e-07,
+      "loss": 1.052,
+      "step": 10700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.062741042189173,
+      "learning_rate": 3.9319393669412733e-07,
+      "loss": 1.0452,
+      "step": 10701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4396026412708762,
+      "learning_rate": 3.924269229594335e-07,
+      "loss": 1.0779,
+      "step": 10702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1605998960022528,
+      "learning_rate": 3.9166064309910545e-07,
+      "loss": 0.9927,
+      "step": 10703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.530277415757424,
+      "learning_rate": 3.908950971716774e-07,
+      "loss": 1.0403,
+      "step": 10704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2650928907246732,
+      "learning_rate": 3.901302852356259e-07,
+      "loss": 0.9839,
+      "step": 10705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.470899942610364,
+      "learning_rate": 3.893662073493698e-07,
+      "loss": 1.1018,
+      "step": 10706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8821595969857268,
+      "learning_rate": 3.8860286357127463e-07,
+      "loss": 1.072,
+      "step": 10707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9728610701330234,
+      "learning_rate": 3.878402539596482e-07,
+      "loss": 1.1021,
+      "step": 10708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3278396751660138,
+      "learning_rate": 3.8707837857274164e-07,
+      "loss": 1.0723,
+      "step": 10709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8596324182679407,
+      "learning_rate": 3.863172374687496e-07,
+      "loss": 1.0654,
+      "step": 10710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.949825446459864,
+      "learning_rate": 3.855568307058155e-07,
+      "loss": 0.9792,
+      "step": 10711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8063175132545055,
+      "learning_rate": 3.847971583420207e-07,
+      "loss": 1.0828,
+      "step": 10712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5054847024302127,
+      "learning_rate": 3.8403822043539097e-07,
+      "loss": 1.1208,
+      "step": 10713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7471933427159363,
+      "learning_rate": 3.8328001704390107e-07,
+      "loss": 1.0054,
+      "step": 10714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.255982498133307,
+      "learning_rate": 3.8252254822546576e-07,
+      "loss": 1.1089,
+      "step": 10715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8958869726036798,
+      "learning_rate": 3.8176581403794323e-07,
+      "loss": 1.0879,
+      "step": 10716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9997880683548845,
+      "learning_rate": 3.810098145391372e-07,
+      "loss": 1.0154,
+      "step": 10717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0742454638223253,
+      "learning_rate": 3.8025454978679267e-07,
+      "loss": 1.0652,
+      "step": 10718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.790882882925107,
+      "learning_rate": 3.7950001983860343e-07,
+      "loss": 1.0869,
+      "step": 10719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.966699028771691,
+      "learning_rate": 3.7874622475220336e-07,
+      "loss": 1.0654,
+      "step": 10720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6306818165335533,
+      "learning_rate": 3.7799316458517086e-07,
+      "loss": 1.0115,
+      "step": 10721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.357394964244549,
+      "learning_rate": 3.772408393950289e-07,
+      "loss": 1.0344,
+      "step": 10722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1864174997078933,
+      "learning_rate": 3.7648924923924137e-07,
+      "loss": 1.1013,
+      "step": 10723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3648790781061044,
+      "learning_rate": 3.757383941752213e-07,
+      "loss": 1.0574,
+      "step": 10724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.08518690867778,
+      "learning_rate": 3.749882742603228e-07,
+      "loss": 1.0342,
+      "step": 10725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.075147159595543,
+      "learning_rate": 3.7423888955184116e-07,
+      "loss": 1.0144,
+      "step": 10726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9646807098510704,
+      "learning_rate": 3.734902401070206e-07,
+      "loss": 1.0217,
+      "step": 10727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3176486973755597,
+      "learning_rate": 3.727423259830465e-07,
+      "loss": 1.0261,
+      "step": 10728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.336860526319415,
+      "learning_rate": 3.7199514723704754e-07,
+      "loss": 1.1021,
+      "step": 10729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.910759591437913,
+      "learning_rate": 3.71248703926097e-07,
+      "loss": 1.0544,
+      "step": 10730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.999095233938532,
+      "learning_rate": 3.705029961072115e-07,
+      "loss": 0.9421,
+      "step": 10731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.16856547593483,
+      "learning_rate": 3.6975802383735324e-07,
+      "loss": 1.0679,
+      "step": 10732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6429536434776,
+      "learning_rate": 3.690137871734256e-07,
+      "loss": 0.9958,
+      "step": 10733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0899947755872246,
+      "learning_rate": 3.6827028617227753e-07,
+      "loss": 1.0662,
+      "step": 10734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1869302814915135,
+      "learning_rate": 3.6752752089070254e-07,
+      "loss": 1.0469,
+      "step": 10735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2552669523365307,
+      "learning_rate": 3.6678549138543405e-07,
+      "loss": 1.0254,
+      "step": 10736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4169324975688826,
+      "learning_rate": 3.6604419771315456e-07,
+      "loss": 1.054,
+      "step": 10737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.665109563924493,
+      "learning_rate": 3.6530363993048654e-07,
+      "loss": 1.0461,
+      "step": 10738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.235390291865906,
+      "learning_rate": 3.64563818093997e-07,
+      "loss": 1.0515,
+      "step": 10739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1491387714158705,
+      "learning_rate": 3.638247322601995e-07,
+      "loss": 1.0449,
+      "step": 10740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1300415518047546,
+      "learning_rate": 3.630863824855468e-07,
+      "loss": 1.0444,
+      "step": 10741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2334076683398867,
+      "learning_rate": 3.6234876882643934e-07,
+      "loss": 1.1272,
+      "step": 10742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1770631201940107,
+      "learning_rate": 3.616118913392186e-07,
+      "loss": 1.0278,
+      "step": 10743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.247423397699476,
+      "learning_rate": 3.6087575008016963e-07,
+      "loss": 1.0002,
+      "step": 10744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8131164767562185,
+      "learning_rate": 3.601403451055252e-07,
+      "loss": 1.0823,
+      "step": 10745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3101116593425175,
+      "learning_rate": 3.5940567647145817e-07,
+      "loss": 1.1108,
+      "step": 10746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3459261103409754,
+      "learning_rate": 3.58671744234087e-07,
+      "loss": 1.0625,
+      "step": 10747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2107463423591085,
+      "learning_rate": 3.5793854844947127e-07,
+      "loss": 1.1077,
+      "step": 10748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.26118092571402,
+      "learning_rate": 3.572060891736184e-07,
+      "loss": 1.1504,
+      "step": 10749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0411288466045563,
+      "learning_rate": 3.564743664624759e-07,
+      "loss": 1.0527,
+      "step": 10750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9855562473091615,
+      "learning_rate": 3.557433803719357e-07,
+      "loss": 1.05,
+      "step": 10751
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2688281043494785,
+      "learning_rate": 3.550131309578353e-07,
+      "loss": 1.1079,
+      "step": 10752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0553609293669455,
+      "learning_rate": 3.542836182759546e-07,
+      "loss": 1.0134,
+      "step": 10753
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.063979725440615,
+      "learning_rate": 3.5355484238201676e-07,
+      "loss": 1.0835,
+      "step": 10754
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.344171000166025,
+      "learning_rate": 3.528268033316928e-07,
+      "loss": 1.0632,
+      "step": 10755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7509715481957744,
+      "learning_rate": 3.5209950118058835e-07,
+      "loss": 1.1008,
+      "step": 10756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5566478282189364,
+      "learning_rate": 3.5137293598426214e-07,
+      "loss": 1.021,
+      "step": 10757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.744592998760418,
+      "learning_rate": 3.50647107798211e-07,
+      "loss": 1.0964,
+      "step": 10758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2855470763079317,
+      "learning_rate": 3.499220166778783e-07,
+      "loss": 1.022,
+      "step": 10759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.18553175241423,
+      "learning_rate": 3.4919766267865086e-07,
+      "loss": 1.0725,
+      "step": 10760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2072443435373046,
+      "learning_rate": 3.4847404585585665e-07,
+      "loss": 1.0264,
+      "step": 10761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8489723042500628,
+      "learning_rate": 3.4775116626477034e-07,
+      "loss": 1.0522,
+      "step": 10762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.383812823081437,
+      "learning_rate": 3.4702902396061e-07,
+      "loss": 1.0894,
+      "step": 10763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.15064522816496,
+      "learning_rate": 3.463076189985326e-07,
+      "loss": 1.0352,
+      "step": 10764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1557926083209393,
+      "learning_rate": 3.455869514336474e-07,
+      "loss": 1.0396,
+      "step": 10765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.086476059370842,
+      "learning_rate": 3.448670213210004e-07,
+      "loss": 1.0566,
+      "step": 10766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.732998632007916,
+      "learning_rate": 3.4414782871558193e-07,
+      "loss": 1.041,
+      "step": 10767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.262055098015049,
+      "learning_rate": 3.4342937367233156e-07,
+      "loss": 1.0876,
+      "step": 10768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.050020499904375,
+      "learning_rate": 3.42711656246123e-07,
+      "loss": 1.0798,
+      "step": 10769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7701653348521047,
+      "learning_rate": 3.4199467649178476e-07,
+      "loss": 1.0696,
+      "step": 10770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2454536094565345,
+      "learning_rate": 3.412784344640796e-07,
+      "loss": 1.0889,
+      "step": 10771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6535132046720773,
+      "learning_rate": 3.4056293021771824e-07,
+      "loss": 1.1196,
+      "step": 10772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9267687693923157,
+      "learning_rate": 3.3984816380735477e-07,
+      "loss": 1.0278,
+      "step": 10773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.696444276098179,
+      "learning_rate": 3.3913413528758877e-07,
+      "loss": 1.1194,
+      "step": 10774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8637084028381854,
+      "learning_rate": 3.3842084471295777e-07,
+      "loss": 1.0273,
+      "step": 10775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1377652740307025,
+      "learning_rate": 3.377082921379482e-07,
+      "loss": 1.0698,
+      "step": 10776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9306311783475807,
+      "learning_rate": 3.3699647761698763e-07,
+      "loss": 1.0925,
+      "step": 10777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.957444904702785,
+      "learning_rate": 3.362854012044503e-07,
+      "loss": 0.9961,
+      "step": 10778
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.179515610056817,
+      "learning_rate": 3.3557506295464724e-07,
+      "loss": 1.0222,
+      "step": 10779
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.195686169762834,
+      "learning_rate": 3.348654629218429e-07,
+      "loss": 1.0498,
+      "step": 10780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3340632776266084,
+      "learning_rate": 3.3415660116023817e-07,
+      "loss": 0.9982,
+      "step": 10781
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.363336152698577,
+      "learning_rate": 3.334484777239777e-07,
+      "loss": 1.0627,
+      "step": 10782
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.057180579956014,
+      "learning_rate": 3.327410926671526e-07,
+      "loss": 1.0747,
+      "step": 10783
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0709307085520443,
+      "learning_rate": 3.320344460437963e-07,
+      "loss": 1.0315,
+      "step": 10784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0742100502845946,
+      "learning_rate": 3.313285379078868e-07,
+      "loss": 1.0613,
+      "step": 10785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4156146482117085,
+      "learning_rate": 3.306233683133442e-07,
+      "loss": 1.0496,
+      "step": 10786
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1167212353125633,
+      "learning_rate": 3.299189373140321e-07,
+      "loss": 1.04,
+      "step": 10787
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2074724900374356,
+      "learning_rate": 3.29215244963762e-07,
+      "loss": 1.0166,
+      "step": 10788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3271840676957463,
+      "learning_rate": 3.285122913162797e-07,
+      "loss": 1.0088,
+      "step": 10789
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.768122973245121,
+      "learning_rate": 3.2781007642528453e-07,
+      "loss": 1.0234,
+      "step": 10790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.965996060645364,
+      "learning_rate": 3.271086003444146e-07,
+      "loss": 1.0649,
+      "step": 10791
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2359936269508274,
+      "learning_rate": 3.264078631272505e-07,
+      "loss": 1.0703,
+      "step": 10792
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.34075699068512,
+      "learning_rate": 3.2570786482732043e-07,
+      "loss": 1.0288,
+      "step": 10793
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3542530359089064,
+      "learning_rate": 3.2500860549809274e-07,
+      "loss": 1.1382,
+      "step": 10794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.924861885667383,
+      "learning_rate": 3.2431008519297927e-07,
+      "loss": 1.1091,
+      "step": 10795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9760280095456757,
+      "learning_rate": 3.2361230396533715e-07,
+      "loss": 1.0532,
+      "step": 10796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.15931853810253,
+      "learning_rate": 3.229152618684661e-07,
+      "loss": 1.0273,
+      "step": 10797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.26407967458956,
+      "learning_rate": 3.222189589556102e-07,
+      "loss": 1.0371,
+      "step": 10798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8995173370390788,
+      "learning_rate": 3.215233952799579e-07,
+      "loss": 1.0767,
+      "step": 10799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.555641679240209,
+      "learning_rate": 3.208285708946357e-07,
+      "loss": 1.0842,
+      "step": 10800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5097858945009706,
+      "learning_rate": 3.2013448585272333e-07,
+      "loss": 1.0923,
+      "step": 10801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.167228009762813,
+      "learning_rate": 3.1944114020723393e-07,
+      "loss": 1.0784,
+      "step": 10802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2442643100671207,
+      "learning_rate": 3.187485340111307e-07,
+      "loss": 1.0847,
+      "step": 10803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.859714697634607,
+      "learning_rate": 3.18056667317318e-07,
+      "loss": 1.0315,
+      "step": 10804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8742925575115104,
+      "learning_rate": 3.173655401786435e-07,
+      "loss": 0.978,
+      "step": 10805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5888347389867015,
+      "learning_rate": 3.166751526478995e-07,
+      "loss": 1.0437,
+      "step": 10806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.259392157471865,
+      "learning_rate": 3.1598550477782263e-07,
+      "loss": 1.1077,
+      "step": 10807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9802824659927643,
+      "learning_rate": 3.1529659662108967e-07,
+      "loss": 1.0217,
+      "step": 10808
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.493361728371603,
+      "learning_rate": 3.14608428230323e-07,
+      "loss": 1.0012,
+      "step": 10809
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.056091307603335,
+      "learning_rate": 3.139209996580872e-07,
+      "loss": 1.1003,
+      "step": 10810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3234257917776326,
+      "learning_rate": 3.1323431095689473e-07,
+      "loss": 1.0815,
+      "step": 10811
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0597018424556963,
+      "learning_rate": 3.125483621791958e-07,
+      "loss": 1.0933,
+      "step": 10812
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.634203795022346,
+      "learning_rate": 3.118631533773886e-07,
+      "loss": 1.0146,
+      "step": 10813
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.004089064671602,
+      "learning_rate": 3.111786846038123e-07,
+      "loss": 1.0066,
+      "step": 10814
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8633313638131925,
+      "learning_rate": 3.104949559107473e-07,
+      "loss": 1.0684,
+      "step": 10815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2366689587137203,
+      "learning_rate": 3.098119673504241e-07,
+      "loss": 1.0222,
+      "step": 10816
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3486709150347616,
+      "learning_rate": 3.0912971897500867e-07,
+      "loss": 1.0249,
+      "step": 10817
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1039107302557283,
+      "learning_rate": 3.0844821083661826e-07,
+      "loss": 1.1082,
+      "step": 10818
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.581424306985502,
+      "learning_rate": 3.07767442987309e-07,
+      "loss": 1.0066,
+      "step": 10819
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.157103497177184,
+      "learning_rate": 3.0708741547907927e-07,
+      "loss": 1.0498,
+      "step": 10820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.207119446673105,
+      "learning_rate": 3.0640812836387645e-07,
+      "loss": 1.0732,
+      "step": 10821
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6929891059830733,
+      "learning_rate": 3.0572958169358455e-07,
+      "loss": 1.0745,
+      "step": 10822
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.200172770260437,
+      "learning_rate": 3.050517755200355e-07,
+      "loss": 1.0898,
+      "step": 10823
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.091170700304502,
+      "learning_rate": 3.043747098950045e-07,
+      "loss": 1.0366,
+      "step": 10824
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1836343451117464,
+      "learning_rate": 3.03698384870208e-07,
+      "loss": 1.0227,
+      "step": 10825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1621407147386607,
+      "learning_rate": 3.0302280049730793e-07,
+      "loss": 1.0498,
+      "step": 10826
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.98540909358885,
+      "learning_rate": 3.0234795682790753e-07,
+      "loss": 1.0999,
+      "step": 10827
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.226981091204981,
+      "learning_rate": 3.016738539135566e-07,
+      "loss": 1.1128,
+      "step": 10828
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.550445157009086,
+      "learning_rate": 3.010004918057452e-07,
+      "loss": 1.02,
+      "step": 10829
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2378562349521998,
+      "learning_rate": 3.003278705559065e-07,
+      "loss": 1.0691,
+      "step": 10830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0049609628779246,
+      "learning_rate": 2.996559902154228e-07,
+      "loss": 1.0173,
+      "step": 10831
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.299315183565061,
+      "learning_rate": 2.989848508356119e-07,
+      "loss": 1.0193,
+      "step": 10832
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5865690885649744,
+      "learning_rate": 2.983144524677406e-07,
+      "loss": 1.0154,
+      "step": 10833
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.979002811267241,
+      "learning_rate": 2.9764479516301794e-07,
+      "loss": 1.0659,
+      "step": 10834
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3002734915983596,
+      "learning_rate": 2.9697587897259296e-07,
+      "loss": 1.0435,
+      "step": 10835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0105060723148047,
+      "learning_rate": 2.963077039475637e-07,
+      "loss": 1.0264,
+      "step": 10836
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0650063258925972,
+      "learning_rate": 2.9564027013896713e-07,
+      "loss": 1.0479,
+      "step": 10837
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.195889984325533,
+      "learning_rate": 2.9497357759778467e-07,
+      "loss": 1.0559,
+      "step": 10838
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.35178579379116,
+      "learning_rate": 2.943076263749445e-07,
+      "loss": 1.0652,
+      "step": 10839
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7068255555010015,
+      "learning_rate": 2.936424165213125e-07,
+      "loss": 1.0742,
+      "step": 10840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.445645355183167,
+      "learning_rate": 2.929779480877015e-07,
+      "loss": 1.0459,
+      "step": 10841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.070067769348209,
+      "learning_rate": 2.923142211248664e-07,
+      "loss": 1.0959,
+      "step": 10842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3042709006128415,
+      "learning_rate": 2.916512356835066e-07,
+      "loss": 0.9768,
+      "step": 10843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1937950519521423,
+      "learning_rate": 2.9098899181426387e-07,
+      "loss": 0.9668,
+      "step": 10844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1286063651820437,
+      "learning_rate": 2.903274895677255e-07,
+      "loss": 0.9924,
+      "step": 10845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.035317813256602,
+      "learning_rate": 2.896667289944166e-07,
+      "loss": 1.051,
+      "step": 10846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9194618372229684,
+      "learning_rate": 2.890067101448135e-07,
+      "loss": 0.9863,
+      "step": 10847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.016756624291387,
+      "learning_rate": 2.8834743306932924e-07,
+      "loss": 1.0127,
+      "step": 10848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4181964358379258,
+      "learning_rate": 2.8768889781832343e-07,
+      "loss": 1.0222,
+      "step": 10849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4062895101997595,
+      "learning_rate": 2.870311044420959e-07,
+      "loss": 1.0623,
+      "step": 10850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0737485841152132,
+      "learning_rate": 2.863740529908965e-07,
+      "loss": 1.0298,
+      "step": 10851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.503697189741288,
+      "learning_rate": 2.8571774351491277e-07,
+      "loss": 1.0464,
+      "step": 10852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.914172003340271,
+      "learning_rate": 2.8506217606427354e-07,
+      "loss": 1.0696,
+      "step": 10853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.208706782094864,
+      "learning_rate": 2.844073506890599e-07,
+      "loss": 1.0525,
+      "step": 10854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1240194977404263,
+      "learning_rate": 2.837532674392862e-07,
+      "loss": 1.0222,
+      "step": 10855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2667036314985207,
+      "learning_rate": 2.830999263649159e-07,
+      "loss": 1.0015,
+      "step": 10856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8952957416601146,
+      "learning_rate": 2.824473275158557e-07,
+      "loss": 1.0596,
+      "step": 10857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4033498235205872,
+      "learning_rate": 2.817954709419513e-07,
+      "loss": 1.0898,
+      "step": 10858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.418080403071035,
+      "learning_rate": 2.811443566929983e-07,
+      "loss": 1.0881,
+      "step": 10859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.149467646712222,
+      "learning_rate": 2.8049398481872933e-07,
+      "loss": 1.0359,
+      "step": 10860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3503143750298374,
+      "learning_rate": 2.798443553688257e-07,
+      "loss": 1.0632,
+      "step": 10861
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.91830149554117,
+      "learning_rate": 2.7919546839290765e-07,
+      "loss": 1.0317,
+      "step": 10862
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.221025526688736,
+      "learning_rate": 2.7854732394053897e-07,
+      "loss": 1.0872,
+      "step": 10863
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.11252706482995,
+      "learning_rate": 2.7789992206123106e-07,
+      "loss": 1.0354,
+      "step": 10864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5770426624414977,
+      "learning_rate": 2.7725326280443444e-07,
+      "loss": 1.074,
+      "step": 10865
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2735793748998985,
+      "learning_rate": 2.766073462195429e-07,
+      "loss": 1.052,
+      "step": 10866
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7330081365638703,
+      "learning_rate": 2.759621723558992e-07,
+      "loss": 1.0095,
+      "step": 10867
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.72340772555249,
+      "learning_rate": 2.7531774126277833e-07,
+      "loss": 1.0911,
+      "step": 10868
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9897999000909707,
+      "learning_rate": 2.7467405298940986e-07,
+      "loss": 1.0137,
+      "step": 10869
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8102068596609975,
+      "learning_rate": 2.740311075849611e-07,
+      "loss": 1.0742,
+      "step": 10870
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4338260200883854,
+      "learning_rate": 2.733889050985417e-07,
+      "loss": 1.0928,
+      "step": 10871
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.400763966904007,
+      "learning_rate": 2.7274744557920896e-07,
+      "loss": 1.0767,
+      "step": 10872
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.69544711155109,
+      "learning_rate": 2.721067290759594e-07,
+      "loss": 1.0657,
+      "step": 10873
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.234956659948589,
+      "learning_rate": 2.714667556377337e-07,
+      "loss": 1.0649,
+      "step": 10874
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1867518428617703,
+      "learning_rate": 2.708275253134174e-07,
+      "loss": 1.0696,
+      "step": 10875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.287641908536754,
+      "learning_rate": 2.701890381518357e-07,
+      "loss": 1.0684,
+      "step": 10876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2170054854334187,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 1.0696,
+      "step": 10877
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.801946737688188,
+      "learning_rate": 2.6891429351190934e-07,
+      "loss": 0.9683,
+      "step": 10878
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4402514879050763,
+      "learning_rate": 2.682780361309345e-07,
+      "loss": 1.0815,
+      "step": 10879
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.20884024780485,
+      "learning_rate": 2.676425221074397e-07,
+      "loss": 1.0686,
+      "step": 10880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0817102582684743,
+      "learning_rate": 2.6700775148996716e-07,
+      "loss": 1.0266,
+      "step": 10881
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0713781896756127,
+      "learning_rate": 2.6637372432700483e-07,
+      "loss": 0.9993,
+      "step": 10882
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9452355378500354,
+      "learning_rate": 2.657404406669817e-07,
+      "loss": 1.0725,
+      "step": 10883
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4455222253317968,
+      "learning_rate": 2.651079005582713e-07,
+      "loss": 1.0146,
+      "step": 10884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.131766656918767,
+      "learning_rate": 2.6447610404919054e-07,
+      "loss": 1.033,
+      "step": 10885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2327064495497986,
+      "learning_rate": 2.638450511879997e-07,
+      "loss": 1.0627,
+      "step": 10886
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2214355770138545,
+      "learning_rate": 2.6321474202290254e-07,
+      "loss": 1.0505,
+      "step": 10887
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.643273528172432,
+      "learning_rate": 2.6258517660204264e-07,
+      "loss": 1.1082,
+      "step": 10888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.602366328913254,
+      "learning_rate": 2.619563549735105e-07,
+      "loss": 1.1401,
+      "step": 10889
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9714017290512342,
+      "learning_rate": 2.6132827718533983e-07,
+      "loss": 1.0669,
+      "step": 10890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8455119606333192,
+      "learning_rate": 2.6070094328550344e-07,
+      "loss": 0.9895,
+      "step": 10891
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3977243959881345,
+      "learning_rate": 2.600743533219241e-07,
+      "loss": 1.0559,
+      "step": 10892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.589241242021508,
+      "learning_rate": 2.594485073424613e-07,
+      "loss": 1.0676,
+      "step": 10893
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9650010694197095,
+      "learning_rate": 2.588234053949212e-07,
+      "loss": 1.0198,
+      "step": 10894
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4408192836283344,
+      "learning_rate": 2.581990475270524e-07,
+      "loss": 1.0691,
+      "step": 10895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9488087541232253,
+      "learning_rate": 2.575754337865444e-07,
+      "loss": 1.1013,
+      "step": 10896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.007770131171912,
+      "learning_rate": 2.5695256422103464e-07,
+      "loss": 1.0332,
+      "step": 10897
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0696767944837404,
+      "learning_rate": 2.563304388780996e-07,
+      "loss": 1.0759,
+      "step": 10898
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.405282116280762,
+      "learning_rate": 2.557090578052601e-07,
+      "loss": 1.1025,
+      "step": 10899
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.090600983278796,
+      "learning_rate": 2.5508842104998265e-07,
+      "loss": 1.0728,
+      "step": 10900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.886928657183601,
+      "learning_rate": 2.544685286596704e-07,
+      "loss": 1.1101,
+      "step": 10901
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.873624416348905,
+      "learning_rate": 2.5384938068167776e-07,
+      "loss": 1.0183,
+      "step": 10902
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3037217316492042,
+      "learning_rate": 2.5323097716329683e-07,
+      "loss": 1.0547,
+      "step": 10903
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.345304463853275,
+      "learning_rate": 2.5261331815176205e-07,
+      "loss": 0.9934,
+      "step": 10904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9734230586083115,
+      "learning_rate": 2.519964036942579e-07,
+      "loss": 1.0374,
+      "step": 10905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2758519496652982,
+      "learning_rate": 2.5138023383790454e-07,
+      "loss": 1.0725,
+      "step": 10906
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1416705455682825,
+      "learning_rate": 2.507648086297687e-07,
+      "loss": 1.0093,
+      "step": 10907
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1444118843613444,
+      "learning_rate": 2.5015012811685946e-07,
+      "loss": 1.0867,
+      "step": 10908
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7904837477274738,
+      "learning_rate": 2.4953619234612816e-07,
+      "loss": 1.0183,
+      "step": 10909
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.52461982071206,
+      "learning_rate": 2.489230013644717e-07,
+      "loss": 1.0525,
+      "step": 10910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0843893389841996,
+      "learning_rate": 2.4831055521872924e-07,
+      "loss": 1.0579,
+      "step": 10911
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6309174735133984,
+      "learning_rate": 2.4769885395568017e-07,
+      "loss": 1.071,
+      "step": 10912
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4894976193187546,
+      "learning_rate": 2.4708789762205143e-07,
+      "loss": 1.0286,
+      "step": 10913
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.281735245198527,
+      "learning_rate": 2.464776862645113e-07,
+      "loss": 1.0085,
+      "step": 10914
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8465996325312277,
+      "learning_rate": 2.45868219929668e-07,
+      "loss": 1.0369,
+      "step": 10915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.219621515021136,
+      "learning_rate": 2.4525949866407885e-07,
+      "loss": 1.0986,
+      "step": 10916
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.103663424795022,
+      "learning_rate": 2.4465152251423765e-07,
+      "loss": 1.0669,
+      "step": 10917
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.086876658086608,
+      "learning_rate": 2.4404429152658726e-07,
+      "loss": 1.073,
+      "step": 10918
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.072208415574105,
+      "learning_rate": 2.434378057475084e-07,
+      "loss": 1.1309,
+      "step": 10919
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0672696231457355,
+      "learning_rate": 2.428320652233318e-07,
+      "loss": 1.0249,
+      "step": 10920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6917803720416202,
+      "learning_rate": 2.4222707000032376e-07,
+      "loss": 0.9927,
+      "step": 10921
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9837814252238393,
+      "learning_rate": 2.416228201246962e-07,
+      "loss": 1.1074,
+      "step": 10922
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1091097595425836,
+      "learning_rate": 2.410193156426077e-07,
+      "loss": 1.0881,
+      "step": 10923
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2961619309250736,
+      "learning_rate": 2.4041655660015374e-07,
+      "loss": 1.0857,
+      "step": 10924
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.496378056724526,
+      "learning_rate": 2.3981454304337845e-07,
+      "loss": 0.9993,
+      "step": 10925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.996041641099268,
+      "learning_rate": 2.392132750182652e-07,
+      "loss": 1.0833,
+      "step": 10926
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.494696179616954,
+      "learning_rate": 2.386127525707427e-07,
+      "loss": 1.0527,
+      "step": 10927
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.170842011903941,
+      "learning_rate": 2.3801297574668203e-07,
+      "loss": 1.1116,
+      "step": 10928
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.558091850320966,
+      "learning_rate": 2.3741394459189436e-07,
+      "loss": 1.0396,
+      "step": 10929
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.390004287983565,
+      "learning_rate": 2.3681565915214088e-07,
+      "loss": 1.0422,
+      "step": 10930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.995801660879741,
+      "learning_rate": 2.3621811947312056e-07,
+      "loss": 0.9875,
+      "step": 10931
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.323472300473045,
+      "learning_rate": 2.3562132560047357e-07,
+      "loss": 1.0364,
+      "step": 10932
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6700801847777926,
+      "learning_rate": 2.35025277579789e-07,
+      "loss": 1.0005,
+      "step": 10933
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2099677429034235,
+      "learning_rate": 2.3442997545659486e-07,
+      "loss": 1.0613,
+      "step": 10934
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4296597443535024,
+      "learning_rate": 2.3383541927636365e-07,
+      "loss": 1.0959,
+      "step": 10935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2910638686812477,
+      "learning_rate": 2.332416090845102e-07,
+      "loss": 1.0764,
+      "step": 10936
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1505631658553983,
+      "learning_rate": 2.3264854492639154e-07,
+      "loss": 1.0459,
+      "step": 10937
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.318677662711675,
+      "learning_rate": 2.3205622684731143e-07,
+      "loss": 1.072,
+      "step": 10938
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0654462356339467,
+      "learning_rate": 2.314646548925137e-07,
+      "loss": 1.0195,
+      "step": 10939
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4942497315811822,
+      "learning_rate": 2.3087382910718325e-07,
+      "loss": 1.1213,
+      "step": 10940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.277853782140676,
+      "learning_rate": 2.3028374953645182e-07,
+      "loss": 1.0498,
+      "step": 10941
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9377441737535515,
+      "learning_rate": 2.2969441622539223e-07,
+      "loss": 1.0464,
+      "step": 10942
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2445317482017897,
+      "learning_rate": 2.291058292190218e-07,
+      "loss": 1.1084,
+      "step": 10943
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0064698366092886,
+      "learning_rate": 2.28517988562299e-07,
+      "loss": 1.0291,
+      "step": 10944
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7702940206898408,
+      "learning_rate": 2.279308943001246e-07,
+      "loss": 1.0593,
+      "step": 10945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2605000138596094,
+      "learning_rate": 2.2734454647734605e-07,
+      "loss": 1.1038,
+      "step": 10946
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.012088768984659,
+      "learning_rate": 2.2675894513875198e-07,
+      "loss": 1.0122,
+      "step": 10947
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.17054590586437,
+      "learning_rate": 2.26174090329071e-07,
+      "loss": 1.051,
+      "step": 10948
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4555456675108887,
+      "learning_rate": 2.2558998209297855e-07,
+      "loss": 1.0908,
+      "step": 10949
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1179736626136405,
+      "learning_rate": 2.2500662047509003e-07,
+      "loss": 1.1094,
+      "step": 10950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9229703202610198,
+      "learning_rate": 2.244240055199687e-07,
+      "loss": 0.9644,
+      "step": 10951
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1398505813225808,
+      "learning_rate": 2.2384213727211558e-07,
+      "loss": 1.0557,
+      "step": 10952
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.070005727701662,
+      "learning_rate": 2.2326101577597737e-07,
+      "loss": 1.0752,
+      "step": 10953
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0987273089730074,
+      "learning_rate": 2.2268064107594302e-07,
+      "loss": 1.0664,
+      "step": 10954
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.706952568521578,
+      "learning_rate": 2.221010132163437e-07,
+      "loss": 1.0857,
+      "step": 10955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.411248337075108,
+      "learning_rate": 2.2152213224145512e-07,
+      "loss": 1.0422,
+      "step": 10956
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9008726909065534,
+      "learning_rate": 2.2094399819549305e-07,
+      "loss": 0.9976,
+      "step": 10957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.02424441888155,
+      "learning_rate": 2.2036661112262104e-07,
+      "loss": 1.0291,
+      "step": 10958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7646634818314206,
+      "learning_rate": 2.1978997106694267e-07,
+      "loss": 1.0532,
+      "step": 10959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.372718970632016,
+      "learning_rate": 2.1921407807250272e-07,
+      "loss": 1.0513,
+      "step": 10960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.484266335496625,
+      "learning_rate": 2.1863893218329157e-07,
+      "loss": 1.1072,
+      "step": 10961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.225878241458125,
+      "learning_rate": 2.1806453344324185e-07,
+      "loss": 1.1006,
+      "step": 10962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.044359529752206,
+      "learning_rate": 2.1749088189622848e-07,
+      "loss": 1.0762,
+      "step": 10963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.209315665401379,
+      "learning_rate": 2.1691797758607192e-07,
+      "loss": 1.0898,
+      "step": 10964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9695274469891446,
+      "learning_rate": 2.1634582055652942e-07,
+      "loss": 1.116,
+      "step": 10965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.809054325821843,
+      "learning_rate": 2.1577441085130934e-07,
+      "loss": 0.924,
+      "step": 10966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5782953772275725,
+      "learning_rate": 2.152037485140568e-07,
+      "loss": 0.9617,
+      "step": 10967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.840118098485679,
+      "learning_rate": 2.1463383358836242e-07,
+      "loss": 1.0605,
+      "step": 10968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9030235120523618,
+      "learning_rate": 2.140646661177581e-07,
+      "loss": 0.9827,
+      "step": 10969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.843487459222938,
+      "learning_rate": 2.1349624614572017e-07,
+      "loss": 1.03,
+      "step": 10970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9951290794095433,
+      "learning_rate": 2.1292857371566833e-07,
+      "loss": 1.0696,
+      "step": 10971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8621704864993283,
+      "learning_rate": 2.1236164887096345e-07,
+      "loss": 0.9993,
+      "step": 10972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.351649969022172,
+      "learning_rate": 2.1179547165491088e-07,
+      "loss": 1.0549,
+      "step": 10973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.175921999327338,
+      "learning_rate": 2.1123004211075603e-07,
+      "loss": 1.1013,
+      "step": 10974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0411759193709162,
+      "learning_rate": 2.1066536028169105e-07,
+      "loss": 0.978,
+      "step": 10975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.656644823474652,
+      "learning_rate": 2.1010142621084807e-07,
+      "loss": 1.1218,
+      "step": 10976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6229038599860806,
+      "learning_rate": 2.0953823994130486e-07,
+      "loss": 0.996,
+      "step": 10977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.459463932239029,
+      "learning_rate": 2.0897580151607811e-07,
+      "loss": 1.0559,
+      "step": 10978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.21468622185552,
+      "learning_rate": 2.0841411097813126e-07,
+      "loss": 1.0195,
+      "step": 10979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2040831396886498,
+      "learning_rate": 2.0785316837036995e-07,
+      "loss": 0.9822,
+      "step": 10980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.274423049859454,
+      "learning_rate": 2.0729297373563995e-07,
+      "loss": 1.0103,
+      "step": 10981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7443865198829065,
+      "learning_rate": 2.0673352711673145e-07,
+      "loss": 1.115,
+      "step": 10982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1559268210862905,
+      "learning_rate": 2.0617482855637805e-07,
+      "loss": 1.0759,
+      "step": 10983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.833758088166802,
+      "learning_rate": 2.0561687809725782e-07,
+      "loss": 1.0405,
+      "step": 10984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.772156066934829,
+      "learning_rate": 2.0505967578198783e-07,
+      "loss": 1.0359,
+      "step": 10985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2727119368894275,
+      "learning_rate": 2.0450322165312953e-07,
+      "loss": 1.092,
+      "step": 10986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6225824024736926,
+      "learning_rate": 2.0394751575319115e-07,
+      "loss": 1.0205,
+      "step": 10987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.029488195026013,
+      "learning_rate": 2.0339255812461656e-07,
+      "loss": 1.0435,
+      "step": 10988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.107503589907999,
+      "learning_rate": 2.0283834880979735e-07,
+      "loss": 1.1443,
+      "step": 10989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.592582665493228,
+      "learning_rate": 2.0228488785106636e-07,
+      "loss": 1.1289,
+      "step": 10990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.943336496457246,
+      "learning_rate": 2.0173217529070088e-07,
+      "loss": 1.0476,
+      "step": 10991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9761155595794917,
+      "learning_rate": 2.0118021117091934e-07,
+      "loss": 1.0408,
+      "step": 10992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.352633889040909,
+      "learning_rate": 2.006289955338836e-07,
+      "loss": 0.9739,
+      "step": 10993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1053409751810603,
+      "learning_rate": 2.0007852842169772e-07,
+      "loss": 1.0554,
+      "step": 10994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.344778320441082,
+      "learning_rate": 1.9952880987640922e-07,
+      "loss": 1.0706,
+      "step": 10995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.177851753757454,
+      "learning_rate": 1.9897983994000892e-07,
+      "loss": 1.1025,
+      "step": 10996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.946823307175083,
+      "learning_rate": 1.9843161865442994e-07,
+      "loss": 1.0496,
+      "step": 10997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.141667476430378,
+      "learning_rate": 1.978841460615466e-07,
+      "loss": 1.0549,
+      "step": 10998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9335328567549612,
+      "learning_rate": 1.9733742220317876e-07,
+      "loss": 1.0576,
+      "step": 10999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.938366476978011,
+      "learning_rate": 1.9679144712108967e-07,
+      "loss": 1.0288,
+      "step": 11000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.265142279492277,
+      "learning_rate": 1.9624622085698042e-07,
+      "loss": 1.0645,
+      "step": 11001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3631326660454754,
+      "learning_rate": 1.9570174345249994e-07,
+      "loss": 1.0452,
+      "step": 11002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7828296482737467,
+      "learning_rate": 1.9515801494923714e-07,
+      "loss": 1.0046,
+      "step": 11003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2573893937532743,
+      "learning_rate": 1.946150353887266e-07,
+      "loss": 1.0508,
+      "step": 11004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.282839068295511,
+      "learning_rate": 1.9407280481244184e-07,
+      "loss": 1.0083,
+      "step": 11005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9959647640932907,
+      "learning_rate": 1.935313232618008e-07,
+      "loss": 0.9785,
+      "step": 11006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4374950220603013,
+      "learning_rate": 1.9299059077816817e-07,
+      "loss": 1.0891,
+      "step": 11007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1513797913728694,
+      "learning_rate": 1.9245060740284316e-07,
+      "loss": 1.0999,
+      "step": 11008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.088518018039636,
+      "learning_rate": 1.9191137317707497e-07,
+      "loss": 1.0818,
+      "step": 11009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1404664182836988,
+      "learning_rate": 1.913728881420529e-07,
+      "loss": 1.0398,
+      "step": 11010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.489328002351663,
+      "learning_rate": 1.9083515233890738e-07,
+      "loss": 1.136,
+      "step": 11011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.04220518297215,
+      "learning_rate": 1.9029816580871552e-07,
+      "loss": 1.0789,
+      "step": 11012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7667784126020676,
+      "learning_rate": 1.8976192859249455e-07,
+      "loss": 1.0403,
+      "step": 11013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7404908422233745,
+      "learning_rate": 1.89226440731205e-07,
+      "loss": 1.0203,
+      "step": 11014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.821728020913728,
+      "learning_rate": 1.8869170226574863e-07,
+      "loss": 1.0496,
+      "step": 11015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9688726558255745,
+      "learning_rate": 1.8815771323697163e-07,
+      "loss": 1.0178,
+      "step": 11016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.87561358523601,
+      "learning_rate": 1.8762447368566582e-07,
+      "loss": 1.0723,
+      "step": 11017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3808476802110183,
+      "learning_rate": 1.8709198365255864e-07,
+      "loss": 1.0442,
+      "step": 11018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0492414728802566,
+      "learning_rate": 1.8656024317832643e-07,
+      "loss": 1.0457,
+      "step": 11019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.051120040626882,
+      "learning_rate": 1.8602925230358782e-07,
+      "loss": 1.1208,
+      "step": 11020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9125469887753215,
+      "learning_rate": 1.854990110688981e-07,
+      "loss": 1.0698,
+      "step": 11021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.505566365647708,
+      "learning_rate": 1.8496951951476384e-07,
+      "loss": 1.0903,
+      "step": 11022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.04671904916731,
+      "learning_rate": 1.8444077768162815e-07,
+      "loss": 1.0735,
+      "step": 11023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2449286546305696,
+      "learning_rate": 1.8391278560987878e-07,
+      "loss": 1.052,
+      "step": 11024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2601649877978236,
+      "learning_rate": 1.8338554333984903e-07,
+      "loss": 1.033,
+      "step": 11025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.787432910366154,
+      "learning_rate": 1.828590509118089e-07,
+      "loss": 1.0522,
+      "step": 11026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9275495310658632,
+      "learning_rate": 1.823333083659773e-07,
+      "loss": 1.0012,
+      "step": 11027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.27984189647045,
+      "learning_rate": 1.81808315742511e-07,
+      "loss": 1.0559,
+      "step": 11028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.760426481923527,
+      "learning_rate": 1.812840730815113e-07,
+      "loss": 1.0796,
+      "step": 11029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.028165913658897,
+      "learning_rate": 1.8076058042302392e-07,
+      "loss": 1.0671,
+      "step": 11030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4379385325729617,
+      "learning_rate": 1.8023783780703575e-07,
+      "loss": 1.0046,
+      "step": 11031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.422726921255537,
+      "learning_rate": 1.7971584527347596e-07,
+      "loss": 1.0522,
+      "step": 11032
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.977793670614632,
+      "learning_rate": 1.7919460286221823e-07,
+      "loss": 0.9398,
+      "step": 11033
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9204292515186387,
+      "learning_rate": 1.7867411061307626e-07,
+      "loss": 1.0112,
+      "step": 11034
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.979324233248477,
+      "learning_rate": 1.7815436856580714e-07,
+      "loss": 1.0222,
+      "step": 11035
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0041541622282346,
+      "learning_rate": 1.7763537676011244e-07,
+      "loss": 1.0151,
+      "step": 11036
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.031091158830772,
+      "learning_rate": 1.7711713523563601e-07,
+      "loss": 1.012,
+      "step": 11037
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.184450185654305,
+      "learning_rate": 1.7659964403196283e-07,
+      "loss": 1.0491,
+      "step": 11038
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9490642301867247,
+      "learning_rate": 1.760829031886202e-07,
+      "loss": 1.011,
+      "step": 11039
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.006827618778537,
+      "learning_rate": 1.7556691274508319e-07,
+      "loss": 1.1248,
+      "step": 11040
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.082043255213284,
+      "learning_rate": 1.7505167274076141e-07,
+      "loss": 1.094,
+      "step": 11041
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.46434172425289,
+      "learning_rate": 1.7453718321501444e-07,
+      "loss": 1.0117,
+      "step": 11042
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1676769073657907,
+      "learning_rate": 1.740234442071409e-07,
+      "loss": 1.0215,
+      "step": 11043
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.025497170447527,
+      "learning_rate": 1.7351045575638047e-07,
+      "loss": 1.0137,
+      "step": 11044
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1322801084378393,
+      "learning_rate": 1.7299821790192073e-07,
+      "loss": 1.062,
+      "step": 11045
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.792599480733136,
+      "learning_rate": 1.7248673068288813e-07,
+      "loss": 0.9768,
+      "step": 11046
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.965858527839501,
+      "learning_rate": 1.7197599413835143e-07,
+      "loss": 1.0459,
+      "step": 11047
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7021652875736675,
+      "learning_rate": 1.7146600830732495e-07,
+      "loss": 1.1187,
+      "step": 11048
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.525722165483015,
+      "learning_rate": 1.7095677322876315e-07,
+      "loss": 1.0488,
+      "step": 11049
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.370966576015153,
+      "learning_rate": 1.7044828894156373e-07,
+      "loss": 1.0383,
+      "step": 11050
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.639514201206445,
+      "learning_rate": 1.6994055548456677e-07,
+      "loss": 1.0974,
+      "step": 11051
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2650092928531538,
+      "learning_rate": 1.694335728965568e-07,
+      "loss": 1.05,
+      "step": 11052
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.980709422119882,
+      "learning_rate": 1.6892734121625954e-07,
+      "loss": 1.0391,
+      "step": 11053
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9971004463127735,
+      "learning_rate": 1.6842186048234177e-07,
+      "loss": 1.0111,
+      "step": 11054
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.729236265102979,
+      "learning_rate": 1.6791713073341709e-07,
+      "loss": 1.0928,
+      "step": 11055
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0467024597654837,
+      "learning_rate": 1.6741315200803799e-07,
+      "loss": 1.0923,
+      "step": 11056
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1645522814362907,
+      "learning_rate": 1.669099243446992e-07,
+      "loss": 0.9934,
+      "step": 11057
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3402162196393457,
+      "learning_rate": 1.6640744778184336e-07,
+      "loss": 1.0803,
+      "step": 11058
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.972201084111034,
+      "learning_rate": 1.6590572235784863e-07,
+      "loss": 0.9674,
+      "step": 11059
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.092778100698898,
+      "learning_rate": 1.6540474811104324e-07,
+      "loss": 1.0369,
+      "step": 11060
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.895442633996885,
+      "learning_rate": 1.64904525079691e-07,
+      "loss": 1.0024,
+      "step": 11061
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.398417163519466,
+      "learning_rate": 1.644050533020014e-07,
+      "loss": 1.1082,
+      "step": 11062
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.639328258866761,
+      "learning_rate": 1.6390633281612834e-07,
+      "loss": 1.0164,
+      "step": 11063
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.293883072999912,
+      "learning_rate": 1.6340836366016466e-07,
+      "loss": 1.0762,
+      "step": 11064
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.565777974891546,
+      "learning_rate": 1.6291114587214884e-07,
+      "loss": 1.0359,
+      "step": 11065
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9147757759554986,
+      "learning_rate": 1.6241467949006273e-07,
+      "loss": 1.0001,
+      "step": 11066
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.159787168274921,
+      "learning_rate": 1.6191896455182488e-07,
+      "loss": 1.0056,
+      "step": 11067
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.004542790547015,
+      "learning_rate": 1.6142400109530275e-07,
+      "loss": 1.0105,
+      "step": 11068
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.728517115926369,
+      "learning_rate": 1.6092978915830392e-07,
+      "loss": 0.9895,
+      "step": 11069
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3238328937828507,
+      "learning_rate": 1.6043632877857928e-07,
+      "loss": 1.1257,
+      "step": 11070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9855790398638287,
+      "learning_rate": 1.5994361999382202e-07,
+      "loss": 1.0764,
+      "step": 11071
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1298405065667763,
+      "learning_rate": 1.5945166284166536e-07,
+      "loss": 1.0598,
+      "step": 11072
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.762544173825647,
+      "learning_rate": 1.5896045735969035e-07,
+      "loss": 1.0298,
+      "step": 11073
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.172501020126051,
+      "learning_rate": 1.5847000358541475e-07,
+      "loss": 1.0217,
+      "step": 11074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0416431490817577,
+      "learning_rate": 1.579803015563053e-07,
+      "loss": 1.0415,
+      "step": 11075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.896639762994499,
+      "learning_rate": 1.5749135130976534e-07,
+      "loss": 1.0798,
+      "step": 11076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2466491786078913,
+      "learning_rate": 1.5700315288314282e-07,
+      "loss": 1.094,
+      "step": 11077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1899587332813892,
+      "learning_rate": 1.5651570631373125e-07,
+      "loss": 1.0479,
+      "step": 11078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5447992345649935,
+      "learning_rate": 1.5602901163876304e-07,
+      "loss": 1.0969,
+      "step": 11079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9178285724668696,
+      "learning_rate": 1.5554306889541514e-07,
+      "loss": 1.0527,
+      "step": 11080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3397406168958894,
+      "learning_rate": 1.5505787812080452e-07,
+      "loss": 1.0952,
+      "step": 11081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8718434636254493,
+      "learning_rate": 1.5457343935199264e-07,
+      "loss": 1.0483,
+      "step": 11082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.179330786758388,
+      "learning_rate": 1.5408975262598546e-07,
+      "loss": 1.0222,
+      "step": 11083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2371576736118106,
+      "learning_rate": 1.5360681797972675e-07,
+      "loss": 1.0959,
+      "step": 11084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0666618811604147,
+      "learning_rate": 1.5312463545010702e-07,
+      "loss": 1.0474,
+      "step": 11085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9473479303998107,
+      "learning_rate": 1.5264320507395903e-07,
+      "loss": 1.1033,
+      "step": 11086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.949681885619538,
+      "learning_rate": 1.5216252688805445e-07,
+      "loss": 1.0681,
+      "step": 11087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1191085461153505,
+      "learning_rate": 1.5168260092911057e-07,
+      "loss": 1.1262,
+      "step": 11088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.91635069036712,
+      "learning_rate": 1.51203427233787e-07,
+      "loss": 1.0642,
+      "step": 11089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.071163337299297,
+      "learning_rate": 1.5072500583868444e-07,
+      "loss": 1.0549,
+      "step": 11090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0781963548583167,
+      "learning_rate": 1.502473367803481e-07,
+      "loss": 0.9976,
+      "step": 11091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4833051377868403,
+      "learning_rate": 1.4977042009526433e-07,
+      "loss": 1.0942,
+      "step": 11092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.393118191440671,
+      "learning_rate": 1.49294255819864e-07,
+      "loss": 1.1086,
+      "step": 11093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0495344745247377,
+      "learning_rate": 1.4881884399051582e-07,
+      "loss": 1.01,
+      "step": 11094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.47426212893079,
+      "learning_rate": 1.4834418464353627e-07,
+      "loss": 1.0527,
+      "step": 11095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4087904950001584,
+      "learning_rate": 1.4787027781518081e-07,
+      "loss": 1.0554,
+      "step": 11096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1630688053231277,
+      "learning_rate": 1.4739712354165047e-07,
+      "loss": 1.1448,
+      "step": 11097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7677981625520527,
+      "learning_rate": 1.4692472185908635e-07,
+      "loss": 0.9858,
+      "step": 11098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4716536578191013,
+      "learning_rate": 1.464530728035729e-07,
+      "loss": 1.0029,
+      "step": 11099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1307457565203647,
+      "learning_rate": 1.4598217641113687e-07,
+      "loss": 1.0908,
+      "step": 11100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.219153555292204,
+      "learning_rate": 1.455120327177473e-07,
+      "loss": 1.011,
+      "step": 11101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.967332455341423,
+      "learning_rate": 1.4504264175931647e-07,
+      "loss": 1.0662,
+      "step": 11102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0488923972214224,
+      "learning_rate": 1.4457400357169916e-07,
+      "loss": 1.0425,
+      "step": 11103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0681708079580883,
+      "learning_rate": 1.441061181906922e-07,
+      "loss": 1.062,
+      "step": 11104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5204641686821123,
+      "learning_rate": 1.436389856520337e-07,
+      "loss": 1.1504,
+      "step": 11105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5238328459357824,
+      "learning_rate": 1.4317260599140847e-07,
+      "loss": 1.0588,
+      "step": 11106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1829318435019514,
+      "learning_rate": 1.42706979244438e-07,
+      "loss": 1.092,
+      "step": 11107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6615503478296234,
+      "learning_rate": 1.4224210544669047e-07,
+      "loss": 1.0234,
+      "step": 11108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.738316967744148,
+      "learning_rate": 1.417779846336753e-07,
+      "loss": 1.0188,
+      "step": 11109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3158520850830873,
+      "learning_rate": 1.4131461684084413e-07,
+      "loss": 1.1057,
+      "step": 11110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4728812434736294,
+      "learning_rate": 1.4085200210359195e-07,
+      "loss": 1.0525,
+      "step": 11111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.282610310131773,
+      "learning_rate": 1.4039014045725386e-07,
+      "loss": 1.0664,
+      "step": 11112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.16507247005582,
+      "learning_rate": 1.399290319371116e-07,
+      "loss": 1.0325,
+      "step": 11113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1864533586187593,
+      "learning_rate": 1.394686765783848e-07,
+      "loss": 1.1213,
+      "step": 11114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.331712819015632,
+      "learning_rate": 1.3900907441623758e-07,
+      "loss": 1.0657,
+      "step": 11115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.516625346145988,
+      "learning_rate": 1.3855022548577845e-07,
+      "loss": 1.0325,
+      "step": 11116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3406428840135525,
+      "learning_rate": 1.380921298220561e-07,
+      "loss": 0.9597,
+      "step": 11117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8103764182750677,
+      "learning_rate": 1.3763478746006142e-07,
+      "loss": 1.0551,
+      "step": 11118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.250963190620747,
+      "learning_rate": 1.371781984347298e-07,
+      "loss": 1.1113,
+      "step": 11119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.375736697853555,
+      "learning_rate": 1.3672236278093554e-07,
+      "loss": 1.0515,
+      "step": 11120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.429301080772552,
+      "learning_rate": 1.362672805334986e-07,
+      "loss": 1.0852,
+      "step": 11121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.926185184996428,
+      "learning_rate": 1.3581295172718224e-07,
+      "loss": 1.0364,
+      "step": 11122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9085240813036086,
+      "learning_rate": 1.353593763966876e-07,
+      "loss": 0.9631,
+      "step": 11123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.177390108425144,
+      "learning_rate": 1.3490655457666257e-07,
+      "loss": 1.1086,
+      "step": 11124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3729817139655,
+      "learning_rate": 1.344544863016961e-07,
+      "loss": 1.0906,
+      "step": 11125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.183683605769009,
+      "learning_rate": 1.340031716063195e-07,
+      "loss": 1.0625,
+      "step": 11126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3137726622641925,
+      "learning_rate": 1.3355261052500512e-07,
+      "loss": 1.0625,
+      "step": 11127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8549708722943166,
+      "learning_rate": 1.3310280309216994e-07,
+      "loss": 1.0466,
+      "step": 11128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4276148840462746,
+      "learning_rate": 1.3265374934217313e-07,
+      "loss": 1.0698,
+      "step": 11129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.283405606336496,
+      "learning_rate": 1.32205449309315e-07,
+      "loss": 1.0554,
+      "step": 11130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1723797643455516,
+      "learning_rate": 1.3175790302783708e-07,
+      "loss": 1.0693,
+      "step": 11131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0109841662923205,
+      "learning_rate": 1.3131111053192868e-07,
+      "loss": 1.0457,
+      "step": 11132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3727136711320522,
+      "learning_rate": 1.3086507185571696e-07,
+      "loss": 1.0134,
+      "step": 11133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.125967008983668,
+      "learning_rate": 1.304197870332713e-07,
+      "loss": 1.0405,
+      "step": 11134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0707714554351897,
+      "learning_rate": 1.299752560986045e-07,
+      "loss": 1.0115,
+      "step": 11135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.095974990015106,
+      "learning_rate": 1.2953147908567388e-07,
+      "loss": 1.1089,
+      "step": 11136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0455328994744724,
+      "learning_rate": 1.2908845602837673e-07,
+      "loss": 1.031,
+      "step": 11137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.872375642211466,
+      "learning_rate": 1.2864618696055263e-07,
+      "loss": 0.9976,
+      "step": 11138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3844435398812918,
+      "learning_rate": 1.2820467191598574e-07,
+      "loss": 1.0691,
+      "step": 11139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9965569022071215,
+      "learning_rate": 1.27763910928399e-07,
+      "loss": 1.032,
+      "step": 11140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.150963826160562,
+      "learning_rate": 1.273239040314611e-07,
+      "loss": 0.9951,
+      "step": 11141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4267507439152065,
+      "learning_rate": 1.268846512587829e-07,
+      "loss": 1.0056,
+      "step": 11142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.986778377679383,
+      "learning_rate": 1.2644615264391536e-07,
+      "loss": 1.0786,
+      "step": 11143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4274732645140333,
+      "learning_rate": 1.2600840822035387e-07,
+      "loss": 1.0229,
+      "step": 11144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.498311256867976,
+      "learning_rate": 1.2557141802153617e-07,
+      "loss": 1.0337,
+      "step": 11145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.053162837574942,
+      "learning_rate": 1.2513518208083996e-07,
+      "loss": 1.0559,
+      "step": 11146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.872307891530179,
+      "learning_rate": 1.2469970043158864e-07,
+      "loss": 1.0737,
+      "step": 11147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.08533778704568,
+      "learning_rate": 1.2426497310704444e-07,
+      "loss": 1.1172,
+      "step": 11148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1156832022650636,
+      "learning_rate": 1.238310001404175e-07,
+      "loss": 1.0029,
+      "step": 11149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.635304904474984,
+      "learning_rate": 1.2339778156485348e-07,
+      "loss": 1.123,
+      "step": 11150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.22003012489608,
+      "learning_rate": 1.2296531741344486e-07,
+      "loss": 1.0425,
+      "step": 11151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.669676728983177,
+      "learning_rate": 1.225336077192274e-07,
+      "loss": 0.9998,
+      "step": 11152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2794224400930347,
+      "learning_rate": 1.2210265251517473e-07,
+      "loss": 1.0447,
+      "step": 11153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8466561009344304,
+      "learning_rate": 1.2167245183420606e-07,
+      "loss": 0.9817,
+      "step": 11154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4629958624778445,
+      "learning_rate": 1.212430057091818e-07,
+      "loss": 1.0835,
+      "step": 11155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.940507053825803,
+      "learning_rate": 1.2081431417290567e-07,
+      "loss": 0.99,
+      "step": 11156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.175310099058536,
+      "learning_rate": 1.2038637725812376e-07,
+      "loss": 1.0591,
+      "step": 11157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9744399912317294,
+      "learning_rate": 1.199591949975243e-07,
+      "loss": 1.063,
+      "step": 11158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3075330114367274,
+      "learning_rate": 1.1953276742373566e-07,
+      "loss": 1.0632,
+      "step": 11159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.825163144055986,
+      "learning_rate": 1.191070945693329e-07,
+      "loss": 1.0947,
+      "step": 11160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9128731568524935,
+      "learning_rate": 1.1868217646682889e-07,
+      "loss": 1.0803,
+      "step": 11161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8566286892187103,
+      "learning_rate": 1.1825801314868212e-07,
+      "loss": 1.0916,
+      "step": 11162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7866612206406236,
+      "learning_rate": 1.1783460464729224e-07,
+      "loss": 1.0464,
+      "step": 11163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8479313774728605,
+      "learning_rate": 1.1741195099500113e-07,
+      "loss": 1.0723,
+      "step": 11164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6962510806189495,
+      "learning_rate": 1.1699005222409299e-07,
+      "loss": 1.1011,
+      "step": 11165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.637643463866148,
+      "learning_rate": 1.1656890836679646e-07,
+      "loss": 0.9919,
+      "step": 11166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.300254920352483,
+      "learning_rate": 1.1614851945527806e-07,
+      "loss": 0.8945,
+      "step": 11167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1280305719914225,
+      "learning_rate": 1.1572888552164984e-07,
+      "loss": 1.0645,
+      "step": 11168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4037169104308256,
+      "learning_rate": 1.1531000659796509e-07,
+      "loss": 1.0603,
+      "step": 11169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.98784922066651,
+      "learning_rate": 1.1489188271622154e-07,
+      "loss": 1.0344,
+      "step": 11170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2985116237631384,
+      "learning_rate": 1.1447451390835585e-07,
+      "loss": 1.0391,
+      "step": 11171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.320834619268083,
+      "learning_rate": 1.1405790020625141e-07,
+      "loss": 1.0198,
+      "step": 11172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1401456603793556,
+      "learning_rate": 1.136420416417272e-07,
+      "loss": 1.0464,
+      "step": 11173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.378344185061084,
+      "learning_rate": 1.1322693824655228e-07,
+      "loss": 1.0349,
+      "step": 11174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6590679524890564,
+      "learning_rate": 1.1281259005243239e-07,
+      "loss": 1.072,
+      "step": 11175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.738897783330504,
+      "learning_rate": 1.1239899709101776e-07,
+      "loss": 1.0742,
+      "step": 11176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9153700746046107,
+      "learning_rate": 1.119861593939009e-07,
+      "loss": 1.0234,
+      "step": 11177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9135525574605334,
+      "learning_rate": 1.1157407699261547e-07,
+      "loss": 1.063,
+      "step": 11178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.147157432511764,
+      "learning_rate": 1.1116274991864073e-07,
+      "loss": 1.0474,
+      "step": 11179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.211133498454005,
+      "learning_rate": 1.107521782033938e-07,
+      "loss": 1.0498,
+      "step": 11180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.259319934638127,
+      "learning_rate": 1.1034236187823622e-07,
+      "loss": 1.0945,
+      "step": 11181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1557566996715543,
+      "learning_rate": 1.0993330097447297e-07,
+      "loss": 1.0447,
+      "step": 11182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2056622617964377,
+      "learning_rate": 1.0952499552334905e-07,
+      "loss": 1.0217,
+      "step": 11183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5116283916436917,
+      "learning_rate": 1.0911744555605397e-07,
+      "loss": 1.053,
+      "step": 11184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2410674600719287,
+      "learning_rate": 1.0871065110371726e-07,
+      "loss": 1.0266,
+      "step": 11185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9256655002649388,
+      "learning_rate": 1.0830461219741406e-07,
+      "loss": 1.0254,
+      "step": 11186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1511402550819523,
+      "learning_rate": 1.0789932886815623e-07,
+      "loss": 1.0776,
+      "step": 11187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.074872914765055,
+      "learning_rate": 1.0749480114690458e-07,
+      "loss": 1.0989,
+      "step": 11188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.712776438034695,
+      "learning_rate": 1.0709102906455659e-07,
+      "loss": 1.053,
+      "step": 11189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1695326234423375,
+      "learning_rate": 1.0668801265195538e-07,
+      "loss": 1.0105,
+      "step": 11190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.197744972585238,
+      "learning_rate": 1.0628575193988633e-07,
+      "loss": 1.012,
+      "step": 11191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3621902766277416,
+      "learning_rate": 1.0588424695907484e-07,
+      "loss": 1.0544,
+      "step": 11192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4638315839688025,
+      "learning_rate": 1.0548349774018973e-07,
+      "loss": 1.0793,
+      "step": 11193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.344005333903372,
+      "learning_rate": 1.0508350431384207e-07,
+      "loss": 1.0164,
+      "step": 11194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3063662807947205,
+      "learning_rate": 1.046842667105874e-07,
+      "loss": 1.0332,
+      "step": 11195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8203785038592466,
+      "learning_rate": 1.0428578496091912e-07,
+      "loss": 1.0676,
+      "step": 11196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3874185904233323,
+      "learning_rate": 1.0388805909527621e-07,
+      "loss": 1.0718,
+      "step": 11197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2205132808161143,
+      "learning_rate": 1.0349108914403883e-07,
+      "loss": 1.0635,
+      "step": 11198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.935968275845305,
+      "learning_rate": 1.0309487513752936e-07,
+      "loss": 1.0715,
+      "step": 11199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.956300562049155,
+      "learning_rate": 1.0269941710601361e-07,
+      "loss": 1.072,
+      "step": 11200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9589087986500773,
+      "learning_rate": 1.0230471507969741e-07,
+      "loss": 1.0334,
+      "step": 11201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.339972408783542,
+      "learning_rate": 1.019107690887311e-07,
+      "loss": 1.1018,
+      "step": 11202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.495491009706097,
+      "learning_rate": 1.0151757916320504e-07,
+      "loss": 1.0283,
+      "step": 11203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.766510381081372,
+      "learning_rate": 1.0112514533315298e-07,
+      "loss": 1.0239,
+      "step": 11204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.312056545498044,
+      "learning_rate": 1.0073346762855318e-07,
+      "loss": 1.0752,
+      "step": 11205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.253534192707722,
+      "learning_rate": 1.003425460793217e-07,
+      "loss": 1.1147,
+      "step": 11206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.878975664231048,
+      "learning_rate": 9.995238071532021e-08,
+      "loss": 1.0674,
+      "step": 11207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0761957619511646,
+      "learning_rate": 9.956297156635043e-08,
+      "loss": 1.1013,
+      "step": 11208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.956808617623372,
+      "learning_rate": 9.917431866215743e-08,
+      "loss": 1.0652,
+      "step": 11209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.289333068332641,
+      "learning_rate": 9.878642203243083e-08,
+      "loss": 1.0374,
+      "step": 11210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2057980347684714,
+      "learning_rate": 9.839928170679691e-08,
+      "loss": 1.0203,
+      "step": 11211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.365259984351256,
+      "learning_rate": 9.801289771482981e-08,
+      "loss": 1.0391,
+      "step": 11212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.112782989485957,
+      "learning_rate": 9.762727008604145e-08,
+      "loss": 1.0906,
+      "step": 11213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2707547988417938,
+      "learning_rate": 9.72423988498894e-08,
+      "loss": 0.9922,
+      "step": 11214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7363818256260006,
+      "learning_rate": 9.685828403577124e-08,
+      "loss": 1.0693,
+      "step": 11215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.98729513375063,
+      "learning_rate": 9.647492567302907e-08,
+      "loss": 1.1077,
+      "step": 11216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.845198186888152,
+      "learning_rate": 9.609232379094391e-08,
+      "loss": 1.0327,
+      "step": 11217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.144612584335745,
+      "learning_rate": 9.571047841874126e-08,
+      "loss": 1.0786,
+      "step": 11218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4337463225663347,
+      "learning_rate": 9.532938958558891e-08,
+      "loss": 1.054,
+      "step": 11219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7958983399007753,
+      "learning_rate": 9.494905732059578e-08,
+      "loss": 0.9078,
+      "step": 11220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1704287985966397,
+      "learning_rate": 9.456948165281421e-08,
+      "loss": 1.0391,
+      "step": 11221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3463520208833053,
+      "learning_rate": 9.419066261123765e-08,
+      "loss": 0.9932,
+      "step": 11222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2941926574352185,
+      "learning_rate": 9.381260022480188e-08,
+      "loss": 1.0486,
+      "step": 11223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.630560773397794,
+      "learning_rate": 9.343529452238598e-08,
+      "loss": 1.073,
+      "step": 11224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6243267256220046,
+      "learning_rate": 9.305874553281025e-08,
+      "loss": 1.0591,
+      "step": 11225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0643444321384865,
+      "learning_rate": 9.268295328483723e-08,
+      "loss": 1.0239,
+      "step": 11226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.37396887612288,
+      "learning_rate": 9.230791780717064e-08,
+      "loss": 1.1245,
+      "step": 11227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0795917397573844,
+      "learning_rate": 9.193363912845866e-08,
+      "loss": 0.99,
+      "step": 11228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1281897043468203,
+      "learning_rate": 9.156011727729175e-08,
+      "loss": 1.0403,
+      "step": 11229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.229901910548815,
+      "learning_rate": 9.118735228219821e-08,
+      "loss": 1.0549,
+      "step": 11230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2243990417731636,
+      "learning_rate": 9.081534417165528e-08,
+      "loss": 1.0425,
+      "step": 11231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.45082941552822,
+      "learning_rate": 9.044409297407685e-08,
+      "loss": 1.0881,
+      "step": 11232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.322204518639244,
+      "learning_rate": 9.00735987178214e-08,
+      "loss": 1.0796,
+      "step": 11233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.145836403201302,
+      "learning_rate": 8.970386143118848e-08,
+      "loss": 1.0142,
+      "step": 11234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0997375639498777,
+      "learning_rate": 8.933488114241995e-08,
+      "loss": 1.0337,
+      "step": 11235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6603071742457503,
+      "learning_rate": 8.896665787970326e-08,
+      "loss": 1.1003,
+      "step": 11236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8152014257432962,
+      "learning_rate": 8.859919167116149e-08,
+      "loss": 1.0149,
+      "step": 11237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.831262083342274,
+      "learning_rate": 8.82324825448666e-08,
+      "loss": 0.9888,
+      "step": 11238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.150051891589196,
+      "learning_rate": 8.786653052882843e-08,
+      "loss": 1.0283,
+      "step": 11239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9945589733681865,
+      "learning_rate": 8.750133565100016e-08,
+      "loss": 1.0974,
+      "step": 11240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.347460427783046,
+      "learning_rate": 8.713689793927615e-08,
+      "loss": 1.0669,
+      "step": 11241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2841129156528317,
+      "learning_rate": 8.677321742149525e-08,
+      "loss": 1.0442,
+      "step": 11242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0484561141835003,
+      "learning_rate": 8.641029412543744e-08,
+      "loss": 1.1074,
+      "step": 11243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.192650710954838,
+      "learning_rate": 8.60481280788239e-08,
+      "loss": 1.0342,
+      "step": 11244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9969075618032526,
+      "learning_rate": 8.568671930931915e-08,
+      "loss": 1.0446,
+      "step": 11245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3337798826164007,
+      "learning_rate": 8.532606784453002e-08,
+      "loss": 1.0186,
+      "step": 11246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1946096158001316,
+      "learning_rate": 8.496617371200222e-08,
+      "loss": 1.0439,
+      "step": 11247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6387936486749535,
+      "learning_rate": 8.460703693922934e-08,
+      "loss": 1.051,
+      "step": 11248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0972143416621947,
+      "learning_rate": 8.424865755364165e-08,
+      "loss": 1.0508,
+      "step": 11249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.881746398985189,
+      "learning_rate": 8.389103558261502e-08,
+      "loss": 1.0239,
+      "step": 11250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9637878025553466,
+      "learning_rate": 8.353417105346762e-08,
+      "loss": 1.019,
+      "step": 11251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.953148331851295,
+      "learning_rate": 8.317806399345651e-08,
+      "loss": 1.0774,
+      "step": 11252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4814363092276643,
+      "learning_rate": 8.28227144297844e-08,
+      "loss": 1.0271,
+      "step": 11253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6462260802498396,
+      "learning_rate": 8.246812238959401e-08,
+      "loss": 1.0742,
+      "step": 11254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.803542239055977,
+      "learning_rate": 8.211428789996922e-08,
+      "loss": 1.0852,
+      "step": 11255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4702142342679227,
+      "learning_rate": 8.176121098794066e-08,
+      "loss": 1.033,
+      "step": 11256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6744771973137755,
+      "learning_rate": 8.140889168047672e-08,
+      "loss": 1.0337,
+      "step": 11257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.168345128474624,
+      "learning_rate": 8.105733000448924e-08,
+      "loss": 1.0312,
+      "step": 11258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9954849834508677,
+      "learning_rate": 8.070652598683337e-08,
+      "loss": 1.0408,
+      "step": 11259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3609995804019532,
+      "learning_rate": 8.035647965430215e-08,
+      "loss": 1.0044,
+      "step": 11260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6883570375695904,
+      "learning_rate": 8.000719103363752e-08,
+      "loss": 1.0259,
+      "step": 11261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.744908001092041,
+      "learning_rate": 7.965866015151814e-08,
+      "loss": 1.0815,
+      "step": 11262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.236650610493,
+      "learning_rate": 7.931088703456601e-08,
+      "loss": 1.0427,
+      "step": 11263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.055040995304293,
+      "learning_rate": 7.89638717093466e-08,
+      "loss": 1.0293,
+      "step": 11264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.189229131055454,
+      "learning_rate": 7.861761420236758e-08,
+      "loss": 1.0466,
+      "step": 11265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.902178648781087,
+      "learning_rate": 7.827211454007666e-08,
+      "loss": 1.0339,
+      "step": 11266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6501958440704425,
+      "learning_rate": 7.792737274886386e-08,
+      "loss": 1.0305,
+      "step": 11267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6028167799195563,
+      "learning_rate": 7.758338885506367e-08,
+      "loss": 1.0132,
+      "step": 11268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0741696187180687,
+      "learning_rate": 7.724016288495173e-08,
+      "loss": 1.0518,
+      "step": 11269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3404590983730684,
+      "learning_rate": 7.689769486474485e-08,
+      "loss": 1.093,
+      "step": 11270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.84690252531152,
+      "learning_rate": 7.6555984820601e-08,
+      "loss": 0.9978,
+      "step": 11271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.008792329320276,
+      "learning_rate": 7.621503277862485e-08,
+      "loss": 1.0745,
+      "step": 11272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2341393790079205,
+      "learning_rate": 7.587483876485669e-08,
+      "loss": 1.0742,
+      "step": 11273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.185230714959327,
+      "learning_rate": 7.55354028052846e-08,
+      "loss": 1.0764,
+      "step": 11274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9410941815996616,
+      "learning_rate": 7.519672492583341e-08,
+      "loss": 1.0208,
+      "step": 11275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0600238195159672,
+      "learning_rate": 7.485880515237686e-08,
+      "loss": 1.1165,
+      "step": 11276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2533211826067485,
+      "learning_rate": 7.452164351072544e-08,
+      "loss": 1.1055,
+      "step": 11277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.733801929332947,
+      "learning_rate": 7.418524002663296e-08,
+      "loss": 1.0117,
+      "step": 11278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.056481813381886,
+      "learning_rate": 7.384959472579445e-08,
+      "loss": 1.0894,
+      "step": 11279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5382648061392112,
+      "learning_rate": 7.351470763384938e-08,
+      "loss": 1.0215,
+      "step": 11280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.842111547981188,
+      "learning_rate": 7.318057877637841e-08,
+      "loss": 1.0364,
+      "step": 11281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.838676421294503,
+      "learning_rate": 7.284720817890445e-08,
+      "loss": 0.9771,
+      "step": 11282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1506237938803316,
+      "learning_rate": 7.251459586689047e-08,
+      "loss": 1.0425,
+      "step": 11283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.415083722091716,
+      "learning_rate": 7.218274186574281e-08,
+      "loss": 1.064,
+      "step": 11284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0391116912079634,
+      "learning_rate": 7.185164620081231e-08,
+      "loss": 1.0347,
+      "step": 11285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.128555458861689,
+      "learning_rate": 7.152130889738873e-08,
+      "loss": 1.0784,
+      "step": 11286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0249074144605697,
+      "learning_rate": 7.119172998070412e-08,
+      "loss": 1.1016,
+      "step": 11287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7114758699847683,
+      "learning_rate": 7.086290947593277e-08,
+      "loss": 0.9922,
+      "step": 11288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9165156029071193,
+      "learning_rate": 7.05348474081946e-08,
+      "loss": 1.0801,
+      "step": 11289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7571630708625188,
+      "learning_rate": 7.020754380254625e-08,
+      "loss": 1.0972,
+      "step": 11290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.200742441243071,
+      "learning_rate": 6.988099868398768e-08,
+      "loss": 1.0981,
+      "step": 11291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1732769480193888,
+      "learning_rate": 6.955521207746563e-08,
+      "loss": 1.0039,
+      "step": 11292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.117107468981301,
+      "learning_rate": 6.923018400786241e-08,
+      "loss": 1.0312,
+      "step": 11293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.077414521256763,
+      "learning_rate": 6.890591450000706e-08,
+      "loss": 1.0581,
+      "step": 11294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.766859513654502,
+      "learning_rate": 6.858240357866752e-08,
+      "loss": 1.0195,
+      "step": 11295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2951826404786098,
+      "learning_rate": 6.825965126855517e-08,
+      "loss": 1.0327,
+      "step": 11296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.181675231729578,
+      "learning_rate": 6.793765759432469e-08,
+      "loss": 1.083,
+      "step": 11297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0537002139060268,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 1.01,
+      "step": 11298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0621092146187165,
+      "learning_rate": 6.729594625183078e-08,
+      "loss": 1.0444,
+      "step": 11299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2147322090826704,
+      "learning_rate": 6.69762286325848e-08,
+      "loss": 0.9907,
+      "step": 11300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.679810150394467,
+      "learning_rate": 6.665726974725339e-08,
+      "loss": 1.0977,
+      "step": 11301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9191564496018643,
+      "learning_rate": 6.633906962020153e-08,
+      "loss": 1.0239,
+      "step": 11302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1645022559098486,
+      "learning_rate": 6.602162827573422e-08,
+      "loss": 1.0474,
+      "step": 11303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8548954292401425,
+      "learning_rate": 6.570494573809983e-08,
+      "loss": 1.0579,
+      "step": 11304
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5992174714755762,
+      "learning_rate": 6.53890220314879e-08,
+      "loss": 1.0923,
+      "step": 11305
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0176180571472506,
+      "learning_rate": 6.507385718003024e-08,
+      "loss": 1.0703,
+      "step": 11306
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.414678324182024,
+      "learning_rate": 6.47594512077998e-08,
+      "loss": 1.0659,
+      "step": 11307
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.622941806638867,
+      "learning_rate": 6.444580413881407e-08,
+      "loss": 1.0076,
+      "step": 11308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0121089822261586,
+      "learning_rate": 6.413291599702942e-08,
+      "loss": 1.019,
+      "step": 11309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.141982253143655,
+      "learning_rate": 6.382078680634673e-08,
+      "loss": 1.0562,
+      "step": 11310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0520617765491553,
+      "learning_rate": 6.350941659060806e-08,
+      "loss": 1.1348,
+      "step": 11311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8935387007779907,
+      "learning_rate": 6.319880537359769e-08,
+      "loss": 1.0215,
+      "step": 11312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9068216273134913,
+      "learning_rate": 6.288895317903998e-08,
+      "loss": 0.9844,
+      "step": 11313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2562932952162083,
+      "learning_rate": 6.25798600306049e-08,
+      "loss": 1.0623,
+      "step": 11314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.501262963124198,
+      "learning_rate": 6.227152595190133e-08,
+      "loss": 1.0554,
+      "step": 11315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.995469309021561,
+      "learning_rate": 6.196395096648156e-08,
+      "loss": 1.0547,
+      "step": 11316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4114078291553143,
+      "learning_rate": 6.165713509784011e-08,
+      "loss": 1.0852,
+      "step": 11317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3800895983669776,
+      "learning_rate": 6.135107836941379e-08,
+      "loss": 1.0625,
+      "step": 11318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.547335699739497,
+      "learning_rate": 6.104578080457946e-08,
+      "loss": 1.0808,
+      "step": 11319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8496514572843106,
+      "learning_rate": 6.074124242665736e-08,
+      "loss": 1.0979,
+      "step": 11320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5894857187604403,
+      "learning_rate": 6.043746325890998e-08,
+      "loss": 1.0532,
+      "step": 11321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9416793909193917,
+      "learning_rate": 6.01344433245421e-08,
+      "loss": 0.9968,
+      "step": 11322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8250137581295074,
+      "learning_rate": 5.983218264669965e-08,
+      "loss": 1.1003,
+      "step": 11323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0783135815039713,
+      "learning_rate": 5.953068124846861e-08,
+      "loss": 1.0964,
+      "step": 11324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0345130386761663,
+      "learning_rate": 5.9229939152882775e-08,
+      "loss": 1.0879,
+      "step": 11325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2349460356581634,
+      "learning_rate": 5.8929956382911545e-08,
+      "loss": 1.0737,
+      "step": 11326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0829959871574637,
+      "learning_rate": 5.863073296147104e-08,
+      "loss": 1.0947,
+      "step": 11327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.357174885348926,
+      "learning_rate": 5.83322689114163e-08,
+      "loss": 1.0759,
+      "step": 11328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.021188657301105,
+      "learning_rate": 5.8034564255544655e-08,
+      "loss": 1.0779,
+      "step": 11329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4515199247939825,
+      "learning_rate": 5.773761901659791e-08,
+      "loss": 1.052,
+      "step": 11330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8668008739840136,
+      "learning_rate": 5.744143321725792e-08,
+      "loss": 0.9844,
+      "step": 11331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1373498752001674,
+      "learning_rate": 5.714600688014882e-08,
+      "loss": 1.0933,
+      "step": 11332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2846545391678683,
+      "learning_rate": 5.685134002783588e-08,
+      "loss": 1.0278,
+      "step": 11333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2350652600696814,
+      "learning_rate": 5.655743268282776e-08,
+      "loss": 1.0156,
+      "step": 11334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3225549043376414,
+      "learning_rate": 5.62642848675754e-08,
+      "loss": 1.0632,
+      "step": 11335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6066768168426893,
+      "learning_rate": 5.5971896604469776e-08,
+      "loss": 1.114,
+      "step": 11336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.859625188100877,
+      "learning_rate": 5.568026791584413e-08,
+      "loss": 1.0137,
+      "step": 11337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.349182007248598,
+      "learning_rate": 5.5389398823978426e-08,
+      "loss": 1.041,
+      "step": 11338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.572390546395112,
+      "learning_rate": 5.5099289351084885e-08,
+      "loss": 1.0413,
+      "step": 11339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.300760732465663,
+      "learning_rate": 5.480993951932911e-08,
+      "loss": 1.0095,
+      "step": 11340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.165289628742844,
+      "learning_rate": 5.4521349350808996e-08,
+      "loss": 1.0576,
+      "step": 11341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0908909215590716,
+      "learning_rate": 5.423351886757022e-08,
+      "loss": 1.0508,
+      "step": 11342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.622622139632628,
+      "learning_rate": 5.3946448091599654e-08,
+      "loss": 1.0518,
+      "step": 11343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0951815085234022,
+      "learning_rate": 5.366013704482309e-08,
+      "loss": 1.0149,
+      "step": 11344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7139458665496226,
+      "learning_rate": 5.337458574911192e-08,
+      "loss": 1.0471,
+      "step": 11345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0493343584862327,
+      "learning_rate": 5.308979422627758e-08,
+      "loss": 1.0249,
+      "step": 11346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.091125971549011,
+      "learning_rate": 5.280576249807268e-08,
+      "loss": 1.0134,
+      "step": 11347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.725710299978369,
+      "learning_rate": 5.252249058619541e-08,
+      "loss": 0.9869,
+      "step": 11348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3442376424037183,
+      "learning_rate": 5.2239978512280685e-08,
+      "loss": 1.0432,
+      "step": 11349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4584735196922995,
+      "learning_rate": 5.195822629791125e-08,
+      "loss": 1.0796,
+      "step": 11350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1108308937783473,
+      "learning_rate": 5.167723396460655e-08,
+      "loss": 1.1177,
+      "step": 11351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.349372087142544,
+      "learning_rate": 5.1397001533830535e-08,
+      "loss": 1.0137,
+      "step": 11352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.501011271018727,
+      "learning_rate": 5.111752902698941e-08,
+      "loss": 1.0874,
+      "step": 11353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7907238479738794,
+      "learning_rate": 5.083881646542943e-08,
+      "loss": 1.0354,
+      "step": 11354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2238896388100775,
+      "learning_rate": 5.056086387044134e-08,
+      "loss": 1.0017,
+      "step": 11355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.557911884480309,
+      "learning_rate": 5.0283671263255954e-08,
+      "loss": 1.0486,
+      "step": 11356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9323955066885694,
+      "learning_rate": 5.000723866504742e-08,
+      "loss": 1.0242,
+      "step": 11357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.207540539407616,
+      "learning_rate": 4.9731566096929976e-08,
+      "loss": 1.0713,
+      "step": 11358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.007478804934532,
+      "learning_rate": 4.945665357996121e-08,
+      "loss": 1.0562,
+      "step": 11359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.946101172553428,
+      "learning_rate": 4.9182501135141004e-08,
+      "loss": 1.0479,
+      "step": 11360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.19613617957687,
+      "learning_rate": 4.8909108783409265e-08,
+      "loss": 1.0145,
+      "step": 11361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9147569082120026,
+      "learning_rate": 4.863647654564929e-08,
+      "loss": 1.0654,
+      "step": 11362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.908066628601317,
+      "learning_rate": 4.8364604442686645e-08,
+      "loss": 1.0334,
+      "step": 11363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.571182656730547,
+      "learning_rate": 4.8093492495288054e-08,
+      "loss": 1.0262,
+      "step": 11364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.14471074978074,
+      "learning_rate": 4.782314072416361e-08,
+      "loss": 1.0825,
+      "step": 11365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.204933713022535,
+      "learning_rate": 4.7553549149962355e-08,
+      "loss": 0.9717,
+      "step": 11366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1857811837531007,
+      "learning_rate": 4.7284717793276705e-08,
+      "loss": 1.0024,
+      "step": 11367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0097447548699052,
+      "learning_rate": 4.701664667464245e-08,
+      "loss": 1.0214,
+      "step": 11368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4866358227209404,
+      "learning_rate": 4.6749335814536556e-08,
+      "loss": 1.0012,
+      "step": 11369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1878105436252597,
+      "learning_rate": 4.6482785233377126e-08,
+      "loss": 1.0876,
+      "step": 11370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.41764979808915,
+      "learning_rate": 4.621699495152454e-08,
+      "loss": 1.0322,
+      "step": 11371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.483823843377262,
+      "learning_rate": 4.595196498928145e-08,
+      "loss": 1.0972,
+      "step": 11372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0930418964610022,
+      "learning_rate": 4.568769536689166e-08,
+      "loss": 1.0706,
+      "step": 11373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.99497014850216,
+      "learning_rate": 4.5424186104541246e-08,
+      "loss": 1.0889,
+      "step": 11374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.279373078457697,
+      "learning_rate": 4.516143722235855e-08,
+      "loss": 1.0669,
+      "step": 11375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.123400015872041,
+      "learning_rate": 4.4899448740414185e-08,
+      "loss": 1.0718,
+      "step": 11376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9984304922001117,
+      "learning_rate": 4.463822067871993e-08,
+      "loss": 0.9565,
+      "step": 11377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9767036476486015,
+      "learning_rate": 4.43777530572298e-08,
+      "loss": 1.0444,
+      "step": 11378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4847730984643484,
+      "learning_rate": 4.4118045895839015e-08,
+      "loss": 1.0537,
+      "step": 11379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0313444316425713,
+      "learning_rate": 4.385909921438502e-08,
+      "loss": 1.0437,
+      "step": 11380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.278963005596141,
+      "learning_rate": 4.360091303264868e-08,
+      "loss": 0.9946,
+      "step": 11381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1594943735913046,
+      "learning_rate": 4.3343487370350875e-08,
+      "loss": 1.0754,
+      "step": 11382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.534643502749503,
+      "learning_rate": 4.308682224715588e-08,
+      "loss": 1.0806,
+      "step": 11383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1642463365766655,
+      "learning_rate": 4.2830917682668e-08,
+      "loss": 1.0798,
+      "step": 11384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.285485740780987,
+      "learning_rate": 4.2575773696433845e-08,
+      "loss": 1.1462,
+      "step": 11385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8974680790396925,
+      "learning_rate": 4.2321390307944464e-08,
+      "loss": 1.0388,
+      "step": 11386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8226838591488654,
+      "learning_rate": 4.206776753662989e-08,
+      "loss": 1.0454,
+      "step": 11387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.048632954970716,
+      "learning_rate": 4.1814905401863506e-08,
+      "loss": 1.009,
+      "step": 11388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.288543608564778,
+      "learning_rate": 4.156280392295986e-08,
+      "loss": 1.0703,
+      "step": 11389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0751175325466,
+      "learning_rate": 4.131146311917467e-08,
+      "loss": 1.1299,
+      "step": 11390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.460651664511216,
+      "learning_rate": 4.1060883009709226e-08,
+      "loss": 1.1489,
+      "step": 11391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1133598258051434,
+      "learning_rate": 4.081106361370157e-08,
+      "loss": 1.0415,
+      "step": 11392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1753862076335717,
+      "learning_rate": 4.056200495023532e-08,
+      "loss": 1.0781,
+      "step": 11393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.454943576940223,
+      "learning_rate": 4.031370703833526e-08,
+      "loss": 1.0098,
+      "step": 11394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.020965497369858,
+      "learning_rate": 4.006616989696621e-08,
+      "loss": 1.0791,
+      "step": 11395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9662319061978413,
+      "learning_rate": 3.98193935450375e-08,
+      "loss": 1.075,
+      "step": 11396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.080997821385854,
+      "learning_rate": 3.9573378001398486e-08,
+      "loss": 1.0991,
+      "step": 11397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8711208721748247,
+      "learning_rate": 3.932812328484081e-08,
+      "loss": 1.0444,
+      "step": 11398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9656632532229263,
+      "learning_rate": 3.908362941409949e-08,
+      "loss": 1.0378,
+      "step": 11399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.198388296822967,
+      "learning_rate": 3.8839896407849576e-08,
+      "loss": 1.1458,
+      "step": 11400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4363508400660088,
+      "learning_rate": 3.85969242847084e-08,
+      "loss": 1.0686,
+      "step": 11401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1143462679009573,
+      "learning_rate": 3.835471306323557e-08,
+      "loss": 0.9636,
+      "step": 11402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9933362346560184,
+      "learning_rate": 3.8113262761931834e-08,
+      "loss": 0.9949,
+      "step": 11403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3075628748512504,
+      "learning_rate": 3.787257339924133e-08,
+      "loss": 1.0339,
+      "step": 11404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.15959736540742,
+      "learning_rate": 3.763264499354713e-08,
+      "loss": 1.0581,
+      "step": 11405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1371897575259102,
+      "learning_rate": 3.739347756317902e-08,
+      "loss": 1.1101,
+      "step": 11406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.443571695412654,
+      "learning_rate": 3.7155071126404596e-08,
+      "loss": 1.021,
+      "step": 11407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1031145398714206,
+      "learning_rate": 3.691742570143375e-08,
+      "loss": 1.0959,
+      "step": 11408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0052641890637566,
+      "learning_rate": 3.668054130641974e-08,
+      "loss": 1.0253,
+      "step": 11409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.273706886688597,
+      "learning_rate": 3.6444417959456966e-08,
+      "loss": 1.002,
+      "step": 11410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9916116832573914,
+      "learning_rate": 3.6209055678582126e-08,
+      "loss": 1.0278,
+      "step": 11411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.844201563978797,
+      "learning_rate": 3.5974454481771946e-08,
+      "loss": 1.0344,
+      "step": 11412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2258897836619758,
+      "learning_rate": 3.5740614386947644e-08,
+      "loss": 1.0635,
+      "step": 11413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1288764686691133,
+      "learning_rate": 3.55075354119705e-08,
+      "loss": 1.0281,
+      "step": 11414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.148160981688125,
+      "learning_rate": 3.527521757464514e-08,
+      "loss": 1.0508,
+      "step": 11415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0950910298548115,
+      "learning_rate": 3.504366089271627e-08,
+      "loss": 1.1118,
+      "step": 11416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0796904451454377,
+      "learning_rate": 3.4812865383870854e-08,
+      "loss": 1.0376,
+      "step": 11417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2173157926563496,
+      "learning_rate": 3.458283106574034e-08,
+      "loss": 1.0828,
+      "step": 11418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8953188561126826,
+      "learning_rate": 3.435355795589401e-08,
+      "loss": 0.9958,
+      "step": 11419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.953899168662231,
+      "learning_rate": 3.4125046071844525e-08,
+      "loss": 1.0569,
+      "step": 11420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4768981941629553,
+      "learning_rate": 3.3897295431049026e-08,
+      "loss": 1.1084,
+      "step": 11421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.519723936420441,
+      "learning_rate": 3.367030605090249e-08,
+      "loss": 1.0488,
+      "step": 11422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.284846096587558,
+      "learning_rate": 3.344407794874327e-08,
+      "loss": 1.0903,
+      "step": 11423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.377977204767478,
+      "learning_rate": 3.32186111418531e-08,
+      "loss": 1.1028,
+      "step": 11424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1979886624263854,
+      "learning_rate": 3.299390564745264e-08,
+      "loss": 1.012,
+      "step": 11425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1967066085049165,
+      "learning_rate": 3.276996148270817e-08,
+      "loss": 1.093,
+      "step": 11426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.670384947671238,
+      "learning_rate": 3.254677866472378e-08,
+      "loss": 1.0479,
+      "step": 11427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7565497309943416,
+      "learning_rate": 3.2324357210548055e-08,
+      "loss": 0.9937,
+      "step": 11428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8469856657210464,
+      "learning_rate": 3.210269713717074e-08,
+      "loss": 1.0708,
+      "step": 11429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.337709483278368,
+      "learning_rate": 3.1881798461522725e-08,
+      "loss": 1.0999,
+      "step": 11430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9405355736903407,
+      "learning_rate": 3.166166120047831e-08,
+      "loss": 1.0229,
+      "step": 11431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.205297267042587,
+      "learning_rate": 3.144228537085181e-08,
+      "loss": 1.062,
+      "step": 11432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.883920886310856,
+      "learning_rate": 3.1223670989399826e-08,
+      "loss": 0.9861,
+      "step": 11433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2537187931444085,
+      "learning_rate": 3.100581807282232e-08,
+      "loss": 1.0444,
+      "step": 11434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.332672429192833,
+      "learning_rate": 3.0788726637760445e-08,
+      "loss": 1.0295,
+      "step": 11435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4777352186155057,
+      "learning_rate": 3.0572396700795374e-08,
+      "loss": 1.0264,
+      "step": 11436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.324811267031143,
+      "learning_rate": 3.0356828278451656e-08,
+      "loss": 1.0242,
+      "step": 11437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3238476593453607,
+      "learning_rate": 3.014202138719613e-08,
+      "loss": 1.0854,
+      "step": 11438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.866405894717956,
+      "learning_rate": 2.9927976043436777e-08,
+      "loss": 1.0708,
+      "step": 11439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.688153242182445,
+      "learning_rate": 2.971469226352275e-08,
+      "loss": 0.9852,
+      "step": 11440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4299749694363943,
+      "learning_rate": 2.950217006374656e-08,
+      "loss": 1.0603,
+      "step": 11441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.251264163370817,
+      "learning_rate": 2.9290409460340785e-08,
+      "loss": 1.0852,
+      "step": 11442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.743709089245993,
+      "learning_rate": 2.9079410469482484e-08,
+      "loss": 1.0027,
+      "step": 11443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3328491654348706,
+      "learning_rate": 2.886917310728654e-08,
+      "loss": 1.0686,
+      "step": 11444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.297933947904051,
+      "learning_rate": 2.8659697389813445e-08,
+      "loss": 1.0144,
+      "step": 11445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1928892935785305,
+      "learning_rate": 2.845098333306262e-08,
+      "loss": 1.0422,
+      "step": 11446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8689185894720026,
+      "learning_rate": 2.824303095297909e-08,
+      "loss": 1.0889,
+      "step": 11447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1485912788163097,
+      "learning_rate": 2.803584026544459e-08,
+      "loss": 1.1006,
+      "step": 11448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2247977305016042,
+      "learning_rate": 2.7829411286287577e-08,
+      "loss": 0.969,
+      "step": 11449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7627818076198287,
+      "learning_rate": 2.7623744031274323e-08,
+      "loss": 1.0159,
+      "step": 11450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.315956473951998,
+      "learning_rate": 2.7418838516115598e-08,
+      "loss": 1.0742,
+      "step": 11451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0466207969059775,
+      "learning_rate": 2.7214694756464433e-08,
+      "loss": 1.0959,
+      "step": 11452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.476816893630027,
+      "learning_rate": 2.7011312767911692e-08,
+      "loss": 1.0808,
+      "step": 11453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.875539011126396,
+      "learning_rate": 2.680869256599383e-08,
+      "loss": 1.0559,
+      "step": 11454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.604245405422918,
+      "learning_rate": 2.6606834166188478e-08,
+      "loss": 1.0603,
+      "step": 11455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6681828667746177,
+      "learning_rate": 2.6405737583914405e-08,
+      "loss": 1.0447,
+      "step": 11456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.152335785273118,
+      "learning_rate": 2.6205402834532655e-08,
+      "loss": 1.0388,
+      "step": 11457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9769988398019556,
+      "learning_rate": 2.600582993334544e-08,
+      "loss": 1.0254,
+      "step": 11458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2819330035371523,
+      "learning_rate": 2.5807018895596115e-08,
+      "loss": 1.0173,
+      "step": 11459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9109968169975002,
+      "learning_rate": 2.5608969736472534e-08,
+      "loss": 1.0359,
+      "step": 11460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.380014895603039,
+      "learning_rate": 2.5411682471101485e-08,
+      "loss": 1.0693,
+      "step": 11461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2405455138402894,
+      "learning_rate": 2.5215157114554246e-08,
+      "loss": 1.0747,
+      "step": 11462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2090957451624247,
+      "learning_rate": 2.501939368183992e-08,
+      "loss": 1.1062,
+      "step": 11463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4703912753980473,
+      "learning_rate": 2.4824392187914326e-08,
+      "loss": 0.9741,
+      "step": 11464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8763578825972567,
+      "learning_rate": 2.463015264767221e-08,
+      "loss": 1.0059,
+      "step": 11465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.527812689889865,
+      "learning_rate": 2.443667507594838e-08,
+      "loss": 1.0884,
+      "step": 11466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9121085434131473,
+      "learning_rate": 2.4243959487524337e-08,
+      "loss": 1.0098,
+      "step": 11467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1334791379412708,
+      "learning_rate": 2.4052005897119422e-08,
+      "loss": 1.0842,
+      "step": 11468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.037611624304342,
+      "learning_rate": 2.3860814319396353e-08,
+      "loss": 1.0603,
+      "step": 11469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2522331228943426,
+      "learning_rate": 2.3670384768959e-08,
+      "loss": 0.9766,
+      "step": 11470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.265263493505386,
+      "learning_rate": 2.3480717260353502e-08,
+      "loss": 1.0859,
+      "step": 11471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1798560993498746,
+      "learning_rate": 2.3291811808068277e-08,
+      "loss": 1.0327,
+      "step": 11472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.89813788216861,
+      "learning_rate": 2.3103668426530667e-08,
+      "loss": 0.9697,
+      "step": 11473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7704163796091645,
+      "learning_rate": 2.2916287130114733e-08,
+      "loss": 1.0986,
+      "step": 11474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5156785386035505,
+      "learning_rate": 2.272966793313125e-08,
+      "loss": 1.0347,
+      "step": 11475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5083099869492496,
+      "learning_rate": 2.254381084983659e-08,
+      "loss": 1.0984,
+      "step": 11476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1343632603232456,
+      "learning_rate": 2.2358715894427174e-08,
+      "loss": 1.0803,
+      "step": 11477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.298426415029444,
+      "learning_rate": 2.2174383081041694e-08,
+      "loss": 1.0417,
+      "step": 11478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3509973253574175,
+      "learning_rate": 2.199081242376e-08,
+      "loss": 1.1184,
+      "step": 11479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1239683434665575,
+      "learning_rate": 2.1808003936604207e-08,
+      "loss": 1.0894,
+      "step": 11480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8541108515194162,
+      "learning_rate": 2.1625957633537586e-08,
+      "loss": 1.0483,
+      "step": 11481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8131850120352024,
+      "learning_rate": 2.1444673528466797e-08,
+      "loss": 1.0505,
+      "step": 11482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4142249920342667,
+      "learning_rate": 2.126415163523854e-08,
+      "loss": 1.0962,
+      "step": 11483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8022663087808652,
+      "learning_rate": 2.1084391967642893e-08,
+      "loss": 1.0059,
+      "step": 11484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.773620715910919,
+      "learning_rate": 2.0905394539409984e-08,
+      "loss": 1.0261,
+      "step": 11485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.102158868060379,
+      "learning_rate": 2.072715936421221e-08,
+      "loss": 1.0347,
+      "step": 11486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.054494533776351,
+      "learning_rate": 2.054968645566424e-08,
+      "loss": 0.9954,
+      "step": 11487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.242735950059371,
+      "learning_rate": 2.0372975827321894e-08,
+      "loss": 1.1326,
+      "step": 11488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.042088714188675,
+      "learning_rate": 2.0197027492685483e-08,
+      "loss": 1.0312,
+      "step": 11489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.755825839558439,
+      "learning_rate": 2.0021841465192038e-08,
+      "loss": 0.9802,
+      "step": 11490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3330892251983486,
+      "learning_rate": 1.9847417758225294e-08,
+      "loss": 1.0283,
+      "step": 11491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.85429581217239,
+      "learning_rate": 1.9673756385106825e-08,
+      "loss": 1.0652,
+      "step": 11492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0402607054258217,
+      "learning_rate": 1.9500857359102676e-08,
+      "loss": 1.0652,
+      "step": 11493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.296652164883357,
+      "learning_rate": 1.9328720693420065e-08,
+      "loss": 1.0408,
+      "step": 11494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.270362378062202,
+      "learning_rate": 1.9157346401207365e-08,
+      "loss": 1.1018,
+      "step": 11495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6448196015202114,
+      "learning_rate": 1.8986734495554106e-08,
+      "loss": 1.0945,
+      "step": 11496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.064415950622607,
+      "learning_rate": 1.88168849894943e-08,
+      "loss": 1.05,
+      "step": 11497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.343830894640324,
+      "learning_rate": 1.8647797895999797e-08,
+      "loss": 1.0369,
+      "step": 11498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.411324935469539,
+      "learning_rate": 1.847947322798804e-08,
+      "loss": 1.0835,
+      "step": 11499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0943931473413864,
+      "learning_rate": 1.8311910998315417e-08,
+      "loss": 1.0884,
+      "step": 11500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.408819701309463,
+      "learning_rate": 1.8145111219781687e-08,
+      "loss": 1.1101,
+      "step": 11501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3032535306244686,
+      "learning_rate": 1.7979073905127763e-08,
+      "loss": 1.1018,
+      "step": 11502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.056950939593433,
+      "learning_rate": 1.781379906703573e-08,
+      "loss": 1.0291,
+      "step": 11503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.519559657154751,
+      "learning_rate": 1.7649286718131043e-08,
+      "loss": 1.1052,
+      "step": 11504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.195081011657545,
+      "learning_rate": 1.7485536870980312e-08,
+      "loss": 1.0071,
+      "step": 11505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.106775841851473,
+      "learning_rate": 1.7322549538091316e-08,
+      "loss": 0.9619,
+      "step": 11506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.522633585963921,
+      "learning_rate": 1.716032473191187e-08,
+      "loss": 1.0349,
+      "step": 11507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1913793061063354,
+      "learning_rate": 1.699886246483651e-08,
+      "loss": 1.0825,
+      "step": 11508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5687719087213154,
+      "learning_rate": 1.6838162749196473e-08,
+      "loss": 0.9795,
+      "step": 11509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0550666054619553,
+      "learning_rate": 1.6678225597266394e-08,
+      "loss": 1.1248,
+      "step": 11510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7563724992272816,
+      "learning_rate": 1.6519051021266496e-08,
+      "loss": 1.0303,
+      "step": 11511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9032685497852095,
+      "learning_rate": 1.636063903335039e-08,
+      "loss": 1.0259,
+      "step": 11512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.008041502081091,
+      "learning_rate": 1.6202989645622835e-08,
+      "loss": 0.9768,
+      "step": 11513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1977053467626466,
+      "learning_rate": 1.604610287012309e-08,
+      "loss": 1.074,
+      "step": 11514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2070567310659133,
+      "learning_rate": 1.5889978718836018e-08,
+      "loss": 1.1277,
+      "step": 11515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.609950585684829,
+      "learning_rate": 1.573461720368763e-08,
+      "loss": 0.9468,
+      "step": 11516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4519181050586214,
+      "learning_rate": 1.5580018336543988e-08,
+      "loss": 1.051,
+      "step": 11517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.240122063962729,
+      "learning_rate": 1.542618212921454e-08,
+      "loss": 1.0488,
+      "step": 11518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6219046853879737,
+      "learning_rate": 1.527310859344988e-08,
+      "loss": 0.99,
+      "step": 11519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3414201762415576,
+      "learning_rate": 1.5120797740943995e-08,
+      "loss": 1.073,
+      "step": 11520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1089449592911933,
+      "learning_rate": 1.496924958332868e-08,
+      "loss": 1.0803,
+      "step": 11521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.935867198978687,
+      "learning_rate": 1.4818464132182464e-08,
+      "loss": 1.0518,
+      "step": 11522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3163899017886966,
+      "learning_rate": 1.4668441399021683e-08,
+      "loss": 1.0889,
+      "step": 11523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.377344642216474,
+      "learning_rate": 1.4519181395306058e-08,
+      "loss": 0.9822,
+      "step": 11524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.408291244324077,
+      "learning_rate": 1.4370684132436474e-08,
+      "loss": 0.998,
+      "step": 11525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0356768033743724,
+      "learning_rate": 1.4222949621756078e-08,
+      "loss": 1.0991,
+      "step": 11526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.957806114918743,
+      "learning_rate": 1.4075977874550284e-08,
+      "loss": 1.0605,
+      "step": 11527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.26900752962146,
+      "learning_rate": 1.3929768902044561e-08,
+      "loss": 1.0349,
+      "step": 11528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3449252904202735,
+      "learning_rate": 1.378432271540775e-08,
+      "loss": 1.0364,
+      "step": 11529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9826475205932685,
+      "learning_rate": 1.3639639325748744e-08,
+      "loss": 1.1074,
+      "step": 11530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0952001713629,
+      "learning_rate": 1.3495718744120922e-08,
+      "loss": 1.1111,
+      "step": 11531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3951766147700173,
+      "learning_rate": 1.3352560981516604e-08,
+      "loss": 0.9956,
+      "step": 11532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8663360581390513,
+      "learning_rate": 1.3210166048870377e-08,
+      "loss": 1.0537,
+      "step": 11533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4336424992011634,
+      "learning_rate": 1.3068533957060204e-08,
+      "loss": 1.0847,
+      "step": 11534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4777849243577563,
+      "learning_rate": 1.2927664716902988e-08,
+      "loss": 1.093,
+      "step": 11535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0012642486350014,
+      "learning_rate": 1.2787558339161233e-08,
+      "loss": 1.0632,
+      "step": 11536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0168389739137242,
+      "learning_rate": 1.2648214834535266e-08,
+      "loss": 0.9714,
+      "step": 11537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5757302448933728,
+      "learning_rate": 1.2509634213669907e-08,
+      "loss": 1.0986,
+      "step": 11538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2011011116854213,
+      "learning_rate": 1.2371816487150023e-08,
+      "loss": 1.0413,
+      "step": 11539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9671956618954076,
+      "learning_rate": 1.2234761665502749e-08,
+      "loss": 1.1187,
+      "step": 11540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.022844423206888,
+      "learning_rate": 1.2098469759197484e-08,
+      "loss": 1.0439,
+      "step": 11541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.362644888336329,
+      "learning_rate": 1.1962940778644794e-08,
+      "loss": 1.0981,
+      "step": 11542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5608596627749116,
+      "learning_rate": 1.1828174734196395e-08,
+      "loss": 1.0889,
+      "step": 11543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2919428987754094,
+      "learning_rate": 1.169417163614739e-08,
+      "loss": 1.1106,
+      "step": 11544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3678991956581186,
+      "learning_rate": 1.1560931494732919e-08,
+      "loss": 0.989,
+      "step": 11545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.436860816494553,
+      "learning_rate": 1.14284543201304e-08,
+      "loss": 1.0515,
+      "step": 11546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1570002033410502,
+      "learning_rate": 1.1296740122459515e-08,
+      "loss": 0.9761,
+      "step": 11547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.276166908493044,
+      "learning_rate": 1.1165788911781106e-08,
+      "loss": 1.0688,
+      "step": 11548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.257879867417487,
+      "learning_rate": 1.103560069809828e-08,
+      "loss": 1.1233,
+      "step": 11549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1320827696623916,
+      "learning_rate": 1.0906175491354199e-08,
+      "loss": 1.0046,
+      "step": 11550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.262242886552209,
+      "learning_rate": 1.0777513301437614e-08,
+      "loss": 1.0312,
+      "step": 11551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.960521180962732,
+      "learning_rate": 1.0649614138172892e-08,
+      "loss": 1.0317,
+      "step": 11552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.015656383998408,
+      "learning_rate": 1.0522478011332216e-08,
+      "loss": 0.9475,
+      "step": 11553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4921433960964325,
+      "learning_rate": 1.0396104930626704e-08,
+      "loss": 1.015,
+      "step": 11554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.393324847842735,
+      "learning_rate": 1.0270494905707529e-08,
+      "loss": 1.0728,
+      "step": 11555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.501803115013598,
+      "learning_rate": 1.0145647946171456e-08,
+      "loss": 1.0635,
+      "step": 11556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1987564857850344,
+      "learning_rate": 1.0021564061554189e-08,
+      "loss": 1.0396,
+      "step": 11557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.357046317687475,
+      "learning_rate": 9.898243261332597e-09,
+      "loss": 1.0955,
+      "step": 11558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.071220835366089,
+      "learning_rate": 9.775685554928027e-09,
+      "loss": 1.0513,
+      "step": 11559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4602697715344966,
+      "learning_rate": 9.653890951701884e-09,
+      "loss": 1.0498,
+      "step": 11560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3771380931761343,
+      "learning_rate": 9.532859460957833e-09,
+      "loss": 1.0344,
+      "step": 11561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.086633429643087,
+      "learning_rate": 9.412591091939593e-09,
+      "loss": 1.0713,
+      "step": 11562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.392503660583902,
+      "learning_rate": 9.293085853835371e-09,
+      "loss": 1.0251,
+      "step": 11563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1534674321596854,
+      "learning_rate": 9.174343755772307e-09,
+      "loss": 1.064,
+      "step": 11564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3288150201327484,
+      "learning_rate": 9.056364806820927e-09,
+      "loss": 1.0166,
+      "step": 11565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3403223986515362,
+      "learning_rate": 8.93914901599291e-09,
+      "loss": 1.0503,
+      "step": 11566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.147335757045514,
+      "learning_rate": 8.822696392243312e-09,
+      "loss": 1.0916,
+      "step": 11567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8179694606221175,
+      "learning_rate": 8.707006944463915e-09,
+      "loss": 0.984,
+      "step": 11568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3421483618619074,
+      "learning_rate": 8.592080681495418e-09,
+      "loss": 1.1125,
+      "step": 11569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.419912872133763,
+      "learning_rate": 8.47791761211414e-09,
+      "loss": 1.1189,
+      "step": 11570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0081502864092085,
+      "learning_rate": 8.364517745040879e-09,
+      "loss": 1.0464,
+      "step": 11571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.01636400133293,
+      "learning_rate": 8.251881088938706e-09,
+      "loss": 1.0837,
+      "step": 11572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.266014693020187,
+      "learning_rate": 8.140007652409632e-09,
+      "loss": 1.0229,
+      "step": 11573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7626262022727652,
+      "learning_rate": 8.02889744400126e-09,
+      "loss": 1.0315,
+      "step": 11574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4922720897006947,
+      "learning_rate": 7.918550472199027e-09,
+      "loss": 0.9944,
+      "step": 11575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8826430778644703,
+      "learning_rate": 7.808966745431745e-09,
+      "loss": 1.0254,
+      "step": 11576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1857005485127408,
+      "learning_rate": 7.700146272070496e-09,
+      "loss": 1.0476,
+      "step": 11577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1387603519706992,
+      "learning_rate": 7.59208906042752e-09,
+      "loss": 1.0925,
+      "step": 11578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6286190433098664,
+      "learning_rate": 7.484795118757326e-09,
+      "loss": 1.1421,
+      "step": 11579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8555152136033497,
+      "learning_rate": 7.37826445525558e-09,
+      "loss": 1.0547,
+      "step": 11580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2172218978853024,
+      "learning_rate": 7.272497078057994e-09,
+      "loss": 1.0813,
+      "step": 11581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8432155359223095,
+      "learning_rate": 7.167492995244773e-09,
+      "loss": 1.095,
+      "step": 11582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.022499207580619,
+      "learning_rate": 7.0632522148372795e-09,
+      "loss": 1.0903,
+      "step": 11583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6495664230580154,
+      "learning_rate": 6.959774744796921e-09,
+      "loss": 0.97,
+      "step": 11584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.578841805236957,
+      "learning_rate": 6.857060593028486e-09,
+      "loss": 1.0649,
+      "step": 11585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5102594144968866,
+      "learning_rate": 6.755109767376811e-09,
+      "loss": 1.0194,
+      "step": 11586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1426755874923917,
+      "learning_rate": 6.653922275630109e-09,
+      "loss": 1.0479,
+      "step": 11587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.251946834180091,
+      "learning_rate": 6.553498125517754e-09,
+      "loss": 1.0315,
+      "step": 11588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.47283162469613,
+      "learning_rate": 6.453837324710277e-09,
+      "loss": 1.0493,
+      "step": 11589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2377507249161463,
+      "learning_rate": 6.354939880820476e-09,
+      "loss": 1.0591,
+      "step": 11590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.034693982925071,
+      "learning_rate": 6.2568058014023105e-09,
+      "loss": 1.0703,
+      "step": 11591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.082558773497364,
+      "learning_rate": 6.159435093952004e-09,
+      "loss": 1.021,
+      "step": 11592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9491642481144686,
+      "learning_rate": 6.062827765906942e-09,
+      "loss": 1.0391,
+      "step": 11593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.119005908576888,
+      "learning_rate": 5.966983824646777e-09,
+      "loss": 0.9998,
+      "step": 11594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7366915499938753,
+      "learning_rate": 5.871903277493429e-09,
+      "loss": 1.031,
+      "step": 11595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0914152729822435,
+      "learning_rate": 5.777586131707758e-09,
+      "loss": 1.0342,
+      "step": 11596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.178796349240796,
+      "learning_rate": 5.684032394496219e-09,
+      "loss": 1.0959,
+      "step": 11597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.165647510170102,
+      "learning_rate": 5.5912420730031e-09,
+      "loss": 1.0671,
+      "step": 11598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.148994898440511,
+      "learning_rate": 5.499215174317174e-09,
+      "loss": 1.0518,
+      "step": 11599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5174898838665025,
+      "learning_rate": 5.4079517054672625e-09,
+      "loss": 1.0605,
+      "step": 11600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1272776429050406,
+      "learning_rate": 5.317451673424456e-09,
+      "loss": 1.0479,
+      "step": 11601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.939444146903005,
+      "learning_rate": 5.227715085103224e-09,
+      "loss": 0.9778,
+      "step": 11602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.176879430348821,
+      "learning_rate": 5.138741947355863e-09,
+      "loss": 1.074,
+      "step": 11603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5992488151403723,
+      "learning_rate": 5.050532266980268e-09,
+      "loss": 1.0913,
+      "step": 11604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2965015619654645,
+      "learning_rate": 4.9630860507132725e-09,
+      "loss": 1.0647,
+      "step": 11605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.49487857718235,
+      "learning_rate": 4.876403305235089e-09,
+      "loss": 1.1167,
+      "step": 11606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0906578904152955,
+      "learning_rate": 4.790484037167087e-09,
+      "loss": 1.0127,
+      "step": 11607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.652852005412053,
+      "learning_rate": 4.705328253071795e-09,
+      "loss": 0.9973,
+      "step": 11608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.037465413649934,
+      "learning_rate": 4.620935959454009e-09,
+      "loss": 1.0493,
+      "step": 11609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9917821485868212,
+      "learning_rate": 4.537307162759685e-09,
+      "loss": 0.9805,
+      "step": 11610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.306457980903054,
+      "learning_rate": 4.454441869377047e-09,
+      "loss": 1.0503,
+      "step": 11611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3837791292724573,
+      "learning_rate": 4.372340085636584e-09,
+      "loss": 1.0857,
+      "step": 11612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1956780693930003,
+      "learning_rate": 4.291001817807727e-09,
+      "loss": 1.1182,
+      "step": 11613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.774880801501033,
+      "learning_rate": 4.210427072105505e-09,
+      "loss": 1.1121,
+      "step": 11614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.461861867336328,
+      "learning_rate": 4.130615854683884e-09,
+      "loss": 1.0981,
+      "step": 11615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8159794565072245,
+      "learning_rate": 4.051568171637987e-09,
+      "loss": 1.043,
+      "step": 11616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3755114466997065,
+      "learning_rate": 3.973284029007429e-09,
+      "loss": 1.0972,
+      "step": 11617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.164678447869263,
+      "learning_rate": 3.895763432771871e-09,
+      "loss": 1.0325,
+      "step": 11618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1229376705226586,
+      "learning_rate": 3.819006388852131e-09,
+      "loss": 1.0352,
+      "step": 11619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.501699308653167,
+      "learning_rate": 3.7430129031124085e-09,
+      "loss": 1.0752,
+      "step": 11620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2418535693966373,
+      "learning_rate": 3.6677829813558387e-09,
+      "loss": 1.074,
+      "step": 11621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.129728466623194,
+      "learning_rate": 3.5933166293289355e-09,
+      "loss": 1.0938,
+      "step": 11622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1675211701086954,
+      "learning_rate": 3.519613852721593e-09,
+      "loss": 0.9792,
+      "step": 11623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5994756870767017,
+      "learning_rate": 3.4466746571615307e-09,
+      "loss": 1.0332,
+      "step": 11624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0427605710172916,
+      "learning_rate": 3.3744990482220687e-09,
+      "loss": 1.0291,
+      "step": 11625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0173631824556377,
+      "learning_rate": 3.3030870314143538e-09,
+      "loss": 1.0559,
+      "step": 11626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4033644225930306,
+      "learning_rate": 3.2324386121951325e-09,
+      "loss": 1.1016,
+      "step": 11627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.04365258374288,
+      "learning_rate": 3.1625537959589782e-09,
+      "loss": 1.0669,
+      "step": 11628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2877330038293846,
+      "learning_rate": 3.0934325880460636e-09,
+      "loss": 1.0471,
+      "step": 11629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8277167086781225,
+      "learning_rate": 3.0250749937354995e-09,
+      "loss": 1.0125,
+      "step": 11630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.303229803231255,
+      "learning_rate": 2.9574810182475546e-09,
+      "loss": 1.0154,
+      "step": 11631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.908721442993646,
+      "learning_rate": 2.8906506667469857e-09,
+      "loss": 1.0271,
+      "step": 11632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3684837141791952,
+      "learning_rate": 2.824583944337489e-09,
+      "loss": 1.073,
+      "step": 11633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0934984402601082,
+      "learning_rate": 2.7592808560661376e-09,
+      "loss": 1.002,
+      "step": 11634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.664522273202468,
+      "learning_rate": 2.6947414069222744e-09,
+      "loss": 1.0598,
+      "step": 11635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.439533026924449,
+      "learning_rate": 2.6309656018330685e-09,
+      "loss": 1.051,
+      "step": 11636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1539294821660984,
+      "learning_rate": 2.5679534456723998e-09,
+      "loss": 1.0676,
+      "step": 11637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8631970307322954,
+      "learning_rate": 2.5057049432519744e-09,
+      "loss": 1.0449,
+      "step": 11638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.96293664783663,
+      "learning_rate": 2.4442200993268773e-09,
+      "loss": 1.0356,
+      "step": 11639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8434829416216965,
+      "learning_rate": 2.383498918594462e-09,
+      "loss": 1.0479,
+      "step": 11640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2779383809745424,
+      "learning_rate": 2.3235414056921314e-09,
+      "loss": 1.085,
+      "step": 11641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.916937624335654,
+      "learning_rate": 2.2643475651995538e-09,
+      "loss": 1.064,
+      "step": 11642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3348242168969753,
+      "learning_rate": 2.205917401639779e-09,
+      "loss": 1.0554,
+      "step": 11643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0266424478111245,
+      "learning_rate": 2.1482509194747923e-09,
+      "loss": 1.0615,
+      "step": 11644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2543329721711274,
+      "learning_rate": 2.0913481231088497e-09,
+      "loss": 1.0229,
+      "step": 11645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.279433131795188,
+      "learning_rate": 2.035209016889583e-09,
+      "loss": 1.0774,
+      "step": 11646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.682656279593887,
+      "learning_rate": 1.979833605104675e-09,
+      "loss": 1.1099,
+      "step": 11647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7999416023622925,
+      "learning_rate": 1.9252218919840746e-09,
+      "loss": 0.9912,
+      "step": 11648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.446544095545234,
+      "learning_rate": 1.8713738817000003e-09,
+      "loss": 1.0571,
+      "step": 11649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.951790030188192,
+      "learning_rate": 1.818289578363608e-09,
+      "loss": 1.0652,
+      "step": 11650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0469282390149117,
+      "learning_rate": 1.7659689860327623e-09,
+      "loss": 1.0098,
+      "step": 11651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9021877497826654,
+      "learning_rate": 1.7144121087009358e-09,
+      "loss": 1.0903,
+      "step": 11652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3908130666833185,
+      "learning_rate": 1.66361895030831e-09,
+      "loss": 1.0542,
+      "step": 11653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4536206850608666,
+      "learning_rate": 1.6135895147340042e-09,
+      "loss": 0.9985,
+      "step": 11654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.941804509665227,
+      "learning_rate": 1.5643238057994059e-09,
+      "loss": 1.0054,
+      "step": 11655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7623599911828403,
+      "learning_rate": 1.5158218272670611e-09,
+      "loss": 1.0818,
+      "step": 11656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.450250662687098,
+      "learning_rate": 1.4680835828440043e-09,
+      "loss": 1.0352,
+      "step": 11657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1189944887093946,
+      "learning_rate": 1.4211090761739876e-09,
+      "loss": 1.1399,
+      "step": 11658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.889360399592933,
+      "learning_rate": 1.3748983108474722e-09,
+      "loss": 1.0713,
+      "step": 11659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7174064787371033,
+      "learning_rate": 1.329451290392747e-09,
+      "loss": 1.0354,
+      "step": 11660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.99359348182009,
+      "learning_rate": 1.2847680182825894e-09,
+      "loss": 1.0046,
+      "step": 11661
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.640728411950442,
+      "learning_rate": 1.2408484979276047e-09,
+      "loss": 1.0977,
+      "step": 11662
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.180163445910769,
+      "learning_rate": 1.1976927326862175e-09,
+      "loss": 0.9919,
+      "step": 11663
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3175107868362557,
+      "learning_rate": 1.1553007258524596e-09,
+      "loss": 1.075,
+      "step": 11664
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2092098946537164,
+      "learning_rate": 1.1136724806637411e-09,
+      "loss": 1.0527,
+      "step": 11665
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0563627592057907,
+      "learning_rate": 1.0728080003019614e-09,
+      "loss": 1.1067,
+      "step": 11666
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.499196941237933,
+      "learning_rate": 1.0327072878879574e-09,
+      "loss": 1.0793,
+      "step": 11667
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.657436311925946,
+      "learning_rate": 9.93370346483724e-10,
+      "loss": 1.0017,
+      "step": 11668
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8138225601527425,
+      "learning_rate": 9.547971790957455e-10,
+      "loss": 1.0388,
+      "step": 11669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.04104327124657,
+      "learning_rate": 9.169877886694434e-10,
+      "loss": 1.0403,
+      "step": 11670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.965674884120844,
+      "learning_rate": 8.799421780925077e-10,
+      "loss": 1.0544,
+      "step": 11671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.959854997836441,
+      "learning_rate": 8.43660350194897e-10,
+      "loss": 1.0168,
+      "step": 11672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.388479971192719,
+      "learning_rate": 8.081423077477279e-10,
+      "loss": 1.0613,
+      "step": 11673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7603601684576513,
+      "learning_rate": 7.733880534654959e-10,
+      "loss": 1.0042,
+      "step": 11674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.203773123383493,
+      "learning_rate": 7.393975900016336e-10,
+      "loss": 1.0642,
+      "step": 11675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5358768945076533,
+      "learning_rate": 7.061709199529532e-10,
+      "loss": 1.0283,
+      "step": 11676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.229440156173173,
+      "learning_rate": 6.737080458574241e-10,
+      "loss": 1.0237,
+      "step": 11677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0607528665128827,
+      "learning_rate": 6.420089701941745e-10,
+      "loss": 1.0747,
+      "step": 11678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.396964014567922,
+      "learning_rate": 6.110736953846008e-10,
+      "loss": 1.0061,
+      "step": 11679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8154130107201807,
+      "learning_rate": 5.809022237923678e-10,
+      "loss": 1.031,
+      "step": 11680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3613713452587786,
+      "learning_rate": 5.514945577222985e-10,
+      "loss": 1.0493,
+      "step": 11681
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7715480353898623,
+      "learning_rate": 5.228506994203741e-10,
+      "loss": 1.0522,
+      "step": 11682
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0440298228729423,
+      "learning_rate": 4.94970651073734e-10,
+      "loss": 1.0759,
+      "step": 11683
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.251311074989096,
+      "learning_rate": 4.678544148128961e-10,
+      "loss": 1.085,
+      "step": 11684
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.296855682388769,
+      "learning_rate": 4.415019927084263e-10,
+      "loss": 1.0791,
+      "step": 11685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.983107092003336,
+      "learning_rate": 4.1591338677426927e-10,
+      "loss": 1.0295,
+      "step": 11686
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2178194610013424,
+      "learning_rate": 3.9108859896441755e-10,
+      "loss": 1.0823,
+      "step": 11687
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.302363163622739,
+      "learning_rate": 3.67027631175132e-10,
+      "loss": 1.0332,
+      "step": 11688
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4014832152801184,
+      "learning_rate": 3.437304852438317e-10,
+      "loss": 0.9526,
+      "step": 11689
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7679840212903235,
+      "learning_rate": 3.211971629513144e-10,
+      "loss": 1.0286,
+      "step": 11690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6079906150570324,
+      "learning_rate": 2.9942766601731563e-10,
+      "loss": 1.0857,
+      "step": 11691
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2027112152745505,
+      "learning_rate": 2.784219961060597e-10,
+      "loss": 1.0166,
+      "step": 11692
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.974475526177777,
+      "learning_rate": 2.581801548218188e-10,
+      "loss": 1.0496,
+      "step": 11693
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.529156215262907,
+      "learning_rate": 2.387021437100234e-10,
+      "loss": 1.0352,
+      "step": 11694
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.062531873113828,
+      "learning_rate": 2.1998796425837244e-10,
+      "loss": 1.0117,
+      "step": 11695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6686900801424205,
+      "learning_rate": 2.0203761789683307e-10,
+      "loss": 1.0339,
+      "step": 11696
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3178676113983356,
+      "learning_rate": 1.8485110599653077e-10,
+      "loss": 1.0598,
+      "step": 11697
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.366010352677635,
+      "learning_rate": 1.684284298708594e-10,
+      "loss": 1.0674,
+      "step": 11698
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.360923359670248,
+      "learning_rate": 1.527695907732607e-10,
+      "loss": 1.0605,
+      "step": 11699
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.335605660668916,
+      "learning_rate": 1.3787458990055514e-10,
+      "loss": 1.083,
+      "step": 11700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9505963441219927,
+      "learning_rate": 1.237434283907213e-10,
+      "loss": 1.009,
+      "step": 11701
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2330095425372245,
+      "learning_rate": 1.1037610732178572e-10,
+      "loss": 1.0972,
+      "step": 11702
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.368056953410495,
+      "learning_rate": 9.777262771515362e-11,
+      "loss": 1.0962,
+      "step": 11703
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.326888598968205,
+      "learning_rate": 8.593299053449855e-11,
+      "loss": 1.0718,
+      "step": 11704
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.83880186612695,
+      "learning_rate": 7.48571966846523e-11,
+      "loss": 1.03,
+      "step": 11705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7784428786091335,
+      "learning_rate": 6.454524700938435e-11,
+      "loss": 1.0271,
+      "step": 11706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.994905673600549,
+      "learning_rate": 5.499714229917352e-11,
+      "loss": 0.9946,
+      "step": 11707
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.108198982794177,
+      "learning_rate": 4.621288328121587e-11,
+      "loss": 1.1113,
+      "step": 11708
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.219843758465381,
+      "learning_rate": 3.819247062608611e-11,
+      "loss": 1.0479,
+      "step": 11709
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.024385994798507,
+      "learning_rate": 3.09359049488478e-11,
+      "loss": 1.0701,
+      "step": 11710
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.927551610718581,
+      "learning_rate": 2.4443186802391993e-11,
+      "loss": 1.0562,
+      "step": 11711
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9033646149531958,
+      "learning_rate": 1.8714316681878175e-11,
+      "loss": 1.0286,
+      "step": 11712
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6004596386846015,
+      "learning_rate": 1.3749295026954657e-11,
+      "loss": 0.9946,
+      "step": 11713
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.52501728096638,
+      "learning_rate": 9.548122215097266e-12,
+      "loss": 1.0593,
+      "step": 11714
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9055498009354874,
+      "learning_rate": 6.110798567160459e-12,
+      "loss": 1.0327,
+      "step": 11715
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.023209572950124,
+      "learning_rate": 3.4373243473773134e-12,
+      "loss": 1.0288,
+      "step": 11716
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.637804571884393,
+      "learning_rate": 1.5276997589186437e-12,
+      "loss": 1.0042,
+      "step": 11717
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0317654077794227,
+      "learning_rate": 3.8192494722366635e-13,
+      "loss": 1.0361,
+      "step": 11718
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7190443451157984,
+      "learning_rate": 0.0,
+      "loss": 1.1354,
+      "step": 11719
+    },
+    {
+      "epoch": 1.0,
+      "step": 11719,
+      "total_flos": 1.149592562375118e+19,
+      "train_loss": 1.152523079507637,
+      "train_runtime": 32977.5443,
+      "train_samples_per_second": 45.486,
+      "train_steps_per_second": 0.355
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11719,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.149592562375118e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}