diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9850 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1401,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0021413276231263384,
+      "grad_norm": 56.592826307068876,
+      "learning_rate": 3.546099290780142e-07,
+      "loss": 3.0396,
+      "step": 1
+    },
+    {
+      "epoch": 0.004282655246252677,
+      "grad_norm": 60.92994210864957,
+      "learning_rate": 7.092198581560284e-07,
+      "loss": 3.4422,
+      "step": 2
+    },
+    {
+      "epoch": 0.006423982869379015,
+      "grad_norm": 58.66990590000166,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 3.2284,
+      "step": 3
+    },
+    {
+      "epoch": 0.008565310492505354,
+      "grad_norm": 57.90854902052254,
+      "learning_rate": 1.4184397163120568e-06,
+      "loss": 3.1414,
+      "step": 4
+    },
+    {
+      "epoch": 0.010706638115631691,
+      "grad_norm": 58.311165642100505,
+      "learning_rate": 1.7730496453900712e-06,
+      "loss": 3.2059,
+      "step": 5
+    },
+    {
+      "epoch": 0.01284796573875803,
+      "grad_norm": 56.80687951003449,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 3.1799,
+      "step": 6
+    },
+    {
+      "epoch": 0.014989293361884369,
+      "grad_norm": 47.7995425239194,
+      "learning_rate": 2.4822695035460995e-06,
+      "loss": 2.7636,
+      "step": 7
+    },
+    {
+      "epoch": 0.017130620985010708,
+      "grad_norm": 49.227409613395615,
+      "learning_rate": 2.8368794326241136e-06,
+      "loss": 2.8918,
+      "step": 8
+    },
+    {
+      "epoch": 0.019271948608137045,
+      "grad_norm": 31.314007438717645,
+      "learning_rate": 3.1914893617021277e-06,
+      "loss": 2.1702,
+      "step": 9
+    },
+    {
+      "epoch": 0.021413276231263382,
+      "grad_norm": 24.736226056643922,
+      "learning_rate": 3.5460992907801423e-06,
+      "loss": 1.9433,
+      "step": 10
+    },
+    {
+      "epoch": 0.023554603854389723,
+      "grad_norm": 5.183793675647891,
+      "learning_rate": 3.9007092198581565e-06,
+      "loss": 1.2843,
+      "step": 11
+    },
+    {
+      "epoch": 0.02569593147751606,
+      "grad_norm": 4.360295224010352,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.2718,
+      "step": 12
+    },
+    {
+      "epoch": 0.027837259100642397,
+      "grad_norm": 3.516837280046668,
+      "learning_rate": 4.609929078014184e-06,
+      "loss": 1.1784,
+      "step": 13
+    },
+    {
+      "epoch": 0.029978586723768737,
+      "grad_norm": 2.9770123258300263,
+      "learning_rate": 4.964539007092199e-06,
+      "loss": 1.2087,
+      "step": 14
+    },
+    {
+      "epoch": 0.032119914346895075,
+      "grad_norm": 3.371301698123799,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 1.1044,
+      "step": 15
+    },
+    {
+      "epoch": 0.034261241970021415,
+      "grad_norm": 2.4929900637756104,
+      "learning_rate": 5.673758865248227e-06,
+      "loss": 1.0305,
+      "step": 16
+    },
+    {
+      "epoch": 0.03640256959314775,
+      "grad_norm": 2.3245385930370683,
+      "learning_rate": 6.028368794326241e-06,
+      "loss": 1.0328,
+      "step": 17
+    },
+    {
+      "epoch": 0.03854389721627409,
+      "grad_norm": 2.1494749257738106,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 0.9676,
+      "step": 18
+    },
+    {
+      "epoch": 0.04068522483940043,
+      "grad_norm": 1.7068403808367762,
+      "learning_rate": 6.73758865248227e-06,
+      "loss": 0.9568,
+      "step": 19
+    },
+    {
+      "epoch": 0.042826552462526764,
+      "grad_norm": 1.3157208379186793,
+      "learning_rate": 7.092198581560285e-06,
+      "loss": 0.8919,
+      "step": 20
+    },
+    {
+      "epoch": 0.044967880085653104,
+      "grad_norm": 1.5755739795069426,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.8786,
+      "step": 21
+    },
+    {
+      "epoch": 0.047109207708779445,
+      "grad_norm": 1.4066443822922425,
+      "learning_rate": 7.801418439716313e-06,
+      "loss": 0.8728,
+      "step": 22
+    },
+    {
+      "epoch": 0.04925053533190578,
+      "grad_norm": 1.232711682082585,
+      "learning_rate": 8.156028368794328e-06,
+      "loss": 0.8089,
+      "step": 23
+    },
+    {
+      "epoch": 0.05139186295503212,
+      "grad_norm": 1.0965349661030956,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.8199,
+      "step": 24
+    },
+    {
+      "epoch": 0.05353319057815846,
+      "grad_norm": 0.779920509833393,
+      "learning_rate": 8.865248226950355e-06,
+      "loss": 0.8012,
+      "step": 25
+    },
+    {
+      "epoch": 0.055674518201284794,
+      "grad_norm": 0.7514108454773875,
+      "learning_rate": 9.219858156028368e-06,
+      "loss": 0.7586,
+      "step": 26
+    },
+    {
+      "epoch": 0.057815845824411134,
+      "grad_norm": 0.8086762191971834,
+      "learning_rate": 9.574468085106383e-06,
+      "loss": 0.7487,
+      "step": 27
+    },
+    {
+      "epoch": 0.059957173447537475,
+      "grad_norm": 0.8878919709734377,
+      "learning_rate": 9.929078014184398e-06,
+      "loss": 0.7034,
+      "step": 28
+    },
+    {
+      "epoch": 0.06209850107066381,
+      "grad_norm": 0.829679313012889,
+      "learning_rate": 1.0283687943262411e-05,
+      "loss": 0.7474,
+      "step": 29
+    },
+    {
+      "epoch": 0.06423982869379015,
+      "grad_norm": 0.6626872942809816,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.7349,
+      "step": 30
+    },
+    {
+      "epoch": 0.06638115631691649,
+      "grad_norm": 0.6753492576166917,
+      "learning_rate": 1.0992907801418441e-05,
+      "loss": 0.7171,
+      "step": 31
+    },
+    {
+      "epoch": 0.06852248394004283,
+      "grad_norm": 0.6704627449665963,
+      "learning_rate": 1.1347517730496454e-05,
+      "loss": 0.7126,
+      "step": 32
+    },
+    {
+      "epoch": 0.07066381156316917,
+      "grad_norm": 0.5562388268523989,
+      "learning_rate": 1.170212765957447e-05,
+      "loss": 0.6868,
+      "step": 33
+    },
+    {
+      "epoch": 0.0728051391862955,
+      "grad_norm": 0.6223987828107421,
+      "learning_rate": 1.2056737588652483e-05,
+      "loss": 0.6845,
+      "step": 34
+    },
+    {
+      "epoch": 0.07494646680942184,
+      "grad_norm": 0.5778135109852265,
+      "learning_rate": 1.2411347517730498e-05,
+      "loss": 0.6766,
+      "step": 35
+    },
+    {
+      "epoch": 0.07708779443254818,
+      "grad_norm": 0.4742088939687226,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 0.6522,
+      "step": 36
+    },
+    {
+      "epoch": 0.07922912205567452,
+      "grad_norm": 0.4554617236825081,
+      "learning_rate": 1.3120567375886524e-05,
+      "loss": 0.6551,
+      "step": 37
+    },
+    {
+      "epoch": 0.08137044967880086,
+      "grad_norm": 0.49432958372427405,
+      "learning_rate": 1.347517730496454e-05,
+      "loss": 0.6351,
+      "step": 38
+    },
+    {
+      "epoch": 0.0835117773019272,
+      "grad_norm": 0.440896644761752,
+      "learning_rate": 1.3829787234042554e-05,
+      "loss": 0.62,
+      "step": 39
+    },
+    {
+      "epoch": 0.08565310492505353,
+      "grad_norm": 0.3847657975966914,
+      "learning_rate": 1.418439716312057e-05,
+      "loss": 0.6449,
+      "step": 40
+    },
+    {
+      "epoch": 0.08779443254817987,
+      "grad_norm": 0.361109158517152,
+      "learning_rate": 1.4539007092198581e-05,
+      "loss": 0.6603,
+      "step": 41
+    },
+    {
+      "epoch": 0.08993576017130621,
+      "grad_norm": 0.35567536401833566,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 0.6005,
+      "step": 42
+    },
+    {
+      "epoch": 0.09207708779443255,
+      "grad_norm": 0.3656052483807157,
+      "learning_rate": 1.5248226950354611e-05,
+      "loss": 0.6126,
+      "step": 43
+    },
+    {
+      "epoch": 0.09421841541755889,
+      "grad_norm": 0.33705396501786156,
+      "learning_rate": 1.5602836879432626e-05,
+      "loss": 0.6066,
+      "step": 44
+    },
+    {
+      "epoch": 0.09635974304068523,
+      "grad_norm": 0.30303119503903936,
+      "learning_rate": 1.595744680851064e-05,
+      "loss": 0.6171,
+      "step": 45
+    },
+    {
+      "epoch": 0.09850107066381156,
+      "grad_norm": 0.2967613905439793,
+      "learning_rate": 1.6312056737588656e-05,
+      "loss": 0.608,
+      "step": 46
+    },
+    {
+      "epoch": 0.1006423982869379,
+      "grad_norm": 0.3767161988103767,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6171,
+      "step": 47
+    },
+    {
+      "epoch": 0.10278372591006424,
+      "grad_norm": 0.314876767137896,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 0.5786,
+      "step": 48
+    },
+    {
+      "epoch": 0.10492505353319058,
+      "grad_norm": 0.2767623351226321,
+      "learning_rate": 1.7375886524822697e-05,
+      "loss": 0.6053,
+      "step": 49
+    },
+    {
+      "epoch": 0.10706638115631692,
+      "grad_norm": 0.29425544708240703,
+      "learning_rate": 1.773049645390071e-05,
+      "loss": 0.5932,
+      "step": 50
+    },
+    {
+      "epoch": 0.10920770877944326,
+      "grad_norm": 0.2867315363666657,
+      "learning_rate": 1.8085106382978724e-05,
+      "loss": 0.5842,
+      "step": 51
+    },
+    {
+      "epoch": 0.11134903640256959,
+      "grad_norm": 0.2585858531022653,
+      "learning_rate": 1.8439716312056736e-05,
+      "loss": 0.5853,
+      "step": 52
+    },
+    {
+      "epoch": 0.11349036402569593,
+      "grad_norm": 0.2655264874976756,
+      "learning_rate": 1.879432624113475e-05,
+      "loss": 0.58,
+      "step": 53
+    },
+    {
+      "epoch": 0.11563169164882227,
+      "grad_norm": 0.30355375200055923,
+      "learning_rate": 1.9148936170212766e-05,
+      "loss": 0.5891,
+      "step": 54
+    },
+    {
+      "epoch": 0.11777301927194861,
+      "grad_norm": 0.30060619559855645,
+      "learning_rate": 1.950354609929078e-05,
+      "loss": 0.5657,
+      "step": 55
+    },
+    {
+      "epoch": 0.11991434689507495,
+      "grad_norm": 0.2544001308153972,
+      "learning_rate": 1.9858156028368796e-05,
+      "loss": 0.6008,
+      "step": 56
+    },
+    {
+      "epoch": 0.12205567451820129,
+      "grad_norm": 0.25996965782644,
+      "learning_rate": 2.0212765957446807e-05,
+      "loss": 0.5703,
+      "step": 57
+    },
+    {
+      "epoch": 0.12419700214132762,
+      "grad_norm": 0.2895469648849053,
+      "learning_rate": 2.0567375886524822e-05,
+      "loss": 0.5817,
+      "step": 58
+    },
+    {
+      "epoch": 0.12633832976445397,
+      "grad_norm": 0.25922509784815095,
+      "learning_rate": 2.0921985815602837e-05,
+      "loss": 0.5665,
+      "step": 59
+    },
+    {
+      "epoch": 0.1284796573875803,
+      "grad_norm": 0.26935959494463335,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 0.5634,
+      "step": 60
+    },
+    {
+      "epoch": 0.13062098501070663,
+      "grad_norm": 0.2472890335094329,
+      "learning_rate": 2.1631205673758867e-05,
+      "loss": 0.5489,
+      "step": 61
+    },
+    {
+      "epoch": 0.13276231263383298,
+      "grad_norm": 0.25220403853856466,
+      "learning_rate": 2.1985815602836882e-05,
+      "loss": 0.5555,
+      "step": 62
+    },
+    {
+      "epoch": 0.1349036402569593,
+      "grad_norm": 0.25227853159521985,
+      "learning_rate": 2.2340425531914894e-05,
+      "loss": 0.5977,
+      "step": 63
+    },
+    {
+      "epoch": 0.13704496788008566,
+      "grad_norm": 0.22662601243259486,
+      "learning_rate": 2.269503546099291e-05,
+      "loss": 0.5481,
+      "step": 64
+    },
+    {
+      "epoch": 0.139186295503212,
+      "grad_norm": 0.2462259567482875,
+      "learning_rate": 2.3049645390070924e-05,
+      "loss": 0.5628,
+      "step": 65
+    },
+    {
+      "epoch": 0.14132762312633834,
+      "grad_norm": 0.24332397161016522,
+      "learning_rate": 2.340425531914894e-05,
+      "loss": 0.5553,
+      "step": 66
+    },
+    {
+      "epoch": 0.14346895074946467,
+      "grad_norm": 0.23477707463573774,
+      "learning_rate": 2.3758865248226954e-05,
+      "loss": 0.5621,
+      "step": 67
+    },
+    {
+      "epoch": 0.145610278372591,
+      "grad_norm": 0.2500380042146856,
+      "learning_rate": 2.4113475177304965e-05,
+      "loss": 0.5661,
+      "step": 68
+    },
+    {
+      "epoch": 0.14775160599571735,
+      "grad_norm": 0.26380612632753875,
+      "learning_rate": 2.446808510638298e-05,
+      "loss": 0.5461,
+      "step": 69
+    },
+    {
+      "epoch": 0.14989293361884368,
+      "grad_norm": 0.2557903047322243,
+      "learning_rate": 2.4822695035460995e-05,
+      "loss": 0.5506,
+      "step": 70
+    },
+    {
+      "epoch": 0.15203426124197003,
+      "grad_norm": 0.25820215490499565,
+      "learning_rate": 2.5177304964539007e-05,
+      "loss": 0.5478,
+      "step": 71
+    },
+    {
+      "epoch": 0.15417558886509636,
+      "grad_norm": 0.24758155471819995,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 0.5518,
+      "step": 72
+    },
+    {
+      "epoch": 0.15631691648822268,
+      "grad_norm": 0.2415443149878719,
+      "learning_rate": 2.5886524822695034e-05,
+      "loss": 0.5505,
+      "step": 73
+    },
+    {
+      "epoch": 0.15845824411134904,
+      "grad_norm": 0.24823560199460706,
+      "learning_rate": 2.624113475177305e-05,
+      "loss": 0.5665,
+      "step": 74
+    },
+    {
+      "epoch": 0.16059957173447537,
+      "grad_norm": 0.23378147422826043,
+      "learning_rate": 2.6595744680851064e-05,
+      "loss": 0.527,
+      "step": 75
+    },
+    {
+      "epoch": 0.16274089935760172,
+      "grad_norm": 0.23375799268334566,
+      "learning_rate": 2.695035460992908e-05,
+      "loss": 0.5522,
+      "step": 76
+    },
+    {
+      "epoch": 0.16488222698072805,
+      "grad_norm": 0.27046995095800896,
+      "learning_rate": 2.7304964539007094e-05,
+      "loss": 0.5317,
+      "step": 77
+    },
+    {
+      "epoch": 0.1670235546038544,
+      "grad_norm": 0.22869648794405983,
+      "learning_rate": 2.765957446808511e-05,
+      "loss": 0.5488,
+      "step": 78
+    },
+    {
+      "epoch": 0.16916488222698073,
+      "grad_norm": 0.2144302125474379,
+      "learning_rate": 2.8014184397163124e-05,
+      "loss": 0.536,
+      "step": 79
+    },
+    {
+      "epoch": 0.17130620985010706,
+      "grad_norm": 0.2573958870247804,
+      "learning_rate": 2.836879432624114e-05,
+      "loss": 0.5261,
+      "step": 80
+    },
+    {
+      "epoch": 0.1734475374732334,
+      "grad_norm": 0.2613906182039433,
+      "learning_rate": 2.8723404255319154e-05,
+      "loss": 0.5418,
+      "step": 81
+    },
+    {
+      "epoch": 0.17558886509635974,
+      "grad_norm": 0.2314728596697455,
+      "learning_rate": 2.9078014184397162e-05,
+      "loss": 0.5258,
+      "step": 82
+    },
+    {
+      "epoch": 0.1777301927194861,
+      "grad_norm": 0.255469146374909,
+      "learning_rate": 2.9432624113475177e-05,
+      "loss": 0.5358,
+      "step": 83
+    },
+    {
+      "epoch": 0.17987152034261242,
+      "grad_norm": 0.269758850025105,
+      "learning_rate": 2.9787234042553192e-05,
+      "loss": 0.5467,
+      "step": 84
+    },
+    {
+      "epoch": 0.18201284796573874,
+      "grad_norm": 0.22108427720048224,
+      "learning_rate": 3.0141843971631207e-05,
+      "loss": 0.5138,
+      "step": 85
+    },
+    {
+      "epoch": 0.1841541755888651,
+      "grad_norm": 0.2918049155831516,
+      "learning_rate": 3.0496453900709222e-05,
+      "loss": 0.5358,
+      "step": 86
+    },
+    {
+      "epoch": 0.18629550321199143,
+      "grad_norm": 0.252878303302326,
+      "learning_rate": 3.085106382978723e-05,
+      "loss": 0.5164,
+      "step": 87
+    },
+    {
+      "epoch": 0.18843683083511778,
+      "grad_norm": 0.28867004146757574,
+      "learning_rate": 3.120567375886525e-05,
+      "loss": 0.54,
+      "step": 88
+    },
+    {
+      "epoch": 0.1905781584582441,
+      "grad_norm": 0.30841177860368885,
+      "learning_rate": 3.156028368794326e-05,
+      "loss": 0.5271,
+      "step": 89
+    },
+    {
+      "epoch": 0.19271948608137046,
+      "grad_norm": 0.29805101294833053,
+      "learning_rate": 3.191489361702128e-05,
+      "loss": 0.5186,
+      "step": 90
+    },
+    {
+      "epoch": 0.1948608137044968,
+      "grad_norm": 0.29287452022732885,
+      "learning_rate": 3.226950354609929e-05,
+      "loss": 0.5288,
+      "step": 91
+    },
+    {
+      "epoch": 0.19700214132762311,
+      "grad_norm": 0.2892573734733911,
+      "learning_rate": 3.262411347517731e-05,
+      "loss": 0.5325,
+      "step": 92
+    },
+    {
+      "epoch": 0.19914346895074947,
+      "grad_norm": 0.241170324930527,
+      "learning_rate": 3.2978723404255317e-05,
+      "loss": 0.5069,
+      "step": 93
+    },
+    {
+      "epoch": 0.2012847965738758,
+      "grad_norm": 0.28618356988432964,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.528,
+      "step": 94
+    },
+    {
+      "epoch": 0.20342612419700215,
+      "grad_norm": 0.2706127494448774,
+      "learning_rate": 3.3687943262411347e-05,
+      "loss": 0.5137,
+      "step": 95
+    },
+    {
+      "epoch": 0.20556745182012848,
+      "grad_norm": 0.2920030135213235,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 0.5279,
+      "step": 96
+    },
+    {
+      "epoch": 0.20770877944325483,
+      "grad_norm": 0.2481402426704902,
+      "learning_rate": 3.4397163120567377e-05,
+      "loss": 0.5001,
+      "step": 97
+    },
+    {
+      "epoch": 0.20985010706638116,
+      "grad_norm": 0.27424390116478947,
+      "learning_rate": 3.4751773049645395e-05,
+      "loss": 0.4932,
+      "step": 98
+    },
+    {
+      "epoch": 0.21199143468950749,
+      "grad_norm": 0.29648670562970614,
+      "learning_rate": 3.5106382978723407e-05,
+      "loss": 0.529,
+      "step": 99
+    },
+    {
+      "epoch": 0.21413276231263384,
+      "grad_norm": 0.2457515803836918,
+      "learning_rate": 3.546099290780142e-05,
+      "loss": 0.5178,
+      "step": 100
+    },
+    {
+      "epoch": 0.21627408993576017,
+      "grad_norm": 0.29926928267286307,
+      "learning_rate": 3.5815602836879437e-05,
+      "loss": 0.5224,
+      "step": 101
+    },
+    {
+      "epoch": 0.21841541755888652,
+      "grad_norm": 0.24321396233390155,
+      "learning_rate": 3.617021276595745e-05,
+      "loss": 0.5296,
+      "step": 102
+    },
+    {
+      "epoch": 0.22055674518201285,
+      "grad_norm": 0.30921888959439514,
+      "learning_rate": 3.6524822695035466e-05,
+      "loss": 0.5179,
+      "step": 103
+    },
+    {
+      "epoch": 0.22269807280513917,
+      "grad_norm": 0.272063115950373,
+      "learning_rate": 3.687943262411347e-05,
+      "loss": 0.5283,
+      "step": 104
+    },
+    {
+      "epoch": 0.22483940042826553,
+      "grad_norm": 0.26613186863197535,
+      "learning_rate": 3.723404255319149e-05,
+      "loss": 0.5242,
+      "step": 105
+    },
+    {
+      "epoch": 0.22698072805139186,
+      "grad_norm": 0.33500951074808644,
+      "learning_rate": 3.75886524822695e-05,
+      "loss": 0.5174,
+      "step": 106
+    },
+    {
+      "epoch": 0.2291220556745182,
+      "grad_norm": 0.2643772639229429,
+      "learning_rate": 3.794326241134752e-05,
+      "loss": 0.5343,
+      "step": 107
+    },
+    {
+      "epoch": 0.23126338329764454,
+      "grad_norm": 0.31115272338829114,
+      "learning_rate": 3.829787234042553e-05,
+      "loss": 0.5052,
+      "step": 108
+    },
+    {
+      "epoch": 0.2334047109207709,
+      "grad_norm": 0.2926558130416796,
+      "learning_rate": 3.865248226950355e-05,
+      "loss": 0.5137,
+      "step": 109
+    },
+    {
+      "epoch": 0.23554603854389722,
+      "grad_norm": 0.3255536325363357,
+      "learning_rate": 3.900709219858156e-05,
+      "loss": 0.5279,
+      "step": 110
+    },
+    {
+      "epoch": 0.23768736616702354,
+      "grad_norm": 0.33989084909132006,
+      "learning_rate": 3.936170212765958e-05,
+      "loss": 0.4988,
+      "step": 111
+    },
+    {
+      "epoch": 0.2398286937901499,
+      "grad_norm": 0.35381236194602544,
+      "learning_rate": 3.971631205673759e-05,
+      "loss": 0.4937,
+      "step": 112
+    },
+    {
+      "epoch": 0.24197002141327623,
+      "grad_norm": 0.3510604214761894,
+      "learning_rate": 4.007092198581561e-05,
+      "loss": 0.5178,
+      "step": 113
+    },
+    {
+      "epoch": 0.24411134903640258,
+      "grad_norm": 0.2974753747795791,
+      "learning_rate": 4.0425531914893614e-05,
+      "loss": 0.5006,
+      "step": 114
+    },
+    {
+      "epoch": 0.2462526766595289,
+      "grad_norm": 0.3391358705953081,
+      "learning_rate": 4.078014184397163e-05,
+      "loss": 0.4991,
+      "step": 115
+    },
+    {
+      "epoch": 0.24839400428265523,
+      "grad_norm": 0.3244454146866931,
+      "learning_rate": 4.1134751773049644e-05,
+      "loss": 0.4992,
+      "step": 116
+    },
+    {
+      "epoch": 0.2505353319057816,
+      "grad_norm": 0.29806045444607565,
+      "learning_rate": 4.148936170212766e-05,
+      "loss": 0.4879,
+      "step": 117
+    },
+    {
+      "epoch": 0.25267665952890794,
+      "grad_norm": 0.34001472132830457,
+      "learning_rate": 4.1843971631205674e-05,
+      "loss": 0.5142,
+      "step": 118
+    },
+    {
+      "epoch": 0.25481798715203424,
+      "grad_norm": 0.3147825744118216,
+      "learning_rate": 4.219858156028369e-05,
+      "loss": 0.4975,
+      "step": 119
+    },
+    {
+      "epoch": 0.2569593147751606,
+      "grad_norm": 0.343655564131983,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 0.5169,
+      "step": 120
+    },
+    {
+      "epoch": 0.25910064239828695,
+      "grad_norm": 0.3165564039834663,
+      "learning_rate": 4.2907801418439716e-05,
+      "loss": 0.5038,
+      "step": 121
+    },
+    {
+      "epoch": 0.26124197002141325,
+      "grad_norm": 0.3421038891467151,
+      "learning_rate": 4.3262411347517734e-05,
+      "loss": 0.5044,
+      "step": 122
+    },
+    {
+      "epoch": 0.2633832976445396,
+      "grad_norm": 0.3236382619911156,
+      "learning_rate": 4.3617021276595746e-05,
+      "loss": 0.5077,
+      "step": 123
+    },
+    {
+      "epoch": 0.26552462526766596,
+      "grad_norm": 0.3482419500485224,
+      "learning_rate": 4.3971631205673764e-05,
+      "loss": 0.4822,
+      "step": 124
+    },
+    {
+      "epoch": 0.2676659528907923,
+      "grad_norm": 0.3967980846994073,
+      "learning_rate": 4.432624113475177e-05,
+      "loss": 0.4947,
+      "step": 125
+    },
+    {
+      "epoch": 0.2698072805139186,
+      "grad_norm": 0.29467264961379613,
+      "learning_rate": 4.468085106382979e-05,
+      "loss": 0.4871,
+      "step": 126
+    },
+    {
+      "epoch": 0.27194860813704497,
+      "grad_norm": 0.35596207729536755,
+      "learning_rate": 4.50354609929078e-05,
+      "loss": 0.4831,
+      "step": 127
+    },
+    {
+      "epoch": 0.2740899357601713,
+      "grad_norm": 0.37395057401049897,
+      "learning_rate": 4.539007092198582e-05,
+      "loss": 0.5122,
+      "step": 128
+    },
+    {
+      "epoch": 0.2762312633832976,
+      "grad_norm": 0.28605544378420467,
+      "learning_rate": 4.574468085106383e-05,
+      "loss": 0.4898,
+      "step": 129
+    },
+    {
+      "epoch": 0.278372591006424,
+      "grad_norm": 0.3331655327731713,
+      "learning_rate": 4.609929078014185e-05,
+      "loss": 0.4912,
+      "step": 130
+    },
+    {
+      "epoch": 0.28051391862955033,
+      "grad_norm": 0.29374454897662605,
+      "learning_rate": 4.645390070921986e-05,
+      "loss": 0.4814,
+      "step": 131
+    },
+    {
+      "epoch": 0.2826552462526767,
+      "grad_norm": 0.30113962440816366,
+      "learning_rate": 4.680851063829788e-05,
+      "loss": 0.4856,
+      "step": 132
+    },
+    {
+      "epoch": 0.284796573875803,
+      "grad_norm": 0.30027399950187866,
+      "learning_rate": 4.716312056737589e-05,
+      "loss": 0.4852,
+      "step": 133
+    },
+    {
+      "epoch": 0.28693790149892934,
+      "grad_norm": 0.3230425204774485,
+      "learning_rate": 4.751773049645391e-05,
+      "loss": 0.4903,
+      "step": 134
+    },
+    {
+      "epoch": 0.2890792291220557,
+      "grad_norm": 0.31234670755297617,
+      "learning_rate": 4.787234042553192e-05,
+      "loss": 0.5041,
+      "step": 135
+    },
+    {
+      "epoch": 0.291220556745182,
+      "grad_norm": 0.30804022118995605,
+      "learning_rate": 4.822695035460993e-05,
+      "loss": 0.4818,
+      "step": 136
+    },
+    {
+      "epoch": 0.29336188436830835,
+      "grad_norm": 0.3524834300849378,
+      "learning_rate": 4.858156028368794e-05,
+      "loss": 0.4935,
+      "step": 137
+    },
+    {
+      "epoch": 0.2955032119914347,
+      "grad_norm": 0.3743768713148858,
+      "learning_rate": 4.893617021276596e-05,
+      "loss": 0.5098,
+      "step": 138
+    },
+    {
+      "epoch": 0.29764453961456105,
+      "grad_norm": 0.34028691668800937,
+      "learning_rate": 4.929078014184397e-05,
+      "loss": 0.4935,
+      "step": 139
+    },
+    {
+      "epoch": 0.29978586723768735,
+      "grad_norm": 0.3432662688866466,
+      "learning_rate": 4.964539007092199e-05,
+      "loss": 0.4906,
+      "step": 140
+    },
+    {
+      "epoch": 0.3019271948608137,
+      "grad_norm": 0.3484159957729913,
+      "learning_rate": 5e-05,
+      "loss": 0.4916,
+      "step": 141
+    },
+    {
+      "epoch": 0.30406852248394006,
+      "grad_norm": 0.2906011601786296,
+      "learning_rate": 4.996031746031746e-05,
+      "loss": 0.4851,
+      "step": 142
+    },
+    {
+      "epoch": 0.30620985010706636,
+      "grad_norm": 0.31602752642817156,
+      "learning_rate": 4.9920634920634924e-05,
+      "loss": 0.4804,
+      "step": 143
+    },
+    {
+      "epoch": 0.3083511777301927,
+      "grad_norm": 0.33931054385910697,
+      "learning_rate": 4.9880952380952385e-05,
+      "loss": 0.4956,
+      "step": 144
+    },
+    {
+      "epoch": 0.31049250535331907,
+      "grad_norm": 0.2673250220712651,
+      "learning_rate": 4.9841269841269845e-05,
+      "loss": 0.474,
+      "step": 145
+    },
+    {
+      "epoch": 0.31263383297644537,
+      "grad_norm": 0.27590244134076686,
+      "learning_rate": 4.9801587301587306e-05,
+      "loss": 0.4652,
+      "step": 146
+    },
+    {
+      "epoch": 0.3147751605995717,
+      "grad_norm": 0.29339636329684166,
+      "learning_rate": 4.976190476190477e-05,
+      "loss": 0.4804,
+      "step": 147
+    },
+    {
+      "epoch": 0.3169164882226981,
+      "grad_norm": 0.2842194678371775,
+      "learning_rate": 4.972222222222223e-05,
+      "loss": 0.4884,
+      "step": 148
+    },
+    {
+      "epoch": 0.31905781584582443,
+      "grad_norm": 0.3177841927362518,
+      "learning_rate": 4.968253968253969e-05,
+      "loss": 0.4887,
+      "step": 149
+    },
+    {
+      "epoch": 0.32119914346895073,
+      "grad_norm": 0.362994905525864,
+      "learning_rate": 4.964285714285715e-05,
+      "loss": 0.4957,
+      "step": 150
+    },
+    {
+      "epoch": 0.3233404710920771,
+      "grad_norm": 0.3134548549120884,
+      "learning_rate": 4.960317460317461e-05,
+      "loss": 0.4746,
+      "step": 151
+    },
+    {
+      "epoch": 0.32548179871520344,
+      "grad_norm": 0.33022141318925424,
+      "learning_rate": 4.956349206349207e-05,
+      "loss": 0.4836,
+      "step": 152
+    },
+    {
+      "epoch": 0.32762312633832974,
+      "grad_norm": 0.2841374060809414,
+      "learning_rate": 4.9523809523809525e-05,
+      "loss": 0.4824,
+      "step": 153
+    },
+    {
+      "epoch": 0.3297644539614561,
+      "grad_norm": 0.3042451933872161,
+      "learning_rate": 4.9484126984126985e-05,
+      "loss": 0.4947,
+      "step": 154
+    },
+    {
+      "epoch": 0.33190578158458245,
+      "grad_norm": 0.3081245822778917,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 0.4882,
+      "step": 155
+    },
+    {
+      "epoch": 0.3340471092077088,
+      "grad_norm": 0.2888403289180772,
+      "learning_rate": 4.940476190476191e-05,
+      "loss": 0.4583,
+      "step": 156
+    },
+    {
+      "epoch": 0.3361884368308351,
+      "grad_norm": 0.49151318304484476,
+      "learning_rate": 4.936507936507937e-05,
+      "loss": 0.5023,
+      "step": 157
+    },
+    {
+      "epoch": 0.33832976445396146,
+      "grad_norm": 0.33926947413240066,
+      "learning_rate": 4.932539682539683e-05,
+      "loss": 0.5,
+      "step": 158
+    },
+    {
+      "epoch": 0.3404710920770878,
+      "grad_norm": 0.3432037588535958,
+      "learning_rate": 4.928571428571429e-05,
+      "loss": 0.4892,
+      "step": 159
+    },
+    {
+      "epoch": 0.3426124197002141,
+      "grad_norm": 0.33357704996030413,
+      "learning_rate": 4.924603174603175e-05,
+      "loss": 0.483,
+      "step": 160
+    },
+    {
+      "epoch": 0.34475374732334046,
+      "grad_norm": 0.3455404282494954,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 0.502,
+      "step": 161
+    },
+    {
+      "epoch": 0.3468950749464668,
+      "grad_norm": 0.3799400968283991,
+      "learning_rate": 4.9166666666666665e-05,
+      "loss": 0.5005,
+      "step": 162
+    },
+    {
+      "epoch": 0.3490364025695932,
+      "grad_norm": 0.28633882509130054,
+      "learning_rate": 4.9126984126984125e-05,
+      "loss": 0.5012,
+      "step": 163
+    },
+    {
+      "epoch": 0.3511777301927195,
+      "grad_norm": 0.3692568848737784,
+      "learning_rate": 4.9087301587301586e-05,
+      "loss": 0.4773,
+      "step": 164
+    },
+    {
+      "epoch": 0.3533190578158458,
+      "grad_norm": 0.28424265463523657,
+      "learning_rate": 4.904761904761905e-05,
+      "loss": 0.4933,
+      "step": 165
+    },
+    {
+      "epoch": 0.3554603854389722,
+      "grad_norm": 0.3433785169927177,
+      "learning_rate": 4.900793650793651e-05,
+      "loss": 0.4708,
+      "step": 166
+    },
+    {
+      "epoch": 0.3576017130620985,
+      "grad_norm": 0.322472121983474,
+      "learning_rate": 4.896825396825397e-05,
+      "loss": 0.484,
+      "step": 167
+    },
+    {
+      "epoch": 0.35974304068522484,
+      "grad_norm": 0.37196647339980105,
+      "learning_rate": 4.892857142857143e-05,
+      "loss": 0.4872,
+      "step": 168
+    },
+    {
+      "epoch": 0.3618843683083512,
+      "grad_norm": 0.3144695427706762,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.4909,
+      "step": 169
+    },
+    {
+      "epoch": 0.3640256959314775,
+      "grad_norm": 0.3399481737215869,
+      "learning_rate": 4.884920634920635e-05,
+      "loss": 0.4851,
+      "step": 170
+    },
+    {
+      "epoch": 0.36616702355460384,
+      "grad_norm": 0.30151182574451896,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 0.4772,
+      "step": 171
+    },
+    {
+      "epoch": 0.3683083511777302,
+      "grad_norm": 0.3128577892502514,
+      "learning_rate": 4.876984126984127e-05,
+      "loss": 0.4848,
+      "step": 172
+    },
+    {
+      "epoch": 0.37044967880085655,
+      "grad_norm": 0.31922897506045883,
+      "learning_rate": 4.873015873015873e-05,
+      "loss": 0.4986,
+      "step": 173
+    },
+    {
+      "epoch": 0.37259100642398285,
+      "grad_norm": 0.3017606389198961,
+      "learning_rate": 4.8690476190476194e-05,
+      "loss": 0.4828,
+      "step": 174
+    },
+    {
+      "epoch": 0.3747323340471092,
+      "grad_norm": 0.3000491473588088,
+      "learning_rate": 4.8650793650793654e-05,
+      "loss": 0.4752,
+      "step": 175
+    },
+    {
+      "epoch": 0.37687366167023556,
+      "grad_norm": 0.3892869618739201,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 0.4892,
+      "step": 176
+    },
+    {
+      "epoch": 0.37901498929336186,
+      "grad_norm": 0.29614553607421257,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.5018,
+      "step": 177
+    },
+    {
+      "epoch": 0.3811563169164882,
+      "grad_norm": 0.3790956924695203,
+      "learning_rate": 4.853174603174604e-05,
+      "loss": 0.5092,
+      "step": 178
+    },
+    {
+      "epoch": 0.38329764453961457,
+      "grad_norm": 0.3316734586371452,
+      "learning_rate": 4.84920634920635e-05,
+      "loss": 0.482,
+      "step": 179
+    },
+    {
+      "epoch": 0.3854389721627409,
+      "grad_norm": 0.3404880131426743,
+      "learning_rate": 4.845238095238095e-05,
+      "loss": 0.4981,
+      "step": 180
+    },
+    {
+      "epoch": 0.3875802997858672,
+      "grad_norm": 0.2900095290587479,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.4668,
+      "step": 181
+    },
+    {
+      "epoch": 0.3897216274089936,
+      "grad_norm": 0.2922900867382386,
+      "learning_rate": 4.837301587301587e-05,
+      "loss": 0.4667,
+      "step": 182
+    },
+    {
+      "epoch": 0.39186295503211993,
+      "grad_norm": 0.3181046448048784,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.4804,
+      "step": 183
+    },
+    {
+      "epoch": 0.39400428265524623,
+      "grad_norm": 0.27687702877774883,
+      "learning_rate": 4.8293650793650794e-05,
+      "loss": 0.4775,
+      "step": 184
+    },
+    {
+      "epoch": 0.3961456102783726,
+      "grad_norm": 0.2836321426330953,
+      "learning_rate": 4.8253968253968255e-05,
+      "loss": 0.4725,
+      "step": 185
+    },
+    {
+      "epoch": 0.39828693790149894,
+      "grad_norm": 0.2913866873418313,
+      "learning_rate": 4.8214285714285716e-05,
+      "loss": 0.4775,
+      "step": 186
+    },
+    {
+      "epoch": 0.4004282655246253,
+      "grad_norm": 0.30835624032708503,
+      "learning_rate": 4.817460317460318e-05,
+      "loss": 0.4878,
+      "step": 187
+    },
+    {
+      "epoch": 0.4025695931477516,
+      "grad_norm": 0.29588677700901966,
+      "learning_rate": 4.813492063492064e-05,
+      "loss": 0.4732,
+      "step": 188
+    },
+    {
+      "epoch": 0.40471092077087795,
+      "grad_norm": 0.30881963199698537,
+      "learning_rate": 4.80952380952381e-05,
+      "loss": 0.4598,
+      "step": 189
+    },
+    {
+      "epoch": 0.4068522483940043,
+      "grad_norm": 0.30298345716545283,
+      "learning_rate": 4.805555555555556e-05,
+      "loss": 0.4717,
+      "step": 190
+    },
+    {
+      "epoch": 0.4089935760171306,
+      "grad_norm": 0.35885982262554333,
+      "learning_rate": 4.801587301587302e-05,
+      "loss": 0.4752,
+      "step": 191
+    },
+    {
+      "epoch": 0.41113490364025695,
+      "grad_norm": 0.3004713627687736,
+      "learning_rate": 4.797619047619048e-05,
+      "loss": 0.4674,
+      "step": 192
+    },
+    {
+      "epoch": 0.4132762312633833,
+      "grad_norm": 0.29124273016906205,
+      "learning_rate": 4.793650793650794e-05,
+      "loss": 0.4883,
+      "step": 193
+    },
+    {
+      "epoch": 0.41541755888650966,
+      "grad_norm": 0.28170083463492585,
+      "learning_rate": 4.78968253968254e-05,
+      "loss": 0.4808,
+      "step": 194
+    },
+    {
+      "epoch": 0.41755888650963596,
+      "grad_norm": 0.2930248614401525,
+      "learning_rate": 4.785714285714286e-05,
+      "loss": 0.4614,
+      "step": 195
+    },
+    {
+      "epoch": 0.4197002141327623,
+      "grad_norm": 0.3725197913239573,
+      "learning_rate": 4.781746031746032e-05,
+      "loss": 0.466,
+      "step": 196
+    },
+    {
+      "epoch": 0.42184154175588867,
+      "grad_norm": 0.3173770734521219,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.4839,
+      "step": 197
+    },
+    {
+      "epoch": 0.42398286937901497,
+      "grad_norm": 0.31792822631775125,
+      "learning_rate": 4.7738095238095245e-05,
+      "loss": 0.4604,
+      "step": 198
+    },
+    {
+      "epoch": 0.4261241970021413,
+      "grad_norm": 0.4547994524865786,
+      "learning_rate": 4.7698412698412706e-05,
+      "loss": 0.458,
+      "step": 199
+    },
+    {
+      "epoch": 0.4282655246252677,
+      "grad_norm": 0.3200904129332156,
+      "learning_rate": 4.7658730158730166e-05,
+      "loss": 0.4722,
+      "step": 200
+    },
+    {
+      "epoch": 0.430406852248394,
+      "grad_norm": 0.33843214154215645,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.4841,
+      "step": 201
+    },
+    {
+      "epoch": 0.43254817987152033,
+      "grad_norm": 0.29671855926132695,
+      "learning_rate": 4.757936507936508e-05,
+      "loss": 0.4896,
+      "step": 202
+    },
+    {
+      "epoch": 0.4346895074946467,
+      "grad_norm": 0.31070750776887823,
+      "learning_rate": 4.753968253968254e-05,
+      "loss": 0.4714,
+      "step": 203
+    },
+    {
+      "epoch": 0.43683083511777304,
+      "grad_norm": 0.3305729189717249,
+      "learning_rate": 4.75e-05,
+      "loss": 0.4553,
+      "step": 204
+    },
+    {
+      "epoch": 0.43897216274089934,
+      "grad_norm": 0.30161254779141056,
+      "learning_rate": 4.746031746031746e-05,
+      "loss": 0.4771,
+      "step": 205
+    },
+    {
+      "epoch": 0.4411134903640257,
+      "grad_norm": 0.4165722704133656,
+      "learning_rate": 4.7420634920634924e-05,
+      "loss": 0.4941,
+      "step": 206
+    },
+    {
+      "epoch": 0.44325481798715205,
+      "grad_norm": 0.3098035646663218,
+      "learning_rate": 4.738095238095238e-05,
+      "loss": 0.4562,
+      "step": 207
+    },
+    {
+      "epoch": 0.44539614561027835,
+      "grad_norm": 0.31217832469780865,
+      "learning_rate": 4.734126984126984e-05,
+      "loss": 0.4593,
+      "step": 208
+    },
+    {
+      "epoch": 0.4475374732334047,
+      "grad_norm": 0.33303037422629134,
+      "learning_rate": 4.73015873015873e-05,
+      "loss": 0.4695,
+      "step": 209
+    },
+    {
+      "epoch": 0.44967880085653106,
+      "grad_norm": 0.3164429971126553,
+      "learning_rate": 4.726190476190476e-05,
+      "loss": 0.4787,
+      "step": 210
+    },
+    {
+      "epoch": 0.4518201284796574,
+      "grad_norm": 0.3667199010122267,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.4741,
+      "step": 211
+    },
+    {
+      "epoch": 0.4539614561027837,
+      "grad_norm": 0.32958761965134165,
+      "learning_rate": 4.718253968253968e-05,
+      "loss": 0.476,
+      "step": 212
+    },
+    {
+      "epoch": 0.45610278372591007,
+      "grad_norm": 0.3847263613923873,
+      "learning_rate": 4.714285714285714e-05,
+      "loss": 0.4699,
+      "step": 213
+    },
+    {
+      "epoch": 0.4582441113490364,
+      "grad_norm": 0.2955982753193451,
+      "learning_rate": 4.71031746031746e-05,
+      "loss": 0.4799,
+      "step": 214
+    },
+    {
+      "epoch": 0.4603854389721627,
+      "grad_norm": 0.38809540646072666,
+      "learning_rate": 4.7063492063492064e-05,
+      "loss": 0.4783,
+      "step": 215
+    },
+    {
+      "epoch": 0.4625267665952891,
+      "grad_norm": 0.33351402005149655,
+      "learning_rate": 4.7023809523809525e-05,
+      "loss": 0.4671,
+      "step": 216
+    },
+    {
+      "epoch": 0.46466809421841543,
+      "grad_norm": 0.3544022577957815,
+      "learning_rate": 4.6984126984126986e-05,
+      "loss": 0.4634,
+      "step": 217
+    },
+    {
+      "epoch": 0.4668094218415418,
+      "grad_norm": 0.3440682763213123,
+      "learning_rate": 4.6944444444444446e-05,
+      "loss": 0.4606,
+      "step": 218
+    },
+    {
+      "epoch": 0.4689507494646681,
+      "grad_norm": 0.35994592001280135,
+      "learning_rate": 4.690476190476191e-05,
+      "loss": 0.4687,
+      "step": 219
+    },
+    {
+      "epoch": 0.47109207708779444,
+      "grad_norm": 0.2792969038742562,
+      "learning_rate": 4.686507936507937e-05,
+      "loss": 0.4675,
+      "step": 220
+    },
+    {
+      "epoch": 0.4732334047109208,
+      "grad_norm": 0.3167594669644963,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.447,
+      "step": 221
+    },
+    {
+      "epoch": 0.4753747323340471,
+      "grad_norm": 0.28698974715014,
+      "learning_rate": 4.678571428571429e-05,
+      "loss": 0.4824,
+      "step": 222
+    },
+    {
+      "epoch": 0.47751605995717344,
+      "grad_norm": 0.2919731601075355,
+      "learning_rate": 4.674603174603175e-05,
+      "loss": 0.4756,
+      "step": 223
+    },
+    {
+      "epoch": 0.4796573875802998,
+      "grad_norm": 0.3961977468030219,
+      "learning_rate": 4.670634920634921e-05,
+      "loss": 0.4988,
+      "step": 224
+    },
+    {
+      "epoch": 0.4817987152034261,
+      "grad_norm": 0.2939889127181417,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.4782,
+      "step": 225
+    },
+    {
+      "epoch": 0.48394004282655245,
+      "grad_norm": 0.38374224244801103,
+      "learning_rate": 4.662698412698413e-05,
+      "loss": 0.4713,
+      "step": 226
+    },
+    {
+      "epoch": 0.4860813704496788,
+      "grad_norm": 0.2783898692581709,
+      "learning_rate": 4.658730158730159e-05,
+      "loss": 0.4683,
+      "step": 227
+    },
+    {
+      "epoch": 0.48822269807280516,
+      "grad_norm": 0.3514854051266336,
+      "learning_rate": 4.6547619047619054e-05,
+      "loss": 0.4745,
+      "step": 228
+    },
+    {
+      "epoch": 0.49036402569593146,
+      "grad_norm": 0.2761349026606947,
+      "learning_rate": 4.6507936507936515e-05,
+      "loss": 0.4777,
+      "step": 229
+    },
+    {
+      "epoch": 0.4925053533190578,
+      "grad_norm": 0.2826076093838578,
+      "learning_rate": 4.646825396825397e-05,
+      "loss": 0.4491,
+      "step": 230
+    },
+    {
+      "epoch": 0.49464668094218417,
+      "grad_norm": 0.2945939088071528,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.4724,
+      "step": 231
+    },
+    {
+      "epoch": 0.49678800856531047,
+      "grad_norm": 0.2750719469025186,
+      "learning_rate": 4.638888888888889e-05,
+      "loss": 0.4592,
+      "step": 232
+    },
+    {
+      "epoch": 0.4989293361884368,
+      "grad_norm": 0.2515310029544384,
+      "learning_rate": 4.634920634920635e-05,
+      "loss": 0.4572,
+      "step": 233
+    },
+    {
+      "epoch": 0.5010706638115632,
+      "grad_norm": 0.29715312851565084,
+      "learning_rate": 4.630952380952381e-05,
+      "loss": 0.4769,
+      "step": 234
+    },
+    {
+      "epoch": 0.5032119914346895,
+      "grad_norm": 0.28840327957350287,
+      "learning_rate": 4.626984126984127e-05,
+      "loss": 0.459,
+      "step": 235
+    },
+    {
+      "epoch": 0.5053533190578159,
+      "grad_norm": 0.38388536035323373,
+      "learning_rate": 4.623015873015873e-05,
+      "loss": 0.4645,
+      "step": 236
+    },
+    {
+      "epoch": 0.5074946466809421,
+      "grad_norm": 0.34005158978288785,
+      "learning_rate": 4.6190476190476194e-05,
+      "loss": 0.4951,
+      "step": 237
+    },
+    {
+      "epoch": 0.5096359743040685,
+      "grad_norm": 0.3135930281832256,
+      "learning_rate": 4.6150793650793655e-05,
+      "loss": 0.4783,
+      "step": 238
+    },
+    {
+      "epoch": 0.5117773019271948,
+      "grad_norm": 0.29153181091541064,
+      "learning_rate": 4.6111111111111115e-05,
+      "loss": 0.4783,
+      "step": 239
+    },
+    {
+      "epoch": 0.5139186295503212,
+      "grad_norm": 0.35211081607111105,
+      "learning_rate": 4.607142857142857e-05,
+      "loss": 0.4784,
+      "step": 240
+    },
+    {
+      "epoch": 0.5160599571734475,
+      "grad_norm": 0.2534327826926188,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.4758,
+      "step": 241
+    },
+    {
+      "epoch": 0.5182012847965739,
+      "grad_norm": 0.291715580605378,
+      "learning_rate": 4.599206349206349e-05,
+      "loss": 0.4613,
+      "step": 242
+    },
+    {
+      "epoch": 0.5203426124197003,
+      "grad_norm": 0.49953531179115856,
+      "learning_rate": 4.595238095238095e-05,
+      "loss": 0.467,
+      "step": 243
+    },
+    {
+      "epoch": 0.5224839400428265,
+      "grad_norm": 0.37988476101063506,
+      "learning_rate": 4.591269841269841e-05,
+      "loss": 0.4668,
+      "step": 244
+    },
+    {
+      "epoch": 0.5246252676659529,
+      "grad_norm": 0.3306624298526701,
+      "learning_rate": 4.587301587301587e-05,
+      "loss": 0.4571,
+      "step": 245
+    },
+    {
+      "epoch": 0.5267665952890792,
+      "grad_norm": 0.292259379733011,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.4747,
+      "step": 246
+    },
+    {
+      "epoch": 0.5289079229122056,
+      "grad_norm": 0.2817284251841534,
+      "learning_rate": 4.5793650793650795e-05,
+      "loss": 0.4601,
+      "step": 247
+    },
+    {
+      "epoch": 0.5310492505353319,
+      "grad_norm": 0.28259998989853713,
+      "learning_rate": 4.5753968253968255e-05,
+      "loss": 0.4489,
+      "step": 248
+    },
+    {
+      "epoch": 0.5331905781584583,
+      "grad_norm": 0.3077230211086655,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.4678,
+      "step": 249
+    },
+    {
+      "epoch": 0.5353319057815846,
+      "grad_norm": 0.2783904187217292,
+      "learning_rate": 4.567460317460318e-05,
+      "loss": 0.4401,
+      "step": 250
+    },
+    {
+      "epoch": 0.5374732334047109,
+      "grad_norm": 0.31896115951211806,
+      "learning_rate": 4.563492063492064e-05,
+      "loss": 0.4572,
+      "step": 251
+    },
+    {
+      "epoch": 0.5396145610278372,
+      "grad_norm": 0.30499818403576123,
+      "learning_rate": 4.55952380952381e-05,
+      "loss": 0.465,
+      "step": 252
+    },
+    {
+      "epoch": 0.5417558886509636,
+      "grad_norm": 0.35575649581615343,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.47,
+      "step": 253
+    },
+    {
+      "epoch": 0.5438972162740899,
+      "grad_norm": 0.2684090185728518,
+      "learning_rate": 4.551587301587302e-05,
+      "loss": 0.4645,
+      "step": 254
+    },
+    {
+      "epoch": 0.5460385438972163,
+      "grad_norm": 0.37921813849517194,
+      "learning_rate": 4.547619047619048e-05,
+      "loss": 0.4678,
+      "step": 255
+    },
+    {
+      "epoch": 0.5481798715203426,
+      "grad_norm": 0.26743329008720806,
+      "learning_rate": 4.543650793650794e-05,
+      "loss": 0.4732,
+      "step": 256
+    },
+    {
+      "epoch": 0.550321199143469,
+      "grad_norm": 0.31026028532831346,
+      "learning_rate": 4.5396825396825395e-05,
+      "loss": 0.4508,
+      "step": 257
+    },
+    {
+      "epoch": 0.5524625267665952,
+      "grad_norm": 0.2581029656532225,
+      "learning_rate": 4.5357142857142856e-05,
+      "loss": 0.471,
+      "step": 258
+    },
+    {
+      "epoch": 0.5546038543897216,
+      "grad_norm": 0.26803290950789216,
+      "learning_rate": 4.531746031746032e-05,
+      "loss": 0.4743,
+      "step": 259
+    },
+    {
+      "epoch": 0.556745182012848,
+      "grad_norm": 0.38556934693664496,
+      "learning_rate": 4.527777777777778e-05,
+      "loss": 0.4768,
+      "step": 260
+    },
+    {
+      "epoch": 0.5588865096359743,
+      "grad_norm": 0.2661542720290909,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.466,
+      "step": 261
+    },
+    {
+      "epoch": 0.5610278372591007,
+      "grad_norm": 0.29067138820514277,
+      "learning_rate": 4.51984126984127e-05,
+      "loss": 0.4702,
+      "step": 262
+    },
+    {
+      "epoch": 0.563169164882227,
+      "grad_norm": 0.2872183782023422,
+      "learning_rate": 4.515873015873016e-05,
+      "loss": 0.4534,
+      "step": 263
+    },
+    {
+      "epoch": 0.5653104925053534,
+      "grad_norm": 0.25367532092796774,
+      "learning_rate": 4.511904761904762e-05,
+      "loss": 0.4551,
+      "step": 264
+    },
+    {
+      "epoch": 0.5674518201284796,
+      "grad_norm": 0.3039430947031411,
+      "learning_rate": 4.507936507936508e-05,
+      "loss": 0.4627,
+      "step": 265
+    },
+    {
+      "epoch": 0.569593147751606,
+      "grad_norm": 0.28351706499792234,
+      "learning_rate": 4.503968253968254e-05,
+      "loss": 0.4574,
+      "step": 266
+    },
+    {
+      "epoch": 0.5717344753747323,
+      "grad_norm": 0.27425219768007014,
+      "learning_rate": 4.5e-05,
+      "loss": 0.4589,
+      "step": 267
+    },
+    {
+      "epoch": 0.5738758029978587,
+      "grad_norm": 0.2869408267882252,
+      "learning_rate": 4.4960317460317464e-05,
+      "loss": 0.4475,
+      "step": 268
+    },
+    {
+      "epoch": 0.576017130620985,
+      "grad_norm": 0.2762345424358576,
+      "learning_rate": 4.4920634920634924e-05,
+      "loss": 0.4708,
+      "step": 269
+    },
+    {
+      "epoch": 0.5781584582441114,
+      "grad_norm": 0.3049597843838145,
+      "learning_rate": 4.4880952380952385e-05,
+      "loss": 0.4689,
+      "step": 270
+    },
+    {
+      "epoch": 0.5802997858672377,
+      "grad_norm": 0.2592725090051658,
+      "learning_rate": 4.4841269841269846e-05,
+      "loss": 0.4656,
+      "step": 271
+    },
+    {
+      "epoch": 0.582441113490364,
+      "grad_norm": 0.26044781623078483,
+      "learning_rate": 4.4801587301587307e-05,
+      "loss": 0.4638,
+      "step": 272
+    },
+    {
+      "epoch": 0.5845824411134903,
+      "grad_norm": 0.2804692407312598,
+      "learning_rate": 4.476190476190477e-05,
+      "loss": 0.4675,
+      "step": 273
+    },
+    {
+      "epoch": 0.5867237687366167,
+      "grad_norm": 0.26684962117113786,
+      "learning_rate": 4.472222222222223e-05,
+      "loss": 0.4467,
+      "step": 274
+    },
+    {
+      "epoch": 0.588865096359743,
+      "grad_norm": 0.31968480635740154,
+      "learning_rate": 4.468253968253969e-05,
+      "loss": 0.474,
+      "step": 275
+    },
+    {
+      "epoch": 0.5910064239828694,
+      "grad_norm": 0.2644188695273227,
+      "learning_rate": 4.464285714285715e-05,
+      "loss": 0.4686,
+      "step": 276
+    },
+    {
+      "epoch": 0.5931477516059958,
+      "grad_norm": 0.2930643108233625,
+      "learning_rate": 4.460317460317461e-05,
+      "loss": 0.4669,
+      "step": 277
+    },
+    {
+      "epoch": 0.5952890792291221,
+      "grad_norm": 0.3015159130498082,
+      "learning_rate": 4.456349206349207e-05,
+      "loss": 0.4648,
+      "step": 278
+    },
+    {
+      "epoch": 0.5974304068522484,
+      "grad_norm": 0.2727279142726974,
+      "learning_rate": 4.4523809523809525e-05,
+      "loss": 0.4612,
+      "step": 279
+    },
+    {
+      "epoch": 0.5995717344753747,
+      "grad_norm": 0.2975236904657764,
+      "learning_rate": 4.4484126984126986e-05,
+      "loss": 0.444,
+      "step": 280
+    },
+    {
+      "epoch": 0.6017130620985011,
+      "grad_norm": 0.2580099237283999,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.4792,
+      "step": 281
+    },
+    {
+      "epoch": 0.6038543897216274,
+      "grad_norm": 0.2961073515555114,
+      "learning_rate": 4.440476190476191e-05,
+      "loss": 0.4534,
+      "step": 282
+    },
+    {
+      "epoch": 0.6059957173447538,
+      "grad_norm": 0.2764889525421342,
+      "learning_rate": 4.436507936507937e-05,
+      "loss": 0.4442,
+      "step": 283
+    },
+    {
+      "epoch": 0.6081370449678801,
+      "grad_norm": 0.26435268929370137,
+      "learning_rate": 4.432539682539683e-05,
+      "loss": 0.4452,
+      "step": 284
+    },
+    {
+      "epoch": 0.6102783725910065,
+      "grad_norm": 0.2882202813256758,
+      "learning_rate": 4.428571428571428e-05,
+      "loss": 0.4448,
+      "step": 285
+    },
+    {
+      "epoch": 0.6124197002141327,
+      "grad_norm": 0.2561848522615123,
+      "learning_rate": 4.4246031746031744e-05,
+      "loss": 0.453,
+      "step": 286
+    },
+    {
+      "epoch": 0.6145610278372591,
+      "grad_norm": 0.32128316367276627,
+      "learning_rate": 4.4206349206349204e-05,
+      "loss": 0.4676,
+      "step": 287
+    },
+    {
+      "epoch": 0.6167023554603854,
+      "grad_norm": 0.305653864447794,
+      "learning_rate": 4.4166666666666665e-05,
+      "loss": 0.457,
+      "step": 288
+    },
+    {
+      "epoch": 0.6188436830835118,
+      "grad_norm": 0.27429918226444405,
+      "learning_rate": 4.4126984126984126e-05,
+      "loss": 0.4592,
+      "step": 289
+    },
+    {
+      "epoch": 0.6209850107066381,
+      "grad_norm": 0.26436848550152375,
+      "learning_rate": 4.4087301587301587e-05,
+      "loss": 0.4343,
+      "step": 290
+    },
+    {
+      "epoch": 0.6231263383297645,
+      "grad_norm": 0.2892075831032154,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 0.4405,
+      "step": 291
+    },
+    {
+      "epoch": 0.6252676659528907,
+      "grad_norm": 0.27340488703312127,
+      "learning_rate": 4.400793650793651e-05,
+      "loss": 0.4754,
+      "step": 292
+    },
+    {
+      "epoch": 0.6274089935760171,
+      "grad_norm": 0.2748574269063886,
+      "learning_rate": 4.396825396825397e-05,
+      "loss": 0.474,
+      "step": 293
+    },
+    {
+      "epoch": 0.6295503211991434,
+      "grad_norm": 0.2941612988883471,
+      "learning_rate": 4.392857142857143e-05,
+      "loss": 0.4564,
+      "step": 294
+    },
+    {
+      "epoch": 0.6316916488222698,
+      "grad_norm": 0.24636592894226964,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 0.4422,
+      "step": 295
+    },
+    {
+      "epoch": 0.6338329764453962,
+      "grad_norm": 0.28582887621940406,
+      "learning_rate": 4.384920634920635e-05,
+      "loss": 0.4588,
+      "step": 296
+    },
+    {
+      "epoch": 0.6359743040685225,
+      "grad_norm": 0.27743437358263945,
+      "learning_rate": 4.380952380952381e-05,
+      "loss": 0.4469,
+      "step": 297
+    },
+    {
+      "epoch": 0.6381156316916489,
+      "grad_norm": 0.26242382702944955,
+      "learning_rate": 4.376984126984127e-05,
+      "loss": 0.4398,
+      "step": 298
+    },
+    {
+      "epoch": 0.6402569593147751,
+      "grad_norm": 0.28165809226942223,
+      "learning_rate": 4.373015873015873e-05,
+      "loss": 0.4413,
+      "step": 299
+    },
+    {
+      "epoch": 0.6423982869379015,
+      "grad_norm": 0.25284838465648923,
+      "learning_rate": 4.3690476190476194e-05,
+      "loss": 0.4342,
+      "step": 300
+    },
+    {
+      "epoch": 0.6445396145610278,
+      "grad_norm": 0.31339700087144234,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.4612,
+      "step": 301
+    },
+    {
+      "epoch": 0.6466809421841542,
+      "grad_norm": 0.3104952345189579,
+      "learning_rate": 4.3611111111111116e-05,
+      "loss": 0.4731,
+      "step": 302
+    },
+    {
+      "epoch": 0.6488222698072805,
+      "grad_norm": 0.2947572345084962,
+      "learning_rate": 4.3571428571428576e-05,
+      "loss": 0.453,
+      "step": 303
+    },
+    {
+      "epoch": 0.6509635974304069,
+      "grad_norm": 0.3455960603661197,
+      "learning_rate": 4.353174603174604e-05,
+      "loss": 0.4644,
+      "step": 304
+    },
+    {
+      "epoch": 0.6531049250535332,
+      "grad_norm": 0.25711343366321854,
+      "learning_rate": 4.34920634920635e-05,
+      "loss": 0.4802,
+      "step": 305
+    },
+    {
+      "epoch": 0.6552462526766595,
+      "grad_norm": 0.31160993789120733,
+      "learning_rate": 4.345238095238096e-05,
+      "loss": 0.4534,
+      "step": 306
+    },
+    {
+      "epoch": 0.6573875802997858,
+      "grad_norm": 0.27484057830146985,
+      "learning_rate": 4.341269841269842e-05,
+      "loss": 0.4529,
+      "step": 307
+    },
+    {
+      "epoch": 0.6595289079229122,
+      "grad_norm": 0.26337065068689747,
+      "learning_rate": 4.337301587301587e-05,
+      "loss": 0.4647,
+      "step": 308
+    },
+    {
+      "epoch": 0.6616702355460385,
+      "grad_norm": 0.271610173490994,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.4445,
+      "step": 309
+    },
+    {
+      "epoch": 0.6638115631691649,
+      "grad_norm": 0.2528923842806531,
+      "learning_rate": 4.3293650793650795e-05,
+      "loss": 0.4571,
+      "step": 310
+    },
+    {
+      "epoch": 0.6659528907922913,
+      "grad_norm": 0.28885045896535533,
+      "learning_rate": 4.3253968253968256e-05,
+      "loss": 0.4554,
+      "step": 311
+    },
+    {
+      "epoch": 0.6680942184154176,
+      "grad_norm": 0.25814187425851787,
+      "learning_rate": 4.3214285714285716e-05,
+      "loss": 0.468,
+      "step": 312
+    },
+    {
+      "epoch": 0.6702355460385439,
+      "grad_norm": 0.26764371303035933,
+      "learning_rate": 4.317460317460318e-05,
+      "loss": 0.4594,
+      "step": 313
+    },
+    {
+      "epoch": 0.6723768736616702,
+      "grad_norm": 0.24837539558393776,
+      "learning_rate": 4.313492063492064e-05,
+      "loss": 0.4371,
+      "step": 314
+    },
+    {
+      "epoch": 0.6745182012847966,
+      "grad_norm": 0.25412398449219464,
+      "learning_rate": 4.30952380952381e-05,
+      "loss": 0.4594,
+      "step": 315
+    },
+    {
+      "epoch": 0.6766595289079229,
+      "grad_norm": 0.23755251323497587,
+      "learning_rate": 4.305555555555556e-05,
+      "loss": 0.4674,
+      "step": 316
+    },
+    {
+      "epoch": 0.6788008565310493,
+      "grad_norm": 0.2924363858039394,
+      "learning_rate": 4.301587301587302e-05,
+      "loss": 0.4688,
+      "step": 317
+    },
+    {
+      "epoch": 0.6809421841541756,
+      "grad_norm": 0.2570229512356918,
+      "learning_rate": 4.297619047619048e-05,
+      "loss": 0.4494,
+      "step": 318
+    },
+    {
+      "epoch": 0.683083511777302,
+      "grad_norm": 0.2764941603084112,
+      "learning_rate": 4.2936507936507935e-05,
+      "loss": 0.4605,
+      "step": 319
+    },
+    {
+      "epoch": 0.6852248394004282,
+      "grad_norm": 0.3153973496882214,
+      "learning_rate": 4.2896825396825396e-05,
+      "loss": 0.4488,
+      "step": 320
+    },
+    {
+      "epoch": 0.6873661670235546,
+      "grad_norm": 0.28153011145522383,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.4705,
+      "step": 321
+    },
+    {
+      "epoch": 0.6895074946466809,
+      "grad_norm": 0.3506781644039644,
+      "learning_rate": 4.281746031746032e-05,
+      "loss": 0.4576,
+      "step": 322
+    },
+    {
+      "epoch": 0.6916488222698073,
+      "grad_norm": 0.3053110048188775,
+      "learning_rate": 4.277777777777778e-05,
+      "loss": 0.4732,
+      "step": 323
+    },
+    {
+      "epoch": 0.6937901498929336,
+      "grad_norm": 0.3240151118248191,
+      "learning_rate": 4.273809523809524e-05,
+      "loss": 0.4554,
+      "step": 324
+    },
+    {
+      "epoch": 0.69593147751606,
+      "grad_norm": 0.29062216265241864,
+      "learning_rate": 4.26984126984127e-05,
+      "loss": 0.4342,
+      "step": 325
+    },
+    {
+      "epoch": 0.6980728051391863,
+      "grad_norm": 0.25283396063398134,
+      "learning_rate": 4.265873015873016e-05,
+      "loss": 0.4464,
+      "step": 326
+    },
+    {
+      "epoch": 0.7002141327623126,
+      "grad_norm": 0.2680651552624257,
+      "learning_rate": 4.261904761904762e-05,
+      "loss": 0.4601,
+      "step": 327
+    },
+    {
+      "epoch": 0.702355460385439,
+      "grad_norm": 0.33752170462860476,
+      "learning_rate": 4.257936507936508e-05,
+      "loss": 0.4478,
+      "step": 328
+    },
+    {
+      "epoch": 0.7044967880085653,
+      "grad_norm": 0.2190411233461398,
+      "learning_rate": 4.253968253968254e-05,
+      "loss": 0.4465,
+      "step": 329
+    },
+    {
+      "epoch": 0.7066381156316917,
+      "grad_norm": 0.2508058536828542,
+      "learning_rate": 4.25e-05,
+      "loss": 0.4452,
+      "step": 330
+    },
+    {
+      "epoch": 0.708779443254818,
+      "grad_norm": 0.2377335448608951,
+      "learning_rate": 4.2460317460317464e-05,
+      "loss": 0.4401,
+      "step": 331
+    },
+    {
+      "epoch": 0.7109207708779444,
+      "grad_norm": 0.2529235872462942,
+      "learning_rate": 4.2420634920634925e-05,
+      "loss": 0.4651,
+      "step": 332
+    },
+    {
+      "epoch": 0.7130620985010707,
+      "grad_norm": 0.2516915529502037,
+      "learning_rate": 4.2380952380952385e-05,
+      "loss": 0.4617,
+      "step": 333
+    },
+    {
+      "epoch": 0.715203426124197,
+      "grad_norm": 0.2364540674065309,
+      "learning_rate": 4.2341269841269846e-05,
+      "loss": 0.4711,
+      "step": 334
+    },
+    {
+      "epoch": 0.7173447537473233,
+      "grad_norm": 0.25560457992791263,
+      "learning_rate": 4.23015873015873e-05,
+      "loss": 0.4558,
+      "step": 335
+    },
+    {
+      "epoch": 0.7194860813704497,
+      "grad_norm": 0.23835034145803924,
+      "learning_rate": 4.226190476190476e-05,
+      "loss": 0.4643,
+      "step": 336
+    },
+    {
+      "epoch": 0.721627408993576,
+      "grad_norm": 0.22675744611839385,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.4477,
+      "step": 337
+    },
+    {
+      "epoch": 0.7237687366167024,
+      "grad_norm": 0.24050077892906585,
+      "learning_rate": 4.218253968253968e-05,
+      "loss": 0.4417,
+      "step": 338
+    },
+    {
+      "epoch": 0.7259100642398287,
+      "grad_norm": 0.25640906605974817,
+      "learning_rate": 4.214285714285714e-05,
+      "loss": 0.4549,
+      "step": 339
+    },
+    {
+      "epoch": 0.728051391862955,
+      "grad_norm": 0.228989440935261,
+      "learning_rate": 4.2103174603174604e-05,
+      "loss": 0.4504,
+      "step": 340
+    },
+    {
+      "epoch": 0.7301927194860813,
+      "grad_norm": 0.2308736253356299,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.4577,
+      "step": 341
+    },
+    {
+      "epoch": 0.7323340471092077,
+      "grad_norm": 0.25703393548284337,
+      "learning_rate": 4.2023809523809525e-05,
+      "loss": 0.4586,
+      "step": 342
+    },
+    {
+      "epoch": 0.734475374732334,
+      "grad_norm": 0.24906898794284202,
+      "learning_rate": 4.1984126984126986e-05,
+      "loss": 0.4457,
+      "step": 343
+    },
+    {
+      "epoch": 0.7366167023554604,
+      "grad_norm": 0.2574124725732392,
+      "learning_rate": 4.194444444444445e-05,
+      "loss": 0.4529,
+      "step": 344
+    },
+    {
+      "epoch": 0.7387580299785867,
+      "grad_norm": 0.243507960931399,
+      "learning_rate": 4.190476190476191e-05,
+      "loss": 0.4438,
+      "step": 345
+    },
+    {
+      "epoch": 0.7408993576017131,
+      "grad_norm": 0.2445092413905703,
+      "learning_rate": 4.186507936507937e-05,
+      "loss": 0.4498,
+      "step": 346
+    },
+    {
+      "epoch": 0.7430406852248393,
+      "grad_norm": 0.30753988028224943,
+      "learning_rate": 4.182539682539683e-05,
+      "loss": 0.4598,
+      "step": 347
+    },
+    {
+      "epoch": 0.7451820128479657,
+      "grad_norm": 0.2505049178661786,
+      "learning_rate": 4.178571428571429e-05,
+      "loss": 0.4553,
+      "step": 348
+    },
+    {
+      "epoch": 0.7473233404710921,
+      "grad_norm": 0.27044349556425973,
+      "learning_rate": 4.174603174603175e-05,
+      "loss": 0.4573,
+      "step": 349
+    },
+    {
+      "epoch": 0.7494646680942184,
+      "grad_norm": 0.25124615427090896,
+      "learning_rate": 4.170634920634921e-05,
+      "loss": 0.476,
+      "step": 350
+    },
+    {
+      "epoch": 0.7516059957173448,
+      "grad_norm": 0.26703556388241734,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4366,
+      "step": 351
+    },
+    {
+      "epoch": 0.7537473233404711,
+      "grad_norm": 0.24998740358371754,
+      "learning_rate": 4.162698412698413e-05,
+      "loss": 0.4467,
+      "step": 352
+    },
+    {
+      "epoch": 0.7558886509635975,
+      "grad_norm": 0.2554947787191526,
+      "learning_rate": 4.1587301587301594e-05,
+      "loss": 0.4414,
+      "step": 353
+    },
+    {
+      "epoch": 0.7580299785867237,
+      "grad_norm": 0.29499953780878213,
+      "learning_rate": 4.1547619047619054e-05,
+      "loss": 0.4668,
+      "step": 354
+    },
+    {
+      "epoch": 0.7601713062098501,
+      "grad_norm": 0.2634686274433604,
+      "learning_rate": 4.1507936507936515e-05,
+      "loss": 0.4501,
+      "step": 355
+    },
+    {
+      "epoch": 0.7623126338329764,
+      "grad_norm": 0.27587805015536104,
+      "learning_rate": 4.1468253968253976e-05,
+      "loss": 0.4345,
+      "step": 356
+    },
+    {
+      "epoch": 0.7644539614561028,
+      "grad_norm": 0.27969963429910366,
+      "learning_rate": 4.1428571428571437e-05,
+      "loss": 0.4517,
+      "step": 357
+    },
+    {
+      "epoch": 0.7665952890792291,
+      "grad_norm": 0.3006002807511561,
+      "learning_rate": 4.138888888888889e-05,
+      "loss": 0.4363,
+      "step": 358
+    },
+    {
+      "epoch": 0.7687366167023555,
+      "grad_norm": 0.24094041832952617,
+      "learning_rate": 4.134920634920635e-05,
+      "loss": 0.4463,
+      "step": 359
+    },
+    {
+      "epoch": 0.7708779443254818,
+      "grad_norm": 0.30068476671680583,
+      "learning_rate": 4.130952380952381e-05,
+      "loss": 0.4394,
+      "step": 360
+    },
+    {
+      "epoch": 0.7730192719486081,
+      "grad_norm": 0.24223112221889828,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 0.4437,
+      "step": 361
+    },
+    {
+      "epoch": 0.7751605995717344,
+      "grad_norm": 0.3306808578708232,
+      "learning_rate": 4.123015873015873e-05,
+      "loss": 0.463,
+      "step": 362
+    },
+    {
+      "epoch": 0.7773019271948608,
+      "grad_norm": 0.22496155789537978,
+      "learning_rate": 4.119047619047619e-05,
+      "loss": 0.4396,
+      "step": 363
+    },
+    {
+      "epoch": 0.7794432548179872,
+      "grad_norm": 0.26980303715734333,
+      "learning_rate": 4.115079365079365e-05,
+      "loss": 0.4502,
+      "step": 364
+    },
+    {
+      "epoch": 0.7815845824411135,
+      "grad_norm": 0.24531612262771849,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 0.4371,
+      "step": 365
+    },
+    {
+      "epoch": 0.7837259100642399,
+      "grad_norm": 0.23508992950961904,
+      "learning_rate": 4.107142857142857e-05,
+      "loss": 0.4351,
+      "step": 366
+    },
+    {
+      "epoch": 0.7858672376873662,
+      "grad_norm": 0.2343557766020594,
+      "learning_rate": 4.103174603174603e-05,
+      "loss": 0.4474,
+      "step": 367
+    },
+    {
+      "epoch": 0.7880085653104925,
+      "grad_norm": 0.24946904187399346,
+      "learning_rate": 4.099206349206349e-05,
+      "loss": 0.4432,
+      "step": 368
+    },
+    {
+      "epoch": 0.7901498929336188,
+      "grad_norm": 0.253014359764136,
+      "learning_rate": 4.095238095238095e-05,
+      "loss": 0.4528,
+      "step": 369
+    },
+    {
+      "epoch": 0.7922912205567452,
+      "grad_norm": 0.22772669050696337,
+      "learning_rate": 4.091269841269841e-05,
+      "loss": 0.4543,
+      "step": 370
+    },
+    {
+      "epoch": 0.7944325481798715,
+      "grad_norm": 0.24881653297292614,
+      "learning_rate": 4.0873015873015874e-05,
+      "loss": 0.4312,
+      "step": 371
+    },
+    {
+      "epoch": 0.7965738758029979,
+      "grad_norm": 0.25995771340020857,
+      "learning_rate": 4.0833333333333334e-05,
+      "loss": 0.4265,
+      "step": 372
+    },
+    {
+      "epoch": 0.7987152034261242,
+      "grad_norm": 0.21693583355693366,
+      "learning_rate": 4.0793650793650795e-05,
+      "loss": 0.4278,
+      "step": 373
+    },
+    {
+      "epoch": 0.8008565310492506,
+      "grad_norm": 0.24159689301797868,
+      "learning_rate": 4.0753968253968256e-05,
+      "loss": 0.4437,
+      "step": 374
+    },
+    {
+      "epoch": 0.8029978586723768,
+      "grad_norm": 0.25009229119018167,
+      "learning_rate": 4.0714285714285717e-05,
+      "loss": 0.4338,
+      "step": 375
+    },
+    {
+      "epoch": 0.8051391862955032,
+      "grad_norm": 0.2486299019264688,
+      "learning_rate": 4.067460317460318e-05,
+      "loss": 0.4506,
+      "step": 376
+    },
+    {
+      "epoch": 0.8072805139186295,
+      "grad_norm": 0.247308996752361,
+      "learning_rate": 4.063492063492064e-05,
+      "loss": 0.4414,
+      "step": 377
+    },
+    {
+      "epoch": 0.8094218415417559,
+      "grad_norm": 0.2590933139992137,
+      "learning_rate": 4.05952380952381e-05,
+      "loss": 0.4297,
+      "step": 378
+    },
+    {
+      "epoch": 0.8115631691648822,
+      "grad_norm": 0.2508458429815614,
+      "learning_rate": 4.055555555555556e-05,
+      "loss": 0.4269,
+      "step": 379
+    },
+    {
+      "epoch": 0.8137044967880086,
+      "grad_norm": 0.2879218570314043,
+      "learning_rate": 4.051587301587302e-05,
+      "loss": 0.4466,
+      "step": 380
+    },
+    {
+      "epoch": 0.815845824411135,
+      "grad_norm": 0.23441911617011937,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.4412,
+      "step": 381
+    },
+    {
+      "epoch": 0.8179871520342612,
+      "grad_norm": 0.32198396205698543,
+      "learning_rate": 4.043650793650794e-05,
+      "loss": 0.4547,
+      "step": 382
+    },
+    {
+      "epoch": 0.8201284796573876,
+      "grad_norm": 0.30161712796978346,
+      "learning_rate": 4.03968253968254e-05,
+      "loss": 0.4515,
+      "step": 383
+    },
+    {
+      "epoch": 0.8222698072805139,
+      "grad_norm": 0.23761583289371385,
+      "learning_rate": 4.035714285714286e-05,
+      "loss": 0.4343,
+      "step": 384
+    },
+    {
+      "epoch": 0.8244111349036403,
+      "grad_norm": 0.27338237171088253,
+      "learning_rate": 4.031746031746032e-05,
+      "loss": 0.445,
+      "step": 385
+    },
+    {
+      "epoch": 0.8265524625267666,
+      "grad_norm": 0.2856895105115863,
+      "learning_rate": 4.027777777777778e-05,
+      "loss": 0.4401,
+      "step": 386
+    },
+    {
+      "epoch": 0.828693790149893,
+      "grad_norm": 0.2379960246228477,
+      "learning_rate": 4.023809523809524e-05,
+      "loss": 0.4411,
+      "step": 387
+    },
+    {
+      "epoch": 0.8308351177730193,
+      "grad_norm": 0.3631643489141991,
+      "learning_rate": 4.01984126984127e-05,
+      "loss": 0.4528,
+      "step": 388
+    },
+    {
+      "epoch": 0.8329764453961456,
+      "grad_norm": 0.2664861994481591,
+      "learning_rate": 4.015873015873016e-05,
+      "loss": 0.454,
+      "step": 389
+    },
+    {
+      "epoch": 0.8351177730192719,
+      "grad_norm": 0.3054363176379015,
+      "learning_rate": 4.011904761904762e-05,
+      "loss": 0.4535,
+      "step": 390
+    },
+    {
+      "epoch": 0.8372591006423983,
+      "grad_norm": 0.31912720459183097,
+      "learning_rate": 4.007936507936508e-05,
+      "loss": 0.4493,
+      "step": 391
+    },
+    {
+      "epoch": 0.8394004282655246,
+      "grad_norm": 0.23251206637574764,
+      "learning_rate": 4.003968253968254e-05,
+      "loss": 0.4431,
+      "step": 392
+    },
+    {
+      "epoch": 0.841541755888651,
+      "grad_norm": 0.28631581267375344,
+      "learning_rate": 4e-05,
+      "loss": 0.4323,
+      "step": 393
+    },
+    {
+      "epoch": 0.8436830835117773,
+      "grad_norm": 0.267723471910929,
+      "learning_rate": 3.9960317460317464e-05,
+      "loss": 0.4473,
+      "step": 394
+    },
+    {
+      "epoch": 0.8458244111349036,
+      "grad_norm": 0.27597440700713444,
+      "learning_rate": 3.9920634920634925e-05,
+      "loss": 0.4542,
+      "step": 395
+    },
+    {
+      "epoch": 0.8479657387580299,
+      "grad_norm": 0.25433002014223405,
+      "learning_rate": 3.9880952380952386e-05,
+      "loss": 0.4444,
+      "step": 396
+    },
+    {
+      "epoch": 0.8501070663811563,
+      "grad_norm": 0.25437045991554746,
+      "learning_rate": 3.984126984126984e-05,
+      "loss": 0.4579,
+      "step": 397
+    },
+    {
+      "epoch": 0.8522483940042827,
+      "grad_norm": 0.24534548283140603,
+      "learning_rate": 3.98015873015873e-05,
+      "loss": 0.4725,
+      "step": 398
+    },
+    {
+      "epoch": 0.854389721627409,
+      "grad_norm": 0.25435785462287974,
+      "learning_rate": 3.976190476190476e-05,
+      "loss": 0.4512,
+      "step": 399
+    },
+    {
+      "epoch": 0.8565310492505354,
+      "grad_norm": 0.24601810950369032,
+      "learning_rate": 3.972222222222222e-05,
+      "loss": 0.448,
+      "step": 400
+    },
+    {
+      "epoch": 0.8586723768736617,
+      "grad_norm": 0.2653361455115733,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.439,
+      "step": 401
+    },
+    {
+      "epoch": 0.860813704496788,
+      "grad_norm": 0.23673358174194264,
+      "learning_rate": 3.964285714285714e-05,
+      "loss": 0.4413,
+      "step": 402
+    },
+    {
+      "epoch": 0.8629550321199143,
+      "grad_norm": 0.24462225822851738,
+      "learning_rate": 3.9603174603174604e-05,
+      "loss": 0.45,
+      "step": 403
+    },
+    {
+      "epoch": 0.8650963597430407,
+      "grad_norm": 0.2788329723246181,
+      "learning_rate": 3.9563492063492065e-05,
+      "loss": 0.454,
+      "step": 404
+    },
+    {
+      "epoch": 0.867237687366167,
+      "grad_norm": 0.24432656732987582,
+      "learning_rate": 3.9523809523809526e-05,
+      "loss": 0.4346,
+      "step": 405
+    },
+    {
+      "epoch": 0.8693790149892934,
+      "grad_norm": 0.3020331491334916,
+      "learning_rate": 3.9484126984126986e-05,
+      "loss": 0.4522,
+      "step": 406
+    },
+    {
+      "epoch": 0.8715203426124197,
+      "grad_norm": 0.2856505508961332,
+      "learning_rate": 3.944444444444445e-05,
+      "loss": 0.4485,
+      "step": 407
+    },
+    {
+      "epoch": 0.8736616702355461,
+      "grad_norm": 0.2656132022406484,
+      "learning_rate": 3.940476190476191e-05,
+      "loss": 0.4601,
+      "step": 408
+    },
+    {
+      "epoch": 0.8758029978586723,
+      "grad_norm": 0.29873704798610634,
+      "learning_rate": 3.936507936507937e-05,
+      "loss": 0.4354,
+      "step": 409
+    },
+    {
+      "epoch": 0.8779443254817987,
+      "grad_norm": 0.28600605020427156,
+      "learning_rate": 3.932539682539683e-05,
+      "loss": 0.4452,
+      "step": 410
+    },
+    {
+      "epoch": 0.880085653104925,
+      "grad_norm": 0.26767141652320475,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.4414,
+      "step": 411
+    },
+    {
+      "epoch": 0.8822269807280514,
+      "grad_norm": 0.35131045572847885,
+      "learning_rate": 3.9246031746031744e-05,
+      "loss": 0.4475,
+      "step": 412
+    },
+    {
+      "epoch": 0.8843683083511777,
+      "grad_norm": 0.2370839823487289,
+      "learning_rate": 3.9206349206349205e-05,
+      "loss": 0.4358,
+      "step": 413
+    },
+    {
+      "epoch": 0.8865096359743041,
+      "grad_norm": 0.3544368048991166,
+      "learning_rate": 3.9166666666666665e-05,
+      "loss": 0.4494,
+      "step": 414
+    },
+    {
+      "epoch": 0.8886509635974305,
+      "grad_norm": 0.3172911304193655,
+      "learning_rate": 3.9126984126984126e-05,
+      "loss": 0.4487,
+      "step": 415
+    },
+    {
+      "epoch": 0.8907922912205567,
+      "grad_norm": 0.275970880794438,
+      "learning_rate": 3.908730158730159e-05,
+      "loss": 0.4192,
+      "step": 416
+    },
+    {
+      "epoch": 0.892933618843683,
+      "grad_norm": 0.3547661108556819,
+      "learning_rate": 3.904761904761905e-05,
+      "loss": 0.4445,
+      "step": 417
+    },
+    {
+      "epoch": 0.8950749464668094,
+      "grad_norm": 0.24666251953784982,
+      "learning_rate": 3.900793650793651e-05,
+      "loss": 0.4486,
+      "step": 418
+    },
+    {
+      "epoch": 0.8972162740899358,
+      "grad_norm": 0.33556985351699165,
+      "learning_rate": 3.896825396825397e-05,
+      "loss": 0.4357,
+      "step": 419
+    },
+    {
+      "epoch": 0.8993576017130621,
+      "grad_norm": 0.29373216787613254,
+      "learning_rate": 3.892857142857143e-05,
+      "loss": 0.4559,
+      "step": 420
+    },
+    {
+      "epoch": 0.9014989293361885,
+      "grad_norm": 0.27236847144104587,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.4509,
+      "step": 421
+    },
+    {
+      "epoch": 0.9036402569593148,
+      "grad_norm": 0.3363146705059116,
+      "learning_rate": 3.884920634920635e-05,
+      "loss": 0.4338,
+      "step": 422
+    },
+    {
+      "epoch": 0.9057815845824411,
+      "grad_norm": 0.23799725549650105,
+      "learning_rate": 3.880952380952381e-05,
+      "loss": 0.4422,
+      "step": 423
+    },
+    {
+      "epoch": 0.9079229122055674,
+      "grad_norm": 0.2828958789244041,
+      "learning_rate": 3.876984126984127e-05,
+      "loss": 0.4597,
+      "step": 424
+    },
+    {
+      "epoch": 0.9100642398286938,
+      "grad_norm": 0.3136305684919964,
+      "learning_rate": 3.8730158730158734e-05,
+      "loss": 0.4361,
+      "step": 425
+    },
+    {
+      "epoch": 0.9122055674518201,
+      "grad_norm": 0.23243444701464022,
+      "learning_rate": 3.8690476190476195e-05,
+      "loss": 0.4337,
+      "step": 426
+    },
+    {
+      "epoch": 0.9143468950749465,
+      "grad_norm": 0.31121643328315307,
+      "learning_rate": 3.8650793650793655e-05,
+      "loss": 0.4429,
+      "step": 427
+    },
+    {
+      "epoch": 0.9164882226980728,
+      "grad_norm": 0.24007066200020147,
+      "learning_rate": 3.8611111111111116e-05,
+      "loss": 0.4412,
+      "step": 428
+    },
+    {
+      "epoch": 0.9186295503211992,
+      "grad_norm": 0.24526627011399632,
+      "learning_rate": 3.857142857142858e-05,
+      "loss": 0.4326,
+      "step": 429
+    },
+    {
+      "epoch": 0.9207708779443254,
+      "grad_norm": 0.2782809072036015,
+      "learning_rate": 3.853174603174604e-05,
+      "loss": 0.435,
+      "step": 430
+    },
+    {
+      "epoch": 0.9229122055674518,
+      "grad_norm": 0.22743806598005653,
+      "learning_rate": 3.84920634920635e-05,
+      "loss": 0.4522,
+      "step": 431
+    },
+    {
+      "epoch": 0.9250535331905781,
+      "grad_norm": 0.2652959665516378,
+      "learning_rate": 3.845238095238096e-05,
+      "loss": 0.4313,
+      "step": 432
+    },
+    {
+      "epoch": 0.9271948608137045,
+      "grad_norm": 0.2320440083455923,
+      "learning_rate": 3.841269841269842e-05,
+      "loss": 0.4336,
+      "step": 433
+    },
+    {
+      "epoch": 0.9293361884368309,
+      "grad_norm": 0.22904699603839362,
+      "learning_rate": 3.837301587301588e-05,
+      "loss": 0.4398,
+      "step": 434
+    },
+    {
+      "epoch": 0.9314775160599572,
+      "grad_norm": 0.28061696790352825,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 0.4398,
+      "step": 435
+    },
+    {
+      "epoch": 0.9336188436830836,
+      "grad_norm": 0.22771552947703397,
+      "learning_rate": 3.8293650793650795e-05,
+      "loss": 0.4427,
+      "step": 436
+    },
+    {
+      "epoch": 0.9357601713062098,
+      "grad_norm": 0.22379899530445632,
+      "learning_rate": 3.8253968253968256e-05,
+      "loss": 0.4292,
+      "step": 437
+    },
+    {
+      "epoch": 0.9379014989293362,
+      "grad_norm": 0.2705759563196105,
+      "learning_rate": 3.821428571428572e-05,
+      "loss": 0.4381,
+      "step": 438
+    },
+    {
+      "epoch": 0.9400428265524625,
+      "grad_norm": 0.25257412614253144,
+      "learning_rate": 3.817460317460317e-05,
+      "loss": 0.4441,
+      "step": 439
+    },
+    {
+      "epoch": 0.9421841541755889,
+      "grad_norm": 0.2230238602545094,
+      "learning_rate": 3.813492063492063e-05,
+      "loss": 0.4382,
+      "step": 440
+    },
+    {
+      "epoch": 0.9443254817987152,
+      "grad_norm": 0.2643425174467501,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.4494,
+      "step": 441
+    },
+    {
+      "epoch": 0.9464668094218416,
+      "grad_norm": 0.22879958232087208,
+      "learning_rate": 3.805555555555555e-05,
+      "loss": 0.4686,
+      "step": 442
+    },
+    {
+      "epoch": 0.9486081370449678,
+      "grad_norm": 0.3234801610814164,
+      "learning_rate": 3.8015873015873014e-05,
+      "loss": 0.4533,
+      "step": 443
+    },
+    {
+      "epoch": 0.9507494646680942,
+      "grad_norm": 0.2238623404459414,
+      "learning_rate": 3.7976190476190474e-05,
+      "loss": 0.4593,
+      "step": 444
+    },
+    {
+      "epoch": 0.9528907922912205,
+      "grad_norm": 0.2854000014313601,
+      "learning_rate": 3.7936507936507935e-05,
+      "loss": 0.4404,
+      "step": 445
+    },
+    {
+      "epoch": 0.9550321199143469,
+      "grad_norm": 0.22982231285103497,
+      "learning_rate": 3.7896825396825396e-05,
+      "loss": 0.447,
+      "step": 446
+    },
+    {
+      "epoch": 0.9571734475374732,
+      "grad_norm": 0.22798223769442008,
+      "learning_rate": 3.785714285714286e-05,
+      "loss": 0.4435,
+      "step": 447
+    },
+    {
+      "epoch": 0.9593147751605996,
+      "grad_norm": 0.2698671605451589,
+      "learning_rate": 3.781746031746032e-05,
+      "loss": 0.4468,
+      "step": 448
+    },
+    {
+      "epoch": 0.961456102783726,
+      "grad_norm": 0.25052757718595015,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.4559,
+      "step": 449
+    },
+    {
+      "epoch": 0.9635974304068522,
+      "grad_norm": 0.2509803927116757,
+      "learning_rate": 3.773809523809524e-05,
+      "loss": 0.4485,
+      "step": 450
+    },
+    {
+      "epoch": 0.9657387580299786,
+      "grad_norm": 0.24678486996154717,
+      "learning_rate": 3.76984126984127e-05,
+      "loss": 0.4469,
+      "step": 451
+    },
+    {
+      "epoch": 0.9678800856531049,
+      "grad_norm": 0.2297303050823692,
+      "learning_rate": 3.765873015873016e-05,
+      "loss": 0.4427,
+      "step": 452
+    },
+    {
+      "epoch": 0.9700214132762313,
+      "grad_norm": 0.24752427508051356,
+      "learning_rate": 3.761904761904762e-05,
+      "loss": 0.4414,
+      "step": 453
+    },
+    {
+      "epoch": 0.9721627408993576,
+      "grad_norm": 0.285731490530565,
+      "learning_rate": 3.757936507936508e-05,
+      "loss": 0.4711,
+      "step": 454
+    },
+    {
+      "epoch": 0.974304068522484,
+      "grad_norm": 0.2623040174515322,
+      "learning_rate": 3.753968253968254e-05,
+      "loss": 0.4297,
+      "step": 455
+    },
+    {
+      "epoch": 0.9764453961456103,
+      "grad_norm": 0.27326887949485235,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.4498,
+      "step": 456
+    },
+    {
+      "epoch": 0.9785867237687366,
+      "grad_norm": 0.2603762615843794,
+      "learning_rate": 3.7460317460317464e-05,
+      "loss": 0.4542,
+      "step": 457
+    },
+    {
+      "epoch": 0.9807280513918629,
+      "grad_norm": 0.32975141148086856,
+      "learning_rate": 3.7420634920634925e-05,
+      "loss": 0.4424,
+      "step": 458
+    },
+    {
+      "epoch": 0.9828693790149893,
+      "grad_norm": 0.29153212143683266,
+      "learning_rate": 3.7380952380952386e-05,
+      "loss": 0.4483,
+      "step": 459
+    },
+    {
+      "epoch": 0.9850107066381156,
+      "grad_norm": 0.29992516288872756,
+      "learning_rate": 3.7341269841269846e-05,
+      "loss": 0.4511,
+      "step": 460
+    },
+    {
+      "epoch": 0.987152034261242,
+      "grad_norm": 0.29958162250392306,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.4263,
+      "step": 461
+    },
+    {
+      "epoch": 0.9892933618843683,
+      "grad_norm": 0.24964751706597657,
+      "learning_rate": 3.726190476190476e-05,
+      "loss": 0.4324,
+      "step": 462
+    },
+    {
+      "epoch": 0.9914346895074947,
+      "grad_norm": 0.2614912237652234,
+      "learning_rate": 3.722222222222222e-05,
+      "loss": 0.4312,
+      "step": 463
+    },
+    {
+      "epoch": 0.9935760171306209,
+      "grad_norm": 0.266348778726537,
+      "learning_rate": 3.718253968253968e-05,
+      "loss": 0.4263,
+      "step": 464
+    },
+    {
+      "epoch": 0.9957173447537473,
+      "grad_norm": 0.23725530062162245,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.4364,
+      "step": 465
+    },
+    {
+      "epoch": 0.9978586723768736,
+      "grad_norm": 0.30914413313910843,
+      "learning_rate": 3.7103174603174604e-05,
+      "loss": 0.4261,
+      "step": 466
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23586300310874772,
+      "learning_rate": 3.7063492063492065e-05,
+      "loss": 0.4211,
+      "step": 467
+    },
+    {
+      "epoch": 1.0021413276231264,
+      "grad_norm": 0.3369816971461586,
+      "learning_rate": 3.7023809523809526e-05,
+      "loss": 0.3762,
+      "step": 468
+    },
+    {
+      "epoch": 1.0042826552462527,
+      "grad_norm": 0.22809858744604314,
+      "learning_rate": 3.6984126984126986e-05,
+      "loss": 0.3646,
+      "step": 469
+    },
+    {
+      "epoch": 1.006423982869379,
+      "grad_norm": 0.266143350315859,
+      "learning_rate": 3.694444444444445e-05,
+      "loss": 0.3746,
+      "step": 470
+    },
+    {
+      "epoch": 1.0085653104925054,
+      "grad_norm": 0.2884518377103039,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 0.3666,
+      "step": 471
+    },
+    {
+      "epoch": 1.0107066381156318,
+      "grad_norm": 0.28068641149516393,
+      "learning_rate": 3.686507936507937e-05,
+      "loss": 0.3743,
+      "step": 472
+    },
+    {
+      "epoch": 1.0128479657387581,
+      "grad_norm": 0.29435491983752216,
+      "learning_rate": 3.682539682539683e-05,
+      "loss": 0.3678,
+      "step": 473
+    },
+    {
+      "epoch": 1.0149892933618843,
+      "grad_norm": 0.29581454893570663,
+      "learning_rate": 3.678571428571429e-05,
+      "loss": 0.382,
+      "step": 474
+    },
+    {
+      "epoch": 1.0171306209850106,
+      "grad_norm": 0.2567395711302861,
+      "learning_rate": 3.674603174603175e-05,
+      "loss": 0.3781,
+      "step": 475
+    },
+    {
+      "epoch": 1.019271948608137,
+      "grad_norm": 0.30179003911878,
+      "learning_rate": 3.6706349206349205e-05,
+      "loss": 0.3815,
+      "step": 476
+    },
+    {
+      "epoch": 1.0214132762312633,
+      "grad_norm": 0.279742858383842,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.3777,
+      "step": 477
+    },
+    {
+      "epoch": 1.0235546038543897,
+      "grad_norm": 0.28030505005478956,
+      "learning_rate": 3.6626984126984126e-05,
+      "loss": 0.3656,
+      "step": 478
+    },
+    {
+      "epoch": 1.025695931477516,
+      "grad_norm": 0.2713443901873829,
+      "learning_rate": 3.658730158730159e-05,
+      "loss": 0.374,
+      "step": 479
+    },
+    {
+      "epoch": 1.0278372591006424,
+      "grad_norm": 0.2643392830096214,
+      "learning_rate": 3.654761904761905e-05,
+      "loss": 0.3649,
+      "step": 480
+    },
+    {
+      "epoch": 1.0299785867237687,
+      "grad_norm": 0.3008739451104665,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.3822,
+      "step": 481
+    },
+    {
+      "epoch": 1.032119914346895,
+      "grad_norm": 0.23862820873191215,
+      "learning_rate": 3.646825396825397e-05,
+      "loss": 0.3632,
+      "step": 482
+    },
+    {
+      "epoch": 1.0342612419700214,
+      "grad_norm": 0.27869337578181386,
+      "learning_rate": 3.642857142857143e-05,
+      "loss": 0.3636,
+      "step": 483
+    },
+    {
+      "epoch": 1.0364025695931478,
+      "grad_norm": 0.2533397054393399,
+      "learning_rate": 3.638888888888889e-05,
+      "loss": 0.3666,
+      "step": 484
+    },
+    {
+      "epoch": 1.0385438972162742,
+      "grad_norm": 0.21014394289523325,
+      "learning_rate": 3.634920634920635e-05,
+      "loss": 0.3609,
+      "step": 485
+    },
+    {
+      "epoch": 1.0406852248394005,
+      "grad_norm": 0.3283463606383022,
+      "learning_rate": 3.630952380952381e-05,
+      "loss": 0.3939,
+      "step": 486
+    },
+    {
+      "epoch": 1.0428265524625269,
+      "grad_norm": 0.22387699439354838,
+      "learning_rate": 3.626984126984127e-05,
+      "loss": 0.3611,
+      "step": 487
+    },
+    {
+      "epoch": 1.044967880085653,
+      "grad_norm": 0.2690411087151262,
+      "learning_rate": 3.6230158730158734e-05,
+      "loss": 0.3869,
+      "step": 488
+    },
+    {
+      "epoch": 1.0471092077087794,
+      "grad_norm": 0.3053411429357243,
+      "learning_rate": 3.619047619047619e-05,
+      "loss": 0.3845,
+      "step": 489
+    },
+    {
+      "epoch": 1.0492505353319057,
+      "grad_norm": 0.2631289951026371,
+      "learning_rate": 3.615079365079365e-05,
+      "loss": 0.3698,
+      "step": 490
+    },
+    {
+      "epoch": 1.051391862955032,
+      "grad_norm": 0.2670111664926503,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.3692,
+      "step": 491
+    },
+    {
+      "epoch": 1.0535331905781584,
+      "grad_norm": 0.2485103124351668,
+      "learning_rate": 3.607142857142857e-05,
+      "loss": 0.3538,
+      "step": 492
+    },
+    {
+      "epoch": 1.0556745182012848,
+      "grad_norm": 0.23420232693366563,
+      "learning_rate": 3.603174603174603e-05,
+      "loss": 0.3756,
+      "step": 493
+    },
+    {
+      "epoch": 1.0578158458244111,
+      "grad_norm": 0.23665666643547903,
+      "learning_rate": 3.599206349206349e-05,
+      "loss": 0.3577,
+      "step": 494
+    },
+    {
+      "epoch": 1.0599571734475375,
+      "grad_norm": 0.2681669094655506,
+      "learning_rate": 3.595238095238095e-05,
+      "loss": 0.3785,
+      "step": 495
+    },
+    {
+      "epoch": 1.0620985010706638,
+      "grad_norm": 0.22646726688527194,
+      "learning_rate": 3.591269841269841e-05,
+      "loss": 0.3657,
+      "step": 496
+    },
+    {
+      "epoch": 1.0642398286937902,
+      "grad_norm": 0.26622378410440317,
+      "learning_rate": 3.5873015873015874e-05,
+      "loss": 0.3879,
+      "step": 497
+    },
+    {
+      "epoch": 1.0663811563169165,
+      "grad_norm": 0.24041744536037002,
+      "learning_rate": 3.5833333333333335e-05,
+      "loss": 0.3799,
+      "step": 498
+    },
+    {
+      "epoch": 1.068522483940043,
+      "grad_norm": 0.27516627734803006,
+      "learning_rate": 3.5793650793650795e-05,
+      "loss": 0.3689,
+      "step": 499
+    },
+    {
+      "epoch": 1.0706638115631693,
+      "grad_norm": 0.23109318989193225,
+      "learning_rate": 3.5753968253968256e-05,
+      "loss": 0.3662,
+      "step": 500
+    },
+    {
+      "epoch": 1.0728051391862956,
+      "grad_norm": 0.2499488422329624,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.3703,
+      "step": 501
+    },
+    {
+      "epoch": 1.0749464668094217,
+      "grad_norm": 0.2706738165446803,
+      "learning_rate": 3.567460317460318e-05,
+      "loss": 0.3641,
+      "step": 502
+    },
+    {
+      "epoch": 1.077087794432548,
+      "grad_norm": 0.22012676483900637,
+      "learning_rate": 3.563492063492064e-05,
+      "loss": 0.3736,
+      "step": 503
+    },
+    {
+      "epoch": 1.0792291220556745,
+      "grad_norm": 0.2650545240192106,
+      "learning_rate": 3.55952380952381e-05,
+      "loss": 0.398,
+      "step": 504
+    },
+    {
+      "epoch": 1.0813704496788008,
+      "grad_norm": 0.2520843476038801,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 0.3597,
+      "step": 505
+    },
+    {
+      "epoch": 1.0835117773019272,
+      "grad_norm": 0.24197390108439093,
+      "learning_rate": 3.551587301587302e-05,
+      "loss": 0.3764,
+      "step": 506
+    },
+    {
+      "epoch": 1.0856531049250535,
+      "grad_norm": 0.24538696032664498,
+      "learning_rate": 3.547619047619048e-05,
+      "loss": 0.3825,
+      "step": 507
+    },
+    {
+      "epoch": 1.0877944325481799,
+      "grad_norm": 0.22888284257244346,
+      "learning_rate": 3.543650793650794e-05,
+      "loss": 0.3745,
+      "step": 508
+    },
+    {
+      "epoch": 1.0899357601713062,
+      "grad_norm": 0.24869463423729066,
+      "learning_rate": 3.53968253968254e-05,
+      "loss": 0.3493,
+      "step": 509
+    },
+    {
+      "epoch": 1.0920770877944326,
+      "grad_norm": 0.23993057946273721,
+      "learning_rate": 3.5357142857142864e-05,
+      "loss": 0.3663,
+      "step": 510
+    },
+    {
+      "epoch": 1.094218415417559,
+      "grad_norm": 0.2657625184616568,
+      "learning_rate": 3.5317460317460324e-05,
+      "loss": 0.3668,
+      "step": 511
+    },
+    {
+      "epoch": 1.0963597430406853,
+      "grad_norm": 0.2309814403862246,
+      "learning_rate": 3.527777777777778e-05,
+      "loss": 0.3787,
+      "step": 512
+    },
+    {
+      "epoch": 1.0985010706638116,
+      "grad_norm": 0.24234237491113314,
+      "learning_rate": 3.523809523809524e-05,
+      "loss": 0.3493,
+      "step": 513
+    },
+    {
+      "epoch": 1.100642398286938,
+      "grad_norm": 0.23974579149766023,
+      "learning_rate": 3.51984126984127e-05,
+      "loss": 0.3641,
+      "step": 514
+    },
+    {
+      "epoch": 1.1027837259100641,
+      "grad_norm": 0.24724340356035635,
+      "learning_rate": 3.515873015873016e-05,
+      "loss": 0.3702,
+      "step": 515
+    },
+    {
+      "epoch": 1.1049250535331905,
+      "grad_norm": 0.28135285517273,
+      "learning_rate": 3.511904761904762e-05,
+      "loss": 0.3695,
+      "step": 516
+    },
+    {
+      "epoch": 1.1070663811563168,
+      "grad_norm": 0.243348659130255,
+      "learning_rate": 3.5079365079365075e-05,
+      "loss": 0.3714,
+      "step": 517
+    },
+    {
+      "epoch": 1.1092077087794432,
+      "grad_norm": 0.24745510283381558,
+      "learning_rate": 3.5039682539682536e-05,
+      "loss": 0.3565,
+      "step": 518
+    },
+    {
+      "epoch": 1.1113490364025695,
+      "grad_norm": 0.2563919456656394,
+      "learning_rate": 3.5e-05,
+      "loss": 0.3766,
+      "step": 519
+    },
+    {
+      "epoch": 1.113490364025696,
+      "grad_norm": 0.22577716624985744,
+      "learning_rate": 3.496031746031746e-05,
+      "loss": 0.3697,
+      "step": 520
+    },
+    {
+      "epoch": 1.1156316916488223,
+      "grad_norm": 0.263849669081119,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.3621,
+      "step": 521
+    },
+    {
+      "epoch": 1.1177730192719486,
+      "grad_norm": 0.2578266915441593,
+      "learning_rate": 3.488095238095238e-05,
+      "loss": 0.3781,
+      "step": 522
+    },
+    {
+      "epoch": 1.119914346895075,
+      "grad_norm": 0.267410790093275,
+      "learning_rate": 3.484126984126984e-05,
+      "loss": 0.3802,
+      "step": 523
+    },
+    {
+      "epoch": 1.1220556745182013,
+      "grad_norm": 0.26720246872107734,
+      "learning_rate": 3.48015873015873e-05,
+      "loss": 0.3686,
+      "step": 524
+    },
+    {
+      "epoch": 1.1241970021413277,
+      "grad_norm": 0.22844067281982589,
+      "learning_rate": 3.476190476190476e-05,
+      "loss": 0.3892,
+      "step": 525
+    },
+    {
+      "epoch": 1.126338329764454,
+      "grad_norm": 0.2522676842739371,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 0.3695,
+      "step": 526
+    },
+    {
+      "epoch": 1.1284796573875804,
+      "grad_norm": 0.20988616282501774,
+      "learning_rate": 3.468253968253968e-05,
+      "loss": 0.3705,
+      "step": 527
+    },
+    {
+      "epoch": 1.1306209850107067,
+      "grad_norm": 0.2452290096455659,
+      "learning_rate": 3.4642857142857144e-05,
+      "loss": 0.3643,
+      "step": 528
+    },
+    {
+      "epoch": 1.132762312633833,
+      "grad_norm": 0.20376413452570455,
+      "learning_rate": 3.4603174603174604e-05,
+      "loss": 0.3621,
+      "step": 529
+    },
+    {
+      "epoch": 1.1349036402569592,
+      "grad_norm": 0.22961373150459624,
+      "learning_rate": 3.4563492063492065e-05,
+      "loss": 0.355,
+      "step": 530
+    },
+    {
+      "epoch": 1.1370449678800856,
+      "grad_norm": 0.2430831935717076,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 0.3689,
+      "step": 531
+    },
+    {
+      "epoch": 1.139186295503212,
+      "grad_norm": 0.2017982939822363,
+      "learning_rate": 3.448412698412699e-05,
+      "loss": 0.3508,
+      "step": 532
+    },
+    {
+      "epoch": 1.1413276231263383,
+      "grad_norm": 0.26746451866780874,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.3695,
+      "step": 533
+    },
+    {
+      "epoch": 1.1434689507494646,
+      "grad_norm": 0.24736040419305078,
+      "learning_rate": 3.440476190476191e-05,
+      "loss": 0.3604,
+      "step": 534
+    },
+    {
+      "epoch": 1.145610278372591,
+      "grad_norm": 0.23139512163454942,
+      "learning_rate": 3.436507936507937e-05,
+      "loss": 0.3597,
+      "step": 535
+    },
+    {
+      "epoch": 1.1477516059957173,
+      "grad_norm": 0.24151228114318973,
+      "learning_rate": 3.432539682539683e-05,
+      "loss": 0.3722,
+      "step": 536
+    },
+    {
+      "epoch": 1.1498929336188437,
+      "grad_norm": 0.225767578077691,
+      "learning_rate": 3.428571428571429e-05,
+      "loss": 0.3852,
+      "step": 537
+    },
+    {
+      "epoch": 1.15203426124197,
+      "grad_norm": 0.2379372088035776,
+      "learning_rate": 3.424603174603175e-05,
+      "loss": 0.377,
+      "step": 538
+    },
+    {
+      "epoch": 1.1541755888650964,
+      "grad_norm": 0.22031199802513296,
+      "learning_rate": 3.420634920634921e-05,
+      "loss": 0.3767,
+      "step": 539
+    },
+    {
+      "epoch": 1.1563169164882228,
+      "grad_norm": 0.23132051552867036,
+      "learning_rate": 3.4166666666666666e-05,
+      "loss": 0.3528,
+      "step": 540
+    },
+    {
+      "epoch": 1.1584582441113491,
+      "grad_norm": 0.24257151344704514,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.3595,
+      "step": 541
+    },
+    {
+      "epoch": 1.1605995717344753,
+      "grad_norm": 0.2261792555412017,
+      "learning_rate": 3.408730158730159e-05,
+      "loss": 0.3633,
+      "step": 542
+    },
+    {
+      "epoch": 1.1627408993576016,
+      "grad_norm": 0.2829561917904997,
+      "learning_rate": 3.404761904761905e-05,
+      "loss": 0.3648,
+      "step": 543
+    },
+    {
+      "epoch": 1.164882226980728,
+      "grad_norm": 0.25098056566365634,
+      "learning_rate": 3.400793650793651e-05,
+      "loss": 0.3702,
+      "step": 544
+    },
+    {
+      "epoch": 1.1670235546038543,
+      "grad_norm": 0.24644300251595483,
+      "learning_rate": 3.396825396825397e-05,
+      "loss": 0.3641,
+      "step": 545
+    },
+    {
+      "epoch": 1.1691648822269807,
+      "grad_norm": 0.23310081121807147,
+      "learning_rate": 3.392857142857143e-05,
+      "loss": 0.3724,
+      "step": 546
+    },
+    {
+      "epoch": 1.171306209850107,
+      "grad_norm": 0.2396918335180637,
+      "learning_rate": 3.388888888888889e-05,
+      "loss": 0.3724,
+      "step": 547
+    },
+    {
+      "epoch": 1.1734475374732334,
+      "grad_norm": 0.26305119086996215,
+      "learning_rate": 3.384920634920635e-05,
+      "loss": 0.3663,
+      "step": 548
+    },
+    {
+      "epoch": 1.1755888650963597,
+      "grad_norm": 0.23766315816254158,
+      "learning_rate": 3.380952380952381e-05,
+      "loss": 0.3543,
+      "step": 549
+    },
+    {
+      "epoch": 1.177730192719486,
+      "grad_norm": 0.24791689978321688,
+      "learning_rate": 3.3769841269841273e-05,
+      "loss": 0.3858,
+      "step": 550
+    },
+    {
+      "epoch": 1.1798715203426124,
+      "grad_norm": 0.2427021752634518,
+      "learning_rate": 3.3730158730158734e-05,
+      "loss": 0.3544,
+      "step": 551
+    },
+    {
+      "epoch": 1.1820128479657388,
+      "grad_norm": 0.2524229855590501,
+      "learning_rate": 3.3690476190476195e-05,
+      "loss": 0.3932,
+      "step": 552
+    },
+    {
+      "epoch": 1.1841541755888652,
+      "grad_norm": 0.2208689079418904,
+      "learning_rate": 3.3650793650793656e-05,
+      "loss": 0.3638,
+      "step": 553
+    },
+    {
+      "epoch": 1.1862955032119915,
+      "grad_norm": 0.24091697133765969,
+      "learning_rate": 3.3611111111111116e-05,
+      "loss": 0.351,
+      "step": 554
+    },
+    {
+      "epoch": 1.1884368308351179,
+      "grad_norm": 0.24197830327776518,
+      "learning_rate": 3.357142857142857e-05,
+      "loss": 0.3596,
+      "step": 555
+    },
+    {
+      "epoch": 1.1905781584582442,
+      "grad_norm": 0.19888129082318382,
+      "learning_rate": 3.353174603174603e-05,
+      "loss": 0.3643,
+      "step": 556
+    },
+    {
+      "epoch": 1.1927194860813706,
+      "grad_norm": 0.26369169905142914,
+      "learning_rate": 3.349206349206349e-05,
+      "loss": 0.3605,
+      "step": 557
+    },
+    {
+      "epoch": 1.1948608137044967,
+      "grad_norm": 0.2532299216339328,
+      "learning_rate": 3.345238095238095e-05,
+      "loss": 0.3878,
+      "step": 558
+    },
+    {
+      "epoch": 1.197002141327623,
+      "grad_norm": 0.23397198291097293,
+      "learning_rate": 3.3412698412698413e-05,
+      "loss": 0.3892,
+      "step": 559
+    },
+    {
+      "epoch": 1.1991434689507494,
+      "grad_norm": 0.3051071879009391,
+      "learning_rate": 3.3373015873015874e-05,
+      "loss": 0.3857,
+      "step": 560
+    },
+    {
+      "epoch": 1.2012847965738758,
+      "grad_norm": 0.2415676269852608,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3529,
+      "step": 561
+    },
+    {
+      "epoch": 1.2034261241970021,
+      "grad_norm": 0.2156997194979025,
+      "learning_rate": 3.3293650793650796e-05,
+      "loss": 0.3618,
+      "step": 562
+    },
+    {
+      "epoch": 1.2055674518201285,
+      "grad_norm": 0.24244616134298602,
+      "learning_rate": 3.3253968253968256e-05,
+      "loss": 0.3528,
+      "step": 563
+    },
+    {
+      "epoch": 1.2077087794432548,
+      "grad_norm": 0.2675996346388178,
+      "learning_rate": 3.321428571428572e-05,
+      "loss": 0.3716,
+      "step": 564
+    },
+    {
+      "epoch": 1.2098501070663812,
+      "grad_norm": 0.22977064909622621,
+      "learning_rate": 3.317460317460318e-05,
+      "loss": 0.3651,
+      "step": 565
+    },
+    {
+      "epoch": 1.2119914346895075,
+      "grad_norm": 0.22289828933891506,
+      "learning_rate": 3.313492063492064e-05,
+      "loss": 0.3578,
+      "step": 566
+    },
+    {
+      "epoch": 1.214132762312634,
+      "grad_norm": 0.24907945525188446,
+      "learning_rate": 3.309523809523809e-05,
+      "loss": 0.3803,
+      "step": 567
+    },
+    {
+      "epoch": 1.2162740899357602,
+      "grad_norm": 0.21260828983218635,
+      "learning_rate": 3.3055555555555553e-05,
+      "loss": 0.3823,
+      "step": 568
+    },
+    {
+      "epoch": 1.2184154175588866,
+      "grad_norm": 0.2444036303892165,
+      "learning_rate": 3.3015873015873014e-05,
+      "loss": 0.3873,
+      "step": 569
+    },
+    {
+      "epoch": 1.2205567451820127,
+      "grad_norm": 0.27044980470468927,
+      "learning_rate": 3.2976190476190475e-05,
+      "loss": 0.3796,
+      "step": 570
+    },
+    {
+      "epoch": 1.222698072805139,
+      "grad_norm": 0.23806387274834903,
+      "learning_rate": 3.2936507936507936e-05,
+      "loss": 0.3606,
+      "step": 571
+    },
+    {
+      "epoch": 1.2248394004282654,
+      "grad_norm": 0.2541644994496726,
+      "learning_rate": 3.2896825396825396e-05,
+      "loss": 0.3795,
+      "step": 572
+    },
+    {
+      "epoch": 1.2269807280513918,
+      "grad_norm": 0.2726575491057458,
+      "learning_rate": 3.285714285714286e-05,
+      "loss": 0.3806,
+      "step": 573
+    },
+    {
+      "epoch": 1.2291220556745182,
+      "grad_norm": 0.23105651670608096,
+      "learning_rate": 3.281746031746032e-05,
+      "loss": 0.3692,
+      "step": 574
+    },
+    {
+      "epoch": 1.2312633832976445,
+      "grad_norm": 0.23853407830325385,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 0.3701,
+      "step": 575
+    },
+    {
+      "epoch": 1.2334047109207709,
+      "grad_norm": 0.2550243654541794,
+      "learning_rate": 3.273809523809524e-05,
+      "loss": 0.3696,
+      "step": 576
+    },
+    {
+      "epoch": 1.2355460385438972,
+      "grad_norm": 0.23511204104744063,
+      "learning_rate": 3.26984126984127e-05,
+      "loss": 0.3853,
+      "step": 577
+    },
+    {
+      "epoch": 1.2376873661670236,
+      "grad_norm": 0.26214275864155384,
+      "learning_rate": 3.265873015873016e-05,
+      "loss": 0.3678,
+      "step": 578
+    },
+    {
+      "epoch": 1.23982869379015,
+      "grad_norm": 0.23149724483459302,
+      "learning_rate": 3.261904761904762e-05,
+      "loss": 0.3584,
+      "step": 579
+    },
+    {
+      "epoch": 1.2419700214132763,
+      "grad_norm": 0.2085157601025498,
+      "learning_rate": 3.257936507936508e-05,
+      "loss": 0.3497,
+      "step": 580
+    },
+    {
+      "epoch": 1.2441113490364026,
+      "grad_norm": 0.24198910705782214,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.3712,
+      "step": 581
+    },
+    {
+      "epoch": 1.246252676659529,
+      "grad_norm": 0.23194060619444556,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.3653,
+      "step": 582
+    },
+    {
+      "epoch": 1.2483940042826553,
+      "grad_norm": 0.21932550329816566,
+      "learning_rate": 3.2460317460317465e-05,
+      "loss": 0.36,
+      "step": 583
+    },
+    {
+      "epoch": 1.2505353319057817,
+      "grad_norm": 0.2485536260156264,
+      "learning_rate": 3.2420634920634925e-05,
+      "loss": 0.3557,
+      "step": 584
+    },
+    {
+      "epoch": 1.252676659528908,
+      "grad_norm": 0.21298281815390674,
+      "learning_rate": 3.2380952380952386e-05,
+      "loss": 0.3768,
+      "step": 585
+    },
+    {
+      "epoch": 1.2548179871520342,
+      "grad_norm": 0.2308383054301672,
+      "learning_rate": 3.234126984126985e-05,
+      "loss": 0.3613,
+      "step": 586
+    },
+    {
+      "epoch": 1.2569593147751605,
+      "grad_norm": 0.2500705573168694,
+      "learning_rate": 3.230158730158731e-05,
+      "loss": 0.3672,
+      "step": 587
+    },
+    {
+      "epoch": 1.259100642398287,
+      "grad_norm": 0.24932919856158853,
+      "learning_rate": 3.226190476190477e-05,
+      "loss": 0.3664,
+      "step": 588
+    },
+    {
+      "epoch": 1.2612419700214133,
+      "grad_norm": 0.21114923566401123,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 0.3654,
+      "step": 589
+    },
+    {
+      "epoch": 1.2633832976445396,
+      "grad_norm": 0.26655876159854425,
+      "learning_rate": 3.218253968253968e-05,
+      "loss": 0.376,
+      "step": 590
+    },
+    {
+      "epoch": 1.265524625267666,
+      "grad_norm": 0.22494122561905386,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.366,
+      "step": 591
+    },
+    {
+      "epoch": 1.2676659528907923,
+      "grad_norm": 0.24850515269385678,
+      "learning_rate": 3.2103174603174605e-05,
+      "loss": 0.3814,
+      "step": 592
+    },
+    {
+      "epoch": 1.2698072805139187,
+      "grad_norm": 0.2359476164199983,
+      "learning_rate": 3.2063492063492065e-05,
+      "loss": 0.3876,
+      "step": 593
+    },
+    {
+      "epoch": 1.271948608137045,
+      "grad_norm": 0.2170060923253427,
+      "learning_rate": 3.202380952380952e-05,
+      "loss": 0.3662,
+      "step": 594
+    },
+    {
+      "epoch": 1.2740899357601714,
+      "grad_norm": 0.2504062463730666,
+      "learning_rate": 3.198412698412698e-05,
+      "loss": 0.3753,
+      "step": 595
+    },
+    {
+      "epoch": 1.2762312633832975,
+      "grad_norm": 0.20106817241986427,
+      "learning_rate": 3.194444444444444e-05,
+      "loss": 0.3592,
+      "step": 596
+    },
+    {
+      "epoch": 1.2783725910064239,
+      "grad_norm": 0.21200714791632813,
+      "learning_rate": 3.19047619047619e-05,
+      "loss": 0.3509,
+      "step": 597
+    },
+    {
+      "epoch": 1.2805139186295502,
+      "grad_norm": 0.21778732718260096,
+      "learning_rate": 3.186507936507936e-05,
+      "loss": 0.3584,
+      "step": 598
+    },
+    {
+      "epoch": 1.2826552462526766,
+      "grad_norm": 0.2189934520948945,
+      "learning_rate": 3.182539682539682e-05,
+      "loss": 0.3549,
+      "step": 599
+    },
+    {
+      "epoch": 1.284796573875803,
+      "grad_norm": 0.2208172986950504,
+      "learning_rate": 3.1785714285714284e-05,
+      "loss": 0.3571,
+      "step": 600
+    },
+    {
+      "epoch": 1.2869379014989293,
+      "grad_norm": 0.22576417970064774,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.3525,
+      "step": 601
+    },
+    {
+      "epoch": 1.2890792291220556,
+      "grad_norm": 0.27692988742074415,
+      "learning_rate": 3.1706349206349205e-05,
+      "loss": 0.3764,
+      "step": 602
+    },
+    {
+      "epoch": 1.291220556745182,
+      "grad_norm": 0.21396523695563535,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.3733,
+      "step": 603
+    },
+    {
+      "epoch": 1.2933618843683083,
+      "grad_norm": 0.23416094155920952,
+      "learning_rate": 3.162698412698413e-05,
+      "loss": 0.3559,
+      "step": 604
+    },
+    {
+      "epoch": 1.2955032119914347,
+      "grad_norm": 0.2486860214774819,
+      "learning_rate": 3.158730158730159e-05,
+      "loss": 0.3562,
+      "step": 605
+    },
+    {
+      "epoch": 1.297644539614561,
+      "grad_norm": 0.2201774327190655,
+      "learning_rate": 3.154761904761905e-05,
+      "loss": 0.3716,
+      "step": 606
+    },
+    {
+      "epoch": 1.2997858672376874,
+      "grad_norm": 0.1964840066863156,
+      "learning_rate": 3.150793650793651e-05,
+      "loss": 0.3542,
+      "step": 607
+    },
+    {
+      "epoch": 1.3019271948608138,
+      "grad_norm": 0.22767951673276937,
+      "learning_rate": 3.146825396825397e-05,
+      "loss": 0.3612,
+      "step": 608
+    },
+    {
+      "epoch": 1.3040685224839401,
+      "grad_norm": 0.22557177823924723,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.3603,
+      "step": 609
+    },
+    {
+      "epoch": 1.3062098501070665,
+      "grad_norm": 0.27527013455606647,
+      "learning_rate": 3.138888888888889e-05,
+      "loss": 0.3502,
+      "step": 610
+    },
+    {
+      "epoch": 1.3083511777301928,
+      "grad_norm": 0.23325579262107884,
+      "learning_rate": 3.134920634920635e-05,
+      "loss": 0.3631,
+      "step": 611
+    },
+    {
+      "epoch": 1.3104925053533192,
+      "grad_norm": 0.22097147014315746,
+      "learning_rate": 3.130952380952381e-05,
+      "loss": 0.3511,
+      "step": 612
+    },
+    {
+      "epoch": 1.3126338329764453,
+      "grad_norm": 0.2174864072196349,
+      "learning_rate": 3.1269841269841274e-05,
+      "loss": 0.3672,
+      "step": 613
+    },
+    {
+      "epoch": 1.3147751605995717,
+      "grad_norm": 0.2106746804771801,
+      "learning_rate": 3.1230158730158734e-05,
+      "loss": 0.3612,
+      "step": 614
+    },
+    {
+      "epoch": 1.316916488222698,
+      "grad_norm": 0.21088050552854007,
+      "learning_rate": 3.1190476190476195e-05,
+      "loss": 0.3471,
+      "step": 615
+    },
+    {
+      "epoch": 1.3190578158458244,
+      "grad_norm": 0.2580125309103289,
+      "learning_rate": 3.1150793650793656e-05,
+      "loss": 0.3741,
+      "step": 616
+    },
+    {
+      "epoch": 1.3211991434689507,
+      "grad_norm": 0.21505898974697335,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.3763,
+      "step": 617
+    },
+    {
+      "epoch": 1.323340471092077,
+      "grad_norm": 0.2465748563264946,
+      "learning_rate": 3.107142857142857e-05,
+      "loss": 0.3766,
+      "step": 618
+    },
+    {
+      "epoch": 1.3254817987152034,
+      "grad_norm": 0.3321013586284405,
+      "learning_rate": 3.103174603174603e-05,
+      "loss": 0.3955,
+      "step": 619
+    },
+    {
+      "epoch": 1.3276231263383298,
+      "grad_norm": 0.22495906786878164,
+      "learning_rate": 3.099206349206349e-05,
+      "loss": 0.3718,
+      "step": 620
+    },
+    {
+      "epoch": 1.3297644539614561,
+      "grad_norm": 0.23719074429579814,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.3544,
+      "step": 621
+    },
+    {
+      "epoch": 1.3319057815845825,
+      "grad_norm": 0.23831973310054141,
+      "learning_rate": 3.0912698412698414e-05,
+      "loss": 0.3795,
+      "step": 622
+    },
+    {
+      "epoch": 1.3340471092077089,
+      "grad_norm": 0.2128688128559845,
+      "learning_rate": 3.0873015873015874e-05,
+      "loss": 0.3668,
+      "step": 623
+    },
+    {
+      "epoch": 1.336188436830835,
+      "grad_norm": 0.22961487193589208,
+      "learning_rate": 3.0833333333333335e-05,
+      "loss": 0.366,
+      "step": 624
+    },
+    {
+      "epoch": 1.3383297644539613,
+      "grad_norm": 0.2214674128035989,
+      "learning_rate": 3.0793650793650796e-05,
+      "loss": 0.365,
+      "step": 625
+    },
+    {
+      "epoch": 1.3404710920770877,
+      "grad_norm": 0.20498199401783132,
+      "learning_rate": 3.075396825396826e-05,
+      "loss": 0.3556,
+      "step": 626
+    },
+    {
+      "epoch": 1.342612419700214,
+      "grad_norm": 0.22359296981746604,
+      "learning_rate": 3.071428571428572e-05,
+      "loss": 0.3812,
+      "step": 627
+    },
+    {
+      "epoch": 1.3447537473233404,
+      "grad_norm": 0.21653621158585087,
+      "learning_rate": 3.067460317460318e-05,
+      "loss": 0.3797,
+      "step": 628
+    },
+    {
+      "epoch": 1.3468950749464668,
+      "grad_norm": 0.22549871106519903,
+      "learning_rate": 3.063492063492064e-05,
+      "loss": 0.3596,
+      "step": 629
+    },
+    {
+      "epoch": 1.3490364025695931,
+      "grad_norm": 0.20475922109140837,
+      "learning_rate": 3.05952380952381e-05,
+      "loss": 0.3735,
+      "step": 630
+    },
+    {
+      "epoch": 1.3511777301927195,
+      "grad_norm": 0.209763601459537,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.3709,
+      "step": 631
+    },
+    {
+      "epoch": 1.3533190578158458,
+      "grad_norm": 0.2176748492097036,
+      "learning_rate": 3.051587301587302e-05,
+      "loss": 0.3808,
+      "step": 632
+    },
+    {
+      "epoch": 1.3554603854389722,
+      "grad_norm": 0.21622906752623547,
+      "learning_rate": 3.0476190476190482e-05,
+      "loss": 0.3604,
+      "step": 633
+    },
+    {
+      "epoch": 1.3576017130620985,
+      "grad_norm": 0.20063962917481207,
+      "learning_rate": 3.0436507936507936e-05,
+      "loss": 0.3546,
+      "step": 634
+    },
+    {
+      "epoch": 1.359743040685225,
+      "grad_norm": 0.20999519152589322,
+      "learning_rate": 3.0396825396825397e-05,
+      "loss": 0.3597,
+      "step": 635
+    },
+    {
+      "epoch": 1.3618843683083512,
+      "grad_norm": 0.20927897537878964,
+      "learning_rate": 3.0357142857142857e-05,
+      "loss": 0.3729,
+      "step": 636
+    },
+    {
+      "epoch": 1.3640256959314776,
+      "grad_norm": 0.21686713042567732,
+      "learning_rate": 3.0317460317460318e-05,
+      "loss": 0.3653,
+      "step": 637
+    },
+    {
+      "epoch": 1.366167023554604,
+      "grad_norm": 0.21754952121551016,
+      "learning_rate": 3.0277777777777776e-05,
+      "loss": 0.3677,
+      "step": 638
+    },
+    {
+      "epoch": 1.3683083511777303,
+      "grad_norm": 0.20450902968020848,
+      "learning_rate": 3.0238095238095236e-05,
+      "loss": 0.3355,
+      "step": 639
+    },
+    {
+      "epoch": 1.3704496788008567,
+      "grad_norm": 0.21623695248881408,
+      "learning_rate": 3.0198412698412697e-05,
+      "loss": 0.3806,
+      "step": 640
+    },
+    {
+      "epoch": 1.3725910064239828,
+      "grad_norm": 0.21537383605412733,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 0.3681,
+      "step": 641
+    },
+    {
+      "epoch": 1.3747323340471092,
+      "grad_norm": 0.19837937617514323,
+      "learning_rate": 3.011904761904762e-05,
+      "loss": 0.3622,
+      "step": 642
+    },
+    {
+      "epoch": 1.3768736616702355,
+      "grad_norm": 0.209108602319792,
+      "learning_rate": 3.007936507936508e-05,
+      "loss": 0.3484,
+      "step": 643
+    },
+    {
+      "epoch": 1.3790149892933619,
+      "grad_norm": 0.2062194560604121,
+      "learning_rate": 3.003968253968254e-05,
+      "loss": 0.3702,
+      "step": 644
+    },
+    {
+      "epoch": 1.3811563169164882,
+      "grad_norm": 0.2080771057237384,
+      "learning_rate": 3e-05,
+      "loss": 0.362,
+      "step": 645
+    },
+    {
+      "epoch": 1.3832976445396146,
+      "grad_norm": 0.2227326233506996,
+      "learning_rate": 2.996031746031746e-05,
+      "loss": 0.3701,
+      "step": 646
+    },
+    {
+      "epoch": 1.385438972162741,
+      "grad_norm": 0.2089744724523155,
+      "learning_rate": 2.9920634920634922e-05,
+      "loss": 0.361,
+      "step": 647
+    },
+    {
+      "epoch": 1.3875802997858673,
+      "grad_norm": 0.25404638727035117,
+      "learning_rate": 2.9880952380952383e-05,
+      "loss": 0.3629,
+      "step": 648
+    },
+    {
+      "epoch": 1.3897216274089936,
+      "grad_norm": 0.21442147488950894,
+      "learning_rate": 2.9841269841269844e-05,
+      "loss": 0.3513,
+      "step": 649
+    },
+    {
+      "epoch": 1.39186295503212,
+      "grad_norm": 0.2481654405500989,
+      "learning_rate": 2.98015873015873e-05,
+      "loss": 0.3662,
+      "step": 650
+    },
+    {
+      "epoch": 1.3940042826552461,
+      "grad_norm": 0.22382398124617614,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 0.3687,
+      "step": 651
+    },
+    {
+      "epoch": 1.3961456102783725,
+      "grad_norm": 0.31356212838373204,
+      "learning_rate": 2.9722222222222223e-05,
+      "loss": 0.3795,
+      "step": 652
+    },
+    {
+      "epoch": 1.3982869379014988,
+      "grad_norm": 0.2429332621043318,
+      "learning_rate": 2.9682539682539683e-05,
+      "loss": 0.373,
+      "step": 653
+    },
+    {
+      "epoch": 1.4004282655246252,
+      "grad_norm": 0.2249979699253431,
+      "learning_rate": 2.9642857142857144e-05,
+      "loss": 0.3676,
+      "step": 654
+    },
+    {
+      "epoch": 1.4025695931477515,
+      "grad_norm": 0.2560531887036681,
+      "learning_rate": 2.9603174603174605e-05,
+      "loss": 0.3637,
+      "step": 655
+    },
+    {
+      "epoch": 1.404710920770878,
+      "grad_norm": 0.24116508821199556,
+      "learning_rate": 2.9563492063492066e-05,
+      "loss": 0.3707,
+      "step": 656
+    },
+    {
+      "epoch": 1.4068522483940042,
+      "grad_norm": 0.2598274205324474,
+      "learning_rate": 2.9523809523809526e-05,
+      "loss": 0.379,
+      "step": 657
+    },
+    {
+      "epoch": 1.4089935760171306,
+      "grad_norm": 0.21575521660148642,
+      "learning_rate": 2.9484126984126987e-05,
+      "loss": 0.376,
+      "step": 658
+    },
+    {
+      "epoch": 1.411134903640257,
+      "grad_norm": 0.22041851374967822,
+      "learning_rate": 2.9444444444444448e-05,
+      "loss": 0.3607,
+      "step": 659
+    },
+    {
+      "epoch": 1.4132762312633833,
+      "grad_norm": 0.2564088225056767,
+      "learning_rate": 2.940476190476191e-05,
+      "loss": 0.3599,
+      "step": 660
+    },
+    {
+      "epoch": 1.4154175588865097,
+      "grad_norm": 0.20549406527051656,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.3657,
+      "step": 661
+    },
+    {
+      "epoch": 1.417558886509636,
+      "grad_norm": 0.2540118149516082,
+      "learning_rate": 2.9325396825396827e-05,
+      "loss": 0.368,
+      "step": 662
+    },
+    {
+      "epoch": 1.4197002141327624,
+      "grad_norm": 0.2648103833326966,
+      "learning_rate": 2.9285714285714288e-05,
+      "loss": 0.3689,
+      "step": 663
+    },
+    {
+      "epoch": 1.4218415417558887,
+      "grad_norm": 0.24077539771530324,
+      "learning_rate": 2.9246031746031748e-05,
+      "loss": 0.351,
+      "step": 664
+    },
+    {
+      "epoch": 1.423982869379015,
+      "grad_norm": 0.24569673366814643,
+      "learning_rate": 2.920634920634921e-05,
+      "loss": 0.3595,
+      "step": 665
+    },
+    {
+      "epoch": 1.4261241970021414,
+      "grad_norm": 0.22436619156977872,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.3636,
+      "step": 666
+    },
+    {
+      "epoch": 1.4282655246252678,
+      "grad_norm": 0.22702096393924773,
+      "learning_rate": 2.912698412698413e-05,
+      "loss": 0.3577,
+      "step": 667
+    },
+    {
+      "epoch": 1.430406852248394,
+      "grad_norm": 0.21223624922722298,
+      "learning_rate": 2.908730158730159e-05,
+      "loss": 0.3677,
+      "step": 668
+    },
+    {
+      "epoch": 1.4325481798715203,
+      "grad_norm": 0.2474621671117468,
+      "learning_rate": 2.9047619047619052e-05,
+      "loss": 0.3632,
+      "step": 669
+    },
+    {
+      "epoch": 1.4346895074946466,
+      "grad_norm": 0.25684655588636496,
+      "learning_rate": 2.9007936507936513e-05,
+      "loss": 0.3592,
+      "step": 670
+    },
+    {
+      "epoch": 1.436830835117773,
+      "grad_norm": 0.25072866337855965,
+      "learning_rate": 2.8968253968253974e-05,
+      "loss": 0.3598,
+      "step": 671
+    },
+    {
+      "epoch": 1.4389721627408993,
+      "grad_norm": 0.22582476526482667,
+      "learning_rate": 2.8928571428571434e-05,
+      "loss": 0.3577,
+      "step": 672
+    },
+    {
+      "epoch": 1.4411134903640257,
+      "grad_norm": 0.28658549817719403,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.3839,
+      "step": 673
+    },
+    {
+      "epoch": 1.443254817987152,
+      "grad_norm": 0.25430283807134485,
+      "learning_rate": 2.884920634920635e-05,
+      "loss": 0.3643,
+      "step": 674
+    },
+    {
+      "epoch": 1.4453961456102784,
+      "grad_norm": 0.22951249670225976,
+      "learning_rate": 2.880952380952381e-05,
+      "loss": 0.3692,
+      "step": 675
+    },
+    {
+      "epoch": 1.4475374732334048,
+      "grad_norm": 0.23388364802048373,
+      "learning_rate": 2.876984126984127e-05,
+      "loss": 0.3708,
+      "step": 676
+    },
+    {
+      "epoch": 1.4496788008565311,
+      "grad_norm": 0.21976733689801523,
+      "learning_rate": 2.8730158730158728e-05,
+      "loss": 0.3802,
+      "step": 677
+    },
+    {
+      "epoch": 1.4518201284796575,
+      "grad_norm": 0.21027240355399024,
+      "learning_rate": 2.869047619047619e-05,
+      "loss": 0.3773,
+      "step": 678
+    },
+    {
+      "epoch": 1.4539614561027836,
+      "grad_norm": 0.20199987883594428,
+      "learning_rate": 2.865079365079365e-05,
+      "loss": 0.3637,
+      "step": 679
+    },
+    {
+      "epoch": 1.45610278372591,
+      "grad_norm": 0.21923422597696604,
+      "learning_rate": 2.861111111111111e-05,
+      "loss": 0.3619,
+      "step": 680
+    },
+    {
+      "epoch": 1.4582441113490363,
+      "grad_norm": 0.21000907335559735,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.3536,
+      "step": 681
+    },
+    {
+      "epoch": 1.4603854389721627,
+      "grad_norm": 0.21711620369351062,
+      "learning_rate": 2.853174603174603e-05,
+      "loss": 0.3738,
+      "step": 682
+    },
+    {
+      "epoch": 1.462526766595289,
+      "grad_norm": 0.22565854928185133,
+      "learning_rate": 2.8492063492063492e-05,
+      "loss": 0.3608,
+      "step": 683
+    },
+    {
+      "epoch": 1.4646680942184154,
+      "grad_norm": 0.24796414653671942,
+      "learning_rate": 2.8452380952380953e-05,
+      "loss": 0.3859,
+      "step": 684
+    },
+    {
+      "epoch": 1.4668094218415417,
+      "grad_norm": 0.19936447407500923,
+      "learning_rate": 2.8412698412698414e-05,
+      "loss": 0.3772,
+      "step": 685
+    },
+    {
+      "epoch": 1.468950749464668,
+      "grad_norm": 0.2659759151283577,
+      "learning_rate": 2.8373015873015875e-05,
+      "loss": 0.3754,
+      "step": 686
+    },
+    {
+      "epoch": 1.4710920770877944,
+      "grad_norm": 0.2142679975770448,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.3705,
+      "step": 687
+    },
+    {
+      "epoch": 1.4732334047109208,
+      "grad_norm": 0.21126669286906707,
+      "learning_rate": 2.8293650793650793e-05,
+      "loss": 0.3831,
+      "step": 688
+    },
+    {
+      "epoch": 1.4753747323340471,
+      "grad_norm": 0.24217810422670202,
+      "learning_rate": 2.8253968253968253e-05,
+      "loss": 0.3549,
+      "step": 689
+    },
+    {
+      "epoch": 1.4775160599571735,
+      "grad_norm": 0.21568420422411547,
+      "learning_rate": 2.8214285714285714e-05,
+      "loss": 0.3693,
+      "step": 690
+    },
+    {
+      "epoch": 1.4796573875802999,
+      "grad_norm": 0.2312299790821551,
+      "learning_rate": 2.8174603174603175e-05,
+      "loss": 0.3764,
+      "step": 691
+    },
+    {
+      "epoch": 1.4817987152034262,
+      "grad_norm": 0.233754289817078,
+      "learning_rate": 2.8134920634920636e-05,
+      "loss": 0.3805,
+      "step": 692
+    },
+    {
+      "epoch": 1.4839400428265526,
+      "grad_norm": 0.2423944727011651,
+      "learning_rate": 2.8095238095238096e-05,
+      "loss": 0.3636,
+      "step": 693
+    },
+    {
+      "epoch": 1.486081370449679,
+      "grad_norm": 0.22905460391929347,
+      "learning_rate": 2.8055555555555557e-05,
+      "loss": 0.3895,
+      "step": 694
+    },
+    {
+      "epoch": 1.4882226980728053,
+      "grad_norm": 0.23214001332154588,
+      "learning_rate": 2.8015873015873018e-05,
+      "loss": 0.3605,
+      "step": 695
+    },
+    {
+      "epoch": 1.4903640256959314,
+      "grad_norm": 0.19690116100293087,
+      "learning_rate": 2.797619047619048e-05,
+      "loss": 0.3582,
+      "step": 696
+    },
+    {
+      "epoch": 1.4925053533190578,
+      "grad_norm": 0.24573029483590722,
+      "learning_rate": 2.793650793650794e-05,
+      "loss": 0.3712,
+      "step": 697
+    },
+    {
+      "epoch": 1.4946466809421841,
+      "grad_norm": 0.19231425648369746,
+      "learning_rate": 2.78968253968254e-05,
+      "loss": 0.3497,
+      "step": 698
+    },
+    {
+      "epoch": 1.4967880085653105,
+      "grad_norm": 0.22612270059401912,
+      "learning_rate": 2.785714285714286e-05,
+      "loss": 0.3417,
+      "step": 699
+    },
+    {
+      "epoch": 1.4989293361884368,
+      "grad_norm": 0.22960092397135454,
+      "learning_rate": 2.781746031746032e-05,
+      "loss": 0.3575,
+      "step": 700
+    },
+    {
+      "epoch": 1.5010706638115632,
+      "grad_norm": 0.23684421828804095,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3786,
+      "step": 701
+    },
+    {
+      "epoch": 1.5032119914346895,
+      "grad_norm": 0.2191044054358468,
+      "learning_rate": 2.773809523809524e-05,
+      "loss": 0.3549,
+      "step": 702
+    },
+    {
+      "epoch": 1.5053533190578159,
+      "grad_norm": 0.24248023653906947,
+      "learning_rate": 2.76984126984127e-05,
+      "loss": 0.3558,
+      "step": 703
+    },
+    {
+      "epoch": 1.507494646680942,
+      "grad_norm": 0.24648400748932908,
+      "learning_rate": 2.765873015873016e-05,
+      "loss": 0.3657,
+      "step": 704
+    },
+    {
+      "epoch": 1.5096359743040684,
+      "grad_norm": 0.23645047152874282,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.3704,
+      "step": 705
+    },
+    {
+      "epoch": 1.5117773019271947,
+      "grad_norm": 0.25487174415303726,
+      "learning_rate": 2.7579365079365083e-05,
+      "loss": 0.3582,
+      "step": 706
+    },
+    {
+      "epoch": 1.513918629550321,
+      "grad_norm": 0.2331120266364236,
+      "learning_rate": 2.7539682539682544e-05,
+      "loss": 0.3699,
+      "step": 707
+    },
+    {
+      "epoch": 1.5160599571734474,
+      "grad_norm": 0.2181485815951784,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.3751,
+      "step": 708
+    },
+    {
+      "epoch": 1.5182012847965738,
+      "grad_norm": 0.24072328359741854,
+      "learning_rate": 2.7460317460317465e-05,
+      "loss": 0.3775,
+      "step": 709
+    },
+    {
+      "epoch": 1.5203426124197001,
+      "grad_norm": 0.22869430663467058,
+      "learning_rate": 2.7420634920634926e-05,
+      "loss": 0.3724,
+      "step": 710
+    },
+    {
+      "epoch": 1.5224839400428265,
+      "grad_norm": 0.21918246335255903,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 0.3547,
+      "step": 711
+    },
+    {
+      "epoch": 1.5246252676659529,
+      "grad_norm": 0.2189822768505496,
+      "learning_rate": 2.734126984126984e-05,
+      "loss": 0.3535,
+      "step": 712
+    },
+    {
+      "epoch": 1.5267665952890792,
+      "grad_norm": 0.2526005237695246,
+      "learning_rate": 2.73015873015873e-05,
+      "loss": 0.3713,
+      "step": 713
+    },
+    {
+      "epoch": 1.5289079229122056,
+      "grad_norm": 0.22601185946675514,
+      "learning_rate": 2.7261904761904762e-05,
+      "loss": 0.3675,
+      "step": 714
+    },
+    {
+      "epoch": 1.531049250535332,
+      "grad_norm": 0.20669359644643262,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 0.3659,
+      "step": 715
+    },
+    {
+      "epoch": 1.5331905781584583,
+      "grad_norm": 0.24031352892120264,
+      "learning_rate": 2.718253968253968e-05,
+      "loss": 0.3602,
+      "step": 716
+    },
+    {
+      "epoch": 1.5353319057815846,
+      "grad_norm": 0.22795507587171882,
+      "learning_rate": 2.714285714285714e-05,
+      "loss": 0.3709,
+      "step": 717
+    },
+    {
+      "epoch": 1.537473233404711,
+      "grad_norm": 0.20536455549527732,
+      "learning_rate": 2.7103174603174602e-05,
+      "loss": 0.3799,
+      "step": 718
+    },
+    {
+      "epoch": 1.5396145610278373,
+      "grad_norm": 0.2943404251055241,
+      "learning_rate": 2.7063492063492062e-05,
+      "loss": 0.3836,
+      "step": 719
+    },
+    {
+      "epoch": 1.5417558886509637,
+      "grad_norm": 0.22205668356720148,
+      "learning_rate": 2.7023809523809523e-05,
+      "loss": 0.3664,
+      "step": 720
+    },
+    {
+      "epoch": 1.54389721627409,
+      "grad_norm": 0.21954728948498192,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.3647,
+      "step": 721
+    },
+    {
+      "epoch": 1.5460385438972164,
+      "grad_norm": 0.2280345060624478,
+      "learning_rate": 2.6944444444444445e-05,
+      "loss": 0.3713,
+      "step": 722
+    },
+    {
+      "epoch": 1.5481798715203428,
+      "grad_norm": 0.22861539448823806,
+      "learning_rate": 2.6904761904761905e-05,
+      "loss": 0.3888,
+      "step": 723
+    },
+    {
+      "epoch": 1.550321199143469,
+      "grad_norm": 0.19425559114407048,
+      "learning_rate": 2.6865079365079366e-05,
+      "loss": 0.3739,
+      "step": 724
+    },
+    {
+      "epoch": 1.5524625267665952,
+      "grad_norm": 0.24288323051735908,
+      "learning_rate": 2.6825396825396827e-05,
+      "loss": 0.355,
+      "step": 725
+    },
+    {
+      "epoch": 1.5546038543897216,
+      "grad_norm": 0.22443141683817264,
+      "learning_rate": 2.6785714285714288e-05,
+      "loss": 0.3888,
+      "step": 726
+    },
+    {
+      "epoch": 1.556745182012848,
+      "grad_norm": 0.2182399943573036,
+      "learning_rate": 2.6746031746031745e-05,
+      "loss": 0.3624,
+      "step": 727
+    },
+    {
+      "epoch": 1.5588865096359743,
+      "grad_norm": 0.23168062600285244,
+      "learning_rate": 2.6706349206349206e-05,
+      "loss": 0.3585,
+      "step": 728
+    },
+    {
+      "epoch": 1.5610278372591007,
+      "grad_norm": 0.21591657004920606,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.3543,
+      "step": 729
+    },
+    {
+      "epoch": 1.563169164882227,
+      "grad_norm": 0.21509414063611806,
+      "learning_rate": 2.6626984126984127e-05,
+      "loss": 0.369,
+      "step": 730
+    },
+    {
+      "epoch": 1.5653104925053534,
+      "grad_norm": 0.21543923122674843,
+      "learning_rate": 2.6587301587301588e-05,
+      "loss": 0.3582,
+      "step": 731
+    },
+    {
+      "epoch": 1.5674518201284795,
+      "grad_norm": 0.20613757156954682,
+      "learning_rate": 2.654761904761905e-05,
+      "loss": 0.3564,
+      "step": 732
+    },
+    {
+      "epoch": 1.5695931477516059,
+      "grad_norm": 0.2068022698111163,
+      "learning_rate": 2.650793650793651e-05,
+      "loss": 0.3714,
+      "step": 733
+    },
+    {
+      "epoch": 1.5717344753747322,
+      "grad_norm": 0.19705404333823645,
+      "learning_rate": 2.646825396825397e-05,
+      "loss": 0.3648,
+      "step": 734
+    },
+    {
+      "epoch": 1.5738758029978586,
+      "grad_norm": 0.20492416067065145,
+      "learning_rate": 2.642857142857143e-05,
+      "loss": 0.36,
+      "step": 735
+    },
+    {
+      "epoch": 1.576017130620985,
+      "grad_norm": 0.20508644827871644,
+      "learning_rate": 2.6388888888888892e-05,
+      "loss": 0.3874,
+      "step": 736
+    },
+    {
+      "epoch": 1.5781584582441113,
+      "grad_norm": 0.20727086230283992,
+      "learning_rate": 2.6349206349206353e-05,
+      "loss": 0.3641,
+      "step": 737
+    },
+    {
+      "epoch": 1.5802997858672376,
+      "grad_norm": 0.21066080929705655,
+      "learning_rate": 2.6309523809523813e-05,
+      "loss": 0.3775,
+      "step": 738
+    },
+    {
+      "epoch": 1.582441113490364,
+      "grad_norm": 0.1858511474413401,
+      "learning_rate": 2.626984126984127e-05,
+      "loss": 0.3493,
+      "step": 739
+    },
+    {
+      "epoch": 1.5845824411134903,
+      "grad_norm": 0.1948727545134536,
+      "learning_rate": 2.623015873015873e-05,
+      "loss": 0.3604,
+      "step": 740
+    },
+    {
+      "epoch": 1.5867237687366167,
+      "grad_norm": 0.18377772006229867,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.3656,
+      "step": 741
+    },
+    {
+      "epoch": 1.588865096359743,
+      "grad_norm": 0.20759171733907542,
+      "learning_rate": 2.6150793650793653e-05,
+      "loss": 0.36,
+      "step": 742
+    },
+    {
+      "epoch": 1.5910064239828694,
+      "grad_norm": 0.20021552648301658,
+      "learning_rate": 2.6111111111111114e-05,
+      "loss": 0.3672,
+      "step": 743
+    },
+    {
+      "epoch": 1.5931477516059958,
+      "grad_norm": 0.1987003787234111,
+      "learning_rate": 2.6071428571428574e-05,
+      "loss": 0.3697,
+      "step": 744
+    },
+    {
+      "epoch": 1.595289079229122,
+      "grad_norm": 0.1992569458099437,
+      "learning_rate": 2.6031746031746035e-05,
+      "loss": 0.3722,
+      "step": 745
+    },
+    {
+      "epoch": 1.5974304068522485,
+      "grad_norm": 0.19490857264680375,
+      "learning_rate": 2.5992063492063496e-05,
+      "loss": 0.3487,
+      "step": 746
+    },
+    {
+      "epoch": 1.5995717344753748,
+      "grad_norm": 0.20992568333120643,
+      "learning_rate": 2.5952380952380957e-05,
+      "loss": 0.3721,
+      "step": 747
+    },
+    {
+      "epoch": 1.6017130620985012,
+      "grad_norm": 0.20903627954075446,
+      "learning_rate": 2.5912698412698417e-05,
+      "loss": 0.3645,
+      "step": 748
+    },
+    {
+      "epoch": 1.6038543897216275,
+      "grad_norm": 0.2206708938047425,
+      "learning_rate": 2.5873015873015878e-05,
+      "loss": 0.3596,
+      "step": 749
+    },
+    {
+      "epoch": 1.6059957173447539,
+      "grad_norm": 0.2315828550558367,
+      "learning_rate": 2.5833333333333336e-05,
+      "loss": 0.3833,
+      "step": 750
+    },
+    {
+      "epoch": 1.6081370449678802,
+      "grad_norm": 0.20729043706506806,
+      "learning_rate": 2.5793650793650796e-05,
+      "loss": 0.3676,
+      "step": 751
+    },
+    {
+      "epoch": 1.6102783725910066,
+      "grad_norm": 0.21304496616203045,
+      "learning_rate": 2.5753968253968254e-05,
+      "loss": 0.3692,
+      "step": 752
+    },
+    {
+      "epoch": 1.6124197002141327,
+      "grad_norm": 0.2066888063019116,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.3513,
+      "step": 753
+    },
+    {
+      "epoch": 1.614561027837259,
+      "grad_norm": 0.20035669627587474,
+      "learning_rate": 2.5674603174603172e-05,
+      "loss": 0.3625,
+      "step": 754
+    },
+    {
+      "epoch": 1.6167023554603854,
+      "grad_norm": 0.21327760276750352,
+      "learning_rate": 2.5634920634920633e-05,
+      "loss": 0.3642,
+      "step": 755
+    },
+    {
+      "epoch": 1.6188436830835118,
+      "grad_norm": 0.2144247994926133,
+      "learning_rate": 2.5595238095238093e-05,
+      "loss": 0.378,
+      "step": 756
+    },
+    {
+      "epoch": 1.6209850107066381,
+      "grad_norm": 0.21826292313986448,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.3682,
+      "step": 757
+    },
+    {
+      "epoch": 1.6231263383297645,
+      "grad_norm": 0.2112350938263051,
+      "learning_rate": 2.5515873015873015e-05,
+      "loss": 0.3737,
+      "step": 758
+    },
+    {
+      "epoch": 1.6252676659528906,
+      "grad_norm": 0.22455328732151775,
+      "learning_rate": 2.5476190476190476e-05,
+      "loss": 0.3654,
+      "step": 759
+    },
+    {
+      "epoch": 1.627408993576017,
+      "grad_norm": 0.18501213734273322,
+      "learning_rate": 2.5436507936507936e-05,
+      "loss": 0.3383,
+      "step": 760
+    },
+    {
+      "epoch": 1.6295503211991433,
+      "grad_norm": 0.20942821521960064,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.3583,
+      "step": 761
+    },
+    {
+      "epoch": 1.6316916488222697,
+      "grad_norm": 0.22203677151968088,
+      "learning_rate": 2.5357142857142858e-05,
+      "loss": 0.37,
+      "step": 762
+    },
+    {
+      "epoch": 1.633832976445396,
+      "grad_norm": 0.22074729844028526,
+      "learning_rate": 2.531746031746032e-05,
+      "loss": 0.3555,
+      "step": 763
+    },
+    {
+      "epoch": 1.6359743040685224,
+      "grad_norm": 0.24745475269473763,
+      "learning_rate": 2.527777777777778e-05,
+      "loss": 0.3913,
+      "step": 764
+    },
+    {
+      "epoch": 1.6381156316916488,
+      "grad_norm": 0.21980493117452848,
+      "learning_rate": 2.523809523809524e-05,
+      "loss": 0.3684,
+      "step": 765
+    },
+    {
+      "epoch": 1.640256959314775,
+      "grad_norm": 0.21850689842540688,
+      "learning_rate": 2.5198412698412697e-05,
+      "loss": 0.3727,
+      "step": 766
+    },
+    {
+      "epoch": 1.6423982869379015,
+      "grad_norm": 0.21446663216573078,
+      "learning_rate": 2.5158730158730158e-05,
+      "loss": 0.3782,
+      "step": 767
+    },
+    {
+      "epoch": 1.6445396145610278,
+      "grad_norm": 0.21778149384624332,
+      "learning_rate": 2.511904761904762e-05,
+      "loss": 0.3683,
+      "step": 768
+    },
+    {
+      "epoch": 1.6466809421841542,
+      "grad_norm": 0.20468677302613195,
+      "learning_rate": 2.507936507936508e-05,
+      "loss": 0.3771,
+      "step": 769
+    },
+    {
+      "epoch": 1.6488222698072805,
+      "grad_norm": 0.23332787991529938,
+      "learning_rate": 2.503968253968254e-05,
+      "loss": 0.367,
+      "step": 770
+    },
+    {
+      "epoch": 1.6509635974304069,
+      "grad_norm": 0.21897816586893398,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3783,
+      "step": 771
+    },
+    {
+      "epoch": 1.6531049250535332,
+      "grad_norm": 0.2030029868201131,
+      "learning_rate": 2.4960317460317462e-05,
+      "loss": 0.3745,
+      "step": 772
+    },
+    {
+      "epoch": 1.6552462526766596,
+      "grad_norm": 0.21310414937938124,
+      "learning_rate": 2.4920634920634923e-05,
+      "loss": 0.3405,
+      "step": 773
+    },
+    {
+      "epoch": 1.657387580299786,
+      "grad_norm": 0.22195298457726914,
+      "learning_rate": 2.4880952380952383e-05,
+      "loss": 0.3874,
+      "step": 774
+    },
+    {
+      "epoch": 1.6595289079229123,
+      "grad_norm": 0.20300473974521194,
+      "learning_rate": 2.4841269841269844e-05,
+      "loss": 0.3663,
+      "step": 775
+    },
+    {
+      "epoch": 1.6616702355460387,
+      "grad_norm": 0.22067636878573646,
+      "learning_rate": 2.4801587301587305e-05,
+      "loss": 0.3887,
+      "step": 776
+    },
+    {
+      "epoch": 1.663811563169165,
+      "grad_norm": 0.21073789338832113,
+      "learning_rate": 2.4761904761904762e-05,
+      "loss": 0.3743,
+      "step": 777
+    },
+    {
+      "epoch": 1.6659528907922914,
+      "grad_norm": 0.22786185959131758,
+      "learning_rate": 2.4722222222222223e-05,
+      "loss": 0.3743,
+      "step": 778
+    },
+    {
+      "epoch": 1.6680942184154177,
+      "grad_norm": 0.20200585214379246,
+      "learning_rate": 2.4682539682539684e-05,
+      "loss": 0.3725,
+      "step": 779
+    },
+    {
+      "epoch": 1.6702355460385439,
+      "grad_norm": 0.21952221945022432,
+      "learning_rate": 2.4642857142857145e-05,
+      "loss": 0.3678,
+      "step": 780
+    },
+    {
+      "epoch": 1.6723768736616702,
+      "grad_norm": 0.49934368068454904,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.3964,
+      "step": 781
+    },
+    {
+      "epoch": 1.6745182012847966,
+      "grad_norm": 0.19757194754481436,
+      "learning_rate": 2.4563492063492063e-05,
+      "loss": 0.3763,
+      "step": 782
+    },
+    {
+      "epoch": 1.676659528907923,
+      "grad_norm": 0.19767924902990838,
+      "learning_rate": 2.4523809523809523e-05,
+      "loss": 0.3543,
+      "step": 783
+    },
+    {
+      "epoch": 1.6788008565310493,
+      "grad_norm": 0.20919876927515002,
+      "learning_rate": 2.4484126984126984e-05,
+      "loss": 0.3735,
+      "step": 784
+    },
+    {
+      "epoch": 1.6809421841541756,
+      "grad_norm": 0.22040578372419578,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.3829,
+      "step": 785
+    },
+    {
+      "epoch": 1.683083511777302,
+      "grad_norm": 0.22430773595492015,
+      "learning_rate": 2.4404761904761906e-05,
+      "loss": 0.3656,
+      "step": 786
+    },
+    {
+      "epoch": 1.685224839400428,
+      "grad_norm": 0.20259143808355987,
+      "learning_rate": 2.4365079365079366e-05,
+      "loss": 0.3647,
+      "step": 787
+    },
+    {
+      "epoch": 1.6873661670235545,
+      "grad_norm": 0.21894252518737073,
+      "learning_rate": 2.4325396825396827e-05,
+      "loss": 0.3601,
+      "step": 788
+    },
+    {
+      "epoch": 1.6895074946466808,
+      "grad_norm": 0.21462981545836501,
+      "learning_rate": 2.4285714285714288e-05,
+      "loss": 0.3629,
+      "step": 789
+    },
+    {
+      "epoch": 1.6916488222698072,
+      "grad_norm": 0.18869575445100586,
+      "learning_rate": 2.424603174603175e-05,
+      "loss": 0.3323,
+      "step": 790
+    },
+    {
+      "epoch": 1.6937901498929335,
+      "grad_norm": 0.20241689753494335,
+      "learning_rate": 2.4206349206349206e-05,
+      "loss": 0.3464,
+      "step": 791
+    },
+    {
+      "epoch": 1.6959314775160599,
+      "grad_norm": 0.19953981094595588,
+      "learning_rate": 2.4166666666666667e-05,
+      "loss": 0.3458,
+      "step": 792
+    },
+    {
+      "epoch": 1.6980728051391862,
+      "grad_norm": 0.19961349385460256,
+      "learning_rate": 2.4126984126984128e-05,
+      "loss": 0.3527,
+      "step": 793
+    },
+    {
+      "epoch": 1.7002141327623126,
+      "grad_norm": 0.21973818696015726,
+      "learning_rate": 2.408730158730159e-05,
+      "loss": 0.3611,
+      "step": 794
+    },
+    {
+      "epoch": 1.702355460385439,
+      "grad_norm": 0.1887578663893525,
+      "learning_rate": 2.404761904761905e-05,
+      "loss": 0.3624,
+      "step": 795
+    },
+    {
+      "epoch": 1.7044967880085653,
+      "grad_norm": 0.19628057987549807,
+      "learning_rate": 2.400793650793651e-05,
+      "loss": 0.3717,
+      "step": 796
+    },
+    {
+      "epoch": 1.7066381156316917,
+      "grad_norm": 0.20905187089247354,
+      "learning_rate": 2.396825396825397e-05,
+      "loss": 0.3797,
+      "step": 797
+    },
+    {
+      "epoch": 1.708779443254818,
+      "grad_norm": 0.210161778786647,
+      "learning_rate": 2.392857142857143e-05,
+      "loss": 0.3648,
+      "step": 798
+    },
+    {
+      "epoch": 1.7109207708779444,
+      "grad_norm": 0.18782590789097356,
+      "learning_rate": 2.3888888888888892e-05,
+      "loss": 0.3672,
+      "step": 799
+    },
+    {
+      "epoch": 1.7130620985010707,
+      "grad_norm": 0.21536059285769518,
+      "learning_rate": 2.3849206349206353e-05,
+      "loss": 0.3556,
+      "step": 800
+    },
+    {
+      "epoch": 1.715203426124197,
+      "grad_norm": 0.19872849614371818,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.3606,
+      "step": 801
+    },
+    {
+      "epoch": 1.7173447537473234,
+      "grad_norm": 0.2190527880750777,
+      "learning_rate": 2.376984126984127e-05,
+      "loss": 0.3778,
+      "step": 802
+    },
+    {
+      "epoch": 1.7194860813704498,
+      "grad_norm": 0.21010529253337107,
+      "learning_rate": 2.373015873015873e-05,
+      "loss": 0.355,
+      "step": 803
+    },
+    {
+      "epoch": 1.7216274089935761,
+      "grad_norm": 0.20516661849890908,
+      "learning_rate": 2.369047619047619e-05,
+      "loss": 0.3595,
+      "step": 804
+    },
+    {
+      "epoch": 1.7237687366167025,
+      "grad_norm": 0.21809159516111487,
+      "learning_rate": 2.365079365079365e-05,
+      "loss": 0.3774,
+      "step": 805
+    },
+    {
+      "epoch": 1.7259100642398288,
+      "grad_norm": 0.21787730326337182,
+      "learning_rate": 2.361111111111111e-05,
+      "loss": 0.3716,
+      "step": 806
+    },
+    {
+      "epoch": 1.728051391862955,
+      "grad_norm": 0.20722387140693752,
+      "learning_rate": 2.357142857142857e-05,
+      "loss": 0.3538,
+      "step": 807
+    },
+    {
+      "epoch": 1.7301927194860813,
+      "grad_norm": 0.19038300413501802,
+      "learning_rate": 2.3531746031746032e-05,
+      "loss": 0.3487,
+      "step": 808
+    },
+    {
+      "epoch": 1.7323340471092077,
+      "grad_norm": 0.21385692991090183,
+      "learning_rate": 2.3492063492063493e-05,
+      "loss": 0.3765,
+      "step": 809
+    },
+    {
+      "epoch": 1.734475374732334,
+      "grad_norm": 0.20161843453069309,
+      "learning_rate": 2.3452380952380954e-05,
+      "loss": 0.3593,
+      "step": 810
+    },
+    {
+      "epoch": 1.7366167023554604,
+      "grad_norm": 0.1991611902428487,
+      "learning_rate": 2.3412698412698414e-05,
+      "loss": 0.3667,
+      "step": 811
+    },
+    {
+      "epoch": 1.7387580299785867,
+      "grad_norm": 0.19751266841983958,
+      "learning_rate": 2.3373015873015875e-05,
+      "loss": 0.3725,
+      "step": 812
+    },
+    {
+      "epoch": 1.740899357601713,
+      "grad_norm": 0.18754293892801968,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.3649,
+      "step": 813
+    },
+    {
+      "epoch": 1.7430406852248392,
+      "grad_norm": 0.22386270344186163,
+      "learning_rate": 2.3293650793650797e-05,
+      "loss": 0.3685,
+      "step": 814
+    },
+    {
+      "epoch": 1.7451820128479656,
+      "grad_norm": 0.1966840040343867,
+      "learning_rate": 2.3253968253968257e-05,
+      "loss": 0.3842,
+      "step": 815
+    },
+    {
+      "epoch": 1.747323340471092,
+      "grad_norm": 0.21657779353698456,
+      "learning_rate": 2.3214285714285715e-05,
+      "loss": 0.3667,
+      "step": 816
+    },
+    {
+      "epoch": 1.7494646680942183,
+      "grad_norm": 0.2288614122016469,
+      "learning_rate": 2.3174603174603175e-05,
+      "loss": 0.3843,
+      "step": 817
+    },
+    {
+      "epoch": 1.7516059957173447,
+      "grad_norm": 0.1836444137030942,
+      "learning_rate": 2.3134920634920636e-05,
+      "loss": 0.3583,
+      "step": 818
+    },
+    {
+      "epoch": 1.753747323340471,
+      "grad_norm": 0.23767773436120954,
+      "learning_rate": 2.3095238095238097e-05,
+      "loss": 0.3863,
+      "step": 819
+    },
+    {
+      "epoch": 1.7558886509635974,
+      "grad_norm": 0.2041079485593283,
+      "learning_rate": 2.3055555555555558e-05,
+      "loss": 0.3522,
+      "step": 820
+    },
+    {
+      "epoch": 1.7580299785867237,
+      "grad_norm": 0.20093188325290554,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.3795,
+      "step": 821
+    },
+    {
+      "epoch": 1.76017130620985,
+      "grad_norm": 0.19995939365331822,
+      "learning_rate": 2.2976190476190476e-05,
+      "loss": 0.3721,
+      "step": 822
+    },
+    {
+      "epoch": 1.7623126338329764,
+      "grad_norm": 0.2166910316666174,
+      "learning_rate": 2.2936507936507937e-05,
+      "loss": 0.3634,
+      "step": 823
+    },
+    {
+      "epoch": 1.7644539614561028,
+      "grad_norm": 0.18670680995845637,
+      "learning_rate": 2.2896825396825397e-05,
+      "loss": 0.344,
+      "step": 824
+    },
+    {
+      "epoch": 1.7665952890792291,
+      "grad_norm": 0.21173006300417144,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.356,
+      "step": 825
+    },
+    {
+      "epoch": 1.7687366167023555,
+      "grad_norm": 0.21040949951223306,
+      "learning_rate": 2.281746031746032e-05,
+      "loss": 0.3816,
+      "step": 826
+    },
+    {
+      "epoch": 1.7708779443254818,
+      "grad_norm": 0.20778505679750192,
+      "learning_rate": 2.277777777777778e-05,
+      "loss": 0.3618,
+      "step": 827
+    },
+    {
+      "epoch": 1.7730192719486082,
+      "grad_norm": 0.213799456066487,
+      "learning_rate": 2.273809523809524e-05,
+      "loss": 0.3589,
+      "step": 828
+    },
+    {
+      "epoch": 1.7751605995717346,
+      "grad_norm": 0.2043033051385102,
+      "learning_rate": 2.2698412698412698e-05,
+      "loss": 0.366,
+      "step": 829
+    },
+    {
+      "epoch": 1.777301927194861,
+      "grad_norm": 0.19754414108874052,
+      "learning_rate": 2.265873015873016e-05,
+      "loss": 0.347,
+      "step": 830
+    },
+    {
+      "epoch": 1.7794432548179873,
+      "grad_norm": 0.2158296684757847,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 0.3915,
+      "step": 831
+    },
+    {
+      "epoch": 1.7815845824411136,
+      "grad_norm": 0.22243362581646817,
+      "learning_rate": 2.257936507936508e-05,
+      "loss": 0.3854,
+      "step": 832
+    },
+    {
+      "epoch": 1.78372591006424,
+      "grad_norm": 0.21571356659346066,
+      "learning_rate": 2.253968253968254e-05,
+      "loss": 0.3409,
+      "step": 833
+    },
+    {
+      "epoch": 1.7858672376873663,
+      "grad_norm": 0.22447191028202343,
+      "learning_rate": 2.25e-05,
+      "loss": 0.3749,
+      "step": 834
+    },
+    {
+      "epoch": 1.7880085653104925,
+      "grad_norm": 0.215504263290714,
+      "learning_rate": 2.2460317460317462e-05,
+      "loss": 0.3586,
+      "step": 835
+    },
+    {
+      "epoch": 1.7901498929336188,
+      "grad_norm": 0.20688541961124565,
+      "learning_rate": 2.2420634920634923e-05,
+      "loss": 0.3723,
+      "step": 836
+    },
+    {
+      "epoch": 1.7922912205567452,
+      "grad_norm": 0.20199140517623562,
+      "learning_rate": 2.2380952380952384e-05,
+      "loss": 0.371,
+      "step": 837
+    },
+    {
+      "epoch": 1.7944325481798715,
+      "grad_norm": 0.22478296242958862,
+      "learning_rate": 2.2341269841269844e-05,
+      "loss": 0.3634,
+      "step": 838
+    },
+    {
+      "epoch": 1.7965738758029979,
+      "grad_norm": 0.19888513887055395,
+      "learning_rate": 2.2301587301587305e-05,
+      "loss": 0.3733,
+      "step": 839
+    },
+    {
+      "epoch": 1.7987152034261242,
+      "grad_norm": 0.19421047575082181,
+      "learning_rate": 2.2261904761904763e-05,
+      "loss": 0.3597,
+      "step": 840
+    },
+    {
+      "epoch": 1.8008565310492506,
+      "grad_norm": 0.20475219316843102,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3708,
+      "step": 841
+    },
+    {
+      "epoch": 1.8029978586723767,
+      "grad_norm": 0.2114777102870167,
+      "learning_rate": 2.2182539682539684e-05,
+      "loss": 0.3677,
+      "step": 842
+    },
+    {
+      "epoch": 1.805139186295503,
+      "grad_norm": 0.19598571633723222,
+      "learning_rate": 2.214285714285714e-05,
+      "loss": 0.3661,
+      "step": 843
+    },
+    {
+      "epoch": 1.8072805139186294,
+      "grad_norm": 0.20252036775324378,
+      "learning_rate": 2.2103174603174602e-05,
+      "loss": 0.3645,
+      "step": 844
+    },
+    {
+      "epoch": 1.8094218415417558,
+      "grad_norm": 0.20511403057320743,
+      "learning_rate": 2.2063492063492063e-05,
+      "loss": 0.3633,
+      "step": 845
+    },
+    {
+      "epoch": 1.8115631691648821,
+      "grad_norm": 0.20558945002253082,
+      "learning_rate": 2.2023809523809524e-05,
+      "loss": 0.3752,
+      "step": 846
+    },
+    {
+      "epoch": 1.8137044967880085,
+      "grad_norm": 0.21428205634679462,
+      "learning_rate": 2.1984126984126984e-05,
+      "loss": 0.3683,
+      "step": 847
+    },
+    {
+      "epoch": 1.8158458244111348,
+      "grad_norm": 0.23331079602832885,
+      "learning_rate": 2.1944444444444445e-05,
+      "loss": 0.3671,
+      "step": 848
+    },
+    {
+      "epoch": 1.8179871520342612,
+      "grad_norm": 0.21486622134643,
+      "learning_rate": 2.1904761904761906e-05,
+      "loss": 0.3652,
+      "step": 849
+    },
+    {
+      "epoch": 1.8201284796573876,
+      "grad_norm": 0.19969930097292724,
+      "learning_rate": 2.1865079365079367e-05,
+      "loss": 0.3689,
+      "step": 850
+    },
+    {
+      "epoch": 1.822269807280514,
+      "grad_norm": 0.21534324934924104,
+      "learning_rate": 2.1825396825396827e-05,
+      "loss": 0.3774,
+      "step": 851
+    },
+    {
+      "epoch": 1.8244111349036403,
+      "grad_norm": 0.21397843196592883,
+      "learning_rate": 2.1785714285714288e-05,
+      "loss": 0.3694,
+      "step": 852
+    },
+    {
+      "epoch": 1.8265524625267666,
+      "grad_norm": 0.23443278557392533,
+      "learning_rate": 2.174603174603175e-05,
+      "loss": 0.3699,
+      "step": 853
+    },
+    {
+      "epoch": 1.828693790149893,
+      "grad_norm": 0.19730634328807958,
+      "learning_rate": 2.170634920634921e-05,
+      "loss": 0.3478,
+      "step": 854
+    },
+    {
+      "epoch": 1.8308351177730193,
+      "grad_norm": 0.21913953376852294,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.3567,
+      "step": 855
+    },
+    {
+      "epoch": 1.8329764453961457,
+      "grad_norm": 0.20501578279293847,
+      "learning_rate": 2.1626984126984128e-05,
+      "loss": 0.3497,
+      "step": 856
+    },
+    {
+      "epoch": 1.835117773019272,
+      "grad_norm": 0.20935735217003343,
+      "learning_rate": 2.158730158730159e-05,
+      "loss": 0.3551,
+      "step": 857
+    },
+    {
+      "epoch": 1.8372591006423984,
+      "grad_norm": 0.19034194770904342,
+      "learning_rate": 2.154761904761905e-05,
+      "loss": 0.3628,
+      "step": 858
+    },
+    {
+      "epoch": 1.8394004282655247,
+      "grad_norm": 0.18870437952631242,
+      "learning_rate": 2.150793650793651e-05,
+      "loss": 0.3513,
+      "step": 859
+    },
+    {
+      "epoch": 1.841541755888651,
+      "grad_norm": 0.1919743307796265,
+      "learning_rate": 2.1468253968253967e-05,
+      "loss": 0.3732,
+      "step": 860
+    },
+    {
+      "epoch": 1.8436830835117775,
+      "grad_norm": 0.20417690639096803,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3554,
+      "step": 861
+    },
+    {
+      "epoch": 1.8458244111349036,
+      "grad_norm": 0.21048046032387568,
+      "learning_rate": 2.138888888888889e-05,
+      "loss": 0.3632,
+      "step": 862
+    },
+    {
+      "epoch": 1.84796573875803,
+      "grad_norm": 0.19340209108489914,
+      "learning_rate": 2.134920634920635e-05,
+      "loss": 0.3691,
+      "step": 863
+    },
+    {
+      "epoch": 1.8501070663811563,
+      "grad_norm": 0.2147750143951101,
+      "learning_rate": 2.130952380952381e-05,
+      "loss": 0.3691,
+      "step": 864
+    },
+    {
+      "epoch": 1.8522483940042827,
+      "grad_norm": 0.19393590878356723,
+      "learning_rate": 2.126984126984127e-05,
+      "loss": 0.3531,
+      "step": 865
+    },
+    {
+      "epoch": 1.854389721627409,
+      "grad_norm": 0.18438206629844184,
+      "learning_rate": 2.1230158730158732e-05,
+      "loss": 0.3529,
+      "step": 866
+    },
+    {
+      "epoch": 1.8565310492505354,
+      "grad_norm": 0.23087737411403764,
+      "learning_rate": 2.1190476190476193e-05,
+      "loss": 0.369,
+      "step": 867
+    },
+    {
+      "epoch": 1.8586723768736617,
+      "grad_norm": 0.20076795586544216,
+      "learning_rate": 2.115079365079365e-05,
+      "loss": 0.3523,
+      "step": 868
+    },
+    {
+      "epoch": 1.8608137044967878,
+      "grad_norm": 0.20026389882619192,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.3623,
+      "step": 869
+    },
+    {
+      "epoch": 1.8629550321199142,
+      "grad_norm": 0.2187940555605095,
+      "learning_rate": 2.107142857142857e-05,
+      "loss": 0.3579,
+      "step": 870
+    },
+    {
+      "epoch": 1.8650963597430406,
+      "grad_norm": 0.22989363140509406,
+      "learning_rate": 2.1031746031746032e-05,
+      "loss": 0.3537,
+      "step": 871
+    },
+    {
+      "epoch": 1.867237687366167,
+      "grad_norm": 0.187165767075903,
+      "learning_rate": 2.0992063492063493e-05,
+      "loss": 0.3734,
+      "step": 872
+    },
+    {
+      "epoch": 1.8693790149892933,
+      "grad_norm": 0.2107266950762665,
+      "learning_rate": 2.0952380952380954e-05,
+      "loss": 0.3672,
+      "step": 873
+    },
+    {
+      "epoch": 1.8715203426124196,
+      "grad_norm": 0.20955760948663935,
+      "learning_rate": 2.0912698412698415e-05,
+      "loss": 0.3584,
+      "step": 874
+    },
+    {
+      "epoch": 1.873661670235546,
+      "grad_norm": 0.20510030528233222,
+      "learning_rate": 2.0873015873015875e-05,
+      "loss": 0.3564,
+      "step": 875
+    },
+    {
+      "epoch": 1.8758029978586723,
+      "grad_norm": 0.22155578745771093,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.3543,
+      "step": 876
+    },
+    {
+      "epoch": 1.8779443254817987,
+      "grad_norm": 0.21075973711020585,
+      "learning_rate": 2.0793650793650797e-05,
+      "loss": 0.344,
+      "step": 877
+    },
+    {
+      "epoch": 1.880085653104925,
+      "grad_norm": 0.21005937923832468,
+      "learning_rate": 2.0753968253968258e-05,
+      "loss": 0.3672,
+      "step": 878
+    },
+    {
+      "epoch": 1.8822269807280514,
+      "grad_norm": 0.20105451656897086,
+      "learning_rate": 2.0714285714285718e-05,
+      "loss": 0.3672,
+      "step": 879
+    },
+    {
+      "epoch": 1.8843683083511777,
+      "grad_norm": 0.245253064751467,
+      "learning_rate": 2.0674603174603176e-05,
+      "loss": 0.3551,
+      "step": 880
+    },
+    {
+      "epoch": 1.886509635974304,
+      "grad_norm": 0.21136163316257608,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.363,
+      "step": 881
+    },
+    {
+      "epoch": 1.8886509635974305,
+      "grad_norm": 0.20602573266640592,
+      "learning_rate": 2.0595238095238094e-05,
+      "loss": 0.3537,
+      "step": 882
+    },
+    {
+      "epoch": 1.8907922912205568,
+      "grad_norm": 0.20197617146682256,
+      "learning_rate": 2.0555555555555555e-05,
+      "loss": 0.3525,
+      "step": 883
+    },
+    {
+      "epoch": 1.8929336188436832,
+      "grad_norm": 0.20854372113696887,
+      "learning_rate": 2.0515873015873015e-05,
+      "loss": 0.3514,
+      "step": 884
+    },
+    {
+      "epoch": 1.8950749464668095,
+      "grad_norm": 0.4220973103981105,
+      "learning_rate": 2.0476190476190476e-05,
+      "loss": 0.3666,
+      "step": 885
+    },
+    {
+      "epoch": 1.8972162740899359,
+      "grad_norm": 0.19662405811862105,
+      "learning_rate": 2.0436507936507937e-05,
+      "loss": 0.3557,
+      "step": 886
+    },
+    {
+      "epoch": 1.8993576017130622,
+      "grad_norm": 0.21424028404578552,
+      "learning_rate": 2.0396825396825398e-05,
+      "loss": 0.3705,
+      "step": 887
+    },
+    {
+      "epoch": 1.9014989293361886,
+      "grad_norm": 0.20444704514356857,
+      "learning_rate": 2.0357142857142858e-05,
+      "loss": 0.3671,
+      "step": 888
+    },
+    {
+      "epoch": 1.903640256959315,
+      "grad_norm": 0.209872146829905,
+      "learning_rate": 2.031746031746032e-05,
+      "loss": 0.3675,
+      "step": 889
+    },
+    {
+      "epoch": 1.905781584582441,
+      "grad_norm": 0.2128263431128925,
+      "learning_rate": 2.027777777777778e-05,
+      "loss": 0.3578,
+      "step": 890
+    },
+    {
+      "epoch": 1.9079229122055674,
+      "grad_norm": 0.21350097793924863,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 0.3579,
+      "step": 891
+    },
+    {
+      "epoch": 1.9100642398286938,
+      "grad_norm": 0.19676311955725848,
+      "learning_rate": 2.01984126984127e-05,
+      "loss": 0.3453,
+      "step": 892
+    },
+    {
+      "epoch": 1.9122055674518201,
+      "grad_norm": 0.18886161439157978,
+      "learning_rate": 2.015873015873016e-05,
+      "loss": 0.3576,
+      "step": 893
+    },
+    {
+      "epoch": 1.9143468950749465,
+      "grad_norm": 0.19774803879409486,
+      "learning_rate": 2.011904761904762e-05,
+      "loss": 0.3534,
+      "step": 894
+    },
+    {
+      "epoch": 1.9164882226980728,
+      "grad_norm": 0.23099750551614034,
+      "learning_rate": 2.007936507936508e-05,
+      "loss": 0.3699,
+      "step": 895
+    },
+    {
+      "epoch": 1.9186295503211992,
+      "grad_norm": 0.2045355403156652,
+      "learning_rate": 2.003968253968254e-05,
+      "loss": 0.359,
+      "step": 896
+    },
+    {
+      "epoch": 1.9207708779443253,
+      "grad_norm": 0.1971397275342457,
+      "learning_rate": 2e-05,
+      "loss": 0.3635,
+      "step": 897
+    },
+    {
+      "epoch": 1.9229122055674517,
+      "grad_norm": 0.21049030917982176,
+      "learning_rate": 1.9960317460317462e-05,
+      "loss": 0.3745,
+      "step": 898
+    },
+    {
+      "epoch": 1.925053533190578,
+      "grad_norm": 0.222721408427582,
+      "learning_rate": 1.992063492063492e-05,
+      "loss": 0.3848,
+      "step": 899
+    },
+    {
+      "epoch": 1.9271948608137044,
+      "grad_norm": 0.21296826592149146,
+      "learning_rate": 1.988095238095238e-05,
+      "loss": 0.3845,
+      "step": 900
+    },
+    {
+      "epoch": 1.9293361884368307,
+      "grad_norm": 0.19916011701018338,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.3653,
+      "step": 901
+    },
+    {
+      "epoch": 1.931477516059957,
+      "grad_norm": 0.2010065237557244,
+      "learning_rate": 1.9801587301587302e-05,
+      "loss": 0.346,
+      "step": 902
+    },
+    {
+      "epoch": 1.9336188436830835,
+      "grad_norm": 0.22109341655275905,
+      "learning_rate": 1.9761904761904763e-05,
+      "loss": 0.3811,
+      "step": 903
+    },
+    {
+      "epoch": 1.9357601713062098,
+      "grad_norm": 0.21704603464265765,
+      "learning_rate": 1.9722222222222224e-05,
+      "loss": 0.3697,
+      "step": 904
+    },
+    {
+      "epoch": 1.9379014989293362,
+      "grad_norm": 0.19589016044717736,
+      "learning_rate": 1.9682539682539684e-05,
+      "loss": 0.3561,
+      "step": 905
+    },
+    {
+      "epoch": 1.9400428265524625,
+      "grad_norm": 0.1844953806039601,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 0.355,
+      "step": 906
+    },
+    {
+      "epoch": 1.9421841541755889,
+      "grad_norm": 0.2450452225863573,
+      "learning_rate": 1.9603174603174602e-05,
+      "loss": 0.34,
+      "step": 907
+    },
+    {
+      "epoch": 1.9443254817987152,
+      "grad_norm": 0.1985635379433612,
+      "learning_rate": 1.9563492063492063e-05,
+      "loss": 0.3724,
+      "step": 908
+    },
+    {
+      "epoch": 1.9464668094218416,
+      "grad_norm": 0.20873253034957434,
+      "learning_rate": 1.9523809523809524e-05,
+      "loss": 0.3693,
+      "step": 909
+    },
+    {
+      "epoch": 1.948608137044968,
+      "grad_norm": 0.22144116581790702,
+      "learning_rate": 1.9484126984126985e-05,
+      "loss": 0.3729,
+      "step": 910
+    },
+    {
+      "epoch": 1.9507494646680943,
+      "grad_norm": 0.23269442568625717,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.3807,
+      "step": 911
+    },
+    {
+      "epoch": 1.9528907922912206,
+      "grad_norm": 0.21675592494303747,
+      "learning_rate": 1.9404761904761906e-05,
+      "loss": 0.3533,
+      "step": 912
+    },
+    {
+      "epoch": 1.955032119914347,
+      "grad_norm": 0.2197742949597075,
+      "learning_rate": 1.9365079365079367e-05,
+      "loss": 0.3607,
+      "step": 913
+    },
+    {
+      "epoch": 1.9571734475374734,
+      "grad_norm": 0.21883406348238188,
+      "learning_rate": 1.9325396825396828e-05,
+      "loss": 0.3635,
+      "step": 914
+    },
+    {
+      "epoch": 1.9593147751605997,
+      "grad_norm": 0.2026715665088903,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 0.358,
+      "step": 915
+    },
+    {
+      "epoch": 1.961456102783726,
+      "grad_norm": 0.1974253749656061,
+      "learning_rate": 1.924603174603175e-05,
+      "loss": 0.3474,
+      "step": 916
+    },
+    {
+      "epoch": 1.9635974304068522,
+      "grad_norm": 0.19294611529574526,
+      "learning_rate": 1.920634920634921e-05,
+      "loss": 0.3625,
+      "step": 917
+    },
+    {
+      "epoch": 1.9657387580299786,
+      "grad_norm": 0.22729056002121276,
+      "learning_rate": 1.9166666666666667e-05,
+      "loss": 0.3548,
+      "step": 918
+    },
+    {
+      "epoch": 1.967880085653105,
+      "grad_norm": 0.20766953098408916,
+      "learning_rate": 1.9126984126984128e-05,
+      "loss": 0.3599,
+      "step": 919
+    },
+    {
+      "epoch": 1.9700214132762313,
+      "grad_norm": 0.23817301013936817,
+      "learning_rate": 1.9087301587301585e-05,
+      "loss": 0.3787,
+      "step": 920
+    },
+    {
+      "epoch": 1.9721627408993576,
+      "grad_norm": 0.21184335304060803,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.356,
+      "step": 921
+    },
+    {
+      "epoch": 1.974304068522484,
+      "grad_norm": 0.19810583708611373,
+      "learning_rate": 1.9007936507936507e-05,
+      "loss": 0.373,
+      "step": 922
+    },
+    {
+      "epoch": 1.9764453961456103,
+      "grad_norm": 0.217549480144923,
+      "learning_rate": 1.8968253968253968e-05,
+      "loss": 0.3611,
+      "step": 923
+    },
+    {
+      "epoch": 1.9785867237687365,
+      "grad_norm": 0.2376544391246724,
+      "learning_rate": 1.892857142857143e-05,
+      "loss": 0.3867,
+      "step": 924
+    },
+    {
+      "epoch": 1.9807280513918628,
+      "grad_norm": 0.19811936382123432,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.3706,
+      "step": 925
+    },
+    {
+      "epoch": 1.9828693790149892,
+      "grad_norm": 0.196442826279356,
+      "learning_rate": 1.884920634920635e-05,
+      "loss": 0.3766,
+      "step": 926
+    },
+    {
+      "epoch": 1.9850107066381155,
+      "grad_norm": 0.21591468021679908,
+      "learning_rate": 1.880952380952381e-05,
+      "loss": 0.3573,
+      "step": 927
+    },
+    {
+      "epoch": 1.9871520342612419,
+      "grad_norm": 0.20004992781857317,
+      "learning_rate": 1.876984126984127e-05,
+      "loss": 0.3573,
+      "step": 928
+    },
+    {
+      "epoch": 1.9892933618843682,
+      "grad_norm": 0.18155989142218865,
+      "learning_rate": 1.8730158730158732e-05,
+      "loss": 0.3508,
+      "step": 929
+    },
+    {
+      "epoch": 1.9914346895074946,
+      "grad_norm": 0.19220975795928547,
+      "learning_rate": 1.8690476190476193e-05,
+      "loss": 0.3665,
+      "step": 930
+    },
+    {
+      "epoch": 1.993576017130621,
+      "grad_norm": 0.20680633105790955,
+      "learning_rate": 1.8650793650793654e-05,
+      "loss": 0.3709,
+      "step": 931
+    },
+    {
+      "epoch": 1.9957173447537473,
+      "grad_norm": 0.19718351617490748,
+      "learning_rate": 1.861111111111111e-05,
+      "loss": 0.3674,
+      "step": 932
+    },
+    {
+      "epoch": 1.9978586723768736,
+      "grad_norm": 0.20385992307338152,
+      "learning_rate": 1.8571428571428572e-05,
+      "loss": 0.3647,
+      "step": 933
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.19952173716760616,
+      "learning_rate": 1.8531746031746032e-05,
+      "loss": 0.3498,
+      "step": 934
+    },
+    {
+      "epoch": 2.0021413276231264,
+      "grad_norm": 0.2887390533219721,
+      "learning_rate": 1.8492063492063493e-05,
+      "loss": 0.2978,
+      "step": 935
+    },
+    {
+      "epoch": 2.0042826552462527,
+      "grad_norm": 0.2261310029497388,
+      "learning_rate": 1.8452380952380954e-05,
+      "loss": 0.2753,
+      "step": 936
+    },
+    {
+      "epoch": 2.006423982869379,
+      "grad_norm": 0.19608695195304876,
+      "learning_rate": 1.8412698412698415e-05,
+      "loss": 0.2894,
+      "step": 937
+    },
+    {
+      "epoch": 2.0085653104925054,
+      "grad_norm": 0.2670758274924303,
+      "learning_rate": 1.8373015873015875e-05,
+      "loss": 0.2855,
+      "step": 938
+    },
+    {
+      "epoch": 2.0107066381156318,
+      "grad_norm": 0.2729391112548352,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.2766,
+      "step": 939
+    },
+    {
+      "epoch": 2.012847965738758,
+      "grad_norm": 0.2268053076448433,
+      "learning_rate": 1.8293650793650794e-05,
+      "loss": 0.2825,
+      "step": 940
+    },
+    {
+      "epoch": 2.0149892933618845,
+      "grad_norm": 0.22891152322973052,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.2704,
+      "step": 941
+    },
+    {
+      "epoch": 2.017130620985011,
+      "grad_norm": 0.24951718116860663,
+      "learning_rate": 1.8214285714285715e-05,
+      "loss": 0.2925,
+      "step": 942
+    },
+    {
+      "epoch": 2.019271948608137,
+      "grad_norm": 0.24317478809263185,
+      "learning_rate": 1.8174603174603176e-05,
+      "loss": 0.2852,
+      "step": 943
+    },
+    {
+      "epoch": 2.0214132762312635,
+      "grad_norm": 0.22035333150669048,
+      "learning_rate": 1.8134920634920637e-05,
+      "loss": 0.28,
+      "step": 944
+    },
+    {
+      "epoch": 2.02355460385439,
+      "grad_norm": 0.22762363253514034,
+      "learning_rate": 1.8095238095238094e-05,
+      "loss": 0.2894,
+      "step": 945
+    },
+    {
+      "epoch": 2.0256959314775163,
+      "grad_norm": 0.22788105608795867,
+      "learning_rate": 1.8055555555555555e-05,
+      "loss": 0.2743,
+      "step": 946
+    },
+    {
+      "epoch": 2.0278372591006426,
+      "grad_norm": 0.20269423040955745,
+      "learning_rate": 1.8015873015873015e-05,
+      "loss": 0.2663,
+      "step": 947
+    },
+    {
+      "epoch": 2.0299785867237685,
+      "grad_norm": 0.22523427656669212,
+      "learning_rate": 1.7976190476190476e-05,
+      "loss": 0.2781,
+      "step": 948
+    },
+    {
+      "epoch": 2.032119914346895,
+      "grad_norm": 0.21588221204986352,
+      "learning_rate": 1.7936507936507937e-05,
+      "loss": 0.2788,
+      "step": 949
+    },
+    {
+      "epoch": 2.0342612419700212,
+      "grad_norm": 0.195728356650988,
+      "learning_rate": 1.7896825396825398e-05,
+      "loss": 0.2769,
+      "step": 950
+    },
+    {
+      "epoch": 2.0364025695931476,
+      "grad_norm": 0.2161839738558612,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.2941,
+      "step": 951
+    },
+    {
+      "epoch": 2.038543897216274,
+      "grad_norm": 0.23767256005231727,
+      "learning_rate": 1.781746031746032e-05,
+      "loss": 0.2895,
+      "step": 952
+    },
+    {
+      "epoch": 2.0406852248394003,
+      "grad_norm": 0.20171938363284947,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.2781,
+      "step": 953
+    },
+    {
+      "epoch": 2.0428265524625266,
+      "grad_norm": 0.1856341413030586,
+      "learning_rate": 1.773809523809524e-05,
+      "loss": 0.2685,
+      "step": 954
+    },
+    {
+      "epoch": 2.044967880085653,
+      "grad_norm": 0.2145432452323137,
+      "learning_rate": 1.76984126984127e-05,
+      "loss": 0.281,
+      "step": 955
+    },
+    {
+      "epoch": 2.0471092077087794,
+      "grad_norm": 0.2112321163664146,
+      "learning_rate": 1.7658730158730162e-05,
+      "loss": 0.2977,
+      "step": 956
+    },
+    {
+      "epoch": 2.0492505353319057,
+      "grad_norm": 0.2067934311888714,
+      "learning_rate": 1.761904761904762e-05,
+      "loss": 0.2793,
+      "step": 957
+    },
+    {
+      "epoch": 2.051391862955032,
+      "grad_norm": 0.21063958513546374,
+      "learning_rate": 1.757936507936508e-05,
+      "loss": 0.2807,
+      "step": 958
+    },
+    {
+      "epoch": 2.0535331905781584,
+      "grad_norm": 0.19678228876621853,
+      "learning_rate": 1.7539682539682538e-05,
+      "loss": 0.2773,
+      "step": 959
+    },
+    {
+      "epoch": 2.0556745182012848,
+      "grad_norm": 0.1926373898265207,
+      "learning_rate": 1.75e-05,
+      "loss": 0.2747,
+      "step": 960
+    },
+    {
+      "epoch": 2.057815845824411,
+      "grad_norm": 0.20704097224369902,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.2722,
+      "step": 961
+    },
+    {
+      "epoch": 2.0599571734475375,
+      "grad_norm": 0.19808242466728723,
+      "learning_rate": 1.742063492063492e-05,
+      "loss": 0.271,
+      "step": 962
+    },
+    {
+      "epoch": 2.062098501070664,
+      "grad_norm": 0.19665606509062533,
+      "learning_rate": 1.738095238095238e-05,
+      "loss": 0.2769,
+      "step": 963
+    },
+    {
+      "epoch": 2.06423982869379,
+      "grad_norm": 0.20159555272227964,
+      "learning_rate": 1.734126984126984e-05,
+      "loss": 0.2754,
+      "step": 964
+    },
+    {
+      "epoch": 2.0663811563169165,
+      "grad_norm": 0.20715083834741455,
+      "learning_rate": 1.7301587301587302e-05,
+      "loss": 0.2819,
+      "step": 965
+    },
+    {
+      "epoch": 2.068522483940043,
+      "grad_norm": 0.1994028177934581,
+      "learning_rate": 1.7261904761904763e-05,
+      "loss": 0.2737,
+      "step": 966
+    },
+    {
+      "epoch": 2.0706638115631693,
+      "grad_norm": 0.21226238723876503,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 0.2755,
+      "step": 967
+    },
+    {
+      "epoch": 2.0728051391862956,
+      "grad_norm": 0.21621507021709063,
+      "learning_rate": 1.7182539682539684e-05,
+      "loss": 0.2871,
+      "step": 968
+    },
+    {
+      "epoch": 2.074946466809422,
+      "grad_norm": 0.2054891028222431,
+      "learning_rate": 1.7142857142857145e-05,
+      "loss": 0.281,
+      "step": 969
+    },
+    {
+      "epoch": 2.0770877944325483,
+      "grad_norm": 0.20802135509652517,
+      "learning_rate": 1.7103174603174606e-05,
+      "loss": 0.2706,
+      "step": 970
+    },
+    {
+      "epoch": 2.0792291220556747,
+      "grad_norm": 0.19737703811731966,
+      "learning_rate": 1.7063492063492063e-05,
+      "loss": 0.2825,
+      "step": 971
+    },
+    {
+      "epoch": 2.081370449678801,
+      "grad_norm": 0.18858696479602974,
+      "learning_rate": 1.7023809523809524e-05,
+      "loss": 0.2631,
+      "step": 972
+    },
+    {
+      "epoch": 2.0835117773019274,
+      "grad_norm": 0.19571849751196396,
+      "learning_rate": 1.6984126984126985e-05,
+      "loss": 0.2878,
+      "step": 973
+    },
+    {
+      "epoch": 2.0856531049250537,
+      "grad_norm": 0.19389617623346578,
+      "learning_rate": 1.6944444444444446e-05,
+      "loss": 0.2821,
+      "step": 974
+    },
+    {
+      "epoch": 2.08779443254818,
+      "grad_norm": 0.19768747519412655,
+      "learning_rate": 1.6904761904761906e-05,
+      "loss": 0.2836,
+      "step": 975
+    },
+    {
+      "epoch": 2.089935760171306,
+      "grad_norm": 0.19370599491498322,
+      "learning_rate": 1.6865079365079367e-05,
+      "loss": 0.2712,
+      "step": 976
+    },
+    {
+      "epoch": 2.0920770877944324,
+      "grad_norm": 0.20240036491722738,
+      "learning_rate": 1.6825396825396828e-05,
+      "loss": 0.287,
+      "step": 977
+    },
+    {
+      "epoch": 2.0942184154175587,
+      "grad_norm": 0.2036262743697464,
+      "learning_rate": 1.6785714285714285e-05,
+      "loss": 0.297,
+      "step": 978
+    },
+    {
+      "epoch": 2.096359743040685,
+      "grad_norm": 0.20250473360666518,
+      "learning_rate": 1.6746031746031746e-05,
+      "loss": 0.2777,
+      "step": 979
+    },
+    {
+      "epoch": 2.0985010706638114,
+      "grad_norm": 0.20192268740605415,
+      "learning_rate": 1.6706349206349207e-05,
+      "loss": 0.2864,
+      "step": 980
+    },
+    {
+      "epoch": 2.1006423982869378,
+      "grad_norm": 0.19075170756571364,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2701,
+      "step": 981
+    },
+    {
+      "epoch": 2.102783725910064,
+      "grad_norm": 0.19697486168056122,
+      "learning_rate": 1.6626984126984128e-05,
+      "loss": 0.2811,
+      "step": 982
+    },
+    {
+      "epoch": 2.1049250535331905,
+      "grad_norm": 0.19096670479918926,
+      "learning_rate": 1.658730158730159e-05,
+      "loss": 0.2769,
+      "step": 983
+    },
+    {
+      "epoch": 2.107066381156317,
+      "grad_norm": 0.19574292129874354,
+      "learning_rate": 1.6547619047619046e-05,
+      "loss": 0.2701,
+      "step": 984
+    },
+    {
+      "epoch": 2.109207708779443,
+      "grad_norm": 0.21404348512022472,
+      "learning_rate": 1.6507936507936507e-05,
+      "loss": 0.2849,
+      "step": 985
+    },
+    {
+      "epoch": 2.1113490364025695,
+      "grad_norm": 0.19567445704922734,
+      "learning_rate": 1.6468253968253968e-05,
+      "loss": 0.2753,
+      "step": 986
+    },
+    {
+      "epoch": 2.113490364025696,
+      "grad_norm": 0.2069043774274054,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 0.2776,
+      "step": 987
+    },
+    {
+      "epoch": 2.1156316916488223,
+      "grad_norm": 0.22126156756781362,
+      "learning_rate": 1.638888888888889e-05,
+      "loss": 0.2979,
+      "step": 988
+    },
+    {
+      "epoch": 2.1177730192719486,
+      "grad_norm": 0.21047914160745446,
+      "learning_rate": 1.634920634920635e-05,
+      "loss": 0.2759,
+      "step": 989
+    },
+    {
+      "epoch": 2.119914346895075,
+      "grad_norm": 0.1866961799813759,
+      "learning_rate": 1.630952380952381e-05,
+      "loss": 0.2734,
+      "step": 990
+    },
+    {
+      "epoch": 2.1220556745182013,
+      "grad_norm": 0.19697846429602314,
+      "learning_rate": 1.626984126984127e-05,
+      "loss": 0.2932,
+      "step": 991
+    },
+    {
+      "epoch": 2.1241970021413277,
+      "grad_norm": 0.1907591280617684,
+      "learning_rate": 1.6230158730158732e-05,
+      "loss": 0.2743,
+      "step": 992
+    },
+    {
+      "epoch": 2.126338329764454,
+      "grad_norm": 0.18785957871532322,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.2797,
+      "step": 993
+    },
+    {
+      "epoch": 2.1284796573875804,
+      "grad_norm": 0.1881059477765473,
+      "learning_rate": 1.6150793650793654e-05,
+      "loss": 0.2795,
+      "step": 994
+    },
+    {
+      "epoch": 2.1306209850107067,
+      "grad_norm": 0.1895829464393248,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 0.2699,
+      "step": 995
+    },
+    {
+      "epoch": 2.132762312633833,
+      "grad_norm": 0.1894894322187353,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 0.2839,
+      "step": 996
+    },
+    {
+      "epoch": 2.1349036402569594,
+      "grad_norm": 0.20942109541121104,
+      "learning_rate": 1.6031746031746033e-05,
+      "loss": 0.276,
+      "step": 997
+    },
+    {
+      "epoch": 2.137044967880086,
+      "grad_norm": 0.1929206423206803,
+      "learning_rate": 1.599206349206349e-05,
+      "loss": 0.2781,
+      "step": 998
+    },
+    {
+      "epoch": 2.139186295503212,
+      "grad_norm": 0.20059791418349765,
+      "learning_rate": 1.595238095238095e-05,
+      "loss": 0.2731,
+      "step": 999
+    },
+    {
+      "epoch": 2.1413276231263385,
+      "grad_norm": 0.19734665993628578,
+      "learning_rate": 1.591269841269841e-05,
+      "loss": 0.2841,
+      "step": 1000
+    },
+    {
+      "epoch": 2.143468950749465,
+      "grad_norm": 0.19634335711712297,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.2861,
+      "step": 1001
+    },
+    {
+      "epoch": 2.145610278372591,
+      "grad_norm": 0.18641688183989846,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 0.2699,
+      "step": 1002
+    },
+    {
+      "epoch": 2.147751605995717,
+      "grad_norm": 0.19839241096997418,
+      "learning_rate": 1.5793650793650794e-05,
+      "loss": 0.2746,
+      "step": 1003
+    },
+    {
+      "epoch": 2.1498929336188435,
+      "grad_norm": 0.1903736062839147,
+      "learning_rate": 1.5753968253968255e-05,
+      "loss": 0.2649,
+      "step": 1004
+    },
+    {
+      "epoch": 2.15203426124197,
+      "grad_norm": 0.19577687838702312,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.2885,
+      "step": 1005
+    },
+    {
+      "epoch": 2.154175588865096,
+      "grad_norm": 0.20084961494956982,
+      "learning_rate": 1.5674603174603176e-05,
+      "loss": 0.277,
+      "step": 1006
+    },
+    {
+      "epoch": 2.1563169164882225,
+      "grad_norm": 0.19044578659203962,
+      "learning_rate": 1.5634920634920637e-05,
+      "loss": 0.2759,
+      "step": 1007
+    },
+    {
+      "epoch": 2.158458244111349,
+      "grad_norm": 0.20168687457620502,
+      "learning_rate": 1.5595238095238098e-05,
+      "loss": 0.2883,
+      "step": 1008
+    },
+    {
+      "epoch": 2.1605995717344753,
+      "grad_norm": 0.1917192379674667,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.2822,
+      "step": 1009
+    },
+    {
+      "epoch": 2.1627408993576016,
+      "grad_norm": 0.21720443733726488,
+      "learning_rate": 1.5515873015873016e-05,
+      "loss": 0.287,
+      "step": 1010
+    },
+    {
+      "epoch": 2.164882226980728,
+      "grad_norm": 0.19059089571719792,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 0.2827,
+      "step": 1011
+    },
+    {
+      "epoch": 2.1670235546038543,
+      "grad_norm": 0.20266248908092088,
+      "learning_rate": 1.5436507936507937e-05,
+      "loss": 0.2944,
+      "step": 1012
+    },
+    {
+      "epoch": 2.1691648822269807,
+      "grad_norm": 0.20424509394375523,
+      "learning_rate": 1.5396825396825398e-05,
+      "loss": 0.2736,
+      "step": 1013
+    },
+    {
+      "epoch": 2.171306209850107,
+      "grad_norm": 0.19086220079548602,
+      "learning_rate": 1.535714285714286e-05,
+      "loss": 0.2772,
+      "step": 1014
+    },
+    {
+      "epoch": 2.1734475374732334,
+      "grad_norm": 0.19412331734644836,
+      "learning_rate": 1.531746031746032e-05,
+      "loss": 0.2748,
+      "step": 1015
+    },
+    {
+      "epoch": 2.1755888650963597,
+      "grad_norm": 0.20630378133521415,
+      "learning_rate": 1.527777777777778e-05,
+      "loss": 0.2938,
+      "step": 1016
+    },
+    {
+      "epoch": 2.177730192719486,
+      "grad_norm": 0.18492977798415206,
+      "learning_rate": 1.5238095238095241e-05,
+      "loss": 0.2745,
+      "step": 1017
+    },
+    {
+      "epoch": 2.1798715203426124,
+      "grad_norm": 0.2039798887458119,
+      "learning_rate": 1.5198412698412698e-05,
+      "loss": 0.2755,
+      "step": 1018
+    },
+    {
+      "epoch": 2.182012847965739,
+      "grad_norm": 0.20121459942322972,
+      "learning_rate": 1.5158730158730159e-05,
+      "loss": 0.2862,
+      "step": 1019
+    },
+    {
+      "epoch": 2.184154175588865,
+      "grad_norm": 0.1827134805798544,
+      "learning_rate": 1.5119047619047618e-05,
+      "loss": 0.2718,
+      "step": 1020
+    },
+    {
+      "epoch": 2.1862955032119915,
+      "grad_norm": 0.17917411734145225,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.2695,
+      "step": 1021
+    },
+    {
+      "epoch": 2.188436830835118,
+      "grad_norm": 0.18767988072287312,
+      "learning_rate": 1.503968253968254e-05,
+      "loss": 0.2684,
+      "step": 1022
+    },
+    {
+      "epoch": 2.190578158458244,
+      "grad_norm": 0.1942652275305181,
+      "learning_rate": 1.5e-05,
+      "loss": 0.2803,
+      "step": 1023
+    },
+    {
+      "epoch": 2.1927194860813706,
+      "grad_norm": 0.20926677447772102,
+      "learning_rate": 1.4960317460317461e-05,
+      "loss": 0.2843,
+      "step": 1024
+    },
+    {
+      "epoch": 2.194860813704497,
+      "grad_norm": 0.1991623873969263,
+      "learning_rate": 1.4920634920634922e-05,
+      "loss": 0.2874,
+      "step": 1025
+    },
+    {
+      "epoch": 2.1970021413276233,
+      "grad_norm": 0.18616156346828455,
+      "learning_rate": 1.4880952380952381e-05,
+      "loss": 0.2836,
+      "step": 1026
+    },
+    {
+      "epoch": 2.1991434689507496,
+      "grad_norm": 0.2001544304384233,
+      "learning_rate": 1.4841269841269842e-05,
+      "loss": 0.292,
+      "step": 1027
+    },
+    {
+      "epoch": 2.201284796573876,
+      "grad_norm": 0.19807876601861757,
+      "learning_rate": 1.4801587301587302e-05,
+      "loss": 0.2815,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2034261241970023,
+      "grad_norm": 0.18403372703681253,
+      "learning_rate": 1.4761904761904763e-05,
+      "loss": 0.2733,
+      "step": 1029
+    },
+    {
+      "epoch": 2.2055674518201283,
+      "grad_norm": 0.1886899672546482,
+      "learning_rate": 1.4722222222222224e-05,
+      "loss": 0.2664,
+      "step": 1030
+    },
+    {
+      "epoch": 2.207708779443255,
+      "grad_norm": 0.19490843580765388,
+      "learning_rate": 1.4682539682539683e-05,
+      "loss": 0.2899,
+      "step": 1031
+    },
+    {
+      "epoch": 2.209850107066381,
+      "grad_norm": 0.18904649297785464,
+      "learning_rate": 1.4642857142857144e-05,
+      "loss": 0.2805,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2119914346895073,
+      "grad_norm": 0.18795150882002612,
+      "learning_rate": 1.4603174603174605e-05,
+      "loss": 0.2855,
+      "step": 1033
+    },
+    {
+      "epoch": 2.2141327623126337,
+      "grad_norm": 0.19796365275830283,
+      "learning_rate": 1.4563492063492065e-05,
+      "loss": 0.2818,
+      "step": 1034
+    },
+    {
+      "epoch": 2.21627408993576,
+      "grad_norm": 0.18219404726440058,
+      "learning_rate": 1.4523809523809526e-05,
+      "loss": 0.287,
+      "step": 1035
+    },
+    {
+      "epoch": 2.2184154175588864,
+      "grad_norm": 0.19237435500214886,
+      "learning_rate": 1.4484126984126987e-05,
+      "loss": 0.2821,
+      "step": 1036
+    },
+    {
+      "epoch": 2.2205567451820127,
+      "grad_norm": 0.21247677887677272,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.2806,
+      "step": 1037
+    },
+    {
+      "epoch": 2.222698072805139,
+      "grad_norm": 0.19613559228915892,
+      "learning_rate": 1.4404761904761905e-05,
+      "loss": 0.2705,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2248394004282654,
+      "grad_norm": 0.20206602453648073,
+      "learning_rate": 1.4365079365079364e-05,
+      "loss": 0.2936,
+      "step": 1039
+    },
+    {
+      "epoch": 2.226980728051392,
+      "grad_norm": 0.21986971551175186,
+      "learning_rate": 1.4325396825396825e-05,
+      "loss": 0.281,
+      "step": 1040
+    },
+    {
+      "epoch": 2.229122055674518,
+      "grad_norm": 0.19544790141801302,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.2772,
+      "step": 1041
+    },
+    {
+      "epoch": 2.2312633832976445,
+      "grad_norm": 0.187999178943079,
+      "learning_rate": 1.4246031746031746e-05,
+      "loss": 0.2712,
+      "step": 1042
+    },
+    {
+      "epoch": 2.233404710920771,
+      "grad_norm": 0.20155673411774347,
+      "learning_rate": 1.4206349206349207e-05,
+      "loss": 0.2796,
+      "step": 1043
+    },
+    {
+      "epoch": 2.235546038543897,
+      "grad_norm": 0.24201763616546484,
+      "learning_rate": 1.4166666666666668e-05,
+      "loss": 0.2821,
+      "step": 1044
+    },
+    {
+      "epoch": 2.2376873661670236,
+      "grad_norm": 0.20108574451228317,
+      "learning_rate": 1.4126984126984127e-05,
+      "loss": 0.2886,
+      "step": 1045
+    },
+    {
+      "epoch": 2.23982869379015,
+      "grad_norm": 0.19434223451617216,
+      "learning_rate": 1.4087301587301587e-05,
+      "loss": 0.2814,
+      "step": 1046
+    },
+    {
+      "epoch": 2.2419700214132763,
+      "grad_norm": 0.19496558913111342,
+      "learning_rate": 1.4047619047619048e-05,
+      "loss": 0.2809,
+      "step": 1047
+    },
+    {
+      "epoch": 2.2441113490364026,
+      "grad_norm": 0.20072045682891357,
+      "learning_rate": 1.4007936507936509e-05,
+      "loss": 0.2791,
+      "step": 1048
+    },
+    {
+      "epoch": 2.246252676659529,
+      "grad_norm": 0.19738112557289852,
+      "learning_rate": 1.396825396825397e-05,
+      "loss": 0.2756,
+      "step": 1049
+    },
+    {
+      "epoch": 2.2483940042826553,
+      "grad_norm": 0.18581301381293844,
+      "learning_rate": 1.392857142857143e-05,
+      "loss": 0.2696,
+      "step": 1050
+    },
+    {
+      "epoch": 2.2505353319057817,
+      "grad_norm": 0.19262455905174403,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.2758,
+      "step": 1051
+    },
+    {
+      "epoch": 2.252676659528908,
+      "grad_norm": 0.2044076909216322,
+      "learning_rate": 1.384920634920635e-05,
+      "loss": 0.2759,
+      "step": 1052
+    },
+    {
+      "epoch": 2.2548179871520344,
+      "grad_norm": 0.20744550401581244,
+      "learning_rate": 1.3809523809523811e-05,
+      "loss": 0.2952,
+      "step": 1053
+    },
+    {
+      "epoch": 2.2569593147751608,
+      "grad_norm": 0.19946431730972056,
+      "learning_rate": 1.3769841269841272e-05,
+      "loss": 0.2776,
+      "step": 1054
+    },
+    {
+      "epoch": 2.259100642398287,
+      "grad_norm": 0.2061227664972564,
+      "learning_rate": 1.3730158730158733e-05,
+      "loss": 0.273,
+      "step": 1055
+    },
+    {
+      "epoch": 2.2612419700214135,
+      "grad_norm": 0.1958943903490819,
+      "learning_rate": 1.3690476190476192e-05,
+      "loss": 0.281,
+      "step": 1056
+    },
+    {
+      "epoch": 2.2633832976445394,
+      "grad_norm": 0.17986444789632078,
+      "learning_rate": 1.365079365079365e-05,
+      "loss": 0.282,
+      "step": 1057
+    },
+    {
+      "epoch": 2.265524625267666,
+      "grad_norm": 0.20331838773796787,
+      "learning_rate": 1.3611111111111111e-05,
+      "loss": 0.2822,
+      "step": 1058
+    },
+    {
+      "epoch": 2.267665952890792,
+      "grad_norm": 0.1854886814182152,
+      "learning_rate": 1.357142857142857e-05,
+      "loss": 0.2871,
+      "step": 1059
+    },
+    {
+      "epoch": 2.2698072805139184,
+      "grad_norm": 0.19038399977635145,
+      "learning_rate": 1.3531746031746031e-05,
+      "loss": 0.2705,
+      "step": 1060
+    },
+    {
+      "epoch": 2.271948608137045,
+      "grad_norm": 0.18815130656204335,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.2756,
+      "step": 1061
+    },
+    {
+      "epoch": 2.274089935760171,
+      "grad_norm": 0.19455215664990538,
+      "learning_rate": 1.3452380952380953e-05,
+      "loss": 0.283,
+      "step": 1062
+    },
+    {
+      "epoch": 2.2762312633832975,
+      "grad_norm": 0.20059126614368925,
+      "learning_rate": 1.3412698412698413e-05,
+      "loss": 0.2785,
+      "step": 1063
+    },
+    {
+      "epoch": 2.278372591006424,
+      "grad_norm": 0.20535076737597538,
+      "learning_rate": 1.3373015873015873e-05,
+      "loss": 0.2827,
+      "step": 1064
+    },
+    {
+      "epoch": 2.28051391862955,
+      "grad_norm": 0.19133006544393558,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.2876,
+      "step": 1065
+    },
+    {
+      "epoch": 2.2826552462526766,
+      "grad_norm": 0.18259775238692044,
+      "learning_rate": 1.3293650793650794e-05,
+      "loss": 0.2778,
+      "step": 1066
+    },
+    {
+      "epoch": 2.284796573875803,
+      "grad_norm": 0.18347983126035391,
+      "learning_rate": 1.3253968253968255e-05,
+      "loss": 0.2761,
+      "step": 1067
+    },
+    {
+      "epoch": 2.2869379014989293,
+      "grad_norm": 0.18777852570980433,
+      "learning_rate": 1.3214285714285716e-05,
+      "loss": 0.2818,
+      "step": 1068
+    },
+    {
+      "epoch": 2.2890792291220556,
+      "grad_norm": 0.18410525589405727,
+      "learning_rate": 1.3174603174603176e-05,
+      "loss": 0.2656,
+      "step": 1069
+    },
+    {
+      "epoch": 2.291220556745182,
+      "grad_norm": 0.19035078229116137,
+      "learning_rate": 1.3134920634920635e-05,
+      "loss": 0.2816,
+      "step": 1070
+    },
+    {
+      "epoch": 2.2933618843683083,
+      "grad_norm": 0.20015240192900935,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 0.3031,
+      "step": 1071
+    },
+    {
+      "epoch": 2.2955032119914347,
+      "grad_norm": 0.18184078914671584,
+      "learning_rate": 1.3055555555555557e-05,
+      "loss": 0.2784,
+      "step": 1072
+    },
+    {
+      "epoch": 2.297644539614561,
+      "grad_norm": 0.19404174251026468,
+      "learning_rate": 1.3015873015873018e-05,
+      "loss": 0.2761,
+      "step": 1073
+    },
+    {
+      "epoch": 2.2997858672376874,
+      "grad_norm": 0.18586683492531947,
+      "learning_rate": 1.2976190476190478e-05,
+      "loss": 0.2781,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3019271948608138,
+      "grad_norm": 0.18361259801329616,
+      "learning_rate": 1.2936507936507939e-05,
+      "loss": 0.2588,
+      "step": 1075
+    },
+    {
+      "epoch": 2.30406852248394,
+      "grad_norm": 0.19747669576089893,
+      "learning_rate": 1.2896825396825398e-05,
+      "loss": 0.2819,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3062098501070665,
+      "grad_norm": 0.1962675355280158,
+      "learning_rate": 1.2857142857142857e-05,
+      "loss": 0.2755,
+      "step": 1077
+    },
+    {
+      "epoch": 2.308351177730193,
+      "grad_norm": 0.19942980028292878,
+      "learning_rate": 1.2817460317460316e-05,
+      "loss": 0.2818,
+      "step": 1078
+    },
+    {
+      "epoch": 2.310492505353319,
+      "grad_norm": 0.19501376392056294,
+      "learning_rate": 1.2777777777777777e-05,
+      "loss": 0.2882,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3126338329764455,
+      "grad_norm": 0.18868853481419254,
+      "learning_rate": 1.2738095238095238e-05,
+      "loss": 0.2762,
+      "step": 1080
+    },
+    {
+      "epoch": 2.314775160599572,
+      "grad_norm": 0.18259220767781897,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.2862,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3169164882226982,
+      "grad_norm": 0.20057460578986203,
+      "learning_rate": 1.265873015873016e-05,
+      "loss": 0.289,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3190578158458246,
+      "grad_norm": 0.20457960878239465,
+      "learning_rate": 1.261904761904762e-05,
+      "loss": 0.2855,
+      "step": 1083
+    },
+    {
+      "epoch": 2.3211991434689505,
+      "grad_norm": 0.19996856442993224,
+      "learning_rate": 1.2579365079365079e-05,
+      "loss": 0.2831,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3233404710920773,
+      "grad_norm": 0.19193162471753858,
+      "learning_rate": 1.253968253968254e-05,
+      "loss": 0.2788,
+      "step": 1085
+    },
+    {
+      "epoch": 2.325481798715203,
+      "grad_norm": 0.20129531906043394,
+      "learning_rate": 1.25e-05,
+      "loss": 0.2653,
+      "step": 1086
+    },
+    {
+      "epoch": 2.3276231263383296,
+      "grad_norm": 0.19714902241567792,
+      "learning_rate": 1.2460317460317461e-05,
+      "loss": 0.2898,
+      "step": 1087
+    },
+    {
+      "epoch": 2.329764453961456,
+      "grad_norm": 0.20515273593194386,
+      "learning_rate": 1.2420634920634922e-05,
+      "loss": 0.2977,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3319057815845823,
+      "grad_norm": 0.19580178817727573,
+      "learning_rate": 1.2380952380952381e-05,
+      "loss": 0.2802,
+      "step": 1089
+    },
+    {
+      "epoch": 2.3340471092077086,
+      "grad_norm": 0.17768572531273108,
+      "learning_rate": 1.2341269841269842e-05,
+      "loss": 0.2703,
+      "step": 1090
+    },
+    {
+      "epoch": 2.336188436830835,
+      "grad_norm": 0.1956542549620933,
+      "learning_rate": 1.2301587301587301e-05,
+      "loss": 0.2757,
+      "step": 1091
+    },
+    {
+      "epoch": 2.3383297644539613,
+      "grad_norm": 0.20127910379978708,
+      "learning_rate": 1.2261904761904762e-05,
+      "loss": 0.2837,
+      "step": 1092
+    },
+    {
+      "epoch": 2.3404710920770877,
+      "grad_norm": 0.20060227683373613,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.2802,
+      "step": 1093
+    },
+    {
+      "epoch": 2.342612419700214,
+      "grad_norm": 0.19431031021153464,
+      "learning_rate": 1.2182539682539683e-05,
+      "loss": 0.291,
+      "step": 1094
+    },
+    {
+      "epoch": 2.3447537473233404,
+      "grad_norm": 0.19260339789181197,
+      "learning_rate": 1.2142857142857144e-05,
+      "loss": 0.2802,
+      "step": 1095
+    },
+    {
+      "epoch": 2.3468950749464668,
+      "grad_norm": 0.20821596374744575,
+      "learning_rate": 1.2103174603174603e-05,
+      "loss": 0.2866,
+      "step": 1096
+    },
+    {
+      "epoch": 2.349036402569593,
+      "grad_norm": 0.20675441129158606,
+      "learning_rate": 1.2063492063492064e-05,
+      "loss": 0.2926,
+      "step": 1097
+    },
+    {
+      "epoch": 2.3511777301927195,
+      "grad_norm": 0.1968034926711773,
+      "learning_rate": 1.2023809523809525e-05,
+      "loss": 0.2757,
+      "step": 1098
+    },
+    {
+      "epoch": 2.353319057815846,
+      "grad_norm": 0.18488353953473421,
+      "learning_rate": 1.1984126984126985e-05,
+      "loss": 0.2868,
+      "step": 1099
+    },
+    {
+      "epoch": 2.355460385438972,
+      "grad_norm": 0.20338110721794142,
+      "learning_rate": 1.1944444444444446e-05,
+      "loss": 0.2882,
+      "step": 1100
+    },
+    {
+      "epoch": 2.3576017130620985,
+      "grad_norm": 0.19166668181356114,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.292,
+      "step": 1101
+    },
+    {
+      "epoch": 2.359743040685225,
+      "grad_norm": 0.1927455443301824,
+      "learning_rate": 1.1865079365079366e-05,
+      "loss": 0.2649,
+      "step": 1102
+    },
+    {
+      "epoch": 2.3618843683083512,
+      "grad_norm": 0.17477212497660993,
+      "learning_rate": 1.1825396825396825e-05,
+      "loss": 0.2752,
+      "step": 1103
+    },
+    {
+      "epoch": 2.3640256959314776,
+      "grad_norm": 0.19987372705216191,
+      "learning_rate": 1.1785714285714286e-05,
+      "loss": 0.2819,
+      "step": 1104
+    },
+    {
+      "epoch": 2.366167023554604,
+      "grad_norm": 0.1922206507729967,
+      "learning_rate": 1.1746031746031746e-05,
+      "loss": 0.2823,
+      "step": 1105
+    },
+    {
+      "epoch": 2.3683083511777303,
+      "grad_norm": 0.18451313459925092,
+      "learning_rate": 1.1706349206349207e-05,
+      "loss": 0.2858,
+      "step": 1106
+    },
+    {
+      "epoch": 2.3704496788008567,
+      "grad_norm": 0.194347680132915,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.2885,
+      "step": 1107
+    },
+    {
+      "epoch": 2.372591006423983,
+      "grad_norm": 0.19502504685509406,
+      "learning_rate": 1.1626984126984129e-05,
+      "loss": 0.2818,
+      "step": 1108
+    },
+    {
+      "epoch": 2.3747323340471094,
+      "grad_norm": 0.18898431759295656,
+      "learning_rate": 1.1587301587301588e-05,
+      "loss": 0.2904,
+      "step": 1109
+    },
+    {
+      "epoch": 2.3768736616702357,
+      "grad_norm": 0.19153164662259894,
+      "learning_rate": 1.1547619047619048e-05,
+      "loss": 0.2726,
+      "step": 1110
+    },
+    {
+      "epoch": 2.3790149892933616,
+      "grad_norm": 0.18268604277337688,
+      "learning_rate": 1.1507936507936508e-05,
+      "loss": 0.279,
+      "step": 1111
+    },
+    {
+      "epoch": 2.3811563169164884,
+      "grad_norm": 0.17533082071419925,
+      "learning_rate": 1.1468253968253968e-05,
+      "loss": 0.2732,
+      "step": 1112
+    },
+    {
+      "epoch": 2.3832976445396143,
+      "grad_norm": 0.2003790102740343,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.2951,
+      "step": 1113
+    },
+    {
+      "epoch": 2.385438972162741,
+      "grad_norm": 0.18705857015717448,
+      "learning_rate": 1.138888888888889e-05,
+      "loss": 0.2793,
+      "step": 1114
+    },
+    {
+      "epoch": 2.387580299785867,
+      "grad_norm": 0.1868482525264962,
+      "learning_rate": 1.1349206349206349e-05,
+      "loss": 0.2747,
+      "step": 1115
+    },
+    {
+      "epoch": 2.3897216274089934,
+      "grad_norm": 0.18021169330552053,
+      "learning_rate": 1.130952380952381e-05,
+      "loss": 0.2812,
+      "step": 1116
+    },
+    {
+      "epoch": 2.3918629550321198,
+      "grad_norm": 0.19301415468138103,
+      "learning_rate": 1.126984126984127e-05,
+      "loss": 0.2886,
+      "step": 1117
+    },
+    {
+      "epoch": 2.394004282655246,
+      "grad_norm": 0.19072636775125612,
+      "learning_rate": 1.1230158730158731e-05,
+      "loss": 0.2806,
+      "step": 1118
+    },
+    {
+      "epoch": 2.3961456102783725,
+      "grad_norm": 0.1840802920208615,
+      "learning_rate": 1.1190476190476192e-05,
+      "loss": 0.2879,
+      "step": 1119
+    },
+    {
+      "epoch": 2.398286937901499,
+      "grad_norm": 0.18258129644157975,
+      "learning_rate": 1.1150793650793653e-05,
+      "loss": 0.2803,
+      "step": 1120
+    },
+    {
+      "epoch": 2.400428265524625,
+      "grad_norm": 0.1827848542873828,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2881,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4025695931477515,
+      "grad_norm": 0.18900064907660674,
+      "learning_rate": 1.107142857142857e-05,
+      "loss": 0.2938,
+      "step": 1122
+    },
+    {
+      "epoch": 2.404710920770878,
+      "grad_norm": 0.18804850484657887,
+      "learning_rate": 1.1031746031746031e-05,
+      "loss": 0.2858,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4068522483940042,
+      "grad_norm": 0.2006000850638827,
+      "learning_rate": 1.0992063492063492e-05,
+      "loss": 0.2876,
+      "step": 1124
+    },
+    {
+      "epoch": 2.4089935760171306,
+      "grad_norm": 0.18714684164420253,
+      "learning_rate": 1.0952380952380953e-05,
+      "loss": 0.2803,
+      "step": 1125
+    },
+    {
+      "epoch": 2.411134903640257,
+      "grad_norm": 0.19792928506355778,
+      "learning_rate": 1.0912698412698414e-05,
+      "loss": 0.2936,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4132762312633833,
+      "grad_norm": 0.18865247304598853,
+      "learning_rate": 1.0873015873015874e-05,
+      "loss": 0.2778,
+      "step": 1127
+    },
+    {
+      "epoch": 2.4154175588865097,
+      "grad_norm": 0.1929275031551357,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 0.2777,
+      "step": 1128
+    },
+    {
+      "epoch": 2.417558886509636,
+      "grad_norm": 0.1806973386093782,
+      "learning_rate": 1.0793650793650794e-05,
+      "loss": 0.2763,
+      "step": 1129
+    },
+    {
+      "epoch": 2.4197002141327624,
+      "grad_norm": 0.19706536820552648,
+      "learning_rate": 1.0753968253968255e-05,
+      "loss": 0.2883,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4218415417558887,
+      "grad_norm": 0.1910372018652057,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.2837,
+      "step": 1131
+    },
+    {
+      "epoch": 2.423982869379015,
+      "grad_norm": 0.1788916292232767,
+      "learning_rate": 1.0674603174603175e-05,
+      "loss": 0.2841,
+      "step": 1132
+    },
+    {
+      "epoch": 2.4261241970021414,
+      "grad_norm": 0.18659149187977309,
+      "learning_rate": 1.0634920634920636e-05,
+      "loss": 0.2894,
+      "step": 1133
+    },
+    {
+      "epoch": 2.428265524625268,
+      "grad_norm": 0.1925123938840746,
+      "learning_rate": 1.0595238095238096e-05,
+      "loss": 0.2838,
+      "step": 1134
+    },
+    {
+      "epoch": 2.430406852248394,
+      "grad_norm": 0.1776215701785684,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 0.2707,
+      "step": 1135
+    },
+    {
+      "epoch": 2.4325481798715205,
+      "grad_norm": 0.20843365977457115,
+      "learning_rate": 1.0515873015873016e-05,
+      "loss": 0.2892,
+      "step": 1136
+    },
+    {
+      "epoch": 2.434689507494647,
+      "grad_norm": 0.19764138815197937,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.2648,
+      "step": 1137
+    },
+    {
+      "epoch": 2.436830835117773,
+      "grad_norm": 0.19252613410306155,
+      "learning_rate": 1.0436507936507938e-05,
+      "loss": 0.2911,
+      "step": 1138
+    },
+    {
+      "epoch": 2.4389721627408996,
+      "grad_norm": 0.20085600997800981,
+      "learning_rate": 1.0396825396825398e-05,
+      "loss": 0.2826,
+      "step": 1139
+    },
+    {
+      "epoch": 2.4411134903640255,
+      "grad_norm": 0.188049018268479,
+      "learning_rate": 1.0357142857142859e-05,
+      "loss": 0.2911,
+      "step": 1140
+    },
+    {
+      "epoch": 2.4432548179871523,
+      "grad_norm": 0.19493824847548077,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.286,
+      "step": 1141
+    },
+    {
+      "epoch": 2.445396145610278,
+      "grad_norm": 0.191504777727128,
+      "learning_rate": 1.0277777777777777e-05,
+      "loss": 0.2933,
+      "step": 1142
+    },
+    {
+      "epoch": 2.4475374732334045,
+      "grad_norm": 0.19108814910860578,
+      "learning_rate": 1.0238095238095238e-05,
+      "loss": 0.2785,
+      "step": 1143
+    },
+    {
+      "epoch": 2.449678800856531,
+      "grad_norm": 0.20799246077044664,
+      "learning_rate": 1.0198412698412699e-05,
+      "loss": 0.2837,
+      "step": 1144
+    },
+    {
+      "epoch": 2.4518201284796572,
+      "grad_norm": 0.1824764184302531,
+      "learning_rate": 1.015873015873016e-05,
+      "loss": 0.2652,
+      "step": 1145
+    },
+    {
+      "epoch": 2.4539614561027836,
+      "grad_norm": 0.1824775079826765,
+      "learning_rate": 1.011904761904762e-05,
+      "loss": 0.2716,
+      "step": 1146
+    },
+    {
+      "epoch": 2.45610278372591,
+      "grad_norm": 0.1932203896888414,
+      "learning_rate": 1.007936507936508e-05,
+      "loss": 0.276,
+      "step": 1147
+    },
+    {
+      "epoch": 2.4582441113490363,
+      "grad_norm": 0.2007871128283347,
+      "learning_rate": 1.003968253968254e-05,
+      "loss": 0.264,
+      "step": 1148
+    },
+    {
+      "epoch": 2.4603854389721627,
+      "grad_norm": 0.21083133706351928,
+      "learning_rate": 1e-05,
+      "loss": 0.2816,
+      "step": 1149
+    },
+    {
+      "epoch": 2.462526766595289,
+      "grad_norm": 0.19278202499501207,
+      "learning_rate": 9.96031746031746e-06,
+      "loss": 0.2823,
+      "step": 1150
+    },
+    {
+      "epoch": 2.4646680942184154,
+      "grad_norm": 0.18797686590775003,
+      "learning_rate": 9.92063492063492e-06,
+      "loss": 0.2879,
+      "step": 1151
+    },
+    {
+      "epoch": 2.4668094218415417,
+      "grad_norm": 0.20454699746081098,
+      "learning_rate": 9.880952380952381e-06,
+      "loss": 0.2672,
+      "step": 1152
+    },
+    {
+      "epoch": 2.468950749464668,
+      "grad_norm": 0.18367484130511327,
+      "learning_rate": 9.841269841269842e-06,
+      "loss": 0.2806,
+      "step": 1153
+    },
+    {
+      "epoch": 2.4710920770877944,
+      "grad_norm": 0.1831686319686948,
+      "learning_rate": 9.801587301587301e-06,
+      "loss": 0.273,
+      "step": 1154
+    },
+    {
+      "epoch": 2.473233404710921,
+      "grad_norm": 0.19694402234544842,
+      "learning_rate": 9.761904761904762e-06,
+      "loss": 0.2878,
+      "step": 1155
+    },
+    {
+      "epoch": 2.475374732334047,
+      "grad_norm": 0.19408589237490334,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.2943,
+      "step": 1156
+    },
+    {
+      "epoch": 2.4775160599571735,
+      "grad_norm": 0.19445588500969652,
+      "learning_rate": 9.682539682539683e-06,
+      "loss": 0.2784,
+      "step": 1157
+    },
+    {
+      "epoch": 2.4796573875803,
+      "grad_norm": 0.17963888726600663,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.2752,
+      "step": 1158
+    },
+    {
+      "epoch": 2.481798715203426,
+      "grad_norm": 0.19347132870813594,
+      "learning_rate": 9.603174603174605e-06,
+      "loss": 0.2867,
+      "step": 1159
+    },
+    {
+      "epoch": 2.4839400428265526,
+      "grad_norm": 0.18730325675923679,
+      "learning_rate": 9.563492063492064e-06,
+      "loss": 0.2762,
+      "step": 1160
+    },
+    {
+      "epoch": 2.486081370449679,
+      "grad_norm": 0.18460049863260458,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.2872,
+      "step": 1161
+    },
+    {
+      "epoch": 2.4882226980728053,
+      "grad_norm": 0.1907698896483033,
+      "learning_rate": 9.484126984126984e-06,
+      "loss": 0.287,
+      "step": 1162
+    },
+    {
+      "epoch": 2.4903640256959316,
+      "grad_norm": 0.18810661893439526,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.2797,
+      "step": 1163
+    },
+    {
+      "epoch": 2.492505353319058,
+      "grad_norm": 0.19192431651136277,
+      "learning_rate": 9.404761904761905e-06,
+      "loss": 0.2779,
+      "step": 1164
+    },
+    {
+      "epoch": 2.4946466809421843,
+      "grad_norm": 0.1791506541350854,
+      "learning_rate": 9.365079365079366e-06,
+      "loss": 0.278,
+      "step": 1165
+    },
+    {
+      "epoch": 2.4967880085653107,
+      "grad_norm": 0.18069237791730375,
+      "learning_rate": 9.325396825396827e-06,
+      "loss": 0.2792,
+      "step": 1166
+    },
+    {
+      "epoch": 2.4989293361884366,
+      "grad_norm": 0.18412348076309296,
+      "learning_rate": 9.285714285714286e-06,
+      "loss": 0.2771,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5010706638115634,
+      "grad_norm": 0.1813880522758016,
+      "learning_rate": 9.246031746031747e-06,
+      "loss": 0.2835,
+      "step": 1168
+    },
+    {
+      "epoch": 2.5032119914346893,
+      "grad_norm": 0.19273095158091466,
+      "learning_rate": 9.206349206349207e-06,
+      "loss": 0.2923,
+      "step": 1169
+    },
+    {
+      "epoch": 2.505353319057816,
+      "grad_norm": 0.18337449464589062,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.2884,
+      "step": 1170
+    },
+    {
+      "epoch": 2.507494646680942,
+      "grad_norm": 0.19024259739381238,
+      "learning_rate": 9.126984126984127e-06,
+      "loss": 0.2956,
+      "step": 1171
+    },
+    {
+      "epoch": 2.5096359743040684,
+      "grad_norm": 0.19901572512340374,
+      "learning_rate": 9.087301587301588e-06,
+      "loss": 0.2737,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5117773019271947,
+      "grad_norm": 0.194998468534562,
+      "learning_rate": 9.047619047619047e-06,
+      "loss": 0.28,
+      "step": 1173
+    },
+    {
+      "epoch": 2.513918629550321,
+      "grad_norm": 0.18412933977392476,
+      "learning_rate": 9.007936507936508e-06,
+      "loss": 0.2819,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5160599571734474,
+      "grad_norm": 0.18374938776839936,
+      "learning_rate": 8.968253968253968e-06,
+      "loss": 0.2813,
+      "step": 1175
+    },
+    {
+      "epoch": 2.518201284796574,
+      "grad_norm": 0.18751207889857766,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.2831,
+      "step": 1176
+    },
+    {
+      "epoch": 2.5203426124197,
+      "grad_norm": 0.17923076174542152,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 0.2682,
+      "step": 1177
+    },
+    {
+      "epoch": 2.5224839400428265,
+      "grad_norm": 0.18035308954107676,
+      "learning_rate": 8.84920634920635e-06,
+      "loss": 0.2677,
+      "step": 1178
+    },
+    {
+      "epoch": 2.524625267665953,
+      "grad_norm": 0.22973038098447218,
+      "learning_rate": 8.80952380952381e-06,
+      "loss": 0.2874,
+      "step": 1179
+    },
+    {
+      "epoch": 2.526766595289079,
+      "grad_norm": 0.18325732719083584,
+      "learning_rate": 8.769841269841269e-06,
+      "loss": 0.271,
+      "step": 1180
+    },
+    {
+      "epoch": 2.5289079229122056,
+      "grad_norm": 0.1838482554089651,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.2743,
+      "step": 1181
+    },
+    {
+      "epoch": 2.531049250535332,
+      "grad_norm": 0.18469422748009917,
+      "learning_rate": 8.69047619047619e-06,
+      "loss": 0.2809,
+      "step": 1182
+    },
+    {
+      "epoch": 2.5331905781584583,
+      "grad_norm": 0.19824918591422117,
+      "learning_rate": 8.650793650793651e-06,
+      "loss": 0.2956,
+      "step": 1183
+    },
+    {
+      "epoch": 2.5353319057815846,
+      "grad_norm": 0.18477009428383825,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 0.2746,
+      "step": 1184
+    },
+    {
+      "epoch": 2.537473233404711,
+      "grad_norm": 0.18367643368070655,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 0.2779,
+      "step": 1185
+    },
+    {
+      "epoch": 2.5396145610278373,
+      "grad_norm": 0.18552020064619668,
+      "learning_rate": 8.531746031746032e-06,
+      "loss": 0.2726,
+      "step": 1186
+    },
+    {
+      "epoch": 2.5417558886509637,
+      "grad_norm": 0.18823754582760785,
+      "learning_rate": 8.492063492063492e-06,
+      "loss": 0.2936,
+      "step": 1187
+    },
+    {
+      "epoch": 2.54389721627409,
+      "grad_norm": 0.1838357043381974,
+      "learning_rate": 8.452380952380953e-06,
+      "loss": 0.2754,
+      "step": 1188
+    },
+    {
+      "epoch": 2.5460385438972164,
+      "grad_norm": 0.1930931282449352,
+      "learning_rate": 8.412698412698414e-06,
+      "loss": 0.2848,
+      "step": 1189
+    },
+    {
+      "epoch": 2.5481798715203428,
+      "grad_norm": 0.18519908919348946,
+      "learning_rate": 8.373015873015873e-06,
+      "loss": 0.2776,
+      "step": 1190
+    },
+    {
+      "epoch": 2.550321199143469,
+      "grad_norm": 0.18754273175059186,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2799,
+      "step": 1191
+    },
+    {
+      "epoch": 2.552462526766595,
+      "grad_norm": 0.18545398781411668,
+      "learning_rate": 8.293650793650794e-06,
+      "loss": 0.2651,
+      "step": 1192
+    },
+    {
+      "epoch": 2.554603854389722,
+      "grad_norm": 0.1847349009972924,
+      "learning_rate": 8.253968253968254e-06,
+      "loss": 0.2714,
+      "step": 1193
+    },
+    {
+      "epoch": 2.5567451820128477,
+      "grad_norm": 0.19268643145654266,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.279,
+      "step": 1194
+    },
+    {
+      "epoch": 2.5588865096359745,
+      "grad_norm": 0.19778277420764503,
+      "learning_rate": 8.174603174603175e-06,
+      "loss": 0.2778,
+      "step": 1195
+    },
+    {
+      "epoch": 2.5610278372591004,
+      "grad_norm": 0.1998660395641265,
+      "learning_rate": 8.134920634920636e-06,
+      "loss": 0.2864,
+      "step": 1196
+    },
+    {
+      "epoch": 2.5631691648822272,
+      "grad_norm": 0.19294158536685677,
+      "learning_rate": 8.095238095238097e-06,
+      "loss": 0.2696,
+      "step": 1197
+    },
+    {
+      "epoch": 2.565310492505353,
+      "grad_norm": 0.18786548806483286,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 0.2692,
+      "step": 1198
+    },
+    {
+      "epoch": 2.5674518201284795,
+      "grad_norm": 0.19390591817924888,
+      "learning_rate": 8.015873015873016e-06,
+      "loss": 0.278,
+      "step": 1199
+    },
+    {
+      "epoch": 2.569593147751606,
+      "grad_norm": 0.18907409735630998,
+      "learning_rate": 7.976190476190475e-06,
+      "loss": 0.2885,
+      "step": 1200
+    },
+    {
+      "epoch": 2.571734475374732,
+      "grad_norm": 0.20258943279374564,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.303,
+      "step": 1201
+    },
+    {
+      "epoch": 2.5738758029978586,
+      "grad_norm": 0.19454350482759478,
+      "learning_rate": 7.896825396825397e-06,
+      "loss": 0.2739,
+      "step": 1202
+    },
+    {
+      "epoch": 2.576017130620985,
+      "grad_norm": 0.18958864960792593,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.2804,
+      "step": 1203
+    },
+    {
+      "epoch": 2.5781584582441113,
+      "grad_norm": 0.1982183376800514,
+      "learning_rate": 7.817460317460318e-06,
+      "loss": 0.2991,
+      "step": 1204
+    },
+    {
+      "epoch": 2.5802997858672376,
+      "grad_norm": 0.1913275931410259,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.2888,
+      "step": 1205
+    },
+    {
+      "epoch": 2.582441113490364,
+      "grad_norm": 0.21019225544474518,
+      "learning_rate": 7.738095238095238e-06,
+      "loss": 0.2887,
+      "step": 1206
+    },
+    {
+      "epoch": 2.5845824411134903,
+      "grad_norm": 0.2042220303342691,
+      "learning_rate": 7.698412698412699e-06,
+      "loss": 0.3031,
+      "step": 1207
+    },
+    {
+      "epoch": 2.5867237687366167,
+      "grad_norm": 0.19819365750871445,
+      "learning_rate": 7.65873015873016e-06,
+      "loss": 0.2896,
+      "step": 1208
+    },
+    {
+      "epoch": 2.588865096359743,
+      "grad_norm": 0.18837665341574075,
+      "learning_rate": 7.6190476190476205e-06,
+      "loss": 0.2775,
+      "step": 1209
+    },
+    {
+      "epoch": 2.5910064239828694,
+      "grad_norm": 0.18886359577025771,
+      "learning_rate": 7.5793650793650795e-06,
+      "loss": 0.2718,
+      "step": 1210
+    },
+    {
+      "epoch": 2.5931477516059958,
+      "grad_norm": 0.19164944758078695,
+      "learning_rate": 7.5396825396825394e-06,
+      "loss": 0.2705,
+      "step": 1211
+    },
+    {
+      "epoch": 2.595289079229122,
+      "grad_norm": 0.1946688680184309,
+      "learning_rate": 7.5e-06,
+      "loss": 0.2817,
+      "step": 1212
+    },
+    {
+      "epoch": 2.5974304068522485,
+      "grad_norm": 0.17995484096025638,
+      "learning_rate": 7.460317460317461e-06,
+      "loss": 0.2763,
+      "step": 1213
+    },
+    {
+      "epoch": 2.599571734475375,
+      "grad_norm": 0.18827882235831994,
+      "learning_rate": 7.420634920634921e-06,
+      "loss": 0.2649,
+      "step": 1214
+    },
+    {
+      "epoch": 2.601713062098501,
+      "grad_norm": 0.1911172359855126,
+      "learning_rate": 7.380952380952382e-06,
+      "loss": 0.2785,
+      "step": 1215
+    },
+    {
+      "epoch": 2.6038543897216275,
+      "grad_norm": 0.20732711252193384,
+      "learning_rate": 7.3412698412698415e-06,
+      "loss": 0.2956,
+      "step": 1216
+    },
+    {
+      "epoch": 2.605995717344754,
+      "grad_norm": 0.19615616622963164,
+      "learning_rate": 7.301587301587302e-06,
+      "loss": 0.2961,
+      "step": 1217
+    },
+    {
+      "epoch": 2.6081370449678802,
+      "grad_norm": 0.18728761641057254,
+      "learning_rate": 7.261904761904763e-06,
+      "loss": 0.2884,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6102783725910066,
+      "grad_norm": 0.18873348095093828,
+      "learning_rate": 7.222222222222222e-06,
+      "loss": 0.2788,
+      "step": 1219
+    },
+    {
+      "epoch": 2.612419700214133,
+      "grad_norm": 0.1895538111432152,
+      "learning_rate": 7.182539682539682e-06,
+      "loss": 0.2874,
+      "step": 1220
+    },
+    {
+      "epoch": 2.614561027837259,
+      "grad_norm": 0.17912222034312614,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.2712,
+      "step": 1221
+    },
+    {
+      "epoch": 2.6167023554603857,
+      "grad_norm": 0.18895587187642374,
+      "learning_rate": 7.1031746031746035e-06,
+      "loss": 0.2777,
+      "step": 1222
+    },
+    {
+      "epoch": 2.6188436830835116,
+      "grad_norm": 0.18078738265388186,
+      "learning_rate": 7.063492063492063e-06,
+      "loss": 0.282,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6209850107066384,
+      "grad_norm": 0.1761591551782532,
+      "learning_rate": 7.023809523809524e-06,
+      "loss": 0.2746,
+      "step": 1224
+    },
+    {
+      "epoch": 2.6231263383297643,
+      "grad_norm": 0.18492923876395373,
+      "learning_rate": 6.984126984126985e-06,
+      "loss": 0.2758,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6252676659528906,
+      "grad_norm": 0.18920352772469154,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.2711,
+      "step": 1226
+    },
+    {
+      "epoch": 2.627408993576017,
+      "grad_norm": 0.18550091010732825,
+      "learning_rate": 6.9047619047619055e-06,
+      "loss": 0.2697,
+      "step": 1227
+    },
+    {
+      "epoch": 2.6295503211991433,
+      "grad_norm": 0.1785541730706682,
+      "learning_rate": 6.865079365079366e-06,
+      "loss": 0.2763,
+      "step": 1228
+    },
+    {
+      "epoch": 2.6316916488222697,
+      "grad_norm": 0.17857492350218257,
+      "learning_rate": 6.825396825396825e-06,
+      "loss": 0.2691,
+      "step": 1229
+    },
+    {
+      "epoch": 2.633832976445396,
+      "grad_norm": 0.1753222130027031,
+      "learning_rate": 6.785714285714285e-06,
+      "loss": 0.2717,
+      "step": 1230
+    },
+    {
+      "epoch": 2.6359743040685224,
+      "grad_norm": 0.19278231524559503,
+      "learning_rate": 6.746031746031746e-06,
+      "loss": 0.2784,
+      "step": 1231
+    },
+    {
+      "epoch": 2.6381156316916488,
+      "grad_norm": 0.20070264553264905,
+      "learning_rate": 6.706349206349207e-06,
+      "loss": 0.2987,
+      "step": 1232
+    },
+    {
+      "epoch": 2.640256959314775,
+      "grad_norm": 0.2010595496902971,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3015,
+      "step": 1233
+    },
+    {
+      "epoch": 2.6423982869379015,
+      "grad_norm": 0.18108631640949424,
+      "learning_rate": 6.626984126984127e-06,
+      "loss": 0.2814,
+      "step": 1234
+    },
+    {
+      "epoch": 2.644539614561028,
+      "grad_norm": 0.18555134865906892,
+      "learning_rate": 6.587301587301588e-06,
+      "loss": 0.2677,
+      "step": 1235
+    },
+    {
+      "epoch": 2.646680942184154,
+      "grad_norm": 0.18862946658278595,
+      "learning_rate": 6.547619047619048e-06,
+      "loss": 0.2738,
+      "step": 1236
+    },
+    {
+      "epoch": 2.6488222698072805,
+      "grad_norm": 0.1859370117993973,
+      "learning_rate": 6.507936507936509e-06,
+      "loss": 0.2845,
+      "step": 1237
+    },
+    {
+      "epoch": 2.650963597430407,
+      "grad_norm": 0.18716469379849815,
+      "learning_rate": 6.4682539682539696e-06,
+      "loss": 0.2716,
+      "step": 1238
+    },
+    {
+      "epoch": 2.6531049250535332,
+      "grad_norm": 0.18552514906798132,
+      "learning_rate": 6.428571428571429e-06,
+      "loss": 0.2742,
+      "step": 1239
+    },
+    {
+      "epoch": 2.6552462526766596,
+      "grad_norm": 0.19953825231892838,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.2964,
+      "step": 1240
+    },
+    {
+      "epoch": 2.657387580299786,
+      "grad_norm": 0.18682163268591973,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.2692,
+      "step": 1241
+    },
+    {
+      "epoch": 2.6595289079229123,
+      "grad_norm": 0.1794385339174016,
+      "learning_rate": 6.30952380952381e-06,
+      "loss": 0.2703,
+      "step": 1242
+    },
+    {
+      "epoch": 2.6616702355460387,
+      "grad_norm": 0.1887882853913073,
+      "learning_rate": 6.26984126984127e-06,
+      "loss": 0.2874,
+      "step": 1243
+    },
+    {
+      "epoch": 2.663811563169165,
+      "grad_norm": 0.1872385299129515,
+      "learning_rate": 6.230158730158731e-06,
+      "loss": 0.2863,
+      "step": 1244
+    },
+    {
+      "epoch": 2.6659528907922914,
+      "grad_norm": 0.1819046482781277,
+      "learning_rate": 6.190476190476191e-06,
+      "loss": 0.2705,
+      "step": 1245
+    },
+    {
+      "epoch": 2.6680942184154177,
+      "grad_norm": 0.19114242870168724,
+      "learning_rate": 6.1507936507936505e-06,
+      "loss": 0.2782,
+      "step": 1246
+    },
+    {
+      "epoch": 2.670235546038544,
+      "grad_norm": 0.19244904517702394,
+      "learning_rate": 6.111111111111111e-06,
+      "loss": 0.2683,
+      "step": 1247
+    },
+    {
+      "epoch": 2.67237687366167,
+      "grad_norm": 0.18118655863654376,
+      "learning_rate": 6.071428571428572e-06,
+      "loss": 0.2776,
+      "step": 1248
+    },
+    {
+      "epoch": 2.674518201284797,
+      "grad_norm": 0.1895477899880136,
+      "learning_rate": 6.031746031746032e-06,
+      "loss": 0.2955,
+      "step": 1249
+    },
+    {
+      "epoch": 2.6766595289079227,
+      "grad_norm": 0.2037032863019117,
+      "learning_rate": 5.992063492063493e-06,
+      "loss": 0.278,
+      "step": 1250
+    },
+    {
+      "epoch": 2.6788008565310495,
+      "grad_norm": 0.215090909915821,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 0.2779,
+      "step": 1251
+    },
+    {
+      "epoch": 2.6809421841541754,
+      "grad_norm": 0.1806317361792395,
+      "learning_rate": 5.9126984126984124e-06,
+      "loss": 0.279,
+      "step": 1252
+    },
+    {
+      "epoch": 2.683083511777302,
+      "grad_norm": 0.17940601877117468,
+      "learning_rate": 5.873015873015873e-06,
+      "loss": 0.2834,
+      "step": 1253
+    },
+    {
+      "epoch": 2.685224839400428,
+      "grad_norm": 0.181650657819111,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.2867,
+      "step": 1254
+    },
+    {
+      "epoch": 2.6873661670235545,
+      "grad_norm": 0.19030036855437318,
+      "learning_rate": 5.793650793650794e-06,
+      "loss": 0.2742,
+      "step": 1255
+    },
+    {
+      "epoch": 2.689507494646681,
+      "grad_norm": 0.18217728969404254,
+      "learning_rate": 5.753968253968254e-06,
+      "loss": 0.2798,
+      "step": 1256
+    },
+    {
+      "epoch": 2.691648822269807,
+      "grad_norm": 0.17476001635638663,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.2736,
+      "step": 1257
+    },
+    {
+      "epoch": 2.6937901498929335,
+      "grad_norm": 0.16928638733834248,
+      "learning_rate": 5.674603174603174e-06,
+      "loss": 0.2695,
+      "step": 1258
+    },
+    {
+      "epoch": 2.69593147751606,
+      "grad_norm": 0.17143125336122542,
+      "learning_rate": 5.634920634920635e-06,
+      "loss": 0.2746,
+      "step": 1259
+    },
+    {
+      "epoch": 2.6980728051391862,
+      "grad_norm": 0.17538867599551303,
+      "learning_rate": 5.595238095238096e-06,
+      "loss": 0.2829,
+      "step": 1260
+    },
+    {
+      "epoch": 2.7002141327623126,
+      "grad_norm": 0.17983290126938117,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.2702,
+      "step": 1261
+    },
+    {
+      "epoch": 2.702355460385439,
+      "grad_norm": 0.17677089972588253,
+      "learning_rate": 5.515873015873016e-06,
+      "loss": 0.2843,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7044967880085653,
+      "grad_norm": 0.1844485018306519,
+      "learning_rate": 5.4761904761904765e-06,
+      "loss": 0.2813,
+      "step": 1263
+    },
+    {
+      "epoch": 2.7066381156316917,
+      "grad_norm": 0.18659859307800958,
+      "learning_rate": 5.436507936507937e-06,
+      "loss": 0.2795,
+      "step": 1264
+    },
+    {
+      "epoch": 2.708779443254818,
+      "grad_norm": 0.17342140616127946,
+      "learning_rate": 5.396825396825397e-06,
+      "loss": 0.2905,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7109207708779444,
+      "grad_norm": 0.1751230736265568,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.2769,
+      "step": 1266
+    },
+    {
+      "epoch": 2.7130620985010707,
+      "grad_norm": 0.18568051901207977,
+      "learning_rate": 5.317460317460318e-06,
+      "loss": 0.2744,
+      "step": 1267
+    },
+    {
+      "epoch": 2.715203426124197,
+      "grad_norm": 0.18188150911305687,
+      "learning_rate": 5.277777777777778e-06,
+      "loss": 0.2811,
+      "step": 1268
+    },
+    {
+      "epoch": 2.7173447537473234,
+      "grad_norm": 0.18160082816537443,
+      "learning_rate": 5.2380952380952384e-06,
+      "loss": 0.2776,
+      "step": 1269
+    },
+    {
+      "epoch": 2.71948608137045,
+      "grad_norm": 0.178758160602488,
+      "learning_rate": 5.198412698412699e-06,
+      "loss": 0.2651,
+      "step": 1270
+    },
+    {
+      "epoch": 2.721627408993576,
+      "grad_norm": 0.17965972231368907,
+      "learning_rate": 5.158730158730159e-06,
+      "loss": 0.2818,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7237687366167025,
+      "grad_norm": 0.19233800455601202,
+      "learning_rate": 5.119047619047619e-06,
+      "loss": 0.2899,
+      "step": 1272
+    },
+    {
+      "epoch": 2.725910064239829,
+      "grad_norm": 0.2193361858358134,
+      "learning_rate": 5.07936507936508e-06,
+      "loss": 0.3,
+      "step": 1273
+    },
+    {
+      "epoch": 2.728051391862955,
+      "grad_norm": 0.19832192608275423,
+      "learning_rate": 5.03968253968254e-06,
+      "loss": 0.2809,
+      "step": 1274
+    },
+    {
+      "epoch": 2.730192719486081,
+      "grad_norm": 0.18205252478668657,
+      "learning_rate": 5e-06,
+      "loss": 0.276,
+      "step": 1275
+    },
+    {
+      "epoch": 2.732334047109208,
+      "grad_norm": 0.18256411369477107,
+      "learning_rate": 4.96031746031746e-06,
+      "loss": 0.2827,
+      "step": 1276
+    },
+    {
+      "epoch": 2.734475374732334,
+      "grad_norm": 0.174235840356806,
+      "learning_rate": 4.920634920634921e-06,
+      "loss": 0.2824,
+      "step": 1277
+    },
+    {
+      "epoch": 2.7366167023554606,
+      "grad_norm": 0.18107414207263575,
+      "learning_rate": 4.880952380952381e-06,
+      "loss": 0.287,
+      "step": 1278
+    },
+    {
+      "epoch": 2.7387580299785865,
+      "grad_norm": 0.1843871245288681,
+      "learning_rate": 4.841269841269842e-06,
+      "loss": 0.2948,
+      "step": 1279
+    },
+    {
+      "epoch": 2.7408993576017133,
+      "grad_norm": 0.18283457450751056,
+      "learning_rate": 4.8015873015873025e-06,
+      "loss": 0.2657,
+      "step": 1280
+    },
+    {
+      "epoch": 2.7430406852248392,
+      "grad_norm": 0.1802921251741436,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.2813,
+      "step": 1281
+    },
+    {
+      "epoch": 2.7451820128479656,
+      "grad_norm": 0.17894436649475992,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.2775,
+      "step": 1282
+    },
+    {
+      "epoch": 2.747323340471092,
+      "grad_norm": 0.1744670375408823,
+      "learning_rate": 4.682539682539683e-06,
+      "loss": 0.2629,
+      "step": 1283
+    },
+    {
+      "epoch": 2.7494646680942183,
+      "grad_norm": 0.18257816801514468,
+      "learning_rate": 4.642857142857143e-06,
+      "loss": 0.2803,
+      "step": 1284
+    },
+    {
+      "epoch": 2.7516059957173447,
+      "grad_norm": 0.19392465112933904,
+      "learning_rate": 4.603174603174604e-06,
+      "loss": 0.2836,
+      "step": 1285
+    },
+    {
+      "epoch": 2.753747323340471,
+      "grad_norm": 0.18598309828859388,
+      "learning_rate": 4.563492063492064e-06,
+      "loss": 0.2703,
+      "step": 1286
+    },
+    {
+      "epoch": 2.7558886509635974,
+      "grad_norm": 0.1844599683925985,
+      "learning_rate": 4.5238095238095235e-06,
+      "loss": 0.2928,
+      "step": 1287
+    },
+    {
+      "epoch": 2.7580299785867237,
+      "grad_norm": 0.17453928950018807,
+      "learning_rate": 4.484126984126984e-06,
+      "loss": 0.2772,
+      "step": 1288
+    },
+    {
+      "epoch": 2.76017130620985,
+      "grad_norm": 0.1825967298341021,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 0.2837,
+      "step": 1289
+    },
+    {
+      "epoch": 2.7623126338329764,
+      "grad_norm": 0.17501549592815271,
+      "learning_rate": 4.404761904761905e-06,
+      "loss": 0.2708,
+      "step": 1290
+    },
+    {
+      "epoch": 2.764453961456103,
+      "grad_norm": 0.17613103586860301,
+      "learning_rate": 4.365079365079365e-06,
+      "loss": 0.2812,
+      "step": 1291
+    },
+    {
+      "epoch": 2.766595289079229,
+      "grad_norm": 0.18330117899392398,
+      "learning_rate": 4.3253968253968256e-06,
+      "loss": 0.2606,
+      "step": 1292
+    },
+    {
+      "epoch": 2.7687366167023555,
+      "grad_norm": 0.18120028818225953,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.284,
+      "step": 1293
+    },
+    {
+      "epoch": 2.770877944325482,
+      "grad_norm": 0.17679996406713563,
+      "learning_rate": 4.246031746031746e-06,
+      "loss": 0.2772,
+      "step": 1294
+    },
+    {
+      "epoch": 2.773019271948608,
+      "grad_norm": 0.17472312513614138,
+      "learning_rate": 4.206349206349207e-06,
+      "loss": 0.281,
+      "step": 1295
+    },
+    {
+      "epoch": 2.7751605995717346,
+      "grad_norm": 0.17058631787650447,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.2867,
+      "step": 1296
+    },
+    {
+      "epoch": 2.777301927194861,
+      "grad_norm": 0.1724204672936674,
+      "learning_rate": 4.126984126984127e-06,
+      "loss": 0.2753,
+      "step": 1297
+    },
+    {
+      "epoch": 2.7794432548179873,
+      "grad_norm": 0.1783817683336482,
+      "learning_rate": 4.0873015873015875e-06,
+      "loss": 0.2734,
+      "step": 1298
+    },
+    {
+      "epoch": 2.7815845824411136,
+      "grad_norm": 0.1768963927750415,
+      "learning_rate": 4.047619047619048e-06,
+      "loss": 0.286,
+      "step": 1299
+    },
+    {
+      "epoch": 2.78372591006424,
+      "grad_norm": 0.1785831360758202,
+      "learning_rate": 4.007936507936508e-06,
+      "loss": 0.288,
+      "step": 1300
+    },
+    {
+      "epoch": 2.7858672376873663,
+      "grad_norm": 0.1973578170792123,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.2851,
+      "step": 1301
+    },
+    {
+      "epoch": 2.7880085653104922,
+      "grad_norm": 0.17966086659509484,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.2764,
+      "step": 1302
+    },
+    {
+      "epoch": 2.790149892933619,
+      "grad_norm": 0.18832071036234518,
+      "learning_rate": 3.888888888888889e-06,
+      "loss": 0.2722,
+      "step": 1303
+    },
+    {
+      "epoch": 2.792291220556745,
+      "grad_norm": 0.17616870935040177,
+      "learning_rate": 3.8492063492063495e-06,
+      "loss": 0.2799,
+      "step": 1304
+    },
+    {
+      "epoch": 2.7944325481798717,
+      "grad_norm": 0.18136072174228499,
+      "learning_rate": 3.8095238095238102e-06,
+      "loss": 0.28,
+      "step": 1305
+    },
+    {
+      "epoch": 2.7965738758029977,
+      "grad_norm": 0.19260217036977118,
+      "learning_rate": 3.7698412698412697e-06,
+      "loss": 0.2795,
+      "step": 1306
+    },
+    {
+      "epoch": 2.7987152034261245,
+      "grad_norm": 0.17281606530646676,
+      "learning_rate": 3.7301587301587305e-06,
+      "loss": 0.2694,
+      "step": 1307
+    },
+    {
+      "epoch": 2.8008565310492504,
+      "grad_norm": 0.17227995573571817,
+      "learning_rate": 3.690476190476191e-06,
+      "loss": 0.2642,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8029978586723767,
+      "grad_norm": 0.17354690029249242,
+      "learning_rate": 3.650793650793651e-06,
+      "loss": 0.2743,
+      "step": 1309
+    },
+    {
+      "epoch": 2.805139186295503,
+      "grad_norm": 0.18596920138748899,
+      "learning_rate": 3.611111111111111e-06,
+      "loss": 0.2666,
+      "step": 1310
+    },
+    {
+      "epoch": 2.8072805139186294,
+      "grad_norm": 0.19988273525925654,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.2939,
+      "step": 1311
+    },
+    {
+      "epoch": 2.809421841541756,
+      "grad_norm": 0.18393678561262797,
+      "learning_rate": 3.5317460317460317e-06,
+      "loss": 0.2927,
+      "step": 1312
+    },
+    {
+      "epoch": 2.811563169164882,
+      "grad_norm": 0.18018691120657057,
+      "learning_rate": 3.4920634920634924e-06,
+      "loss": 0.2664,
+      "step": 1313
+    },
+    {
+      "epoch": 2.8137044967880085,
+      "grad_norm": 0.18530712590206283,
+      "learning_rate": 3.4523809523809528e-06,
+      "loss": 0.277,
+      "step": 1314
+    },
+    {
+      "epoch": 2.815845824411135,
+      "grad_norm": 0.18226232243626736,
+      "learning_rate": 3.4126984126984127e-06,
+      "loss": 0.281,
+      "step": 1315
+    },
+    {
+      "epoch": 2.817987152034261,
+      "grad_norm": 0.19031200668984619,
+      "learning_rate": 3.373015873015873e-06,
+      "loss": 0.2866,
+      "step": 1316
+    },
+    {
+      "epoch": 2.8201284796573876,
+      "grad_norm": 0.1746390900856267,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2805,
+      "step": 1317
+    },
+    {
+      "epoch": 2.822269807280514,
+      "grad_norm": 0.17409052183537008,
+      "learning_rate": 3.293650793650794e-06,
+      "loss": 0.2748,
+      "step": 1318
+    },
+    {
+      "epoch": 2.8244111349036403,
+      "grad_norm": 0.1782862674279465,
+      "learning_rate": 3.2539682539682544e-06,
+      "loss": 0.271,
+      "step": 1319
+    },
+    {
+      "epoch": 2.8265524625267666,
+      "grad_norm": 0.18225874553079588,
+      "learning_rate": 3.2142857142857143e-06,
+      "loss": 0.2602,
+      "step": 1320
+    },
+    {
+      "epoch": 2.828693790149893,
+      "grad_norm": 0.17783389559399063,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.2762,
+      "step": 1321
+    },
+    {
+      "epoch": 2.8308351177730193,
+      "grad_norm": 0.17855207309001997,
+      "learning_rate": 3.134920634920635e-06,
+      "loss": 0.2872,
+      "step": 1322
+    },
+    {
+      "epoch": 2.8329764453961457,
+      "grad_norm": 0.17935885821626021,
+      "learning_rate": 3.0952380952380953e-06,
+      "loss": 0.2794,
+      "step": 1323
+    },
+    {
+      "epoch": 2.835117773019272,
+      "grad_norm": 0.18232652024256538,
+      "learning_rate": 3.0555555555555556e-06,
+      "loss": 0.2831,
+      "step": 1324
+    },
+    {
+      "epoch": 2.8372591006423984,
+      "grad_norm": 0.17948962840127805,
+      "learning_rate": 3.015873015873016e-06,
+      "loss": 0.2776,
+      "step": 1325
+    },
+    {
+      "epoch": 2.8394004282655247,
+      "grad_norm": 0.18464576865432494,
+      "learning_rate": 2.9761904761904763e-06,
+      "loss": 0.2941,
+      "step": 1326
+    },
+    {
+      "epoch": 2.841541755888651,
+      "grad_norm": 0.17028939672073803,
+      "learning_rate": 2.9365079365079366e-06,
+      "loss": 0.2672,
+      "step": 1327
+    },
+    {
+      "epoch": 2.8436830835117775,
+      "grad_norm": 0.16916196038993978,
+      "learning_rate": 2.896825396825397e-06,
+      "loss": 0.2741,
+      "step": 1328
+    },
+    {
+      "epoch": 2.8458244111349034,
+      "grad_norm": 0.175350985222415,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2796,
+      "step": 1329
+    },
+    {
+      "epoch": 2.84796573875803,
+      "grad_norm": 0.17515671194072407,
+      "learning_rate": 2.8174603174603176e-06,
+      "loss": 0.2775,
+      "step": 1330
+    },
+    {
+      "epoch": 2.850107066381156,
+      "grad_norm": 0.17259697407853658,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.2688,
+      "step": 1331
+    },
+    {
+      "epoch": 2.852248394004283,
+      "grad_norm": 0.1739659486779704,
+      "learning_rate": 2.7380952380952382e-06,
+      "loss": 0.2842,
+      "step": 1332
+    },
+    {
+      "epoch": 2.854389721627409,
+      "grad_norm": 0.1897848574714334,
+      "learning_rate": 2.6984126984126986e-06,
+      "loss": 0.2736,
+      "step": 1333
+    },
+    {
+      "epoch": 2.8565310492505356,
+      "grad_norm": 0.17285981110921295,
+      "learning_rate": 2.658730158730159e-06,
+      "loss": 0.2689,
+      "step": 1334
+    },
+    {
+      "epoch": 2.8586723768736615,
+      "grad_norm": 0.18566863960917998,
+      "learning_rate": 2.6190476190476192e-06,
+      "loss": 0.2656,
+      "step": 1335
+    },
+    {
+      "epoch": 2.860813704496788,
+      "grad_norm": 0.17490830341679464,
+      "learning_rate": 2.5793650793650795e-06,
+      "loss": 0.282,
+      "step": 1336
+    },
+    {
+      "epoch": 2.862955032119914,
+      "grad_norm": 0.16847205946436153,
+      "learning_rate": 2.53968253968254e-06,
+      "loss": 0.2686,
+      "step": 1337
+    },
+    {
+      "epoch": 2.8650963597430406,
+      "grad_norm": 0.167671663902774,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2723,
+      "step": 1338
+    },
+    {
+      "epoch": 2.867237687366167,
+      "grad_norm": 0.17505967809467687,
+      "learning_rate": 2.4603174603174605e-06,
+      "loss": 0.2741,
+      "step": 1339
+    },
+    {
+      "epoch": 2.8693790149892933,
+      "grad_norm": 0.18340786011909083,
+      "learning_rate": 2.420634920634921e-06,
+      "loss": 0.2929,
+      "step": 1340
+    },
+    {
+      "epoch": 2.8715203426124196,
+      "grad_norm": 0.17940501576635812,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.2714,
+      "step": 1341
+    },
+    {
+      "epoch": 2.873661670235546,
+      "grad_norm": 0.182484237285869,
+      "learning_rate": 2.3412698412698415e-06,
+      "loss": 0.275,
+      "step": 1342
+    },
+    {
+      "epoch": 2.8758029978586723,
+      "grad_norm": 0.18020814887479553,
+      "learning_rate": 2.301587301587302e-06,
+      "loss": 0.2825,
+      "step": 1343
+    },
+    {
+      "epoch": 2.8779443254817987,
+      "grad_norm": 0.17660806936645942,
+      "learning_rate": 2.2619047619047617e-06,
+      "loss": 0.285,
+      "step": 1344
+    },
+    {
+      "epoch": 2.880085653104925,
+      "grad_norm": 0.17617046995854874,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 0.2754,
+      "step": 1345
+    },
+    {
+      "epoch": 2.8822269807280514,
+      "grad_norm": 0.1744513421687539,
+      "learning_rate": 2.1825396825396824e-06,
+      "loss": 0.2951,
+      "step": 1346
+    },
+    {
+      "epoch": 2.8843683083511777,
+      "grad_norm": 0.17176452515074214,
+      "learning_rate": 2.142857142857143e-06,
+      "loss": 0.2668,
+      "step": 1347
+    },
+    {
+      "epoch": 2.886509635974304,
+      "grad_norm": 0.17904296673736972,
+      "learning_rate": 2.1031746031746035e-06,
+      "loss": 0.2701,
+      "step": 1348
+    },
+    {
+      "epoch": 2.8886509635974305,
+      "grad_norm": 0.17647016428567208,
+      "learning_rate": 2.0634920634920634e-06,
+      "loss": 0.2716,
+      "step": 1349
+    },
+    {
+      "epoch": 2.890792291220557,
+      "grad_norm": 0.17235816276714014,
+      "learning_rate": 2.023809523809524e-06,
+      "loss": 0.265,
+      "step": 1350
+    },
+    {
+      "epoch": 2.892933618843683,
+      "grad_norm": 0.18658165461489193,
+      "learning_rate": 1.984126984126984e-06,
+      "loss": 0.2799,
+      "step": 1351
+    },
+    {
+      "epoch": 2.8950749464668095,
+      "grad_norm": 0.17186186390295335,
+      "learning_rate": 1.9444444444444444e-06,
+      "loss": 0.2714,
+      "step": 1352
+    },
+    {
+      "epoch": 2.897216274089936,
+      "grad_norm": 0.17654423828340407,
+      "learning_rate": 1.9047619047619051e-06,
+      "loss": 0.2748,
+      "step": 1353
+    },
+    {
+      "epoch": 2.8993576017130622,
+      "grad_norm": 0.17779611321463967,
+      "learning_rate": 1.8650793650793652e-06,
+      "loss": 0.2817,
+      "step": 1354
+    },
+    {
+      "epoch": 2.9014989293361886,
+      "grad_norm": 0.16906995082417278,
+      "learning_rate": 1.8253968253968256e-06,
+      "loss": 0.2779,
+      "step": 1355
+    },
+    {
+      "epoch": 2.903640256959315,
+      "grad_norm": 0.1795175517079968,
+      "learning_rate": 1.7857142857142857e-06,
+      "loss": 0.275,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9057815845824413,
+      "grad_norm": 0.24069812689827963,
+      "learning_rate": 1.7460317460317462e-06,
+      "loss": 0.2842,
+      "step": 1357
+    },
+    {
+      "epoch": 2.907922912205567,
+      "grad_norm": 0.18261102094972528,
+      "learning_rate": 1.7063492063492063e-06,
+      "loss": 0.2863,
+      "step": 1358
+    },
+    {
+      "epoch": 2.910064239828694,
+      "grad_norm": 0.17916467521323348,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.2703,
+      "step": 1359
+    },
+    {
+      "epoch": 2.91220556745182,
+      "grad_norm": 0.1684177935210992,
+      "learning_rate": 1.6269841269841272e-06,
+      "loss": 0.2633,
+      "step": 1360
+    },
+    {
+      "epoch": 2.9143468950749467,
+      "grad_norm": 0.17482795526886843,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.2733,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9164882226980726,
+      "grad_norm": 0.17189677626587546,
+      "learning_rate": 1.5476190476190476e-06,
+      "loss": 0.2786,
+      "step": 1362
+    },
+    {
+      "epoch": 2.9186295503211994,
+      "grad_norm": 0.1793311737015449,
+      "learning_rate": 1.507936507936508e-06,
+      "loss": 0.3015,
+      "step": 1363
+    },
+    {
+      "epoch": 2.9207708779443253,
+      "grad_norm": 0.1751443843257153,
+      "learning_rate": 1.4682539682539683e-06,
+      "loss": 0.2701,
+      "step": 1364
+    },
+    {
+      "epoch": 2.9229122055674517,
+      "grad_norm": 0.17885380423708533,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.2732,
+      "step": 1365
+    },
+    {
+      "epoch": 2.925053533190578,
+      "grad_norm": 0.17570565566220792,
+      "learning_rate": 1.388888888888889e-06,
+      "loss": 0.2797,
+      "step": 1366
+    },
+    {
+      "epoch": 2.9271948608137044,
+      "grad_norm": 0.1665971208193023,
+      "learning_rate": 1.3492063492063493e-06,
+      "loss": 0.2726,
+      "step": 1367
+    },
+    {
+      "epoch": 2.9293361884368307,
+      "grad_norm": 0.18707229378818965,
+      "learning_rate": 1.3095238095238096e-06,
+      "loss": 0.3001,
+      "step": 1368
+    },
+    {
+      "epoch": 2.931477516059957,
+      "grad_norm": 0.172875108848975,
+      "learning_rate": 1.26984126984127e-06,
+      "loss": 0.2817,
+      "step": 1369
+    },
+    {
+      "epoch": 2.9336188436830835,
+      "grad_norm": 0.17022686353646632,
+      "learning_rate": 1.2301587301587303e-06,
+      "loss": 0.2776,
+      "step": 1370
+    },
+    {
+      "epoch": 2.93576017130621,
+      "grad_norm": 0.17967995503693934,
+      "learning_rate": 1.1904761904761904e-06,
+      "loss": 0.2683,
+      "step": 1371
+    },
+    {
+      "epoch": 2.937901498929336,
+      "grad_norm": 0.17708372843749126,
+      "learning_rate": 1.150793650793651e-06,
+      "loss": 0.2908,
+      "step": 1372
+    },
+    {
+      "epoch": 2.9400428265524625,
+      "grad_norm": 0.16830115974661627,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 0.2561,
+      "step": 1373
+    },
+    {
+      "epoch": 2.942184154175589,
+      "grad_norm": 0.17929164665415462,
+      "learning_rate": 1.0714285714285716e-06,
+      "loss": 0.2714,
+      "step": 1374
+    },
+    {
+      "epoch": 2.9443254817987152,
+      "grad_norm": 0.16840343711745376,
+      "learning_rate": 1.0317460317460317e-06,
+      "loss": 0.2671,
+      "step": 1375
+    },
+    {
+      "epoch": 2.9464668094218416,
+      "grad_norm": 0.17040496165809252,
+      "learning_rate": 9.92063492063492e-07,
+      "loss": 0.2711,
+      "step": 1376
+    },
+    {
+      "epoch": 2.948608137044968,
+      "grad_norm": 0.17750648853336623,
+      "learning_rate": 9.523809523809526e-07,
+      "loss": 0.2661,
+      "step": 1377
+    },
+    {
+      "epoch": 2.9507494646680943,
+      "grad_norm": 0.17641859608757207,
+      "learning_rate": 9.126984126984128e-07,
+      "loss": 0.282,
+      "step": 1378
+    },
+    {
+      "epoch": 2.9528907922912206,
+      "grad_norm": 0.18027367970879757,
+      "learning_rate": 8.730158730158731e-07,
+      "loss": 0.2795,
+      "step": 1379
+    },
+    {
+      "epoch": 2.955032119914347,
+      "grad_norm": 0.1770116936691827,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.2846,
+      "step": 1380
+    },
+    {
+      "epoch": 2.9571734475374734,
+      "grad_norm": 0.1718006942921136,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.2812,
+      "step": 1381
+    },
+    {
+      "epoch": 2.9593147751605997,
+      "grad_norm": 0.1770273339812149,
+      "learning_rate": 7.53968253968254e-07,
+      "loss": 0.2896,
+      "step": 1382
+    },
+    {
+      "epoch": 2.961456102783726,
+      "grad_norm": 0.17902070118474797,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.2978,
+      "step": 1383
+    },
+    {
+      "epoch": 2.9635974304068524,
+      "grad_norm": 0.17533054931108796,
+      "learning_rate": 6.746031746031746e-07,
+      "loss": 0.2769,
+      "step": 1384
+    },
+    {
+      "epoch": 2.9657387580299783,
+      "grad_norm": 0.16993493071735802,
+      "learning_rate": 6.34920634920635e-07,
+      "loss": 0.2827,
+      "step": 1385
+    },
+    {
+      "epoch": 2.967880085653105,
+      "grad_norm": 0.17684972171074492,
+      "learning_rate": 5.952380952380952e-07,
+      "loss": 0.2917,
+      "step": 1386
+    },
+    {
+      "epoch": 2.970021413276231,
+      "grad_norm": 0.17897804806977222,
+      "learning_rate": 5.555555555555556e-07,
+      "loss": 0.2834,
+      "step": 1387
+    },
+    {
+      "epoch": 2.972162740899358,
+      "grad_norm": 0.19557715172028603,
+      "learning_rate": 5.158730158730158e-07,
+      "loss": 0.3017,
+      "step": 1388
+    },
+    {
+      "epoch": 2.9743040685224837,
+      "grad_norm": 0.17627182121205032,
+      "learning_rate": 4.761904761904763e-07,
+      "loss": 0.2767,
+      "step": 1389
+    },
+    {
+      "epoch": 2.9764453961456105,
+      "grad_norm": 0.17369058443780255,
+      "learning_rate": 4.3650793650793655e-07,
+      "loss": 0.2798,
+      "step": 1390
+    },
+    {
+      "epoch": 2.9785867237687365,
+      "grad_norm": 0.1740941045189046,
+      "learning_rate": 3.9682539682539683e-07,
+      "loss": 0.2683,
+      "step": 1391
+    },
+    {
+      "epoch": 2.980728051391863,
+      "grad_norm": 0.17915684223215836,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 0.2873,
+      "step": 1392
+    },
+    {
+      "epoch": 2.982869379014989,
+      "grad_norm": 0.1679332636215031,
+      "learning_rate": 3.174603174603175e-07,
+      "loss": 0.2745,
+      "step": 1393
+    },
+    {
+      "epoch": 2.9850107066381155,
+      "grad_norm": 0.17729248198099018,
+      "learning_rate": 2.777777777777778e-07,
+      "loss": 0.2756,
+      "step": 1394
+    },
+    {
+      "epoch": 2.987152034261242,
+      "grad_norm": 0.16814830396556607,
+      "learning_rate": 2.3809523809523814e-07,
+      "loss": 0.2713,
+      "step": 1395
+    },
+    {
+      "epoch": 2.9892933618843682,
+      "grad_norm": 0.1765432649703903,
+      "learning_rate": 1.9841269841269841e-07,
+      "loss": 0.2875,
+      "step": 1396
+    },
+    {
+      "epoch": 2.9914346895074946,
+      "grad_norm": 0.1741654729465487,
+      "learning_rate": 1.5873015873015874e-07,
+      "loss": 0.2909,
+      "step": 1397
+    },
+    {
+      "epoch": 2.993576017130621,
+      "grad_norm": 0.1709774041473495,
+      "learning_rate": 1.1904761904761907e-07,
+      "loss": 0.2796,
+      "step": 1398
+    },
+    {
+      "epoch": 2.9957173447537473,
+      "grad_norm": 0.1642668435658554,
+      "learning_rate": 7.936507936507937e-08,
+      "loss": 0.267,
+      "step": 1399
+    },
+    {
+      "epoch": 2.9978586723768736,
+      "grad_norm": 0.17312311494944055,
+      "learning_rate": 3.9682539682539686e-08,
+      "loss": 0.2641,
+      "step": 1400
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.17221087745505198,
+      "learning_rate": 0.0,
+      "loss": 0.2679,
+      "step": 1401
+    },
+    {
+      "epoch": 3.0,
+      "step": 1401,
+      "total_flos": 1.5578375880118895e+19,
+      "train_loss": 0.39975321637486494,
+      "train_runtime": 116840.5239,
+      "train_samples_per_second": 0.192,
+      "train_steps_per_second": 0.012
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1401,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5578375880118895e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}