diff --git "a/training_logs.json" "b/training_logs.json"
new file mode 100644--- /dev/null
+++ "b/training_logs.json"
@@ -0,0 +1,6861 @@
+[
+    {
+        "loss": 30.8234,
+        "grad_norm": 0.5755094885826111,
+        "learning_rate": 0.0009991248796709547,
+        "epoch": 0.0
+    },
+    {
+        "loss": 24.1229,
+        "grad_norm": 0.2920963764190674,
+        "learning_rate": 0.0009982497593419095,
+        "epoch": 0.01
+    },
+    {
+        "loss": 22.7986,
+        "grad_norm": 0.5106011629104614,
+        "learning_rate": 0.0009973746390128642,
+        "epoch": 0.01
+    },
+    {
+        "loss": 21.4924,
+        "grad_norm": 0.9322375059127808,
+        "learning_rate": 0.000996499518683819,
+        "epoch": 0.01
+    },
+    {
+        "loss": 20.7911,
+        "grad_norm": 0.8529098629951477,
+        "learning_rate": 0.0009956243983547737,
+        "epoch": 0.01
+    },
+    {
+        "loss": 19.338,
+        "grad_norm": 0.776152491569519,
+        "learning_rate": 0.0009947492780257286,
+        "epoch": 0.02
+    },
+    {
+        "loss": 19.0175,
+        "grad_norm": 2.11796498298645,
+        "learning_rate": 0.0009938741576966832,
+        "epoch": 0.02
+    },
+    {
+        "loss": 18.2997,
+        "grad_norm": 1.3791886568069458,
+        "learning_rate": 0.0009929990373676381,
+        "epoch": 0.02
+    },
+    {
+        "loss": 17.2791,
+        "grad_norm": 1.3849037885665894,
+        "learning_rate": 0.0009921239170385928,
+        "epoch": 0.02
+    },
+    {
+        "loss": 17.3609,
+        "grad_norm": 1.1861941814422607,
+        "learning_rate": 0.0009912487967095476,
+        "epoch": 0.03
+    },
+    {
+        "loss": 17.1215,
+        "grad_norm": 1.494122862815857,
+        "learning_rate": 0.0009903736763805023,
+        "epoch": 0.03
+    },
+    {
+        "loss": 16.3944,
+        "grad_norm": 1.5872834920883179,
+        "learning_rate": 0.0009894985560514572,
+        "epoch": 0.03
+    },
+    {
+        "loss": 16.0054,
+        "grad_norm": 1.2658979892730713,
+        "learning_rate": 0.0009886234357224118,
+        "epoch": 0.03
+    },
+    {
+        "loss": 15.5523,
+        "grad_norm": 0.8640480041503906,
+        "learning_rate": 0.0009877483153933667,
+        "epoch": 0.04
+    },
+    {
+        "loss": 16.2465,
+        "grad_norm": 0.8946548700332642,
+        "learning_rate": 0.0009868731950643213,
+        "epoch": 0.04
+    },
+    {
+        "loss": 15.0235,
+        "grad_norm": 0.9279372692108154,
+        "learning_rate": 0.0009859980747352762,
+        "epoch": 0.04
+    },
+    {
+        "loss": 15.7517,
+        "grad_norm": 0.8807494044303894,
+        "learning_rate": 0.0009851229544062309,
+        "epoch": 0.04
+    },
+    {
+        "loss": 14.6884,
+        "grad_norm": 0.683822751045227,
+        "learning_rate": 0.0009842478340771857,
+        "epoch": 0.05
+    },
+    {
+        "loss": 14.0949,
+        "grad_norm": 1.1334095001220703,
+        "learning_rate": 0.0009833727137481404,
+        "epoch": 0.05
+    },
+    {
+        "loss": 14.3378,
+        "grad_norm": 1.1247657537460327,
+        "learning_rate": 0.0009824975934190953,
+        "epoch": 0.05
+    },
+    {
+        "loss": 13.7597,
+        "grad_norm": 0.9332773685455322,
+        "learning_rate": 0.00098162247309005,
+        "epoch": 0.06
+    },
+    {
+        "loss": 14.5567,
+        "grad_norm": 0.8742538690567017,
+        "learning_rate": 0.0009807473527610048,
+        "epoch": 0.06
+    },
+    {
+        "loss": 14.0188,
+        "grad_norm": 1.5592143535614014,
+        "learning_rate": 0.0009798722324319594,
+        "epoch": 0.06
+    },
+    {
+        "loss": 13.9401,
+        "grad_norm": 0.9473065733909607,
+        "learning_rate": 0.0009789971121029143,
+        "epoch": 0.06
+    },
+    {
+        "loss": 13.5177,
+        "grad_norm": 0.5469663143157959,
+        "learning_rate": 0.000978121991773869,
+        "epoch": 0.07
+    },
+    {
+        "loss": 13.513,
+        "grad_norm": 1.7497597932815552,
+        "learning_rate": 0.0009772468714448236,
+        "epoch": 0.07
+    },
+    {
+        "loss": 13.6579,
+        "grad_norm": 0.7552927136421204,
+        "learning_rate": 0.0009763717511157785,
+        "epoch": 0.07
+    },
+    {
+        "loss": 13.0899,
+        "grad_norm": 0.5602779984474182,
+        "learning_rate": 0.0009754966307867332,
+        "epoch": 0.07
+    },
+    {
+        "loss": 13.8637,
+        "grad_norm": 0.6577705144882202,
+        "learning_rate": 0.000974621510457688,
+        "epoch": 0.08
+    },
+    {
+        "loss": 14.2909,
+        "grad_norm": 1.0710817575454712,
+        "learning_rate": 0.0009737463901286428,
+        "epoch": 0.08
+    },
+    {
+        "loss": 13.3632,
+        "grad_norm": 0.48803457617759705,
+        "learning_rate": 0.0009728712697995975,
+        "epoch": 0.08
+    },
+    {
+        "loss": 13.5002,
+        "grad_norm": 0.9970788359642029,
+        "learning_rate": 0.0009719961494705523,
+        "epoch": 0.08
+    },
+    {
+        "loss": 13.6276,
+        "grad_norm": 0.9624769687652588,
+        "learning_rate": 0.000971121029141507,
+        "epoch": 0.09
+    },
+    {
+        "loss": 13.7281,
+        "grad_norm": 0.8082631230354309,
+        "learning_rate": 0.0009702459088124618,
+        "epoch": 0.09
+    },
+    {
+        "loss": 13.0793,
+        "grad_norm": 0.6732771396636963,
+        "learning_rate": 0.0009693707884834166,
+        "epoch": 0.09
+    },
+    {
+        "loss": 12.6621,
+        "grad_norm": 0.8451002240180969,
+        "learning_rate": 0.0009684956681543713,
+        "epoch": 0.09
+    },
+    {
+        "loss": 13.2374,
+        "grad_norm": 1.1656385660171509,
+        "learning_rate": 0.0009676205478253261,
+        "epoch": 0.1
+    },
+    {
+        "loss": 12.7625,
+        "grad_norm": 0.9667061567306519,
+        "learning_rate": 0.0009667454274962808,
+        "epoch": 0.1
+    },
+    {
+        "loss": 13.0046,
+        "grad_norm": 0.9311497807502747,
+        "learning_rate": 0.0009658703071672355,
+        "epoch": 0.1
+    },
+    {
+        "loss": 12.9037,
+        "grad_norm": 1.1891040802001953,
+        "learning_rate": 0.0009649951868381903,
+        "epoch": 0.1
+    },
+    {
+        "loss": 12.6521,
+        "grad_norm": 1.1127817630767822,
+        "learning_rate": 0.000964120066509145,
+        "epoch": 0.11
+    },
+    {
+        "loss": 13.2942,
+        "grad_norm": 0.6665758490562439,
+        "learning_rate": 0.0009632449461800998,
+        "epoch": 0.11
+    },
+    {
+        "loss": 12.4443,
+        "grad_norm": 0.8878126740455627,
+        "learning_rate": 0.0009623698258510546,
+        "epoch": 0.11
+    },
+    {
+        "loss": 13.0001,
+        "grad_norm": 1.5000464916229248,
+        "learning_rate": 0.0009614947055220093,
+        "epoch": 0.12
+    },
+    {
+        "loss": 12.2303,
+        "grad_norm": 1.1078687906265259,
+        "learning_rate": 0.0009606195851929641,
+        "epoch": 0.12
+    },
+    {
+        "loss": 12.1915,
+        "grad_norm": 0.8044748306274414,
+        "learning_rate": 0.0009597444648639187,
+        "epoch": 0.12
+    },
+    {
+        "loss": 12.7246,
+        "grad_norm": 0.9232500195503235,
+        "learning_rate": 0.0009588693445348735,
+        "epoch": 0.12
+    },
+    {
+        "loss": 11.9769,
+        "grad_norm": 0.7413458824157715,
+        "learning_rate": 0.0009579942242058283,
+        "epoch": 0.13
+    },
+    {
+        "loss": 12.8006,
+        "grad_norm": 1.1132707595825195,
+        "learning_rate": 0.000957119103876783,
+        "epoch": 0.13
+    },
+    {
+        "loss": 12.4323,
+        "grad_norm": 0.7814503312110901,
+        "learning_rate": 0.0009562439835477378,
+        "epoch": 0.13
+    },
+    {
+        "loss": 12.3482,
+        "grad_norm": 0.8854762315750122,
+        "learning_rate": 0.0009553688632186925,
+        "epoch": 0.13
+    },
+    {
+        "loss": 12.5045,
+        "grad_norm": 0.704131007194519,
+        "learning_rate": 0.0009544937428896473,
+        "epoch": 0.14
+    },
+    {
+        "loss": 12.1405,
+        "grad_norm": 0.7020297050476074,
+        "learning_rate": 0.0009536186225606021,
+        "epoch": 0.14
+    },
+    {
+        "loss": 11.5427,
+        "grad_norm": 0.398807168006897,
+        "learning_rate": 0.0009527435022315568,
+        "epoch": 0.14
+    },
+    {
+        "loss": 12.655,
+        "grad_norm": 1.0002299547195435,
+        "learning_rate": 0.0009518683819025116,
+        "epoch": 0.14
+    },
+    {
+        "loss": 11.9656,
+        "grad_norm": 0.7870428562164307,
+        "learning_rate": 0.0009509932615734664,
+        "epoch": 0.15
+    },
+    {
+        "loss": 12.4639,
+        "grad_norm": 0.9154604077339172,
+        "learning_rate": 0.0009501181412444211,
+        "epoch": 0.15
+    },
+    {
+        "loss": 11.6344,
+        "grad_norm": 1.1896569728851318,
+        "learning_rate": 0.0009492430209153759,
+        "epoch": 0.15
+    },
+    {
+        "loss": 12.4516,
+        "grad_norm": 0.8169024586677551,
+        "learning_rate": 0.0009483679005863306,
+        "epoch": 0.15
+    },
+    {
+        "loss": 12.1848,
+        "grad_norm": 0.8429264426231384,
+        "learning_rate": 0.0009474927802572854,
+        "epoch": 0.16
+    },
+    {
+        "loss": 11.2014,
+        "grad_norm": 0.8499436378479004,
+        "learning_rate": 0.0009466176599282402,
+        "epoch": 0.16
+    },
+    {
+        "loss": 12.2217,
+        "grad_norm": 0.8969743251800537,
+        "learning_rate": 0.0009457425395991948,
+        "epoch": 0.16
+    },
+    {
+        "loss": 11.7729,
+        "grad_norm": 1.0959218740463257,
+        "learning_rate": 0.0009448674192701496,
+        "epoch": 0.17
+    },
+    {
+        "loss": 11.6254,
+        "grad_norm": 1.1692876815795898,
+        "learning_rate": 0.0009439922989411043,
+        "epoch": 0.17
+    },
+    {
+        "loss": 11.5698,
+        "grad_norm": 1.9476372003555298,
+        "learning_rate": 0.0009431171786120591,
+        "epoch": 0.17
+    },
+    {
+        "loss": 11.4321,
+        "grad_norm": 1.1742662191390991,
+        "learning_rate": 0.0009422420582830139,
+        "epoch": 0.17
+    },
+    {
+        "loss": 11.3224,
+        "grad_norm": 0.9839737415313721,
+        "learning_rate": 0.0009413669379539686,
+        "epoch": 0.18
+    },
+    {
+        "loss": 11.8269,
+        "grad_norm": 0.9094179272651672,
+        "learning_rate": 0.0009404918176249234,
+        "epoch": 0.18
+    },
+    {
+        "loss": 11.8652,
+        "grad_norm": 0.9139958620071411,
+        "learning_rate": 0.0009396166972958782,
+        "epoch": 0.18
+    },
+    {
+        "loss": 11.5493,
+        "grad_norm": 0.7938945889472961,
+        "learning_rate": 0.0009387415769668329,
+        "epoch": 0.18
+    },
+    {
+        "loss": 11.413,
+        "grad_norm": 0.8102487921714783,
+        "learning_rate": 0.0009378664566377877,
+        "epoch": 0.19
+    },
+    {
+        "loss": 11.4015,
+        "grad_norm": 0.5892770290374756,
+        "learning_rate": 0.0009369913363087424,
+        "epoch": 0.19
+    },
+    {
+        "loss": 10.8455,
+        "grad_norm": 0.7269143462181091,
+        "learning_rate": 0.0009361162159796972,
+        "epoch": 0.19
+    },
+    {
+        "loss": 11.5612,
+        "grad_norm": 0.8169882893562317,
+        "learning_rate": 0.000935241095650652,
+        "epoch": 0.19
+    },
+    {
+        "loss": 10.545,
+        "grad_norm": 0.8424365520477295,
+        "learning_rate": 0.0009343659753216067,
+        "epoch": 0.2
+    },
+    {
+        "loss": 10.8486,
+        "grad_norm": 0.855518102645874,
+        "learning_rate": 0.0009334908549925615,
+        "epoch": 0.2
+    },
+    {
+        "loss": 10.3733,
+        "grad_norm": 1.1463903188705444,
+        "learning_rate": 0.0009326157346635162,
+        "epoch": 0.2
+    },
+    {
+        "loss": 10.794,
+        "grad_norm": 0.7493767142295837,
+        "learning_rate": 0.000931740614334471,
+        "epoch": 0.2
+    },
+    {
+        "loss": 10.5943,
+        "grad_norm": 0.8767346739768982,
+        "learning_rate": 0.0009308654940054258,
+        "epoch": 0.21
+    },
+    {
+        "loss": 11.4169,
+        "grad_norm": 1.0650781393051147,
+        "learning_rate": 0.0009299903736763805,
+        "epoch": 0.21
+    },
+    {
+        "loss": 10.8176,
+        "grad_norm": 0.8954362869262695,
+        "learning_rate": 0.0009291152533473353,
+        "epoch": 0.21
+    },
+    {
+        "loss": 10.9644,
+        "grad_norm": 0.697245180606842,
+        "learning_rate": 0.0009282401330182901,
+        "epoch": 0.22
+    },
+    {
+        "loss": 11.0427,
+        "grad_norm": 1.5471469163894653,
+        "learning_rate": 0.0009273650126892448,
+        "epoch": 0.22
+    },
+    {
+        "loss": 10.8293,
+        "grad_norm": 0.7173879146575928,
+        "learning_rate": 0.0009264898923601996,
+        "epoch": 0.22
+    },
+    {
+        "loss": 10.744,
+        "grad_norm": 1.1271495819091797,
+        "learning_rate": 0.0009256147720311543,
+        "epoch": 0.22
+    },
+    {
+        "loss": 10.3733,
+        "grad_norm": 0.7106486558914185,
+        "learning_rate": 0.0009247396517021091,
+        "epoch": 0.23
+    },
+    {
+        "loss": 10.9536,
+        "grad_norm": 1.1200592517852783,
+        "learning_rate": 0.0009238645313730638,
+        "epoch": 0.23
+    },
+    {
+        "loss": 10.4749,
+        "grad_norm": 1.0028458833694458,
+        "learning_rate": 0.0009229894110440185,
+        "epoch": 0.23
+    },
+    {
+        "loss": 11.4667,
+        "grad_norm": 1.187585711479187,
+        "learning_rate": 0.0009221142907149733,
+        "epoch": 0.23
+    },
+    {
+        "loss": 10.3349,
+        "grad_norm": 0.8691514134407043,
+        "learning_rate": 0.000921239170385928,
+        "epoch": 0.24
+    },
+    {
+        "loss": 10.6188,
+        "grad_norm": 0.8789599537849426,
+        "learning_rate": 0.0009203640500568828,
+        "epoch": 0.24
+    },
+    {
+        "loss": 10.454,
+        "grad_norm": 0.8376362919807434,
+        "learning_rate": 0.0009194889297278376,
+        "epoch": 0.24
+    },
+    {
+        "loss": 10.2419,
+        "grad_norm": 1.0760575532913208,
+        "learning_rate": 0.0009186138093987923,
+        "epoch": 0.24
+    },
+    {
+        "loss": 10.8593,
+        "grad_norm": 0.709028422832489,
+        "learning_rate": 0.0009177386890697471,
+        "epoch": 0.25
+    },
+    {
+        "loss": 11.073,
+        "grad_norm": 1.0934019088745117,
+        "learning_rate": 0.0009168635687407019,
+        "epoch": 0.25
+    },
+    {
+        "loss": 10.5596,
+        "grad_norm": 0.7833492159843445,
+        "learning_rate": 0.0009159884484116566,
+        "epoch": 0.25
+    },
+    {
+        "loss": 11.2079,
+        "grad_norm": 0.8762934803962708,
+        "learning_rate": 0.0009151133280826114,
+        "epoch": 0.25
+    },
+    {
+        "loss": 11.2229,
+        "grad_norm": 0.8059395551681519,
+        "learning_rate": 0.0009142382077535661,
+        "epoch": 0.26
+    },
+    {
+        "loss": 10.8706,
+        "grad_norm": 1.0892099142074585,
+        "learning_rate": 0.0009133630874245209,
+        "epoch": 0.26
+    },
+    {
+        "loss": 10.9983,
+        "grad_norm": 0.7471132278442383,
+        "learning_rate": 0.0009124879670954757,
+        "epoch": 0.26
+    },
+    {
+        "loss": 11.4291,
+        "grad_norm": 0.9766479730606079,
+        "learning_rate": 0.0009116128467664304,
+        "epoch": 0.27
+    },
+    {
+        "loss": 10.5895,
+        "grad_norm": 0.7469794154167175,
+        "learning_rate": 0.0009107377264373852,
+        "epoch": 0.27
+    },
+    {
+        "loss": 9.9826,
+        "grad_norm": 0.9510082602500916,
+        "learning_rate": 0.00090986260610834,
+        "epoch": 0.27
+    },
+    {
+        "loss": 10.1785,
+        "grad_norm": 0.8061089515686035,
+        "learning_rate": 0.0009089874857792947,
+        "epoch": 0.27
+    },
+    {
+        "loss": 10.5502,
+        "grad_norm": 0.7467952966690063,
+        "learning_rate": 0.0009081123654502495,
+        "epoch": 0.28
+    },
+    {
+        "loss": 10.4848,
+        "grad_norm": 0.9167515635490417,
+        "learning_rate": 0.0009072372451212042,
+        "epoch": 0.28
+    },
+    {
+        "loss": 10.7841,
+        "grad_norm": 1.0157630443572998,
+        "learning_rate": 0.000906362124792159,
+        "epoch": 0.28
+    },
+    {
+        "loss": 10.6985,
+        "grad_norm": 0.8764671087265015,
+        "learning_rate": 0.0009054870044631138,
+        "epoch": 0.28
+    },
+    {
+        "loss": 10.4706,
+        "grad_norm": 0.7716103196144104,
+        "learning_rate": 0.0009046118841340685,
+        "epoch": 0.29
+    },
+    {
+        "loss": 10.4371,
+        "grad_norm": 0.83449387550354,
+        "learning_rate": 0.0009037367638050233,
+        "epoch": 0.29
+    },
+    {
+        "loss": 10.2414,
+        "grad_norm": 0.785839855670929,
+        "learning_rate": 0.000902861643475978,
+        "epoch": 0.29
+    },
+    {
+        "loss": 10.0213,
+        "grad_norm": 0.7405595183372498,
+        "learning_rate": 0.0009019865231469327,
+        "epoch": 0.29
+    },
+    {
+        "loss": 10.2501,
+        "grad_norm": 0.929263710975647,
+        "learning_rate": 0.0009011114028178875,
+        "epoch": 0.3
+    },
+    {
+        "loss": 10.6749,
+        "grad_norm": 0.9185034036636353,
+        "learning_rate": 0.0009002362824888422,
+        "epoch": 0.3
+    },
+    {
+        "loss": 10.4313,
+        "grad_norm": 0.7888991832733154,
+        "learning_rate": 0.000899361162159797,
+        "epoch": 0.3
+    },
+    {
+        "loss": 10.4389,
+        "grad_norm": 0.9736090302467346,
+        "learning_rate": 0.0008984860418307517,
+        "epoch": 0.3
+    },
+    {
+        "loss": 9.9148,
+        "grad_norm": 0.7677895426750183,
+        "learning_rate": 0.0008976109215017065,
+        "epoch": 0.31
+    },
+    {
+        "loss": 9.7635,
+        "grad_norm": 0.9090219736099243,
+        "learning_rate": 0.0008967358011726613,
+        "epoch": 0.31
+    },
+    {
+        "loss": 10.0211,
+        "grad_norm": 0.7184523344039917,
+        "learning_rate": 0.000895860680843616,
+        "epoch": 0.31
+    },
+    {
+        "loss": 9.9932,
+        "grad_norm": 1.0859735012054443,
+        "learning_rate": 0.0008949855605145708,
+        "epoch": 0.31
+    },
+    {
+        "loss": 10.2804,
+        "grad_norm": 1.0252892971038818,
+        "learning_rate": 0.0008941104401855256,
+        "epoch": 0.32
+    },
+    {
+        "loss": 10.0543,
+        "grad_norm": 1.1707403659820557,
+        "learning_rate": 0.0008932353198564803,
+        "epoch": 0.32
+    },
+    {
+        "loss": 10.6658,
+        "grad_norm": 0.6616178750991821,
+        "learning_rate": 0.0008923601995274351,
+        "epoch": 0.32
+    },
+    {
+        "loss": 9.8623,
+        "grad_norm": 1.9947571754455566,
+        "learning_rate": 0.0008914850791983898,
+        "epoch": 0.33
+    },
+    {
+        "loss": 10.1607,
+        "grad_norm": 1.3363871574401855,
+        "learning_rate": 0.0008906099588693446,
+        "epoch": 0.33
+    },
+    {
+        "loss": 10.1063,
+        "grad_norm": 1.0039112567901611,
+        "learning_rate": 0.0008897348385402994,
+        "epoch": 0.33
+    },
+    {
+        "loss": 9.7059,
+        "grad_norm": 1.0225836038589478,
+        "learning_rate": 0.0008888597182112541,
+        "epoch": 0.33
+    },
+    {
+        "loss": 10.2506,
+        "grad_norm": 1.1005779504776,
+        "learning_rate": 0.0008879845978822089,
+        "epoch": 0.34
+    },
+    {
+        "loss": 10.3011,
+        "grad_norm": 1.1654433012008667,
+        "learning_rate": 0.0008871094775531636,
+        "epoch": 0.34
+    },
+    {
+        "loss": 10.088,
+        "grad_norm": 0.9155218601226807,
+        "learning_rate": 0.0008862343572241184,
+        "epoch": 0.34
+    },
+    {
+        "loss": 9.8835,
+        "grad_norm": 1.2090198993682861,
+        "learning_rate": 0.0008853592368950732,
+        "epoch": 0.34
+    },
+    {
+        "loss": 9.6644,
+        "grad_norm": 1.5198620557785034,
+        "learning_rate": 0.0008844841165660279,
+        "epoch": 0.35
+    },
+    {
+        "loss": 9.6799,
+        "grad_norm": 1.0043960809707642,
+        "learning_rate": 0.0008836089962369827,
+        "epoch": 0.35
+    },
+    {
+        "loss": 10.0658,
+        "grad_norm": 1.0404608249664307,
+        "learning_rate": 0.0008827338759079375,
+        "epoch": 0.35
+    },
+    {
+        "loss": 9.9551,
+        "grad_norm": 1.0412163734436035,
+        "learning_rate": 0.0008818587555788922,
+        "epoch": 0.35
+    },
+    {
+        "loss": 9.4082,
+        "grad_norm": 0.9032560586929321,
+        "learning_rate": 0.000880983635249847,
+        "epoch": 0.36
+    },
+    {
+        "loss": 10.2566,
+        "grad_norm": 1.2763034105300903,
+        "learning_rate": 0.0008801085149208016,
+        "epoch": 0.36
+    },
+    {
+        "loss": 9.8585,
+        "grad_norm": 0.8143719434738159,
+        "learning_rate": 0.0008792333945917563,
+        "epoch": 0.36
+    },
+    {
+        "loss": 9.5974,
+        "grad_norm": 1.3916654586791992,
+        "learning_rate": 0.000878358274262711,
+        "epoch": 0.36
+    },
+    {
+        "loss": 10.611,
+        "grad_norm": 1.2270894050598145,
+        "learning_rate": 0.0008774831539336658,
+        "epoch": 0.37
+    },
+    {
+        "loss": 9.4489,
+        "grad_norm": 1.339573621749878,
+        "learning_rate": 0.0008766080336046206,
+        "epoch": 0.37
+    },
+    {
+        "loss": 9.769,
+        "grad_norm": 1.023978352546692,
+        "learning_rate": 0.0008757329132755753,
+        "epoch": 0.37
+    },
+    {
+        "loss": 9.7854,
+        "grad_norm": 1.1513617038726807,
+        "learning_rate": 0.0008748577929465301,
+        "epoch": 0.38
+    },
+    {
+        "loss": 9.4378,
+        "grad_norm": 0.9918627142906189,
+        "learning_rate": 0.0008739826726174849,
+        "epoch": 0.38
+    },
+    {
+        "loss": 9.6902,
+        "grad_norm": 0.9365573525428772,
+        "learning_rate": 0.0008731075522884396,
+        "epoch": 0.38
+    },
+    {
+        "loss": 9.5533,
+        "grad_norm": 1.1697934865951538,
+        "learning_rate": 0.0008722324319593944,
+        "epoch": 0.38
+    },
+    {
+        "loss": 9.5204,
+        "grad_norm": 1.2257342338562012,
+        "learning_rate": 0.0008713573116303491,
+        "epoch": 0.39
+    },
+    {
+        "loss": 9.636,
+        "grad_norm": 1.0158884525299072,
+        "learning_rate": 0.0008704821913013039,
+        "epoch": 0.39
+    },
+    {
+        "loss": 9.8914,
+        "grad_norm": 1.4228135347366333,
+        "learning_rate": 0.0008696070709722587,
+        "epoch": 0.39
+    },
+    {
+        "loss": 9.3714,
+        "grad_norm": 1.2829135656356812,
+        "learning_rate": 0.0008687319506432134,
+        "epoch": 0.39
+    },
+    {
+        "loss": 9.7498,
+        "grad_norm": 1.2624573707580566,
+        "learning_rate": 0.0008678568303141682,
+        "epoch": 0.4
+    },
+    {
+        "loss": 9.8928,
+        "grad_norm": 1.3651659488677979,
+        "learning_rate": 0.000866981709985123,
+        "epoch": 0.4
+    },
+    {
+        "loss": 10.3697,
+        "grad_norm": 1.1383252143859863,
+        "learning_rate": 0.0008661065896560777,
+        "epoch": 0.4
+    },
+    {
+        "loss": 10.1876,
+        "grad_norm": 1.1688463687896729,
+        "learning_rate": 0.0008652314693270325,
+        "epoch": 0.4
+    },
+    {
+        "loss": 9.7974,
+        "grad_norm": 1.1377474069595337,
+        "learning_rate": 0.0008643563489979872,
+        "epoch": 0.41
+    },
+    {
+        "loss": 9.5742,
+        "grad_norm": 1.0107587575912476,
+        "learning_rate": 0.000863481228668942,
+        "epoch": 0.41
+    },
+    {
+        "loss": 9.9821,
+        "grad_norm": 1.3488329648971558,
+        "learning_rate": 0.0008626061083398968,
+        "epoch": 0.41
+    },
+    {
+        "loss": 9.3107,
+        "grad_norm": 1.0305010080337524,
+        "learning_rate": 0.0008617309880108515,
+        "epoch": 0.41
+    },
+    {
+        "loss": 9.3456,
+        "grad_norm": 0.8658286929130554,
+        "learning_rate": 0.0008608558676818063,
+        "epoch": 0.42
+    },
+    {
+        "loss": 9.3709,
+        "grad_norm": 1.1033709049224854,
+        "learning_rate": 0.000859980747352761,
+        "epoch": 0.42
+    },
+    {
+        "loss": 9.5077,
+        "grad_norm": 1.1051572561264038,
+        "learning_rate": 0.0008591056270237157,
+        "epoch": 0.42
+    },
+    {
+        "loss": 9.1458,
+        "grad_norm": 1.3423538208007812,
+        "learning_rate": 0.0008582305066946705,
+        "epoch": 0.43
+    },
+    {
+        "loss": 9.657,
+        "grad_norm": 1.1479153633117676,
+        "learning_rate": 0.0008573553863656252,
+        "epoch": 0.43
+    },
+    {
+        "loss": 10.5804,
+        "grad_norm": 1.1615872383117676,
+        "learning_rate": 0.00085648026603658,
+        "epoch": 0.43
+    },
+    {
+        "loss": 8.2792,
+        "grad_norm": 1.212221384048462,
+        "learning_rate": 0.0008556051457075347,
+        "epoch": 0.43
+    },
+    {
+        "loss": 9.3785,
+        "grad_norm": 1.0849367380142212,
+        "learning_rate": 0.0008547300253784895,
+        "epoch": 0.44
+    },
+    {
+        "loss": 9.4097,
+        "grad_norm": 1.119325041770935,
+        "learning_rate": 0.0008538549050494443,
+        "epoch": 0.44
+    },
+    {
+        "loss": 9.3308,
+        "grad_norm": 1.3356918096542358,
+        "learning_rate": 0.000852979784720399,
+        "epoch": 0.44
+    },
+    {
+        "loss": 9.4548,
+        "grad_norm": 0.9954844117164612,
+        "learning_rate": 0.0008521046643913538,
+        "epoch": 0.44
+    },
+    {
+        "loss": 8.9297,
+        "grad_norm": 0.8752724528312683,
+        "learning_rate": 0.0008512295440623086,
+        "epoch": 0.45
+    },
+    {
+        "loss": 9.1389,
+        "grad_norm": 1.2811753749847412,
+        "learning_rate": 0.0008503544237332633,
+        "epoch": 0.45
+    },
+    {
+        "loss": 9.3155,
+        "grad_norm": 1.253055453300476,
+        "learning_rate": 0.0008494793034042181,
+        "epoch": 0.45
+    },
+    {
+        "loss": 9.548,
+        "grad_norm": 1.2081260681152344,
+        "learning_rate": 0.0008486041830751728,
+        "epoch": 0.45
+    },
+    {
+        "loss": 9.0236,
+        "grad_norm": 1.3752362728118896,
+        "learning_rate": 0.0008477290627461276,
+        "epoch": 0.46
+    },
+    {
+        "loss": 9.0533,
+        "grad_norm": 1.057065725326538,
+        "learning_rate": 0.0008468539424170824,
+        "epoch": 0.46
+    },
+    {
+        "loss": 9.0675,
+        "grad_norm": 1.0036309957504272,
+        "learning_rate": 0.0008459788220880371,
+        "epoch": 0.46
+    },
+    {
+        "loss": 9.5195,
+        "grad_norm": 1.3881008625030518,
+        "learning_rate": 0.0008451037017589919,
+        "epoch": 0.46
+    },
+    {
+        "loss": 9.3519,
+        "grad_norm": 1.4355233907699585,
+        "learning_rate": 0.0008442285814299467,
+        "epoch": 0.47
+    },
+    {
+        "loss": 9.6383,
+        "grad_norm": 0.9438649415969849,
+        "learning_rate": 0.0008433534611009014,
+        "epoch": 0.47
+    },
+    {
+        "loss": 9.2643,
+        "grad_norm": 0.8599776029586792,
+        "learning_rate": 0.0008424783407718562,
+        "epoch": 0.47
+    },
+    {
+        "loss": 8.9869,
+        "grad_norm": 1.1090342998504639,
+        "learning_rate": 0.0008416032204428109,
+        "epoch": 0.48
+    },
+    {
+        "loss": 9.2475,
+        "grad_norm": 1.272929310798645,
+        "learning_rate": 0.0008407281001137657,
+        "epoch": 0.48
+    },
+    {
+        "loss": 9.5772,
+        "grad_norm": 0.9889743328094482,
+        "learning_rate": 0.0008398529797847205,
+        "epoch": 0.48
+    },
+    {
+        "loss": 9.9227,
+        "grad_norm": 1.2748692035675049,
+        "learning_rate": 0.0008389778594556752,
+        "epoch": 0.48
+    },
+    {
+        "loss": 9.9915,
+        "grad_norm": 1.4889165163040161,
+        "learning_rate": 0.00083810273912663,
+        "epoch": 0.49
+    },
+    {
+        "loss": 9.0012,
+        "grad_norm": 1.2172118425369263,
+        "learning_rate": 0.0008372276187975846,
+        "epoch": 0.49
+    },
+    {
+        "loss": 9.2968,
+        "grad_norm": 1.0313849449157715,
+        "learning_rate": 0.0008363524984685394,
+        "epoch": 0.49
+    },
+    {
+        "loss": 8.9158,
+        "grad_norm": 1.3325482606887817,
+        "learning_rate": 0.0008354773781394942,
+        "epoch": 0.49
+    },
+    {
+        "loss": 9.0097,
+        "grad_norm": 1.5407133102416992,
+        "learning_rate": 0.0008346022578104489,
+        "epoch": 0.5
+    },
+    {
+        "loss": 9.0166,
+        "grad_norm": 1.1565685272216797,
+        "learning_rate": 0.0008337271374814037,
+        "epoch": 0.5
+    },
+    {
+        "loss": 9.1856,
+        "grad_norm": 1.0405404567718506,
+        "learning_rate": 0.0008328520171523584,
+        "epoch": 0.5
+    },
+    {
+        "loss": 9.2405,
+        "grad_norm": 1.465058445930481,
+        "learning_rate": 0.0008319768968233132,
+        "epoch": 0.5
+    },
+    {
+        "loss": 8.835,
+        "grad_norm": 0.9321463704109192,
+        "learning_rate": 0.000831101776494268,
+        "epoch": 0.51
+    },
+    {
+        "loss": 9.4076,
+        "grad_norm": 1.1780034303665161,
+        "learning_rate": 0.0008302266561652227,
+        "epoch": 0.51
+    },
+    {
+        "loss": 9.5994,
+        "grad_norm": 1.488897681236267,
+        "learning_rate": 0.0008293515358361775,
+        "epoch": 0.51
+    },
+    {
+        "loss": 8.6378,
+        "grad_norm": 1.0508447885513306,
+        "learning_rate": 0.0008284764155071323,
+        "epoch": 0.51
+    },
+    {
+        "loss": 8.7946,
+        "grad_norm": 1.2236040830612183,
+        "learning_rate": 0.000827601295178087,
+        "epoch": 0.52
+    },
+    {
+        "loss": 9.4619,
+        "grad_norm": 1.0602221488952637,
+        "learning_rate": 0.0008267261748490418,
+        "epoch": 0.52
+    },
+    {
+        "loss": 8.927,
+        "grad_norm": 1.476576328277588,
+        "learning_rate": 0.0008258510545199965,
+        "epoch": 0.52
+    },
+    {
+        "loss": 8.766,
+        "grad_norm": 1.2723809480667114,
+        "learning_rate": 0.0008249759341909513,
+        "epoch": 0.52
+    },
+    {
+        "loss": 9.1577,
+        "grad_norm": 1.2955093383789062,
+        "learning_rate": 0.0008241008138619061,
+        "epoch": 0.53
+    },
+    {
+        "loss": 8.8254,
+        "grad_norm": 1.1421802043914795,
+        "learning_rate": 0.0008232256935328608,
+        "epoch": 0.53
+    },
+    {
+        "loss": 9.3559,
+        "grad_norm": 1.2015204429626465,
+        "learning_rate": 0.0008223505732038156,
+        "epoch": 0.53
+    },
+    {
+        "loss": 8.7055,
+        "grad_norm": 1.02347993850708,
+        "learning_rate": 0.0008214754528747703,
+        "epoch": 0.54
+    },
+    {
+        "loss": 9.1773,
+        "grad_norm": 1.0733789205551147,
+        "learning_rate": 0.0008206003325457251,
+        "epoch": 0.54
+    },
+    {
+        "loss": 9.4909,
+        "grad_norm": 1.140329360961914,
+        "learning_rate": 0.0008197252122166799,
+        "epoch": 0.54
+    },
+    {
+        "loss": 8.4982,
+        "grad_norm": 0.8933946490287781,
+        "learning_rate": 0.0008188500918876346,
+        "epoch": 0.54
+    },
+    {
+        "loss": 9.4497,
+        "grad_norm": 1.3848881721496582,
+        "learning_rate": 0.0008179749715585894,
+        "epoch": 0.55
+    },
+    {
+        "loss": 9.5758,
+        "grad_norm": 1.175162672996521,
+        "learning_rate": 0.0008170998512295442,
+        "epoch": 0.55
+    },
+    {
+        "loss": 9.5138,
+        "grad_norm": 1.1983882188796997,
+        "learning_rate": 0.0008162247309004989,
+        "epoch": 0.55
+    },
+    {
+        "loss": 9.0283,
+        "grad_norm": 0.9055472612380981,
+        "learning_rate": 0.0008153496105714536,
+        "epoch": 0.55
+    },
+    {
+        "loss": 9.2822,
+        "grad_norm": 0.8885380029678345,
+        "learning_rate": 0.0008144744902424083,
+        "epoch": 0.56
+    },
+    {
+        "loss": 8.9084,
+        "grad_norm": 1.0463942289352417,
+        "learning_rate": 0.0008135993699133631,
+        "epoch": 0.56
+    },
+    {
+        "loss": 9.0612,
+        "grad_norm": 1.1517601013183594,
+        "learning_rate": 0.0008127242495843179,
+        "epoch": 0.56
+    },
+    {
+        "loss": 9.7954,
+        "grad_norm": 1.6062026023864746,
+        "learning_rate": 0.0008118491292552726,
+        "epoch": 0.56
+    },
+    {
+        "loss": 8.823,
+        "grad_norm": 1.079883098602295,
+        "learning_rate": 0.0008109740089262274,
+        "epoch": 0.57
+    },
+    {
+        "loss": 8.6287,
+        "grad_norm": 0.8593969345092773,
+        "learning_rate": 0.0008100988885971821,
+        "epoch": 0.57
+    },
+    {
+        "loss": 9.046,
+        "grad_norm": 1.5058172941207886,
+        "learning_rate": 0.0008092237682681369,
+        "epoch": 0.57
+    },
+    {
+        "loss": 8.4422,
+        "grad_norm": 1.0326484441757202,
+        "learning_rate": 0.0008083486479390917,
+        "epoch": 0.57
+    },
+    {
+        "loss": 9.5016,
+        "grad_norm": 0.9177812337875366,
+        "learning_rate": 0.0008074735276100464,
+        "epoch": 0.58
+    },
+    {
+        "loss": 8.4734,
+        "grad_norm": 1.1267443895339966,
+        "learning_rate": 0.0008065984072810012,
+        "epoch": 0.58
+    },
+    {
+        "loss": 8.5878,
+        "grad_norm": 0.9788813591003418,
+        "learning_rate": 0.000805723286951956,
+        "epoch": 0.58
+    },
+    {
+        "loss": 9.1188,
+        "grad_norm": 1.1300309896469116,
+        "learning_rate": 0.0008048481666229107,
+        "epoch": 0.59
+    },
+    {
+        "loss": 8.7167,
+        "grad_norm": 0.9951778650283813,
+        "learning_rate": 0.0008039730462938655,
+        "epoch": 0.59
+    },
+    {
+        "loss": 9.1088,
+        "grad_norm": 0.9415300488471985,
+        "learning_rate": 0.0008030979259648202,
+        "epoch": 0.59
+    },
+    {
+        "loss": 8.4083,
+        "grad_norm": 0.990203320980072,
+        "learning_rate": 0.000802222805635775,
+        "epoch": 0.59
+    },
+    {
+        "loss": 8.4926,
+        "grad_norm": 1.0430456399917603,
+        "learning_rate": 0.0008013476853067298,
+        "epoch": 0.6
+    },
+    {
+        "loss": 9.3307,
+        "grad_norm": 0.9623116254806519,
+        "learning_rate": 0.0008004725649776845,
+        "epoch": 0.6
+    },
+    {
+        "loss": 8.8633,
+        "grad_norm": 1.0354257822036743,
+        "learning_rate": 0.0007995974446486392,
+        "epoch": 0.6
+    },
+    {
+        "loss": 8.7932,
+        "grad_norm": 1.1962673664093018,
+        "learning_rate": 0.0007987223243195939,
+        "epoch": 0.6
+    },
+    {
+        "loss": 8.4265,
+        "grad_norm": 1.0186195373535156,
+        "learning_rate": 0.0007978472039905487,
+        "epoch": 0.61
+    },
+    {
+        "loss": 8.4596,
+        "grad_norm": 1.2448772192001343,
+        "learning_rate": 0.0007969720836615035,
+        "epoch": 0.61
+    },
+    {
+        "loss": 9.0019,
+        "grad_norm": 1.11643385887146,
+        "learning_rate": 0.0007960969633324582,
+        "epoch": 0.61
+    },
+    {
+        "loss": 8.7469,
+        "grad_norm": 1.9622658491134644,
+        "learning_rate": 0.000795221843003413,
+        "epoch": 0.61
+    },
+    {
+        "loss": 8.208,
+        "grad_norm": 0.9547304511070251,
+        "learning_rate": 0.0007943467226743676,
+        "epoch": 0.62
+    },
+    {
+        "loss": 8.3751,
+        "grad_norm": 0.8313985466957092,
+        "learning_rate": 0.0007934716023453224,
+        "epoch": 0.62
+    },
+    {
+        "loss": 8.6238,
+        "grad_norm": 0.9323874711990356,
+        "learning_rate": 0.0007925964820162772,
+        "epoch": 0.62
+    },
+    {
+        "loss": 9.0078,
+        "grad_norm": 1.0662554502487183,
+        "learning_rate": 0.0007917213616872319,
+        "epoch": 0.62
+    },
+    {
+        "loss": 8.7407,
+        "grad_norm": 1.197045087814331,
+        "learning_rate": 0.0007908462413581867,
+        "epoch": 0.63
+    },
+    {
+        "loss": 8.9698,
+        "grad_norm": 1.0494697093963623,
+        "learning_rate": 0.0007899711210291415,
+        "epoch": 0.63
+    },
+    {
+        "loss": 8.56,
+        "grad_norm": 0.9860395789146423,
+        "learning_rate": 0.0007890960007000962,
+        "epoch": 0.63
+    },
+    {
+        "loss": 8.624,
+        "grad_norm": 0.8026842474937439,
+        "learning_rate": 0.000788220880371051,
+        "epoch": 0.64
+    },
+    {
+        "loss": 9.1911,
+        "grad_norm": 1.0249046087265015,
+        "learning_rate": 0.0007873457600420057,
+        "epoch": 0.64
+    },
+    {
+        "loss": 8.552,
+        "grad_norm": 1.3037137985229492,
+        "learning_rate": 0.0007864706397129605,
+        "epoch": 0.64
+    },
+    {
+        "loss": 8.6872,
+        "grad_norm": 1.1018158197402954,
+        "learning_rate": 0.0007855955193839153,
+        "epoch": 0.64
+    },
+    {
+        "loss": 8.5007,
+        "grad_norm": 0.9974724054336548,
+        "learning_rate": 0.00078472039905487,
+        "epoch": 0.65
+    },
+    {
+        "loss": 9.3866,
+        "grad_norm": 1.2537139654159546,
+        "learning_rate": 0.0007838452787258248,
+        "epoch": 0.65
+    },
+    {
+        "loss": 8.9869,
+        "grad_norm": 1.2758492231369019,
+        "learning_rate": 0.0007829701583967795,
+        "epoch": 0.65
+    },
+    {
+        "loss": 8.266,
+        "grad_norm": 0.9684768915176392,
+        "learning_rate": 0.0007820950380677343,
+        "epoch": 0.65
+    },
+    {
+        "loss": 9.0718,
+        "grad_norm": 1.0212547779083252,
+        "learning_rate": 0.0007812199177386891,
+        "epoch": 0.66
+    },
+    {
+        "loss": 8.1438,
+        "grad_norm": 1.2493318319320679,
+        "learning_rate": 0.0007803447974096438,
+        "epoch": 0.66
+    },
+    {
+        "loss": 8.4132,
+        "grad_norm": 0.8168124556541443,
+        "learning_rate": 0.0007794696770805986,
+        "epoch": 0.66
+    },
+    {
+        "loss": 8.4466,
+        "grad_norm": 1.2837003469467163,
+        "learning_rate": 0.0007785945567515534,
+        "epoch": 0.66
+    },
+    {
+        "loss": 8.6008,
+        "grad_norm": 1.1589733362197876,
+        "learning_rate": 0.0007777194364225081,
+        "epoch": 0.67
+    },
+    {
+        "loss": 8.7002,
+        "grad_norm": 1.036216378211975,
+        "learning_rate": 0.0007768443160934629,
+        "epoch": 0.67
+    },
+    {
+        "loss": 8.9616,
+        "grad_norm": 0.9488565921783447,
+        "learning_rate": 0.0007759691957644176,
+        "epoch": 0.67
+    },
+    {
+        "loss": 8.9011,
+        "grad_norm": 1.1349655389785767,
+        "learning_rate": 0.0007750940754353724,
+        "epoch": 0.67
+    },
+    {
+        "loss": 8.7398,
+        "grad_norm": 1.3466508388519287,
+        "learning_rate": 0.0007742189551063272,
+        "epoch": 0.68
+    },
+    {
+        "loss": 8.1787,
+        "grad_norm": 1.1343966722488403,
+        "learning_rate": 0.0007733438347772819,
+        "epoch": 0.68
+    },
+    {
+        "loss": 8.4513,
+        "grad_norm": 0.9983484148979187,
+        "learning_rate": 0.0007724687144482366,
+        "epoch": 0.68
+    },
+    {
+        "loss": 8.6249,
+        "grad_norm": 1.4816855192184448,
+        "learning_rate": 0.0007715935941191913,
+        "epoch": 0.69
+    },
+    {
+        "loss": 8.9094,
+        "grad_norm": 1.0790578126907349,
+        "learning_rate": 0.0007707184737901461,
+        "epoch": 0.69
+    },
+    {
+        "loss": 8.0177,
+        "grad_norm": 1.2572119235992432,
+        "learning_rate": 0.0007698433534611009,
+        "epoch": 0.69
+    },
+    {
+        "loss": 8.5014,
+        "grad_norm": 1.123079776763916,
+        "learning_rate": 0.0007689682331320556,
+        "epoch": 0.69
+    },
+    {
+        "loss": 8.2177,
+        "grad_norm": 0.8789654970169067,
+        "learning_rate": 0.0007680931128030104,
+        "epoch": 0.7
+    },
+    {
+        "loss": 8.3753,
+        "grad_norm": 0.9512013792991638,
+        "learning_rate": 0.0007672179924739651,
+        "epoch": 0.7
+    },
+    {
+        "loss": 8.5434,
+        "grad_norm": 1.929919719696045,
+        "learning_rate": 0.0007663428721449199,
+        "epoch": 0.7
+    },
+    {
+        "loss": 8.5505,
+        "grad_norm": 1.1756147146224976,
+        "learning_rate": 0.0007654677518158747,
+        "epoch": 0.7
+    },
+    {
+        "loss": 8.8823,
+        "grad_norm": 1.1833679676055908,
+        "learning_rate": 0.0007645926314868294,
+        "epoch": 0.71
+    },
+    {
+        "loss": 8.6715,
+        "grad_norm": 1.4701839685440063,
+        "learning_rate": 0.0007637175111577842,
+        "epoch": 0.71
+    },
+    {
+        "loss": 8.7559,
+        "grad_norm": 0.9352959990501404,
+        "learning_rate": 0.0007629299028616435,
+        "epoch": 0.71
+    },
+    {
+        "loss": 9.5594,
+        "grad_norm": 1.0391898155212402,
+        "learning_rate": 0.0007620547825325983,
+        "epoch": 0.71
+    },
+    {
+        "loss": 8.3431,
+        "grad_norm": 1.0766905546188354,
+        "learning_rate": 0.000761179662203553,
+        "epoch": 0.72
+    },
+    {
+        "loss": 8.3928,
+        "grad_norm": 1.10299551486969,
+        "learning_rate": 0.0007603045418745078,
+        "epoch": 0.72
+    },
+    {
+        "loss": 8.9913,
+        "grad_norm": 1.1581339836120605,
+        "learning_rate": 0.0007594294215454624,
+        "epoch": 0.72
+    },
+    {
+        "loss": 8.5142,
+        "grad_norm": 1.086441993713379,
+        "learning_rate": 0.0007585543012164172,
+        "epoch": 0.72
+    },
+    {
+        "loss": 8.7005,
+        "grad_norm": 0.9478667974472046,
+        "learning_rate": 0.000757679180887372,
+        "epoch": 0.73
+    },
+    {
+        "loss": 8.608,
+        "grad_norm": 1.0929220914840698,
+        "learning_rate": 0.0007568040605583267,
+        "epoch": 0.73
+    },
+    {
+        "loss": 8.1125,
+        "grad_norm": 1.217629313468933,
+        "learning_rate": 0.0007559289402292815,
+        "epoch": 0.73
+    },
+    {
+        "loss": 8.4331,
+        "grad_norm": 1.2786823511123657,
+        "learning_rate": 0.0007550538199002362,
+        "epoch": 0.73
+    },
+    {
+        "loss": 9.1985,
+        "grad_norm": 1.0184354782104492,
+        "learning_rate": 0.000754178699571191,
+        "epoch": 0.74
+    },
+    {
+        "loss": 8.6549,
+        "grad_norm": 0.93660968542099,
+        "learning_rate": 0.0007533035792421458,
+        "epoch": 0.74
+    },
+    {
+        "loss": 8.7819,
+        "grad_norm": 1.0092636346817017,
+        "learning_rate": 0.0007524284589131005,
+        "epoch": 0.74
+    },
+    {
+        "loss": 8.3759,
+        "grad_norm": 1.2108792066574097,
+        "learning_rate": 0.0007515533385840553,
+        "epoch": 0.75
+    },
+    {
+        "loss": 8.4973,
+        "grad_norm": 0.9994498491287231,
+        "learning_rate": 0.00075067821825501,
+        "epoch": 0.75
+    },
+    {
+        "loss": 8.3731,
+        "grad_norm": 1.153273344039917,
+        "learning_rate": 0.0007498030979259648,
+        "epoch": 0.75
+    },
+    {
+        "loss": 8.4148,
+        "grad_norm": 1.051223874092102,
+        "learning_rate": 0.0007489279775969196,
+        "epoch": 0.75
+    },
+    {
+        "loss": 8.6672,
+        "grad_norm": 1.4810237884521484,
+        "learning_rate": 0.0007480528572678743,
+        "epoch": 0.76
+    },
+    {
+        "loss": 8.6439,
+        "grad_norm": 1.021606206893921,
+        "learning_rate": 0.0007471777369388291,
+        "epoch": 0.76
+    },
+    {
+        "loss": 8.7591,
+        "grad_norm": 0.8680776357650757,
+        "learning_rate": 0.0007463026166097839,
+        "epoch": 0.76
+    },
+    {
+        "loss": 9.0187,
+        "grad_norm": 1.0177042484283447,
+        "learning_rate": 0.0007454274962807386,
+        "epoch": 0.76
+    },
+    {
+        "loss": 8.9481,
+        "grad_norm": 1.2384392023086548,
+        "learning_rate": 0.0007445523759516934,
+        "epoch": 0.77
+    },
+    {
+        "loss": 8.6184,
+        "grad_norm": 1.3748959302902222,
+        "learning_rate": 0.0007436772556226481,
+        "epoch": 0.77
+    },
+    {
+        "loss": 8.3906,
+        "grad_norm": 1.042493462562561,
+        "learning_rate": 0.0007428021352936029,
+        "epoch": 0.77
+    },
+    {
+        "loss": 9.3308,
+        "grad_norm": 1.0647776126861572,
+        "learning_rate": 0.0007419270149645576,
+        "epoch": 0.77
+    },
+    {
+        "loss": 8.332,
+        "grad_norm": 1.2385993003845215,
+        "learning_rate": 0.0007410518946355123,
+        "epoch": 0.78
+    },
+    {
+        "loss": 8.3127,
+        "grad_norm": 1.0191227197647095,
+        "learning_rate": 0.0007401767743064671,
+        "epoch": 0.78
+    },
+    {
+        "loss": 8.3151,
+        "grad_norm": 0.8735216856002808,
+        "learning_rate": 0.0007393016539774218,
+        "epoch": 0.78
+    },
+    {
+        "loss": 8.701,
+        "grad_norm": 1.202993392944336,
+        "learning_rate": 0.0007384265336483766,
+        "epoch": 0.78
+    },
+    {
+        "loss": 7.8262,
+        "grad_norm": 0.9682905673980713,
+        "learning_rate": 0.0007375514133193314,
+        "epoch": 0.79
+    },
+    {
+        "loss": 8.4729,
+        "grad_norm": 1.2290154695510864,
+        "learning_rate": 0.0007366762929902861,
+        "epoch": 0.79
+    },
+    {
+        "loss": 8.9253,
+        "grad_norm": 1.0369175672531128,
+        "learning_rate": 0.0007358011726612409,
+        "epoch": 0.79
+    },
+    {
+        "loss": 9.2036,
+        "grad_norm": 1.0748445987701416,
+        "learning_rate": 0.0007349260523321957,
+        "epoch": 0.8
+    },
+    {
+        "loss": 8.2364,
+        "grad_norm": 1.147964596748352,
+        "learning_rate": 0.0007340509320031504,
+        "epoch": 0.8
+    },
+    {
+        "loss": 9.006,
+        "grad_norm": 1.0363622903823853,
+        "learning_rate": 0.0007331758116741052,
+        "epoch": 0.8
+    },
+    {
+        "loss": 8.7969,
+        "grad_norm": 1.2576889991760254,
+        "learning_rate": 0.0007323006913450599,
+        "epoch": 0.8
+    },
+    {
+        "loss": 8.4052,
+        "grad_norm": 1.1075588464736938,
+        "learning_rate": 0.0007314255710160147,
+        "epoch": 0.81
+    },
+    {
+        "loss": 8.5912,
+        "grad_norm": 1.0697672367095947,
+        "learning_rate": 0.0007305504506869695,
+        "epoch": 0.81
+    },
+    {
+        "loss": 8.7837,
+        "grad_norm": 1.0865002870559692,
+        "learning_rate": 0.0007296753303579242,
+        "epoch": 0.81
+    },
+    {
+        "loss": 8.0798,
+        "grad_norm": 1.3645957708358765,
+        "learning_rate": 0.000728800210028879,
+        "epoch": 0.81
+    },
+    {
+        "loss": 8.2649,
+        "grad_norm": 1.0889688730239868,
+        "learning_rate": 0.0007279250896998337,
+        "epoch": 0.82
+    },
+    {
+        "loss": 7.902,
+        "grad_norm": 0.9943633675575256,
+        "learning_rate": 0.0007270499693707885,
+        "epoch": 0.82
+    },
+    {
+        "loss": 8.493,
+        "grad_norm": 1.3548861742019653,
+        "learning_rate": 0.0007261748490417433,
+        "epoch": 0.82
+    },
+    {
+        "loss": 9.2024,
+        "grad_norm": 1.1603728532791138,
+        "learning_rate": 0.000725299728712698,
+        "epoch": 0.82
+    },
+    {
+        "loss": 8.7272,
+        "grad_norm": 1.2872350215911865,
+        "learning_rate": 0.0007244246083836528,
+        "epoch": 0.83
+    },
+    {
+        "loss": 8.8292,
+        "grad_norm": 1.0431410074234009,
+        "learning_rate": 0.0007235494880546076,
+        "epoch": 0.83
+    },
+    {
+        "loss": 8.0473,
+        "grad_norm": 0.9648978114128113,
+        "learning_rate": 0.0007226743677255623,
+        "epoch": 0.83
+    },
+    {
+        "loss": 8.134,
+        "grad_norm": 0.8962783217430115,
+        "learning_rate": 0.0007217992473965171,
+        "epoch": 0.83
+    },
+    {
+        "loss": 8.2796,
+        "grad_norm": 0.8879069685935974,
+        "learning_rate": 0.0007209241270674718,
+        "epoch": 0.84
+    },
+    {
+        "loss": 8.6275,
+        "grad_norm": 1.0046008825302124,
+        "learning_rate": 0.0007200490067384265,
+        "epoch": 0.84
+    },
+    {
+        "loss": 8.2847,
+        "grad_norm": 1.1034067869186401,
+        "learning_rate": 0.0007191738864093813,
+        "epoch": 0.84
+    },
+    {
+        "loss": 8.723,
+        "grad_norm": 0.9179050326347351,
+        "learning_rate": 0.000718298766080336,
+        "epoch": 0.85
+    },
+    {
+        "loss": 8.2843,
+        "grad_norm": 1.0402296781539917,
+        "learning_rate": 0.0007174236457512908,
+        "epoch": 0.85
+    },
+    {
+        "loss": 8.2487,
+        "grad_norm": 1.2751373052597046,
+        "learning_rate": 0.0007165485254222455,
+        "epoch": 0.85
+    },
+    {
+        "loss": 8.3491,
+        "grad_norm": 0.8596373200416565,
+        "learning_rate": 0.0007156734050932003,
+        "epoch": 0.85
+    },
+    {
+        "loss": 8.4695,
+        "grad_norm": 1.0553058385849,
+        "learning_rate": 0.0007147982847641551,
+        "epoch": 0.86
+    },
+    {
+        "loss": 8.74,
+        "grad_norm": 1.0505644083023071,
+        "learning_rate": 0.0007139231644351098,
+        "epoch": 0.86
+    },
+    {
+        "loss": 8.3704,
+        "grad_norm": 1.4136569499969482,
+        "learning_rate": 0.0007130480441060646,
+        "epoch": 0.86
+    },
+    {
+        "loss": 7.9998,
+        "grad_norm": 0.9397268295288086,
+        "learning_rate": 0.0007121729237770194,
+        "epoch": 0.86
+    },
+    {
+        "loss": 8.5978,
+        "grad_norm": 1.1479915380477905,
+        "learning_rate": 0.0007112978034479741,
+        "epoch": 0.87
+    },
+    {
+        "loss": 8.6225,
+        "grad_norm": 1.0489866733551025,
+        "learning_rate": 0.0007104226831189289,
+        "epoch": 0.87
+    },
+    {
+        "loss": 8.3155,
+        "grad_norm": 0.9371022582054138,
+        "learning_rate": 0.0007095475627898836,
+        "epoch": 0.87
+    },
+    {
+        "loss": 8.3844,
+        "grad_norm": 1.1981381177902222,
+        "learning_rate": 0.0007086724424608384,
+        "epoch": 0.87
+    },
+    {
+        "loss": 8.5061,
+        "grad_norm": 0.8924277424812317,
+        "learning_rate": 0.0007077973221317932,
+        "epoch": 0.88
+    },
+    {
+        "loss": 8.1918,
+        "grad_norm": 1.4077969789505005,
+        "learning_rate": 0.0007069222018027479,
+        "epoch": 0.88
+    },
+    {
+        "loss": 8.3377,
+        "grad_norm": 1.1926066875457764,
+        "learning_rate": 0.0007060470814737027,
+        "epoch": 0.88
+    },
+    {
+        "loss": 8.4682,
+        "grad_norm": 1.1524171829223633,
+        "learning_rate": 0.0007051719611446574,
+        "epoch": 0.88
+    },
+    {
+        "loss": 8.5678,
+        "grad_norm": 1.0660207271575928,
+        "learning_rate": 0.0007042968408156122,
+        "epoch": 0.89
+    },
+    {
+        "loss": 7.9908,
+        "grad_norm": 1.1786776781082153,
+        "learning_rate": 0.000703421720486567,
+        "epoch": 0.89
+    },
+    {
+        "loss": 9.0339,
+        "grad_norm": 0.9970653057098389,
+        "learning_rate": 0.0007025466001575217,
+        "epoch": 0.89
+    },
+    {
+        "loss": 8.6511,
+        "grad_norm": 1.171247124671936,
+        "learning_rate": 0.0007016714798284765,
+        "epoch": 0.9
+    },
+    {
+        "loss": 8.0249,
+        "grad_norm": 1.1036537885665894,
+        "learning_rate": 0.0007007963594994313,
+        "epoch": 0.9
+    },
+    {
+        "loss": 8.2895,
+        "grad_norm": 1.4363912343978882,
+        "learning_rate": 0.000699921239170386,
+        "epoch": 0.9
+    },
+    {
+        "loss": 8.4263,
+        "grad_norm": 1.2977561950683594,
+        "learning_rate": 0.0006990461188413408,
+        "epoch": 0.9
+    },
+    {
+        "loss": 8.3236,
+        "grad_norm": 1.2732399702072144,
+        "learning_rate": 0.0006981709985122954,
+        "epoch": 0.91
+    },
+    {
+        "loss": 8.0876,
+        "grad_norm": 0.8092446327209473,
+        "learning_rate": 0.0006972958781832502,
+        "epoch": 0.91
+    },
+    {
+        "loss": 8.3052,
+        "grad_norm": 1.0607753992080688,
+        "learning_rate": 0.000696420757854205,
+        "epoch": 0.91
+    },
+    {
+        "loss": 8.2821,
+        "grad_norm": 1.2833763360977173,
+        "learning_rate": 0.0006955456375251597,
+        "epoch": 0.91
+    },
+    {
+        "loss": 8.0437,
+        "grad_norm": 1.2291605472564697,
+        "learning_rate": 0.0006946705171961145,
+        "epoch": 0.92
+    },
+    {
+        "loss": 7.9172,
+        "grad_norm": 0.9950680732727051,
+        "learning_rate": 0.0006937953968670692,
+        "epoch": 0.92
+    },
+    {
+        "loss": 7.8579,
+        "grad_norm": 1.170876145362854,
+        "learning_rate": 0.000692920276538024,
+        "epoch": 0.92
+    },
+    {
+        "loss": 8.7343,
+        "grad_norm": 1.0266340970993042,
+        "learning_rate": 0.0006920451562089788,
+        "epoch": 0.92
+    },
+    {
+        "loss": 8.3685,
+        "grad_norm": 1.1194366216659546,
+        "learning_rate": 0.0006911700358799335,
+        "epoch": 0.93
+    },
+    {
+        "loss": 8.8983,
+        "grad_norm": 1.130362868309021,
+        "learning_rate": 0.0006902949155508883,
+        "epoch": 0.93
+    },
+    {
+        "loss": 8.3624,
+        "grad_norm": 1.2582019567489624,
+        "learning_rate": 0.000689419795221843,
+        "epoch": 0.93
+    },
+    {
+        "loss": 8.5332,
+        "grad_norm": 1.0985493659973145,
+        "learning_rate": 0.0006885446748927978,
+        "epoch": 0.93
+    },
+    {
+        "loss": 8.263,
+        "grad_norm": 1.0480501651763916,
+        "learning_rate": 0.0006876695545637526,
+        "epoch": 0.94
+    },
+    {
+        "loss": 8.1911,
+        "grad_norm": 1.085471510887146,
+        "learning_rate": 0.0006867944342347073,
+        "epoch": 0.94
+    },
+    {
+        "loss": 8.6767,
+        "grad_norm": 1.109959602355957,
+        "learning_rate": 0.0006859193139056621,
+        "epoch": 0.94
+    },
+    {
+        "loss": 8.1904,
+        "grad_norm": 0.9299295544624329,
+        "learning_rate": 0.0006850441935766169,
+        "epoch": 0.94
+    },
+    {
+        "loss": 7.9858,
+        "grad_norm": 1.3819242715835571,
+        "learning_rate": 0.0006841690732475716,
+        "epoch": 0.95
+    },
+    {
+        "loss": 8.3134,
+        "grad_norm": 1.499324083328247,
+        "learning_rate": 0.0006832939529185264,
+        "epoch": 0.95
+    },
+    {
+        "loss": 8.1389,
+        "grad_norm": 1.0068879127502441,
+        "learning_rate": 0.0006824188325894811,
+        "epoch": 0.95
+    },
+    {
+        "loss": 8.0979,
+        "grad_norm": 1.232861876487732,
+        "learning_rate": 0.0006815437122604359,
+        "epoch": 0.96
+    },
+    {
+        "loss": 8.1456,
+        "grad_norm": 1.020922064781189,
+        "learning_rate": 0.0006806685919313907,
+        "epoch": 0.96
+    },
+    {
+        "loss": 8.1438,
+        "grad_norm": 1.2880629301071167,
+        "learning_rate": 0.0006797934716023453,
+        "epoch": 0.96
+    },
+    {
+        "loss": 7.8589,
+        "grad_norm": 1.2720872163772583,
+        "learning_rate": 0.0006789183512733001,
+        "epoch": 0.96
+    },
+    {
+        "loss": 8.338,
+        "grad_norm": 1.1569981575012207,
+        "learning_rate": 0.0006780432309442548,
+        "epoch": 0.97
+    },
+    {
+        "loss": 7.6167,
+        "grad_norm": 1.0755385160446167,
+        "learning_rate": 0.0006771681106152095,
+        "epoch": 0.97
+    },
+    {
+        "loss": 9.1889,
+        "grad_norm": 1.1371173858642578,
+        "learning_rate": 0.0006762929902861643,
+        "epoch": 0.97
+    },
+    {
+        "loss": 8.1603,
+        "grad_norm": 1.2543790340423584,
+        "learning_rate": 0.000675417869957119,
+        "epoch": 0.97
+    },
+    {
+        "loss": 8.1684,
+        "grad_norm": 1.665987491607666,
+        "learning_rate": 0.0006745427496280738,
+        "epoch": 0.98
+    },
+    {
+        "loss": 8.4957,
+        "grad_norm": 1.1479765176773071,
+        "learning_rate": 0.0006736676292990285,
+        "epoch": 0.98
+    },
+    {
+        "loss": 7.998,
+        "grad_norm": 1.1416277885437012,
+        "learning_rate": 0.0006727925089699833,
+        "epoch": 0.98
+    },
+    {
+        "loss": 8.4458,
+        "grad_norm": 1.2610832452774048,
+        "learning_rate": 0.0006719173886409381,
+        "epoch": 0.98
+    },
+    {
+        "loss": 8.2715,
+        "grad_norm": 1.2478748559951782,
+        "learning_rate": 0.0006710422683118928,
+        "epoch": 0.99
+    },
+    {
+        "loss": 8.0882,
+        "grad_norm": 0.9021313190460205,
+        "learning_rate": 0.0006701671479828476,
+        "epoch": 0.99
+    },
+    {
+        "loss": 8.2404,
+        "grad_norm": 1.0023951530456543,
+        "learning_rate": 0.0006692920276538024,
+        "epoch": 0.99
+    },
+    {
+        "loss": 8.681,
+        "grad_norm": 1.3342375755310059,
+        "learning_rate": 0.0006684169073247571,
+        "epoch": 0.99
+    },
+    {
+        "loss": 8.024,
+        "grad_norm": 1.0199118852615356,
+        "learning_rate": 0.0006675417869957119,
+        "epoch": 1.0
+    },
+    {
+        "loss": 8.3688,
+        "grad_norm": 0.893786609172821,
+        "learning_rate": 0.0006666666666666666,
+        "epoch": 1.0
+    },
+    {
+        "loss": 8.0561,
+        "grad_norm": 1.2774296998977661,
+        "learning_rate": 0.0006657915463376214,
+        "epoch": 1.0
+    },
+    {
+        "loss": 7.8444,
+        "grad_norm": 1.0824223756790161,
+        "learning_rate": 0.0006649164260085762,
+        "epoch": 1.01
+    },
+    {
+        "loss": 8.1771,
+        "grad_norm": 0.869452178478241,
+        "learning_rate": 0.0006640413056795309,
+        "epoch": 1.01
+    },
+    {
+        "loss": 7.6838,
+        "grad_norm": 1.1132241487503052,
+        "learning_rate": 0.0006631661853504857,
+        "epoch": 1.01
+    },
+    {
+        "loss": 7.9475,
+        "grad_norm": 1.2853749990463257,
+        "learning_rate": 0.0006622910650214405,
+        "epoch": 1.01
+    },
+    {
+        "loss": 8.8546,
+        "grad_norm": 1.2339048385620117,
+        "learning_rate": 0.0006614159446923952,
+        "epoch": 1.02
+    },
+    {
+        "loss": 8.1339,
+        "grad_norm": 1.2211487293243408,
+        "learning_rate": 0.00066054082436335,
+        "epoch": 1.02
+    },
+    {
+        "loss": 7.402,
+        "grad_norm": 1.0966975688934326,
+        "learning_rate": 0.0006596657040343047,
+        "epoch": 1.02
+    },
+    {
+        "loss": 8.1777,
+        "grad_norm": 1.0253325700759888,
+        "learning_rate": 0.0006587905837052595,
+        "epoch": 1.02
+    },
+    {
+        "loss": 8.2748,
+        "grad_norm": 1.2987836599349976,
+        "learning_rate": 0.0006579154633762143,
+        "epoch": 1.03
+    },
+    {
+        "loss": 8.3225,
+        "grad_norm": 0.945371687412262,
+        "learning_rate": 0.000657040343047169,
+        "epoch": 1.03
+    },
+    {
+        "loss": 8.4416,
+        "grad_norm": 1.0868079662322998,
+        "learning_rate": 0.0006561652227181238,
+        "epoch": 1.03
+    },
+    {
+        "loss": 8.1007,
+        "grad_norm": 1.0190479755401611,
+        "learning_rate": 0.0006552901023890784,
+        "epoch": 1.03
+    },
+    {
+        "loss": 8.1317,
+        "grad_norm": 1.0896625518798828,
+        "learning_rate": 0.0006544149820600332,
+        "epoch": 1.04
+    },
+    {
+        "loss": 7.7364,
+        "grad_norm": 1.1690502166748047,
+        "learning_rate": 0.000653539861730988,
+        "epoch": 1.04
+    },
+    {
+        "loss": 7.8173,
+        "grad_norm": 1.0521645545959473,
+        "learning_rate": 0.0006526647414019427,
+        "epoch": 1.04
+    },
+    {
+        "loss": 7.6212,
+        "grad_norm": 1.3057899475097656,
+        "learning_rate": 0.0006517896210728975,
+        "epoch": 1.04
+    },
+    {
+        "loss": 8.0228,
+        "grad_norm": 0.968885064125061,
+        "learning_rate": 0.0006509145007438522,
+        "epoch": 1.05
+    },
+    {
+        "loss": 7.8535,
+        "grad_norm": 1.1838873624801636,
+        "learning_rate": 0.000650039380414807,
+        "epoch": 1.05
+    },
+    {
+        "loss": 8.1991,
+        "grad_norm": 1.0967016220092773,
+        "learning_rate": 0.0006491642600857618,
+        "epoch": 1.05
+    },
+    {
+        "loss": 8.1515,
+        "grad_norm": 1.0798629522323608,
+        "learning_rate": 0.0006482891397567165,
+        "epoch": 1.06
+    },
+    {
+        "loss": 8.291,
+        "grad_norm": 1.1506596803665161,
+        "learning_rate": 0.0006474140194276713,
+        "epoch": 1.06
+    },
+    {
+        "loss": 7.956,
+        "grad_norm": 1.0459505319595337,
+        "learning_rate": 0.0006465388990986261,
+        "epoch": 1.06
+    },
+    {
+        "loss": 8.4393,
+        "grad_norm": 1.070776343345642,
+        "learning_rate": 0.0006456637787695808,
+        "epoch": 1.06
+    },
+    {
+        "loss": 8.5445,
+        "grad_norm": 1.3064284324645996,
+        "learning_rate": 0.0006447886584405356,
+        "epoch": 1.07
+    },
+    {
+        "loss": 8.701,
+        "grad_norm": 1.0707839727401733,
+        "learning_rate": 0.0006439135381114903,
+        "epoch": 1.07
+    },
+    {
+        "loss": 7.4342,
+        "grad_norm": 1.123377799987793,
+        "learning_rate": 0.0006430384177824451,
+        "epoch": 1.07
+    },
+    {
+        "loss": 8.4883,
+        "grad_norm": 1.7230886220932007,
+        "learning_rate": 0.0006421632974533999,
+        "epoch": 1.07
+    },
+    {
+        "loss": 8.5288,
+        "grad_norm": 0.9721227288246155,
+        "learning_rate": 0.0006412881771243546,
+        "epoch": 1.08
+    },
+    {
+        "loss": 7.8249,
+        "grad_norm": 1.2729851007461548,
+        "learning_rate": 0.0006404130567953094,
+        "epoch": 1.08
+    },
+    {
+        "loss": 8.3277,
+        "grad_norm": 0.9693044424057007,
+        "learning_rate": 0.0006395379364662642,
+        "epoch": 1.08
+    },
+    {
+        "loss": 7.8798,
+        "grad_norm": 1.104020118713379,
+        "learning_rate": 0.0006386628161372189,
+        "epoch": 1.08
+    },
+    {
+        "loss": 7.899,
+        "grad_norm": 1.0556141138076782,
+        "learning_rate": 0.0006377876958081737,
+        "epoch": 1.09
+    },
+    {
+        "loss": 8.6403,
+        "grad_norm": 1.227303147315979,
+        "learning_rate": 0.0006369125754791284,
+        "epoch": 1.09
+    },
+    {
+        "loss": 8.7407,
+        "grad_norm": 1.2486103773117065,
+        "learning_rate": 0.0006360374551500832,
+        "epoch": 1.09
+    },
+    {
+        "loss": 8.226,
+        "grad_norm": 1.1452488899230957,
+        "learning_rate": 0.000635162334821038,
+        "epoch": 1.09
+    },
+    {
+        "loss": 8.5083,
+        "grad_norm": 1.466182827949524,
+        "learning_rate": 0.0006342872144919927,
+        "epoch": 1.1
+    },
+    {
+        "loss": 7.8041,
+        "grad_norm": 1.2693302631378174,
+        "learning_rate": 0.0006334120941629474,
+        "epoch": 1.1
+    },
+    {
+        "loss": 7.918,
+        "grad_norm": 1.1236190795898438,
+        "learning_rate": 0.0006325369738339021,
+        "epoch": 1.1
+    },
+    {
+        "loss": 7.8792,
+        "grad_norm": 0.9166776537895203,
+        "learning_rate": 0.0006316618535048569,
+        "epoch": 1.11
+    },
+    {
+        "loss": 8.3714,
+        "grad_norm": 1.2021427154541016,
+        "learning_rate": 0.0006307867331758117,
+        "epoch": 1.11
+    },
+    {
+        "loss": 8.5282,
+        "grad_norm": 1.1508140563964844,
+        "learning_rate": 0.0006299116128467664,
+        "epoch": 1.11
+    },
+    {
+        "loss": 7.7235,
+        "grad_norm": 1.044027328491211,
+        "learning_rate": 0.0006290364925177212,
+        "epoch": 1.11
+    },
+    {
+        "loss": 8.0483,
+        "grad_norm": 1.00051748752594,
+        "learning_rate": 0.000628161372188676,
+        "epoch": 1.12
+    },
+    {
+        "loss": 8.0003,
+        "grad_norm": 1.0397716760635376,
+        "learning_rate": 0.0006272862518596307,
+        "epoch": 1.12
+    },
+    {
+        "loss": 8.274,
+        "grad_norm": 1.0577192306518555,
+        "learning_rate": 0.0006264111315305855,
+        "epoch": 1.12
+    },
+    {
+        "loss": 7.8435,
+        "grad_norm": 1.1829681396484375,
+        "learning_rate": 0.0006255360112015402,
+        "epoch": 1.12
+    },
+    {
+        "loss": 8.5019,
+        "grad_norm": 1.9353641271591187,
+        "learning_rate": 0.000624660890872495,
+        "epoch": 1.13
+    },
+    {
+        "loss": 8.4582,
+        "grad_norm": 1.237269639968872,
+        "learning_rate": 0.0006237857705434498,
+        "epoch": 1.13
+    },
+    {
+        "loss": 8.0735,
+        "grad_norm": 1.1674834489822388,
+        "learning_rate": 0.0006229106502144045,
+        "epoch": 1.13
+    },
+    {
+        "loss": 8.3781,
+        "grad_norm": 1.32883620262146,
+        "learning_rate": 0.0006220355298853593,
+        "epoch": 1.13
+    },
+    {
+        "loss": 8.723,
+        "grad_norm": 1.3197271823883057,
+        "learning_rate": 0.000621160409556314,
+        "epoch": 1.14
+    },
+    {
+        "loss": 8.414,
+        "grad_norm": 1.137764573097229,
+        "learning_rate": 0.0006202852892272688,
+        "epoch": 1.14
+    },
+    {
+        "loss": 7.9197,
+        "grad_norm": 1.1574738025665283,
+        "learning_rate": 0.0006194101688982236,
+        "epoch": 1.14
+    },
+    {
+        "loss": 8.09,
+        "grad_norm": 1.0444676876068115,
+        "learning_rate": 0.0006185350485691783,
+        "epoch": 1.14
+    },
+    {
+        "loss": 7.3329,
+        "grad_norm": 0.8655235767364502,
+        "learning_rate": 0.0006176599282401331,
+        "epoch": 1.15
+    },
+    {
+        "loss": 8.4163,
+        "grad_norm": 0.9860300421714783,
+        "learning_rate": 0.0006167848079110879,
+        "epoch": 1.15
+    },
+    {
+        "loss": 8.2608,
+        "grad_norm": 1.1680139303207397,
+        "learning_rate": 0.0006159096875820426,
+        "epoch": 1.15
+    },
+    {
+        "loss": 7.9283,
+        "grad_norm": 1.545938491821289,
+        "learning_rate": 0.0006150345672529974,
+        "epoch": 1.15
+    },
+    {
+        "loss": 8.4113,
+        "grad_norm": 1.2768994569778442,
+        "learning_rate": 0.0006141594469239521,
+        "epoch": 1.16
+    },
+    {
+        "loss": 8.2389,
+        "grad_norm": 1.0001721382141113,
+        "learning_rate": 0.0006132843265949069,
+        "epoch": 1.16
+    },
+    {
+        "loss": 8.397,
+        "grad_norm": 1.8651808500289917,
+        "learning_rate": 0.0006124092062658617,
+        "epoch": 1.16
+    },
+    {
+        "loss": 8.003,
+        "grad_norm": 0.947693407535553,
+        "learning_rate": 0.0006115340859368163,
+        "epoch": 1.17
+    },
+    {
+        "loss": 7.5861,
+        "grad_norm": 1.1168384552001953,
+        "learning_rate": 0.0006106589656077711,
+        "epoch": 1.17
+    },
+    {
+        "loss": 8.7788,
+        "grad_norm": 1.1341112852096558,
+        "learning_rate": 0.0006097838452787258,
+        "epoch": 1.17
+    },
+    {
+        "loss": 7.9428,
+        "grad_norm": 1.2905473709106445,
+        "learning_rate": 0.0006089087249496806,
+        "epoch": 1.17
+    },
+    {
+        "loss": 8.6196,
+        "grad_norm": 0.9961435794830322,
+        "learning_rate": 0.0006080336046206354,
+        "epoch": 1.18
+    },
+    {
+        "loss": 8.224,
+        "grad_norm": 1.3134316205978394,
+        "learning_rate": 0.0006071584842915901,
+        "epoch": 1.18
+    },
+    {
+        "loss": 7.9156,
+        "grad_norm": 1.5898418426513672,
+        "learning_rate": 0.0006062833639625449,
+        "epoch": 1.18
+    },
+    {
+        "loss": 8.2147,
+        "grad_norm": 0.99250727891922,
+        "learning_rate": 0.0006054082436334996,
+        "epoch": 1.18
+    },
+    {
+        "loss": 7.6957,
+        "grad_norm": 1.2642431259155273,
+        "learning_rate": 0.0006045331233044544,
+        "epoch": 1.19
+    },
+    {
+        "loss": 7.7926,
+        "grad_norm": 1.314082384109497,
+        "learning_rate": 0.0006036580029754092,
+        "epoch": 1.19
+    },
+    {
+        "loss": 7.9682,
+        "grad_norm": 1.1342573165893555,
+        "learning_rate": 0.0006027828826463639,
+        "epoch": 1.19
+    },
+    {
+        "loss": 8.0208,
+        "grad_norm": 1.3015680313110352,
+        "learning_rate": 0.0006019077623173187,
+        "epoch": 1.19
+    },
+    {
+        "loss": 8.3608,
+        "grad_norm": 0.9990431666374207,
+        "learning_rate": 0.0006010326419882735,
+        "epoch": 1.2
+    },
+    {
+        "loss": 8.2009,
+        "grad_norm": 0.9804344773292542,
+        "learning_rate": 0.0006001575216592282,
+        "epoch": 1.2
+    },
+    {
+        "loss": 8.0484,
+        "grad_norm": 1.1591954231262207,
+        "learning_rate": 0.0005992824013301829,
+        "epoch": 1.2
+    },
+    {
+        "loss": 8.116,
+        "grad_norm": 1.042474627494812,
+        "learning_rate": 0.0005984072810011376,
+        "epoch": 1.2
+    },
+    {
+        "loss": 7.9246,
+        "grad_norm": 1.8579179048538208,
+        "learning_rate": 0.0005975321606720924,
+        "epoch": 1.21
+    },
+    {
+        "loss": 7.9183,
+        "grad_norm": 0.8727061748504639,
+        "learning_rate": 0.0005966570403430472,
+        "epoch": 1.21
+    },
+    {
+        "loss": 7.675,
+        "grad_norm": 1.0189380645751953,
+        "learning_rate": 0.0005957819200140019,
+        "epoch": 1.21
+    },
+    {
+        "loss": 7.6222,
+        "grad_norm": 1.0766206979751587,
+        "learning_rate": 0.0005949067996849567,
+        "epoch": 1.22
+    },
+    {
+        "loss": 7.6455,
+        "grad_norm": 1.121745228767395,
+        "learning_rate": 0.0005940316793559114,
+        "epoch": 1.22
+    },
+    {
+        "loss": 8.1449,
+        "grad_norm": 1.2497507333755493,
+        "learning_rate": 0.0005931565590268662,
+        "epoch": 1.22
+    },
+    {
+        "loss": 8.3586,
+        "grad_norm": 1.301903486251831,
+        "learning_rate": 0.000592281438697821,
+        "epoch": 1.22
+    },
+    {
+        "loss": 8.163,
+        "grad_norm": 1.1964079141616821,
+        "learning_rate": 0.0005914063183687757,
+        "epoch": 1.23
+    },
+    {
+        "loss": 8.2938,
+        "grad_norm": 1.1423827409744263,
+        "learning_rate": 0.0005905311980397304,
+        "epoch": 1.23
+    },
+    {
+        "loss": 8.165,
+        "grad_norm": 1.119884967803955,
+        "learning_rate": 0.0005896560777106851,
+        "epoch": 1.23
+    },
+    {
+        "loss": 7.7234,
+        "grad_norm": 1.4375518560409546,
+        "learning_rate": 0.0005887809573816399,
+        "epoch": 1.23
+    },
+    {
+        "loss": 8.0758,
+        "grad_norm": 1.1417185068130493,
+        "learning_rate": 0.0005879058370525947,
+        "epoch": 1.24
+    },
+    {
+        "loss": 7.9137,
+        "grad_norm": 1.048060417175293,
+        "learning_rate": 0.0005870307167235494,
+        "epoch": 1.24
+    },
+    {
+        "loss": 8.4029,
+        "grad_norm": 0.9880658388137817,
+        "learning_rate": 0.0005861555963945042,
+        "epoch": 1.24
+    },
+    {
+        "loss": 8.4489,
+        "grad_norm": 1.000611424446106,
+        "learning_rate": 0.000585280476065459,
+        "epoch": 1.24
+    },
+    {
+        "loss": 8.2688,
+        "grad_norm": 1.3099920749664307,
+        "learning_rate": 0.0005844053557364137,
+        "epoch": 1.25
+    },
+    {
+        "loss": 7.7948,
+        "grad_norm": 0.8548302054405212,
+        "learning_rate": 0.0005835302354073685,
+        "epoch": 1.25
+    },
+    {
+        "loss": 8.442,
+        "grad_norm": 1.1732860803604126,
+        "learning_rate": 0.0005826551150783232,
+        "epoch": 1.25
+    },
+    {
+        "loss": 7.6346,
+        "grad_norm": 0.803125262260437,
+        "learning_rate": 0.000581779994749278,
+        "epoch": 1.25
+    },
+    {
+        "loss": 8.0567,
+        "grad_norm": 1.258419156074524,
+        "learning_rate": 0.0005809048744202328,
+        "epoch": 1.26
+    },
+    {
+        "loss": 8.1142,
+        "grad_norm": 1.1331418752670288,
+        "learning_rate": 0.0005800297540911875,
+        "epoch": 1.26
+    },
+    {
+        "loss": 8.5457,
+        "grad_norm": 1.5619804859161377,
+        "learning_rate": 0.0005791546337621423,
+        "epoch": 1.26
+    },
+    {
+        "loss": 7.9416,
+        "grad_norm": 1.880534052848816,
+        "learning_rate": 0.000578279513433097,
+        "epoch": 1.27
+    },
+    {
+        "loss": 7.8216,
+        "grad_norm": 1.2279471158981323,
+        "learning_rate": 0.0005774043931040518,
+        "epoch": 1.27
+    },
+    {
+        "loss": 7.8216,
+        "grad_norm": 1.1597974300384521,
+        "learning_rate": 0.0005765292727750066,
+        "epoch": 1.27
+    },
+    {
+        "loss": 7.9033,
+        "grad_norm": 1.1710484027862549,
+        "learning_rate": 0.0005756541524459613,
+        "epoch": 1.27
+    },
+    {
+        "loss": 7.6036,
+        "grad_norm": 1.0655231475830078,
+        "learning_rate": 0.0005747790321169161,
+        "epoch": 1.28
+    },
+    {
+        "loss": 7.5982,
+        "grad_norm": 1.0066710710525513,
+        "learning_rate": 0.0005739039117878709,
+        "epoch": 1.28
+    },
+    {
+        "loss": 7.738,
+        "grad_norm": 1.1333460807800293,
+        "learning_rate": 0.0005730287914588256,
+        "epoch": 1.28
+    },
+    {
+        "loss": 8.0025,
+        "grad_norm": 1.468841791152954,
+        "learning_rate": 0.0005721536711297804,
+        "epoch": 1.28
+    },
+    {
+        "loss": 7.4888,
+        "grad_norm": 1.1363178491592407,
+        "learning_rate": 0.0005712785508007351,
+        "epoch": 1.29
+    },
+    {
+        "loss": 7.3176,
+        "grad_norm": 1.1589970588684082,
+        "learning_rate": 0.0005704034304716899,
+        "epoch": 1.29
+    },
+    {
+        "loss": 7.6323,
+        "grad_norm": 0.9033693075180054,
+        "learning_rate": 0.0005695283101426447,
+        "epoch": 1.29
+    },
+    {
+        "loss": 7.8839,
+        "grad_norm": 1.2384039163589478,
+        "learning_rate": 0.0005686531898135993,
+        "epoch": 1.29
+    },
+    {
+        "loss": 7.8408,
+        "grad_norm": 1.3826912641525269,
+        "learning_rate": 0.0005677780694845541,
+        "epoch": 1.3
+    },
+    {
+        "loss": 7.4433,
+        "grad_norm": 1.1403487920761108,
+        "learning_rate": 0.0005669029491555088,
+        "epoch": 1.3
+    },
+    {
+        "loss": 8.5407,
+        "grad_norm": 1.037423014640808,
+        "learning_rate": 0.0005660278288264636,
+        "epoch": 1.3
+    },
+    {
+        "loss": 8.0943,
+        "grad_norm": 1.4421013593673706,
+        "learning_rate": 0.0005651527084974184,
+        "epoch": 1.3
+    },
+    {
+        "loss": 7.7771,
+        "grad_norm": 1.2977713346481323,
+        "learning_rate": 0.0005642775881683731,
+        "epoch": 1.31
+    },
+    {
+        "loss": 7.54,
+        "grad_norm": 1.049196720123291,
+        "learning_rate": 0.0005634024678393279,
+        "epoch": 1.31
+    },
+    {
+        "loss": 7.4699,
+        "grad_norm": 1.0489652156829834,
+        "learning_rate": 0.0005625273475102827,
+        "epoch": 1.31
+    },
+    {
+        "loss": 7.9441,
+        "grad_norm": 1.1373968124389648,
+        "learning_rate": 0.0005616522271812374,
+        "epoch": 1.32
+    },
+    {
+        "loss": 7.2627,
+        "grad_norm": 1.0570902824401855,
+        "learning_rate": 0.0005607771068521922,
+        "epoch": 1.32
+    },
+    {
+        "loss": 7.7472,
+        "grad_norm": 1.0547776222229004,
+        "learning_rate": 0.0005599019865231469,
+        "epoch": 1.32
+    },
+    {
+        "loss": 7.8815,
+        "grad_norm": 1.2481534481048584,
+        "learning_rate": 0.0005590268661941017,
+        "epoch": 1.32
+    },
+    {
+        "loss": 8.2547,
+        "grad_norm": 1.1728442907333374,
+        "learning_rate": 0.0005581517458650565,
+        "epoch": 1.33
+    },
+    {
+        "loss": 7.5035,
+        "grad_norm": 1.0567808151245117,
+        "learning_rate": 0.0005572766255360112,
+        "epoch": 1.33
+    },
+    {
+        "loss": 7.9982,
+        "grad_norm": 0.8234537243843079,
+        "learning_rate": 0.000556401505206966,
+        "epoch": 1.33
+    },
+    {
+        "loss": 7.5333,
+        "grad_norm": 1.09587824344635,
+        "learning_rate": 0.0005555263848779207,
+        "epoch": 1.33
+    },
+    {
+        "loss": 7.768,
+        "grad_norm": 1.3897008895874023,
+        "learning_rate": 0.0005546512645488755,
+        "epoch": 1.34
+    },
+    {
+        "loss": 7.7645,
+        "grad_norm": 1.1089082956314087,
+        "learning_rate": 0.0005537761442198303,
+        "epoch": 1.34
+    },
+    {
+        "loss": 7.7809,
+        "grad_norm": 1.2678576707839966,
+        "learning_rate": 0.000552901023890785,
+        "epoch": 1.34
+    },
+    {
+        "loss": 7.7376,
+        "grad_norm": 1.3946635723114014,
+        "learning_rate": 0.0005520259035617398,
+        "epoch": 1.34
+    },
+    {
+        "loss": 8.2773,
+        "grad_norm": 1.3742512464523315,
+        "learning_rate": 0.0005511507832326946,
+        "epoch": 1.35
+    },
+    {
+        "loss": 7.7902,
+        "grad_norm": 1.416434645652771,
+        "learning_rate": 0.0005502756629036493,
+        "epoch": 1.35
+    },
+    {
+        "loss": 7.6157,
+        "grad_norm": 1.0419012308120728,
+        "learning_rate": 0.0005494005425746041,
+        "epoch": 1.35
+    },
+    {
+        "loss": 7.5897,
+        "grad_norm": 1.7180145978927612,
+        "learning_rate": 0.0005485254222455588,
+        "epoch": 1.35
+    },
+    {
+        "loss": 8.0068,
+        "grad_norm": 1.6651771068572998,
+        "learning_rate": 0.0005476503019165136,
+        "epoch": 1.36
+    },
+    {
+        "loss": 7.4023,
+        "grad_norm": 1.0715596675872803,
+        "learning_rate": 0.0005467751815874683,
+        "epoch": 1.36
+    },
+    {
+        "loss": 8.0369,
+        "grad_norm": 1.208898901939392,
+        "learning_rate": 0.000545900061258423,
+        "epoch": 1.36
+    },
+    {
+        "loss": 7.6188,
+        "grad_norm": 0.9920070767402649,
+        "learning_rate": 0.0005450249409293778,
+        "epoch": 1.36
+    },
+    {
+        "loss": 8.6854,
+        "grad_norm": 1.174086570739746,
+        "learning_rate": 0.0005441498206003325,
+        "epoch": 1.37
+    },
+    {
+        "loss": 7.5733,
+        "grad_norm": 1.244912028312683,
+        "learning_rate": 0.0005432747002712873,
+        "epoch": 1.37
+    },
+    {
+        "loss": 7.389,
+        "grad_norm": 1.5966273546218872,
+        "learning_rate": 0.0005423995799422421,
+        "epoch": 1.37
+    },
+    {
+        "loss": 8.1756,
+        "grad_norm": 1.0320965051651,
+        "learning_rate": 0.0005415244596131968,
+        "epoch": 1.38
+    },
+    {
+        "loss": 8.897,
+        "grad_norm": 1.2478450536727905,
+        "learning_rate": 0.0005406493392841516,
+        "epoch": 1.38
+    },
+    {
+        "loss": 7.6083,
+        "grad_norm": 1.4347364902496338,
+        "learning_rate": 0.0005397742189551064,
+        "epoch": 1.38
+    },
+    {
+        "loss": 7.9916,
+        "grad_norm": 1.1878119707107544,
+        "learning_rate": 0.0005388990986260611,
+        "epoch": 1.38
+    },
+    {
+        "loss": 8.1032,
+        "grad_norm": 1.3169543743133545,
+        "learning_rate": 0.0005380239782970159,
+        "epoch": 1.39
+    },
+    {
+        "loss": 7.3094,
+        "grad_norm": 1.271192193031311,
+        "learning_rate": 0.0005371488579679706,
+        "epoch": 1.39
+    },
+    {
+        "loss": 7.2947,
+        "grad_norm": 1.484824299812317,
+        "learning_rate": 0.0005362737376389254,
+        "epoch": 1.39
+    },
+    {
+        "loss": 7.7483,
+        "grad_norm": 1.0237884521484375,
+        "learning_rate": 0.0005353986173098802,
+        "epoch": 1.39
+    },
+    {
+        "loss": 7.7284,
+        "grad_norm": 1.141897201538086,
+        "learning_rate": 0.0005345234969808349,
+        "epoch": 1.4
+    },
+    {
+        "loss": 7.9684,
+        "grad_norm": 1.2076783180236816,
+        "learning_rate": 0.0005336483766517897,
+        "epoch": 1.4
+    },
+    {
+        "loss": 7.4731,
+        "grad_norm": 1.0815685987472534,
+        "learning_rate": 0.0005327732563227444,
+        "epoch": 1.4
+    },
+    {
+        "loss": 7.6468,
+        "grad_norm": 1.9115163087844849,
+        "learning_rate": 0.0005318981359936992,
+        "epoch": 1.4
+    },
+    {
+        "loss": 8.179,
+        "grad_norm": 1.1872133016586304,
+        "learning_rate": 0.000531023015664654,
+        "epoch": 1.41
+    },
+    {
+        "loss": 8.1254,
+        "grad_norm": 1.144726037979126,
+        "learning_rate": 0.0005301478953356087,
+        "epoch": 1.41
+    },
+    {
+        "loss": 7.7947,
+        "grad_norm": 1.562495231628418,
+        "learning_rate": 0.0005292727750065635,
+        "epoch": 1.41
+    },
+    {
+        "loss": 7.2917,
+        "grad_norm": 1.20420241355896,
+        "learning_rate": 0.0005283976546775183,
+        "epoch": 1.41
+    },
+    {
+        "loss": 7.9956,
+        "grad_norm": 1.0302613973617554,
+        "learning_rate": 0.000527522534348473,
+        "epoch": 1.42
+    },
+    {
+        "loss": 7.8058,
+        "grad_norm": 1.161452293395996,
+        "learning_rate": 0.0005266474140194278,
+        "epoch": 1.42
+    },
+    {
+        "loss": 8.2652,
+        "grad_norm": 1.2876991033554077,
+        "learning_rate": 0.0005257722936903825,
+        "epoch": 1.42
+    },
+    {
+        "loss": 8.0375,
+        "grad_norm": 1.1002925634384155,
+        "learning_rate": 0.0005248971733613372,
+        "epoch": 1.43
+    },
+    {
+        "loss": 7.82,
+        "grad_norm": 1.0201154947280884,
+        "learning_rate": 0.000524022053032292,
+        "epoch": 1.43
+    },
+    {
+        "loss": 8.3203,
+        "grad_norm": 1.1177037954330444,
+        "learning_rate": 0.0005231469327032467,
+        "epoch": 1.43
+    },
+    {
+        "loss": 7.9789,
+        "grad_norm": 1.4295682907104492,
+        "learning_rate": 0.0005222718123742015,
+        "epoch": 1.43
+    },
+    {
+        "loss": 8.0088,
+        "grad_norm": 1.4420737028121948,
+        "learning_rate": 0.0005213966920451562,
+        "epoch": 1.44
+    },
+    {
+        "loss": 7.8298,
+        "grad_norm": 1.1020231246948242,
+        "learning_rate": 0.000520521571716111,
+        "epoch": 1.44
+    },
+    {
+        "loss": 7.8801,
+        "grad_norm": 1.4339189529418945,
+        "learning_rate": 0.0005196464513870657,
+        "epoch": 1.44
+    },
+    {
+        "loss": 7.6756,
+        "grad_norm": 1.5243607759475708,
+        "learning_rate": 0.0005187713310580204,
+        "epoch": 1.44
+    },
+    {
+        "loss": 8.1007,
+        "grad_norm": 0.9880979657173157,
+        "learning_rate": 0.0005178962107289752,
+        "epoch": 1.45
+    },
+    {
+        "loss": 7.7396,
+        "grad_norm": 1.1447367668151855,
+        "learning_rate": 0.0005170210903999299,
+        "epoch": 1.45
+    },
+    {
+        "loss": 7.8537,
+        "grad_norm": 1.384048342704773,
+        "learning_rate": 0.0005161459700708847,
+        "epoch": 1.45
+    },
+    {
+        "loss": 7.8855,
+        "grad_norm": 1.3757721185684204,
+        "learning_rate": 0.0005152708497418395,
+        "epoch": 1.45
+    },
+    {
+        "loss": 7.8651,
+        "grad_norm": 1.1160024404525757,
+        "learning_rate": 0.0005143957294127942,
+        "epoch": 1.46
+    },
+    {
+        "loss": 7.8378,
+        "grad_norm": 0.9774546027183533,
+        "learning_rate": 0.000513520609083749,
+        "epoch": 1.46
+    },
+    {
+        "loss": 7.9251,
+        "grad_norm": 1.5181477069854736,
+        "learning_rate": 0.0005126454887547038,
+        "epoch": 1.46
+    },
+    {
+        "loss": 8.6781,
+        "grad_norm": 1.203229308128357,
+        "learning_rate": 0.0005117703684256585,
+        "epoch": 1.46
+    },
+    {
+        "loss": 7.6571,
+        "grad_norm": 1.0401496887207031,
+        "learning_rate": 0.0005108952480966133,
+        "epoch": 1.47
+    },
+    {
+        "loss": 7.3908,
+        "grad_norm": 1.3228225708007812,
+        "learning_rate": 0.000510020127767568,
+        "epoch": 1.47
+    },
+    {
+        "loss": 8.1244,
+        "grad_norm": 1.3072296380996704,
+        "learning_rate": 0.0005091450074385228,
+        "epoch": 1.47
+    },
+    {
+        "loss": 7.7535,
+        "grad_norm": 1.9105629920959473,
+        "learning_rate": 0.0005082698871094776,
+        "epoch": 1.48
+    },
+    {
+        "loss": 8.2387,
+        "grad_norm": 1.3035160303115845,
+        "learning_rate": 0.0005073947667804323,
+        "epoch": 1.48
+    },
+    {
+        "loss": 7.998,
+        "grad_norm": 0.9805745482444763,
+        "learning_rate": 0.0005065196464513871,
+        "epoch": 1.48
+    },
+    {
+        "loss": 8.0499,
+        "grad_norm": 1.28218412399292,
+        "learning_rate": 0.0005056445261223418,
+        "epoch": 1.48
+    },
+    {
+        "loss": 8.0939,
+        "grad_norm": 1.289697527885437,
+        "learning_rate": 0.0005047694057932966,
+        "epoch": 1.49
+    },
+    {
+        "loss": 7.8801,
+        "grad_norm": 1.3982206583023071,
+        "learning_rate": 0.0005038942854642513,
+        "epoch": 1.49
+    },
+    {
+        "loss": 7.5012,
+        "grad_norm": 1.1884011030197144,
+        "learning_rate": 0.000503019165135206,
+        "epoch": 1.49
+    },
+    {
+        "loss": 7.7792,
+        "grad_norm": 1.2014328241348267,
+        "learning_rate": 0.0005021440448061608,
+        "epoch": 1.49
+    },
+    {
+        "loss": 8.3151,
+        "grad_norm": 1.2958098649978638,
+        "learning_rate": 0.0005012689244771155,
+        "epoch": 1.5
+    },
+    {
+        "loss": 7.3702,
+        "grad_norm": 1.1195346117019653,
+        "learning_rate": 0.0005003938041480703,
+        "epoch": 1.5
+    },
+    {
+        "loss": 8.0952,
+        "grad_norm": 1.2185337543487549,
+        "learning_rate": 0.0004995186838190251,
+        "epoch": 1.5
+    },
+    {
+        "loss": 7.6605,
+        "grad_norm": 1.1054099798202515,
+        "learning_rate": 0.0004986435634899798,
+        "epoch": 1.5
+    },
+    {
+        "loss": 7.8926,
+        "grad_norm": 1.3183029890060425,
+        "learning_rate": 0.0004977684431609346,
+        "epoch": 1.51
+    },
+    {
+        "loss": 7.8356,
+        "grad_norm": 1.3786067962646484,
+        "learning_rate": 0.0004968933228318894,
+        "epoch": 1.51
+    },
+    {
+        "loss": 7.7605,
+        "grad_norm": 1.3373888731002808,
+        "learning_rate": 0.0004960182025028441,
+        "epoch": 1.51
+    },
+    {
+        "loss": 7.9272,
+        "grad_norm": 1.5524091720581055,
+        "learning_rate": 0.0004951430821737989,
+        "epoch": 1.51
+    },
+    {
+        "loss": 8.1264,
+        "grad_norm": 0.927689790725708,
+        "learning_rate": 0.0004942679618447536,
+        "epoch": 1.52
+    },
+    {
+        "loss": 8.1456,
+        "grad_norm": 1.4429559707641602,
+        "learning_rate": 0.0004933928415157084,
+        "epoch": 1.52
+    },
+    {
+        "loss": 8.5349,
+        "grad_norm": 1.17830228805542,
+        "learning_rate": 0.0004925177211866632,
+        "epoch": 1.52
+    },
+    {
+        "loss": 8.4138,
+        "grad_norm": 1.7398778200149536,
+        "learning_rate": 0.0004916426008576179,
+        "epoch": 1.53
+    },
+    {
+        "loss": 7.6329,
+        "grad_norm": 1.101945161819458,
+        "learning_rate": 0.0004907674805285727,
+        "epoch": 1.53
+    },
+    {
+        "loss": 8.2694,
+        "grad_norm": 1.2424931526184082,
+        "learning_rate": 0.0004898923601995274,
+        "epoch": 1.53
+    },
+    {
+        "loss": 7.2639,
+        "grad_norm": 0.8726850748062134,
+        "learning_rate": 0.0004890172398704822,
+        "epoch": 1.53
+    },
+    {
+        "loss": 7.5542,
+        "grad_norm": 1.020978331565857,
+        "learning_rate": 0.0004881421195414369,
+        "epoch": 1.54
+    },
+    {
+        "loss": 7.3334,
+        "grad_norm": 1.058136224746704,
+        "learning_rate": 0.0004872669992123917,
+        "epoch": 1.54
+    },
+    {
+        "loss": 7.6285,
+        "grad_norm": 1.7856310606002808,
+        "learning_rate": 0.00048639187888334644,
+        "epoch": 1.54
+    },
+    {
+        "loss": 7.8873,
+        "grad_norm": 1.1540299654006958,
+        "learning_rate": 0.0004855167585543012,
+        "epoch": 1.54
+    },
+    {
+        "loss": 7.5676,
+        "grad_norm": 1.4844547510147095,
+        "learning_rate": 0.00048464163822525597,
+        "epoch": 1.55
+    },
+    {
+        "loss": 8.0284,
+        "grad_norm": 1.1018364429473877,
+        "learning_rate": 0.00048376651789621073,
+        "epoch": 1.55
+    },
+    {
+        "loss": 7.8478,
+        "grad_norm": 1.4421080350875854,
+        "learning_rate": 0.0004828913975671655,
+        "epoch": 1.55
+    },
+    {
+        "loss": 8.0614,
+        "grad_norm": 1.322413444519043,
+        "learning_rate": 0.00048201627723812025,
+        "epoch": 1.55
+    },
+    {
+        "loss": 7.9015,
+        "grad_norm": 1.1930081844329834,
+        "learning_rate": 0.000481141156909075,
+        "epoch": 1.56
+    },
+    {
+        "loss": 7.843,
+        "grad_norm": 1.2846688032150269,
+        "learning_rate": 0.0004802660365800298,
+        "epoch": 1.56
+    },
+    {
+        "loss": 7.8268,
+        "grad_norm": 2.0413529872894287,
+        "learning_rate": 0.00047939091625098454,
+        "epoch": 1.56
+    },
+    {
+        "loss": 7.3241,
+        "grad_norm": 1.058362364768982,
+        "learning_rate": 0.0004785157959219393,
+        "epoch": 1.56
+    },
+    {
+        "loss": 7.8329,
+        "grad_norm": 1.725417971611023,
+        "learning_rate": 0.00047764067559289406,
+        "epoch": 1.57
+    },
+    {
+        "loss": 7.6295,
+        "grad_norm": 1.1373404264450073,
+        "learning_rate": 0.00047676555526384877,
+        "epoch": 1.57
+    },
+    {
+        "loss": 7.4763,
+        "grad_norm": 1.1107378005981445,
+        "learning_rate": 0.00047589043493480353,
+        "epoch": 1.57
+    },
+    {
+        "loss": 7.8846,
+        "grad_norm": 1.2450941801071167,
+        "learning_rate": 0.0004750153146057583,
+        "epoch": 1.57
+    },
+    {
+        "loss": 8.4109,
+        "grad_norm": 1.0643541812896729,
+        "learning_rate": 0.00047414019427671305,
+        "epoch": 1.58
+    },
+    {
+        "loss": 7.9126,
+        "grad_norm": 1.2940372228622437,
+        "learning_rate": 0.0004732650739476678,
+        "epoch": 1.58
+    },
+    {
+        "loss": 7.7132,
+        "grad_norm": 2.6067655086517334,
+        "learning_rate": 0.0004723899536186226,
+        "epoch": 1.58
+    },
+    {
+        "loss": 7.5708,
+        "grad_norm": 0.9783037304878235,
+        "learning_rate": 0.00047151483328957734,
+        "epoch": 1.59
+    },
+    {
+        "loss": 7.2771,
+        "grad_norm": 1.037582278251648,
+        "learning_rate": 0.0004706397129605321,
+        "epoch": 1.59
+    },
+    {
+        "loss": 7.7599,
+        "grad_norm": 1.0178707838058472,
+        "learning_rate": 0.00046976459263148686,
+        "epoch": 1.59
+    },
+    {
+        "loss": 7.1538,
+        "grad_norm": 1.558307409286499,
+        "learning_rate": 0.0004688894723024416,
+        "epoch": 1.59
+    },
+    {
+        "loss": 7.5229,
+        "grad_norm": 1.1060800552368164,
+        "learning_rate": 0.0004680143519733964,
+        "epoch": 1.6
+    },
+    {
+        "loss": 7.5813,
+        "grad_norm": 1.8988709449768066,
+        "learning_rate": 0.00046713923164435115,
+        "epoch": 1.6
+    },
+    {
+        "loss": 7.8319,
+        "grad_norm": 1.6066781282424927,
+        "learning_rate": 0.00046626411131530586,
+        "epoch": 1.6
+    },
+    {
+        "loss": 7.8222,
+        "grad_norm": 1.4711729288101196,
+        "learning_rate": 0.0004653889909862606,
+        "epoch": 1.6
+    },
+    {
+        "loss": 7.6115,
+        "grad_norm": 1.3585811853408813,
+        "learning_rate": 0.0004645138706572154,
+        "epoch": 1.61
+    },
+    {
+        "loss": 7.7618,
+        "grad_norm": 1.1487444639205933,
+        "learning_rate": 0.00046363875032817014,
+        "epoch": 1.61
+    },
+    {
+        "loss": 7.868,
+        "grad_norm": 1.4386248588562012,
+        "learning_rate": 0.0004627636299991249,
+        "epoch": 1.61
+    },
+    {
+        "loss": 7.7931,
+        "grad_norm": 1.0714224576950073,
+        "learning_rate": 0.00046188850967007967,
+        "epoch": 1.61
+    },
+    {
+        "loss": 8.2688,
+        "grad_norm": 1.6375863552093506,
+        "learning_rate": 0.00046101338934103443,
+        "epoch": 1.62
+    },
+    {
+        "loss": 7.621,
+        "grad_norm": 1.024120807647705,
+        "learning_rate": 0.0004601382690119892,
+        "epoch": 1.62
+    },
+    {
+        "loss": 8.2226,
+        "grad_norm": 1.2234493494033813,
+        "learning_rate": 0.0004592631486829439,
+        "epoch": 1.62
+    },
+    {
+        "loss": 7.596,
+        "grad_norm": 1.0593066215515137,
+        "learning_rate": 0.00045838802835389866,
+        "epoch": 1.62
+    },
+    {
+        "loss": 7.7407,
+        "grad_norm": 1.2529680728912354,
+        "learning_rate": 0.0004575129080248534,
+        "epoch": 1.63
+    },
+    {
+        "loss": 7.221,
+        "grad_norm": 1.1312929391860962,
+        "learning_rate": 0.0004566377876958082,
+        "epoch": 1.63
+    },
+    {
+        "loss": 7.6136,
+        "grad_norm": 1.4004312753677368,
+        "learning_rate": 0.00045576266736676294,
+        "epoch": 1.63
+    },
+    {
+        "loss": 7.9718,
+        "grad_norm": 1.4514151811599731,
+        "learning_rate": 0.00045488754703771765,
+        "epoch": 1.64
+    },
+    {
+        "loss": 7.3337,
+        "grad_norm": 1.1595350503921509,
+        "learning_rate": 0.0004540124267086724,
+        "epoch": 1.64
+    },
+    {
+        "loss": 7.7414,
+        "grad_norm": 1.1403205394744873,
+        "learning_rate": 0.0004531373063796272,
+        "epoch": 1.64
+    },
+    {
+        "loss": 7.7323,
+        "grad_norm": 1.677051305770874,
+        "learning_rate": 0.00045226218605058194,
+        "epoch": 1.64
+    },
+    {
+        "loss": 8.0048,
+        "grad_norm": 1.338146686553955,
+        "learning_rate": 0.0004513870657215367,
+        "epoch": 1.65
+    },
+    {
+        "loss": 7.9544,
+        "grad_norm": 1.0941588878631592,
+        "learning_rate": 0.00045051194539249146,
+        "epoch": 1.65
+    },
+    {
+        "loss": 8.1043,
+        "grad_norm": 1.224746584892273,
+        "learning_rate": 0.0004496368250634462,
+        "epoch": 1.65
+    },
+    {
+        "loss": 8.0849,
+        "grad_norm": 1.5772489309310913,
+        "learning_rate": 0.000448761704734401,
+        "epoch": 1.65
+    },
+    {
+        "loss": 7.3165,
+        "grad_norm": 1.4434912204742432,
+        "learning_rate": 0.00044788658440535575,
+        "epoch": 1.66
+    },
+    {
+        "loss": 7.8826,
+        "grad_norm": 0.9971029162406921,
+        "learning_rate": 0.0004470114640763105,
+        "epoch": 1.66
+    },
+    {
+        "loss": 7.7822,
+        "grad_norm": 1.061712384223938,
+        "learning_rate": 0.00044613634374726527,
+        "epoch": 1.66
+    },
+    {
+        "loss": 7.8387,
+        "grad_norm": 1.6292518377304077,
+        "learning_rate": 0.00044526122341822003,
+        "epoch": 1.66
+    },
+    {
+        "loss": 7.3463,
+        "grad_norm": 1.0507898330688477,
+        "learning_rate": 0.00044438610308917474,
+        "epoch": 1.67
+    },
+    {
+        "loss": 7.693,
+        "grad_norm": 1.332474708557129,
+        "learning_rate": 0.0004435109827601295,
+        "epoch": 1.67
+    },
+    {
+        "loss": 7.4542,
+        "grad_norm": 1.3393101692199707,
+        "learning_rate": 0.00044263586243108426,
+        "epoch": 1.67
+    },
+    {
+        "loss": 7.4236,
+        "grad_norm": 1.4949504137039185,
+        "learning_rate": 0.000441760742102039,
+        "epoch": 1.67
+    },
+    {
+        "loss": 7.4087,
+        "grad_norm": 1.3824454545974731,
+        "learning_rate": 0.0004408856217729938,
+        "epoch": 1.68
+    },
+    {
+        "loss": 7.3,
+        "grad_norm": 1.3991942405700684,
+        "learning_rate": 0.00044001050144394855,
+        "epoch": 1.68
+    },
+    {
+        "loss": 7.1648,
+        "grad_norm": 1.3270092010498047,
+        "learning_rate": 0.0004391353811149033,
+        "epoch": 1.68
+    },
+    {
+        "loss": 7.753,
+        "grad_norm": 1.1912864446640015,
+        "learning_rate": 0.00043826026078585807,
+        "epoch": 1.69
+    },
+    {
+        "loss": 7.6531,
+        "grad_norm": 1.2112165689468384,
+        "learning_rate": 0.00043738514045681283,
+        "epoch": 1.69
+    },
+    {
+        "loss": 8.0168,
+        "grad_norm": 1.0204828977584839,
+        "learning_rate": 0.0004365100201277676,
+        "epoch": 1.69
+    },
+    {
+        "loss": 7.8334,
+        "grad_norm": 1.8065035343170166,
+        "learning_rate": 0.00043563489979872236,
+        "epoch": 1.69
+    },
+    {
+        "loss": 7.9395,
+        "grad_norm": 1.1826367378234863,
+        "learning_rate": 0.0004347597794696771,
+        "epoch": 1.7
+    },
+    {
+        "loss": 8.0071,
+        "grad_norm": 0.9689782857894897,
+        "learning_rate": 0.00043388465914063183,
+        "epoch": 1.7
+    },
+    {
+        "loss": 7.5284,
+        "grad_norm": 0.9889323115348816,
+        "learning_rate": 0.0004330095388115866,
+        "epoch": 1.7
+    },
+    {
+        "loss": 7.7318,
+        "grad_norm": 1.4257516860961914,
+        "learning_rate": 0.00043213441848254135,
+        "epoch": 1.7
+    },
+    {
+        "loss": 7.6343,
+        "grad_norm": 1.623134970664978,
+        "learning_rate": 0.0004312592981534961,
+        "epoch": 1.71
+    },
+    {
+        "loss": 7.9645,
+        "grad_norm": 1.2686361074447632,
+        "learning_rate": 0.0004303841778244509,
+        "epoch": 1.71
+    },
+    {
+        "loss": 7.5339,
+        "grad_norm": 1.5115247964859009,
+        "learning_rate": 0.00042950905749540564,
+        "epoch": 1.71
+    },
+    {
+        "loss": 7.7401,
+        "grad_norm": 1.285506010055542,
+        "learning_rate": 0.0004286339371663604,
+        "epoch": 1.71
+    },
+    {
+        "loss": 7.6018,
+        "grad_norm": 1.4150651693344116,
+        "learning_rate": 0.00042775881683731516,
+        "epoch": 1.72
+    },
+    {
+        "loss": 7.4015,
+        "grad_norm": 1.485231637954712,
+        "learning_rate": 0.0004268836965082699,
+        "epoch": 1.72
+    },
+    {
+        "loss": 8.4429,
+        "grad_norm": 2.1629021167755127,
+        "learning_rate": 0.0004260085761792247,
+        "epoch": 1.72
+    },
+    {
+        "loss": 7.8298,
+        "grad_norm": 1.1586624383926392,
+        "learning_rate": 0.00042513345585017945,
+        "epoch": 1.72
+    },
+    {
+        "loss": 7.8121,
+        "grad_norm": 1.0134670734405518,
+        "learning_rate": 0.0004242583355211342,
+        "epoch": 1.73
+    },
+    {
+        "loss": 7.8337,
+        "grad_norm": 1.257633090019226,
+        "learning_rate": 0.00042338321519208897,
+        "epoch": 1.73
+    },
+    {
+        "loss": 7.6701,
+        "grad_norm": 1.212266445159912,
+        "learning_rate": 0.0004225080948630437,
+        "epoch": 1.73
+    },
+    {
+        "loss": 7.706,
+        "grad_norm": 1.2191237211227417,
+        "learning_rate": 0.00042163297453399844,
+        "epoch": 1.74
+    },
+    {
+        "loss": 7.4639,
+        "grad_norm": 1.476140022277832,
+        "learning_rate": 0.0004207578542049532,
+        "epoch": 1.74
+    },
+    {
+        "loss": 7.8126,
+        "grad_norm": 1.0655369758605957,
+        "learning_rate": 0.0004198827338759079,
+        "epoch": 1.74
+    },
+    {
+        "loss": 7.4091,
+        "grad_norm": 1.3340696096420288,
+        "learning_rate": 0.00041900761354686267,
+        "epoch": 1.74
+    },
+    {
+        "loss": 7.5701,
+        "grad_norm": 1.3290128707885742,
+        "learning_rate": 0.00041813249321781743,
+        "epoch": 1.75
+    },
+    {
+        "loss": 7.5513,
+        "grad_norm": 1.1993497610092163,
+        "learning_rate": 0.0004172573728887722,
+        "epoch": 1.75
+    },
+    {
+        "loss": 7.5115,
+        "grad_norm": 0.9953559041023254,
+        "learning_rate": 0.00041638225255972696,
+        "epoch": 1.75
+    },
+    {
+        "loss": 8.0513,
+        "grad_norm": 1.1929738521575928,
+        "learning_rate": 0.0004155071322306817,
+        "epoch": 1.75
+    },
+    {
+        "loss": 7.4431,
+        "grad_norm": 1.0211223363876343,
+        "learning_rate": 0.0004146320119016365,
+        "epoch": 1.76
+    },
+    {
+        "loss": 7.4024,
+        "grad_norm": 1.0484708547592163,
+        "learning_rate": 0.00041375689157259124,
+        "epoch": 1.76
+    },
+    {
+        "loss": 7.4321,
+        "grad_norm": 1.2012499570846558,
+        "learning_rate": 0.000412881771243546,
+        "epoch": 1.76
+    },
+    {
+        "loss": 7.2608,
+        "grad_norm": 0.9850478768348694,
+        "learning_rate": 0.0004120066509145007,
+        "epoch": 1.76
+    },
+    {
+        "loss": 7.4744,
+        "grad_norm": 1.1142171621322632,
+        "learning_rate": 0.00041113153058545547,
+        "epoch": 1.77
+    },
+    {
+        "loss": 7.8258,
+        "grad_norm": 1.0107368230819702,
+        "learning_rate": 0.00041025641025641023,
+        "epoch": 1.77
+    },
+    {
+        "loss": 8.0338,
+        "grad_norm": 1.3827756643295288,
+        "learning_rate": 0.000409381289927365,
+        "epoch": 1.77
+    },
+    {
+        "loss": 7.2029,
+        "grad_norm": 1.056078553199768,
+        "learning_rate": 0.00040850616959831976,
+        "epoch": 1.77
+    },
+    {
+        "loss": 7.9763,
+        "grad_norm": 1.3796826601028442,
+        "learning_rate": 0.0004076310492692745,
+        "epoch": 1.78
+    },
+    {
+        "loss": 7.39,
+        "grad_norm": 1.5586506128311157,
+        "learning_rate": 0.0004067559289402293,
+        "epoch": 1.78
+    },
+    {
+        "loss": 7.7479,
+        "grad_norm": 1.3467471599578857,
+        "learning_rate": 0.00040588080861118404,
+        "epoch": 1.78
+    },
+    {
+        "loss": 7.5281,
+        "grad_norm": 1.5824648141860962,
+        "learning_rate": 0.0004050056882821388,
+        "epoch": 1.78
+    },
+    {
+        "loss": 7.4095,
+        "grad_norm": 1.5600448846817017,
+        "learning_rate": 0.00040413056795309357,
+        "epoch": 1.79
+    },
+    {
+        "loss": 7.6296,
+        "grad_norm": 1.4003773927688599,
+        "learning_rate": 0.00040325544762404833,
+        "epoch": 1.79
+    },
+    {
+        "loss": 7.2299,
+        "grad_norm": 1.1784484386444092,
+        "learning_rate": 0.0004023803272950031,
+        "epoch": 1.79
+    },
+    {
+        "loss": 7.2215,
+        "grad_norm": 1.0865730047225952,
+        "learning_rate": 0.0004015052069659578,
+        "epoch": 1.8
+    },
+    {
+        "loss": 7.619,
+        "grad_norm": 1.3708497285842896,
+        "learning_rate": 0.00040063008663691256,
+        "epoch": 1.8
+    },
+    {
+        "loss": 7.6305,
+        "grad_norm": 1.3728278875350952,
+        "learning_rate": 0.0003997549663078673,
+        "epoch": 1.8
+    },
+    {
+        "loss": 7.4218,
+        "grad_norm": 1.385901689529419,
+        "learning_rate": 0.0003988798459788221,
+        "epoch": 1.8
+    },
+    {
+        "loss": 7.7959,
+        "grad_norm": 1.5370672941207886,
+        "learning_rate": 0.00039800472564977685,
+        "epoch": 1.81
+    },
+    {
+        "loss": 7.8249,
+        "grad_norm": 1.039469838142395,
+        "learning_rate": 0.0003971296053207316,
+        "epoch": 1.81
+    },
+    {
+        "loss": 7.4311,
+        "grad_norm": 1.4947952032089233,
+        "learning_rate": 0.00039625448499168637,
+        "epoch": 1.81
+    },
+    {
+        "loss": 7.7797,
+        "grad_norm": 1.2262136936187744,
+        "learning_rate": 0.00039546687669554567,
+        "epoch": 1.81
+    },
+    {
+        "loss": 7.8844,
+        "grad_norm": 1.5757509469985962,
+        "learning_rate": 0.00039459175636650043,
+        "epoch": 1.82
+    },
+    {
+        "loss": 7.8737,
+        "grad_norm": 1.2183258533477783,
+        "learning_rate": 0.0003937166360374552,
+        "epoch": 1.82
+    },
+    {
+        "loss": 7.3515,
+        "grad_norm": 1.3697617053985596,
+        "learning_rate": 0.00039284151570840995,
+        "epoch": 1.82
+    },
+    {
+        "loss": 7.1169,
+        "grad_norm": 1.3007692098617554,
+        "learning_rate": 0.00039196639537936466,
+        "epoch": 1.82
+    },
+    {
+        "loss": 7.2926,
+        "grad_norm": 1.3538720607757568,
+        "learning_rate": 0.0003910912750503194,
+        "epoch": 1.83
+    },
+    {
+        "loss": 7.8445,
+        "grad_norm": 1.4245976209640503,
+        "learning_rate": 0.0003902161547212742,
+        "epoch": 1.83
+    },
+    {
+        "loss": 7.456,
+        "grad_norm": 1.323899269104004,
+        "learning_rate": 0.00038934103439222894,
+        "epoch": 1.83
+    },
+    {
+        "loss": 7.6163,
+        "grad_norm": 1.2635420560836792,
+        "learning_rate": 0.0003884659140631837,
+        "epoch": 1.83
+    },
+    {
+        "loss": 7.5885,
+        "grad_norm": 1.4714936017990112,
+        "learning_rate": 0.00038759079373413847,
+        "epoch": 1.84
+    },
+    {
+        "loss": 7.8382,
+        "grad_norm": 1.1696442365646362,
+        "learning_rate": 0.00038671567340509323,
+        "epoch": 1.84
+    },
+    {
+        "loss": 7.4885,
+        "grad_norm": 1.3797491788864136,
+        "learning_rate": 0.000385840553076048,
+        "epoch": 1.84
+    },
+    {
+        "loss": 7.4614,
+        "grad_norm": 1.0410481691360474,
+        "learning_rate": 0.00038496543274700275,
+        "epoch": 1.85
+    },
+    {
+        "loss": 6.9584,
+        "grad_norm": 1.7356559038162231,
+        "learning_rate": 0.0003840903124179575,
+        "epoch": 1.85
+    },
+    {
+        "loss": 7.8161,
+        "grad_norm": 1.326489806175232,
+        "learning_rate": 0.0003832151920889123,
+        "epoch": 1.85
+    },
+    {
+        "loss": 7.6985,
+        "grad_norm": 1.3822075128555298,
+        "learning_rate": 0.00038234007175986704,
+        "epoch": 1.85
+    },
+    {
+        "loss": 7.9532,
+        "grad_norm": 1.2612171173095703,
+        "learning_rate": 0.00038146495143082175,
+        "epoch": 1.86
+    },
+    {
+        "loss": 7.3309,
+        "grad_norm": 1.8743207454681396,
+        "learning_rate": 0.0003805898311017765,
+        "epoch": 1.86
+    },
+    {
+        "loss": 7.8573,
+        "grad_norm": 1.515641212463379,
+        "learning_rate": 0.0003797147107727312,
+        "epoch": 1.86
+    },
+    {
+        "loss": 8.0815,
+        "grad_norm": 1.970818281173706,
+        "learning_rate": 0.000378839590443686,
+        "epoch": 1.86
+    },
+    {
+        "loss": 7.7197,
+        "grad_norm": 1.6418136358261108,
+        "learning_rate": 0.00037796447011464074,
+        "epoch": 1.87
+    },
+    {
+        "loss": 7.6527,
+        "grad_norm": 1.3693944215774536,
+        "learning_rate": 0.0003770893497855955,
+        "epoch": 1.87
+    },
+    {
+        "loss": 7.7717,
+        "grad_norm": 1.311493992805481,
+        "learning_rate": 0.00037621422945655026,
+        "epoch": 1.87
+    },
+    {
+        "loss": 8.0735,
+        "grad_norm": 1.593992829322815,
+        "learning_rate": 0.000375339109127505,
+        "epoch": 1.87
+    },
+    {
+        "loss": 7.4285,
+        "grad_norm": 1.212729573249817,
+        "learning_rate": 0.0003744639887984598,
+        "epoch": 1.88
+    },
+    {
+        "loss": 7.7873,
+        "grad_norm": 1.1326895952224731,
+        "learning_rate": 0.00037358886846941455,
+        "epoch": 1.88
+    },
+    {
+        "loss": 7.3515,
+        "grad_norm": 1.3937299251556396,
+        "learning_rate": 0.0003727137481403693,
+        "epoch": 1.88
+    },
+    {
+        "loss": 7.353,
+        "grad_norm": 1.5152568817138672,
+        "learning_rate": 0.00037183862781132407,
+        "epoch": 1.88
+    },
+    {
+        "loss": 7.8015,
+        "grad_norm": 1.207973599433899,
+        "learning_rate": 0.0003709635074822788,
+        "epoch": 1.89
+    },
+    {
+        "loss": 7.6713,
+        "grad_norm": 1.003139615058899,
+        "learning_rate": 0.00037008838715323354,
+        "epoch": 1.89
+    },
+    {
+        "loss": 7.9247,
+        "grad_norm": 1.1870025396347046,
+        "learning_rate": 0.0003692132668241883,
+        "epoch": 1.89
+    },
+    {
+        "loss": 7.4496,
+        "grad_norm": 1.237275242805481,
+        "learning_rate": 0.00036833814649514307,
+        "epoch": 1.9
+    },
+    {
+        "loss": 7.2638,
+        "grad_norm": 1.7287304401397705,
+        "learning_rate": 0.00036746302616609783,
+        "epoch": 1.9
+    },
+    {
+        "loss": 7.7464,
+        "grad_norm": 1.5875813961029053,
+        "learning_rate": 0.0003665879058370526,
+        "epoch": 1.9
+    },
+    {
+        "loss": 7.683,
+        "grad_norm": 1.7219480276107788,
+        "learning_rate": 0.00036571278550800735,
+        "epoch": 1.9
+    },
+    {
+        "loss": 7.6059,
+        "grad_norm": 1.3815206289291382,
+        "learning_rate": 0.0003648376651789621,
+        "epoch": 1.91
+    },
+    {
+        "loss": 7.3258,
+        "grad_norm": 1.1902978420257568,
+        "learning_rate": 0.0003639625448499169,
+        "epoch": 1.91
+    },
+    {
+        "loss": 7.4436,
+        "grad_norm": 1.6532816886901855,
+        "learning_rate": 0.00036308742452087164,
+        "epoch": 1.91
+    },
+    {
+        "loss": 7.438,
+        "grad_norm": 1.1358212232589722,
+        "learning_rate": 0.0003622123041918264,
+        "epoch": 1.91
+    },
+    {
+        "loss": 7.9777,
+        "grad_norm": 1.3459230661392212,
+        "learning_rate": 0.00036133718386278116,
+        "epoch": 1.92
+    },
+    {
+        "loss": 7.9087,
+        "grad_norm": 1.0352368354797363,
+        "learning_rate": 0.0003604620635337359,
+        "epoch": 1.92
+    },
+    {
+        "loss": 7.5855,
+        "grad_norm": 1.2582918405532837,
+        "learning_rate": 0.00035958694320469063,
+        "epoch": 1.92
+    },
+    {
+        "loss": 7.4576,
+        "grad_norm": 1.1787996292114258,
+        "learning_rate": 0.0003587118228756454,
+        "epoch": 1.92
+    },
+    {
+        "loss": 7.2572,
+        "grad_norm": 1.2917609214782715,
+        "learning_rate": 0.00035783670254660015,
+        "epoch": 1.93
+    },
+    {
+        "loss": 7.6433,
+        "grad_norm": 1.1689330339431763,
+        "learning_rate": 0.0003569615822175549,
+        "epoch": 1.93
+    },
+    {
+        "loss": 7.6579,
+        "grad_norm": 1.2844352722167969,
+        "learning_rate": 0.0003560864618885097,
+        "epoch": 1.93
+    },
+    {
+        "loss": 7.5178,
+        "grad_norm": 1.498838186264038,
+        "learning_rate": 0.00035521134155946444,
+        "epoch": 1.93
+    },
+    {
+        "loss": 7.0155,
+        "grad_norm": 1.3718552589416504,
+        "learning_rate": 0.0003543362212304192,
+        "epoch": 1.94
+    },
+    {
+        "loss": 7.7558,
+        "grad_norm": 1.2343835830688477,
+        "learning_rate": 0.00035346110090137396,
+        "epoch": 1.94
+    },
+    {
+        "loss": 7.4386,
+        "grad_norm": 1.307979702949524,
+        "learning_rate": 0.0003525859805723287,
+        "epoch": 1.94
+    },
+    {
+        "loss": 7.4287,
+        "grad_norm": 1.46335768699646,
+        "learning_rate": 0.0003517108602432835,
+        "epoch": 1.95
+    },
+    {
+        "loss": 7.0541,
+        "grad_norm": 1.4892301559448242,
+        "learning_rate": 0.00035083573991423825,
+        "epoch": 1.95
+    },
+    {
+        "loss": 7.6458,
+        "grad_norm": 1.3297821283340454,
+        "learning_rate": 0.000349960619585193,
+        "epoch": 1.95
+    },
+    {
+        "loss": 7.3704,
+        "grad_norm": 1.9190036058425903,
+        "learning_rate": 0.0003490854992561477,
+        "epoch": 1.95
+    },
+    {
+        "loss": 7.9292,
+        "grad_norm": 1.1013009548187256,
+        "learning_rate": 0.0003482103789271025,
+        "epoch": 1.96
+    },
+    {
+        "loss": 7.8039,
+        "grad_norm": 1.284121036529541,
+        "learning_rate": 0.00034733525859805724,
+        "epoch": 1.96
+    },
+    {
+        "loss": 7.7188,
+        "grad_norm": 1.118995189666748,
+        "learning_rate": 0.000346460138269012,
+        "epoch": 1.96
+    },
+    {
+        "loss": 7.3617,
+        "grad_norm": 1.5446746349334717,
+        "learning_rate": 0.00034558501793996676,
+        "epoch": 1.96
+    },
+    {
+        "loss": 7.5614,
+        "grad_norm": 1.254835844039917,
+        "learning_rate": 0.0003447098976109215,
+        "epoch": 1.97
+    },
+    {
+        "loss": 7.9923,
+        "grad_norm": 2.215224266052246,
+        "learning_rate": 0.0003438347772818763,
+        "epoch": 1.97
+    },
+    {
+        "loss": 7.6609,
+        "grad_norm": 1.2917975187301636,
+        "learning_rate": 0.00034295965695283105,
+        "epoch": 1.97
+    },
+    {
+        "loss": 6.9695,
+        "grad_norm": 1.3251945972442627,
+        "learning_rate": 0.0003420845366237858,
+        "epoch": 1.97
+    },
+    {
+        "loss": 7.4109,
+        "grad_norm": 1.5397628545761108,
+        "learning_rate": 0.0003412094162947406,
+        "epoch": 1.98
+    },
+    {
+        "loss": 7.3063,
+        "grad_norm": 1.1789202690124512,
+        "learning_rate": 0.00034033429596569534,
+        "epoch": 1.98
+    },
+    {
+        "loss": 7.8137,
+        "grad_norm": 1.6068191528320312,
+        "learning_rate": 0.00033945917563665004,
+        "epoch": 1.98
+    },
+    {
+        "loss": 7.5466,
+        "grad_norm": 1.2397950887680054,
+        "learning_rate": 0.00033858405530760475,
+        "epoch": 1.98
+    },
+    {
+        "loss": 7.9522,
+        "grad_norm": 1.5175119638442993,
+        "learning_rate": 0.0003377089349785595,
+        "epoch": 1.99
+    },
+    {
+        "loss": 7.6781,
+        "grad_norm": 1.315258502960205,
+        "learning_rate": 0.0003368338146495143,
+        "epoch": 1.99
+    },
+    {
+        "loss": 7.7292,
+        "grad_norm": 2.664515256881714,
+        "learning_rate": 0.00033595869432046904,
+        "epoch": 1.99
+    },
+    {
+        "loss": 8.1965,
+        "grad_norm": 1.405129313468933,
+        "learning_rate": 0.0003350835739914238,
+        "epoch": 1.99
+    },
+    {
+        "loss": 7.4133,
+        "grad_norm": 1.0774602890014648,
+        "learning_rate": 0.00033420845366237856,
+        "epoch": 2.0
+    },
+    {
+        "loss": 8.1777,
+        "grad_norm": 1.75553560256958,
+        "learning_rate": 0.0003333333333333333,
+        "epoch": 2.0
+    },
+    {
+        "loss": 7.5693,
+        "grad_norm": 1.857081651687622,
+        "learning_rate": 0.0003324582130042881,
+        "epoch": 2.0
+    },
+    {
+        "loss": 7.4888,
+        "grad_norm": 1.0721529722213745,
+        "learning_rate": 0.00033158309267524285,
+        "epoch": 2.01
+    },
+    {
+        "loss": 7.1311,
+        "grad_norm": 1.0766797065734863,
+        "learning_rate": 0.0003307079723461976,
+        "epoch": 2.01
+    },
+    {
+        "loss": 7.5107,
+        "grad_norm": 1.4615150690078735,
+        "learning_rate": 0.00032983285201715237,
+        "epoch": 2.01
+    },
+    {
+        "loss": 7.5258,
+        "grad_norm": 1.4252068996429443,
+        "learning_rate": 0.00032895773168810713,
+        "epoch": 2.01
+    },
+    {
+        "loss": 7.6049,
+        "grad_norm": 1.2926585674285889,
+        "learning_rate": 0.0003280826113590619,
+        "epoch": 2.02
+    },
+    {
+        "loss": 7.2436,
+        "grad_norm": 1.6630724668502808,
+        "learning_rate": 0.0003272074910300166,
+        "epoch": 2.02
+    },
+    {
+        "loss": 6.951,
+        "grad_norm": 1.2705895900726318,
+        "learning_rate": 0.00032633237070097136,
+        "epoch": 2.02
+    },
+    {
+        "loss": 7.4782,
+        "grad_norm": 1.6801918745040894,
+        "learning_rate": 0.0003254572503719261,
+        "epoch": 2.02
+    },
+    {
+        "loss": 7.7247,
+        "grad_norm": 1.2789455652236938,
+        "learning_rate": 0.0003245821300428809,
+        "epoch": 2.03
+    },
+    {
+        "loss": 7.65,
+        "grad_norm": 1.0772324800491333,
+        "learning_rate": 0.00032370700971383565,
+        "epoch": 2.03
+    },
+    {
+        "loss": 7.3484,
+        "grad_norm": 1.218855857849121,
+        "learning_rate": 0.0003228318893847904,
+        "epoch": 2.03
+    },
+    {
+        "loss": 7.7201,
+        "grad_norm": 1.7484831809997559,
+        "learning_rate": 0.00032195676905574517,
+        "epoch": 2.03
+    },
+    {
+        "loss": 7.606,
+        "grad_norm": 1.4081809520721436,
+        "learning_rate": 0.00032108164872669993,
+        "epoch": 2.04
+    },
+    {
+        "loss": 7.4735,
+        "grad_norm": 1.2214211225509644,
+        "learning_rate": 0.0003202065283976547,
+        "epoch": 2.04
+    },
+    {
+        "loss": 7.3052,
+        "grad_norm": 2.243197441101074,
+        "learning_rate": 0.00031933140806860946,
+        "epoch": 2.04
+    },
+    {
+        "loss": 7.2611,
+        "grad_norm": 1.0560696125030518,
+        "learning_rate": 0.0003184562877395642,
+        "epoch": 2.04
+    },
+    {
+        "loss": 7.3347,
+        "grad_norm": 1.3903985023498535,
+        "learning_rate": 0.000317581167410519,
+        "epoch": 2.05
+    },
+    {
+        "loss": 7.4106,
+        "grad_norm": 1.285888910293579,
+        "learning_rate": 0.0003167060470814737,
+        "epoch": 2.05
+    },
+    {
+        "loss": 7.3237,
+        "grad_norm": 1.6455745697021484,
+        "learning_rate": 0.00031583092675242845,
+        "epoch": 2.05
+    },
+    {
+        "loss": 7.4445,
+        "grad_norm": 1.3552714586257935,
+        "learning_rate": 0.0003149558064233832,
+        "epoch": 2.06
+    },
+    {
+        "loss": 7.3175,
+        "grad_norm": 1.4250375032424927,
+        "learning_rate": 0.000314080686094338,
+        "epoch": 2.06
+    },
+    {
+        "loss": 7.5334,
+        "grad_norm": 1.8445017337799072,
+        "learning_rate": 0.00031320556576529274,
+        "epoch": 2.06
+    },
+    {
+        "loss": 7.7627,
+        "grad_norm": 1.1116868257522583,
+        "learning_rate": 0.0003123304454362475,
+        "epoch": 2.06
+    },
+    {
+        "loss": 7.5347,
+        "grad_norm": 1.1636768579483032,
+        "learning_rate": 0.00031145532510720226,
+        "epoch": 2.07
+    },
+    {
+        "loss": 7.6581,
+        "grad_norm": 1.4612860679626465,
+        "learning_rate": 0.000310580204778157,
+        "epoch": 2.07
+    },
+    {
+        "loss": 7.6164,
+        "grad_norm": 1.4403191804885864,
+        "learning_rate": 0.0003097050844491118,
+        "epoch": 2.07
+    },
+    {
+        "loss": 7.3776,
+        "grad_norm": 1.366955041885376,
+        "learning_rate": 0.00030882996412006655,
+        "epoch": 2.07
+    },
+    {
+        "loss": 7.556,
+        "grad_norm": 1.4476971626281738,
+        "learning_rate": 0.0003079548437910213,
+        "epoch": 2.08
+    },
+    {
+        "loss": 7.6019,
+        "grad_norm": 1.4753084182739258,
+        "learning_rate": 0.00030707972346197607,
+        "epoch": 2.08
+    },
+    {
+        "loss": 7.8493,
+        "grad_norm": 1.2335758209228516,
+        "learning_rate": 0.00030620460313293083,
+        "epoch": 2.08
+    },
+    {
+        "loss": 7.9252,
+        "grad_norm": 1.3958989381790161,
+        "learning_rate": 0.00030532948280388554,
+        "epoch": 2.08
+    },
+    {
+        "loss": 7.2945,
+        "grad_norm": 1.4621672630310059,
+        "learning_rate": 0.0003044543624748403,
+        "epoch": 2.09
+    },
+    {
+        "loss": 7.3977,
+        "grad_norm": 1.428195834159851,
+        "learning_rate": 0.00030357924214579506,
+        "epoch": 2.09
+    },
+    {
+        "loss": 7.74,
+        "grad_norm": 1.363600492477417,
+        "learning_rate": 0.0003027041218167498,
+        "epoch": 2.09
+    },
+    {
+        "loss": 7.4894,
+        "grad_norm": 1.2117736339569092,
+        "learning_rate": 0.0003018290014877046,
+        "epoch": 2.09
+    },
+    {
+        "loss": 7.5678,
+        "grad_norm": 1.9844530820846558,
+        "learning_rate": 0.00030095388115865935,
+        "epoch": 2.1
+    },
+    {
+        "loss": 7.6681,
+        "grad_norm": 1.3558523654937744,
+        "learning_rate": 0.0003000787608296141,
+        "epoch": 2.1
+    },
+    {
+        "loss": 7.9793,
+        "grad_norm": 1.3802049160003662,
+        "learning_rate": 0.0002992036405005688,
+        "epoch": 2.1
+    },
+    {
+        "loss": 8.1848,
+        "grad_norm": 1.845702886581421,
+        "learning_rate": 0.0002983285201715236,
+        "epoch": 2.11
+    },
+    {
+        "loss": 7.2184,
+        "grad_norm": 1.4479707479476929,
+        "learning_rate": 0.00029745339984247834,
+        "epoch": 2.11
+    },
+    {
+        "loss": 7.4373,
+        "grad_norm": 1.9233028888702393,
+        "learning_rate": 0.0002965782795134331,
+        "epoch": 2.11
+    },
+    {
+        "loss": 7.2478,
+        "grad_norm": 1.3621513843536377,
+        "learning_rate": 0.00029570315918438786,
+        "epoch": 2.11
+    },
+    {
+        "loss": 7.5867,
+        "grad_norm": 1.449763298034668,
+        "learning_rate": 0.00029482803885534257,
+        "epoch": 2.12
+    },
+    {
+        "loss": 7.2909,
+        "grad_norm": 1.543834924697876,
+        "learning_rate": 0.00029395291852629733,
+        "epoch": 2.12
+    },
+    {
+        "loss": 7.5481,
+        "grad_norm": 1.2582162618637085,
+        "learning_rate": 0.0002930777981972521,
+        "epoch": 2.12
+    },
+    {
+        "loss": 7.2092,
+        "grad_norm": 1.25532865524292,
+        "learning_rate": 0.00029220267786820686,
+        "epoch": 2.12
+    },
+    {
+        "loss": 7.5117,
+        "grad_norm": 1.4368300437927246,
+        "learning_rate": 0.0002913275575391616,
+        "epoch": 2.13
+    },
+    {
+        "loss": 7.8661,
+        "grad_norm": 1.4054632186889648,
+        "learning_rate": 0.0002904524372101164,
+        "epoch": 2.13
+    },
+    {
+        "loss": 7.7641,
+        "grad_norm": 1.4426825046539307,
+        "learning_rate": 0.00028957731688107114,
+        "epoch": 2.13
+    },
+    {
+        "loss": 6.9808,
+        "grad_norm": 1.6069836616516113,
+        "learning_rate": 0.0002887021965520259,
+        "epoch": 2.13
+    },
+    {
+        "loss": 8.0412,
+        "grad_norm": 1.603289246559143,
+        "learning_rate": 0.00028782707622298067,
+        "epoch": 2.14
+    },
+    {
+        "loss": 7.7541,
+        "grad_norm": 1.2069703340530396,
+        "learning_rate": 0.00028695195589393543,
+        "epoch": 2.14
+    },
+    {
+        "loss": 7.5413,
+        "grad_norm": 1.2976186275482178,
+        "learning_rate": 0.0002860768355648902,
+        "epoch": 2.14
+    },
+    {
+        "loss": 7.6833,
+        "grad_norm": 1.4646226167678833,
+        "learning_rate": 0.00028520171523584495,
+        "epoch": 2.14
+    },
+    {
+        "loss": 7.3603,
+        "grad_norm": 1.3783011436462402,
+        "learning_rate": 0.00028432659490679966,
+        "epoch": 2.15
+    },
+    {
+        "loss": 7.1131,
+        "grad_norm": 1.1677837371826172,
+        "learning_rate": 0.0002834514745777544,
+        "epoch": 2.15
+    },
+    {
+        "loss": 7.8353,
+        "grad_norm": 1.5966696739196777,
+        "learning_rate": 0.0002825763542487092,
+        "epoch": 2.15
+    },
+    {
+        "loss": 7.651,
+        "grad_norm": 1.3074275255203247,
+        "learning_rate": 0.00028170123391966394,
+        "epoch": 2.16
+    },
+    {
+        "loss": 6.8535,
+        "grad_norm": 1.2238943576812744,
+        "learning_rate": 0.0002808261135906187,
+        "epoch": 2.16
+    },
+    {
+        "loss": 7.1677,
+        "grad_norm": 1.2107079029083252,
+        "learning_rate": 0.00027995099326157347,
+        "epoch": 2.16
+    },
+    {
+        "loss": 7.1232,
+        "grad_norm": 1.482686996459961,
+        "learning_rate": 0.00027907587293252823,
+        "epoch": 2.16
+    },
+    {
+        "loss": 7.6958,
+        "grad_norm": 1.9235337972640991,
+        "learning_rate": 0.000278200752603483,
+        "epoch": 2.17
+    },
+    {
+        "loss": 7.5763,
+        "grad_norm": 1.0629470348358154,
+        "learning_rate": 0.00027732563227443775,
+        "epoch": 2.17
+    },
+    {
+        "loss": 7.417,
+        "grad_norm": 1.4404977560043335,
+        "learning_rate": 0.0002764505119453925,
+        "epoch": 2.17
+    },
+    {
+        "loss": 7.4457,
+        "grad_norm": 1.6266590356826782,
+        "learning_rate": 0.0002755753916163473,
+        "epoch": 2.17
+    },
+    {
+        "loss": 7.6768,
+        "grad_norm": 1.4418647289276123,
+        "learning_rate": 0.00027470027128730204,
+        "epoch": 2.18
+    },
+    {
+        "loss": 7.7301,
+        "grad_norm": 1.7269823551177979,
+        "learning_rate": 0.0002738251509582568,
+        "epoch": 2.18
+    },
+    {
+        "loss": 7.1704,
+        "grad_norm": 1.9527968168258667,
+        "learning_rate": 0.0002729500306292115,
+        "epoch": 2.18
+    },
+    {
+        "loss": 8.0284,
+        "grad_norm": 1.1195765733718872,
+        "learning_rate": 0.00027207491030016627,
+        "epoch": 2.18
+    },
+    {
+        "loss": 7.876,
+        "grad_norm": 1.381032109260559,
+        "learning_rate": 0.00027119978997112103,
+        "epoch": 2.19
+    },
+    {
+        "loss": 7.4609,
+        "grad_norm": 2.2558112144470215,
+        "learning_rate": 0.0002703246696420758,
+        "epoch": 2.19
+    },
+    {
+        "loss": 7.524,
+        "grad_norm": 1.0892398357391357,
+        "learning_rate": 0.00026944954931303056,
+        "epoch": 2.19
+    },
+    {
+        "loss": 7.1756,
+        "grad_norm": 1.432793140411377,
+        "learning_rate": 0.0002685744289839853,
+        "epoch": 2.19
+    },
+    {
+        "loss": 7.4677,
+        "grad_norm": 2.4381473064422607,
+        "learning_rate": 0.0002676993086549401,
+        "epoch": 2.2
+    },
+    {
+        "loss": 7.2004,
+        "grad_norm": 1.0947704315185547,
+        "learning_rate": 0.00026682418832589484,
+        "epoch": 2.2
+    },
+    {
+        "loss": 7.6084,
+        "grad_norm": 1.1396403312683105,
+        "learning_rate": 0.0002659490679968496,
+        "epoch": 2.2
+    },
+    {
+        "loss": 7.4592,
+        "grad_norm": 1.7132469415664673,
+        "learning_rate": 0.00026507394766780437,
+        "epoch": 2.2
+    },
+    {
+        "loss": 7.6666,
+        "grad_norm": 1.507416844367981,
+        "learning_rate": 0.00026419882733875913,
+        "epoch": 2.21
+    },
+    {
+        "loss": 7.9483,
+        "grad_norm": 1.997502326965332,
+        "learning_rate": 0.0002633237070097139,
+        "epoch": 2.21
+    },
+    {
+        "loss": 6.8979,
+        "grad_norm": 1.180274486541748,
+        "learning_rate": 0.0002624485866806686,
+        "epoch": 2.21
+    },
+    {
+        "loss": 7.5387,
+        "grad_norm": 1.4130629301071167,
+        "learning_rate": 0.00026157346635162336,
+        "epoch": 2.22
+    },
+    {
+        "loss": 7.7374,
+        "grad_norm": 1.9466407299041748,
+        "learning_rate": 0.0002606983460225781,
+        "epoch": 2.22
+    },
+    {
+        "loss": 7.2489,
+        "grad_norm": 1.2844946384429932,
+        "learning_rate": 0.00025982322569353283,
+        "epoch": 2.22
+    },
+    {
+        "loss": 7.2583,
+        "grad_norm": 1.4728493690490723,
+        "learning_rate": 0.0002589481053644876,
+        "epoch": 2.22
+    },
+    {
+        "loss": 7.1689,
+        "grad_norm": 1.505767583847046,
+        "learning_rate": 0.00025807298503544235,
+        "epoch": 2.23
+    },
+    {
+        "loss": 7.3824,
+        "grad_norm": 1.164609432220459,
+        "learning_rate": 0.0002571978647063971,
+        "epoch": 2.23
+    },
+    {
+        "loss": 8.208,
+        "grad_norm": 1.3337666988372803,
+        "learning_rate": 0.0002563227443773519,
+        "epoch": 2.23
+    },
+    {
+        "loss": 7.1503,
+        "grad_norm": 1.2840052843093872,
+        "learning_rate": 0.00025544762404830664,
+        "epoch": 2.23
+    },
+    {
+        "loss": 7.7838,
+        "grad_norm": 1.6767994165420532,
+        "learning_rate": 0.0002545725037192614,
+        "epoch": 2.24
+    },
+    {
+        "loss": 7.4818,
+        "grad_norm": 1.2790688276290894,
+        "learning_rate": 0.00025369738339021616,
+        "epoch": 2.24
+    },
+    {
+        "loss": 7.1404,
+        "grad_norm": 1.9306037425994873,
+        "learning_rate": 0.0002528222630611709,
+        "epoch": 2.24
+    },
+    {
+        "loss": 6.9151,
+        "grad_norm": 1.0568101406097412,
+        "learning_rate": 0.00025194714273212563,
+        "epoch": 2.24
+    },
+    {
+        "loss": 7.5813,
+        "grad_norm": 1.8494940996170044,
+        "learning_rate": 0.0002510720224030804,
+        "epoch": 2.25
+    },
+    {
+        "loss": 7.1433,
+        "grad_norm": 1.2321641445159912,
+        "learning_rate": 0.00025019690207403515,
+        "epoch": 2.25
+    },
+    {
+        "loss": 7.0211,
+        "grad_norm": 1.5231260061264038,
+        "learning_rate": 0.0002493217817449899,
+        "epoch": 2.25
+    },
+    {
+        "loss": 7.5108,
+        "grad_norm": 1.6787548065185547,
+        "learning_rate": 0.0002484466614159447,
+        "epoch": 2.25
+    },
+    {
+        "loss": 7.5859,
+        "grad_norm": 1.8862128257751465,
+        "learning_rate": 0.00024757154108689944,
+        "epoch": 2.26
+    },
+    {
+        "loss": 7.0871,
+        "grad_norm": 1.5295615196228027,
+        "learning_rate": 0.0002466964207578542,
+        "epoch": 2.26
+    },
+    {
+        "loss": 7.2151,
+        "grad_norm": 1.6439179182052612,
+        "learning_rate": 0.00024582130042880896,
+        "epoch": 2.26
+    },
+    {
+        "loss": 7.851,
+        "grad_norm": 1.5902001857757568,
+        "learning_rate": 0.0002449461800997637,
+        "epoch": 2.27
+    },
+    {
+        "loss": 7.695,
+        "grad_norm": 1.447240948677063,
+        "learning_rate": 0.00024407105977071846,
+        "epoch": 2.27
+    },
+    {
+        "loss": 7.218,
+        "grad_norm": 1.7448298931121826,
+        "learning_rate": 0.00024319593944167322,
+        "epoch": 2.27
+    },
+    {
+        "loss": 7.4559,
+        "grad_norm": 1.7815390825271606,
+        "learning_rate": 0.00024232081911262798,
+        "epoch": 2.27
+    },
+    {
+        "loss": 7.5519,
+        "grad_norm": 1.746805191040039,
+        "learning_rate": 0.00024144569878358275,
+        "epoch": 2.28
+    },
+    {
+        "loss": 7.4818,
+        "grad_norm": 1.771155834197998,
+        "learning_rate": 0.0002405705784545375,
+        "epoch": 2.28
+    },
+    {
+        "loss": 7.8775,
+        "grad_norm": 1.2886364459991455,
+        "learning_rate": 0.00023969545812549227,
+        "epoch": 2.28
+    },
+    {
+        "loss": 7.0862,
+        "grad_norm": 1.3562748432159424,
+        "learning_rate": 0.00023882033779644703,
+        "epoch": 2.28
+    },
+    {
+        "loss": 7.4458,
+        "grad_norm": 1.5549288988113403,
+        "learning_rate": 0.00023794521746740177,
+        "epoch": 2.29
+    },
+    {
+        "loss": 7.5017,
+        "grad_norm": 1.3231199979782104,
+        "learning_rate": 0.00023707009713835653,
+        "epoch": 2.29
+    },
+    {
+        "loss": 6.9317,
+        "grad_norm": 1.0973995923995972,
+        "learning_rate": 0.0002361949768093113,
+        "epoch": 2.29
+    },
+    {
+        "loss": 7.2512,
+        "grad_norm": 1.161665916442871,
+        "learning_rate": 0.00023531985648026605,
+        "epoch": 2.29
+    },
+    {
+        "loss": 7.3376,
+        "grad_norm": 1.1249802112579346,
+        "learning_rate": 0.0002344447361512208,
+        "epoch": 2.3
+    },
+    {
+        "loss": 7.6856,
+        "grad_norm": 1.4549752473831177,
+        "learning_rate": 0.00023356961582217557,
+        "epoch": 2.3
+    },
+    {
+        "loss": 7.6518,
+        "grad_norm": 1.2443310022354126,
+        "learning_rate": 0.0002326944954931303,
+        "epoch": 2.3
+    },
+    {
+        "loss": 7.9287,
+        "grad_norm": 1.2414274215698242,
+        "learning_rate": 0.00023181937516408507,
+        "epoch": 2.3
+    },
+    {
+        "loss": 7.4844,
+        "grad_norm": 1.250632882118225,
+        "learning_rate": 0.00023094425483503983,
+        "epoch": 2.31
+    },
+    {
+        "loss": 6.9439,
+        "grad_norm": 1.5678353309631348,
+        "learning_rate": 0.0002300691345059946,
+        "epoch": 2.31
+    },
+    {
+        "loss": 7.2214,
+        "grad_norm": 1.2777363061904907,
+        "learning_rate": 0.00022919401417694933,
+        "epoch": 2.31
+    },
+    {
+        "loss": 7.6909,
+        "grad_norm": 1.1702243089675903,
+        "learning_rate": 0.0002283188938479041,
+        "epoch": 2.32
+    },
+    {
+        "loss": 7.843,
+        "grad_norm": 1.1647387742996216,
+        "learning_rate": 0.00022744377351885883,
+        "epoch": 2.32
+    },
+    {
+        "loss": 7.5598,
+        "grad_norm": 1.5888360738754272,
+        "learning_rate": 0.0002265686531898136,
+        "epoch": 2.32
+    },
+    {
+        "loss": 7.4084,
+        "grad_norm": 1.2132010459899902,
+        "learning_rate": 0.00022569353286076835,
+        "epoch": 2.32
+    },
+    {
+        "loss": 8.0077,
+        "grad_norm": 1.3676106929779053,
+        "learning_rate": 0.0002248184125317231,
+        "epoch": 2.33
+    },
+    {
+        "loss": 7.4475,
+        "grad_norm": 1.4785172939300537,
+        "learning_rate": 0.00022394329220267787,
+        "epoch": 2.33
+    },
+    {
+        "loss": 7.4934,
+        "grad_norm": 1.6854803562164307,
+        "learning_rate": 0.00022306817187363264,
+        "epoch": 2.33
+    },
+    {
+        "loss": 7.5371,
+        "grad_norm": 1.3336540460586548,
+        "learning_rate": 0.00022219305154458737,
+        "epoch": 2.33
+    },
+    {
+        "loss": 7.091,
+        "grad_norm": 1.5374839305877686,
+        "learning_rate": 0.00022131793121554213,
+        "epoch": 2.34
+    },
+    {
+        "loss": 7.5715,
+        "grad_norm": 1.259857177734375,
+        "learning_rate": 0.0002204428108864969,
+        "epoch": 2.34
+    },
+    {
+        "loss": 7.5012,
+        "grad_norm": 1.435889482498169,
+        "learning_rate": 0.00021956769055745166,
+        "epoch": 2.34
+    },
+    {
+        "loss": 7.5925,
+        "grad_norm": 1.6067544221878052,
+        "learning_rate": 0.00021869257022840642,
+        "epoch": 2.34
+    },
+    {
+        "loss": 7.2756,
+        "grad_norm": 1.2057377099990845,
+        "learning_rate": 0.00021781744989936118,
+        "epoch": 2.35
+    },
+    {
+        "loss": 7.0737,
+        "grad_norm": 1.0249065160751343,
+        "learning_rate": 0.00021694232957031591,
+        "epoch": 2.35
+    },
+    {
+        "loss": 7.2857,
+        "grad_norm": 1.1336891651153564,
+        "learning_rate": 0.00021606720924127068,
+        "epoch": 2.35
+    },
+    {
+        "loss": 7.0709,
+        "grad_norm": 1.1853156089782715,
+        "learning_rate": 0.00021519208891222544,
+        "epoch": 2.35
+    },
+    {
+        "loss": 6.9118,
+        "grad_norm": 1.4682341814041138,
+        "learning_rate": 0.0002143169685831802,
+        "epoch": 2.36
+    },
+    {
+        "loss": 7.3363,
+        "grad_norm": 1.3039721250534058,
+        "learning_rate": 0.00021344184825413496,
+        "epoch": 2.36
+    },
+    {
+        "loss": 7.2827,
+        "grad_norm": 1.28932785987854,
+        "learning_rate": 0.00021256672792508972,
+        "epoch": 2.36
+    },
+    {
+        "loss": 7.6069,
+        "grad_norm": 1.7343271970748901,
+        "learning_rate": 0.00021169160759604448,
+        "epoch": 2.37
+    },
+    {
+        "loss": 7.3543,
+        "grad_norm": 1.9730132818222046,
+        "learning_rate": 0.00021081648726699922,
+        "epoch": 2.37
+    },
+    {
+        "loss": 7.3351,
+        "grad_norm": 2.070822238922119,
+        "learning_rate": 0.00020994136693795395,
+        "epoch": 2.37
+    },
+    {
+        "loss": 7.3199,
+        "grad_norm": 1.1327873468399048,
+        "learning_rate": 0.00020906624660890872,
+        "epoch": 2.37
+    },
+    {
+        "loss": 7.4058,
+        "grad_norm": 1.3796617984771729,
+        "learning_rate": 0.00020819112627986348,
+        "epoch": 2.38
+    },
+    {
+        "loss": 7.3027,
+        "grad_norm": 1.8397942781448364,
+        "learning_rate": 0.00020731600595081824,
+        "epoch": 2.38
+    },
+    {
+        "loss": 7.6354,
+        "grad_norm": 1.4503923654556274,
+        "learning_rate": 0.0002065283976546775,
+        "epoch": 2.38
+    },
+    {
+        "loss": 7.2284,
+        "grad_norm": 1.550950527191162,
+        "learning_rate": 0.00020565327732563227,
+        "epoch": 2.38
+    },
+    {
+        "loss": 7.3061,
+        "grad_norm": 1.5306216478347778,
+        "learning_rate": 0.00020477815699658703,
+        "epoch": 2.39
+    },
+    {
+        "loss": 7.3337,
+        "grad_norm": 1.269167184829712,
+        "learning_rate": 0.0002039030366675418,
+        "epoch": 2.39
+    },
+    {
+        "loss": 7.7686,
+        "grad_norm": 1.600019931793213,
+        "learning_rate": 0.00020302791633849656,
+        "epoch": 2.39
+    },
+    {
+        "loss": 7.35,
+        "grad_norm": 1.5773662328720093,
+        "learning_rate": 0.0002021527960094513,
+        "epoch": 2.39
+    },
+    {
+        "loss": 7.3691,
+        "grad_norm": 1.547160029411316,
+        "learning_rate": 0.00020127767568040605,
+        "epoch": 2.4
+    },
+    {
+        "loss": 7.4863,
+        "grad_norm": 1.4968856573104858,
+        "learning_rate": 0.00020040255535136081,
+        "epoch": 2.4
+    },
+    {
+        "loss": 7.9482,
+        "grad_norm": 1.2087891101837158,
+        "learning_rate": 0.00019952743502231558,
+        "epoch": 2.4
+    },
+    {
+        "loss": 7.0255,
+        "grad_norm": 1.290597677230835,
+        "learning_rate": 0.00019865231469327034,
+        "epoch": 2.4
+    },
+    {
+        "loss": 7.178,
+        "grad_norm": 1.5743247270584106,
+        "learning_rate": 0.0001977771943642251,
+        "epoch": 2.41
+    },
+    {
+        "loss": 7.6474,
+        "grad_norm": 1.5197412967681885,
+        "learning_rate": 0.00019690207403517984,
+        "epoch": 2.41
+    },
+    {
+        "loss": 7.3527,
+        "grad_norm": 1.4716495275497437,
+        "learning_rate": 0.0001960269537061346,
+        "epoch": 2.41
+    },
+    {
+        "loss": 7.6313,
+        "grad_norm": 1.9746785163879395,
+        "learning_rate": 0.00019515183337708936,
+        "epoch": 2.41
+    },
+    {
+        "loss": 7.6972,
+        "grad_norm": 1.2683417797088623,
+        "learning_rate": 0.00019427671304804412,
+        "epoch": 2.42
+    },
+    {
+        "loss": 7.1378,
+        "grad_norm": 1.1373748779296875,
+        "learning_rate": 0.00019340159271899888,
+        "epoch": 2.42
+    },
+    {
+        "loss": 7.0196,
+        "grad_norm": 1.4191349744796753,
+        "learning_rate": 0.00019252647238995364,
+        "epoch": 2.42
+    },
+    {
+        "loss": 6.9102,
+        "grad_norm": 1.6580002307891846,
+        "learning_rate": 0.00019165135206090838,
+        "epoch": 2.43
+    },
+    {
+        "loss": 7.5105,
+        "grad_norm": 1.2877469062805176,
+        "learning_rate": 0.00019077623173186314,
+        "epoch": 2.43
+    },
+    {
+        "loss": 8.0212,
+        "grad_norm": 1.2933236360549927,
+        "learning_rate": 0.00018990111140281788,
+        "epoch": 2.43
+    },
+    {
+        "loss": 7.2108,
+        "grad_norm": 1.6515684127807617,
+        "learning_rate": 0.00018902599107377264,
+        "epoch": 2.43
+    },
+    {
+        "loss": 7.2944,
+        "grad_norm": 1.443547010421753,
+        "learning_rate": 0.0001881508707447274,
+        "epoch": 2.44
+    },
+    {
+        "loss": 6.9623,
+        "grad_norm": 1.5022013187408447,
+        "learning_rate": 0.00018727575041568216,
+        "epoch": 2.44
+    },
+    {
+        "loss": 7.5751,
+        "grad_norm": 1.639228343963623,
+        "learning_rate": 0.0001864006300866369,
+        "epoch": 2.44
+    },
+    {
+        "loss": 7.6183,
+        "grad_norm": 1.3685816526412964,
+        "learning_rate": 0.00018552550975759166,
+        "epoch": 2.44
+    },
+    {
+        "loss": 7.7862,
+        "grad_norm": 1.4008909463882446,
+        "learning_rate": 0.00018465038942854642,
+        "epoch": 2.45
+    },
+    {
+        "loss": 7.3036,
+        "grad_norm": 1.4068384170532227,
+        "learning_rate": 0.00018377526909950118,
+        "epoch": 2.45
+    },
+    {
+        "loss": 7.3222,
+        "grad_norm": 1.4874199628829956,
+        "learning_rate": 0.00018290014877045594,
+        "epoch": 2.45
+    },
+    {
+        "loss": 7.4538,
+        "grad_norm": 2.161606788635254,
+        "learning_rate": 0.0001820250284414107,
+        "epoch": 2.45
+    },
+    {
+        "loss": 7.099,
+        "grad_norm": 1.4761602878570557,
+        "learning_rate": 0.00018114990811236544,
+        "epoch": 2.46
+    },
+    {
+        "loss": 7.6725,
+        "grad_norm": 1.3598577976226807,
+        "learning_rate": 0.0001802747877833202,
+        "epoch": 2.46
+    },
+    {
+        "loss": 7.4651,
+        "grad_norm": 1.352389931678772,
+        "learning_rate": 0.00017939966745427496,
+        "epoch": 2.46
+    },
+    {
+        "loss": 7.0266,
+        "grad_norm": 1.302270770072937,
+        "learning_rate": 0.00017852454712522973,
+        "epoch": 2.46
+    },
+    {
+        "loss": 7.4879,
+        "grad_norm": 1.2166621685028076,
+        "learning_rate": 0.0001776494267961845,
+        "epoch": 2.47
+    },
+    {
+        "loss": 6.7354,
+        "grad_norm": 1.4442105293273926,
+        "learning_rate": 0.00017677430646713925,
+        "epoch": 2.47
+    },
+    {
+        "loss": 7.1184,
+        "grad_norm": 1.6301904916763306,
+        "learning_rate": 0.000175899186138094,
+        "epoch": 2.47
+    },
+    {
+        "loss": 7.4326,
+        "grad_norm": 1.2478090524673462,
+        "learning_rate": 0.00017502406580904875,
+        "epoch": 2.48
+    },
+    {
+        "loss": 7.6185,
+        "grad_norm": 1.2676613330841064,
+        "learning_rate": 0.0001741489454800035,
+        "epoch": 2.48
+    },
+    {
+        "loss": 7.439,
+        "grad_norm": 1.4324458837509155,
+        "learning_rate": 0.00017327382515095827,
+        "epoch": 2.48
+    },
+    {
+        "loss": 7.7999,
+        "grad_norm": 1.634446382522583,
+        "learning_rate": 0.00017239870482191303,
+        "epoch": 2.48
+    },
+    {
+        "loss": 7.3043,
+        "grad_norm": 1.2877479791641235,
+        "learning_rate": 0.0001715235844928678,
+        "epoch": 2.49
+    },
+    {
+        "loss": 7.054,
+        "grad_norm": 1.7003803253173828,
+        "learning_rate": 0.00017064846416382255,
+        "epoch": 2.49
+    },
+    {
+        "loss": 7.1568,
+        "grad_norm": 1.8888310194015503,
+        "learning_rate": 0.00016977334383477726,
+        "epoch": 2.49
+    },
+    {
+        "loss": 7.3495,
+        "grad_norm": 1.2593083381652832,
+        "learning_rate": 0.00016889822350573202,
+        "epoch": 2.49
+    },
+    {
+        "loss": 7.4716,
+        "grad_norm": 1.4410508871078491,
+        "learning_rate": 0.00016802310317668679,
+        "epoch": 2.5
+    },
+    {
+        "loss": 7.5133,
+        "grad_norm": 1.20904541015625,
+        "learning_rate": 0.00016714798284764155,
+        "epoch": 2.5
+    },
+    {
+        "loss": 7.3222,
+        "grad_norm": 1.4503611326217651,
+        "learning_rate": 0.0001662728625185963,
+        "epoch": 2.5
+    },
+    {
+        "loss": 7.6387,
+        "grad_norm": 1.3705183267593384,
+        "learning_rate": 0.00016539774218955107,
+        "epoch": 2.5
+    },
+    {
+        "loss": 7.0609,
+        "grad_norm": 1.2106906175613403,
+        "learning_rate": 0.0001645226218605058,
+        "epoch": 2.51
+    },
+    {
+        "loss": 7.342,
+        "grad_norm": 1.5564229488372803,
+        "learning_rate": 0.00016364750153146057,
+        "epoch": 2.51
+    },
+    {
+        "loss": 7.8121,
+        "grad_norm": 1.6493812799453735,
+        "learning_rate": 0.00016277238120241533,
+        "epoch": 2.51
+    },
+    {
+        "loss": 7.3909,
+        "grad_norm": 1.9025623798370361,
+        "learning_rate": 0.0001618972608733701,
+        "epoch": 2.51
+    },
+    {
+        "loss": 7.0106,
+        "grad_norm": 1.2934685945510864,
+        "learning_rate": 0.00016102214054432485,
+        "epoch": 2.52
+    },
+    {
+        "loss": 7.5199,
+        "grad_norm": 1.2549662590026855,
+        "learning_rate": 0.00016014702021527962,
+        "epoch": 2.52
+    },
+    {
+        "loss": 7.3509,
+        "grad_norm": 1.2111480236053467,
+        "learning_rate": 0.00015927189988623435,
+        "epoch": 2.52
+    },
+    {
+        "loss": 7.5281,
+        "grad_norm": 2.2498984336853027,
+        "learning_rate": 0.0001583967795571891,
+        "epoch": 2.53
+    },
+    {
+        "loss": 7.5218,
+        "grad_norm": 1.4710973501205444,
+        "learning_rate": 0.00015752165922814387,
+        "epoch": 2.53
+    },
+    {
+        "loss": 7.1575,
+        "grad_norm": 1.4040391445159912,
+        "learning_rate": 0.00015664653889909864,
+        "epoch": 2.53
+    },
+    {
+        "loss": 7.3097,
+        "grad_norm": 2.3657708168029785,
+        "learning_rate": 0.0001557714185700534,
+        "epoch": 2.53
+    },
+    {
+        "loss": 7.3235,
+        "grad_norm": 1.8456711769104004,
+        "learning_rate": 0.00015489629824100816,
+        "epoch": 2.54
+    },
+    {
+        "loss": 7.1772,
+        "grad_norm": 1.3032398223876953,
+        "learning_rate": 0.0001540211779119629,
+        "epoch": 2.54
+    },
+    {
+        "loss": 7.331,
+        "grad_norm": 1.2472988367080688,
+        "learning_rate": 0.00015314605758291766,
+        "epoch": 2.54
+    },
+    {
+        "loss": 6.9758,
+        "grad_norm": 1.1861238479614258,
+        "learning_rate": 0.00015227093725387242,
+        "epoch": 2.54
+    },
+    {
+        "loss": 7.357,
+        "grad_norm": 1.2937425374984741,
+        "learning_rate": 0.00015139581692482718,
+        "epoch": 2.55
+    },
+    {
+        "loss": 7.6132,
+        "grad_norm": 1.5241109132766724,
+        "learning_rate": 0.00015052069659578194,
+        "epoch": 2.55
+    },
+    {
+        "loss": 7.1769,
+        "grad_norm": 1.2426915168762207,
+        "learning_rate": 0.00014964557626673668,
+        "epoch": 2.55
+    },
+    {
+        "loss": 7.2242,
+        "grad_norm": 1.5336363315582275,
+        "learning_rate": 0.0001487704559376914,
+        "epoch": 2.55
+    },
+    {
+        "loss": 8.0839,
+        "grad_norm": 1.6944379806518555,
+        "learning_rate": 0.00014789533560864617,
+        "epoch": 2.56
+    },
+    {
+        "loss": 7.2667,
+        "grad_norm": 1.6602429151535034,
+        "learning_rate": 0.00014702021527960093,
+        "epoch": 2.56
+    },
+    {
+        "loss": 7.4821,
+        "grad_norm": 1.331986665725708,
+        "learning_rate": 0.0001461450949505557,
+        "epoch": 2.56
+    },
+    {
+        "loss": 7.4808,
+        "grad_norm": 1.4923409223556519,
+        "learning_rate": 0.00014526997462151046,
+        "epoch": 2.56
+    },
+    {
+        "loss": 7.3579,
+        "grad_norm": 1.5323739051818848,
+        "learning_rate": 0.00014439485429246522,
+        "epoch": 2.57
+    },
+    {
+        "loss": 7.1833,
+        "grad_norm": 1.0281411409378052,
+        "learning_rate": 0.00014351973396341998,
+        "epoch": 2.57
+    },
+    {
+        "loss": 7.521,
+        "grad_norm": 1.777385950088501,
+        "learning_rate": 0.00014264461363437472,
+        "epoch": 2.57
+    },
+    {
+        "loss": 7.5531,
+        "grad_norm": 1.7528423070907593,
+        "learning_rate": 0.00014176949330532948,
+        "epoch": 2.58
+    },
+    {
+        "loss": 7.3295,
+        "grad_norm": 1.665503740310669,
+        "learning_rate": 0.00014089437297628424,
+        "epoch": 2.58
+    },
+    {
+        "loss": 6.9815,
+        "grad_norm": 1.4323763847351074,
+        "learning_rate": 0.000140019252647239,
+        "epoch": 2.58
+    },
+    {
+        "loss": 7.7957,
+        "grad_norm": 1.2623038291931152,
+        "learning_rate": 0.00013914413231819376,
+        "epoch": 2.58
+    },
+    {
+        "loss": 7.2667,
+        "grad_norm": 1.3770829439163208,
+        "learning_rate": 0.00013826901198914853,
+        "epoch": 2.59
+    },
+    {
+        "loss": 7.2641,
+        "grad_norm": 1.495597243309021,
+        "learning_rate": 0.00013739389166010326,
+        "epoch": 2.59
+    },
+    {
+        "loss": 7.6276,
+        "grad_norm": 1.0396783351898193,
+        "learning_rate": 0.00013651877133105802,
+        "epoch": 2.59
+    },
+    {
+        "loss": 7.4811,
+        "grad_norm": 1.5590603351593018,
+        "learning_rate": 0.00013564365100201278,
+        "epoch": 2.59
+    },
+    {
+        "loss": 6.9941,
+        "grad_norm": 1.266262173652649,
+        "learning_rate": 0.00013476853067296755,
+        "epoch": 2.6
+    },
+    {
+        "loss": 7.0138,
+        "grad_norm": 1.3331608772277832,
+        "learning_rate": 0.0001338934103439223,
+        "epoch": 2.6
+    },
+    {
+        "loss": 7.6792,
+        "grad_norm": 1.54330575466156,
+        "learning_rate": 0.00013301829001487707,
+        "epoch": 2.6
+    },
+    {
+        "loss": 7.5151,
+        "grad_norm": 1.266360878944397,
+        "learning_rate": 0.0001321431696858318,
+        "epoch": 2.6
+    },
+    {
+        "loss": 7.6357,
+        "grad_norm": 1.1992617845535278,
+        "learning_rate": 0.00013126804935678657,
+        "epoch": 2.61
+    },
+    {
+        "loss": 7.6848,
+        "grad_norm": 1.6269259452819824,
+        "learning_rate": 0.00013039292902774133,
+        "epoch": 2.61
+    },
+    {
+        "loss": 7.3941,
+        "grad_norm": 1.4221471548080444,
+        "learning_rate": 0.00012951780869869606,
+        "epoch": 2.61
+    },
+    {
+        "loss": 7.5638,
+        "grad_norm": 1.31778085231781,
+        "learning_rate": 0.00012864268836965082,
+        "epoch": 2.61
+    },
+    {
+        "loss": 7.3716,
+        "grad_norm": 1.4217979907989502,
+        "learning_rate": 0.00012776756804060559,
+        "epoch": 2.62
+    },
+    {
+        "loss": 7.7403,
+        "grad_norm": 1.549012541770935,
+        "learning_rate": 0.00012689244771156032,
+        "epoch": 2.62
+    },
+    {
+        "loss": 7.5079,
+        "grad_norm": 1.7808821201324463,
+        "learning_rate": 0.00012601732738251508,
+        "epoch": 2.62
+    },
+    {
+        "loss": 7.338,
+        "grad_norm": 1.6030139923095703,
+        "learning_rate": 0.00012514220705346984,
+        "epoch": 2.62
+    },
+    {
+        "loss": 7.2113,
+        "grad_norm": 1.688103437423706,
+        "learning_rate": 0.0001242670867244246,
+        "epoch": 2.63
+    },
+    {
+        "loss": 7.5297,
+        "grad_norm": 1.4482861757278442,
+        "learning_rate": 0.00012339196639537937,
+        "epoch": 2.63
+    },
+    {
+        "loss": 7.6226,
+        "grad_norm": 1.481149435043335,
+        "learning_rate": 0.00012251684606633413,
+        "epoch": 2.63
+    },
+    {
+        "loss": 7.1199,
+        "grad_norm": 1.5914816856384277,
+        "learning_rate": 0.00012164172573728888,
+        "epoch": 2.64
+    },
+    {
+        "loss": 7.5294,
+        "grad_norm": 1.6436686515808105,
+        "learning_rate": 0.00012076660540824364,
+        "epoch": 2.64
+    },
+    {
+        "loss": 7.7319,
+        "grad_norm": 1.422884225845337,
+        "learning_rate": 0.00011989148507919839,
+        "epoch": 2.64
+    },
+    {
+        "loss": 7.5878,
+        "grad_norm": 1.2468681335449219,
+        "learning_rate": 0.00011901636475015315,
+        "epoch": 2.64
+    },
+    {
+        "loss": 7.4093,
+        "grad_norm": 1.6080206632614136,
+        "learning_rate": 0.00011814124442110791,
+        "epoch": 2.65
+    },
+    {
+        "loss": 6.927,
+        "grad_norm": 1.2568819522857666,
+        "learning_rate": 0.00011726612409206266,
+        "epoch": 2.65
+    },
+    {
+        "loss": 7.524,
+        "grad_norm": 1.4558569192886353,
+        "learning_rate": 0.00011639100376301742,
+        "epoch": 2.65
+    },
+    {
+        "loss": 6.7721,
+        "grad_norm": 1.3554805517196655,
+        "learning_rate": 0.00011551588343397218,
+        "epoch": 2.65
+    },
+    {
+        "loss": 7.5129,
+        "grad_norm": 2.061342239379883,
+        "learning_rate": 0.00011464076310492692,
+        "epoch": 2.66
+    },
+    {
+        "loss": 7.271,
+        "grad_norm": 1.7581554651260376,
+        "learning_rate": 0.00011376564277588168,
+        "epoch": 2.66
+    },
+    {
+        "loss": 7.4605,
+        "grad_norm": 1.3818498849868774,
+        "learning_rate": 0.00011289052244683644,
+        "epoch": 2.66
+    },
+    {
+        "loss": 7.2747,
+        "grad_norm": 1.4640157222747803,
+        "learning_rate": 0.00011201540211779119,
+        "epoch": 2.66
+    },
+    {
+        "loss": 7.4137,
+        "grad_norm": 1.628440499305725,
+        "learning_rate": 0.00011114028178874595,
+        "epoch": 2.67
+    },
+    {
+        "loss": 7.1947,
+        "grad_norm": 2.1291253566741943,
+        "learning_rate": 0.00011026516145970071,
+        "epoch": 2.67
+    },
+    {
+        "loss": 7.3972,
+        "grad_norm": 1.53203284740448,
+        "learning_rate": 0.00010939004113065546,
+        "epoch": 2.67
+    },
+    {
+        "loss": 7.1343,
+        "grad_norm": 1.7009447813034058,
+        "learning_rate": 0.00010851492080161022,
+        "epoch": 2.67
+    },
+    {
+        "loss": 7.4999,
+        "grad_norm": 1.981833815574646,
+        "learning_rate": 0.00010763980047256499,
+        "epoch": 2.68
+    },
+    {
+        "loss": 7.0649,
+        "grad_norm": 1.4151135683059692,
+        "learning_rate": 0.00010676468014351973,
+        "epoch": 2.68
+    },
+    {
+        "loss": 7.4975,
+        "grad_norm": 1.8214997053146362,
+        "learning_rate": 0.0001058895598144745,
+        "epoch": 2.68
+    },
+    {
+        "loss": 7.1928,
+        "grad_norm": 1.475014328956604,
+        "learning_rate": 0.00010501443948542926,
+        "epoch": 2.69
+    },
+    {
+        "loss": 6.7309,
+        "grad_norm": 1.500470757484436,
+        "learning_rate": 0.00010413931915638399,
+        "epoch": 2.69
+    },
+    {
+        "loss": 7.2154,
+        "grad_norm": 1.0923032760620117,
+        "learning_rate": 0.00010326419882733875,
+        "epoch": 2.69
+    },
+    {
+        "loss": 7.4584,
+        "grad_norm": 1.476189136505127,
+        "learning_rate": 0.00010238907849829352,
+        "epoch": 2.69
+    },
+    {
+        "loss": 7.5696,
+        "grad_norm": 1.3299099206924438,
+        "learning_rate": 0.00010151395816924828,
+        "epoch": 2.7
+    },
+    {
+        "loss": 7.4462,
+        "grad_norm": 1.248026967048645,
+        "learning_rate": 0.00010063883784020303,
+        "epoch": 2.7
+    },
+    {
+        "loss": 7.057,
+        "grad_norm": 1.5154845714569092,
+        "learning_rate": 9.976371751115779e-05,
+        "epoch": 2.7
+    },
+    {
+        "loss": 7.4942,
+        "grad_norm": 1.504868745803833,
+        "learning_rate": 9.888859718211255e-05,
+        "epoch": 2.7
+    },
+    {
+        "loss": 7.7042,
+        "grad_norm": 1.2087482213974,
+        "learning_rate": 9.80134768530673e-05,
+        "epoch": 2.71
+    },
+    {
+        "loss": 7.7138,
+        "grad_norm": 2.066254138946533,
+        "learning_rate": 9.713835652402206e-05,
+        "epoch": 2.71
+    },
+    {
+        "loss": 7.4746,
+        "grad_norm": 1.2078548669815063,
+        "learning_rate": 9.626323619497682e-05,
+        "epoch": 2.71
+    },
+    {
+        "loss": 7.5682,
+        "grad_norm": 1.2530779838562012,
+        "learning_rate": 9.538811586593157e-05,
+        "epoch": 2.71
+    },
+    {
+        "loss": 7.4491,
+        "grad_norm": 1.5170719623565674,
+        "learning_rate": 9.451299553688632e-05,
+        "epoch": 2.72
+    },
+    {
+        "loss": 7.2938,
+        "grad_norm": 1.2933870553970337,
+        "learning_rate": 9.363787520784108e-05,
+        "epoch": 2.72
+    },
+    {
+        "loss": 7.1455,
+        "grad_norm": 1.212755799293518,
+        "learning_rate": 9.276275487879583e-05,
+        "epoch": 2.72
+    },
+    {
+        "loss": 7.3702,
+        "grad_norm": 1.4118942022323608,
+        "learning_rate": 9.188763454975059e-05,
+        "epoch": 2.72
+    },
+    {
+        "loss": 7.1194,
+        "grad_norm": 1.575276494026184,
+        "learning_rate": 9.101251422070535e-05,
+        "epoch": 2.73
+    },
+    {
+        "loss": 7.046,
+        "grad_norm": 1.3244752883911133,
+        "learning_rate": 9.01373938916601e-05,
+        "epoch": 2.73
+    },
+    {
+        "loss": 6.875,
+        "grad_norm": 1.369280219078064,
+        "learning_rate": 8.926227356261486e-05,
+        "epoch": 2.73
+    },
+    {
+        "loss": 7.4045,
+        "grad_norm": 1.3210042715072632,
+        "learning_rate": 8.838715323356962e-05,
+        "epoch": 2.74
+    },
+    {
+        "loss": 7.5159,
+        "grad_norm": 1.4352552890777588,
+        "learning_rate": 8.751203290452437e-05,
+        "epoch": 2.74
+    },
+    {
+        "loss": 7.2315,
+        "grad_norm": 1.4860197305679321,
+        "learning_rate": 8.663691257547913e-05,
+        "epoch": 2.74
+    },
+    {
+        "loss": 6.8597,
+        "grad_norm": 1.2331523895263672,
+        "learning_rate": 8.57617922464339e-05,
+        "epoch": 2.74
+    },
+    {
+        "loss": 7.3485,
+        "grad_norm": 1.2187525033950806,
+        "learning_rate": 8.488667191738863e-05,
+        "epoch": 2.75
+    },
+    {
+        "loss": 7.388,
+        "grad_norm": 1.1800241470336914,
+        "learning_rate": 8.401155158834339e-05,
+        "epoch": 2.75
+    },
+    {
+        "loss": 6.9186,
+        "grad_norm": 1.3542723655700684,
+        "learning_rate": 8.313643125929815e-05,
+        "epoch": 2.75
+    },
+    {
+        "loss": 6.9582,
+        "grad_norm": 1.3839143514633179,
+        "learning_rate": 8.22613109302529e-05,
+        "epoch": 2.75
+    },
+    {
+        "loss": 7.4176,
+        "grad_norm": 1.4546840190887451,
+        "learning_rate": 8.138619060120766e-05,
+        "epoch": 2.76
+    },
+    {
+        "loss": 7.2731,
+        "grad_norm": 1.3623560667037964,
+        "learning_rate": 8.051107027216243e-05,
+        "epoch": 2.76
+    },
+    {
+        "loss": 7.1633,
+        "grad_norm": 1.9331005811691284,
+        "learning_rate": 7.963594994311717e-05,
+        "epoch": 2.76
+    },
+    {
+        "loss": 6.8972,
+        "grad_norm": 1.2791029214859009,
+        "learning_rate": 7.876082961407194e-05,
+        "epoch": 2.76
+    },
+    {
+        "loss": 7.1043,
+        "grad_norm": 1.6202424764633179,
+        "learning_rate": 7.78857092850267e-05,
+        "epoch": 2.77
+    },
+    {
+        "loss": 7.0727,
+        "grad_norm": 1.0835381746292114,
+        "learning_rate": 7.701058895598145e-05,
+        "epoch": 2.77
+    },
+    {
+        "loss": 7.0958,
+        "grad_norm": 1.2778371572494507,
+        "learning_rate": 7.613546862693621e-05,
+        "epoch": 2.77
+    },
+    {
+        "loss": 7.2219,
+        "grad_norm": 1.9295389652252197,
+        "learning_rate": 7.526034829789097e-05,
+        "epoch": 2.77
+    },
+    {
+        "loss": 7.0189,
+        "grad_norm": 1.9394477605819702,
+        "learning_rate": 7.43852279688457e-05,
+        "epoch": 2.78
+    },
+    {
+        "loss": 7.0144,
+        "grad_norm": 1.4238934516906738,
+        "learning_rate": 7.351010763980047e-05,
+        "epoch": 2.78
+    },
+    {
+        "loss": 7.2353,
+        "grad_norm": 1.350537657737732,
+        "learning_rate": 7.263498731075523e-05,
+        "epoch": 2.78
+    },
+    {
+        "loss": 6.7353,
+        "grad_norm": 1.3214153051376343,
+        "learning_rate": 7.175986698170999e-05,
+        "epoch": 2.79
+    },
+    {
+        "loss": 7.4143,
+        "grad_norm": 2.469216823577881,
+        "learning_rate": 7.088474665266474e-05,
+        "epoch": 2.79
+    },
+    {
+        "loss": 7.4276,
+        "grad_norm": 1.414184808731079,
+        "learning_rate": 7.00096263236195e-05,
+        "epoch": 2.79
+    },
+    {
+        "loss": 6.9842,
+        "grad_norm": 1.4708011150360107,
+        "learning_rate": 6.913450599457426e-05,
+        "epoch": 2.79
+    },
+    {
+        "loss": 7.572,
+        "grad_norm": 1.449560284614563,
+        "learning_rate": 6.825938566552901e-05,
+        "epoch": 2.8
+    },
+    {
+        "loss": 7.3449,
+        "grad_norm": 1.1261264085769653,
+        "learning_rate": 6.738426533648377e-05,
+        "epoch": 2.8
+    },
+    {
+        "loss": 7.1776,
+        "grad_norm": 1.5502110719680786,
+        "learning_rate": 6.650914500743853e-05,
+        "epoch": 2.8
+    },
+    {
+        "loss": 7.0565,
+        "grad_norm": 1.3916562795639038,
+        "learning_rate": 6.563402467839328e-05,
+        "epoch": 2.8
+    },
+    {
+        "loss": 7.0882,
+        "grad_norm": 1.361229658126831,
+        "learning_rate": 6.475890434934803e-05,
+        "epoch": 2.81
+    },
+    {
+        "loss": 6.981,
+        "grad_norm": 1.6100305318832397,
+        "learning_rate": 6.388378402030279e-05,
+        "epoch": 2.81
+    },
+    {
+        "loss": 7.2502,
+        "grad_norm": 1.5449306964874268,
+        "learning_rate": 6.300866369125754e-05,
+        "epoch": 2.81
+    },
+    {
+        "loss": 7.4208,
+        "grad_norm": 1.3188410997390747,
+        "learning_rate": 6.21335433622123e-05,
+        "epoch": 2.81
+    },
+    {
+        "loss": 7.2957,
+        "grad_norm": 1.543289303779602,
+        "learning_rate": 6.125842303316706e-05,
+        "epoch": 2.82
+    },
+    {
+        "loss": 7.0319,
+        "grad_norm": 1.1590594053268433,
+        "learning_rate": 6.038330270412182e-05,
+        "epoch": 2.82
+    },
+    {
+        "loss": 7.23,
+        "grad_norm": 1.1623939275741577,
+        "learning_rate": 5.9508182375076575e-05,
+        "epoch": 2.82
+    },
+    {
+        "loss": 7.1254,
+        "grad_norm": 1.6204333305358887,
+        "learning_rate": 5.863306204603133e-05,
+        "epoch": 2.82
+    },
+    {
+        "loss": 7.4319,
+        "grad_norm": 1.5845638513565063,
+        "learning_rate": 5.775794171698609e-05,
+        "epoch": 2.83
+    },
+    {
+        "loss": 7.4574,
+        "grad_norm": 1.3281787633895874,
+        "learning_rate": 5.688282138794084e-05,
+        "epoch": 2.83
+    },
+    {
+        "loss": 6.8629,
+        "grad_norm": 1.6502999067306519,
+        "learning_rate": 5.6007701058895595e-05,
+        "epoch": 2.83
+    },
+    {
+        "loss": 7.1493,
+        "grad_norm": 1.7768168449401855,
+        "learning_rate": 5.513258072985036e-05,
+        "epoch": 2.83
+    },
+    {
+        "loss": 7.1971,
+        "grad_norm": 1.1763763427734375,
+        "learning_rate": 5.425746040080511e-05,
+        "epoch": 2.84
+    },
+    {
+        "loss": 7.4182,
+        "grad_norm": 1.4033911228179932,
+        "learning_rate": 5.338234007175987e-05,
+        "epoch": 2.84
+    },
+    {
+        "loss": 6.8175,
+        "grad_norm": 1.5407586097717285,
+        "learning_rate": 5.250721974271463e-05,
+        "epoch": 2.84
+    },
+    {
+        "loss": 7.5091,
+        "grad_norm": 1.5829062461853027,
+        "learning_rate": 5.163209941366938e-05,
+        "epoch": 2.85
+    },
+    {
+        "loss": 7.0728,
+        "grad_norm": 1.3185957670211792,
+        "learning_rate": 5.075697908462414e-05,
+        "epoch": 2.85
+    },
+    {
+        "loss": 7.1931,
+        "grad_norm": 1.1996837854385376,
+        "learning_rate": 4.9881858755578894e-05,
+        "epoch": 2.85
+    },
+    {
+        "loss": 7.2327,
+        "grad_norm": 1.6188883781433105,
+        "learning_rate": 4.900673842653365e-05,
+        "epoch": 2.85
+    },
+    {
+        "loss": 7.2432,
+        "grad_norm": 1.7829197645187378,
+        "learning_rate": 4.813161809748841e-05,
+        "epoch": 2.86
+    },
+    {
+        "loss": 6.8231,
+        "grad_norm": 1.3998175859451294,
+        "learning_rate": 4.725649776844316e-05,
+        "epoch": 2.86
+    },
+    {
+        "loss": 7.5838,
+        "grad_norm": 1.6664845943450928,
+        "learning_rate": 4.6381377439397914e-05,
+        "epoch": 2.86
+    },
+    {
+        "loss": 7.3804,
+        "grad_norm": 1.2328096628189087,
+        "learning_rate": 4.5506257110352676e-05,
+        "epoch": 2.86
+    },
+    {
+        "loss": 7.1497,
+        "grad_norm": 1.5543657541275024,
+        "learning_rate": 4.463113678130743e-05,
+        "epoch": 2.87
+    },
+    {
+        "loss": 7.5067,
+        "grad_norm": 2.0711114406585693,
+        "learning_rate": 4.3756016452262186e-05,
+        "epoch": 2.87
+    },
+    {
+        "loss": 7.1481,
+        "grad_norm": 2.340829372406006,
+        "learning_rate": 4.288089612321695e-05,
+        "epoch": 2.87
+    },
+    {
+        "loss": 7.2767,
+        "grad_norm": 1.3014119863510132,
+        "learning_rate": 4.2005775794171696e-05,
+        "epoch": 2.87
+    },
+    {
+        "loss": 7.2583,
+        "grad_norm": 1.186070442199707,
+        "learning_rate": 4.113065546512645e-05,
+        "epoch": 2.88
+    },
+    {
+        "loss": 7.7179,
+        "grad_norm": 1.4286901950836182,
+        "learning_rate": 4.025553513608121e-05,
+        "epoch": 2.88
+    },
+    {
+        "loss": 6.9271,
+        "grad_norm": 1.561988115310669,
+        "learning_rate": 3.938041480703597e-05,
+        "epoch": 2.88
+    },
+    {
+        "loss": 6.9378,
+        "grad_norm": 1.2756584882736206,
+        "learning_rate": 3.8505294477990723e-05,
+        "epoch": 2.88
+    },
+    {
+        "loss": 7.8091,
+        "grad_norm": 1.5452569723129272,
+        "learning_rate": 3.7630174148945485e-05,
+        "epoch": 2.89
+    },
+    {
+        "loss": 6.7905,
+        "grad_norm": 1.2616968154907227,
+        "learning_rate": 3.6755053819900234e-05,
+        "epoch": 2.89
+    },
+    {
+        "loss": 7.3958,
+        "grad_norm": 1.1684807538986206,
+        "learning_rate": 3.5879933490854995e-05,
+        "epoch": 2.89
+    },
+    {
+        "loss": 6.9238,
+        "grad_norm": 1.351366639137268,
+        "learning_rate": 3.500481316180975e-05,
+        "epoch": 2.9
+    },
+    {
+        "loss": 7.4026,
+        "grad_norm": 1.2473573684692383,
+        "learning_rate": 3.4129692832764505e-05,
+        "epoch": 2.9
+    },
+    {
+        "loss": 7.4247,
+        "grad_norm": 1.5123474597930908,
+        "learning_rate": 3.325457250371927e-05,
+        "epoch": 2.9
+    },
+    {
+        "loss": 7.0967,
+        "grad_norm": 1.1452938318252563,
+        "learning_rate": 3.2379452174674016e-05,
+        "epoch": 2.9
+    },
+    {
+        "loss": 7.0357,
+        "grad_norm": 1.1505627632141113,
+        "learning_rate": 3.150433184562877e-05,
+        "epoch": 2.91
+    },
+    {
+        "loss": 7.4973,
+        "grad_norm": 1.438091516494751,
+        "learning_rate": 3.062921151658353e-05,
+        "epoch": 2.91
+    },
+    {
+        "loss": 7.4715,
+        "grad_norm": 1.1489310264587402,
+        "learning_rate": 2.9754091187538288e-05,
+        "epoch": 2.91
+    },
+    {
+        "loss": 7.0076,
+        "grad_norm": 1.3423534631729126,
+        "learning_rate": 2.8878970858493046e-05,
+        "epoch": 2.91
+    },
+    {
+        "loss": 7.0935,
+        "grad_norm": 1.2484374046325684,
+        "learning_rate": 2.8003850529447798e-05,
+        "epoch": 2.92
+    },
+    {
+        "loss": 7.1792,
+        "grad_norm": 1.310231328010559,
+        "learning_rate": 2.7128730200402556e-05,
+        "epoch": 2.92
+    },
+    {
+        "loss": 7.3469,
+        "grad_norm": 1.417974591255188,
+        "learning_rate": 2.6253609871357314e-05,
+        "epoch": 2.92
+    },
+    {
+        "loss": 7.2473,
+        "grad_norm": 1.3878840208053589,
+        "learning_rate": 2.537848954231207e-05,
+        "epoch": 2.92
+    },
+    {
+        "loss": 7.1321,
+        "grad_norm": 1.6403028964996338,
+        "learning_rate": 2.459088124617135e-05,
+        "epoch": 2.93
+    },
+    {
+        "loss": 7.6076,
+        "grad_norm": 1.2110294103622437,
+        "learning_rate": 2.3715760917126104e-05,
+        "epoch": 2.93
+    },
+    {
+        "loss": 7.3466,
+        "grad_norm": 1.203755497932434,
+        "learning_rate": 2.2840640588080863e-05,
+        "epoch": 2.93
+    },
+    {
+        "loss": 7.4367,
+        "grad_norm": 1.2081892490386963,
+        "learning_rate": 2.1965520259035618e-05,
+        "epoch": 2.93
+    },
+    {
+        "loss": 7.6191,
+        "grad_norm": 1.2515225410461426,
+        "learning_rate": 2.1090399929990373e-05,
+        "epoch": 2.94
+    },
+    {
+        "loss": 7.2915,
+        "grad_norm": 1.2461618185043335,
+        "learning_rate": 2.021527960094513e-05,
+        "epoch": 2.94
+    },
+    {
+        "loss": 7.0825,
+        "grad_norm": 1.3424855470657349,
+        "learning_rate": 1.9340159271899886e-05,
+        "epoch": 2.94
+    },
+    {
+        "loss": 7.6924,
+        "grad_norm": 1.2109103202819824,
+        "learning_rate": 1.846503894285464e-05,
+        "epoch": 2.95
+    },
+    {
+        "loss": 7.531,
+        "grad_norm": 1.2161798477172852,
+        "learning_rate": 1.75899186138094e-05,
+        "epoch": 2.95
+    },
+    {
+        "loss": 7.1992,
+        "grad_norm": 1.347778081893921,
+        "learning_rate": 1.6714798284764158e-05,
+        "epoch": 2.95
+    },
+    {
+        "loss": 7.7785,
+        "grad_norm": 1.2869161367416382,
+        "learning_rate": 1.583967795571891e-05,
+        "epoch": 2.95
+    },
+    {
+        "loss": 7.6703,
+        "grad_norm": 1.1452679634094238,
+        "learning_rate": 1.4964557626673668e-05,
+        "epoch": 2.96
+    },
+    {
+        "loss": 7.3311,
+        "grad_norm": 1.7757437229156494,
+        "learning_rate": 1.4089437297628423e-05,
+        "epoch": 2.96
+    },
+    {
+        "loss": 7.4272,
+        "grad_norm": 1.2730258703231812,
+        "learning_rate": 1.3214316968583182e-05,
+        "epoch": 2.96
+    },
+    {
+        "loss": 6.8195,
+        "grad_norm": 1.0826276540756226,
+        "learning_rate": 1.2339196639537937e-05,
+        "epoch": 2.96
+    },
+    {
+        "loss": 7.1219,
+        "grad_norm": 1.3847414255142212,
+        "learning_rate": 1.1464076310492692e-05,
+        "epoch": 2.97
+    },
+    {
+        "loss": 7.5912,
+        "grad_norm": 1.4612926244735718,
+        "learning_rate": 1.0588955981447449e-05,
+        "epoch": 2.97
+    },
+    {
+        "loss": 6.9373,
+        "grad_norm": 1.5692036151885986,
+        "learning_rate": 9.713835652402205e-06,
+        "epoch": 2.97
+    },
+    {
+        "loss": 7.7104,
+        "grad_norm": 1.4740134477615356,
+        "learning_rate": 8.838715323356962e-06,
+        "epoch": 2.97
+    },
+    {
+        "loss": 7.1918,
+        "grad_norm": 1.026573657989502,
+        "learning_rate": 7.963594994311717e-06,
+        "epoch": 2.98
+    },
+    {
+        "loss": 6.8717,
+        "grad_norm": 1.1959487199783325,
+        "learning_rate": 7.088474665266474e-06,
+        "epoch": 2.98
+    },
+    {
+        "loss": 7.4154,
+        "grad_norm": 1.1354584693908691,
+        "learning_rate": 6.213354336221231e-06,
+        "epoch": 2.98
+    },
+    {
+        "loss": 7.1622,
+        "grad_norm": 1.3372441530227661,
+        "learning_rate": 5.338234007175987e-06,
+        "epoch": 2.98
+    },
+    {
+        "loss": 6.9564,
+        "grad_norm": 1.1713366508483887,
+        "learning_rate": 4.463113678130743e-06,
+        "epoch": 2.99
+    },
+    {
+        "loss": 7.462,
+        "grad_norm": 1.8238294124603271,
+        "learning_rate": 3.587993349085499e-06,
+        "epoch": 2.99
+    },
+    {
+        "loss": 7.5493,
+        "grad_norm": 1.3313993215560913,
+        "learning_rate": 2.7128730200402555e-06,
+        "epoch": 2.99
+    },
+    {
+        "loss": 7.2399,
+        "grad_norm": 1.1780248880386353,
+        "learning_rate": 1.8377526909950118e-06,
+        "epoch": 3.0
+    },
+    {
+        "loss": 6.879,
+        "grad_norm": 1.2703826427459717,
+        "learning_rate": 9.626323619497682e-07,
+        "epoch": 3.0
+    },
+    {
+        "train_runtime": 104781.7564,
+        "train_samples_per_second": 3.49,
+        "train_steps_per_second": 0.109,
+        "train_loss": 8.437174775609405,
+        "epoch": 3.0
+    }
+]
\ No newline at end of file