diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,33648 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.53664,
+  "eval_steps": 500,
+  "global_step": 4802,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00032,
+      "grad_norm": 1.7359182834625244,
+      "learning_rate": 0.0,
+      "loss": 1.9045,
+      "step": 1
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 1.6704541444778442,
+      "learning_rate": 4e-05,
+      "loss": 1.9266,
+      "step": 2
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 1.1031880378723145,
+      "learning_rate": 8e-05,
+      "loss": 1.4249,
+      "step": 3
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 1.3405920267105103,
+      "learning_rate": 0.00012,
+      "loss": 1.5282,
+      "step": 4
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 1.149672508239746,
+      "learning_rate": 0.00016,
+      "loss": 1.4246,
+      "step": 5
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 1.5152997970581055,
+      "learning_rate": 0.0002,
+      "loss": 1.6463,
+      "step": 6
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 1.0266704559326172,
+      "learning_rate": 0.0001999679743795036,
+      "loss": 1.3269,
+      "step": 7
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 0.9892048835754395,
+      "learning_rate": 0.00019993594875900722,
+      "loss": 1.5465,
+      "step": 8
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 0.6304389834403992,
+      "learning_rate": 0.00019990392313851084,
+      "loss": 1.3072,
+      "step": 9
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 0.5412458181381226,
+      "learning_rate": 0.00019987189751801443,
+      "loss": 1.2639,
+      "step": 10
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 0.5660735964775085,
+      "learning_rate": 0.00019983987189751802,
+      "loss": 1.0541,
+      "step": 11
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 0.5487312078475952,
+      "learning_rate": 0.00019980784627702164,
+      "loss": 1.0576,
+      "step": 12
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 0.592808187007904,
+      "learning_rate": 0.00019977582065652523,
+      "loss": 1.1547,
+      "step": 13
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 0.581730306148529,
+      "learning_rate": 0.00019974379503602882,
+      "loss": 1.0445,
+      "step": 14
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 0.695290207862854,
+      "learning_rate": 0.00019971176941553244,
+      "loss": 1.2098,
+      "step": 15
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.4707532227039337,
+      "learning_rate": 0.00019967974379503606,
+      "loss": 0.9183,
+      "step": 16
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 0.4225716292858124,
+      "learning_rate": 0.00019964771817453965,
+      "loss": 0.8521,
+      "step": 17
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 0.3538431227207184,
+      "learning_rate": 0.00019961569255404324,
+      "loss": 0.9504,
+      "step": 18
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 0.41204696893692017,
+      "learning_rate": 0.00019958366693354683,
+      "loss": 1.0759,
+      "step": 19
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.4316575825214386,
+      "learning_rate": 0.00019955164131305045,
+      "loss": 1.0775,
+      "step": 20
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 0.3429096043109894,
+      "learning_rate": 0.00019951961569255407,
+      "loss": 0.9161,
+      "step": 21
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 0.38494449853897095,
+      "learning_rate": 0.00019948759007205766,
+      "loss": 0.9866,
+      "step": 22
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 0.3482462167739868,
+      "learning_rate": 0.00019945556445156125,
+      "loss": 0.9636,
+      "step": 23
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.3135548532009125,
+      "learning_rate": 0.00019942353883106487,
+      "loss": 0.8555,
+      "step": 24
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 0.3369140625,
+      "learning_rate": 0.00019939151321056846,
+      "loss": 0.8487,
+      "step": 25
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 0.3705412447452545,
+      "learning_rate": 0.00019935948759007205,
+      "loss": 0.9237,
+      "step": 26
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 0.3516896963119507,
+      "learning_rate": 0.00019932746196957567,
+      "loss": 0.8189,
+      "step": 27
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.34645694494247437,
+      "learning_rate": 0.0001992954363490793,
+      "loss": 0.9326,
+      "step": 28
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 0.39456260204315186,
+      "learning_rate": 0.00019926341072858288,
+      "loss": 0.992,
+      "step": 29
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 0.3115796744823456,
+      "learning_rate": 0.00019923138510808647,
+      "loss": 0.7624,
+      "step": 30
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 0.3214782178401947,
+      "learning_rate": 0.0001991993594875901,
+      "loss": 0.8499,
+      "step": 31
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.3627919852733612,
+      "learning_rate": 0.00019916733386709368,
+      "loss": 0.9273,
+      "step": 32
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 0.3324357569217682,
+      "learning_rate": 0.0001991353082465973,
+      "loss": 0.8388,
+      "step": 33
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 0.3463478684425354,
+      "learning_rate": 0.0001991032826261009,
+      "loss": 0.8743,
+      "step": 34
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 0.26795780658721924,
+      "learning_rate": 0.0001990712570056045,
+      "loss": 0.7892,
+      "step": 35
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.31018561124801636,
+      "learning_rate": 0.0001990392313851081,
+      "loss": 0.7753,
+      "step": 36
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 0.418283611536026,
+      "learning_rate": 0.0001990072057646117,
+      "loss": 0.9088,
+      "step": 37
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 0.36864641308784485,
+      "learning_rate": 0.00019897518014411528,
+      "loss": 0.849,
+      "step": 38
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 0.3467395305633545,
+      "learning_rate": 0.00019894315452361893,
+      "loss": 0.7657,
+      "step": 39
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.4149913489818573,
+      "learning_rate": 0.00019891112890312252,
+      "loss": 0.8615,
+      "step": 40
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 0.4301813244819641,
+      "learning_rate": 0.0001988791032826261,
+      "loss": 0.9959,
+      "step": 41
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 0.3698485195636749,
+      "learning_rate": 0.0001988470776621297,
+      "loss": 0.8476,
+      "step": 42
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 0.3671010434627533,
+      "learning_rate": 0.00019881505204163332,
+      "loss": 0.8725,
+      "step": 43
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 0.38881540298461914,
+      "learning_rate": 0.0001987830264211369,
+      "loss": 0.7381,
+      "step": 44
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 0.3932352662086487,
+      "learning_rate": 0.00019875100080064053,
+      "loss": 0.7524,
+      "step": 45
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 0.3653354048728943,
+      "learning_rate": 0.00019871897518014412,
+      "loss": 0.9459,
+      "step": 46
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 0.2764647305011749,
+      "learning_rate": 0.00019868694955964774,
+      "loss": 0.648,
+      "step": 47
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 0.28200414776802063,
+      "learning_rate": 0.00019865492393915133,
+      "loss": 0.7517,
+      "step": 48
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 0.4025580883026123,
+      "learning_rate": 0.00019862289831865492,
+      "loss": 0.9034,
+      "step": 49
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.32870325446128845,
+      "learning_rate": 0.00019859087269815854,
+      "loss": 0.9338,
+      "step": 50
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 0.2876620888710022,
+      "learning_rate": 0.00019855884707766216,
+      "loss": 0.7181,
+      "step": 51
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 0.3928801715373993,
+      "learning_rate": 0.00019852682145716575,
+      "loss": 0.8145,
+      "step": 52
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 0.3390360176563263,
+      "learning_rate": 0.00019849479583666934,
+      "loss": 0.8593,
+      "step": 53
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 0.39917275309562683,
+      "learning_rate": 0.00019846277021617296,
+      "loss": 0.9731,
+      "step": 54
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.4004969894886017,
+      "learning_rate": 0.00019843074459567655,
+      "loss": 0.8004,
+      "step": 55
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 0.3302912712097168,
+      "learning_rate": 0.00019839871897518017,
+      "loss": 0.7944,
+      "step": 56
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 0.41863322257995605,
+      "learning_rate": 0.00019836669335468376,
+      "loss": 0.9368,
+      "step": 57
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 0.2867134213447571,
+      "learning_rate": 0.00019833466773418738,
+      "loss": 0.7386,
+      "step": 58
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 0.39955514669418335,
+      "learning_rate": 0.00019830264211369097,
+      "loss": 0.9166,
+      "step": 59
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.32057785987854004,
+      "learning_rate": 0.00019827061649319456,
+      "loss": 0.7185,
+      "step": 60
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 0.39947280287742615,
+      "learning_rate": 0.00019823859087269815,
+      "loss": 0.7844,
+      "step": 61
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 0.37485724687576294,
+      "learning_rate": 0.00019820656525220177,
+      "loss": 0.82,
+      "step": 62
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 0.331901878118515,
+      "learning_rate": 0.00019817453963170539,
+      "loss": 0.8655,
+      "step": 63
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.44090256094932556,
+      "learning_rate": 0.00019814251401120898,
+      "loss": 0.8458,
+      "step": 64
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 0.3575265407562256,
+      "learning_rate": 0.00019811048839071257,
+      "loss": 0.809,
+      "step": 65
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 0.3753909468650818,
+      "learning_rate": 0.0001980784627702162,
+      "loss": 0.8424,
+      "step": 66
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 0.3712959587574005,
+      "learning_rate": 0.00019804643714971978,
+      "loss": 0.8199,
+      "step": 67
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 0.43370306491851807,
+      "learning_rate": 0.0001980144115292234,
+      "loss": 0.7899,
+      "step": 68
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 0.35110247135162354,
+      "learning_rate": 0.000197982385908727,
+      "loss": 0.6704,
+      "step": 69
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.3548092842102051,
+      "learning_rate": 0.0001979503602882306,
+      "loss": 0.9211,
+      "step": 70
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 0.3629775047302246,
+      "learning_rate": 0.0001979183346677342,
+      "loss": 0.8214,
+      "step": 71
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.41008803248405457,
+      "learning_rate": 0.0001978863090472378,
+      "loss": 0.9636,
+      "step": 72
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 0.4360860586166382,
+      "learning_rate": 0.0001978542834267414,
+      "loss": 0.8545,
+      "step": 73
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 0.3629567325115204,
+      "learning_rate": 0.00019782225780624502,
+      "loss": 0.7234,
+      "step": 74
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.4093756079673767,
+      "learning_rate": 0.00019779023218574862,
+      "loss": 0.9529,
+      "step": 75
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.37465348839759827,
+      "learning_rate": 0.0001977582065652522,
+      "loss": 0.7311,
+      "step": 76
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 0.3660638630390167,
+      "learning_rate": 0.00019772618094475583,
+      "loss": 0.7761,
+      "step": 77
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 0.396746426820755,
+      "learning_rate": 0.00019769415532425942,
+      "loss": 0.8572,
+      "step": 78
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 0.32552897930145264,
+      "learning_rate": 0.000197662129703763,
+      "loss": 0.7406,
+      "step": 79
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.41160935163497925,
+      "learning_rate": 0.00019763010408326663,
+      "loss": 0.8938,
+      "step": 80
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 0.4210125207901001,
+      "learning_rate": 0.00019759807846277022,
+      "loss": 0.8589,
+      "step": 81
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 0.3710517883300781,
+      "learning_rate": 0.00019756605284227384,
+      "loss": 0.8776,
+      "step": 82
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 0.4252607226371765,
+      "learning_rate": 0.00019753402722177743,
+      "loss": 0.9195,
+      "step": 83
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.36970028281211853,
+      "learning_rate": 0.00019750200160128102,
+      "loss": 0.6894,
+      "step": 84
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.36201247572898865,
+      "learning_rate": 0.00019746997598078464,
+      "loss": 0.7803,
+      "step": 85
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 0.4238883852958679,
+      "learning_rate": 0.00019743795036028826,
+      "loss": 0.9729,
+      "step": 86
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 0.39085498452186584,
+      "learning_rate": 0.00019740592473979185,
+      "loss": 0.6993,
+      "step": 87
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.3699302673339844,
+      "learning_rate": 0.00019737389911929544,
+      "loss": 0.8322,
+      "step": 88
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 0.4149015545845032,
+      "learning_rate": 0.00019734187349879906,
+      "loss": 0.8532,
+      "step": 89
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.38202205300331116,
+      "learning_rate": 0.00019730984787830265,
+      "loss": 0.766,
+      "step": 90
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 0.4083291292190552,
+      "learning_rate": 0.00019727782225780624,
+      "loss": 0.741,
+      "step": 91
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.47629761695861816,
+      "learning_rate": 0.00019724579663730986,
+      "loss": 0.7736,
+      "step": 92
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 0.36554235219955444,
+      "learning_rate": 0.00019721377101681347,
+      "loss": 0.7396,
+      "step": 93
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 0.43680229783058167,
+      "learning_rate": 0.00019718174539631707,
+      "loss": 0.9695,
+      "step": 94
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 0.3364362418651581,
+      "learning_rate": 0.00019714971977582066,
+      "loss": 0.75,
+      "step": 95
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.36064544320106506,
+      "learning_rate": 0.00019711769415532425,
+      "loss": 0.7843,
+      "step": 96
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 0.42769667506217957,
+      "learning_rate": 0.00019708566853482787,
+      "loss": 0.727,
+      "step": 97
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 0.34345200657844543,
+      "learning_rate": 0.00019705364291433149,
+      "loss": 0.6893,
+      "step": 98
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 0.39843839406967163,
+      "learning_rate": 0.00019702161729383508,
+      "loss": 0.8338,
+      "step": 99
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.3380451500415802,
+      "learning_rate": 0.00019698959167333867,
+      "loss": 0.7143,
+      "step": 100
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 0.449645459651947,
+      "learning_rate": 0.00019695756605284229,
+      "loss": 0.836,
+      "step": 101
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 0.3968237340450287,
+      "learning_rate": 0.00019692554043234588,
+      "loss": 0.7591,
+      "step": 102
+    },
+    {
+      "epoch": 0.03296,
+      "grad_norm": 0.4413282871246338,
+      "learning_rate": 0.0001968935148118495,
+      "loss": 0.7983,
+      "step": 103
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 0.4213366210460663,
+      "learning_rate": 0.0001968614891913531,
+      "loss": 0.7584,
+      "step": 104
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 0.39116552472114563,
+      "learning_rate": 0.0001968294635708567,
+      "loss": 0.83,
+      "step": 105
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 0.45863527059555054,
+      "learning_rate": 0.0001967974379503603,
+      "loss": 0.8189,
+      "step": 106
+    },
+    {
+      "epoch": 0.03424,
+      "grad_norm": 0.3799077868461609,
+      "learning_rate": 0.0001967654123298639,
+      "loss": 0.7526,
+      "step": 107
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 0.35227102041244507,
+      "learning_rate": 0.0001967333867093675,
+      "loss": 0.7267,
+      "step": 108
+    },
+    {
+      "epoch": 0.03488,
+      "grad_norm": 0.4498932659626007,
+      "learning_rate": 0.00019670136108887112,
+      "loss": 0.7758,
+      "step": 109
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 0.4082190990447998,
+      "learning_rate": 0.00019666933546837472,
+      "loss": 0.79,
+      "step": 110
+    },
+    {
+      "epoch": 0.03552,
+      "grad_norm": 0.4846765697002411,
+      "learning_rate": 0.0001966373098478783,
+      "loss": 0.7317,
+      "step": 111
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 0.3948225677013397,
+      "learning_rate": 0.00019660528422738192,
+      "loss": 0.7341,
+      "step": 112
+    },
+    {
+      "epoch": 0.03616,
+      "grad_norm": 0.3919749855995178,
+      "learning_rate": 0.00019657325860688552,
+      "loss": 0.7166,
+      "step": 113
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 0.36422061920166016,
+      "learning_rate": 0.0001965412329863891,
+      "loss": 0.71,
+      "step": 114
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 0.36287641525268555,
+      "learning_rate": 0.00019650920736589273,
+      "loss": 0.8105,
+      "step": 115
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 0.44533467292785645,
+      "learning_rate": 0.00019647718174539634,
+      "loss": 0.6905,
+      "step": 116
+    },
+    {
+      "epoch": 0.03744,
+      "grad_norm": 0.3648212254047394,
+      "learning_rate": 0.00019644515612489993,
+      "loss": 0.7301,
+      "step": 117
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 0.3800407350063324,
+      "learning_rate": 0.00019641313050440353,
+      "loss": 0.7837,
+      "step": 118
+    },
+    {
+      "epoch": 0.03808,
+      "grad_norm": 0.4288330376148224,
+      "learning_rate": 0.00019638110488390712,
+      "loss": 0.7679,
+      "step": 119
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.4401828646659851,
+      "learning_rate": 0.00019634907926341074,
+      "loss": 0.7805,
+      "step": 120
+    },
+    {
+      "epoch": 0.03872,
+      "grad_norm": 0.41975656151771545,
+      "learning_rate": 0.00019631705364291435,
+      "loss": 1.0787,
+      "step": 121
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 0.41932767629623413,
+      "learning_rate": 0.00019628502802241795,
+      "loss": 0.9004,
+      "step": 122
+    },
+    {
+      "epoch": 0.03936,
+      "grad_norm": 0.47830823063850403,
+      "learning_rate": 0.00019625300240192154,
+      "loss": 0.9765,
+      "step": 123
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 0.38607075810432434,
+      "learning_rate": 0.00019622097678142515,
+      "loss": 0.8089,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4077676236629486,
+      "learning_rate": 0.00019618895116092875,
+      "loss": 0.6443,
+      "step": 125
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 0.5592461824417114,
+      "learning_rate": 0.00019615692554043234,
+      "loss": 0.6312,
+      "step": 126
+    },
+    {
+      "epoch": 0.04064,
+      "grad_norm": 0.44260603189468384,
+      "learning_rate": 0.00019612489991993596,
+      "loss": 0.7098,
+      "step": 127
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 0.45345351099967957,
+      "learning_rate": 0.00019609287429943957,
+      "loss": 0.9453,
+      "step": 128
+    },
+    {
+      "epoch": 0.04128,
+      "grad_norm": 0.3703489899635315,
+      "learning_rate": 0.00019606084867894316,
+      "loss": 0.7645,
+      "step": 129
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.37588033080101013,
+      "learning_rate": 0.00019602882305844676,
+      "loss": 0.7328,
+      "step": 130
+    },
+    {
+      "epoch": 0.04192,
+      "grad_norm": 0.346963495016098,
+      "learning_rate": 0.00019599679743795037,
+      "loss": 0.7209,
+      "step": 131
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 0.3816434442996979,
+      "learning_rate": 0.00019596477181745397,
+      "loss": 0.7258,
+      "step": 132
+    },
+    {
+      "epoch": 0.04256,
+      "grad_norm": 0.3990420401096344,
+      "learning_rate": 0.00019593274619695758,
+      "loss": 0.7597,
+      "step": 133
+    },
+    {
+      "epoch": 0.04288,
+      "grad_norm": 0.4126216769218445,
+      "learning_rate": 0.00019590072057646118,
+      "loss": 0.6469,
+      "step": 134
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 0.4426156282424927,
+      "learning_rate": 0.0001958686949559648,
+      "loss": 0.8142,
+      "step": 135
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 0.39366278052330017,
+      "learning_rate": 0.00019583666933546838,
+      "loss": 0.7667,
+      "step": 136
+    },
+    {
+      "epoch": 0.04384,
+      "grad_norm": 0.36906182765960693,
+      "learning_rate": 0.00019580464371497198,
+      "loss": 0.6529,
+      "step": 137
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 0.348753422498703,
+      "learning_rate": 0.00019577261809447557,
+      "loss": 0.7653,
+      "step": 138
+    },
+    {
+      "epoch": 0.04448,
+      "grad_norm": 0.41684240102767944,
+      "learning_rate": 0.0001957405924739792,
+      "loss": 0.8991,
+      "step": 139
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.3756757080554962,
+      "learning_rate": 0.0001957085668534828,
+      "loss": 0.6997,
+      "step": 140
+    },
+    {
+      "epoch": 0.04512,
+      "grad_norm": 0.36973434686660767,
+      "learning_rate": 0.0001956765412329864,
+      "loss": 0.7184,
+      "step": 141
+    },
+    {
+      "epoch": 0.04544,
+      "grad_norm": 0.3714083433151245,
+      "learning_rate": 0.00019564451561248999,
+      "loss": 0.6197,
+      "step": 142
+    },
+    {
+      "epoch": 0.04576,
+      "grad_norm": 0.4686843454837799,
+      "learning_rate": 0.0001956124899919936,
+      "loss": 0.9112,
+      "step": 143
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 0.43380072712898254,
+      "learning_rate": 0.0001955804643714972,
+      "loss": 0.9025,
+      "step": 144
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 0.41534385085105896,
+      "learning_rate": 0.00019554843875100081,
+      "loss": 0.8087,
+      "step": 145
+    },
+    {
+      "epoch": 0.04672,
+      "grad_norm": 0.4158933162689209,
+      "learning_rate": 0.0001955164131305044,
+      "loss": 0.8249,
+      "step": 146
+    },
+    {
+      "epoch": 0.04704,
+      "grad_norm": 0.3631345331668854,
+      "learning_rate": 0.00019548438751000802,
+      "loss": 0.7218,
+      "step": 147
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 0.38313111662864685,
+      "learning_rate": 0.00019545236188951161,
+      "loss": 0.681,
+      "step": 148
+    },
+    {
+      "epoch": 0.04768,
+      "grad_norm": 0.32524222135543823,
+      "learning_rate": 0.0001954203362690152,
+      "loss": 0.6253,
+      "step": 149
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.3964385688304901,
+      "learning_rate": 0.00019538831064851882,
+      "loss": 0.7047,
+      "step": 150
+    },
+    {
+      "epoch": 0.04832,
+      "grad_norm": 0.4611073136329651,
+      "learning_rate": 0.00019535628502802244,
+      "loss": 0.7138,
+      "step": 151
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 0.4682482182979584,
+      "learning_rate": 0.00019532425940752603,
+      "loss": 0.7974,
+      "step": 152
+    },
+    {
+      "epoch": 0.04896,
+      "grad_norm": 0.4045487940311432,
+      "learning_rate": 0.00019529223378702963,
+      "loss": 0.8508,
+      "step": 153
+    },
+    {
+      "epoch": 0.04928,
+      "grad_norm": 0.4062901437282562,
+      "learning_rate": 0.00019526020816653324,
+      "loss": 0.8608,
+      "step": 154
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 0.36533573269844055,
+      "learning_rate": 0.00019522818254603683,
+      "loss": 0.8073,
+      "step": 155
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 0.39547574520111084,
+      "learning_rate": 0.00019519615692554045,
+      "loss": 0.7557,
+      "step": 156
+    },
+    {
+      "epoch": 0.05024,
+      "grad_norm": 0.38503214716911316,
+      "learning_rate": 0.00019516413130504404,
+      "loss": 0.6514,
+      "step": 157
+    },
+    {
+      "epoch": 0.05056,
+      "grad_norm": 0.42307570576667786,
+      "learning_rate": 0.00019513210568454766,
+      "loss": 0.8251,
+      "step": 158
+    },
+    {
+      "epoch": 0.05088,
+      "grad_norm": 0.41058966517448425,
+      "learning_rate": 0.00019510008006405125,
+      "loss": 0.6387,
+      "step": 159
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.3361259400844574,
+      "learning_rate": 0.00019506805444355484,
+      "loss": 0.6494,
+      "step": 160
+    },
+    {
+      "epoch": 0.05152,
+      "grad_norm": 0.41206595301628113,
+      "learning_rate": 0.00019503602882305844,
+      "loss": 0.8764,
+      "step": 161
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 0.4369961619377136,
+      "learning_rate": 0.00019500400320256208,
+      "loss": 0.6985,
+      "step": 162
+    },
+    {
+      "epoch": 0.05216,
+      "grad_norm": 0.36770498752593994,
+      "learning_rate": 0.00019497197758206567,
+      "loss": 0.7004,
+      "step": 163
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 0.4144120514392853,
+      "learning_rate": 0.00019493995196156926,
+      "loss": 0.6527,
+      "step": 164
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 0.46151959896087646,
+      "learning_rate": 0.00019490792634107286,
+      "loss": 0.7693,
+      "step": 165
+    },
+    {
+      "epoch": 0.05312,
+      "grad_norm": 0.35259807109832764,
+      "learning_rate": 0.00019487590072057647,
+      "loss": 0.5998,
+      "step": 166
+    },
+    {
+      "epoch": 0.05344,
+      "grad_norm": 0.49185067415237427,
+      "learning_rate": 0.00019484387510008006,
+      "loss": 0.8901,
+      "step": 167
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 0.36020904779434204,
+      "learning_rate": 0.00019481184947958368,
+      "loss": 0.6609,
+      "step": 168
+    },
+    {
+      "epoch": 0.05408,
+      "grad_norm": 0.41236624121665955,
+      "learning_rate": 0.00019477982385908727,
+      "loss": 0.6528,
+      "step": 169
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 0.39236316084861755,
+      "learning_rate": 0.0001947477982385909,
+      "loss": 0.7555,
+      "step": 170
+    },
+    {
+      "epoch": 0.05472,
+      "grad_norm": 0.3657910227775574,
+      "learning_rate": 0.00019471577261809448,
+      "loss": 0.7651,
+      "step": 171
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 0.43941134214401245,
+      "learning_rate": 0.00019468374699759807,
+      "loss": 0.7719,
+      "step": 172
+    },
+    {
+      "epoch": 0.05536,
+      "grad_norm": 0.4226665198802948,
+      "learning_rate": 0.0001946517213771017,
+      "loss": 0.7059,
+      "step": 173
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 0.3848075866699219,
+      "learning_rate": 0.0001946196957566053,
+      "loss": 0.7856,
+      "step": 174
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.4344492256641388,
+      "learning_rate": 0.0001945876701361089,
+      "loss": 0.7139,
+      "step": 175
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 0.4004977345466614,
+      "learning_rate": 0.0001945556445156125,
+      "loss": 0.7231,
+      "step": 176
+    },
+    {
+      "epoch": 0.05664,
+      "grad_norm": 0.3959861099720001,
+      "learning_rate": 0.0001945236188951161,
+      "loss": 0.6548,
+      "step": 177
+    },
+    {
+      "epoch": 0.05696,
+      "grad_norm": 0.4344013035297394,
+      "learning_rate": 0.0001944915932746197,
+      "loss": 0.8175,
+      "step": 178
+    },
+    {
+      "epoch": 0.05728,
+      "grad_norm": 0.4032697379589081,
+      "learning_rate": 0.0001944595676541233,
+      "loss": 0.8014,
+      "step": 179
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.43058979511260986,
+      "learning_rate": 0.0001944275420336269,
+      "loss": 0.8733,
+      "step": 180
+    },
+    {
+      "epoch": 0.05792,
+      "grad_norm": 0.3620558977127075,
+      "learning_rate": 0.00019439551641313053,
+      "loss": 0.7407,
+      "step": 181
+    },
+    {
+      "epoch": 0.05824,
+      "grad_norm": 0.3888520300388336,
+      "learning_rate": 0.00019436349079263412,
+      "loss": 0.6831,
+      "step": 182
+    },
+    {
+      "epoch": 0.05856,
+      "grad_norm": 0.41351792216300964,
+      "learning_rate": 0.00019433146517213771,
+      "loss": 0.7086,
+      "step": 183
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 0.41693538427352905,
+      "learning_rate": 0.0001942994395516413,
+      "loss": 0.7457,
+      "step": 184
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 0.3726726472377777,
+      "learning_rate": 0.00019426741393114492,
+      "loss": 0.7311,
+      "step": 185
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 0.45244935154914856,
+      "learning_rate": 0.00019423538831064854,
+      "loss": 0.9,
+      "step": 186
+    },
+    {
+      "epoch": 0.05984,
+      "grad_norm": 0.4647224545478821,
+      "learning_rate": 0.00019420336269015213,
+      "loss": 0.7879,
+      "step": 187
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 0.407377153635025,
+      "learning_rate": 0.00019417133706965572,
+      "loss": 0.6731,
+      "step": 188
+    },
+    {
+      "epoch": 0.06048,
+      "grad_norm": 0.4266600012779236,
+      "learning_rate": 0.00019413931144915934,
+      "loss": 0.7934,
+      "step": 189
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 0.41341936588287354,
+      "learning_rate": 0.00019410728582866293,
+      "loss": 0.9278,
+      "step": 190
+    },
+    {
+      "epoch": 0.06112,
+      "grad_norm": 0.3956766724586487,
+      "learning_rate": 0.00019407526020816652,
+      "loss": 0.7756,
+      "step": 191
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 0.4248698353767395,
+      "learning_rate": 0.00019404323458767014,
+      "loss": 0.7697,
+      "step": 192
+    },
+    {
+      "epoch": 0.06176,
+      "grad_norm": 0.3743106722831726,
+      "learning_rate": 0.00019401120896717376,
+      "loss": 0.7578,
+      "step": 193
+    },
+    {
+      "epoch": 0.06208,
+      "grad_norm": 0.4382505714893341,
+      "learning_rate": 0.00019397918334667735,
+      "loss": 0.7677,
+      "step": 194
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 0.36778193712234497,
+      "learning_rate": 0.00019394715772618094,
+      "loss": 0.7313,
+      "step": 195
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 0.40140220522880554,
+      "learning_rate": 0.00019391513210568456,
+      "loss": 0.6793,
+      "step": 196
+    },
+    {
+      "epoch": 0.06304,
+      "grad_norm": 0.39271122217178345,
+      "learning_rate": 0.00019388310648518818,
+      "loss": 0.7619,
+      "step": 197
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 0.4570521116256714,
+      "learning_rate": 0.00019385108086469177,
+      "loss": 0.8262,
+      "step": 198
+    },
+    {
+      "epoch": 0.06368,
+      "grad_norm": 0.47575289011001587,
+      "learning_rate": 0.00019381905524419536,
+      "loss": 0.7959,
+      "step": 199
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.42022374272346497,
+      "learning_rate": 0.00019378702962369898,
+      "loss": 0.8002,
+      "step": 200
+    },
+    {
+      "epoch": 0.06432,
+      "grad_norm": 0.4458564221858978,
+      "learning_rate": 0.00019375500400320257,
+      "loss": 0.7746,
+      "step": 201
+    },
+    {
+      "epoch": 0.06464,
+      "grad_norm": 0.43326252698898315,
+      "learning_rate": 0.00019372297838270616,
+      "loss": 0.8121,
+      "step": 202
+    },
+    {
+      "epoch": 0.06496,
+      "grad_norm": 0.4563269317150116,
+      "learning_rate": 0.00019369095276220978,
+      "loss": 0.6228,
+      "step": 203
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 0.3853350579738617,
+      "learning_rate": 0.00019365892714171337,
+      "loss": 0.7215,
+      "step": 204
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 0.4247748851776123,
+      "learning_rate": 0.000193626901521217,
+      "loss": 0.899,
+      "step": 205
+    },
+    {
+      "epoch": 0.06592,
+      "grad_norm": 0.4751046895980835,
+      "learning_rate": 0.00019359487590072058,
+      "loss": 0.8891,
+      "step": 206
+    },
+    {
+      "epoch": 0.06624,
+      "grad_norm": 0.49774014949798584,
+      "learning_rate": 0.00019356285028022417,
+      "loss": 1.0684,
+      "step": 207
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 0.3435848653316498,
+      "learning_rate": 0.0001935308246597278,
+      "loss": 0.5443,
+      "step": 208
+    },
+    {
+      "epoch": 0.06688,
+      "grad_norm": 0.34130939841270447,
+      "learning_rate": 0.0001934987990392314,
+      "loss": 0.7378,
+      "step": 209
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.5001709461212158,
+      "learning_rate": 0.000193466773418735,
+      "loss": 0.7202,
+      "step": 210
+    },
+    {
+      "epoch": 0.06752,
+      "grad_norm": 0.4151903986930847,
+      "learning_rate": 0.0001934347477982386,
+      "loss": 0.7504,
+      "step": 211
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 0.40069901943206787,
+      "learning_rate": 0.0001934027221777422,
+      "loss": 0.7668,
+      "step": 212
+    },
+    {
+      "epoch": 0.06816,
+      "grad_norm": 0.33539754152297974,
+      "learning_rate": 0.0001933706965572458,
+      "loss": 0.7219,
+      "step": 213
+    },
+    {
+      "epoch": 0.06848,
+      "grad_norm": 0.35829442739486694,
+      "learning_rate": 0.0001933386709367494,
+      "loss": 0.6871,
+      "step": 214
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 0.4014899730682373,
+      "learning_rate": 0.000193306645316253,
+      "loss": 0.6597,
+      "step": 215
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 0.42054837942123413,
+      "learning_rate": 0.00019327461969575663,
+      "loss": 0.6932,
+      "step": 216
+    },
+    {
+      "epoch": 0.06944,
+      "grad_norm": 0.38730472326278687,
+      "learning_rate": 0.00019324259407526022,
+      "loss": 0.7551,
+      "step": 217
+    },
+    {
+      "epoch": 0.06976,
+      "grad_norm": 0.42143189907073975,
+      "learning_rate": 0.0001932105684547638,
+      "loss": 0.9246,
+      "step": 218
+    },
+    {
+      "epoch": 0.07008,
+      "grad_norm": 0.4313777983188629,
+      "learning_rate": 0.00019317854283426743,
+      "loss": 0.8778,
+      "step": 219
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.37183213233947754,
+      "learning_rate": 0.00019314651721377102,
+      "loss": 0.6489,
+      "step": 220
+    },
+    {
+      "epoch": 0.07072,
+      "grad_norm": 0.41780033707618713,
+      "learning_rate": 0.00019311449159327464,
+      "loss": 0.7262,
+      "step": 221
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 0.3593262732028961,
+      "learning_rate": 0.00019308246597277823,
+      "loss": 0.5552,
+      "step": 222
+    },
+    {
+      "epoch": 0.07136,
+      "grad_norm": 0.3701655864715576,
+      "learning_rate": 0.00019305044035228182,
+      "loss": 0.6861,
+      "step": 223
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 0.45401251316070557,
+      "learning_rate": 0.00019301841473178544,
+      "loss": 0.776,
+      "step": 224
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.4911825656890869,
+      "learning_rate": 0.00019298638911128903,
+      "loss": 0.7652,
+      "step": 225
+    },
+    {
+      "epoch": 0.07232,
+      "grad_norm": 0.47178205847740173,
+      "learning_rate": 0.00019295436349079262,
+      "loss": 0.7977,
+      "step": 226
+    },
+    {
+      "epoch": 0.07264,
+      "grad_norm": 0.43735066056251526,
+      "learning_rate": 0.00019292233787029624,
+      "loss": 0.698,
+      "step": 227
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 0.37195974588394165,
+      "learning_rate": 0.00019289031224979986,
+      "loss": 0.6043,
+      "step": 228
+    },
+    {
+      "epoch": 0.07328,
+      "grad_norm": 0.42763856053352356,
+      "learning_rate": 0.00019285828662930345,
+      "loss": 0.8082,
+      "step": 229
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.4227966070175171,
+      "learning_rate": 0.00019282626100880704,
+      "loss": 0.8358,
+      "step": 230
+    },
+    {
+      "epoch": 0.07392,
+      "grad_norm": 0.4282156825065613,
+      "learning_rate": 0.00019279423538831066,
+      "loss": 0.733,
+      "step": 231
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 0.3971414864063263,
+      "learning_rate": 0.00019276220976781425,
+      "loss": 0.7151,
+      "step": 232
+    },
+    {
+      "epoch": 0.07456,
+      "grad_norm": 0.434416264295578,
+      "learning_rate": 0.00019273018414731787,
+      "loss": 0.7154,
+      "step": 233
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 0.42665278911590576,
+      "learning_rate": 0.00019269815852682146,
+      "loss": 0.7654,
+      "step": 234
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 0.4261840283870697,
+      "learning_rate": 0.00019266613290632508,
+      "loss": 0.7173,
+      "step": 235
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 0.35086163878440857,
+      "learning_rate": 0.00019263410728582867,
+      "loss": 0.6749,
+      "step": 236
+    },
+    {
+      "epoch": 0.07584,
+      "grad_norm": 0.4083746671676636,
+      "learning_rate": 0.00019260208166533226,
+      "loss": 0.8571,
+      "step": 237
+    },
+    {
+      "epoch": 0.07616,
+      "grad_norm": 0.43829548358917236,
+      "learning_rate": 0.00019257005604483585,
+      "loss": 0.5742,
+      "step": 238
+    },
+    {
+      "epoch": 0.07648,
+      "grad_norm": 0.3887094557285309,
+      "learning_rate": 0.0001925380304243395,
+      "loss": 0.7571,
+      "step": 239
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.41847652196884155,
+      "learning_rate": 0.0001925060048038431,
+      "loss": 0.7367,
+      "step": 240
+    },
+    {
+      "epoch": 0.07712,
+      "grad_norm": 0.469431608915329,
+      "learning_rate": 0.00019247397918334668,
+      "loss": 0.8599,
+      "step": 241
+    },
+    {
+      "epoch": 0.07744,
+      "grad_norm": 0.3327113389968872,
+      "learning_rate": 0.00019244195356285027,
+      "loss": 0.6346,
+      "step": 242
+    },
+    {
+      "epoch": 0.07776,
+      "grad_norm": 0.3869839012622833,
+      "learning_rate": 0.0001924099279423539,
+      "loss": 0.6673,
+      "step": 243
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 0.44176167249679565,
+      "learning_rate": 0.0001923779023218575,
+      "loss": 0.8653,
+      "step": 244
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 0.36183568835258484,
+      "learning_rate": 0.0001923458767013611,
+      "loss": 0.6873,
+      "step": 245
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 0.47324663400650024,
+      "learning_rate": 0.0001923138510808647,
+      "loss": 0.7747,
+      "step": 246
+    },
+    {
+      "epoch": 0.07904,
+      "grad_norm": 0.5251195430755615,
+      "learning_rate": 0.0001922818254603683,
+      "loss": 0.8511,
+      "step": 247
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 0.37095338106155396,
+      "learning_rate": 0.0001922497998398719,
+      "loss": 0.6834,
+      "step": 248
+    },
+    {
+      "epoch": 0.07968,
+      "grad_norm": 0.3895617425441742,
+      "learning_rate": 0.0001922177742193755,
+      "loss": 0.6648,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4422973096370697,
+      "learning_rate": 0.0001921857485988791,
+      "loss": 0.7937,
+      "step": 250
+    },
+    {
+      "epoch": 0.08032,
+      "grad_norm": 0.42825403809547424,
+      "learning_rate": 0.00019215372297838273,
+      "loss": 0.706,
+      "step": 251
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 0.43490323424339294,
+      "learning_rate": 0.00019212169735788632,
+      "loss": 0.8515,
+      "step": 252
+    },
+    {
+      "epoch": 0.08096,
+      "grad_norm": 0.3549322187900543,
+      "learning_rate": 0.0001920896717373899,
+      "loss": 0.6624,
+      "step": 253
+    },
+    {
+      "epoch": 0.08128,
+      "grad_norm": 0.37234827876091003,
+      "learning_rate": 0.00019205764611689353,
+      "loss": 0.6415,
+      "step": 254
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 0.4193861186504364,
+      "learning_rate": 0.00019202562049639712,
+      "loss": 0.8683,
+      "step": 255
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 0.42537981271743774,
+      "learning_rate": 0.00019199359487590074,
+      "loss": 0.7195,
+      "step": 256
+    },
+    {
+      "epoch": 0.08224,
+      "grad_norm": 0.401801735162735,
+      "learning_rate": 0.00019196156925540433,
+      "loss": 0.7611,
+      "step": 257
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 0.4125421643257141,
+      "learning_rate": 0.00019192954363490795,
+      "loss": 0.6725,
+      "step": 258
+    },
+    {
+      "epoch": 0.08288,
+      "grad_norm": 0.37331387400627136,
+      "learning_rate": 0.00019189751801441154,
+      "loss": 0.5354,
+      "step": 259
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.3985353708267212,
+      "learning_rate": 0.00019186549239391513,
+      "loss": 0.7184,
+      "step": 260
+    },
+    {
+      "epoch": 0.08352,
+      "grad_norm": 0.4259369671344757,
+      "learning_rate": 0.00019183346677341872,
+      "loss": 0.7663,
+      "step": 261
+    },
+    {
+      "epoch": 0.08384,
+      "grad_norm": 0.3976237177848816,
+      "learning_rate": 0.00019180144115292237,
+      "loss": 0.8307,
+      "step": 262
+    },
+    {
+      "epoch": 0.08416,
+      "grad_norm": 0.4754771888256073,
+      "learning_rate": 0.00019176941553242596,
+      "loss": 0.8227,
+      "step": 263
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 0.3591161370277405,
+      "learning_rate": 0.00019173738991192955,
+      "loss": 0.7094,
+      "step": 264
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.3920748233795166,
+      "learning_rate": 0.00019170536429143314,
+      "loss": 0.755,
+      "step": 265
+    },
+    {
+      "epoch": 0.08512,
+      "grad_norm": 0.334701806306839,
+      "learning_rate": 0.00019167333867093676,
+      "loss": 0.7176,
+      "step": 266
+    },
+    {
+      "epoch": 0.08544,
+      "grad_norm": 0.5008222460746765,
+      "learning_rate": 0.00019164131305044035,
+      "loss": 0.7401,
+      "step": 267
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 0.41274645924568176,
+      "learning_rate": 0.00019160928742994397,
+      "loss": 0.7548,
+      "step": 268
+    },
+    {
+      "epoch": 0.08608,
+      "grad_norm": 0.41964104771614075,
+      "learning_rate": 0.00019157726180944756,
+      "loss": 0.8061,
+      "step": 269
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.3484065532684326,
+      "learning_rate": 0.00019154523618895118,
+      "loss": 0.6838,
+      "step": 270
+    },
+    {
+      "epoch": 0.08672,
+      "grad_norm": 0.3875018060207367,
+      "learning_rate": 0.00019151321056845477,
+      "loss": 0.656,
+      "step": 271
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 0.4955156147480011,
+      "learning_rate": 0.00019148118494795836,
+      "loss": 0.8905,
+      "step": 272
+    },
+    {
+      "epoch": 0.08736,
+      "grad_norm": 0.38761159777641296,
+      "learning_rate": 0.00019144915932746198,
+      "loss": 0.7707,
+      "step": 273
+    },
+    {
+      "epoch": 0.08768,
+      "grad_norm": 0.42772185802459717,
+      "learning_rate": 0.0001914171337069656,
+      "loss": 0.8844,
+      "step": 274
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.351605623960495,
+      "learning_rate": 0.0001913851080864692,
+      "loss": 0.649,
+      "step": 275
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 0.36860591173171997,
+      "learning_rate": 0.00019135308246597278,
+      "loss": 0.7766,
+      "step": 276
+    },
+    {
+      "epoch": 0.08864,
+      "grad_norm": 0.5082221031188965,
+      "learning_rate": 0.0001913210568454764,
+      "loss": 0.8105,
+      "step": 277
+    },
+    {
+      "epoch": 0.08896,
+      "grad_norm": 0.4578067362308502,
+      "learning_rate": 0.00019128903122498,
+      "loss": 0.7442,
+      "step": 278
+    },
+    {
+      "epoch": 0.08928,
+      "grad_norm": 0.39385655522346497,
+      "learning_rate": 0.00019125700560448358,
+      "loss": 0.5852,
+      "step": 279
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.43078911304473877,
+      "learning_rate": 0.0001912249799839872,
+      "loss": 0.7648,
+      "step": 280
+    },
+    {
+      "epoch": 0.08992,
+      "grad_norm": 0.3965681195259094,
+      "learning_rate": 0.00019119295436349082,
+      "loss": 0.5901,
+      "step": 281
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 0.3836049437522888,
+      "learning_rate": 0.0001911609287429944,
+      "loss": 0.6863,
+      "step": 282
+    },
+    {
+      "epoch": 0.09056,
+      "grad_norm": 0.4074888527393341,
+      "learning_rate": 0.000191128903122498,
+      "loss": 0.7356,
+      "step": 283
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 0.4345192611217499,
+      "learning_rate": 0.0001910968775020016,
+      "loss": 0.6797,
+      "step": 284
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.4084523916244507,
+      "learning_rate": 0.0001910648518815052,
+      "loss": 0.7159,
+      "step": 285
+    },
+    {
+      "epoch": 0.09152,
+      "grad_norm": 0.39732807874679565,
+      "learning_rate": 0.00019103282626100883,
+      "loss": 0.75,
+      "step": 286
+    },
+    {
+      "epoch": 0.09184,
+      "grad_norm": 0.400614470243454,
+      "learning_rate": 0.00019100080064051242,
+      "loss": 0.8744,
+      "step": 287
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 0.4194415211677551,
+      "learning_rate": 0.000190968775020016,
+      "loss": 0.7278,
+      "step": 288
+    },
+    {
+      "epoch": 0.09248,
+      "grad_norm": 0.3465976417064667,
+      "learning_rate": 0.00019093674939951963,
+      "loss": 0.5832,
+      "step": 289
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.43152135610580444,
+      "learning_rate": 0.00019090472377902322,
+      "loss": 0.7302,
+      "step": 290
+    },
+    {
+      "epoch": 0.09312,
+      "grad_norm": 0.38952887058258057,
+      "learning_rate": 0.00019087269815852684,
+      "loss": 0.7565,
+      "step": 291
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 0.3788875639438629,
+      "learning_rate": 0.00019084067253803043,
+      "loss": 0.6957,
+      "step": 292
+    },
+    {
+      "epoch": 0.09376,
+      "grad_norm": 0.3822607696056366,
+      "learning_rate": 0.00019080864691753405,
+      "loss": 0.7281,
+      "step": 293
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 0.39809250831604004,
+      "learning_rate": 0.00019077662129703764,
+      "loss": 0.6403,
+      "step": 294
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.5097452998161316,
+      "learning_rate": 0.00019074459567654123,
+      "loss": 0.946,
+      "step": 295
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 0.4602653682231903,
+      "learning_rate": 0.00019071257005604485,
+      "loss": 0.7914,
+      "step": 296
+    },
+    {
+      "epoch": 0.09504,
+      "grad_norm": 0.3633464574813843,
+      "learning_rate": 0.00019068054443554847,
+      "loss": 0.7137,
+      "step": 297
+    },
+    {
+      "epoch": 0.09536,
+      "grad_norm": 0.4071849584579468,
+      "learning_rate": 0.00019064851881505206,
+      "loss": 0.6941,
+      "step": 298
+    },
+    {
+      "epoch": 0.09568,
+      "grad_norm": 0.42280247807502747,
+      "learning_rate": 0.00019061649319455565,
+      "loss": 0.7827,
+      "step": 299
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.48201918601989746,
+      "learning_rate": 0.00019058446757405927,
+      "loss": 0.7608,
+      "step": 300
+    },
+    {
+      "epoch": 0.09632,
+      "grad_norm": 0.38005709648132324,
+      "learning_rate": 0.00019055244195356286,
+      "loss": 0.8168,
+      "step": 301
+    },
+    {
+      "epoch": 0.09664,
+      "grad_norm": 0.3695196509361267,
+      "learning_rate": 0.00019052041633306645,
+      "loss": 0.7313,
+      "step": 302
+    },
+    {
+      "epoch": 0.09696,
+      "grad_norm": 0.4019639790058136,
+      "learning_rate": 0.00019048839071257007,
+      "loss": 0.7529,
+      "step": 303
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 0.44755351543426514,
+      "learning_rate": 0.0001904563650920737,
+      "loss": 0.6869,
+      "step": 304
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 0.3883623778820038,
+      "learning_rate": 0.00019042433947157728,
+      "loss": 0.6606,
+      "step": 305
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 0.3925836682319641,
+      "learning_rate": 0.00019039231385108087,
+      "loss": 0.7101,
+      "step": 306
+    },
+    {
+      "epoch": 0.09824,
+      "grad_norm": 0.4106467366218567,
+      "learning_rate": 0.00019036028823058446,
+      "loss": 0.7936,
+      "step": 307
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 0.37678349018096924,
+      "learning_rate": 0.00019032826261008808,
+      "loss": 0.6211,
+      "step": 308
+    },
+    {
+      "epoch": 0.09888,
+      "grad_norm": 0.35536712408065796,
+      "learning_rate": 0.0001902962369895917,
+      "loss": 0.5956,
+      "step": 309
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.3602486550807953,
+      "learning_rate": 0.0001902642113690953,
+      "loss": 0.644,
+      "step": 310
+    },
+    {
+      "epoch": 0.09952,
+      "grad_norm": 0.48227840662002563,
+      "learning_rate": 0.00019023218574859888,
+      "loss": 0.9124,
+      "step": 311
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 0.48681333661079407,
+      "learning_rate": 0.0001902001601281025,
+      "loss": 0.6817,
+      "step": 312
+    },
+    {
+      "epoch": 0.10016,
+      "grad_norm": 0.5185659527778625,
+      "learning_rate": 0.0001901681345076061,
+      "loss": 0.8696,
+      "step": 313
+    },
+    {
+      "epoch": 0.10048,
+      "grad_norm": 0.45130884647369385,
+      "learning_rate": 0.00019013610888710968,
+      "loss": 0.6158,
+      "step": 314
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.3998804986476898,
+      "learning_rate": 0.0001901040832666133,
+      "loss": 0.7709,
+      "step": 315
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 0.40330415964126587,
+      "learning_rate": 0.00019007205764611692,
+      "loss": 0.8208,
+      "step": 316
+    },
+    {
+      "epoch": 0.10144,
+      "grad_norm": 0.46281254291534424,
+      "learning_rate": 0.0001900400320256205,
+      "loss": 0.7066,
+      "step": 317
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 0.460854709148407,
+      "learning_rate": 0.0001900080064051241,
+      "loss": 0.7087,
+      "step": 318
+    },
+    {
+      "epoch": 0.10208,
+      "grad_norm": 0.366912305355072,
+      "learning_rate": 0.00018997598078462772,
+      "loss": 0.7062,
+      "step": 319
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.4225977957248688,
+      "learning_rate": 0.0001899439551641313,
+      "loss": 0.6236,
+      "step": 320
+    },
+    {
+      "epoch": 0.10272,
+      "grad_norm": 0.5245331525802612,
+      "learning_rate": 0.00018991192954363493,
+      "loss": 0.7304,
+      "step": 321
+    },
+    {
+      "epoch": 0.10304,
+      "grad_norm": 0.33363965153694153,
+      "learning_rate": 0.00018987990392313852,
+      "loss": 0.7375,
+      "step": 322
+    },
+    {
+      "epoch": 0.10336,
+      "grad_norm": 0.48624977469444275,
+      "learning_rate": 0.00018984787830264214,
+      "loss": 0.8487,
+      "step": 323
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 0.447396457195282,
+      "learning_rate": 0.00018981585268214573,
+      "loss": 0.712,
+      "step": 324
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.33740419149398804,
+      "learning_rate": 0.00018978382706164932,
+      "loss": 0.5979,
+      "step": 325
+    },
+    {
+      "epoch": 0.10432,
+      "grad_norm": 0.36664626002311707,
+      "learning_rate": 0.0001897518014411529,
+      "loss": 0.6855,
+      "step": 326
+    },
+    {
+      "epoch": 0.10464,
+      "grad_norm": 0.35455721616744995,
+      "learning_rate": 0.00018971977582065656,
+      "loss": 0.6239,
+      "step": 327
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 0.4646221697330475,
+      "learning_rate": 0.00018968775020016015,
+      "loss": 0.793,
+      "step": 328
+    },
+    {
+      "epoch": 0.10528,
+      "grad_norm": 0.4199264645576477,
+      "learning_rate": 0.00018965572457966374,
+      "loss": 0.7491,
+      "step": 329
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.4116901457309723,
+      "learning_rate": 0.00018962369895916733,
+      "loss": 0.6916,
+      "step": 330
+    },
+    {
+      "epoch": 0.10592,
+      "grad_norm": 0.3955143690109253,
+      "learning_rate": 0.00018959167333867095,
+      "loss": 0.6894,
+      "step": 331
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 0.40208229422569275,
+      "learning_rate": 0.00018955964771817454,
+      "loss": 0.7551,
+      "step": 332
+    },
+    {
+      "epoch": 0.10656,
+      "grad_norm": 0.41835013031959534,
+      "learning_rate": 0.00018952762209767816,
+      "loss": 0.7026,
+      "step": 333
+    },
+    {
+      "epoch": 0.10688,
+      "grad_norm": 0.3174933195114136,
+      "learning_rate": 0.00018949559647718175,
+      "loss": 0.705,
+      "step": 334
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.38186565041542053,
+      "learning_rate": 0.00018946357085668537,
+      "loss": 0.7336,
+      "step": 335
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 0.4213207960128784,
+      "learning_rate": 0.00018943154523618896,
+      "loss": 0.7498,
+      "step": 336
+    },
+    {
+      "epoch": 0.10784,
+      "grad_norm": 0.46221476793289185,
+      "learning_rate": 0.00018939951961569255,
+      "loss": 0.7456,
+      "step": 337
+    },
+    {
+      "epoch": 0.10816,
+      "grad_norm": 0.38802433013916016,
+      "learning_rate": 0.00018936749399519617,
+      "loss": 0.7074,
+      "step": 338
+    },
+    {
+      "epoch": 0.10848,
+      "grad_norm": 0.3584643602371216,
+      "learning_rate": 0.00018933546837469979,
+      "loss": 0.7758,
+      "step": 339
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.3827560245990753,
+      "learning_rate": 0.00018930344275420338,
+      "loss": 0.6262,
+      "step": 340
+    },
+    {
+      "epoch": 0.10912,
+      "grad_norm": 0.4872991442680359,
+      "learning_rate": 0.00018927141713370697,
+      "loss": 0.77,
+      "step": 341
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 0.41778433322906494,
+      "learning_rate": 0.00018923939151321059,
+      "loss": 0.7125,
+      "step": 342
+    },
+    {
+      "epoch": 0.10976,
+      "grad_norm": 0.3829624056816101,
+      "learning_rate": 0.00018920736589271418,
+      "loss": 0.675,
+      "step": 343
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 0.3227456510066986,
+      "learning_rate": 0.0001891753402722178,
+      "loss": 0.5629,
+      "step": 344
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 0.3489581346511841,
+      "learning_rate": 0.0001891433146517214,
+      "loss": 0.6324,
+      "step": 345
+    },
+    {
+      "epoch": 0.11072,
+      "grad_norm": 0.3462286591529846,
+      "learning_rate": 0.000189111289031225,
+      "loss": 0.5893,
+      "step": 346
+    },
+    {
+      "epoch": 0.11104,
+      "grad_norm": 0.3770864009857178,
+      "learning_rate": 0.0001890792634107286,
+      "loss": 0.5736,
+      "step": 347
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 0.3786970376968384,
+      "learning_rate": 0.0001890472377902322,
+      "loss": 0.7683,
+      "step": 348
+    },
+    {
+      "epoch": 0.11168,
+      "grad_norm": 0.34881076216697693,
+      "learning_rate": 0.00018901521216973578,
+      "loss": 0.7193,
+      "step": 349
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.4564970135688782,
+      "learning_rate": 0.0001889831865492394,
+      "loss": 0.7069,
+      "step": 350
+    },
+    {
+      "epoch": 0.11232,
+      "grad_norm": 0.49901261925697327,
+      "learning_rate": 0.00018895116092874302,
+      "loss": 0.757,
+      "step": 351
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 0.3927716910839081,
+      "learning_rate": 0.0001889191353082466,
+      "loss": 0.8099,
+      "step": 352
+    },
+    {
+      "epoch": 0.11296,
+      "grad_norm": 0.3926427364349365,
+      "learning_rate": 0.0001888871096877502,
+      "loss": 0.624,
+      "step": 353
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 0.45321208238601685,
+      "learning_rate": 0.00018885508406725382,
+      "loss": 0.802,
+      "step": 354
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.3618319034576416,
+      "learning_rate": 0.0001888230584467574,
+      "loss": 0.78,
+      "step": 355
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 0.36882704496383667,
+      "learning_rate": 0.00018879103282626103,
+      "loss": 0.6001,
+      "step": 356
+    },
+    {
+      "epoch": 0.11424,
+      "grad_norm": 0.4372122287750244,
+      "learning_rate": 0.00018875900720576462,
+      "loss": 0.7617,
+      "step": 357
+    },
+    {
+      "epoch": 0.11456,
+      "grad_norm": 0.46206310391426086,
+      "learning_rate": 0.00018872698158526824,
+      "loss": 0.8061,
+      "step": 358
+    },
+    {
+      "epoch": 0.11488,
+      "grad_norm": 0.37309616804122925,
+      "learning_rate": 0.00018869495596477183,
+      "loss": 0.5836,
+      "step": 359
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.3839647173881531,
+      "learning_rate": 0.00018866293034427542,
+      "loss": 0.708,
+      "step": 360
+    },
+    {
+      "epoch": 0.11552,
+      "grad_norm": 0.351663202047348,
+      "learning_rate": 0.00018863090472377904,
+      "loss": 0.6594,
+      "step": 361
+    },
+    {
+      "epoch": 0.11584,
+      "grad_norm": 0.4527769982814789,
+      "learning_rate": 0.00018859887910328265,
+      "loss": 0.9476,
+      "step": 362
+    },
+    {
+      "epoch": 0.11616,
+      "grad_norm": 0.4207096993923187,
+      "learning_rate": 0.00018856685348278625,
+      "loss": 0.7907,
+      "step": 363
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 0.43379828333854675,
+      "learning_rate": 0.00018853482786228984,
+      "loss": 0.7314,
+      "step": 364
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.36828315258026123,
+      "learning_rate": 0.00018850280224179343,
+      "loss": 0.7741,
+      "step": 365
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 0.4496101438999176,
+      "learning_rate": 0.00018847077662129705,
+      "loss": 0.6492,
+      "step": 366
+    },
+    {
+      "epoch": 0.11744,
+      "grad_norm": 0.44155606627464294,
+      "learning_rate": 0.00018843875100080064,
+      "loss": 0.8095,
+      "step": 367
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 0.4319384694099426,
+      "learning_rate": 0.00018840672538030426,
+      "loss": 0.8047,
+      "step": 368
+    },
+    {
+      "epoch": 0.11808,
+      "grad_norm": 0.3407670259475708,
+      "learning_rate": 0.00018837469975980785,
+      "loss": 0.7584,
+      "step": 369
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.41879862546920776,
+      "learning_rate": 0.00018834267413931147,
+      "loss": 0.7996,
+      "step": 370
+    },
+    {
+      "epoch": 0.11872,
+      "grad_norm": 0.37883302569389343,
+      "learning_rate": 0.00018831064851881506,
+      "loss": 0.6218,
+      "step": 371
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 0.35176724195480347,
+      "learning_rate": 0.00018827862289831865,
+      "loss": 0.6753,
+      "step": 372
+    },
+    {
+      "epoch": 0.11936,
+      "grad_norm": 0.41552597284317017,
+      "learning_rate": 0.00018824659727782227,
+      "loss": 0.7137,
+      "step": 373
+    },
+    {
+      "epoch": 0.11968,
+      "grad_norm": 0.39809852838516235,
+      "learning_rate": 0.00018821457165732588,
+      "loss": 0.7539,
+      "step": 374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40412524342536926,
+      "learning_rate": 0.00018818254603682948,
+      "loss": 0.6483,
+      "step": 375
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 0.39429572224617004,
+      "learning_rate": 0.00018815052041633307,
+      "loss": 0.5671,
+      "step": 376
+    },
+    {
+      "epoch": 0.12064,
+      "grad_norm": 0.3492451012134552,
+      "learning_rate": 0.00018811849479583669,
+      "loss": 0.6798,
+      "step": 377
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 0.42634814977645874,
+      "learning_rate": 0.00018808646917534028,
+      "loss": 0.9117,
+      "step": 378
+    },
+    {
+      "epoch": 0.12128,
+      "grad_norm": 0.4606453478336334,
+      "learning_rate": 0.00018805444355484387,
+      "loss": 0.8637,
+      "step": 379
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.40053874254226685,
+      "learning_rate": 0.00018802241793434749,
+      "loss": 0.6593,
+      "step": 380
+    },
+    {
+      "epoch": 0.12192,
+      "grad_norm": 0.3930060565471649,
+      "learning_rate": 0.0001879903923138511,
+      "loss": 0.7144,
+      "step": 381
+    },
+    {
+      "epoch": 0.12224,
+      "grad_norm": 0.28633129596710205,
+      "learning_rate": 0.0001879583666933547,
+      "loss": 0.6366,
+      "step": 382
+    },
+    {
+      "epoch": 0.12256,
+      "grad_norm": 0.3295207619667053,
+      "learning_rate": 0.0001879263410728583,
+      "loss": 0.5848,
+      "step": 383
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 0.41972410678863525,
+      "learning_rate": 0.00018789431545236188,
+      "loss": 0.7357,
+      "step": 384
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 0.3544144928455353,
+      "learning_rate": 0.00018786228983186552,
+      "loss": 0.6076,
+      "step": 385
+    },
+    {
+      "epoch": 0.12352,
+      "grad_norm": 0.3797321319580078,
+      "learning_rate": 0.00018783026421136911,
+      "loss": 0.7171,
+      "step": 386
+    },
+    {
+      "epoch": 0.12384,
+      "grad_norm": 0.331788033246994,
+      "learning_rate": 0.0001877982385908727,
+      "loss": 0.6702,
+      "step": 387
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 0.38595008850097656,
+      "learning_rate": 0.0001877662129703763,
+      "loss": 0.6233,
+      "step": 388
+    },
+    {
+      "epoch": 0.12448,
+      "grad_norm": 0.4231555759906769,
+      "learning_rate": 0.00018773418734987992,
+      "loss": 0.814,
+      "step": 389
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.38779687881469727,
+      "learning_rate": 0.0001877021617293835,
+      "loss": 0.6543,
+      "step": 390
+    },
+    {
+      "epoch": 0.12512,
+      "grad_norm": 0.3879774212837219,
+      "learning_rate": 0.00018767013610888712,
+      "loss": 0.7365,
+      "step": 391
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 0.42405441403388977,
+      "learning_rate": 0.00018763811048839072,
+      "loss": 0.6211,
+      "step": 392
+    },
+    {
+      "epoch": 0.12576,
+      "grad_norm": 0.4443453550338745,
+      "learning_rate": 0.00018760608486789433,
+      "loss": 0.8195,
+      "step": 393
+    },
+    {
+      "epoch": 0.12608,
+      "grad_norm": 0.40762630105018616,
+      "learning_rate": 0.00018757405924739793,
+      "loss": 0.6994,
+      "step": 394
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.3774224817752838,
+      "learning_rate": 0.00018754203362690152,
+      "loss": 0.5753,
+      "step": 395
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 0.4013447165489197,
+      "learning_rate": 0.00018751000800640514,
+      "loss": 0.5917,
+      "step": 396
+    },
+    {
+      "epoch": 0.12704,
+      "grad_norm": 0.4318583011627197,
+      "learning_rate": 0.00018747798238590875,
+      "loss": 0.6557,
+      "step": 397
+    },
+    {
+      "epoch": 0.12736,
+      "grad_norm": 0.4096556603908539,
+      "learning_rate": 0.00018744595676541234,
+      "loss": 0.7002,
+      "step": 398
+    },
+    {
+      "epoch": 0.12768,
+      "grad_norm": 0.4336279332637787,
+      "learning_rate": 0.00018741393114491594,
+      "loss": 0.6924,
+      "step": 399
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.34838399291038513,
+      "learning_rate": 0.00018738190552441955,
+      "loss": 0.7256,
+      "step": 400
+    },
+    {
+      "epoch": 0.12832,
+      "grad_norm": 0.36352548003196716,
+      "learning_rate": 0.00018734987990392315,
+      "loss": 0.6303,
+      "step": 401
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 0.41436469554901123,
+      "learning_rate": 0.00018731785428342674,
+      "loss": 0.6532,
+      "step": 402
+    },
+    {
+      "epoch": 0.12896,
+      "grad_norm": 0.42088446021080017,
+      "learning_rate": 0.00018728582866293035,
+      "loss": 0.77,
+      "step": 403
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 0.42452794313430786,
+      "learning_rate": 0.00018725380304243397,
+      "loss": 0.7193,
+      "step": 404
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.41519755125045776,
+      "learning_rate": 0.00018722177742193756,
+      "loss": 0.7186,
+      "step": 405
+    },
+    {
+      "epoch": 0.12992,
+      "grad_norm": 0.4906809329986572,
+      "learning_rate": 0.00018718975180144116,
+      "loss": 0.7669,
+      "step": 406
+    },
+    {
+      "epoch": 0.13024,
+      "grad_norm": 0.34763047099113464,
+      "learning_rate": 0.00018715772618094475,
+      "loss": 0.5662,
+      "step": 407
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 0.41002142429351807,
+      "learning_rate": 0.00018712570056044837,
+      "loss": 0.7189,
+      "step": 408
+    },
+    {
+      "epoch": 0.13088,
+      "grad_norm": 0.4564385712146759,
+      "learning_rate": 0.00018709367493995198,
+      "loss": 0.8816,
+      "step": 409
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.38358399271965027,
+      "learning_rate": 0.00018706164931945557,
+      "loss": 0.717,
+      "step": 410
+    },
+    {
+      "epoch": 0.13152,
+      "grad_norm": 0.5077653527259827,
+      "learning_rate": 0.00018702962369895917,
+      "loss": 0.7117,
+      "step": 411
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 0.43017587065696716,
+      "learning_rate": 0.00018699759807846278,
+      "loss": 0.686,
+      "step": 412
+    },
+    {
+      "epoch": 0.13216,
+      "grad_norm": 0.39272361993789673,
+      "learning_rate": 0.00018696557245796638,
+      "loss": 0.6044,
+      "step": 413
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 0.3632037341594696,
+      "learning_rate": 0.00018693354683746997,
+      "loss": 0.7028,
+      "step": 414
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.3879127502441406,
+      "learning_rate": 0.00018690152121697358,
+      "loss": 0.7928,
+      "step": 415
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 0.3738054931163788,
+      "learning_rate": 0.0001868694955964772,
+      "loss": 0.7079,
+      "step": 416
+    },
+    {
+      "epoch": 0.13344,
+      "grad_norm": 0.3670632541179657,
+      "learning_rate": 0.0001868374699759808,
+      "loss": 0.7639,
+      "step": 417
+    },
+    {
+      "epoch": 0.13376,
+      "grad_norm": 0.3228664994239807,
+      "learning_rate": 0.00018680544435548439,
+      "loss": 0.6451,
+      "step": 418
+    },
+    {
+      "epoch": 0.13408,
+      "grad_norm": 0.3685716688632965,
+      "learning_rate": 0.000186773418734988,
+      "loss": 0.7976,
+      "step": 419
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.4210130572319031,
+      "learning_rate": 0.0001867413931144916,
+      "loss": 0.6935,
+      "step": 420
+    },
+    {
+      "epoch": 0.13472,
+      "grad_norm": 0.3161611258983612,
+      "learning_rate": 0.0001867093674939952,
+      "loss": 0.6559,
+      "step": 421
+    },
+    {
+      "epoch": 0.13504,
+      "grad_norm": 0.4087270200252533,
+      "learning_rate": 0.0001866773418734988,
+      "loss": 0.7544,
+      "step": 422
+    },
+    {
+      "epoch": 0.13536,
+      "grad_norm": 0.3436177372932434,
+      "learning_rate": 0.00018664531625300242,
+      "loss": 0.528,
+      "step": 423
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 0.366940975189209,
+      "learning_rate": 0.00018661329063250601,
+      "loss": 0.7055,
+      "step": 424
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.47276127338409424,
+      "learning_rate": 0.0001865812650120096,
+      "loss": 0.5632,
+      "step": 425
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 0.4034774899482727,
+      "learning_rate": 0.0001865492393915132,
+      "loss": 0.7242,
+      "step": 426
+    },
+    {
+      "epoch": 0.13664,
+      "grad_norm": 0.47990959882736206,
+      "learning_rate": 0.00018651721377101684,
+      "loss": 0.7952,
+      "step": 427
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 0.396960586309433,
+      "learning_rate": 0.00018648518815052043,
+      "loss": 0.597,
+      "step": 428
+    },
+    {
+      "epoch": 0.13728,
+      "grad_norm": 0.3714841604232788,
+      "learning_rate": 0.00018645316253002402,
+      "loss": 0.6362,
+      "step": 429
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.3630102574825287,
+      "learning_rate": 0.00018642113690952762,
+      "loss": 0.586,
+      "step": 430
+    },
+    {
+      "epoch": 0.13792,
+      "grad_norm": 0.3690813183784485,
+      "learning_rate": 0.00018638911128903123,
+      "loss": 0.5274,
+      "step": 431
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 0.3681528568267822,
+      "learning_rate": 0.00018635708566853485,
+      "loss": 0.6862,
+      "step": 432
+    },
+    {
+      "epoch": 0.13856,
+      "grad_norm": 0.43585389852523804,
+      "learning_rate": 0.00018632506004803844,
+      "loss": 0.5954,
+      "step": 433
+    },
+    {
+      "epoch": 0.13888,
+      "grad_norm": 0.3885478675365448,
+      "learning_rate": 0.00018629303442754203,
+      "loss": 0.655,
+      "step": 434
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.4392310380935669,
+      "learning_rate": 0.00018626100880704565,
+      "loss": 0.7234,
+      "step": 435
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 0.5326746106147766,
+      "learning_rate": 0.00018622898318654924,
+      "loss": 0.9059,
+      "step": 436
+    },
+    {
+      "epoch": 0.13984,
+      "grad_norm": 0.39512571692466736,
+      "learning_rate": 0.00018619695756605284,
+      "loss": 0.6498,
+      "step": 437
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 0.38490211963653564,
+      "learning_rate": 0.00018616493194555645,
+      "loss": 0.6592,
+      "step": 438
+    },
+    {
+      "epoch": 0.14048,
+      "grad_norm": 0.4570832848548889,
+      "learning_rate": 0.00018613290632506007,
+      "loss": 0.8697,
+      "step": 439
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.41732949018478394,
+      "learning_rate": 0.00018610088070456366,
+      "loss": 0.7024,
+      "step": 440
+    },
+    {
+      "epoch": 0.14112,
+      "grad_norm": 0.3946099877357483,
+      "learning_rate": 0.00018606885508406725,
+      "loss": 0.7958,
+      "step": 441
+    },
+    {
+      "epoch": 0.14144,
+      "grad_norm": 0.3996339440345764,
+      "learning_rate": 0.00018603682946357087,
+      "loss": 0.7218,
+      "step": 442
+    },
+    {
+      "epoch": 0.14176,
+      "grad_norm": 0.35937708616256714,
+      "learning_rate": 0.00018600480384307446,
+      "loss": 0.6794,
+      "step": 443
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 0.44109907746315,
+      "learning_rate": 0.00018597277822257808,
+      "loss": 0.7717,
+      "step": 444
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 0.3278100788593292,
+      "learning_rate": 0.00018594075260208167,
+      "loss": 0.5815,
+      "step": 445
+    },
+    {
+      "epoch": 0.14272,
+      "grad_norm": 0.38432416319847107,
+      "learning_rate": 0.0001859087269815853,
+      "loss": 0.7687,
+      "step": 446
+    },
+    {
+      "epoch": 0.14304,
+      "grad_norm": 0.4637162685394287,
+      "learning_rate": 0.00018587670136108888,
+      "loss": 0.7722,
+      "step": 447
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 0.37541770935058594,
+      "learning_rate": 0.00018584467574059247,
+      "loss": 0.6302,
+      "step": 448
+    },
+    {
+      "epoch": 0.14368,
+      "grad_norm": 0.47312796115875244,
+      "learning_rate": 0.00018581265012009607,
+      "loss": 0.7723,
+      "step": 449
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.35217007994651794,
+      "learning_rate": 0.0001857806244995997,
+      "loss": 0.7025,
+      "step": 450
+    },
+    {
+      "epoch": 0.14432,
+      "grad_norm": 0.4046562910079956,
+      "learning_rate": 0.0001857485988791033,
+      "loss": 0.7592,
+      "step": 451
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 0.34829455614089966,
+      "learning_rate": 0.0001857165732586069,
+      "loss": 0.6055,
+      "step": 452
+    },
+    {
+      "epoch": 0.14496,
+      "grad_norm": 0.4399106502532959,
+      "learning_rate": 0.00018568454763811048,
+      "loss": 0.719,
+      "step": 453
+    },
+    {
+      "epoch": 0.14528,
+      "grad_norm": 0.383394330739975,
+      "learning_rate": 0.0001856525220176141,
+      "loss": 0.7979,
+      "step": 454
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.42221665382385254,
+      "learning_rate": 0.0001856204963971177,
+      "loss": 0.6447,
+      "step": 455
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 0.39467790722846985,
+      "learning_rate": 0.0001855884707766213,
+      "loss": 0.7027,
+      "step": 456
+    },
+    {
+      "epoch": 0.14624,
+      "grad_norm": 0.43809232115745544,
+      "learning_rate": 0.0001855564451561249,
+      "loss": 0.6337,
+      "step": 457
+    },
+    {
+      "epoch": 0.14656,
+      "grad_norm": 0.3936510682106018,
+      "learning_rate": 0.00018552441953562852,
+      "loss": 0.7912,
+      "step": 458
+    },
+    {
+      "epoch": 0.14688,
+      "grad_norm": 0.3876345455646515,
+      "learning_rate": 0.0001854923939151321,
+      "loss": 0.6193,
+      "step": 459
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.36954542994499207,
+      "learning_rate": 0.0001854603682946357,
+      "loss": 0.6132,
+      "step": 460
+    },
+    {
+      "epoch": 0.14752,
+      "grad_norm": 0.34446457028388977,
+      "learning_rate": 0.00018542834267413932,
+      "loss": 0.7239,
+      "step": 461
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 0.44448551535606384,
+      "learning_rate": 0.00018539631705364294,
+      "loss": 0.7335,
+      "step": 462
+    },
+    {
+      "epoch": 0.14816,
+      "grad_norm": 0.38957497477531433,
+      "learning_rate": 0.00018536429143314653,
+      "loss": 0.6479,
+      "step": 463
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 0.4512505829334259,
+      "learning_rate": 0.00018533226581265012,
+      "loss": 0.7041,
+      "step": 464
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.38955575227737427,
+      "learning_rate": 0.00018530024019215374,
+      "loss": 0.9031,
+      "step": 465
+    },
+    {
+      "epoch": 0.14912,
+      "grad_norm": 0.4714851379394531,
+      "learning_rate": 0.00018526821457165733,
+      "loss": 0.6785,
+      "step": 466
+    },
+    {
+      "epoch": 0.14944,
+      "grad_norm": 0.39154791831970215,
+      "learning_rate": 0.00018523618895116092,
+      "loss": 0.6579,
+      "step": 467
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 0.39693886041641235,
+      "learning_rate": 0.00018520416333066454,
+      "loss": 0.6969,
+      "step": 468
+    },
+    {
+      "epoch": 0.15008,
+      "grad_norm": 0.4048503339290619,
+      "learning_rate": 0.00018517213771016816,
+      "loss": 0.7462,
+      "step": 469
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.4949013292789459,
+      "learning_rate": 0.00018514011208967175,
+      "loss": 0.8365,
+      "step": 470
+    },
+    {
+      "epoch": 0.15072,
+      "grad_norm": 0.42741310596466064,
+      "learning_rate": 0.00018510808646917534,
+      "loss": 0.6119,
+      "step": 471
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 0.41830918192863464,
+      "learning_rate": 0.00018507606084867893,
+      "loss": 0.7998,
+      "step": 472
+    },
+    {
+      "epoch": 0.15136,
+      "grad_norm": 0.3437121510505676,
+      "learning_rate": 0.00018504403522818255,
+      "loss": 0.6652,
+      "step": 473
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 0.4198340177536011,
+      "learning_rate": 0.00018501200960768617,
+      "loss": 0.7172,
+      "step": 474
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.43685927987098694,
+      "learning_rate": 0.00018497998398718976,
+      "loss": 0.8308,
+      "step": 475
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 0.34900447726249695,
+      "learning_rate": 0.00018494795836669335,
+      "loss": 0.6435,
+      "step": 476
+    },
+    {
+      "epoch": 0.15264,
+      "grad_norm": 0.3151691257953644,
+      "learning_rate": 0.00018491593274619697,
+      "loss": 0.531,
+      "step": 477
+    },
+    {
+      "epoch": 0.15296,
+      "grad_norm": 0.38561493158340454,
+      "learning_rate": 0.00018488390712570056,
+      "loss": 0.6104,
+      "step": 478
+    },
+    {
+      "epoch": 0.15328,
+      "grad_norm": 0.4135936498641968,
+      "learning_rate": 0.00018485188150520418,
+      "loss": 0.6837,
+      "step": 479
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.4609485864639282,
+      "learning_rate": 0.00018481985588470777,
+      "loss": 0.7747,
+      "step": 480
+    },
+    {
+      "epoch": 0.15392,
+      "grad_norm": 0.3932119905948639,
+      "learning_rate": 0.0001847878302642114,
+      "loss": 0.6516,
+      "step": 481
+    },
+    {
+      "epoch": 0.15424,
+      "grad_norm": 0.4578613340854645,
+      "learning_rate": 0.00018475580464371498,
+      "loss": 0.6174,
+      "step": 482
+    },
+    {
+      "epoch": 0.15456,
+      "grad_norm": 0.37640053033828735,
+      "learning_rate": 0.00018472377902321857,
+      "loss": 0.6821,
+      "step": 483
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 0.45479437708854675,
+      "learning_rate": 0.0001846917534027222,
+      "loss": 0.5377,
+      "step": 484
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.40609532594680786,
+      "learning_rate": 0.0001846597277822258,
+      "loss": 0.6747,
+      "step": 485
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 0.41843003034591675,
+      "learning_rate": 0.0001846277021617294,
+      "loss": 0.7768,
+      "step": 486
+    },
+    {
+      "epoch": 0.15584,
+      "grad_norm": 0.4055987596511841,
+      "learning_rate": 0.000184595676541233,
+      "loss": 0.7311,
+      "step": 487
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 0.33954066038131714,
+      "learning_rate": 0.0001845636509207366,
+      "loss": 0.6543,
+      "step": 488
+    },
+    {
+      "epoch": 0.15648,
+      "grad_norm": 0.422335684299469,
+      "learning_rate": 0.0001845316253002402,
+      "loss": 0.6181,
+      "step": 489
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.4167324900627136,
+      "learning_rate": 0.0001844995996797438,
+      "loss": 0.7401,
+      "step": 490
+    },
+    {
+      "epoch": 0.15712,
+      "grad_norm": 0.4820719063282013,
+      "learning_rate": 0.0001844675740592474,
+      "loss": 0.6522,
+      "step": 491
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 0.36316928267478943,
+      "learning_rate": 0.000184435548438751,
+      "loss": 0.7267,
+      "step": 492
+    },
+    {
+      "epoch": 0.15776,
+      "grad_norm": 0.3780508041381836,
+      "learning_rate": 0.00018440352281825462,
+      "loss": 0.6631,
+      "step": 493
+    },
+    {
+      "epoch": 0.15808,
+      "grad_norm": 0.3649483919143677,
+      "learning_rate": 0.0001843714971977582,
+      "loss": 0.543,
+      "step": 494
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 0.3737383186817169,
+      "learning_rate": 0.0001843394715772618,
+      "loss": 0.7006,
+      "step": 495
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 0.37693876028060913,
+      "learning_rate": 0.00018430744595676542,
+      "loss": 0.7219,
+      "step": 496
+    },
+    {
+      "epoch": 0.15904,
+      "grad_norm": 0.4088519215583801,
+      "learning_rate": 0.00018427542033626904,
+      "loss": 0.6448,
+      "step": 497
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 0.31820815801620483,
+      "learning_rate": 0.00018424339471577263,
+      "loss": 0.5152,
+      "step": 498
+    },
+    {
+      "epoch": 0.15968,
+      "grad_norm": 0.41872063279151917,
+      "learning_rate": 0.00018421136909527622,
+      "loss": 0.8428,
+      "step": 499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4214381277561188,
+      "learning_rate": 0.00018417934347477984,
+      "loss": 0.6745,
+      "step": 500
+    },
+    {
+      "epoch": 0.16032,
+      "grad_norm": 0.4100012481212616,
+      "learning_rate": 0.00018414731785428343,
+      "loss": 0.6849,
+      "step": 501
+    },
+    {
+      "epoch": 0.16064,
+      "grad_norm": 0.42137542366981506,
+      "learning_rate": 0.00018411529223378702,
+      "loss": 0.7824,
+      "step": 502
+    },
+    {
+      "epoch": 0.16096,
+      "grad_norm": 0.3371828496456146,
+      "learning_rate": 0.00018408326661329064,
+      "loss": 0.6419,
+      "step": 503
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 0.3822563588619232,
+      "learning_rate": 0.00018405124099279426,
+      "loss": 0.6689,
+      "step": 504
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.3827621638774872,
+      "learning_rate": 0.00018401921537229785,
+      "loss": 0.6177,
+      "step": 505
+    },
+    {
+      "epoch": 0.16192,
+      "grad_norm": 0.3569611608982086,
+      "learning_rate": 0.00018398718975180144,
+      "loss": 0.6502,
+      "step": 506
+    },
+    {
+      "epoch": 0.16224,
+      "grad_norm": 0.48861172795295715,
+      "learning_rate": 0.00018395516413130503,
+      "loss": 0.771,
+      "step": 507
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 0.4562051296234131,
+      "learning_rate": 0.00018392313851080865,
+      "loss": 0.7734,
+      "step": 508
+    },
+    {
+      "epoch": 0.16288,
+      "grad_norm": 0.38768813014030457,
+      "learning_rate": 0.00018389111289031227,
+      "loss": 0.6714,
+      "step": 509
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.4116150736808777,
+      "learning_rate": 0.00018385908726981586,
+      "loss": 0.6963,
+      "step": 510
+    },
+    {
+      "epoch": 0.16352,
+      "grad_norm": 0.46139904856681824,
+      "learning_rate": 0.00018382706164931945,
+      "loss": 0.7147,
+      "step": 511
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 0.3881680369377136,
+      "learning_rate": 0.00018379503602882307,
+      "loss": 0.8034,
+      "step": 512
+    },
+    {
+      "epoch": 0.16416,
+      "grad_norm": 0.4483790695667267,
+      "learning_rate": 0.00018376301040832666,
+      "loss": 0.7109,
+      "step": 513
+    },
+    {
+      "epoch": 0.16448,
+      "grad_norm": 0.4089036285877228,
+      "learning_rate": 0.00018373098478783025,
+      "loss": 0.6688,
+      "step": 514
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.4461918771266937,
+      "learning_rate": 0.00018369895916733387,
+      "loss": 0.7749,
+      "step": 515
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 0.4111829996109009,
+      "learning_rate": 0.0001836669335468375,
+      "loss": 0.6776,
+      "step": 516
+    },
+    {
+      "epoch": 0.16544,
+      "grad_norm": 0.43297964334487915,
+      "learning_rate": 0.00018363490792634108,
+      "loss": 0.6013,
+      "step": 517
+    },
+    {
+      "epoch": 0.16576,
+      "grad_norm": 0.4114570617675781,
+      "learning_rate": 0.00018360288230584467,
+      "loss": 0.6384,
+      "step": 518
+    },
+    {
+      "epoch": 0.16608,
+      "grad_norm": 0.47909054160118103,
+      "learning_rate": 0.0001835708566853483,
+      "loss": 0.9487,
+      "step": 519
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.36469489336013794,
+      "learning_rate": 0.00018353883106485188,
+      "loss": 0.684,
+      "step": 520
+    },
+    {
+      "epoch": 0.16672,
+      "grad_norm": 0.39657366275787354,
+      "learning_rate": 0.0001835068054443555,
+      "loss": 0.8489,
+      "step": 521
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 0.40450718998908997,
+      "learning_rate": 0.0001834747798238591,
+      "loss": 0.6911,
+      "step": 522
+    },
+    {
+      "epoch": 0.16736,
+      "grad_norm": 0.40527427196502686,
+      "learning_rate": 0.0001834427542033627,
+      "loss": 0.6886,
+      "step": 523
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 0.4055383801460266,
+      "learning_rate": 0.0001834107285828663,
+      "loss": 0.669,
+      "step": 524
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.41727951169013977,
+      "learning_rate": 0.0001833787029623699,
+      "loss": 0.7517,
+      "step": 525
+    },
+    {
+      "epoch": 0.16832,
+      "grad_norm": 0.3840703070163727,
+      "learning_rate": 0.0001833466773418735,
+      "loss": 0.6827,
+      "step": 526
+    },
+    {
+      "epoch": 0.16864,
+      "grad_norm": 0.3988353908061981,
+      "learning_rate": 0.00018331465172137713,
+      "loss": 0.5945,
+      "step": 527
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 0.4323737919330597,
+      "learning_rate": 0.00018328262610088072,
+      "loss": 0.704,
+      "step": 528
+    },
+    {
+      "epoch": 0.16928,
+      "grad_norm": 0.45082369446754456,
+      "learning_rate": 0.0001832506004803843,
+      "loss": 0.6851,
+      "step": 529
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.45386162400245667,
+      "learning_rate": 0.0001832185748598879,
+      "loss": 0.6929,
+      "step": 530
+    },
+    {
+      "epoch": 0.16992,
+      "grad_norm": 0.3560743033885956,
+      "learning_rate": 0.00018318654923939152,
+      "loss": 0.7628,
+      "step": 531
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 0.3851587474346161,
+      "learning_rate": 0.00018315452361889514,
+      "loss": 0.6233,
+      "step": 532
+    },
+    {
+      "epoch": 0.17056,
+      "grad_norm": 0.3453540802001953,
+      "learning_rate": 0.00018312249799839873,
+      "loss": 0.6496,
+      "step": 533
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 0.3629968464374542,
+      "learning_rate": 0.00018309047237790232,
+      "loss": 0.5878,
+      "step": 534
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 0.35973548889160156,
+      "learning_rate": 0.00018305844675740594,
+      "loss": 0.7588,
+      "step": 535
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 0.41032370924949646,
+      "learning_rate": 0.00018302642113690953,
+      "loss": 0.5778,
+      "step": 536
+    },
+    {
+      "epoch": 0.17184,
+      "grad_norm": 0.3912956714630127,
+      "learning_rate": 0.00018299439551641312,
+      "loss": 0.7459,
+      "step": 537
+    },
+    {
+      "epoch": 0.17216,
+      "grad_norm": 0.39061081409454346,
+      "learning_rate": 0.00018296236989591674,
+      "loss": 0.6752,
+      "step": 538
+    },
+    {
+      "epoch": 0.17248,
+      "grad_norm": 0.4324580132961273,
+      "learning_rate": 0.00018293034427542036,
+      "loss": 0.6312,
+      "step": 539
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.36509862542152405,
+      "learning_rate": 0.00018289831865492395,
+      "loss": 0.6054,
+      "step": 540
+    },
+    {
+      "epoch": 0.17312,
+      "grad_norm": 0.30531007051467896,
+      "learning_rate": 0.00018286629303442754,
+      "loss": 0.6236,
+      "step": 541
+    },
+    {
+      "epoch": 0.17344,
+      "grad_norm": 0.3770042061805725,
+      "learning_rate": 0.00018283426741393116,
+      "loss": 0.7658,
+      "step": 542
+    },
+    {
+      "epoch": 0.17376,
+      "grad_norm": 0.5166019201278687,
+      "learning_rate": 0.00018280224179343475,
+      "loss": 0.8123,
+      "step": 543
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 0.37801504135131836,
+      "learning_rate": 0.00018277021617293837,
+      "loss": 0.8322,
+      "step": 544
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 0.38551661372184753,
+      "learning_rate": 0.00018273819055244196,
+      "loss": 0.6054,
+      "step": 545
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 0.3750567138195038,
+      "learning_rate": 0.00018270616493194558,
+      "loss": 0.6804,
+      "step": 546
+    },
+    {
+      "epoch": 0.17504,
+      "grad_norm": 0.4238853454589844,
+      "learning_rate": 0.00018267413931144917,
+      "loss": 0.6559,
+      "step": 547
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 0.3705611228942871,
+      "learning_rate": 0.00018264211369095276,
+      "loss": 0.6832,
+      "step": 548
+    },
+    {
+      "epoch": 0.17568,
+      "grad_norm": 0.38000959157943726,
+      "learning_rate": 0.00018261008807045635,
+      "loss": 0.6127,
+      "step": 549
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.3373851478099823,
+      "learning_rate": 0.00018257806244996,
+      "loss": 0.56,
+      "step": 550
+    },
+    {
+      "epoch": 0.17632,
+      "grad_norm": 0.30364760756492615,
+      "learning_rate": 0.0001825460368294636,
+      "loss": 0.6375,
+      "step": 551
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 0.4089431166648865,
+      "learning_rate": 0.00018251401120896718,
+      "loss": 0.657,
+      "step": 552
+    },
+    {
+      "epoch": 0.17696,
+      "grad_norm": 0.37276798486709595,
+      "learning_rate": 0.00018248198558847077,
+      "loss": 0.7286,
+      "step": 553
+    },
+    {
+      "epoch": 0.17728,
+      "grad_norm": 0.3655564785003662,
+      "learning_rate": 0.0001824499599679744,
+      "loss": 0.6849,
+      "step": 554
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 0.46808362007141113,
+      "learning_rate": 0.00018241793434747798,
+      "loss": 0.8041,
+      "step": 555
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 0.41559815406799316,
+      "learning_rate": 0.0001823859087269816,
+      "loss": 0.8035,
+      "step": 556
+    },
+    {
+      "epoch": 0.17824,
+      "grad_norm": 0.41060441732406616,
+      "learning_rate": 0.0001823538831064852,
+      "loss": 0.803,
+      "step": 557
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 0.4704146981239319,
+      "learning_rate": 0.0001823218574859888,
+      "loss": 0.7679,
+      "step": 558
+    },
+    {
+      "epoch": 0.17888,
+      "grad_norm": 0.3849044144153595,
+      "learning_rate": 0.0001822898318654924,
+      "loss": 0.7153,
+      "step": 559
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.3413396179676056,
+      "learning_rate": 0.000182257806244996,
+      "loss": 0.6978,
+      "step": 560
+    },
+    {
+      "epoch": 0.17952,
+      "grad_norm": 0.41710364818573,
+      "learning_rate": 0.0001822257806244996,
+      "loss": 0.7625,
+      "step": 561
+    },
+    {
+      "epoch": 0.17984,
+      "grad_norm": 0.4305817782878876,
+      "learning_rate": 0.00018219375500400323,
+      "loss": 0.6859,
+      "step": 562
+    },
+    {
+      "epoch": 0.18016,
+      "grad_norm": 0.4472603499889374,
+      "learning_rate": 0.00018216172938350682,
+      "loss": 0.8475,
+      "step": 563
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 0.45414721965789795,
+      "learning_rate": 0.0001821297037630104,
+      "loss": 0.8776,
+      "step": 564
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 0.4056128263473511,
+      "learning_rate": 0.00018209767814251403,
+      "loss": 0.7715,
+      "step": 565
+    },
+    {
+      "epoch": 0.18112,
+      "grad_norm": 0.38937073945999146,
+      "learning_rate": 0.00018206565252201762,
+      "loss": 0.7246,
+      "step": 566
+    },
+    {
+      "epoch": 0.18144,
+      "grad_norm": 0.43704766035079956,
+      "learning_rate": 0.0001820336269015212,
+      "loss": 0.7331,
+      "step": 567
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 0.437811017036438,
+      "learning_rate": 0.00018200160128102483,
+      "loss": 0.6982,
+      "step": 568
+    },
+    {
+      "epoch": 0.18208,
+      "grad_norm": 0.3710220456123352,
+      "learning_rate": 0.00018196957566052845,
+      "loss": 0.7229,
+      "step": 569
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.46758291125297546,
+      "learning_rate": 0.00018193755004003204,
+      "loss": 0.6994,
+      "step": 570
+    },
+    {
+      "epoch": 0.18272,
+      "grad_norm": 0.3821658194065094,
+      "learning_rate": 0.00018190552441953563,
+      "loss": 0.5693,
+      "step": 571
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 0.39665743708610535,
+      "learning_rate": 0.00018187349879903922,
+      "loss": 0.7229,
+      "step": 572
+    },
+    {
+      "epoch": 0.18336,
+      "grad_norm": 0.4053206443786621,
+      "learning_rate": 0.00018184147317854287,
+      "loss": 0.7411,
+      "step": 573
+    },
+    {
+      "epoch": 0.18368,
+      "grad_norm": 0.37660571932792664,
+      "learning_rate": 0.00018180944755804646,
+      "loss": 0.7287,
+      "step": 574
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.38917601108551025,
+      "learning_rate": 0.00018177742193755005,
+      "loss": 0.6978,
+      "step": 575
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 0.45924264192581177,
+      "learning_rate": 0.00018174539631705364,
+      "loss": 0.6774,
+      "step": 576
+    },
+    {
+      "epoch": 0.18464,
+      "grad_norm": 0.43932151794433594,
+      "learning_rate": 0.00018171337069655726,
+      "loss": 0.5981,
+      "step": 577
+    },
+    {
+      "epoch": 0.18496,
+      "grad_norm": 0.3602670431137085,
+      "learning_rate": 0.00018168134507606085,
+      "loss": 0.6196,
+      "step": 578
+    },
+    {
+      "epoch": 0.18528,
+      "grad_norm": 0.3993200659751892,
+      "learning_rate": 0.00018164931945556447,
+      "loss": 0.7674,
+      "step": 579
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.43009382486343384,
+      "learning_rate": 0.00018161729383506806,
+      "loss": 0.585,
+      "step": 580
+    },
+    {
+      "epoch": 0.18592,
+      "grad_norm": 0.48123899102211,
+      "learning_rate": 0.00018158526821457168,
+      "loss": 0.7773,
+      "step": 581
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 0.4809722602367401,
+      "learning_rate": 0.00018155324259407527,
+      "loss": 0.8153,
+      "step": 582
+    },
+    {
+      "epoch": 0.18656,
+      "grad_norm": 0.46205630898475647,
+      "learning_rate": 0.00018152121697357886,
+      "loss": 0.9493,
+      "step": 583
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 0.3634149134159088,
+      "learning_rate": 0.00018148919135308248,
+      "loss": 0.569,
+      "step": 584
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 0.4086751341819763,
+      "learning_rate": 0.0001814571657325861,
+      "loss": 0.7177,
+      "step": 585
+    },
+    {
+      "epoch": 0.18752,
+      "grad_norm": 0.3780278265476227,
+      "learning_rate": 0.0001814251401120897,
+      "loss": 0.6075,
+      "step": 586
+    },
+    {
+      "epoch": 0.18784,
+      "grad_norm": 0.507053017616272,
+      "learning_rate": 0.00018139311449159328,
+      "loss": 0.84,
+      "step": 587
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 0.4489976167678833,
+      "learning_rate": 0.0001813610888710969,
+      "loss": 0.7251,
+      "step": 588
+    },
+    {
+      "epoch": 0.18848,
+      "grad_norm": 0.3980749249458313,
+      "learning_rate": 0.0001813290632506005,
+      "loss": 0.7363,
+      "step": 589
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.3918863534927368,
+      "learning_rate": 0.00018129703763010408,
+      "loss": 0.6084,
+      "step": 590
+    },
+    {
+      "epoch": 0.18912,
+      "grad_norm": 0.36752891540527344,
+      "learning_rate": 0.0001812650120096077,
+      "loss": 0.6989,
+      "step": 591
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 0.4141654372215271,
+      "learning_rate": 0.00018123298638911132,
+      "loss": 0.8004,
+      "step": 592
+    },
+    {
+      "epoch": 0.18976,
+      "grad_norm": 0.4517577886581421,
+      "learning_rate": 0.0001812009607686149,
+      "loss": 0.7002,
+      "step": 593
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 0.3799104392528534,
+      "learning_rate": 0.0001811689351481185,
+      "loss": 0.6949,
+      "step": 594
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 0.41419747471809387,
+      "learning_rate": 0.0001811369095276221,
+      "loss": 0.6759,
+      "step": 595
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 0.36056333780288696,
+      "learning_rate": 0.0001811048839071257,
+      "loss": 0.6747,
+      "step": 596
+    },
+    {
+      "epoch": 0.19104,
+      "grad_norm": 0.38920480012893677,
+      "learning_rate": 0.00018107285828662933,
+      "loss": 0.8081,
+      "step": 597
+    },
+    {
+      "epoch": 0.19136,
+      "grad_norm": 0.47673332691192627,
+      "learning_rate": 0.00018104083266613292,
+      "loss": 0.9897,
+      "step": 598
+    },
+    {
+      "epoch": 0.19168,
+      "grad_norm": 0.403861403465271,
+      "learning_rate": 0.0001810088070456365,
+      "loss": 0.5869,
+      "step": 599
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.40557828545570374,
+      "learning_rate": 0.00018097678142514013,
+      "loss": 0.6723,
+      "step": 600
+    },
+    {
+      "epoch": 0.19232,
+      "grad_norm": 0.39100873470306396,
+      "learning_rate": 0.00018094475580464372,
+      "loss": 0.6776,
+      "step": 601
+    },
+    {
+      "epoch": 0.19264,
+      "grad_norm": 0.4126138389110565,
+      "learning_rate": 0.0001809127301841473,
+      "loss": 0.7344,
+      "step": 602
+    },
+    {
+      "epoch": 0.19296,
+      "grad_norm": 0.4171197712421417,
+      "learning_rate": 0.00018088070456365093,
+      "loss": 0.6591,
+      "step": 603
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 0.34317290782928467,
+      "learning_rate": 0.00018084867894315455,
+      "loss": 0.6765,
+      "step": 604
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 0.4551938474178314,
+      "learning_rate": 0.00018081665332265814,
+      "loss": 0.7206,
+      "step": 605
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 0.426835298538208,
+      "learning_rate": 0.00018078462770216173,
+      "loss": 0.761,
+      "step": 606
+    },
+    {
+      "epoch": 0.19424,
+      "grad_norm": 0.3108452558517456,
+      "learning_rate": 0.00018075260208166535,
+      "loss": 0.5504,
+      "step": 607
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 0.427958607673645,
+      "learning_rate": 0.00018072057646116894,
+      "loss": 0.6461,
+      "step": 608
+    },
+    {
+      "epoch": 0.19488,
+      "grad_norm": 0.3993356227874756,
+      "learning_rate": 0.00018068855084067256,
+      "loss": 0.7032,
+      "step": 609
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.4217299818992615,
+      "learning_rate": 0.00018065652522017615,
+      "loss": 0.6715,
+      "step": 610
+    },
+    {
+      "epoch": 0.19552,
+      "grad_norm": 0.3743104040622711,
+      "learning_rate": 0.00018062449959967977,
+      "loss": 0.6954,
+      "step": 611
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 0.3981457054615021,
+      "learning_rate": 0.00018059247397918336,
+      "loss": 0.7046,
+      "step": 612
+    },
+    {
+      "epoch": 0.19616,
+      "grad_norm": 0.35218438506126404,
+      "learning_rate": 0.00018056044835868695,
+      "loss": 0.5791,
+      "step": 613
+    },
+    {
+      "epoch": 0.19648,
+      "grad_norm": 0.3698236644268036,
+      "learning_rate": 0.00018052842273819054,
+      "loss": 0.7109,
+      "step": 614
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 0.396771639585495,
+      "learning_rate": 0.00018049639711769418,
+      "loss": 0.6762,
+      "step": 615
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 0.33966508507728577,
+      "learning_rate": 0.00018046437149719778,
+      "loss": 0.6238,
+      "step": 616
+    },
+    {
+      "epoch": 0.19744,
+      "grad_norm": 0.42938247323036194,
+      "learning_rate": 0.00018043234587670137,
+      "loss": 0.6286,
+      "step": 617
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 0.45196548104286194,
+      "learning_rate": 0.00018040032025620496,
+      "loss": 0.8097,
+      "step": 618
+    },
+    {
+      "epoch": 0.19808,
+      "grad_norm": 0.43913716077804565,
+      "learning_rate": 0.00018036829463570858,
+      "loss": 0.7829,
+      "step": 619
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.38645610213279724,
+      "learning_rate": 0.0001803362690152122,
+      "loss": 0.6884,
+      "step": 620
+    },
+    {
+      "epoch": 0.19872,
+      "grad_norm": 0.3726969063282013,
+      "learning_rate": 0.00018030424339471579,
+      "loss": 0.7051,
+      "step": 621
+    },
+    {
+      "epoch": 0.19904,
+      "grad_norm": 0.385557621717453,
+      "learning_rate": 0.00018027221777421938,
+      "loss": 0.7055,
+      "step": 622
+    },
+    {
+      "epoch": 0.19936,
+      "grad_norm": 0.4218786060810089,
+      "learning_rate": 0.000180240192153723,
+      "loss": 0.5791,
+      "step": 623
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 0.3564837574958801,
+      "learning_rate": 0.0001802081665332266,
+      "loss": 0.5779,
+      "step": 624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38303253054618835,
+      "learning_rate": 0.00018017614091273018,
+      "loss": 0.8689,
+      "step": 625
+    },
+    {
+      "epoch": 0.20032,
+      "grad_norm": 0.41325557231903076,
+      "learning_rate": 0.0001801441152922338,
+      "loss": 0.6569,
+      "step": 626
+    },
+    {
+      "epoch": 0.20064,
+      "grad_norm": 0.36116936802864075,
+      "learning_rate": 0.00018011208967173741,
+      "loss": 0.6239,
+      "step": 627
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 0.3840174376964569,
+      "learning_rate": 0.000180080064051241,
+      "loss": 0.6159,
+      "step": 628
+    },
+    {
+      "epoch": 0.20128,
+      "grad_norm": 0.40116581320762634,
+      "learning_rate": 0.0001800480384307446,
+      "loss": 0.6666,
+      "step": 629
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.47608301043510437,
+      "learning_rate": 0.00018001601281024822,
+      "loss": 0.559,
+      "step": 630
+    },
+    {
+      "epoch": 0.20192,
+      "grad_norm": 0.410040020942688,
+      "learning_rate": 0.0001799839871897518,
+      "loss": 0.8527,
+      "step": 631
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 0.4649839401245117,
+      "learning_rate": 0.00017995196156925543,
+      "loss": 0.6529,
+      "step": 632
+    },
+    {
+      "epoch": 0.20256,
+      "grad_norm": 0.36767518520355225,
+      "learning_rate": 0.00017991993594875902,
+      "loss": 0.6174,
+      "step": 633
+    },
+    {
+      "epoch": 0.20288,
+      "grad_norm": 0.4459080398082733,
+      "learning_rate": 0.0001798879103282626,
+      "loss": 0.6527,
+      "step": 634
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 0.37570685148239136,
+      "learning_rate": 0.00017985588470776623,
+      "loss": 0.7234,
+      "step": 635
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 0.4316819906234741,
+      "learning_rate": 0.00017982385908726982,
+      "loss": 0.6995,
+      "step": 636
+    },
+    {
+      "epoch": 0.20384,
+      "grad_norm": 0.35585927963256836,
+      "learning_rate": 0.0001797918334667734,
+      "loss": 0.7291,
+      "step": 637
+    },
+    {
+      "epoch": 0.20416,
+      "grad_norm": 0.42792803049087524,
+      "learning_rate": 0.00017975980784627703,
+      "loss": 0.7454,
+      "step": 638
+    },
+    {
+      "epoch": 0.20448,
+      "grad_norm": 0.4363860487937927,
+      "learning_rate": 0.00017972778222578065,
+      "loss": 0.7998,
+      "step": 639
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.419108122587204,
+      "learning_rate": 0.00017969575660528424,
+      "loss": 0.6798,
+      "step": 640
+    },
+    {
+      "epoch": 0.20512,
+      "grad_norm": 0.4206399917602539,
+      "learning_rate": 0.00017966373098478783,
+      "loss": 0.6793,
+      "step": 641
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 0.47212591767311096,
+      "learning_rate": 0.00017963170536429145,
+      "loss": 0.6904,
+      "step": 642
+    },
+    {
+      "epoch": 0.20576,
+      "grad_norm": 0.4287513196468353,
+      "learning_rate": 0.00017959967974379504,
+      "loss": 0.6597,
+      "step": 643
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 0.4516173005104065,
+      "learning_rate": 0.00017956765412329866,
+      "loss": 0.8002,
+      "step": 644
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 0.4005032181739807,
+      "learning_rate": 0.00017953562850280225,
+      "loss": 0.7729,
+      "step": 645
+    },
+    {
+      "epoch": 0.20672,
+      "grad_norm": 0.4218226373195648,
+      "learning_rate": 0.00017950360288230586,
+      "loss": 0.7545,
+      "step": 646
+    },
+    {
+      "epoch": 0.20704,
+      "grad_norm": 0.43795883655548096,
+      "learning_rate": 0.00017947157726180946,
+      "loss": 0.7939,
+      "step": 647
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 0.3841351270675659,
+      "learning_rate": 0.00017943955164131305,
+      "loss": 0.6716,
+      "step": 648
+    },
+    {
+      "epoch": 0.20768,
+      "grad_norm": 0.38691502809524536,
+      "learning_rate": 0.00017940752602081664,
+      "loss": 0.6728,
+      "step": 649
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.3853699564933777,
+      "learning_rate": 0.00017937550040032028,
+      "loss": 0.7923,
+      "step": 650
+    },
+    {
+      "epoch": 0.20832,
+      "grad_norm": 0.4112347662448883,
+      "learning_rate": 0.00017934347477982388,
+      "loss": 0.6725,
+      "step": 651
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 0.37553057074546814,
+      "learning_rate": 0.00017931144915932747,
+      "loss": 0.6909,
+      "step": 652
+    },
+    {
+      "epoch": 0.20896,
+      "grad_norm": 0.40630537271499634,
+      "learning_rate": 0.00017927942353883106,
+      "loss": 0.6848,
+      "step": 653
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 0.3962356150150299,
+      "learning_rate": 0.00017924739791833468,
+      "loss": 0.6605,
+      "step": 654
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 0.3952970802783966,
+      "learning_rate": 0.00017921537229783827,
+      "loss": 0.7151,
+      "step": 655
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 0.38087329268455505,
+      "learning_rate": 0.00017918334667734189,
+      "loss": 0.7516,
+      "step": 656
+    },
+    {
+      "epoch": 0.21024,
+      "grad_norm": 0.4380969703197479,
+      "learning_rate": 0.00017915132105684548,
+      "loss": 0.6332,
+      "step": 657
+    },
+    {
+      "epoch": 0.21056,
+      "grad_norm": 0.5122512578964233,
+      "learning_rate": 0.0001791192954363491,
+      "loss": 0.8723,
+      "step": 658
+    },
+    {
+      "epoch": 0.21088,
+      "grad_norm": 0.44067826867103577,
+      "learning_rate": 0.00017908726981585269,
+      "loss": 0.7286,
+      "step": 659
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.39262887835502625,
+      "learning_rate": 0.00017905524419535628,
+      "loss": 0.8541,
+      "step": 660
+    },
+    {
+      "epoch": 0.21152,
+      "grad_norm": 0.45874860882759094,
+      "learning_rate": 0.0001790232185748599,
+      "loss": 0.7193,
+      "step": 661
+    },
+    {
+      "epoch": 0.21184,
+      "grad_norm": 0.4725707173347473,
+      "learning_rate": 0.00017899119295436351,
+      "loss": 0.7616,
+      "step": 662
+    },
+    {
+      "epoch": 0.21216,
+      "grad_norm": 0.3876330554485321,
+      "learning_rate": 0.0001789591673338671,
+      "loss": 0.657,
+      "step": 663
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 0.32938188314437866,
+      "learning_rate": 0.0001789271417133707,
+      "loss": 0.556,
+      "step": 664
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 0.39855095744132996,
+      "learning_rate": 0.00017889511609287431,
+      "loss": 0.7218,
+      "step": 665
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 0.327328085899353,
+      "learning_rate": 0.0001788630904723779,
+      "loss": 0.6398,
+      "step": 666
+    },
+    {
+      "epoch": 0.21344,
+      "grad_norm": 0.400744765996933,
+      "learning_rate": 0.00017883106485188152,
+      "loss": 0.6254,
+      "step": 667
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 0.4157651364803314,
+      "learning_rate": 0.00017879903923138512,
+      "loss": 0.7405,
+      "step": 668
+    },
+    {
+      "epoch": 0.21408,
+      "grad_norm": 0.4050652086734772,
+      "learning_rate": 0.00017876701361088873,
+      "loss": 0.622,
+      "step": 669
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 0.36614853143692017,
+      "learning_rate": 0.00017873498799039232,
+      "loss": 0.5637,
+      "step": 670
+    },
+    {
+      "epoch": 0.21472,
+      "grad_norm": 0.4172935485839844,
+      "learning_rate": 0.00017870296236989592,
+      "loss": 0.7813,
+      "step": 671
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 0.47844192385673523,
+      "learning_rate": 0.0001786709367493995,
+      "loss": 0.7014,
+      "step": 672
+    },
+    {
+      "epoch": 0.21536,
+      "grad_norm": 0.5072724223136902,
+      "learning_rate": 0.00017863891112890315,
+      "loss": 0.8277,
+      "step": 673
+    },
+    {
+      "epoch": 0.21568,
+      "grad_norm": 0.4182818531990051,
+      "learning_rate": 0.00017860688550840674,
+      "loss": 0.7024,
+      "step": 674
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.40103885531425476,
+      "learning_rate": 0.00017857485988791034,
+      "loss": 0.7092,
+      "step": 675
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 0.5322149991989136,
+      "learning_rate": 0.00017854283426741393,
+      "loss": 0.7963,
+      "step": 676
+    },
+    {
+      "epoch": 0.21664,
+      "grad_norm": 0.44529488682746887,
+      "learning_rate": 0.00017851080864691754,
+      "loss": 0.7575,
+      "step": 677
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 0.3895483911037445,
+      "learning_rate": 0.00017847878302642114,
+      "loss": 0.5707,
+      "step": 678
+    },
+    {
+      "epoch": 0.21728,
+      "grad_norm": 0.47409605979919434,
+      "learning_rate": 0.00017844675740592475,
+      "loss": 0.6124,
+      "step": 679
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.37176355719566345,
+      "learning_rate": 0.00017841473178542835,
+      "loss": 0.615,
+      "step": 680
+    },
+    {
+      "epoch": 0.21792,
+      "grad_norm": 0.4346714913845062,
+      "learning_rate": 0.00017838270616493196,
+      "loss": 0.7648,
+      "step": 681
+    },
+    {
+      "epoch": 0.21824,
+      "grad_norm": 0.44593101739883423,
+      "learning_rate": 0.00017835068054443555,
+      "loss": 0.6302,
+      "step": 682
+    },
+    {
+      "epoch": 0.21856,
+      "grad_norm": 0.4173053801059723,
+      "learning_rate": 0.00017831865492393915,
+      "loss": 0.7322,
+      "step": 683
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 0.3361636698246002,
+      "learning_rate": 0.00017828662930344276,
+      "loss": 0.5949,
+      "step": 684
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 0.4149784743785858,
+      "learning_rate": 0.00017825460368294638,
+      "loss": 0.8984,
+      "step": 685
+    },
+    {
+      "epoch": 0.21952,
+      "grad_norm": 0.4107123911380768,
+      "learning_rate": 0.00017822257806244997,
+      "loss": 0.7398,
+      "step": 686
+    },
+    {
+      "epoch": 0.21984,
+      "grad_norm": 0.37594500184059143,
+      "learning_rate": 0.00017819055244195357,
+      "loss": 0.6405,
+      "step": 687
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 0.46720725297927856,
+      "learning_rate": 0.00017815852682145718,
+      "loss": 0.8146,
+      "step": 688
+    },
+    {
+      "epoch": 0.22048,
+      "grad_norm": 0.422174870967865,
+      "learning_rate": 0.00017812650120096077,
+      "loss": 0.6688,
+      "step": 689
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.35983720421791077,
+      "learning_rate": 0.00017809447558046437,
+      "loss": 0.6664,
+      "step": 690
+    },
+    {
+      "epoch": 0.22112,
+      "grad_norm": 0.48956191539764404,
+      "learning_rate": 0.00017806244995996798,
+      "loss": 0.7019,
+      "step": 691
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 0.32427939772605896,
+      "learning_rate": 0.0001780304243394716,
+      "loss": 0.5623,
+      "step": 692
+    },
+    {
+      "epoch": 0.22176,
+      "grad_norm": 0.3407818377017975,
+      "learning_rate": 0.0001779983987189752,
+      "loss": 0.607,
+      "step": 693
+    },
+    {
+      "epoch": 0.22208,
+      "grad_norm": 0.3687211871147156,
+      "learning_rate": 0.00017796637309847878,
+      "loss": 0.6196,
+      "step": 694
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 0.404786080121994,
+      "learning_rate": 0.00017793434747798238,
+      "loss": 0.6877,
+      "step": 695
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 0.39013659954071045,
+      "learning_rate": 0.000177902321857486,
+      "loss": 0.6424,
+      "step": 696
+    },
+    {
+      "epoch": 0.22304,
+      "grad_norm": 0.514657199382782,
+      "learning_rate": 0.0001778702962369896,
+      "loss": 0.7562,
+      "step": 697
+    },
+    {
+      "epoch": 0.22336,
+      "grad_norm": 0.3729800283908844,
+      "learning_rate": 0.0001778382706164932,
+      "loss": 0.6751,
+      "step": 698
+    },
+    {
+      "epoch": 0.22368,
+      "grad_norm": 0.3254528343677521,
+      "learning_rate": 0.0001778062449959968,
+      "loss": 0.5005,
+      "step": 699
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.4316803216934204,
+      "learning_rate": 0.00017777421937550041,
+      "loss": 0.76,
+      "step": 700
+    },
+    {
+      "epoch": 0.22432,
+      "grad_norm": 0.5405811071395874,
+      "learning_rate": 0.000177742193755004,
+      "loss": 0.9578,
+      "step": 701
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 0.40572044253349304,
+      "learning_rate": 0.0001777101681345076,
+      "loss": 0.6774,
+      "step": 702
+    },
+    {
+      "epoch": 0.22496,
+      "grad_norm": 0.44261807203292847,
+      "learning_rate": 0.00017767814251401121,
+      "loss": 0.7028,
+      "step": 703
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 0.31827402114868164,
+      "learning_rate": 0.00017764611689351483,
+      "loss": 0.5619,
+      "step": 704
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 0.37754201889038086,
+      "learning_rate": 0.00017761409127301842,
+      "loss": 0.5954,
+      "step": 705
+    },
+    {
+      "epoch": 0.22592,
+      "grad_norm": 0.44598087668418884,
+      "learning_rate": 0.00017758206565252202,
+      "loss": 0.7116,
+      "step": 706
+    },
+    {
+      "epoch": 0.22624,
+      "grad_norm": 0.3818089962005615,
+      "learning_rate": 0.00017755004003202563,
+      "loss": 0.6572,
+      "step": 707
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 0.3992815911769867,
+      "learning_rate": 0.00017751801441152922,
+      "loss": 0.5736,
+      "step": 708
+    },
+    {
+      "epoch": 0.22688,
+      "grad_norm": 0.37587547302246094,
+      "learning_rate": 0.00017748598879103284,
+      "loss": 0.576,
+      "step": 709
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.3911394476890564,
+      "learning_rate": 0.00017745396317053643,
+      "loss": 0.6842,
+      "step": 710
+    },
+    {
+      "epoch": 0.22752,
+      "grad_norm": 0.36865049600601196,
+      "learning_rate": 0.00017742193755004005,
+      "loss": 0.6248,
+      "step": 711
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 0.4335592985153198,
+      "learning_rate": 0.00017738991192954364,
+      "loss": 0.7087,
+      "step": 712
+    },
+    {
+      "epoch": 0.22816,
+      "grad_norm": 0.38089385628700256,
+      "learning_rate": 0.00017735788630904723,
+      "loss": 0.6858,
+      "step": 713
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 0.43585726618766785,
+      "learning_rate": 0.00017732586068855085,
+      "loss": 0.7364,
+      "step": 714
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 0.3832882344722748,
+      "learning_rate": 0.00017729383506805447,
+      "loss": 0.5679,
+      "step": 715
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 0.42041921615600586,
+      "learning_rate": 0.00017726180944755806,
+      "loss": 0.5997,
+      "step": 716
+    },
+    {
+      "epoch": 0.22944,
+      "grad_norm": 0.42805084586143494,
+      "learning_rate": 0.00017722978382706165,
+      "loss": 0.6183,
+      "step": 717
+    },
+    {
+      "epoch": 0.22976,
+      "grad_norm": 0.43876177072525024,
+      "learning_rate": 0.00017719775820656525,
+      "loss": 0.6712,
+      "step": 718
+    },
+    {
+      "epoch": 0.23008,
+      "grad_norm": 0.3522574305534363,
+      "learning_rate": 0.00017716573258606886,
+      "loss": 0.5618,
+      "step": 719
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.3886878490447998,
+      "learning_rate": 0.00017713370696557248,
+      "loss": 0.7037,
+      "step": 720
+    },
+    {
+      "epoch": 0.23072,
+      "grad_norm": 0.46576038002967834,
+      "learning_rate": 0.00017710168134507607,
+      "loss": 0.6998,
+      "step": 721
+    },
+    {
+      "epoch": 0.23104,
+      "grad_norm": 0.34737879037857056,
+      "learning_rate": 0.00017706965572457966,
+      "loss": 0.5589,
+      "step": 722
+    },
+    {
+      "epoch": 0.23136,
+      "grad_norm": 0.5799711346626282,
+      "learning_rate": 0.00017703763010408328,
+      "loss": 0.8379,
+      "step": 723
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 0.38293972611427307,
+      "learning_rate": 0.00017700560448358687,
+      "loss": 0.6603,
+      "step": 724
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.5091928243637085,
+      "learning_rate": 0.00017697357886309046,
+      "loss": 0.7844,
+      "step": 725
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 0.390472948551178,
+      "learning_rate": 0.00017694155324259408,
+      "loss": 0.8339,
+      "step": 726
+    },
+    {
+      "epoch": 0.23264,
+      "grad_norm": 0.3927406370639801,
+      "learning_rate": 0.0001769095276220977,
+      "loss": 0.699,
+      "step": 727
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 0.43780747056007385,
+      "learning_rate": 0.0001768775020016013,
+      "loss": 0.6129,
+      "step": 728
+    },
+    {
+      "epoch": 0.23328,
+      "grad_norm": 0.5888350605964661,
+      "learning_rate": 0.00017684547638110488,
+      "loss": 1.0139,
+      "step": 729
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.3547525107860565,
+      "learning_rate": 0.0001768134507606085,
+      "loss": 0.7482,
+      "step": 730
+    },
+    {
+      "epoch": 0.23392,
+      "grad_norm": 0.4281228184700012,
+      "learning_rate": 0.0001767814251401121,
+      "loss": 0.542,
+      "step": 731
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 0.3909932076931,
+      "learning_rate": 0.0001767493995196157,
+      "loss": 0.6335,
+      "step": 732
+    },
+    {
+      "epoch": 0.23456,
+      "grad_norm": 0.36411863565444946,
+      "learning_rate": 0.0001767173738991193,
+      "loss": 0.5635,
+      "step": 733
+    },
+    {
+      "epoch": 0.23488,
+      "grad_norm": 0.5074488520622253,
+      "learning_rate": 0.00017668534827862292,
+      "loss": 0.7085,
+      "step": 734
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 0.3630116581916809,
+      "learning_rate": 0.0001766533226581265,
+      "loss": 0.6254,
+      "step": 735
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 0.38210898637771606,
+      "learning_rate": 0.0001766212970376301,
+      "loss": 0.6145,
+      "step": 736
+    },
+    {
+      "epoch": 0.23584,
+      "grad_norm": 0.34646084904670715,
+      "learning_rate": 0.0001765892714171337,
+      "loss": 0.5463,
+      "step": 737
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 0.4058981239795685,
+      "learning_rate": 0.00017655724579663734,
+      "loss": 0.6746,
+      "step": 738
+    },
+    {
+      "epoch": 0.23648,
+      "grad_norm": 0.4535089433193207,
+      "learning_rate": 0.00017652522017614093,
+      "loss": 0.6496,
+      "step": 739
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.3792216181755066,
+      "learning_rate": 0.00017649319455564452,
+      "loss": 0.5729,
+      "step": 740
+    },
+    {
+      "epoch": 0.23712,
+      "grad_norm": 0.3308945894241333,
+      "learning_rate": 0.00017646116893514811,
+      "loss": 0.5501,
+      "step": 741
+    },
+    {
+      "epoch": 0.23744,
+      "grad_norm": 0.4340299665927887,
+      "learning_rate": 0.00017642914331465173,
+      "loss": 0.6666,
+      "step": 742
+    },
+    {
+      "epoch": 0.23776,
+      "grad_norm": 0.39663732051849365,
+      "learning_rate": 0.00017639711769415532,
+      "loss": 0.7632,
+      "step": 743
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 0.35612136125564575,
+      "learning_rate": 0.00017636509207365894,
+      "loss": 0.628,
+      "step": 744
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 0.45463618636131287,
+      "learning_rate": 0.00017633306645316253,
+      "loss": 0.7386,
+      "step": 745
+    },
+    {
+      "epoch": 0.23872,
+      "grad_norm": 0.390143483877182,
+      "learning_rate": 0.00017630104083266615,
+      "loss": 0.6329,
+      "step": 746
+    },
+    {
+      "epoch": 0.23904,
+      "grad_norm": 0.45698216557502747,
+      "learning_rate": 0.00017626901521216974,
+      "loss": 0.7271,
+      "step": 747
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 0.46281126141548157,
+      "learning_rate": 0.00017623698959167333,
+      "loss": 0.7483,
+      "step": 748
+    },
+    {
+      "epoch": 0.23968,
+      "grad_norm": 0.3986653685569763,
+      "learning_rate": 0.00017620496397117695,
+      "loss": 0.6072,
+      "step": 749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41029831767082214,
+      "learning_rate": 0.00017617293835068057,
+      "loss": 0.6353,
+      "step": 750
+    },
+    {
+      "epoch": 0.24032,
+      "grad_norm": 0.4181952476501465,
+      "learning_rate": 0.00017614091273018416,
+      "loss": 0.6367,
+      "step": 751
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 0.4186416566371918,
+      "learning_rate": 0.00017610888710968775,
+      "loss": 0.6458,
+      "step": 752
+    },
+    {
+      "epoch": 0.24096,
+      "grad_norm": 0.3711709976196289,
+      "learning_rate": 0.00017607686148919137,
+      "loss": 0.6295,
+      "step": 753
+    },
+    {
+      "epoch": 0.24128,
+      "grad_norm": 0.39803627133369446,
+      "learning_rate": 0.00017604483586869496,
+      "loss": 0.7807,
+      "step": 754
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 0.35898879170417786,
+      "learning_rate": 0.00017601281024819855,
+      "loss": 0.6664,
+      "step": 755
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 0.4399225115776062,
+      "learning_rate": 0.00017598078462770217,
+      "loss": 0.6804,
+      "step": 756
+    },
+    {
+      "epoch": 0.24224,
+      "grad_norm": 0.3886784613132477,
+      "learning_rate": 0.0001759487590072058,
+      "loss": 0.7026,
+      "step": 757
+    },
+    {
+      "epoch": 0.24256,
+      "grad_norm": 0.3673941195011139,
+      "learning_rate": 0.00017591673338670938,
+      "loss": 0.64,
+      "step": 758
+    },
+    {
+      "epoch": 0.24288,
+      "grad_norm": 0.4334450960159302,
+      "learning_rate": 0.00017588470776621297,
+      "loss": 0.6396,
+      "step": 759
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.498076468706131,
+      "learning_rate": 0.00017585268214571656,
+      "loss": 0.6846,
+      "step": 760
+    },
+    {
+      "epoch": 0.24352,
+      "grad_norm": 0.40077778697013855,
+      "learning_rate": 0.00017582065652522018,
+      "loss": 0.7004,
+      "step": 761
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 0.46941903233528137,
+      "learning_rate": 0.0001757886309047238,
+      "loss": 0.6504,
+      "step": 762
+    },
+    {
+      "epoch": 0.24416,
+      "grad_norm": 0.379960834980011,
+      "learning_rate": 0.0001757566052842274,
+      "loss": 0.5878,
+      "step": 763
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 0.42586201429367065,
+      "learning_rate": 0.00017572457966373098,
+      "loss": 0.6671,
+      "step": 764
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 0.3924212157726288,
+      "learning_rate": 0.0001756925540432346,
+      "loss": 0.6795,
+      "step": 765
+    },
+    {
+      "epoch": 0.24512,
+      "grad_norm": 0.3634004592895508,
+      "learning_rate": 0.0001756605284227382,
+      "loss": 0.5965,
+      "step": 766
+    },
+    {
+      "epoch": 0.24544,
+      "grad_norm": 0.31967633962631226,
+      "learning_rate": 0.0001756285028022418,
+      "loss": 0.5795,
+      "step": 767
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 0.43673497438430786,
+      "learning_rate": 0.0001755964771817454,
+      "loss": 0.7537,
+      "step": 768
+    },
+    {
+      "epoch": 0.24608,
+      "grad_norm": 0.38086438179016113,
+      "learning_rate": 0.00017556445156124902,
+      "loss": 0.6261,
+      "step": 769
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 0.27452465891838074,
+      "learning_rate": 0.0001755324259407526,
+      "loss": 0.4913,
+      "step": 770
+    },
+    {
+      "epoch": 0.24672,
+      "grad_norm": 0.42407506704330444,
+      "learning_rate": 0.0001755004003202562,
+      "loss": 0.6968,
+      "step": 771
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 0.4043313264846802,
+      "learning_rate": 0.00017546837469975982,
+      "loss": 0.6535,
+      "step": 772
+    },
+    {
+      "epoch": 0.24736,
+      "grad_norm": 0.4044741690158844,
+      "learning_rate": 0.00017543634907926344,
+      "loss": 0.6554,
+      "step": 773
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 0.4057358205318451,
+      "learning_rate": 0.00017540432345876703,
+      "loss": 0.7971,
+      "step": 774
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 0.5016301870346069,
+      "learning_rate": 0.00017537229783827062,
+      "loss": 0.7861,
+      "step": 775
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 0.41519948840141296,
+      "learning_rate": 0.0001753402722177742,
+      "loss": 0.7061,
+      "step": 776
+    },
+    {
+      "epoch": 0.24864,
+      "grad_norm": 0.4074082672595978,
+      "learning_rate": 0.00017530824659727783,
+      "loss": 0.6807,
+      "step": 777
+    },
+    {
+      "epoch": 0.24896,
+      "grad_norm": 0.38981643319129944,
+      "learning_rate": 0.00017527622097678142,
+      "loss": 0.6391,
+      "step": 778
+    },
+    {
+      "epoch": 0.24928,
+      "grad_norm": 0.4005015194416046,
+      "learning_rate": 0.00017524419535628504,
+      "loss": 0.6596,
+      "step": 779
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.38147956132888794,
+      "learning_rate": 0.00017521216973578863,
+      "loss": 0.6491,
+      "step": 780
+    },
+    {
+      "epoch": 0.24992,
+      "grad_norm": 0.4014551341533661,
+      "learning_rate": 0.00017518014411529225,
+      "loss": 0.6992,
+      "step": 781
+    },
+    {
+      "epoch": 0.25024,
+      "grad_norm": 0.41478654742240906,
+      "learning_rate": 0.00017514811849479584,
+      "loss": 0.7458,
+      "step": 782
+    },
+    {
+      "epoch": 0.25056,
+      "grad_norm": 0.40160098671913147,
+      "learning_rate": 0.00017511609287429943,
+      "loss": 0.7008,
+      "step": 783
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 0.4423017203807831,
+      "learning_rate": 0.00017508406725380305,
+      "loss": 0.8041,
+      "step": 784
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 0.3832944929599762,
+      "learning_rate": 0.00017505204163330667,
+      "loss": 0.6524,
+      "step": 785
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 0.4458499252796173,
+      "learning_rate": 0.00017502001601281026,
+      "loss": 0.7802,
+      "step": 786
+    },
+    {
+      "epoch": 0.25184,
+      "grad_norm": 0.3594331741333008,
+      "learning_rate": 0.00017498799039231385,
+      "loss": 0.612,
+      "step": 787
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 0.28718897700309753,
+      "learning_rate": 0.00017495596477181747,
+      "loss": 0.5033,
+      "step": 788
+    },
+    {
+      "epoch": 0.25248,
+      "grad_norm": 0.3251974284648895,
+      "learning_rate": 0.00017492393915132106,
+      "loss": 0.6206,
+      "step": 789
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.4196889102458954,
+      "learning_rate": 0.00017489191353082465,
+      "loss": 0.734,
+      "step": 790
+    },
+    {
+      "epoch": 0.25312,
+      "grad_norm": 0.4176255762577057,
+      "learning_rate": 0.00017485988791032827,
+      "loss": 0.7578,
+      "step": 791
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 0.45544517040252686,
+      "learning_rate": 0.0001748278622898319,
+      "loss": 0.7559,
+      "step": 792
+    },
+    {
+      "epoch": 0.25376,
+      "grad_norm": 0.29982420802116394,
+      "learning_rate": 0.00017479583666933548,
+      "loss": 0.5627,
+      "step": 793
+    },
+    {
+      "epoch": 0.25408,
+      "grad_norm": 0.4078064560890198,
+      "learning_rate": 0.00017476381104883907,
+      "loss": 0.6073,
+      "step": 794
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 0.36697089672088623,
+      "learning_rate": 0.00017473178542834266,
+      "loss": 0.5589,
+      "step": 795
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 0.44719722867012024,
+      "learning_rate": 0.00017469975980784628,
+      "loss": 0.6575,
+      "step": 796
+    },
+    {
+      "epoch": 0.25504,
+      "grad_norm": 0.32681721448898315,
+      "learning_rate": 0.0001746677341873499,
+      "loss": 0.5377,
+      "step": 797
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 0.42954570055007935,
+      "learning_rate": 0.0001746357085668535,
+      "loss": 0.777,
+      "step": 798
+    },
+    {
+      "epoch": 0.25568,
+      "grad_norm": 0.3837994337081909,
+      "learning_rate": 0.00017460368294635708,
+      "loss": 0.6954,
+      "step": 799
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.3895558714866638,
+      "learning_rate": 0.0001745716573258607,
+      "loss": 0.6482,
+      "step": 800
+    },
+    {
+      "epoch": 0.25632,
+      "grad_norm": 0.535961389541626,
+      "learning_rate": 0.0001745396317053643,
+      "loss": 0.8164,
+      "step": 801
+    },
+    {
+      "epoch": 0.25664,
+      "grad_norm": 0.3793974816799164,
+      "learning_rate": 0.00017450760608486788,
+      "loss": 0.64,
+      "step": 802
+    },
+    {
+      "epoch": 0.25696,
+      "grad_norm": 0.37344518303871155,
+      "learning_rate": 0.0001744755804643715,
+      "loss": 0.7098,
+      "step": 803
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 0.3986565172672272,
+      "learning_rate": 0.00017444355484387512,
+      "loss": 0.6232,
+      "step": 804
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 0.4452754557132721,
+      "learning_rate": 0.0001744115292233787,
+      "loss": 0.7809,
+      "step": 805
+    },
+    {
+      "epoch": 0.25792,
+      "grad_norm": 0.4009181559085846,
+      "learning_rate": 0.0001743795036028823,
+      "loss": 0.678,
+      "step": 806
+    },
+    {
+      "epoch": 0.25824,
+      "grad_norm": 0.48988550901412964,
+      "learning_rate": 0.00017434747798238592,
+      "loss": 0.7559,
+      "step": 807
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 0.36506885290145874,
+      "learning_rate": 0.00017431545236188954,
+      "loss": 0.5025,
+      "step": 808
+    },
+    {
+      "epoch": 0.25888,
+      "grad_norm": 0.37770646810531616,
+      "learning_rate": 0.00017428342674139313,
+      "loss": 0.6693,
+      "step": 809
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.34071582555770874,
+      "learning_rate": 0.00017425140112089672,
+      "loss": 0.6408,
+      "step": 810
+    },
+    {
+      "epoch": 0.25952,
+      "grad_norm": 0.43021801114082336,
+      "learning_rate": 0.00017421937550040034,
+      "loss": 0.8079,
+      "step": 811
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 0.32836052775382996,
+      "learning_rate": 0.00017418734987990393,
+      "loss": 0.5369,
+      "step": 812
+    },
+    {
+      "epoch": 0.26016,
+      "grad_norm": 0.4471929371356964,
+      "learning_rate": 0.00017415532425940752,
+      "loss": 0.7047,
+      "step": 813
+    },
+    {
+      "epoch": 0.26048,
+      "grad_norm": 0.3551078736782074,
+      "learning_rate": 0.00017412329863891114,
+      "loss": 0.6221,
+      "step": 814
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 0.3817354738712311,
+      "learning_rate": 0.00017409127301841476,
+      "loss": 0.5619,
+      "step": 815
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 0.47301605343818665,
+      "learning_rate": 0.00017405924739791835,
+      "loss": 0.7652,
+      "step": 816
+    },
+    {
+      "epoch": 0.26144,
+      "grad_norm": 0.45505982637405396,
+      "learning_rate": 0.00017402722177742194,
+      "loss": 0.7635,
+      "step": 817
+    },
+    {
+      "epoch": 0.26176,
+      "grad_norm": 0.5452256202697754,
+      "learning_rate": 0.00017399519615692553,
+      "loss": 0.8499,
+      "step": 818
+    },
+    {
+      "epoch": 0.26208,
+      "grad_norm": 0.35701045393943787,
+      "learning_rate": 0.00017396317053642915,
+      "loss": 0.5767,
+      "step": 819
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.40064558386802673,
+      "learning_rate": 0.00017393114491593277,
+      "loss": 0.6599,
+      "step": 820
+    },
+    {
+      "epoch": 0.26272,
+      "grad_norm": 0.3643248379230499,
+      "learning_rate": 0.00017389911929543636,
+      "loss": 0.6095,
+      "step": 821
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 0.4147706627845764,
+      "learning_rate": 0.00017386709367493995,
+      "loss": 0.6588,
+      "step": 822
+    },
+    {
+      "epoch": 0.26336,
+      "grad_norm": 0.31973665952682495,
+      "learning_rate": 0.00017383506805444357,
+      "loss": 0.6242,
+      "step": 823
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 0.3964540958404541,
+      "learning_rate": 0.00017380304243394716,
+      "loss": 0.6805,
+      "step": 824
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.4710293412208557,
+      "learning_rate": 0.00017377101681345075,
+      "loss": 0.7296,
+      "step": 825
+    },
+    {
+      "epoch": 0.26432,
+      "grad_norm": 0.4644870460033417,
+      "learning_rate": 0.00017373899119295437,
+      "loss": 0.6798,
+      "step": 826
+    },
+    {
+      "epoch": 0.26464,
+      "grad_norm": 0.37687814235687256,
+      "learning_rate": 0.000173706965572458,
+      "loss": 0.6522,
+      "step": 827
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 0.40914878249168396,
+      "learning_rate": 0.00017367493995196158,
+      "loss": 0.8119,
+      "step": 828
+    },
+    {
+      "epoch": 0.26528,
+      "grad_norm": 0.2918422520160675,
+      "learning_rate": 0.00017364291433146517,
+      "loss": 0.5298,
+      "step": 829
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.4405345618724823,
+      "learning_rate": 0.0001736108887109688,
+      "loss": 0.8604,
+      "step": 830
+    },
+    {
+      "epoch": 0.26592,
+      "grad_norm": 0.4246082007884979,
+      "learning_rate": 0.00017357886309047238,
+      "loss": 0.7577,
+      "step": 831
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 0.399863064289093,
+      "learning_rate": 0.000173546837469976,
+      "loss": 0.6873,
+      "step": 832
+    },
+    {
+      "epoch": 0.26656,
+      "grad_norm": 0.42153140902519226,
+      "learning_rate": 0.0001735148118494796,
+      "loss": 0.7111,
+      "step": 833
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 0.42143017053604126,
+      "learning_rate": 0.0001734827862289832,
+      "loss": 0.6593,
+      "step": 834
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 0.41491228342056274,
+      "learning_rate": 0.0001734507606084868,
+      "loss": 0.7719,
+      "step": 835
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 0.4068368971347809,
+      "learning_rate": 0.0001734187349879904,
+      "loss": 0.6352,
+      "step": 836
+    },
+    {
+      "epoch": 0.26784,
+      "grad_norm": 0.38947415351867676,
+      "learning_rate": 0.00017338670936749398,
+      "loss": 0.7182,
+      "step": 837
+    },
+    {
+      "epoch": 0.26816,
+      "grad_norm": 0.3896469175815582,
+      "learning_rate": 0.00017335468374699763,
+      "loss": 0.6087,
+      "step": 838
+    },
+    {
+      "epoch": 0.26848,
+      "grad_norm": 0.3731803894042969,
+      "learning_rate": 0.00017332265812650122,
+      "loss": 0.6858,
+      "step": 839
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.42219144105911255,
+      "learning_rate": 0.0001732906325060048,
+      "loss": 0.592,
+      "step": 840
+    },
+    {
+      "epoch": 0.26912,
+      "grad_norm": 0.40658068656921387,
+      "learning_rate": 0.0001732586068855084,
+      "loss": 0.7077,
+      "step": 841
+    },
+    {
+      "epoch": 0.26944,
+      "grad_norm": 0.43663570284843445,
+      "learning_rate": 0.00017322658126501202,
+      "loss": 0.818,
+      "step": 842
+    },
+    {
+      "epoch": 0.26976,
+      "grad_norm": 0.3785604238510132,
+      "learning_rate": 0.0001731945556445156,
+      "loss": 0.6307,
+      "step": 843
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 0.35595396161079407,
+      "learning_rate": 0.00017316253002401923,
+      "loss": 0.7651,
+      "step": 844
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 0.5012089610099792,
+      "learning_rate": 0.00017313050440352282,
+      "loss": 0.6575,
+      "step": 845
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 0.37576329708099365,
+      "learning_rate": 0.00017309847878302644,
+      "loss": 0.6691,
+      "step": 846
+    },
+    {
+      "epoch": 0.27104,
+      "grad_norm": 0.409168004989624,
+      "learning_rate": 0.00017306645316253003,
+      "loss": 0.7136,
+      "step": 847
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 0.49547410011291504,
+      "learning_rate": 0.00017303442754203362,
+      "loss": 0.7524,
+      "step": 848
+    },
+    {
+      "epoch": 0.27168,
+      "grad_norm": 0.34320831298828125,
+      "learning_rate": 0.00017300240192153724,
+      "loss": 0.5755,
+      "step": 849
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.39681726694107056,
+      "learning_rate": 0.00017297037630104086,
+      "loss": 0.5865,
+      "step": 850
+    },
+    {
+      "epoch": 0.27232,
+      "grad_norm": 0.42252227663993835,
+      "learning_rate": 0.00017293835068054445,
+      "loss": 0.7119,
+      "step": 851
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 0.3091396689414978,
+      "learning_rate": 0.00017290632506004804,
+      "loss": 0.6005,
+      "step": 852
+    },
+    {
+      "epoch": 0.27296,
+      "grad_norm": 0.3978988230228424,
+      "learning_rate": 0.00017287429943955166,
+      "loss": 0.572,
+      "step": 853
+    },
+    {
+      "epoch": 0.27328,
+      "grad_norm": 0.4061945080757141,
+      "learning_rate": 0.00017284227381905525,
+      "loss": 0.8615,
+      "step": 854
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 0.42508235573768616,
+      "learning_rate": 0.00017281024819855887,
+      "loss": 0.6951,
+      "step": 855
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 0.2680080831050873,
+      "learning_rate": 0.00017277822257806246,
+      "loss": 0.4779,
+      "step": 856
+    },
+    {
+      "epoch": 0.27424,
+      "grad_norm": 0.3518430292606354,
+      "learning_rate": 0.00017274619695756608,
+      "loss": 0.5061,
+      "step": 857
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 0.5332779288291931,
+      "learning_rate": 0.00017271417133706967,
+      "loss": 0.7569,
+      "step": 858
+    },
+    {
+      "epoch": 0.27488,
+      "grad_norm": 0.416890949010849,
+      "learning_rate": 0.00017268214571657326,
+      "loss": 0.7254,
+      "step": 859
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.3719886541366577,
+      "learning_rate": 0.00017265012009607685,
+      "loss": 0.7145,
+      "step": 860
+    },
+    {
+      "epoch": 0.27552,
+      "grad_norm": 0.41365548968315125,
+      "learning_rate": 0.0001726180944755805,
+      "loss": 0.7114,
+      "step": 861
+    },
+    {
+      "epoch": 0.27584,
+      "grad_norm": 0.45553886890411377,
+      "learning_rate": 0.0001725860688550841,
+      "loss": 0.6098,
+      "step": 862
+    },
+    {
+      "epoch": 0.27616,
+      "grad_norm": 0.39312708377838135,
+      "learning_rate": 0.00017255404323458768,
+      "loss": 0.6845,
+      "step": 863
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 0.3986629545688629,
+      "learning_rate": 0.00017252201761409127,
+      "loss": 0.7045,
+      "step": 864
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 0.46225693821907043,
+      "learning_rate": 0.0001724899919935949,
+      "loss": 0.7639,
+      "step": 865
+    },
+    {
+      "epoch": 0.27712,
+      "grad_norm": 0.3437390625476837,
+      "learning_rate": 0.00017245796637309848,
+      "loss": 0.5437,
+      "step": 866
+    },
+    {
+      "epoch": 0.27744,
+      "grad_norm": 0.41979777812957764,
+      "learning_rate": 0.0001724259407526021,
+      "loss": 0.7741,
+      "step": 867
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 0.4171474575996399,
+      "learning_rate": 0.0001723939151321057,
+      "loss": 0.7136,
+      "step": 868
+    },
+    {
+      "epoch": 0.27808,
+      "grad_norm": 0.41412535309791565,
+      "learning_rate": 0.0001723618895116093,
+      "loss": 0.5884,
+      "step": 869
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.3445984125137329,
+      "learning_rate": 0.0001723298638911129,
+      "loss": 0.6065,
+      "step": 870
+    },
+    {
+      "epoch": 0.27872,
+      "grad_norm": 0.36672425270080566,
+      "learning_rate": 0.0001722978382706165,
+      "loss": 0.6133,
+      "step": 871
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 0.35718318819999695,
+      "learning_rate": 0.0001722658126501201,
+      "loss": 0.6273,
+      "step": 872
+    },
+    {
+      "epoch": 0.27936,
+      "grad_norm": 0.30761778354644775,
+      "learning_rate": 0.00017223378702962373,
+      "loss": 0.594,
+      "step": 873
+    },
+    {
+      "epoch": 0.27968,
+      "grad_norm": 0.43390321731567383,
+      "learning_rate": 0.00017220176140912732,
+      "loss": 0.7084,
+      "step": 874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46294790506362915,
+      "learning_rate": 0.0001721697357886309,
+      "loss": 0.7841,
+      "step": 875
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 0.4123653471469879,
+      "learning_rate": 0.00017213771016813453,
+      "loss": 0.5712,
+      "step": 876
+    },
+    {
+      "epoch": 0.28064,
+      "grad_norm": 0.43904587626457214,
+      "learning_rate": 0.00017210568454763812,
+      "loss": 0.7911,
+      "step": 877
+    },
+    {
+      "epoch": 0.28096,
+      "grad_norm": 0.4068009555339813,
+      "learning_rate": 0.0001720736589271417,
+      "loss": 0.6254,
+      "step": 878
+    },
+    {
+      "epoch": 0.28128,
+      "grad_norm": 0.3909311592578888,
+      "learning_rate": 0.00017204163330664533,
+      "loss": 0.686,
+      "step": 879
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.3748750686645508,
+      "learning_rate": 0.00017200960768614895,
+      "loss": 0.5778,
+      "step": 880
+    },
+    {
+      "epoch": 0.28192,
+      "grad_norm": 0.4485512673854828,
+      "learning_rate": 0.00017197758206565254,
+      "loss": 0.7171,
+      "step": 881
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 0.4702317714691162,
+      "learning_rate": 0.00017194555644515613,
+      "loss": 0.8194,
+      "step": 882
+    },
+    {
+      "epoch": 0.28256,
+      "grad_norm": 0.4094531536102295,
+      "learning_rate": 0.00017191353082465972,
+      "loss": 0.701,
+      "step": 883
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 0.38572442531585693,
+      "learning_rate": 0.00017188150520416334,
+      "loss": 0.6036,
+      "step": 884
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 0.47247856855392456,
+      "learning_rate": 0.00017184947958366696,
+      "loss": 0.7096,
+      "step": 885
+    },
+    {
+      "epoch": 0.28352,
+      "grad_norm": 0.41771063208580017,
+      "learning_rate": 0.00017181745396317055,
+      "loss": 0.5804,
+      "step": 886
+    },
+    {
+      "epoch": 0.28384,
+      "grad_norm": 0.37608128786087036,
+      "learning_rate": 0.00017178542834267414,
+      "loss": 0.6118,
+      "step": 887
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 0.4910067021846771,
+      "learning_rate": 0.00017175340272217776,
+      "loss": 0.7,
+      "step": 888
+    },
+    {
+      "epoch": 0.28448,
+      "grad_norm": 0.5063808560371399,
+      "learning_rate": 0.00017172137710168135,
+      "loss": 0.7436,
+      "step": 889
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 0.39816007018089294,
+      "learning_rate": 0.00017168935148118494,
+      "loss": 0.5649,
+      "step": 890
+    },
+    {
+      "epoch": 0.28512,
+      "grad_norm": 0.49480873346328735,
+      "learning_rate": 0.00017165732586068856,
+      "loss": 0.7663,
+      "step": 891
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 0.5041493773460388,
+      "learning_rate": 0.00017162530024019218,
+      "loss": 0.8207,
+      "step": 892
+    },
+    {
+      "epoch": 0.28576,
+      "grad_norm": 0.40012243390083313,
+      "learning_rate": 0.00017159327461969577,
+      "loss": 0.7745,
+      "step": 893
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 0.4639826714992523,
+      "learning_rate": 0.00017156124899919936,
+      "loss": 0.7338,
+      "step": 894
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 0.4201996922492981,
+      "learning_rate": 0.00017152922337870298,
+      "loss": 0.7024,
+      "step": 895
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 0.39087820053100586,
+      "learning_rate": 0.00017149719775820657,
+      "loss": 0.6379,
+      "step": 896
+    },
+    {
+      "epoch": 0.28704,
+      "grad_norm": 0.48971229791641235,
+      "learning_rate": 0.00017146517213771019,
+      "loss": 0.8475,
+      "step": 897
+    },
+    {
+      "epoch": 0.28736,
+      "grad_norm": 0.3737387955188751,
+      "learning_rate": 0.00017143314651721378,
+      "loss": 0.6474,
+      "step": 898
+    },
+    {
+      "epoch": 0.28768,
+      "grad_norm": 0.44048476219177246,
+      "learning_rate": 0.0001714011208967174,
+      "loss": 0.6934,
+      "step": 899
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.433660626411438,
+      "learning_rate": 0.000171369095276221,
+      "loss": 0.7798,
+      "step": 900
+    },
+    {
+      "epoch": 0.28832,
+      "grad_norm": 0.3534420132637024,
+      "learning_rate": 0.00017133706965572458,
+      "loss": 0.5922,
+      "step": 901
+    },
+    {
+      "epoch": 0.28864,
+      "grad_norm": 0.4262845516204834,
+      "learning_rate": 0.0001713050440352282,
+      "loss": 0.6944,
+      "step": 902
+    },
+    {
+      "epoch": 0.28896,
+      "grad_norm": 0.3623132109642029,
+      "learning_rate": 0.0001712730184147318,
+      "loss": 0.6009,
+      "step": 903
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 0.43153759837150574,
+      "learning_rate": 0.0001712409927942354,
+      "loss": 0.5872,
+      "step": 904
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 0.4699711501598358,
+      "learning_rate": 0.000171208967173739,
+      "loss": 0.7665,
+      "step": 905
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 0.3532564342021942,
+      "learning_rate": 0.0001711769415532426,
+      "loss": 0.5828,
+      "step": 906
+    },
+    {
+      "epoch": 0.29024,
+      "grad_norm": 0.38483789563179016,
+      "learning_rate": 0.0001711449159327462,
+      "loss": 0.6734,
+      "step": 907
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 0.3330102562904358,
+      "learning_rate": 0.00017111289031224982,
+      "loss": 0.6079,
+      "step": 908
+    },
+    {
+      "epoch": 0.29088,
+      "grad_norm": 0.42011192440986633,
+      "learning_rate": 0.00017108086469175342,
+      "loss": 0.5806,
+      "step": 909
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.3864668309688568,
+      "learning_rate": 0.000171048839071257,
+      "loss": 0.7028,
+      "step": 910
+    },
+    {
+      "epoch": 0.29152,
+      "grad_norm": 0.3378417193889618,
+      "learning_rate": 0.00017101681345076063,
+      "loss": 0.5521,
+      "step": 911
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 0.43647605180740356,
+      "learning_rate": 0.00017098478783026422,
+      "loss": 0.6561,
+      "step": 912
+    },
+    {
+      "epoch": 0.29216,
+      "grad_norm": 0.4504333436489105,
+      "learning_rate": 0.0001709527622097678,
+      "loss": 0.6632,
+      "step": 913
+    },
+    {
+      "epoch": 0.29248,
+      "grad_norm": 0.4258662462234497,
+      "learning_rate": 0.00017092073658927143,
+      "loss": 0.7148,
+      "step": 914
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 0.3908085525035858,
+      "learning_rate": 0.00017088871096877504,
+      "loss": 0.661,
+      "step": 915
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 0.4813253879547119,
+      "learning_rate": 0.00017085668534827864,
+      "loss": 0.6935,
+      "step": 916
+    },
+    {
+      "epoch": 0.29344,
+      "grad_norm": 0.4666612446308136,
+      "learning_rate": 0.00017082465972778223,
+      "loss": 0.7315,
+      "step": 917
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 0.4547072649002075,
+      "learning_rate": 0.00017079263410728582,
+      "loss": 0.9344,
+      "step": 918
+    },
+    {
+      "epoch": 0.29408,
+      "grad_norm": 0.35851532220840454,
+      "learning_rate": 0.00017076060848678944,
+      "loss": 0.5378,
+      "step": 919
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.34555113315582275,
+      "learning_rate": 0.00017072858286629305,
+      "loss": 0.4802,
+      "step": 920
+    },
+    {
+      "epoch": 0.29472,
+      "grad_norm": 0.39396798610687256,
+      "learning_rate": 0.00017069655724579665,
+      "loss": 0.5898,
+      "step": 921
+    },
+    {
+      "epoch": 0.29504,
+      "grad_norm": 0.40469470620155334,
+      "learning_rate": 0.00017066453162530024,
+      "loss": 0.5508,
+      "step": 922
+    },
+    {
+      "epoch": 0.29536,
+      "grad_norm": 0.40479937195777893,
+      "learning_rate": 0.00017063250600480386,
+      "loss": 0.6165,
+      "step": 923
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 0.40274158120155334,
+      "learning_rate": 0.00017060048038430745,
+      "loss": 0.6501,
+      "step": 924
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.40161189436912537,
+      "learning_rate": 0.00017056845476381104,
+      "loss": 0.5344,
+      "step": 925
+    },
+    {
+      "epoch": 0.29632,
+      "grad_norm": 0.43137648701667786,
+      "learning_rate": 0.00017053642914331466,
+      "loss": 0.6859,
+      "step": 926
+    },
+    {
+      "epoch": 0.29664,
+      "grad_norm": 0.4131769835948944,
+      "learning_rate": 0.00017050440352281827,
+      "loss": 0.7532,
+      "step": 927
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 0.41452762484550476,
+      "learning_rate": 0.00017047237790232187,
+      "loss": 0.5441,
+      "step": 928
+    },
+    {
+      "epoch": 0.29728,
+      "grad_norm": 0.46458059549331665,
+      "learning_rate": 0.00017044035228182546,
+      "loss": 0.5902,
+      "step": 929
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.435787171125412,
+      "learning_rate": 0.00017040832666132908,
+      "loss": 0.619,
+      "step": 930
+    },
+    {
+      "epoch": 0.29792,
+      "grad_norm": 0.49662214517593384,
+      "learning_rate": 0.00017037630104083267,
+      "loss": 0.7034,
+      "step": 931
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 0.38956180214881897,
+      "learning_rate": 0.00017034427542033628,
+      "loss": 0.5485,
+      "step": 932
+    },
+    {
+      "epoch": 0.29856,
+      "grad_norm": 0.362155944108963,
+      "learning_rate": 0.00017031224979983988,
+      "loss": 0.5119,
+      "step": 933
+    },
+    {
+      "epoch": 0.29888,
+      "grad_norm": 0.4543428122997284,
+      "learning_rate": 0.0001702802241793435,
+      "loss": 0.7109,
+      "step": 934
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 0.42373567819595337,
+      "learning_rate": 0.00017024819855884709,
+      "loss": 0.7247,
+      "step": 935
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 0.3933662474155426,
+      "learning_rate": 0.00017021617293835068,
+      "loss": 0.6103,
+      "step": 936
+    },
+    {
+      "epoch": 0.29984,
+      "grad_norm": 0.39782580733299255,
+      "learning_rate": 0.00017018414731785427,
+      "loss": 0.6234,
+      "step": 937
+    },
+    {
+      "epoch": 0.30016,
+      "grad_norm": 0.4514563977718353,
+      "learning_rate": 0.0001701521216973579,
+      "loss": 0.7072,
+      "step": 938
+    },
+    {
+      "epoch": 0.30048,
+      "grad_norm": 0.3634970188140869,
+      "learning_rate": 0.0001701200960768615,
+      "loss": 0.5553,
+      "step": 939
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.49074140191078186,
+      "learning_rate": 0.0001700880704563651,
+      "loss": 0.7216,
+      "step": 940
+    },
+    {
+      "epoch": 0.30112,
+      "grad_norm": 0.4429180920124054,
+      "learning_rate": 0.0001700560448358687,
+      "loss": 0.7444,
+      "step": 941
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 0.445959210395813,
+      "learning_rate": 0.0001700240192153723,
+      "loss": 0.7055,
+      "step": 942
+    },
+    {
+      "epoch": 0.30176,
+      "grad_norm": 0.36917537450790405,
+      "learning_rate": 0.0001699919935948759,
+      "loss": 0.6682,
+      "step": 943
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 0.3927396535873413,
+      "learning_rate": 0.00016995996797437951,
+      "loss": 0.6862,
+      "step": 944
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.43677234649658203,
+      "learning_rate": 0.0001699279423538831,
+      "loss": 0.7691,
+      "step": 945
+    },
+    {
+      "epoch": 0.30272,
+      "grad_norm": 0.4811621606349945,
+      "learning_rate": 0.00016989591673338672,
+      "loss": 0.7664,
+      "step": 946
+    },
+    {
+      "epoch": 0.30304,
+      "grad_norm": 0.4000230133533478,
+      "learning_rate": 0.00016986389111289032,
+      "loss": 0.6389,
+      "step": 947
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 0.4446813464164734,
+      "learning_rate": 0.0001698318654923939,
+      "loss": 0.6526,
+      "step": 948
+    },
+    {
+      "epoch": 0.30368,
+      "grad_norm": 0.42321720719337463,
+      "learning_rate": 0.00016979983987189753,
+      "loss": 0.7324,
+      "step": 949
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.44998985528945923,
+      "learning_rate": 0.00016976781425140114,
+      "loss": 0.6239,
+      "step": 950
+    },
+    {
+      "epoch": 0.30432,
+      "grad_norm": 0.3940264880657196,
+      "learning_rate": 0.00016973578863090473,
+      "loss": 0.5563,
+      "step": 951
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 0.42343243956565857,
+      "learning_rate": 0.00016970376301040833,
+      "loss": 0.6326,
+      "step": 952
+    },
+    {
+      "epoch": 0.30496,
+      "grad_norm": 0.3229733407497406,
+      "learning_rate": 0.00016967173738991194,
+      "loss": 0.6085,
+      "step": 953
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 0.47269707918167114,
+      "learning_rate": 0.00016963971176941554,
+      "loss": 0.6501,
+      "step": 954
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 0.4230010509490967,
+      "learning_rate": 0.00016960768614891915,
+      "loss": 0.7023,
+      "step": 955
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 0.4384840130805969,
+      "learning_rate": 0.00016957566052842274,
+      "loss": 0.7518,
+      "step": 956
+    },
+    {
+      "epoch": 0.30624,
+      "grad_norm": 0.47585076093673706,
+      "learning_rate": 0.00016954363490792636,
+      "loss": 0.7521,
+      "step": 957
+    },
+    {
+      "epoch": 0.30656,
+      "grad_norm": 0.4414466917514801,
+      "learning_rate": 0.00016951160928742995,
+      "loss": 0.5663,
+      "step": 958
+    },
+    {
+      "epoch": 0.30688,
+      "grad_norm": 0.498874306678772,
+      "learning_rate": 0.00016947958366693355,
+      "loss": 0.7093,
+      "step": 959
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.3940095007419586,
+      "learning_rate": 0.00016944755804643714,
+      "loss": 0.5876,
+      "step": 960
+    },
+    {
+      "epoch": 0.30752,
+      "grad_norm": 0.403108149766922,
+      "learning_rate": 0.00016941553242594078,
+      "loss": 0.7941,
+      "step": 961
+    },
+    {
+      "epoch": 0.30784,
+      "grad_norm": 0.2765657901763916,
+      "learning_rate": 0.00016938350680544437,
+      "loss": 0.44,
+      "step": 962
+    },
+    {
+      "epoch": 0.30816,
+      "grad_norm": 0.43719595670700073,
+      "learning_rate": 0.00016935148118494796,
+      "loss": 0.6694,
+      "step": 963
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 0.40564650297164917,
+      "learning_rate": 0.00016931945556445156,
+      "loss": 0.5784,
+      "step": 964
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 0.366789847612381,
+      "learning_rate": 0.00016928742994395517,
+      "loss": 0.6088,
+      "step": 965
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 0.5077382922172546,
+      "learning_rate": 0.00016925540432345877,
+      "loss": 0.6424,
+      "step": 966
+    },
+    {
+      "epoch": 0.30944,
+      "grad_norm": 0.5236629247665405,
+      "learning_rate": 0.00016922337870296238,
+      "loss": 0.8114,
+      "step": 967
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 0.420754998922348,
+      "learning_rate": 0.00016919135308246597,
+      "loss": 0.6115,
+      "step": 968
+    },
+    {
+      "epoch": 0.31008,
+      "grad_norm": 0.42690345644950867,
+      "learning_rate": 0.0001691593274619696,
+      "loss": 0.7738,
+      "step": 969
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.3093375861644745,
+      "learning_rate": 0.00016912730184147318,
+      "loss": 0.456,
+      "step": 970
+    },
+    {
+      "epoch": 0.31072,
+      "grad_norm": 0.40798208117485046,
+      "learning_rate": 0.00016909527622097678,
+      "loss": 0.6869,
+      "step": 971
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 0.45086607336997986,
+      "learning_rate": 0.0001690632506004804,
+      "loss": 0.7184,
+      "step": 972
+    },
+    {
+      "epoch": 0.31136,
+      "grad_norm": 0.36980628967285156,
+      "learning_rate": 0.000169031224979984,
+      "loss": 0.6433,
+      "step": 973
+    },
+    {
+      "epoch": 0.31168,
+      "grad_norm": 0.3789350092411041,
+      "learning_rate": 0.0001689991993594876,
+      "loss": 0.7006,
+      "step": 974
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.4214223325252533,
+      "learning_rate": 0.0001689671737389912,
+      "loss": 0.7297,
+      "step": 975
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 0.35496053099632263,
+      "learning_rate": 0.0001689351481184948,
+      "loss": 0.5674,
+      "step": 976
+    },
+    {
+      "epoch": 0.31264,
+      "grad_norm": 0.4099990427494049,
+      "learning_rate": 0.0001689031224979984,
+      "loss": 0.6001,
+      "step": 977
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 0.5112151503562927,
+      "learning_rate": 0.000168871096877502,
+      "loss": 0.7035,
+      "step": 978
+    },
+    {
+      "epoch": 0.31328,
+      "grad_norm": 0.47532418370246887,
+      "learning_rate": 0.00016883907125700561,
+      "loss": 0.6442,
+      "step": 979
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.4672641158103943,
+      "learning_rate": 0.00016880704563650923,
+      "loss": 0.6775,
+      "step": 980
+    },
+    {
+      "epoch": 0.31392,
+      "grad_norm": 0.3189723789691925,
+      "learning_rate": 0.00016877502001601282,
+      "loss": 0.5108,
+      "step": 981
+    },
+    {
+      "epoch": 0.31424,
+      "grad_norm": 0.3973686397075653,
+      "learning_rate": 0.00016874299439551641,
+      "loss": 0.7406,
+      "step": 982
+    },
+    {
+      "epoch": 0.31456,
+      "grad_norm": 0.47230929136276245,
+      "learning_rate": 0.00016871096877502,
+      "loss": 0.7365,
+      "step": 983
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 0.4369361102581024,
+      "learning_rate": 0.00016867894315452362,
+      "loss": 0.6952,
+      "step": 984
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 0.44876450300216675,
+      "learning_rate": 0.00016864691753402724,
+      "loss": 0.6296,
+      "step": 985
+    },
+    {
+      "epoch": 0.31552,
+      "grad_norm": 0.40320950746536255,
+      "learning_rate": 0.00016861489191353083,
+      "loss": 0.6543,
+      "step": 986
+    },
+    {
+      "epoch": 0.31584,
+      "grad_norm": 0.4378165006637573,
+      "learning_rate": 0.00016858286629303442,
+      "loss": 0.7557,
+      "step": 987
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 0.4394074082374573,
+      "learning_rate": 0.00016855084067253804,
+      "loss": 0.8531,
+      "step": 988
+    },
+    {
+      "epoch": 0.31648,
+      "grad_norm": 0.4344981908798218,
+      "learning_rate": 0.00016851881505204163,
+      "loss": 0.6462,
+      "step": 989
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.39521080255508423,
+      "learning_rate": 0.00016848678943154523,
+      "loss": 0.777,
+      "step": 990
+    },
+    {
+      "epoch": 0.31712,
+      "grad_norm": 0.4416322708129883,
+      "learning_rate": 0.00016845476381104884,
+      "loss": 0.6655,
+      "step": 991
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 0.36917340755462646,
+      "learning_rate": 0.00016842273819055246,
+      "loss": 0.6444,
+      "step": 992
+    },
+    {
+      "epoch": 0.31776,
+      "grad_norm": 0.42573603987693787,
+      "learning_rate": 0.00016839071257005605,
+      "loss": 0.6864,
+      "step": 993
+    },
+    {
+      "epoch": 0.31808,
+      "grad_norm": 0.47259828448295593,
+      "learning_rate": 0.00016835868694955964,
+      "loss": 0.5996,
+      "step": 994
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 0.408227801322937,
+      "learning_rate": 0.00016832666132906326,
+      "loss": 0.5774,
+      "step": 995
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 0.4264174699783325,
+      "learning_rate": 0.00016829463570856688,
+      "loss": 0.8324,
+      "step": 996
+    },
+    {
+      "epoch": 0.31904,
+      "grad_norm": 0.39352893829345703,
+      "learning_rate": 0.00016826261008807047,
+      "loss": 0.6803,
+      "step": 997
+    },
+    {
+      "epoch": 0.31936,
+      "grad_norm": 0.39909815788269043,
+      "learning_rate": 0.00016823058446757406,
+      "loss": 0.6467,
+      "step": 998
+    },
+    {
+      "epoch": 0.31968,
+      "grad_norm": 0.4081866145133972,
+      "learning_rate": 0.00016819855884707768,
+      "loss": 0.549,
+      "step": 999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4132840633392334,
+      "learning_rate": 0.00016816653322658127,
+      "loss": 0.7142,
+      "step": 1000
+    },
+    {
+      "epoch": 0.32032,
+      "grad_norm": 0.38809019327163696,
+      "learning_rate": 0.00016813450760608486,
+      "loss": 0.432,
+      "step": 1001
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 0.37467271089553833,
+      "learning_rate": 0.00016810248198558848,
+      "loss": 0.6472,
+      "step": 1002
+    },
+    {
+      "epoch": 0.32096,
+      "grad_norm": 0.3836999237537384,
+      "learning_rate": 0.0001680704563650921,
+      "loss": 0.63,
+      "step": 1003
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 0.2742338478565216,
+      "learning_rate": 0.0001680384307445957,
+      "loss": 0.4817,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 0.35868972539901733,
+      "learning_rate": 0.00016800640512409928,
+      "loss": 0.5375,
+      "step": 1005
+    },
+    {
+      "epoch": 0.32192,
+      "grad_norm": 0.42052707076072693,
+      "learning_rate": 0.00016797437950360287,
+      "loss": 0.5335,
+      "step": 1006
+    },
+    {
+      "epoch": 0.32224,
+      "grad_norm": 0.34040114283561707,
+      "learning_rate": 0.0001679423538831065,
+      "loss": 0.5711,
+      "step": 1007
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 0.5313262939453125,
+      "learning_rate": 0.0001679103282626101,
+      "loss": 0.7721,
+      "step": 1008
+    },
+    {
+      "epoch": 0.32288,
+      "grad_norm": 0.5862169861793518,
+      "learning_rate": 0.0001678783026421137,
+      "loss": 0.6056,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 0.45689085125923157,
+      "learning_rate": 0.0001678462770216173,
+      "loss": 0.6024,
+      "step": 1010
+    },
+    {
+      "epoch": 0.32352,
+      "grad_norm": 0.412634938955307,
+      "learning_rate": 0.0001678142514011209,
+      "loss": 0.5941,
+      "step": 1011
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 0.40237438678741455,
+      "learning_rate": 0.0001677822257806245,
+      "loss": 0.5581,
+      "step": 1012
+    },
+    {
+      "epoch": 0.32416,
+      "grad_norm": 0.4812946617603302,
+      "learning_rate": 0.0001677502001601281,
+      "loss": 0.7668,
+      "step": 1013
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 0.4496348202228546,
+      "learning_rate": 0.0001677181745396317,
+      "loss": 0.5999,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 0.5074209570884705,
+      "learning_rate": 0.00016768614891913533,
+      "loss": 0.6054,
+      "step": 1015
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 0.4500525891780853,
+      "learning_rate": 0.00016765412329863892,
+      "loss": 0.5458,
+      "step": 1016
+    },
+    {
+      "epoch": 0.32544,
+      "grad_norm": 0.3874082863330841,
+      "learning_rate": 0.0001676220976781425,
+      "loss": 0.6135,
+      "step": 1017
+    },
+    {
+      "epoch": 0.32576,
+      "grad_norm": 0.4293360412120819,
+      "learning_rate": 0.00016759007205764613,
+      "loss": 0.689,
+      "step": 1018
+    },
+    {
+      "epoch": 0.32608,
+      "grad_norm": 0.48183199763298035,
+      "learning_rate": 0.00016755804643714972,
+      "loss": 0.6461,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.411663293838501,
+      "learning_rate": 0.00016752602081665334,
+      "loss": 0.6643,
+      "step": 1020
+    },
+    {
+      "epoch": 0.32672,
+      "grad_norm": 0.4574558734893799,
+      "learning_rate": 0.00016749399519615693,
+      "loss": 0.6872,
+      "step": 1021
+    },
+    {
+      "epoch": 0.32704,
+      "grad_norm": 0.48675090074539185,
+      "learning_rate": 0.00016746196957566055,
+      "loss": 0.6852,
+      "step": 1022
+    },
+    {
+      "epoch": 0.32736,
+      "grad_norm": 0.4179941415786743,
+      "learning_rate": 0.00016742994395516414,
+      "loss": 0.6086,
+      "step": 1023
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 0.389563649892807,
+      "learning_rate": 0.00016739791833466773,
+      "loss": 0.5242,
+      "step": 1024
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 0.46075987815856934,
+      "learning_rate": 0.00016736589271417132,
+      "loss": 0.7741,
+      "step": 1025
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 0.3539038300514221,
+      "learning_rate": 0.00016733386709367497,
+      "loss": 0.5516,
+      "step": 1026
+    },
+    {
+      "epoch": 0.32864,
+      "grad_norm": 0.39535990357398987,
+      "learning_rate": 0.00016730184147317856,
+      "loss": 0.5482,
+      "step": 1027
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 0.46230560541152954,
+      "learning_rate": 0.00016726981585268215,
+      "loss": 0.5622,
+      "step": 1028
+    },
+    {
+      "epoch": 0.32928,
+      "grad_norm": 0.3049105703830719,
+      "learning_rate": 0.00016723779023218574,
+      "loss": 0.6376,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 0.4199695587158203,
+      "learning_rate": 0.00016720576461168936,
+      "loss": 0.8096,
+      "step": 1030
+    },
+    {
+      "epoch": 0.32992,
+      "grad_norm": 0.34019169211387634,
+      "learning_rate": 0.00016717373899119295,
+      "loss": 0.5612,
+      "step": 1031
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 0.4193580150604248,
+      "learning_rate": 0.00016714171337069657,
+      "loss": 0.551,
+      "step": 1032
+    },
+    {
+      "epoch": 0.33056,
+      "grad_norm": 0.42196691036224365,
+      "learning_rate": 0.00016710968775020016,
+      "loss": 0.6787,
+      "step": 1033
+    },
+    {
+      "epoch": 0.33088,
+      "grad_norm": 0.3484210669994354,
+      "learning_rate": 0.00016707766212970378,
+      "loss": 0.4743,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 0.38108736276626587,
+      "learning_rate": 0.00016704563650920737,
+      "loss": 0.712,
+      "step": 1035
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 0.6504843831062317,
+      "learning_rate": 0.00016701361088871096,
+      "loss": 0.7672,
+      "step": 1036
+    },
+    {
+      "epoch": 0.33184,
+      "grad_norm": 0.3913379907608032,
+      "learning_rate": 0.00016698158526821458,
+      "loss": 0.6958,
+      "step": 1037
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 0.41113564372062683,
+      "learning_rate": 0.0001669495596477182,
+      "loss": 0.6576,
+      "step": 1038
+    },
+    {
+      "epoch": 0.33248,
+      "grad_norm": 0.4593764543533325,
+      "learning_rate": 0.0001669175340272218,
+      "loss": 0.579,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.35736462473869324,
+      "learning_rate": 0.00016688550840672538,
+      "loss": 0.6763,
+      "step": 1040
+    },
+    {
+      "epoch": 0.33312,
+      "grad_norm": 0.4011344015598297,
+      "learning_rate": 0.000166853482786229,
+      "loss": 0.8068,
+      "step": 1041
+    },
+    {
+      "epoch": 0.33344,
+      "grad_norm": 0.41379818320274353,
+      "learning_rate": 0.0001668214571657326,
+      "loss": 0.7818,
+      "step": 1042
+    },
+    {
+      "epoch": 0.33376,
+      "grad_norm": 0.3271898329257965,
+      "learning_rate": 0.0001667894315452362,
+      "loss": 0.5377,
+      "step": 1043
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 0.40039142966270447,
+      "learning_rate": 0.0001667574059247398,
+      "loss": 0.7142,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 0.42146992683410645,
+      "learning_rate": 0.0001667253803042434,
+      "loss": 0.6993,
+      "step": 1045
+    },
+    {
+      "epoch": 0.33472,
+      "grad_norm": 0.39791202545166016,
+      "learning_rate": 0.000166693354683747,
+      "loss": 0.632,
+      "step": 1046
+    },
+    {
+      "epoch": 0.33504,
+      "grad_norm": 0.34557947516441345,
+      "learning_rate": 0.0001666613290632506,
+      "loss": 0.6114,
+      "step": 1047
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 0.3994189202785492,
+      "learning_rate": 0.0001666293034427542,
+      "loss": 0.617,
+      "step": 1048
+    },
+    {
+      "epoch": 0.33568,
+      "grad_norm": 0.38758665323257446,
+      "learning_rate": 0.0001665972778222578,
+      "loss": 0.5861,
+      "step": 1049
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.4099954664707184,
+      "learning_rate": 0.00016656525220176143,
+      "loss": 0.7154,
+      "step": 1050
+    },
+    {
+      "epoch": 0.33632,
+      "grad_norm": 0.3895832598209381,
+      "learning_rate": 0.00016653322658126502,
+      "loss": 0.7062,
+      "step": 1051
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 0.3152011036872864,
+      "learning_rate": 0.0001665012009607686,
+      "loss": 0.502,
+      "step": 1052
+    },
+    {
+      "epoch": 0.33696,
+      "grad_norm": 0.4073959290981293,
+      "learning_rate": 0.00016646917534027223,
+      "loss": 0.7012,
+      "step": 1053
+    },
+    {
+      "epoch": 0.33728,
+      "grad_norm": 0.4422577917575836,
+      "learning_rate": 0.00016643714971977582,
+      "loss": 0.6974,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 0.3676719069480896,
+      "learning_rate": 0.00016640512409927944,
+      "loss": 0.5862,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 0.38448798656463623,
+      "learning_rate": 0.00016637309847878303,
+      "loss": 0.563,
+      "step": 1056
+    },
+    {
+      "epoch": 0.33824,
+      "grad_norm": 0.45355701446533203,
+      "learning_rate": 0.00016634107285828665,
+      "loss": 0.7017,
+      "step": 1057
+    },
+    {
+      "epoch": 0.33856,
+      "grad_norm": 0.422608882188797,
+      "learning_rate": 0.00016630904723779024,
+      "loss": 0.6684,
+      "step": 1058
+    },
+    {
+      "epoch": 0.33888,
+      "grad_norm": 0.49056288599967957,
+      "learning_rate": 0.00016627702161729383,
+      "loss": 0.7292,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.39333677291870117,
+      "learning_rate": 0.00016624499599679745,
+      "loss": 0.6533,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33952,
+      "grad_norm": 0.43318378925323486,
+      "learning_rate": 0.00016621297037630107,
+      "loss": 0.719,
+      "step": 1061
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 0.40490639209747314,
+      "learning_rate": 0.00016618094475580466,
+      "loss": 0.7369,
+      "step": 1062
+    },
+    {
+      "epoch": 0.34016,
+      "grad_norm": 0.376602441072464,
+      "learning_rate": 0.00016614891913530825,
+      "loss": 0.6418,
+      "step": 1063
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 0.4271795153617859,
+      "learning_rate": 0.00016611689351481184,
+      "loss": 0.6434,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 0.4215434789657593,
+      "learning_rate": 0.00016608486789431546,
+      "loss": 0.6856,
+      "step": 1065
+    },
+    {
+      "epoch": 0.34112,
+      "grad_norm": 0.4221090078353882,
+      "learning_rate": 0.00016605284227381905,
+      "loss": 0.5771,
+      "step": 1066
+    },
+    {
+      "epoch": 0.34144,
+      "grad_norm": 0.38433629274368286,
+      "learning_rate": 0.00016602081665332267,
+      "loss": 0.5829,
+      "step": 1067
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 0.4141973555088043,
+      "learning_rate": 0.00016598879103282626,
+      "loss": 0.6656,
+      "step": 1068
+    },
+    {
+      "epoch": 0.34208,
+      "grad_norm": 0.43190574645996094,
+      "learning_rate": 0.00016595676541232988,
+      "loss": 0.6508,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 0.4205007553100586,
+      "learning_rate": 0.00016592473979183347,
+      "loss": 0.7049,
+      "step": 1070
+    },
+    {
+      "epoch": 0.34272,
+      "grad_norm": 0.36378005146980286,
+      "learning_rate": 0.00016589271417133706,
+      "loss": 0.6277,
+      "step": 1071
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 0.47688454389572144,
+      "learning_rate": 0.00016586068855084068,
+      "loss": 0.7789,
+      "step": 1072
+    },
+    {
+      "epoch": 0.34336,
+      "grad_norm": 0.41992178559303284,
+      "learning_rate": 0.0001658286629303443,
+      "loss": 0.6191,
+      "step": 1073
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 0.511837899684906,
+      "learning_rate": 0.0001657966373098479,
+      "loss": 0.674,
+      "step": 1074
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 0.4289505183696747,
+      "learning_rate": 0.00016576461168935148,
+      "loss": 0.5891,
+      "step": 1075
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 0.4322727918624878,
+      "learning_rate": 0.0001657325860688551,
+      "loss": 0.5998,
+      "step": 1076
+    },
+    {
+      "epoch": 0.34464,
+      "grad_norm": 0.4376969635486603,
+      "learning_rate": 0.0001657005604483587,
+      "loss": 0.6586,
+      "step": 1077
+    },
+    {
+      "epoch": 0.34496,
+      "grad_norm": 0.4900533854961395,
+      "learning_rate": 0.00016566853482786228,
+      "loss": 0.6499,
+      "step": 1078
+    },
+    {
+      "epoch": 0.34528,
+      "grad_norm": 0.41336333751678467,
+      "learning_rate": 0.0001656365092073659,
+      "loss": 0.8146,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.4090036153793335,
+      "learning_rate": 0.00016560448358686952,
+      "loss": 0.6961,
+      "step": 1080
+    },
+    {
+      "epoch": 0.34592,
+      "grad_norm": 0.39981621503829956,
+      "learning_rate": 0.0001655724579663731,
+      "loss": 0.6468,
+      "step": 1081
+    },
+    {
+      "epoch": 0.34624,
+      "grad_norm": 0.42005306482315063,
+      "learning_rate": 0.0001655404323458767,
+      "loss": 0.637,
+      "step": 1082
+    },
+    {
+      "epoch": 0.34656,
+      "grad_norm": 0.400165319442749,
+      "learning_rate": 0.0001655084067253803,
+      "loss": 0.6068,
+      "step": 1083
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 0.40222734212875366,
+      "learning_rate": 0.0001654763811048839,
+      "loss": 0.6269,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 0.3835143744945526,
+      "learning_rate": 0.00016544435548438753,
+      "loss": 0.5255,
+      "step": 1085
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 0.4672694504261017,
+      "learning_rate": 0.00016541232986389112,
+      "loss": 0.5228,
+      "step": 1086
+    },
+    {
+      "epoch": 0.34784,
+      "grad_norm": 0.3609679937362671,
+      "learning_rate": 0.0001653803042433947,
+      "loss": 0.5574,
+      "step": 1087
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 0.45509204268455505,
+      "learning_rate": 0.00016534827862289833,
+      "loss": 0.7522,
+      "step": 1088
+    },
+    {
+      "epoch": 0.34848,
+      "grad_norm": 0.41838210821151733,
+      "learning_rate": 0.00016531625300240192,
+      "loss": 0.6577,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 0.3594484329223633,
+      "learning_rate": 0.00016528422738190554,
+      "loss": 0.6032,
+      "step": 1090
+    },
+    {
+      "epoch": 0.34912,
+      "grad_norm": 0.39003702998161316,
+      "learning_rate": 0.00016525220176140913,
+      "loss": 0.732,
+      "step": 1091
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 0.35818225145339966,
+      "learning_rate": 0.00016522017614091275,
+      "loss": 0.5456,
+      "step": 1092
+    },
+    {
+      "epoch": 0.34976,
+      "grad_norm": 0.42233800888061523,
+      "learning_rate": 0.00016518815052041634,
+      "loss": 0.5765,
+      "step": 1093
+    },
+    {
+      "epoch": 0.35008,
+      "grad_norm": 0.4186314642429352,
+      "learning_rate": 0.00016515612489991993,
+      "loss": 0.7149,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 0.45738333463668823,
+      "learning_rate": 0.00016512409927942355,
+      "loss": 0.7916,
+      "step": 1095
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 0.426444947719574,
+      "learning_rate": 0.00016509207365892717,
+      "loss": 0.7064,
+      "step": 1096
+    },
+    {
+      "epoch": 0.35104,
+      "grad_norm": 0.3971715569496155,
+      "learning_rate": 0.00016506004803843076,
+      "loss": 0.7151,
+      "step": 1097
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 0.42331504821777344,
+      "learning_rate": 0.00016502802241793435,
+      "loss": 0.6745,
+      "step": 1098
+    },
+    {
+      "epoch": 0.35168,
+      "grad_norm": 0.3904751241207123,
+      "learning_rate": 0.00016499599679743797,
+      "loss": 0.4705,
+      "step": 1099
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.4333479106426239,
+      "learning_rate": 0.00016496397117694156,
+      "loss": 0.5748,
+      "step": 1100
+    },
+    {
+      "epoch": 0.35232,
+      "grad_norm": 0.4009312391281128,
+      "learning_rate": 0.00016493194555644515,
+      "loss": 0.6275,
+      "step": 1101
+    },
+    {
+      "epoch": 0.35264,
+      "grad_norm": 0.3873033821582794,
+      "learning_rate": 0.00016489991993594877,
+      "loss": 0.6907,
+      "step": 1102
+    },
+    {
+      "epoch": 0.35296,
+      "grad_norm": 0.46167290210723877,
+      "learning_rate": 0.0001648678943154524,
+      "loss": 0.7712,
+      "step": 1103
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 0.44939619302749634,
+      "learning_rate": 0.00016483586869495598,
+      "loss": 0.6833,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 0.43399548530578613,
+      "learning_rate": 0.00016480384307445957,
+      "loss": 0.6902,
+      "step": 1105
+    },
+    {
+      "epoch": 0.35392,
+      "grad_norm": 0.38739871978759766,
+      "learning_rate": 0.00016477181745396316,
+      "loss": 0.6575,
+      "step": 1106
+    },
+    {
+      "epoch": 0.35424,
+      "grad_norm": 0.3969883918762207,
+      "learning_rate": 0.00016473979183346678,
+      "loss": 0.5675,
+      "step": 1107
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 0.3381742835044861,
+      "learning_rate": 0.0001647077662129704,
+      "loss": 0.5747,
+      "step": 1108
+    },
+    {
+      "epoch": 0.35488,
+      "grad_norm": 0.3955135643482208,
+      "learning_rate": 0.000164675740592474,
+      "loss": 0.6254,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.420635849237442,
+      "learning_rate": 0.00016464371497197758,
+      "loss": 0.7096,
+      "step": 1110
+    },
+    {
+      "epoch": 0.35552,
+      "grad_norm": 0.3653455376625061,
+      "learning_rate": 0.0001646116893514812,
+      "loss": 0.7154,
+      "step": 1111
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 0.3863634765148163,
+      "learning_rate": 0.0001645796637309848,
+      "loss": 0.5747,
+      "step": 1112
+    },
+    {
+      "epoch": 0.35616,
+      "grad_norm": 0.3749634325504303,
+      "learning_rate": 0.00016454763811048838,
+      "loss": 0.6786,
+      "step": 1113
+    },
+    {
+      "epoch": 0.35648,
+      "grad_norm": 0.3677906095981598,
+      "learning_rate": 0.000164515612489992,
+      "loss": 0.6235,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 0.39175957441329956,
+      "learning_rate": 0.00016448358686949562,
+      "loss": 0.5892,
+      "step": 1115
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 0.3403246998786926,
+      "learning_rate": 0.0001644515612489992,
+      "loss": 0.6083,
+      "step": 1116
+    },
+    {
+      "epoch": 0.35744,
+      "grad_norm": 0.3840944766998291,
+      "learning_rate": 0.0001644195356285028,
+      "loss": 0.5997,
+      "step": 1117
+    },
+    {
+      "epoch": 0.35776,
+      "grad_norm": 0.3962903916835785,
+      "learning_rate": 0.00016438751000800642,
+      "loss": 0.5739,
+      "step": 1118
+    },
+    {
+      "epoch": 0.35808,
+      "grad_norm": 0.3809589445590973,
+      "learning_rate": 0.00016435548438751,
+      "loss": 0.645,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.30100080370903015,
+      "learning_rate": 0.00016432345876701363,
+      "loss": 0.5044,
+      "step": 1120
+    },
+    {
+      "epoch": 0.35872,
+      "grad_norm": 0.38952749967575073,
+      "learning_rate": 0.00016429143314651722,
+      "loss": 0.701,
+      "step": 1121
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 0.33911648392677307,
+      "learning_rate": 0.00016425940752602084,
+      "loss": 0.6076,
+      "step": 1122
+    },
+    {
+      "epoch": 0.35936,
+      "grad_norm": 0.43114909529685974,
+      "learning_rate": 0.00016422738190552443,
+      "loss": 0.668,
+      "step": 1123
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 0.4059431552886963,
+      "learning_rate": 0.00016419535628502802,
+      "loss": 0.6524,
+      "step": 1124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4674667716026306,
+      "learning_rate": 0.0001641633306645316,
+      "loss": 0.6255,
+      "step": 1125
+    },
+    {
+      "epoch": 0.36032,
+      "grad_norm": 0.3827831745147705,
+      "learning_rate": 0.00016413130504403526,
+      "loss": 0.5861,
+      "step": 1126
+    },
+    {
+      "epoch": 0.36064,
+      "grad_norm": 0.4293474555015564,
+      "learning_rate": 0.00016409927942353885,
+      "loss": 0.5988,
+      "step": 1127
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 0.3655852675437927,
+      "learning_rate": 0.00016406725380304244,
+      "loss": 0.5339,
+      "step": 1128
+    },
+    {
+      "epoch": 0.36128,
+      "grad_norm": 0.3740267753601074,
+      "learning_rate": 0.00016403522818254603,
+      "loss": 0.6204,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.41032421588897705,
+      "learning_rate": 0.00016400320256204965,
+      "loss": 0.5297,
+      "step": 1130
+    },
+    {
+      "epoch": 0.36192,
+      "grad_norm": 0.5052286386489868,
+      "learning_rate": 0.00016397117694155324,
+      "loss": 0.5971,
+      "step": 1131
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 0.36573678255081177,
+      "learning_rate": 0.00016393915132105686,
+      "loss": 0.4231,
+      "step": 1132
+    },
+    {
+      "epoch": 0.36256,
+      "grad_norm": 0.4147803783416748,
+      "learning_rate": 0.00016390712570056045,
+      "loss": 0.4925,
+      "step": 1133
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 0.4554596543312073,
+      "learning_rate": 0.00016387510008006407,
+      "loss": 0.5479,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 0.44896307587623596,
+      "learning_rate": 0.00016384307445956766,
+      "loss": 0.7876,
+      "step": 1135
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 0.398017555475235,
+      "learning_rate": 0.00016381104883907125,
+      "loss": 0.6091,
+      "step": 1136
+    },
+    {
+      "epoch": 0.36384,
+      "grad_norm": 0.4339744746685028,
+      "learning_rate": 0.00016377902321857487,
+      "loss": 0.717,
+      "step": 1137
+    },
+    {
+      "epoch": 0.36416,
+      "grad_norm": 0.3925042152404785,
+      "learning_rate": 0.00016374699759807849,
+      "loss": 0.7057,
+      "step": 1138
+    },
+    {
+      "epoch": 0.36448,
+      "grad_norm": 0.44340166449546814,
+      "learning_rate": 0.00016371497197758208,
+      "loss": 0.7533,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.6709069013595581,
+      "learning_rate": 0.00016368294635708567,
+      "loss": 0.4548,
+      "step": 1140
+    },
+    {
+      "epoch": 0.36512,
+      "grad_norm": 0.41551288962364197,
+      "learning_rate": 0.0001636509207365893,
+      "loss": 0.5245,
+      "step": 1141
+    },
+    {
+      "epoch": 0.36544,
+      "grad_norm": 0.4433261752128601,
+      "learning_rate": 0.00016361889511609288,
+      "loss": 0.6583,
+      "step": 1142
+    },
+    {
+      "epoch": 0.36576,
+      "grad_norm": 0.45843756198883057,
+      "learning_rate": 0.0001635868694955965,
+      "loss": 0.7845,
+      "step": 1143
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 0.41276583075523376,
+      "learning_rate": 0.0001635548438751001,
+      "loss": 0.662,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 0.4211631119251251,
+      "learning_rate": 0.0001635228182546037,
+      "loss": 0.7264,
+      "step": 1145
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 0.446582168340683,
+      "learning_rate": 0.0001634907926341073,
+      "loss": 0.5521,
+      "step": 1146
+    },
+    {
+      "epoch": 0.36704,
+      "grad_norm": 0.3793337941169739,
+      "learning_rate": 0.0001634587670136109,
+      "loss": 0.548,
+      "step": 1147
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 0.4421611428260803,
+      "learning_rate": 0.00016342674139311448,
+      "loss": 0.6889,
+      "step": 1148
+    },
+    {
+      "epoch": 0.36768,
+      "grad_norm": 0.4752484858036041,
+      "learning_rate": 0.00016339471577261813,
+      "loss": 0.8062,
+      "step": 1149
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.43441566824913025,
+      "learning_rate": 0.00016336269015212172,
+      "loss": 0.738,
+      "step": 1150
+    },
+    {
+      "epoch": 0.36832,
+      "grad_norm": 0.3767280876636505,
+      "learning_rate": 0.0001633306645316253,
+      "loss": 0.6554,
+      "step": 1151
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 0.3409682810306549,
+      "learning_rate": 0.0001632986389111289,
+      "loss": 0.5336,
+      "step": 1152
+    },
+    {
+      "epoch": 0.36896,
+      "grad_norm": 0.40926170349121094,
+      "learning_rate": 0.00016326661329063252,
+      "loss": 0.5541,
+      "step": 1153
+    },
+    {
+      "epoch": 0.36928,
+      "grad_norm": 0.38607579469680786,
+      "learning_rate": 0.0001632345876701361,
+      "loss": 0.6184,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 0.4159526526927948,
+      "learning_rate": 0.00016320256204963973,
+      "loss": 0.6428,
+      "step": 1155
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 0.3682723045349121,
+      "learning_rate": 0.00016317053642914332,
+      "loss": 0.5275,
+      "step": 1156
+    },
+    {
+      "epoch": 0.37024,
+      "grad_norm": 0.43406593799591064,
+      "learning_rate": 0.00016313851080864694,
+      "loss": 0.7019,
+      "step": 1157
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 0.4051198363304138,
+      "learning_rate": 0.00016310648518815053,
+      "loss": 0.5763,
+      "step": 1158
+    },
+    {
+      "epoch": 0.37088,
+      "grad_norm": 0.39975258708000183,
+      "learning_rate": 0.00016307445956765412,
+      "loss": 0.637,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.4424419701099396,
+      "learning_rate": 0.00016304243394715774,
+      "loss": 0.6727,
+      "step": 1160
+    },
+    {
+      "epoch": 0.37152,
+      "grad_norm": 0.37753763794898987,
+      "learning_rate": 0.00016301040832666136,
+      "loss": 0.608,
+      "step": 1161
+    },
+    {
+      "epoch": 0.37184,
+      "grad_norm": 0.39222219586372375,
+      "learning_rate": 0.00016297838270616495,
+      "loss": 0.5399,
+      "step": 1162
+    },
+    {
+      "epoch": 0.37216,
+      "grad_norm": 0.3995283544063568,
+      "learning_rate": 0.00016294635708566854,
+      "loss": 0.6747,
+      "step": 1163
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 0.4672609865665436,
+      "learning_rate": 0.00016291433146517216,
+      "loss": 0.5712,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 0.4909473955631256,
+      "learning_rate": 0.00016288230584467575,
+      "loss": 0.8096,
+      "step": 1165
+    },
+    {
+      "epoch": 0.37312,
+      "grad_norm": 0.3612677752971649,
+      "learning_rate": 0.00016285028022417934,
+      "loss": 0.5342,
+      "step": 1166
+    },
+    {
+      "epoch": 0.37344,
+      "grad_norm": 0.4157281517982483,
+      "learning_rate": 0.00016281825460368296,
+      "loss": 0.6049,
+      "step": 1167
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 0.40507444739341736,
+      "learning_rate": 0.00016278622898318657,
+      "loss": 0.6122,
+      "step": 1168
+    },
+    {
+      "epoch": 0.37408,
+      "grad_norm": 0.38488927483558655,
+      "learning_rate": 0.00016275420336269017,
+      "loss": 0.639,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.4284215569496155,
+      "learning_rate": 0.00016272217774219376,
+      "loss": 0.6768,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37472,
+      "grad_norm": 0.4549587368965149,
+      "learning_rate": 0.00016269015212169735,
+      "loss": 0.4713,
+      "step": 1171
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 0.5018149018287659,
+      "learning_rate": 0.00016265812650120097,
+      "loss": 0.7551,
+      "step": 1172
+    },
+    {
+      "epoch": 0.37536,
+      "grad_norm": 0.4649817943572998,
+      "learning_rate": 0.00016262610088070459,
+      "loss": 0.6426,
+      "step": 1173
+    },
+    {
+      "epoch": 0.37568,
+      "grad_norm": 0.4837520718574524,
+      "learning_rate": 0.00016259407526020818,
+      "loss": 0.6272,
+      "step": 1174
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.4752795994281769,
+      "learning_rate": 0.00016256204963971177,
+      "loss": 0.709,
+      "step": 1175
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 0.465518057346344,
+      "learning_rate": 0.00016253002401921539,
+      "loss": 0.4497,
+      "step": 1176
+    },
+    {
+      "epoch": 0.37664,
+      "grad_norm": 0.3766709268093109,
+      "learning_rate": 0.00016249799839871898,
+      "loss": 0.4884,
+      "step": 1177
+    },
+    {
+      "epoch": 0.37696,
+      "grad_norm": 0.26107364892959595,
+      "learning_rate": 0.00016246597277822257,
+      "loss": 0.3469,
+      "step": 1178
+    },
+    {
+      "epoch": 0.37728,
+      "grad_norm": 0.41186803579330444,
+      "learning_rate": 0.0001624339471577262,
+      "loss": 0.6578,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.46515098214149475,
+      "learning_rate": 0.0001624019215372298,
+      "loss": 0.6122,
+      "step": 1180
+    },
+    {
+      "epoch": 0.37792,
+      "grad_norm": 0.4423938989639282,
+      "learning_rate": 0.0001623698959167334,
+      "loss": 0.6862,
+      "step": 1181
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 0.39101600646972656,
+      "learning_rate": 0.000162337870296237,
+      "loss": 0.6097,
+      "step": 1182
+    },
+    {
+      "epoch": 0.37856,
+      "grad_norm": 0.374570369720459,
+      "learning_rate": 0.0001623058446757406,
+      "loss": 0.6109,
+      "step": 1183
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 0.2974463403224945,
+      "learning_rate": 0.00016227381905524422,
+      "loss": 0.4431,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 0.3574426770210266,
+      "learning_rate": 0.00016224179343474782,
+      "loss": 0.6802,
+      "step": 1185
+    },
+    {
+      "epoch": 0.37952,
+      "grad_norm": 0.3407016694545746,
+      "learning_rate": 0.0001622097678142514,
+      "loss": 0.5145,
+      "step": 1186
+    },
+    {
+      "epoch": 0.37984,
+      "grad_norm": 0.4443570077419281,
+      "learning_rate": 0.000162177742193755,
+      "loss": 0.6595,
+      "step": 1187
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 0.42468732595443726,
+      "learning_rate": 0.00016214571657325862,
+      "loss": 0.6859,
+      "step": 1188
+    },
+    {
+      "epoch": 0.38048,
+      "grad_norm": 0.38645291328430176,
+      "learning_rate": 0.0001621136909527622,
+      "loss": 0.7019,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.35937097668647766,
+      "learning_rate": 0.00016208166533226583,
+      "loss": 0.5775,
+      "step": 1190
+    },
+    {
+      "epoch": 0.38112,
+      "grad_norm": 0.3846529424190521,
+      "learning_rate": 0.00016204963971176942,
+      "loss": 0.5702,
+      "step": 1191
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 0.42936986684799194,
+      "learning_rate": 0.00016201761409127304,
+      "loss": 0.6658,
+      "step": 1192
+    },
+    {
+      "epoch": 0.38176,
+      "grad_norm": 0.3008624017238617,
+      "learning_rate": 0.00016198558847077663,
+      "loss": 0.5501,
+      "step": 1193
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 0.46624132990837097,
+      "learning_rate": 0.00016195356285028022,
+      "loss": 0.6934,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 0.3650088608264923,
+      "learning_rate": 0.00016192153722978384,
+      "loss": 0.6114,
+      "step": 1195
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 0.4070186913013458,
+      "learning_rate": 0.00016188951160928745,
+      "loss": 0.5624,
+      "step": 1196
+    },
+    {
+      "epoch": 0.38304,
+      "grad_norm": 0.3992120325565338,
+      "learning_rate": 0.00016185748598879105,
+      "loss": 0.6582,
+      "step": 1197
+    },
+    {
+      "epoch": 0.38336,
+      "grad_norm": 0.438103586435318,
+      "learning_rate": 0.00016182546036829464,
+      "loss": 0.6165,
+      "step": 1198
+    },
+    {
+      "epoch": 0.38368,
+      "grad_norm": 0.5251441597938538,
+      "learning_rate": 0.00016179343474779825,
+      "loss": 0.7241,
+      "step": 1199
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.4377717673778534,
+      "learning_rate": 0.00016176140912730185,
+      "loss": 0.5749,
+      "step": 1200
+    },
+    {
+      "epoch": 0.38432,
+      "grad_norm": 0.49814409017562866,
+      "learning_rate": 0.00016172938350680544,
+      "loss": 0.743,
+      "step": 1201
+    },
+    {
+      "epoch": 0.38464,
+      "grad_norm": 0.39874687790870667,
+      "learning_rate": 0.00016169735788630906,
+      "loss": 0.6346,
+      "step": 1202
+    },
+    {
+      "epoch": 0.38496,
+      "grad_norm": 0.3396860957145691,
+      "learning_rate": 0.00016166533226581267,
+      "loss": 0.5547,
+      "step": 1203
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 0.32346323132514954,
+      "learning_rate": 0.00016163330664531627,
+      "loss": 0.4511,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 0.347773939371109,
+      "learning_rate": 0.00016160128102481986,
+      "loss": 0.6713,
+      "step": 1205
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 0.41030585765838623,
+      "learning_rate": 0.00016156925540432345,
+      "loss": 0.5934,
+      "step": 1206
+    },
+    {
+      "epoch": 0.38624,
+      "grad_norm": 0.4466695785522461,
+      "learning_rate": 0.00016153722978382707,
+      "loss": 0.642,
+      "step": 1207
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 0.4640570878982544,
+      "learning_rate": 0.00016150520416333068,
+      "loss": 0.6791,
+      "step": 1208
+    },
+    {
+      "epoch": 0.38688,
+      "grad_norm": 0.5479389429092407,
+      "learning_rate": 0.00016147317854283428,
+      "loss": 0.7217,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.2957899570465088,
+      "learning_rate": 0.00016144115292233787,
+      "loss": 0.5237,
+      "step": 1210
+    },
+    {
+      "epoch": 0.38752,
+      "grad_norm": 0.3981134295463562,
+      "learning_rate": 0.00016140912730184148,
+      "loss": 0.574,
+      "step": 1211
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 0.3416503369808197,
+      "learning_rate": 0.00016137710168134508,
+      "loss": 0.6367,
+      "step": 1212
+    },
+    {
+      "epoch": 0.38816,
+      "grad_norm": 0.42653048038482666,
+      "learning_rate": 0.00016134507606084867,
+      "loss": 0.7264,
+      "step": 1213
+    },
+    {
+      "epoch": 0.38848,
+      "grad_norm": 0.4234849512577057,
+      "learning_rate": 0.00016131305044035229,
+      "loss": 0.6071,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 0.4562069773674011,
+      "learning_rate": 0.0001612810248198559,
+      "loss": 0.6981,
+      "step": 1215
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 0.34615105390548706,
+      "learning_rate": 0.0001612489991993595,
+      "loss": 0.5742,
+      "step": 1216
+    },
+    {
+      "epoch": 0.38944,
+      "grad_norm": 0.4311599135398865,
+      "learning_rate": 0.00016121697357886309,
+      "loss": 0.7284,
+      "step": 1217
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 0.3812296688556671,
+      "learning_rate": 0.0001611849479583667,
+      "loss": 0.7295,
+      "step": 1218
+    },
+    {
+      "epoch": 0.39008,
+      "grad_norm": 0.34143874049186707,
+      "learning_rate": 0.0001611529223378703,
+      "loss": 0.5743,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.4767226278781891,
+      "learning_rate": 0.00016112089671737391,
+      "loss": 0.8313,
+      "step": 1220
+    },
+    {
+      "epoch": 0.39072,
+      "grad_norm": 0.40790101885795593,
+      "learning_rate": 0.0001610888710968775,
+      "loss": 0.5693,
+      "step": 1221
+    },
+    {
+      "epoch": 0.39104,
+      "grad_norm": 0.4305535554885864,
+      "learning_rate": 0.00016105684547638112,
+      "loss": 0.7687,
+      "step": 1222
+    },
+    {
+      "epoch": 0.39136,
+      "grad_norm": 0.42215779423713684,
+      "learning_rate": 0.00016102481985588471,
+      "loss": 0.5336,
+      "step": 1223
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 0.41275671124458313,
+      "learning_rate": 0.0001609927942353883,
+      "loss": 0.5618,
+      "step": 1224
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.3569745123386383,
+      "learning_rate": 0.0001609607686148919,
+      "loss": 0.5645,
+      "step": 1225
+    },
+    {
+      "epoch": 0.39232,
+      "grad_norm": 0.4409923553466797,
+      "learning_rate": 0.00016092874299439554,
+      "loss": 0.6594,
+      "step": 1226
+    },
+    {
+      "epoch": 0.39264,
+      "grad_norm": 0.4709078073501587,
+      "learning_rate": 0.00016089671737389913,
+      "loss": 0.7138,
+      "step": 1227
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 0.3747748136520386,
+      "learning_rate": 0.00016086469175340273,
+      "loss": 0.5851,
+      "step": 1228
+    },
+    {
+      "epoch": 0.39328,
+      "grad_norm": 0.3855699896812439,
+      "learning_rate": 0.00016083266613290632,
+      "loss": 0.6557,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.548929750919342,
+      "learning_rate": 0.00016080064051240993,
+      "loss": 0.7376,
+      "step": 1230
+    },
+    {
+      "epoch": 0.39392,
+      "grad_norm": 0.3478051722049713,
+      "learning_rate": 0.00016076861489191355,
+      "loss": 0.5314,
+      "step": 1231
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 0.40939369797706604,
+      "learning_rate": 0.00016073658927141714,
+      "loss": 0.5547,
+      "step": 1232
+    },
+    {
+      "epoch": 0.39456,
+      "grad_norm": 0.445141464471817,
+      "learning_rate": 0.00016070456365092074,
+      "loss": 0.76,
+      "step": 1233
+    },
+    {
+      "epoch": 0.39488,
+      "grad_norm": 0.4018782675266266,
+      "learning_rate": 0.00016067253803042435,
+      "loss": 0.7059,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 0.38671424984931946,
+      "learning_rate": 0.00016064051240992794,
+      "loss": 0.6048,
+      "step": 1235
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 0.4770282804965973,
+      "learning_rate": 0.00016060848678943154,
+      "loss": 0.6729,
+      "step": 1236
+    },
+    {
+      "epoch": 0.39584,
+      "grad_norm": 0.4241800904273987,
+      "learning_rate": 0.00016057646116893515,
+      "loss": 0.697,
+      "step": 1237
+    },
+    {
+      "epoch": 0.39616,
+      "grad_norm": 0.4291219711303711,
+      "learning_rate": 0.00016054443554843877,
+      "loss": 0.6276,
+      "step": 1238
+    },
+    {
+      "epoch": 0.39648,
+      "grad_norm": 0.5139747262001038,
+      "learning_rate": 0.00016051240992794236,
+      "loss": 0.7384,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.3738960027694702,
+      "learning_rate": 0.00016048038430744596,
+      "loss": 0.5885,
+      "step": 1240
+    },
+    {
+      "epoch": 0.39712,
+      "grad_norm": 0.3790477514266968,
+      "learning_rate": 0.00016044835868694957,
+      "loss": 0.6837,
+      "step": 1241
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 0.4405800700187683,
+      "learning_rate": 0.00016041633306645316,
+      "loss": 0.5104,
+      "step": 1242
+    },
+    {
+      "epoch": 0.39776,
+      "grad_norm": 0.42439335584640503,
+      "learning_rate": 0.00016038430744595678,
+      "loss": 0.6338,
+      "step": 1243
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 0.4335632026195526,
+      "learning_rate": 0.00016035228182546037,
+      "loss": 0.662,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 0.5328060984611511,
+      "learning_rate": 0.000160320256204964,
+      "loss": 0.7832,
+      "step": 1245
+    },
+    {
+      "epoch": 0.39872,
+      "grad_norm": 0.3707294464111328,
+      "learning_rate": 0.00016028823058446758,
+      "loss": 0.5526,
+      "step": 1246
+    },
+    {
+      "epoch": 0.39904,
+      "grad_norm": 0.4894300699234009,
+      "learning_rate": 0.00016025620496397117,
+      "loss": 0.6917,
+      "step": 1247
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 0.4552184045314789,
+      "learning_rate": 0.00016022417934347477,
+      "loss": 0.6904,
+      "step": 1248
+    },
+    {
+      "epoch": 0.39968,
+      "grad_norm": 0.40820959210395813,
+      "learning_rate": 0.0001601921537229784,
+      "loss": 0.5407,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3967379331588745,
+      "learning_rate": 0.000160160128102482,
+      "loss": 0.5851,
+      "step": 1250
+    },
+    {
+      "epoch": 0.40032,
+      "grad_norm": 0.43336209654808044,
+      "learning_rate": 0.0001601281024819856,
+      "loss": 0.7027,
+      "step": 1251
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 0.3911953568458557,
+      "learning_rate": 0.00016009607686148919,
+      "loss": 0.6827,
+      "step": 1252
+    },
+    {
+      "epoch": 0.40096,
+      "grad_norm": 0.39412355422973633,
+      "learning_rate": 0.0001600640512409928,
+      "loss": 0.6306,
+      "step": 1253
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 0.41168496012687683,
+      "learning_rate": 0.0001600320256204964,
+      "loss": 0.6173,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 0.35802409052848816,
+      "learning_rate": 0.00016,
+      "loss": 0.4607,
+      "step": 1255
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 0.37500685453414917,
+      "learning_rate": 0.0001599679743795036,
+      "loss": 0.5841,
+      "step": 1256
+    },
+    {
+      "epoch": 0.40224,
+      "grad_norm": 0.3903887867927551,
+      "learning_rate": 0.00015993594875900722,
+      "loss": 0.7388,
+      "step": 1257
+    },
+    {
+      "epoch": 0.40256,
+      "grad_norm": 0.45429715514183044,
+      "learning_rate": 0.00015990392313851081,
+      "loss": 0.647,
+      "step": 1258
+    },
+    {
+      "epoch": 0.40288,
+      "grad_norm": 0.42427289485931396,
+      "learning_rate": 0.0001598718975180144,
+      "loss": 0.6001,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.4580190181732178,
+      "learning_rate": 0.00015983987189751802,
+      "loss": 0.6634,
+      "step": 1260
+    },
+    {
+      "epoch": 0.40352,
+      "grad_norm": 0.4923192262649536,
+      "learning_rate": 0.00015980784627702164,
+      "loss": 0.733,
+      "step": 1261
+    },
+    {
+      "epoch": 0.40384,
+      "grad_norm": 0.3602467477321625,
+      "learning_rate": 0.00015977582065652523,
+      "loss": 0.5205,
+      "step": 1262
+    },
+    {
+      "epoch": 0.40416,
+      "grad_norm": 0.4060787856578827,
+      "learning_rate": 0.00015974379503602882,
+      "loss": 0.6443,
+      "step": 1263
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 0.40143758058547974,
+      "learning_rate": 0.00015971176941553244,
+      "loss": 0.6016,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 0.4037429094314575,
+      "learning_rate": 0.00015967974379503603,
+      "loss": 0.5515,
+      "step": 1265
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 0.4716244637966156,
+      "learning_rate": 0.00015964771817453962,
+      "loss": 0.5863,
+      "step": 1266
+    },
+    {
+      "epoch": 0.40544,
+      "grad_norm": 0.4699464440345764,
+      "learning_rate": 0.00015961569255404324,
+      "loss": 0.767,
+      "step": 1267
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 0.35264500975608826,
+      "learning_rate": 0.00015958366693354686,
+      "loss": 0.6293,
+      "step": 1268
+    },
+    {
+      "epoch": 0.40608,
+      "grad_norm": 0.4608737826347351,
+      "learning_rate": 0.00015955164131305045,
+      "loss": 0.6706,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.3958601653575897,
+      "learning_rate": 0.00015951961569255404,
+      "loss": 0.7082,
+      "step": 1270
+    },
+    {
+      "epoch": 0.40672,
+      "grad_norm": 0.4548644721508026,
+      "learning_rate": 0.00015948759007205764,
+      "loss": 0.6599,
+      "step": 1271
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 0.3886425793170929,
+      "learning_rate": 0.00015945556445156125,
+      "loss": 0.5317,
+      "step": 1272
+    },
+    {
+      "epoch": 0.40736,
+      "grad_norm": 0.42478564381599426,
+      "learning_rate": 0.00015942353883106487,
+      "loss": 0.6601,
+      "step": 1273
+    },
+    {
+      "epoch": 0.40768,
+      "grad_norm": 0.39766359329223633,
+      "learning_rate": 0.00015939151321056846,
+      "loss": 0.5728,
+      "step": 1274
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.392750084400177,
+      "learning_rate": 0.00015935948759007205,
+      "loss": 0.5545,
+      "step": 1275
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 0.4649530053138733,
+      "learning_rate": 0.00015932746196957567,
+      "loss": 0.6952,
+      "step": 1276
+    },
+    {
+      "epoch": 0.40864,
+      "grad_norm": 0.35825079679489136,
+      "learning_rate": 0.00015929543634907926,
+      "loss": 0.5769,
+      "step": 1277
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 0.4876900017261505,
+      "learning_rate": 0.00015926341072858288,
+      "loss": 0.4328,
+      "step": 1278
+    },
+    {
+      "epoch": 0.40928,
+      "grad_norm": 0.40463319420814514,
+      "learning_rate": 0.00015923138510808647,
+      "loss": 0.6702,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.46184614300727844,
+      "learning_rate": 0.0001591993594875901,
+      "loss": 0.6876,
+      "step": 1280
+    },
+    {
+      "epoch": 0.40992,
+      "grad_norm": 0.4325110912322998,
+      "learning_rate": 0.00015916733386709368,
+      "loss": 0.628,
+      "step": 1281
+    },
+    {
+      "epoch": 0.41024,
+      "grad_norm": 0.40107905864715576,
+      "learning_rate": 0.00015913530824659727,
+      "loss": 0.5812,
+      "step": 1282
+    },
+    {
+      "epoch": 0.41056,
+      "grad_norm": 0.3638785183429718,
+      "learning_rate": 0.0001591032826261009,
+      "loss": 0.6329,
+      "step": 1283
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 0.40781810879707336,
+      "learning_rate": 0.0001590712570056045,
+      "loss": 0.7613,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 0.43849310278892517,
+      "learning_rate": 0.0001590392313851081,
+      "loss": 0.6162,
+      "step": 1285
+    },
+    {
+      "epoch": 0.41152,
+      "grad_norm": 0.4211094379425049,
+      "learning_rate": 0.0001590072057646117,
+      "loss": 0.7107,
+      "step": 1286
+    },
+    {
+      "epoch": 0.41184,
+      "grad_norm": 0.40857550501823425,
+      "learning_rate": 0.0001589751801441153,
+      "loss": 0.5833,
+      "step": 1287
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 0.4667288362979889,
+      "learning_rate": 0.0001589431545236189,
+      "loss": 0.6905,
+      "step": 1288
+    },
+    {
+      "epoch": 0.41248,
+      "grad_norm": 0.39145544171333313,
+      "learning_rate": 0.0001589111289031225,
+      "loss": 0.5888,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.36629608273506165,
+      "learning_rate": 0.0001588791032826261,
+      "loss": 0.5352,
+      "step": 1290
+    },
+    {
+      "epoch": 0.41312,
+      "grad_norm": 0.40973952412605286,
+      "learning_rate": 0.00015884707766212973,
+      "loss": 0.7348,
+      "step": 1291
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 0.42966994643211365,
+      "learning_rate": 0.00015881505204163332,
+      "loss": 0.7332,
+      "step": 1292
+    },
+    {
+      "epoch": 0.41376,
+      "grad_norm": 0.4152204096317291,
+      "learning_rate": 0.0001587830264211369,
+      "loss": 0.7017,
+      "step": 1293
+    },
+    {
+      "epoch": 0.41408,
+      "grad_norm": 0.43853870034217834,
+      "learning_rate": 0.0001587510008006405,
+      "loss": 0.5749,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 0.388705313205719,
+      "learning_rate": 0.00015871897518014412,
+      "loss": 0.5462,
+      "step": 1295
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 0.39941397309303284,
+      "learning_rate": 0.00015868694955964774,
+      "loss": 0.5873,
+      "step": 1296
+    },
+    {
+      "epoch": 0.41504,
+      "grad_norm": 0.465790331363678,
+      "learning_rate": 0.00015865492393915133,
+      "loss": 0.7713,
+      "step": 1297
+    },
+    {
+      "epoch": 0.41536,
+      "grad_norm": 0.40599068999290466,
+      "learning_rate": 0.00015862289831865492,
+      "loss": 0.636,
+      "step": 1298
+    },
+    {
+      "epoch": 0.41568,
+      "grad_norm": 0.49515753984451294,
+      "learning_rate": 0.00015859087269815854,
+      "loss": 0.667,
+      "step": 1299
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.48135969042778015,
+      "learning_rate": 0.00015855884707766213,
+      "loss": 0.7522,
+      "step": 1300
+    },
+    {
+      "epoch": 0.41632,
+      "grad_norm": 0.43914127349853516,
+      "learning_rate": 0.00015852682145716572,
+      "loss": 0.522,
+      "step": 1301
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 0.3389265537261963,
+      "learning_rate": 0.00015849479583666934,
+      "loss": 0.5935,
+      "step": 1302
+    },
+    {
+      "epoch": 0.41696,
+      "grad_norm": 0.40796348452568054,
+      "learning_rate": 0.00015846277021617296,
+      "loss": 0.633,
+      "step": 1303
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 0.49675706028938293,
+      "learning_rate": 0.00015843074459567655,
+      "loss": 0.6909,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 0.4124357998371124,
+      "learning_rate": 0.00015839871897518014,
+      "loss": 0.61,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41792,
+      "grad_norm": 0.361156165599823,
+      "learning_rate": 0.00015836669335468376,
+      "loss": 0.5975,
+      "step": 1306
+    },
+    {
+      "epoch": 0.41824,
+      "grad_norm": 0.43280550837516785,
+      "learning_rate": 0.00015833466773418735,
+      "loss": 0.6347,
+      "step": 1307
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 0.3791589140892029,
+      "learning_rate": 0.00015830264211369097,
+      "loss": 0.5766,
+      "step": 1308
+    },
+    {
+      "epoch": 0.41888,
+      "grad_norm": 0.323436439037323,
+      "learning_rate": 0.00015827061649319456,
+      "loss": 0.5047,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.4022815525531769,
+      "learning_rate": 0.00015823859087269818,
+      "loss": 0.6133,
+      "step": 1310
+    },
+    {
+      "epoch": 0.41952,
+      "grad_norm": 0.40150073170661926,
+      "learning_rate": 0.00015820656525220177,
+      "loss": 0.5288,
+      "step": 1311
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 0.3382367491722107,
+      "learning_rate": 0.00015817453963170536,
+      "loss": 0.5171,
+      "step": 1312
+    },
+    {
+      "epoch": 0.42016,
+      "grad_norm": 0.5171733498573303,
+      "learning_rate": 0.00015814251401120895,
+      "loss": 0.6921,
+      "step": 1313
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 0.4533555805683136,
+      "learning_rate": 0.00015811048839071257,
+      "loss": 0.7398,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 0.4044964909553528,
+      "learning_rate": 0.0001580784627702162,
+      "loss": 0.5763,
+      "step": 1315
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 0.43161672353744507,
+      "learning_rate": 0.00015804643714971978,
+      "loss": 0.7246,
+      "step": 1316
+    },
+    {
+      "epoch": 0.42144,
+      "grad_norm": 0.3735911548137665,
+      "learning_rate": 0.00015801441152922337,
+      "loss": 0.5544,
+      "step": 1317
+    },
+    {
+      "epoch": 0.42176,
+      "grad_norm": 0.6016085147857666,
+      "learning_rate": 0.000157982385908727,
+      "loss": 0.7086,
+      "step": 1318
+    },
+    {
+      "epoch": 0.42208,
+      "grad_norm": 0.4195336401462555,
+      "learning_rate": 0.00015795036028823058,
+      "loss": 0.6344,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.3783457577228546,
+      "learning_rate": 0.0001579183346677342,
+      "loss": 0.5197,
+      "step": 1320
+    },
+    {
+      "epoch": 0.42272,
+      "grad_norm": 0.449931800365448,
+      "learning_rate": 0.0001578863090472378,
+      "loss": 0.7087,
+      "step": 1321
+    },
+    {
+      "epoch": 0.42304,
+      "grad_norm": 0.4201870858669281,
+      "learning_rate": 0.0001578542834267414,
+      "loss": 0.6454,
+      "step": 1322
+    },
+    {
+      "epoch": 0.42336,
+      "grad_norm": 0.4194982349872589,
+      "learning_rate": 0.000157822257806245,
+      "loss": 0.7346,
+      "step": 1323
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 0.42280638217926025,
+      "learning_rate": 0.0001577902321857486,
+      "loss": 0.6562,
+      "step": 1324
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.44141513109207153,
+      "learning_rate": 0.0001577582065652522,
+      "loss": 0.8186,
+      "step": 1325
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 0.34653738141059875,
+      "learning_rate": 0.00015772618094475583,
+      "loss": 0.5379,
+      "step": 1326
+    },
+    {
+      "epoch": 0.42464,
+      "grad_norm": 0.3119458556175232,
+      "learning_rate": 0.00015769415532425942,
+      "loss": 0.4565,
+      "step": 1327
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 0.4796963930130005,
+      "learning_rate": 0.000157662129703763,
+      "loss": 0.9034,
+      "step": 1328
+    },
+    {
+      "epoch": 0.42528,
+      "grad_norm": 0.3462713360786438,
+      "learning_rate": 0.00015763010408326663,
+      "loss": 0.4225,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.4905683398246765,
+      "learning_rate": 0.00015759807846277022,
+      "loss": 0.764,
+      "step": 1330
+    },
+    {
+      "epoch": 0.42592,
+      "grad_norm": 0.4774971306324005,
+      "learning_rate": 0.00015756605284227384,
+      "loss": 0.6755,
+      "step": 1331
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 0.43234968185424805,
+      "learning_rate": 0.00015753402722177743,
+      "loss": 0.5833,
+      "step": 1332
+    },
+    {
+      "epoch": 0.42656,
+      "grad_norm": 0.42147600650787354,
+      "learning_rate": 0.00015750200160128102,
+      "loss": 0.6293,
+      "step": 1333
+    },
+    {
+      "epoch": 0.42688,
+      "grad_norm": 0.4219157099723816,
+      "learning_rate": 0.00015746997598078464,
+      "loss": 0.4237,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 0.5773321390151978,
+      "learning_rate": 0.00015743795036028823,
+      "loss": 0.9498,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 0.4606321454048157,
+      "learning_rate": 0.00015740592473979182,
+      "loss": 0.618,
+      "step": 1336
+    },
+    {
+      "epoch": 0.42784,
+      "grad_norm": 0.35409003496170044,
+      "learning_rate": 0.00015737389911929544,
+      "loss": 0.5765,
+      "step": 1337
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 0.40120014548301697,
+      "learning_rate": 0.00015734187349879906,
+      "loss": 0.6096,
+      "step": 1338
+    },
+    {
+      "epoch": 0.42848,
+      "grad_norm": 0.4659929871559143,
+      "learning_rate": 0.00015730984787830265,
+      "loss": 0.7349,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.5044195652008057,
+      "learning_rate": 0.00015727782225780624,
+      "loss": 0.7605,
+      "step": 1340
+    },
+    {
+      "epoch": 0.42912,
+      "grad_norm": 0.39576542377471924,
+      "learning_rate": 0.00015724579663730986,
+      "loss": 0.4895,
+      "step": 1341
+    },
+    {
+      "epoch": 0.42944,
+      "grad_norm": 0.4105573892593384,
+      "learning_rate": 0.00015721377101681345,
+      "loss": 0.6891,
+      "step": 1342
+    },
+    {
+      "epoch": 0.42976,
+      "grad_norm": 0.3827102482318878,
+      "learning_rate": 0.00015718174539631707,
+      "loss": 0.5647,
+      "step": 1343
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 0.43864673376083374,
+      "learning_rate": 0.00015714971977582066,
+      "loss": 0.7352,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 0.4819627106189728,
+      "learning_rate": 0.00015711769415532428,
+      "loss": 0.6848,
+      "step": 1345
+    },
+    {
+      "epoch": 0.43072,
+      "grad_norm": 0.35930827260017395,
+      "learning_rate": 0.00015708566853482787,
+      "loss": 0.4841,
+      "step": 1346
+    },
+    {
+      "epoch": 0.43104,
+      "grad_norm": 0.40655046701431274,
+      "learning_rate": 0.00015705364291433146,
+      "loss": 0.6293,
+      "step": 1347
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 0.5381290316581726,
+      "learning_rate": 0.00015702161729383505,
+      "loss": 0.8441,
+      "step": 1348
+    },
+    {
+      "epoch": 0.43168,
+      "grad_norm": 0.38459235429763794,
+      "learning_rate": 0.0001569895916733387,
+      "loss": 0.4958,
+      "step": 1349
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.36747026443481445,
+      "learning_rate": 0.0001569575660528423,
+      "loss": 0.5618,
+      "step": 1350
+    },
+    {
+      "epoch": 0.43232,
+      "grad_norm": 0.4343733787536621,
+      "learning_rate": 0.00015692554043234588,
+      "loss": 0.732,
+      "step": 1351
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 0.45906516909599304,
+      "learning_rate": 0.00015689351481184947,
+      "loss": 0.6177,
+      "step": 1352
+    },
+    {
+      "epoch": 0.43296,
+      "grad_norm": 0.30340859293937683,
+      "learning_rate": 0.0001568614891913531,
+      "loss": 0.4707,
+      "step": 1353
+    },
+    {
+      "epoch": 0.43328,
+      "grad_norm": 0.48463016748428345,
+      "learning_rate": 0.00015682946357085668,
+      "loss": 0.7262,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 0.41301092505455017,
+      "learning_rate": 0.0001567974379503603,
+      "loss": 0.7356,
+      "step": 1355
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 0.4270017743110657,
+      "learning_rate": 0.0001567654123298639,
+      "loss": 0.7483,
+      "step": 1356
+    },
+    {
+      "epoch": 0.43424,
+      "grad_norm": 0.3573155403137207,
+      "learning_rate": 0.0001567333867093675,
+      "loss": 0.5344,
+      "step": 1357
+    },
+    {
+      "epoch": 0.43456,
+      "grad_norm": 0.43801671266555786,
+      "learning_rate": 0.0001567013610888711,
+      "loss": 0.6712,
+      "step": 1358
+    },
+    {
+      "epoch": 0.43488,
+      "grad_norm": 0.36065322160720825,
+      "learning_rate": 0.0001566693354683747,
+      "loss": 0.5072,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.33040866255760193,
+      "learning_rate": 0.0001566373098478783,
+      "loss": 0.5549,
+      "step": 1360
+    },
+    {
+      "epoch": 0.43552,
+      "grad_norm": 0.396777868270874,
+      "learning_rate": 0.00015660528422738193,
+      "loss": 0.6527,
+      "step": 1361
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 0.35000482201576233,
+      "learning_rate": 0.00015657325860688552,
+      "loss": 0.5272,
+      "step": 1362
+    },
+    {
+      "epoch": 0.43616,
+      "grad_norm": 0.4051343500614166,
+      "learning_rate": 0.0001565412329863891,
+      "loss": 0.6364,
+      "step": 1363
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 0.39511218667030334,
+      "learning_rate": 0.00015650920736589273,
+      "loss": 0.5706,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 0.3674565553665161,
+      "learning_rate": 0.00015647718174539632,
+      "loss": 0.4998,
+      "step": 1365
+    },
+    {
+      "epoch": 0.43712,
+      "grad_norm": 0.4118686616420746,
+      "learning_rate": 0.0001564451561248999,
+      "loss": 0.5945,
+      "step": 1366
+    },
+    {
+      "epoch": 0.43744,
+      "grad_norm": 0.42774638533592224,
+      "learning_rate": 0.00015641313050440353,
+      "loss": 0.6907,
+      "step": 1367
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 0.3459903299808502,
+      "learning_rate": 0.00015638110488390715,
+      "loss": 0.6057,
+      "step": 1368
+    },
+    {
+      "epoch": 0.43808,
+      "grad_norm": 0.4655570983886719,
+      "learning_rate": 0.00015634907926341074,
+      "loss": 0.7305,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 0.4046120345592499,
+      "learning_rate": 0.00015631705364291433,
+      "loss": 0.6229,
+      "step": 1370
+    },
+    {
+      "epoch": 0.43872,
+      "grad_norm": 0.3614268898963928,
+      "learning_rate": 0.00015628502802241792,
+      "loss": 0.4523,
+      "step": 1371
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 0.42226001620292664,
+      "learning_rate": 0.00015625300240192157,
+      "loss": 0.5752,
+      "step": 1372
+    },
+    {
+      "epoch": 0.43936,
+      "grad_norm": 0.3451758027076721,
+      "learning_rate": 0.00015622097678142516,
+      "loss": 0.4817,
+      "step": 1373
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 0.43787944316864014,
+      "learning_rate": 0.00015618895116092875,
+      "loss": 0.6026,
+      "step": 1374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30242031812667847,
+      "learning_rate": 0.00015615692554043234,
+      "loss": 0.5134,
+      "step": 1375
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 0.45698973536491394,
+      "learning_rate": 0.00015612489991993596,
+      "loss": 0.6494,
+      "step": 1376
+    },
+    {
+      "epoch": 0.44064,
+      "grad_norm": 0.37014782428741455,
+      "learning_rate": 0.00015609287429943955,
+      "loss": 0.5498,
+      "step": 1377
+    },
+    {
+      "epoch": 0.44096,
+      "grad_norm": 0.4875912666320801,
+      "learning_rate": 0.00015606084867894317,
+      "loss": 0.7938,
+      "step": 1378
+    },
+    {
+      "epoch": 0.44128,
+      "grad_norm": 0.3790332078933716,
+      "learning_rate": 0.00015602882305844676,
+      "loss": 0.6297,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.355122834444046,
+      "learning_rate": 0.00015599679743795038,
+      "loss": 0.6771,
+      "step": 1380
+    },
+    {
+      "epoch": 0.44192,
+      "grad_norm": 0.3693047761917114,
+      "learning_rate": 0.00015596477181745397,
+      "loss": 0.5606,
+      "step": 1381
+    },
+    {
+      "epoch": 0.44224,
+      "grad_norm": 0.41751736402511597,
+      "learning_rate": 0.00015593274619695756,
+      "loss": 0.6199,
+      "step": 1382
+    },
+    {
+      "epoch": 0.44256,
+      "grad_norm": 0.46476659178733826,
+      "learning_rate": 0.00015590072057646118,
+      "loss": 0.7091,
+      "step": 1383
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 0.42213600873947144,
+      "learning_rate": 0.0001558686949559648,
+      "loss": 0.5758,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 0.4991520643234253,
+      "learning_rate": 0.0001558366693354684,
+      "loss": 0.8049,
+      "step": 1385
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 0.3500632643699646,
+      "learning_rate": 0.00015580464371497198,
+      "loss": 0.5151,
+      "step": 1386
+    },
+    {
+      "epoch": 0.44384,
+      "grad_norm": 0.4511035084724426,
+      "learning_rate": 0.0001557726180944756,
+      "loss": 0.6459,
+      "step": 1387
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 0.5112922191619873,
+      "learning_rate": 0.0001557405924739792,
+      "loss": 0.6818,
+      "step": 1388
+    },
+    {
+      "epoch": 0.44448,
+      "grad_norm": 0.4460408687591553,
+      "learning_rate": 0.00015570856685348278,
+      "loss": 0.6826,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.37085771560668945,
+      "learning_rate": 0.0001556765412329864,
+      "loss": 0.6301,
+      "step": 1390
+    },
+    {
+      "epoch": 0.44512,
+      "grad_norm": 0.4643785357475281,
+      "learning_rate": 0.00015564451561249002,
+      "loss": 0.7577,
+      "step": 1391
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 0.43264371156692505,
+      "learning_rate": 0.0001556124899919936,
+      "loss": 0.6252,
+      "step": 1392
+    },
+    {
+      "epoch": 0.44576,
+      "grad_norm": 0.3916131258010864,
+      "learning_rate": 0.0001555804643714972,
+      "loss": 0.5343,
+      "step": 1393
+    },
+    {
+      "epoch": 0.44608,
+      "grad_norm": 0.3903241455554962,
+      "learning_rate": 0.0001555484387510008,
+      "loss": 0.5378,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 0.42164376378059387,
+      "learning_rate": 0.0001555164131305044,
+      "loss": 0.5515,
+      "step": 1395
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 0.42706766724586487,
+      "learning_rate": 0.00015548438751000803,
+      "loss": 0.65,
+      "step": 1396
+    },
+    {
+      "epoch": 0.44704,
+      "grad_norm": 0.42193999886512756,
+      "learning_rate": 0.00015545236188951162,
+      "loss": 0.6316,
+      "step": 1397
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 0.3912172019481659,
+      "learning_rate": 0.0001554203362690152,
+      "loss": 0.5929,
+      "step": 1398
+    },
+    {
+      "epoch": 0.44768,
+      "grad_norm": 0.4234367907047272,
+      "learning_rate": 0.00015538831064851883,
+      "loss": 0.6742,
+      "step": 1399
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.4279198944568634,
+      "learning_rate": 0.00015535628502802242,
+      "loss": 0.6069,
+      "step": 1400
+    },
+    {
+      "epoch": 0.44832,
+      "grad_norm": 0.3786641061306,
+      "learning_rate": 0.000155324259407526,
+      "loss": 0.7063,
+      "step": 1401
+    },
+    {
+      "epoch": 0.44864,
+      "grad_norm": 0.4302279055118561,
+      "learning_rate": 0.00015529223378702963,
+      "loss": 0.5841,
+      "step": 1402
+    },
+    {
+      "epoch": 0.44896,
+      "grad_norm": 0.41082334518432617,
+      "learning_rate": 0.00015526020816653325,
+      "loss": 0.5532,
+      "step": 1403
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 0.3306528329849243,
+      "learning_rate": 0.00015522818254603684,
+      "loss": 0.4712,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 0.4020909070968628,
+      "learning_rate": 0.00015519615692554043,
+      "loss": 0.7071,
+      "step": 1405
+    },
+    {
+      "epoch": 0.44992,
+      "grad_norm": 0.3991028964519501,
+      "learning_rate": 0.00015516413130504405,
+      "loss": 0.6117,
+      "step": 1406
+    },
+    {
+      "epoch": 0.45024,
+      "grad_norm": 0.40511631965637207,
+      "learning_rate": 0.00015513210568454764,
+      "loss": 0.5537,
+      "step": 1407
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 0.38224756717681885,
+      "learning_rate": 0.00015510008006405126,
+      "loss": 0.5577,
+      "step": 1408
+    },
+    {
+      "epoch": 0.45088,
+      "grad_norm": 0.42746374011039734,
+      "learning_rate": 0.00015506805444355485,
+      "loss": 0.6513,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.40762078762054443,
+      "learning_rate": 0.00015503602882305847,
+      "loss": 0.6499,
+      "step": 1410
+    },
+    {
+      "epoch": 0.45152,
+      "grad_norm": 0.37872472405433655,
+      "learning_rate": 0.00015500400320256206,
+      "loss": 0.5962,
+      "step": 1411
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 0.44815394282341003,
+      "learning_rate": 0.00015497197758206565,
+      "loss": 0.6318,
+      "step": 1412
+    },
+    {
+      "epoch": 0.45216,
+      "grad_norm": 0.3514578640460968,
+      "learning_rate": 0.00015493995196156924,
+      "loss": 0.5951,
+      "step": 1413
+    },
+    {
+      "epoch": 0.45248,
+      "grad_norm": 0.490763783454895,
+      "learning_rate": 0.00015490792634107289,
+      "loss": 0.6813,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 0.3479330837726593,
+      "learning_rate": 0.00015487590072057648,
+      "loss": 0.5769,
+      "step": 1415
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 0.5040646195411682,
+      "learning_rate": 0.00015484387510008007,
+      "loss": 0.7946,
+      "step": 1416
+    },
+    {
+      "epoch": 0.45344,
+      "grad_norm": 0.4628739356994629,
+      "learning_rate": 0.00015481184947958366,
+      "loss": 0.7216,
+      "step": 1417
+    },
+    {
+      "epoch": 0.45376,
+      "grad_norm": 0.4331735670566559,
+      "learning_rate": 0.00015477982385908728,
+      "loss": 0.7846,
+      "step": 1418
+    },
+    {
+      "epoch": 0.45408,
+      "grad_norm": 0.46598461270332336,
+      "learning_rate": 0.0001547477982385909,
+      "loss": 0.5492,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.3700425922870636,
+      "learning_rate": 0.0001547157726180945,
+      "loss": 0.4561,
+      "step": 1420
+    },
+    {
+      "epoch": 0.45472,
+      "grad_norm": 0.41523638367652893,
+      "learning_rate": 0.00015468374699759808,
+      "loss": 0.5707,
+      "step": 1421
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 0.412091463804245,
+      "learning_rate": 0.0001546517213771017,
+      "loss": 0.711,
+      "step": 1422
+    },
+    {
+      "epoch": 0.45536,
+      "grad_norm": 0.51674485206604,
+      "learning_rate": 0.0001546196957566053,
+      "loss": 0.8247,
+      "step": 1423
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 0.4690164029598236,
+      "learning_rate": 0.00015458767013610888,
+      "loss": 0.6927,
+      "step": 1424
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 0.3963469862937927,
+      "learning_rate": 0.0001545556445156125,
+      "loss": 0.6512,
+      "step": 1425
+    },
+    {
+      "epoch": 0.45632,
+      "grad_norm": 0.3646407723426819,
+      "learning_rate": 0.00015452361889511612,
+      "loss": 0.5831,
+      "step": 1426
+    },
+    {
+      "epoch": 0.45664,
+      "grad_norm": 0.43929800391197205,
+      "learning_rate": 0.0001544915932746197,
+      "loss": 0.6782,
+      "step": 1427
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 0.5239555239677429,
+      "learning_rate": 0.0001544595676541233,
+      "loss": 0.9483,
+      "step": 1428
+    },
+    {
+      "epoch": 0.45728,
+      "grad_norm": 0.4606457054615021,
+      "learning_rate": 0.00015442754203362692,
+      "loss": 0.6236,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 0.3686271905899048,
+      "learning_rate": 0.0001543955164131305,
+      "loss": 0.5833,
+      "step": 1430
+    },
+    {
+      "epoch": 0.45792,
+      "grad_norm": 0.41566357016563416,
+      "learning_rate": 0.00015436349079263413,
+      "loss": 0.5531,
+      "step": 1431
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 0.44952067732810974,
+      "learning_rate": 0.00015433146517213772,
+      "loss": 0.6776,
+      "step": 1432
+    },
+    {
+      "epoch": 0.45856,
+      "grad_norm": 0.3611706495285034,
+      "learning_rate": 0.00015429943955164134,
+      "loss": 0.5573,
+      "step": 1433
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 0.35928264260292053,
+      "learning_rate": 0.00015426741393114493,
+      "loss": 0.5796,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 0.41525959968566895,
+      "learning_rate": 0.00015423538831064852,
+      "loss": 0.6282,
+      "step": 1435
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 0.3865865170955658,
+      "learning_rate": 0.0001542033626901521,
+      "loss": 0.5165,
+      "step": 1436
+    },
+    {
+      "epoch": 0.45984,
+      "grad_norm": 0.42326653003692627,
+      "learning_rate": 0.00015417133706965575,
+      "loss": 0.5968,
+      "step": 1437
+    },
+    {
+      "epoch": 0.46016,
+      "grad_norm": 0.4799272418022156,
+      "learning_rate": 0.00015413931144915935,
+      "loss": 0.5763,
+      "step": 1438
+    },
+    {
+      "epoch": 0.46048,
+      "grad_norm": 0.39223918318748474,
+      "learning_rate": 0.00015410728582866294,
+      "loss": 0.5718,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.5210763216018677,
+      "learning_rate": 0.00015407526020816653,
+      "loss": 0.6807,
+      "step": 1440
+    },
+    {
+      "epoch": 0.46112,
+      "grad_norm": 0.43459659814834595,
+      "learning_rate": 0.00015404323458767015,
+      "loss": 0.551,
+      "step": 1441
+    },
+    {
+      "epoch": 0.46144,
+      "grad_norm": 0.5593394637107849,
+      "learning_rate": 0.00015401120896717374,
+      "loss": 0.6837,
+      "step": 1442
+    },
+    {
+      "epoch": 0.46176,
+      "grad_norm": 0.41199105978012085,
+      "learning_rate": 0.00015397918334667736,
+      "loss": 0.588,
+      "step": 1443
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 0.48737412691116333,
+      "learning_rate": 0.00015394715772618095,
+      "loss": 0.6025,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 0.4127195179462433,
+      "learning_rate": 0.00015391513210568457,
+      "loss": 0.5053,
+      "step": 1445
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 0.3279756009578705,
+      "learning_rate": 0.00015388310648518816,
+      "loss": 0.4306,
+      "step": 1446
+    },
+    {
+      "epoch": 0.46304,
+      "grad_norm": 0.4130052626132965,
+      "learning_rate": 0.00015385108086469175,
+      "loss": 0.669,
+      "step": 1447
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 0.4151359498500824,
+      "learning_rate": 0.00015381905524419537,
+      "loss": 0.5116,
+      "step": 1448
+    },
+    {
+      "epoch": 0.46368,
+      "grad_norm": 0.4836346209049225,
+      "learning_rate": 0.00015378702962369898,
+      "loss": 0.628,
+      "step": 1449
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.4735133945941925,
+      "learning_rate": 0.00015375500400320258,
+      "loss": 0.6431,
+      "step": 1450
+    },
+    {
+      "epoch": 0.46432,
+      "grad_norm": 0.4574035704135895,
+      "learning_rate": 0.00015372297838270617,
+      "loss": 0.5858,
+      "step": 1451
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 0.39928683638572693,
+      "learning_rate": 0.00015369095276220979,
+      "loss": 0.5202,
+      "step": 1452
+    },
+    {
+      "epoch": 0.46496,
+      "grad_norm": 0.3989488184452057,
+      "learning_rate": 0.00015365892714171338,
+      "loss": 0.5689,
+      "step": 1453
+    },
+    {
+      "epoch": 0.46528,
+      "grad_norm": 0.4315936267375946,
+      "learning_rate": 0.00015362690152121697,
+      "loss": 0.5574,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 0.42689022421836853,
+      "learning_rate": 0.00015359487590072059,
+      "loss": 0.702,
+      "step": 1455
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 0.45261624455451965,
+      "learning_rate": 0.0001535628502802242,
+      "loss": 0.7021,
+      "step": 1456
+    },
+    {
+      "epoch": 0.46624,
+      "grad_norm": 0.43238353729248047,
+      "learning_rate": 0.0001535308246597278,
+      "loss": 0.663,
+      "step": 1457
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 0.4568151831626892,
+      "learning_rate": 0.0001534987990392314,
+      "loss": 0.6038,
+      "step": 1458
+    },
+    {
+      "epoch": 0.46688,
+      "grad_norm": 0.4050443172454834,
+      "learning_rate": 0.00015346677341873498,
+      "loss": 0.6773,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.5023525357246399,
+      "learning_rate": 0.0001534347477982386,
+      "loss": 0.833,
+      "step": 1460
+    },
+    {
+      "epoch": 0.46752,
+      "grad_norm": 0.41459766030311584,
+      "learning_rate": 0.00015340272217774221,
+      "loss": 0.5958,
+      "step": 1461
+    },
+    {
+      "epoch": 0.46784,
+      "grad_norm": 0.4184015691280365,
+      "learning_rate": 0.0001533706965572458,
+      "loss": 0.6003,
+      "step": 1462
+    },
+    {
+      "epoch": 0.46816,
+      "grad_norm": 0.3837192952632904,
+      "learning_rate": 0.0001533386709367494,
+      "loss": 0.6376,
+      "step": 1463
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 0.4296150505542755,
+      "learning_rate": 0.00015330664531625302,
+      "loss": 0.6223,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 0.4326069951057434,
+      "learning_rate": 0.0001532746196957566,
+      "loss": 0.634,
+      "step": 1465
+    },
+    {
+      "epoch": 0.46912,
+      "grad_norm": 0.48113271594047546,
+      "learning_rate": 0.00015324259407526022,
+      "loss": 0.6665,
+      "step": 1466
+    },
+    {
+      "epoch": 0.46944,
+      "grad_norm": 0.3846622407436371,
+      "learning_rate": 0.00015321056845476382,
+      "loss": 0.6385,
+      "step": 1467
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 0.3542896807193756,
+      "learning_rate": 0.00015317854283426743,
+      "loss": 0.6149,
+      "step": 1468
+    },
+    {
+      "epoch": 0.47008,
+      "grad_norm": 0.4088331162929535,
+      "learning_rate": 0.00015314651721377103,
+      "loss": 0.5569,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.5250117778778076,
+      "learning_rate": 0.00015311449159327462,
+      "loss": 0.5476,
+      "step": 1470
+    },
+    {
+      "epoch": 0.47072,
+      "grad_norm": 0.4015345871448517,
+      "learning_rate": 0.00015308246597277824,
+      "loss": 0.6876,
+      "step": 1471
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 0.4432811141014099,
+      "learning_rate": 0.00015305044035228185,
+      "loss": 0.7126,
+      "step": 1472
+    },
+    {
+      "epoch": 0.47136,
+      "grad_norm": 0.47002461552619934,
+      "learning_rate": 0.00015301841473178544,
+      "loss": 0.7676,
+      "step": 1473
+    },
+    {
+      "epoch": 0.47168,
+      "grad_norm": 0.39825284481048584,
+      "learning_rate": 0.00015298638911128904,
+      "loss": 0.5329,
+      "step": 1474
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.4873245060443878,
+      "learning_rate": 0.00015295436349079263,
+      "loss": 0.6291,
+      "step": 1475
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 0.3872564733028412,
+      "learning_rate": 0.00015292233787029625,
+      "loss": 0.5318,
+      "step": 1476
+    },
+    {
+      "epoch": 0.47264,
+      "grad_norm": 0.4167598783969879,
+      "learning_rate": 0.00015289031224979984,
+      "loss": 0.5351,
+      "step": 1477
+    },
+    {
+      "epoch": 0.47296,
+      "grad_norm": 0.3712940812110901,
+      "learning_rate": 0.00015285828662930345,
+      "loss": 0.496,
+      "step": 1478
+    },
+    {
+      "epoch": 0.47328,
+      "grad_norm": 0.43952324986457825,
+      "learning_rate": 0.00015282626100880705,
+      "loss": 0.6846,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.41462525725364685,
+      "learning_rate": 0.00015279423538831066,
+      "loss": 0.6298,
+      "step": 1480
+    },
+    {
+      "epoch": 0.47392,
+      "grad_norm": 0.43451422452926636,
+      "learning_rate": 0.00015276220976781426,
+      "loss": 0.5606,
+      "step": 1481
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 0.560116171836853,
+      "learning_rate": 0.00015273018414731785,
+      "loss": 0.7843,
+      "step": 1482
+    },
+    {
+      "epoch": 0.47456,
+      "grad_norm": 0.35891005396842957,
+      "learning_rate": 0.00015269815852682147,
+      "loss": 0.4871,
+      "step": 1483
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 0.44833633303642273,
+      "learning_rate": 0.00015266613290632508,
+      "loss": 0.5655,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 0.45219993591308594,
+      "learning_rate": 0.00015263410728582867,
+      "loss": 0.5349,
+      "step": 1485
+    },
+    {
+      "epoch": 0.47552,
+      "grad_norm": 0.4389609396457672,
+      "learning_rate": 0.00015260208166533227,
+      "loss": 0.4827,
+      "step": 1486
+    },
+    {
+      "epoch": 0.47584,
+      "grad_norm": 0.46659988164901733,
+      "learning_rate": 0.00015257005604483588,
+      "loss": 0.4503,
+      "step": 1487
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 0.42142292857170105,
+      "learning_rate": 0.00015253803042433948,
+      "loss": 0.6166,
+      "step": 1488
+    },
+    {
+      "epoch": 0.47648,
+      "grad_norm": 0.43780800700187683,
+      "learning_rate": 0.00015250600480384307,
+      "loss": 0.6049,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.5306322574615479,
+      "learning_rate": 0.00015247397918334668,
+      "loss": 0.7113,
+      "step": 1490
+    },
+    {
+      "epoch": 0.47712,
+      "grad_norm": 0.3918207585811615,
+      "learning_rate": 0.0001524419535628503,
+      "loss": 0.4784,
+      "step": 1491
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 0.41745465993881226,
+      "learning_rate": 0.0001524099279423539,
+      "loss": 0.5778,
+      "step": 1492
+    },
+    {
+      "epoch": 0.47776,
+      "grad_norm": 0.4237818419933319,
+      "learning_rate": 0.00015237790232185749,
+      "loss": 0.5659,
+      "step": 1493
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 0.4207609295845032,
+      "learning_rate": 0.00015234587670136108,
+      "loss": 0.6169,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 0.4056950509548187,
+      "learning_rate": 0.0001523138510808647,
+      "loss": 0.5904,
+      "step": 1495
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 0.4585679769515991,
+      "learning_rate": 0.00015228182546036831,
+      "loss": 0.5572,
+      "step": 1496
+    },
+    {
+      "epoch": 0.47904,
+      "grad_norm": 0.3343838155269623,
+      "learning_rate": 0.0001522497998398719,
+      "loss": 0.5391,
+      "step": 1497
+    },
+    {
+      "epoch": 0.47936,
+      "grad_norm": 0.4628114700317383,
+      "learning_rate": 0.0001522177742193755,
+      "loss": 0.6812,
+      "step": 1498
+    },
+    {
+      "epoch": 0.47968,
+      "grad_norm": 0.4280478358268738,
+      "learning_rate": 0.00015218574859887911,
+      "loss": 0.6509,
+      "step": 1499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49500003457069397,
+      "learning_rate": 0.0001521537229783827,
+      "loss": 0.6431,
+      "step": 1500
+    },
+    {
+      "epoch": 0.48032,
+      "grad_norm": 0.36904376745224,
+      "learning_rate": 0.0001521216973578863,
+      "loss": 0.5925,
+      "step": 1501
+    },
+    {
+      "epoch": 0.48064,
+      "grad_norm": 0.441490113735199,
+      "learning_rate": 0.00015208967173738992,
+      "loss": 0.7162,
+      "step": 1502
+    },
+    {
+      "epoch": 0.48096,
+      "grad_norm": 0.3426421284675598,
+      "learning_rate": 0.00015205764611689353,
+      "loss": 0.5068,
+      "step": 1503
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 0.47684767842292786,
+      "learning_rate": 0.00015202562049639712,
+      "loss": 0.6702,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 0.41973310708999634,
+      "learning_rate": 0.00015199359487590072,
+      "loss": 0.642,
+      "step": 1505
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 0.37280401587486267,
+      "learning_rate": 0.00015196156925540433,
+      "loss": 0.6114,
+      "step": 1506
+    },
+    {
+      "epoch": 0.48224,
+      "grad_norm": 0.4655849039554596,
+      "learning_rate": 0.00015192954363490793,
+      "loss": 0.5936,
+      "step": 1507
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 0.457701176404953,
+      "learning_rate": 0.00015189751801441154,
+      "loss": 0.692,
+      "step": 1508
+    },
+    {
+      "epoch": 0.48288,
+      "grad_norm": 0.5481858253479004,
+      "learning_rate": 0.00015186549239391513,
+      "loss": 0.7339,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.38538238406181335,
+      "learning_rate": 0.00015183346677341875,
+      "loss": 0.4974,
+      "step": 1510
+    },
+    {
+      "epoch": 0.48352,
+      "grad_norm": 0.36177554726600647,
+      "learning_rate": 0.00015180144115292234,
+      "loss": 0.6041,
+      "step": 1511
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 0.38415148854255676,
+      "learning_rate": 0.00015176941553242594,
+      "loss": 0.6285,
+      "step": 1512
+    },
+    {
+      "epoch": 0.48416,
+      "grad_norm": 0.39309027791023254,
+      "learning_rate": 0.00015173738991192955,
+      "loss": 0.6609,
+      "step": 1513
+    },
+    {
+      "epoch": 0.48448,
+      "grad_norm": 0.48184698820114136,
+      "learning_rate": 0.00015170536429143317,
+      "loss": 0.5457,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 0.35899558663368225,
+      "learning_rate": 0.00015167333867093676,
+      "loss": 0.6293,
+      "step": 1515
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 0.43861427903175354,
+      "learning_rate": 0.00015164131305044035,
+      "loss": 0.57,
+      "step": 1516
+    },
+    {
+      "epoch": 0.48544,
+      "grad_norm": 0.45586541295051575,
+      "learning_rate": 0.00015160928742994395,
+      "loss": 0.7333,
+      "step": 1517
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 0.4278445243835449,
+      "learning_rate": 0.00015157726180944756,
+      "loss": 0.6956,
+      "step": 1518
+    },
+    {
+      "epoch": 0.48608,
+      "grad_norm": 0.4398341476917267,
+      "learning_rate": 0.00015154523618895118,
+      "loss": 0.595,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.4537087380886078,
+      "learning_rate": 0.00015151321056845477,
+      "loss": 0.6331,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48672,
+      "grad_norm": 0.452219694852829,
+      "learning_rate": 0.00015148118494795836,
+      "loss": 0.7274,
+      "step": 1521
+    },
+    {
+      "epoch": 0.48704,
+      "grad_norm": 0.399541437625885,
+      "learning_rate": 0.00015144915932746198,
+      "loss": 0.5489,
+      "step": 1522
+    },
+    {
+      "epoch": 0.48736,
+      "grad_norm": 0.4202711284160614,
+      "learning_rate": 0.00015141713370696557,
+      "loss": 0.6275,
+      "step": 1523
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 0.36473676562309265,
+      "learning_rate": 0.00015138510808646917,
+      "loss": 0.505,
+      "step": 1524
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.41317126154899597,
+      "learning_rate": 0.00015135308246597278,
+      "loss": 0.601,
+      "step": 1525
+    },
+    {
+      "epoch": 0.48832,
+      "grad_norm": 0.47797179222106934,
+      "learning_rate": 0.0001513210568454764,
+      "loss": 0.6869,
+      "step": 1526
+    },
+    {
+      "epoch": 0.48864,
+      "grad_norm": 0.4147303104400635,
+      "learning_rate": 0.00015128903122498,
+      "loss": 0.6082,
+      "step": 1527
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 0.5985320210456848,
+      "learning_rate": 0.00015125700560448358,
+      "loss": 0.7299,
+      "step": 1528
+    },
+    {
+      "epoch": 0.48928,
+      "grad_norm": 0.46897050738334656,
+      "learning_rate": 0.0001512249799839872,
+      "loss": 0.6589,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.4058345854282379,
+      "learning_rate": 0.0001511929543634908,
+      "loss": 0.5946,
+      "step": 1530
+    },
+    {
+      "epoch": 0.48992,
+      "grad_norm": 0.3750514090061188,
+      "learning_rate": 0.0001511609287429944,
+      "loss": 0.6293,
+      "step": 1531
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 0.4147687554359436,
+      "learning_rate": 0.000151128903122498,
+      "loss": 0.5674,
+      "step": 1532
+    },
+    {
+      "epoch": 0.49056,
+      "grad_norm": 0.5191084742546082,
+      "learning_rate": 0.00015109687750200162,
+      "loss": 0.6787,
+      "step": 1533
+    },
+    {
+      "epoch": 0.49088,
+      "grad_norm": 0.3888176679611206,
+      "learning_rate": 0.0001510648518815052,
+      "loss": 0.584,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 0.4505578875541687,
+      "learning_rate": 0.0001510328262610088,
+      "loss": 0.6687,
+      "step": 1535
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 0.3876858055591583,
+      "learning_rate": 0.0001510008006405124,
+      "loss": 0.6016,
+      "step": 1536
+    },
+    {
+      "epoch": 0.49184,
+      "grad_norm": 0.3573913872241974,
+      "learning_rate": 0.00015096877502001604,
+      "loss": 0.5396,
+      "step": 1537
+    },
+    {
+      "epoch": 0.49216,
+      "grad_norm": 0.43509185314178467,
+      "learning_rate": 0.00015093674939951963,
+      "loss": 0.5526,
+      "step": 1538
+    },
+    {
+      "epoch": 0.49248,
+      "grad_norm": 0.3340749740600586,
+      "learning_rate": 0.00015090472377902322,
+      "loss": 0.492,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.4673217833042145,
+      "learning_rate": 0.00015087269815852681,
+      "loss": 0.615,
+      "step": 1540
+    },
+    {
+      "epoch": 0.49312,
+      "grad_norm": 0.45726293325424194,
+      "learning_rate": 0.00015084067253803043,
+      "loss": 0.6667,
+      "step": 1541
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 0.3810867667198181,
+      "learning_rate": 0.00015080864691753402,
+      "loss": 0.5195,
+      "step": 1542
+    },
+    {
+      "epoch": 0.49376,
+      "grad_norm": 0.3988247513771057,
+      "learning_rate": 0.00015077662129703764,
+      "loss": 0.577,
+      "step": 1543
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 0.36854708194732666,
+      "learning_rate": 0.00015074459567654123,
+      "loss": 0.4417,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 0.2949543297290802,
+      "learning_rate": 0.00015071257005604485,
+      "loss": 0.5198,
+      "step": 1545
+    },
+    {
+      "epoch": 0.49472,
+      "grad_norm": 0.3645668923854828,
+      "learning_rate": 0.00015068054443554844,
+      "loss": 0.5974,
+      "step": 1546
+    },
+    {
+      "epoch": 0.49504,
+      "grad_norm": 0.4098426401615143,
+      "learning_rate": 0.00015064851881505203,
+      "loss": 0.6777,
+      "step": 1547
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 0.5231944918632507,
+      "learning_rate": 0.00015061649319455565,
+      "loss": 0.7263,
+      "step": 1548
+    },
+    {
+      "epoch": 0.49568,
+      "grad_norm": 0.44490286707878113,
+      "learning_rate": 0.00015058446757405927,
+      "loss": 0.556,
+      "step": 1549
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.40835636854171753,
+      "learning_rate": 0.00015055244195356286,
+      "loss": 0.7233,
+      "step": 1550
+    },
+    {
+      "epoch": 0.49632,
+      "grad_norm": 0.4359986484050751,
+      "learning_rate": 0.00015052041633306645,
+      "loss": 0.7591,
+      "step": 1551
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 0.4524591565132141,
+      "learning_rate": 0.00015048839071257007,
+      "loss": 0.6847,
+      "step": 1552
+    },
+    {
+      "epoch": 0.49696,
+      "grad_norm": 0.38842302560806274,
+      "learning_rate": 0.00015045636509207366,
+      "loss": 0.5877,
+      "step": 1553
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 0.3909432590007782,
+      "learning_rate": 0.00015042433947157725,
+      "loss": 0.5977,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 0.3681485950946808,
+      "learning_rate": 0.00015039231385108087,
+      "loss": 0.5487,
+      "step": 1555
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 0.3967900276184082,
+      "learning_rate": 0.0001503602882305845,
+      "loss": 0.5786,
+      "step": 1556
+    },
+    {
+      "epoch": 0.49824,
+      "grad_norm": 0.5020893812179565,
+      "learning_rate": 0.00015032826261008808,
+      "loss": 0.7933,
+      "step": 1557
+    },
+    {
+      "epoch": 0.49856,
+      "grad_norm": 0.5435899496078491,
+      "learning_rate": 0.00015029623698959167,
+      "loss": 0.8104,
+      "step": 1558
+    },
+    {
+      "epoch": 0.49888,
+      "grad_norm": 0.3473319709300995,
+      "learning_rate": 0.00015026421136909526,
+      "loss": 0.6056,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.38874509930610657,
+      "learning_rate": 0.0001502321857485989,
+      "loss": 0.5864,
+      "step": 1560
+    },
+    {
+      "epoch": 0.49952,
+      "grad_norm": 0.38271698355674744,
+      "learning_rate": 0.0001502001601281025,
+      "loss": 0.6374,
+      "step": 1561
+    },
+    {
+      "epoch": 0.49984,
+      "grad_norm": 0.40099334716796875,
+      "learning_rate": 0.0001501681345076061,
+      "loss": 0.5848,
+      "step": 1562
+    },
+    {
+      "epoch": 0.50016,
+      "grad_norm": 0.36203983426094055,
+      "learning_rate": 0.00015013610888710968,
+      "loss": 0.5741,
+      "step": 1563
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 0.3319595754146576,
+      "learning_rate": 0.0001501040832666133,
+      "loss": 0.5652,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 0.4568646252155304,
+      "learning_rate": 0.0001500720576461169,
+      "loss": 0.6529,
+      "step": 1565
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 0.4290900230407715,
+      "learning_rate": 0.0001500400320256205,
+      "loss": 0.5681,
+      "step": 1566
+    },
+    {
+      "epoch": 0.50144,
+      "grad_norm": 0.348956823348999,
+      "learning_rate": 0.0001500080064051241,
+      "loss": 0.6339,
+      "step": 1567
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 0.46744197607040405,
+      "learning_rate": 0.00014997598078462772,
+      "loss": 0.5531,
+      "step": 1568
+    },
+    {
+      "epoch": 0.50208,
+      "grad_norm": 0.3911832869052887,
+      "learning_rate": 0.0001499439551641313,
+      "loss": 0.7096,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.5339780449867249,
+      "learning_rate": 0.0001499119295436349,
+      "loss": 0.5693,
+      "step": 1570
+    },
+    {
+      "epoch": 0.50272,
+      "grad_norm": 0.40396156907081604,
+      "learning_rate": 0.00014987990392313852,
+      "loss": 0.5991,
+      "step": 1571
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 0.338794082403183,
+      "learning_rate": 0.00014984787830264214,
+      "loss": 0.4688,
+      "step": 1572
+    },
+    {
+      "epoch": 0.50336,
+      "grad_norm": 0.38717058300971985,
+      "learning_rate": 0.00014981585268214573,
+      "loss": 0.5784,
+      "step": 1573
+    },
+    {
+      "epoch": 0.50368,
+      "grad_norm": 0.4268225133419037,
+      "learning_rate": 0.00014978382706164932,
+      "loss": 0.7702,
+      "step": 1574
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 0.41136494278907776,
+      "learning_rate": 0.00014975180144115294,
+      "loss": 0.5413,
+      "step": 1575
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 0.4483546018600464,
+      "learning_rate": 0.00014971977582065653,
+      "loss": 0.6543,
+      "step": 1576
+    },
+    {
+      "epoch": 0.50464,
+      "grad_norm": 0.3953871428966522,
+      "learning_rate": 0.00014968775020016012,
+      "loss": 0.5566,
+      "step": 1577
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 0.5156556367874146,
+      "learning_rate": 0.00014965572457966374,
+      "loss": 0.8931,
+      "step": 1578
+    },
+    {
+      "epoch": 0.50528,
+      "grad_norm": 0.44895628094673157,
+      "learning_rate": 0.00014962369895916736,
+      "loss": 0.5994,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.37594714760780334,
+      "learning_rate": 0.00014959167333867095,
+      "loss": 0.5089,
+      "step": 1580
+    },
+    {
+      "epoch": 0.50592,
+      "grad_norm": 0.40954697132110596,
+      "learning_rate": 0.00014955964771817454,
+      "loss": 0.6289,
+      "step": 1581
+    },
+    {
+      "epoch": 0.50624,
+      "grad_norm": 0.43285027146339417,
+      "learning_rate": 0.00014952762209767813,
+      "loss": 0.6186,
+      "step": 1582
+    },
+    {
+      "epoch": 0.50656,
+      "grad_norm": 0.4416860044002533,
+      "learning_rate": 0.00014949559647718175,
+      "loss": 0.6666,
+      "step": 1583
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 0.4571966826915741,
+      "learning_rate": 0.00014946357085668537,
+      "loss": 0.7006,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 0.3811291754245758,
+      "learning_rate": 0.00014943154523618896,
+      "loss": 0.5971,
+      "step": 1585
+    },
+    {
+      "epoch": 0.50752,
+      "grad_norm": 0.3629266917705536,
+      "learning_rate": 0.00014939951961569255,
+      "loss": 0.5869,
+      "step": 1586
+    },
+    {
+      "epoch": 0.50784,
+      "grad_norm": 0.39941248297691345,
+      "learning_rate": 0.00014936749399519617,
+      "loss": 0.6237,
+      "step": 1587
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 0.4023836851119995,
+      "learning_rate": 0.00014933546837469976,
+      "loss": 0.5955,
+      "step": 1588
+    },
+    {
+      "epoch": 0.50848,
+      "grad_norm": 0.4362434446811676,
+      "learning_rate": 0.00014930344275420335,
+      "loss": 0.5948,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.43958157300949097,
+      "learning_rate": 0.00014927141713370697,
+      "loss": 0.5986,
+      "step": 1590
+    },
+    {
+      "epoch": 0.50912,
+      "grad_norm": 0.366148442029953,
+      "learning_rate": 0.0001492393915132106,
+      "loss": 0.5164,
+      "step": 1591
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 0.4085465669631958,
+      "learning_rate": 0.00014920736589271418,
+      "loss": 0.4489,
+      "step": 1592
+    },
+    {
+      "epoch": 0.50976,
+      "grad_norm": 0.4280279576778412,
+      "learning_rate": 0.00014917534027221777,
+      "loss": 0.6744,
+      "step": 1593
+    },
+    {
+      "epoch": 0.51008,
+      "grad_norm": 0.38611873984336853,
+      "learning_rate": 0.0001491433146517214,
+      "loss": 0.569,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 0.47703760862350464,
+      "learning_rate": 0.00014911128903122498,
+      "loss": 0.6439,
+      "step": 1595
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 0.42542675137519836,
+      "learning_rate": 0.0001490792634107286,
+      "loss": 0.5549,
+      "step": 1596
+    },
+    {
+      "epoch": 0.51104,
+      "grad_norm": 0.42383790016174316,
+      "learning_rate": 0.0001490472377902322,
+      "loss": 0.5157,
+      "step": 1597
+    },
+    {
+      "epoch": 0.51136,
+      "grad_norm": 0.4771800935268402,
+      "learning_rate": 0.0001490152121697358,
+      "loss": 0.7801,
+      "step": 1598
+    },
+    {
+      "epoch": 0.51168,
+      "grad_norm": 0.48154330253601074,
+      "learning_rate": 0.0001489831865492394,
+      "loss": 0.5839,
+      "step": 1599
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.4018516540527344,
+      "learning_rate": 0.000148951160928743,
+      "loss": 0.5543,
+      "step": 1600
+    },
+    {
+      "epoch": 0.51232,
+      "grad_norm": 0.4775291681289673,
+      "learning_rate": 0.00014891913530824658,
+      "loss": 0.6292,
+      "step": 1601
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 0.5031663179397583,
+      "learning_rate": 0.0001488871096877502,
+      "loss": 0.676,
+      "step": 1602
+    },
+    {
+      "epoch": 0.51296,
+      "grad_norm": 0.4088120460510254,
+      "learning_rate": 0.00014885508406725382,
+      "loss": 0.5712,
+      "step": 1603
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 0.4844394624233246,
+      "learning_rate": 0.0001488230584467574,
+      "loss": 0.6865,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 0.4571373462677002,
+      "learning_rate": 0.000148791032826261,
+      "loss": 0.5071,
+      "step": 1605
+    },
+    {
+      "epoch": 0.51392,
+      "grad_norm": 0.41517508029937744,
+      "learning_rate": 0.00014875900720576462,
+      "loss": 0.6985,
+      "step": 1606
+    },
+    {
+      "epoch": 0.51424,
+      "grad_norm": 0.5186550617218018,
+      "learning_rate": 0.00014872698158526824,
+      "loss": 0.5731,
+      "step": 1607
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 0.4299480617046356,
+      "learning_rate": 0.00014869495596477183,
+      "loss": 0.6824,
+      "step": 1608
+    },
+    {
+      "epoch": 0.51488,
+      "grad_norm": 0.4804494380950928,
+      "learning_rate": 0.00014866293034427542,
+      "loss": 0.7593,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 0.42504748702049255,
+      "learning_rate": 0.00014863090472377904,
+      "loss": 0.7173,
+      "step": 1610
+    },
+    {
+      "epoch": 0.51552,
+      "grad_norm": 0.3364151418209076,
+      "learning_rate": 0.00014859887910328263,
+      "loss": 0.61,
+      "step": 1611
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 0.43155744671821594,
+      "learning_rate": 0.00014856685348278622,
+      "loss": 0.6374,
+      "step": 1612
+    },
+    {
+      "epoch": 0.51616,
+      "grad_norm": 0.4159037172794342,
+      "learning_rate": 0.00014853482786228984,
+      "loss": 0.6787,
+      "step": 1613
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 0.4231332242488861,
+      "learning_rate": 0.00014850280224179346,
+      "loss": 0.6093,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 0.4542525112628937,
+      "learning_rate": 0.00014847077662129705,
+      "loss": 0.6462,
+      "step": 1615
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 0.43667614459991455,
+      "learning_rate": 0.00014843875100080064,
+      "loss": 0.5833,
+      "step": 1616
+    },
+    {
+      "epoch": 0.51744,
+      "grad_norm": 0.43001076579093933,
+      "learning_rate": 0.00014840672538030423,
+      "loss": 0.6857,
+      "step": 1617
+    },
+    {
+      "epoch": 0.51776,
+      "grad_norm": 0.37983617186546326,
+      "learning_rate": 0.00014837469975980785,
+      "loss": 0.552,
+      "step": 1618
+    },
+    {
+      "epoch": 0.51808,
+      "grad_norm": 0.513692319393158,
+      "learning_rate": 0.00014834267413931147,
+      "loss": 0.7744,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.515751838684082,
+      "learning_rate": 0.00014831064851881506,
+      "loss": 0.7595,
+      "step": 1620
+    },
+    {
+      "epoch": 0.51872,
+      "grad_norm": 0.38238462805747986,
+      "learning_rate": 0.00014827862289831865,
+      "loss": 0.5659,
+      "step": 1621
+    },
+    {
+      "epoch": 0.51904,
+      "grad_norm": 0.4446403682231903,
+      "learning_rate": 0.00014824659727782227,
+      "loss": 0.5651,
+      "step": 1622
+    },
+    {
+      "epoch": 0.51936,
+      "grad_norm": 0.39717090129852295,
+      "learning_rate": 0.00014821457165732586,
+      "loss": 0.5496,
+      "step": 1623
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 0.3730437755584717,
+      "learning_rate": 0.00014818254603682945,
+      "loss": 0.5872,
+      "step": 1624
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47993630170822144,
+      "learning_rate": 0.00014815052041633307,
+      "loss": 0.7309,
+      "step": 1625
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 0.3931674659252167,
+      "learning_rate": 0.0001481184947958367,
+      "loss": 0.5988,
+      "step": 1626
+    },
+    {
+      "epoch": 0.52064,
+      "grad_norm": 0.42547735571861267,
+      "learning_rate": 0.00014808646917534028,
+      "loss": 0.7006,
+      "step": 1627
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 0.42670825123786926,
+      "learning_rate": 0.00014805444355484387,
+      "loss": 0.6349,
+      "step": 1628
+    },
+    {
+      "epoch": 0.52128,
+      "grad_norm": 0.39265143871307373,
+      "learning_rate": 0.0001480224179343475,
+      "loss": 0.6064,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.4356538951396942,
+      "learning_rate": 0.00014799039231385108,
+      "loss": 0.6593,
+      "step": 1630
+    },
+    {
+      "epoch": 0.52192,
+      "grad_norm": 0.5013262629508972,
+      "learning_rate": 0.0001479583666933547,
+      "loss": 0.6307,
+      "step": 1631
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 0.47995731234550476,
+      "learning_rate": 0.0001479263410728583,
+      "loss": 0.6347,
+      "step": 1632
+    },
+    {
+      "epoch": 0.52256,
+      "grad_norm": 0.37199196219444275,
+      "learning_rate": 0.0001478943154523619,
+      "loss": 0.5399,
+      "step": 1633
+    },
+    {
+      "epoch": 0.52288,
+      "grad_norm": 0.41766753792762756,
+      "learning_rate": 0.0001478622898318655,
+      "loss": 0.5871,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 0.46419382095336914,
+      "learning_rate": 0.0001478302642113691,
+      "loss": 0.649,
+      "step": 1635
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 0.4349631667137146,
+      "learning_rate": 0.00014779823859087268,
+      "loss": 0.4817,
+      "step": 1636
+    },
+    {
+      "epoch": 0.52384,
+      "grad_norm": 0.3878946602344513,
+      "learning_rate": 0.00014776621297037633,
+      "loss": 0.5939,
+      "step": 1637
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 0.40705326199531555,
+      "learning_rate": 0.00014773418734987992,
+      "loss": 0.6121,
+      "step": 1638
+    },
+    {
+      "epoch": 0.52448,
+      "grad_norm": 0.4568576216697693,
+      "learning_rate": 0.0001477021617293835,
+      "loss": 0.6049,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.5750669836997986,
+      "learning_rate": 0.0001476701361088871,
+      "loss": 0.7515,
+      "step": 1640
+    },
+    {
+      "epoch": 0.52512,
+      "grad_norm": 0.48261967301368713,
+      "learning_rate": 0.00014763811048839072,
+      "loss": 0.674,
+      "step": 1641
+    },
+    {
+      "epoch": 0.52544,
+      "grad_norm": 0.4174082577228546,
+      "learning_rate": 0.0001476060848678943,
+      "loss": 0.6256,
+      "step": 1642
+    },
+    {
+      "epoch": 0.52576,
+      "grad_norm": 0.521105170249939,
+      "learning_rate": 0.00014757405924739793,
+      "loss": 0.6919,
+      "step": 1643
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 0.5098194479942322,
+      "learning_rate": 0.00014754203362690152,
+      "loss": 0.7348,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 0.49096381664276123,
+      "learning_rate": 0.00014751000800640514,
+      "loss": 0.5034,
+      "step": 1645
+    },
+    {
+      "epoch": 0.52672,
+      "grad_norm": 0.4650208652019501,
+      "learning_rate": 0.00014747798238590873,
+      "loss": 0.74,
+      "step": 1646
+    },
+    {
+      "epoch": 0.52704,
+      "grad_norm": 0.39836785197257996,
+      "learning_rate": 0.00014744595676541232,
+      "loss": 0.5639,
+      "step": 1647
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 0.4326922595500946,
+      "learning_rate": 0.00014741393114491594,
+      "loss": 0.6063,
+      "step": 1648
+    },
+    {
+      "epoch": 0.52768,
+      "grad_norm": 0.42481115460395813,
+      "learning_rate": 0.00014738190552441956,
+      "loss": 0.5339,
+      "step": 1649
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.445627361536026,
+      "learning_rate": 0.00014734987990392315,
+      "loss": 0.5656,
+      "step": 1650
+    },
+    {
+      "epoch": 0.52832,
+      "grad_norm": 0.47338902950286865,
+      "learning_rate": 0.00014731785428342674,
+      "loss": 0.6351,
+      "step": 1651
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 0.5121362805366516,
+      "learning_rate": 0.00014728582866293036,
+      "loss": 0.6102,
+      "step": 1652
+    },
+    {
+      "epoch": 0.52896,
+      "grad_norm": 0.45235398411750793,
+      "learning_rate": 0.00014725380304243395,
+      "loss": 0.6338,
+      "step": 1653
+    },
+    {
+      "epoch": 0.52928,
+      "grad_norm": 0.45092660188674927,
+      "learning_rate": 0.00014722177742193757,
+      "loss": 0.6857,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 0.42766451835632324,
+      "learning_rate": 0.00014718975180144116,
+      "loss": 0.4965,
+      "step": 1655
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 0.5290222764015198,
+      "learning_rate": 0.00014715772618094478,
+      "loss": 0.7364,
+      "step": 1656
+    },
+    {
+      "epoch": 0.53024,
+      "grad_norm": 0.4118780493736267,
+      "learning_rate": 0.00014712570056044837,
+      "loss": 0.4702,
+      "step": 1657
+    },
+    {
+      "epoch": 0.53056,
+      "grad_norm": 0.48284900188446045,
+      "learning_rate": 0.00014709367493995196,
+      "loss": 0.7274,
+      "step": 1658
+    },
+    {
+      "epoch": 0.53088,
+      "grad_norm": 0.4101698100566864,
+      "learning_rate": 0.00014706164931945555,
+      "loss": 0.4907,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.3862396478652954,
+      "learning_rate": 0.0001470296236989592,
+      "loss": 0.5512,
+      "step": 1660
+    },
+    {
+      "epoch": 0.53152,
+      "grad_norm": 0.40047040581703186,
+      "learning_rate": 0.0001469975980784628,
+      "loss": 0.6572,
+      "step": 1661
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 0.4566109776496887,
+      "learning_rate": 0.00014696557245796638,
+      "loss": 0.5943,
+      "step": 1662
+    },
+    {
+      "epoch": 0.53216,
+      "grad_norm": 0.43418458104133606,
+      "learning_rate": 0.00014693354683746997,
+      "loss": 0.6892,
+      "step": 1663
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 0.4552966356277466,
+      "learning_rate": 0.0001469015212169736,
+      "loss": 0.6584,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 0.4316636621952057,
+      "learning_rate": 0.00014686949559647718,
+      "loss": 0.7559,
+      "step": 1665
+    },
+    {
+      "epoch": 0.53312,
+      "grad_norm": 0.4553171396255493,
+      "learning_rate": 0.0001468374699759808,
+      "loss": 0.5206,
+      "step": 1666
+    },
+    {
+      "epoch": 0.53344,
+      "grad_norm": 0.4388287663459778,
+      "learning_rate": 0.0001468054443554844,
+      "loss": 0.5297,
+      "step": 1667
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 0.5314216017723083,
+      "learning_rate": 0.000146773418734988,
+      "loss": 0.7307,
+      "step": 1668
+    },
+    {
+      "epoch": 0.53408,
+      "grad_norm": 0.4842141270637512,
+      "learning_rate": 0.0001467413931144916,
+      "loss": 0.6355,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.4126124083995819,
+      "learning_rate": 0.0001467093674939952,
+      "loss": 0.5952,
+      "step": 1670
+    },
+    {
+      "epoch": 0.53472,
+      "grad_norm": 0.3560931086540222,
+      "learning_rate": 0.0001466773418734988,
+      "loss": 0.5509,
+      "step": 1671
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 0.4946417212486267,
+      "learning_rate": 0.00014664531625300243,
+      "loss": 0.5342,
+      "step": 1672
+    },
+    {
+      "epoch": 0.53536,
+      "grad_norm": 0.44285520911216736,
+      "learning_rate": 0.00014661329063250602,
+      "loss": 0.7575,
+      "step": 1673
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 0.4728919267654419,
+      "learning_rate": 0.0001465812650120096,
+      "loss": 0.6528,
+      "step": 1674
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 0.40966787934303284,
+      "learning_rate": 0.00014654923939151323,
+      "loss": 0.5089,
+      "step": 1675
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 0.4696500897407532,
+      "learning_rate": 0.00014651721377101682,
+      "loss": 0.6926,
+      "step": 1676
+    },
+    {
+      "epoch": 0.53664,
+      "grad_norm": 0.409135103225708,
+      "learning_rate": 0.0001464851881505204,
+      "loss": 0.5702,
+      "step": 1677
+    },
+    {
+      "epoch": 0.53696,
+      "grad_norm": 0.4043565094470978,
+      "learning_rate": 0.00014645316253002403,
+      "loss": 0.6501,
+      "step": 1678
+    },
+    {
+      "epoch": 0.53728,
+      "grad_norm": 0.4523264765739441,
+      "learning_rate": 0.00014642113690952765,
+      "loss": 0.5603,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.48090124130249023,
+      "learning_rate": 0.00014638911128903124,
+      "loss": 0.6906,
+      "step": 1680
+    },
+    {
+      "epoch": 0.53792,
+      "grad_norm": 0.4024209678173065,
+      "learning_rate": 0.00014635708566853483,
+      "loss": 0.5712,
+      "step": 1681
+    },
+    {
+      "epoch": 0.53824,
+      "grad_norm": 0.4797082841396332,
+      "learning_rate": 0.00014632506004803842,
+      "loss": 0.7643,
+      "step": 1682
+    },
+    {
+      "epoch": 0.53856,
+      "grad_norm": 0.4375150203704834,
+      "learning_rate": 0.00014629303442754204,
+      "loss": 0.67,
+      "step": 1683
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 0.48056143522262573,
+      "learning_rate": 0.00014626100880704566,
+      "loss": 0.5997,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 0.3828798532485962,
+      "learning_rate": 0.00014622898318654925,
+      "loss": 0.5187,
+      "step": 1685
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 0.3476288616657257,
+      "learning_rate": 0.00014619695756605284,
+      "loss": 0.5514,
+      "step": 1686
+    },
+    {
+      "epoch": 0.53984,
+      "grad_norm": 0.44430217146873474,
+      "learning_rate": 0.00014616493194555646,
+      "loss": 0.7041,
+      "step": 1687
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 0.5410586595535278,
+      "learning_rate": 0.00014613290632506005,
+      "loss": 0.6437,
+      "step": 1688
+    },
+    {
+      "epoch": 0.54048,
+      "grad_norm": 0.41584205627441406,
+      "learning_rate": 0.00014610088070456364,
+      "loss": 0.6671,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 0.36268654465675354,
+      "learning_rate": 0.00014606885508406726,
+      "loss": 0.5585,
+      "step": 1690
+    },
+    {
+      "epoch": 0.54112,
+      "grad_norm": 0.420147567987442,
+      "learning_rate": 0.00014603682946357088,
+      "loss": 0.6863,
+      "step": 1691
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 0.4414600431919098,
+      "learning_rate": 0.00014600480384307447,
+      "loss": 0.6573,
+      "step": 1692
+    },
+    {
+      "epoch": 0.54176,
+      "grad_norm": 0.3763164281845093,
+      "learning_rate": 0.00014597277822257806,
+      "loss": 0.539,
+      "step": 1693
+    },
+    {
+      "epoch": 0.54208,
+      "grad_norm": 0.4129893183708191,
+      "learning_rate": 0.00014594075260208168,
+      "loss": 0.622,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 0.4584847092628479,
+      "learning_rate": 0.00014590872698158527,
+      "loss": 0.6959,
+      "step": 1695
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 0.3794647753238678,
+      "learning_rate": 0.0001458767013610889,
+      "loss": 0.6277,
+      "step": 1696
+    },
+    {
+      "epoch": 0.54304,
+      "grad_norm": 0.4008634686470032,
+      "learning_rate": 0.00014584467574059248,
+      "loss": 0.6988,
+      "step": 1697
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 0.4679359495639801,
+      "learning_rate": 0.0001458126501200961,
+      "loss": 0.5581,
+      "step": 1698
+    },
+    {
+      "epoch": 0.54368,
+      "grad_norm": 0.42059269547462463,
+      "learning_rate": 0.0001457806244995997,
+      "loss": 0.6179,
+      "step": 1699
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.38725876808166504,
+      "learning_rate": 0.00014574859887910328,
+      "loss": 0.5545,
+      "step": 1700
+    },
+    {
+      "epoch": 0.54432,
+      "grad_norm": 0.37395504117012024,
+      "learning_rate": 0.0001457165732586069,
+      "loss": 0.4455,
+      "step": 1701
+    },
+    {
+      "epoch": 0.54464,
+      "grad_norm": 0.5110217928886414,
+      "learning_rate": 0.00014568454763811052,
+      "loss": 0.6205,
+      "step": 1702
+    },
+    {
+      "epoch": 0.54496,
+      "grad_norm": 0.3960050046443939,
+      "learning_rate": 0.0001456525220176141,
+      "loss": 0.477,
+      "step": 1703
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 0.45906028151512146,
+      "learning_rate": 0.0001456204963971177,
+      "loss": 0.6088,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 0.4467891454696655,
+      "learning_rate": 0.0001455884707766213,
+      "loss": 0.5971,
+      "step": 1705
+    },
+    {
+      "epoch": 0.54592,
+      "grad_norm": 0.4562687575817108,
+      "learning_rate": 0.0001455564451561249,
+      "loss": 0.477,
+      "step": 1706
+    },
+    {
+      "epoch": 0.54624,
+      "grad_norm": 0.40779098868370056,
+      "learning_rate": 0.00014552441953562853,
+      "loss": 0.5286,
+      "step": 1707
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 0.3931366801261902,
+      "learning_rate": 0.00014549239391513212,
+      "loss": 0.5238,
+      "step": 1708
+    },
+    {
+      "epoch": 0.54688,
+      "grad_norm": 0.5365738868713379,
+      "learning_rate": 0.0001454603682946357,
+      "loss": 0.6509,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.4164142310619354,
+      "learning_rate": 0.00014542834267413933,
+      "loss": 0.584,
+      "step": 1710
+    },
+    {
+      "epoch": 0.54752,
+      "grad_norm": 0.38298431038856506,
+      "learning_rate": 0.00014539631705364292,
+      "loss": 0.6428,
+      "step": 1711
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 0.3999565541744232,
+      "learning_rate": 0.0001453642914331465,
+      "loss": 0.6531,
+      "step": 1712
+    },
+    {
+      "epoch": 0.54816,
+      "grad_norm": 0.4526667296886444,
+      "learning_rate": 0.00014533226581265013,
+      "loss": 0.6357,
+      "step": 1713
+    },
+    {
+      "epoch": 0.54848,
+      "grad_norm": 0.4833303391933441,
+      "learning_rate": 0.00014530024019215375,
+      "loss": 0.6554,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 0.39602264761924744,
+      "learning_rate": 0.00014526821457165734,
+      "loss": 0.6347,
+      "step": 1715
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 0.4341404139995575,
+      "learning_rate": 0.00014523618895116093,
+      "loss": 0.647,
+      "step": 1716
+    },
+    {
+      "epoch": 0.54944,
+      "grad_norm": 0.3716273009777069,
+      "learning_rate": 0.00014520416333066455,
+      "loss": 0.6144,
+      "step": 1717
+    },
+    {
+      "epoch": 0.54976,
+      "grad_norm": 0.4664899408817291,
+      "learning_rate": 0.00014517213771016814,
+      "loss": 0.7534,
+      "step": 1718
+    },
+    {
+      "epoch": 0.55008,
+      "grad_norm": 0.45882874727249146,
+      "learning_rate": 0.00014514011208967176,
+      "loss": 0.7039,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.4413776695728302,
+      "learning_rate": 0.00014510808646917535,
+      "loss": 0.5913,
+      "step": 1720
+    },
+    {
+      "epoch": 0.55072,
+      "grad_norm": 0.40778952836990356,
+      "learning_rate": 0.00014507606084867896,
+      "loss": 0.5868,
+      "step": 1721
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 0.48226913809776306,
+      "learning_rate": 0.00014504403522818256,
+      "loss": 0.6835,
+      "step": 1722
+    },
+    {
+      "epoch": 0.55136,
+      "grad_norm": 0.4796047508716583,
+      "learning_rate": 0.00014501200960768615,
+      "loss": 0.6792,
+      "step": 1723
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 0.4205368459224701,
+      "learning_rate": 0.00014497998398718974,
+      "loss": 0.6487,
+      "step": 1724
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.4827781021595001,
+      "learning_rate": 0.00014494795836669338,
+      "loss": 0.6772,
+      "step": 1725
+    },
+    {
+      "epoch": 0.55232,
+      "grad_norm": 0.3883756697177887,
+      "learning_rate": 0.00014491593274619698,
+      "loss": 0.5581,
+      "step": 1726
+    },
+    {
+      "epoch": 0.55264,
+      "grad_norm": 0.4604884684085846,
+      "learning_rate": 0.00014488390712570057,
+      "loss": 0.5899,
+      "step": 1727
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 0.39551404118537903,
+      "learning_rate": 0.00014485188150520416,
+      "loss": 0.6264,
+      "step": 1728
+    },
+    {
+      "epoch": 0.55328,
+      "grad_norm": 0.4296308755874634,
+      "learning_rate": 0.00014481985588470778,
+      "loss": 0.6819,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.39290544390678406,
+      "learning_rate": 0.00014478783026421137,
+      "loss": 0.5193,
+      "step": 1730
+    },
+    {
+      "epoch": 0.55392,
+      "grad_norm": 0.48716282844543457,
+      "learning_rate": 0.00014475580464371499,
+      "loss": 0.6637,
+      "step": 1731
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 0.4629124104976654,
+      "learning_rate": 0.00014472377902321858,
+      "loss": 0.7196,
+      "step": 1732
+    },
+    {
+      "epoch": 0.55456,
+      "grad_norm": 0.43410563468933105,
+      "learning_rate": 0.0001446917534027222,
+      "loss": 0.5638,
+      "step": 1733
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 0.3920772969722748,
+      "learning_rate": 0.00014465972778222579,
+      "loss": 0.6008,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 0.5046433806419373,
+      "learning_rate": 0.00014462770216172938,
+      "loss": 0.7342,
+      "step": 1735
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 0.3192563056945801,
+      "learning_rate": 0.000144595676541233,
+      "loss": 0.5282,
+      "step": 1736
+    },
+    {
+      "epoch": 0.55584,
+      "grad_norm": 0.4587495028972626,
+      "learning_rate": 0.00014456365092073661,
+      "loss": 0.7029,
+      "step": 1737
+    },
+    {
+      "epoch": 0.55616,
+      "grad_norm": 0.3873540759086609,
+      "learning_rate": 0.0001445316253002402,
+      "loss": 0.58,
+      "step": 1738
+    },
+    {
+      "epoch": 0.55648,
+      "grad_norm": 0.41620585322380066,
+      "learning_rate": 0.0001444995996797438,
+      "loss": 0.5649,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.38239285349845886,
+      "learning_rate": 0.00014446757405924741,
+      "loss": 0.6184,
+      "step": 1740
+    },
+    {
+      "epoch": 0.55712,
+      "grad_norm": 0.37054216861724854,
+      "learning_rate": 0.000144435548438751,
+      "loss": 0.5158,
+      "step": 1741
+    },
+    {
+      "epoch": 0.55744,
+      "grad_norm": 0.44745829701423645,
+      "learning_rate": 0.0001444035228182546,
+      "loss": 0.5241,
+      "step": 1742
+    },
+    {
+      "epoch": 0.55776,
+      "grad_norm": 0.38824939727783203,
+      "learning_rate": 0.00014437149719775822,
+      "loss": 0.5352,
+      "step": 1743
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 0.3455026149749756,
+      "learning_rate": 0.0001443394715772618,
+      "loss": 0.4913,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 0.38081660866737366,
+      "learning_rate": 0.00014430744595676543,
+      "loss": 0.536,
+      "step": 1745
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 0.4633041024208069,
+      "learning_rate": 0.00014427542033626902,
+      "loss": 0.5347,
+      "step": 1746
+    },
+    {
+      "epoch": 0.55904,
+      "grad_norm": 0.4037226438522339,
+      "learning_rate": 0.0001442433947157726,
+      "loss": 0.6332,
+      "step": 1747
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 0.43841251730918884,
+      "learning_rate": 0.00014421136909527623,
+      "loss": 0.5863,
+      "step": 1748
+    },
+    {
+      "epoch": 0.55968,
+      "grad_norm": 0.44607219099998474,
+      "learning_rate": 0.00014417934347477984,
+      "loss": 0.4779,
+      "step": 1749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34019142389297485,
+      "learning_rate": 0.00014414731785428344,
+      "loss": 0.5466,
+      "step": 1750
+    },
+    {
+      "epoch": 0.56032,
+      "grad_norm": 0.37515807151794434,
+      "learning_rate": 0.00014411529223378703,
+      "loss": 0.4955,
+      "step": 1751
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 0.47187891602516174,
+      "learning_rate": 0.00014408326661329064,
+      "loss": 0.67,
+      "step": 1752
+    },
+    {
+      "epoch": 0.56096,
+      "grad_norm": 0.40522485971450806,
+      "learning_rate": 0.00014405124099279424,
+      "loss": 0.4284,
+      "step": 1753
+    },
+    {
+      "epoch": 0.56128,
+      "grad_norm": 0.4068461060523987,
+      "learning_rate": 0.00014401921537229785,
+      "loss": 0.6375,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 0.48030760884284973,
+      "learning_rate": 0.00014398718975180145,
+      "loss": 0.5957,
+      "step": 1755
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 0.41475406289100647,
+      "learning_rate": 0.00014395516413130506,
+      "loss": 0.562,
+      "step": 1756
+    },
+    {
+      "epoch": 0.56224,
+      "grad_norm": 0.4639141857624054,
+      "learning_rate": 0.00014392313851080866,
+      "loss": 0.6586,
+      "step": 1757
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 0.35222023725509644,
+      "learning_rate": 0.00014389111289031225,
+      "loss": 0.5231,
+      "step": 1758
+    },
+    {
+      "epoch": 0.56288,
+      "grad_norm": 0.5137410759925842,
+      "learning_rate": 0.00014385908726981584,
+      "loss": 0.7343,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.48445773124694824,
+      "learning_rate": 0.00014382706164931948,
+      "loss": 0.6133,
+      "step": 1760
+    },
+    {
+      "epoch": 0.56352,
+      "grad_norm": 0.42886993288993835,
+      "learning_rate": 0.00014379503602882307,
+      "loss": 0.5675,
+      "step": 1761
+    },
+    {
+      "epoch": 0.56384,
+      "grad_norm": 0.480501264333725,
+      "learning_rate": 0.00014376301040832667,
+      "loss": 0.6358,
+      "step": 1762
+    },
+    {
+      "epoch": 0.56416,
+      "grad_norm": 0.44278985261917114,
+      "learning_rate": 0.00014373098478783026,
+      "loss": 0.5948,
+      "step": 1763
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 0.4054180085659027,
+      "learning_rate": 0.00014369895916733387,
+      "loss": 0.6896,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 0.4662283658981323,
+      "learning_rate": 0.00014366693354683747,
+      "loss": 0.7067,
+      "step": 1765
+    },
+    {
+      "epoch": 0.56512,
+      "grad_norm": 0.33461177349090576,
+      "learning_rate": 0.00014363490792634108,
+      "loss": 0.565,
+      "step": 1766
+    },
+    {
+      "epoch": 0.56544,
+      "grad_norm": 0.3917163908481598,
+      "learning_rate": 0.00014360288230584468,
+      "loss": 0.6323,
+      "step": 1767
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 0.40968838334083557,
+      "learning_rate": 0.0001435708566853483,
+      "loss": 0.6896,
+      "step": 1768
+    },
+    {
+      "epoch": 0.56608,
+      "grad_norm": 0.36078017950057983,
+      "learning_rate": 0.00014353883106485189,
+      "loss": 0.5385,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.4331192970275879,
+      "learning_rate": 0.00014350680544435548,
+      "loss": 0.5611,
+      "step": 1770
+    },
+    {
+      "epoch": 0.56672,
+      "grad_norm": 0.4258638918399811,
+      "learning_rate": 0.0001434747798238591,
+      "loss": 0.6626,
+      "step": 1771
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 0.353252112865448,
+      "learning_rate": 0.0001434427542033627,
+      "loss": 0.5402,
+      "step": 1772
+    },
+    {
+      "epoch": 0.56736,
+      "grad_norm": 0.4138263165950775,
+      "learning_rate": 0.0001434107285828663,
+      "loss": 0.6232,
+      "step": 1773
+    },
+    {
+      "epoch": 0.56768,
+      "grad_norm": 0.36915355920791626,
+      "learning_rate": 0.0001433787029623699,
+      "loss": 0.5986,
+      "step": 1774
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 0.42587748169898987,
+      "learning_rate": 0.00014334667734187351,
+      "loss": 0.5611,
+      "step": 1775
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 0.4259795546531677,
+      "learning_rate": 0.0001433146517213771,
+      "loss": 0.6804,
+      "step": 1776
+    },
+    {
+      "epoch": 0.56864,
+      "grad_norm": 0.4760611057281494,
+      "learning_rate": 0.0001432826261008807,
+      "loss": 0.609,
+      "step": 1777
+    },
+    {
+      "epoch": 0.56896,
+      "grad_norm": 0.42745864391326904,
+      "learning_rate": 0.00014325060048038431,
+      "loss": 0.6158,
+      "step": 1778
+    },
+    {
+      "epoch": 0.56928,
+      "grad_norm": 0.41684451699256897,
+      "learning_rate": 0.00014321857485988793,
+      "loss": 0.6554,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.4265507459640503,
+      "learning_rate": 0.00014318654923939152,
+      "loss": 0.5261,
+      "step": 1780
+    },
+    {
+      "epoch": 0.56992,
+      "grad_norm": 0.4935876429080963,
+      "learning_rate": 0.00014315452361889512,
+      "loss": 0.6043,
+      "step": 1781
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 0.48855170607566833,
+      "learning_rate": 0.0001431224979983987,
+      "loss": 0.6483,
+      "step": 1782
+    },
+    {
+      "epoch": 0.57056,
+      "grad_norm": 0.49000176787376404,
+      "learning_rate": 0.00014309047237790232,
+      "loss": 0.5694,
+      "step": 1783
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 0.5354358553886414,
+      "learning_rate": 0.00014305844675740594,
+      "loss": 0.6597,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 0.4523873031139374,
+      "learning_rate": 0.00014302642113690953,
+      "loss": 0.8609,
+      "step": 1785
+    },
+    {
+      "epoch": 0.57152,
+      "grad_norm": 0.4434649646282196,
+      "learning_rate": 0.00014299439551641313,
+      "loss": 0.4802,
+      "step": 1786
+    },
+    {
+      "epoch": 0.57184,
+      "grad_norm": 0.5207880139350891,
+      "learning_rate": 0.00014296236989591674,
+      "loss": 0.6431,
+      "step": 1787
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 0.41156113147735596,
+      "learning_rate": 0.00014293034427542033,
+      "loss": 0.5851,
+      "step": 1788
+    },
+    {
+      "epoch": 0.57248,
+      "grad_norm": 0.3727734386920929,
+      "learning_rate": 0.00014289831865492393,
+      "loss": 0.5258,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.4580017924308777,
+      "learning_rate": 0.00014286629303442754,
+      "loss": 0.6369,
+      "step": 1790
+    },
+    {
+      "epoch": 0.57312,
+      "grad_norm": 0.42845404148101807,
+      "learning_rate": 0.00014283426741393116,
+      "loss": 0.6917,
+      "step": 1791
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 0.5139454007148743,
+      "learning_rate": 0.00014280224179343475,
+      "loss": 0.6471,
+      "step": 1792
+    },
+    {
+      "epoch": 0.57376,
+      "grad_norm": 0.352464884519577,
+      "learning_rate": 0.00014277021617293835,
+      "loss": 0.4432,
+      "step": 1793
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 0.47578784823417664,
+      "learning_rate": 0.00014273819055244196,
+      "loss": 0.6324,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 0.4228172302246094,
+      "learning_rate": 0.00014270616493194558,
+      "loss": 0.6564,
+      "step": 1795
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 0.45206505060195923,
+      "learning_rate": 0.00014267413931144917,
+      "loss": 0.6719,
+      "step": 1796
+    },
+    {
+      "epoch": 0.57504,
+      "grad_norm": 0.49961841106414795,
+      "learning_rate": 0.00014264211369095276,
+      "loss": 0.6919,
+      "step": 1797
+    },
+    {
+      "epoch": 0.57536,
+      "grad_norm": 0.3252984881401062,
+      "learning_rate": 0.00014261008807045638,
+      "loss": 0.4752,
+      "step": 1798
+    },
+    {
+      "epoch": 0.57568,
+      "grad_norm": 0.5096743106842041,
+      "learning_rate": 0.00014257806244995997,
+      "loss": 0.7494,
+      "step": 1799
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.46969664096832275,
+      "learning_rate": 0.00014254603682946356,
+      "loss": 0.6144,
+      "step": 1800
+    },
+    {
+      "epoch": 0.57632,
+      "grad_norm": 0.4455832839012146,
+      "learning_rate": 0.00014251401120896718,
+      "loss": 0.5457,
+      "step": 1801
+    },
+    {
+      "epoch": 0.57664,
+      "grad_norm": 0.4240313172340393,
+      "learning_rate": 0.0001424819855884708,
+      "loss": 0.5784,
+      "step": 1802
+    },
+    {
+      "epoch": 0.57696,
+      "grad_norm": 0.48070523142814636,
+      "learning_rate": 0.0001424499599679744,
+      "loss": 0.5803,
+      "step": 1803
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 0.40027230978012085,
+      "learning_rate": 0.00014241793434747798,
+      "loss": 0.5253,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 0.34819960594177246,
+      "learning_rate": 0.00014238590872698158,
+      "loss": 0.4523,
+      "step": 1805
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 0.40567445755004883,
+      "learning_rate": 0.0001423538831064852,
+      "loss": 0.5738,
+      "step": 1806
+    },
+    {
+      "epoch": 0.57824,
+      "grad_norm": 0.5148231983184814,
+      "learning_rate": 0.0001423218574859888,
+      "loss": 0.6284,
+      "step": 1807
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 0.46701791882514954,
+      "learning_rate": 0.0001422898318654924,
+      "loss": 0.5447,
+      "step": 1808
+    },
+    {
+      "epoch": 0.57888,
+      "grad_norm": 0.43240249156951904,
+      "learning_rate": 0.000142257806244996,
+      "loss": 0.5879,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.42619985342025757,
+      "learning_rate": 0.0001422257806244996,
+      "loss": 0.5852,
+      "step": 1810
+    },
+    {
+      "epoch": 0.57952,
+      "grad_norm": 0.44732069969177246,
+      "learning_rate": 0.0001421937550040032,
+      "loss": 0.7388,
+      "step": 1811
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 0.4305241107940674,
+      "learning_rate": 0.0001421617293835068,
+      "loss": 0.6114,
+      "step": 1812
+    },
+    {
+      "epoch": 0.58016,
+      "grad_norm": 0.4546957314014435,
+      "learning_rate": 0.0001421297037630104,
+      "loss": 0.6543,
+      "step": 1813
+    },
+    {
+      "epoch": 0.58048,
+      "grad_norm": 0.41488903760910034,
+      "learning_rate": 0.00014209767814251403,
+      "loss": 0.5938,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 0.44933220744132996,
+      "learning_rate": 0.00014206565252201762,
+      "loss": 0.6074,
+      "step": 1815
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 0.5064064264297485,
+      "learning_rate": 0.00014203362690152121,
+      "loss": 0.6985,
+      "step": 1816
+    },
+    {
+      "epoch": 0.58144,
+      "grad_norm": 0.4278789758682251,
+      "learning_rate": 0.00014200160128102483,
+      "loss": 0.5309,
+      "step": 1817
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 0.47378063201904297,
+      "learning_rate": 0.00014196957566052842,
+      "loss": 0.4672,
+      "step": 1818
+    },
+    {
+      "epoch": 0.58208,
+      "grad_norm": 0.49075987935066223,
+      "learning_rate": 0.00014193755004003204,
+      "loss": 0.7948,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.398196280002594,
+      "learning_rate": 0.00014190552441953563,
+      "loss": 0.6496,
+      "step": 1820
+    },
+    {
+      "epoch": 0.58272,
+      "grad_norm": 0.43981173634529114,
+      "learning_rate": 0.00014187349879903925,
+      "loss": 0.6876,
+      "step": 1821
+    },
+    {
+      "epoch": 0.58304,
+      "grad_norm": 0.43671715259552,
+      "learning_rate": 0.00014184147317854284,
+      "loss": 0.58,
+      "step": 1822
+    },
+    {
+      "epoch": 0.58336,
+      "grad_norm": 0.4633326828479767,
+      "learning_rate": 0.00014180944755804643,
+      "loss": 0.6877,
+      "step": 1823
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 0.4382314383983612,
+      "learning_rate": 0.00014177742193755003,
+      "loss": 0.5914,
+      "step": 1824
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.34588053822517395,
+      "learning_rate": 0.00014174539631705367,
+      "loss": 0.6198,
+      "step": 1825
+    },
+    {
+      "epoch": 0.58432,
+      "grad_norm": 0.3872893452644348,
+      "learning_rate": 0.00014171337069655726,
+      "loss": 0.5636,
+      "step": 1826
+    },
+    {
+      "epoch": 0.58464,
+      "grad_norm": 0.4250887632369995,
+      "learning_rate": 0.00014168134507606085,
+      "loss": 0.7043,
+      "step": 1827
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 0.5872929096221924,
+      "learning_rate": 0.00014164931945556444,
+      "loss": 0.6625,
+      "step": 1828
+    },
+    {
+      "epoch": 0.58528,
+      "grad_norm": 0.3709658682346344,
+      "learning_rate": 0.00014161729383506806,
+      "loss": 0.4868,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.5066165924072266,
+      "learning_rate": 0.00014158526821457165,
+      "loss": 0.6119,
+      "step": 1830
+    },
+    {
+      "epoch": 0.58592,
+      "grad_norm": 0.44675418734550476,
+      "learning_rate": 0.00014155324259407527,
+      "loss": 0.6083,
+      "step": 1831
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 0.47645026445388794,
+      "learning_rate": 0.00014152121697357886,
+      "loss": 0.6441,
+      "step": 1832
+    },
+    {
+      "epoch": 0.58656,
+      "grad_norm": 0.440211683511734,
+      "learning_rate": 0.00014148919135308248,
+      "loss": 0.6357,
+      "step": 1833
+    },
+    {
+      "epoch": 0.58688,
+      "grad_norm": 0.4596514105796814,
+      "learning_rate": 0.00014145716573258607,
+      "loss": 0.7876,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 0.38676130771636963,
+      "learning_rate": 0.00014142514011208966,
+      "loss": 0.6457,
+      "step": 1835
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 0.4354492723941803,
+      "learning_rate": 0.00014139311449159328,
+      "loss": 0.4696,
+      "step": 1836
+    },
+    {
+      "epoch": 0.58784,
+      "grad_norm": 0.4639611542224884,
+      "learning_rate": 0.0001413610888710969,
+      "loss": 0.6994,
+      "step": 1837
+    },
+    {
+      "epoch": 0.58816,
+      "grad_norm": 0.46714600920677185,
+      "learning_rate": 0.0001413290632506005,
+      "loss": 0.5299,
+      "step": 1838
+    },
+    {
+      "epoch": 0.58848,
+      "grad_norm": 0.4295344352722168,
+      "learning_rate": 0.00014129703763010408,
+      "loss": 0.5054,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.4129548966884613,
+      "learning_rate": 0.0001412650120096077,
+      "loss": 0.5197,
+      "step": 1840
+    },
+    {
+      "epoch": 0.58912,
+      "grad_norm": 0.4886884391307831,
+      "learning_rate": 0.0001412329863891113,
+      "loss": 0.6881,
+      "step": 1841
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 0.42168864607810974,
+      "learning_rate": 0.0001412009607686149,
+      "loss": 0.6335,
+      "step": 1842
+    },
+    {
+      "epoch": 0.58976,
+      "grad_norm": 0.4845680594444275,
+      "learning_rate": 0.0001411689351481185,
+      "loss": 0.7042,
+      "step": 1843
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 0.5219905972480774,
+      "learning_rate": 0.00014113690952762212,
+      "loss": 0.6196,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 0.49918410181999207,
+      "learning_rate": 0.0001411048839071257,
+      "loss": 0.7674,
+      "step": 1845
+    },
+    {
+      "epoch": 0.59072,
+      "grad_norm": 0.4350354075431824,
+      "learning_rate": 0.0001410728582866293,
+      "loss": 0.5481,
+      "step": 1846
+    },
+    {
+      "epoch": 0.59104,
+      "grad_norm": 0.506687343120575,
+      "learning_rate": 0.0001410408326661329,
+      "loss": 0.6457,
+      "step": 1847
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 0.4148949682712555,
+      "learning_rate": 0.00014100880704563654,
+      "loss": 0.6811,
+      "step": 1848
+    },
+    {
+      "epoch": 0.59168,
+      "grad_norm": 0.46807464957237244,
+      "learning_rate": 0.00014097678142514013,
+      "loss": 0.5943,
+      "step": 1849
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.48883989453315735,
+      "learning_rate": 0.00014094475580464372,
+      "loss": 0.5435,
+      "step": 1850
+    },
+    {
+      "epoch": 0.59232,
+      "grad_norm": 0.5051497220993042,
+      "learning_rate": 0.0001409127301841473,
+      "loss": 0.6287,
+      "step": 1851
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 0.4339817762374878,
+      "learning_rate": 0.00014088070456365093,
+      "loss": 0.6854,
+      "step": 1852
+    },
+    {
+      "epoch": 0.59296,
+      "grad_norm": 0.45705485343933105,
+      "learning_rate": 0.00014084867894315452,
+      "loss": 0.6843,
+      "step": 1853
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 0.38201722502708435,
+      "learning_rate": 0.00014081665332265814,
+      "loss": 0.5343,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 0.3925459384918213,
+      "learning_rate": 0.00014078462770216173,
+      "loss": 0.4351,
+      "step": 1855
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 0.4138433039188385,
+      "learning_rate": 0.00014075260208166535,
+      "loss": 0.6396,
+      "step": 1856
+    },
+    {
+      "epoch": 0.59424,
+      "grad_norm": 0.4435432255268097,
+      "learning_rate": 0.00014072057646116894,
+      "loss": 0.6056,
+      "step": 1857
+    },
+    {
+      "epoch": 0.59456,
+      "grad_norm": 0.34681591391563416,
+      "learning_rate": 0.00014068855084067253,
+      "loss": 0.4672,
+      "step": 1858
+    },
+    {
+      "epoch": 0.59488,
+      "grad_norm": 0.452079176902771,
+      "learning_rate": 0.00014065652522017615,
+      "loss": 0.7392,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.41505107283592224,
+      "learning_rate": 0.00014062449959967977,
+      "loss": 0.5979,
+      "step": 1860
+    },
+    {
+      "epoch": 0.59552,
+      "grad_norm": 0.43770620226860046,
+      "learning_rate": 0.00014059247397918336,
+      "loss": 0.5554,
+      "step": 1861
+    },
+    {
+      "epoch": 0.59584,
+      "grad_norm": 0.42086535692214966,
+      "learning_rate": 0.00014056044835868695,
+      "loss": 0.6576,
+      "step": 1862
+    },
+    {
+      "epoch": 0.59616,
+      "grad_norm": 0.5748926997184753,
+      "learning_rate": 0.00014052842273819057,
+      "loss": 0.7619,
+      "step": 1863
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 0.5288527607917786,
+      "learning_rate": 0.00014049639711769416,
+      "loss": 0.7379,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 0.4697395861148834,
+      "learning_rate": 0.00014046437149719775,
+      "loss": 0.6785,
+      "step": 1865
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 0.4885197579860687,
+      "learning_rate": 0.00014043234587670137,
+      "loss": 0.6997,
+      "step": 1866
+    },
+    {
+      "epoch": 0.59744,
+      "grad_norm": 0.42511868476867676,
+      "learning_rate": 0.000140400320256205,
+      "loss": 0.648,
+      "step": 1867
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 0.47124049067497253,
+      "learning_rate": 0.00014036829463570858,
+      "loss": 0.638,
+      "step": 1868
+    },
+    {
+      "epoch": 0.59808,
+      "grad_norm": 0.4644719064235687,
+      "learning_rate": 0.00014033626901521217,
+      "loss": 0.6853,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.3428098261356354,
+      "learning_rate": 0.00014030424339471576,
+      "loss": 0.4851,
+      "step": 1870
+    },
+    {
+      "epoch": 0.59872,
+      "grad_norm": 0.35469067096710205,
+      "learning_rate": 0.00014027221777421938,
+      "loss": 0.4961,
+      "step": 1871
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 0.4025283753871918,
+      "learning_rate": 0.000140240192153723,
+      "loss": 0.494,
+      "step": 1872
+    },
+    {
+      "epoch": 0.59936,
+      "grad_norm": 0.4813465178012848,
+      "learning_rate": 0.0001402081665332266,
+      "loss": 0.5643,
+      "step": 1873
+    },
+    {
+      "epoch": 0.59968,
+      "grad_norm": 0.4175402522087097,
+      "learning_rate": 0.00014017614091273018,
+      "loss": 0.6529,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5116497278213501,
+      "learning_rate": 0.0001401441152922338,
+      "loss": 0.6803,
+      "step": 1875
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 0.45636916160583496,
+      "learning_rate": 0.0001401120896717374,
+      "loss": 0.5968,
+      "step": 1876
+    },
+    {
+      "epoch": 0.60064,
+      "grad_norm": 0.3418964147567749,
+      "learning_rate": 0.00014008006405124098,
+      "loss": 0.5019,
+      "step": 1877
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 0.48982444405555725,
+      "learning_rate": 0.0001400480384307446,
+      "loss": 0.5053,
+      "step": 1878
+    },
+    {
+      "epoch": 0.60128,
+      "grad_norm": 0.38959190249443054,
+      "learning_rate": 0.00014001601281024822,
+      "loss": 0.561,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.4799782633781433,
+      "learning_rate": 0.0001399839871897518,
+      "loss": 0.6248,
+      "step": 1880
+    },
+    {
+      "epoch": 0.60192,
+      "grad_norm": 0.3324850797653198,
+      "learning_rate": 0.0001399519615692554,
+      "loss": 0.5022,
+      "step": 1881
+    },
+    {
+      "epoch": 0.60224,
+      "grad_norm": 0.48814281821250916,
+      "learning_rate": 0.00013991993594875902,
+      "loss": 0.7222,
+      "step": 1882
+    },
+    {
+      "epoch": 0.60256,
+      "grad_norm": 0.3707048296928406,
+      "learning_rate": 0.0001398879103282626,
+      "loss": 0.6903,
+      "step": 1883
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 0.41187790036201477,
+      "learning_rate": 0.00013985588470776623,
+      "loss": 0.5952,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 0.38265353441238403,
+      "learning_rate": 0.00013982385908726982,
+      "loss": 0.4851,
+      "step": 1885
+    },
+    {
+      "epoch": 0.60352,
+      "grad_norm": 0.371677428483963,
+      "learning_rate": 0.0001397918334667734,
+      "loss": 0.5994,
+      "step": 1886
+    },
+    {
+      "epoch": 0.60384,
+      "grad_norm": 0.4569857716560364,
+      "learning_rate": 0.00013975980784627703,
+      "loss": 0.641,
+      "step": 1887
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 0.44913116097450256,
+      "learning_rate": 0.00013972778222578062,
+      "loss": 0.5719,
+      "step": 1888
+    },
+    {
+      "epoch": 0.60448,
+      "grad_norm": 0.41148558259010315,
+      "learning_rate": 0.00013969575660528424,
+      "loss": 0.6027,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.5498834848403931,
+      "learning_rate": 0.00013966373098478783,
+      "loss": 0.6585,
+      "step": 1890
+    },
+    {
+      "epoch": 0.60512,
+      "grad_norm": 0.479981392621994,
+      "learning_rate": 0.00013963170536429145,
+      "loss": 0.7498,
+      "step": 1891
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 0.43673673272132874,
+      "learning_rate": 0.00013959967974379504,
+      "loss": 0.5914,
+      "step": 1892
+    },
+    {
+      "epoch": 0.60576,
+      "grad_norm": 0.4892735183238983,
+      "learning_rate": 0.00013956765412329863,
+      "loss": 0.7764,
+      "step": 1893
+    },
+    {
+      "epoch": 0.60608,
+      "grad_norm": 0.38701245188713074,
+      "learning_rate": 0.00013953562850280225,
+      "loss": 0.5933,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 0.4067130982875824,
+      "learning_rate": 0.00013950360288230587,
+      "loss": 0.5968,
+      "step": 1895
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 0.4155884385108948,
+      "learning_rate": 0.00013947157726180946,
+      "loss": 0.6623,
+      "step": 1896
+    },
+    {
+      "epoch": 0.60704,
+      "grad_norm": 0.36536359786987305,
+      "learning_rate": 0.00013943955164131305,
+      "loss": 0.371,
+      "step": 1897
+    },
+    {
+      "epoch": 0.60736,
+      "grad_norm": 0.4521442949771881,
+      "learning_rate": 0.00013940752602081667,
+      "loss": 0.6348,
+      "step": 1898
+    },
+    {
+      "epoch": 0.60768,
+      "grad_norm": 0.5207617282867432,
+      "learning_rate": 0.00013937550040032026,
+      "loss": 0.6536,
+      "step": 1899
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.3429977595806122,
+      "learning_rate": 0.00013934347477982385,
+      "loss": 0.498,
+      "step": 1900
+    },
+    {
+      "epoch": 0.60832,
+      "grad_norm": 0.4333025813102722,
+      "learning_rate": 0.00013931144915932747,
+      "loss": 0.6585,
+      "step": 1901
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 0.48309600353240967,
+      "learning_rate": 0.0001392794235388311,
+      "loss": 0.5972,
+      "step": 1902
+    },
+    {
+      "epoch": 0.60896,
+      "grad_norm": 0.4739081561565399,
+      "learning_rate": 0.00013924739791833468,
+      "loss": 0.754,
+      "step": 1903
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 0.4674888849258423,
+      "learning_rate": 0.00013921537229783827,
+      "loss": 0.7017,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 0.4589008390903473,
+      "learning_rate": 0.00013918334667734186,
+      "loss": 0.6724,
+      "step": 1905
+    },
+    {
+      "epoch": 0.60992,
+      "grad_norm": 0.4239599406719208,
+      "learning_rate": 0.00013915132105684548,
+      "loss": 0.6093,
+      "step": 1906
+    },
+    {
+      "epoch": 0.61024,
+      "grad_norm": 0.4922758936882019,
+      "learning_rate": 0.0001391192954363491,
+      "loss": 0.6643,
+      "step": 1907
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 0.39211738109588623,
+      "learning_rate": 0.0001390872698158527,
+      "loss": 0.5392,
+      "step": 1908
+    },
+    {
+      "epoch": 0.61088,
+      "grad_norm": 0.4257199764251709,
+      "learning_rate": 0.00013905524419535628,
+      "loss": 0.6819,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.45310840010643005,
+      "learning_rate": 0.0001390232185748599,
+      "loss": 0.7065,
+      "step": 1910
+    },
+    {
+      "epoch": 0.61152,
+      "grad_norm": 0.470488041639328,
+      "learning_rate": 0.0001389911929543635,
+      "loss": 0.6652,
+      "step": 1911
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 0.4424392282962799,
+      "learning_rate": 0.00013895916733386708,
+      "loss": 0.5691,
+      "step": 1912
+    },
+    {
+      "epoch": 0.61216,
+      "grad_norm": 0.4267422556877136,
+      "learning_rate": 0.0001389271417133707,
+      "loss": 0.5571,
+      "step": 1913
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 0.5310940742492676,
+      "learning_rate": 0.00013889511609287432,
+      "loss": 0.6034,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 0.3715246915817261,
+      "learning_rate": 0.0001388630904723779,
+      "loss": 0.4711,
+      "step": 1915
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 0.4167306423187256,
+      "learning_rate": 0.0001388310648518815,
+      "loss": 0.5749,
+      "step": 1916
+    },
+    {
+      "epoch": 0.61344,
+      "grad_norm": 0.443233847618103,
+      "learning_rate": 0.00013879903923138512,
+      "loss": 0.6432,
+      "step": 1917
+    },
+    {
+      "epoch": 0.61376,
+      "grad_norm": 0.46675318479537964,
+      "learning_rate": 0.0001387670136108887,
+      "loss": 0.6681,
+      "step": 1918
+    },
+    {
+      "epoch": 0.61408,
+      "grad_norm": 0.44543150067329407,
+      "learning_rate": 0.00013873498799039233,
+      "loss": 0.6333,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.5346324443817139,
+      "learning_rate": 0.00013870296236989592,
+      "loss": 0.7415,
+      "step": 1920
+    },
+    {
+      "epoch": 0.61472,
+      "grad_norm": 0.4127179682254791,
+      "learning_rate": 0.00013867093674939954,
+      "loss": 0.5292,
+      "step": 1921
+    },
+    {
+      "epoch": 0.61504,
+      "grad_norm": 0.4970286190509796,
+      "learning_rate": 0.00013863891112890313,
+      "loss": 0.6438,
+      "step": 1922
+    },
+    {
+      "epoch": 0.61536,
+      "grad_norm": 0.4001183807849884,
+      "learning_rate": 0.00013860688550840672,
+      "loss": 0.6133,
+      "step": 1923
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 0.4613015353679657,
+      "learning_rate": 0.0001385748598879103,
+      "loss": 0.4796,
+      "step": 1924
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.4282706379890442,
+      "learning_rate": 0.00013854283426741396,
+      "loss": 0.4651,
+      "step": 1925
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 0.4365302324295044,
+      "learning_rate": 0.00013851080864691755,
+      "loss": 0.6103,
+      "step": 1926
+    },
+    {
+      "epoch": 0.61664,
+      "grad_norm": 0.4845772087574005,
+      "learning_rate": 0.00013847878302642114,
+      "loss": 0.7988,
+      "step": 1927
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 0.3877650499343872,
+      "learning_rate": 0.00013844675740592473,
+      "loss": 0.5235,
+      "step": 1928
+    },
+    {
+      "epoch": 0.61728,
+      "grad_norm": 0.4662170112133026,
+      "learning_rate": 0.00013841473178542835,
+      "loss": 0.6184,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.3995099365711212,
+      "learning_rate": 0.00013838270616493194,
+      "loss": 0.6162,
+      "step": 1930
+    },
+    {
+      "epoch": 0.61792,
+      "grad_norm": 0.370958149433136,
+      "learning_rate": 0.00013835068054443556,
+      "loss": 0.5479,
+      "step": 1931
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 0.44224318861961365,
+      "learning_rate": 0.00013831865492393915,
+      "loss": 0.6934,
+      "step": 1932
+    },
+    {
+      "epoch": 0.61856,
+      "grad_norm": 0.43156903982162476,
+      "learning_rate": 0.00013828662930344277,
+      "loss": 0.6243,
+      "step": 1933
+    },
+    {
+      "epoch": 0.61888,
+      "grad_norm": 0.461141437292099,
+      "learning_rate": 0.00013825460368294636,
+      "loss": 0.6596,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 0.49115195870399475,
+      "learning_rate": 0.00013822257806244995,
+      "loss": 0.6823,
+      "step": 1935
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 0.4093218445777893,
+      "learning_rate": 0.00013819055244195357,
+      "loss": 0.5317,
+      "step": 1936
+    },
+    {
+      "epoch": 0.61984,
+      "grad_norm": 0.38962507247924805,
+      "learning_rate": 0.0001381585268214572,
+      "loss": 0.5764,
+      "step": 1937
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 0.4571894705295563,
+      "learning_rate": 0.00013812650120096078,
+      "loss": 0.6521,
+      "step": 1938
+    },
+    {
+      "epoch": 0.62048,
+      "grad_norm": 0.501805305480957,
+      "learning_rate": 0.00013809447558046437,
+      "loss": 0.7241,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.44995835423469543,
+      "learning_rate": 0.000138062449959968,
+      "loss": 0.574,
+      "step": 1940
+    },
+    {
+      "epoch": 0.62112,
+      "grad_norm": 0.4828948378562927,
+      "learning_rate": 0.00013803042433947158,
+      "loss": 0.8389,
+      "step": 1941
+    },
+    {
+      "epoch": 0.62144,
+      "grad_norm": 0.5493443608283997,
+      "learning_rate": 0.0001379983987189752,
+      "loss": 0.6028,
+      "step": 1942
+    },
+    {
+      "epoch": 0.62176,
+      "grad_norm": 0.42676103115081787,
+      "learning_rate": 0.0001379663730984788,
+      "loss": 0.5347,
+      "step": 1943
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 0.4061322808265686,
+      "learning_rate": 0.0001379343474779824,
+      "loss": 0.5568,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 0.4080204963684082,
+      "learning_rate": 0.000137902321857486,
+      "loss": 0.5896,
+      "step": 1945
+    },
+    {
+      "epoch": 0.62272,
+      "grad_norm": 0.367042601108551,
+      "learning_rate": 0.0001378702962369896,
+      "loss": 0.5497,
+      "step": 1946
+    },
+    {
+      "epoch": 0.62304,
+      "grad_norm": 0.4456496238708496,
+      "learning_rate": 0.00013783827061649318,
+      "loss": 0.752,
+      "step": 1947
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 0.5800338387489319,
+      "learning_rate": 0.00013780624499599683,
+      "loss": 0.6542,
+      "step": 1948
+    },
+    {
+      "epoch": 0.62368,
+      "grad_norm": 0.6314401626586914,
+      "learning_rate": 0.00013777421937550042,
+      "loss": 0.6491,
+      "step": 1949
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.41931530833244324,
+      "learning_rate": 0.000137742193755004,
+      "loss": 0.5782,
+      "step": 1950
+    },
+    {
+      "epoch": 0.62432,
+      "grad_norm": 0.4826603829860687,
+      "learning_rate": 0.0001377101681345076,
+      "loss": 0.5217,
+      "step": 1951
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 0.5012388229370117,
+      "learning_rate": 0.00013767814251401122,
+      "loss": 0.5689,
+      "step": 1952
+    },
+    {
+      "epoch": 0.62496,
+      "grad_norm": 0.4634721875190735,
+      "learning_rate": 0.0001376461168935148,
+      "loss": 0.5355,
+      "step": 1953
+    },
+    {
+      "epoch": 0.62528,
+      "grad_norm": 0.4352067708969116,
+      "learning_rate": 0.00013761409127301843,
+      "loss": 0.5705,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 0.36974918842315674,
+      "learning_rate": 0.00013758206565252202,
+      "loss": 0.5215,
+      "step": 1955
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 0.4130338132381439,
+      "learning_rate": 0.00013755004003202564,
+      "loss": 0.5025,
+      "step": 1956
+    },
+    {
+      "epoch": 0.62624,
+      "grad_norm": 0.44768890738487244,
+      "learning_rate": 0.00013751801441152923,
+      "loss": 0.5562,
+      "step": 1957
+    },
+    {
+      "epoch": 0.62656,
+      "grad_norm": 0.3820764422416687,
+      "learning_rate": 0.00013748598879103282,
+      "loss": 0.6375,
+      "step": 1958
+    },
+    {
+      "epoch": 0.62688,
+      "grad_norm": 0.42848625779151917,
+      "learning_rate": 0.00013745396317053644,
+      "loss": 0.5742,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.34845873713493347,
+      "learning_rate": 0.00013742193755004006,
+      "loss": 0.5251,
+      "step": 1960
+    },
+    {
+      "epoch": 0.62752,
+      "grad_norm": 0.4441608190536499,
+      "learning_rate": 0.00013738991192954365,
+      "loss": 0.5279,
+      "step": 1961
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 0.5409641861915588,
+      "learning_rate": 0.00013735788630904724,
+      "loss": 0.6263,
+      "step": 1962
+    },
+    {
+      "epoch": 0.62816,
+      "grad_norm": 0.39309078454971313,
+      "learning_rate": 0.00013732586068855086,
+      "loss": 0.5797,
+      "step": 1963
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 0.45649170875549316,
+      "learning_rate": 0.00013729383506805445,
+      "loss": 0.6336,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 0.44312140345573425,
+      "learning_rate": 0.00013726180944755804,
+      "loss": 0.6363,
+      "step": 1965
+    },
+    {
+      "epoch": 0.62912,
+      "grad_norm": 0.5159697532653809,
+      "learning_rate": 0.00013722978382706166,
+      "loss": 0.6261,
+      "step": 1966
+    },
+    {
+      "epoch": 0.62944,
+      "grad_norm": 0.4094468951225281,
+      "learning_rate": 0.00013719775820656528,
+      "loss": 0.5067,
+      "step": 1967
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 0.4448239207267761,
+      "learning_rate": 0.00013716573258606887,
+      "loss": 0.6482,
+      "step": 1968
+    },
+    {
+      "epoch": 0.63008,
+      "grad_norm": 0.35193201899528503,
+      "learning_rate": 0.00013713370696557246,
+      "loss": 0.6143,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.38995805382728577,
+      "learning_rate": 0.00013710168134507605,
+      "loss": 0.5352,
+      "step": 1970
+    },
+    {
+      "epoch": 0.63072,
+      "grad_norm": 0.37460267543792725,
+      "learning_rate": 0.00013706965572457967,
+      "loss": 0.471,
+      "step": 1971
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 0.43503037095069885,
+      "learning_rate": 0.00013703763010408329,
+      "loss": 0.5345,
+      "step": 1972
+    },
+    {
+      "epoch": 0.63136,
+      "grad_norm": 0.39353063702583313,
+      "learning_rate": 0.00013700560448358688,
+      "loss": 0.5316,
+      "step": 1973
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 0.4130716919898987,
+      "learning_rate": 0.00013697357886309047,
+      "loss": 0.5698,
+      "step": 1974
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.46023333072662354,
+      "learning_rate": 0.0001369415532425941,
+      "loss": 0.5057,
+      "step": 1975
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 0.358201801776886,
+      "learning_rate": 0.00013690952762209768,
+      "loss": 0.5473,
+      "step": 1976
+    },
+    {
+      "epoch": 0.63264,
+      "grad_norm": 0.4322684109210968,
+      "learning_rate": 0.00013687750200160127,
+      "loss": 0.6603,
+      "step": 1977
+    },
+    {
+      "epoch": 0.63296,
+      "grad_norm": 0.46719619631767273,
+      "learning_rate": 0.0001368454763811049,
+      "loss": 0.8295,
+      "step": 1978
+    },
+    {
+      "epoch": 0.63328,
+      "grad_norm": 0.38528695702552795,
+      "learning_rate": 0.0001368134507606085,
+      "loss": 0.5276,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.48595985770225525,
+      "learning_rate": 0.0001367814251401121,
+      "loss": 0.5906,
+      "step": 1980
+    },
+    {
+      "epoch": 0.63392,
+      "grad_norm": 0.3348192870616913,
+      "learning_rate": 0.0001367493995196157,
+      "loss": 0.4536,
+      "step": 1981
+    },
+    {
+      "epoch": 0.63424,
+      "grad_norm": 0.40497639775276184,
+      "learning_rate": 0.0001367173738991193,
+      "loss": 0.5218,
+      "step": 1982
+    },
+    {
+      "epoch": 0.63456,
+      "grad_norm": 0.37011080980300903,
+      "learning_rate": 0.00013668534827862292,
+      "loss": 0.5003,
+      "step": 1983
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 0.45395344495773315,
+      "learning_rate": 0.00013665332265812652,
+      "loss": 0.6267,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 0.47634178400039673,
+      "learning_rate": 0.0001366212970376301,
+      "loss": 0.6543,
+      "step": 1985
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 0.4171687066555023,
+      "learning_rate": 0.00013658927141713373,
+      "loss": 0.6019,
+      "step": 1986
+    },
+    {
+      "epoch": 0.63584,
+      "grad_norm": 0.493117094039917,
+      "learning_rate": 0.00013655724579663732,
+      "loss": 0.7218,
+      "step": 1987
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 0.5069145560264587,
+      "learning_rate": 0.0001365252201761409,
+      "loss": 0.5728,
+      "step": 1988
+    },
+    {
+      "epoch": 0.63648,
+      "grad_norm": 0.4489979147911072,
+      "learning_rate": 0.00013649319455564453,
+      "loss": 0.6667,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.36537155508995056,
+      "learning_rate": 0.00013646116893514814,
+      "loss": 0.5084,
+      "step": 1990
+    },
+    {
+      "epoch": 0.63712,
+      "grad_norm": 0.4973263740539551,
+      "learning_rate": 0.00013642914331465174,
+      "loss": 0.5762,
+      "step": 1991
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 0.46971064805984497,
+      "learning_rate": 0.00013639711769415533,
+      "loss": 0.5815,
+      "step": 1992
+    },
+    {
+      "epoch": 0.63776,
+      "grad_norm": 0.47090035676956177,
+      "learning_rate": 0.00013636509207365892,
+      "loss": 0.6138,
+      "step": 1993
+    },
+    {
+      "epoch": 0.63808,
+      "grad_norm": 0.4619061350822449,
+      "learning_rate": 0.00013633306645316254,
+      "loss": 0.6539,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 0.4819643199443817,
+      "learning_rate": 0.00013630104083266615,
+      "loss": 0.5501,
+      "step": 1995
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 0.4908696711063385,
+      "learning_rate": 0.00013626901521216975,
+      "loss": 0.6669,
+      "step": 1996
+    },
+    {
+      "epoch": 0.63904,
+      "grad_norm": 0.3819018304347992,
+      "learning_rate": 0.00013623698959167334,
+      "loss": 0.6213,
+      "step": 1997
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 0.4665070176124573,
+      "learning_rate": 0.00013620496397117696,
+      "loss": 0.5377,
+      "step": 1998
+    },
+    {
+      "epoch": 0.63968,
+      "grad_norm": 0.5026546120643616,
+      "learning_rate": 0.00013617293835068055,
+      "loss": 0.667,
+      "step": 1999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.47140422463417053,
+      "learning_rate": 0.00013614091273018414,
+      "loss": 0.6249,
+      "step": 2000
+    },
+    {
+      "epoch": 0.64032,
+      "grad_norm": 0.4278138279914856,
+      "learning_rate": 0.00013610888710968776,
+      "loss": 0.6239,
+      "step": 2001
+    },
+    {
+      "epoch": 0.64064,
+      "grad_norm": 0.4169631898403168,
+      "learning_rate": 0.00013607686148919137,
+      "loss": 0.5181,
+      "step": 2002
+    },
+    {
+      "epoch": 0.64096,
+      "grad_norm": 0.4102838337421417,
+      "learning_rate": 0.00013604483586869497,
+      "loss": 0.508,
+      "step": 2003
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 0.4587903916835785,
+      "learning_rate": 0.00013601281024819856,
+      "loss": 0.5213,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 0.46816644072532654,
+      "learning_rate": 0.00013598078462770218,
+      "loss": 0.6569,
+      "step": 2005
+    },
+    {
+      "epoch": 0.64192,
+      "grad_norm": 0.5139099359512329,
+      "learning_rate": 0.00013594875900720577,
+      "loss": 0.5641,
+      "step": 2006
+    },
+    {
+      "epoch": 0.64224,
+      "grad_norm": 0.4055844247341156,
+      "learning_rate": 0.00013591673338670938,
+      "loss": 0.5558,
+      "step": 2007
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 0.4605430066585541,
+      "learning_rate": 0.00013588470776621298,
+      "loss": 0.6305,
+      "step": 2008
+    },
+    {
+      "epoch": 0.64288,
+      "grad_norm": 0.4243757426738739,
+      "learning_rate": 0.0001358526821457166,
+      "loss": 0.4721,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.4178702235221863,
+      "learning_rate": 0.00013582065652522019,
+      "loss": 0.5944,
+      "step": 2010
+    },
+    {
+      "epoch": 0.64352,
+      "grad_norm": 0.36094361543655396,
+      "learning_rate": 0.00013578863090472378,
+      "loss": 0.5223,
+      "step": 2011
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 0.47107842564582825,
+      "learning_rate": 0.00013575660528422737,
+      "loss": 0.6521,
+      "step": 2012
+    },
+    {
+      "epoch": 0.64416,
+      "grad_norm": 0.4957985579967499,
+      "learning_rate": 0.00013572457966373099,
+      "loss": 0.7259,
+      "step": 2013
+    },
+    {
+      "epoch": 0.64448,
+      "grad_norm": 0.38843995332717896,
+      "learning_rate": 0.0001356925540432346,
+      "loss": 0.5406,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 0.46218302845954895,
+      "learning_rate": 0.0001356605284227382,
+      "loss": 0.6849,
+      "step": 2015
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 0.4211007058620453,
+      "learning_rate": 0.0001356285028022418,
+      "loss": 0.528,
+      "step": 2016
+    },
+    {
+      "epoch": 0.64544,
+      "grad_norm": 0.5415166616439819,
+      "learning_rate": 0.0001355964771817454,
+      "loss": 0.5529,
+      "step": 2017
+    },
+    {
+      "epoch": 0.64576,
+      "grad_norm": 0.4764891266822815,
+      "learning_rate": 0.000135564451561249,
+      "loss": 0.682,
+      "step": 2018
+    },
+    {
+      "epoch": 0.64608,
+      "grad_norm": 0.46371060609817505,
+      "learning_rate": 0.00013553242594075261,
+      "loss": 0.7513,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.4075927436351776,
+      "learning_rate": 0.0001355004003202562,
+      "loss": 0.6052,
+      "step": 2020
+    },
+    {
+      "epoch": 0.64672,
+      "grad_norm": 0.41813796758651733,
+      "learning_rate": 0.00013546837469975982,
+      "loss": 0.4989,
+      "step": 2021
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 0.4371137022972107,
+      "learning_rate": 0.00013543634907926342,
+      "loss": 0.6443,
+      "step": 2022
+    },
+    {
+      "epoch": 0.64736,
+      "grad_norm": 0.4755712151527405,
+      "learning_rate": 0.000135404323458767,
+      "loss": 0.6523,
+      "step": 2023
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 0.4532909691333771,
+      "learning_rate": 0.00013537229783827063,
+      "loss": 0.7083,
+      "step": 2024
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.40208759903907776,
+      "learning_rate": 0.00013534027221777424,
+      "loss": 0.7155,
+      "step": 2025
+    },
+    {
+      "epoch": 0.64832,
+      "grad_norm": 0.42676377296447754,
+      "learning_rate": 0.00013530824659727783,
+      "loss": 0.644,
+      "step": 2026
+    },
+    {
+      "epoch": 0.64864,
+      "grad_norm": 0.4834870994091034,
+      "learning_rate": 0.00013527622097678143,
+      "loss": 0.7037,
+      "step": 2027
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 0.4260694682598114,
+      "learning_rate": 0.00013524419535628502,
+      "loss": 0.4785,
+      "step": 2028
+    },
+    {
+      "epoch": 0.64928,
+      "grad_norm": 0.4846153259277344,
+      "learning_rate": 0.00013521216973578864,
+      "loss": 0.6848,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.5189780592918396,
+      "learning_rate": 0.00013518014411529225,
+      "loss": 0.5878,
+      "step": 2030
+    },
+    {
+      "epoch": 0.64992,
+      "grad_norm": 0.33686089515686035,
+      "learning_rate": 0.00013514811849479584,
+      "loss": 0.5036,
+      "step": 2031
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 0.5845642685890198,
+      "learning_rate": 0.00013511609287429944,
+      "loss": 0.7478,
+      "step": 2032
+    },
+    {
+      "epoch": 0.65056,
+      "grad_norm": 0.37045320868492126,
+      "learning_rate": 0.00013508406725380305,
+      "loss": 0.6176,
+      "step": 2033
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 0.5263961553573608,
+      "learning_rate": 0.00013505204163330665,
+      "loss": 0.6399,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 0.42873743176460266,
+      "learning_rate": 0.00013502001601281024,
+      "loss": 0.5256,
+      "step": 2035
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 0.42029479146003723,
+      "learning_rate": 0.00013498799039231386,
+      "loss": 0.5799,
+      "step": 2036
+    },
+    {
+      "epoch": 0.65184,
+      "grad_norm": 0.5360143184661865,
+      "learning_rate": 0.00013495596477181747,
+      "loss": 0.7143,
+      "step": 2037
+    },
+    {
+      "epoch": 0.65216,
+      "grad_norm": 0.48051661252975464,
+      "learning_rate": 0.00013492393915132106,
+      "loss": 0.6248,
+      "step": 2038
+    },
+    {
+      "epoch": 0.65248,
+      "grad_norm": 0.49067410826683044,
+      "learning_rate": 0.00013489191353082466,
+      "loss": 0.7479,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.569767415523529,
+      "learning_rate": 0.00013485988791032827,
+      "loss": 0.7349,
+      "step": 2040
+    },
+    {
+      "epoch": 0.65312,
+      "grad_norm": 0.4526996314525604,
+      "learning_rate": 0.00013482786228983187,
+      "loss": 0.6128,
+      "step": 2041
+    },
+    {
+      "epoch": 0.65344,
+      "grad_norm": 0.443215936422348,
+      "learning_rate": 0.00013479583666933548,
+      "loss": 0.6012,
+      "step": 2042
+    },
+    {
+      "epoch": 0.65376,
+      "grad_norm": 0.47472214698791504,
+      "learning_rate": 0.00013476381104883907,
+      "loss": 0.7137,
+      "step": 2043
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 0.41327640414237976,
+      "learning_rate": 0.0001347317854283427,
+      "loss": 0.7339,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 0.37131467461586,
+      "learning_rate": 0.00013469975980784628,
+      "loss": 0.5056,
+      "step": 2045
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 0.3699110746383667,
+      "learning_rate": 0.00013466773418734988,
+      "loss": 0.553,
+      "step": 2046
+    },
+    {
+      "epoch": 0.65504,
+      "grad_norm": 0.4516889154911041,
+      "learning_rate": 0.00013463570856685347,
+      "loss": 0.6718,
+      "step": 2047
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 0.5298005938529968,
+      "learning_rate": 0.0001346036829463571,
+      "loss": 0.7289,
+      "step": 2048
+    },
+    {
+      "epoch": 0.65568,
+      "grad_norm": 0.3994103968143463,
+      "learning_rate": 0.0001345716573258607,
+      "loss": 0.5145,
+      "step": 2049
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.44479718804359436,
+      "learning_rate": 0.0001345396317053643,
+      "loss": 0.6049,
+      "step": 2050
+    },
+    {
+      "epoch": 0.65632,
+      "grad_norm": 0.4166565239429474,
+      "learning_rate": 0.00013450760608486789,
+      "loss": 0.5931,
+      "step": 2051
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 0.5187183022499084,
+      "learning_rate": 0.0001344755804643715,
+      "loss": 0.7235,
+      "step": 2052
+    },
+    {
+      "epoch": 0.65696,
+      "grad_norm": 0.47600769996643066,
+      "learning_rate": 0.0001344435548438751,
+      "loss": 0.6008,
+      "step": 2053
+    },
+    {
+      "epoch": 0.65728,
+      "grad_norm": 0.45247966051101685,
+      "learning_rate": 0.00013441152922337871,
+      "loss": 0.5581,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 0.42817094922065735,
+      "learning_rate": 0.0001343795036028823,
+      "loss": 0.6342,
+      "step": 2055
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 0.4378697872161865,
+      "learning_rate": 0.00013434747798238592,
+      "loss": 0.567,
+      "step": 2056
+    },
+    {
+      "epoch": 0.65824,
+      "grad_norm": 0.4087863862514496,
+      "learning_rate": 0.00013431545236188951,
+      "loss": 0.7097,
+      "step": 2057
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 0.4131960868835449,
+      "learning_rate": 0.0001342834267413931,
+      "loss": 0.5672,
+      "step": 2058
+    },
+    {
+      "epoch": 0.65888,
+      "grad_norm": 0.47496601939201355,
+      "learning_rate": 0.00013425140112089672,
+      "loss": 0.5201,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.44073769450187683,
+      "learning_rate": 0.00013421937550040034,
+      "loss": 0.4826,
+      "step": 2060
+    },
+    {
+      "epoch": 0.65952,
+      "grad_norm": 0.3994128704071045,
+      "learning_rate": 0.00013418734987990393,
+      "loss": 0.5405,
+      "step": 2061
+    },
+    {
+      "epoch": 0.65984,
+      "grad_norm": 0.3527107536792755,
+      "learning_rate": 0.00013415532425940752,
+      "loss": 0.6493,
+      "step": 2062
+    },
+    {
+      "epoch": 0.66016,
+      "grad_norm": 0.4626217484474182,
+      "learning_rate": 0.00013412329863891114,
+      "loss": 0.5437,
+      "step": 2063
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 0.44516685605049133,
+      "learning_rate": 0.00013409127301841473,
+      "loss": 0.6695,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 0.5258999466896057,
+      "learning_rate": 0.00013405924739791833,
+      "loss": 0.6017,
+      "step": 2065
+    },
+    {
+      "epoch": 0.66112,
+      "grad_norm": 0.48528438806533813,
+      "learning_rate": 0.00013402722177742194,
+      "loss": 0.5634,
+      "step": 2066
+    },
+    {
+      "epoch": 0.66144,
+      "grad_norm": 0.4307623505592346,
+      "learning_rate": 0.00013399519615692556,
+      "loss": 0.5285,
+      "step": 2067
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 0.49954140186309814,
+      "learning_rate": 0.00013396317053642915,
+      "loss": 0.7143,
+      "step": 2068
+    },
+    {
+      "epoch": 0.66208,
+      "grad_norm": 0.47828200459480286,
+      "learning_rate": 0.00013393114491593274,
+      "loss": 0.5504,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.3996775150299072,
+      "learning_rate": 0.00013389911929543634,
+      "loss": 0.5988,
+      "step": 2070
+    },
+    {
+      "epoch": 0.66272,
+      "grad_norm": 0.3089894652366638,
+      "learning_rate": 0.00013386709367493995,
+      "loss": 0.486,
+      "step": 2071
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 0.4235036373138428,
+      "learning_rate": 0.00013383506805444357,
+      "loss": 0.5573,
+      "step": 2072
+    },
+    {
+      "epoch": 0.66336,
+      "grad_norm": 0.4191177487373352,
+      "learning_rate": 0.00013380304243394716,
+      "loss": 0.6523,
+      "step": 2073
+    },
+    {
+      "epoch": 0.66368,
+      "grad_norm": 0.40078067779541016,
+      "learning_rate": 0.00013377101681345075,
+      "loss": 0.6045,
+      "step": 2074
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.4212968349456787,
+      "learning_rate": 0.00013373899119295437,
+      "loss": 0.5681,
+      "step": 2075
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 0.37732791900634766,
+      "learning_rate": 0.00013370696557245796,
+      "loss": 0.5013,
+      "step": 2076
+    },
+    {
+      "epoch": 0.66464,
+      "grad_norm": 0.3814069330692291,
+      "learning_rate": 0.00013367493995196158,
+      "loss": 0.6039,
+      "step": 2077
+    },
+    {
+      "epoch": 0.66496,
+      "grad_norm": 0.3444000780582428,
+      "learning_rate": 0.00013364291433146517,
+      "loss": 0.488,
+      "step": 2078
+    },
+    {
+      "epoch": 0.66528,
+      "grad_norm": 0.4447037875652313,
+      "learning_rate": 0.0001336108887109688,
+      "loss": 0.5427,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.383209764957428,
+      "learning_rate": 0.00013357886309047238,
+      "loss": 0.4995,
+      "step": 2080
+    },
+    {
+      "epoch": 0.66592,
+      "grad_norm": 0.43265458941459656,
+      "learning_rate": 0.00013354683746997597,
+      "loss": 0.5623,
+      "step": 2081
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 0.419155478477478,
+      "learning_rate": 0.0001335148118494796,
+      "loss": 0.563,
+      "step": 2082
+    },
+    {
+      "epoch": 0.66656,
+      "grad_norm": 0.4599671959877014,
+      "learning_rate": 0.0001334827862289832,
+      "loss": 0.5837,
+      "step": 2083
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 0.39344948530197144,
+      "learning_rate": 0.0001334507606084868,
+      "loss": 0.5979,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 0.3848174810409546,
+      "learning_rate": 0.0001334187349879904,
+      "loss": 0.6183,
+      "step": 2085
+    },
+    {
+      "epoch": 0.66752,
+      "grad_norm": 0.4039468467235565,
+      "learning_rate": 0.000133386709367494,
+      "loss": 0.6594,
+      "step": 2086
+    },
+    {
+      "epoch": 0.66784,
+      "grad_norm": 0.43314793705940247,
+      "learning_rate": 0.0001333546837469976,
+      "loss": 0.6327,
+      "step": 2087
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 0.4213702380657196,
+      "learning_rate": 0.0001333226581265012,
+      "loss": 0.4516,
+      "step": 2088
+    },
+    {
+      "epoch": 0.66848,
+      "grad_norm": 0.44839322566986084,
+      "learning_rate": 0.0001332906325060048,
+      "loss": 0.5997,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.4965057075023651,
+      "learning_rate": 0.00013325860688550843,
+      "loss": 0.5636,
+      "step": 2090
+    },
+    {
+      "epoch": 0.66912,
+      "grad_norm": 0.36896273493766785,
+      "learning_rate": 0.00013322658126501202,
+      "loss": 0.5004,
+      "step": 2091
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 0.40454965829849243,
+      "learning_rate": 0.0001331945556445156,
+      "loss": 0.5609,
+      "step": 2092
+    },
+    {
+      "epoch": 0.66976,
+      "grad_norm": 0.4581792950630188,
+      "learning_rate": 0.0001331625300240192,
+      "loss": 0.593,
+      "step": 2093
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 0.55596923828125,
+      "learning_rate": 0.00013313050440352282,
+      "loss": 0.6183,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 0.4803834855556488,
+      "learning_rate": 0.00013309847878302644,
+      "loss": 0.5612,
+      "step": 2095
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 0.4868222773075104,
+      "learning_rate": 0.00013306645316253003,
+      "loss": 0.5962,
+      "step": 2096
+    },
+    {
+      "epoch": 0.67104,
+      "grad_norm": 0.4919818043708801,
+      "learning_rate": 0.00013303442754203362,
+      "loss": 0.6326,
+      "step": 2097
+    },
+    {
+      "epoch": 0.67136,
+      "grad_norm": 0.3792168200016022,
+      "learning_rate": 0.00013300240192153724,
+      "loss": 0.534,
+      "step": 2098
+    },
+    {
+      "epoch": 0.67168,
+      "grad_norm": 0.4723796546459198,
+      "learning_rate": 0.00013297037630104083,
+      "loss": 0.5001,
+      "step": 2099
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.4835221469402313,
+      "learning_rate": 0.00013293835068054442,
+      "loss": 0.4951,
+      "step": 2100
+    },
+    {
+      "epoch": 0.67232,
+      "grad_norm": 0.48712873458862305,
+      "learning_rate": 0.00013290632506004804,
+      "loss": 0.6123,
+      "step": 2101
+    },
+    {
+      "epoch": 0.67264,
+      "grad_norm": 0.4781598746776581,
+      "learning_rate": 0.00013287429943955166,
+      "loss": 0.6884,
+      "step": 2102
+    },
+    {
+      "epoch": 0.67296,
+      "grad_norm": 0.4096449315547943,
+      "learning_rate": 0.00013284227381905525,
+      "loss": 0.4898,
+      "step": 2103
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 0.3965255320072174,
+      "learning_rate": 0.00013281024819855884,
+      "loss": 0.442,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 0.49428337812423706,
+      "learning_rate": 0.00013277822257806246,
+      "loss": 0.6321,
+      "step": 2105
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 0.38125157356262207,
+      "learning_rate": 0.00013274619695756605,
+      "loss": 0.5001,
+      "step": 2106
+    },
+    {
+      "epoch": 0.67424,
+      "grad_norm": 0.36895567178726196,
+      "learning_rate": 0.00013271417133706967,
+      "loss": 0.5806,
+      "step": 2107
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 0.5101200938224792,
+      "learning_rate": 0.00013268214571657326,
+      "loss": 0.6135,
+      "step": 2108
+    },
+    {
+      "epoch": 0.67488,
+      "grad_norm": 0.3829675018787384,
+      "learning_rate": 0.00013265012009607688,
+      "loss": 0.5496,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.38745394349098206,
+      "learning_rate": 0.00013261809447558047,
+      "loss": 0.5231,
+      "step": 2110
+    },
+    {
+      "epoch": 0.67552,
+      "grad_norm": 0.39364826679229736,
+      "learning_rate": 0.00013258606885508406,
+      "loss": 0.5077,
+      "step": 2111
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 0.4078015089035034,
+      "learning_rate": 0.00013255404323458765,
+      "loss": 0.4572,
+      "step": 2112
+    },
+    {
+      "epoch": 0.67616,
+      "grad_norm": 0.32650649547576904,
+      "learning_rate": 0.0001325220176140913,
+      "loss": 0.5324,
+      "step": 2113
+    },
+    {
+      "epoch": 0.67648,
+      "grad_norm": 0.5626618266105652,
+      "learning_rate": 0.0001324899919935949,
+      "loss": 0.7078,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 0.39899659156799316,
+      "learning_rate": 0.00013245796637309848,
+      "loss": 0.4162,
+      "step": 2115
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 0.48526933789253235,
+      "learning_rate": 0.00013242594075260207,
+      "loss": 0.7697,
+      "step": 2116
+    },
+    {
+      "epoch": 0.67744,
+      "grad_norm": 0.4414370357990265,
+      "learning_rate": 0.0001323939151321057,
+      "loss": 0.6168,
+      "step": 2117
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 0.4505459666252136,
+      "learning_rate": 0.00013236188951160928,
+      "loss": 0.5978,
+      "step": 2118
+    },
+    {
+      "epoch": 0.67808,
+      "grad_norm": 0.494458943605423,
+      "learning_rate": 0.0001323298638911129,
+      "loss": 0.5266,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.45902279019355774,
+      "learning_rate": 0.0001322978382706165,
+      "loss": 0.5971,
+      "step": 2120
+    },
+    {
+      "epoch": 0.67872,
+      "grad_norm": 0.4916379451751709,
+      "learning_rate": 0.0001322658126501201,
+      "loss": 0.6162,
+      "step": 2121
+    },
+    {
+      "epoch": 0.67904,
+      "grad_norm": 0.4879927933216095,
+      "learning_rate": 0.0001322337870296237,
+      "loss": 0.685,
+      "step": 2122
+    },
+    {
+      "epoch": 0.67936,
+      "grad_norm": 0.48512014746665955,
+      "learning_rate": 0.0001322017614091273,
+      "loss": 0.6766,
+      "step": 2123
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 0.3933636248111725,
+      "learning_rate": 0.0001321697357886309,
+      "loss": 0.5613,
+      "step": 2124
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4161766767501831,
+      "learning_rate": 0.00013213771016813453,
+      "loss": 0.497,
+      "step": 2125
+    },
+    {
+      "epoch": 0.68032,
+      "grad_norm": 0.6172170639038086,
+      "learning_rate": 0.00013210568454763812,
+      "loss": 0.6769,
+      "step": 2126
+    },
+    {
+      "epoch": 0.68064,
+      "grad_norm": 0.496931791305542,
+      "learning_rate": 0.0001320736589271417,
+      "loss": 0.5587,
+      "step": 2127
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 0.4197312593460083,
+      "learning_rate": 0.00013204163330664533,
+      "loss": 0.6871,
+      "step": 2128
+    },
+    {
+      "epoch": 0.68128,
+      "grad_norm": 0.4256606101989746,
+      "learning_rate": 0.00013200960768614892,
+      "loss": 0.6278,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.4077456593513489,
+      "learning_rate": 0.00013197758206565254,
+      "loss": 0.5613,
+      "step": 2130
+    },
+    {
+      "epoch": 0.68192,
+      "grad_norm": 0.46914759278297424,
+      "learning_rate": 0.00013194555644515613,
+      "loss": 0.4838,
+      "step": 2131
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 0.4584275186061859,
+      "learning_rate": 0.00013191353082465975,
+      "loss": 0.4425,
+      "step": 2132
+    },
+    {
+      "epoch": 0.68256,
+      "grad_norm": 0.4782066345214844,
+      "learning_rate": 0.00013188150520416334,
+      "loss": 0.6681,
+      "step": 2133
+    },
+    {
+      "epoch": 0.68288,
+      "grad_norm": 0.47762930393218994,
+      "learning_rate": 0.00013184947958366693,
+      "loss": 0.6537,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 0.5581920742988586,
+      "learning_rate": 0.00013181745396317052,
+      "loss": 0.6924,
+      "step": 2135
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 0.41233792901039124,
+      "learning_rate": 0.00013178542834267417,
+      "loss": 0.5665,
+      "step": 2136
+    },
+    {
+      "epoch": 0.68384,
+      "grad_norm": 0.47497913241386414,
+      "learning_rate": 0.00013175340272217776,
+      "loss": 0.5329,
+      "step": 2137
+    },
+    {
+      "epoch": 0.68416,
+      "grad_norm": 0.4990917444229126,
+      "learning_rate": 0.00013172137710168135,
+      "loss": 0.684,
+      "step": 2138
+    },
+    {
+      "epoch": 0.68448,
+      "grad_norm": 0.44375693798065186,
+      "learning_rate": 0.00013168935148118494,
+      "loss": 0.5926,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.4305485188961029,
+      "learning_rate": 0.00013165732586068856,
+      "loss": 0.4469,
+      "step": 2140
+    },
+    {
+      "epoch": 0.68512,
+      "grad_norm": 0.4515390396118164,
+      "learning_rate": 0.00013162530024019215,
+      "loss": 0.5506,
+      "step": 2141
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 0.4723156988620758,
+      "learning_rate": 0.00013159327461969577,
+      "loss": 0.3566,
+      "step": 2142
+    },
+    {
+      "epoch": 0.68576,
+      "grad_norm": 0.49352312088012695,
+      "learning_rate": 0.00013156124899919936,
+      "loss": 0.5422,
+      "step": 2143
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 0.5811710953712463,
+      "learning_rate": 0.00013152922337870298,
+      "loss": 0.5522,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 0.5452113151550293,
+      "learning_rate": 0.00013149719775820657,
+      "loss": 0.6708,
+      "step": 2145
+    },
+    {
+      "epoch": 0.68672,
+      "grad_norm": 0.4607369601726532,
+      "learning_rate": 0.00013146517213771016,
+      "loss": 0.593,
+      "step": 2146
+    },
+    {
+      "epoch": 0.68704,
+      "grad_norm": 0.5004740357398987,
+      "learning_rate": 0.00013143314651721378,
+      "loss": 0.5914,
+      "step": 2147
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 0.5408174395561218,
+      "learning_rate": 0.0001314011208967174,
+      "loss": 0.7383,
+      "step": 2148
+    },
+    {
+      "epoch": 0.68768,
+      "grad_norm": 0.5353201031684875,
+      "learning_rate": 0.000131369095276221,
+      "loss": 0.6402,
+      "step": 2149
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.5881289839744568,
+      "learning_rate": 0.00013133706965572458,
+      "loss": 0.7408,
+      "step": 2150
+    },
+    {
+      "epoch": 0.68832,
+      "grad_norm": 0.42560943961143494,
+      "learning_rate": 0.0001313050440352282,
+      "loss": 0.5232,
+      "step": 2151
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 0.49179524183273315,
+      "learning_rate": 0.0001312730184147318,
+      "loss": 0.6635,
+      "step": 2152
+    },
+    {
+      "epoch": 0.68896,
+      "grad_norm": 0.4263574481010437,
+      "learning_rate": 0.00013124099279423538,
+      "loss": 0.486,
+      "step": 2153
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 0.4687128961086273,
+      "learning_rate": 0.000131208967173739,
+      "loss": 0.5544,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 0.5224636793136597,
+      "learning_rate": 0.0001311769415532426,
+      "loss": 0.5432,
+      "step": 2155
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 0.5016278624534607,
+      "learning_rate": 0.0001311449159327462,
+      "loss": 0.601,
+      "step": 2156
+    },
+    {
+      "epoch": 0.69024,
+      "grad_norm": 0.42518895864486694,
+      "learning_rate": 0.0001311128903122498,
+      "loss": 0.5623,
+      "step": 2157
+    },
+    {
+      "epoch": 0.69056,
+      "grad_norm": 0.4593464434146881,
+      "learning_rate": 0.0001310808646917534,
+      "loss": 0.5566,
+      "step": 2158
+    },
+    {
+      "epoch": 0.69088,
+      "grad_norm": 0.40130847692489624,
+      "learning_rate": 0.000131048839071257,
+      "loss": 0.5271,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.4974726438522339,
+      "learning_rate": 0.00013101681345076063,
+      "loss": 0.6749,
+      "step": 2160
+    },
+    {
+      "epoch": 0.69152,
+      "grad_norm": 0.4395165741443634,
+      "learning_rate": 0.00013098478783026422,
+      "loss": 0.5542,
+      "step": 2161
+    },
+    {
+      "epoch": 0.69184,
+      "grad_norm": 0.504546046257019,
+      "learning_rate": 0.0001309527622097678,
+      "loss": 0.5711,
+      "step": 2162
+    },
+    {
+      "epoch": 0.69216,
+      "grad_norm": 0.4413326680660248,
+      "learning_rate": 0.00013092073658927143,
+      "loss": 0.5558,
+      "step": 2163
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 0.43103495240211487,
+      "learning_rate": 0.00013088871096877502,
+      "loss": 0.4192,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 0.5270280838012695,
+      "learning_rate": 0.0001308566853482786,
+      "loss": 0.6393,
+      "step": 2165
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 0.49174144864082336,
+      "learning_rate": 0.00013082465972778223,
+      "loss": 0.5841,
+      "step": 2166
+    },
+    {
+      "epoch": 0.69344,
+      "grad_norm": 0.4798937141895294,
+      "learning_rate": 0.00013079263410728585,
+      "loss": 0.5719,
+      "step": 2167
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 0.3771500289440155,
+      "learning_rate": 0.00013076060848678944,
+      "loss": 0.5715,
+      "step": 2168
+    },
+    {
+      "epoch": 0.69408,
+      "grad_norm": 0.46697717905044556,
+      "learning_rate": 0.00013072858286629303,
+      "loss": 0.5457,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.4811326861381531,
+      "learning_rate": 0.00013069655724579662,
+      "loss": 0.7369,
+      "step": 2170
+    },
+    {
+      "epoch": 0.69472,
+      "grad_norm": 0.3164181411266327,
+      "learning_rate": 0.00013066453162530027,
+      "loss": 0.5438,
+      "step": 2171
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 0.44039198756217957,
+      "learning_rate": 0.00013063250600480386,
+      "loss": 0.6366,
+      "step": 2172
+    },
+    {
+      "epoch": 0.69536,
+      "grad_norm": 0.42495593428611755,
+      "learning_rate": 0.00013060048038430745,
+      "loss": 0.6071,
+      "step": 2173
+    },
+    {
+      "epoch": 0.69568,
+      "grad_norm": 0.34734535217285156,
+      "learning_rate": 0.00013056845476381104,
+      "loss": 0.5126,
+      "step": 2174
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.3717395067214966,
+      "learning_rate": 0.00013053642914331466,
+      "loss": 0.5074,
+      "step": 2175
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 0.41764217615127563,
+      "learning_rate": 0.00013050440352281825,
+      "loss": 0.5658,
+      "step": 2176
+    },
+    {
+      "epoch": 0.69664,
+      "grad_norm": 0.4401426613330841,
+      "learning_rate": 0.00013047237790232187,
+      "loss": 0.5032,
+      "step": 2177
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 0.566840648651123,
+      "learning_rate": 0.00013044035228182546,
+      "loss": 0.698,
+      "step": 2178
+    },
+    {
+      "epoch": 0.69728,
+      "grad_norm": 0.5235535502433777,
+      "learning_rate": 0.00013040832666132908,
+      "loss": 0.7232,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.32195767760276794,
+      "learning_rate": 0.00013037630104083267,
+      "loss": 0.6214,
+      "step": 2180
+    },
+    {
+      "epoch": 0.69792,
+      "grad_norm": 0.4797719717025757,
+      "learning_rate": 0.00013034427542033626,
+      "loss": 0.6709,
+      "step": 2181
+    },
+    {
+      "epoch": 0.69824,
+      "grad_norm": 0.41244733333587646,
+      "learning_rate": 0.00013031224979983988,
+      "loss": 0.4862,
+      "step": 2182
+    },
+    {
+      "epoch": 0.69856,
+      "grad_norm": 0.5340002179145813,
+      "learning_rate": 0.0001302802241793435,
+      "loss": 0.618,
+      "step": 2183
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 0.5120226740837097,
+      "learning_rate": 0.0001302481985588471,
+      "loss": 0.7253,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 0.4126931130886078,
+      "learning_rate": 0.00013021617293835068,
+      "loss": 0.6729,
+      "step": 2185
+    },
+    {
+      "epoch": 0.69952,
+      "grad_norm": 0.36855456233024597,
+      "learning_rate": 0.0001301841473178543,
+      "loss": 0.6179,
+      "step": 2186
+    },
+    {
+      "epoch": 0.69984,
+      "grad_norm": 0.6422128677368164,
+      "learning_rate": 0.0001301521216973579,
+      "loss": 0.7441,
+      "step": 2187
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 0.38662171363830566,
+      "learning_rate": 0.00013012009607686148,
+      "loss": 0.4692,
+      "step": 2188
+    },
+    {
+      "epoch": 0.70048,
+      "grad_norm": 0.41101574897766113,
+      "learning_rate": 0.0001300880704563651,
+      "loss": 0.5615,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.40024513006210327,
+      "learning_rate": 0.00013005604483586872,
+      "loss": 0.5658,
+      "step": 2190
+    },
+    {
+      "epoch": 0.70112,
+      "grad_norm": 0.4867980182170868,
+      "learning_rate": 0.0001300240192153723,
+      "loss": 0.7125,
+      "step": 2191
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 0.3838973045349121,
+      "learning_rate": 0.0001299919935948759,
+      "loss": 0.5894,
+      "step": 2192
+    },
+    {
+      "epoch": 0.70176,
+      "grad_norm": 0.40379127860069275,
+      "learning_rate": 0.0001299599679743795,
+      "loss": 0.4529,
+      "step": 2193
+    },
+    {
+      "epoch": 0.70208,
+      "grad_norm": 0.47452229261398315,
+      "learning_rate": 0.0001299279423538831,
+      "loss": 0.6784,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 0.46376460790634155,
+      "learning_rate": 0.00012989591673338673,
+      "loss": 0.5866,
+      "step": 2195
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 0.5061944127082825,
+      "learning_rate": 0.00012986389111289032,
+      "loss": 0.6281,
+      "step": 2196
+    },
+    {
+      "epoch": 0.70304,
+      "grad_norm": 0.4941713511943817,
+      "learning_rate": 0.0001298318654923939,
+      "loss": 0.7152,
+      "step": 2197
+    },
+    {
+      "epoch": 0.70336,
+      "grad_norm": 0.44632747769355774,
+      "learning_rate": 0.00012979983987189753,
+      "loss": 0.5463,
+      "step": 2198
+    },
+    {
+      "epoch": 0.70368,
+      "grad_norm": 0.5706636905670166,
+      "learning_rate": 0.00012976781425140112,
+      "loss": 0.5922,
+      "step": 2199
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.4425111711025238,
+      "learning_rate": 0.0001297357886309047,
+      "loss": 0.5548,
+      "step": 2200
+    },
+    {
+      "epoch": 0.70432,
+      "grad_norm": 0.4452555179595947,
+      "learning_rate": 0.00012970376301040833,
+      "loss": 0.5907,
+      "step": 2201
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 0.41666021943092346,
+      "learning_rate": 0.00012967173738991195,
+      "loss": 0.6158,
+      "step": 2202
+    },
+    {
+      "epoch": 0.70496,
+      "grad_norm": 0.5296522378921509,
+      "learning_rate": 0.00012963971176941554,
+      "loss": 0.6615,
+      "step": 2203
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 0.5152687430381775,
+      "learning_rate": 0.00012960768614891913,
+      "loss": 0.5819,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 0.4767840504646301,
+      "learning_rate": 0.00012957566052842275,
+      "loss": 0.5337,
+      "step": 2205
+    },
+    {
+      "epoch": 0.70592,
+      "grad_norm": 0.3996241092681885,
+      "learning_rate": 0.00012954363490792634,
+      "loss": 0.4848,
+      "step": 2206
+    },
+    {
+      "epoch": 0.70624,
+      "grad_norm": 0.48473528027534485,
+      "learning_rate": 0.00012951160928742996,
+      "loss": 0.7208,
+      "step": 2207
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 0.4734954535961151,
+      "learning_rate": 0.00012947958366693355,
+      "loss": 0.4969,
+      "step": 2208
+    },
+    {
+      "epoch": 0.70688,
+      "grad_norm": 0.4933586120605469,
+      "learning_rate": 0.00012944755804643717,
+      "loss": 0.5894,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.5308293104171753,
+      "learning_rate": 0.00012941553242594076,
+      "loss": 0.6406,
+      "step": 2210
+    },
+    {
+      "epoch": 0.70752,
+      "grad_norm": 0.4713907539844513,
+      "learning_rate": 0.00012938350680544435,
+      "loss": 0.6174,
+      "step": 2211
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 0.3916599154472351,
+      "learning_rate": 0.00012935148118494794,
+      "loss": 0.5551,
+      "step": 2212
+    },
+    {
+      "epoch": 0.70816,
+      "grad_norm": 0.45044389367103577,
+      "learning_rate": 0.00012931945556445159,
+      "loss": 0.5913,
+      "step": 2213
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 0.5091454982757568,
+      "learning_rate": 0.00012928742994395518,
+      "loss": 0.6163,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 0.4751040041446686,
+      "learning_rate": 0.00012925540432345877,
+      "loss": 0.6466,
+      "step": 2215
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 0.5128551721572876,
+      "learning_rate": 0.00012922337870296236,
+      "loss": 0.6955,
+      "step": 2216
+    },
+    {
+      "epoch": 0.70944,
+      "grad_norm": 0.4499015212059021,
+      "learning_rate": 0.00012919135308246598,
+      "loss": 0.5825,
+      "step": 2217
+    },
+    {
+      "epoch": 0.70976,
+      "grad_norm": 0.42009058594703674,
+      "learning_rate": 0.0001291593274619696,
+      "loss": 0.5633,
+      "step": 2218
+    },
+    {
+      "epoch": 0.71008,
+      "grad_norm": 0.40272000432014465,
+      "learning_rate": 0.0001291273018414732,
+      "loss": 0.4687,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.4491935968399048,
+      "learning_rate": 0.00012909527622097678,
+      "loss": 0.6427,
+      "step": 2220
+    },
+    {
+      "epoch": 0.71072,
+      "grad_norm": 0.4429425001144409,
+      "learning_rate": 0.0001290632506004804,
+      "loss": 0.507,
+      "step": 2221
+    },
+    {
+      "epoch": 0.71104,
+      "grad_norm": 0.550117552280426,
+      "learning_rate": 0.000129031224979984,
+      "loss": 0.6188,
+      "step": 2222
+    },
+    {
+      "epoch": 0.71136,
+      "grad_norm": 0.47868460416793823,
+      "learning_rate": 0.00012899919935948758,
+      "loss": 0.5682,
+      "step": 2223
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 0.4514338970184326,
+      "learning_rate": 0.0001289671737389912,
+      "loss": 0.5159,
+      "step": 2224
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.44453972578048706,
+      "learning_rate": 0.00012893514811849482,
+      "loss": 0.6853,
+      "step": 2225
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 0.43636614084243774,
+      "learning_rate": 0.0001289031224979984,
+      "loss": 0.4578,
+      "step": 2226
+    },
+    {
+      "epoch": 0.71264,
+      "grad_norm": 0.5106246471405029,
+      "learning_rate": 0.000128871096877502,
+      "loss": 0.6859,
+      "step": 2227
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 0.352495938539505,
+      "learning_rate": 0.00012883907125700562,
+      "loss": 0.5239,
+      "step": 2228
+    },
+    {
+      "epoch": 0.71328,
+      "grad_norm": 0.5328288078308105,
+      "learning_rate": 0.0001288070456365092,
+      "loss": 0.5841,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.3718686103820801,
+      "learning_rate": 0.00012877502001601283,
+      "loss": 0.5402,
+      "step": 2230
+    },
+    {
+      "epoch": 0.71392,
+      "grad_norm": 0.41665902733802795,
+      "learning_rate": 0.00012874299439551642,
+      "loss": 0.5098,
+      "step": 2231
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 0.637062132358551,
+      "learning_rate": 0.00012871096877502004,
+      "loss": 0.8078,
+      "step": 2232
+    },
+    {
+      "epoch": 0.71456,
+      "grad_norm": 0.4585593342781067,
+      "learning_rate": 0.00012867894315452363,
+      "loss": 0.6369,
+      "step": 2233
+    },
+    {
+      "epoch": 0.71488,
+      "grad_norm": 0.4851985275745392,
+      "learning_rate": 0.00012864691753402722,
+      "loss": 0.6214,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 0.4827353358268738,
+      "learning_rate": 0.0001286148919135308,
+      "loss": 0.5206,
+      "step": 2235
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 0.5215303301811218,
+      "learning_rate": 0.00012858286629303446,
+      "loss": 0.6592,
+      "step": 2236
+    },
+    {
+      "epoch": 0.71584,
+      "grad_norm": 0.4855058789253235,
+      "learning_rate": 0.00012855084067253805,
+      "loss": 0.4633,
+      "step": 2237
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 0.4790673553943634,
+      "learning_rate": 0.00012851881505204164,
+      "loss": 0.6229,
+      "step": 2238
+    },
+    {
+      "epoch": 0.71648,
+      "grad_norm": 0.4728359580039978,
+      "learning_rate": 0.00012848678943154523,
+      "loss": 0.5248,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.43436291813850403,
+      "learning_rate": 0.00012845476381104885,
+      "loss": 0.5972,
+      "step": 2240
+    },
+    {
+      "epoch": 0.71712,
+      "grad_norm": 0.478764146566391,
+      "learning_rate": 0.00012842273819055244,
+      "loss": 0.5906,
+      "step": 2241
+    },
+    {
+      "epoch": 0.71744,
+      "grad_norm": 0.5667250156402588,
+      "learning_rate": 0.00012839071257005606,
+      "loss": 0.5994,
+      "step": 2242
+    },
+    {
+      "epoch": 0.71776,
+      "grad_norm": 0.38697320222854614,
+      "learning_rate": 0.00012835868694955965,
+      "loss": 0.4609,
+      "step": 2243
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 0.4326903820037842,
+      "learning_rate": 0.00012832666132906327,
+      "loss": 0.4849,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 0.4952547550201416,
+      "learning_rate": 0.00012829463570856686,
+      "loss": 0.6439,
+      "step": 2245
+    },
+    {
+      "epoch": 0.71872,
+      "grad_norm": 0.42796915769577026,
+      "learning_rate": 0.00012826261008807045,
+      "loss": 0.4944,
+      "step": 2246
+    },
+    {
+      "epoch": 0.71904,
+      "grad_norm": 0.49013379216194153,
+      "learning_rate": 0.00012823058446757407,
+      "loss": 0.6452,
+      "step": 2247
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 0.4083814322948456,
+      "learning_rate": 0.00012819855884707769,
+      "loss": 0.5979,
+      "step": 2248
+    },
+    {
+      "epoch": 0.71968,
+      "grad_norm": 0.4450615644454956,
+      "learning_rate": 0.00012816653322658128,
+      "loss": 0.4601,
+      "step": 2249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3737667202949524,
+      "learning_rate": 0.00012813450760608487,
+      "loss": 0.5075,
+      "step": 2250
+    },
+    {
+      "epoch": 0.72032,
+      "grad_norm": 0.41689103841781616,
+      "learning_rate": 0.00012810248198558849,
+      "loss": 0.6356,
+      "step": 2251
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 0.41961103677749634,
+      "learning_rate": 0.00012807045636509208,
+      "loss": 0.5462,
+      "step": 2252
+    },
+    {
+      "epoch": 0.72096,
+      "grad_norm": 0.4439800977706909,
+      "learning_rate": 0.00012803843074459567,
+      "loss": 0.4913,
+      "step": 2253
+    },
+    {
+      "epoch": 0.72128,
+      "grad_norm": 0.4446849524974823,
+      "learning_rate": 0.0001280064051240993,
+      "loss": 0.615,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 0.4351779818534851,
+      "learning_rate": 0.0001279743795036029,
+      "loss": 0.5068,
+      "step": 2255
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 0.4712657928466797,
+      "learning_rate": 0.0001279423538831065,
+      "loss": 0.6155,
+      "step": 2256
+    },
+    {
+      "epoch": 0.72224,
+      "grad_norm": 0.4732832610607147,
+      "learning_rate": 0.0001279103282626101,
+      "loss": 0.5999,
+      "step": 2257
+    },
+    {
+      "epoch": 0.72256,
+      "grad_norm": 0.41083431243896484,
+      "learning_rate": 0.00012787830264211368,
+      "loss": 0.5548,
+      "step": 2258
+    },
+    {
+      "epoch": 0.72288,
+      "grad_norm": 0.5003586411476135,
+      "learning_rate": 0.0001278462770216173,
+      "loss": 0.5319,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.524986743927002,
+      "learning_rate": 0.00012781425140112092,
+      "loss": 0.4262,
+      "step": 2260
+    },
+    {
+      "epoch": 0.72352,
+      "grad_norm": 0.47123655676841736,
+      "learning_rate": 0.0001277822257806245,
+      "loss": 0.5338,
+      "step": 2261
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 0.5186426043510437,
+      "learning_rate": 0.0001277502001601281,
+      "loss": 0.7878,
+      "step": 2262
+    },
+    {
+      "epoch": 0.72416,
+      "grad_norm": 0.376624196767807,
+      "learning_rate": 0.00012771817453963172,
+      "loss": 0.4845,
+      "step": 2263
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 0.4609549343585968,
+      "learning_rate": 0.0001276861489191353,
+      "loss": 0.5597,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 0.5924606919288635,
+      "learning_rate": 0.00012765412329863893,
+      "loss": 0.5365,
+      "step": 2265
+    },
+    {
+      "epoch": 0.72512,
+      "grad_norm": 0.540764331817627,
+      "learning_rate": 0.00012762209767814252,
+      "loss": 0.5968,
+      "step": 2266
+    },
+    {
+      "epoch": 0.72544,
+      "grad_norm": 0.40143153071403503,
+      "learning_rate": 0.00012759007205764614,
+      "loss": 0.5275,
+      "step": 2267
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 0.4403091371059418,
+      "learning_rate": 0.00012755804643714973,
+      "loss": 0.5523,
+      "step": 2268
+    },
+    {
+      "epoch": 0.72608,
+      "grad_norm": 0.42376816272735596,
+      "learning_rate": 0.00012752602081665332,
+      "loss": 0.5176,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 0.48698917031288147,
+      "learning_rate": 0.00012749399519615694,
+      "loss": 0.6238,
+      "step": 2270
+    },
+    {
+      "epoch": 0.72672,
+      "grad_norm": 0.46401381492614746,
+      "learning_rate": 0.00012746196957566055,
+      "loss": 0.5935,
+      "step": 2271
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 0.4645531475543976,
+      "learning_rate": 0.00012742994395516415,
+      "loss": 0.506,
+      "step": 2272
+    },
+    {
+      "epoch": 0.72736,
+      "grad_norm": 0.6182917356491089,
+      "learning_rate": 0.00012739791833466774,
+      "loss": 0.659,
+      "step": 2273
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 0.5096768736839294,
+      "learning_rate": 0.00012736589271417135,
+      "loss": 0.6557,
+      "step": 2274
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.5057377815246582,
+      "learning_rate": 0.00012733386709367495,
+      "loss": 0.6631,
+      "step": 2275
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 0.4532689154148102,
+      "learning_rate": 0.00012730184147317854,
+      "loss": 0.5451,
+      "step": 2276
+    },
+    {
+      "epoch": 0.72864,
+      "grad_norm": 0.44215166568756104,
+      "learning_rate": 0.00012726981585268216,
+      "loss": 0.6712,
+      "step": 2277
+    },
+    {
+      "epoch": 0.72896,
+      "grad_norm": 0.43771055340766907,
+      "learning_rate": 0.00012723779023218577,
+      "loss": 0.4622,
+      "step": 2278
+    },
+    {
+      "epoch": 0.72928,
+      "grad_norm": 0.6048862934112549,
+      "learning_rate": 0.00012720576461168937,
+      "loss": 0.7671,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.4657733142375946,
+      "learning_rate": 0.00012717373899119296,
+      "loss": 0.5662,
+      "step": 2280
+    },
+    {
+      "epoch": 0.72992,
+      "grad_norm": 0.48945486545562744,
+      "learning_rate": 0.00012714171337069655,
+      "loss": 0.4917,
+      "step": 2281
+    },
+    {
+      "epoch": 0.73024,
+      "grad_norm": 0.4630267322063446,
+      "learning_rate": 0.00012710968775020017,
+      "loss": 0.5017,
+      "step": 2282
+    },
+    {
+      "epoch": 0.73056,
+      "grad_norm": 0.45915350317955017,
+      "learning_rate": 0.00012707766212970378,
+      "loss": 0.5868,
+      "step": 2283
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 0.4666953384876251,
+      "learning_rate": 0.00012704563650920738,
+      "loss": 0.5659,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 0.4091738164424896,
+      "learning_rate": 0.00012701361088871097,
+      "loss": 0.4226,
+      "step": 2285
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 0.4711691737174988,
+      "learning_rate": 0.00012698158526821458,
+      "loss": 0.6714,
+      "step": 2286
+    },
+    {
+      "epoch": 0.73184,
+      "grad_norm": 0.4315379559993744,
+      "learning_rate": 0.00012694955964771818,
+      "loss": 0.5757,
+      "step": 2287
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 0.5948399305343628,
+      "learning_rate": 0.00012691753402722177,
+      "loss": 0.6563,
+      "step": 2288
+    },
+    {
+      "epoch": 0.73248,
+      "grad_norm": 0.4153330624103546,
+      "learning_rate": 0.00012688550840672539,
+      "loss": 0.5743,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.44154471158981323,
+      "learning_rate": 0.000126853482786229,
+      "loss": 0.5444,
+      "step": 2290
+    },
+    {
+      "epoch": 0.73312,
+      "grad_norm": 0.45439979434013367,
+      "learning_rate": 0.0001268214571657326,
+      "loss": 0.5981,
+      "step": 2291
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 0.5143215656280518,
+      "learning_rate": 0.0001267894315452362,
+      "loss": 0.6744,
+      "step": 2292
+    },
+    {
+      "epoch": 0.73376,
+      "grad_norm": 0.4303244352340698,
+      "learning_rate": 0.0001267574059247398,
+      "loss": 0.5454,
+      "step": 2293
+    },
+    {
+      "epoch": 0.73408,
+      "grad_norm": 0.4696168303489685,
+      "learning_rate": 0.0001267253803042434,
+      "loss": 0.5305,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 0.48198986053466797,
+      "learning_rate": 0.00012669335468374701,
+      "loss": 0.8007,
+      "step": 2295
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 0.47836825251579285,
+      "learning_rate": 0.0001266613290632506,
+      "loss": 0.6134,
+      "step": 2296
+    },
+    {
+      "epoch": 0.73504,
+      "grad_norm": 0.46310582756996155,
+      "learning_rate": 0.0001266293034427542,
+      "loss": 0.5995,
+      "step": 2297
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 0.5071401000022888,
+      "learning_rate": 0.00012659727782225782,
+      "loss": 0.6095,
+      "step": 2298
+    },
+    {
+      "epoch": 0.73568,
+      "grad_norm": 0.44836053252220154,
+      "learning_rate": 0.0001265652522017614,
+      "loss": 0.4036,
+      "step": 2299
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.4242633283138275,
+      "learning_rate": 0.000126533226581265,
+      "loss": 0.4532,
+      "step": 2300
+    },
+    {
+      "epoch": 0.73632,
+      "grad_norm": 0.5059876441955566,
+      "learning_rate": 0.00012650120096076862,
+      "loss": 0.5248,
+      "step": 2301
+    },
+    {
+      "epoch": 0.73664,
+      "grad_norm": 0.5445113182067871,
+      "learning_rate": 0.00012646917534027223,
+      "loss": 0.6311,
+      "step": 2302
+    },
+    {
+      "epoch": 0.73696,
+      "grad_norm": 0.43020540475845337,
+      "learning_rate": 0.00012643714971977583,
+      "loss": 0.5841,
+      "step": 2303
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 0.44270652532577515,
+      "learning_rate": 0.00012640512409927942,
+      "loss": 0.7171,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 0.5372592806816101,
+      "learning_rate": 0.00012637309847878303,
+      "loss": 0.587,
+      "step": 2305
+    },
+    {
+      "epoch": 0.73792,
+      "grad_norm": 0.45795148611068726,
+      "learning_rate": 0.00012634107285828663,
+      "loss": 0.6529,
+      "step": 2306
+    },
+    {
+      "epoch": 0.73824,
+      "grad_norm": 0.459388792514801,
+      "learning_rate": 0.00012630904723779024,
+      "loss": 0.4417,
+      "step": 2307
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 0.4047694504261017,
+      "learning_rate": 0.00012627702161729384,
+      "loss": 0.5031,
+      "step": 2308
+    },
+    {
+      "epoch": 0.73888,
+      "grad_norm": 0.5076388716697693,
+      "learning_rate": 0.00012624499599679745,
+      "loss": 0.6317,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.4946870505809784,
+      "learning_rate": 0.00012621297037630105,
+      "loss": 0.7466,
+      "step": 2310
+    },
+    {
+      "epoch": 0.73952,
+      "grad_norm": 0.4503292441368103,
+      "learning_rate": 0.00012618094475580464,
+      "loss": 0.5088,
+      "step": 2311
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 0.5003451108932495,
+      "learning_rate": 0.00012614891913530825,
+      "loss": 0.6122,
+      "step": 2312
+    },
+    {
+      "epoch": 0.74016,
+      "grad_norm": 0.41444119811058044,
+      "learning_rate": 0.00012611689351481187,
+      "loss": 0.4311,
+      "step": 2313
+    },
+    {
+      "epoch": 0.74048,
+      "grad_norm": 0.47419893741607666,
+      "learning_rate": 0.00012608486789431546,
+      "loss": 0.5338,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 0.4111938178539276,
+      "learning_rate": 0.00012605284227381906,
+      "loss": 0.6227,
+      "step": 2315
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 0.43531155586242676,
+      "learning_rate": 0.00012602081665332265,
+      "loss": 0.5121,
+      "step": 2316
+    },
+    {
+      "epoch": 0.74144,
+      "grad_norm": 0.44196274876594543,
+      "learning_rate": 0.00012598879103282626,
+      "loss": 0.5802,
+      "step": 2317
+    },
+    {
+      "epoch": 0.74176,
+      "grad_norm": 0.500645101070404,
+      "learning_rate": 0.00012595676541232988,
+      "loss": 0.6739,
+      "step": 2318
+    },
+    {
+      "epoch": 0.74208,
+      "grad_norm": 0.5478889346122742,
+      "learning_rate": 0.00012592473979183347,
+      "loss": 0.7046,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.44969475269317627,
+      "learning_rate": 0.00012589271417133707,
+      "loss": 0.6153,
+      "step": 2320
+    },
+    {
+      "epoch": 0.74272,
+      "grad_norm": 0.4962233603000641,
+      "learning_rate": 0.00012586068855084068,
+      "loss": 0.6353,
+      "step": 2321
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 0.4504684507846832,
+      "learning_rate": 0.00012582866293034428,
+      "loss": 0.5622,
+      "step": 2322
+    },
+    {
+      "epoch": 0.74336,
+      "grad_norm": 0.456289678812027,
+      "learning_rate": 0.00012579663730984787,
+      "loss": 0.5627,
+      "step": 2323
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 0.40530160069465637,
+      "learning_rate": 0.00012576461168935148,
+      "loss": 0.4438,
+      "step": 2324
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 0.48793336749076843,
+      "learning_rate": 0.0001257325860688551,
+      "loss": 0.613,
+      "step": 2325
+    },
+    {
+      "epoch": 0.74432,
+      "grad_norm": 0.39526358246803284,
+      "learning_rate": 0.0001257005604483587,
+      "loss": 0.5811,
+      "step": 2326
+    },
+    {
+      "epoch": 0.74464,
+      "grad_norm": 0.41980692744255066,
+      "learning_rate": 0.00012566853482786229,
+      "loss": 0.4667,
+      "step": 2327
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 0.5090502500534058,
+      "learning_rate": 0.0001256365092073659,
+      "loss": 0.7446,
+      "step": 2328
+    },
+    {
+      "epoch": 0.74528,
+      "grad_norm": 0.4514313042163849,
+      "learning_rate": 0.0001256044835868695,
+      "loss": 0.6269,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 0.4526832103729248,
+      "learning_rate": 0.0001255724579663731,
+      "loss": 0.7178,
+      "step": 2330
+    },
+    {
+      "epoch": 0.74592,
+      "grad_norm": 0.4691832959651947,
+      "learning_rate": 0.0001255404323458767,
+      "loss": 0.4997,
+      "step": 2331
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 0.458781898021698,
+      "learning_rate": 0.00012550840672538032,
+      "loss": 0.6396,
+      "step": 2332
+    },
+    {
+      "epoch": 0.74656,
+      "grad_norm": 0.44559213519096375,
+      "learning_rate": 0.00012547638110488391,
+      "loss": 0.5876,
+      "step": 2333
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 0.3975817859172821,
+      "learning_rate": 0.0001254443554843875,
+      "loss": 0.4773,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 0.4195534288883209,
+      "learning_rate": 0.0001254123298638911,
+      "loss": 0.5649,
+      "step": 2335
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 0.49985846877098083,
+      "learning_rate": 0.00012538030424339474,
+      "loss": 0.7171,
+      "step": 2336
+    },
+    {
+      "epoch": 0.74784,
+      "grad_norm": 0.5188073515892029,
+      "learning_rate": 0.00012534827862289833,
+      "loss": 0.6557,
+      "step": 2337
+    },
+    {
+      "epoch": 0.74816,
+      "grad_norm": 0.3692166209220886,
+      "learning_rate": 0.00012531625300240192,
+      "loss": 0.4645,
+      "step": 2338
+    },
+    {
+      "epoch": 0.74848,
+      "grad_norm": 0.49693211913108826,
+      "learning_rate": 0.00012528422738190552,
+      "loss": 0.6414,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.5167421698570251,
+      "learning_rate": 0.00012525220176140913,
+      "loss": 0.6218,
+      "step": 2340
+    },
+    {
+      "epoch": 0.74912,
+      "grad_norm": 0.44577130675315857,
+      "learning_rate": 0.00012522017614091272,
+      "loss": 0.5654,
+      "step": 2341
+    },
+    {
+      "epoch": 0.74944,
+      "grad_norm": 0.47415435314178467,
+      "learning_rate": 0.00012518815052041634,
+      "loss": 0.6522,
+      "step": 2342
+    },
+    {
+      "epoch": 0.74976,
+      "grad_norm": 0.4068785309791565,
+      "learning_rate": 0.00012515612489991993,
+      "loss": 0.5849,
+      "step": 2343
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 0.5082315802574158,
+      "learning_rate": 0.00012512409927942355,
+      "loss": 0.55,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 0.4728572964668274,
+      "learning_rate": 0.00012509207365892714,
+      "loss": 0.4575,
+      "step": 2345
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 0.560402512550354,
+      "learning_rate": 0.00012506004803843074,
+      "loss": 0.6853,
+      "step": 2346
+    },
+    {
+      "epoch": 0.75104,
+      "grad_norm": 0.34757983684539795,
+      "learning_rate": 0.00012502802241793435,
+      "loss": 0.4699,
+      "step": 2347
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 0.3660871684551239,
+      "learning_rate": 0.00012499599679743797,
+      "loss": 0.4953,
+      "step": 2348
+    },
+    {
+      "epoch": 0.75168,
+      "grad_norm": 0.5410837531089783,
+      "learning_rate": 0.00012496397117694156,
+      "loss": 0.6555,
+      "step": 2349
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.632577657699585,
+      "learning_rate": 0.00012493194555644515,
+      "loss": 0.7016,
+      "step": 2350
+    },
+    {
+      "epoch": 0.75232,
+      "grad_norm": 0.4965982735157013,
+      "learning_rate": 0.00012489991993594877,
+      "loss": 0.6231,
+      "step": 2351
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 0.4463093876838684,
+      "learning_rate": 0.00012486789431545236,
+      "loss": 0.5786,
+      "step": 2352
+    },
+    {
+      "epoch": 0.75296,
+      "grad_norm": 0.5159125328063965,
+      "learning_rate": 0.00012483586869495595,
+      "loss": 0.5581,
+      "step": 2353
+    },
+    {
+      "epoch": 0.75328,
+      "grad_norm": 0.43563875555992126,
+      "learning_rate": 0.00012480384307445957,
+      "loss": 0.4895,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 0.41268038749694824,
+      "learning_rate": 0.0001247718174539632,
+      "loss": 0.554,
+      "step": 2355
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 0.4521600902080536,
+      "learning_rate": 0.00012473979183346678,
+      "loss": 0.4407,
+      "step": 2356
+    },
+    {
+      "epoch": 0.75424,
+      "grad_norm": 0.4537864029407501,
+      "learning_rate": 0.00012470776621297037,
+      "loss": 0.6401,
+      "step": 2357
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 0.44591888785362244,
+      "learning_rate": 0.00012467574059247397,
+      "loss": 0.5792,
+      "step": 2358
+    },
+    {
+      "epoch": 0.75488,
+      "grad_norm": 0.49586209654808044,
+      "learning_rate": 0.0001246437149719776,
+      "loss": 0.5334,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.5400336384773254,
+      "learning_rate": 0.0001246116893514812,
+      "loss": 0.6156,
+      "step": 2360
+    },
+    {
+      "epoch": 0.75552,
+      "grad_norm": 0.4186513423919678,
+      "learning_rate": 0.0001245796637309848,
+      "loss": 0.6621,
+      "step": 2361
+    },
+    {
+      "epoch": 0.75584,
+      "grad_norm": 0.43246179819107056,
+      "learning_rate": 0.00012454763811048838,
+      "loss": 0.5841,
+      "step": 2362
+    },
+    {
+      "epoch": 0.75616,
+      "grad_norm": 0.5332304239273071,
+      "learning_rate": 0.000124515612489992,
+      "loss": 0.661,
+      "step": 2363
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 0.4880680739879608,
+      "learning_rate": 0.0001244835868694956,
+      "loss": 0.5244,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 0.5403331518173218,
+      "learning_rate": 0.0001244515612489992,
+      "loss": 0.6093,
+      "step": 2365
+    },
+    {
+      "epoch": 0.75712,
+      "grad_norm": 0.43770018219947815,
+      "learning_rate": 0.0001244195356285028,
+      "loss": 0.6168,
+      "step": 2366
+    },
+    {
+      "epoch": 0.75744,
+      "grad_norm": 0.46149271726608276,
+      "learning_rate": 0.00012438751000800642,
+      "loss": 0.5103,
+      "step": 2367
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 0.5035654306411743,
+      "learning_rate": 0.00012435548438751,
+      "loss": 0.5468,
+      "step": 2368
+    },
+    {
+      "epoch": 0.75808,
+      "grad_norm": 0.5344579815864563,
+      "learning_rate": 0.0001243234587670136,
+      "loss": 0.7372,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.5109825134277344,
+      "learning_rate": 0.00012429143314651722,
+      "loss": 0.5918,
+      "step": 2370
+    },
+    {
+      "epoch": 0.75872,
+      "grad_norm": 0.48059260845184326,
+      "learning_rate": 0.00012425940752602084,
+      "loss": 0.6036,
+      "step": 2371
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 0.40955471992492676,
+      "learning_rate": 0.00012422738190552443,
+      "loss": 0.6039,
+      "step": 2372
+    },
+    {
+      "epoch": 0.75936,
+      "grad_norm": 0.4921233654022217,
+      "learning_rate": 0.00012419535628502802,
+      "loss": 0.5353,
+      "step": 2373
+    },
+    {
+      "epoch": 0.75968,
+      "grad_norm": 0.4590461850166321,
+      "learning_rate": 0.00012416333066453164,
+      "loss": 0.5979,
+      "step": 2374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4449748396873474,
+      "learning_rate": 0.00012413130504403523,
+      "loss": 0.6595,
+      "step": 2375
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 0.6065210103988647,
+      "learning_rate": 0.00012409927942353882,
+      "loss": 0.8274,
+      "step": 2376
+    },
+    {
+      "epoch": 0.76064,
+      "grad_norm": 0.42787057161331177,
+      "learning_rate": 0.00012406725380304244,
+      "loss": 0.5245,
+      "step": 2377
+    },
+    {
+      "epoch": 0.76096,
+      "grad_norm": 0.44588539004325867,
+      "learning_rate": 0.00012403522818254606,
+      "loss": 0.6097,
+      "step": 2378
+    },
+    {
+      "epoch": 0.76128,
+      "grad_norm": 0.3806239366531372,
+      "learning_rate": 0.00012400320256204965,
+      "loss": 0.5867,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 0.47309327125549316,
+      "learning_rate": 0.00012397117694155324,
+      "loss": 0.6604,
+      "step": 2380
+    },
+    {
+      "epoch": 0.76192,
+      "grad_norm": 0.48379841446876526,
+      "learning_rate": 0.00012393915132105683,
+      "loss": 0.5626,
+      "step": 2381
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 0.5105714201927185,
+      "learning_rate": 0.00012390712570056045,
+      "loss": 0.6384,
+      "step": 2382
+    },
+    {
+      "epoch": 0.76256,
+      "grad_norm": 0.4915042817592621,
+      "learning_rate": 0.00012387510008006407,
+      "loss": 0.5627,
+      "step": 2383
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 0.539996862411499,
+      "learning_rate": 0.00012384307445956766,
+      "loss": 0.6858,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 0.4486226737499237,
+      "learning_rate": 0.00012381104883907125,
+      "loss": 0.6267,
+      "step": 2385
+    },
+    {
+      "epoch": 0.76352,
+      "grad_norm": 0.3586444556713104,
+      "learning_rate": 0.00012377902321857487,
+      "loss": 0.4971,
+      "step": 2386
+    },
+    {
+      "epoch": 0.76384,
+      "grad_norm": 0.45715203881263733,
+      "learning_rate": 0.00012374699759807846,
+      "loss": 0.5671,
+      "step": 2387
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 0.4145236909389496,
+      "learning_rate": 0.00012371497197758205,
+      "loss": 0.6071,
+      "step": 2388
+    },
+    {
+      "epoch": 0.76448,
+      "grad_norm": 0.4224635362625122,
+      "learning_rate": 0.00012368294635708567,
+      "loss": 0.5758,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 0.46575304865837097,
+      "learning_rate": 0.0001236509207365893,
+      "loss": 0.5234,
+      "step": 2390
+    },
+    {
+      "epoch": 0.76512,
+      "grad_norm": 0.348375529050827,
+      "learning_rate": 0.00012361889511609288,
+      "loss": 0.4115,
+      "step": 2391
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 0.5129759311676025,
+      "learning_rate": 0.00012358686949559647,
+      "loss": 0.5048,
+      "step": 2392
+    },
+    {
+      "epoch": 0.76576,
+      "grad_norm": 0.3587447702884674,
+      "learning_rate": 0.0001235548438751001,
+      "loss": 0.5216,
+      "step": 2393
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 0.45737236738204956,
+      "learning_rate": 0.00012352281825460368,
+      "loss": 0.5646,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 0.5346295833587646,
+      "learning_rate": 0.0001234907926341073,
+      "loss": 0.5477,
+      "step": 2395
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 0.5019865036010742,
+      "learning_rate": 0.0001234587670136109,
+      "loss": 0.5351,
+      "step": 2396
+    },
+    {
+      "epoch": 0.76704,
+      "grad_norm": 0.5248101353645325,
+      "learning_rate": 0.0001234267413931145,
+      "loss": 0.7658,
+      "step": 2397
+    },
+    {
+      "epoch": 0.76736,
+      "grad_norm": 0.40013912320137024,
+      "learning_rate": 0.0001233947157726181,
+      "loss": 0.5448,
+      "step": 2398
+    },
+    {
+      "epoch": 0.76768,
+      "grad_norm": 0.4300834536552429,
+      "learning_rate": 0.0001233626901521217,
+      "loss": 0.5112,
+      "step": 2399
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.6473549008369446,
+      "learning_rate": 0.00012333066453162528,
+      "loss": 0.5888,
+      "step": 2400
+    },
+    {
+      "epoch": 0.76832,
+      "grad_norm": 0.66737961769104,
+      "learning_rate": 0.00012329863891112893,
+      "loss": 0.7884,
+      "step": 2401
+    },
+    {
+      "epoch": 0.76864,
+      "grad_norm": 0.6415278315544128,
+      "learning_rate": 0.00012326661329063252,
+      "loss": 0.7082,
+      "step": 2402
+    },
+    {
+      "epoch": 0.76896,
+      "grad_norm": 0.48603323101997375,
+      "learning_rate": 0.0001232345876701361,
+      "loss": 0.6351,
+      "step": 2403
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 0.41410043835639954,
+      "learning_rate": 0.0001232025620496397,
+      "loss": 0.4707,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 0.4656958281993866,
+      "learning_rate": 0.00012317053642914332,
+      "loss": 0.5931,
+      "step": 2405
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 0.4776204526424408,
+      "learning_rate": 0.00012313851080864694,
+      "loss": 0.57,
+      "step": 2406
+    },
+    {
+      "epoch": 0.77024,
+      "grad_norm": 0.5044889450073242,
+      "learning_rate": 0.00012310648518815053,
+      "loss": 0.6529,
+      "step": 2407
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 0.43001115322113037,
+      "learning_rate": 0.00012307445956765412,
+      "loss": 0.5934,
+      "step": 2408
+    },
+    {
+      "epoch": 0.77088,
+      "grad_norm": 0.5896954536437988,
+      "learning_rate": 0.00012304243394715774,
+      "loss": 0.7006,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 0.44463640451431274,
+      "learning_rate": 0.00012301040832666133,
+      "loss": 0.4944,
+      "step": 2410
+    },
+    {
+      "epoch": 0.77152,
+      "grad_norm": 0.4469371438026428,
+      "learning_rate": 0.00012297838270616492,
+      "loss": 0.5184,
+      "step": 2411
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 0.5109202861785889,
+      "learning_rate": 0.00012294635708566854,
+      "loss": 0.6615,
+      "step": 2412
+    },
+    {
+      "epoch": 0.77216,
+      "grad_norm": 0.4634099304676056,
+      "learning_rate": 0.00012291433146517216,
+      "loss": 0.5898,
+      "step": 2413
+    },
+    {
+      "epoch": 0.77248,
+      "grad_norm": 0.4335262179374695,
+      "learning_rate": 0.00012288230584467575,
+      "loss": 0.6242,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 0.4996366500854492,
+      "learning_rate": 0.00012285028022417934,
+      "loss": 0.6907,
+      "step": 2415
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 0.5231391787528992,
+      "learning_rate": 0.00012281825460368296,
+      "loss": 0.6715,
+      "step": 2416
+    },
+    {
+      "epoch": 0.77344,
+      "grad_norm": 0.5835326313972473,
+      "learning_rate": 0.00012278622898318655,
+      "loss": 0.5624,
+      "step": 2417
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 0.44012507796287537,
+      "learning_rate": 0.00012275420336269017,
+      "loss": 0.5015,
+      "step": 2418
+    },
+    {
+      "epoch": 0.77408,
+      "grad_norm": 0.4599383473396301,
+      "learning_rate": 0.00012272217774219376,
+      "loss": 0.598,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 0.4844197928905487,
+      "learning_rate": 0.00012269015212169738,
+      "loss": 0.5401,
+      "step": 2420
+    },
+    {
+      "epoch": 0.77472,
+      "grad_norm": 0.3874641954898834,
+      "learning_rate": 0.00012265812650120097,
+      "loss": 0.5801,
+      "step": 2421
+    },
+    {
+      "epoch": 0.77504,
+      "grad_norm": 0.4283032715320587,
+      "learning_rate": 0.00012262610088070456,
+      "loss": 0.6708,
+      "step": 2422
+    },
+    {
+      "epoch": 0.77536,
+      "grad_norm": 0.5407668948173523,
+      "learning_rate": 0.00012259407526020815,
+      "loss": 0.7629,
+      "step": 2423
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 0.4563540518283844,
+      "learning_rate": 0.00012256204963971177,
+      "loss": 0.6008,
+      "step": 2424
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 0.36084774136543274,
+      "learning_rate": 0.0001225300240192154,
+      "loss": 0.511,
+      "step": 2425
+    },
+    {
+      "epoch": 0.77632,
+      "grad_norm": 0.48872217535972595,
+      "learning_rate": 0.00012249799839871898,
+      "loss": 0.7074,
+      "step": 2426
+    },
+    {
+      "epoch": 0.77664,
+      "grad_norm": 0.40319663286209106,
+      "learning_rate": 0.00012246597277822257,
+      "loss": 0.574,
+      "step": 2427
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 0.42561057209968567,
+      "learning_rate": 0.0001224339471577262,
+      "loss": 0.5569,
+      "step": 2428
+    },
+    {
+      "epoch": 0.77728,
+      "grad_norm": 0.5169298052787781,
+      "learning_rate": 0.00012240192153722978,
+      "loss": 0.8593,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 0.4481830894947052,
+      "learning_rate": 0.0001223698959167334,
+      "loss": 0.5624,
+      "step": 2430
+    },
+    {
+      "epoch": 0.77792,
+      "grad_norm": 0.43105268478393555,
+      "learning_rate": 0.000122337870296237,
+      "loss": 0.616,
+      "step": 2431
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 0.47840437293052673,
+      "learning_rate": 0.0001223058446757406,
+      "loss": 0.6032,
+      "step": 2432
+    },
+    {
+      "epoch": 0.77856,
+      "grad_norm": 0.4641217291355133,
+      "learning_rate": 0.0001222738190552442,
+      "loss": 0.5853,
+      "step": 2433
+    },
+    {
+      "epoch": 0.77888,
+      "grad_norm": 0.47455161809921265,
+      "learning_rate": 0.0001222417934347478,
+      "loss": 0.4304,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 0.451654851436615,
+      "learning_rate": 0.0001222097678142514,
+      "loss": 0.6224,
+      "step": 2435
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 0.41325125098228455,
+      "learning_rate": 0.00012217774219375503,
+      "loss": 0.6505,
+      "step": 2436
+    },
+    {
+      "epoch": 0.77984,
+      "grad_norm": 0.4352465271949768,
+      "learning_rate": 0.00012214571657325862,
+      "loss": 0.6393,
+      "step": 2437
+    },
+    {
+      "epoch": 0.78016,
+      "grad_norm": 0.3421870470046997,
+      "learning_rate": 0.0001221136909527622,
+      "loss": 0.5041,
+      "step": 2438
+    },
+    {
+      "epoch": 0.78048,
+      "grad_norm": 0.4289330244064331,
+      "learning_rate": 0.00012208166533226583,
+      "loss": 0.5852,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 0.4517141580581665,
+      "learning_rate": 0.00012204963971176942,
+      "loss": 0.5842,
+      "step": 2440
+    },
+    {
+      "epoch": 0.78112,
+      "grad_norm": 0.3863097131252289,
+      "learning_rate": 0.00012201761409127301,
+      "loss": 0.332,
+      "step": 2441
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 0.530438244342804,
+      "learning_rate": 0.00012198558847077663,
+      "loss": 0.8319,
+      "step": 2442
+    },
+    {
+      "epoch": 0.78176,
+      "grad_norm": 0.47331857681274414,
+      "learning_rate": 0.00012195356285028023,
+      "loss": 0.5824,
+      "step": 2443
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 0.5492784380912781,
+      "learning_rate": 0.00012192153722978384,
+      "loss": 0.575,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 0.41859081387519836,
+      "learning_rate": 0.00012188951160928743,
+      "loss": 0.5004,
+      "step": 2445
+    },
+    {
+      "epoch": 0.78272,
+      "grad_norm": 0.4249667525291443,
+      "learning_rate": 0.00012185748598879104,
+      "loss": 0.4625,
+      "step": 2446
+    },
+    {
+      "epoch": 0.78304,
+      "grad_norm": 0.4485050439834595,
+      "learning_rate": 0.00012182546036829463,
+      "loss": 0.6527,
+      "step": 2447
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 0.5868487358093262,
+      "learning_rate": 0.00012179343474779826,
+      "loss": 0.6937,
+      "step": 2448
+    },
+    {
+      "epoch": 0.78368,
+      "grad_norm": 0.4618946611881256,
+      "learning_rate": 0.00012176140912730185,
+      "loss": 0.4882,
+      "step": 2449
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 0.44787976145744324,
+      "learning_rate": 0.00012172938350680545,
+      "loss": 0.4397,
+      "step": 2450
+    },
+    {
+      "epoch": 0.78432,
+      "grad_norm": 0.4446999132633209,
+      "learning_rate": 0.00012169735788630905,
+      "loss": 0.5707,
+      "step": 2451
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 0.45886170864105225,
+      "learning_rate": 0.00012166533226581265,
+      "loss": 0.5729,
+      "step": 2452
+    },
+    {
+      "epoch": 0.78496,
+      "grad_norm": 0.5526993870735168,
+      "learning_rate": 0.00012163330664531627,
+      "loss": 0.5893,
+      "step": 2453
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 0.47622737288475037,
+      "learning_rate": 0.00012160128102481987,
+      "loss": 0.5003,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 0.4561639726161957,
+      "learning_rate": 0.00012156925540432346,
+      "loss": 0.6228,
+      "step": 2455
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 0.37883880734443665,
+      "learning_rate": 0.00012153722978382707,
+      "loss": 0.3888,
+      "step": 2456
+    },
+    {
+      "epoch": 0.78624,
+      "grad_norm": 0.3688986599445343,
+      "learning_rate": 0.00012150520416333066,
+      "loss": 0.4439,
+      "step": 2457
+    },
+    {
+      "epoch": 0.78656,
+      "grad_norm": 0.5919997692108154,
+      "learning_rate": 0.00012147317854283427,
+      "loss": 0.6483,
+      "step": 2458
+    },
+    {
+      "epoch": 0.78688,
+      "grad_norm": 0.4854859411716461,
+      "learning_rate": 0.00012144115292233788,
+      "loss": 0.5708,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 0.4793454110622406,
+      "learning_rate": 0.00012140912730184149,
+      "loss": 0.5826,
+      "step": 2460
+    },
+    {
+      "epoch": 0.78752,
+      "grad_norm": 0.4953652322292328,
+      "learning_rate": 0.00012137710168134508,
+      "loss": 0.7329,
+      "step": 2461
+    },
+    {
+      "epoch": 0.78784,
+      "grad_norm": 0.41275498270988464,
+      "learning_rate": 0.00012134507606084868,
+      "loss": 0.6367,
+      "step": 2462
+    },
+    {
+      "epoch": 0.78816,
+      "grad_norm": 0.5833121538162231,
+      "learning_rate": 0.00012131305044035229,
+      "loss": 0.6071,
+      "step": 2463
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 0.458523154258728,
+      "learning_rate": 0.00012128102481985588,
+      "loss": 0.6214,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 0.3889986574649811,
+      "learning_rate": 0.0001212489991993595,
+      "loss": 0.4531,
+      "step": 2465
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 0.5388566255569458,
+      "learning_rate": 0.0001212169735788631,
+      "loss": 0.7345,
+      "step": 2466
+    },
+    {
+      "epoch": 0.78944,
+      "grad_norm": 0.5440462827682495,
+      "learning_rate": 0.0001211849479583667,
+      "loss": 0.664,
+      "step": 2467
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 0.3948047459125519,
+      "learning_rate": 0.0001211529223378703,
+      "loss": 0.535,
+      "step": 2468
+    },
+    {
+      "epoch": 0.79008,
+      "grad_norm": 0.4247704744338989,
+      "learning_rate": 0.0001211208967173739,
+      "loss": 0.4059,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 0.4397743046283722,
+      "learning_rate": 0.0001210888710968775,
+      "loss": 0.6703,
+      "step": 2470
+    },
+    {
+      "epoch": 0.79072,
+      "grad_norm": 0.667503833770752,
+      "learning_rate": 0.00012105684547638111,
+      "loss": 0.7472,
+      "step": 2471
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 0.4268604815006256,
+      "learning_rate": 0.00012102481985588472,
+      "loss": 0.4991,
+      "step": 2472
+    },
+    {
+      "epoch": 0.79136,
+      "grad_norm": 0.38424932956695557,
+      "learning_rate": 0.00012099279423538832,
+      "loss": 0.559,
+      "step": 2473
+    },
+    {
+      "epoch": 0.79168,
+      "grad_norm": 0.44833365082740784,
+      "learning_rate": 0.00012096076861489191,
+      "loss": 0.5461,
+      "step": 2474
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 0.4778987169265747,
+      "learning_rate": 0.00012092874299439552,
+      "loss": 0.5247,
+      "step": 2475
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 0.39905059337615967,
+      "learning_rate": 0.00012089671737389911,
+      "loss": 0.5015,
+      "step": 2476
+    },
+    {
+      "epoch": 0.79264,
+      "grad_norm": 0.5216218829154968,
+      "learning_rate": 0.00012086469175340274,
+      "loss": 0.6359,
+      "step": 2477
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 0.4305983781814575,
+      "learning_rate": 0.00012083266613290633,
+      "loss": 0.5592,
+      "step": 2478
+    },
+    {
+      "epoch": 0.79328,
+      "grad_norm": 0.4282768964767456,
+      "learning_rate": 0.00012080064051240994,
+      "loss": 0.5922,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 0.5161150097846985,
+      "learning_rate": 0.00012076861489191353,
+      "loss": 0.6789,
+      "step": 2480
+    },
+    {
+      "epoch": 0.79392,
+      "grad_norm": 0.38567060232162476,
+      "learning_rate": 0.00012073658927141713,
+      "loss": 0.4613,
+      "step": 2481
+    },
+    {
+      "epoch": 0.79424,
+      "grad_norm": 0.37994301319122314,
+      "learning_rate": 0.00012070456365092074,
+      "loss": 0.6617,
+      "step": 2482
+    },
+    {
+      "epoch": 0.79456,
+      "grad_norm": 0.4278795123100281,
+      "learning_rate": 0.00012067253803042436,
+      "loss": 0.5325,
+      "step": 2483
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 0.44483670592308044,
+      "learning_rate": 0.00012064051240992795,
+      "loss": 0.527,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 0.4522266983985901,
+      "learning_rate": 0.00012060848678943155,
+      "loss": 0.5651,
+      "step": 2485
+    },
+    {
+      "epoch": 0.79552,
+      "grad_norm": 0.5145444869995117,
+      "learning_rate": 0.00012057646116893514,
+      "loss": 0.6781,
+      "step": 2486
+    },
+    {
+      "epoch": 0.79584,
+      "grad_norm": 0.3493579626083374,
+      "learning_rate": 0.00012054443554843875,
+      "loss": 0.4793,
+      "step": 2487
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 0.4994441270828247,
+      "learning_rate": 0.00012051240992794235,
+      "loss": 0.6761,
+      "step": 2488
+    },
+    {
+      "epoch": 0.79648,
+      "grad_norm": 0.5358885526657104,
+      "learning_rate": 0.00012048038430744597,
+      "loss": 0.5426,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 0.5260968208312988,
+      "learning_rate": 0.00012044835868694956,
+      "loss": 0.6894,
+      "step": 2490
+    },
+    {
+      "epoch": 0.79712,
+      "grad_norm": 0.5825015306472778,
+      "learning_rate": 0.00012041633306645317,
+      "loss": 0.6924,
+      "step": 2491
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 0.47361257672309875,
+      "learning_rate": 0.00012038430744595677,
+      "loss": 0.5993,
+      "step": 2492
+    },
+    {
+      "epoch": 0.79776,
+      "grad_norm": 0.5334159731864929,
+      "learning_rate": 0.00012035228182546036,
+      "loss": 0.622,
+      "step": 2493
+    },
+    {
+      "epoch": 0.79808,
+      "grad_norm": 0.5701783299446106,
+      "learning_rate": 0.00012032025620496397,
+      "loss": 0.7873,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 0.3862569034099579,
+      "learning_rate": 0.00012028823058446759,
+      "loss": 0.4863,
+      "step": 2495
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 0.5514272451400757,
+      "learning_rate": 0.00012025620496397119,
+      "loss": 0.8013,
+      "step": 2496
+    },
+    {
+      "epoch": 0.79904,
+      "grad_norm": 0.4150988459587097,
+      "learning_rate": 0.00012022417934347478,
+      "loss": 0.5199,
+      "step": 2497
+    },
+    {
+      "epoch": 0.79936,
+      "grad_norm": 0.4717997908592224,
+      "learning_rate": 0.00012019215372297839,
+      "loss": 0.4839,
+      "step": 2498
+    },
+    {
+      "epoch": 0.79968,
+      "grad_norm": 0.5516729950904846,
+      "learning_rate": 0.00012016012810248198,
+      "loss": 0.849,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6162140965461731,
+      "learning_rate": 0.00012012810248198561,
+      "loss": 0.8254,
+      "step": 2500
+    },
+    {
+      "epoch": 0.80032,
+      "grad_norm": 0.5089409947395325,
+      "learning_rate": 0.0001200960768614892,
+      "loss": 0.6197,
+      "step": 2501
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 0.41079384088516235,
+      "learning_rate": 0.00012006405124099281,
+      "loss": 0.6238,
+      "step": 2502
+    },
+    {
+      "epoch": 0.80096,
+      "grad_norm": 0.37976977229118347,
+      "learning_rate": 0.0001200320256204964,
+      "loss": 0.5374,
+      "step": 2503
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 0.4156246483325958,
+      "learning_rate": 0.00012,
+      "loss": 0.5396,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 0.4777294993400574,
+      "learning_rate": 0.0001199679743795036,
+      "loss": 0.6881,
+      "step": 2505
+    },
+    {
+      "epoch": 0.80192,
+      "grad_norm": 0.4389150142669678,
+      "learning_rate": 0.00011993594875900723,
+      "loss": 0.5122,
+      "step": 2506
+    },
+    {
+      "epoch": 0.80224,
+      "grad_norm": 0.5479406118392944,
+      "learning_rate": 0.00011990392313851082,
+      "loss": 0.5179,
+      "step": 2507
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 0.540820837020874,
+      "learning_rate": 0.00011987189751801442,
+      "loss": 0.6656,
+      "step": 2508
+    },
+    {
+      "epoch": 0.80288,
+      "grad_norm": 0.415444940328598,
+      "learning_rate": 0.00011983987189751801,
+      "loss": 0.6124,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 0.40663614869117737,
+      "learning_rate": 0.00011980784627702162,
+      "loss": 0.4782,
+      "step": 2510
+    },
+    {
+      "epoch": 0.80352,
+      "grad_norm": 0.4126022756099701,
+      "learning_rate": 0.00011977582065652522,
+      "loss": 0.5019,
+      "step": 2511
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 0.4230775535106659,
+      "learning_rate": 0.00011974379503602884,
+      "loss": 0.4346,
+      "step": 2512
+    },
+    {
+      "epoch": 0.80416,
+      "grad_norm": 0.4837520122528076,
+      "learning_rate": 0.00011971176941553243,
+      "loss": 0.5638,
+      "step": 2513
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 0.40232163667678833,
+      "learning_rate": 0.00011967974379503604,
+      "loss": 0.5793,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 0.47762757539749146,
+      "learning_rate": 0.00011964771817453964,
+      "loss": 0.6258,
+      "step": 2515
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 0.548795759677887,
+      "learning_rate": 0.00011961569255404323,
+      "loss": 0.7191,
+      "step": 2516
+    },
+    {
+      "epoch": 0.80544,
+      "grad_norm": 0.4455724358558655,
+      "learning_rate": 0.00011958366693354684,
+      "loss": 0.5447,
+      "step": 2517
+    },
+    {
+      "epoch": 0.80576,
+      "grad_norm": 0.5680160522460938,
+      "learning_rate": 0.00011955164131305046,
+      "loss": 0.5823,
+      "step": 2518
+    },
+    {
+      "epoch": 0.80608,
+      "grad_norm": 0.523038387298584,
+      "learning_rate": 0.00011951961569255406,
+      "loss": 0.613,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 0.4193526804447174,
+      "learning_rate": 0.00011948759007205765,
+      "loss": 0.4976,
+      "step": 2520
+    },
+    {
+      "epoch": 0.80672,
+      "grad_norm": 0.4190564453601837,
+      "learning_rate": 0.00011945556445156126,
+      "loss": 0.5483,
+      "step": 2521
+    },
+    {
+      "epoch": 0.80704,
+      "grad_norm": 0.5782697796821594,
+      "learning_rate": 0.00011942353883106485,
+      "loss": 0.5283,
+      "step": 2522
+    },
+    {
+      "epoch": 0.80736,
+      "grad_norm": 0.40420618653297424,
+      "learning_rate": 0.00011939151321056845,
+      "loss": 0.4424,
+      "step": 2523
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 0.5400856733322144,
+      "learning_rate": 0.00011935948759007207,
+      "loss": 0.5762,
+      "step": 2524
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 0.6003260612487793,
+      "learning_rate": 0.00011932746196957568,
+      "loss": 0.6427,
+      "step": 2525
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 0.5052765607833862,
+      "learning_rate": 0.00011929543634907927,
+      "loss": 0.6654,
+      "step": 2526
+    },
+    {
+      "epoch": 0.80864,
+      "grad_norm": 0.46488451957702637,
+      "learning_rate": 0.00011926341072858287,
+      "loss": 0.5783,
+      "step": 2527
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 0.5753117799758911,
+      "learning_rate": 0.00011923138510808646,
+      "loss": 0.6762,
+      "step": 2528
+    },
+    {
+      "epoch": 0.80928,
+      "grad_norm": 0.989375650882721,
+      "learning_rate": 0.00011919935948759007,
+      "loss": 0.6018,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 0.4945506453514099,
+      "learning_rate": 0.00011916733386709369,
+      "loss": 0.6481,
+      "step": 2530
+    },
+    {
+      "epoch": 0.80992,
+      "grad_norm": 0.4688679277896881,
+      "learning_rate": 0.00011913530824659729,
+      "loss": 0.499,
+      "step": 2531
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 0.5091390013694763,
+      "learning_rate": 0.00011910328262610088,
+      "loss": 0.5365,
+      "step": 2532
+    },
+    {
+      "epoch": 0.81056,
+      "grad_norm": 0.47598588466644287,
+      "learning_rate": 0.00011907125700560449,
+      "loss": 0.6279,
+      "step": 2533
+    },
+    {
+      "epoch": 0.81088,
+      "grad_norm": 0.44052502512931824,
+      "learning_rate": 0.00011903923138510809,
+      "loss": 0.436,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 0.40121087431907654,
+      "learning_rate": 0.00011900720576461168,
+      "loss": 0.6077,
+      "step": 2535
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 0.5621079206466675,
+      "learning_rate": 0.0001189751801441153,
+      "loss": 0.5166,
+      "step": 2536
+    },
+    {
+      "epoch": 0.81184,
+      "grad_norm": 0.4539152979850769,
+      "learning_rate": 0.0001189431545236189,
+      "loss": 0.6324,
+      "step": 2537
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 0.4092644453048706,
+      "learning_rate": 0.00011891112890312251,
+      "loss": 0.5349,
+      "step": 2538
+    },
+    {
+      "epoch": 0.81248,
+      "grad_norm": 0.4859711229801178,
+      "learning_rate": 0.0001188791032826261,
+      "loss": 0.5289,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.3599519729614258,
+      "learning_rate": 0.0001188470776621297,
+      "loss": 0.4256,
+      "step": 2540
+    },
+    {
+      "epoch": 0.81312,
+      "grad_norm": 0.7508904933929443,
+      "learning_rate": 0.0001188150520416333,
+      "loss": 0.7651,
+      "step": 2541
+    },
+    {
+      "epoch": 0.81344,
+      "grad_norm": 0.3674793243408203,
+      "learning_rate": 0.00011878302642113692,
+      "loss": 0.4797,
+      "step": 2542
+    },
+    {
+      "epoch": 0.81376,
+      "grad_norm": 0.40884310007095337,
+      "learning_rate": 0.00011875100080064052,
+      "loss": 0.511,
+      "step": 2543
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 0.4544715881347656,
+      "learning_rate": 0.00011871897518014413,
+      "loss": 0.5577,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 0.3420828878879547,
+      "learning_rate": 0.00011868694955964772,
+      "loss": 0.4498,
+      "step": 2545
+    },
+    {
+      "epoch": 0.81472,
+      "grad_norm": 0.4322601556777954,
+      "learning_rate": 0.00011865492393915132,
+      "loss": 0.6304,
+      "step": 2546
+    },
+    {
+      "epoch": 0.81504,
+      "grad_norm": 0.40280285477638245,
+      "learning_rate": 0.00011862289831865494,
+      "loss": 0.5159,
+      "step": 2547
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 0.4443550407886505,
+      "learning_rate": 0.00011859087269815854,
+      "loss": 0.5558,
+      "step": 2548
+    },
+    {
+      "epoch": 0.81568,
+      "grad_norm": 0.4436197280883789,
+      "learning_rate": 0.00011855884707766214,
+      "loss": 0.6019,
+      "step": 2549
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.40161630511283875,
+      "learning_rate": 0.00011852682145716574,
+      "loss": 0.4925,
+      "step": 2550
+    },
+    {
+      "epoch": 0.81632,
+      "grad_norm": 0.4296385645866394,
+      "learning_rate": 0.00011849479583666933,
+      "loss": 0.4026,
+      "step": 2551
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 0.43064355850219727,
+      "learning_rate": 0.00011846277021617294,
+      "loss": 0.5305,
+      "step": 2552
+    },
+    {
+      "epoch": 0.81696,
+      "grad_norm": 0.45453739166259766,
+      "learning_rate": 0.00011843074459567656,
+      "loss": 0.5283,
+      "step": 2553
+    },
+    {
+      "epoch": 0.81728,
+      "grad_norm": 0.47445422410964966,
+      "learning_rate": 0.00011839871897518016,
+      "loss": 0.7073,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 0.43052223324775696,
+      "learning_rate": 0.00011836669335468375,
+      "loss": 0.3612,
+      "step": 2555
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 0.49081093072891235,
+      "learning_rate": 0.00011833466773418736,
+      "loss": 0.6982,
+      "step": 2556
+    },
+    {
+      "epoch": 0.81824,
+      "grad_norm": 0.5123931169509888,
+      "learning_rate": 0.00011830264211369095,
+      "loss": 0.5369,
+      "step": 2557
+    },
+    {
+      "epoch": 0.81856,
+      "grad_norm": 0.4569171071052551,
+      "learning_rate": 0.00011827061649319455,
+      "loss": 0.4626,
+      "step": 2558
+    },
+    {
+      "epoch": 0.81888,
+      "grad_norm": 0.4077035188674927,
+      "learning_rate": 0.00011823859087269817,
+      "loss": 0.4877,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.483029305934906,
+      "learning_rate": 0.00011820656525220177,
+      "loss": 0.5587,
+      "step": 2560
+    },
+    {
+      "epoch": 0.81952,
+      "grad_norm": 0.509975254535675,
+      "learning_rate": 0.00011817453963170537,
+      "loss": 0.642,
+      "step": 2561
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 0.5269249081611633,
+      "learning_rate": 0.00011814251401120897,
+      "loss": 0.6096,
+      "step": 2562
+    },
+    {
+      "epoch": 0.82016,
+      "grad_norm": 0.5224083662033081,
+      "learning_rate": 0.00011811048839071258,
+      "loss": 0.5616,
+      "step": 2563
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 0.45314955711364746,
+      "learning_rate": 0.00011807846277021617,
+      "loss": 0.5788,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 0.4877423346042633,
+      "learning_rate": 0.00011804643714971979,
+      "loss": 0.5891,
+      "step": 2565
+    },
+    {
+      "epoch": 0.82112,
+      "grad_norm": 0.42431941628456116,
+      "learning_rate": 0.00011801441152922339,
+      "loss": 0.5821,
+      "step": 2566
+    },
+    {
+      "epoch": 0.82144,
+      "grad_norm": 0.6022917032241821,
+      "learning_rate": 0.000117982385908727,
+      "loss": 0.6599,
+      "step": 2567
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 0.5362722873687744,
+      "learning_rate": 0.00011795036028823059,
+      "loss": 0.523,
+      "step": 2568
+    },
+    {
+      "epoch": 0.82208,
+      "grad_norm": 0.43076518177986145,
+      "learning_rate": 0.00011791833466773419,
+      "loss": 0.5612,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 0.5106619000434875,
+      "learning_rate": 0.00011788630904723778,
+      "loss": 0.5778,
+      "step": 2570
+    },
+    {
+      "epoch": 0.82272,
+      "grad_norm": 0.47162723541259766,
+      "learning_rate": 0.00011785428342674141,
+      "loss": 0.5295,
+      "step": 2571
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 0.45818406343460083,
+      "learning_rate": 0.000117822257806245,
+      "loss": 0.5217,
+      "step": 2572
+    },
+    {
+      "epoch": 0.82336,
+      "grad_norm": 0.5135090947151184,
+      "learning_rate": 0.00011779023218574861,
+      "loss": 0.7086,
+      "step": 2573
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 0.4599963128566742,
+      "learning_rate": 0.0001177582065652522,
+      "loss": 0.5993,
+      "step": 2574
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.5403341054916382,
+      "learning_rate": 0.0001177261809447558,
+      "loss": 0.4909,
+      "step": 2575
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 0.4658939838409424,
+      "learning_rate": 0.0001176941553242594,
+      "loss": 0.5049,
+      "step": 2576
+    },
+    {
+      "epoch": 0.82464,
+      "grad_norm": 0.4838472008705139,
+      "learning_rate": 0.00011766212970376303,
+      "loss": 0.5244,
+      "step": 2577
+    },
+    {
+      "epoch": 0.82496,
+      "grad_norm": 0.43183133006095886,
+      "learning_rate": 0.00011763010408326662,
+      "loss": 0.5492,
+      "step": 2578
+    },
+    {
+      "epoch": 0.82528,
+      "grad_norm": 0.4954543113708496,
+      "learning_rate": 0.00011759807846277022,
+      "loss": 0.5452,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.5225180387496948,
+      "learning_rate": 0.00011756605284227382,
+      "loss": 0.6652,
+      "step": 2580
+    },
+    {
+      "epoch": 0.82592,
+      "grad_norm": 0.47078919410705566,
+      "learning_rate": 0.00011753402722177742,
+      "loss": 0.7561,
+      "step": 2581
+    },
+    {
+      "epoch": 0.82624,
+      "grad_norm": 0.43209174275398254,
+      "learning_rate": 0.00011750200160128103,
+      "loss": 0.4799,
+      "step": 2582
+    },
+    {
+      "epoch": 0.82656,
+      "grad_norm": 0.49573853611946106,
+      "learning_rate": 0.00011746997598078464,
+      "loss": 0.5131,
+      "step": 2583
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 0.47215840220451355,
+      "learning_rate": 0.00011743795036028823,
+      "loss": 0.584,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 0.404495507478714,
+      "learning_rate": 0.00011740592473979184,
+      "loss": 0.5126,
+      "step": 2585
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 0.4490559995174408,
+      "learning_rate": 0.00011737389911929544,
+      "loss": 0.4686,
+      "step": 2586
+    },
+    {
+      "epoch": 0.82784,
+      "grad_norm": 0.4486583471298218,
+      "learning_rate": 0.00011734187349879904,
+      "loss": 0.5119,
+      "step": 2587
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 0.5352327823638916,
+      "learning_rate": 0.00011730984787830264,
+      "loss": 0.6843,
+      "step": 2588
+    },
+    {
+      "epoch": 0.82848,
+      "grad_norm": 0.4671883285045624,
+      "learning_rate": 0.00011727782225780626,
+      "loss": 0.6492,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 0.46163955330848694,
+      "learning_rate": 0.00011724579663730986,
+      "loss": 0.5872,
+      "step": 2590
+    },
+    {
+      "epoch": 0.82912,
+      "grad_norm": 0.5733055472373962,
+      "learning_rate": 0.00011721377101681345,
+      "loss": 0.6992,
+      "step": 2591
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 0.5973170399665833,
+      "learning_rate": 0.00011718174539631706,
+      "loss": 0.5987,
+      "step": 2592
+    },
+    {
+      "epoch": 0.82976,
+      "grad_norm": 0.5273131132125854,
+      "learning_rate": 0.00011714971977582065,
+      "loss": 0.5569,
+      "step": 2593
+    },
+    {
+      "epoch": 0.83008,
+      "grad_norm": 0.4874577522277832,
+      "learning_rate": 0.00011711769415532427,
+      "loss": 0.6059,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 0.4798988699913025,
+      "learning_rate": 0.00011708566853482787,
+      "loss": 0.5039,
+      "step": 2595
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 0.4751587212085724,
+      "learning_rate": 0.00011705364291433148,
+      "loss": 0.6465,
+      "step": 2596
+    },
+    {
+      "epoch": 0.83104,
+      "grad_norm": 0.509245753288269,
+      "learning_rate": 0.00011702161729383507,
+      "loss": 0.5462,
+      "step": 2597
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 0.44738683104515076,
+      "learning_rate": 0.00011698959167333867,
+      "loss": 0.5605,
+      "step": 2598
+    },
+    {
+      "epoch": 0.83168,
+      "grad_norm": 0.5128889083862305,
+      "learning_rate": 0.00011695756605284227,
+      "loss": 0.5431,
+      "step": 2599
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.44277453422546387,
+      "learning_rate": 0.0001169255404323459,
+      "loss": 0.5217,
+      "step": 2600
+    },
+    {
+      "epoch": 0.83232,
+      "grad_norm": 0.5297160744667053,
+      "learning_rate": 0.00011689351481184949,
+      "loss": 0.6491,
+      "step": 2601
+    },
+    {
+      "epoch": 0.83264,
+      "grad_norm": 0.5843796133995056,
+      "learning_rate": 0.0001168614891913531,
+      "loss": 0.6897,
+      "step": 2602
+    },
+    {
+      "epoch": 0.83296,
+      "grad_norm": 0.4996921718120575,
+      "learning_rate": 0.00011682946357085668,
+      "loss": 0.5288,
+      "step": 2603
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 0.4643668830394745,
+      "learning_rate": 0.00011679743795036029,
+      "loss": 0.5859,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 0.5779374241828918,
+      "learning_rate": 0.0001167654123298639,
+      "loss": 0.6292,
+      "step": 2605
+    },
+    {
+      "epoch": 0.83392,
+      "grad_norm": 0.4328746199607849,
+      "learning_rate": 0.00011673338670936751,
+      "loss": 0.5547,
+      "step": 2606
+    },
+    {
+      "epoch": 0.83424,
+      "grad_norm": 0.4922213852405548,
+      "learning_rate": 0.0001167013610888711,
+      "loss": 0.5525,
+      "step": 2607
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 0.44271114468574524,
+      "learning_rate": 0.00011666933546837471,
+      "loss": 0.604,
+      "step": 2608
+    },
+    {
+      "epoch": 0.83488,
+      "grad_norm": 0.5586934685707092,
+      "learning_rate": 0.00011663730984787831,
+      "loss": 0.584,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.48603761196136475,
+      "learning_rate": 0.0001166052842273819,
+      "loss": 0.5591,
+      "step": 2610
+    },
+    {
+      "epoch": 0.83552,
+      "grad_norm": 0.41124701499938965,
+      "learning_rate": 0.00011657325860688551,
+      "loss": 0.6128,
+      "step": 2611
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 0.3622615337371826,
+      "learning_rate": 0.00011654123298638913,
+      "loss": 0.5129,
+      "step": 2612
+    },
+    {
+      "epoch": 0.83616,
+      "grad_norm": 0.40537506341934204,
+      "learning_rate": 0.00011650920736589272,
+      "loss": 0.51,
+      "step": 2613
+    },
+    {
+      "epoch": 0.83648,
+      "grad_norm": 0.5575379133224487,
+      "learning_rate": 0.00011647718174539632,
+      "loss": 0.6152,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 0.5207103490829468,
+      "learning_rate": 0.00011644515612489993,
+      "loss": 0.5543,
+      "step": 2615
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 0.3939136266708374,
+      "learning_rate": 0.00011641313050440352,
+      "loss": 0.452,
+      "step": 2616
+    },
+    {
+      "epoch": 0.83744,
+      "grad_norm": 0.48216161131858826,
+      "learning_rate": 0.00011638110488390712,
+      "loss": 0.5691,
+      "step": 2617
+    },
+    {
+      "epoch": 0.83776,
+      "grad_norm": 0.5009711980819702,
+      "learning_rate": 0.00011634907926341074,
+      "loss": 0.6283,
+      "step": 2618
+    },
+    {
+      "epoch": 0.83808,
+      "grad_norm": 0.4805917739868164,
+      "learning_rate": 0.00011631705364291435,
+      "loss": 0.6188,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.36027631163597107,
+      "learning_rate": 0.00011628502802241794,
+      "loss": 0.462,
+      "step": 2620
+    },
+    {
+      "epoch": 0.83872,
+      "grad_norm": 0.46544021368026733,
+      "learning_rate": 0.00011625300240192154,
+      "loss": 0.58,
+      "step": 2621
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 0.4826143682003021,
+      "learning_rate": 0.00011622097678142513,
+      "loss": 0.7107,
+      "step": 2622
+    },
+    {
+      "epoch": 0.83936,
+      "grad_norm": 0.5167753100395203,
+      "learning_rate": 0.00011618895116092874,
+      "loss": 0.7418,
+      "step": 2623
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 0.4048837721347809,
+      "learning_rate": 0.00011615692554043236,
+      "loss": 0.5087,
+      "step": 2624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5134204626083374,
+      "learning_rate": 0.00011612489991993596,
+      "loss": 0.6588,
+      "step": 2625
+    },
+    {
+      "epoch": 0.84032,
+      "grad_norm": 0.4732988178730011,
+      "learning_rate": 0.00011609287429943955,
+      "loss": 0.661,
+      "step": 2626
+    },
+    {
+      "epoch": 0.84064,
+      "grad_norm": 0.5023718476295471,
+      "learning_rate": 0.00011606084867894316,
+      "loss": 0.5018,
+      "step": 2627
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 0.45160111784935,
+      "learning_rate": 0.00011602882305844675,
+      "loss": 0.6671,
+      "step": 2628
+    },
+    {
+      "epoch": 0.84128,
+      "grad_norm": 0.5643427968025208,
+      "learning_rate": 0.00011599679743795035,
+      "loss": 0.7239,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 0.44439825415611267,
+      "learning_rate": 0.00011596477181745397,
+      "loss": 0.5268,
+      "step": 2630
+    },
+    {
+      "epoch": 0.84192,
+      "grad_norm": 0.4833722412586212,
+      "learning_rate": 0.00011593274619695758,
+      "loss": 0.5517,
+      "step": 2631
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 0.4305606186389923,
+      "learning_rate": 0.00011590072057646117,
+      "loss": 0.5134,
+      "step": 2632
+    },
+    {
+      "epoch": 0.84256,
+      "grad_norm": 0.5362460613250732,
+      "learning_rate": 0.00011586869495596477,
+      "loss": 0.6467,
+      "step": 2633
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 0.5484098792076111,
+      "learning_rate": 0.00011583666933546838,
+      "loss": 0.5227,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 0.545275866985321,
+      "learning_rate": 0.00011580464371497197,
+      "loss": 0.768,
+      "step": 2635
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 0.49064135551452637,
+      "learning_rate": 0.00011577261809447559,
+      "loss": 0.626,
+      "step": 2636
+    },
+    {
+      "epoch": 0.84384,
+      "grad_norm": 0.4276551902294159,
+      "learning_rate": 0.00011574059247397919,
+      "loss": 0.5301,
+      "step": 2637
+    },
+    {
+      "epoch": 0.84416,
+      "grad_norm": 0.5904160737991333,
+      "learning_rate": 0.0001157085668534828,
+      "loss": 0.6724,
+      "step": 2638
+    },
+    {
+      "epoch": 0.84448,
+      "grad_norm": 1.1107900142669678,
+      "learning_rate": 0.00011567654123298639,
+      "loss": 0.6576,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.5705112814903259,
+      "learning_rate": 0.00011564451561248999,
+      "loss": 0.5908,
+      "step": 2640
+    },
+    {
+      "epoch": 0.84512,
+      "grad_norm": 0.5342549681663513,
+      "learning_rate": 0.00011561248999199361,
+      "loss": 0.7481,
+      "step": 2641
+    },
+    {
+      "epoch": 0.84544,
+      "grad_norm": 0.48843875527381897,
+      "learning_rate": 0.00011558046437149722,
+      "loss": 0.6046,
+      "step": 2642
+    },
+    {
+      "epoch": 0.84576,
+      "grad_norm": 0.4522450268268585,
+      "learning_rate": 0.00011554843875100081,
+      "loss": 0.6946,
+      "step": 2643
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 0.4567502737045288,
+      "learning_rate": 0.00011551641313050441,
+      "loss": 0.5805,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 0.4489455819129944,
+      "learning_rate": 0.000115484387510008,
+      "loss": 0.5692,
+      "step": 2645
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 0.5014421939849854,
+      "learning_rate": 0.00011545236188951161,
+      "loss": 0.4663,
+      "step": 2646
+    },
+    {
+      "epoch": 0.84704,
+      "grad_norm": 0.49470055103302,
+      "learning_rate": 0.00011542033626901523,
+      "loss": 0.555,
+      "step": 2647
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 0.39371562004089355,
+      "learning_rate": 0.00011538831064851883,
+      "loss": 0.484,
+      "step": 2648
+    },
+    {
+      "epoch": 0.84768,
+      "grad_norm": 0.47203773260116577,
+      "learning_rate": 0.00011535628502802242,
+      "loss": 0.5325,
+      "step": 2649
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.5638809204101562,
+      "learning_rate": 0.00011532425940752603,
+      "loss": 0.7671,
+      "step": 2650
+    },
+    {
+      "epoch": 0.84832,
+      "grad_norm": 0.510729968547821,
+      "learning_rate": 0.00011529223378702962,
+      "loss": 0.6251,
+      "step": 2651
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 0.47444963455200195,
+      "learning_rate": 0.00011526020816653322,
+      "loss": 0.5814,
+      "step": 2652
+    },
+    {
+      "epoch": 0.84896,
+      "grad_norm": 0.5941531658172607,
+      "learning_rate": 0.00011522818254603684,
+      "loss": 0.5235,
+      "step": 2653
+    },
+    {
+      "epoch": 0.84928,
+      "grad_norm": 0.46168580651283264,
+      "learning_rate": 0.00011519615692554045,
+      "loss": 0.4831,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 0.4080221354961395,
+      "learning_rate": 0.00011516413130504404,
+      "loss": 0.4589,
+      "step": 2655
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 0.389886736869812,
+      "learning_rate": 0.00011513210568454764,
+      "loss": 0.4927,
+      "step": 2656
+    },
+    {
+      "epoch": 0.85024,
+      "grad_norm": 0.5023879408836365,
+      "learning_rate": 0.00011510008006405125,
+      "loss": 0.5494,
+      "step": 2657
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 0.33860617876052856,
+      "learning_rate": 0.00011506805444355484,
+      "loss": 0.4679,
+      "step": 2658
+    },
+    {
+      "epoch": 0.85088,
+      "grad_norm": 0.5367802977561951,
+      "learning_rate": 0.00011503602882305846,
+      "loss": 0.4694,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.371481716632843,
+      "learning_rate": 0.00011500400320256206,
+      "loss": 0.5342,
+      "step": 2660
+    },
+    {
+      "epoch": 0.85152,
+      "grad_norm": 0.5274410247802734,
+      "learning_rate": 0.00011497197758206567,
+      "loss": 0.6667,
+      "step": 2661
+    },
+    {
+      "epoch": 0.85184,
+      "grad_norm": 0.47757643461227417,
+      "learning_rate": 0.00011493995196156926,
+      "loss": 0.5605,
+      "step": 2662
+    },
+    {
+      "epoch": 0.85216,
+      "grad_norm": 0.5006943345069885,
+      "learning_rate": 0.00011490792634107286,
+      "loss": 0.6509,
+      "step": 2663
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 0.43314650654792786,
+      "learning_rate": 0.00011487590072057645,
+      "loss": 0.516,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 0.5509462356567383,
+      "learning_rate": 0.00011484387510008007,
+      "loss": 0.6095,
+      "step": 2665
+    },
+    {
+      "epoch": 0.85312,
+      "grad_norm": 0.5219637751579285,
+      "learning_rate": 0.00011481184947958368,
+      "loss": 0.6273,
+      "step": 2666
+    },
+    {
+      "epoch": 0.85344,
+      "grad_norm": 0.46904855966567993,
+      "learning_rate": 0.00011477982385908728,
+      "loss": 0.5981,
+      "step": 2667
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 0.5802997946739197,
+      "learning_rate": 0.00011474779823859087,
+      "loss": 0.6021,
+      "step": 2668
+    },
+    {
+      "epoch": 0.85408,
+      "grad_norm": 0.4416864514350891,
+      "learning_rate": 0.00011471577261809448,
+      "loss": 0.5844,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.5157893300056458,
+      "learning_rate": 0.00011468374699759807,
+      "loss": 0.6447,
+      "step": 2670
+    },
+    {
+      "epoch": 0.85472,
+      "grad_norm": 0.47831296920776367,
+      "learning_rate": 0.0001146517213771017,
+      "loss": 0.5188,
+      "step": 2671
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 0.5330260992050171,
+      "learning_rate": 0.00011461969575660529,
+      "loss": 0.6737,
+      "step": 2672
+    },
+    {
+      "epoch": 0.85536,
+      "grad_norm": 0.5019591450691223,
+      "learning_rate": 0.0001145876701361089,
+      "loss": 0.636,
+      "step": 2673
+    },
+    {
+      "epoch": 0.85568,
+      "grad_norm": 0.49352139234542847,
+      "learning_rate": 0.00011455564451561249,
+      "loss": 0.5654,
+      "step": 2674
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.5157530307769775,
+      "learning_rate": 0.00011452361889511609,
+      "loss": 0.6378,
+      "step": 2675
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 0.5100753307342529,
+      "learning_rate": 0.0001144915932746197,
+      "loss": 0.7374,
+      "step": 2676
+    },
+    {
+      "epoch": 0.85664,
+      "grad_norm": 0.4924938380718231,
+      "learning_rate": 0.00011445956765412332,
+      "loss": 0.5068,
+      "step": 2677
+    },
+    {
+      "epoch": 0.85696,
+      "grad_norm": 0.3869999349117279,
+      "learning_rate": 0.0001144275420336269,
+      "loss": 0.5065,
+      "step": 2678
+    },
+    {
+      "epoch": 0.85728,
+      "grad_norm": 0.4689071774482727,
+      "learning_rate": 0.00011439551641313051,
+      "loss": 0.627,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.4585507810115814,
+      "learning_rate": 0.00011436349079263412,
+      "loss": 0.5153,
+      "step": 2680
+    },
+    {
+      "epoch": 0.85792,
+      "grad_norm": 0.5436362028121948,
+      "learning_rate": 0.00011433146517213771,
+      "loss": 0.6833,
+      "step": 2681
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 0.49875640869140625,
+      "learning_rate": 0.00011429943955164131,
+      "loss": 0.5935,
+      "step": 2682
+    },
+    {
+      "epoch": 0.85856,
+      "grad_norm": 0.5966513752937317,
+      "learning_rate": 0.00011426741393114493,
+      "loss": 0.6178,
+      "step": 2683
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 0.49361687898635864,
+      "learning_rate": 0.00011423538831064852,
+      "loss": 0.5726,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 0.5657501220703125,
+      "learning_rate": 0.00011420336269015213,
+      "loss": 0.6583,
+      "step": 2685
+    },
+    {
+      "epoch": 0.85952,
+      "grad_norm": 0.4523167610168457,
+      "learning_rate": 0.00011417133706965573,
+      "loss": 0.6198,
+      "step": 2686
+    },
+    {
+      "epoch": 0.85984,
+      "grad_norm": 0.6706781387329102,
+      "learning_rate": 0.00011413931144915932,
+      "loss": 0.8699,
+      "step": 2687
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 0.5806979537010193,
+      "learning_rate": 0.00011410728582866294,
+      "loss": 0.7102,
+      "step": 2688
+    },
+    {
+      "epoch": 0.86048,
+      "grad_norm": 0.44189244508743286,
+      "learning_rate": 0.00011407526020816655,
+      "loss": 0.6103,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 0.43773627281188965,
+      "learning_rate": 0.00011404323458767015,
+      "loss": 0.5738,
+      "step": 2690
+    },
+    {
+      "epoch": 0.86112,
+      "grad_norm": 0.43098342418670654,
+      "learning_rate": 0.00011401120896717374,
+      "loss": 0.5399,
+      "step": 2691
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 0.45769962668418884,
+      "learning_rate": 0.00011397918334667735,
+      "loss": 0.4945,
+      "step": 2692
+    },
+    {
+      "epoch": 0.86176,
+      "grad_norm": 0.5592633485794067,
+      "learning_rate": 0.00011394715772618094,
+      "loss": 0.6164,
+      "step": 2693
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 0.46443822979927063,
+      "learning_rate": 0.00011391513210568457,
+      "loss": 0.559,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 0.4137526750564575,
+      "learning_rate": 0.00011388310648518816,
+      "loss": 0.5321,
+      "step": 2695
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 0.5033775568008423,
+      "learning_rate": 0.00011385108086469177,
+      "loss": 0.5446,
+      "step": 2696
+    },
+    {
+      "epoch": 0.86304,
+      "grad_norm": 0.44829392433166504,
+      "learning_rate": 0.00011381905524419536,
+      "loss": 0.528,
+      "step": 2697
+    },
+    {
+      "epoch": 0.86336,
+      "grad_norm": 0.49028539657592773,
+      "learning_rate": 0.00011378702962369896,
+      "loss": 0.6045,
+      "step": 2698
+    },
+    {
+      "epoch": 0.86368,
+      "grad_norm": 0.4432521164417267,
+      "learning_rate": 0.00011375500400320255,
+      "loss": 0.5095,
+      "step": 2699
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.47212648391723633,
+      "learning_rate": 0.00011372297838270618,
+      "loss": 0.6144,
+      "step": 2700
+    },
+    {
+      "epoch": 0.86432,
+      "grad_norm": 0.48464998602867126,
+      "learning_rate": 0.00011369095276220978,
+      "loss": 0.5371,
+      "step": 2701
+    },
+    {
+      "epoch": 0.86464,
+      "grad_norm": 0.4802500307559967,
+      "learning_rate": 0.00011365892714171338,
+      "loss": 0.4414,
+      "step": 2702
+    },
+    {
+      "epoch": 0.86496,
+      "grad_norm": 0.40302029252052307,
+      "learning_rate": 0.00011362690152121697,
+      "loss": 0.4521,
+      "step": 2703
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 0.5393025279045105,
+      "learning_rate": 0.00011359487590072058,
+      "loss": 0.4435,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 0.4327370226383209,
+      "learning_rate": 0.00011356285028022418,
+      "loss": 0.5476,
+      "step": 2705
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 0.48276886343955994,
+      "learning_rate": 0.0001135308246597278,
+      "loss": 0.4698,
+      "step": 2706
+    },
+    {
+      "epoch": 0.86624,
+      "grad_norm": 0.4413887560367584,
+      "learning_rate": 0.00011349879903923139,
+      "loss": 0.6004,
+      "step": 2707
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 0.49297529458999634,
+      "learning_rate": 0.000113466773418735,
+      "loss": 0.5905,
+      "step": 2708
+    },
+    {
+      "epoch": 0.86688,
+      "grad_norm": 0.5595799088478088,
+      "learning_rate": 0.0001134347477982386,
+      "loss": 0.7043,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 0.4480595886707306,
+      "learning_rate": 0.00011340272217774219,
+      "loss": 0.5558,
+      "step": 2710
+    },
+    {
+      "epoch": 0.86752,
+      "grad_norm": 0.41989707946777344,
+      "learning_rate": 0.0001133706965572458,
+      "loss": 0.576,
+      "step": 2711
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 0.43645966053009033,
+      "learning_rate": 0.00011333867093674941,
+      "loss": 0.5645,
+      "step": 2712
+    },
+    {
+      "epoch": 0.86816,
+      "grad_norm": 0.4377037286758423,
+      "learning_rate": 0.00011330664531625302,
+      "loss": 0.5973,
+      "step": 2713
+    },
+    {
+      "epoch": 0.86848,
+      "grad_norm": 0.4731050729751587,
+      "learning_rate": 0.00011327461969575661,
+      "loss": 0.547,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 0.5003529191017151,
+      "learning_rate": 0.00011324259407526021,
+      "loss": 0.5592,
+      "step": 2715
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 0.4640887379646301,
+      "learning_rate": 0.0001132105684547638,
+      "loss": 0.7179,
+      "step": 2716
+    },
+    {
+      "epoch": 0.86944,
+      "grad_norm": 0.5412507653236389,
+      "learning_rate": 0.00011317854283426741,
+      "loss": 0.6181,
+      "step": 2717
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 0.5246358513832092,
+      "learning_rate": 0.00011314651721377103,
+      "loss": 0.6154,
+      "step": 2718
+    },
+    {
+      "epoch": 0.87008,
+      "grad_norm": 0.5247026085853577,
+      "learning_rate": 0.00011311449159327463,
+      "loss": 0.6095,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.4215399920940399,
+      "learning_rate": 0.00011308246597277823,
+      "loss": 0.5079,
+      "step": 2720
+    },
+    {
+      "epoch": 0.87072,
+      "grad_norm": 0.4981729984283447,
+      "learning_rate": 0.00011305044035228183,
+      "loss": 0.5286,
+      "step": 2721
+    },
+    {
+      "epoch": 0.87104,
+      "grad_norm": 0.4824583828449249,
+      "learning_rate": 0.00011301841473178542,
+      "loss": 0.6277,
+      "step": 2722
+    },
+    {
+      "epoch": 0.87136,
+      "grad_norm": 0.4891294538974762,
+      "learning_rate": 0.00011298638911128903,
+      "loss": 0.5893,
+      "step": 2723
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 0.42053404450416565,
+      "learning_rate": 0.00011295436349079264,
+      "loss": 0.5393,
+      "step": 2724
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 0.5282824635505676,
+      "learning_rate": 0.00011292233787029625,
+      "loss": 0.7794,
+      "step": 2725
+    },
+    {
+      "epoch": 0.87232,
+      "grad_norm": 0.510481059551239,
+      "learning_rate": 0.00011289031224979984,
+      "loss": 0.7445,
+      "step": 2726
+    },
+    {
+      "epoch": 0.87264,
+      "grad_norm": 0.4764452278614044,
+      "learning_rate": 0.00011285828662930344,
+      "loss": 0.5211,
+      "step": 2727
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 0.5518248677253723,
+      "learning_rate": 0.00011282626100880705,
+      "loss": 0.5829,
+      "step": 2728
+    },
+    {
+      "epoch": 0.87328,
+      "grad_norm": 0.5229703783988953,
+      "learning_rate": 0.00011279423538831064,
+      "loss": 0.5695,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.42388972640037537,
+      "learning_rate": 0.00011276220976781426,
+      "loss": 0.5009,
+      "step": 2730
+    },
+    {
+      "epoch": 0.87392,
+      "grad_norm": 0.4291982352733612,
+      "learning_rate": 0.00011273018414731786,
+      "loss": 0.5879,
+      "step": 2731
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 0.5155397057533264,
+      "learning_rate": 0.00011269815852682147,
+      "loss": 0.6143,
+      "step": 2732
+    },
+    {
+      "epoch": 0.87456,
+      "grad_norm": 0.4703388214111328,
+      "learning_rate": 0.00011266613290632506,
+      "loss": 0.5044,
+      "step": 2733
+    },
+    {
+      "epoch": 0.87488,
+      "grad_norm": 0.5752137303352356,
+      "learning_rate": 0.00011263410728582866,
+      "loss": 0.6418,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 0.514747679233551,
+      "learning_rate": 0.00011260208166533228,
+      "loss": 0.6209,
+      "step": 2735
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 0.5046247243881226,
+      "learning_rate": 0.00011257005604483587,
+      "loss": 0.4723,
+      "step": 2736
+    },
+    {
+      "epoch": 0.87584,
+      "grad_norm": 0.46501097083091736,
+      "learning_rate": 0.00011253803042433948,
+      "loss": 0.6257,
+      "step": 2737
+    },
+    {
+      "epoch": 0.87616,
+      "grad_norm": 0.35089269280433655,
+      "learning_rate": 0.00011250600480384308,
+      "loss": 0.4596,
+      "step": 2738
+    },
+    {
+      "epoch": 0.87648,
+      "grad_norm": 0.34885352849960327,
+      "learning_rate": 0.00011247397918334667,
+      "loss": 0.5093,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.4819800853729248,
+      "learning_rate": 0.00011244195356285028,
+      "loss": 0.5178,
+      "step": 2740
+    },
+    {
+      "epoch": 0.87712,
+      "grad_norm": 0.4165039658546448,
+      "learning_rate": 0.0001124099279423539,
+      "loss": 0.5407,
+      "step": 2741
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 0.4613684415817261,
+      "learning_rate": 0.0001123779023218575,
+      "loss": 0.5677,
+      "step": 2742
+    },
+    {
+      "epoch": 0.87776,
+      "grad_norm": 0.5437636971473694,
+      "learning_rate": 0.0001123458767013611,
+      "loss": 0.5385,
+      "step": 2743
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 0.44850608706474304,
+      "learning_rate": 0.0001123138510808647,
+      "loss": 0.3883,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 0.5939509272575378,
+      "learning_rate": 0.00011228182546036829,
+      "loss": 0.655,
+      "step": 2745
+    },
+    {
+      "epoch": 0.87872,
+      "grad_norm": 0.5534245371818542,
+      "learning_rate": 0.0001122497998398719,
+      "loss": 0.6454,
+      "step": 2746
+    },
+    {
+      "epoch": 0.87904,
+      "grad_norm": 0.5665821433067322,
+      "learning_rate": 0.00011221777421937551,
+      "loss": 0.5838,
+      "step": 2747
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 0.4773208796977997,
+      "learning_rate": 0.00011218574859887912,
+      "loss": 0.449,
+      "step": 2748
+    },
+    {
+      "epoch": 0.87968,
+      "grad_norm": 0.5110756158828735,
+      "learning_rate": 0.00011215372297838271,
+      "loss": 0.5573,
+      "step": 2749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43415164947509766,
+      "learning_rate": 0.00011212169735788631,
+      "loss": 0.5154,
+      "step": 2750
+    },
+    {
+      "epoch": 0.88032,
+      "grad_norm": 0.4802980422973633,
+      "learning_rate": 0.00011208967173738992,
+      "loss": 0.5948,
+      "step": 2751
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 0.42461544275283813,
+      "learning_rate": 0.00011205764611689351,
+      "loss": 0.4193,
+      "step": 2752
+    },
+    {
+      "epoch": 0.88096,
+      "grad_norm": 0.4616337716579437,
+      "learning_rate": 0.00011202562049639713,
+      "loss": 0.4183,
+      "step": 2753
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 0.5223445296287537,
+      "learning_rate": 0.00011199359487590073,
+      "loss": 0.6107,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 0.45762428641319275,
+      "learning_rate": 0.00011196156925540432,
+      "loss": 0.5396,
+      "step": 2755
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 0.5913996696472168,
+      "learning_rate": 0.00011192954363490793,
+      "loss": 0.7051,
+      "step": 2756
+    },
+    {
+      "epoch": 0.88224,
+      "grad_norm": 0.5829991698265076,
+      "learning_rate": 0.00011189751801441153,
+      "loss": 0.6926,
+      "step": 2757
+    },
+    {
+      "epoch": 0.88256,
+      "grad_norm": 0.5960185527801514,
+      "learning_rate": 0.00011186549239391512,
+      "loss": 0.5628,
+      "step": 2758
+    },
+    {
+      "epoch": 0.88288,
+      "grad_norm": 0.5574018955230713,
+      "learning_rate": 0.00011183346677341874,
+      "loss": 0.5269,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.3998291492462158,
+      "learning_rate": 0.00011180144115292235,
+      "loss": 0.4939,
+      "step": 2760
+    },
+    {
+      "epoch": 0.88352,
+      "grad_norm": 0.45303770899772644,
+      "learning_rate": 0.00011176941553242595,
+      "loss": 0.515,
+      "step": 2761
+    },
+    {
+      "epoch": 0.88384,
+      "grad_norm": 0.5125203728675842,
+      "learning_rate": 0.00011173738991192954,
+      "loss": 0.6741,
+      "step": 2762
+    },
+    {
+      "epoch": 0.88416,
+      "grad_norm": 0.5005950331687927,
+      "learning_rate": 0.00011170536429143315,
+      "loss": 0.5515,
+      "step": 2763
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 0.4479672908782959,
+      "learning_rate": 0.00011167333867093674,
+      "loss": 0.5911,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 0.5130813717842102,
+      "learning_rate": 0.00011164131305044037,
+      "loss": 0.5712,
+      "step": 2765
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 0.4910069406032562,
+      "learning_rate": 0.00011160928742994396,
+      "loss": 0.4974,
+      "step": 2766
+    },
+    {
+      "epoch": 0.88544,
+      "grad_norm": 0.4807541072368622,
+      "learning_rate": 0.00011157726180944757,
+      "loss": 0.6424,
+      "step": 2767
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 0.5136614441871643,
+      "learning_rate": 0.00011154523618895116,
+      "loss": 0.7338,
+      "step": 2768
+    },
+    {
+      "epoch": 0.88608,
+      "grad_norm": 0.4988515377044678,
+      "learning_rate": 0.00011151321056845476,
+      "loss": 0.6051,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 0.48109325766563416,
+      "learning_rate": 0.00011148118494795835,
+      "loss": 0.5228,
+      "step": 2770
+    },
+    {
+      "epoch": 0.88672,
+      "grad_norm": 0.5012555122375488,
+      "learning_rate": 0.00011144915932746199,
+      "loss": 0.557,
+      "step": 2771
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 0.6116868257522583,
+      "learning_rate": 0.00011141713370696558,
+      "loss": 0.6668,
+      "step": 2772
+    },
+    {
+      "epoch": 0.88736,
+      "grad_norm": 0.6428093910217285,
+      "learning_rate": 0.00011138510808646918,
+      "loss": 0.6344,
+      "step": 2773
+    },
+    {
+      "epoch": 0.88768,
+      "grad_norm": 0.4118374288082123,
+      "learning_rate": 0.00011135308246597277,
+      "loss": 0.518,
+      "step": 2774
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.3722623288631439,
+      "learning_rate": 0.00011132105684547638,
+      "loss": 0.4598,
+      "step": 2775
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 0.49762818217277527,
+      "learning_rate": 0.00011128903122497998,
+      "loss": 0.633,
+      "step": 2776
+    },
+    {
+      "epoch": 0.88864,
+      "grad_norm": 0.5179993510246277,
+      "learning_rate": 0.0001112570056044836,
+      "loss": 0.6806,
+      "step": 2777
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 0.46002230048179626,
+      "learning_rate": 0.00011122497998398719,
+      "loss": 0.6191,
+      "step": 2778
+    },
+    {
+      "epoch": 0.88928,
+      "grad_norm": 0.5519804358482361,
+      "learning_rate": 0.0001111929543634908,
+      "loss": 0.5629,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.47665491700172424,
+      "learning_rate": 0.0001111609287429944,
+      "loss": 0.6155,
+      "step": 2780
+    },
+    {
+      "epoch": 0.88992,
+      "grad_norm": 0.4250374436378479,
+      "learning_rate": 0.000111128903122498,
+      "loss": 0.5022,
+      "step": 2781
+    },
+    {
+      "epoch": 0.89024,
+      "grad_norm": 0.5410628318786621,
+      "learning_rate": 0.00011109687750200161,
+      "loss": 0.6884,
+      "step": 2782
+    },
+    {
+      "epoch": 0.89056,
+      "grad_norm": 0.536748468875885,
+      "learning_rate": 0.00011106485188150522,
+      "loss": 0.5196,
+      "step": 2783
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 0.5975890159606934,
+      "learning_rate": 0.00011103282626100882,
+      "loss": 0.5164,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 0.4199933111667633,
+      "learning_rate": 0.00011100080064051241,
+      "loss": 0.5162,
+      "step": 2785
+    },
+    {
+      "epoch": 0.89152,
+      "grad_norm": 0.5856785774230957,
+      "learning_rate": 0.00011096877502001602,
+      "loss": 0.5593,
+      "step": 2786
+    },
+    {
+      "epoch": 0.89184,
+      "grad_norm": 0.5275327563285828,
+      "learning_rate": 0.00011093674939951961,
+      "loss": 0.5832,
+      "step": 2787
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 0.44140300154685974,
+      "learning_rate": 0.00011090472377902324,
+      "loss": 0.4383,
+      "step": 2788
+    },
+    {
+      "epoch": 0.89248,
+      "grad_norm": 0.5442470908164978,
+      "learning_rate": 0.00011087269815852683,
+      "loss": 0.6338,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.4676779806613922,
+      "learning_rate": 0.00011084067253803044,
+      "loss": 0.5703,
+      "step": 2790
+    },
+    {
+      "epoch": 0.89312,
+      "grad_norm": 0.6457350850105286,
+      "learning_rate": 0.00011080864691753403,
+      "loss": 0.6383,
+      "step": 2791
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 0.501268208026886,
+      "learning_rate": 0.00011077662129703763,
+      "loss": 0.5561,
+      "step": 2792
+    },
+    {
+      "epoch": 0.89376,
+      "grad_norm": 0.5912644267082214,
+      "learning_rate": 0.00011074459567654122,
+      "loss": 0.7406,
+      "step": 2793
+    },
+    {
+      "epoch": 0.89408,
+      "grad_norm": 0.5004026293754578,
+      "learning_rate": 0.00011071257005604486,
+      "loss": 0.5108,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 0.48542219400405884,
+      "learning_rate": 0.00011068054443554845,
+      "loss": 0.591,
+      "step": 2795
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 0.6127922534942627,
+      "learning_rate": 0.00011064851881505205,
+      "loss": 0.6864,
+      "step": 2796
+    },
+    {
+      "epoch": 0.89504,
+      "grad_norm": 0.4074430465698242,
+      "learning_rate": 0.00011061649319455564,
+      "loss": 0.5876,
+      "step": 2797
+    },
+    {
+      "epoch": 0.89536,
+      "grad_norm": 0.4714474678039551,
+      "learning_rate": 0.00011058446757405925,
+      "loss": 0.4604,
+      "step": 2798
+    },
+    {
+      "epoch": 0.89568,
+      "grad_norm": 0.4374367892742157,
+      "learning_rate": 0.00011055244195356285,
+      "loss": 0.5594,
+      "step": 2799
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.46386805176734924,
+      "learning_rate": 0.00011052041633306647,
+      "loss": 0.5348,
+      "step": 2800
+    },
+    {
+      "epoch": 0.89632,
+      "grad_norm": 0.5254485011100769,
+      "learning_rate": 0.00011048839071257006,
+      "loss": 0.5781,
+      "step": 2801
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 0.5686777234077454,
+      "learning_rate": 0.00011045636509207367,
+      "loss": 0.524,
+      "step": 2802
+    },
+    {
+      "epoch": 0.89696,
+      "grad_norm": 0.42373552918434143,
+      "learning_rate": 0.00011042433947157727,
+      "loss": 0.4861,
+      "step": 2803
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 0.6630972027778625,
+      "learning_rate": 0.00011039231385108086,
+      "loss": 0.7197,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 0.6042394638061523,
+      "learning_rate": 0.00011036028823058447,
+      "loss": 0.5272,
+      "step": 2805
+    },
+    {
+      "epoch": 0.89792,
+      "grad_norm": 0.5705429315567017,
+      "learning_rate": 0.00011032826261008809,
+      "loss": 0.604,
+      "step": 2806
+    },
+    {
+      "epoch": 0.89824,
+      "grad_norm": 0.5933979153633118,
+      "learning_rate": 0.00011029623698959169,
+      "loss": 0.6454,
+      "step": 2807
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 0.5671429634094238,
+      "learning_rate": 0.00011026421136909528,
+      "loss": 0.6421,
+      "step": 2808
+    },
+    {
+      "epoch": 0.89888,
+      "grad_norm": 0.45298755168914795,
+      "learning_rate": 0.00011023218574859889,
+      "loss": 0.5336,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 0.5061096549034119,
+      "learning_rate": 0.00011020016012810248,
+      "loss": 0.5953,
+      "step": 2810
+    },
+    {
+      "epoch": 0.89952,
+      "grad_norm": 0.5046201944351196,
+      "learning_rate": 0.00011016813450760608,
+      "loss": 0.5715,
+      "step": 2811
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 0.5581631660461426,
+      "learning_rate": 0.0001101361088871097,
+      "loss": 0.5972,
+      "step": 2812
+    },
+    {
+      "epoch": 0.90016,
+      "grad_norm": 0.5193576812744141,
+      "learning_rate": 0.0001101040832666133,
+      "loss": 0.6252,
+      "step": 2813
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 0.42851489782333374,
+      "learning_rate": 0.0001100720576461169,
+      "loss": 0.5466,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 0.5270585417747498,
+      "learning_rate": 0.0001100400320256205,
+      "loss": 0.6606,
+      "step": 2815
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 0.3721252679824829,
+      "learning_rate": 0.00011000800640512409,
+      "loss": 0.4207,
+      "step": 2816
+    },
+    {
+      "epoch": 0.90144,
+      "grad_norm": 0.4263012111186981,
+      "learning_rate": 0.0001099759807846277,
+      "loss": 0.4807,
+      "step": 2817
+    },
+    {
+      "epoch": 0.90176,
+      "grad_norm": 0.4945608973503113,
+      "learning_rate": 0.00010994395516413132,
+      "loss": 0.459,
+      "step": 2818
+    },
+    {
+      "epoch": 0.90208,
+      "grad_norm": 0.47278451919555664,
+      "learning_rate": 0.00010991192954363492,
+      "loss": 0.4638,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.47517499327659607,
+      "learning_rate": 0.00010987990392313851,
+      "loss": 0.5182,
+      "step": 2820
+    },
+    {
+      "epoch": 0.90272,
+      "grad_norm": 0.5441727042198181,
+      "learning_rate": 0.00010984787830264212,
+      "loss": 0.585,
+      "step": 2821
+    },
+    {
+      "epoch": 0.90304,
+      "grad_norm": 0.5445504784584045,
+      "learning_rate": 0.00010981585268214572,
+      "loss": 0.7428,
+      "step": 2822
+    },
+    {
+      "epoch": 0.90336,
+      "grad_norm": 0.46833398938179016,
+      "learning_rate": 0.00010978382706164931,
+      "loss": 0.4871,
+      "step": 2823
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 0.39207056164741516,
+      "learning_rate": 0.00010975180144115293,
+      "loss": 0.5623,
+      "step": 2824
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.5444238781929016,
+      "learning_rate": 0.00010971977582065654,
+      "loss": 0.6841,
+      "step": 2825
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 0.45786067843437195,
+      "learning_rate": 0.00010968775020016013,
+      "loss": 0.5065,
+      "step": 2826
+    },
+    {
+      "epoch": 0.90464,
+      "grad_norm": 0.4410295784473419,
+      "learning_rate": 0.00010965572457966373,
+      "loss": 0.5387,
+      "step": 2827
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 0.46998900175094604,
+      "learning_rate": 0.00010962369895916734,
+      "loss": 0.6262,
+      "step": 2828
+    },
+    {
+      "epoch": 0.90528,
+      "grad_norm": 0.401234894990921,
+      "learning_rate": 0.00010959167333867095,
+      "loss": 0.4647,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 0.4436178505420685,
+      "learning_rate": 0.00010955964771817455,
+      "loss": 0.677,
+      "step": 2830
+    },
+    {
+      "epoch": 0.90592,
+      "grad_norm": 0.4939956068992615,
+      "learning_rate": 0.00010952762209767815,
+      "loss": 0.6652,
+      "step": 2831
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 0.5035571455955505,
+      "learning_rate": 0.00010949559647718176,
+      "loss": 0.6003,
+      "step": 2832
+    },
+    {
+      "epoch": 0.90656,
+      "grad_norm": 0.3757145404815674,
+      "learning_rate": 0.00010946357085668535,
+      "loss": 0.4726,
+      "step": 2833
+    },
+    {
+      "epoch": 0.90688,
+      "grad_norm": 0.5764524936676025,
+      "learning_rate": 0.00010943154523618895,
+      "loss": 0.5763,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 0.5526479482650757,
+      "learning_rate": 0.00010939951961569257,
+      "loss": 0.5631,
+      "step": 2835
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 0.4396551847457886,
+      "learning_rate": 0.00010936749399519617,
+      "loss": 0.4808,
+      "step": 2836
+    },
+    {
+      "epoch": 0.90784,
+      "grad_norm": 0.5427368879318237,
+      "learning_rate": 0.00010933546837469977,
+      "loss": 0.6485,
+      "step": 2837
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 0.47680115699768066,
+      "learning_rate": 0.00010930344275420337,
+      "loss": 0.5016,
+      "step": 2838
+    },
+    {
+      "epoch": 0.90848,
+      "grad_norm": 0.5615769624710083,
+      "learning_rate": 0.00010927141713370696,
+      "loss": 0.6611,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.45605769753456116,
+      "learning_rate": 0.00010923939151321057,
+      "loss": 0.4371,
+      "step": 2840
+    },
+    {
+      "epoch": 0.90912,
+      "grad_norm": 0.4144948720932007,
+      "learning_rate": 0.00010920736589271418,
+      "loss": 0.5144,
+      "step": 2841
+    },
+    {
+      "epoch": 0.90944,
+      "grad_norm": 0.5282120108604431,
+      "learning_rate": 0.00010917534027221779,
+      "loss": 0.5932,
+      "step": 2842
+    },
+    {
+      "epoch": 0.90976,
+      "grad_norm": 0.4677801728248596,
+      "learning_rate": 0.00010914331465172138,
+      "loss": 0.5597,
+      "step": 2843
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 0.46219703555107117,
+      "learning_rate": 0.00010911128903122499,
+      "loss": 0.5743,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 0.5815814137458801,
+      "learning_rate": 0.00010907926341072858,
+      "loss": 0.6104,
+      "step": 2845
+    },
+    {
+      "epoch": 0.91072,
+      "grad_norm": 0.567638099193573,
+      "learning_rate": 0.00010904723779023218,
+      "loss": 0.6222,
+      "step": 2846
+    },
+    {
+      "epoch": 0.91104,
+      "grad_norm": 0.42515161633491516,
+      "learning_rate": 0.0001090152121697358,
+      "loss": 0.5735,
+      "step": 2847
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 0.44562283158302307,
+      "learning_rate": 0.0001089831865492394,
+      "loss": 0.5372,
+      "step": 2848
+    },
+    {
+      "epoch": 0.91168,
+      "grad_norm": 0.46798908710479736,
+      "learning_rate": 0.000108951160928743,
+      "loss": 0.5637,
+      "step": 2849
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.5905529856681824,
+      "learning_rate": 0.0001089191353082466,
+      "loss": 0.6131,
+      "step": 2850
+    },
+    {
+      "epoch": 0.91232,
+      "grad_norm": 0.5411832332611084,
+      "learning_rate": 0.0001088871096877502,
+      "loss": 0.5676,
+      "step": 2851
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 0.42661920189857483,
+      "learning_rate": 0.0001088550840672538,
+      "loss": 0.4632,
+      "step": 2852
+    },
+    {
+      "epoch": 0.91296,
+      "grad_norm": 0.4680926203727722,
+      "learning_rate": 0.00010882305844675741,
+      "loss": 0.5334,
+      "step": 2853
+    },
+    {
+      "epoch": 0.91328,
+      "grad_norm": 0.48280808329582214,
+      "learning_rate": 0.00010879103282626102,
+      "loss": 0.5002,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 0.49232375621795654,
+      "learning_rate": 0.00010875900720576462,
+      "loss": 0.6114,
+      "step": 2855
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 0.47566378116607666,
+      "learning_rate": 0.00010872698158526822,
+      "loss": 0.5425,
+      "step": 2856
+    },
+    {
+      "epoch": 0.91424,
+      "grad_norm": 0.5821830630302429,
+      "learning_rate": 0.00010869495596477182,
+      "loss": 0.5539,
+      "step": 2857
+    },
+    {
+      "epoch": 0.91456,
+      "grad_norm": 0.5118959546089172,
+      "learning_rate": 0.00010866293034427541,
+      "loss": 0.5207,
+      "step": 2858
+    },
+    {
+      "epoch": 0.91488,
+      "grad_norm": 0.33722439408302307,
+      "learning_rate": 0.00010863090472377904,
+      "loss": 0.3945,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.5621157884597778,
+      "learning_rate": 0.00010859887910328263,
+      "loss": 0.6993,
+      "step": 2860
+    },
+    {
+      "epoch": 0.91552,
+      "grad_norm": 0.5042179822921753,
+      "learning_rate": 0.00010856685348278624,
+      "loss": 0.5576,
+      "step": 2861
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 0.4368782639503479,
+      "learning_rate": 0.00010853482786228983,
+      "loss": 0.406,
+      "step": 2862
+    },
+    {
+      "epoch": 0.91616,
+      "grad_norm": 0.44221383333206177,
+      "learning_rate": 0.00010850280224179344,
+      "loss": 0.5784,
+      "step": 2863
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 0.46727773547172546,
+      "learning_rate": 0.00010847077662129703,
+      "loss": 0.6097,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 0.4794403314590454,
+      "learning_rate": 0.00010843875100080066,
+      "loss": 0.6097,
+      "step": 2865
+    },
+    {
+      "epoch": 0.91712,
+      "grad_norm": 0.5628266334533691,
+      "learning_rate": 0.00010840672538030425,
+      "loss": 0.6227,
+      "step": 2866
+    },
+    {
+      "epoch": 0.91744,
+      "grad_norm": 0.5198894143104553,
+      "learning_rate": 0.00010837469975980785,
+      "loss": 0.6462,
+      "step": 2867
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 0.515910804271698,
+      "learning_rate": 0.00010834267413931145,
+      "loss": 0.5978,
+      "step": 2868
+    },
+    {
+      "epoch": 0.91808,
+      "grad_norm": 0.4215081036090851,
+      "learning_rate": 0.00010831064851881505,
+      "loss": 0.5539,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 0.4789619743824005,
+      "learning_rate": 0.00010827862289831865,
+      "loss": 0.6045,
+      "step": 2870
+    },
+    {
+      "epoch": 0.91872,
+      "grad_norm": 0.4826526641845703,
+      "learning_rate": 0.00010824659727782227,
+      "loss": 0.5115,
+      "step": 2871
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 0.41785910725593567,
+      "learning_rate": 0.00010821457165732586,
+      "loss": 0.5358,
+      "step": 2872
+    },
+    {
+      "epoch": 0.91936,
+      "grad_norm": 0.5981875658035278,
+      "learning_rate": 0.00010818254603682947,
+      "loss": 0.6032,
+      "step": 2873
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 0.46520647406578064,
+      "learning_rate": 0.00010815052041633307,
+      "loss": 0.5415,
+      "step": 2874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.597777247428894,
+      "learning_rate": 0.00010811849479583667,
+      "loss": 0.5758,
+      "step": 2875
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 0.4953913688659668,
+      "learning_rate": 0.00010808646917534028,
+      "loss": 0.6313,
+      "step": 2876
+    },
+    {
+      "epoch": 0.92064,
+      "grad_norm": 0.557225227355957,
+      "learning_rate": 0.00010805444355484389,
+      "loss": 0.7154,
+      "step": 2877
+    },
+    {
+      "epoch": 0.92096,
+      "grad_norm": 0.4329536259174347,
+      "learning_rate": 0.00010802241793434749,
+      "loss": 0.4992,
+      "step": 2878
+    },
+    {
+      "epoch": 0.92128,
+      "grad_norm": 0.5190502405166626,
+      "learning_rate": 0.00010799039231385108,
+      "loss": 0.6982,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.43534624576568604,
+      "learning_rate": 0.00010795836669335469,
+      "loss": 0.5587,
+      "step": 2880
+    },
+    {
+      "epoch": 0.92192,
+      "grad_norm": 0.5866831541061401,
+      "learning_rate": 0.00010792634107285828,
+      "loss": 0.7869,
+      "step": 2881
+    },
+    {
+      "epoch": 0.92224,
+      "grad_norm": 0.44097772240638733,
+      "learning_rate": 0.0001078943154523619,
+      "loss": 0.5494,
+      "step": 2882
+    },
+    {
+      "epoch": 0.92256,
+      "grad_norm": 0.5052035450935364,
+      "learning_rate": 0.0001078622898318655,
+      "loss": 0.6493,
+      "step": 2883
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 0.4962887763977051,
+      "learning_rate": 0.00010783026421136911,
+      "loss": 0.6853,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 0.46663275361061096,
+      "learning_rate": 0.0001077982385908727,
+      "loss": 0.5308,
+      "step": 2885
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 0.4331126809120178,
+      "learning_rate": 0.0001077662129703763,
+      "loss": 0.6455,
+      "step": 2886
+    },
+    {
+      "epoch": 0.92384,
+      "grad_norm": 0.45057496428489685,
+      "learning_rate": 0.0001077341873498799,
+      "loss": 0.4623,
+      "step": 2887
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 0.4430769979953766,
+      "learning_rate": 0.00010770216172938353,
+      "loss": 0.6104,
+      "step": 2888
+    },
+    {
+      "epoch": 0.92448,
+      "grad_norm": 0.4633301794528961,
+      "learning_rate": 0.00010767013610888712,
+      "loss": 0.5107,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 0.5720324516296387,
+      "learning_rate": 0.00010763811048839072,
+      "loss": 0.6729,
+      "step": 2890
+    },
+    {
+      "epoch": 0.92512,
+      "grad_norm": 0.48331865668296814,
+      "learning_rate": 0.00010760608486789431,
+      "loss": 0.4544,
+      "step": 2891
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 0.38801309466362,
+      "learning_rate": 0.00010757405924739792,
+      "loss": 0.4375,
+      "step": 2892
+    },
+    {
+      "epoch": 0.92576,
+      "grad_norm": 0.5177491903305054,
+      "learning_rate": 0.00010754203362690152,
+      "loss": 0.5108,
+      "step": 2893
+    },
+    {
+      "epoch": 0.92608,
+      "grad_norm": 0.43551984429359436,
+      "learning_rate": 0.00010751000800640514,
+      "loss": 0.4341,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 0.44049718976020813,
+      "learning_rate": 0.00010747798238590873,
+      "loss": 0.4349,
+      "step": 2895
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 0.49996620416641235,
+      "learning_rate": 0.00010744595676541234,
+      "loss": 0.6206,
+      "step": 2896
+    },
+    {
+      "epoch": 0.92704,
+      "grad_norm": 0.4088820517063141,
+      "learning_rate": 0.00010741393114491593,
+      "loss": 0.5507,
+      "step": 2897
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 0.5509172081947327,
+      "learning_rate": 0.00010738190552441953,
+      "loss": 0.4831,
+      "step": 2898
+    },
+    {
+      "epoch": 0.92768,
+      "grad_norm": 0.4930497407913208,
+      "learning_rate": 0.00010734987990392314,
+      "loss": 0.6294,
+      "step": 2899
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.5794312953948975,
+      "learning_rate": 0.00010731785428342676,
+      "loss": 0.6582,
+      "step": 2900
+    },
+    {
+      "epoch": 0.92832,
+      "grad_norm": 0.7082372307777405,
+      "learning_rate": 0.00010728582866293035,
+      "loss": 0.6209,
+      "step": 2901
+    },
+    {
+      "epoch": 0.92864,
+      "grad_norm": 0.5852804183959961,
+      "learning_rate": 0.00010725380304243395,
+      "loss": 0.5428,
+      "step": 2902
+    },
+    {
+      "epoch": 0.92896,
+      "grad_norm": 0.5114850401878357,
+      "learning_rate": 0.00010722177742193756,
+      "loss": 0.6574,
+      "step": 2903
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 0.48023203015327454,
+      "learning_rate": 0.00010718975180144115,
+      "loss": 0.5318,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 0.48197251558303833,
+      "learning_rate": 0.00010715772618094475,
+      "loss": 0.484,
+      "step": 2905
+    },
+    {
+      "epoch": 0.92992,
+      "grad_norm": 0.42469316720962524,
+      "learning_rate": 0.00010712570056044837,
+      "loss": 0.6073,
+      "step": 2906
+    },
+    {
+      "epoch": 0.93024,
+      "grad_norm": 0.5923599004745483,
+      "learning_rate": 0.00010709367493995198,
+      "loss": 0.7541,
+      "step": 2907
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 0.4505990445613861,
+      "learning_rate": 0.00010706164931945557,
+      "loss": 0.6645,
+      "step": 2908
+    },
+    {
+      "epoch": 0.93088,
+      "grad_norm": 0.534603476524353,
+      "learning_rate": 0.00010702962369895917,
+      "loss": 0.7284,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.6674680709838867,
+      "learning_rate": 0.00010699759807846276,
+      "loss": 0.5572,
+      "step": 2910
+    },
+    {
+      "epoch": 0.93152,
+      "grad_norm": 0.5552335381507874,
+      "learning_rate": 0.00010696557245796637,
+      "loss": 0.5188,
+      "step": 2911
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 0.43497607111930847,
+      "learning_rate": 0.00010693354683746999,
+      "loss": 0.4428,
+      "step": 2912
+    },
+    {
+      "epoch": 0.93216,
+      "grad_norm": 0.45341089367866516,
+      "learning_rate": 0.00010690152121697359,
+      "loss": 0.7119,
+      "step": 2913
+    },
+    {
+      "epoch": 0.93248,
+      "grad_norm": 0.46644097566604614,
+      "learning_rate": 0.00010686949559647718,
+      "loss": 0.504,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 0.5412666201591492,
+      "learning_rate": 0.00010683746997598079,
+      "loss": 0.7183,
+      "step": 2915
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 0.49403369426727295,
+      "learning_rate": 0.00010680544435548438,
+      "loss": 0.5463,
+      "step": 2916
+    },
+    {
+      "epoch": 0.93344,
+      "grad_norm": 0.4302772283554077,
+      "learning_rate": 0.00010677341873498798,
+      "loss": 0.5489,
+      "step": 2917
+    },
+    {
+      "epoch": 0.93376,
+      "grad_norm": 0.4592494070529938,
+      "learning_rate": 0.0001067413931144916,
+      "loss": 0.5972,
+      "step": 2918
+    },
+    {
+      "epoch": 0.93408,
+      "grad_norm": 0.4107368588447571,
+      "learning_rate": 0.00010670936749399521,
+      "loss": 0.4155,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.4643627405166626,
+      "learning_rate": 0.0001066773418734988,
+      "loss": 0.5529,
+      "step": 2920
+    },
+    {
+      "epoch": 0.93472,
+      "grad_norm": 0.4478214383125305,
+      "learning_rate": 0.0001066453162530024,
+      "loss": 0.4665,
+      "step": 2921
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 0.40474241971969604,
+      "learning_rate": 0.00010661329063250601,
+      "loss": 0.5964,
+      "step": 2922
+    },
+    {
+      "epoch": 0.93536,
+      "grad_norm": 0.5492748618125916,
+      "learning_rate": 0.00010658126501200963,
+      "loss": 0.5901,
+      "step": 2923
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 0.5788101553916931,
+      "learning_rate": 0.00010654923939151322,
+      "loss": 0.7543,
+      "step": 2924
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 0.668602466583252,
+      "learning_rate": 0.00010651721377101682,
+      "loss": 0.6577,
+      "step": 2925
+    },
+    {
+      "epoch": 0.93632,
+      "grad_norm": 0.4587288796901703,
+      "learning_rate": 0.00010648518815052043,
+      "loss": 0.5497,
+      "step": 2926
+    },
+    {
+      "epoch": 0.93664,
+      "grad_norm": 0.514289379119873,
+      "learning_rate": 0.00010645316253002402,
+      "loss": 0.5327,
+      "step": 2927
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 0.4449906051158905,
+      "learning_rate": 0.00010642113690952762,
+      "loss": 0.529,
+      "step": 2928
+    },
+    {
+      "epoch": 0.93728,
+      "grad_norm": 0.5658524632453918,
+      "learning_rate": 0.00010638911128903124,
+      "loss": 0.5953,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 0.4779348373413086,
+      "learning_rate": 0.00010635708566853485,
+      "loss": 0.6138,
+      "step": 2930
+    },
+    {
+      "epoch": 0.93792,
+      "grad_norm": 0.4292345643043518,
+      "learning_rate": 0.00010632506004803844,
+      "loss": 0.5466,
+      "step": 2931
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 0.47823190689086914,
+      "learning_rate": 0.00010629303442754204,
+      "loss": 0.5815,
+      "step": 2932
+    },
+    {
+      "epoch": 0.93856,
+      "grad_norm": 0.43782174587249756,
+      "learning_rate": 0.00010626100880704563,
+      "loss": 0.6252,
+      "step": 2933
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 0.6349648237228394,
+      "learning_rate": 0.00010622898318654924,
+      "loss": 0.7443,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 0.4943791627883911,
+      "learning_rate": 0.00010619695756605286,
+      "loss": 0.5902,
+      "step": 2935
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 0.5145920515060425,
+      "learning_rate": 0.00010616493194555646,
+      "loss": 0.523,
+      "step": 2936
+    },
+    {
+      "epoch": 0.93984,
+      "grad_norm": 0.4812191426753998,
+      "learning_rate": 0.00010613290632506005,
+      "loss": 0.5607,
+      "step": 2937
+    },
+    {
+      "epoch": 0.94016,
+      "grad_norm": 0.5444737672805786,
+      "learning_rate": 0.00010610088070456366,
+      "loss": 0.6195,
+      "step": 2938
+    },
+    {
+      "epoch": 0.94048,
+      "grad_norm": 0.5037015080451965,
+      "learning_rate": 0.00010606885508406725,
+      "loss": 0.4819,
+      "step": 2939
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.4491082727909088,
+      "learning_rate": 0.00010603682946357085,
+      "loss": 0.4987,
+      "step": 2940
+    },
+    {
+      "epoch": 0.94112,
+      "grad_norm": 0.5125628113746643,
+      "learning_rate": 0.00010600480384307447,
+      "loss": 0.5707,
+      "step": 2941
+    },
+    {
+      "epoch": 0.94144,
+      "grad_norm": 0.5719506740570068,
+      "learning_rate": 0.00010597277822257808,
+      "loss": 0.6663,
+      "step": 2942
+    },
+    {
+      "epoch": 0.94176,
+      "grad_norm": 0.41161808371543884,
+      "learning_rate": 0.00010594075260208167,
+      "loss": 0.4862,
+      "step": 2943
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 0.46917128562927246,
+      "learning_rate": 0.00010590872698158527,
+      "loss": 0.4898,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 0.625738799571991,
+      "learning_rate": 0.00010587670136108888,
+      "loss": 0.8008,
+      "step": 2945
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 0.5155507326126099,
+      "learning_rate": 0.00010584467574059247,
+      "loss": 0.6388,
+      "step": 2946
+    },
+    {
+      "epoch": 0.94304,
+      "grad_norm": 0.5053896903991699,
+      "learning_rate": 0.00010581265012009609,
+      "loss": 0.5858,
+      "step": 2947
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 0.4531412422657013,
+      "learning_rate": 0.00010578062449959969,
+      "loss": 0.4682,
+      "step": 2948
+    },
+    {
+      "epoch": 0.94368,
+      "grad_norm": 0.4912232756614685,
+      "learning_rate": 0.0001057485988791033,
+      "loss": 0.5342,
+      "step": 2949
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.48542502522468567,
+      "learning_rate": 0.00010571657325860689,
+      "loss": 0.551,
+      "step": 2950
+    },
+    {
+      "epoch": 0.94432,
+      "grad_norm": 0.517581045627594,
+      "learning_rate": 0.00010568454763811049,
+      "loss": 0.5227,
+      "step": 2951
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 0.4515475332736969,
+      "learning_rate": 0.00010565252201761408,
+      "loss": 0.5957,
+      "step": 2952
+    },
+    {
+      "epoch": 0.94496,
+      "grad_norm": 0.424979031085968,
+      "learning_rate": 0.0001056204963971177,
+      "loss": 0.5471,
+      "step": 2953
+    },
+    {
+      "epoch": 0.94528,
+      "grad_norm": 0.5898860096931458,
+      "learning_rate": 0.0001055884707766213,
+      "loss": 0.5994,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 0.41437259316444397,
+      "learning_rate": 0.00010555644515612491,
+      "loss": 0.5655,
+      "step": 2955
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 0.4539845585823059,
+      "learning_rate": 0.0001055244195356285,
+      "loss": 0.534,
+      "step": 2956
+    },
+    {
+      "epoch": 0.94624,
+      "grad_norm": 0.5835381150245667,
+      "learning_rate": 0.0001054923939151321,
+      "loss": 0.4727,
+      "step": 2957
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 0.44539856910705566,
+      "learning_rate": 0.0001054603682946357,
+      "loss": 0.4626,
+      "step": 2958
+    },
+    {
+      "epoch": 0.94688,
+      "grad_norm": 0.4014412462711334,
+      "learning_rate": 0.00010542834267413933,
+      "loss": 0.4536,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.4767681360244751,
+      "learning_rate": 0.00010539631705364292,
+      "loss": 0.481,
+      "step": 2960
+    },
+    {
+      "epoch": 0.94752,
+      "grad_norm": 0.5415595173835754,
+      "learning_rate": 0.00010536429143314653,
+      "loss": 0.5467,
+      "step": 2961
+    },
+    {
+      "epoch": 0.94784,
+      "grad_norm": 0.5655352473258972,
+      "learning_rate": 0.00010533226581265012,
+      "loss": 0.6157,
+      "step": 2962
+    },
+    {
+      "epoch": 0.94816,
+      "grad_norm": 0.4475812315940857,
+      "learning_rate": 0.00010530024019215372,
+      "loss": 0.5776,
+      "step": 2963
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 0.5241261720657349,
+      "learning_rate": 0.00010526821457165733,
+      "loss": 0.5416,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 0.4143690764904022,
+      "learning_rate": 0.00010523618895116094,
+      "loss": 0.6207,
+      "step": 2965
+    },
+    {
+      "epoch": 0.94912,
+      "grad_norm": 0.5310909748077393,
+      "learning_rate": 0.00010520416333066454,
+      "loss": 0.5825,
+      "step": 2966
+    },
+    {
+      "epoch": 0.94944,
+      "grad_norm": 0.4872644543647766,
+      "learning_rate": 0.00010517213771016814,
+      "loss": 0.5401,
+      "step": 2967
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 0.6009882092475891,
+      "learning_rate": 0.00010514011208967173,
+      "loss": 0.6539,
+      "step": 2968
+    },
+    {
+      "epoch": 0.95008,
+      "grad_norm": 0.444217711687088,
+      "learning_rate": 0.00010510808646917534,
+      "loss": 0.6559,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.42400676012039185,
+      "learning_rate": 0.00010507606084867895,
+      "loss": 0.5023,
+      "step": 2970
+    },
+    {
+      "epoch": 0.95072,
+      "grad_norm": 0.5807833671569824,
+      "learning_rate": 0.00010504403522818256,
+      "loss": 0.5673,
+      "step": 2971
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 0.5428476333618164,
+      "learning_rate": 0.00010501200960768615,
+      "loss": 0.6296,
+      "step": 2972
+    },
+    {
+      "epoch": 0.95136,
+      "grad_norm": 0.4984135031700134,
+      "learning_rate": 0.00010497998398718976,
+      "loss": 0.6975,
+      "step": 2973
+    },
+    {
+      "epoch": 0.95168,
+      "grad_norm": 0.45177721977233887,
+      "learning_rate": 0.00010494795836669336,
+      "loss": 0.48,
+      "step": 2974
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.4846910536289215,
+      "learning_rate": 0.00010491593274619695,
+      "loss": 0.609,
+      "step": 2975
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 0.402199923992157,
+      "learning_rate": 0.00010488390712570057,
+      "loss": 0.4651,
+      "step": 2976
+    },
+    {
+      "epoch": 0.95264,
+      "grad_norm": 0.5940353870391846,
+      "learning_rate": 0.00010485188150520417,
+      "loss": 0.426,
+      "step": 2977
+    },
+    {
+      "epoch": 0.95296,
+      "grad_norm": 0.5137403011322021,
+      "learning_rate": 0.00010481985588470778,
+      "loss": 0.5667,
+      "step": 2978
+    },
+    {
+      "epoch": 0.95328,
+      "grad_norm": 0.6981788873672485,
+      "learning_rate": 0.00010478783026421137,
+      "loss": 0.5273,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.5496737957000732,
+      "learning_rate": 0.00010475580464371498,
+      "loss": 0.7383,
+      "step": 2980
+    },
+    {
+      "epoch": 0.95392,
+      "grad_norm": 0.4991731643676758,
+      "learning_rate": 0.00010472377902321857,
+      "loss": 0.5071,
+      "step": 2981
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 0.4762144386768341,
+      "learning_rate": 0.0001046917534027222,
+      "loss": 0.5437,
+      "step": 2982
+    },
+    {
+      "epoch": 0.95456,
+      "grad_norm": 0.4508254826068878,
+      "learning_rate": 0.00010465972778222579,
+      "loss": 0.4673,
+      "step": 2983
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 0.4865306317806244,
+      "learning_rate": 0.0001046277021617294,
+      "loss": 0.6073,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 0.5615052580833435,
+      "learning_rate": 0.00010459567654123299,
+      "loss": 0.617,
+      "step": 2985
+    },
+    {
+      "epoch": 0.95552,
+      "grad_norm": 0.49990251660346985,
+      "learning_rate": 0.00010456365092073659,
+      "loss": 0.6765,
+      "step": 2986
+    },
+    {
+      "epoch": 0.95584,
+      "grad_norm": 0.5307373404502869,
+      "learning_rate": 0.00010453162530024018,
+      "loss": 0.7037,
+      "step": 2987
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 0.4959764778614044,
+      "learning_rate": 0.00010449959967974381,
+      "loss": 0.5664,
+      "step": 2988
+    },
+    {
+      "epoch": 0.95648,
+      "grad_norm": 0.5125677585601807,
+      "learning_rate": 0.0001044675740592474,
+      "loss": 0.6212,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 0.45804667472839355,
+      "learning_rate": 0.00010443554843875101,
+      "loss": 0.4588,
+      "step": 2990
+    },
+    {
+      "epoch": 0.95712,
+      "grad_norm": 0.4648250639438629,
+      "learning_rate": 0.0001044035228182546,
+      "loss": 0.64,
+      "step": 2991
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 0.4956052005290985,
+      "learning_rate": 0.0001043714971977582,
+      "loss": 0.6039,
+      "step": 2992
+    },
+    {
+      "epoch": 0.95776,
+      "grad_norm": 0.4743454158306122,
+      "learning_rate": 0.00010433947157726181,
+      "loss": 0.4986,
+      "step": 2993
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 0.3457401990890503,
+      "learning_rate": 0.00010430744595676543,
+      "loss": 0.3939,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 0.6121973991394043,
+      "learning_rate": 0.00010427542033626902,
+      "loss": 0.6249,
+      "step": 2995
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 0.47821199893951416,
+      "learning_rate": 0.00010424339471577262,
+      "loss": 0.4716,
+      "step": 2996
+    },
+    {
+      "epoch": 0.95904,
+      "grad_norm": 0.5126957297325134,
+      "learning_rate": 0.00010421136909527623,
+      "loss": 0.6605,
+      "step": 2997
+    },
+    {
+      "epoch": 0.95936,
+      "grad_norm": 0.5026578903198242,
+      "learning_rate": 0.00010417934347477982,
+      "loss": 0.5429,
+      "step": 2998
+    },
+    {
+      "epoch": 0.95968,
+      "grad_norm": 0.42023271322250366,
+      "learning_rate": 0.00010414731785428343,
+      "loss": 0.5109,
+      "step": 2999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4612977206707001,
+      "learning_rate": 0.00010411529223378704,
+      "loss": 0.5004,
+      "step": 3000
+    },
+    {
+      "epoch": 0.96032,
+      "grad_norm": 0.5849503874778748,
+      "learning_rate": 0.00010408326661329065,
+      "loss": 0.6164,
+      "step": 3001
+    },
+    {
+      "epoch": 0.96064,
+      "grad_norm": 0.6064012050628662,
+      "learning_rate": 0.00010405124099279424,
+      "loss": 0.803,
+      "step": 3002
+    },
+    {
+      "epoch": 0.96096,
+      "grad_norm": 0.5061959028244019,
+      "learning_rate": 0.00010401921537229784,
+      "loss": 0.5987,
+      "step": 3003
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 0.4531479477882385,
+      "learning_rate": 0.00010398718975180144,
+      "loss": 0.561,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 0.4610559046268463,
+      "learning_rate": 0.00010395516413130504,
+      "loss": 0.4916,
+      "step": 3005
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 0.3983750641345978,
+      "learning_rate": 0.00010392313851080866,
+      "loss": 0.4601,
+      "step": 3006
+    },
+    {
+      "epoch": 0.96224,
+      "grad_norm": 0.4578743577003479,
+      "learning_rate": 0.00010389111289031226,
+      "loss": 0.4936,
+      "step": 3007
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 0.5480005145072937,
+      "learning_rate": 0.00010385908726981585,
+      "loss": 0.568,
+      "step": 3008
+    },
+    {
+      "epoch": 0.96288,
+      "grad_norm": 0.42554086446762085,
+      "learning_rate": 0.00010382706164931946,
+      "loss": 0.5484,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 0.4657409191131592,
+      "learning_rate": 0.00010379503602882305,
+      "loss": 0.4953,
+      "step": 3010
+    },
+    {
+      "epoch": 0.96352,
+      "grad_norm": 0.4048027992248535,
+      "learning_rate": 0.00010376301040832666,
+      "loss": 0.5282,
+      "step": 3011
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 0.5440258383750916,
+      "learning_rate": 0.00010373098478783027,
+      "loss": 0.547,
+      "step": 3012
+    },
+    {
+      "epoch": 0.96416,
+      "grad_norm": 0.4555546045303345,
+      "learning_rate": 0.00010369895916733388,
+      "loss": 0.577,
+      "step": 3013
+    },
+    {
+      "epoch": 0.96448,
+      "grad_norm": 0.49337390065193176,
+      "learning_rate": 0.00010366693354683747,
+      "loss": 0.6208,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 0.4110238254070282,
+      "learning_rate": 0.00010363490792634107,
+      "loss": 0.3776,
+      "step": 3015
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 0.5754513144493103,
+      "learning_rate": 0.00010360288230584468,
+      "loss": 0.5408,
+      "step": 3016
+    },
+    {
+      "epoch": 0.96544,
+      "grad_norm": 0.4759342670440674,
+      "learning_rate": 0.0001035708566853483,
+      "loss": 0.5042,
+      "step": 3017
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 0.5122634172439575,
+      "learning_rate": 0.00010353883106485189,
+      "loss": 0.6408,
+      "step": 3018
+    },
+    {
+      "epoch": 0.96608,
+      "grad_norm": 0.582034707069397,
+      "learning_rate": 0.0001035068054443555,
+      "loss": 0.6571,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.47620803117752075,
+      "learning_rate": 0.0001034747798238591,
+      "loss": 0.4842,
+      "step": 3020
+    },
+    {
+      "epoch": 0.96672,
+      "grad_norm": 0.4087372124195099,
+      "learning_rate": 0.00010344275420336269,
+      "loss": 0.5021,
+      "step": 3021
+    },
+    {
+      "epoch": 0.96704,
+      "grad_norm": 0.49236053228378296,
+      "learning_rate": 0.0001034107285828663,
+      "loss": 0.4686,
+      "step": 3022
+    },
+    {
+      "epoch": 0.96736,
+      "grad_norm": 0.4629395604133606,
+      "learning_rate": 0.00010337870296236991,
+      "loss": 0.514,
+      "step": 3023
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 0.5234566330909729,
+      "learning_rate": 0.0001033466773418735,
+      "loss": 0.5392,
+      "step": 3024
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.5151426196098328,
+      "learning_rate": 0.00010331465172137711,
+      "loss": 0.5426,
+      "step": 3025
+    },
+    {
+      "epoch": 0.96832,
+      "grad_norm": 0.5390909314155579,
+      "learning_rate": 0.00010328262610088071,
+      "loss": 0.5533,
+      "step": 3026
+    },
+    {
+      "epoch": 0.96864,
+      "grad_norm": 0.5069742202758789,
+      "learning_rate": 0.0001032506004803843,
+      "loss": 0.6097,
+      "step": 3027
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 0.5396665334701538,
+      "learning_rate": 0.00010321857485988791,
+      "loss": 0.5203,
+      "step": 3028
+    },
+    {
+      "epoch": 0.96928,
+      "grad_norm": 0.448403924703598,
+      "learning_rate": 0.00010318654923939153,
+      "loss": 0.5911,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.6159946322441101,
+      "learning_rate": 0.00010315452361889513,
+      "loss": 0.5785,
+      "step": 3030
+    },
+    {
+      "epoch": 0.96992,
+      "grad_norm": 0.43678876757621765,
+      "learning_rate": 0.00010312249799839872,
+      "loss": 0.4622,
+      "step": 3031
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 0.44540637731552124,
+      "learning_rate": 0.00010309047237790233,
+      "loss": 0.5275,
+      "step": 3032
+    },
+    {
+      "epoch": 0.97056,
+      "grad_norm": 0.5449602603912354,
+      "learning_rate": 0.00010305844675740592,
+      "loss": 0.6689,
+      "step": 3033
+    },
+    {
+      "epoch": 0.97088,
+      "grad_norm": 0.4705146253108978,
+      "learning_rate": 0.00010302642113690952,
+      "loss": 0.4869,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 0.47705960273742676,
+      "learning_rate": 0.00010299439551641314,
+      "loss": 0.483,
+      "step": 3035
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 0.53836590051651,
+      "learning_rate": 0.00010296236989591675,
+      "loss": 0.5484,
+      "step": 3036
+    },
+    {
+      "epoch": 0.97184,
+      "grad_norm": 0.5539997220039368,
+      "learning_rate": 0.00010293034427542034,
+      "loss": 0.5984,
+      "step": 3037
+    },
+    {
+      "epoch": 0.97216,
+      "grad_norm": 0.5214716792106628,
+      "learning_rate": 0.00010289831865492394,
+      "loss": 0.6221,
+      "step": 3038
+    },
+    {
+      "epoch": 0.97248,
+      "grad_norm": 0.4695388972759247,
+      "learning_rate": 0.00010286629303442753,
+      "loss": 0.4553,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.47435787320137024,
+      "learning_rate": 0.00010283426741393114,
+      "loss": 0.4804,
+      "step": 3040
+    },
+    {
+      "epoch": 0.97312,
+      "grad_norm": 0.3936333656311035,
+      "learning_rate": 0.00010280224179343476,
+      "loss": 0.4455,
+      "step": 3041
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 0.520043134689331,
+      "learning_rate": 0.00010277021617293836,
+      "loss": 0.5809,
+      "step": 3042
+    },
+    {
+      "epoch": 0.97376,
+      "grad_norm": 0.5713899731636047,
+      "learning_rate": 0.00010273819055244195,
+      "loss": 0.5408,
+      "step": 3043
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 0.4933745563030243,
+      "learning_rate": 0.00010270616493194556,
+      "loss": 0.7383,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 0.44971492886543274,
+      "learning_rate": 0.00010267413931144916,
+      "loss": 0.5985,
+      "step": 3045
+    },
+    {
+      "epoch": 0.97472,
+      "grad_norm": 0.4704447090625763,
+      "learning_rate": 0.00010264211369095275,
+      "loss": 0.4481,
+      "step": 3046
+    },
+    {
+      "epoch": 0.97504,
+      "grad_norm": 0.5194148421287537,
+      "learning_rate": 0.00010261008807045637,
+      "loss": 0.5955,
+      "step": 3047
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 0.5535142421722412,
+      "learning_rate": 0.00010257806244995998,
+      "loss": 0.5573,
+      "step": 3048
+    },
+    {
+      "epoch": 0.97568,
+      "grad_norm": 0.546876072883606,
+      "learning_rate": 0.00010254603682946358,
+      "loss": 0.4879,
+      "step": 3049
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.3922886252403259,
+      "learning_rate": 0.00010251401120896717,
+      "loss": 0.4996,
+      "step": 3050
+    },
+    {
+      "epoch": 0.97632,
+      "grad_norm": 0.64422607421875,
+      "learning_rate": 0.00010248198558847078,
+      "loss": 0.6216,
+      "step": 3051
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 0.4554085433483124,
+      "learning_rate": 0.00010244995996797437,
+      "loss": 0.6197,
+      "step": 3052
+    },
+    {
+      "epoch": 0.97696,
+      "grad_norm": 0.5509797930717468,
+      "learning_rate": 0.000102417934347478,
+      "loss": 0.5516,
+      "step": 3053
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 0.517092764377594,
+      "learning_rate": 0.00010238590872698159,
+      "loss": 0.6533,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 0.5829808712005615,
+      "learning_rate": 0.0001023538831064852,
+      "loss": 0.6138,
+      "step": 3055
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 0.5089106559753418,
+      "learning_rate": 0.00010232185748598879,
+      "loss": 0.5334,
+      "step": 3056
+    },
+    {
+      "epoch": 0.97824,
+      "grad_norm": 0.45938751101493835,
+      "learning_rate": 0.00010228983186549239,
+      "loss": 0.4805,
+      "step": 3057
+    },
+    {
+      "epoch": 0.97856,
+      "grad_norm": 0.5272316932678223,
+      "learning_rate": 0.00010225780624499598,
+      "loss": 0.6316,
+      "step": 3058
+    },
+    {
+      "epoch": 0.97888,
+      "grad_norm": 0.48077622056007385,
+      "learning_rate": 0.00010222578062449962,
+      "loss": 0.4291,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.483024924993515,
+      "learning_rate": 0.00010219375500400321,
+      "loss": 0.6569,
+      "step": 3060
+    },
+    {
+      "epoch": 0.97952,
+      "grad_norm": 0.5677964091300964,
+      "learning_rate": 0.00010216172938350681,
+      "loss": 0.6188,
+      "step": 3061
+    },
+    {
+      "epoch": 0.97984,
+      "grad_norm": 0.5234769582748413,
+      "learning_rate": 0.0001021297037630104,
+      "loss": 0.4425,
+      "step": 3062
+    },
+    {
+      "epoch": 0.98016,
+      "grad_norm": 0.47875210642814636,
+      "learning_rate": 0.00010209767814251401,
+      "loss": 0.5473,
+      "step": 3063
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 0.5013461112976074,
+      "learning_rate": 0.00010206565252201763,
+      "loss": 0.6343,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 0.4224475622177124,
+      "learning_rate": 0.00010203362690152123,
+      "loss": 0.3932,
+      "step": 3065
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 0.44129201769828796,
+      "learning_rate": 0.00010200160128102482,
+      "loss": 0.4914,
+      "step": 3066
+    },
+    {
+      "epoch": 0.98144,
+      "grad_norm": 0.4299464821815491,
+      "learning_rate": 0.00010196957566052843,
+      "loss": 0.4862,
+      "step": 3067
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 0.43140432238578796,
+      "learning_rate": 0.00010193755004003203,
+      "loss": 0.5241,
+      "step": 3068
+    },
+    {
+      "epoch": 0.98208,
+      "grad_norm": 0.4816804826259613,
+      "learning_rate": 0.00010190552441953562,
+      "loss": 0.5923,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 0.5560425519943237,
+      "learning_rate": 0.00010187349879903924,
+      "loss": 0.4267,
+      "step": 3070
+    },
+    {
+      "epoch": 0.98272,
+      "grad_norm": 0.5039703845977783,
+      "learning_rate": 0.00010184147317854285,
+      "loss": 0.6228,
+      "step": 3071
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 0.512065589427948,
+      "learning_rate": 0.00010180944755804645,
+      "loss": 0.5852,
+      "step": 3072
+    },
+    {
+      "epoch": 0.98336,
+      "grad_norm": 0.5495064854621887,
+      "learning_rate": 0.00010177742193755004,
+      "loss": 0.6361,
+      "step": 3073
+    },
+    {
+      "epoch": 0.98368,
+      "grad_norm": 0.4804792106151581,
+      "learning_rate": 0.00010174539631705365,
+      "loss": 0.5897,
+      "step": 3074
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.47852712869644165,
+      "learning_rate": 0.00010171337069655724,
+      "loss": 0.5235,
+      "step": 3075
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 0.5878530740737915,
+      "learning_rate": 0.00010168134507606087,
+      "loss": 0.524,
+      "step": 3076
+    },
+    {
+      "epoch": 0.98464,
+      "grad_norm": 0.5579085946083069,
+      "learning_rate": 0.00010164931945556446,
+      "loss": 0.5335,
+      "step": 3077
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 0.45107007026672363,
+      "learning_rate": 0.00010161729383506807,
+      "loss": 0.6571,
+      "step": 3078
+    },
+    {
+      "epoch": 0.98528,
+      "grad_norm": 0.512333869934082,
+      "learning_rate": 0.00010158526821457166,
+      "loss": 0.5725,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.6056711077690125,
+      "learning_rate": 0.00010155324259407526,
+      "loss": 0.7043,
+      "step": 3080
+    },
+    {
+      "epoch": 0.98592,
+      "grad_norm": 0.5937466025352478,
+      "learning_rate": 0.00010152121697357885,
+      "loss": 0.624,
+      "step": 3081
+    },
+    {
+      "epoch": 0.98624,
+      "grad_norm": 0.39029189944267273,
+      "learning_rate": 0.00010148919135308248,
+      "loss": 0.4805,
+      "step": 3082
+    },
+    {
+      "epoch": 0.98656,
+      "grad_norm": 0.6359745860099792,
+      "learning_rate": 0.00010145716573258608,
+      "loss": 0.5645,
+      "step": 3083
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 0.6606888771057129,
+      "learning_rate": 0.00010142514011208968,
+      "loss": 0.5904,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 0.46842578053474426,
+      "learning_rate": 0.00010139311449159327,
+      "loss": 0.53,
+      "step": 3085
+    },
+    {
+      "epoch": 0.98752,
+      "grad_norm": 0.5136080384254456,
+      "learning_rate": 0.00010136108887109688,
+      "loss": 0.5516,
+      "step": 3086
+    },
+    {
+      "epoch": 0.98784,
+      "grad_norm": 0.4495748281478882,
+      "learning_rate": 0.00010132906325060048,
+      "loss": 0.5097,
+      "step": 3087
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 0.5386287569999695,
+      "learning_rate": 0.0001012970376301041,
+      "loss": 0.5362,
+      "step": 3088
+    },
+    {
+      "epoch": 0.98848,
+      "grad_norm": 0.4556712210178375,
+      "learning_rate": 0.00010126501200960769,
+      "loss": 0.6031,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.5064831376075745,
+      "learning_rate": 0.0001012329863891113,
+      "loss": 0.5901,
+      "step": 3090
+    },
+    {
+      "epoch": 0.98912,
+      "grad_norm": 0.4786016643047333,
+      "learning_rate": 0.0001012009607686149,
+      "loss": 0.5961,
+      "step": 3091
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 0.46530410647392273,
+      "learning_rate": 0.00010116893514811849,
+      "loss": 0.493,
+      "step": 3092
+    },
+    {
+      "epoch": 0.98976,
+      "grad_norm": 0.5640050172805786,
+      "learning_rate": 0.0001011369095276221,
+      "loss": 0.5777,
+      "step": 3093
+    },
+    {
+      "epoch": 0.99008,
+      "grad_norm": 0.46274426579475403,
+      "learning_rate": 0.00010110488390712572,
+      "loss": 0.4834,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 0.4396246671676636,
+      "learning_rate": 0.0001010728582866293,
+      "loss": 0.4665,
+      "step": 3095
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 0.47993019223213196,
+      "learning_rate": 0.00010104083266613291,
+      "loss": 0.5557,
+      "step": 3096
+    },
+    {
+      "epoch": 0.99104,
+      "grad_norm": 0.2807310223579407,
+      "learning_rate": 0.00010100880704563652,
+      "loss": 0.4123,
+      "step": 3097
+    },
+    {
+      "epoch": 0.99136,
+      "grad_norm": 0.5596112608909607,
+      "learning_rate": 0.00010097678142514011,
+      "loss": 0.563,
+      "step": 3098
+    },
+    {
+      "epoch": 0.99168,
+      "grad_norm": 0.6127992868423462,
+      "learning_rate": 0.00010094475580464371,
+      "loss": 0.5116,
+      "step": 3099
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.5309988856315613,
+      "learning_rate": 0.00010091273018414733,
+      "loss": 0.5995,
+      "step": 3100
+    },
+    {
+      "epoch": 0.99232,
+      "grad_norm": 0.400471568107605,
+      "learning_rate": 0.00010088070456365093,
+      "loss": 0.5477,
+      "step": 3101
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 0.43969669938087463,
+      "learning_rate": 0.00010084867894315453,
+      "loss": 0.5262,
+      "step": 3102
+    },
+    {
+      "epoch": 0.99296,
+      "grad_norm": 0.4806485176086426,
+      "learning_rate": 0.00010081665332265813,
+      "loss": 0.5764,
+      "step": 3103
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 0.7100608944892883,
+      "learning_rate": 0.00010078462770216172,
+      "loss": 0.6877,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 0.43390902876853943,
+      "learning_rate": 0.00010075260208166533,
+      "loss": 0.5598,
+      "step": 3105
+    },
+    {
+      "epoch": 0.99392,
+      "grad_norm": 0.541824460029602,
+      "learning_rate": 0.00010072057646116895,
+      "loss": 0.7639,
+      "step": 3106
+    },
+    {
+      "epoch": 0.99424,
+      "grad_norm": 0.39421340823173523,
+      "learning_rate": 0.00010068855084067255,
+      "loss": 0.6404,
+      "step": 3107
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 0.47497430443763733,
+      "learning_rate": 0.00010065652522017614,
+      "loss": 0.5012,
+      "step": 3108
+    },
+    {
+      "epoch": 0.99488,
+      "grad_norm": 0.4536862075328827,
+      "learning_rate": 0.00010062449959967975,
+      "loss": 0.6008,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 0.48393845558166504,
+      "learning_rate": 0.00010059247397918334,
+      "loss": 0.5392,
+      "step": 3110
+    },
+    {
+      "epoch": 0.99552,
+      "grad_norm": 0.5237390398979187,
+      "learning_rate": 0.00010056044835868697,
+      "loss": 0.5849,
+      "step": 3111
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 0.5962594747543335,
+      "learning_rate": 0.00010052842273819056,
+      "loss": 0.5876,
+      "step": 3112
+    },
+    {
+      "epoch": 0.99616,
+      "grad_norm": 0.4717448353767395,
+      "learning_rate": 0.00010049639711769416,
+      "loss": 0.4952,
+      "step": 3113
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 0.5105111598968506,
+      "learning_rate": 0.00010046437149719776,
+      "loss": 0.5436,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 0.4244648218154907,
+      "learning_rate": 0.00010043234587670136,
+      "loss": 0.4427,
+      "step": 3115
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 0.5558398365974426,
+      "learning_rate": 0.00010040032025620497,
+      "loss": 0.5957,
+      "step": 3116
+    },
+    {
+      "epoch": 0.99744,
+      "grad_norm": 0.5488656163215637,
+      "learning_rate": 0.00010036829463570858,
+      "loss": 0.5151,
+      "step": 3117
+    },
+    {
+      "epoch": 0.99776,
+      "grad_norm": 0.5168829560279846,
+      "learning_rate": 0.00010033626901521218,
+      "loss": 0.5785,
+      "step": 3118
+    },
+    {
+      "epoch": 0.99808,
+      "grad_norm": 0.5107622742652893,
+      "learning_rate": 0.00010030424339471578,
+      "loss": 0.5256,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.4315459132194519,
+      "learning_rate": 0.00010027221777421938,
+      "loss": 0.4312,
+      "step": 3120
+    },
+    {
+      "epoch": 0.99872,
+      "grad_norm": 0.5616961121559143,
+      "learning_rate": 0.00010024019215372298,
+      "loss": 0.6478,
+      "step": 3121
+    },
+    {
+      "epoch": 0.99904,
+      "grad_norm": 0.5263671875,
+      "learning_rate": 0.00010020816653322658,
+      "loss": 0.604,
+      "step": 3122
+    },
+    {
+      "epoch": 0.99936,
+      "grad_norm": 0.48344191908836365,
+      "learning_rate": 0.0001001761409127302,
+      "loss": 0.5484,
+      "step": 3123
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 0.5167639851570129,
+      "learning_rate": 0.0001001441152922338,
+      "loss": 0.538,
+      "step": 3124
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43253278732299805,
+      "learning_rate": 0.0001001120896717374,
+      "loss": 0.5141,
+      "step": 3125
+    },
+    {
+      "epoch": 1.00032,
+      "grad_norm": 0.45655184984207153,
+      "learning_rate": 0.000100080064051241,
+      "loss": 0.638,
+      "step": 3126
+    },
+    {
+      "epoch": 1.00064,
+      "grad_norm": 0.4649527966976166,
+      "learning_rate": 0.00010004803843074459,
+      "loss": 0.5234,
+      "step": 3127
+    },
+    {
+      "epoch": 1.00096,
+      "grad_norm": 0.48761823773384094,
+      "learning_rate": 0.0001000160128102482,
+      "loss": 0.4864,
+      "step": 3128
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 0.4640171229839325,
+      "learning_rate": 9.99839871897518e-05,
+      "loss": 0.4042,
+      "step": 3129
+    },
+    {
+      "epoch": 1.0016,
+      "grad_norm": 0.4474526345729828,
+      "learning_rate": 9.995196156925542e-05,
+      "loss": 0.6553,
+      "step": 3130
+    },
+    {
+      "epoch": 1.00192,
+      "grad_norm": 0.5141965746879578,
+      "learning_rate": 9.991993594875901e-05,
+      "loss": 0.4731,
+      "step": 3131
+    },
+    {
+      "epoch": 1.00224,
+      "grad_norm": 0.4412883520126343,
+      "learning_rate": 9.988791032826261e-05,
+      "loss": 0.4082,
+      "step": 3132
+    },
+    {
+      "epoch": 1.00256,
+      "grad_norm": 0.5625013113021851,
+      "learning_rate": 9.985588470776622e-05,
+      "loss": 0.5194,
+      "step": 3133
+    },
+    {
+      "epoch": 1.00288,
+      "grad_norm": 0.4523092806339264,
+      "learning_rate": 9.982385908726982e-05,
+      "loss": 0.5789,
+      "step": 3134
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 0.4657382071018219,
+      "learning_rate": 9.979183346677342e-05,
+      "loss": 0.4639,
+      "step": 3135
+    },
+    {
+      "epoch": 1.00352,
+      "grad_norm": 0.5668757557868958,
+      "learning_rate": 9.975980784627703e-05,
+      "loss": 0.628,
+      "step": 3136
+    },
+    {
+      "epoch": 1.00384,
+      "grad_norm": 0.5699948668479919,
+      "learning_rate": 9.972778222578062e-05,
+      "loss": 0.496,
+      "step": 3137
+    },
+    {
+      "epoch": 1.00416,
+      "grad_norm": 0.44826385378837585,
+      "learning_rate": 9.969575660528423e-05,
+      "loss": 0.4379,
+      "step": 3138
+    },
+    {
+      "epoch": 1.00448,
+      "grad_norm": 0.5611833930015564,
+      "learning_rate": 9.966373098478783e-05,
+      "loss": 0.489,
+      "step": 3139
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 0.6799756288528442,
+      "learning_rate": 9.963170536429144e-05,
+      "loss": 0.5374,
+      "step": 3140
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 0.44884324073791504,
+      "learning_rate": 9.959967974379504e-05,
+      "loss": 0.4006,
+      "step": 3141
+    },
+    {
+      "epoch": 1.0054400000000001,
+      "grad_norm": 0.5898566246032715,
+      "learning_rate": 9.956765412329865e-05,
+      "loss": 0.6957,
+      "step": 3142
+    },
+    {
+      "epoch": 1.00576,
+      "grad_norm": 0.43386954069137573,
+      "learning_rate": 9.953562850280225e-05,
+      "loss": 0.4413,
+      "step": 3143
+    },
+    {
+      "epoch": 1.00608,
+      "grad_norm": 0.5893003344535828,
+      "learning_rate": 9.950360288230584e-05,
+      "loss": 0.4561,
+      "step": 3144
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 0.4829353988170624,
+      "learning_rate": 9.947157726180946e-05,
+      "loss": 0.4733,
+      "step": 3145
+    },
+    {
+      "epoch": 1.00672,
+      "grad_norm": 0.8538504242897034,
+      "learning_rate": 9.943955164131305e-05,
+      "loss": 0.4748,
+      "step": 3146
+    },
+    {
+      "epoch": 1.00704,
+      "grad_norm": 0.5145028233528137,
+      "learning_rate": 9.940752602081666e-05,
+      "loss": 0.5561,
+      "step": 3147
+    },
+    {
+      "epoch": 1.00736,
+      "grad_norm": 0.5006149411201477,
+      "learning_rate": 9.937550040032026e-05,
+      "loss": 0.3977,
+      "step": 3148
+    },
+    {
+      "epoch": 1.00768,
+      "grad_norm": 0.4364413619041443,
+      "learning_rate": 9.934347477982387e-05,
+      "loss": 0.5005,
+      "step": 3149
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 0.46541646122932434,
+      "learning_rate": 9.931144915932746e-05,
+      "loss": 0.4549,
+      "step": 3150
+    },
+    {
+      "epoch": 1.00832,
+      "grad_norm": 0.4935240149497986,
+      "learning_rate": 9.927942353883108e-05,
+      "loss": 0.4318,
+      "step": 3151
+    },
+    {
+      "epoch": 1.00864,
+      "grad_norm": 0.5767371654510498,
+      "learning_rate": 9.924739791833467e-05,
+      "loss": 0.4868,
+      "step": 3152
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 0.5868511199951172,
+      "learning_rate": 9.921537229783827e-05,
+      "loss": 0.5019,
+      "step": 3153
+    },
+    {
+      "epoch": 1.00928,
+      "grad_norm": 0.3959423899650574,
+      "learning_rate": 9.918334667734188e-05,
+      "loss": 0.4813,
+      "step": 3154
+    },
+    {
+      "epoch": 1.0096,
+      "grad_norm": 0.532892644405365,
+      "learning_rate": 9.915132105684548e-05,
+      "loss": 0.5825,
+      "step": 3155
+    },
+    {
+      "epoch": 1.00992,
+      "grad_norm": 0.5173769593238831,
+      "learning_rate": 9.911929543634907e-05,
+      "loss": 0.5541,
+      "step": 3156
+    },
+    {
+      "epoch": 1.01024,
+      "grad_norm": 0.4442068636417389,
+      "learning_rate": 9.908726981585269e-05,
+      "loss": 0.4641,
+      "step": 3157
+    },
+    {
+      "epoch": 1.01056,
+      "grad_norm": 0.6878951787948608,
+      "learning_rate": 9.905524419535628e-05,
+      "loss": 0.6623,
+      "step": 3158
+    },
+    {
+      "epoch": 1.01088,
+      "grad_norm": 0.39458736777305603,
+      "learning_rate": 9.902321857485989e-05,
+      "loss": 0.4686,
+      "step": 3159
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 0.43198344111442566,
+      "learning_rate": 9.89911929543635e-05,
+      "loss": 0.505,
+      "step": 3160
+    },
+    {
+      "epoch": 1.01152,
+      "grad_norm": 0.5163975954055786,
+      "learning_rate": 9.89591673338671e-05,
+      "loss": 0.435,
+      "step": 3161
+    },
+    {
+      "epoch": 1.01184,
+      "grad_norm": 0.47518056631088257,
+      "learning_rate": 9.89271417133707e-05,
+      "loss": 0.4662,
+      "step": 3162
+    },
+    {
+      "epoch": 1.01216,
+      "grad_norm": 0.6102754473686218,
+      "learning_rate": 9.889511609287431e-05,
+      "loss": 0.6733,
+      "step": 3163
+    },
+    {
+      "epoch": 1.01248,
+      "grad_norm": 0.41785678267478943,
+      "learning_rate": 9.886309047237791e-05,
+      "loss": 0.4402,
+      "step": 3164
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.4334068298339844,
+      "learning_rate": 9.88310648518815e-05,
+      "loss": 0.4971,
+      "step": 3165
+    },
+    {
+      "epoch": 1.01312,
+      "grad_norm": 0.512736976146698,
+      "learning_rate": 9.879903923138511e-05,
+      "loss": 0.4892,
+      "step": 3166
+    },
+    {
+      "epoch": 1.01344,
+      "grad_norm": 0.5426936745643616,
+      "learning_rate": 9.876701361088871e-05,
+      "loss": 0.4249,
+      "step": 3167
+    },
+    {
+      "epoch": 1.01376,
+      "grad_norm": 0.538696825504303,
+      "learning_rate": 9.873498799039232e-05,
+      "loss": 0.5328,
+      "step": 3168
+    },
+    {
+      "epoch": 1.01408,
+      "grad_norm": 0.4124504029750824,
+      "learning_rate": 9.870296236989592e-05,
+      "loss": 0.4043,
+      "step": 3169
+    },
+    {
+      "epoch": 1.0144,
+      "grad_norm": 0.44405296444892883,
+      "learning_rate": 9.867093674939953e-05,
+      "loss": 0.4545,
+      "step": 3170
+    },
+    {
+      "epoch": 1.01472,
+      "grad_norm": 0.4858121871948242,
+      "learning_rate": 9.863891112890312e-05,
+      "loss": 0.4755,
+      "step": 3171
+    },
+    {
+      "epoch": 1.01504,
+      "grad_norm": 0.525364875793457,
+      "learning_rate": 9.860688550840674e-05,
+      "loss": 0.5612,
+      "step": 3172
+    },
+    {
+      "epoch": 1.01536,
+      "grad_norm": 0.45628640055656433,
+      "learning_rate": 9.857485988791033e-05,
+      "loss": 0.3908,
+      "step": 3173
+    },
+    {
+      "epoch": 1.01568,
+      "grad_norm": 0.5402250289916992,
+      "learning_rate": 9.854283426741393e-05,
+      "loss": 0.3855,
+      "step": 3174
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 0.4358239471912384,
+      "learning_rate": 9.851080864691754e-05,
+      "loss": 0.4866,
+      "step": 3175
+    },
+    {
+      "epoch": 1.01632,
+      "grad_norm": 0.5325714349746704,
+      "learning_rate": 9.847878302642114e-05,
+      "loss": 0.4937,
+      "step": 3176
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 0.4827834367752075,
+      "learning_rate": 9.844675740592475e-05,
+      "loss": 0.3844,
+      "step": 3177
+    },
+    {
+      "epoch": 1.01696,
+      "grad_norm": 0.43882691860198975,
+      "learning_rate": 9.841473178542835e-05,
+      "loss": 0.6123,
+      "step": 3178
+    },
+    {
+      "epoch": 1.01728,
+      "grad_norm": 0.46227017045021057,
+      "learning_rate": 9.838270616493194e-05,
+      "loss": 0.5083,
+      "step": 3179
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 0.7243804931640625,
+      "learning_rate": 9.835068054443556e-05,
+      "loss": 0.3952,
+      "step": 3180
+    },
+    {
+      "epoch": 1.01792,
+      "grad_norm": 0.7266965508460999,
+      "learning_rate": 9.831865492393915e-05,
+      "loss": 0.6356,
+      "step": 3181
+    },
+    {
+      "epoch": 1.01824,
+      "grad_norm": 0.49397751688957214,
+      "learning_rate": 9.828662930344276e-05,
+      "loss": 0.4224,
+      "step": 3182
+    },
+    {
+      "epoch": 1.01856,
+      "grad_norm": 0.6040565967559814,
+      "learning_rate": 9.825460368294636e-05,
+      "loss": 0.5117,
+      "step": 3183
+    },
+    {
+      "epoch": 1.01888,
+      "grad_norm": 0.6237596273422241,
+      "learning_rate": 9.822257806244997e-05,
+      "loss": 0.6133,
+      "step": 3184
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 0.5112883448600769,
+      "learning_rate": 9.819055244195356e-05,
+      "loss": 0.3604,
+      "step": 3185
+    },
+    {
+      "epoch": 1.01952,
+      "grad_norm": 0.4346521496772766,
+      "learning_rate": 9.815852682145718e-05,
+      "loss": 0.5072,
+      "step": 3186
+    },
+    {
+      "epoch": 1.01984,
+      "grad_norm": 0.5658510327339172,
+      "learning_rate": 9.812650120096077e-05,
+      "loss": 0.5149,
+      "step": 3187
+    },
+    {
+      "epoch": 1.02016,
+      "grad_norm": 0.4263412356376648,
+      "learning_rate": 9.809447558046437e-05,
+      "loss": 0.4703,
+      "step": 3188
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 0.5694780349731445,
+      "learning_rate": 9.806244995996798e-05,
+      "loss": 0.5141,
+      "step": 3189
+    },
+    {
+      "epoch": 1.0208,
+      "grad_norm": 0.5049242377281189,
+      "learning_rate": 9.803042433947158e-05,
+      "loss": 0.5299,
+      "step": 3190
+    },
+    {
+      "epoch": 1.02112,
+      "grad_norm": 0.545043408870697,
+      "learning_rate": 9.799839871897519e-05,
+      "loss": 0.4386,
+      "step": 3191
+    },
+    {
+      "epoch": 1.02144,
+      "grad_norm": 0.5127218961715698,
+      "learning_rate": 9.796637309847879e-05,
+      "loss": 0.4446,
+      "step": 3192
+    },
+    {
+      "epoch": 1.02176,
+      "grad_norm": 0.47669246792793274,
+      "learning_rate": 9.79343474779824e-05,
+      "loss": 0.4163,
+      "step": 3193
+    },
+    {
+      "epoch": 1.02208,
+      "grad_norm": 0.5546985864639282,
+      "learning_rate": 9.790232185748599e-05,
+      "loss": 0.4512,
+      "step": 3194
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 0.44935575127601624,
+      "learning_rate": 9.78702962369896e-05,
+      "loss": 0.4803,
+      "step": 3195
+    },
+    {
+      "epoch": 1.02272,
+      "grad_norm": 0.6663330793380737,
+      "learning_rate": 9.78382706164932e-05,
+      "loss": 0.5343,
+      "step": 3196
+    },
+    {
+      "epoch": 1.02304,
+      "grad_norm": 0.4638069272041321,
+      "learning_rate": 9.78062449959968e-05,
+      "loss": 0.4416,
+      "step": 3197
+    },
+    {
+      "epoch": 1.02336,
+      "grad_norm": 0.5278282761573792,
+      "learning_rate": 9.777421937550041e-05,
+      "loss": 0.5955,
+      "step": 3198
+    },
+    {
+      "epoch": 1.02368,
+      "grad_norm": 0.5275051593780518,
+      "learning_rate": 9.774219375500401e-05,
+      "loss": 0.4947,
+      "step": 3199
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.4576311707496643,
+      "learning_rate": 9.77101681345076e-05,
+      "loss": 0.3396,
+      "step": 3200
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 0.5539082884788513,
+      "learning_rate": 9.767814251401122e-05,
+      "loss": 0.6097,
+      "step": 3201
+    },
+    {
+      "epoch": 1.02464,
+      "grad_norm": 0.5423197746276855,
+      "learning_rate": 9.764611689351481e-05,
+      "loss": 0.4369,
+      "step": 3202
+    },
+    {
+      "epoch": 1.02496,
+      "grad_norm": 0.5033709406852722,
+      "learning_rate": 9.761409127301842e-05,
+      "loss": 0.502,
+      "step": 3203
+    },
+    {
+      "epoch": 1.02528,
+      "grad_norm": 0.5225690603256226,
+      "learning_rate": 9.758206565252202e-05,
+      "loss": 0.5168,
+      "step": 3204
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 0.5134522914886475,
+      "learning_rate": 9.755004003202563e-05,
+      "loss": 0.507,
+      "step": 3205
+    },
+    {
+      "epoch": 1.02592,
+      "grad_norm": 0.62888103723526,
+      "learning_rate": 9.751801441152922e-05,
+      "loss": 0.5454,
+      "step": 3206
+    },
+    {
+      "epoch": 1.02624,
+      "grad_norm": 0.541764497756958,
+      "learning_rate": 9.748598879103284e-05,
+      "loss": 0.4505,
+      "step": 3207
+    },
+    {
+      "epoch": 1.02656,
+      "grad_norm": 0.5531131625175476,
+      "learning_rate": 9.745396317053643e-05,
+      "loss": 0.4161,
+      "step": 3208
+    },
+    {
+      "epoch": 1.02688,
+      "grad_norm": 0.5210992693901062,
+      "learning_rate": 9.742193755004003e-05,
+      "loss": 0.3907,
+      "step": 3209
+    },
+    {
+      "epoch": 1.0272,
+      "grad_norm": 0.448391318321228,
+      "learning_rate": 9.738991192954364e-05,
+      "loss": 0.3856,
+      "step": 3210
+    },
+    {
+      "epoch": 1.02752,
+      "grad_norm": 0.5762470960617065,
+      "learning_rate": 9.735788630904724e-05,
+      "loss": 0.5003,
+      "step": 3211
+    },
+    {
+      "epoch": 1.02784,
+      "grad_norm": 0.6128514409065247,
+      "learning_rate": 9.732586068855085e-05,
+      "loss": 0.5879,
+      "step": 3212
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 0.6512813568115234,
+      "learning_rate": 9.729383506805445e-05,
+      "loss": 0.4495,
+      "step": 3213
+    },
+    {
+      "epoch": 1.02848,
+      "grad_norm": 0.45827507972717285,
+      "learning_rate": 9.726180944755806e-05,
+      "loss": 0.4976,
+      "step": 3214
+    },
+    {
+      "epoch": 1.0288,
+      "grad_norm": 0.5699587464332581,
+      "learning_rate": 9.722978382706165e-05,
+      "loss": 0.5276,
+      "step": 3215
+    },
+    {
+      "epoch": 1.02912,
+      "grad_norm": 0.5579009056091309,
+      "learning_rate": 9.719775820656527e-05,
+      "loss": 0.5281,
+      "step": 3216
+    },
+    {
+      "epoch": 1.02944,
+      "grad_norm": 0.4773772358894348,
+      "learning_rate": 9.716573258606886e-05,
+      "loss": 0.5266,
+      "step": 3217
+    },
+    {
+      "epoch": 1.02976,
+      "grad_norm": 0.5042899250984192,
+      "learning_rate": 9.713370696557246e-05,
+      "loss": 0.4449,
+      "step": 3218
+    },
+    {
+      "epoch": 1.03008,
+      "grad_norm": 0.5478249788284302,
+      "learning_rate": 9.710168134507607e-05,
+      "loss": 0.4699,
+      "step": 3219
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 0.44113004207611084,
+      "learning_rate": 9.706965572457967e-05,
+      "loss": 0.398,
+      "step": 3220
+    },
+    {
+      "epoch": 1.03072,
+      "grad_norm": 0.5433489084243774,
+      "learning_rate": 9.703763010408326e-05,
+      "loss": 0.4719,
+      "step": 3221
+    },
+    {
+      "epoch": 1.03104,
+      "grad_norm": 0.5132033824920654,
+      "learning_rate": 9.700560448358688e-05,
+      "loss": 0.4403,
+      "step": 3222
+    },
+    {
+      "epoch": 1.03136,
+      "grad_norm": 0.5915552377700806,
+      "learning_rate": 9.697357886309047e-05,
+      "loss": 0.4307,
+      "step": 3223
+    },
+    {
+      "epoch": 1.03168,
+      "grad_norm": 0.6902591586112976,
+      "learning_rate": 9.694155324259409e-05,
+      "loss": 0.4801,
+      "step": 3224
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.5257161855697632,
+      "learning_rate": 9.690952762209768e-05,
+      "loss": 0.4812,
+      "step": 3225
+    },
+    {
+      "epoch": 1.03232,
+      "grad_norm": 0.6038640141487122,
+      "learning_rate": 9.687750200160129e-05,
+      "loss": 0.5043,
+      "step": 3226
+    },
+    {
+      "epoch": 1.03264,
+      "grad_norm": 0.45142850279808044,
+      "learning_rate": 9.684547638110489e-05,
+      "loss": 0.4526,
+      "step": 3227
+    },
+    {
+      "epoch": 1.03296,
+      "grad_norm": 0.5048863291740417,
+      "learning_rate": 9.68134507606085e-05,
+      "loss": 0.4376,
+      "step": 3228
+    },
+    {
+      "epoch": 1.03328,
+      "grad_norm": 0.4574034512042999,
+      "learning_rate": 9.678142514011209e-05,
+      "loss": 0.5356,
+      "step": 3229
+    },
+    {
+      "epoch": 1.0336,
+      "grad_norm": 0.49772360920906067,
+      "learning_rate": 9.67493995196157e-05,
+      "loss": 0.4522,
+      "step": 3230
+    },
+    {
+      "epoch": 1.03392,
+      "grad_norm": 0.5006243586540222,
+      "learning_rate": 9.67173738991193e-05,
+      "loss": 0.4328,
+      "step": 3231
+    },
+    {
+      "epoch": 1.03424,
+      "grad_norm": 0.44997450709342957,
+      "learning_rate": 9.66853482786229e-05,
+      "loss": 0.4517,
+      "step": 3232
+    },
+    {
+      "epoch": 1.03456,
+      "grad_norm": 0.5653454065322876,
+      "learning_rate": 9.66533226581265e-05,
+      "loss": 0.5525,
+      "step": 3233
+    },
+    {
+      "epoch": 1.03488,
+      "grad_norm": 0.5425100922584534,
+      "learning_rate": 9.662129703763011e-05,
+      "loss": 0.5006,
+      "step": 3234
+    },
+    {
+      "epoch": 1.0352,
+      "grad_norm": 0.5540213584899902,
+      "learning_rate": 9.658927141713372e-05,
+      "loss": 0.5523,
+      "step": 3235
+    },
+    {
+      "epoch": 1.03552,
+      "grad_norm": 0.5400300621986389,
+      "learning_rate": 9.655724579663732e-05,
+      "loss": 0.5053,
+      "step": 3236
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 0.49527254700660706,
+      "learning_rate": 9.652522017614091e-05,
+      "loss": 0.5005,
+      "step": 3237
+    },
+    {
+      "epoch": 1.03616,
+      "grad_norm": 0.3400459587574005,
+      "learning_rate": 9.649319455564452e-05,
+      "loss": 0.3677,
+      "step": 3238
+    },
+    {
+      "epoch": 1.03648,
+      "grad_norm": 0.49747148156166077,
+      "learning_rate": 9.646116893514812e-05,
+      "loss": 0.3633,
+      "step": 3239
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.5073350667953491,
+      "learning_rate": 9.642914331465173e-05,
+      "loss": 0.5513,
+      "step": 3240
+    },
+    {
+      "epoch": 1.03712,
+      "grad_norm": 0.48758184909820557,
+      "learning_rate": 9.639711769415533e-05,
+      "loss": 0.4506,
+      "step": 3241
+    },
+    {
+      "epoch": 1.03744,
+      "grad_norm": 0.5085268020629883,
+      "learning_rate": 9.636509207365894e-05,
+      "loss": 0.5138,
+      "step": 3242
+    },
+    {
+      "epoch": 1.03776,
+      "grad_norm": 0.5730687379837036,
+      "learning_rate": 9.633306645316254e-05,
+      "loss": 0.6413,
+      "step": 3243
+    },
+    {
+      "epoch": 1.03808,
+      "grad_norm": 0.45717570185661316,
+      "learning_rate": 9.630104083266613e-05,
+      "loss": 0.4205,
+      "step": 3244
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 0.44563111662864685,
+      "learning_rate": 9.626901521216975e-05,
+      "loss": 0.4824,
+      "step": 3245
+    },
+    {
+      "epoch": 1.03872,
+      "grad_norm": 0.6498026847839355,
+      "learning_rate": 9.623698959167334e-05,
+      "loss": 0.4281,
+      "step": 3246
+    },
+    {
+      "epoch": 1.03904,
+      "grad_norm": 0.5102266669273376,
+      "learning_rate": 9.620496397117695e-05,
+      "loss": 0.5593,
+      "step": 3247
+    },
+    {
+      "epoch": 1.03936,
+      "grad_norm": 0.5262247920036316,
+      "learning_rate": 9.617293835068055e-05,
+      "loss": 0.4394,
+      "step": 3248
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 0.5720447897911072,
+      "learning_rate": 9.614091273018416e-05,
+      "loss": 0.5879,
+      "step": 3249
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6955535411834717,
+      "learning_rate": 9.610888710968775e-05,
+      "loss": 0.4996,
+      "step": 3250
+    },
+    {
+      "epoch": 1.04032,
+      "grad_norm": 0.4705846309661865,
+      "learning_rate": 9.607686148919136e-05,
+      "loss": 0.3749,
+      "step": 3251
+    },
+    {
+      "epoch": 1.04064,
+      "grad_norm": 0.5823104977607727,
+      "learning_rate": 9.604483586869496e-05,
+      "loss": 0.5984,
+      "step": 3252
+    },
+    {
+      "epoch": 1.04096,
+      "grad_norm": 0.595148503780365,
+      "learning_rate": 9.601281024819856e-05,
+      "loss": 0.519,
+      "step": 3253
+    },
+    {
+      "epoch": 1.04128,
+      "grad_norm": 0.43121063709259033,
+      "learning_rate": 9.598078462770217e-05,
+      "loss": 0.4085,
+      "step": 3254
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 0.5400832891464233,
+      "learning_rate": 9.594875900720577e-05,
+      "loss": 0.4962,
+      "step": 3255
+    },
+    {
+      "epoch": 1.04192,
+      "grad_norm": 0.4948318600654602,
+      "learning_rate": 9.591673338670936e-05,
+      "loss": 0.3665,
+      "step": 3256
+    },
+    {
+      "epoch": 1.04224,
+      "grad_norm": 0.5087904334068298,
+      "learning_rate": 9.588470776621298e-05,
+      "loss": 0.4954,
+      "step": 3257
+    },
+    {
+      "epoch": 1.04256,
+      "grad_norm": 0.5108175873756409,
+      "learning_rate": 9.585268214571657e-05,
+      "loss": 0.434,
+      "step": 3258
+    },
+    {
+      "epoch": 1.04288,
+      "grad_norm": 0.5623878240585327,
+      "learning_rate": 9.582065652522018e-05,
+      "loss": 0.4658,
+      "step": 3259
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 0.6789810061454773,
+      "learning_rate": 9.578863090472378e-05,
+      "loss": 0.487,
+      "step": 3260
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 0.5976998805999756,
+      "learning_rate": 9.575660528422739e-05,
+      "loss": 0.5488,
+      "step": 3261
+    },
+    {
+      "epoch": 1.04384,
+      "grad_norm": 0.5862056612968445,
+      "learning_rate": 9.572457966373099e-05,
+      "loss": 0.488,
+      "step": 3262
+    },
+    {
+      "epoch": 1.04416,
+      "grad_norm": 0.7088676691055298,
+      "learning_rate": 9.56925540432346e-05,
+      "loss": 0.5457,
+      "step": 3263
+    },
+    {
+      "epoch": 1.04448,
+      "grad_norm": 0.515631377696991,
+      "learning_rate": 9.56605284227382e-05,
+      "loss": 0.4542,
+      "step": 3264
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 0.4350857138633728,
+      "learning_rate": 9.562850280224179e-05,
+      "loss": 0.433,
+      "step": 3265
+    },
+    {
+      "epoch": 1.04512,
+      "grad_norm": 0.565494954586029,
+      "learning_rate": 9.559647718174541e-05,
+      "loss": 0.5823,
+      "step": 3266
+    },
+    {
+      "epoch": 1.04544,
+      "grad_norm": 0.6151624917984009,
+      "learning_rate": 9.5564451561249e-05,
+      "loss": 0.4479,
+      "step": 3267
+    },
+    {
+      "epoch": 1.04576,
+      "grad_norm": 0.5265957713127136,
+      "learning_rate": 9.55324259407526e-05,
+      "loss": 0.4344,
+      "step": 3268
+    },
+    {
+      "epoch": 1.04608,
+      "grad_norm": 0.5871798396110535,
+      "learning_rate": 9.550040032025621e-05,
+      "loss": 0.5777,
+      "step": 3269
+    },
+    {
+      "epoch": 1.0464,
+      "grad_norm": 0.49267587065696716,
+      "learning_rate": 9.546837469975981e-05,
+      "loss": 0.3743,
+      "step": 3270
+    },
+    {
+      "epoch": 1.04672,
+      "grad_norm": 0.6344022154808044,
+      "learning_rate": 9.543634907926342e-05,
+      "loss": 0.5044,
+      "step": 3271
+    },
+    {
+      "epoch": 1.04704,
+      "grad_norm": 0.5550808906555176,
+      "learning_rate": 9.540432345876702e-05,
+      "loss": 0.5057,
+      "step": 3272
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 0.549550473690033,
+      "learning_rate": 9.537229783827062e-05,
+      "loss": 0.528,
+      "step": 3273
+    },
+    {
+      "epoch": 1.04768,
+      "grad_norm": 0.49615827202796936,
+      "learning_rate": 9.534027221777423e-05,
+      "loss": 0.4322,
+      "step": 3274
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 0.5946497917175293,
+      "learning_rate": 9.530824659727782e-05,
+      "loss": 0.5298,
+      "step": 3275
+    },
+    {
+      "epoch": 1.04832,
+      "grad_norm": 0.7150006294250488,
+      "learning_rate": 9.527622097678143e-05,
+      "loss": 0.4212,
+      "step": 3276
+    },
+    {
+      "epoch": 1.04864,
+      "grad_norm": 0.45385053753852844,
+      "learning_rate": 9.524419535628503e-05,
+      "loss": 0.5083,
+      "step": 3277
+    },
+    {
+      "epoch": 1.04896,
+      "grad_norm": 0.5154277682304382,
+      "learning_rate": 9.521216973578864e-05,
+      "loss": 0.4704,
+      "step": 3278
+    },
+    {
+      "epoch": 1.04928,
+      "grad_norm": 0.501356840133667,
+      "learning_rate": 9.518014411529223e-05,
+      "loss": 0.4181,
+      "step": 3279
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.4946325421333313,
+      "learning_rate": 9.514811849479585e-05,
+      "loss": 0.4233,
+      "step": 3280
+    },
+    {
+      "epoch": 1.04992,
+      "grad_norm": 0.5499904751777649,
+      "learning_rate": 9.511609287429944e-05,
+      "loss": 0.5135,
+      "step": 3281
+    },
+    {
+      "epoch": 1.05024,
+      "grad_norm": 0.5946584343910217,
+      "learning_rate": 9.508406725380304e-05,
+      "loss": 0.54,
+      "step": 3282
+    },
+    {
+      "epoch": 1.05056,
+      "grad_norm": 0.4946201741695404,
+      "learning_rate": 9.505204163330665e-05,
+      "loss": 0.4457,
+      "step": 3283
+    },
+    {
+      "epoch": 1.05088,
+      "grad_norm": 0.4820703864097595,
+      "learning_rate": 9.502001601281025e-05,
+      "loss": 0.5795,
+      "step": 3284
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.45555973052978516,
+      "learning_rate": 9.498799039231386e-05,
+      "loss": 0.3578,
+      "step": 3285
+    },
+    {
+      "epoch": 1.05152,
+      "grad_norm": 0.6241341829299927,
+      "learning_rate": 9.495596477181746e-05,
+      "loss": 0.5596,
+      "step": 3286
+    },
+    {
+      "epoch": 1.0518399999999999,
+      "grad_norm": 0.5309383869171143,
+      "learning_rate": 9.492393915132107e-05,
+      "loss": 0.4671,
+      "step": 3287
+    },
+    {
+      "epoch": 1.05216,
+      "grad_norm": 0.568560779094696,
+      "learning_rate": 9.489191353082466e-05,
+      "loss": 0.4507,
+      "step": 3288
+    },
+    {
+      "epoch": 1.05248,
+      "grad_norm": 0.48555371165275574,
+      "learning_rate": 9.485988791032828e-05,
+      "loss": 0.3913,
+      "step": 3289
+    },
+    {
+      "epoch": 1.0528,
+      "grad_norm": 0.5702592730522156,
+      "learning_rate": 9.482786228983187e-05,
+      "loss": 0.6016,
+      "step": 3290
+    },
+    {
+      "epoch": 1.05312,
+      "grad_norm": 0.5249632000923157,
+      "learning_rate": 9.479583666933547e-05,
+      "loss": 0.5158,
+      "step": 3291
+    },
+    {
+      "epoch": 1.05344,
+      "grad_norm": 0.5645369291305542,
+      "learning_rate": 9.476381104883908e-05,
+      "loss": 0.4687,
+      "step": 3292
+    },
+    {
+      "epoch": 1.05376,
+      "grad_norm": 0.5648180842399597,
+      "learning_rate": 9.473178542834268e-05,
+      "loss": 0.4365,
+      "step": 3293
+    },
+    {
+      "epoch": 1.05408,
+      "grad_norm": 0.5006679892539978,
+      "learning_rate": 9.469975980784627e-05,
+      "loss": 0.4509,
+      "step": 3294
+    },
+    {
+      "epoch": 1.0544,
+      "grad_norm": 0.4492446482181549,
+      "learning_rate": 9.466773418734989e-05,
+      "loss": 0.3783,
+      "step": 3295
+    },
+    {
+      "epoch": 1.05472,
+      "grad_norm": 0.5567432641983032,
+      "learning_rate": 9.463570856685348e-05,
+      "loss": 0.4717,
+      "step": 3296
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 0.6416850090026855,
+      "learning_rate": 9.460368294635709e-05,
+      "loss": 0.5617,
+      "step": 3297
+    },
+    {
+      "epoch": 1.05536,
+      "grad_norm": 0.5513823628425598,
+      "learning_rate": 9.45716573258607e-05,
+      "loss": 0.5906,
+      "step": 3298
+    },
+    {
+      "epoch": 1.05568,
+      "grad_norm": 0.5292769074440002,
+      "learning_rate": 9.45396317053643e-05,
+      "loss": 0.595,
+      "step": 3299
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 0.6147984862327576,
+      "learning_rate": 9.450760608486789e-05,
+      "loss": 0.4259,
+      "step": 3300
+    },
+    {
+      "epoch": 1.05632,
+      "grad_norm": 0.5706994533538818,
+      "learning_rate": 9.447558046437151e-05,
+      "loss": 0.4291,
+      "step": 3301
+    },
+    {
+      "epoch": 1.05664,
+      "grad_norm": 0.47699660062789917,
+      "learning_rate": 9.44435548438751e-05,
+      "loss": 0.4107,
+      "step": 3302
+    },
+    {
+      "epoch": 1.05696,
+      "grad_norm": 0.6384992599487305,
+      "learning_rate": 9.44115292233787e-05,
+      "loss": 0.4354,
+      "step": 3303
+    },
+    {
+      "epoch": 1.05728,
+      "grad_norm": 0.5774004459381104,
+      "learning_rate": 9.437950360288231e-05,
+      "loss": 0.496,
+      "step": 3304
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 0.4981553852558136,
+      "learning_rate": 9.434747798238591e-05,
+      "loss": 0.4272,
+      "step": 3305
+    },
+    {
+      "epoch": 1.05792,
+      "grad_norm": 0.5606350302696228,
+      "learning_rate": 9.431545236188952e-05,
+      "loss": 0.4357,
+      "step": 3306
+    },
+    {
+      "epoch": 1.05824,
+      "grad_norm": 0.7529103755950928,
+      "learning_rate": 9.428342674139312e-05,
+      "loss": 0.5737,
+      "step": 3307
+    },
+    {
+      "epoch": 1.05856,
+      "grad_norm": 0.6347710490226746,
+      "learning_rate": 9.425140112089671e-05,
+      "loss": 0.5074,
+      "step": 3308
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 0.6201152205467224,
+      "learning_rate": 9.421937550040032e-05,
+      "loss": 0.6585,
+      "step": 3309
+    },
+    {
+      "epoch": 1.0592,
+      "grad_norm": 0.5795942544937134,
+      "learning_rate": 9.418734987990392e-05,
+      "loss": 0.4913,
+      "step": 3310
+    },
+    {
+      "epoch": 1.05952,
+      "grad_norm": 0.6341867446899414,
+      "learning_rate": 9.415532425940753e-05,
+      "loss": 0.4974,
+      "step": 3311
+    },
+    {
+      "epoch": 1.05984,
+      "grad_norm": 0.4143245816230774,
+      "learning_rate": 9.412329863891113e-05,
+      "loss": 0.3976,
+      "step": 3312
+    },
+    {
+      "epoch": 1.06016,
+      "grad_norm": 0.46689024567604065,
+      "learning_rate": 9.409127301841474e-05,
+      "loss": 0.4265,
+      "step": 3313
+    },
+    {
+      "epoch": 1.06048,
+      "grad_norm": 0.5957546234130859,
+      "learning_rate": 9.405924739791834e-05,
+      "loss": 0.512,
+      "step": 3314
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 0.5369313955307007,
+      "learning_rate": 9.402722177742193e-05,
+      "loss": 0.5036,
+      "step": 3315
+    },
+    {
+      "epoch": 1.06112,
+      "grad_norm": 0.4967133402824402,
+      "learning_rate": 9.399519615692555e-05,
+      "loss": 0.4245,
+      "step": 3316
+    },
+    {
+      "epoch": 1.06144,
+      "grad_norm": 0.47401732206344604,
+      "learning_rate": 9.396317053642914e-05,
+      "loss": 0.4083,
+      "step": 3317
+    },
+    {
+      "epoch": 1.06176,
+      "grad_norm": 0.42758262157440186,
+      "learning_rate": 9.393114491593276e-05,
+      "loss": 0.3951,
+      "step": 3318
+    },
+    {
+      "epoch": 1.06208,
+      "grad_norm": 0.46983301639556885,
+      "learning_rate": 9.389911929543635e-05,
+      "loss": 0.4384,
+      "step": 3319
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.4966285228729248,
+      "learning_rate": 9.386709367493996e-05,
+      "loss": 0.4871,
+      "step": 3320
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 0.5185624957084656,
+      "learning_rate": 9.383506805444356e-05,
+      "loss": 0.5208,
+      "step": 3321
+    },
+    {
+      "epoch": 1.06304,
+      "grad_norm": 0.6251195073127747,
+      "learning_rate": 9.380304243394717e-05,
+      "loss": 0.5647,
+      "step": 3322
+    },
+    {
+      "epoch": 1.06336,
+      "grad_norm": 0.5663153529167175,
+      "learning_rate": 9.377101681345076e-05,
+      "loss": 0.4351,
+      "step": 3323
+    },
+    {
+      "epoch": 1.06368,
+      "grad_norm": 0.41573289036750793,
+      "learning_rate": 9.373899119295438e-05,
+      "loss": 0.3499,
+      "step": 3324
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.5780336260795593,
+      "learning_rate": 9.370696557245797e-05,
+      "loss": 0.46,
+      "step": 3325
+    },
+    {
+      "epoch": 1.06432,
+      "grad_norm": 0.6110028028488159,
+      "learning_rate": 9.367493995196157e-05,
+      "loss": 0.567,
+      "step": 3326
+    },
+    {
+      "epoch": 1.06464,
+      "grad_norm": 0.5437237620353699,
+      "learning_rate": 9.364291433146518e-05,
+      "loss": 0.4308,
+      "step": 3327
+    },
+    {
+      "epoch": 1.06496,
+      "grad_norm": 0.4595336616039276,
+      "learning_rate": 9.361088871096878e-05,
+      "loss": 0.3195,
+      "step": 3328
+    },
+    {
+      "epoch": 1.06528,
+      "grad_norm": 0.406402587890625,
+      "learning_rate": 9.357886309047237e-05,
+      "loss": 0.4271,
+      "step": 3329
+    },
+    {
+      "epoch": 1.0656,
+      "grad_norm": 0.5166780352592468,
+      "learning_rate": 9.354683746997599e-05,
+      "loss": 0.4892,
+      "step": 3330
+    },
+    {
+      "epoch": 1.06592,
+      "grad_norm": 0.44450658559799194,
+      "learning_rate": 9.351481184947958e-05,
+      "loss": 0.4622,
+      "step": 3331
+    },
+    {
+      "epoch": 1.06624,
+      "grad_norm": 0.5293334722518921,
+      "learning_rate": 9.348278622898319e-05,
+      "loss": 0.5243,
+      "step": 3332
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 0.631640613079071,
+      "learning_rate": 9.345076060848679e-05,
+      "loss": 0.6041,
+      "step": 3333
+    },
+    {
+      "epoch": 1.06688,
+      "grad_norm": 0.5886451601982117,
+      "learning_rate": 9.34187349879904e-05,
+      "loss": 0.4956,
+      "step": 3334
+    },
+    {
+      "epoch": 1.0672,
+      "grad_norm": 0.48117974400520325,
+      "learning_rate": 9.3386709367494e-05,
+      "loss": 0.4183,
+      "step": 3335
+    },
+    {
+      "epoch": 1.06752,
+      "grad_norm": 0.4902232885360718,
+      "learning_rate": 9.33546837469976e-05,
+      "loss": 0.5294,
+      "step": 3336
+    },
+    {
+      "epoch": 1.06784,
+      "grad_norm": 0.6058788299560547,
+      "learning_rate": 9.332265812650121e-05,
+      "loss": 0.5427,
+      "step": 3337
+    },
+    {
+      "epoch": 1.06816,
+      "grad_norm": 0.5566360950469971,
+      "learning_rate": 9.32906325060048e-05,
+      "loss": 0.4481,
+      "step": 3338
+    },
+    {
+      "epoch": 1.06848,
+      "grad_norm": 0.6075891852378845,
+      "learning_rate": 9.325860688550842e-05,
+      "loss": 0.5991,
+      "step": 3339
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 0.5355103611946106,
+      "learning_rate": 9.322658126501201e-05,
+      "loss": 0.4821,
+      "step": 3340
+    },
+    {
+      "epoch": 1.06912,
+      "grad_norm": 0.6230280995368958,
+      "learning_rate": 9.319455564451562e-05,
+      "loss": 0.5766,
+      "step": 3341
+    },
+    {
+      "epoch": 1.06944,
+      "grad_norm": 0.4933406114578247,
+      "learning_rate": 9.316253002401922e-05,
+      "loss": 0.5013,
+      "step": 3342
+    },
+    {
+      "epoch": 1.06976,
+      "grad_norm": 0.46601006388664246,
+      "learning_rate": 9.313050440352283e-05,
+      "loss": 0.4169,
+      "step": 3343
+    },
+    {
+      "epoch": 1.07008,
+      "grad_norm": 0.5325000286102295,
+      "learning_rate": 9.309847878302642e-05,
+      "loss": 0.5059,
+      "step": 3344
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.5610787868499756,
+      "learning_rate": 9.306645316253004e-05,
+      "loss": 0.4542,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0707200000000001,
+      "grad_norm": 0.5495885014533997,
+      "learning_rate": 9.303442754203363e-05,
+      "loss": 0.6251,
+      "step": 3346
+    },
+    {
+      "epoch": 1.07104,
+      "grad_norm": 0.5210787057876587,
+      "learning_rate": 9.300240192153723e-05,
+      "loss": 0.504,
+      "step": 3347
+    },
+    {
+      "epoch": 1.07136,
+      "grad_norm": 0.6026103496551514,
+      "learning_rate": 9.297037630104084e-05,
+      "loss": 0.4497,
+      "step": 3348
+    },
+    {
+      "epoch": 1.07168,
+      "grad_norm": 0.5406185388565063,
+      "learning_rate": 9.293835068054444e-05,
+      "loss": 0.4586,
+      "step": 3349
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 0.4727422893047333,
+      "learning_rate": 9.290632506004803e-05,
+      "loss": 0.5207,
+      "step": 3350
+    },
+    {
+      "epoch": 1.07232,
+      "grad_norm": 0.5732471346855164,
+      "learning_rate": 9.287429943955165e-05,
+      "loss": 0.4476,
+      "step": 3351
+    },
+    {
+      "epoch": 1.07264,
+      "grad_norm": 0.42973798513412476,
+      "learning_rate": 9.284227381905524e-05,
+      "loss": 0.3687,
+      "step": 3352
+    },
+    {
+      "epoch": 1.07296,
+      "grad_norm": 0.535179853439331,
+      "learning_rate": 9.281024819855885e-05,
+      "loss": 0.4209,
+      "step": 3353
+    },
+    {
+      "epoch": 1.07328,
+      "grad_norm": 0.45106327533721924,
+      "learning_rate": 9.277822257806245e-05,
+      "loss": 0.4059,
+      "step": 3354
+    },
+    {
+      "epoch": 1.0735999999999999,
+      "grad_norm": 0.48119714856147766,
+      "learning_rate": 9.274619695756606e-05,
+      "loss": 0.455,
+      "step": 3355
+    },
+    {
+      "epoch": 1.07392,
+      "grad_norm": 0.4793301224708557,
+      "learning_rate": 9.271417133706966e-05,
+      "loss": 0.5226,
+      "step": 3356
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 0.5729004740715027,
+      "learning_rate": 9.268214571657327e-05,
+      "loss": 0.493,
+      "step": 3357
+    },
+    {
+      "epoch": 1.07456,
+      "grad_norm": 0.48106321692466736,
+      "learning_rate": 9.265012009607687e-05,
+      "loss": 0.4576,
+      "step": 3358
+    },
+    {
+      "epoch": 1.07488,
+      "grad_norm": 0.7141630053520203,
+      "learning_rate": 9.261809447558046e-05,
+      "loss": 0.4994,
+      "step": 3359
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.5154789686203003,
+      "learning_rate": 9.258606885508408e-05,
+      "loss": 0.4323,
+      "step": 3360
+    },
+    {
+      "epoch": 1.07552,
+      "grad_norm": 0.4302586019039154,
+      "learning_rate": 9.255404323458767e-05,
+      "loss": 0.354,
+      "step": 3361
+    },
+    {
+      "epoch": 1.07584,
+      "grad_norm": 0.5883063673973083,
+      "learning_rate": 9.252201761409128e-05,
+      "loss": 0.493,
+      "step": 3362
+    },
+    {
+      "epoch": 1.07616,
+      "grad_norm": 0.4734477698802948,
+      "learning_rate": 9.248999199359488e-05,
+      "loss": 0.4767,
+      "step": 3363
+    },
+    {
+      "epoch": 1.07648,
+      "grad_norm": 0.4713309705257416,
+      "learning_rate": 9.245796637309849e-05,
+      "loss": 0.3798,
+      "step": 3364
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 0.44680917263031006,
+      "learning_rate": 9.242594075260209e-05,
+      "loss": 0.3977,
+      "step": 3365
+    },
+    {
+      "epoch": 1.07712,
+      "grad_norm": 0.6004128456115723,
+      "learning_rate": 9.23939151321057e-05,
+      "loss": 0.5857,
+      "step": 3366
+    },
+    {
+      "epoch": 1.07744,
+      "grad_norm": 0.5143663287162781,
+      "learning_rate": 9.236188951160929e-05,
+      "loss": 0.4383,
+      "step": 3367
+    },
+    {
+      "epoch": 1.07776,
+      "grad_norm": 0.5081167817115784,
+      "learning_rate": 9.23298638911129e-05,
+      "loss": 0.4356,
+      "step": 3368
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 0.6379255056381226,
+      "learning_rate": 9.22978382706165e-05,
+      "loss": 0.5892,
+      "step": 3369
+    },
+    {
+      "epoch": 1.0784,
+      "grad_norm": 0.6310507655143738,
+      "learning_rate": 9.22658126501201e-05,
+      "loss": 0.4539,
+      "step": 3370
+    },
+    {
+      "epoch": 1.07872,
+      "grad_norm": 0.5054520964622498,
+      "learning_rate": 9.22337870296237e-05,
+      "loss": 0.4723,
+      "step": 3371
+    },
+    {
+      "epoch": 1.07904,
+      "grad_norm": 0.47777867317199707,
+      "learning_rate": 9.220176140912731e-05,
+      "loss": 0.3731,
+      "step": 3372
+    },
+    {
+      "epoch": 1.07936,
+      "grad_norm": 0.5210418105125427,
+      "learning_rate": 9.21697357886309e-05,
+      "loss": 0.4493,
+      "step": 3373
+    },
+    {
+      "epoch": 1.07968,
+      "grad_norm": 0.5359205007553101,
+      "learning_rate": 9.213771016813452e-05,
+      "loss": 0.46,
+      "step": 3374
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.62983638048172,
+      "learning_rate": 9.210568454763811e-05,
+      "loss": 0.4662,
+      "step": 3375
+    },
+    {
+      "epoch": 1.08032,
+      "grad_norm": 0.4354182183742523,
+      "learning_rate": 9.207365892714172e-05,
+      "loss": 0.3987,
+      "step": 3376
+    },
+    {
+      "epoch": 1.08064,
+      "grad_norm": 0.5192504525184631,
+      "learning_rate": 9.204163330664532e-05,
+      "loss": 0.4285,
+      "step": 3377
+    },
+    {
+      "epoch": 1.08096,
+      "grad_norm": 0.6054072380065918,
+      "learning_rate": 9.200960768614893e-05,
+      "loss": 0.5549,
+      "step": 3378
+    },
+    {
+      "epoch": 1.08128,
+      "grad_norm": 0.47363612055778503,
+      "learning_rate": 9.197758206565252e-05,
+      "loss": 0.4201,
+      "step": 3379
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 0.5465653538703918,
+      "learning_rate": 9.194555644515613e-05,
+      "loss": 0.5519,
+      "step": 3380
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 0.42709025740623474,
+      "learning_rate": 9.191353082465973e-05,
+      "loss": 0.3219,
+      "step": 3381
+    },
+    {
+      "epoch": 1.08224,
+      "grad_norm": 0.5524299740791321,
+      "learning_rate": 9.188150520416333e-05,
+      "loss": 0.515,
+      "step": 3382
+    },
+    {
+      "epoch": 1.08256,
+      "grad_norm": 0.5604859590530396,
+      "learning_rate": 9.184947958366694e-05,
+      "loss": 0.5309,
+      "step": 3383
+    },
+    {
+      "epoch": 1.08288,
+      "grad_norm": 0.533072292804718,
+      "learning_rate": 9.181745396317054e-05,
+      "loss": 0.4988,
+      "step": 3384
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 0.3736531734466553,
+      "learning_rate": 9.178542834267415e-05,
+      "loss": 0.3902,
+      "step": 3385
+    },
+    {
+      "epoch": 1.08352,
+      "grad_norm": 0.5460339188575745,
+      "learning_rate": 9.175340272217775e-05,
+      "loss": 0.4105,
+      "step": 3386
+    },
+    {
+      "epoch": 1.08384,
+      "grad_norm": 0.49991393089294434,
+      "learning_rate": 9.172137710168135e-05,
+      "loss": 0.5078,
+      "step": 3387
+    },
+    {
+      "epoch": 1.08416,
+      "grad_norm": 0.5023396611213684,
+      "learning_rate": 9.168935148118495e-05,
+      "loss": 0.4799,
+      "step": 3388
+    },
+    {
+      "epoch": 1.08448,
+      "grad_norm": 0.5367220640182495,
+      "learning_rate": 9.165732586068856e-05,
+      "loss": 0.6612,
+      "step": 3389
+    },
+    {
+      "epoch": 1.0848,
+      "grad_norm": 0.5035234093666077,
+      "learning_rate": 9.162530024019216e-05,
+      "loss": 0.5357,
+      "step": 3390
+    },
+    {
+      "epoch": 1.08512,
+      "grad_norm": 0.5462536811828613,
+      "learning_rate": 9.159327461969576e-05,
+      "loss": 0.4461,
+      "step": 3391
+    },
+    {
+      "epoch": 1.08544,
+      "grad_norm": 0.46880975365638733,
+      "learning_rate": 9.156124899919936e-05,
+      "loss": 0.5017,
+      "step": 3392
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 0.6162593960762024,
+      "learning_rate": 9.152922337870297e-05,
+      "loss": 0.6016,
+      "step": 3393
+    },
+    {
+      "epoch": 1.08608,
+      "grad_norm": 0.48707520961761475,
+      "learning_rate": 9.149719775820656e-05,
+      "loss": 0.5085,
+      "step": 3394
+    },
+    {
+      "epoch": 1.0864,
+      "grad_norm": 0.4632261395454407,
+      "learning_rate": 9.146517213771018e-05,
+      "loss": 0.5433,
+      "step": 3395
+    },
+    {
+      "epoch": 1.08672,
+      "grad_norm": 0.5744171142578125,
+      "learning_rate": 9.143314651721377e-05,
+      "loss": 0.5134,
+      "step": 3396
+    },
+    {
+      "epoch": 1.08704,
+      "grad_norm": 0.540407121181488,
+      "learning_rate": 9.140112089671738e-05,
+      "loss": 0.5671,
+      "step": 3397
+    },
+    {
+      "epoch": 1.0873599999999999,
+      "grad_norm": 0.5009400248527527,
+      "learning_rate": 9.136909527622098e-05,
+      "loss": 0.5062,
+      "step": 3398
+    },
+    {
+      "epoch": 1.08768,
+      "grad_norm": 0.6993561387062073,
+      "learning_rate": 9.133706965572458e-05,
+      "loss": 0.5381,
+      "step": 3399
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.4620285928249359,
+      "learning_rate": 9.130504403522818e-05,
+      "loss": 0.3826,
+      "step": 3400
+    },
+    {
+      "epoch": 1.08832,
+      "grad_norm": 0.5245941281318665,
+      "learning_rate": 9.12730184147318e-05,
+      "loss": 0.3882,
+      "step": 3401
+    },
+    {
+      "epoch": 1.08864,
+      "grad_norm": 0.6180137395858765,
+      "learning_rate": 9.124099279423539e-05,
+      "loss": 0.3938,
+      "step": 3402
+    },
+    {
+      "epoch": 1.08896,
+      "grad_norm": 0.44115009903907776,
+      "learning_rate": 9.120896717373899e-05,
+      "loss": 0.4642,
+      "step": 3403
+    },
+    {
+      "epoch": 1.08928,
+      "grad_norm": 0.5987918972969055,
+      "learning_rate": 9.11769415532426e-05,
+      "loss": 0.6869,
+      "step": 3404
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.517765462398529,
+      "learning_rate": 9.11449159327462e-05,
+      "loss": 0.5463,
+      "step": 3405
+    },
+    {
+      "epoch": 1.08992,
+      "grad_norm": 0.5471801161766052,
+      "learning_rate": 9.11128903122498e-05,
+      "loss": 0.4761,
+      "step": 3406
+    },
+    {
+      "epoch": 1.09024,
+      "grad_norm": 0.5406278371810913,
+      "learning_rate": 9.108086469175341e-05,
+      "loss": 0.4113,
+      "step": 3407
+    },
+    {
+      "epoch": 1.09056,
+      "grad_norm": 0.7121791839599609,
+      "learning_rate": 9.104883907125701e-05,
+      "loss": 0.4238,
+      "step": 3408
+    },
+    {
+      "epoch": 1.09088,
+      "grad_norm": 0.566218376159668,
+      "learning_rate": 9.10168134507606e-05,
+      "loss": 0.4666,
+      "step": 3409
+    },
+    {
+      "epoch": 1.0912,
+      "grad_norm": 0.6419295072555542,
+      "learning_rate": 9.098478783026422e-05,
+      "loss": 0.4984,
+      "step": 3410
+    },
+    {
+      "epoch": 1.09152,
+      "grad_norm": 0.5499956011772156,
+      "learning_rate": 9.095276220976781e-05,
+      "loss": 0.4398,
+      "step": 3411
+    },
+    {
+      "epoch": 1.09184,
+      "grad_norm": 0.5424249172210693,
+      "learning_rate": 9.092073658927143e-05,
+      "loss": 0.5492,
+      "step": 3412
+    },
+    {
+      "epoch": 1.09216,
+      "grad_norm": 0.43988147377967834,
+      "learning_rate": 9.088871096877502e-05,
+      "loss": 0.4623,
+      "step": 3413
+    },
+    {
+      "epoch": 1.0924800000000001,
+      "grad_norm": 0.5267035365104675,
+      "learning_rate": 9.085668534827863e-05,
+      "loss": 0.4423,
+      "step": 3414
+    },
+    {
+      "epoch": 1.0928,
+      "grad_norm": 0.5450072288513184,
+      "learning_rate": 9.082465972778223e-05,
+      "loss": 0.5256,
+      "step": 3415
+    },
+    {
+      "epoch": 1.09312,
+      "grad_norm": 0.4738229215145111,
+      "learning_rate": 9.079263410728584e-05,
+      "loss": 0.5189,
+      "step": 3416
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 0.5504553318023682,
+      "learning_rate": 9.076060848678943e-05,
+      "loss": 0.4979,
+      "step": 3417
+    },
+    {
+      "epoch": 1.09376,
+      "grad_norm": 0.5103331804275513,
+      "learning_rate": 9.072858286629305e-05,
+      "loss": 0.5271,
+      "step": 3418
+    },
+    {
+      "epoch": 1.09408,
+      "grad_norm": 0.5095517635345459,
+      "learning_rate": 9.069655724579664e-05,
+      "loss": 0.5462,
+      "step": 3419
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 0.5111420750617981,
+      "learning_rate": 9.066453162530024e-05,
+      "loss": 0.401,
+      "step": 3420
+    },
+    {
+      "epoch": 1.09472,
+      "grad_norm": 0.5317994952201843,
+      "learning_rate": 9.063250600480385e-05,
+      "loss": 0.5116,
+      "step": 3421
+    },
+    {
+      "epoch": 1.09504,
+      "grad_norm": 0.5112237334251404,
+      "learning_rate": 9.060048038430745e-05,
+      "loss": 0.5661,
+      "step": 3422
+    },
+    {
+      "epoch": 1.09536,
+      "grad_norm": 0.7028668522834778,
+      "learning_rate": 9.056845476381104e-05,
+      "loss": 0.6129,
+      "step": 3423
+    },
+    {
+      "epoch": 1.09568,
+      "grad_norm": 0.5981730818748474,
+      "learning_rate": 9.053642914331466e-05,
+      "loss": 0.5052,
+      "step": 3424
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 0.7259790301322937,
+      "learning_rate": 9.050440352281825e-05,
+      "loss": 0.4824,
+      "step": 3425
+    },
+    {
+      "epoch": 1.09632,
+      "grad_norm": 0.6909366846084595,
+      "learning_rate": 9.047237790232186e-05,
+      "loss": 0.6628,
+      "step": 3426
+    },
+    {
+      "epoch": 1.09664,
+      "grad_norm": 0.4826480448246002,
+      "learning_rate": 9.044035228182546e-05,
+      "loss": 0.5375,
+      "step": 3427
+    },
+    {
+      "epoch": 1.09696,
+      "grad_norm": 0.5283201336860657,
+      "learning_rate": 9.040832666132907e-05,
+      "loss": 0.4438,
+      "step": 3428
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 0.5730308890342712,
+      "learning_rate": 9.037630104083267e-05,
+      "loss": 0.4185,
+      "step": 3429
+    },
+    {
+      "epoch": 1.0976,
+      "grad_norm": 0.5374807119369507,
+      "learning_rate": 9.034427542033628e-05,
+      "loss": 0.5203,
+      "step": 3430
+    },
+    {
+      "epoch": 1.09792,
+      "grad_norm": 0.6194236874580383,
+      "learning_rate": 9.031224979983988e-05,
+      "loss": 0.5052,
+      "step": 3431
+    },
+    {
+      "epoch": 1.09824,
+      "grad_norm": 0.6470579504966736,
+      "learning_rate": 9.028022417934347e-05,
+      "loss": 0.5304,
+      "step": 3432
+    },
+    {
+      "epoch": 1.09856,
+      "grad_norm": 0.5055282711982727,
+      "learning_rate": 9.024819855884709e-05,
+      "loss": 0.4049,
+      "step": 3433
+    },
+    {
+      "epoch": 1.09888,
+      "grad_norm": 0.49869534373283386,
+      "learning_rate": 9.021617293835068e-05,
+      "loss": 0.4538,
+      "step": 3434
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 0.6213011145591736,
+      "learning_rate": 9.018414731785429e-05,
+      "loss": 0.4522,
+      "step": 3435
+    },
+    {
+      "epoch": 1.09952,
+      "grad_norm": 0.5340168476104736,
+      "learning_rate": 9.015212169735789e-05,
+      "loss": 0.365,
+      "step": 3436
+    },
+    {
+      "epoch": 1.09984,
+      "grad_norm": 0.6297221779823303,
+      "learning_rate": 9.01200960768615e-05,
+      "loss": 0.5252,
+      "step": 3437
+    },
+    {
+      "epoch": 1.10016,
+      "grad_norm": 0.613269567489624,
+      "learning_rate": 9.008807045636509e-05,
+      "loss": 0.4835,
+      "step": 3438
+    },
+    {
+      "epoch": 1.10048,
+      "grad_norm": 0.4739953279495239,
+      "learning_rate": 9.005604483586871e-05,
+      "loss": 0.4333,
+      "step": 3439
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.5764981508255005,
+      "learning_rate": 9.00240192153723e-05,
+      "loss": 0.4612,
+      "step": 3440
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 0.5053103566169739,
+      "learning_rate": 8.99919935948759e-05,
+      "loss": 0.5342,
+      "step": 3441
+    },
+    {
+      "epoch": 1.10144,
+      "grad_norm": 0.4741344749927521,
+      "learning_rate": 8.995996797437951e-05,
+      "loss": 0.4161,
+      "step": 3442
+    },
+    {
+      "epoch": 1.10176,
+      "grad_norm": 0.6536281108856201,
+      "learning_rate": 8.992794235388311e-05,
+      "loss": 0.552,
+      "step": 3443
+    },
+    {
+      "epoch": 1.10208,
+      "grad_norm": 0.6108992695808411,
+      "learning_rate": 8.98959167333867e-05,
+      "loss": 0.4409,
+      "step": 3444
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 0.5455003976821899,
+      "learning_rate": 8.986389111289032e-05,
+      "loss": 0.4675,
+      "step": 3445
+    },
+    {
+      "epoch": 1.10272,
+      "grad_norm": 0.5502614378929138,
+      "learning_rate": 8.983186549239391e-05,
+      "loss": 0.4584,
+      "step": 3446
+    },
+    {
+      "epoch": 1.10304,
+      "grad_norm": 0.4628472328186035,
+      "learning_rate": 8.979983987189752e-05,
+      "loss": 0.4375,
+      "step": 3447
+    },
+    {
+      "epoch": 1.10336,
+      "grad_norm": 0.6305974125862122,
+      "learning_rate": 8.976781425140112e-05,
+      "loss": 0.5091,
+      "step": 3448
+    },
+    {
+      "epoch": 1.10368,
+      "grad_norm": 0.6597842574119568,
+      "learning_rate": 8.973578863090473e-05,
+      "loss": 0.4955,
+      "step": 3449
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 0.5599210858345032,
+      "learning_rate": 8.970376301040832e-05,
+      "loss": 0.4546,
+      "step": 3450
+    },
+    {
+      "epoch": 1.10432,
+      "grad_norm": 0.5577013492584229,
+      "learning_rate": 8.967173738991194e-05,
+      "loss": 0.5663,
+      "step": 3451
+    },
+    {
+      "epoch": 1.10464,
+      "grad_norm": 0.44588881731033325,
+      "learning_rate": 8.963971176941553e-05,
+      "loss": 0.4544,
+      "step": 3452
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 0.6557613015174866,
+      "learning_rate": 8.960768614891913e-05,
+      "loss": 0.5923,
+      "step": 3453
+    },
+    {
+      "epoch": 1.10528,
+      "grad_norm": 0.5165500640869141,
+      "learning_rate": 8.957566052842274e-05,
+      "loss": 0.5103,
+      "step": 3454
+    },
+    {
+      "epoch": 1.1056,
+      "grad_norm": 0.370685875415802,
+      "learning_rate": 8.954363490792634e-05,
+      "loss": 0.3053,
+      "step": 3455
+    },
+    {
+      "epoch": 1.10592,
+      "grad_norm": 0.6548709273338318,
+      "learning_rate": 8.951160928742995e-05,
+      "loss": 0.456,
+      "step": 3456
+    },
+    {
+      "epoch": 1.1062400000000001,
+      "grad_norm": 0.42675843834877014,
+      "learning_rate": 8.947958366693355e-05,
+      "loss": 0.3734,
+      "step": 3457
+    },
+    {
+      "epoch": 1.10656,
+      "grad_norm": 0.5843567252159119,
+      "learning_rate": 8.944755804643716e-05,
+      "loss": 0.4987,
+      "step": 3458
+    },
+    {
+      "epoch": 1.10688,
+      "grad_norm": 0.5447933077812195,
+      "learning_rate": 8.941553242594076e-05,
+      "loss": 0.473,
+      "step": 3459
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 0.615562379360199,
+      "learning_rate": 8.938350680544437e-05,
+      "loss": 0.5464,
+      "step": 3460
+    },
+    {
+      "epoch": 1.10752,
+      "grad_norm": 0.6441030502319336,
+      "learning_rate": 8.935148118494796e-05,
+      "loss": 0.5492,
+      "step": 3461
+    },
+    {
+      "epoch": 1.10784,
+      "grad_norm": 0.572714626789093,
+      "learning_rate": 8.931945556445158e-05,
+      "loss": 0.559,
+      "step": 3462
+    },
+    {
+      "epoch": 1.10816,
+      "grad_norm": 0.51131272315979,
+      "learning_rate": 8.928742994395517e-05,
+      "loss": 0.4819,
+      "step": 3463
+    },
+    {
+      "epoch": 1.10848,
+      "grad_norm": 0.5505568385124207,
+      "learning_rate": 8.925540432345877e-05,
+      "loss": 0.5169,
+      "step": 3464
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.510718047618866,
+      "learning_rate": 8.922337870296238e-05,
+      "loss": 0.3499,
+      "step": 3465
+    },
+    {
+      "epoch": 1.1091199999999999,
+      "grad_norm": 0.6219130158424377,
+      "learning_rate": 8.919135308246598e-05,
+      "loss": 0.6863,
+      "step": 3466
+    },
+    {
+      "epoch": 1.10944,
+      "grad_norm": 0.5729855895042419,
+      "learning_rate": 8.915932746196957e-05,
+      "loss": 0.4294,
+      "step": 3467
+    },
+    {
+      "epoch": 1.10976,
+      "grad_norm": 0.5130463242530823,
+      "learning_rate": 8.912730184147319e-05,
+      "loss": 0.4387,
+      "step": 3468
+    },
+    {
+      "epoch": 1.11008,
+      "grad_norm": 0.6604203581809998,
+      "learning_rate": 8.909527622097678e-05,
+      "loss": 0.5253,
+      "step": 3469
+    },
+    {
+      "epoch": 1.1104,
+      "grad_norm": 0.4954116642475128,
+      "learning_rate": 8.906325060048039e-05,
+      "loss": 0.5093,
+      "step": 3470
+    },
+    {
+      "epoch": 1.11072,
+      "grad_norm": 0.5187800526618958,
+      "learning_rate": 8.903122497998399e-05,
+      "loss": 0.4866,
+      "step": 3471
+    },
+    {
+      "epoch": 1.11104,
+      "grad_norm": 0.566142737865448,
+      "learning_rate": 8.89991993594876e-05,
+      "loss": 0.5513,
+      "step": 3472
+    },
+    {
+      "epoch": 1.11136,
+      "grad_norm": 0.4569252133369446,
+      "learning_rate": 8.896717373899119e-05,
+      "loss": 0.5131,
+      "step": 3473
+    },
+    {
+      "epoch": 1.11168,
+      "grad_norm": 0.5608154535293579,
+      "learning_rate": 8.89351481184948e-05,
+      "loss": 0.4898,
+      "step": 3474
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 0.4263284206390381,
+      "learning_rate": 8.89031224979984e-05,
+      "loss": 0.3635,
+      "step": 3475
+    },
+    {
+      "epoch": 1.11232,
+      "grad_norm": 0.5647116303443909,
+      "learning_rate": 8.8871096877502e-05,
+      "loss": 0.498,
+      "step": 3476
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 0.5429050326347351,
+      "learning_rate": 8.883907125700561e-05,
+      "loss": 0.5015,
+      "step": 3477
+    },
+    {
+      "epoch": 1.11296,
+      "grad_norm": 0.5069053769111633,
+      "learning_rate": 8.880704563650921e-05,
+      "loss": 0.5439,
+      "step": 3478
+    },
+    {
+      "epoch": 1.11328,
+      "grad_norm": 0.45893579721450806,
+      "learning_rate": 8.877502001601282e-05,
+      "loss": 0.3988,
+      "step": 3479
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.4026511311531067,
+      "learning_rate": 8.874299439551642e-05,
+      "loss": 0.4313,
+      "step": 3480
+    },
+    {
+      "epoch": 1.11392,
+      "grad_norm": 0.6719968914985657,
+      "learning_rate": 8.871096877502003e-05,
+      "loss": 0.5147,
+      "step": 3481
+    },
+    {
+      "epoch": 1.11424,
+      "grad_norm": 0.5521899461746216,
+      "learning_rate": 8.867894315452362e-05,
+      "loss": 0.4858,
+      "step": 3482
+    },
+    {
+      "epoch": 1.11456,
+      "grad_norm": 0.6543344855308533,
+      "learning_rate": 8.864691753402724e-05,
+      "loss": 0.6673,
+      "step": 3483
+    },
+    {
+      "epoch": 1.11488,
+      "grad_norm": 0.6155899167060852,
+      "learning_rate": 8.861489191353083e-05,
+      "loss": 0.5164,
+      "step": 3484
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 0.5979326367378235,
+      "learning_rate": 8.858286629303443e-05,
+      "loss": 0.5653,
+      "step": 3485
+    },
+    {
+      "epoch": 1.11552,
+      "grad_norm": 0.6349969506263733,
+      "learning_rate": 8.855084067253804e-05,
+      "loss": 0.5248,
+      "step": 3486
+    },
+    {
+      "epoch": 1.11584,
+      "grad_norm": 0.5014321208000183,
+      "learning_rate": 8.851881505204164e-05,
+      "loss": 0.4733,
+      "step": 3487
+    },
+    {
+      "epoch": 1.11616,
+      "grad_norm": 0.42011910676956177,
+      "learning_rate": 8.848678943154523e-05,
+      "loss": 0.4594,
+      "step": 3488
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 0.45381760597229004,
+      "learning_rate": 8.845476381104885e-05,
+      "loss": 0.4025,
+      "step": 3489
+    },
+    {
+      "epoch": 1.1168,
+      "grad_norm": 0.4848672151565552,
+      "learning_rate": 8.842273819055244e-05,
+      "loss": 0.4518,
+      "step": 3490
+    },
+    {
+      "epoch": 1.11712,
+      "grad_norm": 0.5914720296859741,
+      "learning_rate": 8.839071257005605e-05,
+      "loss": 0.5131,
+      "step": 3491
+    },
+    {
+      "epoch": 1.11744,
+      "grad_norm": 0.5666292905807495,
+      "learning_rate": 8.835868694955965e-05,
+      "loss": 0.528,
+      "step": 3492
+    },
+    {
+      "epoch": 1.11776,
+      "grad_norm": 0.5597184896469116,
+      "learning_rate": 8.832666132906326e-05,
+      "loss": 0.497,
+      "step": 3493
+    },
+    {
+      "epoch": 1.11808,
+      "grad_norm": 0.5898178219795227,
+      "learning_rate": 8.829463570856685e-05,
+      "loss": 0.5101,
+      "step": 3494
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 0.5234562754631042,
+      "learning_rate": 8.826261008807047e-05,
+      "loss": 0.4112,
+      "step": 3495
+    },
+    {
+      "epoch": 1.11872,
+      "grad_norm": 0.5671234130859375,
+      "learning_rate": 8.823058446757406e-05,
+      "loss": 0.5933,
+      "step": 3496
+    },
+    {
+      "epoch": 1.11904,
+      "grad_norm": 0.5762549042701721,
+      "learning_rate": 8.819855884707766e-05,
+      "loss": 0.4901,
+      "step": 3497
+    },
+    {
+      "epoch": 1.11936,
+      "grad_norm": 0.5005612969398499,
+      "learning_rate": 8.816653322658127e-05,
+      "loss": 0.436,
+      "step": 3498
+    },
+    {
+      "epoch": 1.11968,
+      "grad_norm": 0.42232629656791687,
+      "learning_rate": 8.813450760608487e-05,
+      "loss": 0.3756,
+      "step": 3499
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5372239351272583,
+      "learning_rate": 8.810248198558848e-05,
+      "loss": 0.4584,
+      "step": 3500
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 0.6059823036193848,
+      "learning_rate": 8.807045636509208e-05,
+      "loss": 0.5272,
+      "step": 3501
+    },
+    {
+      "epoch": 1.12064,
+      "grad_norm": 0.44775041937828064,
+      "learning_rate": 8.803843074459569e-05,
+      "loss": 0.4102,
+      "step": 3502
+    },
+    {
+      "epoch": 1.12096,
+      "grad_norm": 0.6138641834259033,
+      "learning_rate": 8.800640512409928e-05,
+      "loss": 0.5952,
+      "step": 3503
+    },
+    {
+      "epoch": 1.12128,
+      "grad_norm": 0.6402849555015564,
+      "learning_rate": 8.79743795036029e-05,
+      "loss": 0.6237,
+      "step": 3504
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 0.7078378200531006,
+      "learning_rate": 8.794235388310649e-05,
+      "loss": 0.5868,
+      "step": 3505
+    },
+    {
+      "epoch": 1.12192,
+      "grad_norm": 0.5268515348434448,
+      "learning_rate": 8.791032826261009e-05,
+      "loss": 0.4108,
+      "step": 3506
+    },
+    {
+      "epoch": 1.12224,
+      "grad_norm": 0.7270392775535583,
+      "learning_rate": 8.78783026421137e-05,
+      "loss": 0.5405,
+      "step": 3507
+    },
+    {
+      "epoch": 1.12256,
+      "grad_norm": 0.5641281604766846,
+      "learning_rate": 8.78462770216173e-05,
+      "loss": 0.4649,
+      "step": 3508
+    },
+    {
+      "epoch": 1.12288,
+      "grad_norm": 0.45243874192237854,
+      "learning_rate": 8.78142514011209e-05,
+      "loss": 0.3671,
+      "step": 3509
+    },
+    {
+      "epoch": 1.1232,
+      "grad_norm": 0.5990859866142273,
+      "learning_rate": 8.778222578062451e-05,
+      "loss": 0.4628,
+      "step": 3510
+    },
+    {
+      "epoch": 1.12352,
+      "grad_norm": 0.4442991018295288,
+      "learning_rate": 8.77502001601281e-05,
+      "loss": 0.3214,
+      "step": 3511
+    },
+    {
+      "epoch": 1.12384,
+      "grad_norm": 0.5706406235694885,
+      "learning_rate": 8.771817453963172e-05,
+      "loss": 0.5011,
+      "step": 3512
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 0.5856726765632629,
+      "learning_rate": 8.768614891913531e-05,
+      "loss": 0.5266,
+      "step": 3513
+    },
+    {
+      "epoch": 1.12448,
+      "grad_norm": 0.49093014001846313,
+      "learning_rate": 8.765412329863892e-05,
+      "loss": 0.3929,
+      "step": 3514
+    },
+    {
+      "epoch": 1.1248,
+      "grad_norm": 0.5803288817405701,
+      "learning_rate": 8.762209767814252e-05,
+      "loss": 0.5471,
+      "step": 3515
+    },
+    {
+      "epoch": 1.12512,
+      "grad_norm": 0.6210606694221497,
+      "learning_rate": 8.759007205764613e-05,
+      "loss": 0.4583,
+      "step": 3516
+    },
+    {
+      "epoch": 1.12544,
+      "grad_norm": 0.49231746792793274,
+      "learning_rate": 8.755804643714972e-05,
+      "loss": 0.4269,
+      "step": 3517
+    },
+    {
+      "epoch": 1.12576,
+      "grad_norm": 0.5454963445663452,
+      "learning_rate": 8.752602081665333e-05,
+      "loss": 0.5189,
+      "step": 3518
+    },
+    {
+      "epoch": 1.12608,
+      "grad_norm": 0.545050323009491,
+      "learning_rate": 8.749399519615693e-05,
+      "loss": 0.449,
+      "step": 3519
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.5519834756851196,
+      "learning_rate": 8.746196957566053e-05,
+      "loss": 0.4941,
+      "step": 3520
+    },
+    {
+      "epoch": 1.12672,
+      "grad_norm": 0.49287375807762146,
+      "learning_rate": 8.742994395516414e-05,
+      "loss": 0.3961,
+      "step": 3521
+    },
+    {
+      "epoch": 1.12704,
+      "grad_norm": 0.569459855556488,
+      "learning_rate": 8.739791833466774e-05,
+      "loss": 0.6067,
+      "step": 3522
+    },
+    {
+      "epoch": 1.12736,
+      "grad_norm": 0.574547529220581,
+      "learning_rate": 8.736589271417133e-05,
+      "loss": 0.531,
+      "step": 3523
+    },
+    {
+      "epoch": 1.12768,
+      "grad_norm": 0.4699058532714844,
+      "learning_rate": 8.733386709367495e-05,
+      "loss": 0.3972,
+      "step": 3524
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.6163190007209778,
+      "learning_rate": 8.730184147317854e-05,
+      "loss": 0.5566,
+      "step": 3525
+    },
+    {
+      "epoch": 1.12832,
+      "grad_norm": 0.5265937447547913,
+      "learning_rate": 8.726981585268215e-05,
+      "loss": 0.4912,
+      "step": 3526
+    },
+    {
+      "epoch": 1.12864,
+      "grad_norm": 0.7494415640830994,
+      "learning_rate": 8.723779023218575e-05,
+      "loss": 0.5585,
+      "step": 3527
+    },
+    {
+      "epoch": 1.12896,
+      "grad_norm": 0.5270549058914185,
+      "learning_rate": 8.720576461168936e-05,
+      "loss": 0.5456,
+      "step": 3528
+    },
+    {
+      "epoch": 1.12928,
+      "grad_norm": 0.5301781296730042,
+      "learning_rate": 8.717373899119296e-05,
+      "loss": 0.4705,
+      "step": 3529
+    },
+    {
+      "epoch": 1.1296,
+      "grad_norm": 0.5162952542304993,
+      "learning_rate": 8.714171337069656e-05,
+      "loss": 0.4562,
+      "step": 3530
+    },
+    {
+      "epoch": 1.12992,
+      "grad_norm": 0.43721723556518555,
+      "learning_rate": 8.710968775020017e-05,
+      "loss": 0.4796,
+      "step": 3531
+    },
+    {
+      "epoch": 1.13024,
+      "grad_norm": 0.5706759691238403,
+      "learning_rate": 8.707766212970376e-05,
+      "loss": 0.5254,
+      "step": 3532
+    },
+    {
+      "epoch": 1.13056,
+      "grad_norm": 0.6177812218666077,
+      "learning_rate": 8.704563650920738e-05,
+      "loss": 0.5006,
+      "step": 3533
+    },
+    {
+      "epoch": 1.1308799999999999,
+      "grad_norm": 0.6244451999664307,
+      "learning_rate": 8.701361088871097e-05,
+      "loss": 0.4772,
+      "step": 3534
+    },
+    {
+      "epoch": 1.1312,
+      "grad_norm": 0.5011513233184814,
+      "learning_rate": 8.698158526821457e-05,
+      "loss": 0.5125,
+      "step": 3535
+    },
+    {
+      "epoch": 1.13152,
+      "grad_norm": 0.5688576102256775,
+      "learning_rate": 8.694955964771818e-05,
+      "loss": 0.497,
+      "step": 3536
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 0.66170734167099,
+      "learning_rate": 8.691753402722178e-05,
+      "loss": 0.506,
+      "step": 3537
+    },
+    {
+      "epoch": 1.13216,
+      "grad_norm": 0.5211583971977234,
+      "learning_rate": 8.688550840672538e-05,
+      "loss": 0.4267,
+      "step": 3538
+    },
+    {
+      "epoch": 1.13248,
+      "grad_norm": 0.5361404418945312,
+      "learning_rate": 8.6853482786229e-05,
+      "loss": 0.4896,
+      "step": 3539
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 0.6319483518600464,
+      "learning_rate": 8.682145716573259e-05,
+      "loss": 0.4838,
+      "step": 3540
+    },
+    {
+      "epoch": 1.13312,
+      "grad_norm": 0.5309115648269653,
+      "learning_rate": 8.678943154523619e-05,
+      "loss": 0.4615,
+      "step": 3541
+    },
+    {
+      "epoch": 1.13344,
+      "grad_norm": 0.6791277527809143,
+      "learning_rate": 8.67574059247398e-05,
+      "loss": 0.6137,
+      "step": 3542
+    },
+    {
+      "epoch": 1.13376,
+      "grad_norm": 0.5034888386726379,
+      "learning_rate": 8.67253803042434e-05,
+      "loss": 0.5693,
+      "step": 3543
+    },
+    {
+      "epoch": 1.13408,
+      "grad_norm": 0.6300373077392578,
+      "learning_rate": 8.669335468374699e-05,
+      "loss": 0.5183,
+      "step": 3544
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 0.4940887987613678,
+      "learning_rate": 8.666132906325061e-05,
+      "loss": 0.4638,
+      "step": 3545
+    },
+    {
+      "epoch": 1.13472,
+      "grad_norm": 0.6096082925796509,
+      "learning_rate": 8.66293034427542e-05,
+      "loss": 0.4178,
+      "step": 3546
+    },
+    {
+      "epoch": 1.13504,
+      "grad_norm": 0.6011221408843994,
+      "learning_rate": 8.65972778222578e-05,
+      "loss": 0.557,
+      "step": 3547
+    },
+    {
+      "epoch": 1.13536,
+      "grad_norm": 0.5549004077911377,
+      "learning_rate": 8.656525220176141e-05,
+      "loss": 0.4972,
+      "step": 3548
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 0.66285640001297,
+      "learning_rate": 8.653322658126501e-05,
+      "loss": 0.4511,
+      "step": 3549
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 0.5397788882255554,
+      "learning_rate": 8.650120096076862e-05,
+      "loss": 0.4696,
+      "step": 3550
+    },
+    {
+      "epoch": 1.13632,
+      "grad_norm": 0.5463436245918274,
+      "learning_rate": 8.646917534027222e-05,
+      "loss": 0.4643,
+      "step": 3551
+    },
+    {
+      "epoch": 1.13664,
+      "grad_norm": 0.46885764598846436,
+      "learning_rate": 8.643714971977583e-05,
+      "loss": 0.3208,
+      "step": 3552
+    },
+    {
+      "epoch": 1.13696,
+      "grad_norm": 0.4589516222476959,
+      "learning_rate": 8.640512409927943e-05,
+      "loss": 0.4477,
+      "step": 3553
+    },
+    {
+      "epoch": 1.13728,
+      "grad_norm": 0.5470423698425293,
+      "learning_rate": 8.637309847878304e-05,
+      "loss": 0.4951,
+      "step": 3554
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 0.5328925848007202,
+      "learning_rate": 8.634107285828663e-05,
+      "loss": 0.3955,
+      "step": 3555
+    },
+    {
+      "epoch": 1.13792,
+      "grad_norm": 0.5843022465705872,
+      "learning_rate": 8.630904723779025e-05,
+      "loss": 0.557,
+      "step": 3556
+    },
+    {
+      "epoch": 1.13824,
+      "grad_norm": 0.7209236025810242,
+      "learning_rate": 8.627702161729384e-05,
+      "loss": 0.5552,
+      "step": 3557
+    },
+    {
+      "epoch": 1.13856,
+      "grad_norm": 0.3960036337375641,
+      "learning_rate": 8.624499599679744e-05,
+      "loss": 0.398,
+      "step": 3558
+    },
+    {
+      "epoch": 1.13888,
+      "grad_norm": 0.6268905997276306,
+      "learning_rate": 8.621297037630105e-05,
+      "loss": 0.5751,
+      "step": 3559
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.4649326801300049,
+      "learning_rate": 8.618094475580465e-05,
+      "loss": 0.4651,
+      "step": 3560
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 0.7146782279014587,
+      "learning_rate": 8.614891913530824e-05,
+      "loss": 0.5852,
+      "step": 3561
+    },
+    {
+      "epoch": 1.13984,
+      "grad_norm": 0.5895894765853882,
+      "learning_rate": 8.611689351481186e-05,
+      "loss": 0.5519,
+      "step": 3562
+    },
+    {
+      "epoch": 1.14016,
+      "grad_norm": 0.4463888108730316,
+      "learning_rate": 8.608486789431545e-05,
+      "loss": 0.3507,
+      "step": 3563
+    },
+    {
+      "epoch": 1.14048,
+      "grad_norm": 0.5654453039169312,
+      "learning_rate": 8.605284227381906e-05,
+      "loss": 0.4919,
+      "step": 3564
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 0.421886682510376,
+      "learning_rate": 8.602081665332266e-05,
+      "loss": 0.4226,
+      "step": 3565
+    },
+    {
+      "epoch": 1.14112,
+      "grad_norm": 0.5128637552261353,
+      "learning_rate": 8.598879103282627e-05,
+      "loss": 0.4433,
+      "step": 3566
+    },
+    {
+      "epoch": 1.14144,
+      "grad_norm": 0.5426830053329468,
+      "learning_rate": 8.595676541232986e-05,
+      "loss": 0.4042,
+      "step": 3567
+    },
+    {
+      "epoch": 1.14176,
+      "grad_norm": 0.5870913863182068,
+      "learning_rate": 8.592473979183348e-05,
+      "loss": 0.5562,
+      "step": 3568
+    },
+    {
+      "epoch": 1.14208,
+      "grad_norm": 0.6103060245513916,
+      "learning_rate": 8.589271417133707e-05,
+      "loss": 0.5119,
+      "step": 3569
+    },
+    {
+      "epoch": 1.1424,
+      "grad_norm": 0.5142077803611755,
+      "learning_rate": 8.586068855084067e-05,
+      "loss": 0.4039,
+      "step": 3570
+    },
+    {
+      "epoch": 1.14272,
+      "grad_norm": 0.6240192651748657,
+      "learning_rate": 8.582866293034428e-05,
+      "loss": 0.4809,
+      "step": 3571
+    },
+    {
+      "epoch": 1.14304,
+      "grad_norm": 0.3659708797931671,
+      "learning_rate": 8.579663730984788e-05,
+      "loss": 0.3207,
+      "step": 3572
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 0.5368773341178894,
+      "learning_rate": 8.576461168935149e-05,
+      "loss": 0.511,
+      "step": 3573
+    },
+    {
+      "epoch": 1.14368,
+      "grad_norm": 0.4287433326244354,
+      "learning_rate": 8.573258606885509e-05,
+      "loss": 0.4444,
+      "step": 3574
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 0.6158849000930786,
+      "learning_rate": 8.57005604483587e-05,
+      "loss": 0.5621,
+      "step": 3575
+    },
+    {
+      "epoch": 1.14432,
+      "grad_norm": 0.5045271515846252,
+      "learning_rate": 8.566853482786229e-05,
+      "loss": 0.4532,
+      "step": 3576
+    },
+    {
+      "epoch": 1.1446399999999999,
+      "grad_norm": 0.6697438359260559,
+      "learning_rate": 8.56365092073659e-05,
+      "loss": 0.5207,
+      "step": 3577
+    },
+    {
+      "epoch": 1.14496,
+      "grad_norm": 0.5694824457168579,
+      "learning_rate": 8.56044835868695e-05,
+      "loss": 0.55,
+      "step": 3578
+    },
+    {
+      "epoch": 1.14528,
+      "grad_norm": 0.7119214534759521,
+      "learning_rate": 8.55724579663731e-05,
+      "loss": 0.514,
+      "step": 3579
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 0.6596461534500122,
+      "learning_rate": 8.554043234587671e-05,
+      "loss": 0.5972,
+      "step": 3580
+    },
+    {
+      "epoch": 1.14592,
+      "grad_norm": 0.5102297067642212,
+      "learning_rate": 8.550840672538031e-05,
+      "loss": 0.3898,
+      "step": 3581
+    },
+    {
+      "epoch": 1.14624,
+      "grad_norm": 0.5949417948722839,
+      "learning_rate": 8.54763811048839e-05,
+      "loss": 0.6134,
+      "step": 3582
+    },
+    {
+      "epoch": 1.14656,
+      "grad_norm": 0.5987377166748047,
+      "learning_rate": 8.544435548438752e-05,
+      "loss": 0.4975,
+      "step": 3583
+    },
+    {
+      "epoch": 1.14688,
+      "grad_norm": 0.5608789324760437,
+      "learning_rate": 8.541232986389111e-05,
+      "loss": 0.4939,
+      "step": 3584
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.6142665147781372,
+      "learning_rate": 8.538030424339472e-05,
+      "loss": 0.4541,
+      "step": 3585
+    },
+    {
+      "epoch": 1.14752,
+      "grad_norm": 0.5422946810722351,
+      "learning_rate": 8.534827862289832e-05,
+      "loss": 0.4911,
+      "step": 3586
+    },
+    {
+      "epoch": 1.14784,
+      "grad_norm": 0.46077558398246765,
+      "learning_rate": 8.531625300240193e-05,
+      "loss": 0.3951,
+      "step": 3587
+    },
+    {
+      "epoch": 1.14816,
+      "grad_norm": 0.5165653824806213,
+      "learning_rate": 8.528422738190552e-05,
+      "loss": 0.4735,
+      "step": 3588
+    },
+    {
+      "epoch": 1.14848,
+      "grad_norm": 0.5479727983474731,
+      "learning_rate": 8.525220176140914e-05,
+      "loss": 0.4864,
+      "step": 3589
+    },
+    {
+      "epoch": 1.1488,
+      "grad_norm": 0.4982510209083557,
+      "learning_rate": 8.522017614091273e-05,
+      "loss": 0.5086,
+      "step": 3590
+    },
+    {
+      "epoch": 1.14912,
+      "grad_norm": 0.55315762758255,
+      "learning_rate": 8.518815052041633e-05,
+      "loss": 0.5428,
+      "step": 3591
+    },
+    {
+      "epoch": 1.14944,
+      "grad_norm": 0.5409789085388184,
+      "learning_rate": 8.515612489991994e-05,
+      "loss": 0.4639,
+      "step": 3592
+    },
+    {
+      "epoch": 1.1497600000000001,
+      "grad_norm": 0.61668860912323,
+      "learning_rate": 8.512409927942354e-05,
+      "loss": 0.6045,
+      "step": 3593
+    },
+    {
+      "epoch": 1.15008,
+      "grad_norm": 0.553780198097229,
+      "learning_rate": 8.509207365892713e-05,
+      "loss": 0.4673,
+      "step": 3594
+    },
+    {
+      "epoch": 1.1504,
+      "grad_norm": 0.47794896364212036,
+      "learning_rate": 8.506004803843075e-05,
+      "loss": 0.4769,
+      "step": 3595
+    },
+    {
+      "epoch": 1.15072,
+      "grad_norm": 0.5873520374298096,
+      "learning_rate": 8.502802241793434e-05,
+      "loss": 0.4874,
+      "step": 3596
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 0.4995674788951874,
+      "learning_rate": 8.499599679743795e-05,
+      "loss": 0.4612,
+      "step": 3597
+    },
+    {
+      "epoch": 1.15136,
+      "grad_norm": 0.46315521001815796,
+      "learning_rate": 8.496397117694155e-05,
+      "loss": 0.4268,
+      "step": 3598
+    },
+    {
+      "epoch": 1.15168,
+      "grad_norm": 0.5224704146385193,
+      "learning_rate": 8.493194555644516e-05,
+      "loss": 0.4076,
+      "step": 3599
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.49204784631729126,
+      "learning_rate": 8.489991993594876e-05,
+      "loss": 0.4385,
+      "step": 3600
+    },
+    {
+      "epoch": 1.15232,
+      "grad_norm": 0.47982457280158997,
+      "learning_rate": 8.486789431545237e-05,
+      "loss": 0.5199,
+      "step": 3601
+    },
+    {
+      "epoch": 1.1526399999999999,
+      "grad_norm": 0.4660177528858185,
+      "learning_rate": 8.483586869495597e-05,
+      "loss": 0.4403,
+      "step": 3602
+    },
+    {
+      "epoch": 1.15296,
+      "grad_norm": 0.3291762173175812,
+      "learning_rate": 8.480384307445958e-05,
+      "loss": 0.3507,
+      "step": 3603
+    },
+    {
+      "epoch": 1.15328,
+      "grad_norm": 0.6174018383026123,
+      "learning_rate": 8.477181745396318e-05,
+      "loss": 0.611,
+      "step": 3604
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 0.5713676810264587,
+      "learning_rate": 8.473979183346677e-05,
+      "loss": 0.482,
+      "step": 3605
+    },
+    {
+      "epoch": 1.15392,
+      "grad_norm": 0.5375822186470032,
+      "learning_rate": 8.470776621297039e-05,
+      "loss": 0.4433,
+      "step": 3606
+    },
+    {
+      "epoch": 1.15424,
+      "grad_norm": 0.48860523104667664,
+      "learning_rate": 8.467574059247398e-05,
+      "loss": 0.3673,
+      "step": 3607
+    },
+    {
+      "epoch": 1.15456,
+      "grad_norm": 0.4985749423503876,
+      "learning_rate": 8.464371497197759e-05,
+      "loss": 0.4544,
+      "step": 3608
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 0.5430129170417786,
+      "learning_rate": 8.461168935148119e-05,
+      "loss": 0.5985,
+      "step": 3609
+    },
+    {
+      "epoch": 1.1552,
+      "grad_norm": 0.5166223645210266,
+      "learning_rate": 8.45796637309848e-05,
+      "loss": 0.4176,
+      "step": 3610
+    },
+    {
+      "epoch": 1.15552,
+      "grad_norm": 0.6497548818588257,
+      "learning_rate": 8.454763811048839e-05,
+      "loss": 0.6383,
+      "step": 3611
+    },
+    {
+      "epoch": 1.15584,
+      "grad_norm": 0.46435093879699707,
+      "learning_rate": 8.4515612489992e-05,
+      "loss": 0.4822,
+      "step": 3612
+    },
+    {
+      "epoch": 1.15616,
+      "grad_norm": 0.5361154079437256,
+      "learning_rate": 8.44835868694956e-05,
+      "loss": 0.4385,
+      "step": 3613
+    },
+    {
+      "epoch": 1.15648,
+      "grad_norm": 0.5268837213516235,
+      "learning_rate": 8.44515612489992e-05,
+      "loss": 0.469,
+      "step": 3614
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 0.6825592517852783,
+      "learning_rate": 8.441953562850281e-05,
+      "loss": 0.4908,
+      "step": 3615
+    },
+    {
+      "epoch": 1.15712,
+      "grad_norm": 0.5720934271812439,
+      "learning_rate": 8.438751000800641e-05,
+      "loss": 0.4331,
+      "step": 3616
+    },
+    {
+      "epoch": 1.15744,
+      "grad_norm": 0.5412835478782654,
+      "learning_rate": 8.435548438751e-05,
+      "loss": 0.537,
+      "step": 3617
+    },
+    {
+      "epoch": 1.1577600000000001,
+      "grad_norm": 0.44690626859664917,
+      "learning_rate": 8.432345876701362e-05,
+      "loss": 0.4357,
+      "step": 3618
+    },
+    {
+      "epoch": 1.15808,
+      "grad_norm": 0.5197708606719971,
+      "learning_rate": 8.429143314651721e-05,
+      "loss": 0.4764,
+      "step": 3619
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 0.6178292036056519,
+      "learning_rate": 8.425940752602082e-05,
+      "loss": 0.4865,
+      "step": 3620
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 0.4821197986602783,
+      "learning_rate": 8.422738190552442e-05,
+      "loss": 0.4482,
+      "step": 3621
+    },
+    {
+      "epoch": 1.15904,
+      "grad_norm": 0.71698397397995,
+      "learning_rate": 8.419535628502803e-05,
+      "loss": 0.4597,
+      "step": 3622
+    },
+    {
+      "epoch": 1.15936,
+      "grad_norm": 0.5499432682991028,
+      "learning_rate": 8.416333066453163e-05,
+      "loss": 0.4731,
+      "step": 3623
+    },
+    {
+      "epoch": 1.15968,
+      "grad_norm": 0.5441680550575256,
+      "learning_rate": 8.413130504403524e-05,
+      "loss": 0.4062,
+      "step": 3624
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.40180450677871704,
+      "learning_rate": 8.409927942353884e-05,
+      "loss": 0.304,
+      "step": 3625
+    },
+    {
+      "epoch": 1.16032,
+      "grad_norm": 0.6135880351066589,
+      "learning_rate": 8.406725380304243e-05,
+      "loss": 0.4304,
+      "step": 3626
+    },
+    {
+      "epoch": 1.16064,
+      "grad_norm": 0.5172175168991089,
+      "learning_rate": 8.403522818254605e-05,
+      "loss": 0.5508,
+      "step": 3627
+    },
+    {
+      "epoch": 1.16096,
+      "grad_norm": 0.6770060658454895,
+      "learning_rate": 8.400320256204964e-05,
+      "loss": 0.4565,
+      "step": 3628
+    },
+    {
+      "epoch": 1.16128,
+      "grad_norm": 0.6119065880775452,
+      "learning_rate": 8.397117694155325e-05,
+      "loss": 0.4725,
+      "step": 3629
+    },
+    {
+      "epoch": 1.1616,
+      "grad_norm": 0.5731270909309387,
+      "learning_rate": 8.393915132105685e-05,
+      "loss": 0.5057,
+      "step": 3630
+    },
+    {
+      "epoch": 1.16192,
+      "grad_norm": 0.5125293731689453,
+      "learning_rate": 8.390712570056046e-05,
+      "loss": 0.4754,
+      "step": 3631
+    },
+    {
+      "epoch": 1.16224,
+      "grad_norm": 0.7232589721679688,
+      "learning_rate": 8.387510008006405e-05,
+      "loss": 0.6658,
+      "step": 3632
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 0.6013690829277039,
+      "learning_rate": 8.384307445956767e-05,
+      "loss": 0.4872,
+      "step": 3633
+    },
+    {
+      "epoch": 1.16288,
+      "grad_norm": 0.4750500023365021,
+      "learning_rate": 8.381104883907126e-05,
+      "loss": 0.3992,
+      "step": 3634
+    },
+    {
+      "epoch": 1.1632,
+      "grad_norm": 0.5696016550064087,
+      "learning_rate": 8.377902321857486e-05,
+      "loss": 0.4499,
+      "step": 3635
+    },
+    {
+      "epoch": 1.16352,
+      "grad_norm": 0.44853588938713074,
+      "learning_rate": 8.374699759807847e-05,
+      "loss": 0.4435,
+      "step": 3636
+    },
+    {
+      "epoch": 1.16384,
+      "grad_norm": 0.6849207878112793,
+      "learning_rate": 8.371497197758207e-05,
+      "loss": 0.528,
+      "step": 3637
+    },
+    {
+      "epoch": 1.16416,
+      "grad_norm": 0.5204032063484192,
+      "learning_rate": 8.368294635708566e-05,
+      "loss": 0.4312,
+      "step": 3638
+    },
+    {
+      "epoch": 1.16448,
+      "grad_norm": 0.46114858984947205,
+      "learning_rate": 8.365092073658928e-05,
+      "loss": 0.4951,
+      "step": 3639
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.5416095852851868,
+      "learning_rate": 8.361889511609287e-05,
+      "loss": 0.4702,
+      "step": 3640
+    },
+    {
+      "epoch": 1.16512,
+      "grad_norm": 0.5522504448890686,
+      "learning_rate": 8.358686949559648e-05,
+      "loss": 0.424,
+      "step": 3641
+    },
+    {
+      "epoch": 1.16544,
+      "grad_norm": 0.49364688992500305,
+      "learning_rate": 8.355484387510008e-05,
+      "loss": 0.4731,
+      "step": 3642
+    },
+    {
+      "epoch": 1.16576,
+      "grad_norm": 0.4775312840938568,
+      "learning_rate": 8.352281825460369e-05,
+      "loss": 0.4914,
+      "step": 3643
+    },
+    {
+      "epoch": 1.16608,
+      "grad_norm": 0.5576226711273193,
+      "learning_rate": 8.349079263410729e-05,
+      "loss": 0.5015,
+      "step": 3644
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.4952666759490967,
+      "learning_rate": 8.34587670136109e-05,
+      "loss": 0.5568,
+      "step": 3645
+    },
+    {
+      "epoch": 1.16672,
+      "grad_norm": 0.42779383063316345,
+      "learning_rate": 8.34267413931145e-05,
+      "loss": 0.3697,
+      "step": 3646
+    },
+    {
+      "epoch": 1.16704,
+      "grad_norm": 0.624330461025238,
+      "learning_rate": 8.33947157726181e-05,
+      "loss": 0.5224,
+      "step": 3647
+    },
+    {
+      "epoch": 1.16736,
+      "grad_norm": 0.497090220451355,
+      "learning_rate": 8.33626901521217e-05,
+      "loss": 0.5087,
+      "step": 3648
+    },
+    {
+      "epoch": 1.16768,
+      "grad_norm": 0.6078720092773438,
+      "learning_rate": 8.33306645316253e-05,
+      "loss": 0.445,
+      "step": 3649
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 0.5238363146781921,
+      "learning_rate": 8.32986389111289e-05,
+      "loss": 0.4277,
+      "step": 3650
+    },
+    {
+      "epoch": 1.16832,
+      "grad_norm": 0.6327863335609436,
+      "learning_rate": 8.326661329063251e-05,
+      "loss": 0.6094,
+      "step": 3651
+    },
+    {
+      "epoch": 1.16864,
+      "grad_norm": 0.6833189725875854,
+      "learning_rate": 8.323458767013612e-05,
+      "loss": 0.5216,
+      "step": 3652
+    },
+    {
+      "epoch": 1.16896,
+      "grad_norm": 0.5724887251853943,
+      "learning_rate": 8.320256204963972e-05,
+      "loss": 0.4734,
+      "step": 3653
+    },
+    {
+      "epoch": 1.16928,
+      "grad_norm": 0.637172281742096,
+      "learning_rate": 8.317053642914332e-05,
+      "loss": 0.547,
+      "step": 3654
+    },
+    {
+      "epoch": 1.1696,
+      "grad_norm": 0.5772240161895752,
+      "learning_rate": 8.313851080864692e-05,
+      "loss": 0.4296,
+      "step": 3655
+    },
+    {
+      "epoch": 1.16992,
+      "grad_norm": 0.6011234521865845,
+      "learning_rate": 8.310648518815053e-05,
+      "loss": 0.4745,
+      "step": 3656
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 0.5446835160255432,
+      "learning_rate": 8.307445956765413e-05,
+      "loss": 0.4609,
+      "step": 3657
+    },
+    {
+      "epoch": 1.17056,
+      "grad_norm": 0.6766799688339233,
+      "learning_rate": 8.304243394715773e-05,
+      "loss": 0.5887,
+      "step": 3658
+    },
+    {
+      "epoch": 1.17088,
+      "grad_norm": 0.59932941198349,
+      "learning_rate": 8.301040832666134e-05,
+      "loss": 0.4435,
+      "step": 3659
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 0.5565205216407776,
+      "learning_rate": 8.297838270616494e-05,
+      "loss": 0.508,
+      "step": 3660
+    },
+    {
+      "epoch": 1.1715200000000001,
+      "grad_norm": 0.5711982846260071,
+      "learning_rate": 8.294635708566853e-05,
+      "loss": 0.4668,
+      "step": 3661
+    },
+    {
+      "epoch": 1.17184,
+      "grad_norm": 0.6003556251525879,
+      "learning_rate": 8.291433146517215e-05,
+      "loss": 0.5224,
+      "step": 3662
+    },
+    {
+      "epoch": 1.17216,
+      "grad_norm": 0.5593168139457703,
+      "learning_rate": 8.288230584467574e-05,
+      "loss": 0.4021,
+      "step": 3663
+    },
+    {
+      "epoch": 1.17248,
+      "grad_norm": 0.6171687245368958,
+      "learning_rate": 8.285028022417935e-05,
+      "loss": 0.4595,
+      "step": 3664
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 0.5374482870101929,
+      "learning_rate": 8.281825460368295e-05,
+      "loss": 0.4263,
+      "step": 3665
+    },
+    {
+      "epoch": 1.17312,
+      "grad_norm": 0.4915101230144501,
+      "learning_rate": 8.278622898318655e-05,
+      "loss": 0.4563,
+      "step": 3666
+    },
+    {
+      "epoch": 1.17344,
+      "grad_norm": 0.5575801730155945,
+      "learning_rate": 8.275420336269015e-05,
+      "loss": 0.5411,
+      "step": 3667
+    },
+    {
+      "epoch": 1.17376,
+      "grad_norm": 0.510799765586853,
+      "learning_rate": 8.272217774219376e-05,
+      "loss": 0.4648,
+      "step": 3668
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 0.4329470098018646,
+      "learning_rate": 8.269015212169736e-05,
+      "loss": 0.4513,
+      "step": 3669
+    },
+    {
+      "epoch": 1.1743999999999999,
+      "grad_norm": 0.5663794875144958,
+      "learning_rate": 8.265812650120096e-05,
+      "loss": 0.5456,
+      "step": 3670
+    },
+    {
+      "epoch": 1.17472,
+      "grad_norm": 0.5060170292854309,
+      "learning_rate": 8.262610088070457e-05,
+      "loss": 0.5509,
+      "step": 3671
+    },
+    {
+      "epoch": 1.17504,
+      "grad_norm": 0.4694247245788574,
+      "learning_rate": 8.259407526020817e-05,
+      "loss": 0.3889,
+      "step": 3672
+    },
+    {
+      "epoch": 1.17536,
+      "grad_norm": 0.44502338767051697,
+      "learning_rate": 8.256204963971177e-05,
+      "loss": 0.4391,
+      "step": 3673
+    },
+    {
+      "epoch": 1.17568,
+      "grad_norm": 0.6574710011482239,
+      "learning_rate": 8.253002401921538e-05,
+      "loss": 0.6071,
+      "step": 3674
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.43824538588523865,
+      "learning_rate": 8.249799839871898e-05,
+      "loss": 0.3867,
+      "step": 3675
+    },
+    {
+      "epoch": 1.17632,
+      "grad_norm": 0.48271429538726807,
+      "learning_rate": 8.246597277822258e-05,
+      "loss": 0.5051,
+      "step": 3676
+    },
+    {
+      "epoch": 1.17664,
+      "grad_norm": 0.5330391526222229,
+      "learning_rate": 8.24339471577262e-05,
+      "loss": 0.4193,
+      "step": 3677
+    },
+    {
+      "epoch": 1.17696,
+      "grad_norm": 0.6149617433547974,
+      "learning_rate": 8.240192153722978e-05,
+      "loss": 0.5306,
+      "step": 3678
+    },
+    {
+      "epoch": 1.17728,
+      "grad_norm": 0.4942990839481354,
+      "learning_rate": 8.236989591673339e-05,
+      "loss": 0.4954,
+      "step": 3679
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.5647209882736206,
+      "learning_rate": 8.2337870296237e-05,
+      "loss": 0.4502,
+      "step": 3680
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 0.7517606616020203,
+      "learning_rate": 8.23058446757406e-05,
+      "loss": 0.6006,
+      "step": 3681
+    },
+    {
+      "epoch": 1.17824,
+      "grad_norm": 0.5346749424934387,
+      "learning_rate": 8.227381905524419e-05,
+      "loss": 0.5814,
+      "step": 3682
+    },
+    {
+      "epoch": 1.17856,
+      "grad_norm": 0.5920199751853943,
+      "learning_rate": 8.224179343474781e-05,
+      "loss": 0.4435,
+      "step": 3683
+    },
+    {
+      "epoch": 1.17888,
+      "grad_norm": 0.6421811580657959,
+      "learning_rate": 8.22097678142514e-05,
+      "loss": 0.4765,
+      "step": 3684
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 0.5013575553894043,
+      "learning_rate": 8.2177742193755e-05,
+      "loss": 0.4327,
+      "step": 3685
+    },
+    {
+      "epoch": 1.1795200000000001,
+      "grad_norm": 0.5183260440826416,
+      "learning_rate": 8.214571657325861e-05,
+      "loss": 0.3895,
+      "step": 3686
+    },
+    {
+      "epoch": 1.17984,
+      "grad_norm": 0.5056518912315369,
+      "learning_rate": 8.211369095276221e-05,
+      "loss": 0.5011,
+      "step": 3687
+    },
+    {
+      "epoch": 1.1801599999999999,
+      "grad_norm": 0.507459819316864,
+      "learning_rate": 8.20816653322658e-05,
+      "loss": 0.5223,
+      "step": 3688
+    },
+    {
+      "epoch": 1.18048,
+      "grad_norm": 0.496735543012619,
+      "learning_rate": 8.204963971176942e-05,
+      "loss": 0.3782,
+      "step": 3689
+    },
+    {
+      "epoch": 1.1808,
+      "grad_norm": 0.4158073663711548,
+      "learning_rate": 8.201761409127301e-05,
+      "loss": 0.358,
+      "step": 3690
+    },
+    {
+      "epoch": 1.18112,
+      "grad_norm": 0.598472535610199,
+      "learning_rate": 8.198558847077662e-05,
+      "loss": 0.4017,
+      "step": 3691
+    },
+    {
+      "epoch": 1.18144,
+      "grad_norm": 0.5890576243400574,
+      "learning_rate": 8.195356285028022e-05,
+      "loss": 0.5005,
+      "step": 3692
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 0.5941146016120911,
+      "learning_rate": 8.192153722978383e-05,
+      "loss": 0.4776,
+      "step": 3693
+    },
+    {
+      "epoch": 1.18208,
+      "grad_norm": 0.6465283632278442,
+      "learning_rate": 8.188951160928743e-05,
+      "loss": 0.4449,
+      "step": 3694
+    },
+    {
+      "epoch": 1.1824,
+      "grad_norm": 0.5248129963874817,
+      "learning_rate": 8.185748598879104e-05,
+      "loss": 0.4925,
+      "step": 3695
+    },
+    {
+      "epoch": 1.18272,
+      "grad_norm": 0.4793066680431366,
+      "learning_rate": 8.182546036829464e-05,
+      "loss": 0.4447,
+      "step": 3696
+    },
+    {
+      "epoch": 1.18304,
+      "grad_norm": 0.5477463603019714,
+      "learning_rate": 8.179343474779825e-05,
+      "loss": 0.4688,
+      "step": 3697
+    },
+    {
+      "epoch": 1.18336,
+      "grad_norm": 0.49594202637672424,
+      "learning_rate": 8.176140912730185e-05,
+      "loss": 0.5222,
+      "step": 3698
+    },
+    {
+      "epoch": 1.18368,
+      "grad_norm": 0.5936490297317505,
+      "learning_rate": 8.172938350680544e-05,
+      "loss": 0.5162,
+      "step": 3699
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 0.7380324006080627,
+      "learning_rate": 8.169735788630906e-05,
+      "loss": 0.5338,
+      "step": 3700
+    },
+    {
+      "epoch": 1.18432,
+      "grad_norm": 0.7104972004890442,
+      "learning_rate": 8.166533226581265e-05,
+      "loss": 0.5535,
+      "step": 3701
+    },
+    {
+      "epoch": 1.18464,
+      "grad_norm": 0.6806784868240356,
+      "learning_rate": 8.163330664531626e-05,
+      "loss": 0.4888,
+      "step": 3702
+    },
+    {
+      "epoch": 1.18496,
+      "grad_norm": 0.5115516781806946,
+      "learning_rate": 8.160128102481986e-05,
+      "loss": 0.4719,
+      "step": 3703
+    },
+    {
+      "epoch": 1.1852800000000001,
+      "grad_norm": 0.6195657253265381,
+      "learning_rate": 8.156925540432347e-05,
+      "loss": 0.4244,
+      "step": 3704
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.5008277297019958,
+      "learning_rate": 8.153722978382706e-05,
+      "loss": 0.5184,
+      "step": 3705
+    },
+    {
+      "epoch": 1.18592,
+      "grad_norm": 0.5629845857620239,
+      "learning_rate": 8.150520416333068e-05,
+      "loss": 0.3748,
+      "step": 3706
+    },
+    {
+      "epoch": 1.18624,
+      "grad_norm": 0.6944581866264343,
+      "learning_rate": 8.147317854283427e-05,
+      "loss": 0.5527,
+      "step": 3707
+    },
+    {
+      "epoch": 1.18656,
+      "grad_norm": 0.5493535995483398,
+      "learning_rate": 8.144115292233787e-05,
+      "loss": 0.4955,
+      "step": 3708
+    },
+    {
+      "epoch": 1.18688,
+      "grad_norm": 0.5718292593955994,
+      "learning_rate": 8.140912730184148e-05,
+      "loss": 0.474,
+      "step": 3709
+    },
+    {
+      "epoch": 1.1872,
+      "grad_norm": 0.5691624879837036,
+      "learning_rate": 8.137710168134508e-05,
+      "loss": 0.4553,
+      "step": 3710
+    },
+    {
+      "epoch": 1.18752,
+      "grad_norm": 0.7769295573234558,
+      "learning_rate": 8.134507606084867e-05,
+      "loss": 0.5862,
+      "step": 3711
+    },
+    {
+      "epoch": 1.18784,
+      "grad_norm": 0.5814850330352783,
+      "learning_rate": 8.131305044035229e-05,
+      "loss": 0.4679,
+      "step": 3712
+    },
+    {
+      "epoch": 1.1881599999999999,
+      "grad_norm": 0.5767715573310852,
+      "learning_rate": 8.128102481985588e-05,
+      "loss": 0.5919,
+      "step": 3713
+    },
+    {
+      "epoch": 1.18848,
+      "grad_norm": 0.40498098731040955,
+      "learning_rate": 8.124899919935949e-05,
+      "loss": 0.3729,
+      "step": 3714
+    },
+    {
+      "epoch": 1.1888,
+      "grad_norm": 0.5757904648780823,
+      "learning_rate": 8.12169735788631e-05,
+      "loss": 0.4756,
+      "step": 3715
+    },
+    {
+      "epoch": 1.18912,
+      "grad_norm": 0.5124853849411011,
+      "learning_rate": 8.11849479583667e-05,
+      "loss": 0.5411,
+      "step": 3716
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 0.6095352172851562,
+      "learning_rate": 8.11529223378703e-05,
+      "loss": 0.5516,
+      "step": 3717
+    },
+    {
+      "epoch": 1.18976,
+      "grad_norm": 0.7057108283042908,
+      "learning_rate": 8.112089671737391e-05,
+      "loss": 0.563,
+      "step": 3718
+    },
+    {
+      "epoch": 1.19008,
+      "grad_norm": 0.4753224849700928,
+      "learning_rate": 8.10888710968775e-05,
+      "loss": 0.4062,
+      "step": 3719
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.5392006635665894,
+      "learning_rate": 8.10568454763811e-05,
+      "loss": 0.4524,
+      "step": 3720
+    },
+    {
+      "epoch": 1.19072,
+      "grad_norm": 0.5466881990432739,
+      "learning_rate": 8.102481985588471e-05,
+      "loss": 0.4366,
+      "step": 3721
+    },
+    {
+      "epoch": 1.19104,
+      "grad_norm": 0.561684250831604,
+      "learning_rate": 8.099279423538831e-05,
+      "loss": 0.5849,
+      "step": 3722
+    },
+    {
+      "epoch": 1.19136,
+      "grad_norm": 0.5550235509872437,
+      "learning_rate": 8.096076861489192e-05,
+      "loss": 0.4486,
+      "step": 3723
+    },
+    {
+      "epoch": 1.19168,
+      "grad_norm": 0.6010719537734985,
+      "learning_rate": 8.092874299439552e-05,
+      "loss": 0.5832,
+      "step": 3724
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 0.6676036715507507,
+      "learning_rate": 8.089671737389913e-05,
+      "loss": 0.4839,
+      "step": 3725
+    },
+    {
+      "epoch": 1.19232,
+      "grad_norm": 0.4536895751953125,
+      "learning_rate": 8.086469175340272e-05,
+      "loss": 0.3505,
+      "step": 3726
+    },
+    {
+      "epoch": 1.19264,
+      "grad_norm": 0.6218906044960022,
+      "learning_rate": 8.083266613290634e-05,
+      "loss": 0.504,
+      "step": 3727
+    },
+    {
+      "epoch": 1.19296,
+      "grad_norm": 0.5234502553939819,
+      "learning_rate": 8.080064051240993e-05,
+      "loss": 0.488,
+      "step": 3728
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 0.6374446153640747,
+      "learning_rate": 8.076861489191353e-05,
+      "loss": 0.525,
+      "step": 3729
+    },
+    {
+      "epoch": 1.1936,
+      "grad_norm": 0.5029826164245605,
+      "learning_rate": 8.073658927141714e-05,
+      "loss": 0.4953,
+      "step": 3730
+    },
+    {
+      "epoch": 1.19392,
+      "grad_norm": 0.5930506587028503,
+      "learning_rate": 8.070456365092074e-05,
+      "loss": 0.4939,
+      "step": 3731
+    },
+    {
+      "epoch": 1.19424,
+      "grad_norm": 0.5390010476112366,
+      "learning_rate": 8.067253803042433e-05,
+      "loss": 0.4203,
+      "step": 3732
+    },
+    {
+      "epoch": 1.19456,
+      "grad_norm": 0.5153632164001465,
+      "learning_rate": 8.064051240992795e-05,
+      "loss": 0.4803,
+      "step": 3733
+    },
+    {
+      "epoch": 1.19488,
+      "grad_norm": 0.5181158781051636,
+      "learning_rate": 8.060848678943154e-05,
+      "loss": 0.4774,
+      "step": 3734
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 0.6834588646888733,
+      "learning_rate": 8.057646116893515e-05,
+      "loss": 0.6226,
+      "step": 3735
+    },
+    {
+      "epoch": 1.19552,
+      "grad_norm": 0.6172123551368713,
+      "learning_rate": 8.054443554843875e-05,
+      "loss": 0.4895,
+      "step": 3736
+    },
+    {
+      "epoch": 1.19584,
+      "grad_norm": 0.4453797936439514,
+      "learning_rate": 8.051240992794236e-05,
+      "loss": 0.3904,
+      "step": 3737
+    },
+    {
+      "epoch": 1.19616,
+      "grad_norm": 0.5064626932144165,
+      "learning_rate": 8.048038430744595e-05,
+      "loss": 0.4856,
+      "step": 3738
+    },
+    {
+      "epoch": 1.19648,
+      "grad_norm": 0.6470786333084106,
+      "learning_rate": 8.044835868694957e-05,
+      "loss": 0.4752,
+      "step": 3739
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 0.5168188214302063,
+      "learning_rate": 8.041633306645316e-05,
+      "loss": 0.4115,
+      "step": 3740
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 0.7332826852798462,
+      "learning_rate": 8.038430744595678e-05,
+      "loss": 0.5061,
+      "step": 3741
+    },
+    {
+      "epoch": 1.19744,
+      "grad_norm": 0.5260677337646484,
+      "learning_rate": 8.035228182546037e-05,
+      "loss": 0.4402,
+      "step": 3742
+    },
+    {
+      "epoch": 1.19776,
+      "grad_norm": 0.6136675477027893,
+      "learning_rate": 8.032025620496397e-05,
+      "loss": 0.5253,
+      "step": 3743
+    },
+    {
+      "epoch": 1.19808,
+      "grad_norm": 0.5818415284156799,
+      "learning_rate": 8.028823058446758e-05,
+      "loss": 0.5288,
+      "step": 3744
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 0.5627975463867188,
+      "learning_rate": 8.025620496397118e-05,
+      "loss": 0.4836,
+      "step": 3745
+    },
+    {
+      "epoch": 1.19872,
+      "grad_norm": 0.47365641593933105,
+      "learning_rate": 8.022417934347479e-05,
+      "loss": 0.4907,
+      "step": 3746
+    },
+    {
+      "epoch": 1.19904,
+      "grad_norm": 0.546622633934021,
+      "learning_rate": 8.019215372297839e-05,
+      "loss": 0.569,
+      "step": 3747
+    },
+    {
+      "epoch": 1.19936,
+      "grad_norm": 0.5679376125335693,
+      "learning_rate": 8.0160128102482e-05,
+      "loss": 0.4647,
+      "step": 3748
+    },
+    {
+      "epoch": 1.19968,
+      "grad_norm": 0.6089024543762207,
+      "learning_rate": 8.012810248198559e-05,
+      "loss": 0.5505,
+      "step": 3749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6015721559524536,
+      "learning_rate": 8.00960768614892e-05,
+      "loss": 0.6167,
+      "step": 3750
+    },
+    {
+      "epoch": 1.20032,
+      "grad_norm": 0.586521565914154,
+      "learning_rate": 8.00640512409928e-05,
+      "loss": 0.4307,
+      "step": 3751
+    },
+    {
+      "epoch": 1.20064,
+      "grad_norm": 0.5851666927337646,
+      "learning_rate": 8.00320256204964e-05,
+      "loss": 0.5294,
+      "step": 3752
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 0.5188756585121155,
+      "learning_rate": 8e-05,
+      "loss": 0.4408,
+      "step": 3753
+    },
+    {
+      "epoch": 1.20128,
+      "grad_norm": 0.6986459493637085,
+      "learning_rate": 7.996797437950361e-05,
+      "loss": 0.4961,
+      "step": 3754
+    },
+    {
+      "epoch": 1.2016,
+      "grad_norm": 0.5482494235038757,
+      "learning_rate": 7.99359487590072e-05,
+      "loss": 0.5392,
+      "step": 3755
+    },
+    {
+      "epoch": 1.2019199999999999,
+      "grad_norm": 0.5679329633712769,
+      "learning_rate": 7.990392313851082e-05,
+      "loss": 0.5306,
+      "step": 3756
+    },
+    {
+      "epoch": 1.20224,
+      "grad_norm": 0.43178993463516235,
+      "learning_rate": 7.987189751801441e-05,
+      "loss": 0.4487,
+      "step": 3757
+    },
+    {
+      "epoch": 1.20256,
+      "grad_norm": 0.5856627821922302,
+      "learning_rate": 7.983987189751802e-05,
+      "loss": 0.4573,
+      "step": 3758
+    },
+    {
+      "epoch": 1.20288,
+      "grad_norm": 0.5034602880477905,
+      "learning_rate": 7.980784627702162e-05,
+      "loss": 0.4797,
+      "step": 3759
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.6966837644577026,
+      "learning_rate": 7.977582065652523e-05,
+      "loss": 0.5661,
+      "step": 3760
+    },
+    {
+      "epoch": 1.20352,
+      "grad_norm": 0.5769866704940796,
+      "learning_rate": 7.974379503602882e-05,
+      "loss": 0.6036,
+      "step": 3761
+    },
+    {
+      "epoch": 1.20384,
+      "grad_norm": 0.498653382062912,
+      "learning_rate": 7.971176941553244e-05,
+      "loss": 0.5474,
+      "step": 3762
+    },
+    {
+      "epoch": 1.20416,
+      "grad_norm": 0.5855453014373779,
+      "learning_rate": 7.967974379503603e-05,
+      "loss": 0.5237,
+      "step": 3763
+    },
+    {
+      "epoch": 1.20448,
+      "grad_norm": 0.6261650919914246,
+      "learning_rate": 7.964771817453963e-05,
+      "loss": 0.5184,
+      "step": 3764
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.5929316878318787,
+      "learning_rate": 7.961569255404324e-05,
+      "loss": 0.5917,
+      "step": 3765
+    },
+    {
+      "epoch": 1.20512,
+      "grad_norm": 0.5437177419662476,
+      "learning_rate": 7.958366693354684e-05,
+      "loss": 0.4595,
+      "step": 3766
+    },
+    {
+      "epoch": 1.20544,
+      "grad_norm": 0.5217857956886292,
+      "learning_rate": 7.955164131305045e-05,
+      "loss": 0.4571,
+      "step": 3767
+    },
+    {
+      "epoch": 1.20576,
+      "grad_norm": 0.5040339827537537,
+      "learning_rate": 7.951961569255405e-05,
+      "loss": 0.4765,
+      "step": 3768
+    },
+    {
+      "epoch": 1.20608,
+      "grad_norm": 0.6500065326690674,
+      "learning_rate": 7.948759007205766e-05,
+      "loss": 0.5486,
+      "step": 3769
+    },
+    {
+      "epoch": 1.2064,
+      "grad_norm": 0.5158607363700867,
+      "learning_rate": 7.945556445156125e-05,
+      "loss": 0.4189,
+      "step": 3770
+    },
+    {
+      "epoch": 1.20672,
+      "grad_norm": 0.4876224398612976,
+      "learning_rate": 7.942353883106487e-05,
+      "loss": 0.3671,
+      "step": 3771
+    },
+    {
+      "epoch": 1.2070400000000001,
+      "grad_norm": 0.5719026327133179,
+      "learning_rate": 7.939151321056846e-05,
+      "loss": 0.5437,
+      "step": 3772
+    },
+    {
+      "epoch": 1.20736,
+      "grad_norm": 0.6320881843566895,
+      "learning_rate": 7.935948759007206e-05,
+      "loss": 0.5054,
+      "step": 3773
+    },
+    {
+      "epoch": 1.20768,
+      "grad_norm": 0.5910207629203796,
+      "learning_rate": 7.932746196957567e-05,
+      "loss": 0.4908,
+      "step": 3774
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 0.5409194827079773,
+      "learning_rate": 7.929543634907927e-05,
+      "loss": 0.5248,
+      "step": 3775
+    },
+    {
+      "epoch": 1.20832,
+      "grad_norm": 0.5160664916038513,
+      "learning_rate": 7.926341072858286e-05,
+      "loss": 0.4101,
+      "step": 3776
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 0.575675904750824,
+      "learning_rate": 7.923138510808648e-05,
+      "loss": 0.4597,
+      "step": 3777
+    },
+    {
+      "epoch": 1.20896,
+      "grad_norm": 0.4133860170841217,
+      "learning_rate": 7.919935948759007e-05,
+      "loss": 0.3786,
+      "step": 3778
+    },
+    {
+      "epoch": 1.20928,
+      "grad_norm": 0.7109612822532654,
+      "learning_rate": 7.916733386709368e-05,
+      "loss": 0.4172,
+      "step": 3779
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 0.5843127369880676,
+      "learning_rate": 7.913530824659728e-05,
+      "loss": 0.467,
+      "step": 3780
+    },
+    {
+      "epoch": 1.2099199999999999,
+      "grad_norm": 0.5504019856452942,
+      "learning_rate": 7.910328262610089e-05,
+      "loss": 0.4751,
+      "step": 3781
+    },
+    {
+      "epoch": 1.21024,
+      "grad_norm": 0.6779429316520691,
+      "learning_rate": 7.907125700560448e-05,
+      "loss": 0.4044,
+      "step": 3782
+    },
+    {
+      "epoch": 1.21056,
+      "grad_norm": 0.505162239074707,
+      "learning_rate": 7.90392313851081e-05,
+      "loss": 0.3729,
+      "step": 3783
+    },
+    {
+      "epoch": 1.21088,
+      "grad_norm": 0.5160920023918152,
+      "learning_rate": 7.900720576461169e-05,
+      "loss": 0.4762,
+      "step": 3784
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 0.617420494556427,
+      "learning_rate": 7.897518014411529e-05,
+      "loss": 0.4697,
+      "step": 3785
+    },
+    {
+      "epoch": 1.21152,
+      "grad_norm": 0.6025252342224121,
+      "learning_rate": 7.89431545236189e-05,
+      "loss": 0.4021,
+      "step": 3786
+    },
+    {
+      "epoch": 1.21184,
+      "grad_norm": 0.5325860977172852,
+      "learning_rate": 7.89111289031225e-05,
+      "loss": 0.4538,
+      "step": 3787
+    },
+    {
+      "epoch": 1.21216,
+      "grad_norm": 0.6371051073074341,
+      "learning_rate": 7.88791032826261e-05,
+      "loss": 0.4341,
+      "step": 3788
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 0.5417832732200623,
+      "learning_rate": 7.884707766212971e-05,
+      "loss": 0.5016,
+      "step": 3789
+    },
+    {
+      "epoch": 1.2128,
+      "grad_norm": 0.5250350832939148,
+      "learning_rate": 7.881505204163331e-05,
+      "loss": 0.4414,
+      "step": 3790
+    },
+    {
+      "epoch": 1.21312,
+      "grad_norm": 0.8166633248329163,
+      "learning_rate": 7.878302642113692e-05,
+      "loss": 0.4704,
+      "step": 3791
+    },
+    {
+      "epoch": 1.21344,
+      "grad_norm": 0.48320356011390686,
+      "learning_rate": 7.875100080064051e-05,
+      "loss": 0.3549,
+      "step": 3792
+    },
+    {
+      "epoch": 1.21376,
+      "grad_norm": 0.577102541923523,
+      "learning_rate": 7.871897518014412e-05,
+      "loss": 0.4573,
+      "step": 3793
+    },
+    {
+      "epoch": 1.21408,
+      "grad_norm": 0.5832131505012512,
+      "learning_rate": 7.868694955964772e-05,
+      "loss": 0.4603,
+      "step": 3794
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 0.580754280090332,
+      "learning_rate": 7.865492393915133e-05,
+      "loss": 0.5252,
+      "step": 3795
+    },
+    {
+      "epoch": 1.21472,
+      "grad_norm": 0.4934191107749939,
+      "learning_rate": 7.862289831865493e-05,
+      "loss": 0.4476,
+      "step": 3796
+    },
+    {
+      "epoch": 1.2150400000000001,
+      "grad_norm": 0.6352970600128174,
+      "learning_rate": 7.859087269815853e-05,
+      "loss": 0.5148,
+      "step": 3797
+    },
+    {
+      "epoch": 1.21536,
+      "grad_norm": 0.4379728436470032,
+      "learning_rate": 7.855884707766214e-05,
+      "loss": 0.4546,
+      "step": 3798
+    },
+    {
+      "epoch": 1.21568,
+      "grad_norm": 0.5445491671562195,
+      "learning_rate": 7.852682145716573e-05,
+      "loss": 0.5044,
+      "step": 3799
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.42642468214035034,
+      "learning_rate": 7.849479583666935e-05,
+      "loss": 0.3903,
+      "step": 3800
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 0.5949958562850952,
+      "learning_rate": 7.846277021617294e-05,
+      "loss": 0.5121,
+      "step": 3801
+    },
+    {
+      "epoch": 1.21664,
+      "grad_norm": 0.5158714056015015,
+      "learning_rate": 7.843074459567655e-05,
+      "loss": 0.5598,
+      "step": 3802
+    },
+    {
+      "epoch": 1.21696,
+      "grad_norm": 0.4800281822681427,
+      "learning_rate": 7.839871897518015e-05,
+      "loss": 0.398,
+      "step": 3803
+    },
+    {
+      "epoch": 1.21728,
+      "grad_norm": 0.5790397524833679,
+      "learning_rate": 7.836669335468375e-05,
+      "loss": 0.4995,
+      "step": 3804
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 0.6165149807929993,
+      "learning_rate": 7.833466773418735e-05,
+      "loss": 0.5135,
+      "step": 3805
+    },
+    {
+      "epoch": 1.21792,
+      "grad_norm": 0.5784443020820618,
+      "learning_rate": 7.830264211369096e-05,
+      "loss": 0.3921,
+      "step": 3806
+    },
+    {
+      "epoch": 1.21824,
+      "grad_norm": 0.6065093278884888,
+      "learning_rate": 7.827061649319456e-05,
+      "loss": 0.4194,
+      "step": 3807
+    },
+    {
+      "epoch": 1.21856,
+      "grad_norm": 0.6438487768173218,
+      "learning_rate": 7.823859087269816e-05,
+      "loss": 0.4778,
+      "step": 3808
+    },
+    {
+      "epoch": 1.21888,
+      "grad_norm": 0.6219966411590576,
+      "learning_rate": 7.820656525220176e-05,
+      "loss": 0.5568,
+      "step": 3809
+    },
+    {
+      "epoch": 1.2192,
+      "grad_norm": 0.6371620893478394,
+      "learning_rate": 7.817453963170537e-05,
+      "loss": 0.5072,
+      "step": 3810
+    },
+    {
+      "epoch": 1.21952,
+      "grad_norm": 0.4855334460735321,
+      "learning_rate": 7.814251401120896e-05,
+      "loss": 0.3912,
+      "step": 3811
+    },
+    {
+      "epoch": 1.21984,
+      "grad_norm": 0.7170711755752563,
+      "learning_rate": 7.811048839071258e-05,
+      "loss": 0.6486,
+      "step": 3812
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 0.5640225410461426,
+      "learning_rate": 7.807846277021617e-05,
+      "loss": 0.4523,
+      "step": 3813
+    },
+    {
+      "epoch": 1.22048,
+      "grad_norm": 0.6788540482521057,
+      "learning_rate": 7.804643714971978e-05,
+      "loss": 0.5354,
+      "step": 3814
+    },
+    {
+      "epoch": 1.2208,
+      "grad_norm": 0.6482903957366943,
+      "learning_rate": 7.801441152922338e-05,
+      "loss": 0.4682,
+      "step": 3815
+    },
+    {
+      "epoch": 1.22112,
+      "grad_norm": 0.6528977155685425,
+      "learning_rate": 7.798238590872698e-05,
+      "loss": 0.5106,
+      "step": 3816
+    },
+    {
+      "epoch": 1.22144,
+      "grad_norm": 0.5660985708236694,
+      "learning_rate": 7.795036028823059e-05,
+      "loss": 0.5454,
+      "step": 3817
+    },
+    {
+      "epoch": 1.22176,
+      "grad_norm": 0.6153733730316162,
+      "learning_rate": 7.79183346677342e-05,
+      "loss": 0.4348,
+      "step": 3818
+    },
+    {
+      "epoch": 1.22208,
+      "grad_norm": 0.8831184506416321,
+      "learning_rate": 7.78863090472378e-05,
+      "loss": 0.5975,
+      "step": 3819
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 0.4933340847492218,
+      "learning_rate": 7.785428342674139e-05,
+      "loss": 0.4542,
+      "step": 3820
+    },
+    {
+      "epoch": 1.22272,
+      "grad_norm": 0.6012506484985352,
+      "learning_rate": 7.782225780624501e-05,
+      "loss": 0.4689,
+      "step": 3821
+    },
+    {
+      "epoch": 1.22304,
+      "grad_norm": 0.44566601514816284,
+      "learning_rate": 7.77902321857486e-05,
+      "loss": 0.4342,
+      "step": 3822
+    },
+    {
+      "epoch": 1.22336,
+      "grad_norm": 0.5959725975990295,
+      "learning_rate": 7.77582065652522e-05,
+      "loss": 0.4745,
+      "step": 3823
+    },
+    {
+      "epoch": 1.2236799999999999,
+      "grad_norm": 0.5204259157180786,
+      "learning_rate": 7.772618094475581e-05,
+      "loss": 0.4569,
+      "step": 3824
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.5554651021957397,
+      "learning_rate": 7.769415532425941e-05,
+      "loss": 0.471,
+      "step": 3825
+    },
+    {
+      "epoch": 1.22432,
+      "grad_norm": 0.6630370020866394,
+      "learning_rate": 7.7662129703763e-05,
+      "loss": 0.5007,
+      "step": 3826
+    },
+    {
+      "epoch": 1.22464,
+      "grad_norm": 0.5968466997146606,
+      "learning_rate": 7.763010408326662e-05,
+      "loss": 0.4055,
+      "step": 3827
+    },
+    {
+      "epoch": 1.22496,
+      "grad_norm": 0.837504506111145,
+      "learning_rate": 7.759807846277021e-05,
+      "loss": 0.5325,
+      "step": 3828
+    },
+    {
+      "epoch": 1.22528,
+      "grad_norm": 0.540934681892395,
+      "learning_rate": 7.756605284227382e-05,
+      "loss": 0.4305,
+      "step": 3829
+    },
+    {
+      "epoch": 1.2256,
+      "grad_norm": 0.7246536016464233,
+      "learning_rate": 7.753402722177742e-05,
+      "loss": 0.5909,
+      "step": 3830
+    },
+    {
+      "epoch": 1.22592,
+      "grad_norm": 0.5650377869606018,
+      "learning_rate": 7.750200160128103e-05,
+      "loss": 0.4412,
+      "step": 3831
+    },
+    {
+      "epoch": 1.22624,
+      "grad_norm": 0.6988286972045898,
+      "learning_rate": 7.746997598078462e-05,
+      "loss": 0.538,
+      "step": 3832
+    },
+    {
+      "epoch": 1.22656,
+      "grad_norm": 0.6176184415817261,
+      "learning_rate": 7.743795036028824e-05,
+      "loss": 0.5109,
+      "step": 3833
+    },
+    {
+      "epoch": 1.22688,
+      "grad_norm": 0.4998340904712677,
+      "learning_rate": 7.740592473979183e-05,
+      "loss": 0.4507,
+      "step": 3834
+    },
+    {
+      "epoch": 1.2272,
+      "grad_norm": 0.4501403570175171,
+      "learning_rate": 7.737389911929545e-05,
+      "loss": 0.3597,
+      "step": 3835
+    },
+    {
+      "epoch": 1.22752,
+      "grad_norm": 0.4566451907157898,
+      "learning_rate": 7.734187349879904e-05,
+      "loss": 0.4508,
+      "step": 3836
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 0.4461815655231476,
+      "learning_rate": 7.730984787830264e-05,
+      "loss": 0.4031,
+      "step": 3837
+    },
+    {
+      "epoch": 1.22816,
+      "grad_norm": 0.49793076515197754,
+      "learning_rate": 7.727782225780625e-05,
+      "loss": 0.3918,
+      "step": 3838
+    },
+    {
+      "epoch": 1.22848,
+      "grad_norm": 0.568727970123291,
+      "learning_rate": 7.724579663730985e-05,
+      "loss": 0.4449,
+      "step": 3839
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.5771188735961914,
+      "learning_rate": 7.721377101681346e-05,
+      "loss": 0.4292,
+      "step": 3840
+    },
+    {
+      "epoch": 1.22912,
+      "grad_norm": 0.5841901898384094,
+      "learning_rate": 7.718174539631706e-05,
+      "loss": 0.5295,
+      "step": 3841
+    },
+    {
+      "epoch": 1.22944,
+      "grad_norm": 0.4257286787033081,
+      "learning_rate": 7.714971977582067e-05,
+      "loss": 0.4188,
+      "step": 3842
+    },
+    {
+      "epoch": 1.22976,
+      "grad_norm": 0.4501933157444,
+      "learning_rate": 7.711769415532426e-05,
+      "loss": 0.4745,
+      "step": 3843
+    },
+    {
+      "epoch": 1.23008,
+      "grad_norm": 0.5623858571052551,
+      "learning_rate": 7.708566853482788e-05,
+      "loss": 0.4015,
+      "step": 3844
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 0.4904121160507202,
+      "learning_rate": 7.705364291433147e-05,
+      "loss": 0.4232,
+      "step": 3845
+    },
+    {
+      "epoch": 1.23072,
+      "grad_norm": 0.6172638535499573,
+      "learning_rate": 7.702161729383507e-05,
+      "loss": 0.5373,
+      "step": 3846
+    },
+    {
+      "epoch": 1.23104,
+      "grad_norm": 0.5949060320854187,
+      "learning_rate": 7.698959167333868e-05,
+      "loss": 0.5266,
+      "step": 3847
+    },
+    {
+      "epoch": 1.23136,
+      "grad_norm": 0.6424908638000488,
+      "learning_rate": 7.695756605284228e-05,
+      "loss": 0.4689,
+      "step": 3848
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 0.7444224953651428,
+      "learning_rate": 7.692554043234587e-05,
+      "loss": 0.6062,
+      "step": 3849
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 0.5425844788551331,
+      "learning_rate": 7.689351481184949e-05,
+      "loss": 0.4487,
+      "step": 3850
+    },
+    {
+      "epoch": 1.23232,
+      "grad_norm": 0.6024022102355957,
+      "learning_rate": 7.686148919135308e-05,
+      "loss": 0.6801,
+      "step": 3851
+    },
+    {
+      "epoch": 1.23264,
+      "grad_norm": 0.5581463575363159,
+      "learning_rate": 7.682946357085669e-05,
+      "loss": 0.4202,
+      "step": 3852
+    },
+    {
+      "epoch": 1.23296,
+      "grad_norm": 0.7526379823684692,
+      "learning_rate": 7.679743795036029e-05,
+      "loss": 0.5053,
+      "step": 3853
+    },
+    {
+      "epoch": 1.23328,
+      "grad_norm": 0.7498249411582947,
+      "learning_rate": 7.67654123298639e-05,
+      "loss": 0.6682,
+      "step": 3854
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 0.5759205222129822,
+      "learning_rate": 7.673338670936749e-05,
+      "loss": 0.5068,
+      "step": 3855
+    },
+    {
+      "epoch": 1.23392,
+      "grad_norm": 0.6967865228652954,
+      "learning_rate": 7.670136108887111e-05,
+      "loss": 0.5972,
+      "step": 3856
+    },
+    {
+      "epoch": 1.23424,
+      "grad_norm": 0.6996470093727112,
+      "learning_rate": 7.66693354683747e-05,
+      "loss": 0.5206,
+      "step": 3857
+    },
+    {
+      "epoch": 1.23456,
+      "grad_norm": 0.6150788068771362,
+      "learning_rate": 7.66373098478783e-05,
+      "loss": 0.5165,
+      "step": 3858
+    },
+    {
+      "epoch": 1.23488,
+      "grad_norm": 0.5221530199050903,
+      "learning_rate": 7.660528422738191e-05,
+      "loss": 0.3766,
+      "step": 3859
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 0.4603080749511719,
+      "learning_rate": 7.657325860688551e-05,
+      "loss": 0.3936,
+      "step": 3860
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 0.49804002046585083,
+      "learning_rate": 7.654123298638912e-05,
+      "loss": 0.3662,
+      "step": 3861
+    },
+    {
+      "epoch": 1.23584,
+      "grad_norm": 0.5199658274650574,
+      "learning_rate": 7.650920736589272e-05,
+      "loss": 0.4355,
+      "step": 3862
+    },
+    {
+      "epoch": 1.23616,
+      "grad_norm": 0.4580058157444,
+      "learning_rate": 7.647718174539631e-05,
+      "loss": 0.4209,
+      "step": 3863
+    },
+    {
+      "epoch": 1.23648,
+      "grad_norm": 0.48444145917892456,
+      "learning_rate": 7.644515612489992e-05,
+      "loss": 0.4737,
+      "step": 3864
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 0.48522529006004333,
+      "learning_rate": 7.641313050440352e-05,
+      "loss": 0.4129,
+      "step": 3865
+    },
+    {
+      "epoch": 1.23712,
+      "grad_norm": 0.4880068302154541,
+      "learning_rate": 7.638110488390713e-05,
+      "loss": 0.5372,
+      "step": 3866
+    },
+    {
+      "epoch": 1.23744,
+      "grad_norm": 0.5113685131072998,
+      "learning_rate": 7.634907926341073e-05,
+      "loss": 0.4182,
+      "step": 3867
+    },
+    {
+      "epoch": 1.23776,
+      "grad_norm": 0.621930718421936,
+      "learning_rate": 7.631705364291434e-05,
+      "loss": 0.4601,
+      "step": 3868
+    },
+    {
+      "epoch": 1.23808,
+      "grad_norm": 0.5810978412628174,
+      "learning_rate": 7.628502802241794e-05,
+      "loss": 0.5367,
+      "step": 3869
+    },
+    {
+      "epoch": 1.2384,
+      "grad_norm": 0.5598787069320679,
+      "learning_rate": 7.625300240192153e-05,
+      "loss": 0.4845,
+      "step": 3870
+    },
+    {
+      "epoch": 1.23872,
+      "grad_norm": 0.5194786190986633,
+      "learning_rate": 7.622097678142515e-05,
+      "loss": 0.4764,
+      "step": 3871
+    },
+    {
+      "epoch": 1.23904,
+      "grad_norm": 0.606295108795166,
+      "learning_rate": 7.618895116092874e-05,
+      "loss": 0.5667,
+      "step": 3872
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 0.4259503185749054,
+      "learning_rate": 7.615692554043235e-05,
+      "loss": 0.3509,
+      "step": 3873
+    },
+    {
+      "epoch": 1.23968,
+      "grad_norm": 0.5368761420249939,
+      "learning_rate": 7.612489991993595e-05,
+      "loss": 0.4505,
+      "step": 3874
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5176033973693848,
+      "learning_rate": 7.609287429943956e-05,
+      "loss": 0.4364,
+      "step": 3875
+    },
+    {
+      "epoch": 1.24032,
+      "grad_norm": 0.5274773240089417,
+      "learning_rate": 7.606084867894315e-05,
+      "loss": 0.4798,
+      "step": 3876
+    },
+    {
+      "epoch": 1.24064,
+      "grad_norm": 0.4838143289089203,
+      "learning_rate": 7.602882305844677e-05,
+      "loss": 0.384,
+      "step": 3877
+    },
+    {
+      "epoch": 1.24096,
+      "grad_norm": 0.5674633383750916,
+      "learning_rate": 7.599679743795036e-05,
+      "loss": 0.3736,
+      "step": 3878
+    },
+    {
+      "epoch": 1.24128,
+      "grad_norm": 0.5301576852798462,
+      "learning_rate": 7.596477181745396e-05,
+      "loss": 0.5377,
+      "step": 3879
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.4818125069141388,
+      "learning_rate": 7.593274619695757e-05,
+      "loss": 0.3271,
+      "step": 3880
+    },
+    {
+      "epoch": 1.24192,
+      "grad_norm": 0.8076451420783997,
+      "learning_rate": 7.590072057646117e-05,
+      "loss": 0.7121,
+      "step": 3881
+    },
+    {
+      "epoch": 1.24224,
+      "grad_norm": 0.6082981824874878,
+      "learning_rate": 7.586869495596478e-05,
+      "loss": 0.4606,
+      "step": 3882
+    },
+    {
+      "epoch": 1.24256,
+      "grad_norm": 0.6973363757133484,
+      "learning_rate": 7.583666933546838e-05,
+      "loss": 0.4543,
+      "step": 3883
+    },
+    {
+      "epoch": 1.24288,
+      "grad_norm": 0.8451221585273743,
+      "learning_rate": 7.580464371497197e-05,
+      "loss": 0.5852,
+      "step": 3884
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.5761705040931702,
+      "learning_rate": 7.577261809447559e-05,
+      "loss": 0.5212,
+      "step": 3885
+    },
+    {
+      "epoch": 1.24352,
+      "grad_norm": 0.6205345392227173,
+      "learning_rate": 7.574059247397918e-05,
+      "loss": 0.5324,
+      "step": 3886
+    },
+    {
+      "epoch": 1.24384,
+      "grad_norm": 0.5968483090400696,
+      "learning_rate": 7.570856685348279e-05,
+      "loss": 0.5073,
+      "step": 3887
+    },
+    {
+      "epoch": 1.24416,
+      "grad_norm": 0.6250369548797607,
+      "learning_rate": 7.567654123298639e-05,
+      "loss": 0.6327,
+      "step": 3888
+    },
+    {
+      "epoch": 1.24448,
+      "grad_norm": 0.5104885697364807,
+      "learning_rate": 7.564451561249e-05,
+      "loss": 0.4124,
+      "step": 3889
+    },
+    {
+      "epoch": 1.2448,
+      "grad_norm": 0.47234126925468445,
+      "learning_rate": 7.56124899919936e-05,
+      "loss": 0.466,
+      "step": 3890
+    },
+    {
+      "epoch": 1.24512,
+      "grad_norm": 0.4923839271068573,
+      "learning_rate": 7.55804643714972e-05,
+      "loss": 0.5318,
+      "step": 3891
+    },
+    {
+      "epoch": 1.2454399999999999,
+      "grad_norm": 0.48667198419570923,
+      "learning_rate": 7.554843875100081e-05,
+      "loss": 0.3461,
+      "step": 3892
+    },
+    {
+      "epoch": 1.24576,
+      "grad_norm": 0.568196177482605,
+      "learning_rate": 7.55164131305044e-05,
+      "loss": 0.4678,
+      "step": 3893
+    },
+    {
+      "epoch": 1.24608,
+      "grad_norm": 0.46955716609954834,
+      "learning_rate": 7.548438751000802e-05,
+      "loss": 0.3697,
+      "step": 3894
+    },
+    {
+      "epoch": 1.2464,
+      "grad_norm": 0.4040127098560333,
+      "learning_rate": 7.545236188951161e-05,
+      "loss": 0.3569,
+      "step": 3895
+    },
+    {
+      "epoch": 1.24672,
+      "grad_norm": 0.6271029114723206,
+      "learning_rate": 7.542033626901522e-05,
+      "loss": 0.4432,
+      "step": 3896
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 0.5525457262992859,
+      "learning_rate": 7.538831064851882e-05,
+      "loss": 0.358,
+      "step": 3897
+    },
+    {
+      "epoch": 1.24736,
+      "grad_norm": 0.48730313777923584,
+      "learning_rate": 7.535628502802243e-05,
+      "loss": 0.3167,
+      "step": 3898
+    },
+    {
+      "epoch": 1.24768,
+      "grad_norm": 0.5657276511192322,
+      "learning_rate": 7.532425940752602e-05,
+      "loss": 0.4433,
+      "step": 3899
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 0.5439640879631042,
+      "learning_rate": 7.529223378702964e-05,
+      "loss": 0.4581,
+      "step": 3900
+    },
+    {
+      "epoch": 1.24832,
+      "grad_norm": 0.589332640171051,
+      "learning_rate": 7.526020816653323e-05,
+      "loss": 0.5374,
+      "step": 3901
+    },
+    {
+      "epoch": 1.24864,
+      "grad_norm": 0.7115777730941772,
+      "learning_rate": 7.522818254603683e-05,
+      "loss": 0.4868,
+      "step": 3902
+    },
+    {
+      "epoch": 1.24896,
+      "grad_norm": 0.5039493441581726,
+      "learning_rate": 7.519615692554044e-05,
+      "loss": 0.5137,
+      "step": 3903
+    },
+    {
+      "epoch": 1.24928,
+      "grad_norm": 0.5518924593925476,
+      "learning_rate": 7.516413130504404e-05,
+      "loss": 0.4999,
+      "step": 3904
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 0.5652024149894714,
+      "learning_rate": 7.513210568454763e-05,
+      "loss": 0.4855,
+      "step": 3905
+    },
+    {
+      "epoch": 1.24992,
+      "grad_norm": 0.6513644456863403,
+      "learning_rate": 7.510008006405125e-05,
+      "loss": 0.6493,
+      "step": 3906
+    },
+    {
+      "epoch": 1.25024,
+      "grad_norm": 0.6474030613899231,
+      "learning_rate": 7.506805444355484e-05,
+      "loss": 0.5563,
+      "step": 3907
+    },
+    {
+      "epoch": 1.2505600000000001,
+      "grad_norm": 0.5346472263336182,
+      "learning_rate": 7.503602882305845e-05,
+      "loss": 0.5114,
+      "step": 3908
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 0.7945509552955627,
+      "learning_rate": 7.500400320256205e-05,
+      "loss": 0.565,
+      "step": 3909
+    },
+    {
+      "epoch": 1.2511999999999999,
+      "grad_norm": 0.6321019530296326,
+      "learning_rate": 7.497197758206566e-05,
+      "loss": 0.6036,
+      "step": 3910
+    },
+    {
+      "epoch": 1.25152,
+      "grad_norm": 0.5269639492034912,
+      "learning_rate": 7.493995196156926e-05,
+      "loss": 0.4607,
+      "step": 3911
+    },
+    {
+      "epoch": 1.25184,
+      "grad_norm": 0.554114818572998,
+      "learning_rate": 7.490792634107287e-05,
+      "loss": 0.4606,
+      "step": 3912
+    },
+    {
+      "epoch": 1.25216,
+      "grad_norm": 0.5183867812156677,
+      "learning_rate": 7.487590072057647e-05,
+      "loss": 0.3908,
+      "step": 3913
+    },
+    {
+      "epoch": 1.25248,
+      "grad_norm": 0.5123093128204346,
+      "learning_rate": 7.484387510008006e-05,
+      "loss": 0.3912,
+      "step": 3914
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 0.570334792137146,
+      "learning_rate": 7.481184947958368e-05,
+      "loss": 0.4845,
+      "step": 3915
+    },
+    {
+      "epoch": 1.25312,
+      "grad_norm": 0.590844452381134,
+      "learning_rate": 7.477982385908727e-05,
+      "loss": 0.5232,
+      "step": 3916
+    },
+    {
+      "epoch": 1.2534399999999999,
+      "grad_norm": 0.538836658000946,
+      "learning_rate": 7.474779823859088e-05,
+      "loss": 0.3311,
+      "step": 3917
+    },
+    {
+      "epoch": 1.25376,
+      "grad_norm": 0.5297948122024536,
+      "learning_rate": 7.471577261809448e-05,
+      "loss": 0.3704,
+      "step": 3918
+    },
+    {
+      "epoch": 1.25408,
+      "grad_norm": 0.6930977702140808,
+      "learning_rate": 7.468374699759809e-05,
+      "loss": 0.4652,
+      "step": 3919
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.5363401174545288,
+      "learning_rate": 7.465172137710168e-05,
+      "loss": 0.4529,
+      "step": 3920
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 0.5604110956192017,
+      "learning_rate": 7.46196957566053e-05,
+      "loss": 0.3781,
+      "step": 3921
+    },
+    {
+      "epoch": 1.25504,
+      "grad_norm": 0.609009325504303,
+      "learning_rate": 7.458767013610889e-05,
+      "loss": 0.4804,
+      "step": 3922
+    },
+    {
+      "epoch": 1.25536,
+      "grad_norm": 0.4166547358036041,
+      "learning_rate": 7.455564451561249e-05,
+      "loss": 0.3465,
+      "step": 3923
+    },
+    {
+      "epoch": 1.25568,
+      "grad_norm": 0.6031524538993835,
+      "learning_rate": 7.45236188951161e-05,
+      "loss": 0.4108,
+      "step": 3924
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 0.5179421305656433,
+      "learning_rate": 7.44915932746197e-05,
+      "loss": 0.4183,
+      "step": 3925
+    },
+    {
+      "epoch": 1.25632,
+      "grad_norm": 0.4449617564678192,
+      "learning_rate": 7.445956765412329e-05,
+      "loss": 0.3669,
+      "step": 3926
+    },
+    {
+      "epoch": 1.25664,
+      "grad_norm": 0.5078256130218506,
+      "learning_rate": 7.442754203362691e-05,
+      "loss": 0.4935,
+      "step": 3927
+    },
+    {
+      "epoch": 1.25696,
+      "grad_norm": 0.5212058424949646,
+      "learning_rate": 7.43955164131305e-05,
+      "loss": 0.4163,
+      "step": 3928
+    },
+    {
+      "epoch": 1.25728,
+      "grad_norm": 0.5150148272514343,
+      "learning_rate": 7.436349079263412e-05,
+      "loss": 0.4135,
+      "step": 3929
+    },
+    {
+      "epoch": 1.2576,
+      "grad_norm": 0.4883052706718445,
+      "learning_rate": 7.433146517213771e-05,
+      "loss": 0.3572,
+      "step": 3930
+    },
+    {
+      "epoch": 1.25792,
+      "grad_norm": 0.5666371583938599,
+      "learning_rate": 7.429943955164132e-05,
+      "loss": 0.493,
+      "step": 3931
+    },
+    {
+      "epoch": 1.25824,
+      "grad_norm": 0.6077662706375122,
+      "learning_rate": 7.426741393114492e-05,
+      "loss": 0.4318,
+      "step": 3932
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 0.640296220779419,
+      "learning_rate": 7.423538831064852e-05,
+      "loss": 0.4025,
+      "step": 3933
+    },
+    {
+      "epoch": 1.25888,
+      "grad_norm": 0.5204077363014221,
+      "learning_rate": 7.420336269015212e-05,
+      "loss": 0.3775,
+      "step": 3934
+    },
+    {
+      "epoch": 1.2591999999999999,
+      "grad_norm": 0.6843252778053284,
+      "learning_rate": 7.417133706965573e-05,
+      "loss": 0.5427,
+      "step": 3935
+    },
+    {
+      "epoch": 1.25952,
+      "grad_norm": 0.6886276006698608,
+      "learning_rate": 7.413931144915933e-05,
+      "loss": 0.6394,
+      "step": 3936
+    },
+    {
+      "epoch": 1.25984,
+      "grad_norm": 0.6189560890197754,
+      "learning_rate": 7.410728582866293e-05,
+      "loss": 0.4797,
+      "step": 3937
+    },
+    {
+      "epoch": 1.26016,
+      "grad_norm": 0.5400795936584473,
+      "learning_rate": 7.407526020816654e-05,
+      "loss": 0.3681,
+      "step": 3938
+    },
+    {
+      "epoch": 1.26048,
+      "grad_norm": 0.5682949423789978,
+      "learning_rate": 7.404323458767014e-05,
+      "loss": 0.454,
+      "step": 3939
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 0.47714945673942566,
+      "learning_rate": 7.401120896717374e-05,
+      "loss": 0.4115,
+      "step": 3940
+    },
+    {
+      "epoch": 1.26112,
+      "grad_norm": 0.4488121569156647,
+      "learning_rate": 7.397918334667735e-05,
+      "loss": 0.4225,
+      "step": 3941
+    },
+    {
+      "epoch": 1.26144,
+      "grad_norm": 0.46745532751083374,
+      "learning_rate": 7.394715772618095e-05,
+      "loss": 0.3803,
+      "step": 3942
+    },
+    {
+      "epoch": 1.26176,
+      "grad_norm": 0.5473970770835876,
+      "learning_rate": 7.391513210568455e-05,
+      "loss": 0.3824,
+      "step": 3943
+    },
+    {
+      "epoch": 1.26208,
+      "grad_norm": 0.7033001184463501,
+      "learning_rate": 7.388310648518816e-05,
+      "loss": 0.4776,
+      "step": 3944
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.6013001799583435,
+      "learning_rate": 7.385108086469175e-05,
+      "loss": 0.4908,
+      "step": 3945
+    },
+    {
+      "epoch": 1.26272,
+      "grad_norm": 0.5715627074241638,
+      "learning_rate": 7.381905524419536e-05,
+      "loss": 0.5103,
+      "step": 3946
+    },
+    {
+      "epoch": 1.26304,
+      "grad_norm": 0.5724378824234009,
+      "learning_rate": 7.378702962369896e-05,
+      "loss": 0.4359,
+      "step": 3947
+    },
+    {
+      "epoch": 1.26336,
+      "grad_norm": 0.5238645672798157,
+      "learning_rate": 7.375500400320257e-05,
+      "loss": 0.4322,
+      "step": 3948
+    },
+    {
+      "epoch": 1.26368,
+      "grad_norm": 0.47541797161102295,
+      "learning_rate": 7.372297838270616e-05,
+      "loss": 0.4742,
+      "step": 3949
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 0.5383950471878052,
+      "learning_rate": 7.369095276220978e-05,
+      "loss": 0.4255,
+      "step": 3950
+    },
+    {
+      "epoch": 1.26432,
+      "grad_norm": 0.47851237654685974,
+      "learning_rate": 7.365892714171337e-05,
+      "loss": 0.4793,
+      "step": 3951
+    },
+    {
+      "epoch": 1.26464,
+      "grad_norm": 0.6847167611122131,
+      "learning_rate": 7.362690152121697e-05,
+      "loss": 0.5823,
+      "step": 3952
+    },
+    {
+      "epoch": 1.2649599999999999,
+      "grad_norm": 0.48094600439071655,
+      "learning_rate": 7.359487590072058e-05,
+      "loss": 0.4547,
+      "step": 3953
+    },
+    {
+      "epoch": 1.26528,
+      "grad_norm": 0.7407922744750977,
+      "learning_rate": 7.356285028022418e-05,
+      "loss": 0.5534,
+      "step": 3954
+    },
+    {
+      "epoch": 1.2656,
+      "grad_norm": 0.6665610671043396,
+      "learning_rate": 7.353082465972778e-05,
+      "loss": 0.4588,
+      "step": 3955
+    },
+    {
+      "epoch": 1.26592,
+      "grad_norm": 0.6162267923355103,
+      "learning_rate": 7.34987990392314e-05,
+      "loss": 0.4335,
+      "step": 3956
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 0.518232524394989,
+      "learning_rate": 7.346677341873499e-05,
+      "loss": 0.4236,
+      "step": 3957
+    },
+    {
+      "epoch": 1.2665600000000001,
+      "grad_norm": 0.5285632014274597,
+      "learning_rate": 7.343474779823859e-05,
+      "loss": 0.4433,
+      "step": 3958
+    },
+    {
+      "epoch": 1.26688,
+      "grad_norm": 0.6783008575439453,
+      "learning_rate": 7.34027221777422e-05,
+      "loss": 0.4229,
+      "step": 3959
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.6077352166175842,
+      "learning_rate": 7.33706965572458e-05,
+      "loss": 0.4441,
+      "step": 3960
+    },
+    {
+      "epoch": 1.26752,
+      "grad_norm": 0.8641874194145203,
+      "learning_rate": 7.33386709367494e-05,
+      "loss": 0.4251,
+      "step": 3961
+    },
+    {
+      "epoch": 1.26784,
+      "grad_norm": 0.6230022311210632,
+      "learning_rate": 7.330664531625301e-05,
+      "loss": 0.5224,
+      "step": 3962
+    },
+    {
+      "epoch": 1.26816,
+      "grad_norm": 0.527275800704956,
+      "learning_rate": 7.327461969575661e-05,
+      "loss": 0.4697,
+      "step": 3963
+    },
+    {
+      "epoch": 1.26848,
+      "grad_norm": 0.44392672181129456,
+      "learning_rate": 7.32425940752602e-05,
+      "loss": 0.403,
+      "step": 3964
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 0.58003830909729,
+      "learning_rate": 7.321056845476382e-05,
+      "loss": 0.4779,
+      "step": 3965
+    },
+    {
+      "epoch": 1.26912,
+      "grad_norm": 0.5485794544219971,
+      "learning_rate": 7.317854283426741e-05,
+      "loss": 0.3729,
+      "step": 3966
+    },
+    {
+      "epoch": 1.26944,
+      "grad_norm": 0.7353869080543518,
+      "learning_rate": 7.314651721377102e-05,
+      "loss": 0.4973,
+      "step": 3967
+    },
+    {
+      "epoch": 1.26976,
+      "grad_norm": 0.6707388758659363,
+      "learning_rate": 7.311449159327462e-05,
+      "loss": 0.4991,
+      "step": 3968
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 0.5861333012580872,
+      "learning_rate": 7.308246597277823e-05,
+      "loss": 0.5123,
+      "step": 3969
+    },
+    {
+      "epoch": 1.2704,
+      "grad_norm": 0.5457465052604675,
+      "learning_rate": 7.305044035228182e-05,
+      "loss": 0.3926,
+      "step": 3970
+    },
+    {
+      "epoch": 1.27072,
+      "grad_norm": 0.583090603351593,
+      "learning_rate": 7.301841473178544e-05,
+      "loss": 0.4001,
+      "step": 3971
+    },
+    {
+      "epoch": 1.27104,
+      "grad_norm": 0.5651604533195496,
+      "learning_rate": 7.298638911128903e-05,
+      "loss": 0.5181,
+      "step": 3972
+    },
+    {
+      "epoch": 1.27136,
+      "grad_norm": 0.5570679903030396,
+      "learning_rate": 7.295436349079263e-05,
+      "loss": 0.5434,
+      "step": 3973
+    },
+    {
+      "epoch": 1.27168,
+      "grad_norm": 0.5214605927467346,
+      "learning_rate": 7.292233787029624e-05,
+      "loss": 0.3899,
+      "step": 3974
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.6232374310493469,
+      "learning_rate": 7.289031224979984e-05,
+      "loss": 0.5431,
+      "step": 3975
+    },
+    {
+      "epoch": 1.2723200000000001,
+      "grad_norm": 0.5358322262763977,
+      "learning_rate": 7.285828662930345e-05,
+      "loss": 0.423,
+      "step": 3976
+    },
+    {
+      "epoch": 1.27264,
+      "grad_norm": 0.538800835609436,
+      "learning_rate": 7.282626100880705e-05,
+      "loss": 0.5341,
+      "step": 3977
+    },
+    {
+      "epoch": 1.2729599999999999,
+      "grad_norm": 0.5884597897529602,
+      "learning_rate": 7.279423538831064e-05,
+      "loss": 0.4552,
+      "step": 3978
+    },
+    {
+      "epoch": 1.27328,
+      "grad_norm": 0.5485889315605164,
+      "learning_rate": 7.276220976781426e-05,
+      "loss": 0.4937,
+      "step": 3979
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 0.4951423406600952,
+      "learning_rate": 7.273018414731785e-05,
+      "loss": 0.3608,
+      "step": 3980
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 0.6644945740699768,
+      "learning_rate": 7.269815852682146e-05,
+      "loss": 0.6267,
+      "step": 3981
+    },
+    {
+      "epoch": 1.27424,
+      "grad_norm": 0.4340987205505371,
+      "learning_rate": 7.266613290632506e-05,
+      "loss": 0.3378,
+      "step": 3982
+    },
+    {
+      "epoch": 1.2745600000000001,
+      "grad_norm": 0.48130160570144653,
+      "learning_rate": 7.263410728582867e-05,
+      "loss": 0.4761,
+      "step": 3983
+    },
+    {
+      "epoch": 1.27488,
+      "grad_norm": 0.5665215253829956,
+      "learning_rate": 7.260208166533227e-05,
+      "loss": 0.4103,
+      "step": 3984
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 0.497943252325058,
+      "learning_rate": 7.257005604483588e-05,
+      "loss": 0.5355,
+      "step": 3985
+    },
+    {
+      "epoch": 1.27552,
+      "grad_norm": 0.4780994653701782,
+      "learning_rate": 7.253803042433948e-05,
+      "loss": 0.3782,
+      "step": 3986
+    },
+    {
+      "epoch": 1.27584,
+      "grad_norm": 0.6421124339103699,
+      "learning_rate": 7.250600480384307e-05,
+      "loss": 0.5117,
+      "step": 3987
+    },
+    {
+      "epoch": 1.27616,
+      "grad_norm": 0.616295576095581,
+      "learning_rate": 7.247397918334669e-05,
+      "loss": 0.5078,
+      "step": 3988
+    },
+    {
+      "epoch": 1.27648,
+      "grad_norm": 0.5751025080680847,
+      "learning_rate": 7.244195356285028e-05,
+      "loss": 0.4615,
+      "step": 3989
+    },
+    {
+      "epoch": 1.2768,
+      "grad_norm": 0.6147909760475159,
+      "learning_rate": 7.240992794235389e-05,
+      "loss": 0.5128,
+      "step": 3990
+    },
+    {
+      "epoch": 1.27712,
+      "grad_norm": 0.48079913854599,
+      "learning_rate": 7.237790232185749e-05,
+      "loss": 0.412,
+      "step": 3991
+    },
+    {
+      "epoch": 1.27744,
+      "grad_norm": 0.4749862253665924,
+      "learning_rate": 7.23458767013611e-05,
+      "loss": 0.5874,
+      "step": 3992
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 0.45371851325035095,
+      "learning_rate": 7.231385108086469e-05,
+      "loss": 0.4113,
+      "step": 3993
+    },
+    {
+      "epoch": 1.27808,
+      "grad_norm": 0.39276596903800964,
+      "learning_rate": 7.228182546036831e-05,
+      "loss": 0.4053,
+      "step": 3994
+    },
+    {
+      "epoch": 1.2784,
+      "grad_norm": 0.5170107483863831,
+      "learning_rate": 7.22497998398719e-05,
+      "loss": 0.3885,
+      "step": 3995
+    },
+    {
+      "epoch": 1.27872,
+      "grad_norm": 0.8359499573707581,
+      "learning_rate": 7.22177742193755e-05,
+      "loss": 0.5215,
+      "step": 3996
+    },
+    {
+      "epoch": 1.27904,
+      "grad_norm": 0.5134141445159912,
+      "learning_rate": 7.218574859887911e-05,
+      "loss": 0.3801,
+      "step": 3997
+    },
+    {
+      "epoch": 1.27936,
+      "grad_norm": 0.45856136083602905,
+      "learning_rate": 7.215372297838271e-05,
+      "loss": 0.407,
+      "step": 3998
+    },
+    {
+      "epoch": 1.27968,
+      "grad_norm": 0.5978235602378845,
+      "learning_rate": 7.21216973578863e-05,
+      "loss": 0.3671,
+      "step": 3999
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.4966351389884949,
+      "learning_rate": 7.208967173738992e-05,
+      "loss": 0.4723,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2803200000000001,
+      "grad_norm": 0.5759880542755127,
+      "learning_rate": 7.205764611689351e-05,
+      "loss": 0.4486,
+      "step": 4001
+    },
+    {
+      "epoch": 1.28064,
+      "grad_norm": 0.6123223900794983,
+      "learning_rate": 7.202562049639712e-05,
+      "loss": 0.4655,
+      "step": 4002
+    },
+    {
+      "epoch": 1.2809599999999999,
+      "grad_norm": 0.6664820313453674,
+      "learning_rate": 7.199359487590072e-05,
+      "loss": 0.5929,
+      "step": 4003
+    },
+    {
+      "epoch": 1.28128,
+      "grad_norm": 0.6174222230911255,
+      "learning_rate": 7.196156925540433e-05,
+      "loss": 0.4813,
+      "step": 4004
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.5147210359573364,
+      "learning_rate": 7.192954363490792e-05,
+      "loss": 0.5057,
+      "step": 4005
+    },
+    {
+      "epoch": 1.28192,
+      "grad_norm": 0.6160879731178284,
+      "learning_rate": 7.189751801441154e-05,
+      "loss": 0.4943,
+      "step": 4006
+    },
+    {
+      "epoch": 1.28224,
+      "grad_norm": 0.5017792582511902,
+      "learning_rate": 7.186549239391513e-05,
+      "loss": 0.39,
+      "step": 4007
+    },
+    {
+      "epoch": 1.28256,
+      "grad_norm": 0.5330730676651001,
+      "learning_rate": 7.183346677341873e-05,
+      "loss": 0.4034,
+      "step": 4008
+    },
+    {
+      "epoch": 1.28288,
+      "grad_norm": 0.5036188960075378,
+      "learning_rate": 7.180144115292234e-05,
+      "loss": 0.5001,
+      "step": 4009
+    },
+    {
+      "epoch": 1.2832,
+      "grad_norm": 0.5469613671302795,
+      "learning_rate": 7.176941553242594e-05,
+      "loss": 0.356,
+      "step": 4010
+    },
+    {
+      "epoch": 1.28352,
+      "grad_norm": 0.45891469717025757,
+      "learning_rate": 7.173738991192955e-05,
+      "loss": 0.4313,
+      "step": 4011
+    },
+    {
+      "epoch": 1.28384,
+      "grad_norm": 0.6650918126106262,
+      "learning_rate": 7.170536429143315e-05,
+      "loss": 0.5814,
+      "step": 4012
+    },
+    {
+      "epoch": 1.28416,
+      "grad_norm": 0.4582373797893524,
+      "learning_rate": 7.167333867093676e-05,
+      "loss": 0.3363,
+      "step": 4013
+    },
+    {
+      "epoch": 1.28448,
+      "grad_norm": 0.4977867901325226,
+      "learning_rate": 7.164131305044035e-05,
+      "loss": 0.4068,
+      "step": 4014
+    },
+    {
+      "epoch": 1.2848,
+      "grad_norm": 0.5789068937301636,
+      "learning_rate": 7.160928742994397e-05,
+      "loss": 0.4888,
+      "step": 4015
+    },
+    {
+      "epoch": 1.28512,
+      "grad_norm": 0.45188412070274353,
+      "learning_rate": 7.157726180944756e-05,
+      "loss": 0.3578,
+      "step": 4016
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 0.5131725072860718,
+      "learning_rate": 7.154523618895116e-05,
+      "loss": 0.3843,
+      "step": 4017
+    },
+    {
+      "epoch": 1.28576,
+      "grad_norm": 0.57449871301651,
+      "learning_rate": 7.151321056845477e-05,
+      "loss": 0.4758,
+      "step": 4018
+    },
+    {
+      "epoch": 1.2860800000000001,
+      "grad_norm": 0.5522132515907288,
+      "learning_rate": 7.148118494795837e-05,
+      "loss": 0.3955,
+      "step": 4019
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 0.6256917119026184,
+      "learning_rate": 7.144915932746196e-05,
+      "loss": 0.5476,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2867199999999999,
+      "grad_norm": 0.7434900403022766,
+      "learning_rate": 7.141713370696558e-05,
+      "loss": 0.6138,
+      "step": 4021
+    },
+    {
+      "epoch": 1.28704,
+      "grad_norm": 0.5364322066307068,
+      "learning_rate": 7.138510808646917e-05,
+      "loss": 0.4106,
+      "step": 4022
+    },
+    {
+      "epoch": 1.28736,
+      "grad_norm": 0.6002820134162903,
+      "learning_rate": 7.135308246597279e-05,
+      "loss": 0.3969,
+      "step": 4023
+    },
+    {
+      "epoch": 1.28768,
+      "grad_norm": 0.5548863410949707,
+      "learning_rate": 7.132105684547638e-05,
+      "loss": 0.4437,
+      "step": 4024
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 0.5248554348945618,
+      "learning_rate": 7.128903122497999e-05,
+      "loss": 0.3805,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2883200000000001,
+      "grad_norm": 0.5845619440078735,
+      "learning_rate": 7.125700560448359e-05,
+      "loss": 0.5344,
+      "step": 4026
+    },
+    {
+      "epoch": 1.28864,
+      "grad_norm": 0.4877409338951111,
+      "learning_rate": 7.12249799839872e-05,
+      "loss": 0.3569,
+      "step": 4027
+    },
+    {
+      "epoch": 1.2889599999999999,
+      "grad_norm": 0.5320272445678711,
+      "learning_rate": 7.119295436349079e-05,
+      "loss": 0.3842,
+      "step": 4028
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 0.6298758387565613,
+      "learning_rate": 7.11609287429944e-05,
+      "loss": 0.565,
+      "step": 4029
+    },
+    {
+      "epoch": 1.2896,
+      "grad_norm": 0.6134064793586731,
+      "learning_rate": 7.1128903122498e-05,
+      "loss": 0.5814,
+      "step": 4030
+    },
+    {
+      "epoch": 1.28992,
+      "grad_norm": 0.5807588696479797,
+      "learning_rate": 7.10968775020016e-05,
+      "loss": 0.5038,
+      "step": 4031
+    },
+    {
+      "epoch": 1.29024,
+      "grad_norm": 0.4765007495880127,
+      "learning_rate": 7.10648518815052e-05,
+      "loss": 0.4751,
+      "step": 4032
+    },
+    {
+      "epoch": 1.29056,
+      "grad_norm": 0.7299067378044128,
+      "learning_rate": 7.103282626100881e-05,
+      "loss": 0.4291,
+      "step": 4033
+    },
+    {
+      "epoch": 1.29088,
+      "grad_norm": 0.5947996973991394,
+      "learning_rate": 7.100080064051242e-05,
+      "loss": 0.478,
+      "step": 4034
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 0.5586035251617432,
+      "learning_rate": 7.096877502001602e-05,
+      "loss": 0.3748,
+      "step": 4035
+    },
+    {
+      "epoch": 1.29152,
+      "grad_norm": 0.5497400760650635,
+      "learning_rate": 7.093674939951963e-05,
+      "loss": 0.4324,
+      "step": 4036
+    },
+    {
+      "epoch": 1.29184,
+      "grad_norm": 0.5979791879653931,
+      "learning_rate": 7.090472377902322e-05,
+      "loss": 0.5278,
+      "step": 4037
+    },
+    {
+      "epoch": 1.29216,
+      "grad_norm": 0.5862144231796265,
+      "learning_rate": 7.087269815852684e-05,
+      "loss": 0.4163,
+      "step": 4038
+    },
+    {
+      "epoch": 1.29248,
+      "grad_norm": 0.5308144092559814,
+      "learning_rate": 7.084067253803043e-05,
+      "loss": 0.396,
+      "step": 4039
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.6651210784912109,
+      "learning_rate": 7.080864691753403e-05,
+      "loss": 0.473,
+      "step": 4040
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 0.5166866779327393,
+      "learning_rate": 7.077662129703764e-05,
+      "loss": 0.4188,
+      "step": 4041
+    },
+    {
+      "epoch": 1.29344,
+      "grad_norm": 0.5546486377716064,
+      "learning_rate": 7.074459567654124e-05,
+      "loss": 0.4755,
+      "step": 4042
+    },
+    {
+      "epoch": 1.29376,
+      "grad_norm": 0.5515362024307251,
+      "learning_rate": 7.071257005604483e-05,
+      "loss": 0.4139,
+      "step": 4043
+    },
+    {
+      "epoch": 1.2940800000000001,
+      "grad_norm": 0.5577837824821472,
+      "learning_rate": 7.068054443554845e-05,
+      "loss": 0.4717,
+      "step": 4044
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 0.6456009745597839,
+      "learning_rate": 7.064851881505204e-05,
+      "loss": 0.5032,
+      "step": 4045
+    },
+    {
+      "epoch": 1.2947199999999999,
+      "grad_norm": 0.6643773913383484,
+      "learning_rate": 7.061649319455565e-05,
+      "loss": 0.4709,
+      "step": 4046
+    },
+    {
+      "epoch": 1.29504,
+      "grad_norm": 0.5417097806930542,
+      "learning_rate": 7.058446757405925e-05,
+      "loss": 0.4165,
+      "step": 4047
+    },
+    {
+      "epoch": 1.29536,
+      "grad_norm": 0.5472874641418457,
+      "learning_rate": 7.055244195356286e-05,
+      "loss": 0.4587,
+      "step": 4048
+    },
+    {
+      "epoch": 1.29568,
+      "grad_norm": 0.5550146102905273,
+      "learning_rate": 7.052041633306645e-05,
+      "loss": 0.3799,
+      "step": 4049
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 0.613254964351654,
+      "learning_rate": 7.048839071257007e-05,
+      "loss": 0.5046,
+      "step": 4050
+    },
+    {
+      "epoch": 1.29632,
+      "grad_norm": 0.49978238344192505,
+      "learning_rate": 7.045636509207366e-05,
+      "loss": 0.4725,
+      "step": 4051
+    },
+    {
+      "epoch": 1.29664,
+      "grad_norm": 0.5745127201080322,
+      "learning_rate": 7.042433947157726e-05,
+      "loss": 0.45,
+      "step": 4052
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 0.729048490524292,
+      "learning_rate": 7.039231385108087e-05,
+      "loss": 0.4084,
+      "step": 4053
+    },
+    {
+      "epoch": 1.29728,
+      "grad_norm": 0.5315410494804382,
+      "learning_rate": 7.036028823058447e-05,
+      "loss": 0.3927,
+      "step": 4054
+    },
+    {
+      "epoch": 1.2976,
+      "grad_norm": 0.49169495701789856,
+      "learning_rate": 7.032826261008808e-05,
+      "loss": 0.3984,
+      "step": 4055
+    },
+    {
+      "epoch": 1.29792,
+      "grad_norm": 0.6475728154182434,
+      "learning_rate": 7.029623698959168e-05,
+      "loss": 0.6069,
+      "step": 4056
+    },
+    {
+      "epoch": 1.29824,
+      "grad_norm": 0.6325414776802063,
+      "learning_rate": 7.026421136909529e-05,
+      "loss": 0.4953,
+      "step": 4057
+    },
+    {
+      "epoch": 1.29856,
+      "grad_norm": 0.6087117195129395,
+      "learning_rate": 7.023218574859888e-05,
+      "loss": 0.4486,
+      "step": 4058
+    },
+    {
+      "epoch": 1.29888,
+      "grad_norm": 0.6098604202270508,
+      "learning_rate": 7.02001601281025e-05,
+      "loss": 0.4406,
+      "step": 4059
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 0.5292263031005859,
+      "learning_rate": 7.016813450760609e-05,
+      "loss": 0.3681,
+      "step": 4060
+    },
+    {
+      "epoch": 1.29952,
+      "grad_norm": 0.6145195364952087,
+      "learning_rate": 7.013610888710969e-05,
+      "loss": 0.5401,
+      "step": 4061
+    },
+    {
+      "epoch": 1.29984,
+      "grad_norm": 0.5025644302368164,
+      "learning_rate": 7.01040832666133e-05,
+      "loss": 0.6116,
+      "step": 4062
+    },
+    {
+      "epoch": 1.30016,
+      "grad_norm": 0.46034494042396545,
+      "learning_rate": 7.00720576461169e-05,
+      "loss": 0.4578,
+      "step": 4063
+    },
+    {
+      "epoch": 1.30048,
+      "grad_norm": 0.42470476031303406,
+      "learning_rate": 7.004003202562049e-05,
+      "loss": 0.3315,
+      "step": 4064
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.49637478590011597,
+      "learning_rate": 7.000800640512411e-05,
+      "loss": 0.4175,
+      "step": 4065
+    },
+    {
+      "epoch": 1.30112,
+      "grad_norm": 0.5626251697540283,
+      "learning_rate": 6.99759807846277e-05,
+      "loss": 0.4516,
+      "step": 4066
+    },
+    {
+      "epoch": 1.30144,
+      "grad_norm": 0.6673545837402344,
+      "learning_rate": 6.99439551641313e-05,
+      "loss": 0.6629,
+      "step": 4067
+    },
+    {
+      "epoch": 1.30176,
+      "grad_norm": 0.667945384979248,
+      "learning_rate": 6.991192954363491e-05,
+      "loss": 0.448,
+      "step": 4068
+    },
+    {
+      "epoch": 1.3020800000000001,
+      "grad_norm": 0.5297556519508362,
+      "learning_rate": 6.987990392313852e-05,
+      "loss": 0.3735,
+      "step": 4069
+    },
+    {
+      "epoch": 1.3024,
+      "grad_norm": 0.5738373398780823,
+      "learning_rate": 6.984787830264212e-05,
+      "loss": 0.5218,
+      "step": 4070
+    },
+    {
+      "epoch": 1.3027199999999999,
+      "grad_norm": 0.47596144676208496,
+      "learning_rate": 6.981585268214572e-05,
+      "loss": 0.3919,
+      "step": 4071
+    },
+    {
+      "epoch": 1.30304,
+      "grad_norm": 0.4517759680747986,
+      "learning_rate": 6.978382706164932e-05,
+      "loss": 0.3178,
+      "step": 4072
+    },
+    {
+      "epoch": 1.30336,
+      "grad_norm": 0.5273604989051819,
+      "learning_rate": 6.975180144115293e-05,
+      "loss": 0.3918,
+      "step": 4073
+    },
+    {
+      "epoch": 1.30368,
+      "grad_norm": 0.47720858454704285,
+      "learning_rate": 6.971977582065653e-05,
+      "loss": 0.3288,
+      "step": 4074
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 0.5936060547828674,
+      "learning_rate": 6.968775020016013e-05,
+      "loss": 0.4503,
+      "step": 4075
+    },
+    {
+      "epoch": 1.30432,
+      "grad_norm": 0.47579148411750793,
+      "learning_rate": 6.965572457966373e-05,
+      "loss": 0.3934,
+      "step": 4076
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 0.6289665102958679,
+      "learning_rate": 6.962369895916734e-05,
+      "loss": 0.4261,
+      "step": 4077
+    },
+    {
+      "epoch": 1.30496,
+      "grad_norm": 0.6902977824211121,
+      "learning_rate": 6.959167333867093e-05,
+      "loss": 0.5057,
+      "step": 4078
+    },
+    {
+      "epoch": 1.30528,
+      "grad_norm": 0.5730543732643127,
+      "learning_rate": 6.955964771817455e-05,
+      "loss": 0.4642,
+      "step": 4079
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.5837101340293884,
+      "learning_rate": 6.952762209767814e-05,
+      "loss": 0.4394,
+      "step": 4080
+    },
+    {
+      "epoch": 1.30592,
+      "grad_norm": 0.6272860169410706,
+      "learning_rate": 6.949559647718175e-05,
+      "loss": 0.5726,
+      "step": 4081
+    },
+    {
+      "epoch": 1.30624,
+      "grad_norm": 0.6515441536903381,
+      "learning_rate": 6.946357085668535e-05,
+      "loss": 0.5202,
+      "step": 4082
+    },
+    {
+      "epoch": 1.30656,
+      "grad_norm": 0.8152804970741272,
+      "learning_rate": 6.943154523618895e-05,
+      "loss": 0.6933,
+      "step": 4083
+    },
+    {
+      "epoch": 1.30688,
+      "grad_norm": 0.49030181765556335,
+      "learning_rate": 6.939951961569256e-05,
+      "loss": 0.4466,
+      "step": 4084
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 0.5107598304748535,
+      "learning_rate": 6.936749399519616e-05,
+      "loss": 0.3847,
+      "step": 4085
+    },
+    {
+      "epoch": 1.30752,
+      "grad_norm": 0.4097210764884949,
+      "learning_rate": 6.933546837469977e-05,
+      "loss": 0.4056,
+      "step": 4086
+    },
+    {
+      "epoch": 1.3078400000000001,
+      "grad_norm": 0.624835729598999,
+      "learning_rate": 6.930344275420336e-05,
+      "loss": 0.4759,
+      "step": 4087
+    },
+    {
+      "epoch": 1.30816,
+      "grad_norm": 0.4966779053211212,
+      "learning_rate": 6.927141713370698e-05,
+      "loss": 0.4476,
+      "step": 4088
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 0.6278024911880493,
+      "learning_rate": 6.923939151321057e-05,
+      "loss": 0.4286,
+      "step": 4089
+    },
+    {
+      "epoch": 1.3088,
+      "grad_norm": 0.6017400026321411,
+      "learning_rate": 6.920736589271417e-05,
+      "loss": 0.5026,
+      "step": 4090
+    },
+    {
+      "epoch": 1.30912,
+      "grad_norm": 0.4770606458187103,
+      "learning_rate": 6.917534027221778e-05,
+      "loss": 0.4159,
+      "step": 4091
+    },
+    {
+      "epoch": 1.30944,
+      "grad_norm": 0.6533477902412415,
+      "learning_rate": 6.914331465172138e-05,
+      "loss": 0.4366,
+      "step": 4092
+    },
+    {
+      "epoch": 1.30976,
+      "grad_norm": 0.4905068278312683,
+      "learning_rate": 6.911128903122498e-05,
+      "loss": 0.324,
+      "step": 4093
+    },
+    {
+      "epoch": 1.3100800000000001,
+      "grad_norm": 0.46853113174438477,
+      "learning_rate": 6.90792634107286e-05,
+      "loss": 0.4268,
+      "step": 4094
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 0.521338164806366,
+      "learning_rate": 6.904723779023218e-05,
+      "loss": 0.4694,
+      "step": 4095
+    },
+    {
+      "epoch": 1.3107199999999999,
+      "grad_norm": 0.6344218850135803,
+      "learning_rate": 6.901521216973579e-05,
+      "loss": 0.4909,
+      "step": 4096
+    },
+    {
+      "epoch": 1.31104,
+      "grad_norm": 0.5320963263511658,
+      "learning_rate": 6.89831865492394e-05,
+      "loss": 0.5911,
+      "step": 4097
+    },
+    {
+      "epoch": 1.31136,
+      "grad_norm": 0.7025341391563416,
+      "learning_rate": 6.8951160928743e-05,
+      "loss": 0.6529,
+      "step": 4098
+    },
+    {
+      "epoch": 1.31168,
+      "grad_norm": 0.6425396203994751,
+      "learning_rate": 6.891913530824659e-05,
+      "loss": 0.5337,
+      "step": 4099
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 0.6107209324836731,
+      "learning_rate": 6.888710968775021e-05,
+      "loss": 0.453,
+      "step": 4100
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 0.47300460934638977,
+      "learning_rate": 6.88550840672538e-05,
+      "loss": 0.4155,
+      "step": 4101
+    },
+    {
+      "epoch": 1.31264,
+      "grad_norm": 0.6449198126792908,
+      "learning_rate": 6.88230584467574e-05,
+      "loss": 0.4836,
+      "step": 4102
+    },
+    {
+      "epoch": 1.31296,
+      "grad_norm": 0.535331666469574,
+      "learning_rate": 6.879103282626101e-05,
+      "loss": 0.4973,
+      "step": 4103
+    },
+    {
+      "epoch": 1.31328,
+      "grad_norm": 0.5560728907585144,
+      "learning_rate": 6.875900720576461e-05,
+      "loss": 0.5014,
+      "step": 4104
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 0.6889582276344299,
+      "learning_rate": 6.872698158526822e-05,
+      "loss": 0.4798,
+      "step": 4105
+    },
+    {
+      "epoch": 1.31392,
+      "grad_norm": 0.5346435308456421,
+      "learning_rate": 6.869495596477182e-05,
+      "loss": 0.3462,
+      "step": 4106
+    },
+    {
+      "epoch": 1.31424,
+      "grad_norm": 0.6038232445716858,
+      "learning_rate": 6.866293034427543e-05,
+      "loss": 0.3664,
+      "step": 4107
+    },
+    {
+      "epoch": 1.31456,
+      "grad_norm": 0.6622744798660278,
+      "learning_rate": 6.863090472377902e-05,
+      "loss": 0.5438,
+      "step": 4108
+    },
+    {
+      "epoch": 1.31488,
+      "grad_norm": 0.49299511313438416,
+      "learning_rate": 6.859887910328264e-05,
+      "loss": 0.3772,
+      "step": 4109
+    },
+    {
+      "epoch": 1.3152,
+      "grad_norm": 0.5362630486488342,
+      "learning_rate": 6.856685348278623e-05,
+      "loss": 0.334,
+      "step": 4110
+    },
+    {
+      "epoch": 1.31552,
+      "grad_norm": 0.7626070380210876,
+      "learning_rate": 6.853482786228983e-05,
+      "loss": 0.5655,
+      "step": 4111
+    },
+    {
+      "epoch": 1.3158400000000001,
+      "grad_norm": 0.49453431367874146,
+      "learning_rate": 6.850280224179344e-05,
+      "loss": 0.4404,
+      "step": 4112
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 0.640474796295166,
+      "learning_rate": 6.847077662129704e-05,
+      "loss": 0.505,
+      "step": 4113
+    },
+    {
+      "epoch": 1.3164799999999999,
+      "grad_norm": 0.6205937266349792,
+      "learning_rate": 6.843875100080063e-05,
+      "loss": 0.5349,
+      "step": 4114
+    },
+    {
+      "epoch": 1.3168,
+      "grad_norm": 0.5410770177841187,
+      "learning_rate": 6.840672538030425e-05,
+      "loss": 0.4739,
+      "step": 4115
+    },
+    {
+      "epoch": 1.31712,
+      "grad_norm": 0.5919944643974304,
+      "learning_rate": 6.837469975980784e-05,
+      "loss": 0.6003,
+      "step": 4116
+    },
+    {
+      "epoch": 1.31744,
+      "grad_norm": 0.6065229177474976,
+      "learning_rate": 6.834267413931146e-05,
+      "loss": 0.4788,
+      "step": 4117
+    },
+    {
+      "epoch": 1.31776,
+      "grad_norm": 0.47100746631622314,
+      "learning_rate": 6.831064851881505e-05,
+      "loss": 0.5596,
+      "step": 4118
+    },
+    {
+      "epoch": 1.31808,
+      "grad_norm": 0.5469468235969543,
+      "learning_rate": 6.827862289831866e-05,
+      "loss": 0.4928,
+      "step": 4119
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.5822675228118896,
+      "learning_rate": 6.824659727782226e-05,
+      "loss": 0.4872,
+      "step": 4120
+    },
+    {
+      "epoch": 1.31872,
+      "grad_norm": 0.564139187335968,
+      "learning_rate": 6.821457165732587e-05,
+      "loss": 0.5295,
+      "step": 4121
+    },
+    {
+      "epoch": 1.31904,
+      "grad_norm": 0.7661532163619995,
+      "learning_rate": 6.818254603682946e-05,
+      "loss": 0.4716,
+      "step": 4122
+    },
+    {
+      "epoch": 1.31936,
+      "grad_norm": 0.6032697558403015,
+      "learning_rate": 6.815052041633308e-05,
+      "loss": 0.4544,
+      "step": 4123
+    },
+    {
+      "epoch": 1.31968,
+      "grad_norm": 0.6369198560714722,
+      "learning_rate": 6.811849479583667e-05,
+      "loss": 0.3957,
+      "step": 4124
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5371183753013611,
+      "learning_rate": 6.808646917534027e-05,
+      "loss": 0.4602,
+      "step": 4125
+    },
+    {
+      "epoch": 1.32032,
+      "grad_norm": 0.5470612645149231,
+      "learning_rate": 6.805444355484388e-05,
+      "loss": 0.4764,
+      "step": 4126
+    },
+    {
+      "epoch": 1.32064,
+      "grad_norm": 0.5488792657852173,
+      "learning_rate": 6.802241793434748e-05,
+      "loss": 0.4839,
+      "step": 4127
+    },
+    {
+      "epoch": 1.32096,
+      "grad_norm": 0.6182271242141724,
+      "learning_rate": 6.799039231385109e-05,
+      "loss": 0.3716,
+      "step": 4128
+    },
+    {
+      "epoch": 1.32128,
+      "grad_norm": 0.7525869011878967,
+      "learning_rate": 6.795836669335469e-05,
+      "loss": 0.6771,
+      "step": 4129
+    },
+    {
+      "epoch": 1.3216,
+      "grad_norm": 0.5964409112930298,
+      "learning_rate": 6.79263410728583e-05,
+      "loss": 0.569,
+      "step": 4130
+    },
+    {
+      "epoch": 1.32192,
+      "grad_norm": 0.522990882396698,
+      "learning_rate": 6.789431545236189e-05,
+      "loss": 0.4362,
+      "step": 4131
+    },
+    {
+      "epoch": 1.32224,
+      "grad_norm": 0.5760594606399536,
+      "learning_rate": 6.786228983186549e-05,
+      "loss": 0.4182,
+      "step": 4132
+    },
+    {
+      "epoch": 1.32256,
+      "grad_norm": 0.6751972436904907,
+      "learning_rate": 6.78302642113691e-05,
+      "loss": 0.4552,
+      "step": 4133
+    },
+    {
+      "epoch": 1.32288,
+      "grad_norm": 0.5204690098762512,
+      "learning_rate": 6.77982385908727e-05,
+      "loss": 0.4944,
+      "step": 4134
+    },
+    {
+      "epoch": 1.3232,
+      "grad_norm": 0.47359326481819153,
+      "learning_rate": 6.776621297037631e-05,
+      "loss": 0.4325,
+      "step": 4135
+    },
+    {
+      "epoch": 1.32352,
+      "grad_norm": 0.5132090449333191,
+      "learning_rate": 6.773418734987991e-05,
+      "loss": 0.4078,
+      "step": 4136
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 0.5574632287025452,
+      "learning_rate": 6.77021617293835e-05,
+      "loss": 0.5083,
+      "step": 4137
+    },
+    {
+      "epoch": 1.32416,
+      "grad_norm": 0.476420134305954,
+      "learning_rate": 6.767013610888712e-05,
+      "loss": 0.469,
+      "step": 4138
+    },
+    {
+      "epoch": 1.3244799999999999,
+      "grad_norm": 0.6520438194274902,
+      "learning_rate": 6.763811048839071e-05,
+      "loss": 0.5411,
+      "step": 4139
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 0.5777458548545837,
+      "learning_rate": 6.760608486789432e-05,
+      "loss": 0.3772,
+      "step": 4140
+    },
+    {
+      "epoch": 1.32512,
+      "grad_norm": 0.5628530979156494,
+      "learning_rate": 6.757405924739792e-05,
+      "loss": 0.4063,
+      "step": 4141
+    },
+    {
+      "epoch": 1.32544,
+      "grad_norm": 0.49444296956062317,
+      "learning_rate": 6.754203362690153e-05,
+      "loss": 0.424,
+      "step": 4142
+    },
+    {
+      "epoch": 1.32576,
+      "grad_norm": 0.533329427242279,
+      "learning_rate": 6.751000800640512e-05,
+      "loss": 0.4475,
+      "step": 4143
+    },
+    {
+      "epoch": 1.32608,
+      "grad_norm": 0.5675898790359497,
+      "learning_rate": 6.747798238590874e-05,
+      "loss": 0.6002,
+      "step": 4144
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 0.4656343162059784,
+      "learning_rate": 6.744595676541233e-05,
+      "loss": 0.4416,
+      "step": 4145
+    },
+    {
+      "epoch": 1.32672,
+      "grad_norm": 0.5873148441314697,
+      "learning_rate": 6.741393114491593e-05,
+      "loss": 0.457,
+      "step": 4146
+    },
+    {
+      "epoch": 1.32704,
+      "grad_norm": 0.7302036881446838,
+      "learning_rate": 6.738190552441954e-05,
+      "loss": 0.6921,
+      "step": 4147
+    },
+    {
+      "epoch": 1.32736,
+      "grad_norm": 0.6402762532234192,
+      "learning_rate": 6.734987990392314e-05,
+      "loss": 0.485,
+      "step": 4148
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 0.6144781112670898,
+      "learning_rate": 6.731785428342673e-05,
+      "loss": 0.4558,
+      "step": 4149
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 0.7063935995101929,
+      "learning_rate": 6.728582866293035e-05,
+      "loss": 0.5405,
+      "step": 4150
+    },
+    {
+      "epoch": 1.32832,
+      "grad_norm": 0.6169722676277161,
+      "learning_rate": 6.725380304243394e-05,
+      "loss": 0.565,
+      "step": 4151
+    },
+    {
+      "epoch": 1.32864,
+      "grad_norm": 0.506125807762146,
+      "learning_rate": 6.722177742193755e-05,
+      "loss": 0.5477,
+      "step": 4152
+    },
+    {
+      "epoch": 1.32896,
+      "grad_norm": 0.6336098313331604,
+      "learning_rate": 6.718975180144115e-05,
+      "loss": 0.5418,
+      "step": 4153
+    },
+    {
+      "epoch": 1.32928,
+      "grad_norm": 0.5425248146057129,
+      "learning_rate": 6.715772618094476e-05,
+      "loss": 0.4089,
+      "step": 4154
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 0.6005550026893616,
+      "learning_rate": 6.712570056044836e-05,
+      "loss": 0.5949,
+      "step": 4155
+    },
+    {
+      "epoch": 1.32992,
+      "grad_norm": 0.7321895956993103,
+      "learning_rate": 6.709367493995197e-05,
+      "loss": 0.6245,
+      "step": 4156
+    },
+    {
+      "epoch": 1.3302399999999999,
+      "grad_norm": 0.5213395357131958,
+      "learning_rate": 6.706164931945557e-05,
+      "loss": 0.4529,
+      "step": 4157
+    },
+    {
+      "epoch": 1.33056,
+      "grad_norm": 0.5035428404808044,
+      "learning_rate": 6.702962369895916e-05,
+      "loss": 0.4528,
+      "step": 4158
+    },
+    {
+      "epoch": 1.33088,
+      "grad_norm": 0.49765974283218384,
+      "learning_rate": 6.699759807846278e-05,
+      "loss": 0.3724,
+      "step": 4159
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.5992640256881714,
+      "learning_rate": 6.696557245796637e-05,
+      "loss": 0.4084,
+      "step": 4160
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 0.5742327570915222,
+      "learning_rate": 6.693354683746998e-05,
+      "loss": 0.5149,
+      "step": 4161
+    },
+    {
+      "epoch": 1.3318400000000001,
+      "grad_norm": 0.4939899742603302,
+      "learning_rate": 6.690152121697358e-05,
+      "loss": 0.4358,
+      "step": 4162
+    },
+    {
+      "epoch": 1.33216,
+      "grad_norm": 0.6196455359458923,
+      "learning_rate": 6.686949559647719e-05,
+      "loss": 0.5164,
+      "step": 4163
+    },
+    {
+      "epoch": 1.3324799999999999,
+      "grad_norm": 0.44913846254348755,
+      "learning_rate": 6.683746997598079e-05,
+      "loss": 0.3264,
+      "step": 4164
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 0.6040534973144531,
+      "learning_rate": 6.68054443554844e-05,
+      "loss": 0.4529,
+      "step": 4165
+    },
+    {
+      "epoch": 1.33312,
+      "grad_norm": 0.48971813917160034,
+      "learning_rate": 6.677341873498799e-05,
+      "loss": 0.3185,
+      "step": 4166
+    },
+    {
+      "epoch": 1.33344,
+      "grad_norm": 0.581632673740387,
+      "learning_rate": 6.67413931144916e-05,
+      "loss": 0.5081,
+      "step": 4167
+    },
+    {
+      "epoch": 1.33376,
+      "grad_norm": 0.6162259578704834,
+      "learning_rate": 6.67093674939952e-05,
+      "loss": 0.5784,
+      "step": 4168
+    },
+    {
+      "epoch": 1.33408,
+      "grad_norm": 0.5861248970031738,
+      "learning_rate": 6.66773418734988e-05,
+      "loss": 0.4703,
+      "step": 4169
+    },
+    {
+      "epoch": 1.3344,
+      "grad_norm": 0.6060977578163147,
+      "learning_rate": 6.66453162530024e-05,
+      "loss": 0.3918,
+      "step": 4170
+    },
+    {
+      "epoch": 1.33472,
+      "grad_norm": 0.4701457917690277,
+      "learning_rate": 6.661329063250601e-05,
+      "loss": 0.4547,
+      "step": 4171
+    },
+    {
+      "epoch": 1.33504,
+      "grad_norm": 0.6541212797164917,
+      "learning_rate": 6.65812650120096e-05,
+      "loss": 0.5621,
+      "step": 4172
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 0.7016064524650574,
+      "learning_rate": 6.654923939151322e-05,
+      "loss": 0.5555,
+      "step": 4173
+    },
+    {
+      "epoch": 1.33568,
+      "grad_norm": 0.5501878261566162,
+      "learning_rate": 6.651721377101681e-05,
+      "loss": 0.4852,
+      "step": 4174
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 0.5969823598861694,
+      "learning_rate": 6.648518815052042e-05,
+      "loss": 0.5502,
+      "step": 4175
+    },
+    {
+      "epoch": 1.33632,
+      "grad_norm": 0.7055111527442932,
+      "learning_rate": 6.645316253002402e-05,
+      "loss": 0.4995,
+      "step": 4176
+    },
+    {
+      "epoch": 1.33664,
+      "grad_norm": 0.4353658854961395,
+      "learning_rate": 6.642113690952763e-05,
+      "loss": 0.4197,
+      "step": 4177
+    },
+    {
+      "epoch": 1.33696,
+      "grad_norm": 0.5495374202728271,
+      "learning_rate": 6.638911128903123e-05,
+      "loss": 0.4817,
+      "step": 4178
+    },
+    {
+      "epoch": 1.33728,
+      "grad_norm": 0.4193899631500244,
+      "learning_rate": 6.635708566853484e-05,
+      "loss": 0.3264,
+      "step": 4179
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 0.371205598115921,
+      "learning_rate": 6.632506004803844e-05,
+      "loss": 0.3934,
+      "step": 4180
+    },
+    {
+      "epoch": 1.33792,
+      "grad_norm": 0.6910121440887451,
+      "learning_rate": 6.629303442754203e-05,
+      "loss": 0.536,
+      "step": 4181
+    },
+    {
+      "epoch": 1.3382399999999999,
+      "grad_norm": 0.5252282023429871,
+      "learning_rate": 6.626100880704565e-05,
+      "loss": 0.4552,
+      "step": 4182
+    },
+    {
+      "epoch": 1.33856,
+      "grad_norm": 0.5632966756820679,
+      "learning_rate": 6.622898318654924e-05,
+      "loss": 0.4467,
+      "step": 4183
+    },
+    {
+      "epoch": 1.33888,
+      "grad_norm": 0.5889667868614197,
+      "learning_rate": 6.619695756605285e-05,
+      "loss": 0.4537,
+      "step": 4184
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.6636489629745483,
+      "learning_rate": 6.616493194555645e-05,
+      "loss": 0.5951,
+      "step": 4185
+    },
+    {
+      "epoch": 1.33952,
+      "grad_norm": 0.5922411680221558,
+      "learning_rate": 6.613290632506006e-05,
+      "loss": 0.4391,
+      "step": 4186
+    },
+    {
+      "epoch": 1.33984,
+      "grad_norm": 0.5206584930419922,
+      "learning_rate": 6.610088070456365e-05,
+      "loss": 0.4102,
+      "step": 4187
+    },
+    {
+      "epoch": 1.34016,
+      "grad_norm": 0.5928571820259094,
+      "learning_rate": 6.606885508406726e-05,
+      "loss": 0.5342,
+      "step": 4188
+    },
+    {
+      "epoch": 1.34048,
+      "grad_norm": 0.5485615134239197,
+      "learning_rate": 6.603682946357086e-05,
+      "loss": 0.4717,
+      "step": 4189
+    },
+    {
+      "epoch": 1.3408,
+      "grad_norm": 0.5960674285888672,
+      "learning_rate": 6.600480384307446e-05,
+      "loss": 0.4981,
+      "step": 4190
+    },
+    {
+      "epoch": 1.34112,
+      "grad_norm": 0.5605925917625427,
+      "learning_rate": 6.597277822257807e-05,
+      "loss": 0.514,
+      "step": 4191
+    },
+    {
+      "epoch": 1.34144,
+      "grad_norm": 0.6209437251091003,
+      "learning_rate": 6.594075260208167e-05,
+      "loss": 0.5406,
+      "step": 4192
+    },
+    {
+      "epoch": 1.34176,
+      "grad_norm": 0.47275856137275696,
+      "learning_rate": 6.590872698158526e-05,
+      "loss": 0.3751,
+      "step": 4193
+    },
+    {
+      "epoch": 1.34208,
+      "grad_norm": 0.620380699634552,
+      "learning_rate": 6.587670136108888e-05,
+      "loss": 0.4284,
+      "step": 4194
+    },
+    {
+      "epoch": 1.3424,
+      "grad_norm": 0.4323877692222595,
+      "learning_rate": 6.584467574059247e-05,
+      "loss": 0.3163,
+      "step": 4195
+    },
+    {
+      "epoch": 1.34272,
+      "grad_norm": 0.588239848613739,
+      "learning_rate": 6.581265012009608e-05,
+      "loss": 0.4877,
+      "step": 4196
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 0.5935875177383423,
+      "learning_rate": 6.578062449959968e-05,
+      "loss": 0.4638,
+      "step": 4197
+    },
+    {
+      "epoch": 1.34336,
+      "grad_norm": 0.6226515769958496,
+      "learning_rate": 6.574859887910329e-05,
+      "loss": 0.4441,
+      "step": 4198
+    },
+    {
+      "epoch": 1.34368,
+      "grad_norm": 0.7643909454345703,
+      "learning_rate": 6.571657325860689e-05,
+      "loss": 0.4557,
+      "step": 4199
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.63917475938797,
+      "learning_rate": 6.56845476381105e-05,
+      "loss": 0.472,
+      "step": 4200
+    },
+    {
+      "epoch": 1.34432,
+      "grad_norm": 0.6220800876617432,
+      "learning_rate": 6.56525220176141e-05,
+      "loss": 0.6508,
+      "step": 4201
+    },
+    {
+      "epoch": 1.34464,
+      "grad_norm": 0.49646785855293274,
+      "learning_rate": 6.562049639711769e-05,
+      "loss": 0.397,
+      "step": 4202
+    },
+    {
+      "epoch": 1.34496,
+      "grad_norm": 0.6877707242965698,
+      "learning_rate": 6.55884707766213e-05,
+      "loss": 0.5563,
+      "step": 4203
+    },
+    {
+      "epoch": 1.34528,
+      "grad_norm": 0.44284695386886597,
+      "learning_rate": 6.55564451561249e-05,
+      "loss": 0.3144,
+      "step": 4204
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 0.43508145213127136,
+      "learning_rate": 6.55244195356285e-05,
+      "loss": 0.3547,
+      "step": 4205
+    },
+    {
+      "epoch": 1.34592,
+      "grad_norm": 0.7080327868461609,
+      "learning_rate": 6.549239391513211e-05,
+      "loss": 0.5613,
+      "step": 4206
+    },
+    {
+      "epoch": 1.3462399999999999,
+      "grad_norm": 0.4346381723880768,
+      "learning_rate": 6.546036829463571e-05,
+      "loss": 0.4228,
+      "step": 4207
+    },
+    {
+      "epoch": 1.34656,
+      "grad_norm": 0.7745829224586487,
+      "learning_rate": 6.54283426741393e-05,
+      "loss": 0.5221,
+      "step": 4208
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 0.4667518436908722,
+      "learning_rate": 6.539631705364292e-05,
+      "loss": 0.3811,
+      "step": 4209
+    },
+    {
+      "epoch": 1.3472,
+      "grad_norm": 0.607604444026947,
+      "learning_rate": 6.536429143314652e-05,
+      "loss": 0.4464,
+      "step": 4210
+    },
+    {
+      "epoch": 1.34752,
+      "grad_norm": 0.5866004824638367,
+      "learning_rate": 6.533226581265013e-05,
+      "loss": 0.4927,
+      "step": 4211
+    },
+    {
+      "epoch": 1.34784,
+      "grad_norm": 0.7750657796859741,
+      "learning_rate": 6.530024019215373e-05,
+      "loss": 0.4553,
+      "step": 4212
+    },
+    {
+      "epoch": 1.34816,
+      "grad_norm": 0.7511674761772156,
+      "learning_rate": 6.526821457165733e-05,
+      "loss": 0.514,
+      "step": 4213
+    },
+    {
+      "epoch": 1.34848,
+      "grad_norm": 0.4920574724674225,
+      "learning_rate": 6.523618895116093e-05,
+      "loss": 0.3625,
+      "step": 4214
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 0.5874229669570923,
+      "learning_rate": 6.520416333066454e-05,
+      "loss": 0.4824,
+      "step": 4215
+    },
+    {
+      "epoch": 1.34912,
+      "grad_norm": 0.5159461498260498,
+      "learning_rate": 6.517213771016813e-05,
+      "loss": 0.3605,
+      "step": 4216
+    },
+    {
+      "epoch": 1.34944,
+      "grad_norm": 0.6673134565353394,
+      "learning_rate": 6.514011208967175e-05,
+      "loss": 0.4767,
+      "step": 4217
+    },
+    {
+      "epoch": 1.34976,
+      "grad_norm": 0.4367545247077942,
+      "learning_rate": 6.510808646917534e-05,
+      "loss": 0.3662,
+      "step": 4218
+    },
+    {
+      "epoch": 1.35008,
+      "grad_norm": 0.5759279727935791,
+      "learning_rate": 6.507606084867894e-05,
+      "loss": 0.4347,
+      "step": 4219
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 0.6364127993583679,
+      "learning_rate": 6.504403522818255e-05,
+      "loss": 0.3755,
+      "step": 4220
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 0.8510307669639587,
+      "learning_rate": 6.501200960768615e-05,
+      "loss": 0.57,
+      "step": 4221
+    },
+    {
+      "epoch": 1.35104,
+      "grad_norm": 0.5080515146255493,
+      "learning_rate": 6.497998398718975e-05,
+      "loss": 0.4392,
+      "step": 4222
+    },
+    {
+      "epoch": 1.3513600000000001,
+      "grad_norm": 0.5852148532867432,
+      "learning_rate": 6.494795836669336e-05,
+      "loss": 0.3549,
+      "step": 4223
+    },
+    {
+      "epoch": 1.35168,
+      "grad_norm": 0.3534655272960663,
+      "learning_rate": 6.491593274619696e-05,
+      "loss": 0.328,
+      "step": 4224
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.7357264161109924,
+      "learning_rate": 6.488390712570056e-05,
+      "loss": 0.6498,
+      "step": 4225
+    },
+    {
+      "epoch": 1.35232,
+      "grad_norm": 0.6652783155441284,
+      "learning_rate": 6.485188150520416e-05,
+      "loss": 0.5197,
+      "step": 4226
+    },
+    {
+      "epoch": 1.35264,
+      "grad_norm": 0.6779317855834961,
+      "learning_rate": 6.481985588470777e-05,
+      "loss": 0.4405,
+      "step": 4227
+    },
+    {
+      "epoch": 1.35296,
+      "grad_norm": 0.5549767017364502,
+      "learning_rate": 6.478783026421137e-05,
+      "loss": 0.4266,
+      "step": 4228
+    },
+    {
+      "epoch": 1.35328,
+      "grad_norm": 0.6012527346611023,
+      "learning_rate": 6.475580464371498e-05,
+      "loss": 0.4514,
+      "step": 4229
+    },
+    {
+      "epoch": 1.3536000000000001,
+      "grad_norm": 0.5859169960021973,
+      "learning_rate": 6.472377902321858e-05,
+      "loss": 0.5092,
+      "step": 4230
+    },
+    {
+      "epoch": 1.35392,
+      "grad_norm": 0.5394780039787292,
+      "learning_rate": 6.469175340272217e-05,
+      "loss": 0.4346,
+      "step": 4231
+    },
+    {
+      "epoch": 1.3542399999999999,
+      "grad_norm": 0.49105900526046753,
+      "learning_rate": 6.465972778222579e-05,
+      "loss": 0.3965,
+      "step": 4232
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 0.6491592526435852,
+      "learning_rate": 6.462770216172938e-05,
+      "loss": 0.4969,
+      "step": 4233
+    },
+    {
+      "epoch": 1.35488,
+      "grad_norm": 0.5154183506965637,
+      "learning_rate": 6.459567654123299e-05,
+      "loss": 0.3994,
+      "step": 4234
+    },
+    {
+      "epoch": 1.3552,
+      "grad_norm": 0.5564209222793579,
+      "learning_rate": 6.45636509207366e-05,
+      "loss": 0.4652,
+      "step": 4235
+    },
+    {
+      "epoch": 1.35552,
+      "grad_norm": 0.5094367861747742,
+      "learning_rate": 6.45316253002402e-05,
+      "loss": 0.507,
+      "step": 4236
+    },
+    {
+      "epoch": 1.35584,
+      "grad_norm": 0.7262341976165771,
+      "learning_rate": 6.449959967974379e-05,
+      "loss": 0.4419,
+      "step": 4237
+    },
+    {
+      "epoch": 1.35616,
+      "grad_norm": 0.6106246113777161,
+      "learning_rate": 6.446757405924741e-05,
+      "loss": 0.4173,
+      "step": 4238
+    },
+    {
+      "epoch": 1.35648,
+      "grad_norm": 0.5958698987960815,
+      "learning_rate": 6.4435548438751e-05,
+      "loss": 0.3831,
+      "step": 4239
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.6541719436645508,
+      "learning_rate": 6.44035228182546e-05,
+      "loss": 0.4641,
+      "step": 4240
+    },
+    {
+      "epoch": 1.35712,
+      "grad_norm": 0.614657461643219,
+      "learning_rate": 6.437149719775821e-05,
+      "loss": 0.5447,
+      "step": 4241
+    },
+    {
+      "epoch": 1.35744,
+      "grad_norm": 0.5658915638923645,
+      "learning_rate": 6.433947157726181e-05,
+      "loss": 0.6115,
+      "step": 4242
+    },
+    {
+      "epoch": 1.35776,
+      "grad_norm": 0.6352320909500122,
+      "learning_rate": 6.43074459567654e-05,
+      "loss": 0.5882,
+      "step": 4243
+    },
+    {
+      "epoch": 1.35808,
+      "grad_norm": 0.7996504902839661,
+      "learning_rate": 6.427542033626902e-05,
+      "loss": 0.5578,
+      "step": 4244
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.5849382877349854,
+      "learning_rate": 6.424339471577261e-05,
+      "loss": 0.5425,
+      "step": 4245
+    },
+    {
+      "epoch": 1.35872,
+      "grad_norm": 0.5315607786178589,
+      "learning_rate": 6.421136909527622e-05,
+      "loss": 0.4348,
+      "step": 4246
+    },
+    {
+      "epoch": 1.35904,
+      "grad_norm": 0.6979513168334961,
+      "learning_rate": 6.417934347477982e-05,
+      "loss": 0.6066,
+      "step": 4247
+    },
+    {
+      "epoch": 1.3593600000000001,
+      "grad_norm": 0.557566225528717,
+      "learning_rate": 6.414731785428343e-05,
+      "loss": 0.3961,
+      "step": 4248
+    },
+    {
+      "epoch": 1.35968,
+      "grad_norm": 0.5866889953613281,
+      "learning_rate": 6.411529223378703e-05,
+      "loss": 0.5524,
+      "step": 4249
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.470167338848114,
+      "learning_rate": 6.408326661329064e-05,
+      "loss": 0.3153,
+      "step": 4250
+    },
+    {
+      "epoch": 1.36032,
+      "grad_norm": 0.568352222442627,
+      "learning_rate": 6.405124099279424e-05,
+      "loss": 0.4893,
+      "step": 4251
+    },
+    {
+      "epoch": 1.36064,
+      "grad_norm": 0.6226931214332581,
+      "learning_rate": 6.401921537229783e-05,
+      "loss": 0.4697,
+      "step": 4252
+    },
+    {
+      "epoch": 1.36096,
+      "grad_norm": 0.5779409408569336,
+      "learning_rate": 6.398718975180145e-05,
+      "loss": 0.3839,
+      "step": 4253
+    },
+    {
+      "epoch": 1.36128,
+      "grad_norm": 0.5880792737007141,
+      "learning_rate": 6.395516413130504e-05,
+      "loss": 0.5227,
+      "step": 4254
+    },
+    {
+      "epoch": 1.3616,
+      "grad_norm": 0.5088031888008118,
+      "learning_rate": 6.392313851080865e-05,
+      "loss": 0.5328,
+      "step": 4255
+    },
+    {
+      "epoch": 1.36192,
+      "grad_norm": 0.6202327609062195,
+      "learning_rate": 6.389111289031225e-05,
+      "loss": 0.5217,
+      "step": 4256
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 0.6671051979064941,
+      "learning_rate": 6.385908726981586e-05,
+      "loss": 0.5034,
+      "step": 4257
+    },
+    {
+      "epoch": 1.36256,
+      "grad_norm": 0.4448338449001312,
+      "learning_rate": 6.382706164931946e-05,
+      "loss": 0.2997,
+      "step": 4258
+    },
+    {
+      "epoch": 1.36288,
+      "grad_norm": 0.5557465553283691,
+      "learning_rate": 6.379503602882307e-05,
+      "loss": 0.3477,
+      "step": 4259
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 0.6957972645759583,
+      "learning_rate": 6.376301040832666e-05,
+      "loss": 0.5765,
+      "step": 4260
+    },
+    {
+      "epoch": 1.36352,
+      "grad_norm": 0.682512104511261,
+      "learning_rate": 6.373098478783028e-05,
+      "loss": 0.5121,
+      "step": 4261
+    },
+    {
+      "epoch": 1.36384,
+      "grad_norm": 0.5491620302200317,
+      "learning_rate": 6.369895916733387e-05,
+      "loss": 0.4833,
+      "step": 4262
+    },
+    {
+      "epoch": 1.36416,
+      "grad_norm": 0.46381431818008423,
+      "learning_rate": 6.366693354683747e-05,
+      "loss": 0.5152,
+      "step": 4263
+    },
+    {
+      "epoch": 1.36448,
+      "grad_norm": 0.6158157587051392,
+      "learning_rate": 6.363490792634108e-05,
+      "loss": 0.599,
+      "step": 4264
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 0.545048177242279,
+      "learning_rate": 6.360288230584468e-05,
+      "loss": 0.6447,
+      "step": 4265
+    },
+    {
+      "epoch": 1.3651200000000001,
+      "grad_norm": 0.6328414678573608,
+      "learning_rate": 6.357085668534827e-05,
+      "loss": 0.586,
+      "step": 4266
+    },
+    {
+      "epoch": 1.36544,
+      "grad_norm": 0.5189645886421204,
+      "learning_rate": 6.353883106485189e-05,
+      "loss": 0.3734,
+      "step": 4267
+    },
+    {
+      "epoch": 1.3657599999999999,
+      "grad_norm": 0.6431612968444824,
+      "learning_rate": 6.350680544435548e-05,
+      "loss": 0.4924,
+      "step": 4268
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 0.6202466487884521,
+      "learning_rate": 6.347477982385909e-05,
+      "loss": 0.5545,
+      "step": 4269
+    },
+    {
+      "epoch": 1.3664,
+      "grad_norm": 0.47283607721328735,
+      "learning_rate": 6.344275420336269e-05,
+      "loss": 0.4067,
+      "step": 4270
+    },
+    {
+      "epoch": 1.36672,
+      "grad_norm": 0.5278987884521484,
+      "learning_rate": 6.34107285828663e-05,
+      "loss": 0.3852,
+      "step": 4271
+    },
+    {
+      "epoch": 1.36704,
+      "grad_norm": 0.6135386228561401,
+      "learning_rate": 6.33787029623699e-05,
+      "loss": 0.5494,
+      "step": 4272
+    },
+    {
+      "epoch": 1.3673600000000001,
+      "grad_norm": 0.5948505997657776,
+      "learning_rate": 6.334667734187351e-05,
+      "loss": 0.5114,
+      "step": 4273
+    },
+    {
+      "epoch": 1.36768,
+      "grad_norm": 0.5710611343383789,
+      "learning_rate": 6.33146517213771e-05,
+      "loss": 0.4222,
+      "step": 4274
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.617725133895874,
+      "learning_rate": 6.32826261008807e-05,
+      "loss": 0.3847,
+      "step": 4275
+    },
+    {
+      "epoch": 1.36832,
+      "grad_norm": 0.6604677438735962,
+      "learning_rate": 6.325060048038431e-05,
+      "loss": 0.5048,
+      "step": 4276
+    },
+    {
+      "epoch": 1.36864,
+      "grad_norm": 0.6472762823104858,
+      "learning_rate": 6.321857485988791e-05,
+      "loss": 0.4452,
+      "step": 4277
+    },
+    {
+      "epoch": 1.36896,
+      "grad_norm": 0.6298906207084656,
+      "learning_rate": 6.318654923939152e-05,
+      "loss": 0.4796,
+      "step": 4278
+    },
+    {
+      "epoch": 1.36928,
+      "grad_norm": 0.6809854507446289,
+      "learning_rate": 6.315452361889512e-05,
+      "loss": 0.5075,
+      "step": 4279
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.7926902174949646,
+      "learning_rate": 6.312249799839873e-05,
+      "loss": 0.6054,
+      "step": 4280
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 0.5073546767234802,
+      "learning_rate": 6.309047237790232e-05,
+      "loss": 0.4127,
+      "step": 4281
+    },
+    {
+      "epoch": 1.37024,
+      "grad_norm": 0.6939238905906677,
+      "learning_rate": 6.305844675740594e-05,
+      "loss": 0.4719,
+      "step": 4282
+    },
+    {
+      "epoch": 1.37056,
+      "grad_norm": 0.5936386585235596,
+      "learning_rate": 6.302642113690953e-05,
+      "loss": 0.4826,
+      "step": 4283
+    },
+    {
+      "epoch": 1.37088,
+      "grad_norm": 0.58245450258255,
+      "learning_rate": 6.299439551641313e-05,
+      "loss": 0.4515,
+      "step": 4284
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 0.7509681582450867,
+      "learning_rate": 6.296236989591674e-05,
+      "loss": 0.5975,
+      "step": 4285
+    },
+    {
+      "epoch": 1.37152,
+      "grad_norm": 0.6199663281440735,
+      "learning_rate": 6.293034427542034e-05,
+      "loss": 0.5117,
+      "step": 4286
+    },
+    {
+      "epoch": 1.37184,
+      "grad_norm": 0.5171641707420349,
+      "learning_rate": 6.289831865492393e-05,
+      "loss": 0.4319,
+      "step": 4287
+    },
+    {
+      "epoch": 1.37216,
+      "grad_norm": 0.5066952109336853,
+      "learning_rate": 6.286629303442755e-05,
+      "loss": 0.4349,
+      "step": 4288
+    },
+    {
+      "epoch": 1.37248,
+      "grad_norm": 0.4787362217903137,
+      "learning_rate": 6.283426741393114e-05,
+      "loss": 0.3558,
+      "step": 4289
+    },
+    {
+      "epoch": 1.3728,
+      "grad_norm": 0.5149248242378235,
+      "learning_rate": 6.280224179343475e-05,
+      "loss": 0.488,
+      "step": 4290
+    },
+    {
+      "epoch": 1.3731200000000001,
+      "grad_norm": 0.6201411485671997,
+      "learning_rate": 6.277021617293835e-05,
+      "loss": 0.4653,
+      "step": 4291
+    },
+    {
+      "epoch": 1.37344,
+      "grad_norm": 0.5345122814178467,
+      "learning_rate": 6.273819055244196e-05,
+      "loss": 0.4996,
+      "step": 4292
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 0.6582993865013123,
+      "learning_rate": 6.270616493194555e-05,
+      "loss": 0.5231,
+      "step": 4293
+    },
+    {
+      "epoch": 1.37408,
+      "grad_norm": 0.5686315298080444,
+      "learning_rate": 6.267413931144917e-05,
+      "loss": 0.3601,
+      "step": 4294
+    },
+    {
+      "epoch": 1.3744,
+      "grad_norm": 0.5680921077728271,
+      "learning_rate": 6.264211369095276e-05,
+      "loss": 0.5187,
+      "step": 4295
+    },
+    {
+      "epoch": 1.37472,
+      "grad_norm": 0.5748109221458435,
+      "learning_rate": 6.261008807045636e-05,
+      "loss": 0.4481,
+      "step": 4296
+    },
+    {
+      "epoch": 1.37504,
+      "grad_norm": 0.39371457695961,
+      "learning_rate": 6.257806244995997e-05,
+      "loss": 0.3127,
+      "step": 4297
+    },
+    {
+      "epoch": 1.3753600000000001,
+      "grad_norm": 0.5045334100723267,
+      "learning_rate": 6.254603682946357e-05,
+      "loss": 0.4241,
+      "step": 4298
+    },
+    {
+      "epoch": 1.37568,
+      "grad_norm": 0.5778636932373047,
+      "learning_rate": 6.251401120896718e-05,
+      "loss": 0.4958,
+      "step": 4299
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 0.7636088728904724,
+      "learning_rate": 6.248198558847078e-05,
+      "loss": 0.674,
+      "step": 4300
+    },
+    {
+      "epoch": 1.37632,
+      "grad_norm": 0.6443363428115845,
+      "learning_rate": 6.244995996797439e-05,
+      "loss": 0.5272,
+      "step": 4301
+    },
+    {
+      "epoch": 1.37664,
+      "grad_norm": 0.6160416603088379,
+      "learning_rate": 6.241793434747798e-05,
+      "loss": 0.4324,
+      "step": 4302
+    },
+    {
+      "epoch": 1.37696,
+      "grad_norm": 0.6208785772323608,
+      "learning_rate": 6.23859087269816e-05,
+      "loss": 0.4703,
+      "step": 4303
+    },
+    {
+      "epoch": 1.37728,
+      "grad_norm": 0.5367745161056519,
+      "learning_rate": 6.235388310648519e-05,
+      "loss": 0.4327,
+      "step": 4304
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.5434665679931641,
+      "learning_rate": 6.23218574859888e-05,
+      "loss": 0.3625,
+      "step": 4305
+    },
+    {
+      "epoch": 1.37792,
+      "grad_norm": 0.6903082132339478,
+      "learning_rate": 6.22898318654924e-05,
+      "loss": 0.563,
+      "step": 4306
+    },
+    {
+      "epoch": 1.37824,
+      "grad_norm": 0.5592443346977234,
+      "learning_rate": 6.2257806244996e-05,
+      "loss": 0.4944,
+      "step": 4307
+    },
+    {
+      "epoch": 1.37856,
+      "grad_norm": 0.4864504337310791,
+      "learning_rate": 6.22257806244996e-05,
+      "loss": 0.4683,
+      "step": 4308
+    },
+    {
+      "epoch": 1.37888,
+      "grad_norm": 0.6063712239265442,
+      "learning_rate": 6.219375500400321e-05,
+      "loss": 0.4789,
+      "step": 4309
+    },
+    {
+      "epoch": 1.3792,
+      "grad_norm": 0.6612493991851807,
+      "learning_rate": 6.21617293835068e-05,
+      "loss": 0.482,
+      "step": 4310
+    },
+    {
+      "epoch": 1.37952,
+      "grad_norm": 0.6121131181716919,
+      "learning_rate": 6.212970376301042e-05,
+      "loss": 0.4595,
+      "step": 4311
+    },
+    {
+      "epoch": 1.37984,
+      "grad_norm": 0.6313183903694153,
+      "learning_rate": 6.209767814251401e-05,
+      "loss": 0.4382,
+      "step": 4312
+    },
+    {
+      "epoch": 1.38016,
+      "grad_norm": 0.6658868789672852,
+      "learning_rate": 6.206565252201762e-05,
+      "loss": 0.4613,
+      "step": 4313
+    },
+    {
+      "epoch": 1.38048,
+      "grad_norm": 0.5198780298233032,
+      "learning_rate": 6.203362690152122e-05,
+      "loss": 0.3986,
+      "step": 4314
+    },
+    {
+      "epoch": 1.3808,
+      "grad_norm": 0.5781584978103638,
+      "learning_rate": 6.200160128102483e-05,
+      "loss": 0.4251,
+      "step": 4315
+    },
+    {
+      "epoch": 1.3811200000000001,
+      "grad_norm": 0.6319796442985535,
+      "learning_rate": 6.196957566052842e-05,
+      "loss": 0.5108,
+      "step": 4316
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 0.6411386132240295,
+      "learning_rate": 6.193755004003204e-05,
+      "loss": 0.4601,
+      "step": 4317
+    },
+    {
+      "epoch": 1.3817599999999999,
+      "grad_norm": 0.7092155814170837,
+      "learning_rate": 6.190552441953563e-05,
+      "loss": 0.6924,
+      "step": 4318
+    },
+    {
+      "epoch": 1.38208,
+      "grad_norm": 0.656281054019928,
+      "learning_rate": 6.187349879903923e-05,
+      "loss": 0.4843,
+      "step": 4319
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.6296519041061401,
+      "learning_rate": 6.184147317854284e-05,
+      "loss": 0.4152,
+      "step": 4320
+    },
+    {
+      "epoch": 1.38272,
+      "grad_norm": 0.4845150113105774,
+      "learning_rate": 6.180944755804644e-05,
+      "loss": 0.4509,
+      "step": 4321
+    },
+    {
+      "epoch": 1.38304,
+      "grad_norm": 0.5086080431938171,
+      "learning_rate": 6.177742193755005e-05,
+      "loss": 0.42,
+      "step": 4322
+    },
+    {
+      "epoch": 1.38336,
+      "grad_norm": 0.5649260878562927,
+      "learning_rate": 6.174539631705365e-05,
+      "loss": 0.3605,
+      "step": 4323
+    },
+    {
+      "epoch": 1.38368,
+      "grad_norm": 0.4789144694805145,
+      "learning_rate": 6.171337069655726e-05,
+      "loss": 0.4076,
+      "step": 4324
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 0.5432643294334412,
+      "learning_rate": 6.168134507606085e-05,
+      "loss": 0.4278,
+      "step": 4325
+    },
+    {
+      "epoch": 1.38432,
+      "grad_norm": 0.5996572375297546,
+      "learning_rate": 6.164931945556446e-05,
+      "loss": 0.4882,
+      "step": 4326
+    },
+    {
+      "epoch": 1.38464,
+      "grad_norm": 0.6800515651702881,
+      "learning_rate": 6.161729383506806e-05,
+      "loss": 0.4215,
+      "step": 4327
+    },
+    {
+      "epoch": 1.38496,
+      "grad_norm": 0.5945009589195251,
+      "learning_rate": 6.158526821457166e-05,
+      "loss": 0.4899,
+      "step": 4328
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 0.5767536163330078,
+      "learning_rate": 6.155324259407527e-05,
+      "loss": 0.4433,
+      "step": 4329
+    },
+    {
+      "epoch": 1.3856,
+      "grad_norm": 0.5418493747711182,
+      "learning_rate": 6.152121697357887e-05,
+      "loss": 0.4557,
+      "step": 4330
+    },
+    {
+      "epoch": 1.38592,
+      "grad_norm": 0.44668111205101013,
+      "learning_rate": 6.148919135308246e-05,
+      "loss": 0.4229,
+      "step": 4331
+    },
+    {
+      "epoch": 1.38624,
+      "grad_norm": 0.5222465395927429,
+      "learning_rate": 6.145716573258608e-05,
+      "loss": 0.4758,
+      "step": 4332
+    },
+    {
+      "epoch": 1.38656,
+      "grad_norm": 0.4250122606754303,
+      "learning_rate": 6.142514011208967e-05,
+      "loss": 0.4484,
+      "step": 4333
+    },
+    {
+      "epoch": 1.3868800000000001,
+      "grad_norm": 0.5456657409667969,
+      "learning_rate": 6.139311449159328e-05,
+      "loss": 0.4096,
+      "step": 4334
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 0.6357489824295044,
+      "learning_rate": 6.136108887109688e-05,
+      "loss": 0.58,
+      "step": 4335
+    },
+    {
+      "epoch": 1.3875199999999999,
+      "grad_norm": 0.6107792258262634,
+      "learning_rate": 6.132906325060049e-05,
+      "loss": 0.3775,
+      "step": 4336
+    },
+    {
+      "epoch": 1.38784,
+      "grad_norm": 0.6473737359046936,
+      "learning_rate": 6.129703763010408e-05,
+      "loss": 0.3679,
+      "step": 4337
+    },
+    {
+      "epoch": 1.38816,
+      "grad_norm": 0.5850574374198914,
+      "learning_rate": 6.12650120096077e-05,
+      "loss": 0.4921,
+      "step": 4338
+    },
+    {
+      "epoch": 1.38848,
+      "grad_norm": 0.5659459233283997,
+      "learning_rate": 6.123298638911129e-05,
+      "loss": 0.5748,
+      "step": 4339
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 0.5972387790679932,
+      "learning_rate": 6.120096076861489e-05,
+      "loss": 0.4941,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 0.652532696723938,
+      "learning_rate": 6.11689351481185e-05,
+      "loss": 0.5155,
+      "step": 4341
+    },
+    {
+      "epoch": 1.38944,
+      "grad_norm": 0.6232966780662537,
+      "learning_rate": 6.11369095276221e-05,
+      "loss": 0.4648,
+      "step": 4342
+    },
+    {
+      "epoch": 1.3897599999999999,
+      "grad_norm": 0.520963191986084,
+      "learning_rate": 6.11048839071257e-05,
+      "loss": 0.3833,
+      "step": 4343
+    },
+    {
+      "epoch": 1.39008,
+      "grad_norm": 0.5985981822013855,
+      "learning_rate": 6.107285828662931e-05,
+      "loss": 0.3431,
+      "step": 4344
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 0.490218847990036,
+      "learning_rate": 6.104083266613291e-05,
+      "loss": 0.3523,
+      "step": 4345
+    },
+    {
+      "epoch": 1.39072,
+      "grad_norm": 0.5149635076522827,
+      "learning_rate": 6.1008807045636506e-05,
+      "loss": 0.461,
+      "step": 4346
+    },
+    {
+      "epoch": 1.39104,
+      "grad_norm": 0.6838841438293457,
+      "learning_rate": 6.097678142514012e-05,
+      "loss": 0.4291,
+      "step": 4347
+    },
+    {
+      "epoch": 1.39136,
+      "grad_norm": 0.6817654967308044,
+      "learning_rate": 6.0944755804643715e-05,
+      "loss": 0.4803,
+      "step": 4348
+    },
+    {
+      "epoch": 1.39168,
+      "grad_norm": 0.5479546785354614,
+      "learning_rate": 6.091273018414731e-05,
+      "loss": 0.3531,
+      "step": 4349
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 0.6924534440040588,
+      "learning_rate": 6.0880704563650925e-05,
+      "loss": 0.3949,
+      "step": 4350
+    },
+    {
+      "epoch": 1.39232,
+      "grad_norm": 0.5422678589820862,
+      "learning_rate": 6.084867894315452e-05,
+      "loss": 0.3247,
+      "step": 4351
+    },
+    {
+      "epoch": 1.39264,
+      "grad_norm": 0.5831544995307922,
+      "learning_rate": 6.0816653322658134e-05,
+      "loss": 0.4456,
+      "step": 4352
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 0.5816759467124939,
+      "learning_rate": 6.078462770216173e-05,
+      "loss": 0.3698,
+      "step": 4353
+    },
+    {
+      "epoch": 1.39328,
+      "grad_norm": 0.6610085964202881,
+      "learning_rate": 6.075260208166533e-05,
+      "loss": 0.4912,
+      "step": 4354
+    },
+    {
+      "epoch": 1.3936,
+      "grad_norm": 0.5608308911323547,
+      "learning_rate": 6.072057646116894e-05,
+      "loss": 0.4072,
+      "step": 4355
+    },
+    {
+      "epoch": 1.39392,
+      "grad_norm": 0.666467010974884,
+      "learning_rate": 6.068855084067254e-05,
+      "loss": 0.5034,
+      "step": 4356
+    },
+    {
+      "epoch": 1.39424,
+      "grad_norm": 0.6449629068374634,
+      "learning_rate": 6.0656525220176145e-05,
+      "loss": 0.5168,
+      "step": 4357
+    },
+    {
+      "epoch": 1.39456,
+      "grad_norm": 0.7970860004425049,
+      "learning_rate": 6.062449959967975e-05,
+      "loss": 0.5176,
+      "step": 4358
+    },
+    {
+      "epoch": 1.3948800000000001,
+      "grad_norm": 0.628826916217804,
+      "learning_rate": 6.059247397918335e-05,
+      "loss": 0.4861,
+      "step": 4359
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.6630723476409912,
+      "learning_rate": 6.056044835868695e-05,
+      "loss": 0.4885,
+      "step": 4360
+    },
+    {
+      "epoch": 1.3955199999999999,
+      "grad_norm": 0.6067970395088196,
+      "learning_rate": 6.052842273819056e-05,
+      "loss": 0.4487,
+      "step": 4361
+    },
+    {
+      "epoch": 1.39584,
+      "grad_norm": 0.7208156585693359,
+      "learning_rate": 6.049639711769416e-05,
+      "loss": 0.4918,
+      "step": 4362
+    },
+    {
+      "epoch": 1.39616,
+      "grad_norm": 0.499834805727005,
+      "learning_rate": 6.046437149719776e-05,
+      "loss": 0.4,
+      "step": 4363
+    },
+    {
+      "epoch": 1.39648,
+      "grad_norm": 0.5993565320968628,
+      "learning_rate": 6.043234587670137e-05,
+      "loss": 0.6285,
+      "step": 4364
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.7196264863014221,
+      "learning_rate": 6.040032025620497e-05,
+      "loss": 0.4234,
+      "step": 4365
+    },
+    {
+      "epoch": 1.39712,
+      "grad_norm": 0.5306333899497986,
+      "learning_rate": 6.036829463570857e-05,
+      "loss": 0.4362,
+      "step": 4366
+    },
+    {
+      "epoch": 1.39744,
+      "grad_norm": 0.6958664059638977,
+      "learning_rate": 6.033626901521218e-05,
+      "loss": 0.5449,
+      "step": 4367
+    },
+    {
+      "epoch": 1.39776,
+      "grad_norm": 0.5227981209754944,
+      "learning_rate": 6.0304243394715777e-05,
+      "loss": 0.33,
+      "step": 4368
+    },
+    {
+      "epoch": 1.39808,
+      "grad_norm": 0.5341054201126099,
+      "learning_rate": 6.0272217774219375e-05,
+      "loss": 0.3646,
+      "step": 4369
+    },
+    {
+      "epoch": 1.3984,
+      "grad_norm": 0.6668089628219604,
+      "learning_rate": 6.0240192153722986e-05,
+      "loss": 0.5049,
+      "step": 4370
+    },
+    {
+      "epoch": 1.39872,
+      "grad_norm": 0.43527287244796753,
+      "learning_rate": 6.0208166533226584e-05,
+      "loss": 0.3753,
+      "step": 4371
+    },
+    {
+      "epoch": 1.39904,
+      "grad_norm": 0.6120976209640503,
+      "learning_rate": 6.017614091273018e-05,
+      "loss": 0.4342,
+      "step": 4372
+    },
+    {
+      "epoch": 1.39936,
+      "grad_norm": 0.6567062139511108,
+      "learning_rate": 6.0144115292233794e-05,
+      "loss": 0.5439,
+      "step": 4373
+    },
+    {
+      "epoch": 1.39968,
+      "grad_norm": 0.5744851231575012,
+      "learning_rate": 6.011208967173739e-05,
+      "loss": 0.3842,
+      "step": 4374
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5018812417984009,
+      "learning_rate": 6.008006405124099e-05,
+      "loss": 0.3665,
+      "step": 4375
+    },
+    {
+      "epoch": 1.40032,
+      "grad_norm": 0.5533809661865234,
+      "learning_rate": 6.00480384307446e-05,
+      "loss": 0.513,
+      "step": 4376
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 0.601315438747406,
+      "learning_rate": 6.00160128102482e-05,
+      "loss": 0.4048,
+      "step": 4377
+    },
+    {
+      "epoch": 1.40096,
+      "grad_norm": 0.7409510016441345,
+      "learning_rate": 5.99839871897518e-05,
+      "loss": 0.5394,
+      "step": 4378
+    },
+    {
+      "epoch": 1.40128,
+      "grad_norm": 0.6688363552093506,
+      "learning_rate": 5.995196156925541e-05,
+      "loss": 0.4923,
+      "step": 4379
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 0.737493634223938,
+      "learning_rate": 5.991993594875901e-05,
+      "loss": 0.5457,
+      "step": 4380
+    },
+    {
+      "epoch": 1.40192,
+      "grad_norm": 0.5680164694786072,
+      "learning_rate": 5.988791032826261e-05,
+      "loss": 0.3926,
+      "step": 4381
+    },
+    {
+      "epoch": 1.40224,
+      "grad_norm": 0.6135359406471252,
+      "learning_rate": 5.9855884707766216e-05,
+      "loss": 0.4696,
+      "step": 4382
+    },
+    {
+      "epoch": 1.40256,
+      "grad_norm": 0.5679519176483154,
+      "learning_rate": 5.982385908726982e-05,
+      "loss": 0.4794,
+      "step": 4383
+    },
+    {
+      "epoch": 1.4028800000000001,
+      "grad_norm": 0.6083953380584717,
+      "learning_rate": 5.979183346677342e-05,
+      "loss": 0.445,
+      "step": 4384
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 0.5582169890403748,
+      "learning_rate": 5.975980784627703e-05,
+      "loss": 0.4885,
+      "step": 4385
+    },
+    {
+      "epoch": 1.4035199999999999,
+      "grad_norm": 0.582603931427002,
+      "learning_rate": 5.972778222578063e-05,
+      "loss": 0.3729,
+      "step": 4386
+    },
+    {
+      "epoch": 1.40384,
+      "grad_norm": 0.4788751006126404,
+      "learning_rate": 5.9695756605284226e-05,
+      "loss": 0.3591,
+      "step": 4387
+    },
+    {
+      "epoch": 1.40416,
+      "grad_norm": 0.48596471548080444,
+      "learning_rate": 5.966373098478784e-05,
+      "loss": 0.4233,
+      "step": 4388
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 0.5770334005355835,
+      "learning_rate": 5.9631705364291436e-05,
+      "loss": 0.409,
+      "step": 4389
+    },
+    {
+      "epoch": 1.4048,
+      "grad_norm": 0.7798293828964233,
+      "learning_rate": 5.9599679743795034e-05,
+      "loss": 0.5914,
+      "step": 4390
+    },
+    {
+      "epoch": 1.40512,
+      "grad_norm": 0.7811009883880615,
+      "learning_rate": 5.9567654123298645e-05,
+      "loss": 0.516,
+      "step": 4391
+    },
+    {
+      "epoch": 1.40544,
+      "grad_norm": 0.5320774912834167,
+      "learning_rate": 5.9535628502802243e-05,
+      "loss": 0.5864,
+      "step": 4392
+    },
+    {
+      "epoch": 1.40576,
+      "grad_norm": 0.7281491756439209,
+      "learning_rate": 5.950360288230584e-05,
+      "loss": 0.5435,
+      "step": 4393
+    },
+    {
+      "epoch": 1.40608,
+      "grad_norm": 0.5826244950294495,
+      "learning_rate": 5.947157726180945e-05,
+      "loss": 0.5991,
+      "step": 4394
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 0.6399038434028625,
+      "learning_rate": 5.943955164131305e-05,
+      "loss": 0.4998,
+      "step": 4395
+    },
+    {
+      "epoch": 1.40672,
+      "grad_norm": 0.669920027256012,
+      "learning_rate": 5.940752602081665e-05,
+      "loss": 0.4334,
+      "step": 4396
+    },
+    {
+      "epoch": 1.40704,
+      "grad_norm": 0.5803078413009644,
+      "learning_rate": 5.937550040032026e-05,
+      "loss": 0.4293,
+      "step": 4397
+    },
+    {
+      "epoch": 1.40736,
+      "grad_norm": 0.5405992269515991,
+      "learning_rate": 5.934347477982386e-05,
+      "loss": 0.4837,
+      "step": 4398
+    },
+    {
+      "epoch": 1.40768,
+      "grad_norm": 0.4799081087112427,
+      "learning_rate": 5.931144915932747e-05,
+      "loss": 0.3707,
+      "step": 4399
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.5888561606407166,
+      "learning_rate": 5.927942353883107e-05,
+      "loss": 0.4224,
+      "step": 4400
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 0.537301242351532,
+      "learning_rate": 5.9247397918334666e-05,
+      "loss": 0.4304,
+      "step": 4401
+    },
+    {
+      "epoch": 1.4086400000000001,
+      "grad_norm": 0.6038801670074463,
+      "learning_rate": 5.921537229783828e-05,
+      "loss": 0.5311,
+      "step": 4402
+    },
+    {
+      "epoch": 1.40896,
+      "grad_norm": 0.6198928952217102,
+      "learning_rate": 5.9183346677341876e-05,
+      "loss": 0.4797,
+      "step": 4403
+    },
+    {
+      "epoch": 1.4092799999999999,
+      "grad_norm": 0.6444536447525024,
+      "learning_rate": 5.9151321056845474e-05,
+      "loss": 0.5518,
+      "step": 4404
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 0.6430035829544067,
+      "learning_rate": 5.9119295436349085e-05,
+      "loss": 0.5378,
+      "step": 4405
+    },
+    {
+      "epoch": 1.40992,
+      "grad_norm": 0.6329089999198914,
+      "learning_rate": 5.908726981585268e-05,
+      "loss": 0.5717,
+      "step": 4406
+    },
+    {
+      "epoch": 1.41024,
+      "grad_norm": 0.5763000249862671,
+      "learning_rate": 5.905524419535629e-05,
+      "loss": 0.3346,
+      "step": 4407
+    },
+    {
+      "epoch": 1.41056,
+      "grad_norm": 0.45008665323257446,
+      "learning_rate": 5.902321857485989e-05,
+      "loss": 0.3641,
+      "step": 4408
+    },
+    {
+      "epoch": 1.4108800000000001,
+      "grad_norm": 0.5143076181411743,
+      "learning_rate": 5.89911929543635e-05,
+      "loss": 0.47,
+      "step": 4409
+    },
+    {
+      "epoch": 1.4112,
+      "grad_norm": 0.6585938334465027,
+      "learning_rate": 5.8959167333867095e-05,
+      "loss": 0.4528,
+      "step": 4410
+    },
+    {
+      "epoch": 1.4115199999999999,
+      "grad_norm": 0.5737709403038025,
+      "learning_rate": 5.892714171337071e-05,
+      "loss": 0.4689,
+      "step": 4411
+    },
+    {
+      "epoch": 1.41184,
+      "grad_norm": 0.622248649597168,
+      "learning_rate": 5.8895116092874305e-05,
+      "loss": 0.452,
+      "step": 4412
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 0.5721738338470459,
+      "learning_rate": 5.88630904723779e-05,
+      "loss": 0.4325,
+      "step": 4413
+    },
+    {
+      "epoch": 1.41248,
+      "grad_norm": 0.6255063414573669,
+      "learning_rate": 5.8831064851881514e-05,
+      "loss": 0.4947,
+      "step": 4414
+    },
+    {
+      "epoch": 1.4128,
+      "grad_norm": 0.7641318440437317,
+      "learning_rate": 5.879903923138511e-05,
+      "loss": 0.5716,
+      "step": 4415
+    },
+    {
+      "epoch": 1.41312,
+      "grad_norm": 0.7587405443191528,
+      "learning_rate": 5.876701361088871e-05,
+      "loss": 0.4426,
+      "step": 4416
+    },
+    {
+      "epoch": 1.41344,
+      "grad_norm": 0.5103662610054016,
+      "learning_rate": 5.873498799039232e-05,
+      "loss": 0.4368,
+      "step": 4417
+    },
+    {
+      "epoch": 1.41376,
+      "grad_norm": 0.5309995412826538,
+      "learning_rate": 5.870296236989592e-05,
+      "loss": 0.3616,
+      "step": 4418
+    },
+    {
+      "epoch": 1.41408,
+      "grad_norm": 0.724609375,
+      "learning_rate": 5.867093674939952e-05,
+      "loss": 0.6171,
+      "step": 4419
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 0.5943878889083862,
+      "learning_rate": 5.863891112890313e-05,
+      "loss": 0.4979,
+      "step": 4420
+    },
+    {
+      "epoch": 1.41472,
+      "grad_norm": 0.6654344201087952,
+      "learning_rate": 5.860688550840673e-05,
+      "loss": 0.5365,
+      "step": 4421
+    },
+    {
+      "epoch": 1.41504,
+      "grad_norm": 0.6180540323257446,
+      "learning_rate": 5.8574859887910325e-05,
+      "loss": 0.4771,
+      "step": 4422
+    },
+    {
+      "epoch": 1.41536,
+      "grad_norm": 0.6325697898864746,
+      "learning_rate": 5.854283426741394e-05,
+      "loss": 0.4751,
+      "step": 4423
+    },
+    {
+      "epoch": 1.41568,
+      "grad_norm": 0.6093636155128479,
+      "learning_rate": 5.8510808646917535e-05,
+      "loss": 0.4219,
+      "step": 4424
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.6004346013069153,
+      "learning_rate": 5.847878302642113e-05,
+      "loss": 0.5,
+      "step": 4425
+    },
+    {
+      "epoch": 1.41632,
+      "grad_norm": 0.5888866186141968,
+      "learning_rate": 5.8446757405924744e-05,
+      "loss": 0.4487,
+      "step": 4426
+    },
+    {
+      "epoch": 1.4166400000000001,
+      "grad_norm": 0.48045670986175537,
+      "learning_rate": 5.841473178542834e-05,
+      "loss": 0.5125,
+      "step": 4427
+    },
+    {
+      "epoch": 1.41696,
+      "grad_norm": 0.45355987548828125,
+      "learning_rate": 5.838270616493195e-05,
+      "loss": 0.3804,
+      "step": 4428
+    },
+    {
+      "epoch": 1.4172799999999999,
+      "grad_norm": 0.48373493552207947,
+      "learning_rate": 5.835068054443555e-05,
+      "loss": 0.4225,
+      "step": 4429
+    },
+    {
+      "epoch": 1.4176,
+      "grad_norm": 0.5258024334907532,
+      "learning_rate": 5.831865492393916e-05,
+      "loss": 0.4737,
+      "step": 4430
+    },
+    {
+      "epoch": 1.41792,
+      "grad_norm": 0.6683601140975952,
+      "learning_rate": 5.8286629303442755e-05,
+      "loss": 0.4862,
+      "step": 4431
+    },
+    {
+      "epoch": 1.41824,
+      "grad_norm": 0.48072221875190735,
+      "learning_rate": 5.825460368294636e-05,
+      "loss": 0.4184,
+      "step": 4432
+    },
+    {
+      "epoch": 1.41856,
+      "grad_norm": 0.5251738429069519,
+      "learning_rate": 5.8222578062449964e-05,
+      "loss": 0.3407,
+      "step": 4433
+    },
+    {
+      "epoch": 1.41888,
+      "grad_norm": 0.6298186182975769,
+      "learning_rate": 5.819055244195356e-05,
+      "loss": 0.3441,
+      "step": 4434
+    },
+    {
+      "epoch": 1.4192,
+      "grad_norm": 0.6958298683166504,
+      "learning_rate": 5.8158526821457174e-05,
+      "loss": 0.4867,
+      "step": 4435
+    },
+    {
+      "epoch": 1.41952,
+      "grad_norm": 0.6056778430938721,
+      "learning_rate": 5.812650120096077e-05,
+      "loss": 0.4624,
+      "step": 4436
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 0.5757900476455688,
+      "learning_rate": 5.809447558046437e-05,
+      "loss": 0.4108,
+      "step": 4437
+    },
+    {
+      "epoch": 1.42016,
+      "grad_norm": 0.6501722931861877,
+      "learning_rate": 5.806244995996798e-05,
+      "loss": 0.4439,
+      "step": 4438
+    },
+    {
+      "epoch": 1.42048,
+      "grad_norm": 0.5730315446853638,
+      "learning_rate": 5.803042433947158e-05,
+      "loss": 0.4742,
+      "step": 4439
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.6266297101974487,
+      "learning_rate": 5.799839871897518e-05,
+      "loss": 0.4849,
+      "step": 4440
+    },
+    {
+      "epoch": 1.42112,
+      "grad_norm": 0.6129741072654724,
+      "learning_rate": 5.796637309847879e-05,
+      "loss": 0.5364,
+      "step": 4441
+    },
+    {
+      "epoch": 1.42144,
+      "grad_norm": 0.5090457797050476,
+      "learning_rate": 5.793434747798239e-05,
+      "loss": 0.361,
+      "step": 4442
+    },
+    {
+      "epoch": 1.42176,
+      "grad_norm": 0.830126941204071,
+      "learning_rate": 5.7902321857485985e-05,
+      "loss": 0.4203,
+      "step": 4443
+    },
+    {
+      "epoch": 1.42208,
+      "grad_norm": 0.6436464786529541,
+      "learning_rate": 5.7870296236989596e-05,
+      "loss": 0.4594,
+      "step": 4444
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 0.6578628420829773,
+      "learning_rate": 5.7838270616493194e-05,
+      "loss": 0.434,
+      "step": 4445
+    },
+    {
+      "epoch": 1.42272,
+      "grad_norm": 0.6424952745437622,
+      "learning_rate": 5.7806244995996806e-05,
+      "loss": 0.4485,
+      "step": 4446
+    },
+    {
+      "epoch": 1.42304,
+      "grad_norm": 0.6138772964477539,
+      "learning_rate": 5.7774219375500404e-05,
+      "loss": 0.3846,
+      "step": 4447
+    },
+    {
+      "epoch": 1.42336,
+      "grad_norm": 0.7532479166984558,
+      "learning_rate": 5.7742193755004e-05,
+      "loss": 0.3969,
+      "step": 4448
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 0.5583685636520386,
+      "learning_rate": 5.771016813450761e-05,
+      "loss": 0.4379,
+      "step": 4449
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 0.5552833676338196,
+      "learning_rate": 5.767814251401121e-05,
+      "loss": 0.4844,
+      "step": 4450
+    },
+    {
+      "epoch": 1.42432,
+      "grad_norm": 0.5559164881706238,
+      "learning_rate": 5.764611689351481e-05,
+      "loss": 0.3339,
+      "step": 4451
+    },
+    {
+      "epoch": 1.4246400000000001,
+      "grad_norm": 0.4828597605228424,
+      "learning_rate": 5.761409127301842e-05,
+      "loss": 0.3923,
+      "step": 4452
+    },
+    {
+      "epoch": 1.42496,
+      "grad_norm": 0.4528234899044037,
+      "learning_rate": 5.758206565252202e-05,
+      "loss": 0.3774,
+      "step": 4453
+    },
+    {
+      "epoch": 1.4252799999999999,
+      "grad_norm": 0.5662843585014343,
+      "learning_rate": 5.7550040032025624e-05,
+      "loss": 0.3798,
+      "step": 4454
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 0.5732978582382202,
+      "learning_rate": 5.751801441152923e-05,
+      "loss": 0.5666,
+      "step": 4455
+    },
+    {
+      "epoch": 1.42592,
+      "grad_norm": 0.5776765942573547,
+      "learning_rate": 5.748598879103283e-05,
+      "loss": 0.4426,
+      "step": 4456
+    },
+    {
+      "epoch": 1.42624,
+      "grad_norm": 0.6499331593513489,
+      "learning_rate": 5.745396317053643e-05,
+      "loss": 0.3874,
+      "step": 4457
+    },
+    {
+      "epoch": 1.42656,
+      "grad_norm": 0.6307076215744019,
+      "learning_rate": 5.7421937550040036e-05,
+      "loss": 0.5379,
+      "step": 4458
+    },
+    {
+      "epoch": 1.42688,
+      "grad_norm": 0.7925885319709778,
+      "learning_rate": 5.738991192954364e-05,
+      "loss": 0.6813,
+      "step": 4459
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 0.525276243686676,
+      "learning_rate": 5.735788630904724e-05,
+      "loss": 0.3949,
+      "step": 4460
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 0.5513490438461304,
+      "learning_rate": 5.732586068855085e-05,
+      "loss": 0.3814,
+      "step": 4461
+    },
+    {
+      "epoch": 1.42784,
+      "grad_norm": 0.5896931290626526,
+      "learning_rate": 5.729383506805445e-05,
+      "loss": 0.4505,
+      "step": 4462
+    },
+    {
+      "epoch": 1.42816,
+      "grad_norm": 0.5354005098342896,
+      "learning_rate": 5.7261809447558046e-05,
+      "loss": 0.3424,
+      "step": 4463
+    },
+    {
+      "epoch": 1.42848,
+      "grad_norm": 0.7042950391769409,
+      "learning_rate": 5.722978382706166e-05,
+      "loss": 0.5062,
+      "step": 4464
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 0.5830133557319641,
+      "learning_rate": 5.7197758206565256e-05,
+      "loss": 0.4719,
+      "step": 4465
+    },
+    {
+      "epoch": 1.42912,
+      "grad_norm": 0.6468369364738464,
+      "learning_rate": 5.7165732586068854e-05,
+      "loss": 0.5918,
+      "step": 4466
+    },
+    {
+      "epoch": 1.42944,
+      "grad_norm": 0.6201424598693848,
+      "learning_rate": 5.7133706965572465e-05,
+      "loss": 0.5041,
+      "step": 4467
+    },
+    {
+      "epoch": 1.42976,
+      "grad_norm": 0.5592491626739502,
+      "learning_rate": 5.710168134507606e-05,
+      "loss": 0.4563,
+      "step": 4468
+    },
+    {
+      "epoch": 1.43008,
+      "grad_norm": 0.5462591648101807,
+      "learning_rate": 5.706965572457966e-05,
+      "loss": 0.3381,
+      "step": 4469
+    },
+    {
+      "epoch": 1.4304000000000001,
+      "grad_norm": 0.6040661931037903,
+      "learning_rate": 5.703763010408327e-05,
+      "loss": 0.5645,
+      "step": 4470
+    },
+    {
+      "epoch": 1.43072,
+      "grad_norm": 0.5020428895950317,
+      "learning_rate": 5.700560448358687e-05,
+      "loss": 0.4363,
+      "step": 4471
+    },
+    {
+      "epoch": 1.4310399999999999,
+      "grad_norm": 0.5660825371742249,
+      "learning_rate": 5.697357886309047e-05,
+      "loss": 0.4228,
+      "step": 4472
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 0.6847643852233887,
+      "learning_rate": 5.694155324259408e-05,
+      "loss": 0.4536,
+      "step": 4473
+    },
+    {
+      "epoch": 1.43168,
+      "grad_norm": 0.542328953742981,
+      "learning_rate": 5.690952762209768e-05,
+      "loss": 0.4981,
+      "step": 4474
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 0.7237474322319031,
+      "learning_rate": 5.6877502001601276e-05,
+      "loss": 0.678,
+      "step": 4475
+    },
+    {
+      "epoch": 1.43232,
+      "grad_norm": 0.46561938524246216,
+      "learning_rate": 5.684547638110489e-05,
+      "loss": 0.3728,
+      "step": 4476
+    },
+    {
+      "epoch": 1.4326400000000001,
+      "grad_norm": 0.5004613399505615,
+      "learning_rate": 5.6813450760608486e-05,
+      "loss": 0.5658,
+      "step": 4477
+    },
+    {
+      "epoch": 1.43296,
+      "grad_norm": 0.5856306552886963,
+      "learning_rate": 5.678142514011209e-05,
+      "loss": 0.3862,
+      "step": 4478
+    },
+    {
+      "epoch": 1.4332799999999999,
+      "grad_norm": 0.6972157955169678,
+      "learning_rate": 5.6749399519615695e-05,
+      "loss": 0.4558,
+      "step": 4479
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.5785418748855591,
+      "learning_rate": 5.67173738991193e-05,
+      "loss": 0.4926,
+      "step": 4480
+    },
+    {
+      "epoch": 1.43392,
+      "grad_norm": 0.6639676690101624,
+      "learning_rate": 5.66853482786229e-05,
+      "loss": 0.2983,
+      "step": 4481
+    },
+    {
+      "epoch": 1.43424,
+      "grad_norm": 0.469874769449234,
+      "learning_rate": 5.665332265812651e-05,
+      "loss": 0.3585,
+      "step": 4482
+    },
+    {
+      "epoch": 1.43456,
+      "grad_norm": 0.6708923578262329,
+      "learning_rate": 5.662129703763011e-05,
+      "loss": 0.4915,
+      "step": 4483
+    },
+    {
+      "epoch": 1.43488,
+      "grad_norm": 0.58640456199646,
+      "learning_rate": 5.6589271417133705e-05,
+      "loss": 0.544,
+      "step": 4484
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.6817744374275208,
+      "learning_rate": 5.655724579663732e-05,
+      "loss": 0.5268,
+      "step": 4485
+    },
+    {
+      "epoch": 1.43552,
+      "grad_norm": 0.515275239944458,
+      "learning_rate": 5.6525220176140915e-05,
+      "loss": 0.4933,
+      "step": 4486
+    },
+    {
+      "epoch": 1.43584,
+      "grad_norm": 0.6546064615249634,
+      "learning_rate": 5.649319455564451e-05,
+      "loss": 0.4799,
+      "step": 4487
+    },
+    {
+      "epoch": 1.43616,
+      "grad_norm": 0.5942469239234924,
+      "learning_rate": 5.6461168935148124e-05,
+      "loss": 0.3742,
+      "step": 4488
+    },
+    {
+      "epoch": 1.43648,
+      "grad_norm": 0.8755815029144287,
+      "learning_rate": 5.642914331465172e-05,
+      "loss": 0.4442,
+      "step": 4489
+    },
+    {
+      "epoch": 1.4368,
+      "grad_norm": 0.5819509625434875,
+      "learning_rate": 5.639711769415532e-05,
+      "loss": 0.4902,
+      "step": 4490
+    },
+    {
+      "epoch": 1.43712,
+      "grad_norm": 0.6894660592079163,
+      "learning_rate": 5.636509207365893e-05,
+      "loss": 0.4429,
+      "step": 4491
+    },
+    {
+      "epoch": 1.43744,
+      "grad_norm": 0.6062670946121216,
+      "learning_rate": 5.633306645316253e-05,
+      "loss": 0.5295,
+      "step": 4492
+    },
+    {
+      "epoch": 1.43776,
+      "grad_norm": 0.6698014140129089,
+      "learning_rate": 5.630104083266614e-05,
+      "loss": 0.3999,
+      "step": 4493
+    },
+    {
+      "epoch": 1.43808,
+      "grad_norm": 0.7085646390914917,
+      "learning_rate": 5.626901521216974e-05,
+      "loss": 0.5319,
+      "step": 4494
+    },
+    {
+      "epoch": 1.4384000000000001,
+      "grad_norm": 0.6298987865447998,
+      "learning_rate": 5.623698959167334e-05,
+      "loss": 0.4694,
+      "step": 4495
+    },
+    {
+      "epoch": 1.43872,
+      "grad_norm": 0.8260484933853149,
+      "learning_rate": 5.620496397117695e-05,
+      "loss": 0.5,
+      "step": 4496
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 0.6551100015640259,
+      "learning_rate": 5.617293835068055e-05,
+      "loss": 0.4446,
+      "step": 4497
+    },
+    {
+      "epoch": 1.43936,
+      "grad_norm": 0.4837891757488251,
+      "learning_rate": 5.6140912730184145e-05,
+      "loss": 0.481,
+      "step": 4498
+    },
+    {
+      "epoch": 1.43968,
+      "grad_norm": 0.5718587636947632,
+      "learning_rate": 5.6108887109687757e-05,
+      "loss": 0.4929,
+      "step": 4499
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7033084034919739,
+      "learning_rate": 5.6076861489191355e-05,
+      "loss": 0.4745,
+      "step": 4500
+    },
+    {
+      "epoch": 1.44032,
+      "grad_norm": 0.6523274183273315,
+      "learning_rate": 5.604483586869496e-05,
+      "loss": 0.5172,
+      "step": 4501
+    },
+    {
+      "epoch": 1.44064,
+      "grad_norm": 0.5832304358482361,
+      "learning_rate": 5.6012810248198564e-05,
+      "loss": 0.4757,
+      "step": 4502
+    },
+    {
+      "epoch": 1.44096,
+      "grad_norm": 0.6817187666893005,
+      "learning_rate": 5.598078462770216e-05,
+      "loss": 0.4605,
+      "step": 4503
+    },
+    {
+      "epoch": 1.44128,
+      "grad_norm": 0.5034158825874329,
+      "learning_rate": 5.594875900720577e-05,
+      "loss": 0.386,
+      "step": 4504
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 0.6272976398468018,
+      "learning_rate": 5.591673338670937e-05,
+      "loss": 0.414,
+      "step": 4505
+    },
+    {
+      "epoch": 1.44192,
+      "grad_norm": 0.7927446961402893,
+      "learning_rate": 5.5884707766212976e-05,
+      "loss": 0.5032,
+      "step": 4506
+    },
+    {
+      "epoch": 1.44224,
+      "grad_norm": 0.6259413957595825,
+      "learning_rate": 5.5852682145716574e-05,
+      "loss": 0.3802,
+      "step": 4507
+    },
+    {
+      "epoch": 1.44256,
+      "grad_norm": 0.6340528130531311,
+      "learning_rate": 5.5820656525220186e-05,
+      "loss": 0.4938,
+      "step": 4508
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 0.5937731862068176,
+      "learning_rate": 5.5788630904723784e-05,
+      "loss": 0.4942,
+      "step": 4509
+    },
+    {
+      "epoch": 1.4432,
+      "grad_norm": 0.5028270483016968,
+      "learning_rate": 5.575660528422738e-05,
+      "loss": 0.5206,
+      "step": 4510
+    },
+    {
+      "epoch": 1.44352,
+      "grad_norm": 0.7219099998474121,
+      "learning_rate": 5.572457966373099e-05,
+      "loss": 0.455,
+      "step": 4511
+    },
+    {
+      "epoch": 1.44384,
+      "grad_norm": 0.6976882815361023,
+      "learning_rate": 5.569255404323459e-05,
+      "loss": 0.4991,
+      "step": 4512
+    },
+    {
+      "epoch": 1.44416,
+      "grad_norm": 0.7104635834693909,
+      "learning_rate": 5.566052842273819e-05,
+      "loss": 0.5059,
+      "step": 4513
+    },
+    {
+      "epoch": 1.44448,
+      "grad_norm": 0.654168426990509,
+      "learning_rate": 5.56285028022418e-05,
+      "loss": 0.5389,
+      "step": 4514
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 0.6217471957206726,
+      "learning_rate": 5.55964771817454e-05,
+      "loss": 0.4112,
+      "step": 4515
+    },
+    {
+      "epoch": 1.44512,
+      "grad_norm": 0.5812258124351501,
+      "learning_rate": 5.5564451561249e-05,
+      "loss": 0.4945,
+      "step": 4516
+    },
+    {
+      "epoch": 1.44544,
+      "grad_norm": 0.5329819917678833,
+      "learning_rate": 5.553242594075261e-05,
+      "loss": 0.3745,
+      "step": 4517
+    },
+    {
+      "epoch": 1.44576,
+      "grad_norm": 0.5345543026924133,
+      "learning_rate": 5.5500400320256206e-05,
+      "loss": 0.4579,
+      "step": 4518
+    },
+    {
+      "epoch": 1.44608,
+      "grad_norm": 0.5636152625083923,
+      "learning_rate": 5.5468374699759804e-05,
+      "loss": 0.4902,
+      "step": 4519
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.6059844493865967,
+      "learning_rate": 5.5436349079263416e-05,
+      "loss": 0.3986,
+      "step": 4520
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 0.6664760112762451,
+      "learning_rate": 5.5404323458767014e-05,
+      "loss": 0.4429,
+      "step": 4521
+    },
+    {
+      "epoch": 1.4470399999999999,
+      "grad_norm": 0.5227361917495728,
+      "learning_rate": 5.537229783827061e-05,
+      "loss": 0.4505,
+      "step": 4522
+    },
+    {
+      "epoch": 1.44736,
+      "grad_norm": 0.44634684920310974,
+      "learning_rate": 5.534027221777422e-05,
+      "loss": 0.3718,
+      "step": 4523
+    },
+    {
+      "epoch": 1.44768,
+      "grad_norm": 0.6564716696739197,
+      "learning_rate": 5.530824659727782e-05,
+      "loss": 0.4526,
+      "step": 4524
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.4731645882129669,
+      "learning_rate": 5.5276220976781426e-05,
+      "loss": 0.3117,
+      "step": 4525
+    },
+    {
+      "epoch": 1.44832,
+      "grad_norm": 0.4127015769481659,
+      "learning_rate": 5.524419535628503e-05,
+      "loss": 0.4559,
+      "step": 4526
+    },
+    {
+      "epoch": 1.44864,
+      "grad_norm": 0.5261837840080261,
+      "learning_rate": 5.5212169735788636e-05,
+      "loss": 0.4512,
+      "step": 4527
+    },
+    {
+      "epoch": 1.44896,
+      "grad_norm": 0.6370715498924255,
+      "learning_rate": 5.5180144115292234e-05,
+      "loss": 0.3995,
+      "step": 4528
+    },
+    {
+      "epoch": 1.44928,
+      "grad_norm": 0.6476557850837708,
+      "learning_rate": 5.5148118494795845e-05,
+      "loss": 0.4915,
+      "step": 4529
+    },
+    {
+      "epoch": 1.4496,
+      "grad_norm": 0.6141683459281921,
+      "learning_rate": 5.511609287429944e-05,
+      "loss": 0.3994,
+      "step": 4530
+    },
+    {
+      "epoch": 1.44992,
+      "grad_norm": 0.6061080694198608,
+      "learning_rate": 5.508406725380304e-05,
+      "loss": 0.4365,
+      "step": 4531
+    },
+    {
+      "epoch": 1.45024,
+      "grad_norm": 0.6886715888977051,
+      "learning_rate": 5.505204163330665e-05,
+      "loss": 0.4442,
+      "step": 4532
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 0.736088752746582,
+      "learning_rate": 5.502001601281025e-05,
+      "loss": 0.5084,
+      "step": 4533
+    },
+    {
+      "epoch": 1.45088,
+      "grad_norm": 0.5913689136505127,
+      "learning_rate": 5.498799039231385e-05,
+      "loss": 0.4507,
+      "step": 4534
+    },
+    {
+      "epoch": 1.4512,
+      "grad_norm": 0.5910004377365112,
+      "learning_rate": 5.495596477181746e-05,
+      "loss": 0.4777,
+      "step": 4535
+    },
+    {
+      "epoch": 1.45152,
+      "grad_norm": 0.5905410647392273,
+      "learning_rate": 5.492393915132106e-05,
+      "loss": 0.3701,
+      "step": 4536
+    },
+    {
+      "epoch": 1.45184,
+      "grad_norm": 0.5186659693717957,
+      "learning_rate": 5.4891913530824656e-05,
+      "loss": 0.4651,
+      "step": 4537
+    },
+    {
+      "epoch": 1.4521600000000001,
+      "grad_norm": 0.6414391994476318,
+      "learning_rate": 5.485988791032827e-05,
+      "loss": 0.4396,
+      "step": 4538
+    },
+    {
+      "epoch": 1.45248,
+      "grad_norm": 0.6242100596427917,
+      "learning_rate": 5.4827862289831866e-05,
+      "loss": 0.4706,
+      "step": 4539
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 0.4785717725753784,
+      "learning_rate": 5.479583666933548e-05,
+      "loss": 0.4301,
+      "step": 4540
+    },
+    {
+      "epoch": 1.45312,
+      "grad_norm": 0.6395258903503418,
+      "learning_rate": 5.4763811048839075e-05,
+      "loss": 0.3756,
+      "step": 4541
+    },
+    {
+      "epoch": 1.45344,
+      "grad_norm": 0.671898603439331,
+      "learning_rate": 5.473178542834267e-05,
+      "loss": 0.5403,
+      "step": 4542
+    },
+    {
+      "epoch": 1.45376,
+      "grad_norm": 0.747259795665741,
+      "learning_rate": 5.4699759807846285e-05,
+      "loss": 0.4939,
+      "step": 4543
+    },
+    {
+      "epoch": 1.45408,
+      "grad_norm": 0.6956653594970703,
+      "learning_rate": 5.466773418734988e-05,
+      "loss": 0.522,
+      "step": 4544
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.559464693069458,
+      "learning_rate": 5.463570856685348e-05,
+      "loss": 0.4039,
+      "step": 4545
+    },
+    {
+      "epoch": 1.45472,
+      "grad_norm": 0.531466007232666,
+      "learning_rate": 5.460368294635709e-05,
+      "loss": 0.3751,
+      "step": 4546
+    },
+    {
+      "epoch": 1.45504,
+      "grad_norm": 0.6742259860038757,
+      "learning_rate": 5.457165732586069e-05,
+      "loss": 0.4293,
+      "step": 4547
+    },
+    {
+      "epoch": 1.45536,
+      "grad_norm": 0.5527973175048828,
+      "learning_rate": 5.453963170536429e-05,
+      "loss": 0.3935,
+      "step": 4548
+    },
+    {
+      "epoch": 1.45568,
+      "grad_norm": 0.6308045983314514,
+      "learning_rate": 5.45076060848679e-05,
+      "loss": 0.454,
+      "step": 4549
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 0.48133769631385803,
+      "learning_rate": 5.44755804643715e-05,
+      "loss": 0.3898,
+      "step": 4550
+    },
+    {
+      "epoch": 1.45632,
+      "grad_norm": 0.5714696049690247,
+      "learning_rate": 5.44435548438751e-05,
+      "loss": 0.4236,
+      "step": 4551
+    },
+    {
+      "epoch": 1.45664,
+      "grad_norm": 0.6017426252365112,
+      "learning_rate": 5.441152922337871e-05,
+      "loss": 0.6046,
+      "step": 4552
+    },
+    {
+      "epoch": 1.45696,
+      "grad_norm": 0.5690589547157288,
+      "learning_rate": 5.437950360288231e-05,
+      "loss": 0.5676,
+      "step": 4553
+    },
+    {
+      "epoch": 1.45728,
+      "grad_norm": 0.5385337471961975,
+      "learning_rate": 5.434747798238591e-05,
+      "loss": 0.452,
+      "step": 4554
+    },
+    {
+      "epoch": 1.4576,
+      "grad_norm": 0.7362189888954163,
+      "learning_rate": 5.431545236188952e-05,
+      "loss": 0.5268,
+      "step": 4555
+    },
+    {
+      "epoch": 1.45792,
+      "grad_norm": 0.5309774279594421,
+      "learning_rate": 5.428342674139312e-05,
+      "loss": 0.5038,
+      "step": 4556
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 0.6002863645553589,
+      "learning_rate": 5.425140112089672e-05,
+      "loss": 0.4816,
+      "step": 4557
+    },
+    {
+      "epoch": 1.45856,
+      "grad_norm": 0.5581322908401489,
+      "learning_rate": 5.421937550040033e-05,
+      "loss": 0.4386,
+      "step": 4558
+    },
+    {
+      "epoch": 1.45888,
+      "grad_norm": 0.6121633052825928,
+      "learning_rate": 5.418734987990393e-05,
+      "loss": 0.4697,
+      "step": 4559
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.683131754398346,
+      "learning_rate": 5.4155324259407525e-05,
+      "loss": 0.4715,
+      "step": 4560
+    },
+    {
+      "epoch": 1.45952,
+      "grad_norm": 0.5169825553894043,
+      "learning_rate": 5.4123298638911137e-05,
+      "loss": 0.4128,
+      "step": 4561
+    },
+    {
+      "epoch": 1.45984,
+      "grad_norm": 0.6622176170349121,
+      "learning_rate": 5.4091273018414735e-05,
+      "loss": 0.5006,
+      "step": 4562
+    },
+    {
+      "epoch": 1.4601600000000001,
+      "grad_norm": 0.5511406064033508,
+      "learning_rate": 5.405924739791833e-05,
+      "loss": 0.4833,
+      "step": 4563
+    },
+    {
+      "epoch": 1.46048,
+      "grad_norm": 0.8210783004760742,
+      "learning_rate": 5.4027221777421944e-05,
+      "loss": 0.4499,
+      "step": 4564
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 0.41943299770355225,
+      "learning_rate": 5.399519615692554e-05,
+      "loss": 0.3452,
+      "step": 4565
+    },
+    {
+      "epoch": 1.46112,
+      "grad_norm": 0.5936564803123474,
+      "learning_rate": 5.396317053642914e-05,
+      "loss": 0.5531,
+      "step": 4566
+    },
+    {
+      "epoch": 1.46144,
+      "grad_norm": 0.4646463096141815,
+      "learning_rate": 5.393114491593275e-05,
+      "loss": 0.4087,
+      "step": 4567
+    },
+    {
+      "epoch": 1.46176,
+      "grad_norm": 0.6657764911651611,
+      "learning_rate": 5.389911929543635e-05,
+      "loss": 0.5091,
+      "step": 4568
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 0.5820514559745789,
+      "learning_rate": 5.386709367493995e-05,
+      "loss": 0.4856,
+      "step": 4569
+    },
+    {
+      "epoch": 1.4624,
+      "grad_norm": 0.8756990432739258,
+      "learning_rate": 5.383506805444356e-05,
+      "loss": 0.6458,
+      "step": 4570
+    },
+    {
+      "epoch": 1.46272,
+      "grad_norm": 0.6349920034408569,
+      "learning_rate": 5.380304243394716e-05,
+      "loss": 0.4463,
+      "step": 4571
+    },
+    {
+      "epoch": 1.46304,
+      "grad_norm": 0.5552802085876465,
+      "learning_rate": 5.377101681345076e-05,
+      "loss": 0.5285,
+      "step": 4572
+    },
+    {
+      "epoch": 1.46336,
+      "grad_norm": 0.4094937741756439,
+      "learning_rate": 5.373899119295437e-05,
+      "loss": 0.3393,
+      "step": 4573
+    },
+    {
+      "epoch": 1.46368,
+      "grad_norm": 0.5855910778045654,
+      "learning_rate": 5.3706965572457965e-05,
+      "loss": 0.4408,
+      "step": 4574
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.7400329113006592,
+      "learning_rate": 5.367493995196157e-05,
+      "loss": 0.4395,
+      "step": 4575
+    },
+    {
+      "epoch": 1.46432,
+      "grad_norm": 0.5521350502967834,
+      "learning_rate": 5.3642914331465174e-05,
+      "loss": 0.4692,
+      "step": 4576
+    },
+    {
+      "epoch": 1.46464,
+      "grad_norm": 0.7321921586990356,
+      "learning_rate": 5.361088871096878e-05,
+      "loss": 0.4594,
+      "step": 4577
+    },
+    {
+      "epoch": 1.46496,
+      "grad_norm": 0.45471087098121643,
+      "learning_rate": 5.357886309047238e-05,
+      "loss": 0.4015,
+      "step": 4578
+    },
+    {
+      "epoch": 1.46528,
+      "grad_norm": 0.5852909684181213,
+      "learning_rate": 5.354683746997599e-05,
+      "loss": 0.4322,
+      "step": 4579
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 0.6456478238105774,
+      "learning_rate": 5.3514811849479586e-05,
+      "loss": 0.4674,
+      "step": 4580
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 0.5626813173294067,
+      "learning_rate": 5.3482786228983184e-05,
+      "loss": 0.5739,
+      "step": 4581
+    },
+    {
+      "epoch": 1.46624,
+      "grad_norm": 0.7250415086746216,
+      "learning_rate": 5.3450760608486796e-05,
+      "loss": 0.5712,
+      "step": 4582
+    },
+    {
+      "epoch": 1.4665599999999999,
+      "grad_norm": 0.4512862265110016,
+      "learning_rate": 5.3418734987990394e-05,
+      "loss": 0.3803,
+      "step": 4583
+    },
+    {
+      "epoch": 1.46688,
+      "grad_norm": 0.593744695186615,
+      "learning_rate": 5.338670936749399e-05,
+      "loss": 0.5944,
+      "step": 4584
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 0.7022010087966919,
+      "learning_rate": 5.3354683746997603e-05,
+      "loss": 0.459,
+      "step": 4585
+    },
+    {
+      "epoch": 1.46752,
+      "grad_norm": 0.5782161951065063,
+      "learning_rate": 5.33226581265012e-05,
+      "loss": 0.4018,
+      "step": 4586
+    },
+    {
+      "epoch": 1.46784,
+      "grad_norm": 0.4795737564563751,
+      "learning_rate": 5.329063250600481e-05,
+      "loss": 0.3457,
+      "step": 4587
+    },
+    {
+      "epoch": 1.4681600000000001,
+      "grad_norm": 0.754830539226532,
+      "learning_rate": 5.325860688550841e-05,
+      "loss": 0.5189,
+      "step": 4588
+    },
+    {
+      "epoch": 1.46848,
+      "grad_norm": 0.5310376286506653,
+      "learning_rate": 5.322658126501201e-05,
+      "loss": 0.4301,
+      "step": 4589
+    },
+    {
+      "epoch": 1.4687999999999999,
+      "grad_norm": 0.6590364575386047,
+      "learning_rate": 5.319455564451562e-05,
+      "loss": 0.5043,
+      "step": 4590
+    },
+    {
+      "epoch": 1.46912,
+      "grad_norm": 0.5170981287956238,
+      "learning_rate": 5.316253002401922e-05,
+      "loss": 0.3755,
+      "step": 4591
+    },
+    {
+      "epoch": 1.46944,
+      "grad_norm": 0.5791333913803101,
+      "learning_rate": 5.3130504403522816e-05,
+      "loss": 0.3817,
+      "step": 4592
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 0.7031758427619934,
+      "learning_rate": 5.309847878302643e-05,
+      "loss": 0.5121,
+      "step": 4593
+    },
+    {
+      "epoch": 1.47008,
+      "grad_norm": 0.5815871953964233,
+      "learning_rate": 5.3066453162530026e-05,
+      "loss": 0.4952,
+      "step": 4594
+    },
+    {
+      "epoch": 1.4704,
+      "grad_norm": 0.6108233332633972,
+      "learning_rate": 5.3034427542033624e-05,
+      "loss": 0.3897,
+      "step": 4595
+    },
+    {
+      "epoch": 1.47072,
+      "grad_norm": 0.8479376435279846,
+      "learning_rate": 5.3002401921537236e-05,
+      "loss": 0.5376,
+      "step": 4596
+    },
+    {
+      "epoch": 1.47104,
+      "grad_norm": 0.5435357093811035,
+      "learning_rate": 5.2970376301040833e-05,
+      "loss": 0.3799,
+      "step": 4597
+    },
+    {
+      "epoch": 1.47136,
+      "grad_norm": 0.6324735879898071,
+      "learning_rate": 5.293835068054444e-05,
+      "loss": 0.452,
+      "step": 4598
+    },
+    {
+      "epoch": 1.47168,
+      "grad_norm": 0.6560746431350708,
+      "learning_rate": 5.290632506004804e-05,
+      "loss": 0.3659,
+      "step": 4599
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.7542440891265869,
+      "learning_rate": 5.287429943955165e-05,
+      "loss": 0.5332,
+      "step": 4600
+    },
+    {
+      "epoch": 1.47232,
+      "grad_norm": 0.6559101343154907,
+      "learning_rate": 5.2842273819055246e-05,
+      "loss": 0.3916,
+      "step": 4601
+    },
+    {
+      "epoch": 1.47264,
+      "grad_norm": 0.6501086950302124,
+      "learning_rate": 5.281024819855885e-05,
+      "loss": 0.5051,
+      "step": 4602
+    },
+    {
+      "epoch": 1.47296,
+      "grad_norm": 0.629135251045227,
+      "learning_rate": 5.2778222578062455e-05,
+      "loss": 0.4188,
+      "step": 4603
+    },
+    {
+      "epoch": 1.47328,
+      "grad_norm": 0.6322762370109558,
+      "learning_rate": 5.274619695756605e-05,
+      "loss": 0.4636,
+      "step": 4604
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.6656152606010437,
+      "learning_rate": 5.2714171337069665e-05,
+      "loss": 0.5364,
+      "step": 4605
+    },
+    {
+      "epoch": 1.4739200000000001,
+      "grad_norm": 0.5788151621818542,
+      "learning_rate": 5.268214571657326e-05,
+      "loss": 0.4046,
+      "step": 4606
+    },
+    {
+      "epoch": 1.47424,
+      "grad_norm": 0.42944079637527466,
+      "learning_rate": 5.265012009607686e-05,
+      "loss": 0.3887,
+      "step": 4607
+    },
+    {
+      "epoch": 1.4745599999999999,
+      "grad_norm": 0.3526090085506439,
+      "learning_rate": 5.261809447558047e-05,
+      "loss": 0.2822,
+      "step": 4608
+    },
+    {
+      "epoch": 1.47488,
+      "grad_norm": 0.5638303756713867,
+      "learning_rate": 5.258606885508407e-05,
+      "loss": 0.518,
+      "step": 4609
+    },
+    {
+      "epoch": 1.4752,
+      "grad_norm": 0.7163906693458557,
+      "learning_rate": 5.255404323458767e-05,
+      "loss": 0.6412,
+      "step": 4610
+    },
+    {
+      "epoch": 1.47552,
+      "grad_norm": 0.5565876960754395,
+      "learning_rate": 5.252201761409128e-05,
+      "loss": 0.3841,
+      "step": 4611
+    },
+    {
+      "epoch": 1.47584,
+      "grad_norm": 0.43775632977485657,
+      "learning_rate": 5.248999199359488e-05,
+      "loss": 0.3793,
+      "step": 4612
+    },
+    {
+      "epoch": 1.4761600000000001,
+      "grad_norm": 0.6294745206832886,
+      "learning_rate": 5.2457966373098476e-05,
+      "loss": 0.5221,
+      "step": 4613
+    },
+    {
+      "epoch": 1.47648,
+      "grad_norm": 0.6238160729408264,
+      "learning_rate": 5.242594075260209e-05,
+      "loss": 0.4443,
+      "step": 4614
+    },
+    {
+      "epoch": 1.4768,
+      "grad_norm": 0.5789990425109863,
+      "learning_rate": 5.2393915132105685e-05,
+      "loss": 0.4487,
+      "step": 4615
+    },
+    {
+      "epoch": 1.47712,
+      "grad_norm": 0.5943078994750977,
+      "learning_rate": 5.236188951160928e-05,
+      "loss": 0.4657,
+      "step": 4616
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 0.6775705814361572,
+      "learning_rate": 5.2329863891112895e-05,
+      "loss": 0.4156,
+      "step": 4617
+    },
+    {
+      "epoch": 1.47776,
+      "grad_norm": 0.4691201150417328,
+      "learning_rate": 5.229783827061649e-05,
+      "loss": 0.2934,
+      "step": 4618
+    },
+    {
+      "epoch": 1.47808,
+      "grad_norm": 0.47522494196891785,
+      "learning_rate": 5.226581265012009e-05,
+      "loss": 0.4494,
+      "step": 4619
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 0.5614331960678101,
+      "learning_rate": 5.22337870296237e-05,
+      "loss": 0.4696,
+      "step": 4620
+    },
+    {
+      "epoch": 1.47872,
+      "grad_norm": 0.6151960492134094,
+      "learning_rate": 5.22017614091273e-05,
+      "loss": 0.4846,
+      "step": 4621
+    },
+    {
+      "epoch": 1.47904,
+      "grad_norm": 0.5327469110488892,
+      "learning_rate": 5.2169735788630905e-05,
+      "loss": 0.4189,
+      "step": 4622
+    },
+    {
+      "epoch": 1.47936,
+      "grad_norm": 0.541114866733551,
+      "learning_rate": 5.213771016813451e-05,
+      "loss": 0.5286,
+      "step": 4623
+    },
+    {
+      "epoch": 1.47968,
+      "grad_norm": 0.5364333987236023,
+      "learning_rate": 5.2105684547638115e-05,
+      "loss": 0.426,
+      "step": 4624
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4468080997467041,
+      "learning_rate": 5.207365892714171e-05,
+      "loss": 0.3449,
+      "step": 4625
+    },
+    {
+      "epoch": 1.48032,
+      "grad_norm": 0.5454398393630981,
+      "learning_rate": 5.2041633306645324e-05,
+      "loss": 0.3667,
+      "step": 4626
+    },
+    {
+      "epoch": 1.48064,
+      "grad_norm": 0.5788027048110962,
+      "learning_rate": 5.200960768614892e-05,
+      "loss": 0.3927,
+      "step": 4627
+    },
+    {
+      "epoch": 1.48096,
+      "grad_norm": 0.7096971273422241,
+      "learning_rate": 5.197758206565252e-05,
+      "loss": 0.6011,
+      "step": 4628
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 0.6093466877937317,
+      "learning_rate": 5.194555644515613e-05,
+      "loss": 0.4613,
+      "step": 4629
+    },
+    {
+      "epoch": 1.4816,
+      "grad_norm": 0.5724893808364868,
+      "learning_rate": 5.191353082465973e-05,
+      "loss": 0.4223,
+      "step": 4630
+    },
+    {
+      "epoch": 1.4819200000000001,
+      "grad_norm": 0.46192899346351624,
+      "learning_rate": 5.188150520416333e-05,
+      "loss": 0.5454,
+      "step": 4631
+    },
+    {
+      "epoch": 1.48224,
+      "grad_norm": 0.4870675206184387,
+      "learning_rate": 5.184947958366694e-05,
+      "loss": 0.3225,
+      "step": 4632
+    },
+    {
+      "epoch": 1.4825599999999999,
+      "grad_norm": 0.4584518373012543,
+      "learning_rate": 5.181745396317054e-05,
+      "loss": 0.4286,
+      "step": 4633
+    },
+    {
+      "epoch": 1.48288,
+      "grad_norm": 0.6937483549118042,
+      "learning_rate": 5.178542834267415e-05,
+      "loss": 0.4816,
+      "step": 4634
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 0.5708996653556824,
+      "learning_rate": 5.175340272217775e-05,
+      "loss": 0.4571,
+      "step": 4635
+    },
+    {
+      "epoch": 1.48352,
+      "grad_norm": 0.5734128355979919,
+      "learning_rate": 5.1721377101681345e-05,
+      "loss": 0.4177,
+      "step": 4636
+    },
+    {
+      "epoch": 1.48384,
+      "grad_norm": 0.7229241728782654,
+      "learning_rate": 5.1689351481184956e-05,
+      "loss": 0.3746,
+      "step": 4637
+    },
+    {
+      "epoch": 1.48416,
+      "grad_norm": 0.650858461856842,
+      "learning_rate": 5.1657325860688554e-05,
+      "loss": 0.5332,
+      "step": 4638
+    },
+    {
+      "epoch": 1.48448,
+      "grad_norm": 0.66302889585495,
+      "learning_rate": 5.162530024019215e-05,
+      "loss": 0.5144,
+      "step": 4639
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.6414022445678711,
+      "learning_rate": 5.1593274619695764e-05,
+      "loss": 0.5165,
+      "step": 4640
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 0.572726309299469,
+      "learning_rate": 5.156124899919936e-05,
+      "loss": 0.3913,
+      "step": 4641
+    },
+    {
+      "epoch": 1.48544,
+      "grad_norm": 0.5697221755981445,
+      "learning_rate": 5.152922337870296e-05,
+      "loss": 0.4317,
+      "step": 4642
+    },
+    {
+      "epoch": 1.48576,
+      "grad_norm": 0.589024007320404,
+      "learning_rate": 5.149719775820657e-05,
+      "loss": 0.4811,
+      "step": 4643
+    },
+    {
+      "epoch": 1.48608,
+      "grad_norm": 0.7843692898750305,
+      "learning_rate": 5.146517213771017e-05,
+      "loss": 0.5779,
+      "step": 4644
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 0.5853954553604126,
+      "learning_rate": 5.143314651721377e-05,
+      "loss": 0.4285,
+      "step": 4645
+    },
+    {
+      "epoch": 1.48672,
+      "grad_norm": 0.600922167301178,
+      "learning_rate": 5.140112089671738e-05,
+      "loss": 0.5817,
+      "step": 4646
+    },
+    {
+      "epoch": 1.48704,
+      "grad_norm": 0.4518871605396271,
+      "learning_rate": 5.136909527622098e-05,
+      "loss": 0.3916,
+      "step": 4647
+    },
+    {
+      "epoch": 1.48736,
+      "grad_norm": 0.459471195936203,
+      "learning_rate": 5.133706965572458e-05,
+      "loss": 0.3691,
+      "step": 4648
+    },
+    {
+      "epoch": 1.4876800000000001,
+      "grad_norm": 0.5650181174278259,
+      "learning_rate": 5.1305044035228186e-05,
+      "loss": 0.4263,
+      "step": 4649
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 0.5123778581619263,
+      "learning_rate": 5.127301841473179e-05,
+      "loss": 0.3985,
+      "step": 4650
+    },
+    {
+      "epoch": 1.4883199999999999,
+      "grad_norm": 0.5773478746414185,
+      "learning_rate": 5.124099279423539e-05,
+      "loss": 0.3851,
+      "step": 4651
+    },
+    {
+      "epoch": 1.48864,
+      "grad_norm": 0.5894458889961243,
+      "learning_rate": 5.1208967173739e-05,
+      "loss": 0.4654,
+      "step": 4652
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 0.4974169433116913,
+      "learning_rate": 5.11769415532426e-05,
+      "loss": 0.4133,
+      "step": 4653
+    },
+    {
+      "epoch": 1.48928,
+      "grad_norm": 0.764571487903595,
+      "learning_rate": 5.1144915932746197e-05,
+      "loss": 0.5191,
+      "step": 4654
+    },
+    {
+      "epoch": 1.4896,
+      "grad_norm": 0.8447023034095764,
+      "learning_rate": 5.111289031224981e-05,
+      "loss": 0.5264,
+      "step": 4655
+    },
+    {
+      "epoch": 1.4899200000000001,
+      "grad_norm": 0.5139899253845215,
+      "learning_rate": 5.1080864691753406e-05,
+      "loss": 0.3903,
+      "step": 4656
+    },
+    {
+      "epoch": 1.49024,
+      "grad_norm": 0.6266457438468933,
+      "learning_rate": 5.1048839071257004e-05,
+      "loss": 0.4656,
+      "step": 4657
+    },
+    {
+      "epoch": 1.4905599999999999,
+      "grad_norm": 0.5180706977844238,
+      "learning_rate": 5.1016813450760616e-05,
+      "loss": 0.4004,
+      "step": 4658
+    },
+    {
+      "epoch": 1.49088,
+      "grad_norm": 0.7048482298851013,
+      "learning_rate": 5.0984787830264214e-05,
+      "loss": 0.5068,
+      "step": 4659
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 0.5550336837768555,
+      "learning_rate": 5.095276220976781e-05,
+      "loss": 0.4347,
+      "step": 4660
+    },
+    {
+      "epoch": 1.49152,
+      "grad_norm": 0.6786839365959167,
+      "learning_rate": 5.092073658927142e-05,
+      "loss": 0.5396,
+      "step": 4661
+    },
+    {
+      "epoch": 1.49184,
+      "grad_norm": 0.6224067211151123,
+      "learning_rate": 5.088871096877502e-05,
+      "loss": 0.3935,
+      "step": 4662
+    },
+    {
+      "epoch": 1.49216,
+      "grad_norm": 0.7139783501625061,
+      "learning_rate": 5.085668534827862e-05,
+      "loss": 0.3797,
+      "step": 4663
+    },
+    {
+      "epoch": 1.49248,
+      "grad_norm": 0.5376801490783691,
+      "learning_rate": 5.082465972778223e-05,
+      "loss": 0.4584,
+      "step": 4664
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.5712575316429138,
+      "learning_rate": 5.079263410728583e-05,
+      "loss": 0.4249,
+      "step": 4665
+    },
+    {
+      "epoch": 1.49312,
+      "grad_norm": 0.6840248703956604,
+      "learning_rate": 5.0760608486789427e-05,
+      "loss": 0.4333,
+      "step": 4666
+    },
+    {
+      "epoch": 1.49344,
+      "grad_norm": 0.5576247572898865,
+      "learning_rate": 5.072858286629304e-05,
+      "loss": 0.4743,
+      "step": 4667
+    },
+    {
+      "epoch": 1.49376,
+      "grad_norm": 0.5650597214698792,
+      "learning_rate": 5.0696557245796636e-05,
+      "loss": 0.4569,
+      "step": 4668
+    },
+    {
+      "epoch": 1.49408,
+      "grad_norm": 0.6256127953529358,
+      "learning_rate": 5.066453162530024e-05,
+      "loss": 0.4423,
+      "step": 4669
+    },
+    {
+      "epoch": 1.4944,
+      "grad_norm": 0.6049395203590393,
+      "learning_rate": 5.0632506004803846e-05,
+      "loss": 0.4813,
+      "step": 4670
+    },
+    {
+      "epoch": 1.49472,
+      "grad_norm": 0.6966089606285095,
+      "learning_rate": 5.060048038430745e-05,
+      "loss": 0.4816,
+      "step": 4671
+    },
+    {
+      "epoch": 1.49504,
+      "grad_norm": 0.5890588760375977,
+      "learning_rate": 5.056845476381105e-05,
+      "loss": 0.3811,
+      "step": 4672
+    },
+    {
+      "epoch": 1.49536,
+      "grad_norm": 0.6752182245254517,
+      "learning_rate": 5.053642914331465e-05,
+      "loss": 0.3882,
+      "step": 4673
+    },
+    {
+      "epoch": 1.4956800000000001,
+      "grad_norm": 0.6183276772499084,
+      "learning_rate": 5.050440352281826e-05,
+      "loss": 0.5137,
+      "step": 4674
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 0.5086974501609802,
+      "learning_rate": 5.0472377902321856e-05,
+      "loss": 0.4338,
+      "step": 4675
+    },
+    {
+      "epoch": 1.4963199999999999,
+      "grad_norm": 0.7701089978218079,
+      "learning_rate": 5.044035228182547e-05,
+      "loss": 0.4132,
+      "step": 4676
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 0.5129746794700623,
+      "learning_rate": 5.0408326661329065e-05,
+      "loss": 0.515,
+      "step": 4677
+    },
+    {
+      "epoch": 1.49696,
+      "grad_norm": 0.5945625901222229,
+      "learning_rate": 5.037630104083266e-05,
+      "loss": 0.4821,
+      "step": 4678
+    },
+    {
+      "epoch": 1.49728,
+      "grad_norm": 0.6290009021759033,
+      "learning_rate": 5.0344275420336275e-05,
+      "loss": 0.6032,
+      "step": 4679
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.4726774990558624,
+      "learning_rate": 5.031224979983987e-05,
+      "loss": 0.4229,
+      "step": 4680
+    },
+    {
+      "epoch": 1.49792,
+      "grad_norm": 0.517573356628418,
+      "learning_rate": 5.0280224179343484e-05,
+      "loss": 0.3381,
+      "step": 4681
+    },
+    {
+      "epoch": 1.49824,
+      "grad_norm": 0.5200397372245789,
+      "learning_rate": 5.024819855884708e-05,
+      "loss": 0.3843,
+      "step": 4682
+    },
+    {
+      "epoch": 1.49856,
+      "grad_norm": 0.609065055847168,
+      "learning_rate": 5.021617293835068e-05,
+      "loss": 0.4621,
+      "step": 4683
+    },
+    {
+      "epoch": 1.49888,
+      "grad_norm": 0.6200820803642273,
+      "learning_rate": 5.018414731785429e-05,
+      "loss": 0.3989,
+      "step": 4684
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 0.5215333104133606,
+      "learning_rate": 5.015212169735789e-05,
+      "loss": 0.4891,
+      "step": 4685
+    },
+    {
+      "epoch": 1.49952,
+      "grad_norm": 0.4983515441417694,
+      "learning_rate": 5.012009607686149e-05,
+      "loss": 0.3596,
+      "step": 4686
+    },
+    {
+      "epoch": 1.49984,
+      "grad_norm": 1.0814238786697388,
+      "learning_rate": 5.00880704563651e-05,
+      "loss": 0.3943,
+      "step": 4687
+    },
+    {
+      "epoch": 1.5001600000000002,
+      "grad_norm": 0.6068863272666931,
+      "learning_rate": 5.00560448358687e-05,
+      "loss": 0.4715,
+      "step": 4688
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 0.6194530129432678,
+      "learning_rate": 5.0024019215372295e-05,
+      "loss": 0.4514,
+      "step": 4689
+    },
+    {
+      "epoch": 1.5008,
+      "grad_norm": 0.6756887435913086,
+      "learning_rate": 4.99919935948759e-05,
+      "loss": 0.4238,
+      "step": 4690
+    },
+    {
+      "epoch": 1.50112,
+      "grad_norm": 0.6971200704574585,
+      "learning_rate": 4.9959967974379505e-05,
+      "loss": 0.4616,
+      "step": 4691
+    },
+    {
+      "epoch": 1.50144,
+      "grad_norm": 0.6896497011184692,
+      "learning_rate": 4.992794235388311e-05,
+      "loss": 0.5125,
+      "step": 4692
+    },
+    {
+      "epoch": 1.50176,
+      "grad_norm": 0.6113795042037964,
+      "learning_rate": 4.989591673338671e-05,
+      "loss": 0.3059,
+      "step": 4693
+    },
+    {
+      "epoch": 1.5020799999999999,
+      "grad_norm": 0.7536332607269287,
+      "learning_rate": 4.986389111289031e-05,
+      "loss": 0.64,
+      "step": 4694
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 0.6560052633285522,
+      "learning_rate": 4.983186549239392e-05,
+      "loss": 0.4671,
+      "step": 4695
+    },
+    {
+      "epoch": 1.50272,
+      "grad_norm": 0.7675296068191528,
+      "learning_rate": 4.979983987189752e-05,
+      "loss": 0.497,
+      "step": 4696
+    },
+    {
+      "epoch": 1.50304,
+      "grad_norm": 0.4816417396068573,
+      "learning_rate": 4.976781425140113e-05,
+      "loss": 0.5247,
+      "step": 4697
+    },
+    {
+      "epoch": 1.50336,
+      "grad_norm": 0.45526352524757385,
+      "learning_rate": 4.973578863090473e-05,
+      "loss": 0.4203,
+      "step": 4698
+    },
+    {
+      "epoch": 1.5036800000000001,
+      "grad_norm": 0.3615989089012146,
+      "learning_rate": 4.970376301040833e-05,
+      "loss": 0.3009,
+      "step": 4699
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 0.7391599416732788,
+      "learning_rate": 4.9671737389911934e-05,
+      "loss": 0.3729,
+      "step": 4700
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 0.6363121867179871,
+      "learning_rate": 4.963971176941554e-05,
+      "loss": 0.4034,
+      "step": 4701
+    },
+    {
+      "epoch": 1.50464,
+      "grad_norm": 0.48599278926849365,
+      "learning_rate": 4.960768614891914e-05,
+      "loss": 0.4233,
+      "step": 4702
+    },
+    {
+      "epoch": 1.50496,
+      "grad_norm": 0.7773295044898987,
+      "learning_rate": 4.957566052842274e-05,
+      "loss": 0.5556,
+      "step": 4703
+    },
+    {
+      "epoch": 1.50528,
+      "grad_norm": 0.6584400534629822,
+      "learning_rate": 4.9543634907926347e-05,
+      "loss": 0.4984,
+      "step": 4704
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 0.6793642640113831,
+      "learning_rate": 4.9511609287429945e-05,
+      "loss": 0.4822,
+      "step": 4705
+    },
+    {
+      "epoch": 1.5059200000000001,
+      "grad_norm": 0.6728963851928711,
+      "learning_rate": 4.947958366693355e-05,
+      "loss": 0.4397,
+      "step": 4706
+    },
+    {
+      "epoch": 1.50624,
+      "grad_norm": 0.6381680369377136,
+      "learning_rate": 4.9447558046437154e-05,
+      "loss": 0.4312,
+      "step": 4707
+    },
+    {
+      "epoch": 1.50656,
+      "grad_norm": 0.5153624415397644,
+      "learning_rate": 4.941553242594075e-05,
+      "loss": 0.3273,
+      "step": 4708
+    },
+    {
+      "epoch": 1.50688,
+      "grad_norm": 0.4793238639831543,
+      "learning_rate": 4.938350680544436e-05,
+      "loss": 0.3998,
+      "step": 4709
+    },
+    {
+      "epoch": 1.5072,
+      "grad_norm": 1.935788869857788,
+      "learning_rate": 4.935148118494796e-05,
+      "loss": 0.4367,
+      "step": 4710
+    },
+    {
+      "epoch": 1.50752,
+      "grad_norm": 0.5231462121009827,
+      "learning_rate": 4.931945556445156e-05,
+      "loss": 0.4221,
+      "step": 4711
+    },
+    {
+      "epoch": 1.5078399999999998,
+      "grad_norm": 0.6204705834388733,
+      "learning_rate": 4.9287429943955164e-05,
+      "loss": 0.4788,
+      "step": 4712
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 0.5853046774864197,
+      "learning_rate": 4.925540432345877e-05,
+      "loss": 0.4049,
+      "step": 4713
+    },
+    {
+      "epoch": 1.50848,
+      "grad_norm": 0.6093079447746277,
+      "learning_rate": 4.9223378702962374e-05,
+      "loss": 0.5135,
+      "step": 4714
+    },
+    {
+      "epoch": 1.5088,
+      "grad_norm": 0.5129292607307434,
+      "learning_rate": 4.919135308246597e-05,
+      "loss": 0.4217,
+      "step": 4715
+    },
+    {
+      "epoch": 1.50912,
+      "grad_norm": 0.8098514080047607,
+      "learning_rate": 4.9159327461969577e-05,
+      "loss": 0.6667,
+      "step": 4716
+    },
+    {
+      "epoch": 1.5094400000000001,
+      "grad_norm": 0.8154210448265076,
+      "learning_rate": 4.912730184147318e-05,
+      "loss": 0.4401,
+      "step": 4717
+    },
+    {
+      "epoch": 1.50976,
+      "grad_norm": 0.5972124934196472,
+      "learning_rate": 4.909527622097678e-05,
+      "loss": 0.4636,
+      "step": 4718
+    },
+    {
+      "epoch": 1.5100799999999999,
+      "grad_norm": 0.877556324005127,
+      "learning_rate": 4.9063250600480384e-05,
+      "loss": 0.5114,
+      "step": 4719
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.6360782980918884,
+      "learning_rate": 4.903122497998399e-05,
+      "loss": 0.3234,
+      "step": 4720
+    },
+    {
+      "epoch": 1.51072,
+      "grad_norm": 0.5418350696563721,
+      "learning_rate": 4.8999199359487594e-05,
+      "loss": 0.4227,
+      "step": 4721
+    },
+    {
+      "epoch": 1.51104,
+      "grad_norm": 0.6790722608566284,
+      "learning_rate": 4.89671737389912e-05,
+      "loss": 0.6342,
+      "step": 4722
+    },
+    {
+      "epoch": 1.51136,
+      "grad_norm": 0.7138352394104004,
+      "learning_rate": 4.89351481184948e-05,
+      "loss": 0.5324,
+      "step": 4723
+    },
+    {
+      "epoch": 1.5116800000000001,
+      "grad_norm": 0.5497035384178162,
+      "learning_rate": 4.89031224979984e-05,
+      "loss": 0.4578,
+      "step": 4724
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.5523759722709656,
+      "learning_rate": 4.8871096877502006e-05,
+      "loss": 0.4552,
+      "step": 4725
+    },
+    {
+      "epoch": 1.5123199999999999,
+      "grad_norm": 0.6901834607124329,
+      "learning_rate": 4.883907125700561e-05,
+      "loss": 0.4887,
+      "step": 4726
+    },
+    {
+      "epoch": 1.51264,
+      "grad_norm": 0.5638870596885681,
+      "learning_rate": 4.880704563650921e-05,
+      "loss": 0.3834,
+      "step": 4727
+    },
+    {
+      "epoch": 1.51296,
+      "grad_norm": 0.6519521474838257,
+      "learning_rate": 4.8775020016012813e-05,
+      "loss": 0.5064,
+      "step": 4728
+    },
+    {
+      "epoch": 1.51328,
+      "grad_norm": 0.6024044156074524,
+      "learning_rate": 4.874299439551642e-05,
+      "loss": 0.4611,
+      "step": 4729
+    },
+    {
+      "epoch": 1.5135999999999998,
+      "grad_norm": 0.6598643064498901,
+      "learning_rate": 4.8710968775020016e-05,
+      "loss": 0.4283,
+      "step": 4730
+    },
+    {
+      "epoch": 1.5139200000000002,
+      "grad_norm": 0.5644962787628174,
+      "learning_rate": 4.867894315452362e-05,
+      "loss": 0.4585,
+      "step": 4731
+    },
+    {
+      "epoch": 1.51424,
+      "grad_norm": 0.6287744045257568,
+      "learning_rate": 4.8646917534027226e-05,
+      "loss": 0.4175,
+      "step": 4732
+    },
+    {
+      "epoch": 1.51456,
+      "grad_norm": 0.48011350631713867,
+      "learning_rate": 4.8614891913530824e-05,
+      "loss": 0.2883,
+      "step": 4733
+    },
+    {
+      "epoch": 1.51488,
+      "grad_norm": 0.6403210163116455,
+      "learning_rate": 4.858286629303443e-05,
+      "loss": 0.4065,
+      "step": 4734
+    },
+    {
+      "epoch": 1.5152,
+      "grad_norm": 0.6085357666015625,
+      "learning_rate": 4.855084067253803e-05,
+      "loss": 0.4439,
+      "step": 4735
+    },
+    {
+      "epoch": 1.51552,
+      "grad_norm": 0.5913042426109314,
+      "learning_rate": 4.851881505204163e-05,
+      "loss": 0.361,
+      "step": 4736
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 0.5406367778778076,
+      "learning_rate": 4.8486789431545236e-05,
+      "loss": 0.4501,
+      "step": 4737
+    },
+    {
+      "epoch": 1.51616,
+      "grad_norm": 0.6727082133293152,
+      "learning_rate": 4.845476381104884e-05,
+      "loss": 0.5082,
+      "step": 4738
+    },
+    {
+      "epoch": 1.51648,
+      "grad_norm": 0.5205621123313904,
+      "learning_rate": 4.8422738190552445e-05,
+      "loss": 0.3463,
+      "step": 4739
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 0.646427571773529,
+      "learning_rate": 4.8390712570056043e-05,
+      "loss": 0.515,
+      "step": 4740
+    },
+    {
+      "epoch": 1.51712,
+      "grad_norm": 0.5468747019767761,
+      "learning_rate": 4.835868694955965e-05,
+      "loss": 0.4752,
+      "step": 4741
+    },
+    {
+      "epoch": 1.5174400000000001,
+      "grad_norm": 0.4609379768371582,
+      "learning_rate": 4.832666132906325e-05,
+      "loss": 0.337,
+      "step": 4742
+    },
+    {
+      "epoch": 1.51776,
+      "grad_norm": 0.6662101149559021,
+      "learning_rate": 4.829463570856686e-05,
+      "loss": 0.3898,
+      "step": 4743
+    },
+    {
+      "epoch": 1.5180799999999999,
+      "grad_norm": 0.5603944659233093,
+      "learning_rate": 4.8262610088070456e-05,
+      "loss": 0.3815,
+      "step": 4744
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 0.8107700943946838,
+      "learning_rate": 4.823058446757406e-05,
+      "loss": 0.4411,
+      "step": 4745
+    },
+    {
+      "epoch": 1.51872,
+      "grad_norm": 0.6246293783187866,
+      "learning_rate": 4.8198558847077665e-05,
+      "loss": 0.4984,
+      "step": 4746
+    },
+    {
+      "epoch": 1.51904,
+      "grad_norm": 0.622407853603363,
+      "learning_rate": 4.816653322658127e-05,
+      "loss": 0.3969,
+      "step": 4747
+    },
+    {
+      "epoch": 1.51936,
+      "grad_norm": 0.6821801066398621,
+      "learning_rate": 4.8134507606084875e-05,
+      "loss": 0.5042,
+      "step": 4748
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 0.6860357522964478,
+      "learning_rate": 4.810248198558847e-05,
+      "loss": 0.3774,
+      "step": 4749
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6028900146484375,
+      "learning_rate": 4.807045636509208e-05,
+      "loss": 0.3985,
+      "step": 4750
+    },
+    {
+      "epoch": 1.52032,
+      "grad_norm": 0.5081947445869446,
+      "learning_rate": 4.803843074459568e-05,
+      "loss": 0.4281,
+      "step": 4751
+    },
+    {
+      "epoch": 1.52064,
+      "grad_norm": 0.7579690217971802,
+      "learning_rate": 4.800640512409928e-05,
+      "loss": 0.6054,
+      "step": 4752
+    },
+    {
+      "epoch": 1.52096,
+      "grad_norm": 0.7496891021728516,
+      "learning_rate": 4.7974379503602885e-05,
+      "loss": 0.4639,
+      "step": 4753
+    },
+    {
+      "epoch": 1.52128,
+      "grad_norm": 0.6637994647026062,
+      "learning_rate": 4.794235388310649e-05,
+      "loss": 0.481,
+      "step": 4754
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 0.6115459203720093,
+      "learning_rate": 4.791032826261009e-05,
+      "loss": 0.4612,
+      "step": 4755
+    },
+    {
+      "epoch": 1.5219200000000002,
+      "grad_norm": 0.5393297672271729,
+      "learning_rate": 4.787830264211369e-05,
+      "loss": 0.3213,
+      "step": 4756
+    },
+    {
+      "epoch": 1.52224,
+      "grad_norm": 0.4168643057346344,
+      "learning_rate": 4.78462770216173e-05,
+      "loss": 0.4854,
+      "step": 4757
+    },
+    {
+      "epoch": 1.52256,
+      "grad_norm": 0.6368916630744934,
+      "learning_rate": 4.7814251401120895e-05,
+      "loss": 0.4374,
+      "step": 4758
+    },
+    {
+      "epoch": 1.52288,
+      "grad_norm": 0.5795953869819641,
+      "learning_rate": 4.77822257806245e-05,
+      "loss": 0.417,
+      "step": 4759
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.5326150059700012,
+      "learning_rate": 4.7750200160128105e-05,
+      "loss": 0.4598,
+      "step": 4760
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 0.7057052850723267,
+      "learning_rate": 4.771817453963171e-05,
+      "loss": 0.4922,
+      "step": 4761
+    },
+    {
+      "epoch": 1.5238399999999999,
+      "grad_norm": 0.6505028605461121,
+      "learning_rate": 4.768614891913531e-05,
+      "loss": 0.5484,
+      "step": 4762
+    },
+    {
+      "epoch": 1.52416,
+      "grad_norm": 0.3800669312477112,
+      "learning_rate": 4.765412329863891e-05,
+      "loss": 0.3272,
+      "step": 4763
+    },
+    {
+      "epoch": 1.52448,
+      "grad_norm": 0.6643231511116028,
+      "learning_rate": 4.762209767814252e-05,
+      "loss": 0.4412,
+      "step": 4764
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 0.5546316504478455,
+      "learning_rate": 4.7590072057646115e-05,
+      "loss": 0.3108,
+      "step": 4765
+    },
+    {
+      "epoch": 1.52512,
+      "grad_norm": 0.5025911927223206,
+      "learning_rate": 4.755804643714972e-05,
+      "loss": 0.4159,
+      "step": 4766
+    },
+    {
+      "epoch": 1.5254400000000001,
+      "grad_norm": 0.4912903904914856,
+      "learning_rate": 4.7526020816653325e-05,
+      "loss": 0.3339,
+      "step": 4767
+    },
+    {
+      "epoch": 1.52576,
+      "grad_norm": 0.6346544623374939,
+      "learning_rate": 4.749399519615693e-05,
+      "loss": 0.4419,
+      "step": 4768
+    },
+    {
+      "epoch": 1.5260799999999999,
+      "grad_norm": 0.4998990595340729,
+      "learning_rate": 4.7461969575660534e-05,
+      "loss": 0.4015,
+      "step": 4769
+    },
+    {
+      "epoch": 1.5264,
+      "grad_norm": 0.5806402564048767,
+      "learning_rate": 4.742994395516414e-05,
+      "loss": 0.3572,
+      "step": 4770
+    },
+    {
+      "epoch": 1.52672,
+      "grad_norm": 0.6345856785774231,
+      "learning_rate": 4.739791833466774e-05,
+      "loss": 0.4909,
+      "step": 4771
+    },
+    {
+      "epoch": 1.52704,
+      "grad_norm": 0.6190733313560486,
+      "learning_rate": 4.736589271417134e-05,
+      "loss": 0.524,
+      "step": 4772
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 0.6988130807876587,
+      "learning_rate": 4.7333867093674946e-05,
+      "loss": 0.5248,
+      "step": 4773
+    },
+    {
+      "epoch": 1.5276800000000001,
+      "grad_norm": 0.6218895316123962,
+      "learning_rate": 4.7301841473178544e-05,
+      "loss": 0.457,
+      "step": 4774
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 0.6245235204696655,
+      "learning_rate": 4.726981585268215e-05,
+      "loss": 0.6209,
+      "step": 4775
+    },
+    {
+      "epoch": 1.52832,
+      "grad_norm": 0.7377709746360779,
+      "learning_rate": 4.7237790232185754e-05,
+      "loss": 0.4433,
+      "step": 4776
+    },
+    {
+      "epoch": 1.52864,
+      "grad_norm": 0.6552140116691589,
+      "learning_rate": 4.720576461168935e-05,
+      "loss": 0.4698,
+      "step": 4777
+    },
+    {
+      "epoch": 1.52896,
+      "grad_norm": 0.5452696084976196,
+      "learning_rate": 4.717373899119296e-05,
+      "loss": 0.4306,
+      "step": 4778
+    },
+    {
+      "epoch": 1.52928,
+      "grad_norm": 0.6197740435600281,
+      "learning_rate": 4.714171337069656e-05,
+      "loss": 0.4235,
+      "step": 4779
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 0.6225225925445557,
+      "learning_rate": 4.710968775020016e-05,
+      "loss": 0.4392,
+      "step": 4780
+    },
+    {
+      "epoch": 1.52992,
+      "grad_norm": 0.6523774862289429,
+      "learning_rate": 4.7077662129703764e-05,
+      "loss": 0.4559,
+      "step": 4781
+    },
+    {
+      "epoch": 1.53024,
+      "grad_norm": 0.7311254143714905,
+      "learning_rate": 4.704563650920737e-05,
+      "loss": 0.375,
+      "step": 4782
+    },
+    {
+      "epoch": 1.53056,
+      "grad_norm": 0.604383111000061,
+      "learning_rate": 4.701361088871097e-05,
+      "loss": 0.4708,
+      "step": 4783
+    },
+    {
+      "epoch": 1.53088,
+      "grad_norm": 0.5650271773338318,
+      "learning_rate": 4.698158526821457e-05,
+      "loss": 0.524,
+      "step": 4784
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.5686578750610352,
+      "learning_rate": 4.6949559647718176e-05,
+      "loss": 0.3506,
+      "step": 4785
+    },
+    {
+      "epoch": 1.53152,
+      "grad_norm": 0.5630221962928772,
+      "learning_rate": 4.691753402722178e-05,
+      "loss": 0.4055,
+      "step": 4786
+    },
+    {
+      "epoch": 1.5318399999999999,
+      "grad_norm": 0.8550300002098083,
+      "learning_rate": 4.688550840672538e-05,
+      "loss": 0.4525,
+      "step": 4787
+    },
+    {
+      "epoch": 1.53216,
+      "grad_norm": 0.7214192152023315,
+      "learning_rate": 4.6853482786228984e-05,
+      "loss": 0.4349,
+      "step": 4788
+    },
+    {
+      "epoch": 1.53248,
+      "grad_norm": 0.607349693775177,
+      "learning_rate": 4.682145716573259e-05,
+      "loss": 0.4474,
+      "step": 4789
+    },
+    {
+      "epoch": 1.5328,
+      "grad_norm": 0.5198094844818115,
+      "learning_rate": 4.678943154523619e-05,
+      "loss": 0.3754,
+      "step": 4790
+    },
+    {
+      "epoch": 1.53312,
+      "grad_norm": 0.5183441042900085,
+      "learning_rate": 4.675740592473979e-05,
+      "loss": 0.3981,
+      "step": 4791
+    },
+    {
+      "epoch": 1.5334400000000001,
+      "grad_norm": 0.5722155570983887,
+      "learning_rate": 4.6725380304243396e-05,
+      "loss": 0.373,
+      "step": 4792
+    },
+    {
+      "epoch": 1.53376,
+      "grad_norm": 0.5191179513931274,
+      "learning_rate": 4.6693354683747e-05,
+      "loss": 0.4532,
+      "step": 4793
+    },
+    {
+      "epoch": 1.5340799999999999,
+      "grad_norm": 0.6414000988006592,
+      "learning_rate": 4.6661329063250606e-05,
+      "loss": 0.6135,
+      "step": 4794
+    },
+    {
+      "epoch": 1.5344,
+      "grad_norm": 0.5325633883476257,
+      "learning_rate": 4.662930344275421e-05,
+      "loss": 0.5042,
+      "step": 4795
+    },
+    {
+      "epoch": 1.53472,
+      "grad_norm": 0.7569255828857422,
+      "learning_rate": 4.659727782225781e-05,
+      "loss": 0.4124,
+      "step": 4796
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 0.8150873780250549,
+      "learning_rate": 4.656525220176141e-05,
+      "loss": 0.437,
+      "step": 4797
+    },
+    {
+      "epoch": 1.5353599999999998,
+      "grad_norm": 0.7377203106880188,
+      "learning_rate": 4.653322658126502e-05,
+      "loss": 0.5527,
+      "step": 4798
+    },
+    {
+      "epoch": 1.5356800000000002,
+      "grad_norm": 0.4963825047016144,
+      "learning_rate": 4.6501200960768616e-05,
+      "loss": 0.3495,
+      "step": 4799
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.5330488085746765,
+      "learning_rate": 4.646917534027222e-05,
+      "loss": 0.5026,
+      "step": 4800
+    },
+    {
+      "epoch": 1.53632,
+      "grad_norm": 0.7084789276123047,
+      "learning_rate": 4.6437149719775826e-05,
+      "loss": 0.6065,
+      "step": 4801
+    },
+    {
+      "epoch": 1.53664,
+      "grad_norm": 0.5714585781097412,
+      "learning_rate": 4.6405124099279424e-05,
+      "loss": 0.3648,
+      "step": 4802
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.412158773855191e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}