diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7801 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.49992924461551524,
+  "eval_steps": 276,
+  "global_step": 1104,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00045283446070245947,
+      "grad_norm": 4.009160995483398,
+      "learning_rate": 1.5e-06,
+      "loss": 1.777,
+      "step": 1
+    },
+    {
+      "epoch": 0.00045283446070245947,
+      "eval_loss": 1.9080619812011719,
+      "eval_runtime": 19.3772,
+      "eval_samples_per_second": 5.883,
+      "eval_steps_per_second": 0.774,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009056689214049189,
+      "grad_norm": 3.954909086227417,
+      "learning_rate": 3e-06,
+      "loss": 1.7658,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013585033821073783,
+      "grad_norm": 3.939889907836914,
+      "learning_rate": 4.5e-06,
+      "loss": 1.7708,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018113378428098379,
+      "grad_norm": 2.411039352416992,
+      "learning_rate": 6e-06,
+      "loss": 1.7529,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022641723035122974,
+      "grad_norm": 1.929020643234253,
+      "learning_rate": 7.5e-06,
+      "loss": 1.6849,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027170067642147567,
+      "grad_norm": 2.1016135215759277,
+      "learning_rate": 9e-06,
+      "loss": 1.7142,
+      "step": 6
+    },
+    {
+      "epoch": 0.003169841224917216,
+      "grad_norm": 2.308230400085449,
+      "learning_rate": 1.05e-05,
+      "loss": 1.7518,
+      "step": 7
+    },
+    {
+      "epoch": 0.0036226756856196757,
+      "grad_norm": 1.8725852966308594,
+      "learning_rate": 1.2e-05,
+      "loss": 1.7033,
+      "step": 8
+    },
+    {
+      "epoch": 0.004075510146322135,
+      "grad_norm": 2.1712286472320557,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.6949,
+      "step": 9
+    },
+    {
+      "epoch": 0.004528344607024595,
+      "grad_norm": 1.3823119401931763,
+      "learning_rate": 1.5e-05,
+      "loss": 1.647,
+      "step": 10
+    },
+    {
+      "epoch": 0.004981179067727054,
+      "grad_norm": 1.3454489707946777,
+      "learning_rate": 1.4999999153937153e-05,
+      "loss": 1.6661,
+      "step": 11
+    },
+    {
+      "epoch": 0.005434013528429513,
+      "grad_norm": 1.0993516445159912,
+      "learning_rate": 1.4999996615748808e-05,
+      "loss": 1.5813,
+      "step": 12
+    },
+    {
+      "epoch": 0.005886847989131973,
+      "grad_norm": 1.3689382076263428,
+      "learning_rate": 1.4999992385435533e-05,
+      "loss": 1.5947,
+      "step": 13
+    },
+    {
+      "epoch": 0.006339682449834432,
+      "grad_norm": 1.1131012439727783,
+      "learning_rate": 1.4999986462998284e-05,
+      "loss": 1.6002,
+      "step": 14
+    },
+    {
+      "epoch": 0.006792516910536892,
+      "grad_norm": 1.140008807182312,
+      "learning_rate": 1.4999978848438399e-05,
+      "loss": 1.5594,
+      "step": 15
+    },
+    {
+      "epoch": 0.0072453513712393515,
+      "grad_norm": 0.921599268913269,
+      "learning_rate": 1.4999969541757593e-05,
+      "loss": 1.5514,
+      "step": 16
+    },
+    {
+      "epoch": 0.007698185831941811,
+      "grad_norm": 0.9127187132835388,
+      "learning_rate": 1.4999958542957966e-05,
+      "loss": 1.5352,
+      "step": 17
+    },
+    {
+      "epoch": 0.00815102029264427,
+      "grad_norm": 0.7846677303314209,
+      "learning_rate": 1.4999945852042004e-05,
+      "loss": 1.5188,
+      "step": 18
+    },
+    {
+      "epoch": 0.00860385475334673,
+      "grad_norm": 0.843215823173523,
+      "learning_rate": 1.4999931469012563e-05,
+      "loss": 1.5185,
+      "step": 19
+    },
+    {
+      "epoch": 0.00905668921404919,
+      "grad_norm": 0.8540645241737366,
+      "learning_rate": 1.4999915393872895e-05,
+      "loss": 1.5068,
+      "step": 20
+    },
+    {
+      "epoch": 0.00950952367475165,
+      "grad_norm": 0.7134804725646973,
+      "learning_rate": 1.4999897626626621e-05,
+      "loss": 1.515,
+      "step": 21
+    },
+    {
+      "epoch": 0.009962358135454107,
+      "grad_norm": 0.870681643486023,
+      "learning_rate": 1.4999878167277757e-05,
+      "loss": 1.4962,
+      "step": 22
+    },
+    {
+      "epoch": 0.010415192596156567,
+      "grad_norm": 0.7976385951042175,
+      "learning_rate": 1.4999857015830684e-05,
+      "loss": 1.4902,
+      "step": 23
+    },
+    {
+      "epoch": 0.010868027056859027,
+      "grad_norm": 0.6572742462158203,
+      "learning_rate": 1.4999834172290182e-05,
+      "loss": 1.4845,
+      "step": 24
+    },
+    {
+      "epoch": 0.011320861517561486,
+      "grad_norm": 0.7284879684448242,
+      "learning_rate": 1.49998096366614e-05,
+      "loss": 1.4778,
+      "step": 25
+    },
+    {
+      "epoch": 0.011773695978263946,
+      "grad_norm": 0.7442657351493835,
+      "learning_rate": 1.4999783408949875e-05,
+      "loss": 1.4279,
+      "step": 26
+    },
+    {
+      "epoch": 0.012226530438966406,
+      "grad_norm": 0.7158825993537903,
+      "learning_rate": 1.4999755489161527e-05,
+      "loss": 1.4394,
+      "step": 27
+    },
+    {
+      "epoch": 0.012679364899668864,
+      "grad_norm": 0.6678286790847778,
+      "learning_rate": 1.4999725877302652e-05,
+      "loss": 1.4384,
+      "step": 28
+    },
+    {
+      "epoch": 0.013132199360371324,
+      "grad_norm": 0.6753094792366028,
+      "learning_rate": 1.4999694573379931e-05,
+      "loss": 1.4938,
+      "step": 29
+    },
+    {
+      "epoch": 0.013585033821073783,
+      "grad_norm": 0.8098469972610474,
+      "learning_rate": 1.499966157740043e-05,
+      "loss": 1.413,
+      "step": 30
+    },
+    {
+      "epoch": 0.014037868281776243,
+      "grad_norm": 0.6817219257354736,
+      "learning_rate": 1.4999626889371588e-05,
+      "loss": 1.4387,
+      "step": 31
+    },
+    {
+      "epoch": 0.014490702742478703,
+      "grad_norm": 0.6587890982627869,
+      "learning_rate": 1.4999590509301237e-05,
+      "loss": 1.3972,
+      "step": 32
+    },
+    {
+      "epoch": 0.014943537203181163,
+      "grad_norm": 0.6853736042976379,
+      "learning_rate": 1.4999552437197582e-05,
+      "loss": 1.4371,
+      "step": 33
+    },
+    {
+      "epoch": 0.015396371663883622,
+      "grad_norm": 0.6654878854751587,
+      "learning_rate": 1.4999512673069211e-05,
+      "loss": 1.3717,
+      "step": 34
+    },
+    {
+      "epoch": 0.015849206124586082,
+      "grad_norm": 0.6802534461021423,
+      "learning_rate": 1.4999471216925097e-05,
+      "loss": 1.3965,
+      "step": 35
+    },
+    {
+      "epoch": 0.01630204058528854,
+      "grad_norm": 0.638325035572052,
+      "learning_rate": 1.4999428068774597e-05,
+      "loss": 1.3658,
+      "step": 36
+    },
+    {
+      "epoch": 0.016754875045991,
+      "grad_norm": 0.6029091477394104,
+      "learning_rate": 1.499938322862744e-05,
+      "loss": 1.3593,
+      "step": 37
+    },
+    {
+      "epoch": 0.01720770950669346,
+      "grad_norm": 0.6502748727798462,
+      "learning_rate": 1.4999336696493746e-05,
+      "loss": 1.3406,
+      "step": 38
+    },
+    {
+      "epoch": 0.017660543967395918,
+      "grad_norm": 0.5874539613723755,
+      "learning_rate": 1.4999288472384011e-05,
+      "loss": 1.3959,
+      "step": 39
+    },
+    {
+      "epoch": 0.01811337842809838,
+      "grad_norm": 0.5745269060134888,
+      "learning_rate": 1.499923855630912e-05,
+      "loss": 1.3568,
+      "step": 40
+    },
+    {
+      "epoch": 0.018566212888800837,
+      "grad_norm": 0.6025118827819824,
+      "learning_rate": 1.4999186948280329e-05,
+      "loss": 1.3679,
+      "step": 41
+    },
+    {
+      "epoch": 0.0190190473495033,
+      "grad_norm": 0.6154868006706238,
+      "learning_rate": 1.4999133648309286e-05,
+      "loss": 1.362,
+      "step": 42
+    },
+    {
+      "epoch": 0.019471881810205757,
+      "grad_norm": 0.5797818303108215,
+      "learning_rate": 1.4999078656408013e-05,
+      "loss": 1.3357,
+      "step": 43
+    },
+    {
+      "epoch": 0.019924716270908215,
+      "grad_norm": 0.6571888327598572,
+      "learning_rate": 1.499902197258892e-05,
+      "loss": 1.3554,
+      "step": 44
+    },
+    {
+      "epoch": 0.020377550731610676,
+      "grad_norm": 0.6282890439033508,
+      "learning_rate": 1.4998963596864794e-05,
+      "loss": 1.3436,
+      "step": 45
+    },
+    {
+      "epoch": 0.020830385192313134,
+      "grad_norm": 0.541675865650177,
+      "learning_rate": 1.4998903529248806e-05,
+      "loss": 1.3465,
+      "step": 46
+    },
+    {
+      "epoch": 0.021283219653015596,
+      "grad_norm": 0.629550576210022,
+      "learning_rate": 1.499884176975451e-05,
+      "loss": 1.3405,
+      "step": 47
+    },
+    {
+      "epoch": 0.021736054113718054,
+      "grad_norm": 0.6273293495178223,
+      "learning_rate": 1.4998778318395837e-05,
+      "loss": 1.3425,
+      "step": 48
+    },
+    {
+      "epoch": 0.022188888574420515,
+      "grad_norm": 0.5797371864318848,
+      "learning_rate": 1.4998713175187105e-05,
+      "loss": 1.3066,
+      "step": 49
+    },
+    {
+      "epoch": 0.022641723035122973,
+      "grad_norm": 0.6254112720489502,
+      "learning_rate": 1.499864634014301e-05,
+      "loss": 1.287,
+      "step": 50
+    },
+    {
+      "epoch": 0.02309455749582543,
+      "grad_norm": 0.5634319186210632,
+      "learning_rate": 1.4998577813278631e-05,
+      "loss": 1.3163,
+      "step": 51
+    },
+    {
+      "epoch": 0.023547391956527892,
+      "grad_norm": 0.634515106678009,
+      "learning_rate": 1.4998507594609432e-05,
+      "loss": 1.293,
+      "step": 52
+    },
+    {
+      "epoch": 0.02400022641723035,
+      "grad_norm": 0.6572280526161194,
+      "learning_rate": 1.499843568415125e-05,
+      "loss": 1.3049,
+      "step": 53
+    },
+    {
+      "epoch": 0.024453060877932812,
+      "grad_norm": 0.627619206905365,
+      "learning_rate": 1.4998362081920313e-05,
+      "loss": 1.293,
+      "step": 54
+    },
+    {
+      "epoch": 0.02490589533863527,
+      "grad_norm": 0.5970916152000427,
+      "learning_rate": 1.4998286787933226e-05,
+      "loss": 1.3552,
+      "step": 55
+    },
+    {
+      "epoch": 0.025358729799337728,
+      "grad_norm": 0.5540199875831604,
+      "learning_rate": 1.4998209802206977e-05,
+      "loss": 1.3124,
+      "step": 56
+    },
+    {
+      "epoch": 0.02581156426004019,
+      "grad_norm": 0.5490045547485352,
+      "learning_rate": 1.4998131124758935e-05,
+      "loss": 1.2801,
+      "step": 57
+    },
+    {
+      "epoch": 0.026264398720742647,
+      "grad_norm": 0.6576207280158997,
+      "learning_rate": 1.4998050755606851e-05,
+      "loss": 1.2807,
+      "step": 58
+    },
+    {
+      "epoch": 0.02671723318144511,
+      "grad_norm": 0.617120623588562,
+      "learning_rate": 1.4997968694768856e-05,
+      "loss": 1.2434,
+      "step": 59
+    },
+    {
+      "epoch": 0.027170067642147567,
+      "grad_norm": 0.6149354577064514,
+      "learning_rate": 1.4997884942263468e-05,
+      "loss": 1.2947,
+      "step": 60
+    },
+    {
+      "epoch": 0.02762290210285003,
+      "grad_norm": 0.6856628060340881,
+      "learning_rate": 1.499779949810958e-05,
+      "loss": 1.304,
+      "step": 61
+    },
+    {
+      "epoch": 0.028075736563552486,
+      "grad_norm": 0.5651207566261292,
+      "learning_rate": 1.4997712362326468e-05,
+      "loss": 1.2328,
+      "step": 62
+    },
+    {
+      "epoch": 0.028528571024254944,
+      "grad_norm": 0.6854087114334106,
+      "learning_rate": 1.4997623534933796e-05,
+      "loss": 1.2766,
+      "step": 63
+    },
+    {
+      "epoch": 0.028981405484957406,
+      "grad_norm": 0.609596848487854,
+      "learning_rate": 1.4997533015951603e-05,
+      "loss": 1.2682,
+      "step": 64
+    },
+    {
+      "epoch": 0.029434239945659864,
+      "grad_norm": 0.575474202632904,
+      "learning_rate": 1.499744080540031e-05,
+      "loss": 1.1962,
+      "step": 65
+    },
+    {
+      "epoch": 0.029887074406362325,
+      "grad_norm": 0.6126519441604614,
+      "learning_rate": 1.4997346903300723e-05,
+      "loss": 1.2239,
+      "step": 66
+    },
+    {
+      "epoch": 0.030339908867064783,
+      "grad_norm": 0.5966687798500061,
+      "learning_rate": 1.4997251309674028e-05,
+      "loss": 1.2515,
+      "step": 67
+    },
+    {
+      "epoch": 0.030792743327767245,
+      "grad_norm": 0.7234688997268677,
+      "learning_rate": 1.499715402454179e-05,
+      "loss": 1.2173,
+      "step": 68
+    },
+    {
+      "epoch": 0.031245577788469703,
+      "grad_norm": 0.5534050464630127,
+      "learning_rate": 1.4997055047925962e-05,
+      "loss": 1.2269,
+      "step": 69
+    },
+    {
+      "epoch": 0.031698412249172164,
+      "grad_norm": 0.7033390402793884,
+      "learning_rate": 1.499695437984887e-05,
+      "loss": 1.2371,
+      "step": 70
+    },
+    {
+      "epoch": 0.03215124670987462,
+      "grad_norm": 0.7099838852882385,
+      "learning_rate": 1.4996852020333232e-05,
+      "loss": 1.2639,
+      "step": 71
+    },
+    {
+      "epoch": 0.03260408117057708,
+      "grad_norm": 0.5796670913696289,
+      "learning_rate": 1.4996747969402139e-05,
+      "loss": 1.2291,
+      "step": 72
+    },
+    {
+      "epoch": 0.03305691563127954,
+      "grad_norm": 0.6576344966888428,
+      "learning_rate": 1.4996642227079065e-05,
+      "loss": 1.1869,
+      "step": 73
+    },
+    {
+      "epoch": 0.033509750091982,
+      "grad_norm": 0.6396910548210144,
+      "learning_rate": 1.4996534793387871e-05,
+      "loss": 1.2227,
+      "step": 74
+    },
+    {
+      "epoch": 0.03396258455268446,
+      "grad_norm": 0.6648237109184265,
+      "learning_rate": 1.4996425668352793e-05,
+      "loss": 1.2715,
+      "step": 75
+    },
+    {
+      "epoch": 0.03441541901338692,
+      "grad_norm": 0.6780776381492615,
+      "learning_rate": 1.4996314851998453e-05,
+      "loss": 1.226,
+      "step": 76
+    },
+    {
+      "epoch": 0.03486825347408938,
+      "grad_norm": 0.5589417815208435,
+      "learning_rate": 1.4996202344349853e-05,
+      "loss": 1.2053,
+      "step": 77
+    },
+    {
+      "epoch": 0.035321087934791835,
+      "grad_norm": 0.5351216793060303,
+      "learning_rate": 1.4996088145432375e-05,
+      "loss": 1.2103,
+      "step": 78
+    },
+    {
+      "epoch": 0.0357739223954943,
+      "grad_norm": 0.4985811114311218,
+      "learning_rate": 1.4995972255271787e-05,
+      "loss": 1.2327,
+      "step": 79
+    },
+    {
+      "epoch": 0.03622675685619676,
+      "grad_norm": 0.5302553176879883,
+      "learning_rate": 1.499585467389423e-05,
+      "loss": 1.1468,
+      "step": 80
+    },
+    {
+      "epoch": 0.03667959131689921,
+      "grad_norm": 0.6337281465530396,
+      "learning_rate": 1.499573540132624e-05,
+      "loss": 1.2117,
+      "step": 81
+    },
+    {
+      "epoch": 0.037132425777601674,
+      "grad_norm": 0.8564434051513672,
+      "learning_rate": 1.4995614437594721e-05,
+      "loss": 1.1857,
+      "step": 82
+    },
+    {
+      "epoch": 0.037585260238304136,
+      "grad_norm": 0.7630729675292969,
+      "learning_rate": 1.4995491782726968e-05,
+      "loss": 1.2389,
+      "step": 83
+    },
+    {
+      "epoch": 0.0380380946990066,
+      "grad_norm": 0.7095062732696533,
+      "learning_rate": 1.499536743675065e-05,
+      "loss": 1.162,
+      "step": 84
+    },
+    {
+      "epoch": 0.03849092915970905,
+      "grad_norm": 0.7915695309638977,
+      "learning_rate": 1.4995241399693827e-05,
+      "loss": 1.2049,
+      "step": 85
+    },
+    {
+      "epoch": 0.03894376362041151,
+      "grad_norm": 0.6757071018218994,
+      "learning_rate": 1.4995113671584933e-05,
+      "loss": 1.2344,
+      "step": 86
+    },
+    {
+      "epoch": 0.039396598081113975,
+      "grad_norm": 0.8018056750297546,
+      "learning_rate": 1.4994984252452782e-05,
+      "loss": 1.1507,
+      "step": 87
+    },
+    {
+      "epoch": 0.03984943254181643,
+      "grad_norm": 0.8417953848838806,
+      "learning_rate": 1.4994853142326578e-05,
+      "loss": 1.1799,
+      "step": 88
+    },
+    {
+      "epoch": 0.04030226700251889,
+      "grad_norm": 0.7193191051483154,
+      "learning_rate": 1.4994720341235898e-05,
+      "loss": 1.1755,
+      "step": 89
+    },
+    {
+      "epoch": 0.04075510146322135,
+      "grad_norm": 1.2904716730117798,
+      "learning_rate": 1.4994585849210707e-05,
+      "loss": 1.1643,
+      "step": 90
+    },
+    {
+      "epoch": 0.041207935923923814,
+      "grad_norm": 0.7058910727500916,
+      "learning_rate": 1.4994449666281348e-05,
+      "loss": 1.1737,
+      "step": 91
+    },
+    {
+      "epoch": 0.04166077038462627,
+      "grad_norm": 0.9418469667434692,
+      "learning_rate": 1.4994311792478543e-05,
+      "loss": 1.1885,
+      "step": 92
+    },
+    {
+      "epoch": 0.04211360484532873,
+      "grad_norm": 0.8901066780090332,
+      "learning_rate": 1.4994172227833402e-05,
+      "loss": 1.1702,
+      "step": 93
+    },
+    {
+      "epoch": 0.04256643930603119,
+      "grad_norm": 0.5510401129722595,
+      "learning_rate": 1.4994030972377414e-05,
+      "loss": 1.1792,
+      "step": 94
+    },
+    {
+      "epoch": 0.043019273766733646,
+      "grad_norm": 0.7584707140922546,
+      "learning_rate": 1.4993888026142448e-05,
+      "loss": 1.1627,
+      "step": 95
+    },
+    {
+      "epoch": 0.04347210822743611,
+      "grad_norm": 0.7134795188903809,
+      "learning_rate": 1.499374338916075e-05,
+      "loss": 1.236,
+      "step": 96
+    },
+    {
+      "epoch": 0.04392494268813857,
+      "grad_norm": 0.6071385145187378,
+      "learning_rate": 1.499359706146496e-05,
+      "loss": 1.1651,
+      "step": 97
+    },
+    {
+      "epoch": 0.04437777714884103,
+      "grad_norm": 0.6142598390579224,
+      "learning_rate": 1.4993449043088088e-05,
+      "loss": 1.129,
+      "step": 98
+    },
+    {
+      "epoch": 0.044830611609543485,
+      "grad_norm": 0.713774561882019,
+      "learning_rate": 1.4993299334063528e-05,
+      "loss": 1.1959,
+      "step": 99
+    },
+    {
+      "epoch": 0.045283446070245946,
+      "grad_norm": 0.6220033764839172,
+      "learning_rate": 1.499314793442506e-05,
+      "loss": 1.1464,
+      "step": 100
+    },
+    {
+      "epoch": 0.04573628053094841,
+      "grad_norm": 0.9813457727432251,
+      "learning_rate": 1.4992994844206843e-05,
+      "loss": 1.1228,
+      "step": 101
+    },
+    {
+      "epoch": 0.04618911499165086,
+      "grad_norm": 0.9431173205375671,
+      "learning_rate": 1.499284006344341e-05,
+      "loss": 1.1478,
+      "step": 102
+    },
+    {
+      "epoch": 0.04664194945235332,
+      "grad_norm": 0.5895239114761353,
+      "learning_rate": 1.4992683592169691e-05,
+      "loss": 1.1035,
+      "step": 103
+    },
+    {
+      "epoch": 0.047094783913055785,
+      "grad_norm": 1.2074604034423828,
+      "learning_rate": 1.4992525430420984e-05,
+      "loss": 1.1472,
+      "step": 104
+    },
+    {
+      "epoch": 0.047547618373758246,
+      "grad_norm": 0.8212599158287048,
+      "learning_rate": 1.499236557823297e-05,
+      "loss": 1.1222,
+      "step": 105
+    },
+    {
+      "epoch": 0.0480004528344607,
+      "grad_norm": 0.7460811138153076,
+      "learning_rate": 1.4992204035641721e-05,
+      "loss": 1.1551,
+      "step": 106
+    },
+    {
+      "epoch": 0.04845328729516316,
+      "grad_norm": 0.9674639105796814,
+      "learning_rate": 1.499204080268368e-05,
+      "loss": 1.15,
+      "step": 107
+    },
+    {
+      "epoch": 0.048906121755865624,
+      "grad_norm": 0.7379394173622131,
+      "learning_rate": 1.4991875879395677e-05,
+      "loss": 1.096,
+      "step": 108
+    },
+    {
+      "epoch": 0.04935895621656808,
+      "grad_norm": 0.7826714515686035,
+      "learning_rate": 1.4991709265814918e-05,
+      "loss": 1.1086,
+      "step": 109
+    },
+    {
+      "epoch": 0.04981179067727054,
+      "grad_norm": 0.7406508326530457,
+      "learning_rate": 1.4991540961978997e-05,
+      "loss": 1.1203,
+      "step": 110
+    },
+    {
+      "epoch": 0.050264625137973,
+      "grad_norm": 0.6145086884498596,
+      "learning_rate": 1.4991370967925882e-05,
+      "loss": 1.1688,
+      "step": 111
+    },
+    {
+      "epoch": 0.050717459598675456,
+      "grad_norm": 0.7194088697433472,
+      "learning_rate": 1.4991199283693933e-05,
+      "loss": 1.0994,
+      "step": 112
+    },
+    {
+      "epoch": 0.05117029405937792,
+      "grad_norm": 0.6828657984733582,
+      "learning_rate": 1.499102590932188e-05,
+      "loss": 1.1362,
+      "step": 113
+    },
+    {
+      "epoch": 0.05162312852008038,
+      "grad_norm": 0.6402639150619507,
+      "learning_rate": 1.4990850844848841e-05,
+      "loss": 1.1146,
+      "step": 114
+    },
+    {
+      "epoch": 0.05207596298078284,
+      "grad_norm": 0.7166230082511902,
+      "learning_rate": 1.4990674090314313e-05,
+      "loss": 1.085,
+      "step": 115
+    },
+    {
+      "epoch": 0.052528797441485295,
+      "grad_norm": 0.6384372711181641,
+      "learning_rate": 1.4990495645758174e-05,
+      "loss": 1.0982,
+      "step": 116
+    },
+    {
+      "epoch": 0.052981631902187756,
+      "grad_norm": 0.6809221506118774,
+      "learning_rate": 1.4990315511220685e-05,
+      "loss": 1.1414,
+      "step": 117
+    },
+    {
+      "epoch": 0.05343446636289022,
+      "grad_norm": 0.6124834418296814,
+      "learning_rate": 1.4990133686742488e-05,
+      "loss": 1.0886,
+      "step": 118
+    },
+    {
+      "epoch": 0.05388730082359267,
+      "grad_norm": 0.7249141335487366,
+      "learning_rate": 1.4989950172364603e-05,
+      "loss": 1.1568,
+      "step": 119
+    },
+    {
+      "epoch": 0.054340135284295134,
+      "grad_norm": 0.6189077496528625,
+      "learning_rate": 1.498976496812844e-05,
+      "loss": 1.0731,
+      "step": 120
+    },
+    {
+      "epoch": 0.054792969744997595,
+      "grad_norm": 0.8025009036064148,
+      "learning_rate": 1.4989578074075777e-05,
+      "loss": 1.1252,
+      "step": 121
+    },
+    {
+      "epoch": 0.05524580420570006,
+      "grad_norm": 0.6181627511978149,
+      "learning_rate": 1.4989389490248783e-05,
+      "loss": 1.0713,
+      "step": 122
+    },
+    {
+      "epoch": 0.05569863866640251,
+      "grad_norm": 0.8665769696235657,
+      "learning_rate": 1.4989199216690006e-05,
+      "loss": 1.1481,
+      "step": 123
+    },
+    {
+      "epoch": 0.05615147312710497,
+      "grad_norm": 0.7832885384559631,
+      "learning_rate": 1.4989007253442377e-05,
+      "loss": 1.0775,
+      "step": 124
+    },
+    {
+      "epoch": 0.056604307587807434,
+      "grad_norm": 0.7431057095527649,
+      "learning_rate": 1.4988813600549202e-05,
+      "loss": 1.1013,
+      "step": 125
+    },
+    {
+      "epoch": 0.05705714204850989,
+      "grad_norm": 1.1883816719055176,
+      "learning_rate": 1.4988618258054176e-05,
+      "loss": 1.1003,
+      "step": 126
+    },
+    {
+      "epoch": 0.05750997650921235,
+      "grad_norm": 0.6886972188949585,
+      "learning_rate": 1.498842122600137e-05,
+      "loss": 1.0754,
+      "step": 127
+    },
+    {
+      "epoch": 0.05796281096991481,
+      "grad_norm": 0.8941016793251038,
+      "learning_rate": 1.4988222504435235e-05,
+      "loss": 1.0577,
+      "step": 128
+    },
+    {
+      "epoch": 0.05841564543061727,
+      "grad_norm": 0.9873454570770264,
+      "learning_rate": 1.498802209340061e-05,
+      "loss": 1.089,
+      "step": 129
+    },
+    {
+      "epoch": 0.05886847989131973,
+      "grad_norm": 0.6695548892021179,
+      "learning_rate": 1.4987819992942712e-05,
+      "loss": 1.0799,
+      "step": 130
+    },
+    {
+      "epoch": 0.05932131435202219,
+      "grad_norm": 0.9906266331672668,
+      "learning_rate": 1.4987616203107134e-05,
+      "loss": 1.0753,
+      "step": 131
+    },
+    {
+      "epoch": 0.05977414881272465,
+      "grad_norm": 0.6169360876083374,
+      "learning_rate": 1.4987410723939857e-05,
+      "loss": 1.0863,
+      "step": 132
+    },
+    {
+      "epoch": 0.060226983273427105,
+      "grad_norm": 1.0210751295089722,
+      "learning_rate": 1.4987203555487242e-05,
+      "loss": 1.0493,
+      "step": 133
+    },
+    {
+      "epoch": 0.06067981773412957,
+      "grad_norm": 0.675287663936615,
+      "learning_rate": 1.4986994697796022e-05,
+      "loss": 1.0669,
+      "step": 134
+    },
+    {
+      "epoch": 0.06113265219483203,
+      "grad_norm": 0.8825522661209106,
+      "learning_rate": 1.4986784150913329e-05,
+      "loss": 1.0482,
+      "step": 135
+    },
+    {
+      "epoch": 0.06158548665553449,
+      "grad_norm": 0.6397138237953186,
+      "learning_rate": 1.4986571914886662e-05,
+      "loss": 1.0544,
+      "step": 136
+    },
+    {
+      "epoch": 0.062038321116236944,
+      "grad_norm": 1.0741486549377441,
+      "learning_rate": 1.4986357989763901e-05,
+      "loss": 1.037,
+      "step": 137
+    },
+    {
+      "epoch": 0.062491155576939406,
+      "grad_norm": 0.6419923901557922,
+      "learning_rate": 1.4986142375593315e-05,
+      "loss": 1.0751,
+      "step": 138
+    },
+    {
+      "epoch": 0.06294399003764187,
+      "grad_norm": 0.8368469476699829,
+      "learning_rate": 1.498592507242355e-05,
+      "loss": 1.0357,
+      "step": 139
+    },
+    {
+      "epoch": 0.06339682449834433,
+      "grad_norm": 0.7393881678581238,
+      "learning_rate": 1.4985706080303634e-05,
+      "loss": 1.0733,
+      "step": 140
+    },
+    {
+      "epoch": 0.06384965895904679,
+      "grad_norm": 0.8628896474838257,
+      "learning_rate": 1.4985485399282973e-05,
+      "loss": 1.0808,
+      "step": 141
+    },
+    {
+      "epoch": 0.06430249341974924,
+      "grad_norm": 0.6672838926315308,
+      "learning_rate": 1.4985263029411356e-05,
+      "loss": 1.0266,
+      "step": 142
+    },
+    {
+      "epoch": 0.0647553278804517,
+      "grad_norm": 0.7643623352050781,
+      "learning_rate": 1.4985038970738959e-05,
+      "loss": 1.0044,
+      "step": 143
+    },
+    {
+      "epoch": 0.06520816234115416,
+      "grad_norm": 0.8692808747291565,
+      "learning_rate": 1.4984813223316326e-05,
+      "loss": 1.0128,
+      "step": 144
+    },
+    {
+      "epoch": 0.06566099680185662,
+      "grad_norm": 0.7556530833244324,
+      "learning_rate": 1.4984585787194392e-05,
+      "loss": 1.0772,
+      "step": 145
+    },
+    {
+      "epoch": 0.06611383126255908,
+      "grad_norm": 0.8598385453224182,
+      "learning_rate": 1.4984356662424473e-05,
+      "loss": 1.0511,
+      "step": 146
+    },
+    {
+      "epoch": 0.06656666572326154,
+      "grad_norm": 0.6620029807090759,
+      "learning_rate": 1.498412584905826e-05,
+      "loss": 1.0401,
+      "step": 147
+    },
+    {
+      "epoch": 0.067019500183964,
+      "grad_norm": 0.8549764752388,
+      "learning_rate": 1.498389334714783e-05,
+      "loss": 1.0172,
+      "step": 148
+    },
+    {
+      "epoch": 0.06747233464466645,
+      "grad_norm": 0.6814466714859009,
+      "learning_rate": 1.498365915674564e-05,
+      "loss": 1.0638,
+      "step": 149
+    },
+    {
+      "epoch": 0.06792516910536892,
+      "grad_norm": 0.7864624261856079,
+      "learning_rate": 1.4983423277904526e-05,
+      "loss": 1.0069,
+      "step": 150
+    },
+    {
+      "epoch": 0.06837800356607138,
+      "grad_norm": 0.7862728238105774,
+      "learning_rate": 1.4983185710677705e-05,
+      "loss": 1.0092,
+      "step": 151
+    },
+    {
+      "epoch": 0.06883083802677384,
+      "grad_norm": 0.656066358089447,
+      "learning_rate": 1.498294645511878e-05,
+      "loss": 0.9948,
+      "step": 152
+    },
+    {
+      "epoch": 0.0692836724874763,
+      "grad_norm": 0.8510375022888184,
+      "learning_rate": 1.4982705511281728e-05,
+      "loss": 1.0411,
+      "step": 153
+    },
+    {
+      "epoch": 0.06973650694817876,
+      "grad_norm": 0.6606243848800659,
+      "learning_rate": 1.4982462879220911e-05,
+      "loss": 1.0054,
+      "step": 154
+    },
+    {
+      "epoch": 0.07018934140888121,
+      "grad_norm": 0.7300797700881958,
+      "learning_rate": 1.498221855899107e-05,
+      "loss": 0.9801,
+      "step": 155
+    },
+    {
+      "epoch": 0.07064217586958367,
+      "grad_norm": 0.8046183586120605,
+      "learning_rate": 1.4981972550647328e-05,
+      "loss": 0.9794,
+      "step": 156
+    },
+    {
+      "epoch": 0.07109501033028613,
+      "grad_norm": 0.6511421203613281,
+      "learning_rate": 1.4981724854245189e-05,
+      "loss": 1.065,
+      "step": 157
+    },
+    {
+      "epoch": 0.0715478447909886,
+      "grad_norm": 0.8642223477363586,
+      "learning_rate": 1.4981475469840538e-05,
+      "loss": 0.9808,
+      "step": 158
+    },
+    {
+      "epoch": 0.07200067925169105,
+      "grad_norm": 0.7534111738204956,
+      "learning_rate": 1.498122439748964e-05,
+      "loss": 1.0371,
+      "step": 159
+    },
+    {
+      "epoch": 0.07245351371239352,
+      "grad_norm": 0.7241458892822266,
+      "learning_rate": 1.4980971637249141e-05,
+      "loss": 0.9568,
+      "step": 160
+    },
+    {
+      "epoch": 0.07290634817309598,
+      "grad_norm": 0.6701076030731201,
+      "learning_rate": 1.4980717189176066e-05,
+      "loss": 1.0158,
+      "step": 161
+    },
+    {
+      "epoch": 0.07335918263379843,
+      "grad_norm": 0.6040867567062378,
+      "learning_rate": 1.4980461053327829e-05,
+      "loss": 0.9966,
+      "step": 162
+    },
+    {
+      "epoch": 0.07381201709450089,
+      "grad_norm": 0.7389962077140808,
+      "learning_rate": 1.4980203229762208e-05,
+      "loss": 0.962,
+      "step": 163
+    },
+    {
+      "epoch": 0.07426485155520335,
+      "grad_norm": 0.6864858269691467,
+      "learning_rate": 1.4979943718537383e-05,
+      "loss": 1.0563,
+      "step": 164
+    },
+    {
+      "epoch": 0.07471768601590581,
+      "grad_norm": 0.6471516489982605,
+      "learning_rate": 1.4979682519711897e-05,
+      "loss": 1.0448,
+      "step": 165
+    },
+    {
+      "epoch": 0.07517052047660827,
+      "grad_norm": 0.7508234977722168,
+      "learning_rate": 1.4979419633344686e-05,
+      "loss": 1.0353,
+      "step": 166
+    },
+    {
+      "epoch": 0.07562335493731073,
+      "grad_norm": 0.6425490379333496,
+      "learning_rate": 1.4979155059495056e-05,
+      "loss": 0.9819,
+      "step": 167
+    },
+    {
+      "epoch": 0.0760761893980132,
+      "grad_norm": 0.6431488990783691,
+      "learning_rate": 1.4978888798222703e-05,
+      "loss": 1.0394,
+      "step": 168
+    },
+    {
+      "epoch": 0.07652902385871564,
+      "grad_norm": 0.5777195692062378,
+      "learning_rate": 1.49786208495877e-05,
+      "loss": 0.9881,
+      "step": 169
+    },
+    {
+      "epoch": 0.0769818583194181,
+      "grad_norm": 0.7131640315055847,
+      "learning_rate": 1.4978351213650498e-05,
+      "loss": 1.0054,
+      "step": 170
+    },
+    {
+      "epoch": 0.07743469278012056,
+      "grad_norm": 0.7139365673065186,
+      "learning_rate": 1.4978079890471935e-05,
+      "loss": 1.0341,
+      "step": 171
+    },
+    {
+      "epoch": 0.07788752724082303,
+      "grad_norm": 0.6020310521125793,
+      "learning_rate": 1.4977806880113223e-05,
+      "loss": 0.9785,
+      "step": 172
+    },
+    {
+      "epoch": 0.07834036170152549,
+      "grad_norm": 0.7464612126350403,
+      "learning_rate": 1.4977532182635963e-05,
+      "loss": 0.9976,
+      "step": 173
+    },
+    {
+      "epoch": 0.07879319616222795,
+      "grad_norm": 0.8518685698509216,
+      "learning_rate": 1.4977255798102122e-05,
+      "loss": 0.9976,
+      "step": 174
+    },
+    {
+      "epoch": 0.07924603062293041,
+      "grad_norm": 0.6868304014205933,
+      "learning_rate": 1.4976977726574065e-05,
+      "loss": 0.9989,
+      "step": 175
+    },
+    {
+      "epoch": 0.07969886508363286,
+      "grad_norm": 0.7775158882141113,
+      "learning_rate": 1.4976697968114529e-05,
+      "loss": 0.9655,
+      "step": 176
+    },
+    {
+      "epoch": 0.08015169954433532,
+      "grad_norm": 0.9228640794754028,
+      "learning_rate": 1.4976416522786626e-05,
+      "loss": 0.9497,
+      "step": 177
+    },
+    {
+      "epoch": 0.08060453400503778,
+      "grad_norm": 0.6497952342033386,
+      "learning_rate": 1.4976133390653861e-05,
+      "loss": 0.94,
+      "step": 178
+    },
+    {
+      "epoch": 0.08105736846574024,
+      "grad_norm": 0.853291928768158,
+      "learning_rate": 1.497584857178011e-05,
+      "loss": 0.9267,
+      "step": 179
+    },
+    {
+      "epoch": 0.0815102029264427,
+      "grad_norm": 1.0284061431884766,
+      "learning_rate": 1.4975562066229635e-05,
+      "loss": 0.9772,
+      "step": 180
+    },
+    {
+      "epoch": 0.08196303738714517,
+      "grad_norm": 0.7046072483062744,
+      "learning_rate": 1.4975273874067078e-05,
+      "loss": 1.0039,
+      "step": 181
+    },
+    {
+      "epoch": 0.08241587184784763,
+      "grad_norm": 0.8511521220207214,
+      "learning_rate": 1.4974983995357456e-05,
+      "loss": 0.935,
+      "step": 182
+    },
+    {
+      "epoch": 0.08286870630855007,
+      "grad_norm": 0.9605690240859985,
+      "learning_rate": 1.497469243016617e-05,
+      "loss": 0.9885,
+      "step": 183
+    },
+    {
+      "epoch": 0.08332154076925254,
+      "grad_norm": 0.7198604345321655,
+      "learning_rate": 1.4974399178559007e-05,
+      "loss": 0.9757,
+      "step": 184
+    },
+    {
+      "epoch": 0.083774375229955,
+      "grad_norm": 0.8301750421524048,
+      "learning_rate": 1.4974104240602127e-05,
+      "loss": 0.9232,
+      "step": 185
+    },
+    {
+      "epoch": 0.08422720969065746,
+      "grad_norm": 0.8821402788162231,
+      "learning_rate": 1.497380761636207e-05,
+      "loss": 0.9657,
+      "step": 186
+    },
+    {
+      "epoch": 0.08468004415135992,
+      "grad_norm": 1.1182292699813843,
+      "learning_rate": 1.4973509305905762e-05,
+      "loss": 1.005,
+      "step": 187
+    },
+    {
+      "epoch": 0.08513287861206238,
+      "grad_norm": 0.7202256917953491,
+      "learning_rate": 1.4973209309300508e-05,
+      "loss": 0.916,
+      "step": 188
+    },
+    {
+      "epoch": 0.08558571307276484,
+      "grad_norm": 1.0314316749572754,
+      "learning_rate": 1.4972907626613993e-05,
+      "loss": 0.9722,
+      "step": 189
+    },
+    {
+      "epoch": 0.08603854753346729,
+      "grad_norm": 0.8789144158363342,
+      "learning_rate": 1.4972604257914277e-05,
+      "loss": 0.9835,
+      "step": 190
+    },
+    {
+      "epoch": 0.08649138199416975,
+      "grad_norm": 0.8964241743087769,
+      "learning_rate": 1.497229920326981e-05,
+      "loss": 0.9562,
+      "step": 191
+    },
+    {
+      "epoch": 0.08694421645487221,
+      "grad_norm": 0.9407457709312439,
+      "learning_rate": 1.4971992462749413e-05,
+      "loss": 0.935,
+      "step": 192
+    },
+    {
+      "epoch": 0.08739705091557468,
+      "grad_norm": 0.7482190132141113,
+      "learning_rate": 1.4971684036422295e-05,
+      "loss": 0.9561,
+      "step": 193
+    },
+    {
+      "epoch": 0.08784988537627714,
+      "grad_norm": 0.7358856201171875,
+      "learning_rate": 1.4971373924358041e-05,
+      "loss": 0.9106,
+      "step": 194
+    },
+    {
+      "epoch": 0.0883027198369796,
+      "grad_norm": 0.7911369204521179,
+      "learning_rate": 1.4971062126626617e-05,
+      "loss": 0.9625,
+      "step": 195
+    },
+    {
+      "epoch": 0.08875555429768206,
+      "grad_norm": 0.6739444136619568,
+      "learning_rate": 1.497074864329837e-05,
+      "loss": 0.936,
+      "step": 196
+    },
+    {
+      "epoch": 0.08920838875838451,
+      "grad_norm": 0.7796293497085571,
+      "learning_rate": 1.497043347444403e-05,
+      "loss": 0.9645,
+      "step": 197
+    },
+    {
+      "epoch": 0.08966122321908697,
+      "grad_norm": 0.7281574010848999,
+      "learning_rate": 1.4970116620134701e-05,
+      "loss": 0.985,
+      "step": 198
+    },
+    {
+      "epoch": 0.09011405767978943,
+      "grad_norm": 0.7205932140350342,
+      "learning_rate": 1.4969798080441872e-05,
+      "loss": 0.9253,
+      "step": 199
+    },
+    {
+      "epoch": 0.09056689214049189,
+      "grad_norm": 0.8215734958648682,
+      "learning_rate": 1.496947785543741e-05,
+      "loss": 0.9391,
+      "step": 200
+    },
+    {
+      "epoch": 0.09101972660119435,
+      "grad_norm": 0.6455587148666382,
+      "learning_rate": 1.4969155945193562e-05,
+      "loss": 0.8646,
+      "step": 201
+    },
+    {
+      "epoch": 0.09147256106189681,
+      "grad_norm": 0.6836207509040833,
+      "learning_rate": 1.496883234978296e-05,
+      "loss": 0.9728,
+      "step": 202
+    },
+    {
+      "epoch": 0.09192539552259928,
+      "grad_norm": 0.6182245016098022,
+      "learning_rate": 1.496850706927861e-05,
+      "loss": 0.8912,
+      "step": 203
+    },
+    {
+      "epoch": 0.09237822998330172,
+      "grad_norm": 1.1609197854995728,
+      "learning_rate": 1.4968180103753901e-05,
+      "loss": 0.9035,
+      "step": 204
+    },
+    {
+      "epoch": 0.09283106444400419,
+      "grad_norm": 0.7769509553909302,
+      "learning_rate": 1.4967851453282601e-05,
+      "loss": 0.8651,
+      "step": 205
+    },
+    {
+      "epoch": 0.09328389890470665,
+      "grad_norm": 0.8045849204063416,
+      "learning_rate": 1.4967521117938861e-05,
+      "loss": 0.9432,
+      "step": 206
+    },
+    {
+      "epoch": 0.09373673336540911,
+      "grad_norm": 0.747963547706604,
+      "learning_rate": 1.496718909779721e-05,
+      "loss": 0.9142,
+      "step": 207
+    },
+    {
+      "epoch": 0.09418956782611157,
+      "grad_norm": 0.8653061389923096,
+      "learning_rate": 1.4966855392932558e-05,
+      "loss": 0.9138,
+      "step": 208
+    },
+    {
+      "epoch": 0.09464240228681403,
+      "grad_norm": 0.7377287745475769,
+      "learning_rate": 1.4966520003420191e-05,
+      "loss": 0.9411,
+      "step": 209
+    },
+    {
+      "epoch": 0.09509523674751649,
+      "grad_norm": 0.8704388737678528,
+      "learning_rate": 1.496618292933578e-05,
+      "loss": 0.9223,
+      "step": 210
+    },
+    {
+      "epoch": 0.09554807120821894,
+      "grad_norm": 0.8722823262214661,
+      "learning_rate": 1.4965844170755376e-05,
+      "loss": 0.9261,
+      "step": 211
+    },
+    {
+      "epoch": 0.0960009056689214,
+      "grad_norm": 0.6991630792617798,
+      "learning_rate": 1.4965503727755408e-05,
+      "loss": 0.9053,
+      "step": 212
+    },
+    {
+      "epoch": 0.09645374012962386,
+      "grad_norm": 0.762243926525116,
+      "learning_rate": 1.4965161600412686e-05,
+      "loss": 0.8989,
+      "step": 213
+    },
+    {
+      "epoch": 0.09690657459032632,
+      "grad_norm": 0.6228227019309998,
+      "learning_rate": 1.4964817788804398e-05,
+      "loss": 0.897,
+      "step": 214
+    },
+    {
+      "epoch": 0.09735940905102879,
+      "grad_norm": 0.871157705783844,
+      "learning_rate": 1.4964472293008114e-05,
+      "loss": 0.9551,
+      "step": 215
+    },
+    {
+      "epoch": 0.09781224351173125,
+      "grad_norm": 0.7100203037261963,
+      "learning_rate": 1.4964125113101787e-05,
+      "loss": 0.9334,
+      "step": 216
+    },
+    {
+      "epoch": 0.0982650779724337,
+      "grad_norm": 0.6398739218711853,
+      "learning_rate": 1.4963776249163742e-05,
+      "loss": 0.9256,
+      "step": 217
+    },
+    {
+      "epoch": 0.09871791243313616,
+      "grad_norm": 0.7361502051353455,
+      "learning_rate": 1.496342570127269e-05,
+      "loss": 0.9207,
+      "step": 218
+    },
+    {
+      "epoch": 0.09917074689383862,
+      "grad_norm": 0.6598243713378906,
+      "learning_rate": 1.4963073469507722e-05,
+      "loss": 0.9029,
+      "step": 219
+    },
+    {
+      "epoch": 0.09962358135454108,
+      "grad_norm": 0.7405319213867188,
+      "learning_rate": 1.4962719553948306e-05,
+      "loss": 0.9661,
+      "step": 220
+    },
+    {
+      "epoch": 0.10007641581524354,
+      "grad_norm": 0.7439802289009094,
+      "learning_rate": 1.4962363954674294e-05,
+      "loss": 0.9133,
+      "step": 221
+    },
+    {
+      "epoch": 0.100529250275946,
+      "grad_norm": 0.7841500639915466,
+      "learning_rate": 1.4962006671765911e-05,
+      "loss": 0.8958,
+      "step": 222
+    },
+    {
+      "epoch": 0.10098208473664846,
+      "grad_norm": 0.8495052456855774,
+      "learning_rate": 1.4961647705303765e-05,
+      "loss": 0.9755,
+      "step": 223
+    },
+    {
+      "epoch": 0.10143491919735091,
+      "grad_norm": 0.7986276149749756,
+      "learning_rate": 1.4961287055368853e-05,
+      "loss": 0.994,
+      "step": 224
+    },
+    {
+      "epoch": 0.10188775365805337,
+      "grad_norm": 0.8611103892326355,
+      "learning_rate": 1.4960924722042536e-05,
+      "loss": 0.8454,
+      "step": 225
+    },
+    {
+      "epoch": 0.10234058811875583,
+      "grad_norm": 1.167007565498352,
+      "learning_rate": 1.4960560705406563e-05,
+      "loss": 0.9154,
+      "step": 226
+    },
+    {
+      "epoch": 0.1027934225794583,
+      "grad_norm": 0.7711225748062134,
+      "learning_rate": 1.4960195005543066e-05,
+      "loss": 0.8714,
+      "step": 227
+    },
+    {
+      "epoch": 0.10324625704016076,
+      "grad_norm": 0.7879083752632141,
+      "learning_rate": 1.4959827622534551e-05,
+      "loss": 0.8909,
+      "step": 228
+    },
+    {
+      "epoch": 0.10369909150086322,
+      "grad_norm": 0.964492678642273,
+      "learning_rate": 1.4959458556463905e-05,
+      "loss": 0.8881,
+      "step": 229
+    },
+    {
+      "epoch": 0.10415192596156568,
+      "grad_norm": 1.0750067234039307,
+      "learning_rate": 1.4959087807414397e-05,
+      "loss": 0.8857,
+      "step": 230
+    },
+    {
+      "epoch": 0.10460476042226813,
+      "grad_norm": 0.7446492910385132,
+      "learning_rate": 1.4958715375469674e-05,
+      "loss": 0.9002,
+      "step": 231
+    },
+    {
+      "epoch": 0.10505759488297059,
+      "grad_norm": 1.0399891138076782,
+      "learning_rate": 1.4958341260713762e-05,
+      "loss": 0.9695,
+      "step": 232
+    },
+    {
+      "epoch": 0.10551042934367305,
+      "grad_norm": 0.7288771867752075,
+      "learning_rate": 1.495796546323107e-05,
+      "loss": 0.9033,
+      "step": 233
+    },
+    {
+      "epoch": 0.10596326380437551,
+      "grad_norm": 1.0252315998077393,
+      "learning_rate": 1.495758798310638e-05,
+      "loss": 0.8462,
+      "step": 234
+    },
+    {
+      "epoch": 0.10641609826507797,
+      "grad_norm": 0.7485429644584656,
+      "learning_rate": 1.4957208820424859e-05,
+      "loss": 0.8822,
+      "step": 235
+    },
+    {
+      "epoch": 0.10686893272578044,
+      "grad_norm": 0.9481981992721558,
+      "learning_rate": 1.4956827975272054e-05,
+      "loss": 0.8892,
+      "step": 236
+    },
+    {
+      "epoch": 0.1073217671864829,
+      "grad_norm": 0.8089492917060852,
+      "learning_rate": 1.495644544773389e-05,
+      "loss": 0.8245,
+      "step": 237
+    },
+    {
+      "epoch": 0.10777460164718534,
+      "grad_norm": 0.9913555383682251,
+      "learning_rate": 1.4956061237896671e-05,
+      "loss": 0.8442,
+      "step": 238
+    },
+    {
+      "epoch": 0.1082274361078878,
+      "grad_norm": 0.8869263529777527,
+      "learning_rate": 1.4955675345847084e-05,
+      "loss": 0.9323,
+      "step": 239
+    },
+    {
+      "epoch": 0.10868027056859027,
+      "grad_norm": 0.8241413235664368,
+      "learning_rate": 1.4955287771672187e-05,
+      "loss": 0.8985,
+      "step": 240
+    },
+    {
+      "epoch": 0.10913310502929273,
+      "grad_norm": 0.7058445811271667,
+      "learning_rate": 1.4954898515459428e-05,
+      "loss": 0.817,
+      "step": 241
+    },
+    {
+      "epoch": 0.10958593948999519,
+      "grad_norm": 0.7939592599868774,
+      "learning_rate": 1.4954507577296628e-05,
+      "loss": 0.7496,
+      "step": 242
+    },
+    {
+      "epoch": 0.11003877395069765,
+      "grad_norm": 0.8440166115760803,
+      "learning_rate": 1.4954114957271988e-05,
+      "loss": 0.9368,
+      "step": 243
+    },
+    {
+      "epoch": 0.11049160841140011,
+      "grad_norm": 0.8787747621536255,
+      "learning_rate": 1.4953720655474092e-05,
+      "loss": 0.8562,
+      "step": 244
+    },
+    {
+      "epoch": 0.11094444287210256,
+      "grad_norm": 0.7784985899925232,
+      "learning_rate": 1.4953324671991897e-05,
+      "loss": 0.8718,
+      "step": 245
+    },
+    {
+      "epoch": 0.11139727733280502,
+      "grad_norm": 0.9408149719238281,
+      "learning_rate": 1.495292700691475e-05,
+      "loss": 0.9013,
+      "step": 246
+    },
+    {
+      "epoch": 0.11185011179350748,
+      "grad_norm": 0.8225075602531433,
+      "learning_rate": 1.4952527660332368e-05,
+      "loss": 0.8769,
+      "step": 247
+    },
+    {
+      "epoch": 0.11230294625420995,
+      "grad_norm": 0.7701689004898071,
+      "learning_rate": 1.4952126632334847e-05,
+      "loss": 0.8824,
+      "step": 248
+    },
+    {
+      "epoch": 0.11275578071491241,
+      "grad_norm": 0.7905251979827881,
+      "learning_rate": 1.4951723923012667e-05,
+      "loss": 0.8753,
+      "step": 249
+    },
+    {
+      "epoch": 0.11320861517561487,
+      "grad_norm": 0.7921510338783264,
+      "learning_rate": 1.495131953245669e-05,
+      "loss": 0.8127,
+      "step": 250
+    },
+    {
+      "epoch": 0.11366144963631733,
+      "grad_norm": 0.8098307251930237,
+      "learning_rate": 1.495091346075815e-05,
+      "loss": 0.9178,
+      "step": 251
+    },
+    {
+      "epoch": 0.11411428409701978,
+      "grad_norm": 0.789347767829895,
+      "learning_rate": 1.4950505708008665e-05,
+      "loss": 0.8268,
+      "step": 252
+    },
+    {
+      "epoch": 0.11456711855772224,
+      "grad_norm": 0.8333251476287842,
+      "learning_rate": 1.495009627430023e-05,
+      "loss": 0.8422,
+      "step": 253
+    },
+    {
+      "epoch": 0.1150199530184247,
+      "grad_norm": 0.9456163048744202,
+      "learning_rate": 1.4949685159725219e-05,
+      "loss": 0.8445,
+      "step": 254
+    },
+    {
+      "epoch": 0.11547278747912716,
+      "grad_norm": 0.7319315075874329,
+      "learning_rate": 1.4949272364376389e-05,
+      "loss": 0.8195,
+      "step": 255
+    },
+    {
+      "epoch": 0.11592562193982962,
+      "grad_norm": 0.9746232628822327,
+      "learning_rate": 1.494885788834687e-05,
+      "loss": 0.8909,
+      "step": 256
+    },
+    {
+      "epoch": 0.11637845640053208,
+      "grad_norm": 0.9130688905715942,
+      "learning_rate": 1.4948441731730177e-05,
+      "loss": 0.8685,
+      "step": 257
+    },
+    {
+      "epoch": 0.11683129086123455,
+      "grad_norm": 0.7087278962135315,
+      "learning_rate": 1.4948023894620201e-05,
+      "loss": 0.9413,
+      "step": 258
+    },
+    {
+      "epoch": 0.117284125321937,
+      "grad_norm": 0.9176363348960876,
+      "learning_rate": 1.4947604377111216e-05,
+      "loss": 0.8186,
+      "step": 259
+    },
+    {
+      "epoch": 0.11773695978263946,
+      "grad_norm": 0.8463951945304871,
+      "learning_rate": 1.494718317929787e-05,
+      "loss": 0.8986,
+      "step": 260
+    },
+    {
+      "epoch": 0.11818979424334192,
+      "grad_norm": 0.8302780389785767,
+      "learning_rate": 1.494676030127519e-05,
+      "loss": 0.8204,
+      "step": 261
+    },
+    {
+      "epoch": 0.11864262870404438,
+      "grad_norm": 0.7446224093437195,
+      "learning_rate": 1.4946335743138587e-05,
+      "loss": 0.8383,
+      "step": 262
+    },
+    {
+      "epoch": 0.11909546316474684,
+      "grad_norm": 0.7819026112556458,
+      "learning_rate": 1.4945909504983848e-05,
+      "loss": 0.8248,
+      "step": 263
+    },
+    {
+      "epoch": 0.1195482976254493,
+      "grad_norm": 0.7390825152397156,
+      "learning_rate": 1.4945481586907141e-05,
+      "loss": 0.8604,
+      "step": 264
+    },
+    {
+      "epoch": 0.12000113208615176,
+      "grad_norm": 0.7803400158882141,
+      "learning_rate": 1.494505198900501e-05,
+      "loss": 0.8302,
+      "step": 265
+    },
+    {
+      "epoch": 0.12045396654685421,
+      "grad_norm": 0.9145985841751099,
+      "learning_rate": 1.4944620711374377e-05,
+      "loss": 0.8732,
+      "step": 266
+    },
+    {
+      "epoch": 0.12090680100755667,
+      "grad_norm": 0.7339340448379517,
+      "learning_rate": 1.494418775411255e-05,
+      "loss": 0.8042,
+      "step": 267
+    },
+    {
+      "epoch": 0.12135963546825913,
+      "grad_norm": 0.7937312722206116,
+      "learning_rate": 1.4943753117317208e-05,
+      "loss": 0.8347,
+      "step": 268
+    },
+    {
+      "epoch": 0.1218124699289616,
+      "grad_norm": 0.7886615991592407,
+      "learning_rate": 1.4943316801086414e-05,
+      "loss": 0.8511,
+      "step": 269
+    },
+    {
+      "epoch": 0.12226530438966406,
+      "grad_norm": 0.8435689210891724,
+      "learning_rate": 1.4942878805518608e-05,
+      "loss": 0.8187,
+      "step": 270
+    },
+    {
+      "epoch": 0.12271813885036652,
+      "grad_norm": 0.6933366060256958,
+      "learning_rate": 1.494243913071261e-05,
+      "loss": 0.762,
+      "step": 271
+    },
+    {
+      "epoch": 0.12317097331106898,
+      "grad_norm": 0.743684709072113,
+      "learning_rate": 1.4941997776767616e-05,
+      "loss": 0.8686,
+      "step": 272
+    },
+    {
+      "epoch": 0.12362380777177143,
+      "grad_norm": 0.772577702999115,
+      "learning_rate": 1.4941554743783204e-05,
+      "loss": 0.8271,
+      "step": 273
+    },
+    {
+      "epoch": 0.12407664223247389,
+      "grad_norm": 0.7139607667922974,
+      "learning_rate": 1.4941110031859327e-05,
+      "loss": 0.7724,
+      "step": 274
+    },
+    {
+      "epoch": 0.12452947669317635,
+      "grad_norm": 0.6617469191551208,
+      "learning_rate": 1.4940663641096325e-05,
+      "loss": 0.8092,
+      "step": 275
+    },
+    {
+      "epoch": 0.12498231115387881,
+      "grad_norm": 0.841748833656311,
+      "learning_rate": 1.4940215571594908e-05,
+      "loss": 0.9361,
+      "step": 276
+    },
+    {
+      "epoch": 0.12498231115387881,
+      "eval_loss": 0.8086027503013611,
+      "eval_runtime": 19.4718,
+      "eval_samples_per_second": 5.855,
+      "eval_steps_per_second": 0.77,
+      "step": 276
+    },
+    {
+      "epoch": 0.12543514561458127,
+      "grad_norm": 0.6689454913139343,
+      "learning_rate": 1.4939765823456168e-05,
+      "loss": 0.8363,
+      "step": 277
+    },
+    {
+      "epoch": 0.12588798007528373,
+      "grad_norm": 0.7700071334838867,
+      "learning_rate": 1.4939314396781575e-05,
+      "loss": 0.8393,
+      "step": 278
+    },
+    {
+      "epoch": 0.1263408145359862,
+      "grad_norm": 0.7219974994659424,
+      "learning_rate": 1.4938861291672983e-05,
+      "loss": 0.8039,
+      "step": 279
+    },
+    {
+      "epoch": 0.12679364899668866,
+      "grad_norm": 0.6459200382232666,
+      "learning_rate": 1.4938406508232613e-05,
+      "loss": 0.819,
+      "step": 280
+    },
+    {
+      "epoch": 0.12724648345739112,
+      "grad_norm": 0.7447069883346558,
+      "learning_rate": 1.4937950046563075e-05,
+      "loss": 0.8567,
+      "step": 281
+    },
+    {
+      "epoch": 0.12769931791809358,
+      "grad_norm": 0.8317793011665344,
+      "learning_rate": 1.4937491906767357e-05,
+      "loss": 0.8555,
+      "step": 282
+    },
+    {
+      "epoch": 0.128152152378796,
+      "grad_norm": 0.6332393884658813,
+      "learning_rate": 1.4937032088948819e-05,
+      "loss": 0.851,
+      "step": 283
+    },
+    {
+      "epoch": 0.12860498683949848,
+      "grad_norm": 0.6920958757400513,
+      "learning_rate": 1.4936570593211203e-05,
+      "loss": 0.7737,
+      "step": 284
+    },
+    {
+      "epoch": 0.12905782130020094,
+      "grad_norm": 0.7432802319526672,
+      "learning_rate": 1.4936107419658635e-05,
+      "loss": 0.8177,
+      "step": 285
+    },
+    {
+      "epoch": 0.1295106557609034,
+      "grad_norm": 0.7308900952339172,
+      "learning_rate": 1.4935642568395613e-05,
+      "loss": 0.751,
+      "step": 286
+    },
+    {
+      "epoch": 0.12996349022160586,
+      "grad_norm": 0.9042837619781494,
+      "learning_rate": 1.4935176039527014e-05,
+      "loss": 0.8192,
+      "step": 287
+    },
+    {
+      "epoch": 0.13041632468230832,
+      "grad_norm": 0.7064964175224304,
+      "learning_rate": 1.4934707833158094e-05,
+      "loss": 0.7918,
+      "step": 288
+    },
+    {
+      "epoch": 0.13086915914301078,
+      "grad_norm": 1.04722261428833,
+      "learning_rate": 1.4934237949394492e-05,
+      "loss": 0.8003,
+      "step": 289
+    },
+    {
+      "epoch": 0.13132199360371324,
+      "grad_norm": 0.6335585117340088,
+      "learning_rate": 1.4933766388342215e-05,
+      "loss": 0.7773,
+      "step": 290
+    },
+    {
+      "epoch": 0.1317748280644157,
+      "grad_norm": 0.8007393479347229,
+      "learning_rate": 1.4933293150107663e-05,
+      "loss": 0.8016,
+      "step": 291
+    },
+    {
+      "epoch": 0.13222766252511817,
+      "grad_norm": 0.7477773427963257,
+      "learning_rate": 1.49328182347976e-05,
+      "loss": 0.7772,
+      "step": 292
+    },
+    {
+      "epoch": 0.13268049698582063,
+      "grad_norm": 0.8117051124572754,
+      "learning_rate": 1.4932341642519178e-05,
+      "loss": 0.8021,
+      "step": 293
+    },
+    {
+      "epoch": 0.1331333314465231,
+      "grad_norm": 1.086212158203125,
+      "learning_rate": 1.4931863373379923e-05,
+      "loss": 0.8271,
+      "step": 294
+    },
+    {
+      "epoch": 0.13358616590722555,
+      "grad_norm": 0.6811758279800415,
+      "learning_rate": 1.4931383427487741e-05,
+      "loss": 0.7714,
+      "step": 295
+    },
+    {
+      "epoch": 0.134039000367928,
+      "grad_norm": 0.810197114944458,
+      "learning_rate": 1.4930901804950918e-05,
+      "loss": 0.7556,
+      "step": 296
+    },
+    {
+      "epoch": 0.13449183482863045,
+      "grad_norm": 0.656467080116272,
+      "learning_rate": 1.4930418505878113e-05,
+      "loss": 0.7308,
+      "step": 297
+    },
+    {
+      "epoch": 0.1349446692893329,
+      "grad_norm": 0.7282403111457825,
+      "learning_rate": 1.4929933530378367e-05,
+      "loss": 0.7529,
+      "step": 298
+    },
+    {
+      "epoch": 0.13539750375003537,
+      "grad_norm": 0.7933338284492493,
+      "learning_rate": 1.4929446878561098e-05,
+      "loss": 0.8193,
+      "step": 299
+    },
+    {
+      "epoch": 0.13585033821073783,
+      "grad_norm": 0.7075956463813782,
+      "learning_rate": 1.4928958550536107e-05,
+      "loss": 0.7736,
+      "step": 300
+    },
+    {
+      "epoch": 0.1363031726714403,
+      "grad_norm": 0.6675400137901306,
+      "learning_rate": 1.4928468546413563e-05,
+      "loss": 0.7477,
+      "step": 301
+    },
+    {
+      "epoch": 0.13675600713214275,
+      "grad_norm": 0.8855708241462708,
+      "learning_rate": 1.4927976866304024e-05,
+      "loss": 0.7988,
+      "step": 302
+    },
+    {
+      "epoch": 0.13720884159284522,
+      "grad_norm": 0.7367263436317444,
+      "learning_rate": 1.4927483510318417e-05,
+      "loss": 0.8018,
+      "step": 303
+    },
+    {
+      "epoch": 0.13766167605354768,
+      "grad_norm": 0.7198050022125244,
+      "learning_rate": 1.4926988478568055e-05,
+      "loss": 0.8311,
+      "step": 304
+    },
+    {
+      "epoch": 0.13811451051425014,
+      "grad_norm": 0.6773768067359924,
+      "learning_rate": 1.4926491771164623e-05,
+      "loss": 0.8388,
+      "step": 305
+    },
+    {
+      "epoch": 0.1385673449749526,
+      "grad_norm": 0.7650995850563049,
+      "learning_rate": 1.4925993388220189e-05,
+      "loss": 0.7779,
+      "step": 306
+    },
+    {
+      "epoch": 0.13902017943565506,
+      "grad_norm": 0.7893260717391968,
+      "learning_rate": 1.4925493329847195e-05,
+      "loss": 0.7559,
+      "step": 307
+    },
+    {
+      "epoch": 0.13947301389635752,
+      "grad_norm": 0.7186692357063293,
+      "learning_rate": 1.4924991596158462e-05,
+      "loss": 0.8107,
+      "step": 308
+    },
+    {
+      "epoch": 0.13992584835705998,
+      "grad_norm": 0.7286010980606079,
+      "learning_rate": 1.492448818726719e-05,
+      "loss": 0.6999,
+      "step": 309
+    },
+    {
+      "epoch": 0.14037868281776242,
+      "grad_norm": 0.7160384058952332,
+      "learning_rate": 1.4923983103286957e-05,
+      "loss": 0.7403,
+      "step": 310
+    },
+    {
+      "epoch": 0.14083151727846488,
+      "grad_norm": 0.858772337436676,
+      "learning_rate": 1.492347634433172e-05,
+      "loss": 0.7424,
+      "step": 311
+    },
+    {
+      "epoch": 0.14128435173916734,
+      "grad_norm": 0.7329276204109192,
+      "learning_rate": 1.4922967910515809e-05,
+      "loss": 0.7232,
+      "step": 312
+    },
+    {
+      "epoch": 0.1417371861998698,
+      "grad_norm": 0.8894422054290771,
+      "learning_rate": 1.4922457801953934e-05,
+      "loss": 0.7737,
+      "step": 313
+    },
+    {
+      "epoch": 0.14219002066057226,
+      "grad_norm": 0.7823007106781006,
+      "learning_rate": 1.492194601876119e-05,
+      "loss": 0.7835,
+      "step": 314
+    },
+    {
+      "epoch": 0.14264285512127473,
+      "grad_norm": 0.8172221779823303,
+      "learning_rate": 1.492143256105304e-05,
+      "loss": 0.7721,
+      "step": 315
+    },
+    {
+      "epoch": 0.1430956895819772,
+      "grad_norm": 0.9521918892860413,
+      "learning_rate": 1.492091742894533e-05,
+      "loss": 0.7202,
+      "step": 316
+    },
+    {
+      "epoch": 0.14354852404267965,
+      "grad_norm": 0.6912105679512024,
+      "learning_rate": 1.492040062255428e-05,
+      "loss": 0.7555,
+      "step": 317
+    },
+    {
+      "epoch": 0.1440013585033821,
+      "grad_norm": 0.713433563709259,
+      "learning_rate": 1.4919882141996493e-05,
+      "loss": 0.7422,
+      "step": 318
+    },
+    {
+      "epoch": 0.14445419296408457,
+      "grad_norm": 0.6809953451156616,
+      "learning_rate": 1.4919361987388942e-05,
+      "loss": 0.7298,
+      "step": 319
+    },
+    {
+      "epoch": 0.14490702742478703,
+      "grad_norm": 0.8265045285224915,
+      "learning_rate": 1.491884015884899e-05,
+      "loss": 0.8056,
+      "step": 320
+    },
+    {
+      "epoch": 0.1453598618854895,
+      "grad_norm": 0.7027236819267273,
+      "learning_rate": 1.491831665649437e-05,
+      "loss": 0.7647,
+      "step": 321
+    },
+    {
+      "epoch": 0.14581269634619196,
+      "grad_norm": 0.6985670328140259,
+      "learning_rate": 1.4917791480443183e-05,
+      "loss": 0.7859,
+      "step": 322
+    },
+    {
+      "epoch": 0.14626553080689442,
+      "grad_norm": 0.6507831811904907,
+      "learning_rate": 1.4917264630813925e-05,
+      "loss": 0.7143,
+      "step": 323
+    },
+    {
+      "epoch": 0.14671836526759685,
+      "grad_norm": 0.8120051026344299,
+      "learning_rate": 1.4916736107725463e-05,
+      "loss": 0.7825,
+      "step": 324
+    },
+    {
+      "epoch": 0.1471711997282993,
+      "grad_norm": 0.7564501166343689,
+      "learning_rate": 1.4916205911297039e-05,
+      "loss": 0.8156,
+      "step": 325
+    },
+    {
+      "epoch": 0.14762403418900177,
+      "grad_norm": 0.7937654852867126,
+      "learning_rate": 1.4915674041648274e-05,
+      "loss": 0.7779,
+      "step": 326
+    },
+    {
+      "epoch": 0.14807686864970424,
+      "grad_norm": 0.6390685439109802,
+      "learning_rate": 1.4915140498899165e-05,
+      "loss": 0.7297,
+      "step": 327
+    },
+    {
+      "epoch": 0.1485297031104067,
+      "grad_norm": 0.8413376212120056,
+      "learning_rate": 1.4914605283170092e-05,
+      "loss": 0.7459,
+      "step": 328
+    },
+    {
+      "epoch": 0.14898253757110916,
+      "grad_norm": 0.8402100801467896,
+      "learning_rate": 1.4914068394581805e-05,
+      "loss": 0.7285,
+      "step": 329
+    },
+    {
+      "epoch": 0.14943537203181162,
+      "grad_norm": 0.8007932901382446,
+      "learning_rate": 1.4913529833255436e-05,
+      "loss": 0.8058,
+      "step": 330
+    },
+    {
+      "epoch": 0.14988820649251408,
+      "grad_norm": 0.913092851638794,
+      "learning_rate": 1.4912989599312496e-05,
+      "loss": 0.7506,
+      "step": 331
+    },
+    {
+      "epoch": 0.15034104095321654,
+      "grad_norm": 0.9280517101287842,
+      "learning_rate": 1.4912447692874865e-05,
+      "loss": 0.8402,
+      "step": 332
+    },
+    {
+      "epoch": 0.150793875413919,
+      "grad_norm": 0.8286890387535095,
+      "learning_rate": 1.4911904114064816e-05,
+      "loss": 0.6831,
+      "step": 333
+    },
+    {
+      "epoch": 0.15124670987462147,
+      "grad_norm": 1.3162356615066528,
+      "learning_rate": 1.491135886300498e-05,
+      "loss": 0.7654,
+      "step": 334
+    },
+    {
+      "epoch": 0.15169954433532393,
+      "grad_norm": 0.6534262895584106,
+      "learning_rate": 1.4910811939818376e-05,
+      "loss": 0.7818,
+      "step": 335
+    },
+    {
+      "epoch": 0.1521523787960264,
+      "grad_norm": 1.4352598190307617,
+      "learning_rate": 1.4910263344628405e-05,
+      "loss": 0.7497,
+      "step": 336
+    },
+    {
+      "epoch": 0.15260521325672885,
+      "grad_norm": 0.7645979523658752,
+      "learning_rate": 1.4909713077558834e-05,
+      "loss": 0.7298,
+      "step": 337
+    },
+    {
+      "epoch": 0.15305804771743128,
+      "grad_norm": 1.1644870042800903,
+      "learning_rate": 1.4909161138733815e-05,
+      "loss": 0.7725,
+      "step": 338
+    },
+    {
+      "epoch": 0.15351088217813375,
+      "grad_norm": 0.7654008865356445,
+      "learning_rate": 1.4908607528277873e-05,
+      "loss": 0.6841,
+      "step": 339
+    },
+    {
+      "epoch": 0.1539637166388362,
+      "grad_norm": 0.9176437258720398,
+      "learning_rate": 1.4908052246315915e-05,
+      "loss": 0.7965,
+      "step": 340
+    },
+    {
+      "epoch": 0.15441655109953867,
+      "grad_norm": 0.8297829031944275,
+      "learning_rate": 1.4907495292973215e-05,
+      "loss": 0.7624,
+      "step": 341
+    },
+    {
+      "epoch": 0.15486938556024113,
+      "grad_norm": 0.7733472585678101,
+      "learning_rate": 1.490693666837544e-05,
+      "loss": 0.7679,
+      "step": 342
+    },
+    {
+      "epoch": 0.1553222200209436,
+      "grad_norm": 0.8300292491912842,
+      "learning_rate": 1.4906376372648618e-05,
+      "loss": 0.7582,
+      "step": 343
+    },
+    {
+      "epoch": 0.15577505448164605,
+      "grad_norm": 0.7907745838165283,
+      "learning_rate": 1.4905814405919163e-05,
+      "loss": 0.6966,
+      "step": 344
+    },
+    {
+      "epoch": 0.15622788894234851,
+      "grad_norm": 1.0779638290405273,
+      "learning_rate": 1.4905250768313865e-05,
+      "loss": 0.698,
+      "step": 345
+    },
+    {
+      "epoch": 0.15668072340305098,
+      "grad_norm": 0.7394607663154602,
+      "learning_rate": 1.490468545995989e-05,
+      "loss": 0.7973,
+      "step": 346
+    },
+    {
+      "epoch": 0.15713355786375344,
+      "grad_norm": 0.9171550869941711,
+      "learning_rate": 1.4904118480984782e-05,
+      "loss": 0.7185,
+      "step": 347
+    },
+    {
+      "epoch": 0.1575863923244559,
+      "grad_norm": 0.7766843438148499,
+      "learning_rate": 1.4903549831516459e-05,
+      "loss": 0.692,
+      "step": 348
+    },
+    {
+      "epoch": 0.15803922678515836,
+      "grad_norm": 0.8801367282867432,
+      "learning_rate": 1.490297951168322e-05,
+      "loss": 0.7797,
+      "step": 349
+    },
+    {
+      "epoch": 0.15849206124586082,
+      "grad_norm": 0.8276110887527466,
+      "learning_rate": 1.4902407521613735e-05,
+      "loss": 0.7405,
+      "step": 350
+    },
+    {
+      "epoch": 0.15894489570656328,
+      "grad_norm": 0.8306746482849121,
+      "learning_rate": 1.4901833861437058e-05,
+      "loss": 0.7268,
+      "step": 351
+    },
+    {
+      "epoch": 0.15939773016726572,
+      "grad_norm": 0.9609989523887634,
+      "learning_rate": 1.4901258531282616e-05,
+      "loss": 0.7233,
+      "step": 352
+    },
+    {
+      "epoch": 0.15985056462796818,
+      "grad_norm": 0.8433880805969238,
+      "learning_rate": 1.4900681531280212e-05,
+      "loss": 0.7076,
+      "step": 353
+    },
+    {
+      "epoch": 0.16030339908867064,
+      "grad_norm": 0.7875558137893677,
+      "learning_rate": 1.4900102861560027e-05,
+      "loss": 0.6939,
+      "step": 354
+    },
+    {
+      "epoch": 0.1607562335493731,
+      "grad_norm": 0.7184025645256042,
+      "learning_rate": 1.4899522522252618e-05,
+      "loss": 0.7195,
+      "step": 355
+    },
+    {
+      "epoch": 0.16120906801007556,
+      "grad_norm": 0.7774180173873901,
+      "learning_rate": 1.4898940513488921e-05,
+      "loss": 0.713,
+      "step": 356
+    },
+    {
+      "epoch": 0.16166190247077802,
+      "grad_norm": 0.8321492075920105,
+      "learning_rate": 1.4898356835400247e-05,
+      "loss": 0.7299,
+      "step": 357
+    },
+    {
+      "epoch": 0.16211473693148049,
+      "grad_norm": 0.8344357013702393,
+      "learning_rate": 1.4897771488118281e-05,
+      "loss": 0.723,
+      "step": 358
+    },
+    {
+      "epoch": 0.16256757139218295,
+      "grad_norm": 0.8810961842536926,
+      "learning_rate": 1.489718447177509e-05,
+      "loss": 0.723,
+      "step": 359
+    },
+    {
+      "epoch": 0.1630204058528854,
+      "grad_norm": 0.6815357804298401,
+      "learning_rate": 1.4896595786503114e-05,
+      "loss": 0.7178,
+      "step": 360
+    },
+    {
+      "epoch": 0.16347324031358787,
+      "grad_norm": 0.7081362009048462,
+      "learning_rate": 1.4896005432435168e-05,
+      "loss": 0.6913,
+      "step": 361
+    },
+    {
+      "epoch": 0.16392607477429033,
+      "grad_norm": 0.712780237197876,
+      "learning_rate": 1.4895413409704448e-05,
+      "loss": 0.7063,
+      "step": 362
+    },
+    {
+      "epoch": 0.1643789092349928,
+      "grad_norm": 0.7857349514961243,
+      "learning_rate": 1.4894819718444525e-05,
+      "loss": 0.6999,
+      "step": 363
+    },
+    {
+      "epoch": 0.16483174369569525,
+      "grad_norm": 0.7282727360725403,
+      "learning_rate": 1.4894224358789344e-05,
+      "loss": 0.7216,
+      "step": 364
+    },
+    {
+      "epoch": 0.16528457815639772,
+      "grad_norm": 0.7465353012084961,
+      "learning_rate": 1.4893627330873227e-05,
+      "loss": 0.7342,
+      "step": 365
+    },
+    {
+      "epoch": 0.16573741261710015,
+      "grad_norm": 0.8547371029853821,
+      "learning_rate": 1.4893028634830877e-05,
+      "loss": 0.6788,
+      "step": 366
+    },
+    {
+      "epoch": 0.1661902470778026,
+      "grad_norm": 0.7159935832023621,
+      "learning_rate": 1.4892428270797368e-05,
+      "loss": 0.6689,
+      "step": 367
+    },
+    {
+      "epoch": 0.16664308153850507,
+      "grad_norm": 0.8785684108734131,
+      "learning_rate": 1.489182623890815e-05,
+      "loss": 0.7402,
+      "step": 368
+    },
+    {
+      "epoch": 0.16709591599920753,
+      "grad_norm": 0.8778204917907715,
+      "learning_rate": 1.4891222539299058e-05,
+      "loss": 0.6354,
+      "step": 369
+    },
+    {
+      "epoch": 0.16754875045991,
+      "grad_norm": 0.8214927315711975,
+      "learning_rate": 1.4890617172106286e-05,
+      "loss": 0.7109,
+      "step": 370
+    },
+    {
+      "epoch": 0.16800158492061246,
+      "grad_norm": 0.7724870443344116,
+      "learning_rate": 1.4890010137466428e-05,
+      "loss": 0.7082,
+      "step": 371
+    },
+    {
+      "epoch": 0.16845441938131492,
+      "grad_norm": 0.7708100080490112,
+      "learning_rate": 1.4889401435516431e-05,
+      "loss": 0.6824,
+      "step": 372
+    },
+    {
+      "epoch": 0.16890725384201738,
+      "grad_norm": 0.6783519387245178,
+      "learning_rate": 1.4888791066393632e-05,
+      "loss": 0.733,
+      "step": 373
+    },
+    {
+      "epoch": 0.16936008830271984,
+      "grad_norm": 0.7518247961997986,
+      "learning_rate": 1.488817903023574e-05,
+      "loss": 0.7549,
+      "step": 374
+    },
+    {
+      "epoch": 0.1698129227634223,
+      "grad_norm": 0.7920773029327393,
+      "learning_rate": 1.4887565327180842e-05,
+      "loss": 0.7505,
+      "step": 375
+    },
+    {
+      "epoch": 0.17026575722412476,
+      "grad_norm": 0.7498000860214233,
+      "learning_rate": 1.4886949957367398e-05,
+      "loss": 0.738,
+      "step": 376
+    },
+    {
+      "epoch": 0.17071859168482723,
+      "grad_norm": 0.682931661605835,
+      "learning_rate": 1.4886332920934247e-05,
+      "loss": 0.6858,
+      "step": 377
+    },
+    {
+      "epoch": 0.1711714261455297,
+      "grad_norm": 0.705777108669281,
+      "learning_rate": 1.4885714218020604e-05,
+      "loss": 0.7319,
+      "step": 378
+    },
+    {
+      "epoch": 0.17162426060623212,
+      "grad_norm": 0.7821267247200012,
+      "learning_rate": 1.4885093848766055e-05,
+      "loss": 0.6815,
+      "step": 379
+    },
+    {
+      "epoch": 0.17207709506693458,
+      "grad_norm": 0.7366465330123901,
+      "learning_rate": 1.4884471813310567e-05,
+      "loss": 0.7158,
+      "step": 380
+    },
+    {
+      "epoch": 0.17252992952763704,
+      "grad_norm": 0.7276479601860046,
+      "learning_rate": 1.4883848111794484e-05,
+      "loss": 0.6813,
+      "step": 381
+    },
+    {
+      "epoch": 0.1729827639883395,
+      "grad_norm": 0.6883667707443237,
+      "learning_rate": 1.488322274435852e-05,
+      "loss": 0.6472,
+      "step": 382
+    },
+    {
+      "epoch": 0.17343559844904197,
+      "grad_norm": 0.6681039333343506,
+      "learning_rate": 1.4882595711143772e-05,
+      "loss": 0.6808,
+      "step": 383
+    },
+    {
+      "epoch": 0.17388843290974443,
+      "grad_norm": 0.9255726933479309,
+      "learning_rate": 1.4881967012291707e-05,
+      "loss": 0.6653,
+      "step": 384
+    },
+    {
+      "epoch": 0.1743412673704469,
+      "grad_norm": 0.7656347155570984,
+      "learning_rate": 1.488133664794417e-05,
+      "loss": 0.6502,
+      "step": 385
+    },
+    {
+      "epoch": 0.17479410183114935,
+      "grad_norm": 0.6879505515098572,
+      "learning_rate": 1.4880704618243382e-05,
+      "loss": 0.6785,
+      "step": 386
+    },
+    {
+      "epoch": 0.1752469362918518,
+      "grad_norm": 0.7322476506233215,
+      "learning_rate": 1.4880070923331942e-05,
+      "loss": 0.6697,
+      "step": 387
+    },
+    {
+      "epoch": 0.17569977075255427,
+      "grad_norm": 0.7081751823425293,
+      "learning_rate": 1.4879435563352815e-05,
+      "loss": 0.6706,
+      "step": 388
+    },
+    {
+      "epoch": 0.17615260521325674,
+      "grad_norm": 0.9385679364204407,
+      "learning_rate": 1.4878798538449358e-05,
+      "loss": 0.6689,
+      "step": 389
+    },
+    {
+      "epoch": 0.1766054396739592,
+      "grad_norm": 1.0547739267349243,
+      "learning_rate": 1.4878159848765288e-05,
+      "loss": 0.7344,
+      "step": 390
+    },
+    {
+      "epoch": 0.17705827413466166,
+      "grad_norm": 0.772883951663971,
+      "learning_rate": 1.4877519494444707e-05,
+      "loss": 0.7109,
+      "step": 391
+    },
+    {
+      "epoch": 0.17751110859536412,
+      "grad_norm": 0.9401255249977112,
+      "learning_rate": 1.4876877475632089e-05,
+      "loss": 0.6846,
+      "step": 392
+    },
+    {
+      "epoch": 0.17796394305606655,
+      "grad_norm": 0.6944053769111633,
+      "learning_rate": 1.4876233792472284e-05,
+      "loss": 0.6735,
+      "step": 393
+    },
+    {
+      "epoch": 0.17841677751676902,
+      "grad_norm": 1.1043004989624023,
+      "learning_rate": 1.4875588445110517e-05,
+      "loss": 0.6569,
+      "step": 394
+    },
+    {
+      "epoch": 0.17886961197747148,
+      "grad_norm": 0.6451114416122437,
+      "learning_rate": 1.4874941433692393e-05,
+      "loss": 0.6748,
+      "step": 395
+    },
+    {
+      "epoch": 0.17932244643817394,
+      "grad_norm": 0.8441924452781677,
+      "learning_rate": 1.4874292758363882e-05,
+      "loss": 0.729,
+      "step": 396
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 0.7901274561882019,
+      "learning_rate": 1.4873642419271342e-05,
+      "loss": 0.7124,
+      "step": 397
+    },
+    {
+      "epoch": 0.18022811535957886,
+      "grad_norm": 1.0363836288452148,
+      "learning_rate": 1.4872990416561499e-05,
+      "loss": 0.6816,
+      "step": 398
+    },
+    {
+      "epoch": 0.18068094982028132,
+      "grad_norm": 0.7908300161361694,
+      "learning_rate": 1.4872336750381452e-05,
+      "loss": 0.6721,
+      "step": 399
+    },
+    {
+      "epoch": 0.18113378428098378,
+      "grad_norm": 0.8819090127944946,
+      "learning_rate": 1.4871681420878683e-05,
+      "loss": 0.6521,
+      "step": 400
+    },
+    {
+      "epoch": 0.18158661874168625,
+      "grad_norm": 1.1245052814483643,
+      "learning_rate": 1.4871024428201043e-05,
+      "loss": 0.7597,
+      "step": 401
+    },
+    {
+      "epoch": 0.1820394532023887,
+      "grad_norm": 0.962951123714447,
+      "learning_rate": 1.4870365772496764e-05,
+      "loss": 0.7093,
+      "step": 402
+    },
+    {
+      "epoch": 0.18249228766309117,
+      "grad_norm": 1.236630916595459,
+      "learning_rate": 1.4869705453914446e-05,
+      "loss": 0.7013,
+      "step": 403
+    },
+    {
+      "epoch": 0.18294512212379363,
+      "grad_norm": 0.8700730204582214,
+      "learning_rate": 1.486904347260307e-05,
+      "loss": 0.6539,
+      "step": 404
+    },
+    {
+      "epoch": 0.1833979565844961,
+      "grad_norm": 1.0793607234954834,
+      "learning_rate": 1.4868379828711991e-05,
+      "loss": 0.6784,
+      "step": 405
+    },
+    {
+      "epoch": 0.18385079104519855,
+      "grad_norm": 0.9725819826126099,
+      "learning_rate": 1.4867714522390934e-05,
+      "loss": 0.7212,
+      "step": 406
+    },
+    {
+      "epoch": 0.184303625505901,
+      "grad_norm": 0.9909386038780212,
+      "learning_rate": 1.4867047553790007e-05,
+      "loss": 0.67,
+      "step": 407
+    },
+    {
+      "epoch": 0.18475645996660345,
+      "grad_norm": 0.993817925453186,
+      "learning_rate": 1.4866378923059687e-05,
+      "loss": 0.6186,
+      "step": 408
+    },
+    {
+      "epoch": 0.1852092944273059,
+      "grad_norm": 1.0599784851074219,
+      "learning_rate": 1.4865708630350834e-05,
+      "loss": 0.674,
+      "step": 409
+    },
+    {
+      "epoch": 0.18566212888800837,
+      "grad_norm": 1.0941828489303589,
+      "learning_rate": 1.4865036675814669e-05,
+      "loss": 0.6587,
+      "step": 410
+    },
+    {
+      "epoch": 0.18611496334871083,
+      "grad_norm": 3.841005802154541,
+      "learning_rate": 1.48643630596028e-05,
+      "loss": 0.729,
+      "step": 411
+    },
+    {
+      "epoch": 0.1865677978094133,
+      "grad_norm": 1.5443971157073975,
+      "learning_rate": 1.4863687781867209e-05,
+      "loss": 0.6768,
+      "step": 412
+    },
+    {
+      "epoch": 0.18702063227011576,
+      "grad_norm": 0.9371129870414734,
+      "learning_rate": 1.4863010842760246e-05,
+      "loss": 0.6169,
+      "step": 413
+    },
+    {
+      "epoch": 0.18747346673081822,
+      "grad_norm": 1.437848448753357,
+      "learning_rate": 1.4862332242434639e-05,
+      "loss": 0.7218,
+      "step": 414
+    },
+    {
+      "epoch": 0.18792630119152068,
+      "grad_norm": 1.4850507974624634,
+      "learning_rate": 1.4861651981043495e-05,
+      "loss": 0.6684,
+      "step": 415
+    },
+    {
+      "epoch": 0.18837913565222314,
+      "grad_norm": 0.7159014940261841,
+      "learning_rate": 1.486097005874029e-05,
+      "loss": 0.6782,
+      "step": 416
+    },
+    {
+      "epoch": 0.1888319701129256,
+      "grad_norm": 1.2282770872116089,
+      "learning_rate": 1.4860286475678877e-05,
+      "loss": 0.7224,
+      "step": 417
+    },
+    {
+      "epoch": 0.18928480457362806,
+      "grad_norm": 1.0234930515289307,
+      "learning_rate": 1.4859601232013488e-05,
+      "loss": 0.693,
+      "step": 418
+    },
+    {
+      "epoch": 0.18973763903433052,
+      "grad_norm": 0.8315566778182983,
+      "learning_rate": 1.485891432789872e-05,
+      "loss": 0.6348,
+      "step": 419
+    },
+    {
+      "epoch": 0.19019047349503299,
+      "grad_norm": 1.353355884552002,
+      "learning_rate": 1.4858225763489552e-05,
+      "loss": 0.7412,
+      "step": 420
+    },
+    {
+      "epoch": 0.19064330795573542,
+      "grad_norm": 0.8341365456581116,
+      "learning_rate": 1.4857535538941339e-05,
+      "loss": 0.6768,
+      "step": 421
+    },
+    {
+      "epoch": 0.19109614241643788,
+      "grad_norm": 0.8955153226852417,
+      "learning_rate": 1.4856843654409802e-05,
+      "loss": 0.65,
+      "step": 422
+    },
+    {
+      "epoch": 0.19154897687714034,
+      "grad_norm": 0.8571435809135437,
+      "learning_rate": 1.4856150110051044e-05,
+      "loss": 0.6437,
+      "step": 423
+    },
+    {
+      "epoch": 0.1920018113378428,
+      "grad_norm": 0.76616370677948,
+      "learning_rate": 1.485545490602154e-05,
+      "loss": 0.6599,
+      "step": 424
+    },
+    {
+      "epoch": 0.19245464579854527,
+      "grad_norm": 0.7934353947639465,
+      "learning_rate": 1.485475804247814e-05,
+      "loss": 0.6271,
+      "step": 425
+    },
+    {
+      "epoch": 0.19290748025924773,
+      "grad_norm": 0.8939810395240784,
+      "learning_rate": 1.4854059519578069e-05,
+      "loss": 0.659,
+      "step": 426
+    },
+    {
+      "epoch": 0.1933603147199502,
+      "grad_norm": 0.6890648007392883,
+      "learning_rate": 1.4853359337478923e-05,
+      "loss": 0.7456,
+      "step": 427
+    },
+    {
+      "epoch": 0.19381314918065265,
+      "grad_norm": 0.8249607682228088,
+      "learning_rate": 1.4852657496338678e-05,
+      "loss": 0.6701,
+      "step": 428
+    },
+    {
+      "epoch": 0.1942659836413551,
+      "grad_norm": 0.8182385563850403,
+      "learning_rate": 1.4851953996315678e-05,
+      "loss": 0.6081,
+      "step": 429
+    },
+    {
+      "epoch": 0.19471881810205757,
+      "grad_norm": 0.7631829977035522,
+      "learning_rate": 1.4851248837568646e-05,
+      "loss": 0.5855,
+      "step": 430
+    },
+    {
+      "epoch": 0.19517165256276003,
+      "grad_norm": 0.9246631860733032,
+      "learning_rate": 1.4850542020256677e-05,
+      "loss": 0.6469,
+      "step": 431
+    },
+    {
+      "epoch": 0.1956244870234625,
+      "grad_norm": 0.7482291460037231,
+      "learning_rate": 1.4849833544539242e-05,
+      "loss": 0.6801,
+      "step": 432
+    },
+    {
+      "epoch": 0.19607732148416496,
+      "grad_norm": 0.8039886951446533,
+      "learning_rate": 1.4849123410576183e-05,
+      "loss": 0.71,
+      "step": 433
+    },
+    {
+      "epoch": 0.1965301559448674,
+      "grad_norm": 0.7532901763916016,
+      "learning_rate": 1.484841161852772e-05,
+      "loss": 0.6619,
+      "step": 434
+    },
+    {
+      "epoch": 0.19698299040556985,
+      "grad_norm": 0.7386736273765564,
+      "learning_rate": 1.4847698168554447e-05,
+      "loss": 0.69,
+      "step": 435
+    },
+    {
+      "epoch": 0.1974358248662723,
+      "grad_norm": 0.6776562929153442,
+      "learning_rate": 1.4846983060817324e-05,
+      "loss": 0.6575,
+      "step": 436
+    },
+    {
+      "epoch": 0.19788865932697478,
+      "grad_norm": 0.7260106205940247,
+      "learning_rate": 1.4846266295477698e-05,
+      "loss": 0.6311,
+      "step": 437
+    },
+    {
+      "epoch": 0.19834149378767724,
+      "grad_norm": 0.7524548172950745,
+      "learning_rate": 1.4845547872697279e-05,
+      "loss": 0.6287,
+      "step": 438
+    },
+    {
+      "epoch": 0.1987943282483797,
+      "grad_norm": 0.8071796298027039,
+      "learning_rate": 1.4844827792638158e-05,
+      "loss": 0.6258,
+      "step": 439
+    },
+    {
+      "epoch": 0.19924716270908216,
+      "grad_norm": 0.8051711320877075,
+      "learning_rate": 1.4844106055462793e-05,
+      "loss": 0.6966,
+      "step": 440
+    },
+    {
+      "epoch": 0.19969999716978462,
+      "grad_norm": 0.8135205507278442,
+      "learning_rate": 1.4843382661334025e-05,
+      "loss": 0.6637,
+      "step": 441
+    },
+    {
+      "epoch": 0.20015283163048708,
+      "grad_norm": 0.8501531481742859,
+      "learning_rate": 1.4842657610415061e-05,
+      "loss": 0.6969,
+      "step": 442
+    },
+    {
+      "epoch": 0.20060566609118954,
+      "grad_norm": 0.7646431922912598,
+      "learning_rate": 1.4841930902869486e-05,
+      "loss": 0.6786,
+      "step": 443
+    },
+    {
+      "epoch": 0.201058500551892,
+      "grad_norm": 0.7486881613731384,
+      "learning_rate": 1.4841202538861255e-05,
+      "loss": 0.6009,
+      "step": 444
+    },
+    {
+      "epoch": 0.20151133501259447,
+      "grad_norm": 0.7672154307365417,
+      "learning_rate": 1.4840472518554702e-05,
+      "loss": 0.6538,
+      "step": 445
+    },
+    {
+      "epoch": 0.20196416947329693,
+      "grad_norm": 0.7490094900131226,
+      "learning_rate": 1.4839740842114529e-05,
+      "loss": 0.624,
+      "step": 446
+    },
+    {
+      "epoch": 0.2024170039339994,
+      "grad_norm": 0.6690830588340759,
+      "learning_rate": 1.4839007509705819e-05,
+      "loss": 0.6335,
+      "step": 447
+    },
+    {
+      "epoch": 0.20286983839470182,
+      "grad_norm": 0.7209163308143616,
+      "learning_rate": 1.483827252149402e-05,
+      "loss": 0.596,
+      "step": 448
+    },
+    {
+      "epoch": 0.20332267285540429,
+      "grad_norm": 0.7429115772247314,
+      "learning_rate": 1.4837535877644957e-05,
+      "loss": 0.6344,
+      "step": 449
+    },
+    {
+      "epoch": 0.20377550731610675,
+      "grad_norm": 0.7298805713653564,
+      "learning_rate": 1.4836797578324833e-05,
+      "loss": 0.6197,
+      "step": 450
+    },
+    {
+      "epoch": 0.2042283417768092,
+      "grad_norm": 0.750446081161499,
+      "learning_rate": 1.4836057623700218e-05,
+      "loss": 0.5782,
+      "step": 451
+    },
+    {
+      "epoch": 0.20468117623751167,
+      "grad_norm": 0.811336100101471,
+      "learning_rate": 1.483531601393806e-05,
+      "loss": 0.6411,
+      "step": 452
+    },
+    {
+      "epoch": 0.20513401069821413,
+      "grad_norm": 0.7158386707305908,
+      "learning_rate": 1.4834572749205675e-05,
+      "loss": 0.6706,
+      "step": 453
+    },
+    {
+      "epoch": 0.2055868451589166,
+      "grad_norm": 0.6980316042900085,
+      "learning_rate": 1.483382782967076e-05,
+      "loss": 0.607,
+      "step": 454
+    },
+    {
+      "epoch": 0.20603967961961905,
+      "grad_norm": 0.7448428273200989,
+      "learning_rate": 1.483308125550138e-05,
+      "loss": 0.666,
+      "step": 455
+    },
+    {
+      "epoch": 0.20649251408032152,
+      "grad_norm": 0.6907296776771545,
+      "learning_rate": 1.4832333026865973e-05,
+      "loss": 0.6145,
+      "step": 456
+    },
+    {
+      "epoch": 0.20694534854102398,
+      "grad_norm": 0.7163165211677551,
+      "learning_rate": 1.4831583143933353e-05,
+      "loss": 0.6306,
+      "step": 457
+    },
+    {
+      "epoch": 0.20739818300172644,
+      "grad_norm": 0.6344444155693054,
+      "learning_rate": 1.483083160687271e-05,
+      "loss": 0.619,
+      "step": 458
+    },
+    {
+      "epoch": 0.2078510174624289,
+      "grad_norm": 0.6727322936058044,
+      "learning_rate": 1.4830078415853596e-05,
+      "loss": 0.6618,
+      "step": 459
+    },
+    {
+      "epoch": 0.20830385192313136,
+      "grad_norm": 0.7381990551948547,
+      "learning_rate": 1.482932357104595e-05,
+      "loss": 0.5994,
+      "step": 460
+    },
+    {
+      "epoch": 0.20875668638383382,
+      "grad_norm": 0.7138816118240356,
+      "learning_rate": 1.4828567072620074e-05,
+      "loss": 0.6549,
+      "step": 461
+    },
+    {
+      "epoch": 0.20920952084453626,
+      "grad_norm": 0.708865761756897,
+      "learning_rate": 1.4827808920746646e-05,
+      "loss": 0.6603,
+      "step": 462
+    },
+    {
+      "epoch": 0.20966235530523872,
+      "grad_norm": 0.7122054696083069,
+      "learning_rate": 1.4827049115596722e-05,
+      "loss": 0.6423,
+      "step": 463
+    },
+    {
+      "epoch": 0.21011518976594118,
+      "grad_norm": 0.7929750680923462,
+      "learning_rate": 1.4826287657341723e-05,
+      "loss": 0.6772,
+      "step": 464
+    },
+    {
+      "epoch": 0.21056802422664364,
+      "grad_norm": 0.7480528354644775,
+      "learning_rate": 1.4825524546153449e-05,
+      "loss": 0.6118,
+      "step": 465
+    },
+    {
+      "epoch": 0.2110208586873461,
+      "grad_norm": 0.8834244608879089,
+      "learning_rate": 1.4824759782204068e-05,
+      "loss": 0.622,
+      "step": 466
+    },
+    {
+      "epoch": 0.21147369314804856,
+      "grad_norm": 0.6908849477767944,
+      "learning_rate": 1.4823993365666127e-05,
+      "loss": 0.6625,
+      "step": 467
+    },
+    {
+      "epoch": 0.21192652760875103,
+      "grad_norm": 0.7661241888999939,
+      "learning_rate": 1.4823225296712539e-05,
+      "loss": 0.6409,
+      "step": 468
+    },
+    {
+      "epoch": 0.2123793620694535,
+      "grad_norm": 0.6954145431518555,
+      "learning_rate": 1.4822455575516595e-05,
+      "loss": 0.5925,
+      "step": 469
+    },
+    {
+      "epoch": 0.21283219653015595,
+      "grad_norm": 0.6936039924621582,
+      "learning_rate": 1.4821684202251958e-05,
+      "loss": 0.6555,
+      "step": 470
+    },
+    {
+      "epoch": 0.2132850309908584,
+      "grad_norm": 0.8155099749565125,
+      "learning_rate": 1.482091117709266e-05,
+      "loss": 0.6868,
+      "step": 471
+    },
+    {
+      "epoch": 0.21373786545156087,
+      "grad_norm": 0.6670776605606079,
+      "learning_rate": 1.4820136500213112e-05,
+      "loss": 0.602,
+      "step": 472
+    },
+    {
+      "epoch": 0.21419069991226333,
+      "grad_norm": 0.6749069690704346,
+      "learning_rate": 1.4819360171788092e-05,
+      "loss": 0.5757,
+      "step": 473
+    },
+    {
+      "epoch": 0.2146435343729658,
+      "grad_norm": 0.8901804685592651,
+      "learning_rate": 1.4818582191992752e-05,
+      "loss": 0.6199,
+      "step": 474
+    },
+    {
+      "epoch": 0.21509636883366826,
+      "grad_norm": 1.0478390455245972,
+      "learning_rate": 1.4817802561002619e-05,
+      "loss": 0.5611,
+      "step": 475
+    },
+    {
+      "epoch": 0.2155492032943707,
+      "grad_norm": 0.679456353187561,
+      "learning_rate": 1.4817021278993588e-05,
+      "loss": 0.5717,
+      "step": 476
+    },
+    {
+      "epoch": 0.21600203775507315,
+      "grad_norm": 1.1234318017959595,
+      "learning_rate": 1.4816238346141934e-05,
+      "loss": 0.6477,
+      "step": 477
+    },
+    {
+      "epoch": 0.2164548722157756,
+      "grad_norm": 0.7910907864570618,
+      "learning_rate": 1.4815453762624295e-05,
+      "loss": 0.5973,
+      "step": 478
+    },
+    {
+      "epoch": 0.21690770667647807,
+      "grad_norm": 0.8617794513702393,
+      "learning_rate": 1.481466752861769e-05,
+      "loss": 0.5653,
+      "step": 479
+    },
+    {
+      "epoch": 0.21736054113718054,
+      "grad_norm": 0.7190772294998169,
+      "learning_rate": 1.4813879644299502e-05,
+      "loss": 0.6063,
+      "step": 480
+    },
+    {
+      "epoch": 0.217813375597883,
+      "grad_norm": 0.9855300188064575,
+      "learning_rate": 1.4813090109847495e-05,
+      "loss": 0.6402,
+      "step": 481
+    },
+    {
+      "epoch": 0.21826621005858546,
+      "grad_norm": 0.7089081406593323,
+      "learning_rate": 1.4812298925439799e-05,
+      "loss": 0.6129,
+      "step": 482
+    },
+    {
+      "epoch": 0.21871904451928792,
+      "grad_norm": 0.8527977466583252,
+      "learning_rate": 1.4811506091254922e-05,
+      "loss": 0.6116,
+      "step": 483
+    },
+    {
+      "epoch": 0.21917187897999038,
+      "grad_norm": 0.8156578540802002,
+      "learning_rate": 1.4810711607471737e-05,
+      "loss": 0.6608,
+      "step": 484
+    },
+    {
+      "epoch": 0.21962471344069284,
+      "grad_norm": 0.7411708235740662,
+      "learning_rate": 1.4809915474269493e-05,
+      "loss": 0.5877,
+      "step": 485
+    },
+    {
+      "epoch": 0.2200775479013953,
+      "grad_norm": 0.9931669235229492,
+      "learning_rate": 1.4809117691827812e-05,
+      "loss": 0.6118,
+      "step": 486
+    },
+    {
+      "epoch": 0.22053038236209777,
+      "grad_norm": 0.79165118932724,
+      "learning_rate": 1.4808318260326687e-05,
+      "loss": 0.5565,
+      "step": 487
+    },
+    {
+      "epoch": 0.22098321682280023,
+      "grad_norm": 0.8213533163070679,
+      "learning_rate": 1.4807517179946483e-05,
+      "loss": 0.6048,
+      "step": 488
+    },
+    {
+      "epoch": 0.2214360512835027,
+      "grad_norm": 0.7628992795944214,
+      "learning_rate": 1.4806714450867937e-05,
+      "loss": 0.5837,
+      "step": 489
+    },
+    {
+      "epoch": 0.22188888574420512,
+      "grad_norm": 0.7759009003639221,
+      "learning_rate": 1.4805910073272159e-05,
+      "loss": 0.6224,
+      "step": 490
+    },
+    {
+      "epoch": 0.22234172020490758,
+      "grad_norm": 0.7115432024002075,
+      "learning_rate": 1.4805104047340628e-05,
+      "loss": 0.6427,
+      "step": 491
+    },
+    {
+      "epoch": 0.22279455466561005,
+      "grad_norm": 0.756812334060669,
+      "learning_rate": 1.4804296373255201e-05,
+      "loss": 0.6536,
+      "step": 492
+    },
+    {
+      "epoch": 0.2232473891263125,
+      "grad_norm": 0.7481681108474731,
+      "learning_rate": 1.48034870511981e-05,
+      "loss": 0.6257,
+      "step": 493
+    },
+    {
+      "epoch": 0.22370022358701497,
+      "grad_norm": 0.7567417621612549,
+      "learning_rate": 1.4802676081351922e-05,
+      "loss": 0.5999,
+      "step": 494
+    },
+    {
+      "epoch": 0.22415305804771743,
+      "grad_norm": 0.8136939406394958,
+      "learning_rate": 1.4801863463899634e-05,
+      "loss": 0.6432,
+      "step": 495
+    },
+    {
+      "epoch": 0.2246058925084199,
+      "grad_norm": 0.7277612686157227,
+      "learning_rate": 1.4801049199024577e-05,
+      "loss": 0.6153,
+      "step": 496
+    },
+    {
+      "epoch": 0.22505872696912235,
+      "grad_norm": 0.8077271580696106,
+      "learning_rate": 1.4800233286910462e-05,
+      "loss": 0.5886,
+      "step": 497
+    },
+    {
+      "epoch": 0.22551156142982481,
+      "grad_norm": 0.8224574327468872,
+      "learning_rate": 1.4799415727741376e-05,
+      "loss": 0.5642,
+      "step": 498
+    },
+    {
+      "epoch": 0.22596439589052728,
+      "grad_norm": 0.7653172612190247,
+      "learning_rate": 1.479859652170177e-05,
+      "loss": 0.605,
+      "step": 499
+    },
+    {
+      "epoch": 0.22641723035122974,
+      "grad_norm": 0.74990314245224,
+      "learning_rate": 1.4797775668976473e-05,
+      "loss": 0.619,
+      "step": 500
+    },
+    {
+      "epoch": 0.2268700648119322,
+      "grad_norm": 0.8035469055175781,
+      "learning_rate": 1.4796953169750684e-05,
+      "loss": 0.5863,
+      "step": 501
+    },
+    {
+      "epoch": 0.22732289927263466,
+      "grad_norm": 0.7987021207809448,
+      "learning_rate": 1.4796129024209968e-05,
+      "loss": 0.5786,
+      "step": 502
+    },
+    {
+      "epoch": 0.2277757337333371,
+      "grad_norm": 0.7279565334320068,
+      "learning_rate": 1.4795303232540272e-05,
+      "loss": 0.646,
+      "step": 503
+    },
+    {
+      "epoch": 0.22822856819403955,
+      "grad_norm": 0.753925621509552,
+      "learning_rate": 1.4794475794927902e-05,
+      "loss": 0.6106,
+      "step": 504
+    },
+    {
+      "epoch": 0.22868140265474202,
+      "grad_norm": 0.8243687152862549,
+      "learning_rate": 1.4793646711559549e-05,
+      "loss": 0.5846,
+      "step": 505
+    },
+    {
+      "epoch": 0.22913423711544448,
+      "grad_norm": 0.7034870982170105,
+      "learning_rate": 1.4792815982622264e-05,
+      "loss": 0.6537,
+      "step": 506
+    },
+    {
+      "epoch": 0.22958707157614694,
+      "grad_norm": 0.7104423642158508,
+      "learning_rate": 1.4791983608303472e-05,
+      "loss": 0.6009,
+      "step": 507
+    },
+    {
+      "epoch": 0.2300399060368494,
+      "grad_norm": 0.8007641434669495,
+      "learning_rate": 1.4791149588790973e-05,
+      "loss": 0.5634,
+      "step": 508
+    },
+    {
+      "epoch": 0.23049274049755186,
+      "grad_norm": 1.0584112405776978,
+      "learning_rate": 1.4790313924272935e-05,
+      "loss": 0.5732,
+      "step": 509
+    },
+    {
+      "epoch": 0.23094557495825432,
+      "grad_norm": 0.7870271801948547,
+      "learning_rate": 1.4789476614937899e-05,
+      "loss": 0.6032,
+      "step": 510
+    },
+    {
+      "epoch": 0.23139840941895679,
+      "grad_norm": 0.8935226202011108,
+      "learning_rate": 1.4788637660974773e-05,
+      "loss": 0.5845,
+      "step": 511
+    },
+    {
+      "epoch": 0.23185124387965925,
+      "grad_norm": 0.7717947363853455,
+      "learning_rate": 1.4787797062572842e-05,
+      "loss": 0.6126,
+      "step": 512
+    },
+    {
+      "epoch": 0.2323040783403617,
+      "grad_norm": 0.8335773944854736,
+      "learning_rate": 1.4786954819921759e-05,
+      "loss": 0.6342,
+      "step": 513
+    },
+    {
+      "epoch": 0.23275691280106417,
+      "grad_norm": 0.7727064490318298,
+      "learning_rate": 1.4786110933211548e-05,
+      "loss": 0.5998,
+      "step": 514
+    },
+    {
+      "epoch": 0.23320974726176663,
+      "grad_norm": 0.8422814011573792,
+      "learning_rate": 1.4785265402632602e-05,
+      "loss": 0.6381,
+      "step": 515
+    },
+    {
+      "epoch": 0.2336625817224691,
+      "grad_norm": 0.8143975138664246,
+      "learning_rate": 1.4784418228375688e-05,
+      "loss": 0.5694,
+      "step": 516
+    },
+    {
+      "epoch": 0.23411541618317153,
+      "grad_norm": 0.764120876789093,
+      "learning_rate": 1.4783569410631942e-05,
+      "loss": 0.566,
+      "step": 517
+    },
+    {
+      "epoch": 0.234568250643874,
+      "grad_norm": 0.8712199926376343,
+      "learning_rate": 1.4782718949592873e-05,
+      "loss": 0.6068,
+      "step": 518
+    },
+    {
+      "epoch": 0.23502108510457645,
+      "grad_norm": 0.8048343062400818,
+      "learning_rate": 1.4781866845450361e-05,
+      "loss": 0.5919,
+      "step": 519
+    },
+    {
+      "epoch": 0.2354739195652789,
+      "grad_norm": 0.8512641191482544,
+      "learning_rate": 1.478101309839665e-05,
+      "loss": 0.6131,
+      "step": 520
+    },
+    {
+      "epoch": 0.23592675402598137,
+      "grad_norm": 0.7001386880874634,
+      "learning_rate": 1.4780157708624364e-05,
+      "loss": 0.5971,
+      "step": 521
+    },
+    {
+      "epoch": 0.23637958848668383,
+      "grad_norm": 0.7913312911987305,
+      "learning_rate": 1.477930067632649e-05,
+      "loss": 0.6156,
+      "step": 522
+    },
+    {
+      "epoch": 0.2368324229473863,
+      "grad_norm": 0.7557013034820557,
+      "learning_rate": 1.4778442001696392e-05,
+      "loss": 0.5684,
+      "step": 523
+    },
+    {
+      "epoch": 0.23728525740808876,
+      "grad_norm": 0.6847025156021118,
+      "learning_rate": 1.4777581684927797e-05,
+      "loss": 0.6304,
+      "step": 524
+    },
+    {
+      "epoch": 0.23773809186879122,
+      "grad_norm": 0.7465041875839233,
+      "learning_rate": 1.4776719726214811e-05,
+      "loss": 0.5996,
+      "step": 525
+    },
+    {
+      "epoch": 0.23819092632949368,
+      "grad_norm": 0.776152491569519,
+      "learning_rate": 1.4775856125751907e-05,
+      "loss": 0.6457,
+      "step": 526
+    },
+    {
+      "epoch": 0.23864376079019614,
+      "grad_norm": 0.7176225185394287,
+      "learning_rate": 1.4774990883733924e-05,
+      "loss": 0.6206,
+      "step": 527
+    },
+    {
+      "epoch": 0.2390965952508986,
+      "grad_norm": 0.6723232865333557,
+      "learning_rate": 1.4774124000356074e-05,
+      "loss": 0.5649,
+      "step": 528
+    },
+    {
+      "epoch": 0.23954942971160106,
+      "grad_norm": 0.7886945605278015,
+      "learning_rate": 1.4773255475813948e-05,
+      "loss": 0.6057,
+      "step": 529
+    },
+    {
+      "epoch": 0.24000226417230353,
+      "grad_norm": 0.8538393974304199,
+      "learning_rate": 1.4772385310303492e-05,
+      "loss": 0.5672,
+      "step": 530
+    },
+    {
+      "epoch": 0.24045509863300596,
+      "grad_norm": 0.8311111927032471,
+      "learning_rate": 1.4771513504021034e-05,
+      "loss": 0.6067,
+      "step": 531
+    },
+    {
+      "epoch": 0.24090793309370842,
+      "grad_norm": 0.6744326949119568,
+      "learning_rate": 1.4770640057163265e-05,
+      "loss": 0.5392,
+      "step": 532
+    },
+    {
+      "epoch": 0.24136076755441088,
+      "grad_norm": 0.8035153150558472,
+      "learning_rate": 1.4769764969927255e-05,
+      "loss": 0.5968,
+      "step": 533
+    },
+    {
+      "epoch": 0.24181360201511334,
+      "grad_norm": 0.9007952213287354,
+      "learning_rate": 1.4768888242510433e-05,
+      "loss": 0.5998,
+      "step": 534
+    },
+    {
+      "epoch": 0.2422664364758158,
+      "grad_norm": 1.0340895652770996,
+      "learning_rate": 1.4768009875110603e-05,
+      "loss": 0.6269,
+      "step": 535
+    },
+    {
+      "epoch": 0.24271927093651827,
+      "grad_norm": 0.7285747528076172,
+      "learning_rate": 1.4767129867925942e-05,
+      "loss": 0.6182,
+      "step": 536
+    },
+    {
+      "epoch": 0.24317210539722073,
+      "grad_norm": 0.7769334316253662,
+      "learning_rate": 1.4766248221154996e-05,
+      "loss": 0.6133,
+      "step": 537
+    },
+    {
+      "epoch": 0.2436249398579232,
+      "grad_norm": 0.8429601788520813,
+      "learning_rate": 1.4765364934996674e-05,
+      "loss": 0.5653,
+      "step": 538
+    },
+    {
+      "epoch": 0.24407777431862565,
+      "grad_norm": 0.7434231638908386,
+      "learning_rate": 1.4764480009650264e-05,
+      "loss": 0.5946,
+      "step": 539
+    },
+    {
+      "epoch": 0.2445306087793281,
+      "grad_norm": 0.8999215960502625,
+      "learning_rate": 1.4763593445315422e-05,
+      "loss": 0.5909,
+      "step": 540
+    },
+    {
+      "epoch": 0.24498344324003057,
+      "grad_norm": 0.749334454536438,
+      "learning_rate": 1.4762705242192165e-05,
+      "loss": 0.6052,
+      "step": 541
+    },
+    {
+      "epoch": 0.24543627770073304,
+      "grad_norm": 0.8988975882530212,
+      "learning_rate": 1.4761815400480892e-05,
+      "loss": 0.5615,
+      "step": 542
+    },
+    {
+      "epoch": 0.2458891121614355,
+      "grad_norm": 0.7588968276977539,
+      "learning_rate": 1.4760923920382364e-05,
+      "loss": 0.5623,
+      "step": 543
+    },
+    {
+      "epoch": 0.24634194662213796,
+      "grad_norm": 0.8823333978652954,
+      "learning_rate": 1.4760030802097715e-05,
+      "loss": 0.5761,
+      "step": 544
+    },
+    {
+      "epoch": 0.2467947810828404,
+      "grad_norm": 0.6606689095497131,
+      "learning_rate": 1.4759136045828447e-05,
+      "loss": 0.5995,
+      "step": 545
+    },
+    {
+      "epoch": 0.24724761554354285,
+      "grad_norm": 0.9664527773857117,
+      "learning_rate": 1.4758239651776434e-05,
+      "loss": 0.6049,
+      "step": 546
+    },
+    {
+      "epoch": 0.24770045000424532,
+      "grad_norm": 0.6853055357933044,
+      "learning_rate": 1.4757341620143913e-05,
+      "loss": 0.6096,
+      "step": 547
+    },
+    {
+      "epoch": 0.24815328446494778,
+      "grad_norm": 0.845271110534668,
+      "learning_rate": 1.4756441951133497e-05,
+      "loss": 0.607,
+      "step": 548
+    },
+    {
+      "epoch": 0.24860611892565024,
+      "grad_norm": 0.7086828947067261,
+      "learning_rate": 1.4755540644948168e-05,
+      "loss": 0.5752,
+      "step": 549
+    },
+    {
+      "epoch": 0.2490589533863527,
+      "grad_norm": 0.7899136543273926,
+      "learning_rate": 1.4754637701791273e-05,
+      "loss": 0.5464,
+      "step": 550
+    },
+    {
+      "epoch": 0.24951178784705516,
+      "grad_norm": 0.7464292645454407,
+      "learning_rate": 1.4753733121866534e-05,
+      "loss": 0.5789,
+      "step": 551
+    },
+    {
+      "epoch": 0.24996462230775762,
+      "grad_norm": 0.8088657259941101,
+      "learning_rate": 1.4752826905378039e-05,
+      "loss": 0.6007,
+      "step": 552
+    },
+    {
+      "epoch": 0.24996462230775762,
+      "eval_loss": 0.5881260633468628,
+      "eval_runtime": 19.4408,
+      "eval_samples_per_second": 5.864,
+      "eval_steps_per_second": 0.772,
+      "step": 552
+    },
+    {
+      "epoch": 0.25041745676846006,
+      "grad_norm": 0.8311989307403564,
+      "learning_rate": 1.4751919052530245e-05,
+      "loss": 0.5806,
+      "step": 553
+    },
+    {
+      "epoch": 0.25087029122916255,
+      "grad_norm": 0.9002801179885864,
+      "learning_rate": 1.4751009563527977e-05,
+      "loss": 0.5398,
+      "step": 554
+    },
+    {
+      "epoch": 0.251323125689865,
+      "grad_norm": 0.8495317697525024,
+      "learning_rate": 1.4750098438576434e-05,
+      "loss": 0.6073,
+      "step": 555
+    },
+    {
+      "epoch": 0.25177596015056747,
+      "grad_norm": 0.7978039979934692,
+      "learning_rate": 1.474918567788118e-05,
+      "loss": 0.6079,
+      "step": 556
+    },
+    {
+      "epoch": 0.2522287946112699,
+      "grad_norm": 0.7676923274993896,
+      "learning_rate": 1.4748271281648145e-05,
+      "loss": 0.633,
+      "step": 557
+    },
+    {
+      "epoch": 0.2526816290719724,
+      "grad_norm": 0.7194515466690063,
+      "learning_rate": 1.4747355250083639e-05,
+      "loss": 0.5659,
+      "step": 558
+    },
+    {
+      "epoch": 0.2531344635326748,
+      "grad_norm": 0.8001179099082947,
+      "learning_rate": 1.474643758339433e-05,
+      "loss": 0.56,
+      "step": 559
+    },
+    {
+      "epoch": 0.2535872979933773,
+      "grad_norm": 0.7714060544967651,
+      "learning_rate": 1.4745518281787259e-05,
+      "loss": 0.553,
+      "step": 560
+    },
+    {
+      "epoch": 0.25404013245407975,
+      "grad_norm": 0.8208261132240295,
+      "learning_rate": 1.474459734546984e-05,
+      "loss": 0.614,
+      "step": 561
+    },
+    {
+      "epoch": 0.25449296691478224,
+      "grad_norm": 0.6942229866981506,
+      "learning_rate": 1.4743674774649845e-05,
+      "loss": 0.5483,
+      "step": 562
+    },
+    {
+      "epoch": 0.25494580137548467,
+      "grad_norm": 0.7569373250007629,
+      "learning_rate": 1.4742750569535425e-05,
+      "loss": 0.615,
+      "step": 563
+    },
+    {
+      "epoch": 0.25539863583618716,
+      "grad_norm": 0.7039315700531006,
+      "learning_rate": 1.4741824730335098e-05,
+      "loss": 0.5252,
+      "step": 564
+    },
+    {
+      "epoch": 0.2558514702968896,
+      "grad_norm": 0.8170211315155029,
+      "learning_rate": 1.4740897257257745e-05,
+      "loss": 0.6154,
+      "step": 565
+    },
+    {
+      "epoch": 0.256304304757592,
+      "grad_norm": 0.688449501991272,
+      "learning_rate": 1.4739968150512622e-05,
+      "loss": 0.5537,
+      "step": 566
+    },
+    {
+      "epoch": 0.2567571392182945,
+      "grad_norm": 0.74798983335495,
+      "learning_rate": 1.473903741030935e-05,
+      "loss": 0.5854,
+      "step": 567
+    },
+    {
+      "epoch": 0.25720997367899695,
+      "grad_norm": 0.7894384264945984,
+      "learning_rate": 1.4738105036857921e-05,
+      "loss": 0.5017,
+      "step": 568
+    },
+    {
+      "epoch": 0.25766280813969944,
+      "grad_norm": 0.8515793085098267,
+      "learning_rate": 1.4737171030368693e-05,
+      "loss": 0.5644,
+      "step": 569
+    },
+    {
+      "epoch": 0.2581156426004019,
+      "grad_norm": 0.9285520315170288,
+      "learning_rate": 1.4736235391052393e-05,
+      "loss": 0.5714,
+      "step": 570
+    },
+    {
+      "epoch": 0.25856847706110436,
+      "grad_norm": 0.7274629473686218,
+      "learning_rate": 1.4735298119120116e-05,
+      "loss": 0.4978,
+      "step": 571
+    },
+    {
+      "epoch": 0.2590213115218068,
+      "grad_norm": 1.1050032377243042,
+      "learning_rate": 1.473435921478333e-05,
+      "loss": 0.5982,
+      "step": 572
+    },
+    {
+      "epoch": 0.2594741459825093,
+      "grad_norm": 0.8430832028388977,
+      "learning_rate": 1.4733418678253865e-05,
+      "loss": 0.5126,
+      "step": 573
+    },
+    {
+      "epoch": 0.2599269804432117,
+      "grad_norm": 0.7927145957946777,
+      "learning_rate": 1.4732476509743919e-05,
+      "loss": 0.5066,
+      "step": 574
+    },
+    {
+      "epoch": 0.2603798149039142,
+      "grad_norm": 0.8087596893310547,
+      "learning_rate": 1.473153270946607e-05,
+      "loss": 0.6461,
+      "step": 575
+    },
+    {
+      "epoch": 0.26083264936461664,
+      "grad_norm": 1.0339022874832153,
+      "learning_rate": 1.4730587277633245e-05,
+      "loss": 0.6035,
+      "step": 576
+    },
+    {
+      "epoch": 0.26128548382531913,
+      "grad_norm": 0.8909124732017517,
+      "learning_rate": 1.4729640214458754e-05,
+      "loss": 0.573,
+      "step": 577
+    },
+    {
+      "epoch": 0.26173831828602157,
+      "grad_norm": 1.0848407745361328,
+      "learning_rate": 1.4728691520156269e-05,
+      "loss": 0.5968,
+      "step": 578
+    },
+    {
+      "epoch": 0.262191152746724,
+      "grad_norm": 0.8249455094337463,
+      "learning_rate": 1.4727741194939832e-05,
+      "loss": 0.5785,
+      "step": 579
+    },
+    {
+      "epoch": 0.2626439872074265,
+      "grad_norm": 1.190679669380188,
+      "learning_rate": 1.4726789239023856e-05,
+      "loss": 0.597,
+      "step": 580
+    },
+    {
+      "epoch": 0.2630968216681289,
+      "grad_norm": 0.784512460231781,
+      "learning_rate": 1.4725835652623112e-05,
+      "loss": 0.534,
+      "step": 581
+    },
+    {
+      "epoch": 0.2635496561288314,
+      "grad_norm": 0.8939950466156006,
+      "learning_rate": 1.4724880435952747e-05,
+      "loss": 0.5271,
+      "step": 582
+    },
+    {
+      "epoch": 0.26400249058953384,
+      "grad_norm": 0.7243357300758362,
+      "learning_rate": 1.4723923589228276e-05,
+      "loss": 0.5108,
+      "step": 583
+    },
+    {
+      "epoch": 0.26445532505023633,
+      "grad_norm": 0.7685579657554626,
+      "learning_rate": 1.4722965112665577e-05,
+      "loss": 0.5129,
+      "step": 584
+    },
+    {
+      "epoch": 0.26490815951093877,
+      "grad_norm": 0.8200727105140686,
+      "learning_rate": 1.47220050064809e-05,
+      "loss": 0.5488,
+      "step": 585
+    },
+    {
+      "epoch": 0.26536099397164126,
+      "grad_norm": 0.8088552355766296,
+      "learning_rate": 1.4721043270890862e-05,
+      "loss": 0.503,
+      "step": 586
+    },
+    {
+      "epoch": 0.2658138284323437,
+      "grad_norm": 0.7941896915435791,
+      "learning_rate": 1.4720079906112444e-05,
+      "loss": 0.5661,
+      "step": 587
+    },
+    {
+      "epoch": 0.2662666628930462,
+      "grad_norm": 0.7164261937141418,
+      "learning_rate": 1.4719114912363e-05,
+      "loss": 0.5767,
+      "step": 588
+    },
+    {
+      "epoch": 0.2667194973537486,
+      "grad_norm": 0.92717444896698,
+      "learning_rate": 1.4718148289860247e-05,
+      "loss": 0.5572,
+      "step": 589
+    },
+    {
+      "epoch": 0.2671723318144511,
+      "grad_norm": 0.7904344797134399,
+      "learning_rate": 1.471718003882227e-05,
+      "loss": 0.5009,
+      "step": 590
+    },
+    {
+      "epoch": 0.26762516627515354,
+      "grad_norm": 0.8072219491004944,
+      "learning_rate": 1.4716210159467527e-05,
+      "loss": 0.5609,
+      "step": 591
+    },
+    {
+      "epoch": 0.268078000735856,
+      "grad_norm": 0.7373447418212891,
+      "learning_rate": 1.4715238652014834e-05,
+      "loss": 0.5378,
+      "step": 592
+    },
+    {
+      "epoch": 0.26853083519655846,
+      "grad_norm": 0.8908963203430176,
+      "learning_rate": 1.4714265516683383e-05,
+      "loss": 0.5076,
+      "step": 593
+    },
+    {
+      "epoch": 0.2689836696572609,
+      "grad_norm": 0.7174452543258667,
+      "learning_rate": 1.4713290753692728e-05,
+      "loss": 0.5344,
+      "step": 594
+    },
+    {
+      "epoch": 0.2694365041179634,
+      "grad_norm": 0.7640471458435059,
+      "learning_rate": 1.471231436326279e-05,
+      "loss": 0.5525,
+      "step": 595
+    },
+    {
+      "epoch": 0.2698893385786658,
+      "grad_norm": 0.8122929334640503,
+      "learning_rate": 1.4711336345613864e-05,
+      "loss": 0.5774,
+      "step": 596
+    },
+    {
+      "epoch": 0.2703421730393683,
+      "grad_norm": 0.740322470664978,
+      "learning_rate": 1.4710356700966603e-05,
+      "loss": 0.5363,
+      "step": 597
+    },
+    {
+      "epoch": 0.27079500750007074,
+      "grad_norm": 0.7563768625259399,
+      "learning_rate": 1.4709375429542034e-05,
+      "loss": 0.5507,
+      "step": 598
+    },
+    {
+      "epoch": 0.27124784196077323,
+      "grad_norm": 0.6652012467384338,
+      "learning_rate": 1.4708392531561544e-05,
+      "loss": 0.538,
+      "step": 599
+    },
+    {
+      "epoch": 0.27170067642147566,
+      "grad_norm": 0.785305917263031,
+      "learning_rate": 1.4707408007246898e-05,
+      "loss": 0.6097,
+      "step": 600
+    },
+    {
+      "epoch": 0.27215351088217815,
+      "grad_norm": 0.7112635970115662,
+      "learning_rate": 1.4706421856820214e-05,
+      "loss": 0.5993,
+      "step": 601
+    },
+    {
+      "epoch": 0.2726063453428806,
+      "grad_norm": 0.7882818579673767,
+      "learning_rate": 1.4705434080503988e-05,
+      "loss": 0.5176,
+      "step": 602
+    },
+    {
+      "epoch": 0.2730591798035831,
+      "grad_norm": 1.0064949989318848,
+      "learning_rate": 1.470444467852108e-05,
+      "loss": 0.5876,
+      "step": 603
+    },
+    {
+      "epoch": 0.2735120142642855,
+      "grad_norm": 0.8223081827163696,
+      "learning_rate": 1.4703453651094713e-05,
+      "loss": 0.5901,
+      "step": 604
+    },
+    {
+      "epoch": 0.273964848724988,
+      "grad_norm": 0.9124444723129272,
+      "learning_rate": 1.4702460998448482e-05,
+      "loss": 0.5698,
+      "step": 605
+    },
+    {
+      "epoch": 0.27441768318569043,
+      "grad_norm": 0.8722511529922485,
+      "learning_rate": 1.4701466720806342e-05,
+      "loss": 0.6134,
+      "step": 606
+    },
+    {
+      "epoch": 0.27487051764639286,
+      "grad_norm": 0.6496700048446655,
+      "learning_rate": 1.4700470818392621e-05,
+      "loss": 0.5107,
+      "step": 607
+    },
+    {
+      "epoch": 0.27532335210709535,
+      "grad_norm": 0.696812629699707,
+      "learning_rate": 1.4699473291432012e-05,
+      "loss": 0.5341,
+      "step": 608
+    },
+    {
+      "epoch": 0.2757761865677978,
+      "grad_norm": 0.7689089775085449,
+      "learning_rate": 1.4698474140149576e-05,
+      "loss": 0.5457,
+      "step": 609
+    },
+    {
+      "epoch": 0.2762290210285003,
+      "grad_norm": 0.7031781673431396,
+      "learning_rate": 1.4697473364770731e-05,
+      "loss": 0.5142,
+      "step": 610
+    },
+    {
+      "epoch": 0.2766818554892027,
+      "grad_norm": 0.8002460598945618,
+      "learning_rate": 1.4696470965521276e-05,
+      "loss": 0.585,
+      "step": 611
+    },
+    {
+      "epoch": 0.2771346899499052,
+      "grad_norm": 0.8532388210296631,
+      "learning_rate": 1.4695466942627365e-05,
+      "loss": 0.5656,
+      "step": 612
+    },
+    {
+      "epoch": 0.27758752441060763,
+      "grad_norm": 0.7644308805465698,
+      "learning_rate": 1.4694461296315525e-05,
+      "loss": 0.5082,
+      "step": 613
+    },
+    {
+      "epoch": 0.2780403588713101,
+      "grad_norm": 0.8224950432777405,
+      "learning_rate": 1.4693454026812643e-05,
+      "loss": 0.5058,
+      "step": 614
+    },
+    {
+      "epoch": 0.27849319333201256,
+      "grad_norm": 0.6774716973304749,
+      "learning_rate": 1.4692445134345981e-05,
+      "loss": 0.4961,
+      "step": 615
+    },
+    {
+      "epoch": 0.27894602779271505,
+      "grad_norm": 0.8179674744606018,
+      "learning_rate": 1.4691434619143159e-05,
+      "loss": 0.4869,
+      "step": 616
+    },
+    {
+      "epoch": 0.2793988622534175,
+      "grad_norm": 0.6929678320884705,
+      "learning_rate": 1.4690422481432164e-05,
+      "loss": 0.5274,
+      "step": 617
+    },
+    {
+      "epoch": 0.27985169671411997,
+      "grad_norm": 0.7661529779434204,
+      "learning_rate": 1.4689408721441357e-05,
+      "loss": 0.4941,
+      "step": 618
+    },
+    {
+      "epoch": 0.2803045311748224,
+      "grad_norm": 0.7640644907951355,
+      "learning_rate": 1.4688393339399453e-05,
+      "loss": 0.5352,
+      "step": 619
+    },
+    {
+      "epoch": 0.28075736563552484,
+      "grad_norm": 1.1076126098632812,
+      "learning_rate": 1.4687376335535543e-05,
+      "loss": 0.5875,
+      "step": 620
+    },
+    {
+      "epoch": 0.2812102000962273,
+      "grad_norm": 0.7371631860733032,
+      "learning_rate": 1.468635771007908e-05,
+      "loss": 0.524,
+      "step": 621
+    },
+    {
+      "epoch": 0.28166303455692976,
+      "grad_norm": 0.7571913003921509,
+      "learning_rate": 1.468533746325988e-05,
+      "loss": 0.5778,
+      "step": 622
+    },
+    {
+      "epoch": 0.28211586901763225,
+      "grad_norm": 0.7033771276473999,
+      "learning_rate": 1.4684315595308132e-05,
+      "loss": 0.5334,
+      "step": 623
+    },
+    {
+      "epoch": 0.2825687034783347,
+      "grad_norm": 1.0408718585968018,
+      "learning_rate": 1.4683292106454383e-05,
+      "loss": 0.5654,
+      "step": 624
+    },
+    {
+      "epoch": 0.28302153793903717,
+      "grad_norm": 0.7843034267425537,
+      "learning_rate": 1.4682266996929551e-05,
+      "loss": 0.5262,
+      "step": 625
+    },
+    {
+      "epoch": 0.2834743723997396,
+      "grad_norm": 0.8728121519088745,
+      "learning_rate": 1.4681240266964919e-05,
+      "loss": 0.51,
+      "step": 626
+    },
+    {
+      "epoch": 0.2839272068604421,
+      "grad_norm": 0.640317976474762,
+      "learning_rate": 1.4680211916792133e-05,
+      "loss": 0.5431,
+      "step": 627
+    },
+    {
+      "epoch": 0.2843800413211445,
+      "grad_norm": 0.9380479454994202,
+      "learning_rate": 1.4679181946643205e-05,
+      "loss": 0.539,
+      "step": 628
+    },
+    {
+      "epoch": 0.284832875781847,
+      "grad_norm": 0.7024577856063843,
+      "learning_rate": 1.4678150356750514e-05,
+      "loss": 0.4875,
+      "step": 629
+    },
+    {
+      "epoch": 0.28528571024254945,
+      "grad_norm": 0.7149611711502075,
+      "learning_rate": 1.4677117147346806e-05,
+      "loss": 0.4813,
+      "step": 630
+    },
+    {
+      "epoch": 0.28573854470325194,
+      "grad_norm": 0.6758112907409668,
+      "learning_rate": 1.4676082318665188e-05,
+      "loss": 0.5813,
+      "step": 631
+    },
+    {
+      "epoch": 0.2861913791639544,
+      "grad_norm": 0.733073890209198,
+      "learning_rate": 1.4675045870939137e-05,
+      "loss": 0.5057,
+      "step": 632
+    },
+    {
+      "epoch": 0.28664421362465686,
+      "grad_norm": 0.9430336356163025,
+      "learning_rate": 1.467400780440249e-05,
+      "loss": 0.5309,
+      "step": 633
+    },
+    {
+      "epoch": 0.2870970480853593,
+      "grad_norm": 0.779950737953186,
+      "learning_rate": 1.4672968119289455e-05,
+      "loss": 0.5352,
+      "step": 634
+    },
+    {
+      "epoch": 0.28754988254606173,
+      "grad_norm": 0.804973840713501,
+      "learning_rate": 1.4671926815834602e-05,
+      "loss": 0.5636,
+      "step": 635
+    },
+    {
+      "epoch": 0.2880027170067642,
+      "grad_norm": 0.6549612283706665,
+      "learning_rate": 1.4670883894272865e-05,
+      "loss": 0.5356,
+      "step": 636
+    },
+    {
+      "epoch": 0.28845555146746665,
+      "grad_norm": 0.7596065998077393,
+      "learning_rate": 1.4669839354839545e-05,
+      "loss": 0.5719,
+      "step": 637
+    },
+    {
+      "epoch": 0.28890838592816914,
+      "grad_norm": 0.8659469485282898,
+      "learning_rate": 1.4668793197770308e-05,
+      "loss": 0.5028,
+      "step": 638
+    },
+    {
+      "epoch": 0.2893612203888716,
+      "grad_norm": 0.8582921028137207,
+      "learning_rate": 1.4667745423301185e-05,
+      "loss": 0.509,
+      "step": 639
+    },
+    {
+      "epoch": 0.28981405484957407,
+      "grad_norm": 0.751609206199646,
+      "learning_rate": 1.466669603166857e-05,
+      "loss": 0.5419,
+      "step": 640
+    },
+    {
+      "epoch": 0.2902668893102765,
+      "grad_norm": 1.0048867464065552,
+      "learning_rate": 1.4665645023109226e-05,
+      "loss": 0.5745,
+      "step": 641
+    },
+    {
+      "epoch": 0.290719723770979,
+      "grad_norm": 0.7976627945899963,
+      "learning_rate": 1.4664592397860274e-05,
+      "loss": 0.4956,
+      "step": 642
+    },
+    {
+      "epoch": 0.2911725582316814,
+      "grad_norm": 0.7978916168212891,
+      "learning_rate": 1.4663538156159208e-05,
+      "loss": 0.5252,
+      "step": 643
+    },
+    {
+      "epoch": 0.2916253926923839,
+      "grad_norm": 0.824474573135376,
+      "learning_rate": 1.466248229824388e-05,
+      "loss": 0.5338,
+      "step": 644
+    },
+    {
+      "epoch": 0.29207822715308634,
+      "grad_norm": 0.7580894827842712,
+      "learning_rate": 1.4661424824352512e-05,
+      "loss": 0.5074,
+      "step": 645
+    },
+    {
+      "epoch": 0.29253106161378883,
+      "grad_norm": 0.7472822070121765,
+      "learning_rate": 1.4660365734723684e-05,
+      "loss": 0.4755,
+      "step": 646
+    },
+    {
+      "epoch": 0.29298389607449127,
+      "grad_norm": 0.7068562507629395,
+      "learning_rate": 1.465930502959635e-05,
+      "loss": 0.5129,
+      "step": 647
+    },
+    {
+      "epoch": 0.2934367305351937,
+      "grad_norm": 1.2026171684265137,
+      "learning_rate": 1.4658242709209815e-05,
+      "loss": 0.538,
+      "step": 648
+    },
+    {
+      "epoch": 0.2938895649958962,
+      "grad_norm": 0.7542515397071838,
+      "learning_rate": 1.4657178773803761e-05,
+      "loss": 0.5344,
+      "step": 649
+    },
+    {
+      "epoch": 0.2943423994565986,
+      "grad_norm": 1.083905816078186,
+      "learning_rate": 1.4656113223618232e-05,
+      "loss": 0.5295,
+      "step": 650
+    },
+    {
+      "epoch": 0.2947952339173011,
+      "grad_norm": 0.7265732884407043,
+      "learning_rate": 1.4655046058893629e-05,
+      "loss": 0.5152,
+      "step": 651
+    },
+    {
+      "epoch": 0.29524806837800355,
+      "grad_norm": 0.8517547249794006,
+      "learning_rate": 1.4653977279870725e-05,
+      "loss": 0.5316,
+      "step": 652
+    },
+    {
+      "epoch": 0.29570090283870604,
+      "grad_norm": 0.7927708625793457,
+      "learning_rate": 1.4652906886790654e-05,
+      "loss": 0.5193,
+      "step": 653
+    },
+    {
+      "epoch": 0.29615373729940847,
+      "grad_norm": 0.661551296710968,
+      "learning_rate": 1.4651834879894914e-05,
+      "loss": 0.5148,
+      "step": 654
+    },
+    {
+      "epoch": 0.29660657176011096,
+      "grad_norm": 0.7670228481292725,
+      "learning_rate": 1.4650761259425369e-05,
+      "loss": 0.5274,
+      "step": 655
+    },
+    {
+      "epoch": 0.2970594062208134,
+      "grad_norm": 0.7227180004119873,
+      "learning_rate": 1.4649686025624244e-05,
+      "loss": 0.5116,
+      "step": 656
+    },
+    {
+      "epoch": 0.2975122406815159,
+      "grad_norm": 0.8273769617080688,
+      "learning_rate": 1.4648609178734132e-05,
+      "loss": 0.4723,
+      "step": 657
+    },
+    {
+      "epoch": 0.2979650751422183,
+      "grad_norm": 0.6958010196685791,
+      "learning_rate": 1.4647530718997987e-05,
+      "loss": 0.5077,
+      "step": 658
+    },
+    {
+      "epoch": 0.2984179096029208,
+      "grad_norm": 0.8402749300003052,
+      "learning_rate": 1.4646450646659126e-05,
+      "loss": 0.5319,
+      "step": 659
+    },
+    {
+      "epoch": 0.29887074406362324,
+      "grad_norm": 0.7076095938682556,
+      "learning_rate": 1.4645368961961232e-05,
+      "loss": 0.5306,
+      "step": 660
+    },
+    {
+      "epoch": 0.29932357852432573,
+      "grad_norm": 0.8241704702377319,
+      "learning_rate": 1.4644285665148352e-05,
+      "loss": 0.5652,
+      "step": 661
+    },
+    {
+      "epoch": 0.29977641298502816,
+      "grad_norm": 0.6605017781257629,
+      "learning_rate": 1.4643200756464895e-05,
+      "loss": 0.4739,
+      "step": 662
+    },
+    {
+      "epoch": 0.3002292474457306,
+      "grad_norm": 0.7162265181541443,
+      "learning_rate": 1.4642114236155635e-05,
+      "loss": 0.5339,
+      "step": 663
+    },
+    {
+      "epoch": 0.3006820819064331,
+      "grad_norm": 0.920584499835968,
+      "learning_rate": 1.464102610446571e-05,
+      "loss": 0.477,
+      "step": 664
+    },
+    {
+      "epoch": 0.3011349163671355,
+      "grad_norm": 1.0123887062072754,
+      "learning_rate": 1.463993636164062e-05,
+      "loss": 0.5005,
+      "step": 665
+    },
+    {
+      "epoch": 0.301587750827838,
+      "grad_norm": 0.8752802610397339,
+      "learning_rate": 1.4638845007926232e-05,
+      "loss": 0.5035,
+      "step": 666
+    },
+    {
+      "epoch": 0.30204058528854044,
+      "grad_norm": 0.9289608001708984,
+      "learning_rate": 1.4637752043568767e-05,
+      "loss": 0.4911,
+      "step": 667
+    },
+    {
+      "epoch": 0.30249341974924293,
+      "grad_norm": 0.8423551321029663,
+      "learning_rate": 1.463665746881482e-05,
+      "loss": 0.5333,
+      "step": 668
+    },
+    {
+      "epoch": 0.30294625420994536,
+      "grad_norm": 1.0525884628295898,
+      "learning_rate": 1.463556128391135e-05,
+      "loss": 0.5097,
+      "step": 669
+    },
+    {
+      "epoch": 0.30339908867064785,
+      "grad_norm": 0.909200131893158,
+      "learning_rate": 1.4634463489105666e-05,
+      "loss": 0.5584,
+      "step": 670
+    },
+    {
+      "epoch": 0.3038519231313503,
+      "grad_norm": 1.0725759267807007,
+      "learning_rate": 1.4633364084645457e-05,
+      "loss": 0.5722,
+      "step": 671
+    },
+    {
+      "epoch": 0.3043047575920528,
+      "grad_norm": 0.7647678256034851,
+      "learning_rate": 1.4632263070778758e-05,
+      "loss": 0.5064,
+      "step": 672
+    },
+    {
+      "epoch": 0.3047575920527552,
+      "grad_norm": 1.1567573547363281,
+      "learning_rate": 1.4631160447753985e-05,
+      "loss": 0.5108,
+      "step": 673
+    },
+    {
+      "epoch": 0.3052104265134577,
+      "grad_norm": 1.012374758720398,
+      "learning_rate": 1.4630056215819904e-05,
+      "loss": 0.5589,
+      "step": 674
+    },
+    {
+      "epoch": 0.30566326097416013,
+      "grad_norm": 1.4147967100143433,
+      "learning_rate": 1.4628950375225648e-05,
+      "loss": 0.5355,
+      "step": 675
+    },
+    {
+      "epoch": 0.30611609543486257,
+      "grad_norm": 0.9129096269607544,
+      "learning_rate": 1.4627842926220715e-05,
+      "loss": 0.5529,
+      "step": 676
+    },
+    {
+      "epoch": 0.30656892989556506,
+      "grad_norm": 0.9252450466156006,
+      "learning_rate": 1.4626733869054964e-05,
+      "loss": 0.4604,
+      "step": 677
+    },
+    {
+      "epoch": 0.3070217643562675,
+      "grad_norm": 1.167197585105896,
+      "learning_rate": 1.4625623203978614e-05,
+      "loss": 0.5238,
+      "step": 678
+    },
+    {
+      "epoch": 0.30747459881697,
+      "grad_norm": 0.672999918460846,
+      "learning_rate": 1.4624510931242252e-05,
+      "loss": 0.4724,
+      "step": 679
+    },
+    {
+      "epoch": 0.3079274332776724,
+      "grad_norm": 0.8699397444725037,
+      "learning_rate": 1.4623397051096825e-05,
+      "loss": 0.5158,
+      "step": 680
+    },
+    {
+      "epoch": 0.3083802677383749,
+      "grad_norm": 0.7273291945457458,
+      "learning_rate": 1.4622281563793643e-05,
+      "loss": 0.5246,
+      "step": 681
+    },
+    {
+      "epoch": 0.30883310219907734,
+      "grad_norm": 0.9489529728889465,
+      "learning_rate": 1.4621164469584378e-05,
+      "loss": 0.5135,
+      "step": 682
+    },
+    {
+      "epoch": 0.3092859366597798,
+      "grad_norm": 0.786284863948822,
+      "learning_rate": 1.4620045768721065e-05,
+      "loss": 0.4938,
+      "step": 683
+    },
+    {
+      "epoch": 0.30973877112048226,
+      "grad_norm": 0.894683837890625,
+      "learning_rate": 1.4618925461456106e-05,
+      "loss": 0.5306,
+      "step": 684
+    },
+    {
+      "epoch": 0.31019160558118475,
+      "grad_norm": 0.9071069955825806,
+      "learning_rate": 1.4617803548042254e-05,
+      "loss": 0.5439,
+      "step": 685
+    },
+    {
+      "epoch": 0.3106444400418872,
+      "grad_norm": 0.8423410654067993,
+      "learning_rate": 1.4616680028732637e-05,
+      "loss": 0.4799,
+      "step": 686
+    },
+    {
+      "epoch": 0.31109727450258967,
+      "grad_norm": 0.7595556378364563,
+      "learning_rate": 1.4615554903780738e-05,
+      "loss": 0.4923,
+      "step": 687
+    },
+    {
+      "epoch": 0.3115501089632921,
+      "grad_norm": 0.7766554355621338,
+      "learning_rate": 1.4614428173440403e-05,
+      "loss": 0.5725,
+      "step": 688
+    },
+    {
+      "epoch": 0.31200294342399454,
+      "grad_norm": 0.8758528828620911,
+      "learning_rate": 1.4613299837965843e-05,
+      "loss": 0.5016,
+      "step": 689
+    },
+    {
+      "epoch": 0.31245577788469703,
+      "grad_norm": 0.8516387343406677,
+      "learning_rate": 1.4612169897611626e-05,
+      "loss": 0.511,
+      "step": 690
+    },
+    {
+      "epoch": 0.31290861234539946,
+      "grad_norm": 0.7681635022163391,
+      "learning_rate": 1.461103835263269e-05,
+      "loss": 0.4992,
+      "step": 691
+    },
+    {
+      "epoch": 0.31336144680610195,
+      "grad_norm": 0.7421332597732544,
+      "learning_rate": 1.4609905203284329e-05,
+      "loss": 0.4717,
+      "step": 692
+    },
+    {
+      "epoch": 0.3138142812668044,
+      "grad_norm": 0.7615931630134583,
+      "learning_rate": 1.46087704498222e-05,
+      "loss": 0.505,
+      "step": 693
+    },
+    {
+      "epoch": 0.3142671157275069,
+      "grad_norm": 0.7673407793045044,
+      "learning_rate": 1.460763409250232e-05,
+      "loss": 0.4748,
+      "step": 694
+    },
+    {
+      "epoch": 0.3147199501882093,
+      "grad_norm": 0.8670549988746643,
+      "learning_rate": 1.4606496131581075e-05,
+      "loss": 0.496,
+      "step": 695
+    },
+    {
+      "epoch": 0.3151727846489118,
+      "grad_norm": 0.8095021843910217,
+      "learning_rate": 1.4605356567315204e-05,
+      "loss": 0.4385,
+      "step": 696
+    },
+    {
+      "epoch": 0.31562561910961423,
+      "grad_norm": 0.7151927947998047,
+      "learning_rate": 1.4604215399961815e-05,
+      "loss": 0.5262,
+      "step": 697
+    },
+    {
+      "epoch": 0.3160784535703167,
+      "grad_norm": 1.0873347520828247,
+      "learning_rate": 1.4603072629778372e-05,
+      "loss": 0.444,
+      "step": 698
+    },
+    {
+      "epoch": 0.31653128803101915,
+      "grad_norm": 0.8482450246810913,
+      "learning_rate": 1.4601928257022705e-05,
+      "loss": 0.5739,
+      "step": 699
+    },
+    {
+      "epoch": 0.31698412249172164,
+      "grad_norm": 0.7151737213134766,
+      "learning_rate": 1.4600782281953e-05,
+      "loss": 0.4712,
+      "step": 700
+    },
+    {
+      "epoch": 0.3174369569524241,
+      "grad_norm": 0.685892641544342,
+      "learning_rate": 1.4599634704827813e-05,
+      "loss": 0.4681,
+      "step": 701
+    },
+    {
+      "epoch": 0.31788979141312657,
+      "grad_norm": 0.8122037649154663,
+      "learning_rate": 1.4598485525906055e-05,
+      "loss": 0.4768,
+      "step": 702
+    },
+    {
+      "epoch": 0.318342625873829,
+      "grad_norm": 0.6169959902763367,
+      "learning_rate": 1.4597334745447e-05,
+      "loss": 0.4869,
+      "step": 703
+    },
+    {
+      "epoch": 0.31879546033453143,
+      "grad_norm": 0.7096338272094727,
+      "learning_rate": 1.4596182363710281e-05,
+      "loss": 0.4805,
+      "step": 704
+    },
+    {
+      "epoch": 0.3192482947952339,
+      "grad_norm": 0.6837482452392578,
+      "learning_rate": 1.4595028380955897e-05,
+      "loss": 0.5615,
+      "step": 705
+    },
+    {
+      "epoch": 0.31970112925593636,
+      "grad_norm": 0.7366883158683777,
+      "learning_rate": 1.4593872797444206e-05,
+      "loss": 0.4706,
+      "step": 706
+    },
+    {
+      "epoch": 0.32015396371663885,
+      "grad_norm": 0.8051954507827759,
+      "learning_rate": 1.4592715613435927e-05,
+      "loss": 0.4847,
+      "step": 707
+    },
+    {
+      "epoch": 0.3206067981773413,
+      "grad_norm": 0.8133533596992493,
+      "learning_rate": 1.4591556829192137e-05,
+      "loss": 0.4954,
+      "step": 708
+    },
+    {
+      "epoch": 0.32105963263804377,
+      "grad_norm": 0.7073555588722229,
+      "learning_rate": 1.4590396444974284e-05,
+      "loss": 0.4308,
+      "step": 709
+    },
+    {
+      "epoch": 0.3215124670987462,
+      "grad_norm": 0.8296173810958862,
+      "learning_rate": 1.4589234461044162e-05,
+      "loss": 0.5252,
+      "step": 710
+    },
+    {
+      "epoch": 0.3219653015594487,
+      "grad_norm": 0.7163283824920654,
+      "learning_rate": 1.458807087766394e-05,
+      "loss": 0.5206,
+      "step": 711
+    },
+    {
+      "epoch": 0.3224181360201511,
+      "grad_norm": 0.8291743993759155,
+      "learning_rate": 1.4586905695096137e-05,
+      "loss": 0.4864,
+      "step": 712
+    },
+    {
+      "epoch": 0.3228709704808536,
+      "grad_norm": 0.7736324667930603,
+      "learning_rate": 1.4585738913603644e-05,
+      "loss": 0.5039,
+      "step": 713
+    },
+    {
+      "epoch": 0.32332380494155605,
+      "grad_norm": 0.7839064002037048,
+      "learning_rate": 1.45845705334497e-05,
+      "loss": 0.4585,
+      "step": 714
+    },
+    {
+      "epoch": 0.32377663940225854,
+      "grad_norm": 0.782141923904419,
+      "learning_rate": 1.4583400554897916e-05,
+      "loss": 0.5271,
+      "step": 715
+    },
+    {
+      "epoch": 0.32422947386296097,
+      "grad_norm": 0.7646268010139465,
+      "learning_rate": 1.4582228978212256e-05,
+      "loss": 0.5123,
+      "step": 716
+    },
+    {
+      "epoch": 0.3246823083236634,
+      "grad_norm": 0.8933156728744507,
+      "learning_rate": 1.458105580365705e-05,
+      "loss": 0.4291,
+      "step": 717
+    },
+    {
+      "epoch": 0.3251351427843659,
+      "grad_norm": 0.796796441078186,
+      "learning_rate": 1.4579881031496982e-05,
+      "loss": 0.5057,
+      "step": 718
+    },
+    {
+      "epoch": 0.3255879772450683,
+      "grad_norm": 0.8032305836677551,
+      "learning_rate": 1.45787046619971e-05,
+      "loss": 0.5003,
+      "step": 719
+    },
+    {
+      "epoch": 0.3260408117057708,
+      "grad_norm": 0.788154125213623,
+      "learning_rate": 1.457752669542282e-05,
+      "loss": 0.4792,
+      "step": 720
+    },
+    {
+      "epoch": 0.32649364616647325,
+      "grad_norm": 0.75339674949646,
+      "learning_rate": 1.4576347132039902e-05,
+      "loss": 0.4287,
+      "step": 721
+    },
+    {
+      "epoch": 0.32694648062717574,
+      "grad_norm": 1.0328612327575684,
+      "learning_rate": 1.457516597211448e-05,
+      "loss": 0.5249,
+      "step": 722
+    },
+    {
+      "epoch": 0.3273993150878782,
+      "grad_norm": 1.01662278175354,
+      "learning_rate": 1.4573983215913043e-05,
+      "loss": 0.5471,
+      "step": 723
+    },
+    {
+      "epoch": 0.32785214954858066,
+      "grad_norm": 0.8054636120796204,
+      "learning_rate": 1.4572798863702442e-05,
+      "loss": 0.4786,
+      "step": 724
+    },
+    {
+      "epoch": 0.3283049840092831,
+      "grad_norm": 0.8108761310577393,
+      "learning_rate": 1.4571612915749885e-05,
+      "loss": 0.4995,
+      "step": 725
+    },
+    {
+      "epoch": 0.3287578184699856,
+      "grad_norm": 1.0297499895095825,
+      "learning_rate": 1.457042537232294e-05,
+      "loss": 0.5023,
+      "step": 726
+    },
+    {
+      "epoch": 0.329210652930688,
+      "grad_norm": 0.8287637829780579,
+      "learning_rate": 1.4569236233689537e-05,
+      "loss": 0.4844,
+      "step": 727
+    },
+    {
+      "epoch": 0.3296634873913905,
+      "grad_norm": 0.8962579965591431,
+      "learning_rate": 1.456804550011797e-05,
+      "loss": 0.4787,
+      "step": 728
+    },
+    {
+      "epoch": 0.33011632185209294,
+      "grad_norm": 0.7159314751625061,
+      "learning_rate": 1.4566853171876886e-05,
+      "loss": 0.4616,
+      "step": 729
+    },
+    {
+      "epoch": 0.33056915631279543,
+      "grad_norm": 0.9780300855636597,
+      "learning_rate": 1.456565924923529e-05,
+      "loss": 0.4748,
+      "step": 730
+    },
+    {
+      "epoch": 0.33102199077349787,
+      "grad_norm": 0.8780568242073059,
+      "learning_rate": 1.4564463732462559e-05,
+      "loss": 0.5119,
+      "step": 731
+    },
+    {
+      "epoch": 0.3314748252342003,
+      "grad_norm": 0.7616591453552246,
+      "learning_rate": 1.4563266621828413e-05,
+      "loss": 0.4763,
+      "step": 732
+    },
+    {
+      "epoch": 0.3319276596949028,
+      "grad_norm": 0.913154661655426,
+      "learning_rate": 1.4562067917602947e-05,
+      "loss": 0.5038,
+      "step": 733
+    },
+    {
+      "epoch": 0.3323804941556052,
+      "grad_norm": 0.7419286370277405,
+      "learning_rate": 1.4560867620056608e-05,
+      "loss": 0.5218,
+      "step": 734
+    },
+    {
+      "epoch": 0.3328333286163077,
+      "grad_norm": 0.759284496307373,
+      "learning_rate": 1.4559665729460198e-05,
+      "loss": 0.4869,
+      "step": 735
+    },
+    {
+      "epoch": 0.33328616307701014,
+      "grad_norm": 0.885164201259613,
+      "learning_rate": 1.455846224608489e-05,
+      "loss": 0.5,
+      "step": 736
+    },
+    {
+      "epoch": 0.33373899753771263,
+      "grad_norm": 0.7694923281669617,
+      "learning_rate": 1.4557257170202207e-05,
+      "loss": 0.4784,
+      "step": 737
+    },
+    {
+      "epoch": 0.33419183199841507,
+      "grad_norm": 0.7753260731697083,
+      "learning_rate": 1.4556050502084034e-05,
+      "loss": 0.4699,
+      "step": 738
+    },
+    {
+      "epoch": 0.33464466645911756,
+      "grad_norm": 0.6947426795959473,
+      "learning_rate": 1.4554842242002614e-05,
+      "loss": 0.4374,
+      "step": 739
+    },
+    {
+      "epoch": 0.33509750091982,
+      "grad_norm": 0.796204686164856,
+      "learning_rate": 1.4553632390230554e-05,
+      "loss": 0.5163,
+      "step": 740
+    },
+    {
+      "epoch": 0.3355503353805225,
+      "grad_norm": 0.7171505093574524,
+      "learning_rate": 1.4552420947040817e-05,
+      "loss": 0.484,
+      "step": 741
+    },
+    {
+      "epoch": 0.3360031698412249,
+      "grad_norm": 0.7220250964164734,
+      "learning_rate": 1.4551207912706725e-05,
+      "loss": 0.4531,
+      "step": 742
+    },
+    {
+      "epoch": 0.3364560043019274,
+      "grad_norm": 0.7847347855567932,
+      "learning_rate": 1.4549993287501955e-05,
+      "loss": 0.5114,
+      "step": 743
+    },
+    {
+      "epoch": 0.33690883876262984,
+      "grad_norm": 0.7722388505935669,
+      "learning_rate": 1.4548777071700549e-05,
+      "loss": 0.4938,
+      "step": 744
+    },
+    {
+      "epoch": 0.33736167322333227,
+      "grad_norm": 0.7462444305419922,
+      "learning_rate": 1.4547559265576907e-05,
+      "loss": 0.4641,
+      "step": 745
+    },
+    {
+      "epoch": 0.33781450768403476,
+      "grad_norm": 0.8149163722991943,
+      "learning_rate": 1.4546339869405783e-05,
+      "loss": 0.4787,
+      "step": 746
+    },
+    {
+      "epoch": 0.3382673421447372,
+      "grad_norm": 0.8563286066055298,
+      "learning_rate": 1.4545118883462298e-05,
+      "loss": 0.4962,
+      "step": 747
+    },
+    {
+      "epoch": 0.3387201766054397,
+      "grad_norm": 0.7843456268310547,
+      "learning_rate": 1.4543896308021924e-05,
+      "loss": 0.4664,
+      "step": 748
+    },
+    {
+      "epoch": 0.3391730110661421,
+      "grad_norm": 0.7675468325614929,
+      "learning_rate": 1.4542672143360493e-05,
+      "loss": 0.4841,
+      "step": 749
+    },
+    {
+      "epoch": 0.3396258455268446,
+      "grad_norm": 0.8849632740020752,
+      "learning_rate": 1.4541446389754201e-05,
+      "loss": 0.5167,
+      "step": 750
+    },
+    {
+      "epoch": 0.34007867998754704,
+      "grad_norm": 0.778249979019165,
+      "learning_rate": 1.4540219047479597e-05,
+      "loss": 0.4565,
+      "step": 751
+    },
+    {
+      "epoch": 0.34053151444824953,
+      "grad_norm": 0.7831906080245972,
+      "learning_rate": 1.4538990116813589e-05,
+      "loss": 0.4558,
+      "step": 752
+    },
+    {
+      "epoch": 0.34098434890895196,
+      "grad_norm": 0.777973473072052,
+      "learning_rate": 1.4537759598033446e-05,
+      "loss": 0.4822,
+      "step": 753
+    },
+    {
+      "epoch": 0.34143718336965445,
+      "grad_norm": 0.6689452528953552,
+      "learning_rate": 1.453652749141679e-05,
+      "loss": 0.4402,
+      "step": 754
+    },
+    {
+      "epoch": 0.3418900178303569,
+      "grad_norm": 0.6991743445396423,
+      "learning_rate": 1.453529379724161e-05,
+      "loss": 0.4195,
+      "step": 755
+    },
+    {
+      "epoch": 0.3423428522910594,
+      "grad_norm": 0.8085342049598694,
+      "learning_rate": 1.4534058515786245e-05,
+      "loss": 0.4789,
+      "step": 756
+    },
+    {
+      "epoch": 0.3427956867517618,
+      "grad_norm": 0.7133055329322815,
+      "learning_rate": 1.4532821647329397e-05,
+      "loss": 0.4706,
+      "step": 757
+    },
+    {
+      "epoch": 0.34324852121246424,
+      "grad_norm": 0.7966130375862122,
+      "learning_rate": 1.4531583192150119e-05,
+      "loss": 0.4782,
+      "step": 758
+    },
+    {
+      "epoch": 0.34370135567316673,
+      "grad_norm": 0.7512375712394714,
+      "learning_rate": 1.4530343150527836e-05,
+      "loss": 0.4674,
+      "step": 759
+    },
+    {
+      "epoch": 0.34415419013386916,
+      "grad_norm": 0.9407236576080322,
+      "learning_rate": 1.4529101522742311e-05,
+      "loss": 0.4598,
+      "step": 760
+    },
+    {
+      "epoch": 0.34460702459457165,
+      "grad_norm": 0.8138443231582642,
+      "learning_rate": 1.4527858309073688e-05,
+      "loss": 0.4619,
+      "step": 761
+    },
+    {
+      "epoch": 0.3450598590552741,
+      "grad_norm": 0.6962406039237976,
+      "learning_rate": 1.452661350980245e-05,
+      "loss": 0.4936,
+      "step": 762
+    },
+    {
+      "epoch": 0.3455126935159766,
+      "grad_norm": 0.7135865092277527,
+      "learning_rate": 1.4525367125209445e-05,
+      "loss": 0.4329,
+      "step": 763
+    },
+    {
+      "epoch": 0.345965527976679,
+      "grad_norm": 0.8204096555709839,
+      "learning_rate": 1.452411915557588e-05,
+      "loss": 0.4502,
+      "step": 764
+    },
+    {
+      "epoch": 0.3464183624373815,
+      "grad_norm": 0.7781338691711426,
+      "learning_rate": 1.4522869601183318e-05,
+      "loss": 0.4148,
+      "step": 765
+    },
+    {
+      "epoch": 0.34687119689808393,
+      "grad_norm": 0.7633736729621887,
+      "learning_rate": 1.4521618462313679e-05,
+      "loss": 0.4297,
+      "step": 766
+    },
+    {
+      "epoch": 0.3473240313587864,
+      "grad_norm": 0.904326856136322,
+      "learning_rate": 1.4520365739249236e-05,
+      "loss": 0.4972,
+      "step": 767
+    },
+    {
+      "epoch": 0.34777686581948886,
+      "grad_norm": 0.7865485548973083,
+      "learning_rate": 1.4519111432272632e-05,
+      "loss": 0.4718,
+      "step": 768
+    },
+    {
+      "epoch": 0.34822970028019135,
+      "grad_norm": 0.9248954057693481,
+      "learning_rate": 1.4517855541666855e-05,
+      "loss": 0.4736,
+      "step": 769
+    },
+    {
+      "epoch": 0.3486825347408938,
+      "grad_norm": 0.8533254861831665,
+      "learning_rate": 1.4516598067715257e-05,
+      "loss": 0.4452,
+      "step": 770
+    },
+    {
+      "epoch": 0.34913536920159627,
+      "grad_norm": 0.9142417311668396,
+      "learning_rate": 1.4515339010701544e-05,
+      "loss": 0.3997,
+      "step": 771
+    },
+    {
+      "epoch": 0.3495882036622987,
+      "grad_norm": 0.9811474680900574,
+      "learning_rate": 1.4514078370909782e-05,
+      "loss": 0.4839,
+      "step": 772
+    },
+    {
+      "epoch": 0.35004103812300114,
+      "grad_norm": 0.8523197770118713,
+      "learning_rate": 1.451281614862439e-05,
+      "loss": 0.5039,
+      "step": 773
+    },
+    {
+      "epoch": 0.3504938725837036,
+      "grad_norm": 0.9271774888038635,
+      "learning_rate": 1.4511552344130147e-05,
+      "loss": 0.526,
+      "step": 774
+    },
+    {
+      "epoch": 0.35094670704440606,
+      "grad_norm": 0.7626346349716187,
+      "learning_rate": 1.4510286957712192e-05,
+      "loss": 0.4031,
+      "step": 775
+    },
+    {
+      "epoch": 0.35139954150510855,
+      "grad_norm": 0.9117299318313599,
+      "learning_rate": 1.4509019989656013e-05,
+      "loss": 0.4756,
+      "step": 776
+    },
+    {
+      "epoch": 0.351852375965811,
+      "grad_norm": 0.7210586071014404,
+      "learning_rate": 1.4507751440247461e-05,
+      "loss": 0.4667,
+      "step": 777
+    },
+    {
+      "epoch": 0.35230521042651347,
+      "grad_norm": 0.7991892695426941,
+      "learning_rate": 1.4506481309772743e-05,
+      "loss": 0.4789,
+      "step": 778
+    },
+    {
+      "epoch": 0.3527580448872159,
+      "grad_norm": 0.7282071113586426,
+      "learning_rate": 1.4505209598518418e-05,
+      "loss": 0.4288,
+      "step": 779
+    },
+    {
+      "epoch": 0.3532108793479184,
+      "grad_norm": 0.8729153275489807,
+      "learning_rate": 1.450393630677141e-05,
+      "loss": 0.4637,
+      "step": 780
+    },
+    {
+      "epoch": 0.3536637138086208,
+      "grad_norm": 0.8181951642036438,
+      "learning_rate": 1.4502661434818992e-05,
+      "loss": 0.4957,
+      "step": 781
+    },
+    {
+      "epoch": 0.3541165482693233,
+      "grad_norm": 0.8652138710021973,
+      "learning_rate": 1.4501384982948799e-05,
+      "loss": 0.4715,
+      "step": 782
+    },
+    {
+      "epoch": 0.35456938273002575,
+      "grad_norm": 0.9875885248184204,
+      "learning_rate": 1.4500106951448816e-05,
+      "loss": 0.4996,
+      "step": 783
+    },
+    {
+      "epoch": 0.35502221719072824,
+      "grad_norm": 0.7315845489501953,
+      "learning_rate": 1.4498827340607393e-05,
+      "loss": 0.46,
+      "step": 784
+    },
+    {
+      "epoch": 0.3554750516514307,
+      "grad_norm": 0.93685382604599,
+      "learning_rate": 1.4497546150713227e-05,
+      "loss": 0.4265,
+      "step": 785
+    },
+    {
+      "epoch": 0.3559278861121331,
+      "grad_norm": 0.7243474125862122,
+      "learning_rate": 1.449626338205538e-05,
+      "loss": 0.4844,
+      "step": 786
+    },
+    {
+      "epoch": 0.3563807205728356,
+      "grad_norm": 0.8693944215774536,
+      "learning_rate": 1.4494979034923265e-05,
+      "loss": 0.5018,
+      "step": 787
+    },
+    {
+      "epoch": 0.35683355503353803,
+      "grad_norm": 0.7501546144485474,
+      "learning_rate": 1.4493693109606649e-05,
+      "loss": 0.4865,
+      "step": 788
+    },
+    {
+      "epoch": 0.3572863894942405,
+      "grad_norm": 0.8437265157699585,
+      "learning_rate": 1.4492405606395663e-05,
+      "loss": 0.495,
+      "step": 789
+    },
+    {
+      "epoch": 0.35773922395494295,
+      "grad_norm": 0.782289981842041,
+      "learning_rate": 1.4491116525580786e-05,
+      "loss": 0.4567,
+      "step": 790
+    },
+    {
+      "epoch": 0.35819205841564544,
+      "grad_norm": 0.755082368850708,
+      "learning_rate": 1.4489825867452858e-05,
+      "loss": 0.4891,
+      "step": 791
+    },
+    {
+      "epoch": 0.3586448928763479,
+      "grad_norm": 0.7294925451278687,
+      "learning_rate": 1.4488533632303072e-05,
+      "loss": 0.4613,
+      "step": 792
+    },
+    {
+      "epoch": 0.35909772733705037,
+      "grad_norm": 0.8621608018875122,
+      "learning_rate": 1.448723982042298e-05,
+      "loss": 0.4479,
+      "step": 793
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 0.7976484298706055,
+      "learning_rate": 1.4485944432104483e-05,
+      "loss": 0.4942,
+      "step": 794
+    },
+    {
+      "epoch": 0.3600033962584553,
+      "grad_norm": 0.6837937235832214,
+      "learning_rate": 1.4484647467639848e-05,
+      "loss": 0.4596,
+      "step": 795
+    },
+    {
+      "epoch": 0.3604562307191577,
+      "grad_norm": 0.7584710717201233,
+      "learning_rate": 1.4483348927321688e-05,
+      "loss": 0.4767,
+      "step": 796
+    },
+    {
+      "epoch": 0.3609090651798602,
+      "grad_norm": 0.8236560225486755,
+      "learning_rate": 1.4482048811442978e-05,
+      "loss": 0.4923,
+      "step": 797
+    },
+    {
+      "epoch": 0.36136189964056264,
+      "grad_norm": 0.7171486020088196,
+      "learning_rate": 1.4480747120297043e-05,
+      "loss": 0.4381,
+      "step": 798
+    },
+    {
+      "epoch": 0.3618147341012651,
+      "grad_norm": 0.7314976453781128,
+      "learning_rate": 1.447944385417757e-05,
+      "loss": 0.4573,
+      "step": 799
+    },
+    {
+      "epoch": 0.36226756856196757,
+      "grad_norm": 0.7224847674369812,
+      "learning_rate": 1.4478139013378597e-05,
+      "loss": 0.5023,
+      "step": 800
+    },
+    {
+      "epoch": 0.36272040302267,
+      "grad_norm": 0.7627956867218018,
+      "learning_rate": 1.4476832598194514e-05,
+      "loss": 0.4326,
+      "step": 801
+    },
+    {
+      "epoch": 0.3631732374833725,
+      "grad_norm": 0.8306934237480164,
+      "learning_rate": 1.4475524608920071e-05,
+      "loss": 0.5251,
+      "step": 802
+    },
+    {
+      "epoch": 0.3636260719440749,
+      "grad_norm": 0.8524587750434875,
+      "learning_rate": 1.4474215045850377e-05,
+      "loss": 0.4546,
+      "step": 803
+    },
+    {
+      "epoch": 0.3640789064047774,
+      "grad_norm": 0.6962935924530029,
+      "learning_rate": 1.4472903909280888e-05,
+      "loss": 0.4495,
+      "step": 804
+    },
+    {
+      "epoch": 0.36453174086547985,
+      "grad_norm": 0.847118616104126,
+      "learning_rate": 1.4471591199507417e-05,
+      "loss": 0.4508,
+      "step": 805
+    },
+    {
+      "epoch": 0.36498457532618234,
+      "grad_norm": 0.7857109308242798,
+      "learning_rate": 1.4470276916826138e-05,
+      "loss": 0.456,
+      "step": 806
+    },
+    {
+      "epoch": 0.36543740978688477,
+      "grad_norm": 0.9544152021408081,
+      "learning_rate": 1.446896106153357e-05,
+      "loss": 0.4344,
+      "step": 807
+    },
+    {
+      "epoch": 0.36589024424758726,
+      "grad_norm": 0.6988970041275024,
+      "learning_rate": 1.4467643633926598e-05,
+      "loss": 0.4755,
+      "step": 808
+    },
+    {
+      "epoch": 0.3663430787082897,
+      "grad_norm": 0.9396758675575256,
+      "learning_rate": 1.4466324634302448e-05,
+      "loss": 0.4444,
+      "step": 809
+    },
+    {
+      "epoch": 0.3667959131689922,
+      "grad_norm": 0.813913106918335,
+      "learning_rate": 1.4465004062958716e-05,
+      "loss": 0.4326,
+      "step": 810
+    },
+    {
+      "epoch": 0.3672487476296946,
+      "grad_norm": 0.8034855723381042,
+      "learning_rate": 1.4463681920193341e-05,
+      "loss": 0.4905,
+      "step": 811
+    },
+    {
+      "epoch": 0.3677015820903971,
+      "grad_norm": 0.7723264694213867,
+      "learning_rate": 1.446235820630462e-05,
+      "loss": 0.376,
+      "step": 812
+    },
+    {
+      "epoch": 0.36815441655109954,
+      "grad_norm": 0.8094017505645752,
+      "learning_rate": 1.4461032921591207e-05,
+      "loss": 0.5052,
+      "step": 813
+    },
+    {
+      "epoch": 0.368607251011802,
+      "grad_norm": 0.7081817984580994,
+      "learning_rate": 1.4459706066352105e-05,
+      "loss": 0.4714,
+      "step": 814
+    },
+    {
+      "epoch": 0.36906008547250446,
+      "grad_norm": 0.7871371507644653,
+      "learning_rate": 1.4458377640886682e-05,
+      "loss": 0.4692,
+      "step": 815
+    },
+    {
+      "epoch": 0.3695129199332069,
+      "grad_norm": 0.8632165789604187,
+      "learning_rate": 1.4457047645494645e-05,
+      "loss": 0.4475,
+      "step": 816
+    },
+    {
+      "epoch": 0.3699657543939094,
+      "grad_norm": 0.875834047794342,
+      "learning_rate": 1.4455716080476068e-05,
+      "loss": 0.4508,
+      "step": 817
+    },
+    {
+      "epoch": 0.3704185888546118,
+      "grad_norm": 0.7545289397239685,
+      "learning_rate": 1.4454382946131373e-05,
+      "loss": 0.4841,
+      "step": 818
+    },
+    {
+      "epoch": 0.3708714233153143,
+      "grad_norm": 0.8614468574523926,
+      "learning_rate": 1.4453048242761337e-05,
+      "loss": 0.477,
+      "step": 819
+    },
+    {
+      "epoch": 0.37132425777601674,
+      "grad_norm": 0.9642696380615234,
+      "learning_rate": 1.4451711970667094e-05,
+      "loss": 0.4879,
+      "step": 820
+    },
+    {
+      "epoch": 0.37177709223671923,
+      "grad_norm": 0.7030223608016968,
+      "learning_rate": 1.4450374130150124e-05,
+      "loss": 0.4212,
+      "step": 821
+    },
+    {
+      "epoch": 0.37222992669742166,
+      "grad_norm": 0.9804565906524658,
+      "learning_rate": 1.4449034721512271e-05,
+      "loss": 0.4439,
+      "step": 822
+    },
+    {
+      "epoch": 0.37268276115812415,
+      "grad_norm": 0.732269287109375,
+      "learning_rate": 1.4447693745055728e-05,
+      "loss": 0.4869,
+      "step": 823
+    },
+    {
+      "epoch": 0.3731355956188266,
+      "grad_norm": 0.914057195186615,
+      "learning_rate": 1.444635120108304e-05,
+      "loss": 0.4113,
+      "step": 824
+    },
+    {
+      "epoch": 0.3735884300795291,
+      "grad_norm": 0.8915912508964539,
+      "learning_rate": 1.4445007089897108e-05,
+      "loss": 0.4566,
+      "step": 825
+    },
+    {
+      "epoch": 0.3740412645402315,
+      "grad_norm": 0.803848147392273,
+      "learning_rate": 1.4443661411801183e-05,
+      "loss": 0.4411,
+      "step": 826
+    },
+    {
+      "epoch": 0.37449409900093394,
+      "grad_norm": 0.8054463267326355,
+      "learning_rate": 1.4442314167098878e-05,
+      "loss": 0.4521,
+      "step": 827
+    },
+    {
+      "epoch": 0.37494693346163643,
+      "grad_norm": 0.6809094548225403,
+      "learning_rate": 1.444096535609415e-05,
+      "loss": 0.4083,
+      "step": 828
+    },
+    {
+      "epoch": 0.37494693346163643,
+      "eval_loss": 0.4509795606136322,
+      "eval_runtime": 19.4619,
+      "eval_samples_per_second": 5.858,
+      "eval_steps_per_second": 0.771,
+      "step": 828
+    },
+    {
+      "epoch": 0.37539976792233887,
+      "grad_norm": 0.759226381778717,
+      "learning_rate": 1.4439614979091316e-05,
+      "loss": 0.3904,
+      "step": 829
+    },
+    {
+      "epoch": 0.37585260238304136,
+      "grad_norm": 0.7871491312980652,
+      "learning_rate": 1.4438263036395045e-05,
+      "loss": 0.4172,
+      "step": 830
+    },
+    {
+      "epoch": 0.3763054368437438,
+      "grad_norm": 0.6610317230224609,
+      "learning_rate": 1.4436909528310352e-05,
+      "loss": 0.4328,
+      "step": 831
+    },
+    {
+      "epoch": 0.3767582713044463,
+      "grad_norm": 0.7470182776451111,
+      "learning_rate": 1.4435554455142614e-05,
+      "loss": 0.4792,
+      "step": 832
+    },
+    {
+      "epoch": 0.3772111057651487,
+      "grad_norm": 0.7115527391433716,
+      "learning_rate": 1.4434197817197558e-05,
+      "loss": 0.4253,
+      "step": 833
+    },
+    {
+      "epoch": 0.3776639402258512,
+      "grad_norm": 0.8027327656745911,
+      "learning_rate": 1.4432839614781268e-05,
+      "loss": 0.4349,
+      "step": 834
+    },
+    {
+      "epoch": 0.37811677468655364,
+      "grad_norm": 0.6858651041984558,
+      "learning_rate": 1.4431479848200174e-05,
+      "loss": 0.464,
+      "step": 835
+    },
+    {
+      "epoch": 0.3785696091472561,
+      "grad_norm": 0.780019998550415,
+      "learning_rate": 1.443011851776106e-05,
+      "loss": 0.4301,
+      "step": 836
+    },
+    {
+      "epoch": 0.37902244360795856,
+      "grad_norm": 0.8875818252563477,
+      "learning_rate": 1.4428755623771068e-05,
+      "loss": 0.4638,
+      "step": 837
+    },
+    {
+      "epoch": 0.37947527806866105,
+      "grad_norm": 0.7595096230506897,
+      "learning_rate": 1.442739116653769e-05,
+      "loss": 0.4761,
+      "step": 838
+    },
+    {
+      "epoch": 0.3799281125293635,
+      "grad_norm": 0.7235725522041321,
+      "learning_rate": 1.4426025146368768e-05,
+      "loss": 0.4596,
+      "step": 839
+    },
+    {
+      "epoch": 0.38038094699006597,
+      "grad_norm": 0.8074544072151184,
+      "learning_rate": 1.4424657563572499e-05,
+      "loss": 0.5049,
+      "step": 840
+    },
+    {
+      "epoch": 0.3808337814507684,
+      "grad_norm": 0.784801185131073,
+      "learning_rate": 1.4423288418457437e-05,
+      "loss": 0.4387,
+      "step": 841
+    },
+    {
+      "epoch": 0.38128661591147084,
+      "grad_norm": 0.6974912285804749,
+      "learning_rate": 1.442191771133248e-05,
+      "loss": 0.4004,
+      "step": 842
+    },
+    {
+      "epoch": 0.38173945037217333,
+      "grad_norm": 0.6632540822029114,
+      "learning_rate": 1.4420545442506885e-05,
+      "loss": 0.4515,
+      "step": 843
+    },
+    {
+      "epoch": 0.38219228483287576,
+      "grad_norm": 0.7197034358978271,
+      "learning_rate": 1.4419171612290256e-05,
+      "loss": 0.4033,
+      "step": 844
+    },
+    {
+      "epoch": 0.38264511929357825,
+      "grad_norm": 0.6470990180969238,
+      "learning_rate": 1.4417796220992553e-05,
+      "loss": 0.4453,
+      "step": 845
+    },
+    {
+      "epoch": 0.3830979537542807,
+      "grad_norm": 0.7469446659088135,
+      "learning_rate": 1.441641926892409e-05,
+      "loss": 0.4657,
+      "step": 846
+    },
+    {
+      "epoch": 0.3835507882149832,
+      "grad_norm": 0.7464520335197449,
+      "learning_rate": 1.4415040756395527e-05,
+      "loss": 0.4376,
+      "step": 847
+    },
+    {
+      "epoch": 0.3840036226756856,
+      "grad_norm": 0.800579309463501,
+      "learning_rate": 1.441366068371788e-05,
+      "loss": 0.4434,
+      "step": 848
+    },
+    {
+      "epoch": 0.3844564571363881,
+      "grad_norm": 0.6948139071464539,
+      "learning_rate": 1.4412279051202518e-05,
+      "loss": 0.3879,
+      "step": 849
+    },
+    {
+      "epoch": 0.38490929159709053,
+      "grad_norm": 0.6828701496124268,
+      "learning_rate": 1.441089585916116e-05,
+      "loss": 0.4162,
+      "step": 850
+    },
+    {
+      "epoch": 0.385362126057793,
+      "grad_norm": 0.9038822054862976,
+      "learning_rate": 1.4409511107905878e-05,
+      "loss": 0.457,
+      "step": 851
+    },
+    {
+      "epoch": 0.38581496051849545,
+      "grad_norm": 0.7000515460968018,
+      "learning_rate": 1.4408124797749093e-05,
+      "loss": 0.4054,
+      "step": 852
+    },
+    {
+      "epoch": 0.38626779497919794,
+      "grad_norm": 0.8235939145088196,
+      "learning_rate": 1.4406736929003582e-05,
+      "loss": 0.4283,
+      "step": 853
+    },
+    {
+      "epoch": 0.3867206294399004,
+      "grad_norm": 0.824226975440979,
+      "learning_rate": 1.4405347501982473e-05,
+      "loss": 0.4499,
+      "step": 854
+    },
+    {
+      "epoch": 0.3871734639006028,
+      "grad_norm": 0.7194985151290894,
+      "learning_rate": 1.4403956516999235e-05,
+      "loss": 0.4278,
+      "step": 855
+    },
+    {
+      "epoch": 0.3876262983613053,
+      "grad_norm": 0.7407351732254028,
+      "learning_rate": 1.4402563974367709e-05,
+      "loss": 0.4495,
+      "step": 856
+    },
+    {
+      "epoch": 0.38807913282200773,
+      "grad_norm": 0.7273187041282654,
+      "learning_rate": 1.4401169874402071e-05,
+      "loss": 0.389,
+      "step": 857
+    },
+    {
+      "epoch": 0.3885319672827102,
+      "grad_norm": 0.6498844623565674,
+      "learning_rate": 1.4399774217416852e-05,
+      "loss": 0.3919,
+      "step": 858
+    },
+    {
+      "epoch": 0.38898480174341266,
+      "grad_norm": 0.7037122845649719,
+      "learning_rate": 1.4398377003726936e-05,
+      "loss": 0.4601,
+      "step": 859
+    },
+    {
+      "epoch": 0.38943763620411515,
+      "grad_norm": 0.7716974020004272,
+      "learning_rate": 1.4396978233647561e-05,
+      "loss": 0.4147,
+      "step": 860
+    },
+    {
+      "epoch": 0.3898904706648176,
+      "grad_norm": 0.7838717699050903,
+      "learning_rate": 1.439557790749431e-05,
+      "loss": 0.4185,
+      "step": 861
+    },
+    {
+      "epoch": 0.39034330512552007,
+      "grad_norm": 0.7486699819564819,
+      "learning_rate": 1.439417602558312e-05,
+      "loss": 0.4397,
+      "step": 862
+    },
+    {
+      "epoch": 0.3907961395862225,
+      "grad_norm": 0.7644836902618408,
+      "learning_rate": 1.4392772588230281e-05,
+      "loss": 0.4814,
+      "step": 863
+    },
+    {
+      "epoch": 0.391248974046925,
+      "grad_norm": 0.8007204532623291,
+      "learning_rate": 1.439136759575243e-05,
+      "loss": 0.47,
+      "step": 864
+    },
+    {
+      "epoch": 0.3917018085076274,
+      "grad_norm": 0.7783535122871399,
+      "learning_rate": 1.4389961048466557e-05,
+      "loss": 0.3774,
+      "step": 865
+    },
+    {
+      "epoch": 0.3921546429683299,
+      "grad_norm": 0.7233867645263672,
+      "learning_rate": 1.4388552946690004e-05,
+      "loss": 0.4344,
+      "step": 866
+    },
+    {
+      "epoch": 0.39260747742903235,
+      "grad_norm": 0.6840167045593262,
+      "learning_rate": 1.4387143290740462e-05,
+      "loss": 0.4511,
+      "step": 867
+    },
+    {
+      "epoch": 0.3930603118897348,
+      "grad_norm": 0.8994132280349731,
+      "learning_rate": 1.4385732080935974e-05,
+      "loss": 0.4716,
+      "step": 868
+    },
+    {
+      "epoch": 0.39351314635043727,
+      "grad_norm": 0.8595420122146606,
+      "learning_rate": 1.4384319317594926e-05,
+      "loss": 0.4388,
+      "step": 869
+    },
+    {
+      "epoch": 0.3939659808111397,
+      "grad_norm": 0.675515353679657,
+      "learning_rate": 1.4382905001036073e-05,
+      "loss": 0.4134,
+      "step": 870
+    },
+    {
+      "epoch": 0.3944188152718422,
+      "grad_norm": 1.7675050497055054,
+      "learning_rate": 1.4381489131578499e-05,
+      "loss": 0.4454,
+      "step": 871
+    },
+    {
+      "epoch": 0.3948716497325446,
+      "grad_norm": 1.0099380016326904,
+      "learning_rate": 1.438007170954165e-05,
+      "loss": 0.4738,
+      "step": 872
+    },
+    {
+      "epoch": 0.3953244841932471,
+      "grad_norm": 0.7462349534034729,
+      "learning_rate": 1.437865273524532e-05,
+      "loss": 0.3707,
+      "step": 873
+    },
+    {
+      "epoch": 0.39577731865394955,
+      "grad_norm": 0.7054492235183716,
+      "learning_rate": 1.4377232209009657e-05,
+      "loss": 0.4068,
+      "step": 874
+    },
+    {
+      "epoch": 0.39623015311465204,
+      "grad_norm": 0.7035594582557678,
+      "learning_rate": 1.4375810131155152e-05,
+      "loss": 0.4456,
+      "step": 875
+    },
+    {
+      "epoch": 0.3966829875753545,
+      "grad_norm": 0.691320538520813,
+      "learning_rate": 1.4374386502002649e-05,
+      "loss": 0.4188,
+      "step": 876
+    },
+    {
+      "epoch": 0.39713582203605696,
+      "grad_norm": 0.7369733452796936,
+      "learning_rate": 1.4372961321873344e-05,
+      "loss": 0.4572,
+      "step": 877
+    },
+    {
+      "epoch": 0.3975886564967594,
+      "grad_norm": 0.687791109085083,
+      "learning_rate": 1.4371534591088785e-05,
+      "loss": 0.4136,
+      "step": 878
+    },
+    {
+      "epoch": 0.3980414909574619,
+      "grad_norm": 0.7917645573616028,
+      "learning_rate": 1.437010630997086e-05,
+      "loss": 0.4474,
+      "step": 879
+    },
+    {
+      "epoch": 0.3984943254181643,
+      "grad_norm": 0.7191473245620728,
+      "learning_rate": 1.4368676478841818e-05,
+      "loss": 0.4608,
+      "step": 880
+    },
+    {
+      "epoch": 0.3989471598788668,
+      "grad_norm": 0.7102738618850708,
+      "learning_rate": 1.4367245098024248e-05,
+      "loss": 0.4014,
+      "step": 881
+    },
+    {
+      "epoch": 0.39939999433956924,
+      "grad_norm": 0.8336265683174133,
+      "learning_rate": 1.4365812167841096e-05,
+      "loss": 0.4597,
+      "step": 882
+    },
+    {
+      "epoch": 0.3998528288002717,
+      "grad_norm": 0.6486582159996033,
+      "learning_rate": 1.436437768861566e-05,
+      "loss": 0.4273,
+      "step": 883
+    },
+    {
+      "epoch": 0.40030566326097416,
+      "grad_norm": 0.779009222984314,
+      "learning_rate": 1.4362941660671574e-05,
+      "loss": 0.4285,
+      "step": 884
+    },
+    {
+      "epoch": 0.4007584977216766,
+      "grad_norm": 0.8129340410232544,
+      "learning_rate": 1.4361504084332835e-05,
+      "loss": 0.4729,
+      "step": 885
+    },
+    {
+      "epoch": 0.4012113321823791,
+      "grad_norm": 0.765630841255188,
+      "learning_rate": 1.4360064959923784e-05,
+      "loss": 0.4146,
+      "step": 886
+    },
+    {
+      "epoch": 0.4016641666430815,
+      "grad_norm": 0.7774229049682617,
+      "learning_rate": 1.435862428776911e-05,
+      "loss": 0.442,
+      "step": 887
+    },
+    {
+      "epoch": 0.402117001103784,
+      "grad_norm": 0.7265660166740417,
+      "learning_rate": 1.4357182068193856e-05,
+      "loss": 0.4161,
+      "step": 888
+    },
+    {
+      "epoch": 0.40256983556448644,
+      "grad_norm": 0.6984989643096924,
+      "learning_rate": 1.4355738301523405e-05,
+      "loss": 0.4508,
+      "step": 889
+    },
+    {
+      "epoch": 0.40302267002518893,
+      "grad_norm": 0.8146981596946716,
+      "learning_rate": 1.4354292988083503e-05,
+      "loss": 0.3812,
+      "step": 890
+    },
+    {
+      "epoch": 0.40347550448589137,
+      "grad_norm": 0.7826554775238037,
+      "learning_rate": 1.435284612820023e-05,
+      "loss": 0.4889,
+      "step": 891
+    },
+    {
+      "epoch": 0.40392833894659386,
+      "grad_norm": 0.7819204926490784,
+      "learning_rate": 1.4351397722200023e-05,
+      "loss": 0.4392,
+      "step": 892
+    },
+    {
+      "epoch": 0.4043811734072963,
+      "grad_norm": 0.851521909236908,
+      "learning_rate": 1.4349947770409669e-05,
+      "loss": 0.457,
+      "step": 893
+    },
+    {
+      "epoch": 0.4048340078679988,
+      "grad_norm": 0.751393735408783,
+      "learning_rate": 1.4348496273156302e-05,
+      "loss": 0.4016,
+      "step": 894
+    },
+    {
+      "epoch": 0.4052868423287012,
+      "grad_norm": 0.7438313961029053,
+      "learning_rate": 1.4347043230767402e-05,
+      "loss": 0.4529,
+      "step": 895
+    },
+    {
+      "epoch": 0.40573967678940365,
+      "grad_norm": 0.8485840559005737,
+      "learning_rate": 1.43455886435708e-05,
+      "loss": 0.4588,
+      "step": 896
+    },
+    {
+      "epoch": 0.40619251125010614,
+      "grad_norm": 0.6871113777160645,
+      "learning_rate": 1.4344132511894673e-05,
+      "loss": 0.4313,
+      "step": 897
+    },
+    {
+      "epoch": 0.40664534571080857,
+      "grad_norm": 0.7333969473838806,
+      "learning_rate": 1.4342674836067553e-05,
+      "loss": 0.4039,
+      "step": 898
+    },
+    {
+      "epoch": 0.40709818017151106,
+      "grad_norm": 0.7326341867446899,
+      "learning_rate": 1.4341215616418316e-05,
+      "loss": 0.4024,
+      "step": 899
+    },
+    {
+      "epoch": 0.4075510146322135,
+      "grad_norm": 0.754438042640686,
+      "learning_rate": 1.4339754853276182e-05,
+      "loss": 0.4381,
+      "step": 900
+    },
+    {
+      "epoch": 0.408003849092916,
+      "grad_norm": 0.7537654638290405,
+      "learning_rate": 1.4338292546970729e-05,
+      "loss": 0.3864,
+      "step": 901
+    },
+    {
+      "epoch": 0.4084566835536184,
+      "grad_norm": 0.7983335852622986,
+      "learning_rate": 1.4336828697831873e-05,
+      "loss": 0.465,
+      "step": 902
+    },
+    {
+      "epoch": 0.4089095180143209,
+      "grad_norm": 0.6547751426696777,
+      "learning_rate": 1.4335363306189887e-05,
+      "loss": 0.4017,
+      "step": 903
+    },
+    {
+      "epoch": 0.40936235247502334,
+      "grad_norm": 0.8735558390617371,
+      "learning_rate": 1.4333896372375383e-05,
+      "loss": 0.4062,
+      "step": 904
+    },
+    {
+      "epoch": 0.40981518693572583,
+      "grad_norm": 0.6709791421890259,
+      "learning_rate": 1.4332427896719331e-05,
+      "loss": 0.4209,
+      "step": 905
+    },
+    {
+      "epoch": 0.41026802139642826,
+      "grad_norm": 1.0297253131866455,
+      "learning_rate": 1.433095787955304e-05,
+      "loss": 0.4074,
+      "step": 906
+    },
+    {
+      "epoch": 0.41072085585713075,
+      "grad_norm": 0.6717653274536133,
+      "learning_rate": 1.4329486321208176e-05,
+      "loss": 0.4291,
+      "step": 907
+    },
+    {
+      "epoch": 0.4111736903178332,
+      "grad_norm": 0.789406418800354,
+      "learning_rate": 1.4328013222016739e-05,
+      "loss": 0.4289,
+      "step": 908
+    },
+    {
+      "epoch": 0.4116265247785357,
+      "grad_norm": 0.7695425152778625,
+      "learning_rate": 1.432653858231109e-05,
+      "loss": 0.3966,
+      "step": 909
+    },
+    {
+      "epoch": 0.4120793592392381,
+      "grad_norm": 0.8856887817382812,
+      "learning_rate": 1.4325062402423932e-05,
+      "loss": 0.3951,
+      "step": 910
+    },
+    {
+      "epoch": 0.41253219369994054,
+      "grad_norm": 0.8140108585357666,
+      "learning_rate": 1.4323584682688317e-05,
+      "loss": 0.4092,
+      "step": 911
+    },
+    {
+      "epoch": 0.41298502816064303,
+      "grad_norm": 0.8315860033035278,
+      "learning_rate": 1.4322105423437639e-05,
+      "loss": 0.4223,
+      "step": 912
+    },
+    {
+      "epoch": 0.41343786262134546,
+      "grad_norm": 0.7078157067298889,
+      "learning_rate": 1.4320624625005648e-05,
+      "loss": 0.4425,
+      "step": 913
+    },
+    {
+      "epoch": 0.41389069708204795,
+      "grad_norm": 0.740716814994812,
+      "learning_rate": 1.4319142287726436e-05,
+      "loss": 0.417,
+      "step": 914
+    },
+    {
+      "epoch": 0.4143435315427504,
+      "grad_norm": 0.7005321979522705,
+      "learning_rate": 1.4317658411934442e-05,
+      "loss": 0.3627,
+      "step": 915
+    },
+    {
+      "epoch": 0.4147963660034529,
+      "grad_norm": 0.6697372794151306,
+      "learning_rate": 1.4316172997964452e-05,
+      "loss": 0.4012,
+      "step": 916
+    },
+    {
+      "epoch": 0.4152492004641553,
+      "grad_norm": 0.9033950567245483,
+      "learning_rate": 1.4314686046151605e-05,
+      "loss": 0.4569,
+      "step": 917
+    },
+    {
+      "epoch": 0.4157020349248578,
+      "grad_norm": 0.7096682786941528,
+      "learning_rate": 1.4313197556831378e-05,
+      "loss": 0.4487,
+      "step": 918
+    },
+    {
+      "epoch": 0.41615486938556023,
+      "grad_norm": 0.7417932152748108,
+      "learning_rate": 1.4311707530339601e-05,
+      "loss": 0.4104,
+      "step": 919
+    },
+    {
+      "epoch": 0.4166077038462627,
+      "grad_norm": 0.6767085790634155,
+      "learning_rate": 1.4310215967012449e-05,
+      "loss": 0.4512,
+      "step": 920
+    },
+    {
+      "epoch": 0.41706053830696516,
+      "grad_norm": 0.9227702617645264,
+      "learning_rate": 1.4308722867186442e-05,
+      "loss": 0.3907,
+      "step": 921
+    },
+    {
+      "epoch": 0.41751337276766765,
+      "grad_norm": 0.7554175853729248,
+      "learning_rate": 1.430722823119845e-05,
+      "loss": 0.4507,
+      "step": 922
+    },
+    {
+      "epoch": 0.4179662072283701,
+      "grad_norm": 1.1104964017868042,
+      "learning_rate": 1.4305732059385687e-05,
+      "loss": 0.3809,
+      "step": 923
+    },
+    {
+      "epoch": 0.4184190416890725,
+      "grad_norm": 0.6951049566268921,
+      "learning_rate": 1.4304234352085715e-05,
+      "loss": 0.4229,
+      "step": 924
+    },
+    {
+      "epoch": 0.418871876149775,
+      "grad_norm": 0.9023833870887756,
+      "learning_rate": 1.4302735109636442e-05,
+      "loss": 0.4166,
+      "step": 925
+    },
+    {
+      "epoch": 0.41932471061047744,
+      "grad_norm": 0.6952568888664246,
+      "learning_rate": 1.430123433237612e-05,
+      "loss": 0.3956,
+      "step": 926
+    },
+    {
+      "epoch": 0.4197775450711799,
+      "grad_norm": 0.8569507002830505,
+      "learning_rate": 1.4299732020643354e-05,
+      "loss": 0.4372,
+      "step": 927
+    },
+    {
+      "epoch": 0.42023037953188236,
+      "grad_norm": 0.8564557433128357,
+      "learning_rate": 1.4298228174777086e-05,
+      "loss": 0.4311,
+      "step": 928
+    },
+    {
+      "epoch": 0.42068321399258485,
+      "grad_norm": 0.9847443699836731,
+      "learning_rate": 1.4296722795116614e-05,
+      "loss": 0.4198,
+      "step": 929
+    },
+    {
+      "epoch": 0.4211360484532873,
+      "grad_norm": 0.9145928621292114,
+      "learning_rate": 1.4295215882001572e-05,
+      "loss": 0.4092,
+      "step": 930
+    },
+    {
+      "epoch": 0.42158888291398977,
+      "grad_norm": 1.0060977935791016,
+      "learning_rate": 1.4293707435771947e-05,
+      "loss": 0.38,
+      "step": 931
+    },
+    {
+      "epoch": 0.4220417173746922,
+      "grad_norm": 0.8660867810249329,
+      "learning_rate": 1.4292197456768069e-05,
+      "loss": 0.4302,
+      "step": 932
+    },
+    {
+      "epoch": 0.4224945518353947,
+      "grad_norm": 0.7605267763137817,
+      "learning_rate": 1.4290685945330615e-05,
+      "loss": 0.4346,
+      "step": 933
+    },
+    {
+      "epoch": 0.4229473862960971,
+      "grad_norm": 0.9517015814781189,
+      "learning_rate": 1.4289172901800606e-05,
+      "loss": 0.4111,
+      "step": 934
+    },
+    {
+      "epoch": 0.4234002207567996,
+      "grad_norm": 0.7545219659805298,
+      "learning_rate": 1.4287658326519414e-05,
+      "loss": 0.4231,
+      "step": 935
+    },
+    {
+      "epoch": 0.42385305521750205,
+      "grad_norm": 0.7649204134941101,
+      "learning_rate": 1.4286142219828748e-05,
+      "loss": 0.4532,
+      "step": 936
+    },
+    {
+      "epoch": 0.4243058896782045,
+      "grad_norm": 0.8143602013587952,
+      "learning_rate": 1.428462458207067e-05,
+      "loss": 0.4212,
+      "step": 937
+    },
+    {
+      "epoch": 0.424758724138907,
+      "grad_norm": 0.9045494794845581,
+      "learning_rate": 1.4283105413587581e-05,
+      "loss": 0.4313,
+      "step": 938
+    },
+    {
+      "epoch": 0.4252115585996094,
+      "grad_norm": 0.9623934030532837,
+      "learning_rate": 1.4281584714722237e-05,
+      "loss": 0.4485,
+      "step": 939
+    },
+    {
+      "epoch": 0.4256643930603119,
+      "grad_norm": 0.7864290475845337,
+      "learning_rate": 1.4280062485817727e-05,
+      "loss": 0.4093,
+      "step": 940
+    },
+    {
+      "epoch": 0.42611722752101433,
+      "grad_norm": 0.6734517812728882,
+      "learning_rate": 1.4278538727217495e-05,
+      "loss": 0.4376,
+      "step": 941
+    },
+    {
+      "epoch": 0.4265700619817168,
+      "grad_norm": 0.7887140512466431,
+      "learning_rate": 1.4277013439265325e-05,
+      "loss": 0.3778,
+      "step": 942
+    },
+    {
+      "epoch": 0.42702289644241925,
+      "grad_norm": 0.8578718900680542,
+      "learning_rate": 1.4275486622305348e-05,
+      "loss": 0.4567,
+      "step": 943
+    },
+    {
+      "epoch": 0.42747573090312174,
+      "grad_norm": 0.775020956993103,
+      "learning_rate": 1.427395827668204e-05,
+      "loss": 0.4177,
+      "step": 944
+    },
+    {
+      "epoch": 0.4279285653638242,
+      "grad_norm": 0.7445123791694641,
+      "learning_rate": 1.427242840274022e-05,
+      "loss": 0.4109,
+      "step": 945
+    },
+    {
+      "epoch": 0.42838139982452667,
+      "grad_norm": 0.660199761390686,
+      "learning_rate": 1.4270897000825052e-05,
+      "loss": 0.3661,
+      "step": 946
+    },
+    {
+      "epoch": 0.4288342342852291,
+      "grad_norm": 0.960015594959259,
+      "learning_rate": 1.426936407128205e-05,
+      "loss": 0.4006,
+      "step": 947
+    },
+    {
+      "epoch": 0.4292870687459316,
+      "grad_norm": 0.7209600806236267,
+      "learning_rate": 1.4267829614457067e-05,
+      "loss": 0.4041,
+      "step": 948
+    },
+    {
+      "epoch": 0.429739903206634,
+      "grad_norm": 0.7578160762786865,
+      "learning_rate": 1.42662936306963e-05,
+      "loss": 0.3883,
+      "step": 949
+    },
+    {
+      "epoch": 0.4301927376673365,
+      "grad_norm": 0.7112842798233032,
+      "learning_rate": 1.4264756120346294e-05,
+      "loss": 0.4147,
+      "step": 950
+    },
+    {
+      "epoch": 0.43064557212803894,
+      "grad_norm": 0.8537875413894653,
+      "learning_rate": 1.4263217083753938e-05,
+      "loss": 0.4498,
+      "step": 951
+    },
+    {
+      "epoch": 0.4310984065887414,
+      "grad_norm": 0.6418316960334778,
+      "learning_rate": 1.4261676521266466e-05,
+      "loss": 0.3858,
+      "step": 952
+    },
+    {
+      "epoch": 0.43155124104944387,
+      "grad_norm": 0.8194690346717834,
+      "learning_rate": 1.4260134433231449e-05,
+      "loss": 0.3749,
+      "step": 953
+    },
+    {
+      "epoch": 0.4320040755101463,
+      "grad_norm": 1.0442348718643188,
+      "learning_rate": 1.4258590819996813e-05,
+      "loss": 0.3945,
+      "step": 954
+    },
+    {
+      "epoch": 0.4324569099708488,
+      "grad_norm": 0.778914213180542,
+      "learning_rate": 1.4257045681910822e-05,
+      "loss": 0.4199,
+      "step": 955
+    },
+    {
+      "epoch": 0.4329097444315512,
+      "grad_norm": 0.7745048999786377,
+      "learning_rate": 1.4255499019322084e-05,
+      "loss": 0.4514,
+      "step": 956
+    },
+    {
+      "epoch": 0.4333625788922537,
+      "grad_norm": 0.7798154354095459,
+      "learning_rate": 1.4253950832579552e-05,
+      "loss": 0.4377,
+      "step": 957
+    },
+    {
+      "epoch": 0.43381541335295615,
+      "grad_norm": 0.7294392585754395,
+      "learning_rate": 1.4252401122032525e-05,
+      "loss": 0.3638,
+      "step": 958
+    },
+    {
+      "epoch": 0.43426824781365864,
+      "grad_norm": 0.795789361000061,
+      "learning_rate": 1.4250849888030641e-05,
+      "loss": 0.4311,
+      "step": 959
+    },
+    {
+      "epoch": 0.43472108227436107,
+      "grad_norm": 1.5630682706832886,
+      "learning_rate": 1.4249297130923885e-05,
+      "loss": 0.396,
+      "step": 960
+    },
+    {
+      "epoch": 0.43517391673506356,
+      "grad_norm": 0.7337138056755066,
+      "learning_rate": 1.4247742851062587e-05,
+      "loss": 0.4561,
+      "step": 961
+    },
+    {
+      "epoch": 0.435626751195766,
+      "grad_norm": 0.7088420987129211,
+      "learning_rate": 1.4246187048797414e-05,
+      "loss": 0.4247,
+      "step": 962
+    },
+    {
+      "epoch": 0.4360795856564685,
+      "grad_norm": 0.9262889623641968,
+      "learning_rate": 1.4244629724479387e-05,
+      "loss": 0.4006,
+      "step": 963
+    },
+    {
+      "epoch": 0.4365324201171709,
+      "grad_norm": 0.8591921329498291,
+      "learning_rate": 1.4243070878459862e-05,
+      "loss": 0.4003,
+      "step": 964
+    },
+    {
+      "epoch": 0.43698525457787335,
+      "grad_norm": 1.0036776065826416,
+      "learning_rate": 1.4241510511090539e-05,
+      "loss": 0.4098,
+      "step": 965
+    },
+    {
+      "epoch": 0.43743808903857584,
+      "grad_norm": 0.6864471435546875,
+      "learning_rate": 1.4239948622723465e-05,
+      "loss": 0.4326,
+      "step": 966
+    },
+    {
+      "epoch": 0.4378909234992783,
+      "grad_norm": 1.0244603157043457,
+      "learning_rate": 1.4238385213711027e-05,
+      "loss": 0.4419,
+      "step": 967
+    },
+    {
+      "epoch": 0.43834375795998076,
+      "grad_norm": 0.7093055844306946,
+      "learning_rate": 1.4236820284405956e-05,
+      "loss": 0.3951,
+      "step": 968
+    },
+    {
+      "epoch": 0.4387965924206832,
+      "grad_norm": 0.8618746995925903,
+      "learning_rate": 1.4235253835161329e-05,
+      "loss": 0.4756,
+      "step": 969
+    },
+    {
+      "epoch": 0.4392494268813857,
+      "grad_norm": 0.7317725419998169,
+      "learning_rate": 1.4233685866330561e-05,
+      "loss": 0.3789,
+      "step": 970
+    },
+    {
+      "epoch": 0.4397022613420881,
+      "grad_norm": 0.8535686731338501,
+      "learning_rate": 1.4232116378267412e-05,
+      "loss": 0.4096,
+      "step": 971
+    },
+    {
+      "epoch": 0.4401550958027906,
+      "grad_norm": 0.7266016006469727,
+      "learning_rate": 1.4230545371325985e-05,
+      "loss": 0.3706,
+      "step": 972
+    },
+    {
+      "epoch": 0.44060793026349304,
+      "grad_norm": 0.7952896356582642,
+      "learning_rate": 1.4228972845860725e-05,
+      "loss": 0.3392,
+      "step": 973
+    },
+    {
+      "epoch": 0.44106076472419553,
+      "grad_norm": 0.8086087107658386,
+      "learning_rate": 1.422739880222642e-05,
+      "loss": 0.3972,
+      "step": 974
+    },
+    {
+      "epoch": 0.44151359918489796,
+      "grad_norm": 0.8605599403381348,
+      "learning_rate": 1.4225823240778207e-05,
+      "loss": 0.422,
+      "step": 975
+    },
+    {
+      "epoch": 0.44196643364560045,
+      "grad_norm": 0.752622663974762,
+      "learning_rate": 1.4224246161871548e-05,
+      "loss": 0.3965,
+      "step": 976
+    },
+    {
+      "epoch": 0.4424192681063029,
+      "grad_norm": 0.8178388476371765,
+      "learning_rate": 1.4222667565862264e-05,
+      "loss": 0.4438,
+      "step": 977
+    },
+    {
+      "epoch": 0.4428721025670054,
+      "grad_norm": 0.7991887927055359,
+      "learning_rate": 1.4221087453106515e-05,
+      "loss": 0.3748,
+      "step": 978
+    },
+    {
+      "epoch": 0.4433249370277078,
+      "grad_norm": 0.7863441705703735,
+      "learning_rate": 1.4219505823960798e-05,
+      "loss": 0.3352,
+      "step": 979
+    },
+    {
+      "epoch": 0.44377777148841024,
+      "grad_norm": 0.7784522175788879,
+      "learning_rate": 1.4217922678781955e-05,
+      "loss": 0.4319,
+      "step": 980
+    },
+    {
+      "epoch": 0.44423060594911273,
+      "grad_norm": 0.8807750344276428,
+      "learning_rate": 1.4216338017927171e-05,
+      "loss": 0.4165,
+      "step": 981
+    },
+    {
+      "epoch": 0.44468344040981517,
+      "grad_norm": 0.6990790367126465,
+      "learning_rate": 1.4214751841753971e-05,
+      "loss": 0.4141,
+      "step": 982
+    },
+    {
+      "epoch": 0.44513627487051766,
+      "grad_norm": 0.9423395991325378,
+      "learning_rate": 1.4213164150620227e-05,
+      "loss": 0.3669,
+      "step": 983
+    },
+    {
+      "epoch": 0.4455891093312201,
+      "grad_norm": 0.7699115872383118,
+      "learning_rate": 1.4211574944884143e-05,
+      "loss": 0.4115,
+      "step": 984
+    },
+    {
+      "epoch": 0.4460419437919226,
+      "grad_norm": 0.8870624899864197,
+      "learning_rate": 1.4209984224904272e-05,
+      "loss": 0.3953,
+      "step": 985
+    },
+    {
+      "epoch": 0.446494778252625,
+      "grad_norm": 0.8935746550559998,
+      "learning_rate": 1.4208391991039509e-05,
+      "loss": 0.4049,
+      "step": 986
+    },
+    {
+      "epoch": 0.4469476127133275,
+      "grad_norm": 0.647849440574646,
+      "learning_rate": 1.4206798243649085e-05,
+      "loss": 0.3873,
+      "step": 987
+    },
+    {
+      "epoch": 0.44740044717402994,
+      "grad_norm": 0.7980324625968933,
+      "learning_rate": 1.4205202983092583e-05,
+      "loss": 0.4253,
+      "step": 988
+    },
+    {
+      "epoch": 0.4478532816347324,
+      "grad_norm": 0.6993192434310913,
+      "learning_rate": 1.4203606209729912e-05,
+      "loss": 0.409,
+      "step": 989
+    },
+    {
+      "epoch": 0.44830611609543486,
+      "grad_norm": 0.7574604153633118,
+      "learning_rate": 1.4202007923921338e-05,
+      "loss": 0.345,
+      "step": 990
+    },
+    {
+      "epoch": 0.44875895055613735,
+      "grad_norm": 0.8472474217414856,
+      "learning_rate": 1.4200408126027458e-05,
+      "loss": 0.3847,
+      "step": 991
+    },
+    {
+      "epoch": 0.4492117850168398,
+      "grad_norm": 0.7414373159408569,
+      "learning_rate": 1.4198806816409211e-05,
+      "loss": 0.3923,
+      "step": 992
+    },
+    {
+      "epoch": 0.4496646194775422,
+      "grad_norm": 0.6760424971580505,
+      "learning_rate": 1.4197203995427883e-05,
+      "loss": 0.3471,
+      "step": 993
+    },
+    {
+      "epoch": 0.4501174539382447,
+      "grad_norm": 0.6407291889190674,
+      "learning_rate": 1.4195599663445096e-05,
+      "loss": 0.4312,
+      "step": 994
+    },
+    {
+      "epoch": 0.45057028839894714,
+      "grad_norm": 0.6934922337532043,
+      "learning_rate": 1.4193993820822815e-05,
+      "loss": 0.4067,
+      "step": 995
+    },
+    {
+      "epoch": 0.45102312285964963,
+      "grad_norm": 0.7909060716629028,
+      "learning_rate": 1.4192386467923343e-05,
+      "loss": 0.4303,
+      "step": 996
+    },
+    {
+      "epoch": 0.45147595732035206,
+      "grad_norm": 0.7516681551933289,
+      "learning_rate": 1.4190777605109326e-05,
+      "loss": 0.4274,
+      "step": 997
+    },
+    {
+      "epoch": 0.45192879178105455,
+      "grad_norm": 0.7599420547485352,
+      "learning_rate": 1.4189167232743754e-05,
+      "loss": 0.3949,
+      "step": 998
+    },
+    {
+      "epoch": 0.452381626241757,
+      "grad_norm": 0.7314419150352478,
+      "learning_rate": 1.4187555351189948e-05,
+      "loss": 0.4176,
+      "step": 999
+    },
+    {
+      "epoch": 0.4528344607024595,
+      "grad_norm": 0.6400272846221924,
+      "learning_rate": 1.418594196081158e-05,
+      "loss": 0.3949,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4532872951631619,
+      "grad_norm": 0.8566839098930359,
+      "learning_rate": 1.4184327061972657e-05,
+      "loss": 0.4034,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4537401296238644,
+      "grad_norm": 0.7478398084640503,
+      "learning_rate": 1.4182710655037527e-05,
+      "loss": 0.4593,
+      "step": 1002
+    },
+    {
+      "epoch": 0.45419296408456683,
+      "grad_norm": 0.8677337169647217,
+      "learning_rate": 1.4181092740370878e-05,
+      "loss": 0.432,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4546457985452693,
+      "grad_norm": 0.7963004112243652,
+      "learning_rate": 1.4179473318337738e-05,
+      "loss": 0.3899,
+      "step": 1004
+    },
+    {
+      "epoch": 0.45509863300597175,
+      "grad_norm": 0.7104138135910034,
+      "learning_rate": 1.4177852389303479e-05,
+      "loss": 0.3823,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4555514674666742,
+      "grad_norm": 0.6873833537101746,
+      "learning_rate": 1.4176229953633806e-05,
+      "loss": 0.3799,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4560043019273767,
+      "grad_norm": 0.8924839496612549,
+      "learning_rate": 1.4174606011694767e-05,
+      "loss": 0.4167,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4564571363880791,
+      "grad_norm": 0.8046056032180786,
+      "learning_rate": 1.4172980563852757e-05,
+      "loss": 0.3823,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4569099708487816,
+      "grad_norm": 0.7857614159584045,
+      "learning_rate": 1.4171353610474496e-05,
+      "loss": 0.4105,
+      "step": 1009
+    },
+    {
+      "epoch": 0.45736280530948403,
+      "grad_norm": 0.715185284614563,
+      "learning_rate": 1.416972515192706e-05,
+      "loss": 0.3557,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4578156397701865,
+      "grad_norm": 0.7968368530273438,
+      "learning_rate": 1.416809518857785e-05,
+      "loss": 0.3438,
+      "step": 1011
+    },
+    {
+      "epoch": 0.45826847423088896,
+      "grad_norm": 0.9046311974525452,
+      "learning_rate": 1.4166463720794615e-05,
+      "loss": 0.4533,
+      "step": 1012
+    },
+    {
+      "epoch": 0.45872130869159145,
+      "grad_norm": 0.6228207349777222,
+      "learning_rate": 1.4164830748945445e-05,
+      "loss": 0.3698,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4591741431522939,
+      "grad_norm": 0.8371472954750061,
+      "learning_rate": 1.416319627339876e-05,
+      "loss": 0.4385,
+      "step": 1014
+    },
+    {
+      "epoch": 0.45962697761299637,
+      "grad_norm": 0.6872038841247559,
+      "learning_rate": 1.416156029452333e-05,
+      "loss": 0.3892,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4600798120736988,
+      "grad_norm": 0.7857869863510132,
+      "learning_rate": 1.4159922812688255e-05,
+      "loss": 0.4179,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4605326465344013,
+      "grad_norm": 0.8071677684783936,
+      "learning_rate": 1.4158283828262982e-05,
+      "loss": 0.4028,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4609854809951037,
+      "grad_norm": 0.7820652723312378,
+      "learning_rate": 1.415664334161729e-05,
+      "loss": 0.379,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4614383154558062,
+      "grad_norm": 0.7160912752151489,
+      "learning_rate": 1.4155001353121304e-05,
+      "loss": 0.393,
+      "step": 1019
+    },
+    {
+      "epoch": 0.46189114991650865,
+      "grad_norm": 0.7849195003509521,
+      "learning_rate": 1.415335786314548e-05,
+      "loss": 0.4273,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4623439843772111,
+      "grad_norm": 0.7774779200553894,
+      "learning_rate": 1.4151712872060623e-05,
+      "loss": 0.356,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46279681883791357,
+      "grad_norm": 0.7844209671020508,
+      "learning_rate": 1.4150066380237864e-05,
+      "loss": 0.4261,
+      "step": 1022
+    },
+    {
+      "epoch": 0.463249653298616,
+      "grad_norm": 0.840189516544342,
+      "learning_rate": 1.4148418388048683e-05,
+      "loss": 0.3801,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4637024877593185,
+      "grad_norm": 0.8453295230865479,
+      "learning_rate": 1.4146768895864892e-05,
+      "loss": 0.4049,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4641553222200209,
+      "grad_norm": 0.7747737169265747,
+      "learning_rate": 1.4145117904058645e-05,
+      "loss": 0.4002,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4646081566807234,
+      "grad_norm": 0.7298173904418945,
+      "learning_rate": 1.4143465413002435e-05,
+      "loss": 0.3816,
+      "step": 1026
+    },
+    {
+      "epoch": 0.46506099114142585,
+      "grad_norm": 0.8087685704231262,
+      "learning_rate": 1.4141811423069092e-05,
+      "loss": 0.3819,
+      "step": 1027
+    },
+    {
+      "epoch": 0.46551382560212834,
+      "grad_norm": 0.6976639032363892,
+      "learning_rate": 1.4140155934631783e-05,
+      "loss": 0.3915,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4659666600628308,
+      "grad_norm": 0.6582549214363098,
+      "learning_rate": 1.413849894806401e-05,
+      "loss": 0.3571,
+      "step": 1029
+    },
+    {
+      "epoch": 0.46641949452353326,
+      "grad_norm": 0.7097831964492798,
+      "learning_rate": 1.4136840463739623e-05,
+      "loss": 0.4004,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4668723289842357,
+      "grad_norm": 0.8044737577438354,
+      "learning_rate": 1.4135180482032801e-05,
+      "loss": 0.4087,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4673251634449382,
+      "grad_norm": 0.7350354194641113,
+      "learning_rate": 1.4133519003318063e-05,
+      "loss": 0.4101,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4677779979056406,
+      "grad_norm": 0.7550399899482727,
+      "learning_rate": 1.4131856027970269e-05,
+      "loss": 0.3657,
+      "step": 1033
+    },
+    {
+      "epoch": 0.46823083236634305,
+      "grad_norm": 0.65655118227005,
+      "learning_rate": 1.413019155636461e-05,
+      "loss": 0.373,
+      "step": 1034
+    },
+    {
+      "epoch": 0.46868366682704554,
+      "grad_norm": 0.7206447124481201,
+      "learning_rate": 1.4128525588876624e-05,
+      "loss": 0.353,
+      "step": 1035
+    },
+    {
+      "epoch": 0.469136501287748,
+      "grad_norm": 0.7007904052734375,
+      "learning_rate": 1.4126858125882176e-05,
+      "loss": 0.3668,
+      "step": 1036
+    },
+    {
+      "epoch": 0.46958933574845046,
+      "grad_norm": 0.9158049821853638,
+      "learning_rate": 1.4125189167757477e-05,
+      "loss": 0.3746,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4700421702091529,
+      "grad_norm": 0.8141727447509766,
+      "learning_rate": 1.412351871487907e-05,
+      "loss": 0.3728,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4704950046698554,
+      "grad_norm": 0.696324348449707,
+      "learning_rate": 1.4121846767623839e-05,
+      "loss": 0.3415,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4709478391305578,
+      "grad_norm": 0.8677844405174255,
+      "learning_rate": 1.4120173326369003e-05,
+      "loss": 0.3584,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4714006735912603,
+      "grad_norm": 0.8021166324615479,
+      "learning_rate": 1.4118498391492118e-05,
+      "loss": 0.3832,
+      "step": 1041
+    },
+    {
+      "epoch": 0.47185350805196274,
+      "grad_norm": 0.9213194847106934,
+      "learning_rate": 1.4116821963371075e-05,
+      "loss": 0.4097,
+      "step": 1042
+    },
+    {
+      "epoch": 0.47230634251266523,
+      "grad_norm": 0.8504592776298523,
+      "learning_rate": 1.4115144042384107e-05,
+      "loss": 0.4031,
+      "step": 1043
+    },
+    {
+      "epoch": 0.47275917697336767,
+      "grad_norm": 0.746096134185791,
+      "learning_rate": 1.411346462890978e-05,
+      "loss": 0.4064,
+      "step": 1044
+    },
+    {
+      "epoch": 0.47321201143407016,
+      "grad_norm": 0.9829161167144775,
+      "learning_rate": 1.4111783723327e-05,
+      "loss": 0.409,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4736648458947726,
+      "grad_norm": 0.7933005690574646,
+      "learning_rate": 1.4110101326015007e-05,
+      "loss": 0.4209,
+      "step": 1046
+    },
+    {
+      "epoch": 0.474117680355475,
+      "grad_norm": 0.8225368857383728,
+      "learning_rate": 1.4108417437353374e-05,
+      "loss": 0.3688,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4745705148161775,
+      "grad_norm": 0.6937693357467651,
+      "learning_rate": 1.4106732057722018e-05,
+      "loss": 0.4077,
+      "step": 1048
+    },
+    {
+      "epoch": 0.47502334927687995,
+      "grad_norm": 0.7931678891181946,
+      "learning_rate": 1.410504518750119e-05,
+      "loss": 0.3273,
+      "step": 1049
+    },
+    {
+      "epoch": 0.47547618373758244,
+      "grad_norm": 0.7295666933059692,
+      "learning_rate": 1.4103356827071473e-05,
+      "loss": 0.3798,
+      "step": 1050
+    },
+    {
+      "epoch": 0.47592901819828487,
+      "grad_norm": 0.7447585463523865,
+      "learning_rate": 1.410166697681379e-05,
+      "loss": 0.3923,
+      "step": 1051
+    },
+    {
+      "epoch": 0.47638185265898736,
+      "grad_norm": 0.732587993144989,
+      "learning_rate": 1.4099975637109401e-05,
+      "loss": 0.3618,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4768346871196898,
+      "grad_norm": 0.6432980895042419,
+      "learning_rate": 1.4098282808339899e-05,
+      "loss": 0.3782,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4772875215803923,
+      "grad_norm": 0.9398018717765808,
+      "learning_rate": 1.4096588490887218e-05,
+      "loss": 0.4002,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4777403560410947,
+      "grad_norm": 0.845302164554596,
+      "learning_rate": 1.409489268513362e-05,
+      "loss": 0.3981,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4781931905017972,
+      "grad_norm": 0.7976647615432739,
+      "learning_rate": 1.409319539146171e-05,
+      "loss": 0.3828,
+      "step": 1056
+    },
+    {
+      "epoch": 0.47864602496249964,
+      "grad_norm": 0.7224624156951904,
+      "learning_rate": 1.4091496610254426e-05,
+      "loss": 0.4466,
+      "step": 1057
+    },
+    {
+      "epoch": 0.47909885942320213,
+      "grad_norm": 0.6982882618904114,
+      "learning_rate": 1.4089796341895036e-05,
+      "loss": 0.3732,
+      "step": 1058
+    },
+    {
+      "epoch": 0.47955169388390456,
+      "grad_norm": 0.6954671144485474,
+      "learning_rate": 1.408809458676716e-05,
+      "loss": 0.3646,
+      "step": 1059
+    },
+    {
+      "epoch": 0.48000452834460705,
+      "grad_norm": 0.6358652710914612,
+      "learning_rate": 1.4086391345254732e-05,
+      "loss": 0.382,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4804573628053095,
+      "grad_norm": 0.7169029116630554,
+      "learning_rate": 1.4084686617742037e-05,
+      "loss": 0.384,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4809101972660119,
+      "grad_norm": 0.7383198738098145,
+      "learning_rate": 1.4082980404613687e-05,
+      "loss": 0.3487,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4813630317267144,
+      "grad_norm": 0.6351798176765442,
+      "learning_rate": 1.4081272706254637e-05,
+      "loss": 0.3988,
+      "step": 1063
+    },
+    {
+      "epoch": 0.48181586618741684,
+      "grad_norm": 0.7320924401283264,
+      "learning_rate": 1.4079563523050169e-05,
+      "loss": 0.3706,
+      "step": 1064
+    },
+    {
+      "epoch": 0.48226870064811933,
+      "grad_norm": 0.7490411996841431,
+      "learning_rate": 1.4077852855385902e-05,
+      "loss": 0.3777,
+      "step": 1065
+    },
+    {
+      "epoch": 0.48272153510882176,
+      "grad_norm": 0.8109067678451538,
+      "learning_rate": 1.4076140703647794e-05,
+      "loss": 0.391,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48317436956952425,
+      "grad_norm": 0.7815183401107788,
+      "learning_rate": 1.4074427068222135e-05,
+      "loss": 0.3854,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4836272040302267,
+      "grad_norm": 0.8142592906951904,
+      "learning_rate": 1.407271194949555e-05,
+      "loss": 0.3907,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4840800384909292,
+      "grad_norm": 0.7322883009910583,
+      "learning_rate": 1.4070995347854997e-05,
+      "loss": 0.3879,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4845328729516316,
+      "grad_norm": 0.7850584983825684,
+      "learning_rate": 1.4069277263687772e-05,
+      "loss": 0.3516,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4849857074123341,
+      "grad_norm": 0.6609899401664734,
+      "learning_rate": 1.40675576973815e-05,
+      "loss": 0.3677,
+      "step": 1071
+    },
+    {
+      "epoch": 0.48543854187303653,
+      "grad_norm": 0.6708543300628662,
+      "learning_rate": 1.4065836649324148e-05,
+      "loss": 0.3708,
+      "step": 1072
+    },
+    {
+      "epoch": 0.485891376333739,
+      "grad_norm": 1.1243339776992798,
+      "learning_rate": 1.4064114119904012e-05,
+      "loss": 0.4169,
+      "step": 1073
+    },
+    {
+      "epoch": 0.48634421079444146,
+      "grad_norm": 0.7623947858810425,
+      "learning_rate": 1.4062390109509722e-05,
+      "loss": 0.3518,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4867970452551439,
+      "grad_norm": 0.7470114827156067,
+      "learning_rate": 1.4060664618530247e-05,
+      "loss": 0.3892,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4872498797158464,
+      "grad_norm": 0.8398137092590332,
+      "learning_rate": 1.4058937647354882e-05,
+      "loss": 0.4376,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4877027141765488,
+      "grad_norm": 0.7564401626586914,
+      "learning_rate": 1.4057209196373265e-05,
+      "loss": 0.3588,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4881555486372513,
+      "grad_norm": 0.7910377383232117,
+      "learning_rate": 1.4055479265975361e-05,
+      "loss": 0.3916,
+      "step": 1078
+    },
+    {
+      "epoch": 0.48860838309795374,
+      "grad_norm": 0.720329999923706,
+      "learning_rate": 1.405374785655147e-05,
+      "loss": 0.3732,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4890612175586562,
+      "grad_norm": 0.7369146347045898,
+      "learning_rate": 1.4052014968492232e-05,
+      "loss": 0.397,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48951405201935866,
+      "grad_norm": 0.6554039120674133,
+      "learning_rate": 1.4050280602188613e-05,
+      "loss": 0.3219,
+      "step": 1081
+    },
+    {
+      "epoch": 0.48996688648006115,
+      "grad_norm": 0.7499859929084778,
+      "learning_rate": 1.4048544758031912e-05,
+      "loss": 0.3816,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4904197209407636,
+      "grad_norm": 0.855495810508728,
+      "learning_rate": 1.4046807436413768e-05,
+      "loss": 0.3978,
+      "step": 1083
+    },
+    {
+      "epoch": 0.49087255540146607,
+      "grad_norm": 0.6773454546928406,
+      "learning_rate": 1.404506863772615e-05,
+      "loss": 0.389,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4913253898621685,
+      "grad_norm": 0.7844047546386719,
+      "learning_rate": 1.4043328362361359e-05,
+      "loss": 0.3594,
+      "step": 1085
+    },
+    {
+      "epoch": 0.491778224322871,
+      "grad_norm": 0.7174739241600037,
+      "learning_rate": 1.404158661071203e-05,
+      "loss": 0.3667,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4922310587835734,
+      "grad_norm": 0.6895620226860046,
+      "learning_rate": 1.4039843383171132e-05,
+      "loss": 0.4006,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4926838932442759,
+      "grad_norm": 0.6506296992301941,
+      "learning_rate": 1.4038098680131966e-05,
+      "loss": 0.3277,
+      "step": 1088
+    },
+    {
+      "epoch": 0.49313672770497835,
+      "grad_norm": 0.7601318359375,
+      "learning_rate": 1.4036352501988168e-05,
+      "loss": 0.3736,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4935895621656808,
+      "grad_norm": 0.6839517951011658,
+      "learning_rate": 1.4034604849133703e-05,
+      "loss": 0.3365,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4940423966263833,
+      "grad_norm": 0.6496476531028748,
+      "learning_rate": 1.4032855721962869e-05,
+      "loss": 0.3697,
+      "step": 1091
+    },
+    {
+      "epoch": 0.4944952310870857,
+      "grad_norm": 0.7376840710639954,
+      "learning_rate": 1.4031105120870303e-05,
+      "loss": 0.4017,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4949480655477882,
+      "grad_norm": 0.7958086133003235,
+      "learning_rate": 1.4029353046250965e-05,
+      "loss": 0.3657,
+      "step": 1093
+    },
+    {
+      "epoch": 0.49540090000849063,
+      "grad_norm": 0.6121701002120972,
+      "learning_rate": 1.4027599498500157e-05,
+      "loss": 0.3189,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4958537344691931,
+      "grad_norm": 0.7224218249320984,
+      "learning_rate": 1.4025844478013505e-05,
+      "loss": 0.3443,
+      "step": 1095
+    },
+    {
+      "epoch": 0.49630656892989555,
+      "grad_norm": 0.8465177416801453,
+      "learning_rate": 1.4024087985186975e-05,
+      "loss": 0.3789,
+      "step": 1096
+    },
+    {
+      "epoch": 0.49675940339059804,
+      "grad_norm": 0.6270837783813477,
+      "learning_rate": 1.4022330020416858e-05,
+      "loss": 0.3388,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4972122378513005,
+      "grad_norm": 0.6935149431228638,
+      "learning_rate": 1.4020570584099782e-05,
+      "loss": 0.336,
+      "step": 1098
+    },
+    {
+      "epoch": 0.49766507231200297,
+      "grad_norm": 0.7307567596435547,
+      "learning_rate": 1.4018809676632703e-05,
+      "loss": 0.4503,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4981179067727054,
+      "grad_norm": 0.7322489023208618,
+      "learning_rate": 1.4017047298412914e-05,
+      "loss": 0.3521,
+      "step": 1100
+    },
+    {
+      "epoch": 0.4985707412334079,
+      "grad_norm": 0.7978783249855042,
+      "learning_rate": 1.4015283449838037e-05,
+      "loss": 0.3992,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4990235756941103,
+      "grad_norm": 0.6273020505905151,
+      "learning_rate": 1.4013518131306024e-05,
+      "loss": 0.3647,
+      "step": 1102
+    },
+    {
+      "epoch": 0.49947641015481276,
+      "grad_norm": 0.6884061098098755,
+      "learning_rate": 1.401175134321516e-05,
+      "loss": 0.3903,
+      "step": 1103
+    },
+    {
+      "epoch": 0.49992924461551524,
+      "grad_norm": 0.7667953968048096,
+      "learning_rate": 1.4009983085964064e-05,
+      "loss": 0.3562,
+      "step": 1104
+    },
+    {
+      "epoch": 0.49992924461551524,
+      "eval_loss": 0.3898637592792511,
+      "eval_runtime": 19.4358,
+      "eval_samples_per_second": 5.865,
+      "eval_steps_per_second": 0.772,
+      "step": 1104
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6624,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 552,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.82217497538577e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}