diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.8139057755470276,
-  "best_model_checkpoint": "miner_id_24/checkpoint-50",
-  "epoch": 0.016080402010050253,
+  "best_metric": 0.7127780914306641,
+  "best_model_checkpoint": "miner_id_24/checkpoint-750",
+  "epoch": 0.2733668341708543,
   "eval_steps": 50,
-  "global_step": 50,
+  "global_step": 850,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -373,6 +373,5734 @@
       "eval_samples_per_second": 3.623,
       "eval_steps_per_second": 1.812,
       "step": 50
+    },
+    {
+      "epoch": 0.016402010050251256,
+      "grad_norm": 0.23711921274662018,
+      "learning_rate": 0.0001999904439512822,
+      "loss": 0.5316,
+      "step": 51
+    },
+    {
+      "epoch": 0.016723618090452263,
+      "grad_norm": 0.3278312683105469,
+      "learning_rate": 0.00019998997212570974,
+      "loss": 0.6492,
+      "step": 52
+    },
+    {
+      "epoch": 0.017045226130653266,
+      "grad_norm": 0.27989721298217773,
+      "learning_rate": 0.00019998948893161554,
+      "loss": 0.6844,
+      "step": 53
+    },
+    {
+      "epoch": 0.017366834170854273,
+      "grad_norm": 0.25756481289863586,
+      "learning_rate": 0.0001999889943690545,
+      "loss": 0.7775,
+      "step": 54
+    },
+    {
+      "epoch": 0.017688442211055276,
+      "grad_norm": 0.2182818502187729,
+      "learning_rate": 0.00019998848843808288,
+      "loss": 0.7043,
+      "step": 55
+    },
+    {
+      "epoch": 0.018010050251256283,
+      "grad_norm": 0.23244501650333405,
+      "learning_rate": 0.0001999879711387582,
+      "loss": 0.6046,
+      "step": 56
+    },
+    {
+      "epoch": 0.018331658291457286,
+      "grad_norm": 0.24752679467201233,
+      "learning_rate": 0.0001999874424711392,
+      "loss": 0.7545,
+      "step": 57
+    },
+    {
+      "epoch": 0.018653266331658293,
+      "grad_norm": 0.2313208132982254,
+      "learning_rate": 0.0001999869024352861,
+      "loss": 0.5923,
+      "step": 58
+    },
+    {
+      "epoch": 0.018974874371859296,
+      "grad_norm": 0.23669332265853882,
+      "learning_rate": 0.0001999863510312602,
+      "loss": 0.6044,
+      "step": 59
+    },
+    {
+      "epoch": 0.019296482412060303,
+      "grad_norm": 0.24155811965465546,
+      "learning_rate": 0.00019998578825912428,
+      "loss": 0.7805,
+      "step": 60
+    },
+    {
+      "epoch": 0.019618090452261306,
+      "grad_norm": 0.25742971897125244,
+      "learning_rate": 0.00019998521411894227,
+      "loss": 0.7484,
+      "step": 61
+    },
+    {
+      "epoch": 0.019939698492462313,
+      "grad_norm": 0.2815692126750946,
+      "learning_rate": 0.00019998462861077948,
+      "loss": 0.6499,
+      "step": 62
+    },
+    {
+      "epoch": 0.020261306532663316,
+      "grad_norm": 0.2727331221103668,
+      "learning_rate": 0.00019998403173470243,
+      "loss": 0.806,
+      "step": 63
+    },
+    {
+      "epoch": 0.020582914572864323,
+      "grad_norm": 0.27911487221717834,
+      "learning_rate": 0.00019998342349077905,
+      "loss": 0.9338,
+      "step": 64
+    },
+    {
+      "epoch": 0.020904522613065326,
+      "grad_norm": 0.26240643858909607,
+      "learning_rate": 0.00019998280387907843,
+      "loss": 0.7632,
+      "step": 65
+    },
+    {
+      "epoch": 0.021226130653266333,
+      "grad_norm": 0.2765254080295563,
+      "learning_rate": 0.00019998217289967109,
+      "loss": 0.833,
+      "step": 66
+    },
+    {
+      "epoch": 0.021547738693467336,
+      "grad_norm": 0.28112301230430603,
+      "learning_rate": 0.0001999815305526287,
+      "loss": 0.7094,
+      "step": 67
+    },
+    {
+      "epoch": 0.021869346733668343,
+      "grad_norm": 0.275939404964447,
+      "learning_rate": 0.00019998087683802433,
+      "loss": 0.8635,
+      "step": 68
+    },
+    {
+      "epoch": 0.022190954773869346,
+      "grad_norm": 0.26946502923965454,
+      "learning_rate": 0.0001999802117559323,
+      "loss": 0.7489,
+      "step": 69
+    },
+    {
+      "epoch": 0.022512562814070353,
+      "grad_norm": 0.2632855176925659,
+      "learning_rate": 0.00019997953530642823,
+      "loss": 0.755,
+      "step": 70
+    },
+    {
+      "epoch": 0.022834170854271356,
+      "grad_norm": 0.2278454601764679,
+      "learning_rate": 0.00019997884748958902,
+      "loss": 0.6671,
+      "step": 71
+    },
+    {
+      "epoch": 0.023155778894472363,
+      "grad_norm": 0.2758639454841614,
+      "learning_rate": 0.00019997814830549287,
+      "loss": 0.7463,
+      "step": 72
+    },
+    {
+      "epoch": 0.023477386934673366,
+      "grad_norm": 0.27494460344314575,
+      "learning_rate": 0.00019997743775421932,
+      "loss": 0.7974,
+      "step": 73
+    },
+    {
+      "epoch": 0.023798994974874373,
+      "grad_norm": 0.2553146481513977,
+      "learning_rate": 0.00019997671583584908,
+      "loss": 0.6709,
+      "step": 74
+    },
+    {
+      "epoch": 0.024120603015075376,
+      "grad_norm": 0.2966288924217224,
+      "learning_rate": 0.00019997598255046427,
+      "loss": 0.8363,
+      "step": 75
+    },
+    {
+      "epoch": 0.024442211055276383,
+      "grad_norm": 0.27822479605674744,
+      "learning_rate": 0.00019997523789814828,
+      "loss": 0.7242,
+      "step": 76
+    },
+    {
+      "epoch": 0.024763819095477386,
+      "grad_norm": 0.30041515827178955,
+      "learning_rate": 0.00019997448187898572,
+      "loss": 0.6518,
+      "step": 77
+    },
+    {
+      "epoch": 0.025085427135678393,
+      "grad_norm": 0.2713053524494171,
+      "learning_rate": 0.00019997371449306263,
+      "loss": 0.6944,
+      "step": 78
+    },
+    {
+      "epoch": 0.025407035175879396,
+      "grad_norm": 0.2938779890537262,
+      "learning_rate": 0.00019997293574046618,
+      "loss": 0.6322,
+      "step": 79
+    },
+    {
+      "epoch": 0.025728643216080403,
+      "grad_norm": 0.3085947036743164,
+      "learning_rate": 0.00019997214562128497,
+      "loss": 0.6734,
+      "step": 80
+    },
+    {
+      "epoch": 0.026050251256281406,
+      "grad_norm": 0.29282763600349426,
+      "learning_rate": 0.00019997134413560878,
+      "loss": 0.6303,
+      "step": 81
+    },
+    {
+      "epoch": 0.026371859296482413,
+      "grad_norm": 0.3260582685470581,
+      "learning_rate": 0.00019997053128352878,
+      "loss": 0.6201,
+      "step": 82
+    },
+    {
+      "epoch": 0.026693467336683416,
+      "grad_norm": 0.30943652987480164,
+      "learning_rate": 0.00019996970706513735,
+      "loss": 0.6981,
+      "step": 83
+    },
+    {
+      "epoch": 0.027015075376884422,
+      "grad_norm": 0.2850770652294159,
+      "learning_rate": 0.00019996887148052825,
+      "loss": 0.675,
+      "step": 84
+    },
+    {
+      "epoch": 0.027336683417085426,
+      "grad_norm": 0.313479483127594,
+      "learning_rate": 0.00019996802452979645,
+      "loss": 0.808,
+      "step": 85
+    },
+    {
+      "epoch": 0.027658291457286432,
+      "grad_norm": 0.32705870270729065,
+      "learning_rate": 0.00019996716621303827,
+      "loss": 0.709,
+      "step": 86
+    },
+    {
+      "epoch": 0.027979899497487436,
+      "grad_norm": 0.3565938174724579,
+      "learning_rate": 0.00019996629653035126,
+      "loss": 0.8668,
+      "step": 87
+    },
+    {
+      "epoch": 0.028301507537688442,
+      "grad_norm": 0.3635787069797516,
+      "learning_rate": 0.00019996541548183432,
+      "loss": 0.9387,
+      "step": 88
+    },
+    {
+      "epoch": 0.028623115577889446,
+      "grad_norm": 0.3373911380767822,
+      "learning_rate": 0.00019996452306758763,
+      "loss": 0.6269,
+      "step": 89
+    },
+    {
+      "epoch": 0.028944723618090452,
+      "grad_norm": 0.5400757789611816,
+      "learning_rate": 0.0001999636192877126,
+      "loss": 0.6728,
+      "step": 90
+    },
+    {
+      "epoch": 0.029266331658291456,
+      "grad_norm": 0.35397493839263916,
+      "learning_rate": 0.0001999627041423121,
+      "loss": 0.6705,
+      "step": 91
+    },
+    {
+      "epoch": 0.029587939698492462,
+      "grad_norm": 0.4623129963874817,
+      "learning_rate": 0.00019996177763149006,
+      "loss": 0.9886,
+      "step": 92
+    },
+    {
+      "epoch": 0.029909547738693466,
+      "grad_norm": 0.33066219091415405,
+      "learning_rate": 0.0001999608397553519,
+      "loss": 0.528,
+      "step": 93
+    },
+    {
+      "epoch": 0.030231155778894472,
+      "grad_norm": 0.37556031346321106,
+      "learning_rate": 0.00019995989051400424,
+      "loss": 0.7389,
+      "step": 94
+    },
+    {
+      "epoch": 0.03055276381909548,
+      "grad_norm": 0.4391907751560211,
+      "learning_rate": 0.00019995892990755496,
+      "loss": 0.6496,
+      "step": 95
+    },
+    {
+      "epoch": 0.030874371859296482,
+      "grad_norm": 0.3602975010871887,
+      "learning_rate": 0.0001999579579361133,
+      "loss": 0.6117,
+      "step": 96
+    },
+    {
+      "epoch": 0.03119597989949749,
+      "grad_norm": 0.46833890676498413,
+      "learning_rate": 0.00019995697459978984,
+      "loss": 0.8511,
+      "step": 97
+    },
+    {
+      "epoch": 0.03151758793969849,
+      "grad_norm": 0.46151864528656006,
+      "learning_rate": 0.00019995597989869626,
+      "loss": 0.8557,
+      "step": 98
+    },
+    {
+      "epoch": 0.031839195979899496,
+      "grad_norm": 0.626875102519989,
+      "learning_rate": 0.00019995497383294575,
+      "loss": 1.1563,
+      "step": 99
+    },
+    {
+      "epoch": 0.032160804020100506,
+      "grad_norm": 0.7481655478477478,
+      "learning_rate": 0.00019995395640265263,
+      "loss": 1.2427,
+      "step": 100
+    },
+    {
+      "epoch": 0.032160804020100506,
+      "eval_loss": 0.7932191491127014,
+      "eval_runtime": 69.0763,
+      "eval_samples_per_second": 3.619,
+      "eval_steps_per_second": 1.81,
+      "step": 100
+    },
+    {
+      "epoch": 0.03248241206030151,
+      "grad_norm": 0.24802717566490173,
+      "learning_rate": 0.00019995292760793264,
+      "loss": 0.4894,
+      "step": 101
+    },
+    {
+      "epoch": 0.03280402010050251,
+      "grad_norm": 0.22657950222492218,
+      "learning_rate": 0.00019995188744890272,
+      "loss": 0.515,
+      "step": 102
+    },
+    {
+      "epoch": 0.033125628140703516,
+      "grad_norm": 0.26874077320098877,
+      "learning_rate": 0.00019995083592568111,
+      "loss": 0.5728,
+      "step": 103
+    },
+    {
+      "epoch": 0.033447236180904526,
+      "grad_norm": 0.24371561408042908,
+      "learning_rate": 0.00019994977303838742,
+      "loss": 0.6202,
+      "step": 104
+    },
+    {
+      "epoch": 0.03376884422110553,
+      "grad_norm": 0.23280157148838043,
+      "learning_rate": 0.00019994869878714244,
+      "loss": 0.6558,
+      "step": 105
+    },
+    {
+      "epoch": 0.03409045226130653,
+      "grad_norm": 0.24496659636497498,
+      "learning_rate": 0.00019994761317206834,
+      "loss": 0.6961,
+      "step": 106
+    },
+    {
+      "epoch": 0.034412060301507535,
+      "grad_norm": 0.260040283203125,
+      "learning_rate": 0.00019994651619328853,
+      "loss": 0.816,
+      "step": 107
+    },
+    {
+      "epoch": 0.034733668341708546,
+      "grad_norm": 0.29280975461006165,
+      "learning_rate": 0.00019994540785092776,
+      "loss": 0.7362,
+      "step": 108
+    },
+    {
+      "epoch": 0.03505527638190955,
+      "grad_norm": 0.2618405818939209,
+      "learning_rate": 0.00019994428814511204,
+      "loss": 0.788,
+      "step": 109
+    },
+    {
+      "epoch": 0.03537688442211055,
+      "grad_norm": 0.27010878920555115,
+      "learning_rate": 0.00019994315707596867,
+      "loss": 0.8279,
+      "step": 110
+    },
+    {
+      "epoch": 0.035698492462311555,
+      "grad_norm": 0.25788235664367676,
+      "learning_rate": 0.00019994201464362623,
+      "loss": 0.7271,
+      "step": 111
+    },
+    {
+      "epoch": 0.036020100502512566,
+      "grad_norm": 0.26228904724121094,
+      "learning_rate": 0.00019994086084821464,
+      "loss": 0.6475,
+      "step": 112
+    },
+    {
+      "epoch": 0.03634170854271357,
+      "grad_norm": 0.28028932213783264,
+      "learning_rate": 0.00019993969568986506,
+      "loss": 0.7783,
+      "step": 113
+    },
+    {
+      "epoch": 0.03666331658291457,
+      "grad_norm": 0.37087297439575195,
+      "learning_rate": 0.00019993851916870998,
+      "loss": 0.8131,
+      "step": 114
+    },
+    {
+      "epoch": 0.036984924623115575,
+      "grad_norm": 0.31174468994140625,
+      "learning_rate": 0.00019993733128488317,
+      "loss": 0.8285,
+      "step": 115
+    },
+    {
+      "epoch": 0.037306532663316586,
+      "grad_norm": 0.2531948983669281,
+      "learning_rate": 0.00019993613203851966,
+      "loss": 0.5535,
+      "step": 116
+    },
+    {
+      "epoch": 0.03762814070351759,
+      "grad_norm": 0.2401345819234848,
+      "learning_rate": 0.00019993492142975585,
+      "loss": 0.6799,
+      "step": 117
+    },
+    {
+      "epoch": 0.03794974874371859,
+      "grad_norm": 0.2507750988006592,
+      "learning_rate": 0.0001999336994587293,
+      "loss": 0.6501,
+      "step": 118
+    },
+    {
+      "epoch": 0.038271356783919595,
+      "grad_norm": 0.28330686688423157,
+      "learning_rate": 0.00019993246612557905,
+      "loss": 0.6049,
+      "step": 119
+    },
+    {
+      "epoch": 0.038592964824120606,
+      "grad_norm": 0.2696840167045593,
+      "learning_rate": 0.00019993122143044528,
+      "loss": 0.7149,
+      "step": 120
+    },
+    {
+      "epoch": 0.03891457286432161,
+      "grad_norm": 0.2695451080799103,
+      "learning_rate": 0.00019992996537346945,
+      "loss": 0.7895,
+      "step": 121
+    },
+    {
+      "epoch": 0.03923618090452261,
+      "grad_norm": 0.27120038866996765,
+      "learning_rate": 0.00019992869795479442,
+      "loss": 0.6716,
+      "step": 122
+    },
+    {
+      "epoch": 0.039557788944723615,
+      "grad_norm": 0.2795969843864441,
+      "learning_rate": 0.00019992741917456433,
+      "loss": 0.6951,
+      "step": 123
+    },
+    {
+      "epoch": 0.039879396984924625,
+      "grad_norm": 0.3036459982395172,
+      "learning_rate": 0.00019992612903292452,
+      "loss": 0.7524,
+      "step": 124
+    },
+    {
+      "epoch": 0.04020100502512563,
+      "grad_norm": 0.2911680340766907,
+      "learning_rate": 0.00019992482753002167,
+      "loss": 0.6264,
+      "step": 125
+    },
+    {
+      "epoch": 0.04052261306532663,
+      "grad_norm": 0.3199528455734253,
+      "learning_rate": 0.00019992351466600381,
+      "loss": 0.7127,
+      "step": 126
+    },
+    {
+      "epoch": 0.040844221105527635,
+      "grad_norm": 0.287069171667099,
+      "learning_rate": 0.00019992219044102013,
+      "loss": 0.6761,
+      "step": 127
+    },
+    {
+      "epoch": 0.041165829145728645,
+      "grad_norm": 0.3182419240474701,
+      "learning_rate": 0.00019992085485522125,
+      "loss": 0.7587,
+      "step": 128
+    },
+    {
+      "epoch": 0.04148743718592965,
+      "grad_norm": 0.27255985140800476,
+      "learning_rate": 0.00019991950790875903,
+      "loss": 0.6954,
+      "step": 129
+    },
+    {
+      "epoch": 0.04180904522613065,
+      "grad_norm": 0.29751449823379517,
+      "learning_rate": 0.00019991814960178657,
+      "loss": 0.6468,
+      "step": 130
+    },
+    {
+      "epoch": 0.042130653266331655,
+      "grad_norm": 0.28682902455329895,
+      "learning_rate": 0.0001999167799344583,
+      "loss": 0.6072,
+      "step": 131
+    },
+    {
+      "epoch": 0.042452261306532665,
+      "grad_norm": 0.31165406107902527,
+      "learning_rate": 0.00019991539890692998,
+      "loss": 0.6788,
+      "step": 132
+    },
+    {
+      "epoch": 0.04277386934673367,
+      "grad_norm": 0.2839391231536865,
+      "learning_rate": 0.00019991400651935863,
+      "loss": 0.6709,
+      "step": 133
+    },
+    {
+      "epoch": 0.04309547738693467,
+      "grad_norm": 0.321027547121048,
+      "learning_rate": 0.00019991260277190255,
+      "loss": 0.8005,
+      "step": 134
+    },
+    {
+      "epoch": 0.043417085427135675,
+      "grad_norm": 0.33298325538635254,
+      "learning_rate": 0.00019991118766472132,
+      "loss": 0.6781,
+      "step": 135
+    },
+    {
+      "epoch": 0.043738693467336685,
+      "grad_norm": 0.3268507122993469,
+      "learning_rate": 0.00019990976119797584,
+      "loss": 0.6648,
+      "step": 136
+    },
+    {
+      "epoch": 0.04406030150753769,
+      "grad_norm": 0.3155120313167572,
+      "learning_rate": 0.0001999083233718283,
+      "loss": 0.6108,
+      "step": 137
+    },
+    {
+      "epoch": 0.04438190954773869,
+      "grad_norm": 0.3536016047000885,
+      "learning_rate": 0.00019990687418644222,
+      "loss": 0.6936,
+      "step": 138
+    },
+    {
+      "epoch": 0.044703517587939695,
+      "grad_norm": 0.33633413910865784,
+      "learning_rate": 0.00019990541364198235,
+      "loss": 0.6927,
+      "step": 139
+    },
+    {
+      "epoch": 0.045025125628140705,
+      "grad_norm": 0.3715938329696655,
+      "learning_rate": 0.0001999039417386147,
+      "loss": 0.7768,
+      "step": 140
+    },
+    {
+      "epoch": 0.04534673366834171,
+      "grad_norm": 0.36584120988845825,
+      "learning_rate": 0.0001999024584765066,
+      "loss": 0.7078,
+      "step": 141
+    },
+    {
+      "epoch": 0.04566834170854271,
+      "grad_norm": 0.30421900749206543,
+      "learning_rate": 0.0001999009638558268,
+      "loss": 0.5985,
+      "step": 142
+    },
+    {
+      "epoch": 0.04598994974874372,
+      "grad_norm": 0.37903034687042236,
+      "learning_rate": 0.00019989945787674516,
+      "loss": 0.6492,
+      "step": 143
+    },
+    {
+      "epoch": 0.046311557788944725,
+      "grad_norm": 0.38960736989974976,
+      "learning_rate": 0.00019989794053943292,
+      "loss": 0.6972,
+      "step": 144
+    },
+    {
+      "epoch": 0.04663316582914573,
+      "grad_norm": 0.3871925175189972,
+      "learning_rate": 0.00019989641184406257,
+      "loss": 0.6399,
+      "step": 145
+    },
+    {
+      "epoch": 0.04695477386934673,
+      "grad_norm": 0.38902440667152405,
+      "learning_rate": 0.00019989487179080797,
+      "loss": 0.7674,
+      "step": 146
+    },
+    {
+      "epoch": 0.04727638190954774,
+      "grad_norm": 0.4706658124923706,
+      "learning_rate": 0.0001998933203798442,
+      "loss": 0.8804,
+      "step": 147
+    },
+    {
+      "epoch": 0.047597989949748745,
+      "grad_norm": 0.807559609413147,
+      "learning_rate": 0.0001998917576113476,
+      "loss": 0.9446,
+      "step": 148
+    },
+    {
+      "epoch": 0.04791959798994975,
+      "grad_norm": 0.8040673732757568,
+      "learning_rate": 0.0001998901834854959,
+      "loss": 1.2804,
+      "step": 149
+    },
+    {
+      "epoch": 0.04824120603015075,
+      "grad_norm": 0.7172465324401855,
+      "learning_rate": 0.0001998885980024681,
+      "loss": 1.1203,
+      "step": 150
+    },
+    {
+      "epoch": 0.04824120603015075,
+      "eval_loss": 0.7792608737945557,
+      "eval_runtime": 68.7355,
+      "eval_samples_per_second": 3.637,
+      "eval_steps_per_second": 1.819,
+      "step": 150
+    },
+    {
+      "epoch": 0.04856281407035176,
+      "grad_norm": 0.23332886397838593,
+      "learning_rate": 0.0001998870011624444,
+      "loss": 0.548,
+      "step": 151
+    },
+    {
+      "epoch": 0.048884422110552765,
+      "grad_norm": 0.28656452894210815,
+      "learning_rate": 0.00019988539296560639,
+      "loss": 0.6616,
+      "step": 152
+    },
+    {
+      "epoch": 0.04920603015075377,
+      "grad_norm": 0.26958590745925903,
+      "learning_rate": 0.00019988377341213692,
+      "loss": 0.5592,
+      "step": 153
+    },
+    {
+      "epoch": 0.04952763819095477,
+      "grad_norm": 0.3448958396911621,
+      "learning_rate": 0.00019988214250222012,
+      "loss": 0.8696,
+      "step": 154
+    },
+    {
+      "epoch": 0.04984924623115578,
+      "grad_norm": 0.3263605535030365,
+      "learning_rate": 0.0001998805002360414,
+      "loss": 0.8408,
+      "step": 155
+    },
+    {
+      "epoch": 0.050170854271356785,
+      "grad_norm": 0.24680665135383606,
+      "learning_rate": 0.00019987884661378752,
+      "loss": 0.7482,
+      "step": 156
+    },
+    {
+      "epoch": 0.05049246231155779,
+      "grad_norm": 0.2885967493057251,
+      "learning_rate": 0.00019987718163564644,
+      "loss": 0.7518,
+      "step": 157
+    },
+    {
+      "epoch": 0.05081407035175879,
+      "grad_norm": 0.26518917083740234,
+      "learning_rate": 0.00019987550530180752,
+      "loss": 0.6402,
+      "step": 158
+    },
+    {
+      "epoch": 0.0511356783919598,
+      "grad_norm": 0.28461605310440063,
+      "learning_rate": 0.0001998738176124613,
+      "loss": 0.8303,
+      "step": 159
+    },
+    {
+      "epoch": 0.051457286432160805,
+      "grad_norm": 0.26935309171676636,
+      "learning_rate": 0.00019987211856779973,
+      "loss": 0.7031,
+      "step": 160
+    },
+    {
+      "epoch": 0.05177889447236181,
+      "grad_norm": 2.391087055206299,
+      "learning_rate": 0.00019987040816801593,
+      "loss": 0.9029,
+      "step": 161
+    },
+    {
+      "epoch": 0.05210050251256281,
+      "grad_norm": 0.28461599349975586,
+      "learning_rate": 0.00019986868641330436,
+      "loss": 0.6788,
+      "step": 162
+    },
+    {
+      "epoch": 0.05242211055276382,
+      "grad_norm": 0.3029423654079437,
+      "learning_rate": 0.0001998669533038608,
+      "loss": 0.6295,
+      "step": 163
+    },
+    {
+      "epoch": 0.052743718592964825,
+      "grad_norm": 0.44214892387390137,
+      "learning_rate": 0.00019986520883988232,
+      "loss": 0.6156,
+      "step": 164
+    },
+    {
+      "epoch": 0.05306532663316583,
+      "grad_norm": 0.4488637149333954,
+      "learning_rate": 0.00019986345302156725,
+      "loss": 0.8151,
+      "step": 165
+    },
+    {
+      "epoch": 0.05338693467336683,
+      "grad_norm": 0.4122033417224884,
+      "learning_rate": 0.00019986168584911515,
+      "loss": 0.7768,
+      "step": 166
+    },
+    {
+      "epoch": 0.05370854271356784,
+      "grad_norm": 0.2943034768104553,
+      "learning_rate": 0.00019985990732272705,
+      "loss": 0.7546,
+      "step": 167
+    },
+    {
+      "epoch": 0.054030150753768845,
+      "grad_norm": 0.28280842304229736,
+      "learning_rate": 0.0001998581174426051,
+      "loss": 0.7192,
+      "step": 168
+    },
+    {
+      "epoch": 0.05435175879396985,
+      "grad_norm": 0.287214070558548,
+      "learning_rate": 0.00019985631620895279,
+      "loss": 0.7051,
+      "step": 169
+    },
+    {
+      "epoch": 0.05467336683417085,
+      "grad_norm": 0.2786403298377991,
+      "learning_rate": 0.000199854503621975,
+      "loss": 0.7829,
+      "step": 170
+    },
+    {
+      "epoch": 0.05499497487437186,
+      "grad_norm": 0.281941294670105,
+      "learning_rate": 0.0001998526796818777,
+      "loss": 0.6975,
+      "step": 171
+    },
+    {
+      "epoch": 0.055316582914572865,
+      "grad_norm": 0.2721412777900696,
+      "learning_rate": 0.00019985084438886835,
+      "loss": 0.7675,
+      "step": 172
+    },
+    {
+      "epoch": 0.05563819095477387,
+      "grad_norm": 0.30607619881629944,
+      "learning_rate": 0.00019984899774315557,
+      "loss": 0.7518,
+      "step": 173
+    },
+    {
+      "epoch": 0.05595979899497487,
+      "grad_norm": 0.2951827049255371,
+      "learning_rate": 0.00019984713974494935,
+      "loss": 0.5959,
+      "step": 174
+    },
+    {
+      "epoch": 0.05628140703517588,
+      "grad_norm": 0.30903956294059753,
+      "learning_rate": 0.00019984527039446093,
+      "loss": 0.6139,
+      "step": 175
+    },
+    {
+      "epoch": 0.056603015075376885,
+      "grad_norm": 0.32232603430747986,
+      "learning_rate": 0.00019984338969190283,
+      "loss": 0.836,
+      "step": 176
+    },
+    {
+      "epoch": 0.05692462311557789,
+      "grad_norm": 0.3350957930088043,
+      "learning_rate": 0.00019984149763748886,
+      "loss": 0.6985,
+      "step": 177
+    },
+    {
+      "epoch": 0.05724623115577889,
+      "grad_norm": 0.34590116143226624,
+      "learning_rate": 0.0001998395942314342,
+      "loss": 0.7911,
+      "step": 178
+    },
+    {
+      "epoch": 0.0575678391959799,
+      "grad_norm": 0.3729935586452484,
+      "learning_rate": 0.00019983767947395525,
+      "loss": 0.7067,
+      "step": 179
+    },
+    {
+      "epoch": 0.057889447236180905,
+      "grad_norm": 0.3160020709037781,
+      "learning_rate": 0.00019983575336526968,
+      "loss": 0.7479,
+      "step": 180
+    },
+    {
+      "epoch": 0.05821105527638191,
+      "grad_norm": 0.3247075080871582,
+      "learning_rate": 0.00019983381590559649,
+      "loss": 0.7209,
+      "step": 181
+    },
+    {
+      "epoch": 0.05853266331658291,
+      "grad_norm": 0.3256763815879822,
+      "learning_rate": 0.00019983186709515596,
+      "loss": 0.7591,
+      "step": 182
+    },
+    {
+      "epoch": 0.05885427135678392,
+      "grad_norm": 0.3437361717224121,
+      "learning_rate": 0.0001998299069341697,
+      "loss": 0.7768,
+      "step": 183
+    },
+    {
+      "epoch": 0.059175879396984925,
+      "grad_norm": 0.3492826819419861,
+      "learning_rate": 0.00019982793542286053,
+      "loss": 0.8856,
+      "step": 184
+    },
+    {
+      "epoch": 0.05949748743718593,
+      "grad_norm": 0.35625192523002625,
+      "learning_rate": 0.00019982595256145261,
+      "loss": 0.6869,
+      "step": 185
+    },
+    {
+      "epoch": 0.05981909547738693,
+      "grad_norm": 0.3370972275733948,
+      "learning_rate": 0.0001998239583501714,
+      "loss": 0.6598,
+      "step": 186
+    },
+    {
+      "epoch": 0.06014070351758794,
+      "grad_norm": 0.3618307113647461,
+      "learning_rate": 0.00019982195278924362,
+      "loss": 0.7644,
+      "step": 187
+    },
+    {
+      "epoch": 0.060462311557788945,
+      "grad_norm": 0.3327068090438843,
+      "learning_rate": 0.00019981993587889733,
+      "loss": 0.5518,
+      "step": 188
+    },
+    {
+      "epoch": 0.06078391959798995,
+      "grad_norm": 0.3569602370262146,
+      "learning_rate": 0.0001998179076193618,
+      "loss": 0.662,
+      "step": 189
+    },
+    {
+      "epoch": 0.06110552763819096,
+      "grad_norm": 0.3791339099407196,
+      "learning_rate": 0.00019981586801086766,
+      "loss": 0.7292,
+      "step": 190
+    },
+    {
+      "epoch": 0.06142713567839196,
+      "grad_norm": 0.3519534170627594,
+      "learning_rate": 0.00019981381705364682,
+      "loss": 0.628,
+      "step": 191
+    },
+    {
+      "epoch": 0.061748743718592965,
+      "grad_norm": 0.4257189929485321,
+      "learning_rate": 0.00019981175474793243,
+      "loss": 0.8068,
+      "step": 192
+    },
+    {
+      "epoch": 0.06207035175879397,
+      "grad_norm": 0.37364888191223145,
+      "learning_rate": 0.00019980968109395902,
+      "loss": 0.6726,
+      "step": 193
+    },
+    {
+      "epoch": 0.06239195979899498,
+      "grad_norm": 0.3985297977924347,
+      "learning_rate": 0.00019980759609196228,
+      "loss": 0.6154,
+      "step": 194
+    },
+    {
+      "epoch": 0.06271356783919597,
+      "grad_norm": 0.3966808021068573,
+      "learning_rate": 0.00019980549974217934,
+      "loss": 0.616,
+      "step": 195
+    },
+    {
+      "epoch": 0.06303517587939698,
+      "grad_norm": 0.44981667399406433,
+      "learning_rate": 0.00019980339204484851,
+      "loss": 0.8705,
+      "step": 196
+    },
+    {
+      "epoch": 0.063356783919598,
+      "grad_norm": 0.43207600712776184,
+      "learning_rate": 0.00019980127300020945,
+      "loss": 0.7247,
+      "step": 197
+    },
+    {
+      "epoch": 0.06367839195979899,
+      "grad_norm": 0.6379528641700745,
+      "learning_rate": 0.00019979914260850307,
+      "loss": 0.965,
+      "step": 198
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.614347517490387,
+      "learning_rate": 0.0001997970008699716,
+      "loss": 1.1311,
+      "step": 199
+    },
+    {
+      "epoch": 0.06432160804020101,
+      "grad_norm": 0.6902658343315125,
+      "learning_rate": 0.00019979484778485853,
+      "loss": 1.2486,
+      "step": 200
+    },
+    {
+      "epoch": 0.06432160804020101,
+      "eval_loss": 0.780014157295227,
+      "eval_runtime": 69.2804,
+      "eval_samples_per_second": 3.609,
+      "eval_steps_per_second": 1.804,
+      "step": 200
+    },
+    {
+      "epoch": 0.06464321608040201,
+      "grad_norm": 0.21630311012268066,
+      "learning_rate": 0.00019979268335340868,
+      "loss": 0.4658,
+      "step": 201
+    },
+    {
+      "epoch": 0.06496482412060302,
+      "grad_norm": 0.2785795032978058,
+      "learning_rate": 0.00019979050757586812,
+      "loss": 0.5905,
+      "step": 202
+    },
+    {
+      "epoch": 0.06528643216080401,
+      "grad_norm": 0.32486870884895325,
+      "learning_rate": 0.00019978832045248427,
+      "loss": 0.6047,
+      "step": 203
+    },
+    {
+      "epoch": 0.06560804020100502,
+      "grad_norm": 0.27402544021606445,
+      "learning_rate": 0.00019978612198350574,
+      "loss": 0.7414,
+      "step": 204
+    },
+    {
+      "epoch": 0.06592964824120603,
+      "grad_norm": 0.251230925321579,
+      "learning_rate": 0.0001997839121691825,
+      "loss": 0.7023,
+      "step": 205
+    },
+    {
+      "epoch": 0.06625125628140703,
+      "grad_norm": 0.2718803584575653,
+      "learning_rate": 0.00019978169100976585,
+      "loss": 0.9002,
+      "step": 206
+    },
+    {
+      "epoch": 0.06657286432160804,
+      "grad_norm": 0.2742752730846405,
+      "learning_rate": 0.0001997794585055083,
+      "loss": 0.6531,
+      "step": 207
+    },
+    {
+      "epoch": 0.06689447236180905,
+      "grad_norm": 0.25602105259895325,
+      "learning_rate": 0.00019977721465666364,
+      "loss": 0.713,
+      "step": 208
+    },
+    {
+      "epoch": 0.06721608040201005,
+      "grad_norm": 0.26393651962280273,
+      "learning_rate": 0.00019977495946348702,
+      "loss": 0.6108,
+      "step": 209
+    },
+    {
+      "epoch": 0.06753768844221106,
+      "grad_norm": 0.2806963622570038,
+      "learning_rate": 0.00019977269292623486,
+      "loss": 0.7211,
+      "step": 210
+    },
+    {
+      "epoch": 0.06785929648241205,
+      "grad_norm": 0.26890912652015686,
+      "learning_rate": 0.00019977041504516484,
+      "loss": 0.5881,
+      "step": 211
+    },
+    {
+      "epoch": 0.06818090452261306,
+      "grad_norm": 0.30698782205581665,
+      "learning_rate": 0.00019976812582053594,
+      "loss": 0.8162,
+      "step": 212
+    },
+    {
+      "epoch": 0.06850251256281407,
+      "grad_norm": 0.2880939841270447,
+      "learning_rate": 0.0001997658252526085,
+      "loss": 0.8265,
+      "step": 213
+    },
+    {
+      "epoch": 0.06882412060301507,
+      "grad_norm": 0.2918909788131714,
+      "learning_rate": 0.00019976351334164398,
+      "loss": 0.8076,
+      "step": 214
+    },
+    {
+      "epoch": 0.06914572864321608,
+      "grad_norm": 0.2898813784122467,
+      "learning_rate": 0.00019976119008790534,
+      "loss": 0.7452,
+      "step": 215
+    },
+    {
+      "epoch": 0.06946733668341709,
+      "grad_norm": 0.29868465662002563,
+      "learning_rate": 0.00019975885549165662,
+      "loss": 0.8174,
+      "step": 216
+    },
+    {
+      "epoch": 0.06978894472361809,
+      "grad_norm": 0.29010850191116333,
+      "learning_rate": 0.00019975650955316335,
+      "loss": 0.6722,
+      "step": 217
+    },
+    {
+      "epoch": 0.0701105527638191,
+      "grad_norm": 0.27701061964035034,
+      "learning_rate": 0.00019975415227269224,
+      "loss": 0.6746,
+      "step": 218
+    },
+    {
+      "epoch": 0.0704321608040201,
+      "grad_norm": 0.29137811064720154,
+      "learning_rate": 0.00019975178365051123,
+      "loss": 0.7736,
+      "step": 219
+    },
+    {
+      "epoch": 0.0707537688442211,
+      "grad_norm": 0.3063967525959015,
+      "learning_rate": 0.0001997494036868897,
+      "loss": 0.7349,
+      "step": 220
+    },
+    {
+      "epoch": 0.07107537688442211,
+      "grad_norm": 0.290232390165329,
+      "learning_rate": 0.00019974701238209823,
+      "loss": 0.6274,
+      "step": 221
+    },
+    {
+      "epoch": 0.07139698492462311,
+      "grad_norm": 0.30241337418556213,
+      "learning_rate": 0.0001997446097364087,
+      "loss": 0.7362,
+      "step": 222
+    },
+    {
+      "epoch": 0.07171859296482412,
+      "grad_norm": 0.3109579086303711,
+      "learning_rate": 0.00019974219575009426,
+      "loss": 0.7945,
+      "step": 223
+    },
+    {
+      "epoch": 0.07204020100502513,
+      "grad_norm": 0.31153643131256104,
+      "learning_rate": 0.00019973977042342942,
+      "loss": 0.8714,
+      "step": 224
+    },
+    {
+      "epoch": 0.07236180904522613,
+      "grad_norm": 0.31827324628829956,
+      "learning_rate": 0.0001997373337566899,
+      "loss": 0.8222,
+      "step": 225
+    },
+    {
+      "epoch": 0.07268341708542714,
+      "grad_norm": 0.30141428112983704,
+      "learning_rate": 0.00019973488575015272,
+      "loss": 0.728,
+      "step": 226
+    },
+    {
+      "epoch": 0.07300502512562813,
+      "grad_norm": 0.3078593313694,
+      "learning_rate": 0.00019973242640409624,
+      "loss": 0.7031,
+      "step": 227
+    },
+    {
+      "epoch": 0.07332663316582914,
+      "grad_norm": 0.32512515783309937,
+      "learning_rate": 0.00019972995571880008,
+      "loss": 0.6755,
+      "step": 228
+    },
+    {
+      "epoch": 0.07364824120603015,
+      "grad_norm": 0.30891963839530945,
+      "learning_rate": 0.00019972747369454512,
+      "loss": 0.6169,
+      "step": 229
+    },
+    {
+      "epoch": 0.07396984924623115,
+      "grad_norm": 0.33461982011795044,
+      "learning_rate": 0.0001997249803316136,
+      "loss": 0.7505,
+      "step": 230
+    },
+    {
+      "epoch": 0.07429145728643216,
+      "grad_norm": 0.3086506128311157,
+      "learning_rate": 0.00019972247563028898,
+      "loss": 0.5963,
+      "step": 231
+    },
+    {
+      "epoch": 0.07461306532663317,
+      "grad_norm": 0.310531347990036,
+      "learning_rate": 0.000199719959590856,
+      "loss": 0.6049,
+      "step": 232
+    },
+    {
+      "epoch": 0.07493467336683417,
+      "grad_norm": 0.32778555154800415,
+      "learning_rate": 0.00019971743221360085,
+      "loss": 0.6116,
+      "step": 233
+    },
+    {
+      "epoch": 0.07525628140703518,
+      "grad_norm": 0.3377425968647003,
+      "learning_rate": 0.00019971489349881075,
+      "loss": 0.8245,
+      "step": 234
+    },
+    {
+      "epoch": 0.07557788944723617,
+      "grad_norm": 0.31767261028289795,
+      "learning_rate": 0.00019971234344677439,
+      "loss": 0.6879,
+      "step": 235
+    },
+    {
+      "epoch": 0.07589949748743718,
+      "grad_norm": 0.3302140235900879,
+      "learning_rate": 0.00019970978205778172,
+      "loss": 0.6762,
+      "step": 236
+    },
+    {
+      "epoch": 0.0762211055276382,
+      "grad_norm": 0.3524129092693329,
+      "learning_rate": 0.00019970720933212396,
+      "loss": 0.7051,
+      "step": 237
+    },
+    {
+      "epoch": 0.07654271356783919,
+      "grad_norm": 0.30854153633117676,
+      "learning_rate": 0.00019970462527009358,
+      "loss": 0.5861,
+      "step": 238
+    },
+    {
+      "epoch": 0.0768643216080402,
+      "grad_norm": 0.32968831062316895,
+      "learning_rate": 0.0001997020298719844,
+      "loss": 0.6365,
+      "step": 239
+    },
+    {
+      "epoch": 0.07718592964824121,
+      "grad_norm": 0.3842124342918396,
+      "learning_rate": 0.00019969942313809153,
+      "loss": 0.5958,
+      "step": 240
+    },
+    {
+      "epoch": 0.07750753768844221,
+      "grad_norm": 0.35582709312438965,
+      "learning_rate": 0.00019969680506871137,
+      "loss": 0.6175,
+      "step": 241
+    },
+    {
+      "epoch": 0.07782914572864322,
+      "grad_norm": 0.34835779666900635,
+      "learning_rate": 0.00019969417566414148,
+      "loss": 0.5049,
+      "step": 242
+    },
+    {
+      "epoch": 0.07815075376884423,
+      "grad_norm": 0.4019937515258789,
+      "learning_rate": 0.00019969153492468094,
+      "loss": 0.7996,
+      "step": 243
+    },
+    {
+      "epoch": 0.07847236180904522,
+      "grad_norm": 0.39154139161109924,
+      "learning_rate": 0.0001996888828506299,
+      "loss": 0.7461,
+      "step": 244
+    },
+    {
+      "epoch": 0.07879396984924623,
+      "grad_norm": 0.4105947017669678,
+      "learning_rate": 0.00019968621944228994,
+      "loss": 0.8028,
+      "step": 245
+    },
+    {
+      "epoch": 0.07911557788944723,
+      "grad_norm": 0.3965797424316406,
+      "learning_rate": 0.00019968354469996385,
+      "loss": 0.7984,
+      "step": 246
+    },
+    {
+      "epoch": 0.07943718592964824,
+      "grad_norm": 0.4477737843990326,
+      "learning_rate": 0.00019968085862395575,
+      "loss": 0.7477,
+      "step": 247
+    },
+    {
+      "epoch": 0.07975879396984925,
+      "grad_norm": 0.48387402296066284,
+      "learning_rate": 0.00019967816121457108,
+      "loss": 1.0214,
+      "step": 248
+    },
+    {
+      "epoch": 0.08008040201005025,
+      "grad_norm": 0.5509622097015381,
+      "learning_rate": 0.00019967545247211648,
+      "loss": 1.1386,
+      "step": 249
+    },
+    {
+      "epoch": 0.08040201005025126,
+      "grad_norm": 0.5555517673492432,
+      "learning_rate": 0.00019967273239689993,
+      "loss": 1.0426,
+      "step": 250
+    },
+    {
+      "epoch": 0.08040201005025126,
+      "eval_loss": 0.7569652199745178,
+      "eval_runtime": 69.2811,
+      "eval_samples_per_second": 3.608,
+      "eval_steps_per_second": 1.804,
+      "step": 250
+    },
+    {
+      "epoch": 0.08072361809045227,
+      "grad_norm": 0.19764406979084015,
+      "learning_rate": 0.0001996700009892307,
+      "loss": 0.424,
+      "step": 251
+    },
+    {
+      "epoch": 0.08104522613065326,
+      "grad_norm": 0.27977269887924194,
+      "learning_rate": 0.00019966725824941932,
+      "loss": 0.6377,
+      "step": 252
+    },
+    {
+      "epoch": 0.08136683417085427,
+      "grad_norm": 0.2845287322998047,
+      "learning_rate": 0.00019966450417777768,
+      "loss": 0.5798,
+      "step": 253
+    },
+    {
+      "epoch": 0.08168844221105527,
+      "grad_norm": 0.2786215841770172,
+      "learning_rate": 0.00019966173877461886,
+      "loss": 0.6743,
+      "step": 254
+    },
+    {
+      "epoch": 0.08201005025125628,
+      "grad_norm": 0.2676329016685486,
+      "learning_rate": 0.00019965896204025732,
+      "loss": 0.7719,
+      "step": 255
+    },
+    {
+      "epoch": 0.08233165829145729,
+      "grad_norm": 0.24040904641151428,
+      "learning_rate": 0.00019965617397500873,
+      "loss": 0.5955,
+      "step": 256
+    },
+    {
+      "epoch": 0.08265326633165829,
+      "grad_norm": 0.2543548345565796,
+      "learning_rate": 0.00019965337457919012,
+      "loss": 0.6594,
+      "step": 257
+    },
+    {
+      "epoch": 0.0829748743718593,
+      "grad_norm": 0.3393791913986206,
+      "learning_rate": 0.0001996505638531197,
+      "loss": 0.7516,
+      "step": 258
+    },
+    {
+      "epoch": 0.08329648241206031,
+      "grad_norm": 0.2649451494216919,
+      "learning_rate": 0.00019964774179711712,
+      "loss": 0.6708,
+      "step": 259
+    },
+    {
+      "epoch": 0.0836180904522613,
+      "grad_norm": 0.28479498624801636,
+      "learning_rate": 0.00019964490841150315,
+      "loss": 0.7705,
+      "step": 260
+    },
+    {
+      "epoch": 0.08393969849246231,
+      "grad_norm": 0.26171791553497314,
+      "learning_rate": 0.00019964206369660006,
+      "loss": 0.6867,
+      "step": 261
+    },
+    {
+      "epoch": 0.08426130653266331,
+      "grad_norm": 0.28884440660476685,
+      "learning_rate": 0.00019963920765273118,
+      "loss": 0.6801,
+      "step": 262
+    },
+    {
+      "epoch": 0.08458291457286432,
+      "grad_norm": 0.305441290140152,
+      "learning_rate": 0.00019963634028022132,
+      "loss": 0.7607,
+      "step": 263
+    },
+    {
+      "epoch": 0.08490452261306533,
+      "grad_norm": 0.28244510293006897,
+      "learning_rate": 0.00019963346157939637,
+      "loss": 0.7498,
+      "step": 264
+    },
+    {
+      "epoch": 0.08522613065326633,
+      "grad_norm": 0.30209043622016907,
+      "learning_rate": 0.00019963057155058373,
+      "loss": 0.682,
+      "step": 265
+    },
+    {
+      "epoch": 0.08554773869346734,
+      "grad_norm": 0.29869410395622253,
+      "learning_rate": 0.00019962767019411195,
+      "loss": 0.7103,
+      "step": 266
+    },
+    {
+      "epoch": 0.08586934673366835,
+      "grad_norm": 0.27020129561424255,
+      "learning_rate": 0.0001996247575103109,
+      "loss": 0.5624,
+      "step": 267
+    },
+    {
+      "epoch": 0.08619095477386934,
+      "grad_norm": 0.2982742488384247,
+      "learning_rate": 0.0001996218334995118,
+      "loss": 0.6912,
+      "step": 268
+    },
+    {
+      "epoch": 0.08651256281407035,
+      "grad_norm": 0.2759303152561188,
+      "learning_rate": 0.000199618898162047,
+      "loss": 0.707,
+      "step": 269
+    },
+    {
+      "epoch": 0.08683417085427135,
+      "grad_norm": 0.2757350206375122,
+      "learning_rate": 0.0001996159514982503,
+      "loss": 0.6531,
+      "step": 270
+    },
+    {
+      "epoch": 0.08715577889447236,
+      "grad_norm": 0.25845128297805786,
+      "learning_rate": 0.0001996129935084567,
+      "loss": 0.4924,
+      "step": 271
+    },
+    {
+      "epoch": 0.08747738693467337,
+      "grad_norm": 0.297620564699173,
+      "learning_rate": 0.00019961002419300258,
+      "loss": 0.622,
+      "step": 272
+    },
+    {
+      "epoch": 0.08779899497487437,
+      "grad_norm": 0.3007553815841675,
+      "learning_rate": 0.00019960704355222544,
+      "loss": 0.5083,
+      "step": 273
+    },
+    {
+      "epoch": 0.08812060301507538,
+      "grad_norm": 0.31905415654182434,
+      "learning_rate": 0.00019960405158646425,
+      "loss": 0.661,
+      "step": 274
+    },
+    {
+      "epoch": 0.08844221105527639,
+      "grad_norm": 0.3016757071018219,
+      "learning_rate": 0.00019960104829605912,
+      "loss": 0.7183,
+      "step": 275
+    },
+    {
+      "epoch": 0.08876381909547738,
+      "grad_norm": 0.28524714708328247,
+      "learning_rate": 0.00019959803368135157,
+      "loss": 0.5427,
+      "step": 276
+    },
+    {
+      "epoch": 0.0890854271356784,
+      "grad_norm": 0.3103770911693573,
+      "learning_rate": 0.00019959500774268434,
+      "loss": 0.5872,
+      "step": 277
+    },
+    {
+      "epoch": 0.08940703517587939,
+      "grad_norm": 0.3064669072628021,
+      "learning_rate": 0.00019959197048040147,
+      "loss": 0.6241,
+      "step": 278
+    },
+    {
+      "epoch": 0.0897286432160804,
+      "grad_norm": 0.3114566206932068,
+      "learning_rate": 0.00019958892189484822,
+      "loss": 0.5936,
+      "step": 279
+    },
+    {
+      "epoch": 0.09005025125628141,
+      "grad_norm": 0.37707945704460144,
+      "learning_rate": 0.0001995858619863713,
+      "loss": 0.6873,
+      "step": 280
+    },
+    {
+      "epoch": 0.0903718592964824,
+      "grad_norm": 0.3506699204444885,
+      "learning_rate": 0.00019958279075531855,
+      "loss": 0.7335,
+      "step": 281
+    },
+    {
+      "epoch": 0.09069346733668342,
+      "grad_norm": 0.33031418919563293,
+      "learning_rate": 0.0001995797082020392,
+      "loss": 0.6712,
+      "step": 282
+    },
+    {
+      "epoch": 0.09101507537688443,
+      "grad_norm": 0.34307384490966797,
+      "learning_rate": 0.00019957661432688372,
+      "loss": 0.732,
+      "step": 283
+    },
+    {
+      "epoch": 0.09133668341708542,
+      "grad_norm": 0.38271409273147583,
+      "learning_rate": 0.00019957350913020382,
+      "loss": 0.6968,
+      "step": 284
+    },
+    {
+      "epoch": 0.09165829145728643,
+      "grad_norm": 0.35980135202407837,
+      "learning_rate": 0.0001995703926123526,
+      "loss": 0.6921,
+      "step": 285
+    },
+    {
+      "epoch": 0.09197989949748744,
+      "grad_norm": 0.3397926092147827,
+      "learning_rate": 0.00019956726477368442,
+      "loss": 0.6275,
+      "step": 286
+    },
+    {
+      "epoch": 0.09230150753768844,
+      "grad_norm": 0.34540098905563354,
+      "learning_rate": 0.0001995641256145548,
+      "loss": 0.6403,
+      "step": 287
+    },
+    {
+      "epoch": 0.09262311557788945,
+      "grad_norm": 0.3400019705295563,
+      "learning_rate": 0.0001995609751353208,
+      "loss": 0.5852,
+      "step": 288
+    },
+    {
+      "epoch": 0.09294472361809045,
+      "grad_norm": 0.3435593247413635,
+      "learning_rate": 0.0001995578133363405,
+      "loss": 0.6042,
+      "step": 289
+    },
+    {
+      "epoch": 0.09326633165829146,
+      "grad_norm": 0.353845477104187,
+      "learning_rate": 0.00019955464021797342,
+      "loss": 0.676,
+      "step": 290
+    },
+    {
+      "epoch": 0.09358793969849247,
+      "grad_norm": 0.37604397535324097,
+      "learning_rate": 0.00019955145578058037,
+      "loss": 0.698,
+      "step": 291
+    },
+    {
+      "epoch": 0.09390954773869346,
+      "grad_norm": 0.46072182059288025,
+      "learning_rate": 0.00019954826002452336,
+      "loss": 0.7182,
+      "step": 292
+    },
+    {
+      "epoch": 0.09423115577889447,
+      "grad_norm": 0.4154977798461914,
+      "learning_rate": 0.00019954505295016576,
+      "loss": 0.7492,
+      "step": 293
+    },
+    {
+      "epoch": 0.09455276381909548,
+      "grad_norm": 0.397504597902298,
+      "learning_rate": 0.0001995418345578722,
+      "loss": 0.7201,
+      "step": 294
+    },
+    {
+      "epoch": 0.09487437185929648,
+      "grad_norm": 0.44909676909446716,
+      "learning_rate": 0.00019953860484800862,
+      "loss": 0.8279,
+      "step": 295
+    },
+    {
+      "epoch": 0.09519597989949749,
+      "grad_norm": 0.4281664788722992,
+      "learning_rate": 0.00019953536382094218,
+      "loss": 0.7666,
+      "step": 296
+    },
+    {
+      "epoch": 0.09551758793969849,
+      "grad_norm": 0.4675356149673462,
+      "learning_rate": 0.0001995321114770414,
+      "loss": 0.786,
+      "step": 297
+    },
+    {
+      "epoch": 0.0958391959798995,
+      "grad_norm": 0.5142356753349304,
+      "learning_rate": 0.00019952884781667608,
+      "loss": 0.7897,
+      "step": 298
+    },
+    {
+      "epoch": 0.09616080402010051,
+      "grad_norm": 0.5967409014701843,
+      "learning_rate": 0.00019952557284021728,
+      "loss": 0.9809,
+      "step": 299
+    },
+    {
+      "epoch": 0.0964824120603015,
+      "grad_norm": 0.8204864263534546,
+      "learning_rate": 0.0001995222865480373,
+      "loss": 1.1053,
+      "step": 300
+    },
+    {
+      "epoch": 0.0964824120603015,
+      "eval_loss": 0.7589501142501831,
+      "eval_runtime": 69.0782,
+      "eval_samples_per_second": 3.619,
+      "eval_steps_per_second": 1.81,
+      "step": 300
+    },
+    {
+      "epoch": 0.09680402010050251,
+      "grad_norm": 0.22103621065616608,
+      "learning_rate": 0.00019951898894050987,
+      "loss": 0.4517,
+      "step": 301
+    },
+    {
+      "epoch": 0.09712562814070352,
+      "grad_norm": 0.2650449573993683,
+      "learning_rate": 0.00019951568001800986,
+      "loss": 0.5882,
+      "step": 302
+    },
+    {
+      "epoch": 0.09744723618090452,
+      "grad_norm": 0.30080127716064453,
+      "learning_rate": 0.00019951235978091347,
+      "loss": 0.6827,
+      "step": 303
+    },
+    {
+      "epoch": 0.09776884422110553,
+      "grad_norm": 0.27281373739242554,
+      "learning_rate": 0.00019950902822959821,
+      "loss": 0.6373,
+      "step": 304
+    },
+    {
+      "epoch": 0.09809045226130653,
+      "grad_norm": 0.2672353684902191,
+      "learning_rate": 0.00019950568536444293,
+      "loss": 0.7638,
+      "step": 305
+    },
+    {
+      "epoch": 0.09841206030150754,
+      "grad_norm": 0.2505207657814026,
+      "learning_rate": 0.00019950233118582762,
+      "loss": 0.6339,
+      "step": 306
+    },
+    {
+      "epoch": 0.09873366834170855,
+      "grad_norm": 0.2656727433204651,
+      "learning_rate": 0.0001994989656941337,
+      "loss": 0.7168,
+      "step": 307
+    },
+    {
+      "epoch": 0.09905527638190954,
+      "grad_norm": 0.25751009583473206,
+      "learning_rate": 0.00019949558888974375,
+      "loss": 0.6443,
+      "step": 308
+    },
+    {
+      "epoch": 0.09937688442211055,
+      "grad_norm": 0.26401689648628235,
+      "learning_rate": 0.00019949220077304173,
+      "loss": 0.591,
+      "step": 309
+    },
+    {
+      "epoch": 0.09969849246231156,
+      "grad_norm": 0.3080660104751587,
+      "learning_rate": 0.0001994888013444129,
+      "loss": 0.7722,
+      "step": 310
+    },
+    {
+      "epoch": 0.10002010050251256,
+      "grad_norm": 0.2839815020561218,
+      "learning_rate": 0.0001994853906042437,
+      "loss": 0.7159,
+      "step": 311
+    },
+    {
+      "epoch": 0.10034170854271357,
+      "grad_norm": 0.3083145320415497,
+      "learning_rate": 0.00019948196855292194,
+      "loss": 0.7903,
+      "step": 312
+    },
+    {
+      "epoch": 0.10066331658291457,
+      "grad_norm": 0.2846497893333435,
+      "learning_rate": 0.00019947853519083672,
+      "loss": 0.6206,
+      "step": 313
+    },
+    {
+      "epoch": 0.10098492462311558,
+      "grad_norm": 0.2835051417350769,
+      "learning_rate": 0.0001994750905183784,
+      "loss": 0.7105,
+      "step": 314
+    },
+    {
+      "epoch": 0.10130653266331659,
+      "grad_norm": 0.29267215728759766,
+      "learning_rate": 0.0001994716345359386,
+      "loss": 0.6162,
+      "step": 315
+    },
+    {
+      "epoch": 0.10162814070351758,
+      "grad_norm": 0.3165644109249115,
+      "learning_rate": 0.00019946816724391027,
+      "loss": 0.7514,
+      "step": 316
+    },
+    {
+      "epoch": 0.1019497487437186,
+      "grad_norm": 0.2821200489997864,
+      "learning_rate": 0.00019946468864268758,
+      "loss": 0.6697,
+      "step": 317
+    },
+    {
+      "epoch": 0.1022713567839196,
+      "grad_norm": 0.2952963709831238,
+      "learning_rate": 0.00019946119873266613,
+      "loss": 0.6412,
+      "step": 318
+    },
+    {
+      "epoch": 0.1025929648241206,
+      "grad_norm": 0.2964838743209839,
+      "learning_rate": 0.00019945769751424268,
+      "loss": 0.8092,
+      "step": 319
+    },
+    {
+      "epoch": 0.10291457286432161,
+      "grad_norm": 0.2943262755870819,
+      "learning_rate": 0.00019945418498781522,
+      "loss": 0.7361,
+      "step": 320
+    },
+    {
+      "epoch": 0.1032361809045226,
+      "grad_norm": 0.2877391576766968,
+      "learning_rate": 0.00019945066115378325,
+      "loss": 0.6889,
+      "step": 321
+    },
+    {
+      "epoch": 0.10355778894472362,
+      "grad_norm": 0.31130778789520264,
+      "learning_rate": 0.00019944712601254732,
+      "loss": 0.6854,
+      "step": 322
+    },
+    {
+      "epoch": 0.10387939698492463,
+      "grad_norm": 0.33947786688804626,
+      "learning_rate": 0.00019944357956450936,
+      "loss": 0.5782,
+      "step": 323
+    },
+    {
+      "epoch": 0.10420100502512562,
+      "grad_norm": 0.2787618339061737,
+      "learning_rate": 0.00019944002181007266,
+      "loss": 0.6023,
+      "step": 324
+    },
+    {
+      "epoch": 0.10452261306532663,
+      "grad_norm": 0.3044894337654114,
+      "learning_rate": 0.0001994364527496417,
+      "loss": 0.6425,
+      "step": 325
+    },
+    {
+      "epoch": 0.10484422110552764,
+      "grad_norm": 0.29626861214637756,
+      "learning_rate": 0.0001994328723836222,
+      "loss": 0.6735,
+      "step": 326
+    },
+    {
+      "epoch": 0.10516582914572864,
+      "grad_norm": 0.32953017950057983,
+      "learning_rate": 0.00019942928071242132,
+      "loss": 0.6903,
+      "step": 327
+    },
+    {
+      "epoch": 0.10548743718592965,
+      "grad_norm": 0.33926501870155334,
+      "learning_rate": 0.0001994256777364474,
+      "loss": 0.7772,
+      "step": 328
+    },
+    {
+      "epoch": 0.10580904522613065,
+      "grad_norm": 0.3403639793395996,
+      "learning_rate": 0.00019942206345611007,
+      "loss": 0.5947,
+      "step": 329
+    },
+    {
+      "epoch": 0.10613065326633166,
+      "grad_norm": 0.3651147186756134,
+      "learning_rate": 0.0001994184378718203,
+      "loss": 0.7705,
+      "step": 330
+    },
+    {
+      "epoch": 0.10645226130653267,
+      "grad_norm": 0.3135772943496704,
+      "learning_rate": 0.00019941480098399023,
+      "loss": 0.5679,
+      "step": 331
+    },
+    {
+      "epoch": 0.10677386934673366,
+      "grad_norm": 0.37232285737991333,
+      "learning_rate": 0.00019941115279303344,
+      "loss": 0.8304,
+      "step": 332
+    },
+    {
+      "epoch": 0.10709547738693467,
+      "grad_norm": 0.337146133184433,
+      "learning_rate": 0.00019940749329936465,
+      "loss": 0.663,
+      "step": 333
+    },
+    {
+      "epoch": 0.10741708542713568,
+      "grad_norm": 0.3423645496368408,
+      "learning_rate": 0.00019940382250339998,
+      "loss": 0.7017,
+      "step": 334
+    },
+    {
+      "epoch": 0.10773869346733668,
+      "grad_norm": 0.34679216146469116,
+      "learning_rate": 0.0001994001404055568,
+      "loss": 0.6616,
+      "step": 335
+    },
+    {
+      "epoch": 0.10806030150753769,
+      "grad_norm": 0.3776894807815552,
+      "learning_rate": 0.0001993964470062537,
+      "loss": 0.7515,
+      "step": 336
+    },
+    {
+      "epoch": 0.1083819095477387,
+      "grad_norm": 0.3490152359008789,
+      "learning_rate": 0.00019939274230591067,
+      "loss": 0.6315,
+      "step": 337
+    },
+    {
+      "epoch": 0.1087035175879397,
+      "grad_norm": 0.33338648080825806,
+      "learning_rate": 0.00019938902630494884,
+      "loss": 0.6549,
+      "step": 338
+    },
+    {
+      "epoch": 0.1090251256281407,
+      "grad_norm": 0.3925664722919464,
+      "learning_rate": 0.00019938529900379078,
+      "loss": 0.6558,
+      "step": 339
+    },
+    {
+      "epoch": 0.1093467336683417,
+      "grad_norm": 0.3501339852809906,
+      "learning_rate": 0.00019938156040286023,
+      "loss": 0.6318,
+      "step": 340
+    },
+    {
+      "epoch": 0.10966834170854271,
+      "grad_norm": 0.36366525292396545,
+      "learning_rate": 0.0001993778105025823,
+      "loss": 0.6506,
+      "step": 341
+    },
+    {
+      "epoch": 0.10998994974874372,
+      "grad_norm": 0.3830644190311432,
+      "learning_rate": 0.00019937404930338328,
+      "loss": 0.6479,
+      "step": 342
+    },
+    {
+      "epoch": 0.11031155778894472,
+      "grad_norm": 0.39970287680625916,
+      "learning_rate": 0.00019937027680569082,
+      "loss": 0.7504,
+      "step": 343
+    },
+    {
+      "epoch": 0.11063316582914573,
+      "grad_norm": 0.4591592848300934,
+      "learning_rate": 0.00019936649300993388,
+      "loss": 0.5898,
+      "step": 344
+    },
+    {
+      "epoch": 0.11095477386934674,
+      "grad_norm": 0.4169829189777374,
+      "learning_rate": 0.00019936269791654264,
+      "loss": 0.7069,
+      "step": 345
+    },
+    {
+      "epoch": 0.11127638190954774,
+      "grad_norm": 0.5072042346000671,
+      "learning_rate": 0.0001993588915259486,
+      "loss": 0.7677,
+      "step": 346
+    },
+    {
+      "epoch": 0.11159798994974875,
+      "grad_norm": 0.4963483512401581,
+      "learning_rate": 0.00019935507383858454,
+      "loss": 0.8027,
+      "step": 347
+    },
+    {
+      "epoch": 0.11191959798994974,
+      "grad_norm": 0.5276997089385986,
+      "learning_rate": 0.00019935124485488448,
+      "loss": 0.8585,
+      "step": 348
+    },
+    {
+      "epoch": 0.11224120603015075,
+      "grad_norm": 0.6158155798912048,
+      "learning_rate": 0.00019934740457528378,
+      "loss": 1.0391,
+      "step": 349
+    },
+    {
+      "epoch": 0.11256281407035176,
+      "grad_norm": 0.6715731620788574,
+      "learning_rate": 0.0001993435530002191,
+      "loss": 1.0803,
+      "step": 350
+    },
+    {
+      "epoch": 0.11256281407035176,
+      "eval_loss": 0.7560379505157471,
+      "eval_runtime": 68.9789,
+      "eval_samples_per_second": 3.624,
+      "eval_steps_per_second": 1.812,
+      "step": 350
+    },
+    {
+      "epoch": 0.11288442211055276,
+      "grad_norm": 0.2856181263923645,
+      "learning_rate": 0.00019933969013012827,
+      "loss": 0.4902,
+      "step": 351
+    },
+    {
+      "epoch": 0.11320603015075377,
+      "grad_norm": 0.443899929523468,
+      "learning_rate": 0.0001993358159654506,
+      "loss": 0.5711,
+      "step": 352
+    },
+    {
+      "epoch": 0.11352763819095478,
+      "grad_norm": 0.31197893619537354,
+      "learning_rate": 0.00019933193050662647,
+      "loss": 0.5219,
+      "step": 353
+    },
+    {
+      "epoch": 0.11384924623115578,
+      "grad_norm": 0.29089024662971497,
+      "learning_rate": 0.0001993280337540977,
+      "loss": 0.7393,
+      "step": 354
+    },
+    {
+      "epoch": 0.11417085427135679,
+      "grad_norm": 0.26892736554145813,
+      "learning_rate": 0.0001993241257083073,
+      "loss": 0.6291,
+      "step": 355
+    },
+    {
+      "epoch": 0.11449246231155778,
+      "grad_norm": 0.36393070220947266,
+      "learning_rate": 0.00019932020636969963,
+      "loss": 0.6547,
+      "step": 356
+    },
+    {
+      "epoch": 0.11481407035175879,
+      "grad_norm": 0.270700067281723,
+      "learning_rate": 0.0001993162757387203,
+      "loss": 0.7666,
+      "step": 357
+    },
+    {
+      "epoch": 0.1151356783919598,
+      "grad_norm": 0.2868196368217468,
+      "learning_rate": 0.00019931233381581618,
+      "loss": 0.6871,
+      "step": 358
+    },
+    {
+      "epoch": 0.1154572864321608,
+      "grad_norm": 0.2757340669631958,
+      "learning_rate": 0.0001993083806014355,
+      "loss": 0.6314,
+      "step": 359
+    },
+    {
+      "epoch": 0.11577889447236181,
+      "grad_norm": 0.3017682433128357,
+      "learning_rate": 0.00019930441609602772,
+      "loss": 0.6838,
+      "step": 360
+    },
+    {
+      "epoch": 0.11610050251256282,
+      "grad_norm": 0.3901962339878082,
+      "learning_rate": 0.00019930044030004357,
+      "loss": 0.6864,
+      "step": 361
+    },
+    {
+      "epoch": 0.11642211055276382,
+      "grad_norm": 0.3153689205646515,
+      "learning_rate": 0.00019929645321393506,
+      "loss": 0.6073,
+      "step": 362
+    },
+    {
+      "epoch": 0.11674371859296483,
+      "grad_norm": 0.3271941542625427,
+      "learning_rate": 0.00019929245483815559,
+      "loss": 0.8836,
+      "step": 363
+    },
+    {
+      "epoch": 0.11706532663316582,
+      "grad_norm": 0.28788936138153076,
+      "learning_rate": 0.00019928844517315967,
+      "loss": 0.6658,
+      "step": 364
+    },
+    {
+      "epoch": 0.11738693467336683,
+      "grad_norm": 0.3083438575267792,
+      "learning_rate": 0.00019928442421940324,
+      "loss": 0.6869,
+      "step": 365
+    },
+    {
+      "epoch": 0.11770854271356784,
+      "grad_norm": 0.32216233015060425,
+      "learning_rate": 0.00019928039197734346,
+      "loss": 0.8318,
+      "step": 366
+    },
+    {
+      "epoch": 0.11803015075376884,
+      "grad_norm": 0.3430779278278351,
+      "learning_rate": 0.00019927634844743878,
+      "loss": 0.6668,
+      "step": 367
+    },
+    {
+      "epoch": 0.11835175879396985,
+      "grad_norm": 0.2945297658443451,
+      "learning_rate": 0.00019927229363014894,
+      "loss": 0.6847,
+      "step": 368
+    },
+    {
+      "epoch": 0.11867336683417086,
+      "grad_norm": 0.32731348276138306,
+      "learning_rate": 0.00019926822752593495,
+      "loss": 0.7668,
+      "step": 369
+    },
+    {
+      "epoch": 0.11899497487437186,
+      "grad_norm": 0.40527111291885376,
+      "learning_rate": 0.0001992641501352591,
+      "loss": 0.7195,
+      "step": 370
+    },
+    {
+      "epoch": 0.11931658291457287,
+      "grad_norm": 0.35145846009254456,
+      "learning_rate": 0.000199260061458585,
+      "loss": 0.6594,
+      "step": 371
+    },
+    {
+      "epoch": 0.11963819095477386,
+      "grad_norm": 0.45268985629081726,
+      "learning_rate": 0.00019925596149637751,
+      "loss": 0.7827,
+      "step": 372
+    },
+    {
+      "epoch": 0.11995979899497487,
+      "grad_norm": 10.479754447937012,
+      "learning_rate": 0.00019925185024910277,
+      "loss": 0.6755,
+      "step": 373
+    },
+    {
+      "epoch": 0.12028140703517588,
+      "grad_norm": 1.0545467138290405,
+      "learning_rate": 0.00019924772771722827,
+      "loss": 0.6545,
+      "step": 374
+    },
+    {
+      "epoch": 0.12060301507537688,
+      "grad_norm": 0.35511791706085205,
+      "learning_rate": 0.00019924359390122262,
+      "loss": 0.6901,
+      "step": 375
+    },
+    {
+      "epoch": 0.12092462311557789,
+      "grad_norm": 0.34161028265953064,
+      "learning_rate": 0.0001992394488015559,
+      "loss": 0.6453,
+      "step": 376
+    },
+    {
+      "epoch": 0.1212462311557789,
+      "grad_norm": 0.32422423362731934,
+      "learning_rate": 0.00019923529241869942,
+      "loss": 0.5965,
+      "step": 377
+    },
+    {
+      "epoch": 0.1215678391959799,
+      "grad_norm": 0.34200823307037354,
+      "learning_rate": 0.00019923112475312568,
+      "loss": 0.7232,
+      "step": 378
+    },
+    {
+      "epoch": 0.1218894472361809,
+      "grad_norm": 0.43807172775268555,
+      "learning_rate": 0.00019922694580530855,
+      "loss": 0.8435,
+      "step": 379
+    },
+    {
+      "epoch": 0.12221105527638192,
+      "grad_norm": 0.3634149730205536,
+      "learning_rate": 0.00019922275557572317,
+      "loss": 0.6582,
+      "step": 380
+    },
+    {
+      "epoch": 0.12253266331658291,
+      "grad_norm": 0.34131136536598206,
+      "learning_rate": 0.00019921855406484595,
+      "loss": 0.5502,
+      "step": 381
+    },
+    {
+      "epoch": 0.12285427135678392,
+      "grad_norm": 0.3941468298435211,
+      "learning_rate": 0.0001992143412731546,
+      "loss": 0.7153,
+      "step": 382
+    },
+    {
+      "epoch": 0.12317587939698492,
+      "grad_norm": 0.33790889382362366,
+      "learning_rate": 0.00019921011720112807,
+      "loss": 0.6727,
+      "step": 383
+    },
+    {
+      "epoch": 0.12349748743718593,
+      "grad_norm": 0.38417166471481323,
+      "learning_rate": 0.00019920588184924664,
+      "loss": 0.82,
+      "step": 384
+    },
+    {
+      "epoch": 0.12381909547738694,
+      "grad_norm": 0.3321269750595093,
+      "learning_rate": 0.0001992016352179919,
+      "loss": 0.5474,
+      "step": 385
+    },
+    {
+      "epoch": 0.12414070351758794,
+      "grad_norm": 0.382433146238327,
+      "learning_rate": 0.0001991973773078466,
+      "loss": 0.6736,
+      "step": 386
+    },
+    {
+      "epoch": 0.12446231155778895,
+      "grad_norm": 0.3955988883972168,
+      "learning_rate": 0.0001991931081192949,
+      "loss": 0.738,
+      "step": 387
+    },
+    {
+      "epoch": 0.12478391959798996,
+      "grad_norm": 0.3920446038246155,
+      "learning_rate": 0.00019918882765282216,
+      "loss": 0.6773,
+      "step": 388
+    },
+    {
+      "epoch": 0.12510552763819097,
+      "grad_norm": 0.3775482773780823,
+      "learning_rate": 0.00019918453590891507,
+      "loss": 0.5593,
+      "step": 389
+    },
+    {
+      "epoch": 0.12542713567839195,
+      "grad_norm": 0.43792831897735596,
+      "learning_rate": 0.0001991802328880616,
+      "loss": 0.7129,
+      "step": 390
+    },
+    {
+      "epoch": 0.12574874371859296,
+      "grad_norm": 0.3579481542110443,
+      "learning_rate": 0.000199175918590751,
+      "loss": 0.6452,
+      "step": 391
+    },
+    {
+      "epoch": 0.12607035175879397,
+      "grad_norm": 0.4246240258216858,
+      "learning_rate": 0.00019917159301747374,
+      "loss": 0.7879,
+      "step": 392
+    },
+    {
+      "epoch": 0.12639195979899498,
+      "grad_norm": 0.3748584985733032,
+      "learning_rate": 0.00019916725616872166,
+      "loss": 0.6923,
+      "step": 393
+    },
+    {
+      "epoch": 0.126713567839196,
+      "grad_norm": 0.4105924069881439,
+      "learning_rate": 0.00019916290804498785,
+      "loss": 0.6694,
+      "step": 394
+    },
+    {
+      "epoch": 0.12703517587939697,
+      "grad_norm": 0.47539928555488586,
+      "learning_rate": 0.00019915854864676664,
+      "loss": 0.7075,
+      "step": 395
+    },
+    {
+      "epoch": 0.12735678391959798,
+      "grad_norm": 0.4180709421634674,
+      "learning_rate": 0.0001991541779745537,
+      "loss": 0.6507,
+      "step": 396
+    },
+    {
+      "epoch": 0.127678391959799,
+      "grad_norm": 0.5322070717811584,
+      "learning_rate": 0.00019914979602884598,
+      "loss": 0.8236,
+      "step": 397
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.5464707612991333,
+      "learning_rate": 0.00019914540281014168,
+      "loss": 0.8185,
+      "step": 398
+    },
+    {
+      "epoch": 0.128321608040201,
+      "grad_norm": 0.5445318818092346,
+      "learning_rate": 0.00019914099831894029,
+      "loss": 0.9853,
+      "step": 399
+    },
+    {
+      "epoch": 0.12864321608040202,
+      "grad_norm": 0.6355411410331726,
+      "learning_rate": 0.00019913658255574256,
+      "loss": 0.9983,
+      "step": 400
+    },
+    {
+      "epoch": 0.12864321608040202,
+      "eval_loss": 0.7501429915428162,
+      "eval_runtime": 69.3484,
+      "eval_samples_per_second": 3.605,
+      "eval_steps_per_second": 1.802,
+      "step": 400
+    },
+    {
+      "epoch": 0.128964824120603,
+      "grad_norm": 0.22814010083675385,
+      "learning_rate": 0.00019913215552105063,
+      "loss": 0.4693,
+      "step": 401
+    },
+    {
+      "epoch": 0.12928643216080402,
+      "grad_norm": 0.30054304003715515,
+      "learning_rate": 0.00019912771721536775,
+      "loss": 0.5335,
+      "step": 402
+    },
+    {
+      "epoch": 0.12960804020100503,
+      "grad_norm": 0.2936047315597534,
+      "learning_rate": 0.00019912326763919853,
+      "loss": 0.6358,
+      "step": 403
+    },
+    {
+      "epoch": 0.12992964824120604,
+      "grad_norm": 0.3044109046459198,
+      "learning_rate": 0.00019911880679304895,
+      "loss": 0.7196,
+      "step": 404
+    },
+    {
+      "epoch": 0.13025125628140705,
+      "grad_norm": 0.2635999023914337,
+      "learning_rate": 0.00019911433467742614,
+      "loss": 0.5881,
+      "step": 405
+    },
+    {
+      "epoch": 0.13057286432160803,
+      "grad_norm": 0.2742912173271179,
+      "learning_rate": 0.00019910985129283862,
+      "loss": 0.6828,
+      "step": 406
+    },
+    {
+      "epoch": 0.13089447236180904,
+      "grad_norm": 0.2926326394081116,
+      "learning_rate": 0.00019910535663979608,
+      "loss": 0.6252,
+      "step": 407
+    },
+    {
+      "epoch": 0.13121608040201005,
+      "grad_norm": 0.27853521704673767,
+      "learning_rate": 0.00019910085071880954,
+      "loss": 0.683,
+      "step": 408
+    },
+    {
+      "epoch": 0.13153768844221106,
+      "grad_norm": 0.3189549148082733,
+      "learning_rate": 0.00019909633353039135,
+      "loss": 0.697,
+      "step": 409
+    },
+    {
+      "epoch": 0.13185929648241207,
+      "grad_norm": 0.3010167181491852,
+      "learning_rate": 0.0001990918050750551,
+      "loss": 0.6505,
+      "step": 410
+    },
+    {
+      "epoch": 0.13218090452261305,
+      "grad_norm": 0.2827252149581909,
+      "learning_rate": 0.0001990872653533156,
+      "loss": 0.6656,
+      "step": 411
+    },
+    {
+      "epoch": 0.13250251256281406,
+      "grad_norm": 0.3428483307361603,
+      "learning_rate": 0.00019908271436568908,
+      "loss": 0.8136,
+      "step": 412
+    },
+    {
+      "epoch": 0.13282412060301507,
+      "grad_norm": 0.29689159989356995,
+      "learning_rate": 0.00019907815211269293,
+      "loss": 0.644,
+      "step": 413
+    },
+    {
+      "epoch": 0.13314572864321608,
+      "grad_norm": 0.3537396490573883,
+      "learning_rate": 0.00019907357859484586,
+      "loss": 0.8007,
+      "step": 414
+    },
+    {
+      "epoch": 0.1334673366834171,
+      "grad_norm": 0.34163200855255127,
+      "learning_rate": 0.00019906899381266787,
+      "loss": 0.7568,
+      "step": 415
+    },
+    {
+      "epoch": 0.1337889447236181,
+      "grad_norm": 0.3060307800769806,
+      "learning_rate": 0.00019906439776668023,
+      "loss": 0.6302,
+      "step": 416
+    },
+    {
+      "epoch": 0.13411055276381909,
+      "grad_norm": 0.3451438844203949,
+      "learning_rate": 0.0001990597904574055,
+      "loss": 0.9706,
+      "step": 417
+    },
+    {
+      "epoch": 0.1344321608040201,
+      "grad_norm": 0.3006080985069275,
+      "learning_rate": 0.00019905517188536752,
+      "loss": 0.717,
+      "step": 418
+    },
+    {
+      "epoch": 0.1347537688442211,
+      "grad_norm": 0.3454923927783966,
+      "learning_rate": 0.0001990505420510914,
+      "loss": 0.5597,
+      "step": 419
+    },
+    {
+      "epoch": 0.13507537688442212,
+      "grad_norm": 0.3017250895500183,
+      "learning_rate": 0.00019904590095510353,
+      "loss": 0.5252,
+      "step": 420
+    },
+    {
+      "epoch": 0.13539698492462313,
+      "grad_norm": 0.36881935596466064,
+      "learning_rate": 0.0001990412485979316,
+      "loss": 0.9081,
+      "step": 421
+    },
+    {
+      "epoch": 0.1357185929648241,
+      "grad_norm": 0.35308098793029785,
+      "learning_rate": 0.00019903658498010457,
+      "loss": 0.8458,
+      "step": 422
+    },
+    {
+      "epoch": 0.13604020100502512,
+      "grad_norm": 0.3268420994281769,
+      "learning_rate": 0.00019903191010215264,
+      "loss": 0.6632,
+      "step": 423
+    },
+    {
+      "epoch": 0.13636180904522613,
+      "grad_norm": 0.31583157181739807,
+      "learning_rate": 0.0001990272239646074,
+      "loss": 0.6208,
+      "step": 424
+    },
+    {
+      "epoch": 0.13668341708542714,
+      "grad_norm": 0.3173026144504547,
+      "learning_rate": 0.00019902252656800158,
+      "loss": 0.653,
+      "step": 425
+    },
+    {
+      "epoch": 0.13700502512562815,
+      "grad_norm": 0.341020792722702,
+      "learning_rate": 0.00019901781791286925,
+      "loss": 0.7264,
+      "step": 426
+    },
+    {
+      "epoch": 0.13732663316582916,
+      "grad_norm": 0.32180771231651306,
+      "learning_rate": 0.00019901309799974585,
+      "loss": 0.588,
+      "step": 427
+    },
+    {
+      "epoch": 0.13764824120603014,
+      "grad_norm": 0.3036893308162689,
+      "learning_rate": 0.00019900836682916796,
+      "loss": 0.5435,
+      "step": 428
+    },
+    {
+      "epoch": 0.13796984924623115,
+      "grad_norm": 0.3573414981365204,
+      "learning_rate": 0.00019900362440167352,
+      "loss": 0.6872,
+      "step": 429
+    },
+    {
+      "epoch": 0.13829145728643216,
+      "grad_norm": 0.3649362623691559,
+      "learning_rate": 0.0001989988707178017,
+      "loss": 0.6698,
+      "step": 430
+    },
+    {
+      "epoch": 0.13861306532663317,
+      "grad_norm": 0.39232033491134644,
+      "learning_rate": 0.000198994105778093,
+      "loss": 0.6365,
+      "step": 431
+    },
+    {
+      "epoch": 0.13893467336683418,
+      "grad_norm": 0.3411564528942108,
+      "learning_rate": 0.00019898932958308915,
+      "loss": 0.534,
+      "step": 432
+    },
+    {
+      "epoch": 0.13925628140703517,
+      "grad_norm": 0.32758474349975586,
+      "learning_rate": 0.0001989845421333332,
+      "loss": 0.5234,
+      "step": 433
+    },
+    {
+      "epoch": 0.13957788944723618,
+      "grad_norm": 0.3756495416164398,
+      "learning_rate": 0.0001989797434293695,
+      "loss": 0.6011,
+      "step": 434
+    },
+    {
+      "epoch": 0.13989949748743719,
+      "grad_norm": 0.4003129005432129,
+      "learning_rate": 0.00019897493347174365,
+      "loss": 0.5618,
+      "step": 435
+    },
+    {
+      "epoch": 0.1402211055276382,
+      "grad_norm": 0.36236846446990967,
+      "learning_rate": 0.00019897011226100247,
+      "loss": 0.7106,
+      "step": 436
+    },
+    {
+      "epoch": 0.1405427135678392,
+      "grad_norm": 0.3520951271057129,
+      "learning_rate": 0.00019896527979769416,
+      "loss": 0.6343,
+      "step": 437
+    },
+    {
+      "epoch": 0.1408643216080402,
+      "grad_norm": 0.36788055300712585,
+      "learning_rate": 0.00019896043608236808,
+      "loss": 0.5582,
+      "step": 438
+    },
+    {
+      "epoch": 0.1411859296482412,
+      "grad_norm": 0.35627469420433044,
+      "learning_rate": 0.00019895558111557505,
+      "loss": 0.57,
+      "step": 439
+    },
+    {
+      "epoch": 0.1415075376884422,
+      "grad_norm": 0.3308781087398529,
+      "learning_rate": 0.00019895071489786702,
+      "loss": 0.5501,
+      "step": 440
+    },
+    {
+      "epoch": 0.14182914572864322,
+      "grad_norm": 0.39617815613746643,
+      "learning_rate": 0.00019894583742979723,
+      "loss": 0.7328,
+      "step": 441
+    },
+    {
+      "epoch": 0.14215075376884423,
+      "grad_norm": 0.37014466524124146,
+      "learning_rate": 0.00019894094871192026,
+      "loss": 0.5943,
+      "step": 442
+    },
+    {
+      "epoch": 0.14247236180904524,
+      "grad_norm": 0.4164433777332306,
+      "learning_rate": 0.00019893604874479196,
+      "loss": 0.6102,
+      "step": 443
+    },
+    {
+      "epoch": 0.14279396984924622,
+      "grad_norm": 0.4099109470844269,
+      "learning_rate": 0.0001989311375289694,
+      "loss": 0.6041,
+      "step": 444
+    },
+    {
+      "epoch": 0.14311557788944723,
+      "grad_norm": 0.44024282693862915,
+      "learning_rate": 0.00019892621506501098,
+      "loss": 0.6486,
+      "step": 445
+    },
+    {
+      "epoch": 0.14343718592964824,
+      "grad_norm": 0.42002180218696594,
+      "learning_rate": 0.00019892128135347642,
+      "loss": 0.5941,
+      "step": 446
+    },
+    {
+      "epoch": 0.14375879396984925,
+      "grad_norm": 0.5036507844924927,
+      "learning_rate": 0.0001989163363949266,
+      "loss": 0.7309,
+      "step": 447
+    },
+    {
+      "epoch": 0.14408040201005026,
+      "grad_norm": 0.5919266939163208,
+      "learning_rate": 0.00019891138018992375,
+      "loss": 1.0796,
+      "step": 448
+    },
+    {
+      "epoch": 0.14440201005025124,
+      "grad_norm": 0.5990227460861206,
+      "learning_rate": 0.0001989064127390314,
+      "loss": 0.9935,
+      "step": 449
+    },
+    {
+      "epoch": 0.14472361809045226,
+      "grad_norm": 0.8288195729255676,
+      "learning_rate": 0.00019890143404281433,
+      "loss": 1.1895,
+      "step": 450
+    },
+    {
+      "epoch": 0.14472361809045226,
+      "eval_loss": 0.7556187510490417,
+      "eval_runtime": 68.9983,
+      "eval_samples_per_second": 3.623,
+      "eval_steps_per_second": 1.812,
+      "step": 450
+    },
+    {
+      "epoch": 0.14504522613065327,
+      "grad_norm": 0.32703897356987,
+      "learning_rate": 0.00019889644410183856,
+      "loss": 0.5789,
+      "step": 451
+    },
+    {
+      "epoch": 0.14536683417085428,
+      "grad_norm": 0.3096190392971039,
+      "learning_rate": 0.0001988914429166715,
+      "loss": 0.5582,
+      "step": 452
+    },
+    {
+      "epoch": 0.14568844221105529,
+      "grad_norm": 0.3071274757385254,
+      "learning_rate": 0.00019888643048788173,
+      "loss": 0.6454,
+      "step": 453
+    },
+    {
+      "epoch": 0.14601005025125627,
+      "grad_norm": 0.27530086040496826,
+      "learning_rate": 0.00019888140681603912,
+      "loss": 0.6707,
+      "step": 454
+    },
+    {
+      "epoch": 0.14633165829145728,
+      "grad_norm": 0.2821219265460968,
+      "learning_rate": 0.00019887637190171492,
+      "loss": 0.6623,
+      "step": 455
+    },
+    {
+      "epoch": 0.1466532663316583,
+      "grad_norm": 0.24380551278591156,
+      "learning_rate": 0.00019887132574548152,
+      "loss": 0.4981,
+      "step": 456
+    },
+    {
+      "epoch": 0.1469748743718593,
+      "grad_norm": 0.25048956274986267,
+      "learning_rate": 0.00019886626834791265,
+      "loss": 0.453,
+      "step": 457
+    },
+    {
+      "epoch": 0.1472964824120603,
+      "grad_norm": 0.274775892496109,
+      "learning_rate": 0.00019886119970958334,
+      "loss": 0.5301,
+      "step": 458
+    },
+    {
+      "epoch": 0.14761809045226132,
+      "grad_norm": 0.3389226496219635,
+      "learning_rate": 0.00019885611983106984,
+      "loss": 0.7759,
+      "step": 459
+    },
+    {
+      "epoch": 0.1479396984924623,
+      "grad_norm": 0.31445416808128357,
+      "learning_rate": 0.0001988510287129498,
+      "loss": 0.7493,
+      "step": 460
+    },
+    {
+      "epoch": 0.1482613065326633,
+      "grad_norm": 0.29652178287506104,
+      "learning_rate": 0.00019884592635580197,
+      "loss": 0.6816,
+      "step": 461
+    },
+    {
+      "epoch": 0.14858291457286432,
+      "grad_norm": 0.3157201409339905,
+      "learning_rate": 0.0001988408127602065,
+      "loss": 0.7551,
+      "step": 462
+    },
+    {
+      "epoch": 0.14890452261306533,
+      "grad_norm": 0.3251638412475586,
+      "learning_rate": 0.00019883568792674483,
+      "loss": 0.6796,
+      "step": 463
+    },
+    {
+      "epoch": 0.14922613065326634,
+      "grad_norm": 0.31846338510513306,
+      "learning_rate": 0.0001988305518559996,
+      "loss": 0.6469,
+      "step": 464
+    },
+    {
+      "epoch": 0.14954773869346732,
+      "grad_norm": 0.35254862904548645,
+      "learning_rate": 0.00019882540454855478,
+      "loss": 0.8744,
+      "step": 465
+    },
+    {
+      "epoch": 0.14986934673366833,
+      "grad_norm": 0.3078581988811493,
+      "learning_rate": 0.0001988202460049956,
+      "loss": 0.6734,
+      "step": 466
+    },
+    {
+      "epoch": 0.15019095477386935,
+      "grad_norm": 0.3021988868713379,
+      "learning_rate": 0.0001988150762259085,
+      "loss": 0.6144,
+      "step": 467
+    },
+    {
+      "epoch": 0.15051256281407036,
+      "grad_norm": 0.31421002745628357,
+      "learning_rate": 0.0001988098952118814,
+      "loss": 0.6865,
+      "step": 468
+    },
+    {
+      "epoch": 0.15083417085427137,
+      "grad_norm": 0.30060771107673645,
+      "learning_rate": 0.00019880470296350323,
+      "loss": 0.7211,
+      "step": 469
+    },
+    {
+      "epoch": 0.15115577889447235,
+      "grad_norm": 0.36412227153778076,
+      "learning_rate": 0.0001987994994813644,
+      "loss": 0.7156,
+      "step": 470
+    },
+    {
+      "epoch": 0.15147738693467336,
+      "grad_norm": 0.31202271580696106,
+      "learning_rate": 0.00019879428476605654,
+      "loss": 0.5825,
+      "step": 471
+    },
+    {
+      "epoch": 0.15179899497487437,
+      "grad_norm": 0.3653612434864044,
+      "learning_rate": 0.00019878905881817252,
+      "loss": 0.7243,
+      "step": 472
+    },
+    {
+      "epoch": 0.15212060301507538,
+      "grad_norm": 0.3118390440940857,
+      "learning_rate": 0.0001987838216383065,
+      "loss": 0.6086,
+      "step": 473
+    },
+    {
+      "epoch": 0.1524422110552764,
+      "grad_norm": 0.35151809453964233,
+      "learning_rate": 0.00019877857322705397,
+      "loss": 0.7143,
+      "step": 474
+    },
+    {
+      "epoch": 0.1527638190954774,
+      "grad_norm": 0.33639010787010193,
+      "learning_rate": 0.0001987733135850116,
+      "loss": 0.6557,
+      "step": 475
+    },
+    {
+      "epoch": 0.15308542713567838,
+      "grad_norm": 0.3119165301322937,
+      "learning_rate": 0.00019876804271277748,
+      "loss": 0.691,
+      "step": 476
+    },
+    {
+      "epoch": 0.1534070351758794,
+      "grad_norm": 0.3577800691127777,
+      "learning_rate": 0.0001987627606109508,
+      "loss": 0.7274,
+      "step": 477
+    },
+    {
+      "epoch": 0.1537286432160804,
+      "grad_norm": 0.3867248296737671,
+      "learning_rate": 0.00019875746728013219,
+      "loss": 0.8895,
+      "step": 478
+    },
+    {
+      "epoch": 0.1540502512562814,
+      "grad_norm": 0.3830859065055847,
+      "learning_rate": 0.00019875216272092342,
+      "loss": 0.7446,
+      "step": 479
+    },
+    {
+      "epoch": 0.15437185929648242,
+      "grad_norm": 0.3739246129989624,
+      "learning_rate": 0.0001987468469339276,
+      "loss": 0.6663,
+      "step": 480
+    },
+    {
+      "epoch": 0.1546934673366834,
+      "grad_norm": 0.35838425159454346,
+      "learning_rate": 0.00019874151991974918,
+      "loss": 0.5943,
+      "step": 481
+    },
+    {
+      "epoch": 0.15501507537688441,
+      "grad_norm": 0.3706784248352051,
+      "learning_rate": 0.0001987361816789938,
+      "loss": 0.6672,
+      "step": 482
+    },
+    {
+      "epoch": 0.15533668341708542,
+      "grad_norm": 0.34062620997428894,
+      "learning_rate": 0.00019873083221226833,
+      "loss": 0.5448,
+      "step": 483
+    },
+    {
+      "epoch": 0.15565829145728644,
+      "grad_norm": 0.37892401218414307,
+      "learning_rate": 0.0001987254715201811,
+      "loss": 0.7095,
+      "step": 484
+    },
+    {
+      "epoch": 0.15597989949748745,
+      "grad_norm": 0.3556469976902008,
+      "learning_rate": 0.00019872009960334155,
+      "loss": 0.6105,
+      "step": 485
+    },
+    {
+      "epoch": 0.15630150753768846,
+      "grad_norm": 0.35378313064575195,
+      "learning_rate": 0.00019871471646236045,
+      "loss": 0.6477,
+      "step": 486
+    },
+    {
+      "epoch": 0.15662311557788944,
+      "grad_norm": 0.35008472204208374,
+      "learning_rate": 0.00019870932209784981,
+      "loss": 0.5435,
+      "step": 487
+    },
+    {
+      "epoch": 0.15694472361809045,
+      "grad_norm": 0.3734475076198578,
+      "learning_rate": 0.00019870391651042297,
+      "loss": 0.6492,
+      "step": 488
+    },
+    {
+      "epoch": 0.15726633165829146,
+      "grad_norm": 0.42147645354270935,
+      "learning_rate": 0.00019869849970069456,
+      "loss": 0.6551,
+      "step": 489
+    },
+    {
+      "epoch": 0.15758793969849247,
+      "grad_norm": 0.4381759464740753,
+      "learning_rate": 0.00019869307166928043,
+      "loss": 0.7757,
+      "step": 490
+    },
+    {
+      "epoch": 0.15790954773869348,
+      "grad_norm": 0.3745943009853363,
+      "learning_rate": 0.00019868763241679777,
+      "loss": 0.6508,
+      "step": 491
+    },
+    {
+      "epoch": 0.15823115577889446,
+      "grad_norm": 0.4166151285171509,
+      "learning_rate": 0.00019868218194386494,
+      "loss": 0.7684,
+      "step": 492
+    },
+    {
+      "epoch": 0.15855276381909547,
+      "grad_norm": 0.39115533232688904,
+      "learning_rate": 0.00019867672025110163,
+      "loss": 0.6156,
+      "step": 493
+    },
+    {
+      "epoch": 0.15887437185929648,
+      "grad_norm": 0.39986225962638855,
+      "learning_rate": 0.0001986712473391289,
+      "loss": 0.6176,
+      "step": 494
+    },
+    {
+      "epoch": 0.1591959798994975,
+      "grad_norm": 0.44499534368515015,
+      "learning_rate": 0.00019866576320856893,
+      "loss": 0.7105,
+      "step": 495
+    },
+    {
+      "epoch": 0.1595175879396985,
+      "grad_norm": 0.568112850189209,
+      "learning_rate": 0.0001986602678600453,
+      "loss": 0.98,
+      "step": 496
+    },
+    {
+      "epoch": 0.15983919597989948,
+      "grad_norm": 0.5521255731582642,
+      "learning_rate": 0.00019865476129418274,
+      "loss": 0.949,
+      "step": 497
+    },
+    {
+      "epoch": 0.1601608040201005,
+      "grad_norm": 0.5176576375961304,
+      "learning_rate": 0.0001986492435116074,
+      "loss": 0.9036,
+      "step": 498
+    },
+    {
+      "epoch": 0.1604824120603015,
+      "grad_norm": 0.6676732301712036,
+      "learning_rate": 0.0001986437145129466,
+      "loss": 1.0617,
+      "step": 499
+    },
+    {
+      "epoch": 0.16080402010050251,
+      "grad_norm": 0.728453516960144,
+      "learning_rate": 0.00019863817429882898,
+      "loss": 1.1736,
+      "step": 500
+    },
+    {
+      "epoch": 0.16080402010050251,
+      "eval_loss": 0.7435534596443176,
+      "eval_runtime": 69.033,
+      "eval_samples_per_second": 3.621,
+      "eval_steps_per_second": 1.811,
+      "step": 500
+    },
+    {
+      "epoch": 0.16112562814070353,
+      "grad_norm": 0.2767173945903778,
+      "learning_rate": 0.00019863262286988443,
+      "loss": 0.4322,
+      "step": 501
+    },
+    {
+      "epoch": 0.16144723618090454,
+      "grad_norm": 0.3126039206981659,
+      "learning_rate": 0.0001986270602267441,
+      "loss": 0.6167,
+      "step": 502
+    },
+    {
+      "epoch": 0.16176884422110552,
+      "grad_norm": 0.3135744631290436,
+      "learning_rate": 0.00019862148637004055,
+      "loss": 0.6951,
+      "step": 503
+    },
+    {
+      "epoch": 0.16209045226130653,
+      "grad_norm": 0.2798527181148529,
+      "learning_rate": 0.00019861590130040738,
+      "loss": 0.6645,
+      "step": 504
+    },
+    {
+      "epoch": 0.16241206030150754,
+      "grad_norm": 0.26567572355270386,
+      "learning_rate": 0.00019861030501847968,
+      "loss": 0.6042,
+      "step": 505
+    },
+    {
+      "epoch": 0.16273366834170855,
+      "grad_norm": 0.33532339334487915,
+      "learning_rate": 0.00019860469752489368,
+      "loss": 0.8365,
+      "step": 506
+    },
+    {
+      "epoch": 0.16305527638190956,
+      "grad_norm": 0.2620905637741089,
+      "learning_rate": 0.00019859907882028695,
+      "loss": 0.737,
+      "step": 507
+    },
+    {
+      "epoch": 0.16337688442211054,
+      "grad_norm": 0.2757718861103058,
+      "learning_rate": 0.00019859344890529835,
+      "loss": 0.6191,
+      "step": 508
+    },
+    {
+      "epoch": 0.16369849246231155,
+      "grad_norm": 0.30086466670036316,
+      "learning_rate": 0.00019858780778056793,
+      "loss": 0.7429,
+      "step": 509
+    },
+    {
+      "epoch": 0.16402010050251256,
+      "grad_norm": 0.2759198844432831,
+      "learning_rate": 0.0001985821554467371,
+      "loss": 0.6942,
+      "step": 510
+    },
+    {
+      "epoch": 0.16434170854271357,
+      "grad_norm": 0.28879162669181824,
+      "learning_rate": 0.00019857649190444852,
+      "loss": 0.7349,
+      "step": 511
+    },
+    {
+      "epoch": 0.16466331658291458,
+      "grad_norm": 0.2868627905845642,
+      "learning_rate": 0.00019857081715434608,
+      "loss": 0.7593,
+      "step": 512
+    },
+    {
+      "epoch": 0.16498492462311556,
+      "grad_norm": 0.310342013835907,
+      "learning_rate": 0.00019856513119707495,
+      "loss": 0.8321,
+      "step": 513
+    },
+    {
+      "epoch": 0.16530653266331657,
+      "grad_norm": 0.2921718955039978,
+      "learning_rate": 0.0001985594340332817,
+      "loss": 0.6556,
+      "step": 514
+    },
+    {
+      "epoch": 0.16562814070351758,
+      "grad_norm": 0.31096580624580383,
+      "learning_rate": 0.00019855372566361405,
+      "loss": 0.7388,
+      "step": 515
+    },
+    {
+      "epoch": 0.1659497487437186,
+      "grad_norm": 0.3317711651325226,
+      "learning_rate": 0.00019854800608872094,
+      "loss": 0.6861,
+      "step": 516
+    },
+    {
+      "epoch": 0.1662713567839196,
+      "grad_norm": 0.3107955753803253,
+      "learning_rate": 0.00019854227530925278,
+      "loss": 0.5996,
+      "step": 517
+    },
+    {
+      "epoch": 0.16659296482412062,
+      "grad_norm": 0.3388027250766754,
+      "learning_rate": 0.00019853653332586104,
+      "loss": 0.7158,
+      "step": 518
+    },
+    {
+      "epoch": 0.1669145728643216,
+      "grad_norm": 0.31410157680511475,
+      "learning_rate": 0.00019853078013919864,
+      "loss": 0.6481,
+      "step": 519
+    },
+    {
+      "epoch": 0.1672361809045226,
+      "grad_norm": 0.32940009236335754,
+      "learning_rate": 0.00019852501574991967,
+      "loss": 0.6889,
+      "step": 520
+    },
+    {
+      "epoch": 0.16755778894472362,
+      "grad_norm": 0.3290988802909851,
+      "learning_rate": 0.0001985192401586795,
+      "loss": 0.6266,
+      "step": 521
+    },
+    {
+      "epoch": 0.16787939698492463,
+      "grad_norm": 0.28928038477897644,
+      "learning_rate": 0.00019851345336613485,
+      "loss": 0.5806,
+      "step": 522
+    },
+    {
+      "epoch": 0.16820100502512564,
+      "grad_norm": 0.361422061920166,
+      "learning_rate": 0.0001985076553729436,
+      "loss": 0.7427,
+      "step": 523
+    },
+    {
+      "epoch": 0.16852261306532662,
+      "grad_norm": 0.31583306193351746,
+      "learning_rate": 0.00019850184617976503,
+      "loss": 0.6734,
+      "step": 524
+    },
+    {
+      "epoch": 0.16884422110552763,
+      "grad_norm": 0.2929077744483948,
+      "learning_rate": 0.00019849602578725955,
+      "loss": 0.5857,
+      "step": 525
+    },
+    {
+      "epoch": 0.16916582914572864,
+      "grad_norm": 0.32200345396995544,
+      "learning_rate": 0.00019849019419608893,
+      "loss": 0.627,
+      "step": 526
+    },
+    {
+      "epoch": 0.16948743718592965,
+      "grad_norm": 0.329223096370697,
+      "learning_rate": 0.00019848435140691625,
+      "loss": 0.5954,
+      "step": 527
+    },
+    {
+      "epoch": 0.16980904522613066,
+      "grad_norm": 0.3635766804218292,
+      "learning_rate": 0.0001984784974204058,
+      "loss": 0.58,
+      "step": 528
+    },
+    {
+      "epoch": 0.17013065326633167,
+      "grad_norm": 0.33528000116348267,
+      "learning_rate": 0.0001984726322372231,
+      "loss": 0.5735,
+      "step": 529
+    },
+    {
+      "epoch": 0.17045226130653265,
+      "grad_norm": 0.4059266448020935,
+      "learning_rate": 0.0001984667558580351,
+      "loss": 0.8664,
+      "step": 530
+    },
+    {
+      "epoch": 0.17077386934673366,
+      "grad_norm": 0.35762912034988403,
+      "learning_rate": 0.00019846086828350984,
+      "loss": 0.6193,
+      "step": 531
+    },
+    {
+      "epoch": 0.17109547738693467,
+      "grad_norm": 0.43446946144104004,
+      "learning_rate": 0.0001984549695143168,
+      "loss": 0.721,
+      "step": 532
+    },
+    {
+      "epoch": 0.17141708542713568,
+      "grad_norm": 0.33650630712509155,
+      "learning_rate": 0.00019844905955112655,
+      "loss": 0.6517,
+      "step": 533
+    },
+    {
+      "epoch": 0.1717386934673367,
+      "grad_norm": 0.39659634232521057,
+      "learning_rate": 0.00019844313839461112,
+      "loss": 0.7081,
+      "step": 534
+    },
+    {
+      "epoch": 0.17206030150753768,
+      "grad_norm": 0.3652941584587097,
+      "learning_rate": 0.00019843720604544366,
+      "loss": 0.6525,
+      "step": 535
+    },
+    {
+      "epoch": 0.1723819095477387,
+      "grad_norm": 0.3738958537578583,
+      "learning_rate": 0.00019843126250429868,
+      "loss": 0.6773,
+      "step": 536
+    },
+    {
+      "epoch": 0.1727035175879397,
+      "grad_norm": 0.37862101197242737,
+      "learning_rate": 0.000198425307771852,
+      "loss": 0.817,
+      "step": 537
+    },
+    {
+      "epoch": 0.1730251256281407,
+      "grad_norm": 0.4404222369194031,
+      "learning_rate": 0.00019841934184878054,
+      "loss": 0.7331,
+      "step": 538
+    },
+    {
+      "epoch": 0.17334673366834172,
+      "grad_norm": 0.34747540950775146,
+      "learning_rate": 0.0001984133647357627,
+      "loss": 0.5716,
+      "step": 539
+    },
+    {
+      "epoch": 0.1736683417085427,
+      "grad_norm": 0.3861823081970215,
+      "learning_rate": 0.00019840737643347803,
+      "loss": 0.542,
+      "step": 540
+    },
+    {
+      "epoch": 0.1739899497487437,
+      "grad_norm": 0.4850083887577057,
+      "learning_rate": 0.00019840137694260738,
+      "loss": 0.6521,
+      "step": 541
+    },
+    {
+      "epoch": 0.17431155778894472,
+      "grad_norm": 0.4260999858379364,
+      "learning_rate": 0.00019839536626383285,
+      "loss": 0.7769,
+      "step": 542
+    },
+    {
+      "epoch": 0.17463316582914573,
+      "grad_norm": 0.3641606867313385,
+      "learning_rate": 0.00019838934439783782,
+      "loss": 0.5607,
+      "step": 543
+    },
+    {
+      "epoch": 0.17495477386934674,
+      "grad_norm": 0.407595157623291,
+      "learning_rate": 0.00019838331134530704,
+      "loss": 0.6719,
+      "step": 544
+    },
+    {
+      "epoch": 0.17527638190954775,
+      "grad_norm": 0.6115728616714478,
+      "learning_rate": 0.00019837726710692635,
+      "loss": 0.7112,
+      "step": 545
+    },
+    {
+      "epoch": 0.17559798994974873,
+      "grad_norm": 0.5357325077056885,
+      "learning_rate": 0.000198371211683383,
+      "loss": 0.794,
+      "step": 546
+    },
+    {
+      "epoch": 0.17591959798994974,
+      "grad_norm": 0.4672856330871582,
+      "learning_rate": 0.0001983651450753655,
+      "loss": 0.7251,
+      "step": 547
+    },
+    {
+      "epoch": 0.17624120603015075,
+      "grad_norm": 0.5625156760215759,
+      "learning_rate": 0.00019835906728356357,
+      "loss": 0.8617,
+      "step": 548
+    },
+    {
+      "epoch": 0.17656281407035176,
+      "grad_norm": 0.6480356454849243,
+      "learning_rate": 0.00019835297830866826,
+      "loss": 1.0773,
+      "step": 549
+    },
+    {
+      "epoch": 0.17688442211055277,
+      "grad_norm": 0.8110795021057129,
+      "learning_rate": 0.0001983468781513718,
+      "loss": 1.1411,
+      "step": 550
+    },
+    {
+      "epoch": 0.17688442211055277,
+      "eval_loss": 0.7360964417457581,
+      "eval_runtime": 69.2953,
+      "eval_samples_per_second": 3.608,
+      "eval_steps_per_second": 1.804,
+      "step": 550
+    },
+    {
+      "epoch": 0.17720603015075376,
+      "grad_norm": 0.24573051929473877,
+      "learning_rate": 0.00019834076681236784,
+      "loss": 0.4445,
+      "step": 551
+    },
+    {
+      "epoch": 0.17752763819095477,
+      "grad_norm": 0.31515994668006897,
+      "learning_rate": 0.00019833464429235118,
+      "loss": 0.5222,
+      "step": 552
+    },
+    {
+      "epoch": 0.17784924623115578,
+      "grad_norm": 0.32813042402267456,
+      "learning_rate": 0.00019832851059201794,
+      "loss": 0.6117,
+      "step": 553
+    },
+    {
+      "epoch": 0.1781708542713568,
+      "grad_norm": 0.2774469256401062,
+      "learning_rate": 0.00019832236571206547,
+      "loss": 0.5915,
+      "step": 554
+    },
+    {
+      "epoch": 0.1784924623115578,
+      "grad_norm": 0.3163706064224243,
+      "learning_rate": 0.00019831620965319243,
+      "loss": 0.7147,
+      "step": 555
+    },
+    {
+      "epoch": 0.17881407035175878,
+      "grad_norm": 0.31922855973243713,
+      "learning_rate": 0.0001983100424160988,
+      "loss": 0.6554,
+      "step": 556
+    },
+    {
+      "epoch": 0.1791356783919598,
+      "grad_norm": 0.27969813346862793,
+      "learning_rate": 0.00019830386400148571,
+      "loss": 0.6454,
+      "step": 557
+    },
+    {
+      "epoch": 0.1794572864321608,
+      "grad_norm": 0.29126226902008057,
+      "learning_rate": 0.00019829767441005566,
+      "loss": 0.6616,
+      "step": 558
+    },
+    {
+      "epoch": 0.1797788944723618,
+      "grad_norm": 0.3020893633365631,
+      "learning_rate": 0.00019829147364251238,
+      "loss": 0.7017,
+      "step": 559
+    },
+    {
+      "epoch": 0.18010050251256282,
+      "grad_norm": 0.2971557676792145,
+      "learning_rate": 0.00019828526169956084,
+      "loss": 0.6774,
+      "step": 560
+    },
+    {
+      "epoch": 0.18042211055276383,
+      "grad_norm": 0.3015565276145935,
+      "learning_rate": 0.00019827903858190735,
+      "loss": 0.7456,
+      "step": 561
+    },
+    {
+      "epoch": 0.1807437185929648,
+      "grad_norm": 0.26172611117362976,
+      "learning_rate": 0.00019827280429025946,
+      "loss": 0.5977,
+      "step": 562
+    },
+    {
+      "epoch": 0.18106532663316582,
+      "grad_norm": 0.3006877601146698,
+      "learning_rate": 0.000198266558825326,
+      "loss": 0.6765,
+      "step": 563
+    },
+    {
+      "epoch": 0.18138693467336683,
+      "grad_norm": 0.323259174823761,
+      "learning_rate": 0.00019826030218781702,
+      "loss": 0.6914,
+      "step": 564
+    },
+    {
+      "epoch": 0.18170854271356784,
+      "grad_norm": 0.29998642206192017,
+      "learning_rate": 0.0001982540343784439,
+      "loss": 0.719,
+      "step": 565
+    },
+    {
+      "epoch": 0.18203015075376885,
+      "grad_norm": 0.31888896226882935,
+      "learning_rate": 0.00019824775539791927,
+      "loss": 0.7416,
+      "step": 566
+    },
+    {
+      "epoch": 0.18235175879396984,
+      "grad_norm": 0.3023034334182739,
+      "learning_rate": 0.00019824146524695702,
+      "loss": 0.6543,
+      "step": 567
+    },
+    {
+      "epoch": 0.18267336683417085,
+      "grad_norm": 0.31559985876083374,
+      "learning_rate": 0.00019823516392627232,
+      "loss": 0.7375,
+      "step": 568
+    },
+    {
+      "epoch": 0.18299497487437186,
+      "grad_norm": 0.33618640899658203,
+      "learning_rate": 0.00019822885143658164,
+      "loss": 0.6168,
+      "step": 569
+    },
+    {
+      "epoch": 0.18331658291457287,
+      "grad_norm": 0.3048398196697235,
+      "learning_rate": 0.00019822252777860263,
+      "loss": 0.6244,
+      "step": 570
+    },
+    {
+      "epoch": 0.18363819095477388,
+      "grad_norm": 0.31295210123062134,
+      "learning_rate": 0.0001982161929530543,
+      "loss": 0.6661,
+      "step": 571
+    },
+    {
+      "epoch": 0.1839597989949749,
+      "grad_norm": 0.32959219813346863,
+      "learning_rate": 0.00019820984696065694,
+      "loss": 0.6783,
+      "step": 572
+    },
+    {
+      "epoch": 0.18428140703517587,
+      "grad_norm": 0.3208419382572174,
+      "learning_rate": 0.000198203489802132,
+      "loss": 0.5732,
+      "step": 573
+    },
+    {
+      "epoch": 0.18460301507537688,
+      "grad_norm": 0.33770766854286194,
+      "learning_rate": 0.00019819712147820228,
+      "loss": 0.7576,
+      "step": 574
+    },
+    {
+      "epoch": 0.1849246231155779,
+      "grad_norm": 0.3535097539424896,
+      "learning_rate": 0.00019819074198959186,
+      "loss": 0.6628,
+      "step": 575
+    },
+    {
+      "epoch": 0.1852462311557789,
+      "grad_norm": 0.3424528241157532,
+      "learning_rate": 0.00019818435133702608,
+      "loss": 0.6406,
+      "step": 576
+    },
+    {
+      "epoch": 0.1855678391959799,
+      "grad_norm": 0.3320842385292053,
+      "learning_rate": 0.0001981779495212315,
+      "loss": 0.6295,
+      "step": 577
+    },
+    {
+      "epoch": 0.1858894472361809,
+      "grad_norm": 0.3642159104347229,
+      "learning_rate": 0.000198171536542936,
+      "loss": 0.7184,
+      "step": 578
+    },
+    {
+      "epoch": 0.1862110552763819,
+      "grad_norm": 0.3486365079879761,
+      "learning_rate": 0.0001981651124028687,
+      "loss": 0.6008,
+      "step": 579
+    },
+    {
+      "epoch": 0.18653266331658291,
+      "grad_norm": 0.37552589178085327,
+      "learning_rate": 0.00019815867710176,
+      "loss": 0.6776,
+      "step": 580
+    },
+    {
+      "epoch": 0.18685427135678392,
+      "grad_norm": 0.3602588176727295,
+      "learning_rate": 0.00019815223064034165,
+      "loss": 0.6486,
+      "step": 581
+    },
+    {
+      "epoch": 0.18717587939698493,
+      "grad_norm": 0.315745085477829,
+      "learning_rate": 0.00019814577301934647,
+      "loss": 0.5767,
+      "step": 582
+    },
+    {
+      "epoch": 0.18749748743718592,
+      "grad_norm": 0.36800962686538696,
+      "learning_rate": 0.00019813930423950876,
+      "loss": 0.722,
+      "step": 583
+    },
+    {
+      "epoch": 0.18781909547738693,
+      "grad_norm": 0.6591741442680359,
+      "learning_rate": 0.00019813282430156392,
+      "loss": 0.636,
+      "step": 584
+    },
+    {
+      "epoch": 0.18814070351758794,
+      "grad_norm": 0.33843374252319336,
+      "learning_rate": 0.0001981263332062488,
+      "loss": 0.522,
+      "step": 585
+    },
+    {
+      "epoch": 0.18846231155778895,
+      "grad_norm": 0.3784267008304596,
+      "learning_rate": 0.00019811983095430132,
+      "loss": 0.6057,
+      "step": 586
+    },
+    {
+      "epoch": 0.18878391959798996,
+      "grad_norm": 0.3550659120082855,
+      "learning_rate": 0.0001981133175464608,
+      "loss": 0.5726,
+      "step": 587
+    },
+    {
+      "epoch": 0.18910552763819097,
+      "grad_norm": 0.42098724842071533,
+      "learning_rate": 0.00019810679298346782,
+      "loss": 0.7908,
+      "step": 588
+    },
+    {
+      "epoch": 0.18942713567839195,
+      "grad_norm": 0.3616294264793396,
+      "learning_rate": 0.00019810025726606418,
+      "loss": 0.5779,
+      "step": 589
+    },
+    {
+      "epoch": 0.18974874371859296,
+      "grad_norm": 0.34586772322654724,
+      "learning_rate": 0.00019809371039499296,
+      "loss": 0.5285,
+      "step": 590
+    },
+    {
+      "epoch": 0.19007035175879397,
+      "grad_norm": 0.40045374631881714,
+      "learning_rate": 0.0001980871523709985,
+      "loss": 0.6228,
+      "step": 591
+    },
+    {
+      "epoch": 0.19039195979899498,
+      "grad_norm": 0.3896762728691101,
+      "learning_rate": 0.00019808058319482644,
+      "loss": 0.6087,
+      "step": 592
+    },
+    {
+      "epoch": 0.190713567839196,
+      "grad_norm": 0.41355323791503906,
+      "learning_rate": 0.0001980740028672237,
+      "loss": 0.5575,
+      "step": 593
+    },
+    {
+      "epoch": 0.19103517587939697,
+      "grad_norm": 0.6239345073699951,
+      "learning_rate": 0.00019806741138893842,
+      "loss": 0.819,
+      "step": 594
+    },
+    {
+      "epoch": 0.19135678391959798,
+      "grad_norm": 0.4455684423446655,
+      "learning_rate": 0.00019806080876072003,
+      "loss": 0.6154,
+      "step": 595
+    },
+    {
+      "epoch": 0.191678391959799,
+      "grad_norm": 0.5026252865791321,
+      "learning_rate": 0.00019805419498331923,
+      "loss": 0.7243,
+      "step": 596
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.439420223236084,
+      "learning_rate": 0.00019804757005748797,
+      "loss": 0.5931,
+      "step": 597
+    },
+    {
+      "epoch": 0.19232160804020101,
+      "grad_norm": 0.5247384309768677,
+      "learning_rate": 0.0001980409339839795,
+      "loss": 0.844,
+      "step": 598
+    },
+    {
+      "epoch": 0.192643216080402,
+      "grad_norm": 0.5893823504447937,
+      "learning_rate": 0.00019803428676354833,
+      "loss": 0.9437,
+      "step": 599
+    },
+    {
+      "epoch": 0.192964824120603,
+      "grad_norm": 0.7636343836784363,
+      "learning_rate": 0.00019802762839695018,
+      "loss": 1.0162,
+      "step": 600
+    },
+    {
+      "epoch": 0.192964824120603,
+      "eval_loss": 0.730941891670227,
+      "eval_runtime": 69.3206,
+      "eval_samples_per_second": 3.606,
+      "eval_steps_per_second": 1.803,
+      "step": 600
+    },
+    {
+      "epoch": 0.19328643216080402,
+      "grad_norm": 0.2747584581375122,
+      "learning_rate": 0.00019802095888494213,
+      "loss": 0.6001,
+      "step": 601
+    },
+    {
+      "epoch": 0.19360804020100503,
+      "grad_norm": 0.27816104888916016,
+      "learning_rate": 0.00019801427822828242,
+      "loss": 0.4687,
+      "step": 602
+    },
+    {
+      "epoch": 0.19392964824120604,
+      "grad_norm": 0.3029923737049103,
+      "learning_rate": 0.0001980075864277307,
+      "loss": 0.6215,
+      "step": 603
+    },
+    {
+      "epoch": 0.19425125628140705,
+      "grad_norm": 0.332209974527359,
+      "learning_rate": 0.00019800088348404776,
+      "loss": 0.7154,
+      "step": 604
+    },
+    {
+      "epoch": 0.19457286432160803,
+      "grad_norm": 0.3040963411331177,
+      "learning_rate": 0.00019799416939799571,
+      "loss": 0.6492,
+      "step": 605
+    },
+    {
+      "epoch": 0.19489447236180904,
+      "grad_norm": 0.2812245488166809,
+      "learning_rate": 0.0001979874441703379,
+      "loss": 0.6286,
+      "step": 606
+    },
+    {
+      "epoch": 0.19521608040201005,
+      "grad_norm": 0.2704682946205139,
+      "learning_rate": 0.00019798070780183903,
+      "loss": 0.6843,
+      "step": 607
+    },
+    {
+      "epoch": 0.19553768844221106,
+      "grad_norm": 0.28553757071495056,
+      "learning_rate": 0.00019797396029326488,
+      "loss": 0.6978,
+      "step": 608
+    },
+    {
+      "epoch": 0.19585929648241207,
+      "grad_norm": 0.3061988353729248,
+      "learning_rate": 0.00019796720164538275,
+      "loss": 0.8749,
+      "step": 609
+    },
+    {
+      "epoch": 0.19618090452261305,
+      "grad_norm": 0.278898686170578,
+      "learning_rate": 0.000197960431858961,
+      "loss": 0.6523,
+      "step": 610
+    },
+    {
+      "epoch": 0.19650251256281406,
+      "grad_norm": 0.28180062770843506,
+      "learning_rate": 0.00019795365093476934,
+      "loss": 0.7035,
+      "step": 611
+    },
+    {
+      "epoch": 0.19682412060301507,
+      "grad_norm": 0.28678154945373535,
+      "learning_rate": 0.00019794685887357876,
+      "loss": 0.6689,
+      "step": 612
+    },
+    {
+      "epoch": 0.19714572864321608,
+      "grad_norm": 0.30030882358551025,
+      "learning_rate": 0.0001979400556761615,
+      "loss": 0.6335,
+      "step": 613
+    },
+    {
+      "epoch": 0.1974673366834171,
+      "grad_norm": 0.34381571412086487,
+      "learning_rate": 0.00019793324134329102,
+      "loss": 0.7895,
+      "step": 614
+    },
+    {
+      "epoch": 0.1977889447236181,
+      "grad_norm": 0.32943135499954224,
+      "learning_rate": 0.00019792641587574212,
+      "loss": 0.7107,
+      "step": 615
+    },
+    {
+      "epoch": 0.1981105527638191,
+      "grad_norm": 0.35190483927726746,
+      "learning_rate": 0.00019791957927429082,
+      "loss": 0.5575,
+      "step": 616
+    },
+    {
+      "epoch": 0.1984321608040201,
+      "grad_norm": 0.3087659776210785,
+      "learning_rate": 0.00019791273153971442,
+      "loss": 0.5614,
+      "step": 617
+    },
+    {
+      "epoch": 0.1987537688442211,
+      "grad_norm": 0.3321646749973297,
+      "learning_rate": 0.00019790587267279152,
+      "loss": 0.6411,
+      "step": 618
+    },
+    {
+      "epoch": 0.19907537688442212,
+      "grad_norm": 0.30465972423553467,
+      "learning_rate": 0.00019789900267430186,
+      "loss": 0.6202,
+      "step": 619
+    },
+    {
+      "epoch": 0.19939698492462313,
+      "grad_norm": 0.3003741204738617,
+      "learning_rate": 0.00019789212154502662,
+      "loss": 0.6478,
+      "step": 620
+    },
+    {
+      "epoch": 0.1997185929648241,
+      "grad_norm": 0.28635311126708984,
+      "learning_rate": 0.00019788522928574812,
+      "loss": 0.5527,
+      "step": 621
+    },
+    {
+      "epoch": 0.20004020100502512,
+      "grad_norm": 0.31752485036849976,
+      "learning_rate": 0.00019787832589725,
+      "loss": 0.6033,
+      "step": 622
+    },
+    {
+      "epoch": 0.20036180904522613,
+      "grad_norm": 0.347943514585495,
+      "learning_rate": 0.00019787141138031714,
+      "loss": 0.7225,
+      "step": 623
+    },
+    {
+      "epoch": 0.20068341708542714,
+      "grad_norm": 0.34638723731040955,
+      "learning_rate": 0.00019786448573573575,
+      "loss": 0.6746,
+      "step": 624
+    },
+    {
+      "epoch": 0.20100502512562815,
+      "grad_norm": 0.31864744424819946,
+      "learning_rate": 0.00019785754896429317,
+      "loss": 0.5765,
+      "step": 625
+    },
+    {
+      "epoch": 0.20132663316582913,
+      "grad_norm": 0.3359425663948059,
+      "learning_rate": 0.00019785060106677818,
+      "loss": 0.5529,
+      "step": 626
+    },
+    {
+      "epoch": 0.20164824120603014,
+      "grad_norm": 0.3207460343837738,
+      "learning_rate": 0.00019784364204398063,
+      "loss": 0.5835,
+      "step": 627
+    },
+    {
+      "epoch": 0.20196984924623115,
+      "grad_norm": 0.32770800590515137,
+      "learning_rate": 0.0001978366718966918,
+      "loss": 0.5927,
+      "step": 628
+    },
+    {
+      "epoch": 0.20229145728643216,
+      "grad_norm": 0.355775386095047,
+      "learning_rate": 0.00019782969062570417,
+      "loss": 0.6627,
+      "step": 629
+    },
+    {
+      "epoch": 0.20261306532663317,
+      "grad_norm": 0.38056692481040955,
+      "learning_rate": 0.00019782269823181148,
+      "loss": 0.6786,
+      "step": 630
+    },
+    {
+      "epoch": 0.20293467336683418,
+      "grad_norm": 0.37571579217910767,
+      "learning_rate": 0.0001978156947158087,
+      "loss": 0.7283,
+      "step": 631
+    },
+    {
+      "epoch": 0.20325628140703517,
+      "grad_norm": 0.37134334444999695,
+      "learning_rate": 0.0001978086800784922,
+      "loss": 0.7769,
+      "step": 632
+    },
+    {
+      "epoch": 0.20357788944723618,
+      "grad_norm": 0.36520472168922424,
+      "learning_rate": 0.00019780165432065942,
+      "loss": 0.7124,
+      "step": 633
+    },
+    {
+      "epoch": 0.2038994974874372,
+      "grad_norm": 0.36571255326271057,
+      "learning_rate": 0.00019779461744310922,
+      "loss": 0.6637,
+      "step": 634
+    },
+    {
+      "epoch": 0.2042211055276382,
+      "grad_norm": 0.3406449258327484,
+      "learning_rate": 0.00019778756944664166,
+      "loss": 0.6136,
+      "step": 635
+    },
+    {
+      "epoch": 0.2045427135678392,
+      "grad_norm": 0.37028196454048157,
+      "learning_rate": 0.0001977805103320581,
+      "loss": 0.6806,
+      "step": 636
+    },
+    {
+      "epoch": 0.2048643216080402,
+      "grad_norm": 0.3542538583278656,
+      "learning_rate": 0.00019777344010016109,
+      "loss": 0.5092,
+      "step": 637
+    },
+    {
+      "epoch": 0.2051859296482412,
+      "grad_norm": 0.34225964546203613,
+      "learning_rate": 0.0001977663587517545,
+      "loss": 0.4874,
+      "step": 638
+    },
+    {
+      "epoch": 0.2055075376884422,
+      "grad_norm": 0.4580456614494324,
+      "learning_rate": 0.00019775926628764348,
+      "loss": 0.7452,
+      "step": 639
+    },
+    {
+      "epoch": 0.20582914572864322,
+      "grad_norm": 0.4165748059749603,
+      "learning_rate": 0.00019775216270863442,
+      "loss": 0.7624,
+      "step": 640
+    },
+    {
+      "epoch": 0.20615075376884423,
+      "grad_norm": 0.38376113772392273,
+      "learning_rate": 0.00019774504801553492,
+      "loss": 0.544,
+      "step": 641
+    },
+    {
+      "epoch": 0.2064723618090452,
+      "grad_norm": 0.42880958318710327,
+      "learning_rate": 0.00019773792220915398,
+      "loss": 0.6986,
+      "step": 642
+    },
+    {
+      "epoch": 0.20679396984924622,
+      "grad_norm": 0.3905481696128845,
+      "learning_rate": 0.0001977307852903017,
+      "loss": 0.6267,
+      "step": 643
+    },
+    {
+      "epoch": 0.20711557788944723,
+      "grad_norm": 0.4827076196670532,
+      "learning_rate": 0.00019772363725978958,
+      "loss": 0.7443,
+      "step": 644
+    },
+    {
+      "epoch": 0.20743718592964824,
+      "grad_norm": 0.6492273211479187,
+      "learning_rate": 0.00019771647811843028,
+      "loss": 0.6731,
+      "step": 645
+    },
+    {
+      "epoch": 0.20775879396984925,
+      "grad_norm": 0.5485308170318604,
+      "learning_rate": 0.00019770930786703784,
+      "loss": 0.8492,
+      "step": 646
+    },
+    {
+      "epoch": 0.20808040201005026,
+      "grad_norm": 0.4789660573005676,
+      "learning_rate": 0.00019770212650642742,
+      "loss": 0.6435,
+      "step": 647
+    },
+    {
+      "epoch": 0.20840201005025125,
+      "grad_norm": 0.5435038208961487,
+      "learning_rate": 0.00019769493403741555,
+      "loss": 0.815,
+      "step": 648
+    },
+    {
+      "epoch": 0.20872361809045226,
+      "grad_norm": 0.5580148100852966,
+      "learning_rate": 0.00019768773046081997,
+      "loss": 0.8987,
+      "step": 649
+    },
+    {
+      "epoch": 0.20904522613065327,
+      "grad_norm": 0.6992361545562744,
+      "learning_rate": 0.00019768051577745974,
+      "loss": 1.0623,
+      "step": 650
+    },
+    {
+      "epoch": 0.20904522613065327,
+      "eval_loss": 0.73699551820755,
+      "eval_runtime": 69.3004,
+      "eval_samples_per_second": 3.607,
+      "eval_steps_per_second": 1.804,
+      "step": 650
+    },
+    {
+      "epoch": 0.20936683417085428,
+      "grad_norm": 0.2621065378189087,
+      "learning_rate": 0.00019767328998815514,
+      "loss": 0.3645,
+      "step": 651
+    },
+    {
+      "epoch": 0.2096884422110553,
+      "grad_norm": 0.3653203248977661,
+      "learning_rate": 0.00019766605309372764,
+      "loss": 0.5244,
+      "step": 652
+    },
+    {
+      "epoch": 0.21001005025125627,
+      "grad_norm": 0.34660616517066956,
+      "learning_rate": 0.00019765880509500016,
+      "loss": 0.5817,
+      "step": 653
+    },
+    {
+      "epoch": 0.21033165829145728,
+      "grad_norm": 0.3863109350204468,
+      "learning_rate": 0.0001976515459927967,
+      "loss": 0.6396,
+      "step": 654
+    },
+    {
+      "epoch": 0.2106532663316583,
+      "grad_norm": 0.2637709975242615,
+      "learning_rate": 0.00019764427578794263,
+      "loss": 0.5253,
+      "step": 655
+    },
+    {
+      "epoch": 0.2109748743718593,
+      "grad_norm": 0.3334469795227051,
+      "learning_rate": 0.00019763699448126457,
+      "loss": 0.6983,
+      "step": 656
+    },
+    {
+      "epoch": 0.2112964824120603,
+      "grad_norm": 0.29878470301628113,
+      "learning_rate": 0.00019762970207359032,
+      "loss": 0.7233,
+      "step": 657
+    },
+    {
+      "epoch": 0.2116180904522613,
+      "grad_norm": 0.30041465163230896,
+      "learning_rate": 0.000197622398565749,
+      "loss": 0.6604,
+      "step": 658
+    },
+    {
+      "epoch": 0.2119396984924623,
+      "grad_norm": 0.3030201196670532,
+      "learning_rate": 0.00019761508395857105,
+      "loss": 0.6303,
+      "step": 659
+    },
+    {
+      "epoch": 0.2122613065326633,
+      "grad_norm": 0.30520695447921753,
+      "learning_rate": 0.00019760775825288808,
+      "loss": 0.6191,
+      "step": 660
+    },
+    {
+      "epoch": 0.21258291457286432,
+      "grad_norm": 0.34923115372657776,
+      "learning_rate": 0.00019760042144953302,
+      "loss": 1.0316,
+      "step": 661
+    },
+    {
+      "epoch": 0.21290452261306533,
+      "grad_norm": 0.3282504975795746,
+      "learning_rate": 0.00019759307354934002,
+      "loss": 0.5188,
+      "step": 662
+    },
+    {
+      "epoch": 0.21322613065326634,
+      "grad_norm": 0.34390103816986084,
+      "learning_rate": 0.00019758571455314453,
+      "loss": 0.7182,
+      "step": 663
+    },
+    {
+      "epoch": 0.21354773869346733,
+      "grad_norm": 0.3625928461551666,
+      "learning_rate": 0.00019757834446178322,
+      "loss": 0.9392,
+      "step": 664
+    },
+    {
+      "epoch": 0.21386934673366834,
+      "grad_norm": 0.295958548784256,
+      "learning_rate": 0.00019757096327609406,
+      "loss": 0.6365,
+      "step": 665
+    },
+    {
+      "epoch": 0.21419095477386935,
+      "grad_norm": 0.32226479053497314,
+      "learning_rate": 0.00019756357099691625,
+      "loss": 0.7058,
+      "step": 666
+    },
+    {
+      "epoch": 0.21451256281407036,
+      "grad_norm": 0.3180067837238312,
+      "learning_rate": 0.00019755616762509029,
+      "loss": 0.6882,
+      "step": 667
+    },
+    {
+      "epoch": 0.21483417085427137,
+      "grad_norm": 0.3434496819972992,
+      "learning_rate": 0.0001975487531614579,
+      "loss": 0.6827,
+      "step": 668
+    },
+    {
+      "epoch": 0.21515577889447235,
+      "grad_norm": 0.3300999104976654,
+      "learning_rate": 0.0001975413276068621,
+      "loss": 0.777,
+      "step": 669
+    },
+    {
+      "epoch": 0.21547738693467336,
+      "grad_norm": 0.319518506526947,
+      "learning_rate": 0.00019753389096214714,
+      "loss": 0.5921,
+      "step": 670
+    },
+    {
+      "epoch": 0.21579899497487437,
+      "grad_norm": 0.34864890575408936,
+      "learning_rate": 0.00019752644322815853,
+      "loss": 0.7804,
+      "step": 671
+    },
+    {
+      "epoch": 0.21612060301507538,
+      "grad_norm": 0.33051055669784546,
+      "learning_rate": 0.00019751898440574305,
+      "loss": 0.7051,
+      "step": 672
+    },
+    {
+      "epoch": 0.2164422110552764,
+      "grad_norm": 0.3371402621269226,
+      "learning_rate": 0.00019751151449574875,
+      "loss": 0.7157,
+      "step": 673
+    },
+    {
+      "epoch": 0.2167638190954774,
+      "grad_norm": 0.3452366292476654,
+      "learning_rate": 0.00019750403349902496,
+      "loss": 0.6905,
+      "step": 674
+    },
+    {
+      "epoch": 0.21708542713567838,
+      "grad_norm": 0.3126762807369232,
+      "learning_rate": 0.00019749654141642221,
+      "loss": 0.556,
+      "step": 675
+    },
+    {
+      "epoch": 0.2174070351758794,
+      "grad_norm": 0.372152179479599,
+      "learning_rate": 0.00019748903824879234,
+      "loss": 0.6833,
+      "step": 676
+    },
+    {
+      "epoch": 0.2177286432160804,
+      "grad_norm": 0.3173413574695587,
+      "learning_rate": 0.0001974815239969884,
+      "loss": 0.5334,
+      "step": 677
+    },
+    {
+      "epoch": 0.2180502512562814,
+      "grad_norm": 0.38659507036209106,
+      "learning_rate": 0.00019747399866186483,
+      "loss": 0.8614,
+      "step": 678
+    },
+    {
+      "epoch": 0.21837185929648242,
+      "grad_norm": 0.3856159746646881,
+      "learning_rate": 0.00019746646224427713,
+      "loss": 0.7007,
+      "step": 679
+    },
+    {
+      "epoch": 0.2186934673366834,
+      "grad_norm": 0.37818577885627747,
+      "learning_rate": 0.0001974589147450822,
+      "loss": 0.8032,
+      "step": 680
+    },
+    {
+      "epoch": 0.21901507537688442,
+      "grad_norm": 0.31758254766464233,
+      "learning_rate": 0.00019745135616513817,
+      "loss": 0.5828,
+      "step": 681
+    },
+    {
+      "epoch": 0.21933668341708543,
+      "grad_norm": 0.3696713447570801,
+      "learning_rate": 0.00019744378650530443,
+      "loss": 0.6231,
+      "step": 682
+    },
+    {
+      "epoch": 0.21965829145728644,
+      "grad_norm": 0.36858007311820984,
+      "learning_rate": 0.00019743620576644162,
+      "loss": 0.586,
+      "step": 683
+    },
+    {
+      "epoch": 0.21997989949748745,
+      "grad_norm": 0.38380831480026245,
+      "learning_rate": 0.00019742861394941166,
+      "loss": 0.7247,
+      "step": 684
+    },
+    {
+      "epoch": 0.22030150753768843,
+      "grad_norm": 0.36575767397880554,
+      "learning_rate": 0.00019742101105507768,
+      "loss": 0.7121,
+      "step": 685
+    },
+    {
+      "epoch": 0.22062311557788944,
+      "grad_norm": 0.3324841856956482,
+      "learning_rate": 0.00019741339708430412,
+      "loss": 0.5908,
+      "step": 686
+    },
+    {
+      "epoch": 0.22094472361809045,
+      "grad_norm": 0.4107503890991211,
+      "learning_rate": 0.0001974057720379567,
+      "loss": 0.7281,
+      "step": 687
+    },
+    {
+      "epoch": 0.22126633165829146,
+      "grad_norm": 0.4330814778804779,
+      "learning_rate": 0.0001973981359169023,
+      "loss": 0.7701,
+      "step": 688
+    },
+    {
+      "epoch": 0.22158793969849247,
+      "grad_norm": 0.3725590407848358,
+      "learning_rate": 0.00019739048872200916,
+      "loss": 0.5899,
+      "step": 689
+    },
+    {
+      "epoch": 0.22190954773869348,
+      "grad_norm": 0.39335477352142334,
+      "learning_rate": 0.0001973828304541467,
+      "loss": 0.6796,
+      "step": 690
+    },
+    {
+      "epoch": 0.22223115577889446,
+      "grad_norm": 0.426072359085083,
+      "learning_rate": 0.0001973751611141857,
+      "loss": 0.7336,
+      "step": 691
+    },
+    {
+      "epoch": 0.22255276381909547,
+      "grad_norm": 0.4547775983810425,
+      "learning_rate": 0.0001973674807029981,
+      "loss": 0.7772,
+      "step": 692
+    },
+    {
+      "epoch": 0.22287437185929648,
+      "grad_norm": 0.40736913681030273,
+      "learning_rate": 0.00019735978922145715,
+      "loss": 0.6907,
+      "step": 693
+    },
+    {
+      "epoch": 0.2231959798994975,
+      "grad_norm": 0.4014805853366852,
+      "learning_rate": 0.0001973520866704373,
+      "loss": 0.6369,
+      "step": 694
+    },
+    {
+      "epoch": 0.2235175879396985,
+      "grad_norm": 0.5079379677772522,
+      "learning_rate": 0.0001973443730508144,
+      "loss": 0.9761,
+      "step": 695
+    },
+    {
+      "epoch": 0.22383919597989949,
+      "grad_norm": 0.4839482605457306,
+      "learning_rate": 0.00019733664836346534,
+      "loss": 0.7344,
+      "step": 696
+    },
+    {
+      "epoch": 0.2241608040201005,
+      "grad_norm": 0.4573533535003662,
+      "learning_rate": 0.0001973289126092685,
+      "loss": 0.6988,
+      "step": 697
+    },
+    {
+      "epoch": 0.2244824120603015,
+      "grad_norm": 0.5197833776473999,
+      "learning_rate": 0.00019732116578910334,
+      "loss": 0.6634,
+      "step": 698
+    },
+    {
+      "epoch": 0.22480402010050252,
+      "grad_norm": 0.5656108856201172,
+      "learning_rate": 0.00019731340790385068,
+      "loss": 0.8744,
+      "step": 699
+    },
+    {
+      "epoch": 0.22512562814070353,
+      "grad_norm": 0.6746645569801331,
+      "learning_rate": 0.0001973056389543926,
+      "loss": 0.8775,
+      "step": 700
+    },
+    {
+      "epoch": 0.22512562814070353,
+      "eval_loss": 0.7240564227104187,
+      "eval_runtime": 68.6913,
+      "eval_samples_per_second": 3.639,
+      "eval_steps_per_second": 1.82,
+      "step": 700
+    },
+    {
+      "epoch": 0.2254472361809045,
+      "grad_norm": 0.23933616280555725,
+      "learning_rate": 0.00019729785894161229,
+      "loss": 0.4413,
+      "step": 701
+    },
+    {
+      "epoch": 0.22576884422110552,
+      "grad_norm": 0.3225289285182953,
+      "learning_rate": 0.0001972900678663944,
+      "loss": 0.5962,
+      "step": 702
+    },
+    {
+      "epoch": 0.22609045226130653,
+      "grad_norm": 0.2967303395271301,
+      "learning_rate": 0.00019728226572962473,
+      "loss": 0.5765,
+      "step": 703
+    },
+    {
+      "epoch": 0.22641206030150754,
+      "grad_norm": 0.29661962389945984,
+      "learning_rate": 0.0001972744525321904,
+      "loss": 0.6207,
+      "step": 704
+    },
+    {
+      "epoch": 0.22673366834170855,
+      "grad_norm": 0.2974497079849243,
+      "learning_rate": 0.00019726662827497963,
+      "loss": 0.6578,
+      "step": 705
+    },
+    {
+      "epoch": 0.22705527638190956,
+      "grad_norm": 0.3111231029033661,
+      "learning_rate": 0.00019725879295888214,
+      "loss": 0.6738,
+      "step": 706
+    },
+    {
+      "epoch": 0.22737688442211054,
+      "grad_norm": 0.28051963448524475,
+      "learning_rate": 0.0001972509465847887,
+      "loss": 0.6053,
+      "step": 707
+    },
+    {
+      "epoch": 0.22769849246231155,
+      "grad_norm": 0.2954857051372528,
+      "learning_rate": 0.00019724308915359143,
+      "loss": 0.7797,
+      "step": 708
+    },
+    {
+      "epoch": 0.22802010050251256,
+      "grad_norm": 0.3252151310443878,
+      "learning_rate": 0.0001972352206661837,
+      "loss": 0.8844,
+      "step": 709
+    },
+    {
+      "epoch": 0.22834170854271357,
+      "grad_norm": 0.3027791678905487,
+      "learning_rate": 0.00019722734112346014,
+      "loss": 0.815,
+      "step": 710
+    },
+    {
+      "epoch": 0.22866331658291458,
+      "grad_norm": 0.35143449902534485,
+      "learning_rate": 0.00019721945052631664,
+      "loss": 0.7936,
+      "step": 711
+    },
+    {
+      "epoch": 0.22898492462311557,
+      "grad_norm": 0.34293296933174133,
+      "learning_rate": 0.00019721154887565029,
+      "loss": 0.9168,
+      "step": 712
+    },
+    {
+      "epoch": 0.22930653266331658,
+      "grad_norm": 0.3325900137424469,
+      "learning_rate": 0.00019720363617235952,
+      "loss": 0.8303,
+      "step": 713
+    },
+    {
+      "epoch": 0.22962814070351759,
+      "grad_norm": 0.34882882237434387,
+      "learning_rate": 0.00019719571241734395,
+      "loss": 0.6909,
+      "step": 714
+    },
+    {
+      "epoch": 0.2299497487437186,
+      "grad_norm": 0.3530502915382385,
+      "learning_rate": 0.0001971877776115045,
+      "loss": 0.8686,
+      "step": 715
+    },
+    {
+      "epoch": 0.2302713567839196,
+      "grad_norm": 0.3415836691856384,
+      "learning_rate": 0.00019717983175574332,
+      "loss": 0.7804,
+      "step": 716
+    },
+    {
+      "epoch": 0.23059296482412062,
+      "grad_norm": 0.31379079818725586,
+      "learning_rate": 0.0001971718748509638,
+      "loss": 0.7782,
+      "step": 717
+    },
+    {
+      "epoch": 0.2309145728643216,
+      "grad_norm": 0.36317524313926697,
+      "learning_rate": 0.0001971639068980707,
+      "loss": 0.7346,
+      "step": 718
+    },
+    {
+      "epoch": 0.2312361809045226,
+      "grad_norm": 0.31568434834480286,
+      "learning_rate": 0.00019715592789796985,
+      "loss": 0.5545,
+      "step": 719
+    },
+    {
+      "epoch": 0.23155778894472362,
+      "grad_norm": 0.3523799777030945,
+      "learning_rate": 0.00019714793785156852,
+      "loss": 0.6621,
+      "step": 720
+    },
+    {
+      "epoch": 0.23187939698492463,
+      "grad_norm": 0.3274880647659302,
+      "learning_rate": 0.0001971399367597751,
+      "loss": 0.5808,
+      "step": 721
+    },
+    {
+      "epoch": 0.23220100502512564,
+      "grad_norm": 0.3030402958393097,
+      "learning_rate": 0.0001971319246234993,
+      "loss": 0.522,
+      "step": 722
+    },
+    {
+      "epoch": 0.23252261306532662,
+      "grad_norm": 0.3760279715061188,
+      "learning_rate": 0.00019712390144365207,
+      "loss": 0.7773,
+      "step": 723
+    },
+    {
+      "epoch": 0.23284422110552763,
+      "grad_norm": 0.3315511643886566,
+      "learning_rate": 0.00019711586722114563,
+      "loss": 0.5261,
+      "step": 724
+    },
+    {
+      "epoch": 0.23316582914572864,
+      "grad_norm": 0.37824514508247375,
+      "learning_rate": 0.0001971078219568934,
+      "loss": 0.7277,
+      "step": 725
+    },
+    {
+      "epoch": 0.23348743718592965,
+      "grad_norm": 0.33694887161254883,
+      "learning_rate": 0.0001970997656518102,
+      "loss": 0.6192,
+      "step": 726
+    },
+    {
+      "epoch": 0.23380904522613066,
+      "grad_norm": 0.37134644389152527,
+      "learning_rate": 0.00019709169830681188,
+      "loss": 0.7626,
+      "step": 727
+    },
+    {
+      "epoch": 0.23413065326633165,
+      "grad_norm": 0.3716621398925781,
+      "learning_rate": 0.00019708361992281574,
+      "loss": 0.7607,
+      "step": 728
+    },
+    {
+      "epoch": 0.23445226130653266,
+      "grad_norm": 0.3269186019897461,
+      "learning_rate": 0.00019707553050074027,
+      "loss": 0.5907,
+      "step": 729
+    },
+    {
+      "epoch": 0.23477386934673367,
+      "grad_norm": 0.35407814383506775,
+      "learning_rate": 0.0001970674300415052,
+      "loss": 0.6102,
+      "step": 730
+    },
+    {
+      "epoch": 0.23509547738693468,
+      "grad_norm": 0.3680501878261566,
+      "learning_rate": 0.00019705931854603148,
+      "loss": 0.768,
+      "step": 731
+    },
+    {
+      "epoch": 0.23541708542713569,
+      "grad_norm": 0.3638216257095337,
+      "learning_rate": 0.00019705119601524143,
+      "loss": 0.6788,
+      "step": 732
+    },
+    {
+      "epoch": 0.2357386934673367,
+      "grad_norm": 0.3657969534397125,
+      "learning_rate": 0.00019704306245005852,
+      "loss": 0.6614,
+      "step": 733
+    },
+    {
+      "epoch": 0.23606030150753768,
+      "grad_norm": 0.3710695505142212,
+      "learning_rate": 0.0001970349178514075,
+      "loss": 0.6015,
+      "step": 734
+    },
+    {
+      "epoch": 0.2363819095477387,
+      "grad_norm": 0.35924914479255676,
+      "learning_rate": 0.0001970267622202144,
+      "loss": 0.6426,
+      "step": 735
+    },
+    {
+      "epoch": 0.2367035175879397,
+      "grad_norm": 0.40453553199768066,
+      "learning_rate": 0.00019701859555740648,
+      "loss": 0.6819,
+      "step": 736
+    },
+    {
+      "epoch": 0.2370251256281407,
+      "grad_norm": 0.3846459686756134,
+      "learning_rate": 0.00019701041786391224,
+      "loss": 0.6807,
+      "step": 737
+    },
+    {
+      "epoch": 0.23734673366834172,
+      "grad_norm": 0.38189271092414856,
+      "learning_rate": 0.00019700222914066152,
+      "loss": 0.5759,
+      "step": 738
+    },
+    {
+      "epoch": 0.2376683417085427,
+      "grad_norm": 0.41226956248283386,
+      "learning_rate": 0.0001969940293885853,
+      "loss": 0.7136,
+      "step": 739
+    },
+    {
+      "epoch": 0.2379899497487437,
+      "grad_norm": 0.4381696879863739,
+      "learning_rate": 0.00019698581860861584,
+      "loss": 0.6793,
+      "step": 740
+    },
+    {
+      "epoch": 0.23831155778894472,
+      "grad_norm": 0.4089527130126953,
+      "learning_rate": 0.00019697759680168675,
+      "loss": 0.7081,
+      "step": 741
+    },
+    {
+      "epoch": 0.23863316582914573,
+      "grad_norm": 0.3800293207168579,
+      "learning_rate": 0.00019696936396873275,
+      "loss": 0.5417,
+      "step": 742
+    },
+    {
+      "epoch": 0.23895477386934674,
+      "grad_norm": 0.4303738474845886,
+      "learning_rate": 0.00019696112011068998,
+      "loss": 0.6672,
+      "step": 743
+    },
+    {
+      "epoch": 0.23927638190954773,
+      "grad_norm": 0.4650689959526062,
+      "learning_rate": 0.00019695286522849561,
+      "loss": 0.7111,
+      "step": 744
+    },
+    {
+      "epoch": 0.23959798994974874,
+      "grad_norm": 0.48023107647895813,
+      "learning_rate": 0.00019694459932308829,
+      "loss": 0.7158,
+      "step": 745
+    },
+    {
+      "epoch": 0.23991959798994975,
+      "grad_norm": 0.4477865397930145,
+      "learning_rate": 0.0001969363223954078,
+      "loss": 0.4948,
+      "step": 746
+    },
+    {
+      "epoch": 0.24024120603015076,
+      "grad_norm": 0.6077302098274231,
+      "learning_rate": 0.0001969280344463952,
+      "loss": 0.9597,
+      "step": 747
+    },
+    {
+      "epoch": 0.24056281407035177,
+      "grad_norm": 0.5349471569061279,
+      "learning_rate": 0.00019691973547699277,
+      "loss": 0.826,
+      "step": 748
+    },
+    {
+      "epoch": 0.24088442211055278,
+      "grad_norm": 0.6425649523735046,
+      "learning_rate": 0.00019691142548814412,
+      "loss": 1.0061,
+      "step": 749
+    },
+    {
+      "epoch": 0.24120603015075376,
+      "grad_norm": 0.779952883720398,
+      "learning_rate": 0.00019690310448079403,
+      "loss": 1.0963,
+      "step": 750
+    },
+    {
+      "epoch": 0.24120603015075376,
+      "eval_loss": 0.7127780914306641,
+      "eval_runtime": 68.9871,
+      "eval_samples_per_second": 3.624,
+      "eval_steps_per_second": 1.812,
+      "step": 750
+    },
+    {
+      "epoch": 0.24152763819095477,
+      "grad_norm": 0.2459581047296524,
+      "learning_rate": 0.00019689477245588862,
+      "loss": 0.4914,
+      "step": 751
+    },
+    {
+      "epoch": 0.24184924623115578,
+      "grad_norm": 0.32978856563568115,
+      "learning_rate": 0.00019688642941437516,
+      "loss": 0.6168,
+      "step": 752
+    },
+    {
+      "epoch": 0.2421708542713568,
+      "grad_norm": 0.28750288486480713,
+      "learning_rate": 0.00019687807535720224,
+      "loss": 0.6053,
+      "step": 753
+    },
+    {
+      "epoch": 0.2424924623115578,
+      "grad_norm": 0.3740372955799103,
+      "learning_rate": 0.00019686971028531973,
+      "loss": 0.8734,
+      "step": 754
+    },
+    {
+      "epoch": 0.24281407035175878,
+      "grad_norm": 0.2906869351863861,
+      "learning_rate": 0.00019686133419967864,
+      "loss": 0.6702,
+      "step": 755
+    },
+    {
+      "epoch": 0.2431356783919598,
+      "grad_norm": 0.280679851770401,
+      "learning_rate": 0.00019685294710123134,
+      "loss": 0.6146,
+      "step": 756
+    },
+    {
+      "epoch": 0.2434572864321608,
+      "grad_norm": 0.25428593158721924,
+      "learning_rate": 0.00019684454899093146,
+      "loss": 0.4541,
+      "step": 757
+    },
+    {
+      "epoch": 0.2437788944723618,
+      "grad_norm": 0.2923628091812134,
+      "learning_rate": 0.00019683613986973373,
+      "loss": 0.7239,
+      "step": 758
+    },
+    {
+      "epoch": 0.24410050251256282,
+      "grad_norm": 0.32716384530067444,
+      "learning_rate": 0.00019682771973859433,
+      "loss": 0.7477,
+      "step": 759
+    },
+    {
+      "epoch": 0.24442211055276383,
+      "grad_norm": 0.32640540599823,
+      "learning_rate": 0.00019681928859847057,
+      "loss": 0.6733,
+      "step": 760
+    },
+    {
+      "epoch": 0.24474371859296482,
+      "grad_norm": 0.3050812780857086,
+      "learning_rate": 0.00019681084645032103,
+      "loss": 0.6756,
+      "step": 761
+    },
+    {
+      "epoch": 0.24506532663316583,
+      "grad_norm": 0.31835633516311646,
+      "learning_rate": 0.00019680239329510557,
+      "loss": 0.7471,
+      "step": 762
+    },
+    {
+      "epoch": 0.24538693467336684,
+      "grad_norm": 0.3159918189048767,
+      "learning_rate": 0.0001967939291337853,
+      "loss": 0.6487,
+      "step": 763
+    },
+    {
+      "epoch": 0.24570854271356785,
+      "grad_norm": 0.3212244510650635,
+      "learning_rate": 0.00019678545396732252,
+      "loss": 0.6725,
+      "step": 764
+    },
+    {
+      "epoch": 0.24603015075376886,
+      "grad_norm": 0.34220248460769653,
+      "learning_rate": 0.00019677696779668084,
+      "loss": 0.7764,
+      "step": 765
+    },
+    {
+      "epoch": 0.24635175879396984,
+      "grad_norm": 0.3299233019351959,
+      "learning_rate": 0.00019676847062282516,
+      "loss": 0.7397,
+      "step": 766
+    },
+    {
+      "epoch": 0.24667336683417085,
+      "grad_norm": 0.2972373962402344,
+      "learning_rate": 0.0001967599624467215,
+      "loss": 0.589,
+      "step": 767
+    },
+    {
+      "epoch": 0.24699497487437186,
+      "grad_norm": 0.32379043102264404,
+      "learning_rate": 0.0001967514432693373,
+      "loss": 0.6589,
+      "step": 768
+    },
+    {
+      "epoch": 0.24731658291457287,
+      "grad_norm": 0.3253432512283325,
+      "learning_rate": 0.00019674291309164108,
+      "loss": 0.6962,
+      "step": 769
+    },
+    {
+      "epoch": 0.24763819095477388,
+      "grad_norm": 0.2925957441329956,
+      "learning_rate": 0.00019673437191460274,
+      "loss": 0.512,
+      "step": 770
+    },
+    {
+      "epoch": 0.24795979899497486,
+      "grad_norm": 0.32865580916404724,
+      "learning_rate": 0.0001967258197391934,
+      "loss": 0.6461,
+      "step": 771
+    },
+    {
+      "epoch": 0.24828140703517587,
+      "grad_norm": 0.36776721477508545,
+      "learning_rate": 0.00019671725656638536,
+      "loss": 0.748,
+      "step": 772
+    },
+    {
+      "epoch": 0.24860301507537688,
+      "grad_norm": 0.32894015312194824,
+      "learning_rate": 0.00019670868239715224,
+      "loss": 0.5128,
+      "step": 773
+    },
+    {
+      "epoch": 0.2489246231155779,
+      "grad_norm": 0.3957439363002777,
+      "learning_rate": 0.0001967000972324689,
+      "loss": 0.78,
+      "step": 774
+    },
+    {
+      "epoch": 0.2492462311557789,
+      "grad_norm": 0.3211399018764496,
+      "learning_rate": 0.00019669150107331146,
+      "loss": 0.6415,
+      "step": 775
+    },
+    {
+      "epoch": 0.2495678391959799,
+      "grad_norm": 0.3474048972129822,
+      "learning_rate": 0.00019668289392065728,
+      "loss": 0.6126,
+      "step": 776
+    },
+    {
+      "epoch": 0.2498894472361809,
+      "grad_norm": 0.33722245693206787,
+      "learning_rate": 0.00019667427577548492,
+      "loss": 0.5279,
+      "step": 777
+    },
+    {
+      "epoch": 0.25021105527638193,
+      "grad_norm": 0.36115992069244385,
+      "learning_rate": 0.00019666564663877427,
+      "loss": 0.6245,
+      "step": 778
+    },
+    {
+      "epoch": 0.2505326633165829,
+      "grad_norm": 0.39823293685913086,
+      "learning_rate": 0.0001966570065115064,
+      "loss": 0.8214,
+      "step": 779
+    },
+    {
+      "epoch": 0.2508542713567839,
+      "grad_norm": 0.3755451738834381,
+      "learning_rate": 0.0001966483553946637,
+      "loss": 0.5899,
+      "step": 780
+    },
+    {
+      "epoch": 0.25117587939698494,
+      "grad_norm": 0.3944664001464844,
+      "learning_rate": 0.0001966396932892298,
+      "loss": 0.6685,
+      "step": 781
+    },
+    {
+      "epoch": 0.2514974874371859,
+      "grad_norm": 0.3741309940814972,
+      "learning_rate": 0.00019663102019618943,
+      "loss": 0.6372,
+      "step": 782
+    },
+    {
+      "epoch": 0.25181909547738696,
+      "grad_norm": 0.42738616466522217,
+      "learning_rate": 0.00019662233611652882,
+      "loss": 0.6339,
+      "step": 783
+    },
+    {
+      "epoch": 0.25214070351758794,
+      "grad_norm": 0.42854607105255127,
+      "learning_rate": 0.00019661364105123528,
+      "loss": 0.7273,
+      "step": 784
+    },
+    {
+      "epoch": 0.2524623115577889,
+      "grad_norm": 0.3921884596347809,
+      "learning_rate": 0.00019660493500129738,
+      "loss": 0.6284,
+      "step": 785
+    },
+    {
+      "epoch": 0.25278391959798996,
+      "grad_norm": 0.3690395653247833,
+      "learning_rate": 0.000196596217967705,
+      "loss": 0.597,
+      "step": 786
+    },
+    {
+      "epoch": 0.25310552763819094,
+      "grad_norm": 0.38130655884742737,
+      "learning_rate": 0.00019658748995144922,
+      "loss": 0.6345,
+      "step": 787
+    },
+    {
+      "epoch": 0.253427135678392,
+      "grad_norm": 0.36159974336624146,
+      "learning_rate": 0.00019657875095352242,
+      "loss": 0.501,
+      "step": 788
+    },
+    {
+      "epoch": 0.25374874371859296,
+      "grad_norm": 0.4033864736557007,
+      "learning_rate": 0.00019657000097491811,
+      "loss": 0.6673,
+      "step": 789
+    },
+    {
+      "epoch": 0.25407035175879394,
+      "grad_norm": 0.36105579137802124,
+      "learning_rate": 0.00019656124001663123,
+      "loss": 0.5244,
+      "step": 790
+    },
+    {
+      "epoch": 0.254391959798995,
+      "grad_norm": 0.5150249600410461,
+      "learning_rate": 0.00019655246807965784,
+      "loss": 0.7443,
+      "step": 791
+    },
+    {
+      "epoch": 0.25471356783919596,
+      "grad_norm": 0.4228598177433014,
+      "learning_rate": 0.00019654368516499527,
+      "loss": 0.6324,
+      "step": 792
+    },
+    {
+      "epoch": 0.255035175879397,
+      "grad_norm": 0.44264519214630127,
+      "learning_rate": 0.00019653489127364207,
+      "loss": 0.648,
+      "step": 793
+    },
+    {
+      "epoch": 0.255356783919598,
+      "grad_norm": 0.5047042369842529,
+      "learning_rate": 0.00019652608640659817,
+      "loss": 0.6736,
+      "step": 794
+    },
+    {
+      "epoch": 0.25567839195979897,
+      "grad_norm": 1.237250804901123,
+      "learning_rate": 0.00019651727056486457,
+      "loss": 0.6045,
+      "step": 795
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.49821993708610535,
+      "learning_rate": 0.00019650844374944363,
+      "loss": 0.7299,
+      "step": 796
+    },
+    {
+      "epoch": 0.256321608040201,
+      "grad_norm": 0.5537844300270081,
+      "learning_rate": 0.00019649960596133896,
+      "loss": 0.9434,
+      "step": 797
+    },
+    {
+      "epoch": 0.256643216080402,
+      "grad_norm": 0.6757361888885498,
+      "learning_rate": 0.00019649075720155535,
+      "loss": 0.9217,
+      "step": 798
+    },
+    {
+      "epoch": 0.256964824120603,
+      "grad_norm": 0.6626216769218445,
+      "learning_rate": 0.00019648189747109885,
+      "loss": 0.9162,
+      "step": 799
+    },
+    {
+      "epoch": 0.25728643216080405,
+      "grad_norm": 0.7006928324699402,
+      "learning_rate": 0.0001964730267709768,
+      "loss": 1.0963,
+      "step": 800
+    },
+    {
+      "epoch": 0.25728643216080405,
+      "eval_loss": 0.7180275321006775,
+      "eval_runtime": 69.4382,
+      "eval_samples_per_second": 3.6,
+      "eval_steps_per_second": 1.8,
+      "step": 800
+    },
+    {
+      "epoch": 0.25760804020100503,
+      "grad_norm": 0.2432035207748413,
+      "learning_rate": 0.00019646414510219784,
+      "loss": 0.4394,
+      "step": 801
+    },
+    {
+      "epoch": 0.257929648241206,
+      "grad_norm": 0.297922283411026,
+      "learning_rate": 0.00019645525246577168,
+      "loss": 0.4748,
+      "step": 802
+    },
+    {
+      "epoch": 0.25825125628140705,
+      "grad_norm": 0.30467143654823303,
+      "learning_rate": 0.00019644634886270944,
+      "loss": 0.6198,
+      "step": 803
+    },
+    {
+      "epoch": 0.25857286432160803,
+      "grad_norm": 0.31693345308303833,
+      "learning_rate": 0.00019643743429402342,
+      "loss": 0.6611,
+      "step": 804
+    },
+    {
+      "epoch": 0.25889447236180907,
+      "grad_norm": 0.29568272829055786,
+      "learning_rate": 0.00019642850876072716,
+      "loss": 0.5466,
+      "step": 805
+    },
+    {
+      "epoch": 0.25921608040201005,
+      "grad_norm": 0.2799466848373413,
+      "learning_rate": 0.00019641957226383553,
+      "loss": 0.594,
+      "step": 806
+    },
+    {
+      "epoch": 0.25953768844221103,
+      "grad_norm": 0.32903438806533813,
+      "learning_rate": 0.00019641062480436444,
+      "loss": 0.7572,
+      "step": 807
+    },
+    {
+      "epoch": 0.25985929648241207,
+      "grad_norm": 0.2904883921146393,
+      "learning_rate": 0.00019640166638333135,
+      "loss": 0.6412,
+      "step": 808
+    },
+    {
+      "epoch": 0.26018090452261305,
+      "grad_norm": 0.28040972352027893,
+      "learning_rate": 0.0001963926970017547,
+      "loss": 0.5891,
+      "step": 809
+    },
+    {
+      "epoch": 0.2605025125628141,
+      "grad_norm": 0.27400726079940796,
+      "learning_rate": 0.0001963837166606543,
+      "loss": 0.5316,
+      "step": 810
+    },
+    {
+      "epoch": 0.2608241206030151,
+      "grad_norm": 0.3047313392162323,
+      "learning_rate": 0.0001963747253610512,
+      "loss": 0.5951,
+      "step": 811
+    },
+    {
+      "epoch": 0.26114572864321606,
+      "grad_norm": 0.34562045335769653,
+      "learning_rate": 0.00019636572310396765,
+      "loss": 0.869,
+      "step": 812
+    },
+    {
+      "epoch": 0.2614673366834171,
+      "grad_norm": 0.3248054087162018,
+      "learning_rate": 0.0001963567098904272,
+      "loss": 0.6431,
+      "step": 813
+    },
+    {
+      "epoch": 0.2617889447236181,
+      "grad_norm": 0.3427451550960541,
+      "learning_rate": 0.00019634768572145462,
+      "loss": 0.8292,
+      "step": 814
+    },
+    {
+      "epoch": 0.2621105527638191,
+      "grad_norm": 0.34792545437812805,
+      "learning_rate": 0.0001963386505980759,
+      "loss": 0.8332,
+      "step": 815
+    },
+    {
+      "epoch": 0.2624321608040201,
+      "grad_norm": 0.3241157829761505,
+      "learning_rate": 0.00019632960452131836,
+      "loss": 0.6129,
+      "step": 816
+    },
+    {
+      "epoch": 0.2627537688442211,
+      "grad_norm": 0.3326020836830139,
+      "learning_rate": 0.00019632054749221047,
+      "loss": 0.8611,
+      "step": 817
+    },
+    {
+      "epoch": 0.2630753768844221,
+      "grad_norm": 0.31111210584640503,
+      "learning_rate": 0.00019631147951178196,
+      "loss": 0.5626,
+      "step": 818
+    },
+    {
+      "epoch": 0.2633969849246231,
+      "grad_norm": 0.39161941409111023,
+      "learning_rate": 0.00019630240058106388,
+      "loss": 0.8717,
+      "step": 819
+    },
+    {
+      "epoch": 0.26371859296482414,
+      "grad_norm": 0.31833720207214355,
+      "learning_rate": 0.00019629331070108847,
+      "loss": 0.6047,
+      "step": 820
+    },
+    {
+      "epoch": 0.2640402010050251,
+      "grad_norm": 0.35861510038375854,
+      "learning_rate": 0.00019628420987288916,
+      "loss": 0.7178,
+      "step": 821
+    },
+    {
+      "epoch": 0.2643618090452261,
+      "grad_norm": 0.2984914779663086,
+      "learning_rate": 0.00019627509809750076,
+      "loss": 0.6054,
+      "step": 822
+    },
+    {
+      "epoch": 0.26468341708542714,
+      "grad_norm": 0.3404580056667328,
+      "learning_rate": 0.0001962659753759592,
+      "loss": 0.7709,
+      "step": 823
+    },
+    {
+      "epoch": 0.2650050251256281,
+      "grad_norm": 0.34336867928504944,
+      "learning_rate": 0.0001962568417093017,
+      "loss": 0.6115,
+      "step": 824
+    },
+    {
+      "epoch": 0.26532663316582916,
+      "grad_norm": 0.41466858983039856,
+      "learning_rate": 0.00019624769709856673,
+      "loss": 0.8954,
+      "step": 825
+    },
+    {
+      "epoch": 0.26564824120603014,
+      "grad_norm": 0.3718617558479309,
+      "learning_rate": 0.00019623854154479403,
+      "loss": 0.6685,
+      "step": 826
+    },
+    {
+      "epoch": 0.2659698492462312,
+      "grad_norm": 0.3724638819694519,
+      "learning_rate": 0.00019622937504902455,
+      "loss": 0.6604,
+      "step": 827
+    },
+    {
+      "epoch": 0.26629145728643216,
+      "grad_norm": 0.3711127042770386,
+      "learning_rate": 0.00019622019761230045,
+      "loss": 0.6845,
+      "step": 828
+    },
+    {
+      "epoch": 0.26661306532663315,
+      "grad_norm": 0.37369856238365173,
+      "learning_rate": 0.00019621100923566522,
+      "loss": 0.775,
+      "step": 829
+    },
+    {
+      "epoch": 0.2669346733668342,
+      "grad_norm": 0.35682374238967896,
+      "learning_rate": 0.00019620180992016356,
+      "loss": 0.4879,
+      "step": 830
+    },
+    {
+      "epoch": 0.26725628140703517,
+      "grad_norm": 0.4311290979385376,
+      "learning_rate": 0.0001961925996668413,
+      "loss": 0.8225,
+      "step": 831
+    },
+    {
+      "epoch": 0.2675778894472362,
+      "grad_norm": 0.3813476860523224,
+      "learning_rate": 0.00019618337847674571,
+      "loss": 0.6876,
+      "step": 832
+    },
+    {
+      "epoch": 0.2678994974874372,
+      "grad_norm": 0.33009132742881775,
+      "learning_rate": 0.00019617414635092524,
+      "loss": 0.5987,
+      "step": 833
+    },
+    {
+      "epoch": 0.26822110552763817,
+      "grad_norm": 0.3474346995353699,
+      "learning_rate": 0.00019616490329042943,
+      "loss": 0.5411,
+      "step": 834
+    },
+    {
+      "epoch": 0.2685427135678392,
+      "grad_norm": 0.4289398491382599,
+      "learning_rate": 0.00019615564929630926,
+      "loss": 0.7756,
+      "step": 835
+    },
+    {
+      "epoch": 0.2688643216080402,
+      "grad_norm": 0.3504607081413269,
+      "learning_rate": 0.00019614638436961688,
+      "loss": 0.6187,
+      "step": 836
+    },
+    {
+      "epoch": 0.26918592964824123,
+      "grad_norm": 0.3657979965209961,
+      "learning_rate": 0.00019613710851140565,
+      "loss": 0.661,
+      "step": 837
+    },
+    {
+      "epoch": 0.2695075376884422,
+      "grad_norm": 0.3971709609031677,
+      "learning_rate": 0.0001961278217227302,
+      "loss": 0.7287,
+      "step": 838
+    },
+    {
+      "epoch": 0.2698291457286432,
+      "grad_norm": 0.4091176688671112,
+      "learning_rate": 0.00019611852400464649,
+      "loss": 0.7261,
+      "step": 839
+    },
+    {
+      "epoch": 0.27015075376884423,
+      "grad_norm": 0.35221415758132935,
+      "learning_rate": 0.00019610921535821154,
+      "loss": 0.4774,
+      "step": 840
+    },
+    {
+      "epoch": 0.2704723618090452,
+      "grad_norm": 0.4134233593940735,
+      "learning_rate": 0.00019609989578448373,
+      "loss": 0.5539,
+      "step": 841
+    },
+    {
+      "epoch": 0.27079396984924625,
+      "grad_norm": 0.4528118371963501,
+      "learning_rate": 0.0001960905652845227,
+      "loss": 0.6422,
+      "step": 842
+    },
+    {
+      "epoch": 0.27111557788944723,
+      "grad_norm": 0.4416557252407074,
+      "learning_rate": 0.0001960812238593893,
+      "loss": 0.6141,
+      "step": 843
+    },
+    {
+      "epoch": 0.2714371859296482,
+      "grad_norm": 0.43463316559791565,
+      "learning_rate": 0.00019607187151014554,
+      "loss": 0.6315,
+      "step": 844
+    },
+    {
+      "epoch": 0.27175879396984925,
+      "grad_norm": 0.4529591202735901,
+      "learning_rate": 0.00019606250823785487,
+      "loss": 0.7021,
+      "step": 845
+    },
+    {
+      "epoch": 0.27208040201005024,
+      "grad_norm": 0.4524073600769043,
+      "learning_rate": 0.00019605313404358175,
+      "loss": 0.6384,
+      "step": 846
+    },
+    {
+      "epoch": 0.2724020100502513,
+      "grad_norm": 0.5789950489997864,
+      "learning_rate": 0.0001960437489283921,
+      "loss": 0.8001,
+      "step": 847
+    },
+    {
+      "epoch": 0.27272361809045226,
+      "grad_norm": 0.5944107174873352,
+      "learning_rate": 0.00019603435289335285,
+      "loss": 0.8846,
+      "step": 848
+    },
+    {
+      "epoch": 0.27304522613065324,
+      "grad_norm": 0.7075732946395874,
+      "learning_rate": 0.00019602494593953243,
+      "loss": 0.9798,
+      "step": 849
+    },
+    {
+      "epoch": 0.2733668341708543,
+      "grad_norm": 0.8297386169433594,
+      "learning_rate": 0.00019601552806800028,
+      "loss": 1.0461,
+      "step": 850
+    },
+    {
+      "epoch": 0.2733668341708543,
+      "eval_loss": 0.713762104511261,
+      "eval_runtime": 69.2864,
+      "eval_samples_per_second": 3.608,
+      "eval_steps_per_second": 1.804,
+      "step": 850
     }
   ],
   "logging_steps": 1,
@@ -387,7 +6115,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 0
+        "early_stopping_patience_counter": 1
       }
     },
     "TrainerControl": {
@@ -401,7 +6129,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.8005052044017664e+16,
+  "total_flos": 4.7327686737002496e+17,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null