diff --git "a/checkpoint-1500/trainer_state.json" "b/checkpoint-1500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1500/trainer_state.json"
@@ -0,0 +1,10533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.5004219936857242,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001000281329123816,
+      "grad_norm": 5.35972785949707,
+      "learning_rate": 0.0001999999450590425,
+      "loss": 2.2302,
+      "step": 1
+    },
+    {
+      "epoch": 0.002000562658247632,
+      "grad_norm": 2.222695827484131,
+      "learning_rate": 0.00019999978023623033,
+      "loss": 1.5115,
+      "step": 2
+    },
+    {
+      "epoch": 0.003000843987371448,
+      "grad_norm": 2.513378858566284,
+      "learning_rate": 0.0001999995055317446,
+      "loss": 1.3975,
+      "step": 3
+    },
+    {
+      "epoch": 0.004001125316495264,
+      "grad_norm": 1.3351486921310425,
+      "learning_rate": 0.00019999912094588717,
+      "loss": 1.1971,
+      "step": 4
+    },
+    {
+      "epoch": 0.005001406645619081,
+      "grad_norm": 2.562424659729004,
+      "learning_rate": 0.00019999862647908064,
+      "loss": 1.7113,
+      "step": 5
+    },
+    {
+      "epoch": 0.006001687974742896,
+      "grad_norm": 2.8064351081848145,
+      "learning_rate": 0.00019999802213186834,
+      "loss": 1.9351,
+      "step": 6
+    },
+    {
+      "epoch": 0.007001969303866712,
+      "grad_norm": 1.640201449394226,
+      "learning_rate": 0.0001999973079049143,
+      "loss": 1.1837,
+      "step": 7
+    },
+    {
+      "epoch": 0.008002250632990529,
+      "grad_norm": 0.86270672082901,
+      "learning_rate": 0.00019999648379900338,
+      "loss": 0.9277,
+      "step": 8
+    },
+    {
+      "epoch": 0.009002531962114344,
+      "grad_norm": 0.9115725755691528,
+      "learning_rate": 0.0001999955498150411,
+      "loss": 1.0205,
+      "step": 9
+    },
+    {
+      "epoch": 0.010002813291238161,
+      "grad_norm": 0.8274697065353394,
+      "learning_rate": 0.00019999450595405374,
+      "loss": 0.959,
+      "step": 10
+    },
+    {
+      "epoch": 0.011003094620361977,
+      "grad_norm": 0.8405666947364807,
+      "learning_rate": 0.0001999933522171883,
+      "loss": 0.7265,
+      "step": 11
+    },
+    {
+      "epoch": 0.012003375949485792,
+      "grad_norm": 0.6411307454109192,
+      "learning_rate": 0.00019999208860571255,
+      "loss": 0.7987,
+      "step": 12
+    },
+    {
+      "epoch": 0.01300365727860961,
+      "grad_norm": 0.5139630436897278,
+      "learning_rate": 0.00019999071512101496,
+      "loss": 0.7672,
+      "step": 13
+    },
+    {
+      "epoch": 0.014003938607733425,
+      "grad_norm": 0.5102379322052002,
+      "learning_rate": 0.00019998923176460474,
+      "loss": 0.7856,
+      "step": 14
+    },
+    {
+      "epoch": 0.01500421993685724,
+      "grad_norm": 0.41639742255210876,
+      "learning_rate": 0.00019998763853811184,
+      "loss": 0.6591,
+      "step": 15
+    },
+    {
+      "epoch": 0.016004501265981057,
+      "grad_norm": 0.5631232857704163,
+      "learning_rate": 0.00019998593544328692,
+      "loss": 0.6823,
+      "step": 16
+    },
+    {
+      "epoch": 0.017004782595104875,
+      "grad_norm": 0.601658046245575,
+      "learning_rate": 0.00019998412248200138,
+      "loss": 0.8059,
+      "step": 17
+    },
+    {
+      "epoch": 0.01800506392422869,
+      "grad_norm": 0.5272611379623413,
+      "learning_rate": 0.00019998219965624734,
+      "loss": 0.7871,
+      "step": 18
+    },
+    {
+      "epoch": 0.019005345253352506,
+      "grad_norm": 0.4801584780216217,
+      "learning_rate": 0.0001999801669681376,
+      "loss": 0.69,
+      "step": 19
+    },
+    {
+      "epoch": 0.020005626582476323,
+      "grad_norm": 0.5749354362487793,
+      "learning_rate": 0.00019997802441990573,
+      "loss": 0.669,
+      "step": 20
+    },
+    {
+      "epoch": 0.021005907911600136,
+      "grad_norm": 0.5993685126304626,
+      "learning_rate": 0.00019997577201390606,
+      "loss": 0.7249,
+      "step": 21
+    },
+    {
+      "epoch": 0.022006189240723954,
+      "grad_norm": 0.5108559727668762,
+      "learning_rate": 0.00019997340975261353,
+      "loss": 0.6593,
+      "step": 22
+    },
+    {
+      "epoch": 0.02300647056984777,
+      "grad_norm": 0.5030386447906494,
+      "learning_rate": 0.00019997093763862383,
+      "loss": 0.6637,
+      "step": 23
+    },
+    {
+      "epoch": 0.024006751898971584,
+      "grad_norm": 0.4953747093677521,
+      "learning_rate": 0.0001999683556746534,
+      "loss": 0.6865,
+      "step": 24
+    },
+    {
+      "epoch": 0.0250070332280954,
+      "grad_norm": 0.44772374629974365,
+      "learning_rate": 0.0001999656638635393,
+      "loss": 0.6848,
+      "step": 25
+    },
+    {
+      "epoch": 0.02600731455721922,
+      "grad_norm": 0.5174048542976379,
+      "learning_rate": 0.0001999628622082394,
+      "loss": 0.6173,
+      "step": 26
+    },
+    {
+      "epoch": 0.027007595886343033,
+      "grad_norm": 0.525750994682312,
+      "learning_rate": 0.0001999599507118322,
+      "loss": 0.7309,
+      "step": 27
+    },
+    {
+      "epoch": 0.02800787721546685,
+      "grad_norm": 0.5904054641723633,
+      "learning_rate": 0.00019995692937751683,
+      "loss": 0.7459,
+      "step": 28
+    },
+    {
+      "epoch": 0.029008158544590667,
+      "grad_norm": 0.4662536084651947,
+      "learning_rate": 0.0001999537982086133,
+      "loss": 0.6309,
+      "step": 29
+    },
+    {
+      "epoch": 0.03000843987371448,
+      "grad_norm": 0.5540624856948853,
+      "learning_rate": 0.00019995055720856218,
+      "loss": 0.6604,
+      "step": 30
+    },
+    {
+      "epoch": 0.031008721202838298,
+      "grad_norm": 0.4457445442676544,
+      "learning_rate": 0.00019994720638092468,
+      "loss": 0.636,
+      "step": 31
+    },
+    {
+      "epoch": 0.032009002531962115,
+      "grad_norm": 0.4571971893310547,
+      "learning_rate": 0.00019994374572938277,
+      "loss": 0.6506,
+      "step": 32
+    },
+    {
+      "epoch": 0.03300928386108593,
+      "grad_norm": 0.5759263634681702,
+      "learning_rate": 0.00019994017525773913,
+      "loss": 0.5958,
+      "step": 33
+    },
+    {
+      "epoch": 0.03400956519020975,
+      "grad_norm": 0.4859147071838379,
+      "learning_rate": 0.00019993649496991705,
+      "loss": 0.6042,
+      "step": 34
+    },
+    {
+      "epoch": 0.03500984651933356,
+      "grad_norm": 0.3346123695373535,
+      "learning_rate": 0.00019993270486996046,
+      "loss": 0.5615,
+      "step": 35
+    },
+    {
+      "epoch": 0.03601012784845738,
+      "grad_norm": 0.4086807668209076,
+      "learning_rate": 0.000199928804962034,
+      "loss": 0.5048,
+      "step": 36
+    },
+    {
+      "epoch": 0.037010409177581194,
+      "grad_norm": 0.45720070600509644,
+      "learning_rate": 0.00019992479525042303,
+      "loss": 0.6318,
+      "step": 37
+    },
+    {
+      "epoch": 0.03801069050670501,
+      "grad_norm": 0.3767400085926056,
+      "learning_rate": 0.00019992067573953342,
+      "loss": 0.6242,
+      "step": 38
+    },
+    {
+      "epoch": 0.03901097183582883,
+      "grad_norm": 0.4942374527454376,
+      "learning_rate": 0.0001999164464338918,
+      "loss": 0.5696,
+      "step": 39
+    },
+    {
+      "epoch": 0.040011253164952645,
+      "grad_norm": 0.6337136626243591,
+      "learning_rate": 0.0001999121073381454,
+      "loss": 0.6565,
+      "step": 40
+    },
+    {
+      "epoch": 0.041011534494076456,
+      "grad_norm": 0.5405489802360535,
+      "learning_rate": 0.0001999076584570621,
+      "loss": 0.6096,
+      "step": 41
+    },
+    {
+      "epoch": 0.04201181582320027,
+      "grad_norm": 0.7466319799423218,
+      "learning_rate": 0.00019990309979553045,
+      "loss": 0.5782,
+      "step": 42
+    },
+    {
+      "epoch": 0.04301209715232409,
+      "grad_norm": 0.691272497177124,
+      "learning_rate": 0.00019989843135855958,
+      "loss": 0.622,
+      "step": 43
+    },
+    {
+      "epoch": 0.04401237848144791,
+      "grad_norm": 0.563149094581604,
+      "learning_rate": 0.00019989365315127922,
+      "loss": 0.6198,
+      "step": 44
+    },
+    {
+      "epoch": 0.045012659810571724,
+      "grad_norm": 0.47361648082733154,
+      "learning_rate": 0.0001998887651789398,
+      "loss": 0.5823,
+      "step": 45
+    },
+    {
+      "epoch": 0.04601294113969554,
+      "grad_norm": 0.5451771020889282,
+      "learning_rate": 0.0001998837674469123,
+      "loss": 0.6963,
+      "step": 46
+    },
+    {
+      "epoch": 0.04701322246881936,
+      "grad_norm": 0.4364232122898102,
+      "learning_rate": 0.00019987865996068833,
+      "loss": 0.5565,
+      "step": 47
+    },
+    {
+      "epoch": 0.04801350379794317,
+      "grad_norm": 0.4986230134963989,
+      "learning_rate": 0.00019987344272588006,
+      "loss": 0.5707,
+      "step": 48
+    },
+    {
+      "epoch": 0.049013785127066986,
+      "grad_norm": 0.46246567368507385,
+      "learning_rate": 0.00019986811574822033,
+      "loss": 0.5864,
+      "step": 49
+    },
+    {
+      "epoch": 0.0500140664561908,
+      "grad_norm": 0.41016271710395813,
+      "learning_rate": 0.00019986267903356254,
+      "loss": 0.552,
+      "step": 50
+    },
+    {
+      "epoch": 0.05101434778531462,
+      "grad_norm": 0.3554680049419403,
+      "learning_rate": 0.0001998571325878806,
+      "loss": 0.5582,
+      "step": 51
+    },
+    {
+      "epoch": 0.05201462911443844,
+      "grad_norm": 0.4468945562839508,
+      "learning_rate": 0.0001998514764172691,
+      "loss": 0.5108,
+      "step": 52
+    },
+    {
+      "epoch": 0.053014910443562255,
+      "grad_norm": 0.4482102692127228,
+      "learning_rate": 0.00019984571052794313,
+      "loss": 0.564,
+      "step": 53
+    },
+    {
+      "epoch": 0.054015191772686065,
+      "grad_norm": 0.5015258193016052,
+      "learning_rate": 0.00019983983492623833,
+      "loss": 0.6323,
+      "step": 54
+    },
+    {
+      "epoch": 0.05501547310180988,
+      "grad_norm": 0.4177694618701935,
+      "learning_rate": 0.00019983384961861096,
+      "loss": 0.565,
+      "step": 55
+    },
+    {
+      "epoch": 0.0560157544309337,
+      "grad_norm": 0.3855839669704437,
+      "learning_rate": 0.0001998277546116378,
+      "loss": 0.4947,
+      "step": 56
+    },
+    {
+      "epoch": 0.05701603576005752,
+      "grad_norm": 0.40609753131866455,
+      "learning_rate": 0.00019982154991201608,
+      "loss": 0.5428,
+      "step": 57
+    },
+    {
+      "epoch": 0.058016317089181334,
+      "grad_norm": 0.41122737526893616,
+      "learning_rate": 0.00019981523552656377,
+      "loss": 0.5374,
+      "step": 58
+    },
+    {
+      "epoch": 0.05901659841830515,
+      "grad_norm": 0.42212188243865967,
+      "learning_rate": 0.00019980881146221914,
+      "loss": 0.5192,
+      "step": 59
+    },
+    {
+      "epoch": 0.06001687974742896,
+      "grad_norm": 0.40538090467453003,
+      "learning_rate": 0.00019980227772604112,
+      "loss": 0.568,
+      "step": 60
+    },
+    {
+      "epoch": 0.06101716107655278,
+      "grad_norm": 0.3973814845085144,
+      "learning_rate": 0.0001997956343252091,
+      "loss": 0.5156,
+      "step": 61
+    },
+    {
+      "epoch": 0.062017442405676595,
+      "grad_norm": 0.36684533953666687,
+      "learning_rate": 0.00019978888126702296,
+      "loss": 0.4978,
+      "step": 62
+    },
+    {
+      "epoch": 0.06301772373480041,
+      "grad_norm": 0.45237308740615845,
+      "learning_rate": 0.00019978201855890308,
+      "loss": 0.5377,
+      "step": 63
+    },
+    {
+      "epoch": 0.06401800506392423,
+      "grad_norm": 0.4670639634132385,
+      "learning_rate": 0.00019977504620839035,
+      "loss": 0.5208,
+      "step": 64
+    },
+    {
+      "epoch": 0.06501828639304805,
+      "grad_norm": 0.4872264564037323,
+      "learning_rate": 0.00019976796422314615,
+      "loss": 0.5248,
+      "step": 65
+    },
+    {
+      "epoch": 0.06601856772217186,
+      "grad_norm": 0.5238344073295593,
+      "learning_rate": 0.00019976077261095226,
+      "loss": 0.6493,
+      "step": 66
+    },
+    {
+      "epoch": 0.06701884905129568,
+      "grad_norm": 0.6553749442100525,
+      "learning_rate": 0.00019975347137971098,
+      "loss": 0.5854,
+      "step": 67
+    },
+    {
+      "epoch": 0.0680191303804195,
+      "grad_norm": 0.49785664677619934,
+      "learning_rate": 0.00019974606053744503,
+      "loss": 0.551,
+      "step": 68
+    },
+    {
+      "epoch": 0.06901941170954332,
+      "grad_norm": 0.4122195839881897,
+      "learning_rate": 0.00019973854009229763,
+      "loss": 0.491,
+      "step": 69
+    },
+    {
+      "epoch": 0.07001969303866712,
+      "grad_norm": 0.4016016721725464,
+      "learning_rate": 0.00019973091005253232,
+      "loss": 0.5082,
+      "step": 70
+    },
+    {
+      "epoch": 0.07101997436779094,
+      "grad_norm": 0.3816310465335846,
+      "learning_rate": 0.0001997231704265332,
+      "loss": 0.5183,
+      "step": 71
+    },
+    {
+      "epoch": 0.07202025569691475,
+      "grad_norm": 0.3194684684276581,
+      "learning_rate": 0.00019971532122280464,
+      "loss": 0.4754,
+      "step": 72
+    },
+    {
+      "epoch": 0.07302053702603857,
+      "grad_norm": 0.33635130524635315,
+      "learning_rate": 0.0001997073624499716,
+      "loss": 0.5171,
+      "step": 73
+    },
+    {
+      "epoch": 0.07402081835516239,
+      "grad_norm": 0.4141254723072052,
+      "learning_rate": 0.0001996992941167792,
+      "loss": 0.5794,
+      "step": 74
+    },
+    {
+      "epoch": 0.0750210996842862,
+      "grad_norm": 0.381892591714859,
+      "learning_rate": 0.00019969111623209323,
+      "loss": 0.4623,
+      "step": 75
+    },
+    {
+      "epoch": 0.07602138101341002,
+      "grad_norm": 0.5439775586128235,
+      "learning_rate": 0.00019968282880489957,
+      "loss": 0.5866,
+      "step": 76
+    },
+    {
+      "epoch": 0.07702166234253384,
+      "grad_norm": 0.418949693441391,
+      "learning_rate": 0.00019967443184430467,
+      "loss": 0.4915,
+      "step": 77
+    },
+    {
+      "epoch": 0.07802194367165766,
+      "grad_norm": 0.41112858057022095,
+      "learning_rate": 0.0001996659253595353,
+      "loss": 0.5748,
+      "step": 78
+    },
+    {
+      "epoch": 0.07902222500078147,
+      "grad_norm": 0.44345736503601074,
+      "learning_rate": 0.0001996573093599385,
+      "loss": 0.5249,
+      "step": 79
+    },
+    {
+      "epoch": 0.08002250632990529,
+      "grad_norm": 0.37021568417549133,
+      "learning_rate": 0.00019964858385498172,
+      "loss": 0.4544,
+      "step": 80
+    },
+    {
+      "epoch": 0.08102278765902911,
+      "grad_norm": 0.38031843304634094,
+      "learning_rate": 0.00019963974885425266,
+      "loss": 0.4866,
+      "step": 81
+    },
+    {
+      "epoch": 0.08202306898815291,
+      "grad_norm": 0.35023489594459534,
+      "learning_rate": 0.00019963080436745945,
+      "loss": 0.5048,
+      "step": 82
+    },
+    {
+      "epoch": 0.08302335031727673,
+      "grad_norm": 0.4405301511287689,
+      "learning_rate": 0.00019962175040443044,
+      "loss": 0.4983,
+      "step": 83
+    },
+    {
+      "epoch": 0.08402363164640055,
+      "grad_norm": 0.28912487626075745,
+      "learning_rate": 0.0001996125869751143,
+      "loss": 0.4442,
+      "step": 84
+    },
+    {
+      "epoch": 0.08502391297552436,
+      "grad_norm": 0.3885224759578705,
+      "learning_rate": 0.00019960331408957997,
+      "loss": 0.4661,
+      "step": 85
+    },
+    {
+      "epoch": 0.08602419430464818,
+      "grad_norm": 0.34638479351997375,
+      "learning_rate": 0.00019959393175801671,
+      "loss": 0.5045,
+      "step": 86
+    },
+    {
+      "epoch": 0.087024475633772,
+      "grad_norm": 0.36377573013305664,
+      "learning_rate": 0.00019958443999073397,
+      "loss": 0.4827,
+      "step": 87
+    },
+    {
+      "epoch": 0.08802475696289581,
+      "grad_norm": 0.3212088942527771,
+      "learning_rate": 0.00019957483879816151,
+      "loss": 0.5012,
+      "step": 88
+    },
+    {
+      "epoch": 0.08902503829201963,
+      "grad_norm": 0.37304019927978516,
+      "learning_rate": 0.00019956512819084928,
+      "loss": 0.5184,
+      "step": 89
+    },
+    {
+      "epoch": 0.09002531962114345,
+      "grad_norm": 0.4289352297782898,
+      "learning_rate": 0.00019955530817946748,
+      "loss": 0.4852,
+      "step": 90
+    },
+    {
+      "epoch": 0.09102560095026727,
+      "grad_norm": 0.37521710991859436,
+      "learning_rate": 0.00019954537877480655,
+      "loss": 0.4492,
+      "step": 91
+    },
+    {
+      "epoch": 0.09202588227939108,
+      "grad_norm": 0.40025514364242554,
+      "learning_rate": 0.00019953533998777706,
+      "loss": 0.4724,
+      "step": 92
+    },
+    {
+      "epoch": 0.0930261636085149,
+      "grad_norm": 0.3497840464115143,
+      "learning_rate": 0.00019952519182940993,
+      "loss": 0.5133,
+      "step": 93
+    },
+    {
+      "epoch": 0.09402644493763872,
+      "grad_norm": 0.3495926558971405,
+      "learning_rate": 0.00019951493431085603,
+      "loss": 0.4623,
+      "step": 94
+    },
+    {
+      "epoch": 0.09502672626676252,
+      "grad_norm": 0.40398499369621277,
+      "learning_rate": 0.00019950456744338658,
+      "loss": 0.475,
+      "step": 95
+    },
+    {
+      "epoch": 0.09602700759588634,
+      "grad_norm": 0.37891319394111633,
+      "learning_rate": 0.00019949409123839288,
+      "loss": 0.5273,
+      "step": 96
+    },
+    {
+      "epoch": 0.09702728892501016,
+      "grad_norm": 0.37730953097343445,
+      "learning_rate": 0.00019948350570738642,
+      "loss": 0.5177,
+      "step": 97
+    },
+    {
+      "epoch": 0.09802757025413397,
+      "grad_norm": 0.3501202464103699,
+      "learning_rate": 0.0001994728108619987,
+      "loss": 0.5032,
+      "step": 98
+    },
+    {
+      "epoch": 0.09902785158325779,
+      "grad_norm": 0.35575389862060547,
+      "learning_rate": 0.0001994620067139815,
+      "loss": 0.4533,
+      "step": 99
+    },
+    {
+      "epoch": 0.1000281329123816,
+      "grad_norm": 0.3112006187438965,
+      "learning_rate": 0.00019945109327520658,
+      "loss": 0.461,
+      "step": 100
+    },
+    {
+      "epoch": 0.10102841424150542,
+      "grad_norm": 0.37702956795692444,
+      "learning_rate": 0.00019944007055766586,
+      "loss": 0.4929,
+      "step": 101
+    },
+    {
+      "epoch": 0.10202869557062924,
+      "grad_norm": 0.3292798101902008,
+      "learning_rate": 0.00019942893857347128,
+      "loss": 0.5031,
+      "step": 102
+    },
+    {
+      "epoch": 0.10302897689975306,
+      "grad_norm": 0.36839979887008667,
+      "learning_rate": 0.00019941769733485494,
+      "loss": 0.5204,
+      "step": 103
+    },
+    {
+      "epoch": 0.10402925822887688,
+      "grad_norm": 0.4038218557834625,
+      "learning_rate": 0.00019940634685416888,
+      "loss": 0.5036,
+      "step": 104
+    },
+    {
+      "epoch": 0.10502953955800069,
+      "grad_norm": 0.3477931022644043,
+      "learning_rate": 0.00019939488714388524,
+      "loss": 0.4547,
+      "step": 105
+    },
+    {
+      "epoch": 0.10602982088712451,
+      "grad_norm": 0.3681749105453491,
+      "learning_rate": 0.00019938331821659614,
+      "loss": 0.4588,
+      "step": 106
+    },
+    {
+      "epoch": 0.10703010221624833,
+      "grad_norm": 0.295214980840683,
+      "learning_rate": 0.0001993716400850138,
+      "loss": 0.4626,
+      "step": 107
+    },
+    {
+      "epoch": 0.10803038354537213,
+      "grad_norm": 0.35674479603767395,
+      "learning_rate": 0.0001993598527619703,
+      "loss": 0.4726,
+      "step": 108
+    },
+    {
+      "epoch": 0.10903066487449595,
+      "grad_norm": 0.3253282606601715,
+      "learning_rate": 0.00019934795626041783,
+      "loss": 0.4847,
+      "step": 109
+    },
+    {
+      "epoch": 0.11003094620361976,
+      "grad_norm": 0.3049859404563904,
+      "learning_rate": 0.0001993359505934285,
+      "loss": 0.4037,
+      "step": 110
+    },
+    {
+      "epoch": 0.11103122753274358,
+      "grad_norm": 0.3004118502140045,
+      "learning_rate": 0.00019932383577419432,
+      "loss": 0.4445,
+      "step": 111
+    },
+    {
+      "epoch": 0.1120315088618674,
+      "grad_norm": 0.32134872674942017,
+      "learning_rate": 0.0001993116118160273,
+      "loss": 0.4729,
+      "step": 112
+    },
+    {
+      "epoch": 0.11303179019099122,
+      "grad_norm": 0.4950062930583954,
+      "learning_rate": 0.00019929927873235938,
+      "loss": 0.5163,
+      "step": 113
+    },
+    {
+      "epoch": 0.11403207152011503,
+      "grad_norm": 0.34567245841026306,
+      "learning_rate": 0.00019928683653674237,
+      "loss": 0.4865,
+      "step": 114
+    },
+    {
+      "epoch": 0.11503235284923885,
+      "grad_norm": 0.4753624200820923,
+      "learning_rate": 0.00019927428524284805,
+      "loss": 0.5264,
+      "step": 115
+    },
+    {
+      "epoch": 0.11603263417836267,
+      "grad_norm": 0.3215334117412567,
+      "learning_rate": 0.00019926162486446792,
+      "loss": 0.4876,
+      "step": 116
+    },
+    {
+      "epoch": 0.11703291550748648,
+      "grad_norm": 0.3329266905784607,
+      "learning_rate": 0.0001992488554155135,
+      "loss": 0.4813,
+      "step": 117
+    },
+    {
+      "epoch": 0.1180331968366103,
+      "grad_norm": 0.3493446707725525,
+      "learning_rate": 0.00019923597691001615,
+      "loss": 0.4675,
+      "step": 118
+    },
+    {
+      "epoch": 0.11903347816573412,
+      "grad_norm": 0.336968332529068,
+      "learning_rate": 0.0001992229893621269,
+      "loss": 0.4712,
+      "step": 119
+    },
+    {
+      "epoch": 0.12003375949485792,
+      "grad_norm": 0.32647812366485596,
+      "learning_rate": 0.00019920989278611687,
+      "loss": 0.475,
+      "step": 120
+    },
+    {
+      "epoch": 0.12103404082398174,
+      "grad_norm": 0.36252933740615845,
+      "learning_rate": 0.0001991966871963767,
+      "loss": 0.5218,
+      "step": 121
+    },
+    {
+      "epoch": 0.12203432215310556,
+      "grad_norm": 0.31632480025291443,
+      "learning_rate": 0.000199183372607417,
+      "loss": 0.4798,
+      "step": 122
+    },
+    {
+      "epoch": 0.12303460348222937,
+      "grad_norm": 0.38675904273986816,
+      "learning_rate": 0.0001991699490338681,
+      "loss": 0.4993,
+      "step": 123
+    },
+    {
+      "epoch": 0.12403488481135319,
+      "grad_norm": 0.3446194529533386,
+      "learning_rate": 0.00019915641649048005,
+      "loss": 0.4299,
+      "step": 124
+    },
+    {
+      "epoch": 0.12503516614047702,
+      "grad_norm": 0.35425180196762085,
+      "learning_rate": 0.0001991427749921227,
+      "loss": 0.4235,
+      "step": 125
+    },
+    {
+      "epoch": 0.12603544746960083,
+      "grad_norm": 0.3712017834186554,
+      "learning_rate": 0.00019912902455378556,
+      "loss": 0.4756,
+      "step": 126
+    },
+    {
+      "epoch": 0.12703572879872463,
+      "grad_norm": 0.37233901023864746,
+      "learning_rate": 0.00019911516519057788,
+      "loss": 0.4519,
+      "step": 127
+    },
+    {
+      "epoch": 0.12803601012784846,
+      "grad_norm": 0.3591810464859009,
+      "learning_rate": 0.00019910119691772863,
+      "loss": 0.4598,
+      "step": 128
+    },
+    {
+      "epoch": 0.12903629145697226,
+      "grad_norm": 0.3153529465198517,
+      "learning_rate": 0.00019908711975058637,
+      "loss": 0.4534,
+      "step": 129
+    },
+    {
+      "epoch": 0.1300365727860961,
+      "grad_norm": 0.2840005159378052,
+      "learning_rate": 0.0001990729337046194,
+      "loss": 0.3998,
+      "step": 130
+    },
+    {
+      "epoch": 0.1310368541152199,
+      "grad_norm": 0.31850433349609375,
+      "learning_rate": 0.0001990586387954156,
+      "loss": 0.4448,
+      "step": 131
+    },
+    {
+      "epoch": 0.13203713544434373,
+      "grad_norm": 0.35457924008369446,
+      "learning_rate": 0.00019904423503868247,
+      "loss": 0.4629,
+      "step": 132
+    },
+    {
+      "epoch": 0.13303741677346753,
+      "grad_norm": 0.30500829219818115,
+      "learning_rate": 0.00019902972245024715,
+      "loss": 0.4671,
+      "step": 133
+    },
+    {
+      "epoch": 0.13403769810259136,
+      "grad_norm": 0.28144538402557373,
+      "learning_rate": 0.00019901510104605637,
+      "loss": 0.4292,
+      "step": 134
+    },
+    {
+      "epoch": 0.13503797943171517,
+      "grad_norm": 0.32505279779434204,
+      "learning_rate": 0.00019900037084217637,
+      "loss": 0.4862,
+      "step": 135
+    },
+    {
+      "epoch": 0.136038260760839,
+      "grad_norm": 0.30766561627388,
+      "learning_rate": 0.00019898553185479303,
+      "loss": 0.4031,
+      "step": 136
+    },
+    {
+      "epoch": 0.1370385420899628,
+      "grad_norm": 0.3485235273838043,
+      "learning_rate": 0.00019897058410021167,
+      "loss": 0.4298,
+      "step": 137
+    },
+    {
+      "epoch": 0.13803882341908663,
+      "grad_norm": 0.3085039556026459,
+      "learning_rate": 0.00019895552759485722,
+      "loss": 0.4799,
+      "step": 138
+    },
+    {
+      "epoch": 0.13903910474821043,
+      "grad_norm": 0.32276931405067444,
+      "learning_rate": 0.00019894036235527395,
+      "loss": 0.4488,
+      "step": 139
+    },
+    {
+      "epoch": 0.14003938607733424,
+      "grad_norm": 0.3475525677204132,
+      "learning_rate": 0.00019892508839812584,
+      "loss": 0.4838,
+      "step": 140
+    },
+    {
+      "epoch": 0.14103966740645807,
+      "grad_norm": 0.3139823079109192,
+      "learning_rate": 0.00019890970574019617,
+      "loss": 0.4428,
+      "step": 141
+    },
+    {
+      "epoch": 0.14203994873558187,
+      "grad_norm": 0.3611343801021576,
+      "learning_rate": 0.00019889421439838763,
+      "loss": 0.4276,
+      "step": 142
+    },
+    {
+      "epoch": 0.1430402300647057,
+      "grad_norm": 0.25845348834991455,
+      "learning_rate": 0.00019887861438972246,
+      "loss": 0.4118,
+      "step": 143
+    },
+    {
+      "epoch": 0.1440405113938295,
+      "grad_norm": 0.3170836567878723,
+      "learning_rate": 0.00019886290573134228,
+      "loss": 0.4679,
+      "step": 144
+    },
+    {
+      "epoch": 0.14504079272295334,
+      "grad_norm": 0.28241029381752014,
+      "learning_rate": 0.000198847088440508,
+      "loss": 0.415,
+      "step": 145
+    },
+    {
+      "epoch": 0.14604107405207714,
+      "grad_norm": 0.3596460223197937,
+      "learning_rate": 0.0001988311625346,
+      "loss": 0.4619,
+      "step": 146
+    },
+    {
+      "epoch": 0.14704135538120097,
+      "grad_norm": 0.2995036244392395,
+      "learning_rate": 0.00019881512803111796,
+      "loss": 0.4465,
+      "step": 147
+    },
+    {
+      "epoch": 0.14804163671032478,
+      "grad_norm": 0.3156537413597107,
+      "learning_rate": 0.00019879898494768093,
+      "loss": 0.4638,
+      "step": 148
+    },
+    {
+      "epoch": 0.1490419180394486,
+      "grad_norm": 0.39860132336616516,
+      "learning_rate": 0.00019878273330202717,
+      "loss": 0.4385,
+      "step": 149
+    },
+    {
+      "epoch": 0.1500421993685724,
+      "grad_norm": 0.2727566659450531,
+      "learning_rate": 0.00019876637311201433,
+      "loss": 0.4114,
+      "step": 150
+    },
+    {
+      "epoch": 0.15104248069769624,
+      "grad_norm": 0.2695155739784241,
+      "learning_rate": 0.00019874990439561934,
+      "loss": 0.4154,
+      "step": 151
+    },
+    {
+      "epoch": 0.15204276202682004,
+      "grad_norm": 0.33344969153404236,
+      "learning_rate": 0.0001987333271709383,
+      "loss": 0.477,
+      "step": 152
+    },
+    {
+      "epoch": 0.15304304335594385,
+      "grad_norm": 0.283865749835968,
+      "learning_rate": 0.00019871664145618657,
+      "loss": 0.4346,
+      "step": 153
+    },
+    {
+      "epoch": 0.15404332468506768,
+      "grad_norm": 0.27019625902175903,
+      "learning_rate": 0.00019869984726969878,
+      "loss": 0.4061,
+      "step": 154
+    },
+    {
+      "epoch": 0.15504360601419148,
+      "grad_norm": 0.2840537428855896,
+      "learning_rate": 0.00019868294462992866,
+      "loss": 0.4521,
+      "step": 155
+    },
+    {
+      "epoch": 0.1560438873433153,
+      "grad_norm": 0.2851097583770752,
+      "learning_rate": 0.00019866593355544922,
+      "loss": 0.3906,
+      "step": 156
+    },
+    {
+      "epoch": 0.15704416867243912,
+      "grad_norm": 0.3750353753566742,
+      "learning_rate": 0.00019864881406495246,
+      "loss": 0.475,
+      "step": 157
+    },
+    {
+      "epoch": 0.15804445000156295,
+      "grad_norm": 0.32451707124710083,
+      "learning_rate": 0.00019863158617724967,
+      "loss": 0.4191,
+      "step": 158
+    },
+    {
+      "epoch": 0.15904473133068675,
+      "grad_norm": 0.32818925380706787,
+      "learning_rate": 0.00019861424991127115,
+      "loss": 0.4004,
+      "step": 159
+    },
+    {
+      "epoch": 0.16004501265981058,
+      "grad_norm": 0.3559636175632477,
+      "learning_rate": 0.00019859680528606637,
+      "loss": 0.5077,
+      "step": 160
+    },
+    {
+      "epoch": 0.16104529398893438,
+      "grad_norm": 0.3427540063858032,
+      "learning_rate": 0.00019857925232080373,
+      "loss": 0.4617,
+      "step": 161
+    },
+    {
+      "epoch": 0.16204557531805822,
+      "grad_norm": 0.3394433856010437,
+      "learning_rate": 0.00019856159103477086,
+      "loss": 0.4849,
+      "step": 162
+    },
+    {
+      "epoch": 0.16304585664718202,
+      "grad_norm": 0.34075990319252014,
+      "learning_rate": 0.00019854382144737426,
+      "loss": 0.4486,
+      "step": 163
+    },
+    {
+      "epoch": 0.16404613797630582,
+      "grad_norm": 0.3583144545555115,
+      "learning_rate": 0.00019852594357813952,
+      "loss": 0.4616,
+      "step": 164
+    },
+    {
+      "epoch": 0.16504641930542965,
+      "grad_norm": 0.31794264912605286,
+      "learning_rate": 0.00019850795744671116,
+      "loss": 0.4964,
+      "step": 165
+    },
+    {
+      "epoch": 0.16604670063455346,
+      "grad_norm": 0.3856930732727051,
+      "learning_rate": 0.0001984898630728527,
+      "loss": 0.4186,
+      "step": 166
+    },
+    {
+      "epoch": 0.1670469819636773,
+      "grad_norm": 0.2840869724750519,
+      "learning_rate": 0.0001984716604764466,
+      "loss": 0.4166,
+      "step": 167
+    },
+    {
+      "epoch": 0.1680472632928011,
+      "grad_norm": 0.2664308249950409,
+      "learning_rate": 0.0001984533496774942,
+      "loss": 0.3982,
+      "step": 168
+    },
+    {
+      "epoch": 0.16904754462192492,
+      "grad_norm": 0.30538210272789,
+      "learning_rate": 0.0001984349306961158,
+      "loss": 0.4286,
+      "step": 169
+    },
+    {
+      "epoch": 0.17004782595104873,
+      "grad_norm": 0.32451754808425903,
+      "learning_rate": 0.00019841640355255043,
+      "loss": 0.4558,
+      "step": 170
+    },
+    {
+      "epoch": 0.17104810728017256,
+      "grad_norm": 0.29396554827690125,
+      "learning_rate": 0.00019839776826715614,
+      "loss": 0.3862,
+      "step": 171
+    },
+    {
+      "epoch": 0.17204838860929636,
+      "grad_norm": 0.337141752243042,
+      "learning_rate": 0.00019837902486040978,
+      "loss": 0.4805,
+      "step": 172
+    },
+    {
+      "epoch": 0.1730486699384202,
+      "grad_norm": 0.27084705233573914,
+      "learning_rate": 0.0001983601733529069,
+      "loss": 0.4354,
+      "step": 173
+    },
+    {
+      "epoch": 0.174048951267544,
+      "grad_norm": 0.31350868940353394,
+      "learning_rate": 0.00019834121376536187,
+      "loss": 0.4606,
+      "step": 174
+    },
+    {
+      "epoch": 0.17504923259666783,
+      "grad_norm": 0.33653292059898376,
+      "learning_rate": 0.00019832214611860793,
+      "loss": 0.5141,
+      "step": 175
+    },
+    {
+      "epoch": 0.17604951392579163,
+      "grad_norm": 0.28151023387908936,
+      "learning_rate": 0.00019830297043359692,
+      "loss": 0.4386,
+      "step": 176
+    },
+    {
+      "epoch": 0.17704979525491543,
+      "grad_norm": 0.3067042827606201,
+      "learning_rate": 0.00019828368673139947,
+      "loss": 0.4084,
+      "step": 177
+    },
+    {
+      "epoch": 0.17805007658403926,
+      "grad_norm": 0.2893977463245392,
+      "learning_rate": 0.0001982642950332049,
+      "loss": 0.4395,
+      "step": 178
+    },
+    {
+      "epoch": 0.17905035791316307,
+      "grad_norm": 0.2908482253551483,
+      "learning_rate": 0.00019824479536032112,
+      "loss": 0.3923,
+      "step": 179
+    },
+    {
+      "epoch": 0.1800506392422869,
+      "grad_norm": 0.30002954602241516,
+      "learning_rate": 0.0001982251877341748,
+      "loss": 0.3975,
+      "step": 180
+    },
+    {
+      "epoch": 0.1810509205714107,
+      "grad_norm": 0.31258004903793335,
+      "learning_rate": 0.00019820547217631117,
+      "loss": 0.4244,
+      "step": 181
+    },
+    {
+      "epoch": 0.18205120190053453,
+      "grad_norm": 0.4338636100292206,
+      "learning_rate": 0.00019818564870839405,
+      "loss": 0.4361,
+      "step": 182
+    },
+    {
+      "epoch": 0.18305148322965833,
+      "grad_norm": 0.2990376949310303,
+      "learning_rate": 0.00019816571735220583,
+      "loss": 0.4257,
+      "step": 183
+    },
+    {
+      "epoch": 0.18405176455878217,
+      "grad_norm": 0.3014718294143677,
+      "learning_rate": 0.00019814567812964748,
+      "loss": 0.3985,
+      "step": 184
+    },
+    {
+      "epoch": 0.18505204588790597,
+      "grad_norm": 0.2983604073524475,
+      "learning_rate": 0.00019812553106273847,
+      "loss": 0.413,
+      "step": 185
+    },
+    {
+      "epoch": 0.1860523272170298,
+      "grad_norm": 0.26064085960388184,
+      "learning_rate": 0.00019810527617361681,
+      "loss": 0.3713,
+      "step": 186
+    },
+    {
+      "epoch": 0.1870526085461536,
+      "grad_norm": 0.317233681678772,
+      "learning_rate": 0.00019808491348453894,
+      "loss": 0.4123,
+      "step": 187
+    },
+    {
+      "epoch": 0.18805288987527743,
+      "grad_norm": 0.32353103160858154,
+      "learning_rate": 0.00019806444301787978,
+      "loss": 0.4241,
+      "step": 188
+    },
+    {
+      "epoch": 0.18905317120440124,
+      "grad_norm": 0.3154282569885254,
+      "learning_rate": 0.0001980438647961327,
+      "loss": 0.464,
+      "step": 189
+    },
+    {
+      "epoch": 0.19005345253352504,
+      "grad_norm": 0.5833514928817749,
+      "learning_rate": 0.00019802317884190935,
+      "loss": 0.4107,
+      "step": 190
+    },
+    {
+      "epoch": 0.19105373386264887,
+      "grad_norm": 0.26230475306510925,
+      "learning_rate": 0.00019800238517793996,
+      "loss": 0.4036,
+      "step": 191
+    },
+    {
+      "epoch": 0.19205401519177268,
+      "grad_norm": 0.2554140090942383,
+      "learning_rate": 0.00019798148382707296,
+      "loss": 0.3799,
+      "step": 192
+    },
+    {
+      "epoch": 0.1930542965208965,
+      "grad_norm": 0.2918940484523773,
+      "learning_rate": 0.00019796047481227515,
+      "loss": 0.4366,
+      "step": 193
+    },
+    {
+      "epoch": 0.1940545778500203,
+      "grad_norm": 0.27494698762893677,
+      "learning_rate": 0.00019793935815663163,
+      "loss": 0.4452,
+      "step": 194
+    },
+    {
+      "epoch": 0.19505485917914414,
+      "grad_norm": 0.31667977571487427,
+      "learning_rate": 0.00019791813388334581,
+      "loss": 0.4058,
+      "step": 195
+    },
+    {
+      "epoch": 0.19605514050826794,
+      "grad_norm": 0.8476918339729309,
+      "learning_rate": 0.00019789680201573933,
+      "loss": 0.4026,
+      "step": 196
+    },
+    {
+      "epoch": 0.19705542183739178,
+      "grad_norm": 0.38591355085372925,
+      "learning_rate": 0.00019787536257725202,
+      "loss": 0.4508,
+      "step": 197
+    },
+    {
+      "epoch": 0.19805570316651558,
+      "grad_norm": 0.565131664276123,
+      "learning_rate": 0.00019785381559144196,
+      "loss": 0.405,
+      "step": 198
+    },
+    {
+      "epoch": 0.1990559844956394,
+      "grad_norm": 0.32413753867149353,
+      "learning_rate": 0.00019783216108198542,
+      "loss": 0.3934,
+      "step": 199
+    },
+    {
+      "epoch": 0.2000562658247632,
+      "grad_norm": 0.3276815712451935,
+      "learning_rate": 0.00019781039907267677,
+      "loss": 0.448,
+      "step": 200
+    },
+    {
+      "epoch": 0.20105654715388704,
+      "grad_norm": 0.3068343698978424,
+      "learning_rate": 0.00019778852958742853,
+      "loss": 0.4454,
+      "step": 201
+    },
+    {
+      "epoch": 0.20205682848301085,
+      "grad_norm": 0.29425284266471863,
+      "learning_rate": 0.00019776655265027127,
+      "loss": 0.4473,
+      "step": 202
+    },
+    {
+      "epoch": 0.20305710981213465,
+      "grad_norm": 0.26625314354896545,
+      "learning_rate": 0.00019774446828535371,
+      "loss": 0.3749,
+      "step": 203
+    },
+    {
+      "epoch": 0.20405739114125848,
+      "grad_norm": 0.29396072030067444,
+      "learning_rate": 0.00019772227651694256,
+      "loss": 0.4101,
+      "step": 204
+    },
+    {
+      "epoch": 0.20505767247038229,
+      "grad_norm": 0.301468163728714,
+      "learning_rate": 0.00019769997736942258,
+      "loss": 0.3681,
+      "step": 205
+    },
+    {
+      "epoch": 0.20605795379950612,
+      "grad_norm": 0.3525398373603821,
+      "learning_rate": 0.00019767757086729647,
+      "loss": 0.3903,
+      "step": 206
+    },
+    {
+      "epoch": 0.20705823512862992,
+      "grad_norm": 0.3096582889556885,
+      "learning_rate": 0.00019765505703518496,
+      "loss": 0.4027,
+      "step": 207
+    },
+    {
+      "epoch": 0.20805851645775375,
+      "grad_norm": 0.47655290365219116,
+      "learning_rate": 0.00019763243589782662,
+      "loss": 0.4027,
+      "step": 208
+    },
+    {
+      "epoch": 0.20905879778687755,
+      "grad_norm": 0.30790695548057556,
+      "learning_rate": 0.00019760970748007803,
+      "loss": 0.3621,
+      "step": 209
+    },
+    {
+      "epoch": 0.21005907911600138,
+      "grad_norm": 0.32526037096977234,
+      "learning_rate": 0.0001975868718069136,
+      "loss": 0.4366,
+      "step": 210
+    },
+    {
+      "epoch": 0.2110593604451252,
+      "grad_norm": 0.28599101305007935,
+      "learning_rate": 0.00019756392890342563,
+      "loss": 0.3801,
+      "step": 211
+    },
+    {
+      "epoch": 0.21205964177424902,
+      "grad_norm": 0.3474257290363312,
+      "learning_rate": 0.00019754087879482422,
+      "loss": 0.4245,
+      "step": 212
+    },
+    {
+      "epoch": 0.21305992310337282,
+      "grad_norm": 0.34767356514930725,
+      "learning_rate": 0.00019751772150643722,
+      "loss": 0.4571,
+      "step": 213
+    },
+    {
+      "epoch": 0.21406020443249665,
+      "grad_norm": 0.3164922297000885,
+      "learning_rate": 0.00019749445706371038,
+      "loss": 0.4161,
+      "step": 214
+    },
+    {
+      "epoch": 0.21506048576162046,
+      "grad_norm": 0.3492523729801178,
+      "learning_rate": 0.00019747108549220702,
+      "loss": 0.4551,
+      "step": 215
+    },
+    {
+      "epoch": 0.21606076709074426,
+      "grad_norm": 0.31708961725234985,
+      "learning_rate": 0.00019744760681760832,
+      "loss": 0.4416,
+      "step": 216
+    },
+    {
+      "epoch": 0.2170610484198681,
+      "grad_norm": 0.2811904847621918,
+      "learning_rate": 0.00019742402106571314,
+      "loss": 0.4187,
+      "step": 217
+    },
+    {
+      "epoch": 0.2180613297489919,
+      "grad_norm": 0.31777599453926086,
+      "learning_rate": 0.00019740032826243788,
+      "loss": 0.4135,
+      "step": 218
+    },
+    {
+      "epoch": 0.21906161107811573,
+      "grad_norm": 0.2507490813732147,
+      "learning_rate": 0.0001973765284338167,
+      "loss": 0.371,
+      "step": 219
+    },
+    {
+      "epoch": 0.22006189240723953,
+      "grad_norm": 0.2680091857910156,
+      "learning_rate": 0.00019735262160600127,
+      "loss": 0.416,
+      "step": 220
+    },
+    {
+      "epoch": 0.22106217373636336,
+      "grad_norm": 0.359147846698761,
+      "learning_rate": 0.00019732860780526088,
+      "loss": 0.4347,
+      "step": 221
+    },
+    {
+      "epoch": 0.22206245506548716,
+      "grad_norm": 0.29516977071762085,
+      "learning_rate": 0.00019730448705798239,
+      "loss": 0.4553,
+      "step": 222
+    },
+    {
+      "epoch": 0.223062736394611,
+      "grad_norm": 0.3992477059364319,
+      "learning_rate": 0.00019728025939067008,
+      "loss": 0.3778,
+      "step": 223
+    },
+    {
+      "epoch": 0.2240630177237348,
+      "grad_norm": 0.25359246134757996,
+      "learning_rate": 0.00019725592482994583,
+      "loss": 0.3797,
+      "step": 224
+    },
+    {
+      "epoch": 0.22506329905285863,
+      "grad_norm": 0.4242265522480011,
+      "learning_rate": 0.00019723148340254892,
+      "loss": 0.3718,
+      "step": 225
+    },
+    {
+      "epoch": 0.22606358038198243,
+      "grad_norm": 0.26787999272346497,
+      "learning_rate": 0.00019720693513533598,
+      "loss": 0.3857,
+      "step": 226
+    },
+    {
+      "epoch": 0.22706386171110624,
+      "grad_norm": 0.23111741244792938,
+      "learning_rate": 0.00019718228005528122,
+      "loss": 0.349,
+      "step": 227
+    },
+    {
+      "epoch": 0.22806414304023007,
+      "grad_norm": 0.2645491063594818,
+      "learning_rate": 0.00019715751818947603,
+      "loss": 0.413,
+      "step": 228
+    },
+    {
+      "epoch": 0.22906442436935387,
+      "grad_norm": 0.30736207962036133,
+      "learning_rate": 0.0001971326495651293,
+      "loss": 0.4129,
+      "step": 229
+    },
+    {
+      "epoch": 0.2300647056984777,
+      "grad_norm": 0.30426740646362305,
+      "learning_rate": 0.00019710767420956705,
+      "loss": 0.3977,
+      "step": 230
+    },
+    {
+      "epoch": 0.2310649870276015,
+      "grad_norm": 0.3317965865135193,
+      "learning_rate": 0.0001970825921502328,
+      "loss": 0.4309,
+      "step": 231
+    },
+    {
+      "epoch": 0.23206526835672533,
+      "grad_norm": 0.31584450602531433,
+      "learning_rate": 0.0001970574034146871,
+      "loss": 0.3855,
+      "step": 232
+    },
+    {
+      "epoch": 0.23306554968584914,
+      "grad_norm": 0.2980955243110657,
+      "learning_rate": 0.00019703210803060782,
+      "loss": 0.415,
+      "step": 233
+    },
+    {
+      "epoch": 0.23406583101497297,
+      "grad_norm": 0.24319759011268616,
+      "learning_rate": 0.00019700670602579008,
+      "loss": 0.4051,
+      "step": 234
+    },
+    {
+      "epoch": 0.23506611234409677,
+      "grad_norm": 0.3305763900279999,
+      "learning_rate": 0.00019698119742814606,
+      "loss": 0.4734,
+      "step": 235
+    },
+    {
+      "epoch": 0.2360663936732206,
+      "grad_norm": 0.26589512825012207,
+      "learning_rate": 0.00019695558226570507,
+      "loss": 0.4197,
+      "step": 236
+    },
+    {
+      "epoch": 0.2370666750023444,
+      "grad_norm": 0.3133488893508911,
+      "learning_rate": 0.00019692986056661356,
+      "loss": 0.4478,
+      "step": 237
+    },
+    {
+      "epoch": 0.23806695633146824,
+      "grad_norm": 0.27163878083229065,
+      "learning_rate": 0.00019690403235913504,
+      "loss": 0.3991,
+      "step": 238
+    },
+    {
+      "epoch": 0.23906723766059204,
+      "grad_norm": 0.256094753742218,
+      "learning_rate": 0.00019687809767165,
+      "loss": 0.4073,
+      "step": 239
+    },
+    {
+      "epoch": 0.24006751898971584,
+      "grad_norm": 0.31229445338249207,
+      "learning_rate": 0.000196852056532656,
+      "loss": 0.3997,
+      "step": 240
+    },
+    {
+      "epoch": 0.24106780031883968,
+      "grad_norm": 0.2771662771701813,
+      "learning_rate": 0.00019682590897076752,
+      "loss": 0.4268,
+      "step": 241
+    },
+    {
+      "epoch": 0.24206808164796348,
+      "grad_norm": 0.25749140977859497,
+      "learning_rate": 0.00019679965501471608,
+      "loss": 0.3355,
+      "step": 242
+    },
+    {
+      "epoch": 0.2430683629770873,
+      "grad_norm": 0.29009538888931274,
+      "learning_rate": 0.0001967732946933499,
+      "loss": 0.444,
+      "step": 243
+    },
+    {
+      "epoch": 0.2440686443062111,
+      "grad_norm": 0.2621695101261139,
+      "learning_rate": 0.00019674682803563428,
+      "loss": 0.4083,
+      "step": 244
+    },
+    {
+      "epoch": 0.24506892563533494,
+      "grad_norm": 0.342739462852478,
+      "learning_rate": 0.00019672025507065131,
+      "loss": 0.4375,
+      "step": 245
+    },
+    {
+      "epoch": 0.24606920696445875,
+      "grad_norm": 0.3089183270931244,
+      "learning_rate": 0.00019669357582759983,
+      "loss": 0.4151,
+      "step": 246
+    },
+    {
+      "epoch": 0.24706948829358258,
+      "grad_norm": 0.4531654119491577,
+      "learning_rate": 0.00019666679033579552,
+      "loss": 0.4266,
+      "step": 247
+    },
+    {
+      "epoch": 0.24806976962270638,
+      "grad_norm": 0.27016133069992065,
+      "learning_rate": 0.00019663989862467082,
+      "loss": 0.4274,
+      "step": 248
+    },
+    {
+      "epoch": 0.2490700509518302,
+      "grad_norm": 0.27403491735458374,
+      "learning_rate": 0.00019661290072377482,
+      "loss": 0.383,
+      "step": 249
+    },
+    {
+      "epoch": 0.25007033228095404,
+      "grad_norm": 0.2885066866874695,
+      "learning_rate": 0.00019658579666277334,
+      "loss": 0.4072,
+      "step": 250
+    },
+    {
+      "epoch": 0.2510706136100778,
+      "grad_norm": 0.24861767888069153,
+      "learning_rate": 0.0001965585864714488,
+      "loss": 0.3717,
+      "step": 251
+    },
+    {
+      "epoch": 0.25207089493920165,
+      "grad_norm": 0.2628852128982544,
+      "learning_rate": 0.00019653127017970034,
+      "loss": 0.3876,
+      "step": 252
+    },
+    {
+      "epoch": 0.2530711762683255,
+      "grad_norm": 0.2990763783454895,
+      "learning_rate": 0.0001965038478175436,
+      "loss": 0.4279,
+      "step": 253
+    },
+    {
+      "epoch": 0.25407145759744926,
+      "grad_norm": 0.36784109473228455,
+      "learning_rate": 0.00019647631941511082,
+      "loss": 0.3788,
+      "step": 254
+    },
+    {
+      "epoch": 0.2550717389265731,
+      "grad_norm": 0.34909719228744507,
+      "learning_rate": 0.0001964486850026507,
+      "loss": 0.4315,
+      "step": 255
+    },
+    {
+      "epoch": 0.2560720202556969,
+      "grad_norm": 0.3366207480430603,
+      "learning_rate": 0.00019642094461052852,
+      "loss": 0.4038,
+      "step": 256
+    },
+    {
+      "epoch": 0.25707230158482075,
+      "grad_norm": 0.27798953652381897,
+      "learning_rate": 0.00019639309826922585,
+      "loss": 0.4142,
+      "step": 257
+    },
+    {
+      "epoch": 0.2580725829139445,
+      "grad_norm": 0.23876960575580597,
+      "learning_rate": 0.0001963651460093409,
+      "loss": 0.3745,
+      "step": 258
+    },
+    {
+      "epoch": 0.25907286424306836,
+      "grad_norm": 0.8659366369247437,
+      "learning_rate": 0.00019633708786158806,
+      "loss": 0.4195,
+      "step": 259
+    },
+    {
+      "epoch": 0.2600731455721922,
+      "grad_norm": 0.2651257812976837,
+      "learning_rate": 0.00019630892385679818,
+      "loss": 0.4289,
+      "step": 260
+    },
+    {
+      "epoch": 0.261073426901316,
+      "grad_norm": 0.23504844307899475,
+      "learning_rate": 0.00019628065402591845,
+      "loss": 0.3659,
+      "step": 261
+    },
+    {
+      "epoch": 0.2620737082304398,
+      "grad_norm": 0.2684794068336487,
+      "learning_rate": 0.00019625227840001225,
+      "loss": 0.3854,
+      "step": 262
+    },
+    {
+      "epoch": 0.2630739895595636,
+      "grad_norm": 0.25387704372406006,
+      "learning_rate": 0.0001962237970102593,
+      "loss": 0.3877,
+      "step": 263
+    },
+    {
+      "epoch": 0.26407427088868746,
+      "grad_norm": 0.24009455740451813,
+      "learning_rate": 0.0001961952098879555,
+      "loss": 0.3785,
+      "step": 264
+    },
+    {
+      "epoch": 0.26507455221781123,
+      "grad_norm": 0.27629122138023376,
+      "learning_rate": 0.00019616651706451287,
+      "loss": 0.3474,
+      "step": 265
+    },
+    {
+      "epoch": 0.26607483354693506,
+      "grad_norm": 0.2714496850967407,
+      "learning_rate": 0.0001961377185714597,
+      "loss": 0.3708,
+      "step": 266
+    },
+    {
+      "epoch": 0.2670751148760589,
+      "grad_norm": 0.2567186951637268,
+      "learning_rate": 0.0001961088144404403,
+      "loss": 0.3725,
+      "step": 267
+    },
+    {
+      "epoch": 0.2680753962051827,
+      "grad_norm": 0.2741433382034302,
+      "learning_rate": 0.00019607980470321505,
+      "loss": 0.4145,
+      "step": 268
+    },
+    {
+      "epoch": 0.2690756775343065,
+      "grad_norm": 0.28943076729774475,
+      "learning_rate": 0.00019605068939166045,
+      "loss": 0.4202,
+      "step": 269
+    },
+    {
+      "epoch": 0.27007595886343033,
+      "grad_norm": 0.2776735723018646,
+      "learning_rate": 0.00019602146853776894,
+      "loss": 0.3991,
+      "step": 270
+    },
+    {
+      "epoch": 0.27107624019255416,
+      "grad_norm": 0.23200170695781708,
+      "learning_rate": 0.000195992142173649,
+      "loss": 0.3167,
+      "step": 271
+    },
+    {
+      "epoch": 0.272076521521678,
+      "grad_norm": 0.44690176844596863,
+      "learning_rate": 0.0001959627103315249,
+      "loss": 0.3782,
+      "step": 272
+    },
+    {
+      "epoch": 0.27307680285080177,
+      "grad_norm": 0.28088852763175964,
+      "learning_rate": 0.00019593317304373705,
+      "loss": 0.3672,
+      "step": 273
+    },
+    {
+      "epoch": 0.2740770841799256,
+      "grad_norm": 0.29305198788642883,
+      "learning_rate": 0.00019590353034274144,
+      "loss": 0.4075,
+      "step": 274
+    },
+    {
+      "epoch": 0.27507736550904943,
+      "grad_norm": 0.30940836668014526,
+      "learning_rate": 0.00019587378226111014,
+      "loss": 0.4669,
+      "step": 275
+    },
+    {
+      "epoch": 0.27607764683817326,
+      "grad_norm": 0.2809953987598419,
+      "learning_rate": 0.00019584392883153088,
+      "loss": 0.3778,
+      "step": 276
+    },
+    {
+      "epoch": 0.27707792816729704,
+      "grad_norm": 0.32367703318595886,
+      "learning_rate": 0.00019581397008680717,
+      "loss": 0.3945,
+      "step": 277
+    },
+    {
+      "epoch": 0.27807820949642087,
+      "grad_norm": 0.3615182042121887,
+      "learning_rate": 0.00019578390605985826,
+      "loss": 0.4112,
+      "step": 278
+    },
+    {
+      "epoch": 0.2790784908255447,
+      "grad_norm": 0.3386395275592804,
+      "learning_rate": 0.00019575373678371909,
+      "loss": 0.4026,
+      "step": 279
+    },
+    {
+      "epoch": 0.2800787721546685,
+      "grad_norm": 0.30399829149246216,
+      "learning_rate": 0.00019572346229154025,
+      "loss": 0.3748,
+      "step": 280
+    },
+    {
+      "epoch": 0.2810790534837923,
+      "grad_norm": 0.24133749306201935,
+      "learning_rate": 0.00019569308261658787,
+      "loss": 0.3774,
+      "step": 281
+    },
+    {
+      "epoch": 0.28207933481291614,
+      "grad_norm": 0.24240224063396454,
+      "learning_rate": 0.00019566259779224378,
+      "loss": 0.4204,
+      "step": 282
+    },
+    {
+      "epoch": 0.28307961614203997,
+      "grad_norm": 0.278439998626709,
+      "learning_rate": 0.00019563200785200526,
+      "loss": 0.375,
+      "step": 283
+    },
+    {
+      "epoch": 0.28407989747116374,
+      "grad_norm": 0.24788439273834229,
+      "learning_rate": 0.00019560131282948516,
+      "loss": 0.3577,
+      "step": 284
+    },
+    {
+      "epoch": 0.2850801788002876,
+      "grad_norm": 0.31011322140693665,
+      "learning_rate": 0.0001955705127584117,
+      "loss": 0.4538,
+      "step": 285
+    },
+    {
+      "epoch": 0.2860804601294114,
+      "grad_norm": 0.257071316242218,
+      "learning_rate": 0.00019553960767262863,
+      "loss": 0.3881,
+      "step": 286
+    },
+    {
+      "epoch": 0.28708074145853524,
+      "grad_norm": 0.5129604339599609,
+      "learning_rate": 0.00019550859760609503,
+      "loss": 0.3833,
+      "step": 287
+    },
+    {
+      "epoch": 0.288081022787659,
+      "grad_norm": 0.234080508351326,
+      "learning_rate": 0.00019547748259288536,
+      "loss": 0.3843,
+      "step": 288
+    },
+    {
+      "epoch": 0.28908130411678284,
+      "grad_norm": 0.4176006615161896,
+      "learning_rate": 0.0001954462626671894,
+      "loss": 0.431,
+      "step": 289
+    },
+    {
+      "epoch": 0.2900815854459067,
+      "grad_norm": 0.30611085891723633,
+      "learning_rate": 0.0001954149378633122,
+      "loss": 0.4091,
+      "step": 290
+    },
+    {
+      "epoch": 0.29108186677503045,
+      "grad_norm": 0.26238131523132324,
+      "learning_rate": 0.00019538350821567404,
+      "loss": 0.346,
+      "step": 291
+    },
+    {
+      "epoch": 0.2920821481041543,
+      "grad_norm": 0.28092989325523376,
+      "learning_rate": 0.00019535197375881045,
+      "loss": 0.4251,
+      "step": 292
+    },
+    {
+      "epoch": 0.2930824294332781,
+      "grad_norm": 0.27506351470947266,
+      "learning_rate": 0.00019532033452737205,
+      "loss": 0.3944,
+      "step": 293
+    },
+    {
+      "epoch": 0.29408271076240194,
+      "grad_norm": 0.2865377366542816,
+      "learning_rate": 0.00019528859055612468,
+      "loss": 0.4214,
+      "step": 294
+    },
+    {
+      "epoch": 0.2950829920915257,
+      "grad_norm": 0.2901553809642792,
+      "learning_rate": 0.0001952567418799492,
+      "loss": 0.4395,
+      "step": 295
+    },
+    {
+      "epoch": 0.29608327342064955,
+      "grad_norm": 0.22749190032482147,
+      "learning_rate": 0.00019522478853384155,
+      "loss": 0.36,
+      "step": 296
+    },
+    {
+      "epoch": 0.2970835547497734,
+      "grad_norm": 0.26204293966293335,
+      "learning_rate": 0.00019519273055291266,
+      "loss": 0.4006,
+      "step": 297
+    },
+    {
+      "epoch": 0.2980838360788972,
+      "grad_norm": 0.25596141815185547,
+      "learning_rate": 0.00019516056797238846,
+      "loss": 0.3789,
+      "step": 298
+    },
+    {
+      "epoch": 0.299084117408021,
+      "grad_norm": 0.2562818229198456,
+      "learning_rate": 0.00019512830082760987,
+      "loss": 0.404,
+      "step": 299
+    },
+    {
+      "epoch": 0.3000843987371448,
+      "grad_norm": 0.3109433352947235,
+      "learning_rate": 0.00019509592915403255,
+      "loss": 0.4498,
+      "step": 300
+    },
+    {
+      "epoch": 0.30108468006626865,
+      "grad_norm": 0.27133166790008545,
+      "learning_rate": 0.00019506345298722717,
+      "loss": 0.3634,
+      "step": 301
+    },
+    {
+      "epoch": 0.3020849613953925,
+      "grad_norm": 0.29989707469940186,
+      "learning_rate": 0.00019503087236287913,
+      "loss": 0.4089,
+      "step": 302
+    },
+    {
+      "epoch": 0.30308524272451626,
+      "grad_norm": 0.31667646765708923,
+      "learning_rate": 0.00019499818731678873,
+      "loss": 0.3368,
+      "step": 303
+    },
+    {
+      "epoch": 0.3040855240536401,
+      "grad_norm": 0.2644802927970886,
+      "learning_rate": 0.00019496539788487082,
+      "loss": 0.39,
+      "step": 304
+    },
+    {
+      "epoch": 0.3050858053827639,
+      "grad_norm": 0.36831945180892944,
+      "learning_rate": 0.0001949325041031551,
+      "loss": 0.4179,
+      "step": 305
+    },
+    {
+      "epoch": 0.3060860867118877,
+      "grad_norm": 0.2724269926548004,
+      "learning_rate": 0.0001948995060077859,
+      "loss": 0.4387,
+      "step": 306
+    },
+    {
+      "epoch": 0.3070863680410115,
+      "grad_norm": 0.24974443018436432,
+      "learning_rate": 0.0001948664036350221,
+      "loss": 0.3933,
+      "step": 307
+    },
+    {
+      "epoch": 0.30808664937013536,
+      "grad_norm": 0.26349300146102905,
+      "learning_rate": 0.00019483319702123732,
+      "loss": 0.3895,
+      "step": 308
+    },
+    {
+      "epoch": 0.3090869306992592,
+      "grad_norm": 0.3007661998271942,
+      "learning_rate": 0.00019479988620291956,
+      "loss": 0.3507,
+      "step": 309
+    },
+    {
+      "epoch": 0.31008721202838296,
+      "grad_norm": 0.2490624487400055,
+      "learning_rate": 0.00019476647121667137,
+      "loss": 0.3334,
+      "step": 310
+    },
+    {
+      "epoch": 0.3110874933575068,
+      "grad_norm": 0.2903260886669159,
+      "learning_rate": 0.00019473295209920983,
+      "loss": 0.4624,
+      "step": 311
+    },
+    {
+      "epoch": 0.3120877746866306,
+      "grad_norm": 0.2873629033565521,
+      "learning_rate": 0.00019469932888736632,
+      "loss": 0.4099,
+      "step": 312
+    },
+    {
+      "epoch": 0.31308805601575446,
+      "grad_norm": 0.2918339967727661,
+      "learning_rate": 0.00019466560161808674,
+      "loss": 0.3617,
+      "step": 313
+    },
+    {
+      "epoch": 0.31408833734487823,
+      "grad_norm": 0.2637473940849304,
+      "learning_rate": 0.00019463177032843124,
+      "loss": 0.3906,
+      "step": 314
+    },
+    {
+      "epoch": 0.31508861867400206,
+      "grad_norm": 0.364161878824234,
+      "learning_rate": 0.00019459783505557424,
+      "loss": 0.402,
+      "step": 315
+    },
+    {
+      "epoch": 0.3160889000031259,
+      "grad_norm": 0.27097612619400024,
+      "learning_rate": 0.00019456379583680452,
+      "loss": 0.3748,
+      "step": 316
+    },
+    {
+      "epoch": 0.31708918133224967,
+      "grad_norm": 0.29231584072113037,
+      "learning_rate": 0.000194529652709525,
+      "loss": 0.4232,
+      "step": 317
+    },
+    {
+      "epoch": 0.3180894626613735,
+      "grad_norm": 0.2763817310333252,
+      "learning_rate": 0.00019449540571125286,
+      "loss": 0.3895,
+      "step": 318
+    },
+    {
+      "epoch": 0.31908974399049733,
+      "grad_norm": 0.27088895440101624,
+      "learning_rate": 0.00019446105487961926,
+      "loss": 0.369,
+      "step": 319
+    },
+    {
+      "epoch": 0.32009002531962116,
+      "grad_norm": 0.37634366750717163,
+      "learning_rate": 0.0001944266002523696,
+      "loss": 0.3474,
+      "step": 320
+    },
+    {
+      "epoch": 0.32109030664874494,
+      "grad_norm": 0.26676034927368164,
+      "learning_rate": 0.0001943920418673633,
+      "loss": 0.3793,
+      "step": 321
+    },
+    {
+      "epoch": 0.32209058797786877,
+      "grad_norm": 0.27637040615081787,
+      "learning_rate": 0.00019435737976257377,
+      "loss": 0.3914,
+      "step": 322
+    },
+    {
+      "epoch": 0.3230908693069926,
+      "grad_norm": 0.2805480360984802,
+      "learning_rate": 0.00019432261397608834,
+      "loss": 0.3985,
+      "step": 323
+    },
+    {
+      "epoch": 0.32409115063611643,
+      "grad_norm": 0.23319090902805328,
+      "learning_rate": 0.00019428774454610843,
+      "loss": 0.3664,
+      "step": 324
+    },
+    {
+      "epoch": 0.3250914319652402,
+      "grad_norm": 0.23325973749160767,
+      "learning_rate": 0.00019425277151094913,
+      "loss": 0.3393,
+      "step": 325
+    },
+    {
+      "epoch": 0.32609171329436404,
+      "grad_norm": 0.26448819041252136,
+      "learning_rate": 0.00019421769490903957,
+      "loss": 0.3972,
+      "step": 326
+    },
+    {
+      "epoch": 0.32709199462348787,
+      "grad_norm": 0.2579127848148346,
+      "learning_rate": 0.0001941825147789225,
+      "loss": 0.317,
+      "step": 327
+    },
+    {
+      "epoch": 0.32809227595261165,
+      "grad_norm": 0.2921576499938965,
+      "learning_rate": 0.00019414723115925456,
+      "loss": 0.3781,
+      "step": 328
+    },
+    {
+      "epoch": 0.3290925572817355,
+      "grad_norm": 0.2583731710910797,
+      "learning_rate": 0.0001941118440888061,
+      "loss": 0.4428,
+      "step": 329
+    },
+    {
+      "epoch": 0.3300928386108593,
+      "grad_norm": 0.30170106887817383,
+      "learning_rate": 0.0001940763536064611,
+      "loss": 0.3839,
+      "step": 330
+    },
+    {
+      "epoch": 0.33109311993998314,
+      "grad_norm": 0.35121557116508484,
+      "learning_rate": 0.00019404075975121716,
+      "loss": 0.4176,
+      "step": 331
+    },
+    {
+      "epoch": 0.3320934012691069,
+      "grad_norm": 0.29237017035484314,
+      "learning_rate": 0.0001940050625621855,
+      "loss": 0.3906,
+      "step": 332
+    },
+    {
+      "epoch": 0.33309368259823074,
+      "grad_norm": 0.22922085225582123,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 0.36,
+      "step": 333
+    },
+    {
+      "epoch": 0.3340939639273546,
+      "grad_norm": 0.2649683952331543,
+      "learning_rate": 0.0001939333583397715,
+      "loss": 0.3781,
+      "step": 334
+    },
+    {
+      "epoch": 0.3350942452564784,
+      "grad_norm": 0.22827917337417603,
+      "learning_rate": 0.00019389735138517915,
+      "loss": 0.343,
+      "step": 335
+    },
+    {
+      "epoch": 0.3360945265856022,
+      "grad_norm": 0.26274073123931885,
+      "learning_rate": 0.00019386124125437895,
+      "loss": 0.3464,
+      "step": 336
+    },
+    {
+      "epoch": 0.337094807914726,
+      "grad_norm": 0.27798980474472046,
+      "learning_rate": 0.00019382502798704935,
+      "loss": 0.3553,
+      "step": 337
+    },
+    {
+      "epoch": 0.33809508924384984,
+      "grad_norm": 0.3262801170349121,
+      "learning_rate": 0.00019378871162298227,
+      "loss": 0.4019,
+      "step": 338
+    },
+    {
+      "epoch": 0.3390953705729737,
+      "grad_norm": 0.27483808994293213,
+      "learning_rate": 0.00019375229220208276,
+      "loss": 0.3487,
+      "step": 339
+    },
+    {
+      "epoch": 0.34009565190209745,
+      "grad_norm": 0.38662320375442505,
+      "learning_rate": 0.00019371576976436917,
+      "loss": 0.4287,
+      "step": 340
+    },
+    {
+      "epoch": 0.3410959332312213,
+      "grad_norm": 0.2590675354003906,
+      "learning_rate": 0.00019367914434997312,
+      "loss": 0.3743,
+      "step": 341
+    },
+    {
+      "epoch": 0.3420962145603451,
+      "grad_norm": 0.2873009741306305,
+      "learning_rate": 0.00019364241599913924,
+      "loss": 0.3902,
+      "step": 342
+    },
+    {
+      "epoch": 0.3430964958894689,
+      "grad_norm": 0.3668614327907562,
+      "learning_rate": 0.0001936055847522254,
+      "loss": 0.3645,
+      "step": 343
+    },
+    {
+      "epoch": 0.3440967772185927,
+      "grad_norm": 0.306484580039978,
+      "learning_rate": 0.00019356865064970244,
+      "loss": 0.3919,
+      "step": 344
+    },
+    {
+      "epoch": 0.34509705854771655,
+      "grad_norm": 0.3255596458911896,
+      "learning_rate": 0.0001935316137321543,
+      "loss": 0.4359,
+      "step": 345
+    },
+    {
+      "epoch": 0.3460973398768404,
+      "grad_norm": 0.4163992702960968,
+      "learning_rate": 0.00019349447404027782,
+      "loss": 0.3574,
+      "step": 346
+    },
+    {
+      "epoch": 0.34709762120596416,
+      "grad_norm": 0.344198614358902,
+      "learning_rate": 0.00019345723161488283,
+      "loss": 0.4011,
+      "step": 347
+    },
+    {
+      "epoch": 0.348097902535088,
+      "grad_norm": 0.2813580334186554,
+      "learning_rate": 0.000193419886496892,
+      "loss": 0.3766,
+      "step": 348
+    },
+    {
+      "epoch": 0.3490981838642118,
+      "grad_norm": 0.27424079179763794,
+      "learning_rate": 0.00019338243872734086,
+      "loss": 0.3241,
+      "step": 349
+    },
+    {
+      "epoch": 0.35009846519333565,
+      "grad_norm": 0.2421925663948059,
+      "learning_rate": 0.00019334488834737775,
+      "loss": 0.3852,
+      "step": 350
+    },
+    {
+      "epoch": 0.3510987465224594,
+      "grad_norm": 0.22247883677482605,
+      "learning_rate": 0.00019330723539826375,
+      "loss": 0.3595,
+      "step": 351
+    },
+    {
+      "epoch": 0.35209902785158326,
+      "grad_norm": 0.26099297404289246,
+      "learning_rate": 0.00019326947992137262,
+      "loss": 0.3739,
+      "step": 352
+    },
+    {
+      "epoch": 0.3530993091807071,
+      "grad_norm": 0.28967341780662537,
+      "learning_rate": 0.00019323162195819082,
+      "loss": 0.3831,
+      "step": 353
+    },
+    {
+      "epoch": 0.35409959050983086,
+      "grad_norm": 0.2714492678642273,
+      "learning_rate": 0.0001931936615503174,
+      "loss": 0.3413,
+      "step": 354
+    },
+    {
+      "epoch": 0.3550998718389547,
+      "grad_norm": 0.27229174971580505,
+      "learning_rate": 0.000193155598739464,
+      "loss": 0.3825,
+      "step": 355
+    },
+    {
+      "epoch": 0.3561001531680785,
+      "grad_norm": 0.23672465980052948,
+      "learning_rate": 0.0001931174335674547,
+      "loss": 0.3694,
+      "step": 356
+    },
+    {
+      "epoch": 0.35710043449720236,
+      "grad_norm": 0.23558713495731354,
+      "learning_rate": 0.0001930791660762262,
+      "loss": 0.3629,
+      "step": 357
+    },
+    {
+      "epoch": 0.35810071582632613,
+      "grad_norm": 0.46677660942077637,
+      "learning_rate": 0.00019304079630782752,
+      "loss": 0.3946,
+      "step": 358
+    },
+    {
+      "epoch": 0.35910099715544996,
+      "grad_norm": 0.24028021097183228,
+      "learning_rate": 0.0001930023243044201,
+      "loss": 0.3605,
+      "step": 359
+    },
+    {
+      "epoch": 0.3601012784845738,
+      "grad_norm": 0.25813183188438416,
+      "learning_rate": 0.00019296375010827773,
+      "loss": 0.3983,
+      "step": 360
+    },
+    {
+      "epoch": 0.3611015598136976,
+      "grad_norm": 0.23249755799770355,
+      "learning_rate": 0.00019292507376178643,
+      "loss": 0.3963,
+      "step": 361
+    },
+    {
+      "epoch": 0.3621018411428214,
+      "grad_norm": 0.31035348773002625,
+      "learning_rate": 0.00019288629530744454,
+      "loss": 0.4103,
+      "step": 362
+    },
+    {
+      "epoch": 0.36310212247194523,
+      "grad_norm": 0.2520258128643036,
+      "learning_rate": 0.0001928474147878626,
+      "loss": 0.3408,
+      "step": 363
+    },
+    {
+      "epoch": 0.36410240380106906,
+      "grad_norm": 0.25639891624450684,
+      "learning_rate": 0.0001928084322457632,
+      "loss": 0.3771,
+      "step": 364
+    },
+    {
+      "epoch": 0.3651026851301929,
+      "grad_norm": 0.2763039171695709,
+      "learning_rate": 0.00019276934772398114,
+      "loss": 0.3385,
+      "step": 365
+    },
+    {
+      "epoch": 0.36610296645931667,
+      "grad_norm": 0.2602714002132416,
+      "learning_rate": 0.00019273016126546323,
+      "loss": 0.3836,
+      "step": 366
+    },
+    {
+      "epoch": 0.3671032477884405,
+      "grad_norm": 0.33483123779296875,
+      "learning_rate": 0.00019269087291326833,
+      "loss": 0.4129,
+      "step": 367
+    },
+    {
+      "epoch": 0.36810352911756433,
+      "grad_norm": 0.23084044456481934,
+      "learning_rate": 0.00019265148271056722,
+      "loss": 0.326,
+      "step": 368
+    },
+    {
+      "epoch": 0.3691038104466881,
+      "grad_norm": 0.26484981179237366,
+      "learning_rate": 0.0001926119907006426,
+      "loss": 0.3881,
+      "step": 369
+    },
+    {
+      "epoch": 0.37010409177581194,
+      "grad_norm": 0.24120277166366577,
+      "learning_rate": 0.00019257239692688907,
+      "loss": 0.3705,
+      "step": 370
+    },
+    {
+      "epoch": 0.37110437310493577,
+      "grad_norm": 0.23136964440345764,
+      "learning_rate": 0.00019253270143281296,
+      "loss": 0.3632,
+      "step": 371
+    },
+    {
+      "epoch": 0.3721046544340596,
+      "grad_norm": 0.24877014756202698,
+      "learning_rate": 0.00019249290426203252,
+      "loss": 0.3409,
+      "step": 372
+    },
+    {
+      "epoch": 0.3731049357631834,
+      "grad_norm": 0.23634536564350128,
+      "learning_rate": 0.0001924530054582776,
+      "loss": 0.3415,
+      "step": 373
+    },
+    {
+      "epoch": 0.3741052170923072,
+      "grad_norm": 0.6095355153083801,
+      "learning_rate": 0.0001924130050653898,
+      "loss": 0.3556,
+      "step": 374
+    },
+    {
+      "epoch": 0.37510549842143104,
+      "grad_norm": 0.2479625940322876,
+      "learning_rate": 0.00019237290312732226,
+      "loss": 0.345,
+      "step": 375
+    },
+    {
+      "epoch": 0.37610577975055487,
+      "grad_norm": 0.22326619923114777,
+      "learning_rate": 0.00019233269968813984,
+      "loss": 0.3442,
+      "step": 376
+    },
+    {
+      "epoch": 0.37710606107967864,
+      "grad_norm": 0.2376389503479004,
+      "learning_rate": 0.00019229239479201876,
+      "loss": 0.3251,
+      "step": 377
+    },
+    {
+      "epoch": 0.3781063424088025,
+      "grad_norm": 0.2835697829723358,
+      "learning_rate": 0.0001922519884832469,
+      "loss": 0.3834,
+      "step": 378
+    },
+    {
+      "epoch": 0.3791066237379263,
+      "grad_norm": 0.27211618423461914,
+      "learning_rate": 0.0001922114808062234,
+      "loss": 0.3642,
+      "step": 379
+    },
+    {
+      "epoch": 0.3801069050670501,
+      "grad_norm": 0.3268214166164398,
+      "learning_rate": 0.00019217087180545893,
+      "loss": 0.3909,
+      "step": 380
+    },
+    {
+      "epoch": 0.3811071863961739,
+      "grad_norm": 0.2260449230670929,
+      "learning_rate": 0.0001921301615255754,
+      "loss": 0.3796,
+      "step": 381
+    },
+    {
+      "epoch": 0.38210746772529774,
+      "grad_norm": 0.21646197140216827,
+      "learning_rate": 0.0001920893500113061,
+      "loss": 0.3455,
+      "step": 382
+    },
+    {
+      "epoch": 0.3831077490544216,
+      "grad_norm": 0.27988046407699585,
+      "learning_rate": 0.00019204843730749547,
+      "loss": 0.3443,
+      "step": 383
+    },
+    {
+      "epoch": 0.38410803038354535,
+      "grad_norm": 0.2434566169977188,
+      "learning_rate": 0.00019200742345909915,
+      "loss": 0.36,
+      "step": 384
+    },
+    {
+      "epoch": 0.3851083117126692,
+      "grad_norm": 0.21464578807353973,
+      "learning_rate": 0.00019196630851118398,
+      "loss": 0.3434,
+      "step": 385
+    },
+    {
+      "epoch": 0.386108593041793,
+      "grad_norm": 0.2649652063846588,
+      "learning_rate": 0.0001919250925089278,
+      "loss": 0.3438,
+      "step": 386
+    },
+    {
+      "epoch": 0.38710887437091684,
+      "grad_norm": 0.21409133076667786,
+      "learning_rate": 0.00019188377549761963,
+      "loss": 0.3675,
+      "step": 387
+    },
+    {
+      "epoch": 0.3881091557000406,
+      "grad_norm": 0.27577823400497437,
+      "learning_rate": 0.00019184235752265928,
+      "loss": 0.3486,
+      "step": 388
+    },
+    {
+      "epoch": 0.38910943702916445,
+      "grad_norm": 0.3003029227256775,
+      "learning_rate": 0.00019180083862955772,
+      "loss": 0.3968,
+      "step": 389
+    },
+    {
+      "epoch": 0.3901097183582883,
+      "grad_norm": 0.2916281521320343,
+      "learning_rate": 0.00019175921886393666,
+      "loss": 0.3835,
+      "step": 390
+    },
+    {
+      "epoch": 0.39110999968741206,
+      "grad_norm": 0.27574488520622253,
+      "learning_rate": 0.00019171749827152869,
+      "loss": 0.3958,
+      "step": 391
+    },
+    {
+      "epoch": 0.3921102810165359,
+      "grad_norm": 0.25350049138069153,
+      "learning_rate": 0.0001916756768981772,
+      "loss": 0.3579,
+      "step": 392
+    },
+    {
+      "epoch": 0.3931105623456597,
+      "grad_norm": 0.2273736596107483,
+      "learning_rate": 0.00019163375478983632,
+      "loss": 0.3348,
+      "step": 393
+    },
+    {
+      "epoch": 0.39411084367478355,
+      "grad_norm": 0.31927919387817383,
+      "learning_rate": 0.00019159173199257085,
+      "loss": 0.3714,
+      "step": 394
+    },
+    {
+      "epoch": 0.3951111250039073,
+      "grad_norm": 0.22905245423316956,
+      "learning_rate": 0.00019154960855255628,
+      "loss": 0.3707,
+      "step": 395
+    },
+    {
+      "epoch": 0.39611140633303116,
+      "grad_norm": 0.22161217033863068,
+      "learning_rate": 0.0001915073845160786,
+      "loss": 0.3591,
+      "step": 396
+    },
+    {
+      "epoch": 0.397111687662155,
+      "grad_norm": 0.24688982963562012,
+      "learning_rate": 0.00019146505992953446,
+      "loss": 0.3113,
+      "step": 397
+    },
+    {
+      "epoch": 0.3981119689912788,
+      "grad_norm": 0.3700356185436249,
+      "learning_rate": 0.00019142263483943085,
+      "loss": 0.3675,
+      "step": 398
+    },
+    {
+      "epoch": 0.3991122503204026,
+      "grad_norm": 0.2571461796760559,
+      "learning_rate": 0.00019138010929238534,
+      "loss": 0.3741,
+      "step": 399
+    },
+    {
+      "epoch": 0.4001125316495264,
+      "grad_norm": 0.25479117035865784,
+      "learning_rate": 0.00019133748333512575,
+      "loss": 0.3374,
+      "step": 400
+    },
+    {
+      "epoch": 0.40111281297865026,
+      "grad_norm": 0.300425261259079,
+      "learning_rate": 0.00019129475701449035,
+      "loss": 0.3541,
+      "step": 401
+    },
+    {
+      "epoch": 0.4021130943077741,
+      "grad_norm": 0.2358659952878952,
+      "learning_rate": 0.0001912519303774276,
+      "loss": 0.3808,
+      "step": 402
+    },
+    {
+      "epoch": 0.40311337563689786,
+      "grad_norm": 0.22720776498317719,
+      "learning_rate": 0.0001912090034709963,
+      "loss": 0.3359,
+      "step": 403
+    },
+    {
+      "epoch": 0.4041136569660217,
+      "grad_norm": 0.2396584004163742,
+      "learning_rate": 0.00019116597634236525,
+      "loss": 0.3397,
+      "step": 404
+    },
+    {
+      "epoch": 0.4051139382951455,
+      "grad_norm": 0.3370227813720703,
+      "learning_rate": 0.0001911228490388136,
+      "loss": 0.347,
+      "step": 405
+    },
+    {
+      "epoch": 0.4061142196242693,
+      "grad_norm": 0.22302855551242828,
+      "learning_rate": 0.00019107962160773035,
+      "loss": 0.3055,
+      "step": 406
+    },
+    {
+      "epoch": 0.40711450095339313,
+      "grad_norm": 0.2422298789024353,
+      "learning_rate": 0.0001910362940966147,
+      "loss": 0.3662,
+      "step": 407
+    },
+    {
+      "epoch": 0.40811478228251696,
+      "grad_norm": 0.24443741142749786,
+      "learning_rate": 0.00019099286655307568,
+      "loss": 0.3929,
+      "step": 408
+    },
+    {
+      "epoch": 0.4091150636116408,
+      "grad_norm": 0.2055700719356537,
+      "learning_rate": 0.0001909493390248324,
+      "loss": 0.3078,
+      "step": 409
+    },
+    {
+      "epoch": 0.41011534494076457,
+      "grad_norm": 0.2374294251203537,
+      "learning_rate": 0.00019090571155971366,
+      "loss": 0.353,
+      "step": 410
+    },
+    {
+      "epoch": 0.4111156262698884,
+      "grad_norm": 0.29268407821655273,
+      "learning_rate": 0.00019086198420565823,
+      "loss": 0.3204,
+      "step": 411
+    },
+    {
+      "epoch": 0.41211590759901223,
+      "grad_norm": 0.26289287209510803,
+      "learning_rate": 0.00019081815701071445,
+      "loss": 0.292,
+      "step": 412
+    },
+    {
+      "epoch": 0.41311618892813606,
+      "grad_norm": 0.2546895742416382,
+      "learning_rate": 0.0001907742300230406,
+      "loss": 0.3856,
+      "step": 413
+    },
+    {
+      "epoch": 0.41411647025725984,
+      "grad_norm": 0.23252642154693604,
+      "learning_rate": 0.00019073020329090444,
+      "loss": 0.3482,
+      "step": 414
+    },
+    {
+      "epoch": 0.41511675158638367,
+      "grad_norm": 0.2277359515428543,
+      "learning_rate": 0.0001906860768626834,
+      "loss": 0.3185,
+      "step": 415
+    },
+    {
+      "epoch": 0.4161170329155075,
+      "grad_norm": 0.2553969919681549,
+      "learning_rate": 0.00019064185078686443,
+      "loss": 0.3476,
+      "step": 416
+    },
+    {
+      "epoch": 0.4171173142446313,
+      "grad_norm": 0.26211032271385193,
+      "learning_rate": 0.000190597525112044,
+      "loss": 0.3506,
+      "step": 417
+    },
+    {
+      "epoch": 0.4181175955737551,
+      "grad_norm": 0.32664135098457336,
+      "learning_rate": 0.000190553099886928,
+      "loss": 0.3519,
+      "step": 418
+    },
+    {
+      "epoch": 0.41911787690287894,
+      "grad_norm": 0.2600671947002411,
+      "learning_rate": 0.00019050857516033173,
+      "loss": 0.3422,
+      "step": 419
+    },
+    {
+      "epoch": 0.42011815823200277,
+      "grad_norm": 0.24694529175758362,
+      "learning_rate": 0.00019046395098117983,
+      "loss": 0.3411,
+      "step": 420
+    },
+    {
+      "epoch": 0.42111843956112655,
+      "grad_norm": 0.25047555565834045,
+      "learning_rate": 0.00019041922739850616,
+      "loss": 0.3654,
+      "step": 421
+    },
+    {
+      "epoch": 0.4221187208902504,
+      "grad_norm": 0.2995339035987854,
+      "learning_rate": 0.00019037440446145385,
+      "loss": 0.4087,
+      "step": 422
+    },
+    {
+      "epoch": 0.4231190022193742,
+      "grad_norm": 0.25556957721710205,
+      "learning_rate": 0.00019032948221927524,
+      "loss": 0.3692,
+      "step": 423
+    },
+    {
+      "epoch": 0.42411928354849804,
+      "grad_norm": 0.26491332054138184,
+      "learning_rate": 0.00019028446072133175,
+      "loss": 0.3866,
+      "step": 424
+    },
+    {
+      "epoch": 0.4251195648776218,
+      "grad_norm": 0.23855310678482056,
+      "learning_rate": 0.00019023934001709383,
+      "loss": 0.329,
+      "step": 425
+    },
+    {
+      "epoch": 0.42611984620674564,
+      "grad_norm": 0.22823669016361237,
+      "learning_rate": 0.00019019412015614098,
+      "loss": 0.3594,
+      "step": 426
+    },
+    {
+      "epoch": 0.4271201275358695,
+      "grad_norm": 0.3833972215652466,
+      "learning_rate": 0.00019014880118816164,
+      "loss": 0.3786,
+      "step": 427
+    },
+    {
+      "epoch": 0.4281204088649933,
+      "grad_norm": 0.2663598954677582,
+      "learning_rate": 0.0001901033831629532,
+      "loss": 0.3619,
+      "step": 428
+    },
+    {
+      "epoch": 0.4291206901941171,
+      "grad_norm": 0.28205564618110657,
+      "learning_rate": 0.00019005786613042185,
+      "loss": 0.3879,
+      "step": 429
+    },
+    {
+      "epoch": 0.4301209715232409,
+      "grad_norm": 0.2598055899143219,
+      "learning_rate": 0.00019001225014058255,
+      "loss": 0.3448,
+      "step": 430
+    },
+    {
+      "epoch": 0.43112125285236474,
+      "grad_norm": 0.279117614030838,
+      "learning_rate": 0.00018996653524355902,
+      "loss": 0.3787,
+      "step": 431
+    },
+    {
+      "epoch": 0.4321215341814885,
+      "grad_norm": 0.22764593362808228,
+      "learning_rate": 0.00018992072148958368,
+      "loss": 0.3862,
+      "step": 432
+    },
+    {
+      "epoch": 0.43312181551061235,
+      "grad_norm": 0.38709068298339844,
+      "learning_rate": 0.00018987480892899758,
+      "loss": 0.3459,
+      "step": 433
+    },
+    {
+      "epoch": 0.4341220968397362,
+      "grad_norm": 0.2217923253774643,
+      "learning_rate": 0.00018982879761225027,
+      "loss": 0.3517,
+      "step": 434
+    },
+    {
+      "epoch": 0.43512237816886,
+      "grad_norm": 0.23652343451976776,
+      "learning_rate": 0.00018978268758989991,
+      "loss": 0.3644,
+      "step": 435
+    },
+    {
+      "epoch": 0.4361226594979838,
+      "grad_norm": 0.5657463669776917,
+      "learning_rate": 0.00018973647891261307,
+      "loss": 0.3325,
+      "step": 436
+    },
+    {
+      "epoch": 0.4371229408271076,
+      "grad_norm": 0.32180288434028625,
+      "learning_rate": 0.00018969017163116472,
+      "loss": 0.3864,
+      "step": 437
+    },
+    {
+      "epoch": 0.43812322215623145,
+      "grad_norm": 0.3013526201248169,
+      "learning_rate": 0.0001896437657964382,
+      "loss": 0.4002,
+      "step": 438
+    },
+    {
+      "epoch": 0.4391235034853553,
+      "grad_norm": 0.26990970969200134,
+      "learning_rate": 0.00018959726145942508,
+      "loss": 0.3331,
+      "step": 439
+    },
+    {
+      "epoch": 0.44012378481447906,
+      "grad_norm": 0.19621394574642181,
+      "learning_rate": 0.00018955065867122528,
+      "loss": 0.2931,
+      "step": 440
+    },
+    {
+      "epoch": 0.4411240661436029,
+      "grad_norm": 0.2245367169380188,
+      "learning_rate": 0.00018950395748304678,
+      "loss": 0.333,
+      "step": 441
+    },
+    {
+      "epoch": 0.4421243474727267,
+      "grad_norm": 0.2619412839412689,
+      "learning_rate": 0.0001894571579462058,
+      "loss": 0.3442,
+      "step": 442
+    },
+    {
+      "epoch": 0.4431246288018505,
+      "grad_norm": 0.26194262504577637,
+      "learning_rate": 0.00018941026011212654,
+      "loss": 0.3369,
+      "step": 443
+    },
+    {
+      "epoch": 0.4441249101309743,
+      "grad_norm": 0.2473272681236267,
+      "learning_rate": 0.00018936326403234125,
+      "loss": 0.3724,
+      "step": 444
+    },
+    {
+      "epoch": 0.44512519146009816,
+      "grad_norm": 0.32609519362449646,
+      "learning_rate": 0.00018931616975849006,
+      "loss": 0.3933,
+      "step": 445
+    },
+    {
+      "epoch": 0.446125472789222,
+      "grad_norm": 0.3470418155193329,
+      "learning_rate": 0.00018926897734232115,
+      "loss": 0.4049,
+      "step": 446
+    },
+    {
+      "epoch": 0.44712575411834576,
+      "grad_norm": 0.2971716821193695,
+      "learning_rate": 0.0001892216868356904,
+      "loss": 0.3469,
+      "step": 447
+    },
+    {
+      "epoch": 0.4481260354474696,
+      "grad_norm": 0.24874372780323029,
+      "learning_rate": 0.0001891742982905615,
+      "loss": 0.3531,
+      "step": 448
+    },
+    {
+      "epoch": 0.4491263167765934,
+      "grad_norm": 0.2729557752609253,
+      "learning_rate": 0.00018912681175900598,
+      "loss": 0.3627,
+      "step": 449
+    },
+    {
+      "epoch": 0.45012659810571726,
+      "grad_norm": 0.2681429982185364,
+      "learning_rate": 0.00018907922729320285,
+      "loss": 0.3675,
+      "step": 450
+    },
+    {
+      "epoch": 0.45112687943484103,
+      "grad_norm": 0.2941266894340515,
+      "learning_rate": 0.00018903154494543889,
+      "loss": 0.4168,
+      "step": 451
+    },
+    {
+      "epoch": 0.45212716076396486,
+      "grad_norm": 0.24707025289535522,
+      "learning_rate": 0.00018898376476810834,
+      "loss": 0.3437,
+      "step": 452
+    },
+    {
+      "epoch": 0.4531274420930887,
+      "grad_norm": 0.3138103187084198,
+      "learning_rate": 0.00018893588681371303,
+      "loss": 0.3738,
+      "step": 453
+    },
+    {
+      "epoch": 0.45412772342221247,
+      "grad_norm": 0.32942667603492737,
+      "learning_rate": 0.00018888791113486213,
+      "loss": 0.4186,
+      "step": 454
+    },
+    {
+      "epoch": 0.4551280047513363,
+      "grad_norm": 0.26734960079193115,
+      "learning_rate": 0.00018883983778427227,
+      "loss": 0.3519,
+      "step": 455
+    },
+    {
+      "epoch": 0.45612828608046013,
+      "grad_norm": 0.23201091587543488,
+      "learning_rate": 0.0001887916668147673,
+      "loss": 0.3613,
+      "step": 456
+    },
+    {
+      "epoch": 0.45712856740958396,
+      "grad_norm": 0.2691316306591034,
+      "learning_rate": 0.00018874339827927846,
+      "loss": 0.3924,
+      "step": 457
+    },
+    {
+      "epoch": 0.45812884873870774,
+      "grad_norm": 0.27234920859336853,
+      "learning_rate": 0.00018869503223084414,
+      "loss": 0.3704,
+      "step": 458
+    },
+    {
+      "epoch": 0.45912913006783157,
+      "grad_norm": 0.24237160384655,
+      "learning_rate": 0.00018864656872260985,
+      "loss": 0.3009,
+      "step": 459
+    },
+    {
+      "epoch": 0.4601294113969554,
+      "grad_norm": 0.2326439768075943,
+      "learning_rate": 0.00018859800780782828,
+      "loss": 0.3243,
+      "step": 460
+    },
+    {
+      "epoch": 0.46112969272607923,
+      "grad_norm": 0.250873863697052,
+      "learning_rate": 0.000188549349539859,
+      "loss": 0.3778,
+      "step": 461
+    },
+    {
+      "epoch": 0.462129974055203,
+      "grad_norm": 0.22086897492408752,
+      "learning_rate": 0.00018850059397216876,
+      "loss": 0.3322,
+      "step": 462
+    },
+    {
+      "epoch": 0.46313025538432684,
+      "grad_norm": 0.20893535017967224,
+      "learning_rate": 0.00018845174115833099,
+      "loss": 0.3278,
+      "step": 463
+    },
+    {
+      "epoch": 0.46413053671345067,
+      "grad_norm": 0.20041684806346893,
+      "learning_rate": 0.0001884027911520262,
+      "loss": 0.3055,
+      "step": 464
+    },
+    {
+      "epoch": 0.4651308180425745,
+      "grad_norm": 0.30201271176338196,
+      "learning_rate": 0.00018835374400704154,
+      "loss": 0.3257,
+      "step": 465
+    },
+    {
+      "epoch": 0.4661310993716983,
+      "grad_norm": 0.23991478979587555,
+      "learning_rate": 0.00018830459977727096,
+      "loss": 0.3396,
+      "step": 466
+    },
+    {
+      "epoch": 0.4671313807008221,
+      "grad_norm": 0.22463761270046234,
+      "learning_rate": 0.0001882553585167151,
+      "loss": 0.2841,
+      "step": 467
+    },
+    {
+      "epoch": 0.46813166202994594,
+      "grad_norm": 0.24309033155441284,
+      "learning_rate": 0.00018820602027948114,
+      "loss": 0.332,
+      "step": 468
+    },
+    {
+      "epoch": 0.4691319433590697,
+      "grad_norm": 0.24418720602989197,
+      "learning_rate": 0.00018815658511978298,
+      "loss": 0.3274,
+      "step": 469
+    },
+    {
+      "epoch": 0.47013222468819355,
+      "grad_norm": 0.2473025619983673,
+      "learning_rate": 0.00018810705309194083,
+      "loss": 0.3641,
+      "step": 470
+    },
+    {
+      "epoch": 0.4711325060173174,
+      "grad_norm": 0.22359922528266907,
+      "learning_rate": 0.00018805742425038145,
+      "loss": 0.3158,
+      "step": 471
+    },
+    {
+      "epoch": 0.4721327873464412,
+      "grad_norm": 0.25090882182121277,
+      "learning_rate": 0.00018800769864963802,
+      "loss": 0.3622,
+      "step": 472
+    },
+    {
+      "epoch": 0.473133068675565,
+      "grad_norm": 0.22978092730045319,
+      "learning_rate": 0.00018795787634434994,
+      "loss": 0.3546,
+      "step": 473
+    },
+    {
+      "epoch": 0.4741333500046888,
+      "grad_norm": 0.2290024757385254,
+      "learning_rate": 0.0001879079573892629,
+      "loss": 0.3277,
+      "step": 474
+    },
+    {
+      "epoch": 0.47513363133381264,
+      "grad_norm": 0.2646024525165558,
+      "learning_rate": 0.00018785794183922883,
+      "loss": 0.3446,
+      "step": 475
+    },
+    {
+      "epoch": 0.4761339126629365,
+      "grad_norm": 0.3361208140850067,
+      "learning_rate": 0.00018780782974920572,
+      "loss": 0.3321,
+      "step": 476
+    },
+    {
+      "epoch": 0.47713419399206025,
+      "grad_norm": 0.23654255270957947,
+      "learning_rate": 0.00018775762117425777,
+      "loss": 0.3294,
+      "step": 477
+    },
+    {
+      "epoch": 0.4781344753211841,
+      "grad_norm": 0.266557902097702,
+      "learning_rate": 0.0001877073161695551,
+      "loss": 0.363,
+      "step": 478
+    },
+    {
+      "epoch": 0.4791347566503079,
+      "grad_norm": 0.32463106513023376,
+      "learning_rate": 0.00018765691479037376,
+      "loss": 0.3751,
+      "step": 479
+    },
+    {
+      "epoch": 0.4801350379794317,
+      "grad_norm": 0.24274414777755737,
+      "learning_rate": 0.00018760641709209583,
+      "loss": 0.3638,
+      "step": 480
+    },
+    {
+      "epoch": 0.4811353193085555,
+      "grad_norm": 0.2479117512702942,
+      "learning_rate": 0.0001875558231302091,
+      "loss": 0.3269,
+      "step": 481
+    },
+    {
+      "epoch": 0.48213560063767935,
+      "grad_norm": 0.24928362667560577,
+      "learning_rate": 0.00018750513296030718,
+      "loss": 0.3452,
+      "step": 482
+    },
+    {
+      "epoch": 0.4831358819668032,
+      "grad_norm": 0.2932121753692627,
+      "learning_rate": 0.00018745434663808942,
+      "loss": 0.4229,
+      "step": 483
+    },
+    {
+      "epoch": 0.48413616329592696,
+      "grad_norm": 0.3067155182361603,
+      "learning_rate": 0.0001874034642193608,
+      "loss": 0.3521,
+      "step": 484
+    },
+    {
+      "epoch": 0.4851364446250508,
+      "grad_norm": 0.29176706075668335,
+      "learning_rate": 0.0001873524857600319,
+      "loss": 0.3618,
+      "step": 485
+    },
+    {
+      "epoch": 0.4861367259541746,
+      "grad_norm": 0.22693835198879242,
+      "learning_rate": 0.00018730141131611882,
+      "loss": 0.3389,
+      "step": 486
+    },
+    {
+      "epoch": 0.48713700728329845,
+      "grad_norm": 0.2182817906141281,
+      "learning_rate": 0.00018725024094374315,
+      "loss": 0.3478,
+      "step": 487
+    },
+    {
+      "epoch": 0.4881372886124222,
+      "grad_norm": 0.2273852825164795,
+      "learning_rate": 0.00018719897469913184,
+      "loss": 0.3236,
+      "step": 488
+    },
+    {
+      "epoch": 0.48913756994154606,
+      "grad_norm": 0.24996232986450195,
+      "learning_rate": 0.00018714761263861728,
+      "loss": 0.3005,
+      "step": 489
+    },
+    {
+      "epoch": 0.4901378512706699,
+      "grad_norm": 0.47928324341773987,
+      "learning_rate": 0.000187096154818637,
+      "loss": 0.3412,
+      "step": 490
+    },
+    {
+      "epoch": 0.4911381325997937,
+      "grad_norm": 0.22586360573768616,
+      "learning_rate": 0.00018704460129573391,
+      "loss": 0.3113,
+      "step": 491
+    },
+    {
+      "epoch": 0.4921384139289175,
+      "grad_norm": 0.23548360168933868,
+      "learning_rate": 0.00018699295212655596,
+      "loss": 0.3246,
+      "step": 492
+    },
+    {
+      "epoch": 0.4931386952580413,
+      "grad_norm": 0.26637163758277893,
+      "learning_rate": 0.00018694120736785632,
+      "loss": 0.3434,
+      "step": 493
+    },
+    {
+      "epoch": 0.49413897658716516,
+      "grad_norm": 0.2589665949344635,
+      "learning_rate": 0.00018688936707649304,
+      "loss": 0.3459,
+      "step": 494
+    },
+    {
+      "epoch": 0.49513925791628893,
+      "grad_norm": 0.24287012219429016,
+      "learning_rate": 0.00018683743130942928,
+      "loss": 0.3509,
+      "step": 495
+    },
+    {
+      "epoch": 0.49613953924541276,
+      "grad_norm": 0.2158239781856537,
+      "learning_rate": 0.00018678540012373302,
+      "loss": 0.3393,
+      "step": 496
+    },
+    {
+      "epoch": 0.4971398205745366,
+      "grad_norm": 0.2196134328842163,
+      "learning_rate": 0.00018673327357657715,
+      "loss": 0.3316,
+      "step": 497
+    },
+    {
+      "epoch": 0.4981401019036604,
+      "grad_norm": 0.29125913977622986,
+      "learning_rate": 0.0001866810517252393,
+      "loss": 0.3905,
+      "step": 498
+    },
+    {
+      "epoch": 0.4991403832327842,
+      "grad_norm": 0.31574273109436035,
+      "learning_rate": 0.00018662873462710184,
+      "loss": 0.356,
+      "step": 499
+    },
+    {
+      "epoch": 0.5001406645619081,
+      "grad_norm": 0.2551417946815491,
+      "learning_rate": 0.0001865763223396518,
+      "loss": 0.354,
+      "step": 500
+    },
+    {
+      "epoch": 0.5011409458910319,
+      "grad_norm": 0.24650706350803375,
+      "learning_rate": 0.00018652381492048083,
+      "loss": 0.3501,
+      "step": 501
+    },
+    {
+      "epoch": 0.5021412272201556,
+      "grad_norm": 0.2927659749984741,
+      "learning_rate": 0.00018647121242728506,
+      "loss": 0.345,
+      "step": 502
+    },
+    {
+      "epoch": 0.5031415085492795,
+      "grad_norm": 0.24049070477485657,
+      "learning_rate": 0.00018641851491786512,
+      "loss": 0.3399,
+      "step": 503
+    },
+    {
+      "epoch": 0.5041417898784033,
+      "grad_norm": 0.31743761897087097,
+      "learning_rate": 0.00018636572245012606,
+      "loss": 0.3688,
+      "step": 504
+    },
+    {
+      "epoch": 0.5051420712075271,
+      "grad_norm": 0.22832700610160828,
+      "learning_rate": 0.00018631283508207725,
+      "loss": 0.3222,
+      "step": 505
+    },
+    {
+      "epoch": 0.506142352536651,
+      "grad_norm": 0.2232360690832138,
+      "learning_rate": 0.00018625985287183233,
+      "loss": 0.3686,
+      "step": 506
+    },
+    {
+      "epoch": 0.5071426338657747,
+      "grad_norm": 0.280877023935318,
+      "learning_rate": 0.00018620677587760916,
+      "loss": 0.351,
+      "step": 507
+    },
+    {
+      "epoch": 0.5081429151948985,
+      "grad_norm": 0.2949641942977905,
+      "learning_rate": 0.00018615360415772978,
+      "loss": 0.395,
+      "step": 508
+    },
+    {
+      "epoch": 0.5091431965240224,
+      "grad_norm": 0.2326454520225525,
+      "learning_rate": 0.00018610033777062025,
+      "loss": 0.3218,
+      "step": 509
+    },
+    {
+      "epoch": 0.5101434778531462,
+      "grad_norm": 0.2455054670572281,
+      "learning_rate": 0.0001860469767748108,
+      "loss": 0.3788,
+      "step": 510
+    },
+    {
+      "epoch": 0.5111437591822701,
+      "grad_norm": 0.23465023934841156,
+      "learning_rate": 0.00018599352122893539,
+      "loss": 0.3552,
+      "step": 511
+    },
+    {
+      "epoch": 0.5121440405113938,
+      "grad_norm": 0.23088668286800385,
+      "learning_rate": 0.00018593997119173205,
+      "loss": 0.3532,
+      "step": 512
+    },
+    {
+      "epoch": 0.5131443218405176,
+      "grad_norm": 0.23052315413951874,
+      "learning_rate": 0.00018588632672204264,
+      "loss": 0.3019,
+      "step": 513
+    },
+    {
+      "epoch": 0.5141446031696415,
+      "grad_norm": 0.22070975601673126,
+      "learning_rate": 0.0001858325878788126,
+      "loss": 0.3158,
+      "step": 514
+    },
+    {
+      "epoch": 0.5151448844987653,
+      "grad_norm": 0.23837679624557495,
+      "learning_rate": 0.00018577875472109134,
+      "loss": 0.3306,
+      "step": 515
+    },
+    {
+      "epoch": 0.516145165827889,
+      "grad_norm": 0.28564974665641785,
+      "learning_rate": 0.0001857248273080317,
+      "loss": 0.3297,
+      "step": 516
+    },
+    {
+      "epoch": 0.5171454471570129,
+      "grad_norm": 0.2735040783882141,
+      "learning_rate": 0.00018567080569889015,
+      "loss": 0.3713,
+      "step": 517
+    },
+    {
+      "epoch": 0.5181457284861367,
+      "grad_norm": 0.26505136489868164,
+      "learning_rate": 0.00018561668995302667,
+      "loss": 0.338,
+      "step": 518
+    },
+    {
+      "epoch": 0.5191460098152605,
+      "grad_norm": 0.2347506731748581,
+      "learning_rate": 0.00018556248012990468,
+      "loss": 0.3131,
+      "step": 519
+    },
+    {
+      "epoch": 0.5201462911443844,
+      "grad_norm": 0.2417895495891571,
+      "learning_rate": 0.000185508176289091,
+      "loss": 0.3264,
+      "step": 520
+    },
+    {
+      "epoch": 0.5211465724735082,
+      "grad_norm": 0.25715285539627075,
+      "learning_rate": 0.00018545377849025566,
+      "loss": 0.3195,
+      "step": 521
+    },
+    {
+      "epoch": 0.522146853802632,
+      "grad_norm": 0.24137462675571442,
+      "learning_rate": 0.0001853992867931721,
+      "loss": 0.3445,
+      "step": 522
+    },
+    {
+      "epoch": 0.5231471351317558,
+      "grad_norm": 0.22217655181884766,
+      "learning_rate": 0.00018534470125771674,
+      "loss": 0.3438,
+      "step": 523
+    },
+    {
+      "epoch": 0.5241474164608796,
+      "grad_norm": 0.2817876935005188,
+      "learning_rate": 0.0001852900219438693,
+      "loss": 0.3443,
+      "step": 524
+    },
+    {
+      "epoch": 0.5251476977900035,
+      "grad_norm": 0.49497222900390625,
+      "learning_rate": 0.0001852352489117124,
+      "loss": 0.313,
+      "step": 525
+    },
+    {
+      "epoch": 0.5261479791191273,
+      "grad_norm": 0.20260442793369293,
+      "learning_rate": 0.00018518038222143174,
+      "loss": 0.3224,
+      "step": 526
+    },
+    {
+      "epoch": 0.527148260448251,
+      "grad_norm": 0.23660187423229218,
+      "learning_rate": 0.00018512542193331583,
+      "loss": 0.3615,
+      "step": 527
+    },
+    {
+      "epoch": 0.5281485417773749,
+      "grad_norm": 0.2505345642566681,
+      "learning_rate": 0.00018507036810775615,
+      "loss": 0.3294,
+      "step": 528
+    },
+    {
+      "epoch": 0.5291488231064987,
+      "grad_norm": 0.22744148969650269,
+      "learning_rate": 0.00018501522080524688,
+      "loss": 0.3687,
+      "step": 529
+    },
+    {
+      "epoch": 0.5301491044356225,
+      "grad_norm": 0.23085086047649384,
+      "learning_rate": 0.0001849599800863849,
+      "loss": 0.3681,
+      "step": 530
+    },
+    {
+      "epoch": 0.5311493857647464,
+      "grad_norm": 0.25152313709259033,
+      "learning_rate": 0.0001849046460118698,
+      "loss": 0.3505,
+      "step": 531
+    },
+    {
+      "epoch": 0.5321496670938701,
+      "grad_norm": 0.2753462791442871,
+      "learning_rate": 0.0001848492186425037,
+      "loss": 0.294,
+      "step": 532
+    },
+    {
+      "epoch": 0.533149948422994,
+      "grad_norm": 0.2085839956998825,
+      "learning_rate": 0.0001847936980391913,
+      "loss": 0.3122,
+      "step": 533
+    },
+    {
+      "epoch": 0.5341502297521178,
+      "grad_norm": 0.28155985474586487,
+      "learning_rate": 0.00018473808426293964,
+      "loss": 0.3436,
+      "step": 534
+    },
+    {
+      "epoch": 0.5351505110812416,
+      "grad_norm": 0.22050590813159943,
+      "learning_rate": 0.00018468237737485823,
+      "loss": 0.3378,
+      "step": 535
+    },
+    {
+      "epoch": 0.5361507924103655,
+      "grad_norm": 1.025260329246521,
+      "learning_rate": 0.00018462657743615888,
+      "loss": 0.3143,
+      "step": 536
+    },
+    {
+      "epoch": 0.5371510737394892,
+      "grad_norm": 0.2612319886684418,
+      "learning_rate": 0.00018457068450815562,
+      "loss": 0.3553,
+      "step": 537
+    },
+    {
+      "epoch": 0.538151355068613,
+      "grad_norm": 0.23278780281543732,
+      "learning_rate": 0.00018451469865226464,
+      "loss": 0.3525,
+      "step": 538
+    },
+    {
+      "epoch": 0.5391516363977369,
+      "grad_norm": 0.22587038576602936,
+      "learning_rate": 0.00018445861993000436,
+      "loss": 0.3388,
+      "step": 539
+    },
+    {
+      "epoch": 0.5401519177268607,
+      "grad_norm": 0.2605971097946167,
+      "learning_rate": 0.00018440244840299506,
+      "loss": 0.3552,
+      "step": 540
+    },
+    {
+      "epoch": 0.5411521990559846,
+      "grad_norm": 0.2513737678527832,
+      "learning_rate": 0.0001843461841329591,
+      "loss": 0.3519,
+      "step": 541
+    },
+    {
+      "epoch": 0.5421524803851083,
+      "grad_norm": 0.2181878238916397,
+      "learning_rate": 0.0001842898271817208,
+      "loss": 0.2949,
+      "step": 542
+    },
+    {
+      "epoch": 0.5431527617142321,
+      "grad_norm": 0.23953957855701447,
+      "learning_rate": 0.00018423337761120618,
+      "loss": 0.3083,
+      "step": 543
+    },
+    {
+      "epoch": 0.544153043043356,
+      "grad_norm": 0.23563261330127716,
+      "learning_rate": 0.00018417683548344318,
+      "loss": 0.3721,
+      "step": 544
+    },
+    {
+      "epoch": 0.5451533243724798,
+      "grad_norm": 0.22247374057769775,
+      "learning_rate": 0.00018412020086056133,
+      "loss": 0.3142,
+      "step": 545
+    },
+    {
+      "epoch": 0.5461536057016035,
+      "grad_norm": 0.21039196848869324,
+      "learning_rate": 0.0001840634738047918,
+      "loss": 0.3168,
+      "step": 546
+    },
+    {
+      "epoch": 0.5471538870307274,
+      "grad_norm": 0.28400999307632446,
+      "learning_rate": 0.0001840066543784675,
+      "loss": 0.3317,
+      "step": 547
+    },
+    {
+      "epoch": 0.5481541683598512,
+      "grad_norm": 0.2570148706436157,
+      "learning_rate": 0.00018394974264402257,
+      "loss": 0.3511,
+      "step": 548
+    },
+    {
+      "epoch": 0.549154449688975,
+      "grad_norm": 0.3795485198497772,
+      "learning_rate": 0.00018389273866399275,
+      "loss": 0.3787,
+      "step": 549
+    },
+    {
+      "epoch": 0.5501547310180989,
+      "grad_norm": 0.20879346132278442,
+      "learning_rate": 0.00018383564250101512,
+      "loss": 0.3108,
+      "step": 550
+    },
+    {
+      "epoch": 0.5511550123472226,
+      "grad_norm": 0.2465057373046875,
+      "learning_rate": 0.000183778454217828,
+      "loss": 0.352,
+      "step": 551
+    },
+    {
+      "epoch": 0.5521552936763465,
+      "grad_norm": 0.21287333965301514,
+      "learning_rate": 0.0001837211738772711,
+      "loss": 0.3322,
+      "step": 552
+    },
+    {
+      "epoch": 0.5531555750054703,
+      "grad_norm": 0.24229544401168823,
+      "learning_rate": 0.000183663801542285,
+      "loss": 0.3524,
+      "step": 553
+    },
+    {
+      "epoch": 0.5541558563345941,
+      "grad_norm": 0.2168058156967163,
+      "learning_rate": 0.00018360633727591155,
+      "loss": 0.3234,
+      "step": 554
+    },
+    {
+      "epoch": 0.555156137663718,
+      "grad_norm": 0.27349379658699036,
+      "learning_rate": 0.00018354878114129367,
+      "loss": 0.3142,
+      "step": 555
+    },
+    {
+      "epoch": 0.5561564189928417,
+      "grad_norm": 0.19927184283733368,
+      "learning_rate": 0.00018349113320167504,
+      "loss": 0.3241,
+      "step": 556
+    },
+    {
+      "epoch": 0.5571567003219655,
+      "grad_norm": 0.245859295129776,
+      "learning_rate": 0.00018343339352040042,
+      "loss": 0.3619,
+      "step": 557
+    },
+    {
+      "epoch": 0.5581569816510894,
+      "grad_norm": 0.21804504096508026,
+      "learning_rate": 0.00018337556216091517,
+      "loss": 0.314,
+      "step": 558
+    },
+    {
+      "epoch": 0.5591572629802132,
+      "grad_norm": 0.2590177357196808,
+      "learning_rate": 0.00018331763918676556,
+      "loss": 0.3607,
+      "step": 559
+    },
+    {
+      "epoch": 0.560157544309337,
+      "grad_norm": 0.2252151519060135,
+      "learning_rate": 0.00018325962466159848,
+      "loss": 0.3274,
+      "step": 560
+    },
+    {
+      "epoch": 0.5611578256384608,
+      "grad_norm": 0.2471918910741806,
+      "learning_rate": 0.00018320151864916135,
+      "loss": 0.3736,
+      "step": 561
+    },
+    {
+      "epoch": 0.5621581069675846,
+      "grad_norm": 0.22999118268489838,
+      "learning_rate": 0.00018314332121330225,
+      "loss": 0.3339,
+      "step": 562
+    },
+    {
+      "epoch": 0.5631583882967085,
+      "grad_norm": 0.2884382903575897,
+      "learning_rate": 0.0001830850324179695,
+      "loss": 0.354,
+      "step": 563
+    },
+    {
+      "epoch": 0.5641586696258323,
+      "grad_norm": 0.28535163402557373,
+      "learning_rate": 0.00018302665232721208,
+      "loss": 0.3752,
+      "step": 564
+    },
+    {
+      "epoch": 0.565158950954956,
+      "grad_norm": 0.2396842986345291,
+      "learning_rate": 0.0001829681810051791,
+      "loss": 0.35,
+      "step": 565
+    },
+    {
+      "epoch": 0.5661592322840799,
+      "grad_norm": 0.23295581340789795,
+      "learning_rate": 0.00018290961851611995,
+      "loss": 0.3341,
+      "step": 566
+    },
+    {
+      "epoch": 0.5671595136132037,
+      "grad_norm": 0.21623344719409943,
+      "learning_rate": 0.00018285096492438424,
+      "loss": 0.3304,
+      "step": 567
+    },
+    {
+      "epoch": 0.5681597949423275,
+      "grad_norm": 0.2423194944858551,
+      "learning_rate": 0.00018279222029442163,
+      "loss": 0.3272,
+      "step": 568
+    },
+    {
+      "epoch": 0.5691600762714514,
+      "grad_norm": 0.2847912311553955,
+      "learning_rate": 0.00018273338469078186,
+      "loss": 0.3501,
+      "step": 569
+    },
+    {
+      "epoch": 0.5701603576005752,
+      "grad_norm": 0.2251608967781067,
+      "learning_rate": 0.00018267445817811466,
+      "loss": 0.3368,
+      "step": 570
+    },
+    {
+      "epoch": 0.5711606389296989,
+      "grad_norm": 0.22282929718494415,
+      "learning_rate": 0.00018261544082116954,
+      "loss": 0.3269,
+      "step": 571
+    },
+    {
+      "epoch": 0.5721609202588228,
+      "grad_norm": 0.23604530096054077,
+      "learning_rate": 0.00018255633268479595,
+      "loss": 0.3411,
+      "step": 572
+    },
+    {
+      "epoch": 0.5731612015879466,
+      "grad_norm": 0.26834771037101746,
+      "learning_rate": 0.00018249713383394303,
+      "loss": 0.3504,
+      "step": 573
+    },
+    {
+      "epoch": 0.5741614829170705,
+      "grad_norm": 0.2110867202281952,
+      "learning_rate": 0.0001824378443336596,
+      "loss": 0.2986,
+      "step": 574
+    },
+    {
+      "epoch": 0.5751617642461943,
+      "grad_norm": 0.30523523688316345,
+      "learning_rate": 0.00018237846424909413,
+      "loss": 0.3564,
+      "step": 575
+    },
+    {
+      "epoch": 0.576162045575318,
+      "grad_norm": 0.24037593603134155,
+      "learning_rate": 0.00018231899364549455,
+      "loss": 0.3258,
+      "step": 576
+    },
+    {
+      "epoch": 0.5771623269044419,
+      "grad_norm": 0.2534974217414856,
+      "learning_rate": 0.00018225943258820833,
+      "loss": 0.3154,
+      "step": 577
+    },
+    {
+      "epoch": 0.5781626082335657,
+      "grad_norm": 0.30275171995162964,
+      "learning_rate": 0.00018219978114268227,
+      "loss": 0.4134,
+      "step": 578
+    },
+    {
+      "epoch": 0.5791628895626895,
+      "grad_norm": 0.23564928770065308,
+      "learning_rate": 0.00018214003937446253,
+      "loss": 0.3006,
+      "step": 579
+    },
+    {
+      "epoch": 0.5801631708918134,
+      "grad_norm": 0.2837570309638977,
+      "learning_rate": 0.00018208020734919455,
+      "loss": 0.3188,
+      "step": 580
+    },
+    {
+      "epoch": 0.5811634522209371,
+      "grad_norm": 0.23933278024196625,
+      "learning_rate": 0.00018202028513262288,
+      "loss": 0.3817,
+      "step": 581
+    },
+    {
+      "epoch": 0.5821637335500609,
+      "grad_norm": 0.23992188274860382,
+      "learning_rate": 0.00018196027279059117,
+      "loss": 0.3394,
+      "step": 582
+    },
+    {
+      "epoch": 0.5831640148791848,
+      "grad_norm": 1.3730578422546387,
+      "learning_rate": 0.00018190017038904215,
+      "loss": 0.3569,
+      "step": 583
+    },
+    {
+      "epoch": 0.5841642962083086,
+      "grad_norm": 0.26228848099708557,
+      "learning_rate": 0.0001818399779940175,
+      "loss": 0.3353,
+      "step": 584
+    },
+    {
+      "epoch": 0.5851645775374325,
+      "grad_norm": 0.20937691628932953,
+      "learning_rate": 0.0001817796956716578,
+      "loss": 0.3263,
+      "step": 585
+    },
+    {
+      "epoch": 0.5861648588665562,
+      "grad_norm": 0.20607903599739075,
+      "learning_rate": 0.00018171932348820234,
+      "loss": 0.2773,
+      "step": 586
+    },
+    {
+      "epoch": 0.58716514019568,
+      "grad_norm": 0.22859016060829163,
+      "learning_rate": 0.0001816588615099893,
+      "loss": 0.356,
+      "step": 587
+    },
+    {
+      "epoch": 0.5881654215248039,
+      "grad_norm": 0.2487163245677948,
+      "learning_rate": 0.00018159830980345548,
+      "loss": 0.3185,
+      "step": 588
+    },
+    {
+      "epoch": 0.5891657028539277,
+      "grad_norm": 0.21327054500579834,
+      "learning_rate": 0.0001815376684351362,
+      "loss": 0.3512,
+      "step": 589
+    },
+    {
+      "epoch": 0.5901659841830514,
+      "grad_norm": 0.2244402915239334,
+      "learning_rate": 0.00018147693747166534,
+      "loss": 0.3415,
+      "step": 590
+    },
+    {
+      "epoch": 0.5911662655121753,
+      "grad_norm": 0.23631636798381805,
+      "learning_rate": 0.00018141611697977529,
+      "loss": 0.3283,
+      "step": 591
+    },
+    {
+      "epoch": 0.5921665468412991,
+      "grad_norm": 0.24881725013256073,
+      "learning_rate": 0.00018135520702629675,
+      "loss": 0.3544,
+      "step": 592
+    },
+    {
+      "epoch": 0.5931668281704229,
+      "grad_norm": 0.2511225938796997,
+      "learning_rate": 0.0001812942076781588,
+      "loss": 0.3285,
+      "step": 593
+    },
+    {
+      "epoch": 0.5941671094995468,
+      "grad_norm": 0.24289599061012268,
+      "learning_rate": 0.0001812331190023886,
+      "loss": 0.3156,
+      "step": 594
+    },
+    {
+      "epoch": 0.5951673908286705,
+      "grad_norm": 0.34566813707351685,
+      "learning_rate": 0.0001811719410661116,
+      "loss": 0.3561,
+      "step": 595
+    },
+    {
+      "epoch": 0.5961676721577944,
+      "grad_norm": 0.22271698713302612,
+      "learning_rate": 0.00018111067393655132,
+      "loss": 0.2966,
+      "step": 596
+    },
+    {
+      "epoch": 0.5971679534869182,
+      "grad_norm": 0.24736548960208893,
+      "learning_rate": 0.0001810493176810292,
+      "loss": 0.3319,
+      "step": 597
+    },
+    {
+      "epoch": 0.598168234816042,
+      "grad_norm": 0.2430410534143448,
+      "learning_rate": 0.00018098787236696474,
+      "loss": 0.3423,
+      "step": 598
+    },
+    {
+      "epoch": 0.5991685161451659,
+      "grad_norm": 0.23629088699817657,
+      "learning_rate": 0.00018092633806187513,
+      "loss": 0.3462,
+      "step": 599
+    },
+    {
+      "epoch": 0.6001687974742896,
+      "grad_norm": 0.23937566578388214,
+      "learning_rate": 0.0001808647148333755,
+      "loss": 0.3127,
+      "step": 600
+    },
+    {
+      "epoch": 0.6011690788034134,
+      "grad_norm": 0.20777089893817902,
+      "learning_rate": 0.00018080300274917862,
+      "loss": 0.3094,
+      "step": 601
+    },
+    {
+      "epoch": 0.6021693601325373,
+      "grad_norm": 0.23153486847877502,
+      "learning_rate": 0.00018074120187709495,
+      "loss": 0.3603,
+      "step": 602
+    },
+    {
+      "epoch": 0.6031696414616611,
+      "grad_norm": 0.20608308911323547,
+      "learning_rate": 0.00018067931228503246,
+      "loss": 0.3515,
+      "step": 603
+    },
+    {
+      "epoch": 0.604169922790785,
+      "grad_norm": 0.24065174162387848,
+      "learning_rate": 0.00018061733404099655,
+      "loss": 0.3543,
+      "step": 604
+    },
+    {
+      "epoch": 0.6051702041199087,
+      "grad_norm": 0.22055767476558685,
+      "learning_rate": 0.00018055526721309016,
+      "loss": 0.317,
+      "step": 605
+    },
+    {
+      "epoch": 0.6061704854490325,
+      "grad_norm": 0.22590354084968567,
+      "learning_rate": 0.0001804931118695135,
+      "loss": 0.3184,
+      "step": 606
+    },
+    {
+      "epoch": 0.6071707667781564,
+      "grad_norm": 0.24891667068004608,
+      "learning_rate": 0.00018043086807856403,
+      "loss": 0.3342,
+      "step": 607
+    },
+    {
+      "epoch": 0.6081710481072802,
+      "grad_norm": 0.22794680297374725,
+      "learning_rate": 0.00018036853590863648,
+      "loss": 0.3235,
+      "step": 608
+    },
+    {
+      "epoch": 0.609171329436404,
+      "grad_norm": 0.28197333216667175,
+      "learning_rate": 0.00018030611542822257,
+      "loss": 0.3146,
+      "step": 609
+    },
+    {
+      "epoch": 0.6101716107655278,
+      "grad_norm": 0.2463848888874054,
+      "learning_rate": 0.00018024360670591114,
+      "loss": 0.353,
+      "step": 610
+    },
+    {
+      "epoch": 0.6111718920946516,
+      "grad_norm": 0.29584237933158875,
+      "learning_rate": 0.00018018100981038798,
+      "loss": 0.3163,
+      "step": 611
+    },
+    {
+      "epoch": 0.6121721734237754,
+      "grad_norm": 0.2657766342163086,
+      "learning_rate": 0.00018011832481043576,
+      "loss": 0.3201,
+      "step": 612
+    },
+    {
+      "epoch": 0.6131724547528993,
+      "grad_norm": 0.24318768084049225,
+      "learning_rate": 0.00018005555177493394,
+      "loss": 0.342,
+      "step": 613
+    },
+    {
+      "epoch": 0.614172736082023,
+      "grad_norm": 0.2235187441110611,
+      "learning_rate": 0.00017999269077285875,
+      "loss": 0.2909,
+      "step": 614
+    },
+    {
+      "epoch": 0.6151730174111469,
+      "grad_norm": 0.22863905131816864,
+      "learning_rate": 0.00017992974187328305,
+      "loss": 0.3279,
+      "step": 615
+    },
+    {
+      "epoch": 0.6161732987402707,
+      "grad_norm": 0.21534810960292816,
+      "learning_rate": 0.00017986670514537627,
+      "loss": 0.3308,
+      "step": 616
+    },
+    {
+      "epoch": 0.6171735800693945,
+      "grad_norm": 0.19034357368946075,
+      "learning_rate": 0.00017980358065840444,
+      "loss": 0.3039,
+      "step": 617
+    },
+    {
+      "epoch": 0.6181738613985184,
+      "grad_norm": 0.2606121301651001,
+      "learning_rate": 0.0001797403684817299,
+      "loss": 0.3421,
+      "step": 618
+    },
+    {
+      "epoch": 0.6191741427276422,
+      "grad_norm": 0.2246251106262207,
+      "learning_rate": 0.00017967706868481144,
+      "loss": 0.3883,
+      "step": 619
+    },
+    {
+      "epoch": 0.6201744240567659,
+      "grad_norm": 0.23439204692840576,
+      "learning_rate": 0.00017961368133720407,
+      "loss": 0.3251,
+      "step": 620
+    },
+    {
+      "epoch": 0.6211747053858898,
+      "grad_norm": 0.2128812074661255,
+      "learning_rate": 0.000179550206508559,
+      "loss": 0.3584,
+      "step": 621
+    },
+    {
+      "epoch": 0.6221749867150136,
+      "grad_norm": 0.2302810102701187,
+      "learning_rate": 0.00017948664426862364,
+      "loss": 0.3227,
+      "step": 622
+    },
+    {
+      "epoch": 0.6231752680441374,
+      "grad_norm": 0.3257722556591034,
+      "learning_rate": 0.00017942299468724134,
+      "loss": 0.3334,
+      "step": 623
+    },
+    {
+      "epoch": 0.6241755493732613,
+      "grad_norm": 0.27489688992500305,
+      "learning_rate": 0.0001793592578343515,
+      "loss": 0.3069,
+      "step": 624
+    },
+    {
+      "epoch": 0.625175830702385,
+      "grad_norm": 0.22322577238082886,
+      "learning_rate": 0.0001792954337799894,
+      "loss": 0.3347,
+      "step": 625
+    },
+    {
+      "epoch": 0.6261761120315089,
+      "grad_norm": 0.20777520537376404,
+      "learning_rate": 0.00017923152259428612,
+      "loss": 0.3237,
+      "step": 626
+    },
+    {
+      "epoch": 0.6271763933606327,
+      "grad_norm": 0.25841495394706726,
+      "learning_rate": 0.00017916752434746856,
+      "loss": 0.3485,
+      "step": 627
+    },
+    {
+      "epoch": 0.6281766746897565,
+      "grad_norm": 0.21619772911071777,
+      "learning_rate": 0.0001791034391098591,
+      "loss": 0.3357,
+      "step": 628
+    },
+    {
+      "epoch": 0.6291769560188804,
+      "grad_norm": 0.20234306156635284,
+      "learning_rate": 0.00017903926695187595,
+      "loss": 0.331,
+      "step": 629
+    },
+    {
+      "epoch": 0.6301772373480041,
+      "grad_norm": 0.19359396398067474,
+      "learning_rate": 0.0001789750079440326,
+      "loss": 0.3169,
+      "step": 630
+    },
+    {
+      "epoch": 0.6311775186771279,
+      "grad_norm": 0.23527684807777405,
+      "learning_rate": 0.00017891066215693817,
+      "loss": 0.3341,
+      "step": 631
+    },
+    {
+      "epoch": 0.6321778000062518,
+      "grad_norm": 0.27282392978668213,
+      "learning_rate": 0.00017884622966129695,
+      "loss": 0.3447,
+      "step": 632
+    },
+    {
+      "epoch": 0.6331780813353756,
+      "grad_norm": 0.20408576726913452,
+      "learning_rate": 0.00017878171052790868,
+      "loss": 0.3118,
+      "step": 633
+    },
+    {
+      "epoch": 0.6341783626644993,
+      "grad_norm": 0.23893193900585175,
+      "learning_rate": 0.00017871710482766817,
+      "loss": 0.3326,
+      "step": 634
+    },
+    {
+      "epoch": 0.6351786439936232,
+      "grad_norm": 0.19590994715690613,
+      "learning_rate": 0.00017865241263156546,
+      "loss": 0.2967,
+      "step": 635
+    },
+    {
+      "epoch": 0.636178925322747,
+      "grad_norm": 0.27993151545524597,
+      "learning_rate": 0.0001785876340106855,
+      "loss": 0.3287,
+      "step": 636
+    },
+    {
+      "epoch": 0.6371792066518709,
+      "grad_norm": 0.25192657113075256,
+      "learning_rate": 0.0001785227690362083,
+      "loss": 0.3437,
+      "step": 637
+    },
+    {
+      "epoch": 0.6381794879809947,
+      "grad_norm": 0.24973343312740326,
+      "learning_rate": 0.00017845781777940878,
+      "loss": 0.3582,
+      "step": 638
+    },
+    {
+      "epoch": 0.6391797693101184,
+      "grad_norm": 0.22954507172107697,
+      "learning_rate": 0.00017839278031165658,
+      "loss": 0.3126,
+      "step": 639
+    },
+    {
+      "epoch": 0.6401800506392423,
+      "grad_norm": 0.19363701343536377,
+      "learning_rate": 0.00017832765670441612,
+      "loss": 0.322,
+      "step": 640
+    },
+    {
+      "epoch": 0.6411803319683661,
+      "grad_norm": 0.3236650228500366,
+      "learning_rate": 0.0001782624470292465,
+      "loss": 0.3645,
+      "step": 641
+    },
+    {
+      "epoch": 0.6421806132974899,
+      "grad_norm": 0.299369752407074,
+      "learning_rate": 0.0001781971513578013,
+      "loss": 0.3166,
+      "step": 642
+    },
+    {
+      "epoch": 0.6431808946266138,
+      "grad_norm": 0.26070523262023926,
+      "learning_rate": 0.00017813176976182873,
+      "loss": 0.3527,
+      "step": 643
+    },
+    {
+      "epoch": 0.6441811759557375,
+      "grad_norm": 0.24070867896080017,
+      "learning_rate": 0.00017806630231317127,
+      "loss": 0.3361,
+      "step": 644
+    },
+    {
+      "epoch": 0.6451814572848613,
+      "grad_norm": 0.20702619850635529,
+      "learning_rate": 0.00017800074908376584,
+      "loss": 0.3337,
+      "step": 645
+    },
+    {
+      "epoch": 0.6461817386139852,
+      "grad_norm": 0.21480228006839752,
+      "learning_rate": 0.00017793511014564358,
+      "loss": 0.2922,
+      "step": 646
+    },
+    {
+      "epoch": 0.647182019943109,
+      "grad_norm": 0.22880414128303528,
+      "learning_rate": 0.00017786938557092983,
+      "loss": 0.3195,
+      "step": 647
+    },
+    {
+      "epoch": 0.6481823012722329,
+      "grad_norm": 0.23580582439899445,
+      "learning_rate": 0.00017780357543184397,
+      "loss": 0.3334,
+      "step": 648
+    },
+    {
+      "epoch": 0.6491825826013566,
+      "grad_norm": 0.2207668125629425,
+      "learning_rate": 0.00017773767980069945,
+      "loss": 0.2695,
+      "step": 649
+    },
+    {
+      "epoch": 0.6501828639304804,
+      "grad_norm": 0.1960780769586563,
+      "learning_rate": 0.0001776716987499037,
+      "loss": 0.2766,
+      "step": 650
+    },
+    {
+      "epoch": 0.6511831452596043,
+      "grad_norm": 0.22533899545669556,
+      "learning_rate": 0.0001776056323519579,
+      "loss": 0.3196,
+      "step": 651
+    },
+    {
+      "epoch": 0.6521834265887281,
+      "grad_norm": 0.22704724967479706,
+      "learning_rate": 0.00017753948067945712,
+      "loss": 0.2933,
+      "step": 652
+    },
+    {
+      "epoch": 0.6531837079178519,
+      "grad_norm": 0.21153584122657776,
+      "learning_rate": 0.00017747324380509006,
+      "loss": 0.3128,
+      "step": 653
+    },
+    {
+      "epoch": 0.6541839892469757,
+      "grad_norm": 0.19762000441551208,
+      "learning_rate": 0.00017740692180163908,
+      "loss": 0.2809,
+      "step": 654
+    },
+    {
+      "epoch": 0.6551842705760995,
+      "grad_norm": 0.2962697148323059,
+      "learning_rate": 0.00017734051474198003,
+      "loss": 0.3279,
+      "step": 655
+    },
+    {
+      "epoch": 0.6561845519052233,
+      "grad_norm": 0.20812612771987915,
+      "learning_rate": 0.0001772740226990823,
+      "loss": 0.2978,
+      "step": 656
+    },
+    {
+      "epoch": 0.6571848332343472,
+      "grad_norm": 0.18996229767799377,
+      "learning_rate": 0.00017720744574600863,
+      "loss": 0.3043,
+      "step": 657
+    },
+    {
+      "epoch": 0.658185114563471,
+      "grad_norm": 0.19563351571559906,
+      "learning_rate": 0.00017714078395591502,
+      "loss": 0.3056,
+      "step": 658
+    },
+    {
+      "epoch": 0.6591853958925948,
+      "grad_norm": 0.22013229131698608,
+      "learning_rate": 0.00017707403740205071,
+      "loss": 0.3307,
+      "step": 659
+    },
+    {
+      "epoch": 0.6601856772217186,
+      "grad_norm": 0.23212532699108124,
+      "learning_rate": 0.00017700720615775812,
+      "loss": 0.3191,
+      "step": 660
+    },
+    {
+      "epoch": 0.6611859585508424,
+      "grad_norm": 0.21490336954593658,
+      "learning_rate": 0.0001769402902964727,
+      "loss": 0.2875,
+      "step": 661
+    },
+    {
+      "epoch": 0.6621862398799663,
+      "grad_norm": 0.21017922461032867,
+      "learning_rate": 0.00017687328989172288,
+      "loss": 0.2782,
+      "step": 662
+    },
+    {
+      "epoch": 0.66318652120909,
+      "grad_norm": 0.20409469306468964,
+      "learning_rate": 0.00017680620501712996,
+      "loss": 0.31,
+      "step": 663
+    },
+    {
+      "epoch": 0.6641868025382138,
+      "grad_norm": 0.2117970734834671,
+      "learning_rate": 0.00017673903574640814,
+      "loss": 0.3358,
+      "step": 664
+    },
+    {
+      "epoch": 0.6651870838673377,
+      "grad_norm": 0.22394441068172455,
+      "learning_rate": 0.00017667178215336423,
+      "loss": 0.2816,
+      "step": 665
+    },
+    {
+      "epoch": 0.6661873651964615,
+      "grad_norm": 0.2687826156616211,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 0.3295,
+      "step": 666
+    },
+    {
+      "epoch": 0.6671876465255854,
+      "grad_norm": 0.42887550592422485,
+      "learning_rate": 0.000176537022296001,
+      "loss": 0.3804,
+      "step": 667
+    },
+    {
+      "epoch": 0.6681879278547092,
+      "grad_norm": 0.20850545167922974,
+      "learning_rate": 0.00017646951617975837,
+      "loss": 0.2975,
+      "step": 668
+    },
+    {
+      "epoch": 0.6691882091838329,
+      "grad_norm": 0.29046183824539185,
+      "learning_rate": 0.00017640192603734692,
+      "loss": 0.3875,
+      "step": 669
+    },
+    {
+      "epoch": 0.6701884905129568,
+      "grad_norm": 0.3161238133907318,
+      "learning_rate": 0.00017633425194303606,
+      "loss": 0.3701,
+      "step": 670
+    },
+    {
+      "epoch": 0.6711887718420806,
+      "grad_norm": 0.2167435884475708,
+      "learning_rate": 0.00017626649397118734,
+      "loss": 0.3254,
+      "step": 671
+    },
+    {
+      "epoch": 0.6721890531712044,
+      "grad_norm": 0.21845142543315887,
+      "learning_rate": 0.00017619865219625452,
+      "loss": 0.3138,
+      "step": 672
+    },
+    {
+      "epoch": 0.6731893345003283,
+      "grad_norm": 0.2281215488910675,
+      "learning_rate": 0.00017613072669278343,
+      "loss": 0.307,
+      "step": 673
+    },
+    {
+      "epoch": 0.674189615829452,
+      "grad_norm": 0.2242210954427719,
+      "learning_rate": 0.00017606271753541192,
+      "loss": 0.3356,
+      "step": 674
+    },
+    {
+      "epoch": 0.6751898971585758,
+      "grad_norm": 0.21816246211528778,
+      "learning_rate": 0.00017599462479886974,
+      "loss": 0.3168,
+      "step": 675
+    },
+    {
+      "epoch": 0.6761901784876997,
+      "grad_norm": 0.21995382010936737,
+      "learning_rate": 0.00017592644855797854,
+      "loss": 0.3334,
+      "step": 676
+    },
+    {
+      "epoch": 0.6771904598168235,
+      "grad_norm": 0.235803484916687,
+      "learning_rate": 0.00017585818888765168,
+      "loss": 0.352,
+      "step": 677
+    },
+    {
+      "epoch": 0.6781907411459474,
+      "grad_norm": 0.261046439409256,
+      "learning_rate": 0.0001757898458628941,
+      "loss": 0.3191,
+      "step": 678
+    },
+    {
+      "epoch": 0.6791910224750711,
+      "grad_norm": 0.22552995383739471,
+      "learning_rate": 0.00017572141955880252,
+      "loss": 0.3214,
+      "step": 679
+    },
+    {
+      "epoch": 0.6801913038041949,
+      "grad_norm": 0.2219519317150116,
+      "learning_rate": 0.00017565291005056504,
+      "loss": 0.2891,
+      "step": 680
+    },
+    {
+      "epoch": 0.6811915851333188,
+      "grad_norm": 0.20971596240997314,
+      "learning_rate": 0.00017558431741346122,
+      "loss": 0.2962,
+      "step": 681
+    },
+    {
+      "epoch": 0.6821918664624426,
+      "grad_norm": 0.197629913687706,
+      "learning_rate": 0.00017551564172286197,
+      "loss": 0.2952,
+      "step": 682
+    },
+    {
+      "epoch": 0.6831921477915663,
+      "grad_norm": 0.21036967635154724,
+      "learning_rate": 0.00017544688305422943,
+      "loss": 0.3051,
+      "step": 683
+    },
+    {
+      "epoch": 0.6841924291206902,
+      "grad_norm": 0.22057054936885834,
+      "learning_rate": 0.00017537804148311695,
+      "loss": 0.31,
+      "step": 684
+    },
+    {
+      "epoch": 0.685192710449814,
+      "grad_norm": 0.24886031448841095,
+      "learning_rate": 0.00017530911708516902,
+      "loss": 0.3421,
+      "step": 685
+    },
+    {
+      "epoch": 0.6861929917789378,
+      "grad_norm": 0.22189615666866302,
+      "learning_rate": 0.00017524010993612098,
+      "loss": 0.3274,
+      "step": 686
+    },
+    {
+      "epoch": 0.6871932731080617,
+      "grad_norm": 0.21390300989151,
+      "learning_rate": 0.00017517102011179933,
+      "loss": 0.304,
+      "step": 687
+    },
+    {
+      "epoch": 0.6881935544371854,
+      "grad_norm": 0.19355562329292297,
+      "learning_rate": 0.0001751018476881212,
+      "loss": 0.2837,
+      "step": 688
+    },
+    {
+      "epoch": 0.6891938357663093,
+      "grad_norm": 0.21197232604026794,
+      "learning_rate": 0.00017503259274109464,
+      "loss": 0.2925,
+      "step": 689
+    },
+    {
+      "epoch": 0.6901941170954331,
+      "grad_norm": 0.23380397260189056,
+      "learning_rate": 0.00017496325534681825,
+      "loss": 0.3101,
+      "step": 690
+    },
+    {
+      "epoch": 0.6911943984245569,
+      "grad_norm": 0.21716703474521637,
+      "learning_rate": 0.00017489383558148136,
+      "loss": 0.3325,
+      "step": 691
+    },
+    {
+      "epoch": 0.6921946797536808,
+      "grad_norm": 0.21835178136825562,
+      "learning_rate": 0.00017482433352136365,
+      "loss": 0.2924,
+      "step": 692
+    },
+    {
+      "epoch": 0.6931949610828045,
+      "grad_norm": 0.2123212218284607,
+      "learning_rate": 0.00017475474924283536,
+      "loss": 0.3113,
+      "step": 693
+    },
+    {
+      "epoch": 0.6941952424119283,
+      "grad_norm": 0.21609410643577576,
+      "learning_rate": 0.00017468508282235704,
+      "loss": 0.3299,
+      "step": 694
+    },
+    {
+      "epoch": 0.6951955237410522,
+      "grad_norm": 0.23166966438293457,
+      "learning_rate": 0.00017461533433647946,
+      "loss": 0.333,
+      "step": 695
+    },
+    {
+      "epoch": 0.696195805070176,
+      "grad_norm": 0.2625264525413513,
+      "learning_rate": 0.00017454550386184362,
+      "loss": 0.3226,
+      "step": 696
+    },
+    {
+      "epoch": 0.6971960863992998,
+      "grad_norm": 0.21635104715824127,
+      "learning_rate": 0.00017447559147518055,
+      "loss": 0.3311,
+      "step": 697
+    },
+    {
+      "epoch": 0.6981963677284236,
+      "grad_norm": 0.1935082972049713,
+      "learning_rate": 0.00017440559725331135,
+      "loss": 0.2981,
+      "step": 698
+    },
+    {
+      "epoch": 0.6991966490575474,
+      "grad_norm": 0.25610676407814026,
+      "learning_rate": 0.000174335521273147,
+      "loss": 0.3477,
+      "step": 699
+    },
+    {
+      "epoch": 0.7001969303866713,
+      "grad_norm": 0.19709213078022003,
+      "learning_rate": 0.00017426536361168834,
+      "loss": 0.2946,
+      "step": 700
+    },
+    {
+      "epoch": 0.7011972117157951,
+      "grad_norm": 0.31819432973861694,
+      "learning_rate": 0.00017419512434602594,
+      "loss": 0.3624,
+      "step": 701
+    },
+    {
+      "epoch": 0.7021974930449189,
+      "grad_norm": 0.2429499626159668,
+      "learning_rate": 0.00017412480355334005,
+      "loss": 0.2983,
+      "step": 702
+    },
+    {
+      "epoch": 0.7031977743740427,
+      "grad_norm": 0.20953412353992462,
+      "learning_rate": 0.00017405440131090048,
+      "loss": 0.3071,
+      "step": 703
+    },
+    {
+      "epoch": 0.7041980557031665,
+      "grad_norm": 0.2496146410703659,
+      "learning_rate": 0.00017398391769606658,
+      "loss": 0.2915,
+      "step": 704
+    },
+    {
+      "epoch": 0.7051983370322903,
+      "grad_norm": 0.21096958220005035,
+      "learning_rate": 0.00017391335278628712,
+      "loss": 0.3472,
+      "step": 705
+    },
+    {
+      "epoch": 0.7061986183614142,
+      "grad_norm": 0.20255278050899506,
+      "learning_rate": 0.00017384270665910014,
+      "loss": 0.3194,
+      "step": 706
+    },
+    {
+      "epoch": 0.707198899690538,
+      "grad_norm": 0.22382409870624542,
+      "learning_rate": 0.000173771979392133,
+      "loss": 0.3071,
+      "step": 707
+    },
+    {
+      "epoch": 0.7081991810196617,
+      "grad_norm": 0.21205894649028778,
+      "learning_rate": 0.00017370117106310214,
+      "loss": 0.3141,
+      "step": 708
+    },
+    {
+      "epoch": 0.7091994623487856,
+      "grad_norm": 0.24390722811222076,
+      "learning_rate": 0.0001736302817498131,
+      "loss": 0.3215,
+      "step": 709
+    },
+    {
+      "epoch": 0.7101997436779094,
+      "grad_norm": 0.32894018292427063,
+      "learning_rate": 0.00017355931153016044,
+      "loss": 0.3414,
+      "step": 710
+    },
+    {
+      "epoch": 0.7112000250070333,
+      "grad_norm": 0.20446595549583435,
+      "learning_rate": 0.0001734882604821276,
+      "loss": 0.3104,
+      "step": 711
+    },
+    {
+      "epoch": 0.712200306336157,
+      "grad_norm": 0.1856837272644043,
+      "learning_rate": 0.0001734171286837868,
+      "loss": 0.2912,
+      "step": 712
+    },
+    {
+      "epoch": 0.7132005876652808,
+      "grad_norm": 0.20234134793281555,
+      "learning_rate": 0.00017334591621329906,
+      "loss": 0.2682,
+      "step": 713
+    },
+    {
+      "epoch": 0.7142008689944047,
+      "grad_norm": 0.2752034664154053,
+      "learning_rate": 0.00017327462314891402,
+      "loss": 0.3302,
+      "step": 714
+    },
+    {
+      "epoch": 0.7152011503235285,
+      "grad_norm": 0.24601981043815613,
+      "learning_rate": 0.00017320324956896977,
+      "loss": 0.3282,
+      "step": 715
+    },
+    {
+      "epoch": 0.7162014316526523,
+      "grad_norm": 0.21674604713916779,
+      "learning_rate": 0.00017313179555189306,
+      "loss": 0.3069,
+      "step": 716
+    },
+    {
+      "epoch": 0.7172017129817762,
+      "grad_norm": 0.28281110525131226,
+      "learning_rate": 0.00017306026117619889,
+      "loss": 0.3505,
+      "step": 717
+    },
+    {
+      "epoch": 0.7182019943108999,
+      "grad_norm": 0.22063292562961578,
+      "learning_rate": 0.0001729886465204906,
+      "loss": 0.3075,
+      "step": 718
+    },
+    {
+      "epoch": 0.7192022756400237,
+      "grad_norm": 0.24117493629455566,
+      "learning_rate": 0.0001729169516634598,
+      "loss": 0.3649,
+      "step": 719
+    },
+    {
+      "epoch": 0.7202025569691476,
+      "grad_norm": 0.18200711905956268,
+      "learning_rate": 0.0001728451766838861,
+      "loss": 0.2833,
+      "step": 720
+    },
+    {
+      "epoch": 0.7212028382982714,
+      "grad_norm": 0.20031879842281342,
+      "learning_rate": 0.00017277332166063726,
+      "loss": 0.3171,
+      "step": 721
+    },
+    {
+      "epoch": 0.7222031196273953,
+      "grad_norm": 0.24200114607810974,
+      "learning_rate": 0.00017270138667266894,
+      "loss": 0.345,
+      "step": 722
+    },
+    {
+      "epoch": 0.723203400956519,
+      "grad_norm": 0.27134808897972107,
+      "learning_rate": 0.00017262937179902472,
+      "loss": 0.3442,
+      "step": 723
+    },
+    {
+      "epoch": 0.7242036822856428,
+      "grad_norm": 0.20982469618320465,
+      "learning_rate": 0.00017255727711883588,
+      "loss": 0.3057,
+      "step": 724
+    },
+    {
+      "epoch": 0.7252039636147667,
+      "grad_norm": 0.2001883089542389,
+      "learning_rate": 0.00017248510271132144,
+      "loss": 0.3101,
+      "step": 725
+    },
+    {
+      "epoch": 0.7262042449438905,
+      "grad_norm": 0.20036129653453827,
+      "learning_rate": 0.00017241284865578802,
+      "loss": 0.2912,
+      "step": 726
+    },
+    {
+      "epoch": 0.7272045262730142,
+      "grad_norm": 0.20709696412086487,
+      "learning_rate": 0.00017234051503162978,
+      "loss": 0.3367,
+      "step": 727
+    },
+    {
+      "epoch": 0.7282048076021381,
+      "grad_norm": 0.2235272228717804,
+      "learning_rate": 0.0001722681019183283,
+      "loss": 0.3069,
+      "step": 728
+    },
+    {
+      "epoch": 0.7292050889312619,
+      "grad_norm": 0.24653470516204834,
+      "learning_rate": 0.00017219560939545246,
+      "loss": 0.3523,
+      "step": 729
+    },
+    {
+      "epoch": 0.7302053702603858,
+      "grad_norm": 0.22146138548851013,
+      "learning_rate": 0.00017212303754265843,
+      "loss": 0.3427,
+      "step": 730
+    },
+    {
+      "epoch": 0.7312056515895096,
+      "grad_norm": 0.21458594501018524,
+      "learning_rate": 0.0001720503864396896,
+      "loss": 0.3265,
+      "step": 731
+    },
+    {
+      "epoch": 0.7322059329186333,
+      "grad_norm": 0.2700503468513489,
+      "learning_rate": 0.00017197765616637636,
+      "loss": 0.3291,
+      "step": 732
+    },
+    {
+      "epoch": 0.7332062142477572,
+      "grad_norm": 0.3798985779285431,
+      "learning_rate": 0.0001719048468026361,
+      "loss": 0.3103,
+      "step": 733
+    },
+    {
+      "epoch": 0.734206495576881,
+      "grad_norm": 0.23312173783779144,
+      "learning_rate": 0.00017183195842847322,
+      "loss": 0.3061,
+      "step": 734
+    },
+    {
+      "epoch": 0.7352067769060048,
+      "grad_norm": 0.23339532315731049,
+      "learning_rate": 0.0001717589911239788,
+      "loss": 0.3065,
+      "step": 735
+    },
+    {
+      "epoch": 0.7362070582351287,
+      "grad_norm": 0.21831004321575165,
+      "learning_rate": 0.00017168594496933074,
+      "loss": 0.3392,
+      "step": 736
+    },
+    {
+      "epoch": 0.7372073395642524,
+      "grad_norm": 0.26539188623428345,
+      "learning_rate": 0.00017161282004479351,
+      "loss": 0.3644,
+      "step": 737
+    },
+    {
+      "epoch": 0.7382076208933762,
+      "grad_norm": 0.23967157304286957,
+      "learning_rate": 0.0001715396164307182,
+      "loss": 0.298,
+      "step": 738
+    },
+    {
+      "epoch": 0.7392079022225001,
+      "grad_norm": 0.19692367315292358,
+      "learning_rate": 0.0001714663342075424,
+      "loss": 0.2706,
+      "step": 739
+    },
+    {
+      "epoch": 0.7402081835516239,
+      "grad_norm": 0.2776857912540436,
+      "learning_rate": 0.00017139297345578994,
+      "loss": 0.3293,
+      "step": 740
+    },
+    {
+      "epoch": 0.7412084648807478,
+      "grad_norm": 0.22425878047943115,
+      "learning_rate": 0.00017131953425607104,
+      "loss": 0.3614,
+      "step": 741
+    },
+    {
+      "epoch": 0.7422087462098715,
+      "grad_norm": 0.3110194802284241,
+      "learning_rate": 0.00017124601668908212,
+      "loss": 0.3608,
+      "step": 742
+    },
+    {
+      "epoch": 0.7432090275389953,
+      "grad_norm": 0.2057676762342453,
+      "learning_rate": 0.00017117242083560568,
+      "loss": 0.3152,
+      "step": 743
+    },
+    {
+      "epoch": 0.7442093088681192,
+      "grad_norm": 0.31609559059143066,
+      "learning_rate": 0.00017109874677651024,
+      "loss": 0.3022,
+      "step": 744
+    },
+    {
+      "epoch": 0.745209590197243,
+      "grad_norm": 0.2300511598587036,
+      "learning_rate": 0.0001710249945927503,
+      "loss": 0.333,
+      "step": 745
+    },
+    {
+      "epoch": 0.7462098715263668,
+      "grad_norm": 0.27964073419570923,
+      "learning_rate": 0.00017095116436536612,
+      "loss": 0.3232,
+      "step": 746
+    },
+    {
+      "epoch": 0.7472101528554906,
+      "grad_norm": 0.19513124227523804,
+      "learning_rate": 0.00017087725617548385,
+      "loss": 0.3145,
+      "step": 747
+    },
+    {
+      "epoch": 0.7482104341846144,
+      "grad_norm": 0.18826736509799957,
+      "learning_rate": 0.00017080327010431513,
+      "loss": 0.2824,
+      "step": 748
+    },
+    {
+      "epoch": 0.7492107155137382,
+      "grad_norm": 0.25046253204345703,
+      "learning_rate": 0.00017072920623315734,
+      "loss": 0.3606,
+      "step": 749
+    },
+    {
+      "epoch": 0.7502109968428621,
+      "grad_norm": 0.21152693033218384,
+      "learning_rate": 0.00017065506464339326,
+      "loss": 0.2763,
+      "step": 750
+    },
+    {
+      "epoch": 0.7512112781719859,
+      "grad_norm": 0.4696851968765259,
+      "learning_rate": 0.00017058084541649106,
+      "loss": 0.2811,
+      "step": 751
+    },
+    {
+      "epoch": 0.7522115595011097,
+      "grad_norm": 0.18902602791786194,
+      "learning_rate": 0.00017050654863400429,
+      "loss": 0.3121,
+      "step": 752
+    },
+    {
+      "epoch": 0.7532118408302335,
+      "grad_norm": 0.22310034930706024,
+      "learning_rate": 0.00017043217437757164,
+      "loss": 0.2997,
+      "step": 753
+    },
+    {
+      "epoch": 0.7542121221593573,
+      "grad_norm": 0.2153480052947998,
+      "learning_rate": 0.00017035772272891702,
+      "loss": 0.2981,
+      "step": 754
+    },
+    {
+      "epoch": 0.7552124034884812,
+      "grad_norm": 0.2857193052768707,
+      "learning_rate": 0.00017028319376984928,
+      "loss": 0.2946,
+      "step": 755
+    },
+    {
+      "epoch": 0.756212684817605,
+      "grad_norm": 0.322998046875,
+      "learning_rate": 0.00017020858758226229,
+      "loss": 0.3223,
+      "step": 756
+    },
+    {
+      "epoch": 0.7572129661467287,
+      "grad_norm": 0.21400819718837738,
+      "learning_rate": 0.0001701339042481347,
+      "loss": 0.2804,
+      "step": 757
+    },
+    {
+      "epoch": 0.7582132474758526,
+      "grad_norm": 0.2679826319217682,
+      "learning_rate": 0.00017005914384953007,
+      "loss": 0.3514,
+      "step": 758
+    },
+    {
+      "epoch": 0.7592135288049764,
+      "grad_norm": 0.210464745759964,
+      "learning_rate": 0.00016998430646859654,
+      "loss": 0.3203,
+      "step": 759
+    },
+    {
+      "epoch": 0.7602138101341002,
+      "grad_norm": 0.27516427636146545,
+      "learning_rate": 0.00016990939218756683,
+      "loss": 0.2673,
+      "step": 760
+    },
+    {
+      "epoch": 0.761214091463224,
+      "grad_norm": 0.6008299589157104,
+      "learning_rate": 0.0001698344010887582,
+      "loss": 0.3482,
+      "step": 761
+    },
+    {
+      "epoch": 0.7622143727923478,
+      "grad_norm": 0.2188321352005005,
+      "learning_rate": 0.0001697593332545723,
+      "loss": 0.3034,
+      "step": 762
+    },
+    {
+      "epoch": 0.7632146541214717,
+      "grad_norm": 0.2251119315624237,
+      "learning_rate": 0.0001696841887674951,
+      "loss": 0.2865,
+      "step": 763
+    },
+    {
+      "epoch": 0.7642149354505955,
+      "grad_norm": 0.24993588030338287,
+      "learning_rate": 0.00016960896771009684,
+      "loss": 0.3164,
+      "step": 764
+    },
+    {
+      "epoch": 0.7652152167797193,
+      "grad_norm": 0.2394706904888153,
+      "learning_rate": 0.00016953367016503182,
+      "loss": 0.3283,
+      "step": 765
+    },
+    {
+      "epoch": 0.7662154981088432,
+      "grad_norm": 0.24101629853248596,
+      "learning_rate": 0.00016945829621503838,
+      "loss": 0.3249,
+      "step": 766
+    },
+    {
+      "epoch": 0.7672157794379669,
+      "grad_norm": 0.23974081873893738,
+      "learning_rate": 0.00016938284594293897,
+      "loss": 0.3332,
+      "step": 767
+    },
+    {
+      "epoch": 0.7682160607670907,
+      "grad_norm": 0.21005812287330627,
+      "learning_rate": 0.00016930731943163972,
+      "loss": 0.3263,
+      "step": 768
+    },
+    {
+      "epoch": 0.7692163420962146,
+      "grad_norm": 0.22122922539710999,
+      "learning_rate": 0.00016923171676413063,
+      "loss": 0.3237,
+      "step": 769
+    },
+    {
+      "epoch": 0.7702166234253384,
+      "grad_norm": 0.1977543979883194,
+      "learning_rate": 0.00016915603802348535,
+      "loss": 0.277,
+      "step": 770
+    },
+    {
+      "epoch": 0.7712169047544621,
+      "grad_norm": 0.2360074371099472,
+      "learning_rate": 0.00016908028329286112,
+      "loss": 0.3029,
+      "step": 771
+    },
+    {
+      "epoch": 0.772217186083586,
+      "grad_norm": 0.20942842960357666,
+      "learning_rate": 0.0001690044526554987,
+      "loss": 0.3048,
+      "step": 772
+    },
+    {
+      "epoch": 0.7732174674127098,
+      "grad_norm": 0.19742004573345184,
+      "learning_rate": 0.00016892854619472223,
+      "loss": 0.2657,
+      "step": 773
+    },
+    {
+      "epoch": 0.7742177487418337,
+      "grad_norm": 0.211008682847023,
+      "learning_rate": 0.00016885256399393924,
+      "loss": 0.3017,
+      "step": 774
+    },
+    {
+      "epoch": 0.7752180300709575,
+      "grad_norm": 0.23676526546478271,
+      "learning_rate": 0.00016877650613664034,
+      "loss": 0.265,
+      "step": 775
+    },
+    {
+      "epoch": 0.7762183114000812,
+      "grad_norm": 0.20885905623435974,
+      "learning_rate": 0.00016870037270639942,
+      "loss": 0.3108,
+      "step": 776
+    },
+    {
+      "epoch": 0.7772185927292051,
+      "grad_norm": 0.20989930629730225,
+      "learning_rate": 0.0001686241637868734,
+      "loss": 0.3089,
+      "step": 777
+    },
+    {
+      "epoch": 0.7782188740583289,
+      "grad_norm": 0.2795998156070709,
+      "learning_rate": 0.00016854787946180198,
+      "loss": 0.3664,
+      "step": 778
+    },
+    {
+      "epoch": 0.7792191553874527,
+      "grad_norm": 0.20469176769256592,
+      "learning_rate": 0.00016847151981500789,
+      "loss": 0.3529,
+      "step": 779
+    },
+    {
+      "epoch": 0.7802194367165766,
+      "grad_norm": 0.19085997343063354,
+      "learning_rate": 0.00016839508493039657,
+      "loss": 0.3047,
+      "step": 780
+    },
+    {
+      "epoch": 0.7812197180457003,
+      "grad_norm": 0.21169711649417877,
+      "learning_rate": 0.00016831857489195618,
+      "loss": 0.2937,
+      "step": 781
+    },
+    {
+      "epoch": 0.7822199993748241,
+      "grad_norm": 0.2310487926006317,
+      "learning_rate": 0.00016824198978375736,
+      "loss": 0.3164,
+      "step": 782
+    },
+    {
+      "epoch": 0.783220280703948,
+      "grad_norm": 0.2429760843515396,
+      "learning_rate": 0.00016816532968995328,
+      "loss": 0.3293,
+      "step": 783
+    },
+    {
+      "epoch": 0.7842205620330718,
+      "grad_norm": 0.41659173369407654,
+      "learning_rate": 0.0001680885946947796,
+      "loss": 0.2967,
+      "step": 784
+    },
+    {
+      "epoch": 0.7852208433621957,
+      "grad_norm": 0.1966882199048996,
+      "learning_rate": 0.00016801178488255413,
+      "loss": 0.2968,
+      "step": 785
+    },
+    {
+      "epoch": 0.7862211246913194,
+      "grad_norm": 0.19591602683067322,
+      "learning_rate": 0.00016793490033767698,
+      "loss": 0.3043,
+      "step": 786
+    },
+    {
+      "epoch": 0.7872214060204432,
+      "grad_norm": 0.180844247341156,
+      "learning_rate": 0.00016785794114463037,
+      "loss": 0.2758,
+      "step": 787
+    },
+    {
+      "epoch": 0.7882216873495671,
+      "grad_norm": 0.17917554080486298,
+      "learning_rate": 0.00016778090738797853,
+      "loss": 0.302,
+      "step": 788
+    },
+    {
+      "epoch": 0.7892219686786909,
+      "grad_norm": 0.2658692002296448,
+      "learning_rate": 0.00016770379915236766,
+      "loss": 0.3128,
+      "step": 789
+    },
+    {
+      "epoch": 0.7902222500078147,
+      "grad_norm": 0.21031467616558075,
+      "learning_rate": 0.00016762661652252567,
+      "loss": 0.3044,
+      "step": 790
+    },
+    {
+      "epoch": 0.7912225313369385,
+      "grad_norm": 0.30306217074394226,
+      "learning_rate": 0.00016754935958326244,
+      "loss": 0.3506,
+      "step": 791
+    },
+    {
+      "epoch": 0.7922228126660623,
+      "grad_norm": 0.24374417960643768,
+      "learning_rate": 0.00016747202841946928,
+      "loss": 0.3255,
+      "step": 792
+    },
+    {
+      "epoch": 0.7932230939951862,
+      "grad_norm": 0.21545256674289703,
+      "learning_rate": 0.00016739462311611919,
+      "loss": 0.2927,
+      "step": 793
+    },
+    {
+      "epoch": 0.79422337532431,
+      "grad_norm": 0.2924513518810272,
+      "learning_rate": 0.00016731714375826657,
+      "loss": 0.3137,
+      "step": 794
+    },
+    {
+      "epoch": 0.7952236566534338,
+      "grad_norm": 0.21599429845809937,
+      "learning_rate": 0.00016723959043104728,
+      "loss": 0.3341,
+      "step": 795
+    },
+    {
+      "epoch": 0.7962239379825576,
+      "grad_norm": 0.21233899891376495,
+      "learning_rate": 0.00016716196321967832,
+      "loss": 0.2925,
+      "step": 796
+    },
+    {
+      "epoch": 0.7972242193116814,
+      "grad_norm": 0.24538256227970123,
+      "learning_rate": 0.00016708426220945802,
+      "loss": 0.3438,
+      "step": 797
+    },
+    {
+      "epoch": 0.7982245006408052,
+      "grad_norm": 0.2090294510126114,
+      "learning_rate": 0.00016700648748576574,
+      "loss": 0.3039,
+      "step": 798
+    },
+    {
+      "epoch": 0.7992247819699291,
+      "grad_norm": 0.30814480781555176,
+      "learning_rate": 0.0001669286391340618,
+      "loss": 0.3153,
+      "step": 799
+    },
+    {
+      "epoch": 0.8002250632990529,
+      "grad_norm": 0.22740478813648224,
+      "learning_rate": 0.00016685071723988748,
+      "loss": 0.3268,
+      "step": 800
+    },
+    {
+      "epoch": 0.8012253446281766,
+      "grad_norm": 0.2445048987865448,
+      "learning_rate": 0.00016677272188886483,
+      "loss": 0.3258,
+      "step": 801
+    },
+    {
+      "epoch": 0.8022256259573005,
+      "grad_norm": 0.7179467082023621,
+      "learning_rate": 0.00016669465316669667,
+      "loss": 0.2985,
+      "step": 802
+    },
+    {
+      "epoch": 0.8032259072864243,
+      "grad_norm": 0.1969257891178131,
+      "learning_rate": 0.00016661651115916642,
+      "loss": 0.3014,
+      "step": 803
+    },
+    {
+      "epoch": 0.8042261886155482,
+      "grad_norm": 0.2130976766347885,
+      "learning_rate": 0.00016653829595213794,
+      "loss": 0.283,
+      "step": 804
+    },
+    {
+      "epoch": 0.805226469944672,
+      "grad_norm": 0.2909943461418152,
+      "learning_rate": 0.00016646000763155568,
+      "loss": 0.3614,
+      "step": 805
+    },
+    {
+      "epoch": 0.8062267512737957,
+      "grad_norm": 0.2181178331375122,
+      "learning_rate": 0.00016638164628344425,
+      "loss": 0.3472,
+      "step": 806
+    },
+    {
+      "epoch": 0.8072270326029196,
+      "grad_norm": 0.2834266722202301,
+      "learning_rate": 0.00016630321199390867,
+      "loss": 0.349,
+      "step": 807
+    },
+    {
+      "epoch": 0.8082273139320434,
+      "grad_norm": 0.19750535488128662,
+      "learning_rate": 0.00016622470484913406,
+      "loss": 0.312,
+      "step": 808
+    },
+    {
+      "epoch": 0.8092275952611672,
+      "grad_norm": 0.2314111590385437,
+      "learning_rate": 0.00016614612493538551,
+      "loss": 0.3038,
+      "step": 809
+    },
+    {
+      "epoch": 0.810227876590291,
+      "grad_norm": 0.18575963377952576,
+      "learning_rate": 0.00016606747233900815,
+      "loss": 0.2609,
+      "step": 810
+    },
+    {
+      "epoch": 0.8112281579194148,
+      "grad_norm": 0.19067701697349548,
+      "learning_rate": 0.00016598874714642697,
+      "loss": 0.2773,
+      "step": 811
+    },
+    {
+      "epoch": 0.8122284392485386,
+      "grad_norm": 0.24707461893558502,
+      "learning_rate": 0.00016590994944414678,
+      "loss": 0.326,
+      "step": 812
+    },
+    {
+      "epoch": 0.8132287205776625,
+      "grad_norm": 0.2056681215763092,
+      "learning_rate": 0.00016583107931875192,
+      "loss": 0.2953,
+      "step": 813
+    },
+    {
+      "epoch": 0.8142290019067863,
+      "grad_norm": 0.2088230401277542,
+      "learning_rate": 0.0001657521368569064,
+      "loss": 0.2887,
+      "step": 814
+    },
+    {
+      "epoch": 0.8152292832359102,
+      "grad_norm": 0.19099104404449463,
+      "learning_rate": 0.0001656731221453537,
+      "loss": 0.2961,
+      "step": 815
+    },
+    {
+      "epoch": 0.8162295645650339,
+      "grad_norm": 0.19913238286972046,
+      "learning_rate": 0.00016559403527091675,
+      "loss": 0.2976,
+      "step": 816
+    },
+    {
+      "epoch": 0.8172298458941577,
+      "grad_norm": 0.30864930152893066,
+      "learning_rate": 0.0001655148763204977,
+      "loss": 0.287,
+      "step": 817
+    },
+    {
+      "epoch": 0.8182301272232816,
+      "grad_norm": 0.2696269452571869,
+      "learning_rate": 0.00016543564538107797,
+      "loss": 0.3543,
+      "step": 818
+    },
+    {
+      "epoch": 0.8192304085524054,
+      "grad_norm": 0.20706598460674286,
+      "learning_rate": 0.00016535634253971794,
+      "loss": 0.3057,
+      "step": 819
+    },
+    {
+      "epoch": 0.8202306898815291,
+      "grad_norm": 0.20164966583251953,
+      "learning_rate": 0.00016527696788355714,
+      "loss": 0.2871,
+      "step": 820
+    },
+    {
+      "epoch": 0.821230971210653,
+      "grad_norm": 0.2445070594549179,
+      "learning_rate": 0.00016519752149981397,
+      "loss": 0.3211,
+      "step": 821
+    },
+    {
+      "epoch": 0.8222312525397768,
+      "grad_norm": 0.2110719233751297,
+      "learning_rate": 0.0001651180034757856,
+      "loss": 0.3042,
+      "step": 822
+    },
+    {
+      "epoch": 0.8232315338689006,
+      "grad_norm": 0.23555681109428406,
+      "learning_rate": 0.00016503841389884798,
+      "loss": 0.3003,
+      "step": 823
+    },
+    {
+      "epoch": 0.8242318151980245,
+      "grad_norm": 0.1950465440750122,
+      "learning_rate": 0.00016495875285645566,
+      "loss": 0.2461,
+      "step": 824
+    },
+    {
+      "epoch": 0.8252320965271482,
+      "grad_norm": 0.1924734264612198,
+      "learning_rate": 0.00016487902043614173,
+      "loss": 0.2847,
+      "step": 825
+    },
+    {
+      "epoch": 0.8262323778562721,
+      "grad_norm": 0.20394857227802277,
+      "learning_rate": 0.0001647992167255177,
+      "loss": 0.3343,
+      "step": 826
+    },
+    {
+      "epoch": 0.8272326591853959,
+      "grad_norm": 0.20623324811458588,
+      "learning_rate": 0.0001647193418122734,
+      "loss": 0.3162,
+      "step": 827
+    },
+    {
+      "epoch": 0.8282329405145197,
+      "grad_norm": 0.19690217077732086,
+      "learning_rate": 0.00016463939578417692,
+      "loss": 0.3232,
+      "step": 828
+    },
+    {
+      "epoch": 0.8292332218436436,
+      "grad_norm": 0.18426956236362457,
+      "learning_rate": 0.0001645593787290745,
+      "loss": 0.2931,
+      "step": 829
+    },
+    {
+      "epoch": 0.8302335031727673,
+      "grad_norm": 0.18333686888217926,
+      "learning_rate": 0.0001644792907348904,
+      "loss": 0.2577,
+      "step": 830
+    },
+    {
+      "epoch": 0.8312337845018911,
+      "grad_norm": 0.23116953670978546,
+      "learning_rate": 0.00016439913188962685,
+      "loss": 0.3194,
+      "step": 831
+    },
+    {
+      "epoch": 0.832234065831015,
+      "grad_norm": 0.2502155601978302,
+      "learning_rate": 0.0001643189022813639,
+      "loss": 0.2686,
+      "step": 832
+    },
+    {
+      "epoch": 0.8332343471601388,
+      "grad_norm": 0.2191857546567917,
+      "learning_rate": 0.0001642386019982594,
+      "loss": 0.2989,
+      "step": 833
+    },
+    {
+      "epoch": 0.8342346284892626,
+      "grad_norm": 0.19816067814826965,
+      "learning_rate": 0.00016415823112854883,
+      "loss": 0.321,
+      "step": 834
+    },
+    {
+      "epoch": 0.8352349098183864,
+      "grad_norm": 0.20771951973438263,
+      "learning_rate": 0.00016407778976054526,
+      "loss": 0.3122,
+      "step": 835
+    },
+    {
+      "epoch": 0.8362351911475102,
+      "grad_norm": 0.19089916348457336,
+      "learning_rate": 0.0001639972779826392,
+      "loss": 0.2801,
+      "step": 836
+    },
+    {
+      "epoch": 0.8372354724766341,
+      "grad_norm": 0.19718053936958313,
+      "learning_rate": 0.0001639166958832985,
+      "loss": 0.2992,
+      "step": 837
+    },
+    {
+      "epoch": 0.8382357538057579,
+      "grad_norm": 0.1814364790916443,
+      "learning_rate": 0.00016383604355106837,
+      "loss": 0.3045,
+      "step": 838
+    },
+    {
+      "epoch": 0.8392360351348817,
+      "grad_norm": 0.21328268945217133,
+      "learning_rate": 0.00016375532107457108,
+      "loss": 0.3057,
+      "step": 839
+    },
+    {
+      "epoch": 0.8402363164640055,
+      "grad_norm": 0.22306102514266968,
+      "learning_rate": 0.00016367452854250603,
+      "loss": 0.3138,
+      "step": 840
+    },
+    {
+      "epoch": 0.8412365977931293,
+      "grad_norm": 0.23547765612602234,
+      "learning_rate": 0.00016359366604364972,
+      "loss": 0.3235,
+      "step": 841
+    },
+    {
+      "epoch": 0.8422368791222531,
+      "grad_norm": 0.2214178442955017,
+      "learning_rate": 0.00016351273366685526,
+      "loss": 0.3172,
+      "step": 842
+    },
+    {
+      "epoch": 0.843237160451377,
+      "grad_norm": 0.19480574131011963,
+      "learning_rate": 0.00016343173150105278,
+      "loss": 0.2859,
+      "step": 843
+    },
+    {
+      "epoch": 0.8442374417805008,
+      "grad_norm": 0.21132993698120117,
+      "learning_rate": 0.00016335065963524897,
+      "loss": 0.2701,
+      "step": 844
+    },
+    {
+      "epoch": 0.8452377231096245,
+      "grad_norm": 0.24767053127288818,
+      "learning_rate": 0.0001632695181585272,
+      "loss": 0.2852,
+      "step": 845
+    },
+    {
+      "epoch": 0.8462380044387484,
+      "grad_norm": 0.47689205408096313,
+      "learning_rate": 0.00016318830716004722,
+      "loss": 0.2959,
+      "step": 846
+    },
+    {
+      "epoch": 0.8472382857678722,
+      "grad_norm": 0.17976781725883484,
+      "learning_rate": 0.00016310702672904528,
+      "loss": 0.2866,
+      "step": 847
+    },
+    {
+      "epoch": 0.8482385670969961,
+      "grad_norm": 0.2098311185836792,
+      "learning_rate": 0.00016302567695483382,
+      "loss": 0.3111,
+      "step": 848
+    },
+    {
+      "epoch": 0.8492388484261199,
+      "grad_norm": 0.2239324301481247,
+      "learning_rate": 0.0001629442579268016,
+      "loss": 0.2564,
+      "step": 849
+    },
+    {
+      "epoch": 0.8502391297552436,
+      "grad_norm": 0.26249971985816956,
+      "learning_rate": 0.00016286276973441333,
+      "loss": 0.3033,
+      "step": 850
+    },
+    {
+      "epoch": 0.8512394110843675,
+      "grad_norm": 0.22129863500595093,
+      "learning_rate": 0.00016278121246720987,
+      "loss": 0.3152,
+      "step": 851
+    },
+    {
+      "epoch": 0.8522396924134913,
+      "grad_norm": 0.22528328001499176,
+      "learning_rate": 0.00016269958621480788,
+      "loss": 0.3163,
+      "step": 852
+    },
+    {
+      "epoch": 0.8532399737426151,
+      "grad_norm": 0.23196867108345032,
+      "learning_rate": 0.0001626178910668998,
+      "loss": 0.3082,
+      "step": 853
+    },
+    {
+      "epoch": 0.854240255071739,
+      "grad_norm": 0.19330881536006927,
+      "learning_rate": 0.00016253612711325386,
+      "loss": 0.2772,
+      "step": 854
+    },
+    {
+      "epoch": 0.8552405364008627,
+      "grad_norm": 0.2197023630142212,
+      "learning_rate": 0.0001624542944437139,
+      "loss": 0.2958,
+      "step": 855
+    },
+    {
+      "epoch": 0.8562408177299866,
+      "grad_norm": 0.21971316635608673,
+      "learning_rate": 0.00016237239314819917,
+      "loss": 0.2997,
+      "step": 856
+    },
+    {
+      "epoch": 0.8572410990591104,
+      "grad_norm": 0.24324870109558105,
+      "learning_rate": 0.0001622904233167044,
+      "loss": 0.2924,
+      "step": 857
+    },
+    {
+      "epoch": 0.8582413803882342,
+      "grad_norm": 0.2716549336910248,
+      "learning_rate": 0.0001622083850392996,
+      "loss": 0.3693,
+      "step": 858
+    },
+    {
+      "epoch": 0.859241661717358,
+      "grad_norm": 0.1929221898317337,
+      "learning_rate": 0.00016212627840613003,
+      "loss": 0.3011,
+      "step": 859
+    },
+    {
+      "epoch": 0.8602419430464818,
+      "grad_norm": 0.2041851282119751,
+      "learning_rate": 0.000162044103507416,
+      "loss": 0.3062,
+      "step": 860
+    },
+    {
+      "epoch": 0.8612422243756056,
+      "grad_norm": 0.2032964527606964,
+      "learning_rate": 0.00016196186043345288,
+      "loss": 0.2896,
+      "step": 861
+    },
+    {
+      "epoch": 0.8622425057047295,
+      "grad_norm": 0.22376392781734467,
+      "learning_rate": 0.00016187954927461093,
+      "loss": 0.2901,
+      "step": 862
+    },
+    {
+      "epoch": 0.8632427870338533,
+      "grad_norm": 0.23736202716827393,
+      "learning_rate": 0.00016179717012133521,
+      "loss": 0.2924,
+      "step": 863
+    },
+    {
+      "epoch": 0.864243068362977,
+      "grad_norm": 0.2162121683359146,
+      "learning_rate": 0.00016171472306414554,
+      "loss": 0.2655,
+      "step": 864
+    },
+    {
+      "epoch": 0.8652433496921009,
+      "grad_norm": 0.24308191239833832,
+      "learning_rate": 0.00016163220819363628,
+      "loss": 0.3062,
+      "step": 865
+    },
+    {
+      "epoch": 0.8662436310212247,
+      "grad_norm": 0.2229555994272232,
+      "learning_rate": 0.00016154962560047643,
+      "loss": 0.2748,
+      "step": 866
+    },
+    {
+      "epoch": 0.8672439123503486,
+      "grad_norm": 0.20984022319316864,
+      "learning_rate": 0.00016146697537540924,
+      "loss": 0.2777,
+      "step": 867
+    },
+    {
+      "epoch": 0.8682441936794724,
+      "grad_norm": 0.21790897846221924,
+      "learning_rate": 0.0001613842576092524,
+      "loss": 0.3167,
+      "step": 868
+    },
+    {
+      "epoch": 0.8692444750085961,
+      "grad_norm": 0.2066025286912918,
+      "learning_rate": 0.00016130147239289778,
+      "loss": 0.2897,
+      "step": 869
+    },
+    {
+      "epoch": 0.87024475633772,
+      "grad_norm": 0.20541371405124664,
+      "learning_rate": 0.00016121861981731135,
+      "loss": 0.2942,
+      "step": 870
+    },
+    {
+      "epoch": 0.8712450376668438,
+      "grad_norm": 0.25480279326438904,
+      "learning_rate": 0.00016113569997353312,
+      "loss": 0.3508,
+      "step": 871
+    },
+    {
+      "epoch": 0.8722453189959676,
+      "grad_norm": 0.2800876200199127,
+      "learning_rate": 0.000161052712952677,
+      "loss": 0.3008,
+      "step": 872
+    },
+    {
+      "epoch": 0.8732456003250915,
+      "grad_norm": 0.20582012832164764,
+      "learning_rate": 0.0001609696588459307,
+      "loss": 0.3009,
+      "step": 873
+    },
+    {
+      "epoch": 0.8742458816542152,
+      "grad_norm": 0.21027050912380219,
+      "learning_rate": 0.00016088653774455568,
+      "loss": 0.2857,
+      "step": 874
+    },
+    {
+      "epoch": 0.875246162983339,
+      "grad_norm": 0.3266025483608246,
+      "learning_rate": 0.00016080334973988695,
+      "loss": 0.3018,
+      "step": 875
+    },
+    {
+      "epoch": 0.8762464443124629,
+      "grad_norm": 0.2748381495475769,
+      "learning_rate": 0.00016072009492333318,
+      "loss": 0.3902,
+      "step": 876
+    },
+    {
+      "epoch": 0.8772467256415867,
+      "grad_norm": 0.36255359649658203,
+      "learning_rate": 0.0001606367733863763,
+      "loss": 0.2897,
+      "step": 877
+    },
+    {
+      "epoch": 0.8782470069707106,
+      "grad_norm": 0.19695764780044556,
+      "learning_rate": 0.00016055338522057158,
+      "loss": 0.3233,
+      "step": 878
+    },
+    {
+      "epoch": 0.8792472882998343,
+      "grad_norm": 0.21529537439346313,
+      "learning_rate": 0.00016046993051754756,
+      "loss": 0.2955,
+      "step": 879
+    },
+    {
+      "epoch": 0.8802475696289581,
+      "grad_norm": 0.19709599018096924,
+      "learning_rate": 0.00016038640936900586,
+      "loss": 0.2914,
+      "step": 880
+    },
+    {
+      "epoch": 0.881247850958082,
+      "grad_norm": 0.19307830929756165,
+      "learning_rate": 0.00016030282186672116,
+      "loss": 0.2867,
+      "step": 881
+    },
+    {
+      "epoch": 0.8822481322872058,
+      "grad_norm": 0.2391085922718048,
+      "learning_rate": 0.00016021916810254097,
+      "loss": 0.2666,
+      "step": 882
+    },
+    {
+      "epoch": 0.8832484136163296,
+      "grad_norm": 0.18861837685108185,
+      "learning_rate": 0.00016013544816838565,
+      "loss": 0.2658,
+      "step": 883
+    },
+    {
+      "epoch": 0.8842486949454534,
+      "grad_norm": 0.22086991369724274,
+      "learning_rate": 0.00016005166215624827,
+      "loss": 0.3262,
+      "step": 884
+    },
+    {
+      "epoch": 0.8852489762745772,
+      "grad_norm": 0.21743939816951752,
+      "learning_rate": 0.0001599678101581945,
+      "loss": 0.2986,
+      "step": 885
+    },
+    {
+      "epoch": 0.886249257603701,
+      "grad_norm": 0.2183314710855484,
+      "learning_rate": 0.00015988389226636253,
+      "loss": 0.3179,
+      "step": 886
+    },
+    {
+      "epoch": 0.8872495389328249,
+      "grad_norm": 0.19507034122943878,
+      "learning_rate": 0.00015979990857296295,
+      "loss": 0.3036,
+      "step": 887
+    },
+    {
+      "epoch": 0.8882498202619487,
+      "grad_norm": 0.20520102977752686,
+      "learning_rate": 0.00015971585917027862,
+      "loss": 0.2755,
+      "step": 888
+    },
+    {
+      "epoch": 0.8892501015910725,
+      "grad_norm": 0.17556002736091614,
+      "learning_rate": 0.00015963174415066468,
+      "loss": 0.2464,
+      "step": 889
+    },
+    {
+      "epoch": 0.8902503829201963,
+      "grad_norm": 0.2668997645378113,
+      "learning_rate": 0.0001595475636065483,
+      "loss": 0.3473,
+      "step": 890
+    },
+    {
+      "epoch": 0.8912506642493201,
+      "grad_norm": 0.18344111740589142,
+      "learning_rate": 0.00015946331763042867,
+      "loss": 0.2849,
+      "step": 891
+    },
+    {
+      "epoch": 0.892250945578444,
+      "grad_norm": 0.20132502913475037,
+      "learning_rate": 0.00015937900631487686,
+      "loss": 0.2914,
+      "step": 892
+    },
+    {
+      "epoch": 0.8932512269075678,
+      "grad_norm": 0.26806825399398804,
+      "learning_rate": 0.00015929462975253585,
+      "loss": 0.2491,
+      "step": 893
+    },
+    {
+      "epoch": 0.8942515082366915,
+      "grad_norm": 0.20963765680789948,
+      "learning_rate": 0.00015921018803612014,
+      "loss": 0.298,
+      "step": 894
+    },
+    {
+      "epoch": 0.8952517895658154,
+      "grad_norm": 0.18469758331775665,
+      "learning_rate": 0.0001591256812584159,
+      "loss": 0.2344,
+      "step": 895
+    },
+    {
+      "epoch": 0.8962520708949392,
+      "grad_norm": 0.2107250988483429,
+      "learning_rate": 0.00015904110951228082,
+      "loss": 0.311,
+      "step": 896
+    },
+    {
+      "epoch": 0.897252352224063,
+      "grad_norm": 0.20409642159938812,
+      "learning_rate": 0.00015895647289064396,
+      "loss": 0.3288,
+      "step": 897
+    },
+    {
+      "epoch": 0.8982526335531869,
+      "grad_norm": 0.21404282748699188,
+      "learning_rate": 0.00015887177148650564,
+      "loss": 0.316,
+      "step": 898
+    },
+    {
+      "epoch": 0.8992529148823106,
+      "grad_norm": 0.20457734167575836,
+      "learning_rate": 0.0001587870053929374,
+      "loss": 0.2643,
+      "step": 899
+    },
+    {
+      "epoch": 0.9002531962114345,
+      "grad_norm": 0.1898927241563797,
+      "learning_rate": 0.00015870217470308188,
+      "loss": 0.2944,
+      "step": 900
+    },
+    {
+      "epoch": 0.9012534775405583,
+      "grad_norm": 0.18605229258537292,
+      "learning_rate": 0.0001586172795101526,
+      "loss": 0.2657,
+      "step": 901
+    },
+    {
+      "epoch": 0.9022537588696821,
+      "grad_norm": 0.25137388706207275,
+      "learning_rate": 0.00015853231990743406,
+      "loss": 0.2839,
+      "step": 902
+    },
+    {
+      "epoch": 0.903254040198806,
+      "grad_norm": 0.20574310421943665,
+      "learning_rate": 0.0001584472959882815,
+      "loss": 0.2923,
+      "step": 903
+    },
+    {
+      "epoch": 0.9042543215279297,
+      "grad_norm": 0.23242512345314026,
+      "learning_rate": 0.00015836220784612085,
+      "loss": 0.3062,
+      "step": 904
+    },
+    {
+      "epoch": 0.9052546028570535,
+      "grad_norm": 0.2369641214609146,
+      "learning_rate": 0.00015827705557444852,
+      "loss": 0.3301,
+      "step": 905
+    },
+    {
+      "epoch": 0.9062548841861774,
+      "grad_norm": 0.22508639097213745,
+      "learning_rate": 0.00015819183926683153,
+      "loss": 0.2826,
+      "step": 906
+    },
+    {
+      "epoch": 0.9072551655153012,
+      "grad_norm": 0.17570531368255615,
+      "learning_rate": 0.00015810655901690715,
+      "loss": 0.2963,
+      "step": 907
+    },
+    {
+      "epoch": 0.9082554468444249,
+      "grad_norm": 0.19561176002025604,
+      "learning_rate": 0.00015802121491838297,
+      "loss": 0.2882,
+      "step": 908
+    },
+    {
+      "epoch": 0.9092557281735488,
+      "grad_norm": 0.2130153924226761,
+      "learning_rate": 0.0001579358070650367,
+      "loss": 0.2761,
+      "step": 909
+    },
+    {
+      "epoch": 0.9102560095026726,
+      "grad_norm": 0.21141602098941803,
+      "learning_rate": 0.00015785033555071616,
+      "loss": 0.2816,
+      "step": 910
+    },
+    {
+      "epoch": 0.9112562908317965,
+      "grad_norm": 0.22545745968818665,
+      "learning_rate": 0.00015776480046933905,
+      "loss": 0.2737,
+      "step": 911
+    },
+    {
+      "epoch": 0.9122565721609203,
+      "grad_norm": 0.18807393312454224,
+      "learning_rate": 0.000157679201914893,
+      "loss": 0.2588,
+      "step": 912
+    },
+    {
+      "epoch": 0.913256853490044,
+      "grad_norm": 0.2146979570388794,
+      "learning_rate": 0.00015759353998143528,
+      "loss": 0.3129,
+      "step": 913
+    },
+    {
+      "epoch": 0.9142571348191679,
+      "grad_norm": 0.21207363903522491,
+      "learning_rate": 0.00015750781476309288,
+      "loss": 0.3245,
+      "step": 914
+    },
+    {
+      "epoch": 0.9152574161482917,
+      "grad_norm": 0.1911744475364685,
+      "learning_rate": 0.00015742202635406235,
+      "loss": 0.2846,
+      "step": 915
+    },
+    {
+      "epoch": 0.9162576974774155,
+      "grad_norm": 0.21280640363693237,
+      "learning_rate": 0.00015733617484860963,
+      "loss": 0.3086,
+      "step": 916
+    },
+    {
+      "epoch": 0.9172579788065394,
+      "grad_norm": 0.19109156727790833,
+      "learning_rate": 0.00015725026034106996,
+      "loss": 0.2856,
+      "step": 917
+    },
+    {
+      "epoch": 0.9182582601356631,
+      "grad_norm": 0.3014916777610779,
+      "learning_rate": 0.00015716428292584787,
+      "loss": 0.2946,
+      "step": 918
+    },
+    {
+      "epoch": 0.919258541464787,
+      "grad_norm": 0.5512151718139648,
+      "learning_rate": 0.00015707824269741702,
+      "loss": 0.2777,
+      "step": 919
+    },
+    {
+      "epoch": 0.9202588227939108,
+      "grad_norm": 0.20209389925003052,
+      "learning_rate": 0.00015699213975031996,
+      "loss": 0.2888,
+      "step": 920
+    },
+    {
+      "epoch": 0.9212591041230346,
+      "grad_norm": 0.20380209386348724,
+      "learning_rate": 0.0001569059741791684,
+      "loss": 0.3165,
+      "step": 921
+    },
+    {
+      "epoch": 0.9222593854521585,
+      "grad_norm": 0.18393391370773315,
+      "learning_rate": 0.0001568197460786426,
+      "loss": 0.2892,
+      "step": 922
+    },
+    {
+      "epoch": 0.9232596667812822,
+      "grad_norm": 0.18897509574890137,
+      "learning_rate": 0.0001567334555434917,
+      "loss": 0.2666,
+      "step": 923
+    },
+    {
+      "epoch": 0.924259948110406,
+      "grad_norm": 0.2418288290500641,
+      "learning_rate": 0.0001566471026685334,
+      "loss": 0.32,
+      "step": 924
+    },
+    {
+      "epoch": 0.9252602294395299,
+      "grad_norm": 0.22516793012619019,
+      "learning_rate": 0.00015656068754865387,
+      "loss": 0.3005,
+      "step": 925
+    },
+    {
+      "epoch": 0.9262605107686537,
+      "grad_norm": 0.1989380270242691,
+      "learning_rate": 0.00015647421027880772,
+      "loss": 0.2946,
+      "step": 926
+    },
+    {
+      "epoch": 0.9272607920977775,
+      "grad_norm": 0.1808221936225891,
+      "learning_rate": 0.0001563876709540178,
+      "loss": 0.2778,
+      "step": 927
+    },
+    {
+      "epoch": 0.9282610734269013,
+      "grad_norm": 0.17342513799667358,
+      "learning_rate": 0.0001563010696693752,
+      "loss": 0.2686,
+      "step": 928
+    },
+    {
+      "epoch": 0.9292613547560251,
+      "grad_norm": 0.21851111948490143,
+      "learning_rate": 0.00015621440652003907,
+      "loss": 0.2956,
+      "step": 929
+    },
+    {
+      "epoch": 0.930261636085149,
+      "grad_norm": 0.20432516932487488,
+      "learning_rate": 0.00015612768160123652,
+      "loss": 0.3276,
+      "step": 930
+    },
+    {
+      "epoch": 0.9312619174142728,
+      "grad_norm": 0.19698838889598846,
+      "learning_rate": 0.00015604089500826257,
+      "loss": 0.2823,
+      "step": 931
+    },
+    {
+      "epoch": 0.9322621987433966,
+      "grad_norm": 0.22568564116954803,
+      "learning_rate": 0.00015595404683648,
+      "loss": 0.2992,
+      "step": 932
+    },
+    {
+      "epoch": 0.9332624800725204,
+      "grad_norm": 0.17305763065814972,
+      "learning_rate": 0.00015586713718131922,
+      "loss": 0.2596,
+      "step": 933
+    },
+    {
+      "epoch": 0.9342627614016442,
+      "grad_norm": 0.2285163849592209,
+      "learning_rate": 0.0001557801661382782,
+      "loss": 0.3317,
+      "step": 934
+    },
+    {
+      "epoch": 0.935263042730768,
+      "grad_norm": 0.23833929002285004,
+      "learning_rate": 0.00015569313380292248,
+      "loss": 0.3479,
+      "step": 935
+    },
+    {
+      "epoch": 0.9362633240598919,
+      "grad_norm": 0.26495903730392456,
+      "learning_rate": 0.00015560604027088477,
+      "loss": 0.3403,
+      "step": 936
+    },
+    {
+      "epoch": 0.9372636053890157,
+      "grad_norm": 0.19236880540847778,
+      "learning_rate": 0.00015551888563786515,
+      "loss": 0.2995,
+      "step": 937
+    },
+    {
+      "epoch": 0.9382638867181394,
+      "grad_norm": 0.23580637574195862,
+      "learning_rate": 0.00015543166999963076,
+      "loss": 0.2828,
+      "step": 938
+    },
+    {
+      "epoch": 0.9392641680472633,
+      "grad_norm": 0.2398647964000702,
+      "learning_rate": 0.0001553443934520159,
+      "loss": 0.3128,
+      "step": 939
+    },
+    {
+      "epoch": 0.9402644493763871,
+      "grad_norm": 0.2201540470123291,
+      "learning_rate": 0.00015525705609092157,
+      "loss": 0.3219,
+      "step": 940
+    },
+    {
+      "epoch": 0.941264730705511,
+      "grad_norm": 0.19159969687461853,
+      "learning_rate": 0.00015516965801231586,
+      "loss": 0.255,
+      "step": 941
+    },
+    {
+      "epoch": 0.9422650120346348,
+      "grad_norm": 0.22276592254638672,
+      "learning_rate": 0.0001550821993122334,
+      "loss": 0.3208,
+      "step": 942
+    },
+    {
+      "epoch": 0.9432652933637585,
+      "grad_norm": 0.1987476497888565,
+      "learning_rate": 0.0001549946800867755,
+      "loss": 0.3068,
+      "step": 943
+    },
+    {
+      "epoch": 0.9442655746928824,
+      "grad_norm": 0.18668995797634125,
+      "learning_rate": 0.00015490710043210997,
+      "loss": 0.295,
+      "step": 944
+    },
+    {
+      "epoch": 0.9452658560220062,
+      "grad_norm": 0.2638959288597107,
+      "learning_rate": 0.00015481946044447099,
+      "loss": 0.3201,
+      "step": 945
+    },
+    {
+      "epoch": 0.94626613735113,
+      "grad_norm": 0.8796263933181763,
+      "learning_rate": 0.00015473176022015906,
+      "loss": 0.2667,
+      "step": 946
+    },
+    {
+      "epoch": 0.9472664186802539,
+      "grad_norm": 0.19040840864181519,
+      "learning_rate": 0.0001546439998555409,
+      "loss": 0.269,
+      "step": 947
+    },
+    {
+      "epoch": 0.9482667000093776,
+      "grad_norm": 0.15700708329677582,
+      "learning_rate": 0.0001545561794470492,
+      "loss": 0.2426,
+      "step": 948
+    },
+    {
+      "epoch": 0.9492669813385014,
+      "grad_norm": 0.28926882147789,
+      "learning_rate": 0.00015446829909118275,
+      "loss": 0.2785,
+      "step": 949
+    },
+    {
+      "epoch": 0.9502672626676253,
+      "grad_norm": 0.24721843004226685,
+      "learning_rate": 0.00015438035888450623,
+      "loss": 0.2548,
+      "step": 950
+    },
+    {
+      "epoch": 0.9512675439967491,
+      "grad_norm": 0.22793874144554138,
+      "learning_rate": 0.00015429235892364994,
+      "loss": 0.2897,
+      "step": 951
+    },
+    {
+      "epoch": 0.952267825325873,
+      "grad_norm": 0.19813783466815948,
+      "learning_rate": 0.00015420429930530996,
+      "loss": 0.2933,
+      "step": 952
+    },
+    {
+      "epoch": 0.9532681066549967,
+      "grad_norm": 0.2070704698562622,
+      "learning_rate": 0.00015411618012624786,
+      "loss": 0.299,
+      "step": 953
+    },
+    {
+      "epoch": 0.9542683879841205,
+      "grad_norm": 0.19984149932861328,
+      "learning_rate": 0.00015402800148329071,
+      "loss": 0.3293,
+      "step": 954
+    },
+    {
+      "epoch": 0.9552686693132444,
+      "grad_norm": 0.21448279917240143,
+      "learning_rate": 0.00015393976347333088,
+      "loss": 0.2679,
+      "step": 955
+    },
+    {
+      "epoch": 0.9562689506423682,
+      "grad_norm": 0.19312746822834015,
+      "learning_rate": 0.00015385146619332596,
+      "loss": 0.2829,
+      "step": 956
+    },
+    {
+      "epoch": 0.9572692319714919,
+      "grad_norm": 0.2931583821773529,
+      "learning_rate": 0.00015376310974029873,
+      "loss": 0.2786,
+      "step": 957
+    },
+    {
+      "epoch": 0.9582695133006158,
+      "grad_norm": 0.2195436805486679,
+      "learning_rate": 0.00015367469421133695,
+      "loss": 0.2761,
+      "step": 958
+    },
+    {
+      "epoch": 0.9592697946297396,
+      "grad_norm": 0.20441684126853943,
+      "learning_rate": 0.00015358621970359325,
+      "loss": 0.2926,
+      "step": 959
+    },
+    {
+      "epoch": 0.9602700759588634,
+      "grad_norm": 0.1912553310394287,
+      "learning_rate": 0.00015349768631428519,
+      "loss": 0.2807,
+      "step": 960
+    },
+    {
+      "epoch": 0.9612703572879873,
+      "grad_norm": 0.1885378062725067,
+      "learning_rate": 0.00015340909414069488,
+      "loss": 0.3156,
+      "step": 961
+    },
+    {
+      "epoch": 0.962270638617111,
+      "grad_norm": 0.20673014223575592,
+      "learning_rate": 0.00015332044328016914,
+      "loss": 0.2812,
+      "step": 962
+    },
+    {
+      "epoch": 0.9632709199462349,
+      "grad_norm": 0.22105248272418976,
+      "learning_rate": 0.0001532317338301192,
+      "loss": 0.2856,
+      "step": 963
+    },
+    {
+      "epoch": 0.9642712012753587,
+      "grad_norm": 0.29232075810432434,
+      "learning_rate": 0.00015314296588802076,
+      "loss": 0.3532,
+      "step": 964
+    },
+    {
+      "epoch": 0.9652714826044825,
+      "grad_norm": 0.22949157655239105,
+      "learning_rate": 0.00015305413955141365,
+      "loss": 0.3059,
+      "step": 965
+    },
+    {
+      "epoch": 0.9662717639336064,
+      "grad_norm": 0.20151565968990326,
+      "learning_rate": 0.00015296525491790205,
+      "loss": 0.2736,
+      "step": 966
+    },
+    {
+      "epoch": 0.9672720452627301,
+      "grad_norm": 0.31306812167167664,
+      "learning_rate": 0.00015287631208515406,
+      "loss": 0.3002,
+      "step": 967
+    },
+    {
+      "epoch": 0.9682723265918539,
+      "grad_norm": 0.17731359601020813,
+      "learning_rate": 0.00015278731115090171,
+      "loss": 0.2624,
+      "step": 968
+    },
+    {
+      "epoch": 0.9692726079209778,
+      "grad_norm": 0.3789874017238617,
+      "learning_rate": 0.00015269825221294098,
+      "loss": 0.2818,
+      "step": 969
+    },
+    {
+      "epoch": 0.9702728892501016,
+      "grad_norm": 0.24409924447536469,
+      "learning_rate": 0.00015260913536913154,
+      "loss": 0.3363,
+      "step": 970
+    },
+    {
+      "epoch": 0.9712731705792254,
+      "grad_norm": 0.21300776302814484,
+      "learning_rate": 0.00015251996071739664,
+      "loss": 0.3053,
+      "step": 971
+    },
+    {
+      "epoch": 0.9722734519083492,
+      "grad_norm": 0.2754755914211273,
+      "learning_rate": 0.00015243072835572318,
+      "loss": 0.2633,
+      "step": 972
+    },
+    {
+      "epoch": 0.973273733237473,
+      "grad_norm": 0.21204949915409088,
+      "learning_rate": 0.0001523414383821613,
+      "loss": 0.2976,
+      "step": 973
+    },
+    {
+      "epoch": 0.9742740145665969,
+      "grad_norm": 0.2121218591928482,
+      "learning_rate": 0.00015225209089482462,
+      "loss": 0.2698,
+      "step": 974
+    },
+    {
+      "epoch": 0.9752742958957207,
+      "grad_norm": 0.19355735182762146,
+      "learning_rate": 0.0001521626859918898,
+      "loss": 0.2931,
+      "step": 975
+    },
+    {
+      "epoch": 0.9762745772248445,
+      "grad_norm": 0.22311675548553467,
+      "learning_rate": 0.00015207322377159668,
+      "loss": 0.3175,
+      "step": 976
+    },
+    {
+      "epoch": 0.9772748585539683,
+      "grad_norm": 0.17831672728061676,
+      "learning_rate": 0.00015198370433224805,
+      "loss": 0.2727,
+      "step": 977
+    },
+    {
+      "epoch": 0.9782751398830921,
+      "grad_norm": 0.24245727062225342,
+      "learning_rate": 0.00015189412777220958,
+      "loss": 0.3186,
+      "step": 978
+    },
+    {
+      "epoch": 0.9792754212122159,
+      "grad_norm": 0.182240828871727,
+      "learning_rate": 0.00015180449418990976,
+      "loss": 0.3001,
+      "step": 979
+    },
+    {
+      "epoch": 0.9802757025413398,
+      "grad_norm": 0.1840401440858841,
+      "learning_rate": 0.00015171480368383964,
+      "loss": 0.2987,
+      "step": 980
+    },
+    {
+      "epoch": 0.9812759838704636,
+      "grad_norm": 0.23978300392627716,
+      "learning_rate": 0.00015162505635255287,
+      "loss": 0.2963,
+      "step": 981
+    },
+    {
+      "epoch": 0.9822762651995874,
+      "grad_norm": 0.22814355790615082,
+      "learning_rate": 0.00015153525229466555,
+      "loss": 0.2765,
+      "step": 982
+    },
+    {
+      "epoch": 0.9832765465287112,
+      "grad_norm": 0.18747608363628387,
+      "learning_rate": 0.00015144539160885613,
+      "loss": 0.259,
+      "step": 983
+    },
+    {
+      "epoch": 0.984276827857835,
+      "grad_norm": 0.22687505185604095,
+      "learning_rate": 0.00015135547439386516,
+      "loss": 0.3158,
+      "step": 984
+    },
+    {
+      "epoch": 0.9852771091869589,
+      "grad_norm": 0.2270641028881073,
+      "learning_rate": 0.0001512655007484955,
+      "loss": 0.2649,
+      "step": 985
+    },
+    {
+      "epoch": 0.9862773905160827,
+      "grad_norm": 0.18852733075618744,
+      "learning_rate": 0.00015117547077161185,
+      "loss": 0.3217,
+      "step": 986
+    },
+    {
+      "epoch": 0.9872776718452064,
+      "grad_norm": 0.17261897027492523,
+      "learning_rate": 0.0001510853845621409,
+      "loss": 0.2583,
+      "step": 987
+    },
+    {
+      "epoch": 0.9882779531743303,
+      "grad_norm": 0.1948390007019043,
+      "learning_rate": 0.00015099524221907107,
+      "loss": 0.3033,
+      "step": 988
+    },
+    {
+      "epoch": 0.9892782345034541,
+      "grad_norm": 0.20993740856647491,
+      "learning_rate": 0.0001509050438414525,
+      "loss": 0.2888,
+      "step": 989
+    },
+    {
+      "epoch": 0.9902785158325779,
+      "grad_norm": 0.23788276314735413,
+      "learning_rate": 0.00015081478952839693,
+      "loss": 0.3152,
+      "step": 990
+    },
+    {
+      "epoch": 0.9912787971617018,
+      "grad_norm": 0.18581551313400269,
+      "learning_rate": 0.00015072447937907753,
+      "loss": 0.2962,
+      "step": 991
+    },
+    {
+      "epoch": 0.9922790784908255,
+      "grad_norm": 0.21751491725444794,
+      "learning_rate": 0.00015063411349272877,
+      "loss": 0.2899,
+      "step": 992
+    },
+    {
+      "epoch": 0.9932793598199494,
+      "grad_norm": 0.16804128885269165,
+      "learning_rate": 0.00015054369196864644,
+      "loss": 0.2378,
+      "step": 993
+    },
+    {
+      "epoch": 0.9942796411490732,
+      "grad_norm": 0.22858916223049164,
+      "learning_rate": 0.00015045321490618748,
+      "loss": 0.3072,
+      "step": 994
+    },
+    {
+      "epoch": 0.995279922478197,
+      "grad_norm": 0.21114645898342133,
+      "learning_rate": 0.00015036268240476978,
+      "loss": 0.2869,
+      "step": 995
+    },
+    {
+      "epoch": 0.9962802038073209,
+      "grad_norm": 0.23004184663295746,
+      "learning_rate": 0.00015027209456387218,
+      "loss": 0.262,
+      "step": 996
+    },
+    {
+      "epoch": 0.9972804851364446,
+      "grad_norm": 0.19193367660045624,
+      "learning_rate": 0.00015018145148303438,
+      "loss": 0.261,
+      "step": 997
+    },
+    {
+      "epoch": 0.9982807664655684,
+      "grad_norm": 0.21400488913059235,
+      "learning_rate": 0.00015009075326185667,
+      "loss": 0.3506,
+      "step": 998
+    },
+    {
+      "epoch": 0.9992810477946923,
+      "grad_norm": 0.18997825682163239,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.2638,
+      "step": 999
+    },
+    {
+      "epoch": 1.0002813291238162,
+      "grad_norm": 0.17756420373916626,
+      "learning_rate": 0.00014990919179718584,
+      "loss": 0.2457,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0012816104529398,
+      "grad_norm": 0.17825394868850708,
+      "learning_rate": 0.00014981832875319597,
+      "loss": 0.2287,
+      "step": 1001
+    },
+    {
+      "epoch": 1.0022818917820637,
+      "grad_norm": 0.1440640389919281,
+      "learning_rate": 0.00014972741096787242,
+      "loss": 0.1833,
+      "step": 1002
+    },
+    {
+      "epoch": 1.0032821731111876,
+      "grad_norm": 0.21111971139907837,
+      "learning_rate": 0.0001496364385411174,
+      "loss": 0.1999,
+      "step": 1003
+    },
+    {
+      "epoch": 1.0042824544403113,
+      "grad_norm": 0.17393074929714203,
+      "learning_rate": 0.0001495454115728932,
+      "loss": 0.2021,
+      "step": 1004
+    },
+    {
+      "epoch": 1.0052827357694352,
+      "grad_norm": 0.15784429013729095,
+      "learning_rate": 0.0001494543301632219,
+      "loss": 0.202,
+      "step": 1005
+    },
+    {
+      "epoch": 1.006283017098559,
+      "grad_norm": 0.20182862877845764,
+      "learning_rate": 0.00014936319441218555,
+      "loss": 0.2061,
+      "step": 1006
+    },
+    {
+      "epoch": 1.0072832984276827,
+      "grad_norm": 0.16590118408203125,
+      "learning_rate": 0.0001492720044199259,
+      "loss": 0.1997,
+      "step": 1007
+    },
+    {
+      "epoch": 1.0082835797568066,
+      "grad_norm": 0.14610984921455383,
+      "learning_rate": 0.0001491807602866442,
+      "loss": 0.1819,
+      "step": 1008
+    },
+    {
+      "epoch": 1.0092838610859305,
+      "grad_norm": 0.18120163679122925,
+      "learning_rate": 0.00014908946211260123,
+      "loss": 0.1992,
+      "step": 1009
+    },
+    {
+      "epoch": 1.0102841424150542,
+      "grad_norm": 0.16269902884960175,
+      "learning_rate": 0.00014899810999811726,
+      "loss": 0.1812,
+      "step": 1010
+    },
+    {
+      "epoch": 1.011284423744178,
+      "grad_norm": 0.19697803258895874,
+      "learning_rate": 0.0001489067040435717,
+      "loss": 0.1978,
+      "step": 1011
+    },
+    {
+      "epoch": 1.012284705073302,
+      "grad_norm": 0.1755535751581192,
+      "learning_rate": 0.00014881524434940313,
+      "loss": 0.2263,
+      "step": 1012
+    },
+    {
+      "epoch": 1.0132849864024256,
+      "grad_norm": 0.16891999542713165,
+      "learning_rate": 0.0001487237310161093,
+      "loss": 0.2039,
+      "step": 1013
+    },
+    {
+      "epoch": 1.0142852677315495,
+      "grad_norm": 0.19249218702316284,
+      "learning_rate": 0.0001486321641442467,
+      "loss": 0.2069,
+      "step": 1014
+    },
+    {
+      "epoch": 1.0152855490606734,
+      "grad_norm": 0.2377549409866333,
+      "learning_rate": 0.00014854054383443081,
+      "loss": 0.2192,
+      "step": 1015
+    },
+    {
+      "epoch": 1.016285830389797,
+      "grad_norm": 0.17392681539058685,
+      "learning_rate": 0.00014844887018733582,
+      "loss": 0.2221,
+      "step": 1016
+    },
+    {
+      "epoch": 1.017286111718921,
+      "grad_norm": 0.24536015093326569,
+      "learning_rate": 0.00014835714330369446,
+      "loss": 0.2174,
+      "step": 1017
+    },
+    {
+      "epoch": 1.0182863930480448,
+      "grad_norm": 0.19177570939064026,
+      "learning_rate": 0.00014826536328429795,
+      "loss": 0.2125,
+      "step": 1018
+    },
+    {
+      "epoch": 1.0192866743771685,
+      "grad_norm": 0.1779998242855072,
+      "learning_rate": 0.000148173530229996,
+      "loss": 0.2146,
+      "step": 1019
+    },
+    {
+      "epoch": 1.0202869557062924,
+      "grad_norm": 0.17234370112419128,
+      "learning_rate": 0.00014808164424169647,
+      "loss": 0.2017,
+      "step": 1020
+    },
+    {
+      "epoch": 1.0212872370354162,
+      "grad_norm": 0.16123813390731812,
+      "learning_rate": 0.0001479897054203655,
+      "loss": 0.1816,
+      "step": 1021
+    },
+    {
+      "epoch": 1.0222875183645401,
+      "grad_norm": 0.2026429921388626,
+      "learning_rate": 0.00014789771386702717,
+      "loss": 0.1863,
+      "step": 1022
+    },
+    {
+      "epoch": 1.0232877996936638,
+      "grad_norm": 0.182438462972641,
+      "learning_rate": 0.0001478056696827636,
+      "loss": 0.1731,
+      "step": 1023
+    },
+    {
+      "epoch": 1.0242880810227877,
+      "grad_norm": 0.24581795930862427,
+      "learning_rate": 0.0001477135729687147,
+      "loss": 0.2082,
+      "step": 1024
+    },
+    {
+      "epoch": 1.0252883623519116,
+      "grad_norm": 0.2977474331855774,
+      "learning_rate": 0.0001476214238260781,
+      "loss": 0.2201,
+      "step": 1025
+    },
+    {
+      "epoch": 1.0262886436810352,
+      "grad_norm": 0.1638759821653366,
+      "learning_rate": 0.000147529222356109,
+      "loss": 0.176,
+      "step": 1026
+    },
+    {
+      "epoch": 1.0272889250101591,
+      "grad_norm": 0.5346854329109192,
+      "learning_rate": 0.0001474369686601202,
+      "loss": 0.2209,
+      "step": 1027
+    },
+    {
+      "epoch": 1.028289206339283,
+      "grad_norm": 0.20119376480579376,
+      "learning_rate": 0.0001473446628394818,
+      "loss": 0.1938,
+      "step": 1028
+    },
+    {
+      "epoch": 1.0292894876684067,
+      "grad_norm": 0.187885120511055,
+      "learning_rate": 0.00014725230499562119,
+      "loss": 0.2383,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0302897689975306,
+      "grad_norm": 0.4424345791339874,
+      "learning_rate": 0.00014715989523002296,
+      "loss": 0.2074,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0312900503266544,
+      "grad_norm": 0.1824462115764618,
+      "learning_rate": 0.00014706743364422878,
+      "loss": 0.2062,
+      "step": 1031
+    },
+    {
+      "epoch": 1.032290331655778,
+      "grad_norm": 0.1551816165447235,
+      "learning_rate": 0.00014697492033983707,
+      "loss": 0.1933,
+      "step": 1032
+    },
+    {
+      "epoch": 1.033290612984902,
+      "grad_norm": 0.1997290551662445,
+      "learning_rate": 0.00014688235541850337,
+      "loss": 0.2042,
+      "step": 1033
+    },
+    {
+      "epoch": 1.0342908943140259,
+      "grad_norm": 0.2991243004798889,
+      "learning_rate": 0.0001467897389819397,
+      "loss": 0.1909,
+      "step": 1034
+    },
+    {
+      "epoch": 1.0352911756431495,
+      "grad_norm": 0.19960065186023712,
+      "learning_rate": 0.00014669707113191483,
+      "loss": 0.2173,
+      "step": 1035
+    },
+    {
+      "epoch": 1.0362914569722734,
+      "grad_norm": 0.22429999709129333,
+      "learning_rate": 0.0001466043519702539,
+      "loss": 0.1971,
+      "step": 1036
+    },
+    {
+      "epoch": 1.0372917383013973,
+      "grad_norm": 0.17386357486248016,
+      "learning_rate": 0.00014651158159883855,
+      "loss": 0.1928,
+      "step": 1037
+    },
+    {
+      "epoch": 1.038292019630521,
+      "grad_norm": 0.17837831377983093,
+      "learning_rate": 0.0001464187601196066,
+      "loss": 0.2118,
+      "step": 1038
+    },
+    {
+      "epoch": 1.0392923009596449,
+      "grad_norm": 0.18373790383338928,
+      "learning_rate": 0.00014632588763455212,
+      "loss": 0.1941,
+      "step": 1039
+    },
+    {
+      "epoch": 1.0402925822887688,
+      "grad_norm": 0.16558334231376648,
+      "learning_rate": 0.00014623296424572517,
+      "loss": 0.183,
+      "step": 1040
+    },
+    {
+      "epoch": 1.0412928636178926,
+      "grad_norm": 0.21426749229431152,
+      "learning_rate": 0.00014613999005523174,
+      "loss": 0.2239,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0422931449470163,
+      "grad_norm": 0.20320329070091248,
+      "learning_rate": 0.00014604696516523361,
+      "loss": 0.2186,
+      "step": 1042
+    },
+    {
+      "epoch": 1.0432934262761402,
+      "grad_norm": 0.1887698471546173,
+      "learning_rate": 0.00014595388967794835,
+      "loss": 0.2052,
+      "step": 1043
+    },
+    {
+      "epoch": 1.044293707605264,
+      "grad_norm": 0.22706736624240875,
+      "learning_rate": 0.00014586076369564908,
+      "loss": 0.2334,
+      "step": 1044
+    },
+    {
+      "epoch": 1.0452939889343877,
+      "grad_norm": 0.20627042651176453,
+      "learning_rate": 0.00014576758732066442,
+      "loss": 0.2071,
+      "step": 1045
+    },
+    {
+      "epoch": 1.0462942702635116,
+      "grad_norm": 0.2628501355648041,
+      "learning_rate": 0.00014567436065537835,
+      "loss": 0.2071,
+      "step": 1046
+    },
+    {
+      "epoch": 1.0472945515926355,
+      "grad_norm": 0.24056898057460785,
+      "learning_rate": 0.00014558108380223012,
+      "loss": 0.1892,
+      "step": 1047
+    },
+    {
+      "epoch": 1.0482948329217592,
+      "grad_norm": 0.14923742413520813,
+      "learning_rate": 0.00014548775686371412,
+      "loss": 0.1759,
+      "step": 1048
+    },
+    {
+      "epoch": 1.049295114250883,
+      "grad_norm": 0.18112002313137054,
+      "learning_rate": 0.00014539437994237977,
+      "loss": 0.2123,
+      "step": 1049
+    },
+    {
+      "epoch": 1.050295395580007,
+      "grad_norm": 0.18290945887565613,
+      "learning_rate": 0.00014530095314083143,
+      "loss": 0.1786,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0512956769091306,
+      "grad_norm": 0.2807084321975708,
+      "learning_rate": 0.00014520747656172824,
+      "loss": 0.2074,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0522959582382545,
+      "grad_norm": 0.2327062338590622,
+      "learning_rate": 0.00014511395030778406,
+      "loss": 0.1823,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0532962395673784,
+      "grad_norm": 0.17464283108711243,
+      "learning_rate": 0.00014502037448176734,
+      "loss": 0.1877,
+      "step": 1053
+    },
+    {
+      "epoch": 1.054296520896502,
+      "grad_norm": 0.17308488488197327,
+      "learning_rate": 0.000144926749186501,
+      "loss": 0.2139,
+      "step": 1054
+    },
+    {
+      "epoch": 1.055296802225626,
+      "grad_norm": 0.19831931591033936,
+      "learning_rate": 0.00014483307452486227,
+      "loss": 0.2044,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0562970835547498,
+      "grad_norm": 0.1914462447166443,
+      "learning_rate": 0.0001447393505997827,
+      "loss": 0.2284,
+      "step": 1056
+    },
+    {
+      "epoch": 1.0572973648838735,
+      "grad_norm": 0.37563416361808777,
+      "learning_rate": 0.00014464557751424793,
+      "loss": 0.1837,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0582976462129974,
+      "grad_norm": 0.25262922048568726,
+      "learning_rate": 0.00014455175537129758,
+      "loss": 0.1962,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0592979275421213,
+      "grad_norm": 0.15078376233577728,
+      "learning_rate": 0.00014445788427402528,
+      "loss": 0.173,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0602982088712452,
+      "grad_norm": 0.17346541583538055,
+      "learning_rate": 0.00014436396432557835,
+      "loss": 0.1834,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0612984902003688,
+      "grad_norm": 0.17446568608283997,
+      "learning_rate": 0.00014426999562915782,
+      "loss": 0.1883,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0622987715294927,
+      "grad_norm": 0.20120836794376373,
+      "learning_rate": 0.00014417597828801832,
+      "loss": 0.2135,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0632990528586166,
+      "grad_norm": 0.2236095517873764,
+      "learning_rate": 0.0001440819124054679,
+      "loss": 0.2565,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0642993341877403,
+      "grad_norm": 0.1920352578163147,
+      "learning_rate": 0.00014398779808486793,
+      "loss": 0.194,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0652996155168641,
+      "grad_norm": 0.15097807347774506,
+      "learning_rate": 0.00014389363542963306,
+      "loss": 0.1897,
+      "step": 1065
+    },
+    {
+      "epoch": 1.066299896845988,
+      "grad_norm": 0.27446869015693665,
+      "learning_rate": 0.000143799424543231,
+      "loss": 0.1961,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0673001781751117,
+      "grad_norm": 0.18010550737380981,
+      "learning_rate": 0.0001437051655291825,
+      "loss": 0.2012,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0683004595042356,
+      "grad_norm": 0.1741923838853836,
+      "learning_rate": 0.0001436108584910611,
+      "loss": 0.195,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0693007408333595,
+      "grad_norm": 0.17451412975788116,
+      "learning_rate": 0.0001435165035324933,
+      "loss": 0.2168,
+      "step": 1069
+    },
+    {
+      "epoch": 1.0703010221624831,
+      "grad_norm": 0.20139080286026,
+      "learning_rate": 0.000143422100757158,
+      "loss": 0.2332,
+      "step": 1070
+    },
+    {
+      "epoch": 1.071301303491607,
+      "grad_norm": 0.18896177411079407,
+      "learning_rate": 0.00014332765026878687,
+      "loss": 0.1924,
+      "step": 1071
+    },
+    {
+      "epoch": 1.072301584820731,
+      "grad_norm": 0.3877180516719818,
+      "learning_rate": 0.0001432331521711639,
+      "loss": 0.2356,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0733018661498546,
+      "grad_norm": 0.16627468168735504,
+      "learning_rate": 0.00014313860656812536,
+      "loss": 0.1887,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0743021474789785,
+      "grad_norm": 0.20352627336978912,
+      "learning_rate": 0.00014304401356355983,
+      "loss": 0.225,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0753024288081023,
+      "grad_norm": 0.17726773023605347,
+      "learning_rate": 0.00014294937326140788,
+      "loss": 0.2165,
+      "step": 1075
+    },
+    {
+      "epoch": 1.076302710137226,
+      "grad_norm": 0.17296278476715088,
+      "learning_rate": 0.00014285468576566207,
+      "loss": 0.1806,
+      "step": 1076
+    },
+    {
+      "epoch": 1.07730299146635,
+      "grad_norm": 0.16486625373363495,
+      "learning_rate": 0.00014275995118036693,
+      "loss": 0.1814,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0783032727954738,
+      "grad_norm": 0.19006824493408203,
+      "learning_rate": 0.00014266516960961852,
+      "loss": 0.1883,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0793035541245974,
+      "grad_norm": 0.16644811630249023,
+      "learning_rate": 0.00014257034115756472,
+      "loss": 0.1858,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0803038354537213,
+      "grad_norm": 0.3777883052825928,
+      "learning_rate": 0.0001424754659284048,
+      "loss": 0.2252,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0813041167828452,
+      "grad_norm": 0.18761883676052094,
+      "learning_rate": 0.0001423805440263895,
+      "loss": 0.1998,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0823043981119689,
+      "grad_norm": 0.261598140001297,
+      "learning_rate": 0.0001422855755558208,
+      "loss": 0.2061,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0833046794410928,
+      "grad_norm": 0.18047110736370087,
+      "learning_rate": 0.00014219056062105193,
+      "loss": 0.1983,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0843049607702167,
+      "grad_norm": 0.21733161807060242,
+      "learning_rate": 0.0001420954993264871,
+      "loss": 0.1735,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0853052420993405,
+      "grad_norm": 0.16773627698421478,
+      "learning_rate": 0.00014200039177658145,
+      "loss": 0.1907,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0863055234284642,
+      "grad_norm": 0.408469021320343,
+      "learning_rate": 0.000141905238075841,
+      "loss": 0.2174,
+      "step": 1086
+    },
+    {
+      "epoch": 1.087305804757588,
+      "grad_norm": 0.1632164865732193,
+      "learning_rate": 0.00014181003832882248,
+      "loss": 0.1727,
+      "step": 1087
+    },
+    {
+      "epoch": 1.088306086086712,
+      "grad_norm": 0.1682378202676773,
+      "learning_rate": 0.00014171479264013311,
+      "loss": 0.193,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0893063674158356,
+      "grad_norm": 0.20175616443157196,
+      "learning_rate": 0.00014161950111443077,
+      "loss": 0.1777,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0903066487449595,
+      "grad_norm": 0.20562951266765594,
+      "learning_rate": 0.00014152416385642357,
+      "loss": 0.2237,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0913069300740834,
+      "grad_norm": 0.17610840499401093,
+      "learning_rate": 0.00014142878097086995,
+      "loss": 0.1969,
+      "step": 1091
+    },
+    {
+      "epoch": 1.092307211403207,
+      "grad_norm": 0.1920180320739746,
+      "learning_rate": 0.0001413333525625784,
+      "loss": 0.2046,
+      "step": 1092
+    },
+    {
+      "epoch": 1.093307492732331,
+      "grad_norm": 0.19373036921024323,
+      "learning_rate": 0.00014123787873640754,
+      "loss": 0.1731,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0943077740614549,
+      "grad_norm": 0.2086368203163147,
+      "learning_rate": 0.00014114235959726575,
+      "loss": 0.2134,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0953080553905785,
+      "grad_norm": 0.1742224097251892,
+      "learning_rate": 0.0001410467952501114,
+      "loss": 0.2025,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0963083367197024,
+      "grad_norm": 0.19609488546848297,
+      "learning_rate": 0.00014095118579995235,
+      "loss": 0.2154,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0973086180488263,
+      "grad_norm": 0.18279018998146057,
+      "learning_rate": 0.0001408555313518461,
+      "loss": 0.203,
+      "step": 1097
+    },
+    {
+      "epoch": 1.09830889937795,
+      "grad_norm": 0.1721639335155487,
+      "learning_rate": 0.00014075983201089964,
+      "loss": 0.1883,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0993091807070738,
+      "grad_norm": 0.19042350351810455,
+      "learning_rate": 0.0001406640878822692,
+      "loss": 0.191,
+      "step": 1099
+    },
+    {
+      "epoch": 1.1003094620361977,
+      "grad_norm": 0.1903952807188034,
+      "learning_rate": 0.00014056829907116024,
+      "loss": 0.2079,
+      "step": 1100
+    },
+    {
+      "epoch": 1.1013097433653214,
+      "grad_norm": 0.14835655689239502,
+      "learning_rate": 0.00014047246568282736,
+      "loss": 0.1918,
+      "step": 1101
+    },
+    {
+      "epoch": 1.1023100246944453,
+      "grad_norm": 0.19426876306533813,
+      "learning_rate": 0.00014037658782257414,
+      "loss": 0.2219,
+      "step": 1102
+    },
+    {
+      "epoch": 1.1033103060235692,
+      "grad_norm": 0.16929534077644348,
+      "learning_rate": 0.00014028066559575302,
+      "loss": 0.2205,
+      "step": 1103
+    },
+    {
+      "epoch": 1.104310587352693,
+      "grad_norm": 0.15848885476589203,
+      "learning_rate": 0.00014018469910776513,
+      "loss": 0.1772,
+      "step": 1104
+    },
+    {
+      "epoch": 1.1053108686818167,
+      "grad_norm": 0.1809755563735962,
+      "learning_rate": 0.0001400886884640603,
+      "loss": 0.1725,
+      "step": 1105
+    },
+    {
+      "epoch": 1.1063111500109406,
+      "grad_norm": 0.16270798444747925,
+      "learning_rate": 0.00013999263377013693,
+      "loss": 0.1821,
+      "step": 1106
+    },
+    {
+      "epoch": 1.1073114313400645,
+      "grad_norm": 0.19316376745700836,
+      "learning_rate": 0.00013989653513154165,
+      "loss": 0.216,
+      "step": 1107
+    },
+    {
+      "epoch": 1.1083117126691882,
+      "grad_norm": 0.18078681826591492,
+      "learning_rate": 0.00013980039265386955,
+      "loss": 0.1988,
+      "step": 1108
+    },
+    {
+      "epoch": 1.109311993998312,
+      "grad_norm": 0.15593406558036804,
+      "learning_rate": 0.00013970420644276383,
+      "loss": 0.1709,
+      "step": 1109
+    },
+    {
+      "epoch": 1.110312275327436,
+      "grad_norm": 0.18162721395492554,
+      "learning_rate": 0.0001396079766039157,
+      "loss": 0.1908,
+      "step": 1110
+    },
+    {
+      "epoch": 1.1113125566565596,
+      "grad_norm": 0.1557709127664566,
+      "learning_rate": 0.00013951170324306435,
+      "loss": 0.1724,
+      "step": 1111
+    },
+    {
+      "epoch": 1.1123128379856835,
+      "grad_norm": 0.17272527515888214,
+      "learning_rate": 0.00013941538646599687,
+      "loss": 0.1875,
+      "step": 1112
+    },
+    {
+      "epoch": 1.1133131193148074,
+      "grad_norm": 0.17165982723236084,
+      "learning_rate": 0.0001393190263785479,
+      "loss": 0.201,
+      "step": 1113
+    },
+    {
+      "epoch": 1.114313400643931,
+      "grad_norm": 0.17316922545433044,
+      "learning_rate": 0.0001392226230865998,
+      "loss": 0.2224,
+      "step": 1114
+    },
+    {
+      "epoch": 1.115313681973055,
+      "grad_norm": 0.15698738396167755,
+      "learning_rate": 0.0001391261766960823,
+      "loss": 0.1751,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1163139633021788,
+      "grad_norm": 0.16118797659873962,
+      "learning_rate": 0.00013902968731297255,
+      "loss": 0.1825,
+      "step": 1116
+    },
+    {
+      "epoch": 1.1173142446313025,
+      "grad_norm": 0.2679535746574402,
+      "learning_rate": 0.00013893315504329498,
+      "loss": 0.2209,
+      "step": 1117
+    },
+    {
+      "epoch": 1.1183145259604264,
+      "grad_norm": 0.16494572162628174,
+      "learning_rate": 0.00013883657999312109,
+      "loss": 0.1924,
+      "step": 1118
+    },
+    {
+      "epoch": 1.1193148072895502,
+      "grad_norm": 0.21941478550434113,
+      "learning_rate": 0.00013873996226856933,
+      "loss": 0.2352,
+      "step": 1119
+    },
+    {
+      "epoch": 1.120315088618674,
+      "grad_norm": 0.17025230824947357,
+      "learning_rate": 0.00013864330197580513,
+      "loss": 0.2101,
+      "step": 1120
+    },
+    {
+      "epoch": 1.1213153699477978,
+      "grad_norm": 0.15950845181941986,
+      "learning_rate": 0.0001385465992210407,
+      "loss": 0.1722,
+      "step": 1121
+    },
+    {
+      "epoch": 1.1223156512769217,
+      "grad_norm": 0.246660977602005,
+      "learning_rate": 0.00013844985411053492,
+      "loss": 0.2169,
+      "step": 1122
+    },
+    {
+      "epoch": 1.1233159326060456,
+      "grad_norm": 0.16967430710792542,
+      "learning_rate": 0.00013835306675059308,
+      "loss": 0.1952,
+      "step": 1123
+    },
+    {
+      "epoch": 1.1243162139351692,
+      "grad_norm": 0.16052035987377167,
+      "learning_rate": 0.00013825623724756704,
+      "loss": 0.1806,
+      "step": 1124
+    },
+    {
+      "epoch": 1.1253164952642931,
+      "grad_norm": 0.252022922039032,
+      "learning_rate": 0.00013815936570785487,
+      "loss": 0.1836,
+      "step": 1125
+    },
+    {
+      "epoch": 1.1263167765934168,
+      "grad_norm": 0.1942882388830185,
+      "learning_rate": 0.00013806245223790088,
+      "loss": 0.2009,
+      "step": 1126
+    },
+    {
+      "epoch": 1.1273170579225407,
+      "grad_norm": 0.21579767763614655,
+      "learning_rate": 0.0001379654969441955,
+      "loss": 0.2092,
+      "step": 1127
+    },
+    {
+      "epoch": 1.1283173392516646,
+      "grad_norm": 0.24866971373558044,
+      "learning_rate": 0.000137868499933275,
+      "loss": 0.1858,
+      "step": 1128
+    },
+    {
+      "epoch": 1.1293176205807884,
+      "grad_norm": 0.20886608958244324,
+      "learning_rate": 0.00013777146131172162,
+      "loss": 0.1979,
+      "step": 1129
+    },
+    {
+      "epoch": 1.130317901909912,
+      "grad_norm": 0.1537419557571411,
+      "learning_rate": 0.00013767438118616318,
+      "loss": 0.1895,
+      "step": 1130
+    },
+    {
+      "epoch": 1.131318183239036,
+      "grad_norm": 0.20272980630397797,
+      "learning_rate": 0.00013757725966327322,
+      "loss": 0.1854,
+      "step": 1131
+    },
+    {
+      "epoch": 1.1323184645681599,
+      "grad_norm": 0.2039756327867508,
+      "learning_rate": 0.00013748009684977073,
+      "loss": 0.253,
+      "step": 1132
+    },
+    {
+      "epoch": 1.1333187458972835,
+      "grad_norm": 0.15201859176158905,
+      "learning_rate": 0.0001373828928524201,
+      "loss": 0.1764,
+      "step": 1133
+    },
+    {
+      "epoch": 1.1343190272264074,
+      "grad_norm": 0.21233822405338287,
+      "learning_rate": 0.00013728564777803088,
+      "loss": 0.1971,
+      "step": 1134
+    },
+    {
+      "epoch": 1.1353193085555313,
+      "grad_norm": 0.18667684495449066,
+      "learning_rate": 0.00013718836173345783,
+      "loss": 0.1887,
+      "step": 1135
+    },
+    {
+      "epoch": 1.136319589884655,
+      "grad_norm": 0.18006233870983124,
+      "learning_rate": 0.00013709103482560078,
+      "loss": 0.1948,
+      "step": 1136
+    },
+    {
+      "epoch": 1.1373198712137789,
+      "grad_norm": 0.14822353422641754,
+      "learning_rate": 0.00013699366716140435,
+      "loss": 0.1707,
+      "step": 1137
+    },
+    {
+      "epoch": 1.1383201525429028,
+      "grad_norm": 0.23304615914821625,
+      "learning_rate": 0.00013689625884785798,
+      "loss": 0.1872,
+      "step": 1138
+    },
+    {
+      "epoch": 1.1393204338720264,
+      "grad_norm": 0.2797177731990814,
+      "learning_rate": 0.00013679880999199583,
+      "loss": 0.2073,
+      "step": 1139
+    },
+    {
+      "epoch": 1.1403207152011503,
+      "grad_norm": 0.18375234305858612,
+      "learning_rate": 0.00013670132070089653,
+      "loss": 0.2234,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1413209965302742,
+      "grad_norm": 0.1607845425605774,
+      "learning_rate": 0.00013660379108168324,
+      "loss": 0.1754,
+      "step": 1141
+    },
+    {
+      "epoch": 1.142321277859398,
+      "grad_norm": 0.16823294758796692,
+      "learning_rate": 0.00013650622124152334,
+      "loss": 0.1973,
+      "step": 1142
+    },
+    {
+      "epoch": 1.1433215591885217,
+      "grad_norm": 0.17448188364505768,
+      "learning_rate": 0.0001364086112876284,
+      "loss": 0.1983,
+      "step": 1143
+    },
+    {
+      "epoch": 1.1443218405176456,
+      "grad_norm": 0.4084351062774658,
+      "learning_rate": 0.00013631096132725413,
+      "loss": 0.2082,
+      "step": 1144
+    },
+    {
+      "epoch": 1.1453221218467693,
+      "grad_norm": 0.21218331158161163,
+      "learning_rate": 0.00013621327146770025,
+      "loss": 0.2118,
+      "step": 1145
+    },
+    {
+      "epoch": 1.1463224031758932,
+      "grad_norm": 0.2552039623260498,
+      "learning_rate": 0.00013611554181631013,
+      "loss": 0.2101,
+      "step": 1146
+    },
+    {
+      "epoch": 1.147322684505017,
+      "grad_norm": 0.21880224347114563,
+      "learning_rate": 0.00013601777248047105,
+      "loss": 0.216,
+      "step": 1147
+    },
+    {
+      "epoch": 1.148322965834141,
+      "grad_norm": 0.18925979733467102,
+      "learning_rate": 0.0001359199635676138,
+      "loss": 0.1956,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1493232471632646,
+      "grad_norm": 0.2236345410346985,
+      "learning_rate": 0.00013582211518521273,
+      "loss": 0.1826,
+      "step": 1149
+    },
+    {
+      "epoch": 1.1503235284923885,
+      "grad_norm": 0.19078874588012695,
+      "learning_rate": 0.00013572422744078551,
+      "loss": 0.1874,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1513238098215124,
+      "grad_norm": 0.17836186289787292,
+      "learning_rate": 0.00013562630044189304,
+      "loss": 0.1978,
+      "step": 1151
+    },
+    {
+      "epoch": 1.152324091150636,
+      "grad_norm": 0.17768734693527222,
+      "learning_rate": 0.00013552833429613938,
+      "loss": 0.2108,
+      "step": 1152
+    },
+    {
+      "epoch": 1.15332437247976,
+      "grad_norm": 0.17896975576877594,
+      "learning_rate": 0.0001354303291111716,
+      "loss": 0.1899,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1543246538088838,
+      "grad_norm": 0.20729735493659973,
+      "learning_rate": 0.0001353322849946797,
+      "loss": 0.2046,
+      "step": 1154
+    },
+    {
+      "epoch": 1.1553249351380075,
+      "grad_norm": 0.2111363559961319,
+      "learning_rate": 0.00013523420205439646,
+      "loss": 0.2219,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1563252164671314,
+      "grad_norm": 0.19003383815288544,
+      "learning_rate": 0.0001351360803980972,
+      "loss": 0.2137,
+      "step": 1156
+    },
+    {
+      "epoch": 1.1573254977962553,
+      "grad_norm": 0.17860478162765503,
+      "learning_rate": 0.00013503792013359997,
+      "loss": 0.2101,
+      "step": 1157
+    },
+    {
+      "epoch": 1.158325779125379,
+      "grad_norm": 0.15965665876865387,
+      "learning_rate": 0.00013493972136876509,
+      "loss": 0.1827,
+      "step": 1158
+    },
+    {
+      "epoch": 1.1593260604545028,
+      "grad_norm": 0.17400485277175903,
+      "learning_rate": 0.00013484148421149527,
+      "loss": 0.1936,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1603263417836267,
+      "grad_norm": 0.19293096661567688,
+      "learning_rate": 0.0001347432087697354,
+      "loss": 0.1984,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1613266231127504,
+      "grad_norm": 0.30938926339149475,
+      "learning_rate": 0.00013464489515147238,
+      "loss": 0.1779,
+      "step": 1161
+    },
+    {
+      "epoch": 1.1623269044418743,
+      "grad_norm": 0.20373430848121643,
+      "learning_rate": 0.0001345465434647351,
+      "loss": 0.2354,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1633271857709981,
+      "grad_norm": 0.16392414271831512,
+      "learning_rate": 0.00013444815381759425,
+      "loss": 0.1662,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1643274671001218,
+      "grad_norm": 0.17659424245357513,
+      "learning_rate": 0.00013434972631816235,
+      "loss": 0.1684,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1653277484292457,
+      "grad_norm": 0.19293954968452454,
+      "learning_rate": 0.0001342512610745933,
+      "loss": 0.2194,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1663280297583696,
+      "grad_norm": 0.1835450679063797,
+      "learning_rate": 0.0001341527581950827,
+      "loss": 0.1843,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1673283110874935,
+      "grad_norm": 0.1926405429840088,
+      "learning_rate": 0.00013405421778786737,
+      "loss": 0.2037,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1683285924166171,
+      "grad_norm": 0.15698739886283875,
+      "learning_rate": 0.00013395563996122537,
+      "loss": 0.1707,
+      "step": 1168
+    },
+    {
+      "epoch": 1.169328873745741,
+      "grad_norm": 0.1622772216796875,
+      "learning_rate": 0.00013385702482347593,
+      "loss": 0.201,
+      "step": 1169
+    },
+    {
+      "epoch": 1.170329155074865,
+      "grad_norm": 0.15372814238071442,
+      "learning_rate": 0.00013375837248297926,
+      "loss": 0.1609,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1713294364039886,
+      "grad_norm": 0.1947375386953354,
+      "learning_rate": 0.0001336596830481364,
+      "loss": 0.1973,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1723297177331125,
+      "grad_norm": 0.5508653521537781,
+      "learning_rate": 0.0001335609566273892,
+      "loss": 0.1981,
+      "step": 1172
+    },
+    {
+      "epoch": 1.1733299990622363,
+      "grad_norm": 0.19331258535385132,
+      "learning_rate": 0.00013346219332922016,
+      "loss": 0.2128,
+      "step": 1173
+    },
+    {
+      "epoch": 1.17433028039136,
+      "grad_norm": 0.1639270782470703,
+      "learning_rate": 0.00013336339326215228,
+      "loss": 0.1781,
+      "step": 1174
+    },
+    {
+      "epoch": 1.175330561720484,
+      "grad_norm": 0.39454808831214905,
+      "learning_rate": 0.00013326455653474897,
+      "loss": 0.1918,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1763308430496078,
+      "grad_norm": 0.1775580495595932,
+      "learning_rate": 0.00013316568325561393,
+      "loss": 0.197,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1773311243787314,
+      "grad_norm": 0.18332943320274353,
+      "learning_rate": 0.00013306677353339098,
+      "loss": 0.2078,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1783314057078553,
+      "grad_norm": 0.16777454316616058,
+      "learning_rate": 0.000132967827476764,
+      "loss": 0.197,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1793316870369792,
+      "grad_norm": 0.1485321968793869,
+      "learning_rate": 0.0001328688451944569,
+      "loss": 0.1647,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1803319683661029,
+      "grad_norm": 0.19244162738323212,
+      "learning_rate": 0.00013276982679523322,
+      "loss": 0.1799,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1813322496952268,
+      "grad_norm": 0.1723315566778183,
+      "learning_rate": 0.00013267077238789633,
+      "loss": 0.1932,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1823325310243507,
+      "grad_norm": 0.17928031086921692,
+      "learning_rate": 0.00013257168208128908,
+      "loss": 0.2081,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1833328123534743,
+      "grad_norm": 0.18451717495918274,
+      "learning_rate": 0.00013247255598429378,
+      "loss": 0.2077,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1843330936825982,
+      "grad_norm": 0.18119074404239655,
+      "learning_rate": 0.00013237339420583212,
+      "loss": 0.189,
+      "step": 1184
+    },
+    {
+      "epoch": 1.185333375011722,
+      "grad_norm": 0.15664182603359222,
+      "learning_rate": 0.00013227419685486492,
+      "loss": 0.163,
+      "step": 1185
+    },
+    {
+      "epoch": 1.186333656340846,
+      "grad_norm": 0.24777516722679138,
+      "learning_rate": 0.00013217496404039218,
+      "loss": 0.2108,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1873339376699696,
+      "grad_norm": 0.16925553977489471,
+      "learning_rate": 0.0001320756958714528,
+      "loss": 0.2077,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1883342189990935,
+      "grad_norm": 0.17016278207302094,
+      "learning_rate": 0.00013197639245712454,
+      "loss": 0.1771,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1893345003282172,
+      "grad_norm": 0.2483927607536316,
+      "learning_rate": 0.00013187705390652388,
+      "loss": 0.192,
+      "step": 1189
+    },
+    {
+      "epoch": 1.190334781657341,
+      "grad_norm": 0.25870785117149353,
+      "learning_rate": 0.00013177768032880593,
+      "loss": 0.1868,
+      "step": 1190
+    },
+    {
+      "epoch": 1.191335062986465,
+      "grad_norm": 0.3790164291858673,
+      "learning_rate": 0.0001316782718331643,
+      "loss": 0.1947,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1923353443155889,
+      "grad_norm": 0.20143865048885345,
+      "learning_rate": 0.0001315788285288309,
+      "loss": 0.1945,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1933356256447125,
+      "grad_norm": 0.18720094859600067,
+      "learning_rate": 0.00013147935052507597,
+      "loss": 0.2133,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1943359069738364,
+      "grad_norm": 0.17185567319393158,
+      "learning_rate": 0.00013137983793120786,
+      "loss": 0.1882,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1953361883029603,
+      "grad_norm": 0.17893126606941223,
+      "learning_rate": 0.0001312802908565729,
+      "loss": 0.1861,
+      "step": 1195
+    },
+    {
+      "epoch": 1.196336469632084,
+      "grad_norm": 0.17477837204933167,
+      "learning_rate": 0.0001311807094105553,
+      "loss": 0.1907,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1973367509612078,
+      "grad_norm": 0.1644591987133026,
+      "learning_rate": 0.00013108109370257712,
+      "loss": 0.1689,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1983370322903317,
+      "grad_norm": 0.3162935674190521,
+      "learning_rate": 0.00013098144384209796,
+      "loss": 0.1924,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1993373136194554,
+      "grad_norm": 0.21296380460262299,
+      "learning_rate": 0.000130881759938615,
+      "loss": 0.2019,
+      "step": 1199
+    },
+    {
+      "epoch": 1.2003375949485793,
+      "grad_norm": 0.15518273413181305,
+      "learning_rate": 0.00013078204210166278,
+      "loss": 0.1877,
+      "step": 1200
+    },
+    {
+      "epoch": 1.2013378762777032,
+      "grad_norm": 0.18660126626491547,
+      "learning_rate": 0.00013068229044081324,
+      "loss": 0.1943,
+      "step": 1201
+    },
+    {
+      "epoch": 1.2023381576068268,
+      "grad_norm": 0.1803644597530365,
+      "learning_rate": 0.0001305825050656754,
+      "loss": 0.2068,
+      "step": 1202
+    },
+    {
+      "epoch": 1.2033384389359507,
+      "grad_norm": 0.1444634348154068,
+      "learning_rate": 0.00013048268608589533,
+      "loss": 0.1735,
+      "step": 1203
+    },
+    {
+      "epoch": 1.2043387202650746,
+      "grad_norm": 0.17907103896141052,
+      "learning_rate": 0.00013038283361115603,
+      "loss": 0.1829,
+      "step": 1204
+    },
+    {
+      "epoch": 1.2053390015941985,
+      "grad_norm": 0.19631987810134888,
+      "learning_rate": 0.0001302829477511773,
+      "loss": 0.2035,
+      "step": 1205
+    },
+    {
+      "epoch": 1.2063392829233222,
+      "grad_norm": 0.16690067946910858,
+      "learning_rate": 0.0001301830286157157,
+      "loss": 0.1777,
+      "step": 1206
+    },
+    {
+      "epoch": 1.207339564252446,
+      "grad_norm": 0.16956549882888794,
+      "learning_rate": 0.0001300830763145642,
+      "loss": 0.1662,
+      "step": 1207
+    },
+    {
+      "epoch": 1.2083398455815697,
+      "grad_norm": 0.19812379777431488,
+      "learning_rate": 0.00012998309095755235,
+      "loss": 0.2247,
+      "step": 1208
+    },
+    {
+      "epoch": 1.2093401269106936,
+      "grad_norm": 0.1658792346715927,
+      "learning_rate": 0.00012988307265454597,
+      "loss": 0.1934,
+      "step": 1209
+    },
+    {
+      "epoch": 1.2103404082398175,
+      "grad_norm": 0.18409915268421173,
+      "learning_rate": 0.0001297830215154471,
+      "loss": 0.1926,
+      "step": 1210
+    },
+    {
+      "epoch": 1.2113406895689414,
+      "grad_norm": 0.16715079545974731,
+      "learning_rate": 0.00012968293765019384,
+      "loss": 0.1759,
+      "step": 1211
+    },
+    {
+      "epoch": 1.212340970898065,
+      "grad_norm": 0.19099463522434235,
+      "learning_rate": 0.00012958282116876026,
+      "loss": 0.2091,
+      "step": 1212
+    },
+    {
+      "epoch": 1.213341252227189,
+      "grad_norm": 0.17515191435813904,
+      "learning_rate": 0.00012948267218115624,
+      "loss": 0.1924,
+      "step": 1213
+    },
+    {
+      "epoch": 1.2143415335563128,
+      "grad_norm": 0.2107793390750885,
+      "learning_rate": 0.00012938249079742743,
+      "loss": 0.1981,
+      "step": 1214
+    },
+    {
+      "epoch": 1.2153418148854365,
+      "grad_norm": 0.19806204736232758,
+      "learning_rate": 0.00012928227712765504,
+      "loss": 0.1966,
+      "step": 1215
+    },
+    {
+      "epoch": 1.2163420962145604,
+      "grad_norm": 0.17539450526237488,
+      "learning_rate": 0.0001291820312819558,
+      "loss": 0.1749,
+      "step": 1216
+    },
+    {
+      "epoch": 1.2173423775436842,
+      "grad_norm": 0.19170427322387695,
+      "learning_rate": 0.00012908175337048174,
+      "loss": 0.2226,
+      "step": 1217
+    },
+    {
+      "epoch": 1.218342658872808,
+      "grad_norm": 0.16263103485107422,
+      "learning_rate": 0.00012898144350342015,
+      "loss": 0.1839,
+      "step": 1218
+    },
+    {
+      "epoch": 1.2193429402019318,
+      "grad_norm": 0.18162356317043304,
+      "learning_rate": 0.0001288811017909934,
+      "loss": 0.1904,
+      "step": 1219
+    },
+    {
+      "epoch": 1.2203432215310557,
+      "grad_norm": 0.1786147803068161,
+      "learning_rate": 0.00012878072834345895,
+      "loss": 0.1878,
+      "step": 1220
+    },
+    {
+      "epoch": 1.2213435028601793,
+      "grad_norm": 0.20931778848171234,
+      "learning_rate": 0.00012868032327110904,
+      "loss": 0.2032,
+      "step": 1221
+    },
+    {
+      "epoch": 1.2223437841893032,
+      "grad_norm": 0.19125699996948242,
+      "learning_rate": 0.00012857988668427066,
+      "loss": 0.2154,
+      "step": 1222
+    },
+    {
+      "epoch": 1.2233440655184271,
+      "grad_norm": 0.18324574828147888,
+      "learning_rate": 0.0001284794186933055,
+      "loss": 0.1954,
+      "step": 1223
+    },
+    {
+      "epoch": 1.2243443468475508,
+      "grad_norm": 0.17114463448524475,
+      "learning_rate": 0.00012837891940860972,
+      "loss": 0.1738,
+      "step": 1224
+    },
+    {
+      "epoch": 1.2253446281766747,
+      "grad_norm": 0.17293782532215118,
+      "learning_rate": 0.00012827838894061377,
+      "loss": 0.1814,
+      "step": 1225
+    },
+    {
+      "epoch": 1.2263449095057986,
+      "grad_norm": 0.1815311312675476,
+      "learning_rate": 0.00012817782739978255,
+      "loss": 0.1946,
+      "step": 1226
+    },
+    {
+      "epoch": 1.2273451908349222,
+      "grad_norm": 0.222374826669693,
+      "learning_rate": 0.00012807723489661495,
+      "loss": 0.2012,
+      "step": 1227
+    },
+    {
+      "epoch": 1.228345472164046,
+      "grad_norm": 0.18566803634166718,
+      "learning_rate": 0.00012797661154164395,
+      "loss": 0.2177,
+      "step": 1228
+    },
+    {
+      "epoch": 1.22934575349317,
+      "grad_norm": 0.1825873851776123,
+      "learning_rate": 0.00012787595744543647,
+      "loss": 0.2153,
+      "step": 1229
+    },
+    {
+      "epoch": 1.2303460348222939,
+      "grad_norm": 0.192878857254982,
+      "learning_rate": 0.00012777527271859307,
+      "loss": 0.2019,
+      "step": 1230
+    },
+    {
+      "epoch": 1.2313463161514175,
+      "grad_norm": 0.16210146248340607,
+      "learning_rate": 0.0001276745574717481,
+      "loss": 0.1925,
+      "step": 1231
+    },
+    {
+      "epoch": 1.2323465974805414,
+      "grad_norm": 0.19294248521327972,
+      "learning_rate": 0.00012757381181556943,
+      "loss": 0.1992,
+      "step": 1232
+    },
+    {
+      "epoch": 1.2333468788096653,
+      "grad_norm": 0.1765872687101364,
+      "learning_rate": 0.0001274730358607583,
+      "loss": 0.2151,
+      "step": 1233
+    },
+    {
+      "epoch": 1.234347160138789,
+      "grad_norm": 0.21067166328430176,
+      "learning_rate": 0.00012737222971804924,
+      "loss": 0.1957,
+      "step": 1234
+    },
+    {
+      "epoch": 1.2353474414679129,
+      "grad_norm": 0.20086489617824554,
+      "learning_rate": 0.00012727139349821,
+      "loss": 0.2178,
+      "step": 1235
+    },
+    {
+      "epoch": 1.2363477227970368,
+      "grad_norm": 0.16447032988071442,
+      "learning_rate": 0.0001271705273120413,
+      "loss": 0.1859,
+      "step": 1236
+    },
+    {
+      "epoch": 1.2373480041261604,
+      "grad_norm": 0.17268379032611847,
+      "learning_rate": 0.00012706963127037685,
+      "loss": 0.1937,
+      "step": 1237
+    },
+    {
+      "epoch": 1.2383482854552843,
+      "grad_norm": 0.17328353226184845,
+      "learning_rate": 0.00012696870548408316,
+      "loss": 0.2004,
+      "step": 1238
+    },
+    {
+      "epoch": 1.2393485667844082,
+      "grad_norm": 0.21452516317367554,
+      "learning_rate": 0.00012686775006405946,
+      "loss": 0.2122,
+      "step": 1239
+    },
+    {
+      "epoch": 1.2403488481135319,
+      "grad_norm": 0.1722584366798401,
+      "learning_rate": 0.00012676676512123747,
+      "loss": 0.1725,
+      "step": 1240
+    },
+    {
+      "epoch": 1.2413491294426557,
+      "grad_norm": 0.2064584642648697,
+      "learning_rate": 0.00012666575076658134,
+      "loss": 0.1861,
+      "step": 1241
+    },
+    {
+      "epoch": 1.2423494107717796,
+      "grad_norm": 0.18923532962799072,
+      "learning_rate": 0.00012656470711108764,
+      "loss": 0.2122,
+      "step": 1242
+    },
+    {
+      "epoch": 1.2433496921009033,
+      "grad_norm": 0.18311384320259094,
+      "learning_rate": 0.00012646363426578505,
+      "loss": 0.1996,
+      "step": 1243
+    },
+    {
+      "epoch": 1.2443499734300272,
+      "grad_norm": 0.17757707834243774,
+      "learning_rate": 0.0001263625323417343,
+      "loss": 0.1827,
+      "step": 1244
+    },
+    {
+      "epoch": 1.245350254759151,
+      "grad_norm": 0.18281926214694977,
+      "learning_rate": 0.0001262614014500282,
+      "loss": 0.1853,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2463505360882747,
+      "grad_norm": 0.18698139488697052,
+      "learning_rate": 0.00012616024170179126,
+      "loss": 0.2033,
+      "step": 1246
+    },
+    {
+      "epoch": 1.2473508174173986,
+      "grad_norm": 0.17655418813228607,
+      "learning_rate": 0.00012605905320817976,
+      "loss": 0.2049,
+      "step": 1247
+    },
+    {
+      "epoch": 1.2483510987465225,
+      "grad_norm": 0.16421499848365784,
+      "learning_rate": 0.00012595783608038155,
+      "loss": 0.1733,
+      "step": 1248
+    },
+    {
+      "epoch": 1.2493513800756464,
+      "grad_norm": 0.19940869510173798,
+      "learning_rate": 0.00012585659042961596,
+      "loss": 0.1864,
+      "step": 1249
+    },
+    {
+      "epoch": 1.25035166140477,
+      "grad_norm": 0.17662669718265533,
+      "learning_rate": 0.00012575531636713368,
+      "loss": 0.1952,
+      "step": 1250
+    },
+    {
+      "epoch": 1.251351942733894,
+      "grad_norm": 0.17970876395702362,
+      "learning_rate": 0.00012565401400421651,
+      "loss": 0.2012,
+      "step": 1251
+    },
+    {
+      "epoch": 1.2523522240630176,
+      "grad_norm": 0.1915876716375351,
+      "learning_rate": 0.0001255526834521775,
+      "loss": 0.2038,
+      "step": 1252
+    },
+    {
+      "epoch": 1.2533525053921415,
+      "grad_norm": 0.1632014960050583,
+      "learning_rate": 0.00012545132482236055,
+      "loss": 0.1639,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2543527867212654,
+      "grad_norm": 0.17216497659683228,
+      "learning_rate": 0.0001253499382261405,
+      "loss": 0.1809,
+      "step": 1254
+    },
+    {
+      "epoch": 1.2553530680503893,
+      "grad_norm": 0.1890319287776947,
+      "learning_rate": 0.00012524852377492285,
+      "loss": 0.2038,
+      "step": 1255
+    },
+    {
+      "epoch": 1.256353349379513,
+      "grad_norm": 0.2119457870721817,
+      "learning_rate": 0.00012514708158014378,
+      "loss": 0.2259,
+      "step": 1256
+    },
+    {
+      "epoch": 1.2573536307086368,
+      "grad_norm": 0.16909374296665192,
+      "learning_rate": 0.00012504561175326985,
+      "loss": 0.1767,
+      "step": 1257
+    },
+    {
+      "epoch": 1.2583539120377607,
+      "grad_norm": 0.31165897846221924,
+      "learning_rate": 0.00012494411440579814,
+      "loss": 0.2153,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2593541933668844,
+      "grad_norm": 0.16875721514225006,
+      "learning_rate": 0.0001248425896492558,
+      "loss": 0.2201,
+      "step": 1259
+    },
+    {
+      "epoch": 1.2603544746960083,
+      "grad_norm": 0.19854949414730072,
+      "learning_rate": 0.00012474103759520027,
+      "loss": 0.2137,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2613547560251321,
+      "grad_norm": 0.19432970881462097,
+      "learning_rate": 0.00012463945835521878,
+      "loss": 0.1751,
+      "step": 1261
+    },
+    {
+      "epoch": 1.2623550373542558,
+      "grad_norm": 0.16364595293998718,
+      "learning_rate": 0.0001245378520409286,
+      "loss": 0.1903,
+      "step": 1262
+    },
+    {
+      "epoch": 1.2633553186833797,
+      "grad_norm": 0.1770123541355133,
+      "learning_rate": 0.0001244362187639767,
+      "loss": 0.1948,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2643556000125036,
+      "grad_norm": 0.27067604660987854,
+      "learning_rate": 0.00012433455863603967,
+      "loss": 0.2025,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2653558813416272,
+      "grad_norm": 0.2278088480234146,
+      "learning_rate": 0.00012423287176882358,
+      "loss": 0.2173,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2663561626707511,
+      "grad_norm": 0.18657363951206207,
+      "learning_rate": 0.00012413115827406392,
+      "loss": 0.211,
+      "step": 1266
+    },
+    {
+      "epoch": 1.267356443999875,
+      "grad_norm": 0.15636909008026123,
+      "learning_rate": 0.00012402941826352546,
+      "loss": 0.1814,
+      "step": 1267
+    },
+    {
+      "epoch": 1.268356725328999,
+      "grad_norm": 0.161123588681221,
+      "learning_rate": 0.00012392765184900202,
+      "loss": 0.183,
+      "step": 1268
+    },
+    {
+      "epoch": 1.2693570066581226,
+      "grad_norm": 0.15669114887714386,
+      "learning_rate": 0.0001238258591423165,
+      "loss": 0.1881,
+      "step": 1269
+    },
+    {
+      "epoch": 1.2703572879872465,
+      "grad_norm": 0.17591902613639832,
+      "learning_rate": 0.00012372404025532072,
+      "loss": 0.2106,
+      "step": 1270
+    },
+    {
+      "epoch": 1.2713575693163701,
+      "grad_norm": 0.18734882771968842,
+      "learning_rate": 0.00012362219529989514,
+      "loss": 0.2026,
+      "step": 1271
+    },
+    {
+      "epoch": 1.272357850645494,
+      "grad_norm": 0.16859911382198334,
+      "learning_rate": 0.00012352032438794902,
+      "loss": 0.1911,
+      "step": 1272
+    },
+    {
+      "epoch": 1.273358131974618,
+      "grad_norm": 0.16812944412231445,
+      "learning_rate": 0.00012341842763142005,
+      "loss": 0.1843,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2743584133037418,
+      "grad_norm": 0.19603118300437927,
+      "learning_rate": 0.00012331650514227425,
+      "loss": 0.2181,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2753586946328654,
+      "grad_norm": 0.1578560322523117,
+      "learning_rate": 0.00012321455703250616,
+      "loss": 0.2025,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2763589759619893,
+      "grad_norm": 0.16386160254478455,
+      "learning_rate": 0.00012311258341413822,
+      "loss": 0.1749,
+      "step": 1276
+    },
+    {
+      "epoch": 1.277359257291113,
+      "grad_norm": 0.22084735333919525,
+      "learning_rate": 0.00012301058439922102,
+      "loss": 0.1934,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2783595386202369,
+      "grad_norm": 0.17984852194786072,
+      "learning_rate": 0.000122908560099833,
+      "loss": 0.2098,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2793598199493608,
+      "grad_norm": 0.17951016128063202,
+      "learning_rate": 0.00012280651062808047,
+      "loss": 0.2087,
+      "step": 1279
+    },
+    {
+      "epoch": 1.2803601012784847,
+      "grad_norm": 0.18283426761627197,
+      "learning_rate": 0.00012270443609609729,
+      "loss": 0.1862,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2813603826076083,
+      "grad_norm": 0.21533912420272827,
+      "learning_rate": 0.0001226023366160449,
+      "loss": 0.2034,
+      "step": 1281
+    },
+    {
+      "epoch": 1.2823606639367322,
+      "grad_norm": 0.2440841943025589,
+      "learning_rate": 0.00012250021230011225,
+      "loss": 0.2069,
+      "step": 1282
+    },
+    {
+      "epoch": 1.283360945265856,
+      "grad_norm": 0.2983700633049011,
+      "learning_rate": 0.00012239806326051539,
+      "loss": 0.1898,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2843612265949798,
+      "grad_norm": 0.17114244401454926,
+      "learning_rate": 0.00012229588960949771,
+      "loss": 0.2158,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2853615079241036,
+      "grad_norm": 0.1767871081829071,
+      "learning_rate": 0.00012219369145932959,
+      "loss": 0.2081,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2863617892532275,
+      "grad_norm": 0.16608798503875732,
+      "learning_rate": 0.00012209146892230822,
+      "loss": 0.1795,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2873620705823514,
+      "grad_norm": 0.18475057184696198,
+      "learning_rate": 0.00012198922211075778,
+      "loss": 0.1943,
+      "step": 1287
+    },
+    {
+      "epoch": 1.288362351911475,
+      "grad_norm": 0.19820700585842133,
+      "learning_rate": 0.00012188695113702896,
+      "loss": 0.2192,
+      "step": 1288
+    },
+    {
+      "epoch": 1.289362633240599,
+      "grad_norm": 0.2036782205104828,
+      "learning_rate": 0.00012178465611349911,
+      "loss": 0.196,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2903629145697226,
+      "grad_norm": 0.18136808276176453,
+      "learning_rate": 0.00012168233715257194,
+      "loss": 0.2072,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2913631958988465,
+      "grad_norm": 0.2058573067188263,
+      "learning_rate": 0.00012157999436667747,
+      "loss": 0.2065,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2923634772279704,
+      "grad_norm": 0.17427518963813782,
+      "learning_rate": 0.00012147762786827193,
+      "loss": 0.1814,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2933637585570943,
+      "grad_norm": 0.254863440990448,
+      "learning_rate": 0.00012137523776983757,
+      "loss": 0.2208,
+      "step": 1293
+    },
+    {
+      "epoch": 1.294364039886218,
+      "grad_norm": 0.20321603119373322,
+      "learning_rate": 0.00012127282418388264,
+      "loss": 0.2082,
+      "step": 1294
+    },
+    {
+      "epoch": 1.2953643212153418,
+      "grad_norm": 0.21121065318584442,
+      "learning_rate": 0.0001211703872229411,
+      "loss": 0.2205,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2963646025444655,
+      "grad_norm": 0.21764880418777466,
+      "learning_rate": 0.00012106792699957263,
+      "loss": 0.2104,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2973648838735894,
+      "grad_norm": 0.17313605546951294,
+      "learning_rate": 0.00012096544362636255,
+      "loss": 0.1849,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2983651652027133,
+      "grad_norm": 0.19549302756786346,
+      "learning_rate": 0.00012086293721592152,
+      "loss": 0.1986,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2993654465318372,
+      "grad_norm": 0.1744566261768341,
+      "learning_rate": 0.00012076040788088554,
+      "loss": 0.1721,
+      "step": 1299
+    },
+    {
+      "epoch": 1.3003657278609608,
+      "grad_norm": 0.18029043078422546,
+      "learning_rate": 0.00012065785573391581,
+      "loss": 0.1982,
+      "step": 1300
+    },
+    {
+      "epoch": 1.3013660091900847,
+      "grad_norm": 0.19969074428081512,
+      "learning_rate": 0.00012055528088769861,
+      "loss": 0.2021,
+      "step": 1301
+    },
+    {
+      "epoch": 1.3023662905192086,
+      "grad_norm": 0.15677644312381744,
+      "learning_rate": 0.00012045268345494511,
+      "loss": 0.1907,
+      "step": 1302
+    },
+    {
+      "epoch": 1.3033665718483323,
+      "grad_norm": 0.16569839417934418,
+      "learning_rate": 0.00012035006354839133,
+      "loss": 0.1747,
+      "step": 1303
+    },
+    {
+      "epoch": 1.3043668531774562,
+      "grad_norm": 0.17134904861450195,
+      "learning_rate": 0.00012024742128079805,
+      "loss": 0.1978,
+      "step": 1304
+    },
+    {
+      "epoch": 1.30536713450658,
+      "grad_norm": 0.14358048141002655,
+      "learning_rate": 0.00012014475676495052,
+      "loss": 0.1752,
+      "step": 1305
+    },
+    {
+      "epoch": 1.306367415835704,
+      "grad_norm": 0.19968698918819427,
+      "learning_rate": 0.00012004207011365849,
+      "loss": 0.2056,
+      "step": 1306
+    },
+    {
+      "epoch": 1.3073676971648276,
+      "grad_norm": 0.28996333479881287,
+      "learning_rate": 0.00011993936143975599,
+      "loss": 0.1874,
+      "step": 1307
+    },
+    {
+      "epoch": 1.3083679784939515,
+      "grad_norm": 0.22329270839691162,
+      "learning_rate": 0.00011983663085610131,
+      "loss": 0.1799,
+      "step": 1308
+    },
+    {
+      "epoch": 1.3093682598230751,
+      "grad_norm": 0.1631443351507187,
+      "learning_rate": 0.00011973387847557676,
+      "loss": 0.1799,
+      "step": 1309
+    },
+    {
+      "epoch": 1.310368541152199,
+      "grad_norm": 0.1755417287349701,
+      "learning_rate": 0.00011963110441108863,
+      "loss": 0.2083,
+      "step": 1310
+    },
+    {
+      "epoch": 1.311368822481323,
+      "grad_norm": 0.1902787834405899,
+      "learning_rate": 0.000119528308775567,
+      "loss": 0.2004,
+      "step": 1311
+    },
+    {
+      "epoch": 1.3123691038104468,
+      "grad_norm": 0.1753208041191101,
+      "learning_rate": 0.00011942549168196575,
+      "loss": 0.2036,
+      "step": 1312
+    },
+    {
+      "epoch": 1.3133693851395705,
+      "grad_norm": 0.15781405568122864,
+      "learning_rate": 0.00011932265324326221,
+      "loss": 0.1904,
+      "step": 1313
+    },
+    {
+      "epoch": 1.3143696664686944,
+      "grad_norm": 0.20198071002960205,
+      "learning_rate": 0.0001192197935724573,
+      "loss": 0.201,
+      "step": 1314
+    },
+    {
+      "epoch": 1.315369947797818,
+      "grad_norm": 0.2038806825876236,
+      "learning_rate": 0.00011911691278257511,
+      "loss": 0.2215,
+      "step": 1315
+    },
+    {
+      "epoch": 1.316370229126942,
+      "grad_norm": 0.1787245273590088,
+      "learning_rate": 0.0001190140109866631,
+      "loss": 0.222,
+      "step": 1316
+    },
+    {
+      "epoch": 1.3173705104560658,
+      "grad_norm": 0.322486013174057,
+      "learning_rate": 0.00011891108829779165,
+      "loss": 0.1436,
+      "step": 1317
+    },
+    {
+      "epoch": 1.3183707917851897,
+      "grad_norm": 0.22720371186733246,
+      "learning_rate": 0.00011880814482905422,
+      "loss": 0.1775,
+      "step": 1318
+    },
+    {
+      "epoch": 1.3193710731143133,
+      "grad_norm": 0.19523940980434418,
+      "learning_rate": 0.00011870518069356709,
+      "loss": 0.1958,
+      "step": 1319
+    },
+    {
+      "epoch": 1.3203713544434372,
+      "grad_norm": 0.1870850771665573,
+      "learning_rate": 0.0001186021960044692,
+      "loss": 0.1938,
+      "step": 1320
+    },
+    {
+      "epoch": 1.3213716357725611,
+      "grad_norm": 0.1700391322374344,
+      "learning_rate": 0.00011849919087492211,
+      "loss": 0.1944,
+      "step": 1321
+    },
+    {
+      "epoch": 1.3223719171016848,
+      "grad_norm": 0.19307498633861542,
+      "learning_rate": 0.00011839616541810983,
+      "loss": 0.23,
+      "step": 1322
+    },
+    {
+      "epoch": 1.3233721984308087,
+      "grad_norm": 0.20897121727466583,
+      "learning_rate": 0.00011829311974723867,
+      "loss": 0.1975,
+      "step": 1323
+    },
+    {
+      "epoch": 1.3243724797599326,
+      "grad_norm": 0.1749228537082672,
+      "learning_rate": 0.00011819005397553723,
+      "loss": 0.175,
+      "step": 1324
+    },
+    {
+      "epoch": 1.3253727610890562,
+      "grad_norm": 0.1899823695421219,
+      "learning_rate": 0.00011808696821625613,
+      "loss": 0.1979,
+      "step": 1325
+    },
+    {
+      "epoch": 1.32637304241818,
+      "grad_norm": 0.1583666056394577,
+      "learning_rate": 0.000117983862582668,
+      "loss": 0.1891,
+      "step": 1326
+    },
+    {
+      "epoch": 1.327373323747304,
+      "grad_norm": 0.16197249293327332,
+      "learning_rate": 0.00011788073718806725,
+      "loss": 0.1789,
+      "step": 1327
+    },
+    {
+      "epoch": 1.3283736050764277,
+      "grad_norm": 0.20761188864707947,
+      "learning_rate": 0.00011777759214577006,
+      "loss": 0.1837,
+      "step": 1328
+    },
+    {
+      "epoch": 1.3293738864055515,
+      "grad_norm": 0.19915562868118286,
+      "learning_rate": 0.00011767442756911417,
+      "loss": 0.2072,
+      "step": 1329
+    },
+    {
+      "epoch": 1.3303741677346754,
+      "grad_norm": 0.17245349287986755,
+      "learning_rate": 0.00011757124357145881,
+      "loss": 0.1829,
+      "step": 1330
+    },
+    {
+      "epoch": 1.3313744490637993,
+      "grad_norm": 0.17270657420158386,
+      "learning_rate": 0.00011746804026618452,
+      "loss": 0.1947,
+      "step": 1331
+    },
+    {
+      "epoch": 1.332374730392923,
+      "grad_norm": 0.15636077523231506,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.1809,
+      "step": 1332
+    },
+    {
+      "epoch": 1.3333750117220469,
+      "grad_norm": 0.17239832878112793,
+      "learning_rate": 0.00011726157618640728,
+      "loss": 0.1991,
+      "step": 1333
+    },
+    {
+      "epoch": 1.3343752930511705,
+      "grad_norm": 0.16701045632362366,
+      "learning_rate": 0.00011715831563877104,
+      "loss": 0.1826,
+      "step": 1334
+    },
+    {
+      "epoch": 1.3353755743802944,
+      "grad_norm": 0.17416009306907654,
+      "learning_rate": 0.00011705503623724898,
+      "loss": 0.2179,
+      "step": 1335
+    },
+    {
+      "epoch": 1.3363758557094183,
+      "grad_norm": 0.1712433099746704,
+      "learning_rate": 0.00011695173809532652,
+      "loss": 0.1632,
+      "step": 1336
+    },
+    {
+      "epoch": 1.3373761370385422,
+      "grad_norm": 0.30863678455352783,
+      "learning_rate": 0.00011684842132650957,
+      "loss": 0.1833,
+      "step": 1337
+    },
+    {
+      "epoch": 1.3383764183676659,
+      "grad_norm": 0.17909862101078033,
+      "learning_rate": 0.00011674508604432464,
+      "loss": 0.1739,
+      "step": 1338
+    },
+    {
+      "epoch": 1.3393766996967897,
+      "grad_norm": 0.1853017956018448,
+      "learning_rate": 0.00011664173236231848,
+      "loss": 0.1953,
+      "step": 1339
+    },
+    {
+      "epoch": 1.3403769810259134,
+      "grad_norm": 0.20876812934875488,
+      "learning_rate": 0.0001165383603940581,
+      "loss": 0.2132,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3413772623550373,
+      "grad_norm": 0.20856992900371552,
+      "learning_rate": 0.00011643497025313061,
+      "loss": 0.1964,
+      "step": 1341
+    },
+    {
+      "epoch": 1.3423775436841612,
+      "grad_norm": 0.18442268669605255,
+      "learning_rate": 0.00011633156205314309,
+      "loss": 0.1889,
+      "step": 1342
+    },
+    {
+      "epoch": 1.343377825013285,
+      "grad_norm": 0.16034339368343353,
+      "learning_rate": 0.00011622813590772244,
+      "loss": 0.1864,
+      "step": 1343
+    },
+    {
+      "epoch": 1.3443781063424087,
+      "grad_norm": 0.1836758852005005,
+      "learning_rate": 0.00011612469193051525,
+      "loss": 0.1957,
+      "step": 1344
+    },
+    {
+      "epoch": 1.3453783876715326,
+      "grad_norm": 0.1490117758512497,
+      "learning_rate": 0.00011602123023518779,
+      "loss": 0.1747,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3463786690006565,
+      "grad_norm": 0.17359519004821777,
+      "learning_rate": 0.00011591775093542572,
+      "loss": 0.2054,
+      "step": 1346
+    },
+    {
+      "epoch": 1.3473789503297802,
+      "grad_norm": 0.20816031098365784,
+      "learning_rate": 0.0001158142541449341,
+      "loss": 0.2331,
+      "step": 1347
+    },
+    {
+      "epoch": 1.348379231658904,
+      "grad_norm": 0.18390358984470367,
+      "learning_rate": 0.00011571073997743716,
+      "loss": 0.1992,
+      "step": 1348
+    },
+    {
+      "epoch": 1.349379512988028,
+      "grad_norm": 0.13852614164352417,
+      "learning_rate": 0.0001156072085466783,
+      "loss": 0.1526,
+      "step": 1349
+    },
+    {
+      "epoch": 1.3503797943171518,
+      "grad_norm": 0.18713389337062836,
+      "learning_rate": 0.00011550365996641979,
+      "loss": 0.1872,
+      "step": 1350
+    },
+    {
+      "epoch": 1.3513800756462755,
+      "grad_norm": 0.2727030813694,
+      "learning_rate": 0.00011540009435044281,
+      "loss": 0.1843,
+      "step": 1351
+    },
+    {
+      "epoch": 1.3523803569753994,
+      "grad_norm": 0.24894751608371735,
+      "learning_rate": 0.00011529651181254723,
+      "loss": 0.1943,
+      "step": 1352
+    },
+    {
+      "epoch": 1.353380638304523,
+      "grad_norm": 0.1970788687467575,
+      "learning_rate": 0.0001151929124665516,
+      "loss": 0.1961,
+      "step": 1353
+    },
+    {
+      "epoch": 1.354380919633647,
+      "grad_norm": 0.18298476934432983,
+      "learning_rate": 0.00011508929642629274,
+      "loss": 0.1961,
+      "step": 1354
+    },
+    {
+      "epoch": 1.3553812009627708,
+      "grad_norm": 0.24999108910560608,
+      "learning_rate": 0.00011498566380562601,
+      "loss": 0.2149,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3563814822918947,
+      "grad_norm": 0.1643693745136261,
+      "learning_rate": 0.0001148820147184249,
+      "loss": 0.1835,
+      "step": 1356
+    },
+    {
+      "epoch": 1.3573817636210184,
+      "grad_norm": 0.16976682841777802,
+      "learning_rate": 0.00011477834927858104,
+      "loss": 0.1751,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3583820449501423,
+      "grad_norm": 0.19875158369541168,
+      "learning_rate": 0.00011467466760000399,
+      "loss": 0.2048,
+      "step": 1358
+    },
+    {
+      "epoch": 1.359382326279266,
+      "grad_norm": 0.18397216498851776,
+      "learning_rate": 0.00011457096979662114,
+      "loss": 0.1723,
+      "step": 1359
+    },
+    {
+      "epoch": 1.3603826076083898,
+      "grad_norm": 0.22913973033428192,
+      "learning_rate": 0.00011446725598237767,
+      "loss": 0.2257,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3613828889375137,
+      "grad_norm": 0.1700964868068695,
+      "learning_rate": 0.00011436352627123623,
+      "loss": 0.1934,
+      "step": 1361
+    },
+    {
+      "epoch": 1.3623831702666376,
+      "grad_norm": 0.16378380358219147,
+      "learning_rate": 0.00011425978077717709,
+      "loss": 0.1856,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3633834515957612,
+      "grad_norm": 0.19228540360927582,
+      "learning_rate": 0.00011415601961419775,
+      "loss": 0.1934,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3643837329248851,
+      "grad_norm": 0.29254934191703796,
+      "learning_rate": 0.00011405224289631295,
+      "loss": 0.1806,
+      "step": 1364
+    },
+    {
+      "epoch": 1.365384014254009,
+      "grad_norm": 0.18236131966114044,
+      "learning_rate": 0.00011394845073755455,
+      "loss": 0.2027,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3663842955831327,
+      "grad_norm": 0.16580255329608917,
+      "learning_rate": 0.0001138446432519714,
+      "loss": 0.2062,
+      "step": 1366
+    },
+    {
+      "epoch": 1.3673845769122566,
+      "grad_norm": 0.19767959415912628,
+      "learning_rate": 0.00011374082055362909,
+      "loss": 0.2425,
+      "step": 1367
+    },
+    {
+      "epoch": 1.3683848582413805,
+      "grad_norm": 0.1766742318868637,
+      "learning_rate": 0.00011363698275661001,
+      "loss": 0.2314,
+      "step": 1368
+    },
+    {
+      "epoch": 1.3693851395705043,
+      "grad_norm": 0.19101130962371826,
+      "learning_rate": 0.00011353312997501313,
+      "loss": 0.2262,
+      "step": 1369
+    },
+    {
+      "epoch": 1.370385420899628,
+      "grad_norm": 0.15882061421871185,
+      "learning_rate": 0.00011342926232295386,
+      "loss": 0.1896,
+      "step": 1370
+    },
+    {
+      "epoch": 1.371385702228752,
+      "grad_norm": 0.1638452559709549,
+      "learning_rate": 0.00011332537991456398,
+      "loss": 0.1651,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3723859835578756,
+      "grad_norm": 0.17697995901107788,
+      "learning_rate": 0.00011322148286399147,
+      "loss": 0.1867,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3733862648869994,
+      "grad_norm": 0.22525693476200104,
+      "learning_rate": 0.0001131175712854004,
+      "loss": 0.2256,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3743865462161233,
+      "grad_norm": 0.17651285231113434,
+      "learning_rate": 0.00011301364529297079,
+      "loss": 0.1989,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3753868275452472,
+      "grad_norm": 0.20560400187969208,
+      "learning_rate": 0.0001129097050008985,
+      "loss": 0.192,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3763871088743709,
+      "grad_norm": 0.17010372877120972,
+      "learning_rate": 0.00011280575052339514,
+      "loss": 0.1925,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3773873902034948,
+      "grad_norm": 0.17165188491344452,
+      "learning_rate": 0.00011270178197468789,
+      "loss": 0.2056,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3783876715326184,
+      "grad_norm": 0.17581522464752197,
+      "learning_rate": 0.00011259779946901934,
+      "loss": 0.1889,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3793879528617423,
+      "grad_norm": 0.16023679077625275,
+      "learning_rate": 0.0001124938031206475,
+      "loss": 0.1562,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3803882341908662,
+      "grad_norm": 0.17573261260986328,
+      "learning_rate": 0.00011238979304384554,
+      "loss": 0.191,
+      "step": 1380
+    },
+    {
+      "epoch": 1.38138851551999,
+      "grad_norm": 0.16171301901340485,
+      "learning_rate": 0.0001122857693529017,
+      "loss": 0.1804,
+      "step": 1381
+    },
+    {
+      "epoch": 1.3823887968491138,
+      "grad_norm": 0.19086718559265137,
+      "learning_rate": 0.0001121817321621192,
+      "loss": 0.2175,
+      "step": 1382
+    },
+    {
+      "epoch": 1.3833890781782376,
+      "grad_norm": 0.18050417304039001,
+      "learning_rate": 0.00011207768158581613,
+      "loss": 0.1693,
+      "step": 1383
+    },
+    {
+      "epoch": 1.3843893595073615,
+      "grad_norm": 0.18947653472423553,
+      "learning_rate": 0.00011197361773832525,
+      "loss": 0.2077,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3853896408364852,
+      "grad_norm": 0.1669187843799591,
+      "learning_rate": 0.00011186954073399387,
+      "loss": 0.202,
+      "step": 1385
+    },
+    {
+      "epoch": 1.386389922165609,
+      "grad_norm": 0.156162828207016,
+      "learning_rate": 0.00011176545068718385,
+      "loss": 0.1723,
+      "step": 1386
+    },
+    {
+      "epoch": 1.387390203494733,
+      "grad_norm": 0.16977345943450928,
+      "learning_rate": 0.0001116613477122713,
+      "loss": 0.1913,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3883904848238566,
+      "grad_norm": 0.19801703095436096,
+      "learning_rate": 0.00011155723192364658,
+      "loss": 0.1985,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3893907661529805,
+      "grad_norm": 0.24575640261173248,
+      "learning_rate": 0.00011145310343571411,
+      "loss": 0.2073,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3903910474821044,
+      "grad_norm": 0.21002817153930664,
+      "learning_rate": 0.00011134896236289224,
+      "loss": 0.1726,
+      "step": 1390
+    },
+    {
+      "epoch": 1.391391328811228,
+      "grad_norm": 0.15609630942344666,
+      "learning_rate": 0.0001112448088196132,
+      "loss": 0.181,
+      "step": 1391
+    },
+    {
+      "epoch": 1.392391610140352,
+      "grad_norm": 0.18997639417648315,
+      "learning_rate": 0.00011114064292032282,
+      "loss": 0.1972,
+      "step": 1392
+    },
+    {
+      "epoch": 1.3933918914694758,
+      "grad_norm": 0.17887260019779205,
+      "learning_rate": 0.0001110364647794807,
+      "loss": 0.1881,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3943921727985997,
+      "grad_norm": 0.1690485030412674,
+      "learning_rate": 0.00011093227451155974,
+      "loss": 0.18,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3953924541277234,
+      "grad_norm": 0.167836993932724,
+      "learning_rate": 0.0001108280722310462,
+      "loss": 0.195,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3963927354568473,
+      "grad_norm": 0.18403775990009308,
+      "learning_rate": 0.0001107238580524395,
+      "loss": 0.2048,
+      "step": 1396
+    },
+    {
+      "epoch": 1.397393016785971,
+      "grad_norm": 0.16556072235107422,
+      "learning_rate": 0.00011061963209025223,
+      "loss": 0.1761,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3983932981150948,
+      "grad_norm": 0.249759703874588,
+      "learning_rate": 0.00011051539445900983,
+      "loss": 0.1772,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3993935794442187,
+      "grad_norm": 0.14909394085407257,
+      "learning_rate": 0.00011041114527325065,
+      "loss": 0.1577,
+      "step": 1399
+    },
+    {
+      "epoch": 1.4003938607733426,
+      "grad_norm": 0.17442744970321655,
+      "learning_rate": 0.00011030688464752566,
+      "loss": 0.1894,
+      "step": 1400
+    },
+    {
+      "epoch": 1.4013941421024663,
+      "grad_norm": 0.160821795463562,
+      "learning_rate": 0.00011020261269639842,
+      "loss": 0.1905,
+      "step": 1401
+    },
+    {
+      "epoch": 1.4023944234315902,
+      "grad_norm": 0.1798688769340515,
+      "learning_rate": 0.000110098329534445,
+      "loss": 0.1879,
+      "step": 1402
+    },
+    {
+      "epoch": 1.4033947047607138,
+      "grad_norm": 0.18705101311206818,
+      "learning_rate": 0.00010999403527625367,
+      "loss": 0.1798,
+      "step": 1403
+    },
+    {
+      "epoch": 1.4043949860898377,
+      "grad_norm": 0.17905879020690918,
+      "learning_rate": 0.00010988973003642499,
+      "loss": 0.2032,
+      "step": 1404
+    },
+    {
+      "epoch": 1.4053952674189616,
+      "grad_norm": 0.18656395375728607,
+      "learning_rate": 0.00010978541392957156,
+      "loss": 0.1996,
+      "step": 1405
+    },
+    {
+      "epoch": 1.4063955487480855,
+      "grad_norm": 0.18419604003429413,
+      "learning_rate": 0.00010968108707031792,
+      "loss": 0.211,
+      "step": 1406
+    },
+    {
+      "epoch": 1.4073958300772091,
+      "grad_norm": 0.17336289584636688,
+      "learning_rate": 0.00010957674957330042,
+      "loss": 0.1981,
+      "step": 1407
+    },
+    {
+      "epoch": 1.408396111406333,
+      "grad_norm": 0.22460317611694336,
+      "learning_rate": 0.00010947240155316707,
+      "loss": 0.1862,
+      "step": 1408
+    },
+    {
+      "epoch": 1.409396392735457,
+      "grad_norm": 0.1774190068244934,
+      "learning_rate": 0.00010936804312457749,
+      "loss": 0.173,
+      "step": 1409
+    },
+    {
+      "epoch": 1.4103966740645806,
+      "grad_norm": 0.18285249173641205,
+      "learning_rate": 0.00010926367440220276,
+      "loss": 0.1956,
+      "step": 1410
+    },
+    {
+      "epoch": 1.4113969553937045,
+      "grad_norm": 0.18456754088401794,
+      "learning_rate": 0.00010915929550072517,
+      "loss": 0.1697,
+      "step": 1411
+    },
+    {
+      "epoch": 1.4123972367228284,
+      "grad_norm": 0.19202814996242523,
+      "learning_rate": 0.00010905490653483827,
+      "loss": 0.2337,
+      "step": 1412
+    },
+    {
+      "epoch": 1.4133975180519522,
+      "grad_norm": 0.24877160787582397,
+      "learning_rate": 0.00010895050761924668,
+      "loss": 0.1991,
+      "step": 1413
+    },
+    {
+      "epoch": 1.414397799381076,
+      "grad_norm": 0.16730110347270966,
+      "learning_rate": 0.00010884609886866588,
+      "loss": 0.175,
+      "step": 1414
+    },
+    {
+      "epoch": 1.4153980807101998,
+      "grad_norm": 0.1627979874610901,
+      "learning_rate": 0.00010874168039782227,
+      "loss": 0.1754,
+      "step": 1415
+    },
+    {
+      "epoch": 1.4163983620393235,
+      "grad_norm": 0.22207699716091156,
+      "learning_rate": 0.00010863725232145286,
+      "loss": 0.227,
+      "step": 1416
+    },
+    {
+      "epoch": 1.4173986433684473,
+      "grad_norm": 0.15587595105171204,
+      "learning_rate": 0.00010853281475430517,
+      "loss": 0.1911,
+      "step": 1417
+    },
+    {
+      "epoch": 1.4183989246975712,
+      "grad_norm": 0.17471295595169067,
+      "learning_rate": 0.0001084283678111372,
+      "loss": 0.1853,
+      "step": 1418
+    },
+    {
+      "epoch": 1.4193992060266951,
+      "grad_norm": 0.18903140723705292,
+      "learning_rate": 0.00010832391160671729,
+      "loss": 0.2185,
+      "step": 1419
+    },
+    {
+      "epoch": 1.4203994873558188,
+      "grad_norm": 0.18616153299808502,
+      "learning_rate": 0.00010821944625582392,
+      "loss": 0.2103,
+      "step": 1420
+    },
+    {
+      "epoch": 1.4213997686849427,
+      "grad_norm": 0.18851277232170105,
+      "learning_rate": 0.00010811497187324555,
+      "loss": 0.2154,
+      "step": 1421
+    },
+    {
+      "epoch": 1.4224000500140663,
+      "grad_norm": 0.19893857836723328,
+      "learning_rate": 0.00010801048857378071,
+      "loss": 0.2242,
+      "step": 1422
+    },
+    {
+      "epoch": 1.4234003313431902,
+      "grad_norm": 0.1547599732875824,
+      "learning_rate": 0.00010790599647223763,
+      "loss": 0.2081,
+      "step": 1423
+    },
+    {
+      "epoch": 1.424400612672314,
+      "grad_norm": 0.15343767404556274,
+      "learning_rate": 0.0001078014956834342,
+      "loss": 0.1768,
+      "step": 1424
+    },
+    {
+      "epoch": 1.425400894001438,
+      "grad_norm": 0.16398976743221283,
+      "learning_rate": 0.00010769698632219794,
+      "loss": 0.1842,
+      "step": 1425
+    },
+    {
+      "epoch": 1.4264011753305617,
+      "grad_norm": 0.1657567173242569,
+      "learning_rate": 0.00010759246850336572,
+      "loss": 0.1655,
+      "step": 1426
+    },
+    {
+      "epoch": 1.4274014566596855,
+      "grad_norm": 0.1761874556541443,
+      "learning_rate": 0.0001074879423417837,
+      "loss": 0.1997,
+      "step": 1427
+    },
+    {
+      "epoch": 1.4284017379888094,
+      "grad_norm": 0.1825510412454605,
+      "learning_rate": 0.00010738340795230721,
+      "loss": 0.1846,
+      "step": 1428
+    },
+    {
+      "epoch": 1.429402019317933,
+      "grad_norm": 0.20601406693458557,
+      "learning_rate": 0.00010727886544980068,
+      "loss": 0.19,
+      "step": 1429
+    },
+    {
+      "epoch": 1.430402300647057,
+      "grad_norm": 0.1835612654685974,
+      "learning_rate": 0.00010717431494913741,
+      "loss": 0.191,
+      "step": 1430
+    },
+    {
+      "epoch": 1.4314025819761809,
+      "grad_norm": 0.17501139640808105,
+      "learning_rate": 0.00010706975656519946,
+      "loss": 0.1633,
+      "step": 1431
+    },
+    {
+      "epoch": 1.4324028633053048,
+      "grad_norm": 0.18430274724960327,
+      "learning_rate": 0.00010696519041287765,
+      "loss": 0.1917,
+      "step": 1432
+    },
+    {
+      "epoch": 1.4334031446344284,
+      "grad_norm": 0.178082674741745,
+      "learning_rate": 0.0001068606166070712,
+      "loss": 0.1783,
+      "step": 1433
+    },
+    {
+      "epoch": 1.4344034259635523,
+      "grad_norm": 0.17774569988250732,
+      "learning_rate": 0.00010675603526268785,
+      "loss": 0.193,
+      "step": 1434
+    },
+    {
+      "epoch": 1.435403707292676,
+      "grad_norm": 0.1701330691576004,
+      "learning_rate": 0.00010665144649464356,
+      "loss": 0.2064,
+      "step": 1435
+    },
+    {
+      "epoch": 1.4364039886217999,
+      "grad_norm": 0.2015557885169983,
+      "learning_rate": 0.00010654685041786249,
+      "loss": 0.2575,
+      "step": 1436
+    },
+    {
+      "epoch": 1.4374042699509237,
+      "grad_norm": 0.15367308259010315,
+      "learning_rate": 0.00010644224714727681,
+      "loss": 0.1742,
+      "step": 1437
+    },
+    {
+      "epoch": 1.4384045512800476,
+      "grad_norm": 0.3574206531047821,
+      "learning_rate": 0.0001063376367978266,
+      "loss": 0.2037,
+      "step": 1438
+    },
+    {
+      "epoch": 1.4394048326091713,
+      "grad_norm": 0.19089138507843018,
+      "learning_rate": 0.00010623301948445971,
+      "loss": 0.2037,
+      "step": 1439
+    },
+    {
+      "epoch": 1.4404051139382952,
+      "grad_norm": 0.16550390422344208,
+      "learning_rate": 0.00010612839532213164,
+      "loss": 0.196,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4414053952674188,
+      "grad_norm": 0.17209358513355255,
+      "learning_rate": 0.00010602376442580544,
+      "loss": 0.1799,
+      "step": 1441
+    },
+    {
+      "epoch": 1.4424056765965427,
+      "grad_norm": 0.17247600853443146,
+      "learning_rate": 0.00010591912691045152,
+      "loss": 0.1908,
+      "step": 1442
+    },
+    {
+      "epoch": 1.4434059579256666,
+      "grad_norm": 0.17946626245975494,
+      "learning_rate": 0.00010581448289104758,
+      "loss": 0.1934,
+      "step": 1443
+    },
+    {
+      "epoch": 1.4444062392547905,
+      "grad_norm": 0.18610835075378418,
+      "learning_rate": 0.00010570983248257853,
+      "loss": 0.1921,
+      "step": 1444
+    },
+    {
+      "epoch": 1.4454065205839142,
+      "grad_norm": 0.16108182072639465,
+      "learning_rate": 0.00010560517580003617,
+      "loss": 0.1578,
+      "step": 1445
+    },
+    {
+      "epoch": 1.446406801913038,
+      "grad_norm": 0.21424968540668488,
+      "learning_rate": 0.00010550051295841931,
+      "loss": 0.183,
+      "step": 1446
+    },
+    {
+      "epoch": 1.447407083242162,
+      "grad_norm": 0.19121809303760529,
+      "learning_rate": 0.00010539584407273349,
+      "loss": 0.1973,
+      "step": 1447
+    },
+    {
+      "epoch": 1.4484073645712856,
+      "grad_norm": 0.16361166536808014,
+      "learning_rate": 0.00010529116925799085,
+      "loss": 0.1669,
+      "step": 1448
+    },
+    {
+      "epoch": 1.4494076459004095,
+      "grad_norm": 0.17561350762844086,
+      "learning_rate": 0.00010518648862921012,
+      "loss": 0.1766,
+      "step": 1449
+    },
+    {
+      "epoch": 1.4504079272295334,
+      "grad_norm": 0.17351047694683075,
+      "learning_rate": 0.00010508180230141635,
+      "loss": 0.2103,
+      "step": 1450
+    },
+    {
+      "epoch": 1.451408208558657,
+      "grad_norm": 0.15463320910930634,
+      "learning_rate": 0.00010497711038964086,
+      "loss": 0.1744,
+      "step": 1451
+    },
+    {
+      "epoch": 1.452408489887781,
+      "grad_norm": 0.18692006170749664,
+      "learning_rate": 0.0001048724130089212,
+      "loss": 0.202,
+      "step": 1452
+    },
+    {
+      "epoch": 1.4534087712169048,
+      "grad_norm": 0.16209864616394043,
+      "learning_rate": 0.00010476771027430086,
+      "loss": 0.168,
+      "step": 1453
+    },
+    {
+      "epoch": 1.4544090525460285,
+      "grad_norm": 0.3008090555667877,
+      "learning_rate": 0.00010466300230082911,
+      "loss": 0.1985,
+      "step": 1454
+    },
+    {
+      "epoch": 1.4554093338751524,
+      "grad_norm": 0.16463787853717804,
+      "learning_rate": 0.00010455828920356115,
+      "loss": 0.1922,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4564096152042763,
+      "grad_norm": 0.1805451661348343,
+      "learning_rate": 0.00010445357109755771,
+      "loss": 0.1967,
+      "step": 1456
+    },
+    {
+      "epoch": 1.4574098965334001,
+      "grad_norm": 0.19637352228164673,
+      "learning_rate": 0.00010434884809788508,
+      "loss": 0.197,
+      "step": 1457
+    },
+    {
+      "epoch": 1.4584101778625238,
+      "grad_norm": 0.17542944848537445,
+      "learning_rate": 0.00010424412031961484,
+      "loss": 0.1802,
+      "step": 1458
+    },
+    {
+      "epoch": 1.4594104591916477,
+      "grad_norm": 0.17630788683891296,
+      "learning_rate": 0.00010413938787782394,
+      "loss": 0.1843,
+      "step": 1459
+    },
+    {
+      "epoch": 1.4604107405207714,
+      "grad_norm": 0.3292463421821594,
+      "learning_rate": 0.00010403465088759437,
+      "loss": 0.188,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4614110218498952,
+      "grad_norm": 0.18971534073352814,
+      "learning_rate": 0.00010392990946401313,
+      "loss": 0.1873,
+      "step": 1461
+    },
+    {
+      "epoch": 1.4624113031790191,
+      "grad_norm": 0.17018955945968628,
+      "learning_rate": 0.00010382516372217215,
+      "loss": 0.2125,
+      "step": 1462
+    },
+    {
+      "epoch": 1.463411584508143,
+      "grad_norm": 0.18013770878314972,
+      "learning_rate": 0.000103720413777168,
+      "loss": 0.2277,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4644118658372667,
+      "grad_norm": 0.15812082588672638,
+      "learning_rate": 0.00010361565974410192,
+      "loss": 0.1715,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4654121471663906,
+      "grad_norm": 0.1643846333026886,
+      "learning_rate": 0.00010351090173807969,
+      "loss": 0.1799,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4664124284955142,
+      "grad_norm": 0.1803956776857376,
+      "learning_rate": 0.00010340613987421137,
+      "loss": 0.181,
+      "step": 1466
+    },
+    {
+      "epoch": 1.4674127098246381,
+      "grad_norm": 0.37751659750938416,
+      "learning_rate": 0.00010330137426761135,
+      "loss": 0.1822,
+      "step": 1467
+    },
+    {
+      "epoch": 1.468412991153762,
+      "grad_norm": 0.16566865146160126,
+      "learning_rate": 0.00010319660503339808,
+      "loss": 0.1969,
+      "step": 1468
+    },
+    {
+      "epoch": 1.469413272482886,
+      "grad_norm": 0.17209063470363617,
+      "learning_rate": 0.00010309183228669397,
+      "loss": 0.1897,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4704135538120096,
+      "grad_norm": 0.20157137513160706,
+      "learning_rate": 0.00010298705614262532,
+      "loss": 0.1951,
+      "step": 1470
+    },
+    {
+      "epoch": 1.4714138351411334,
+      "grad_norm": 0.1866229921579361,
+      "learning_rate": 0.0001028822767163222,
+      "loss": 0.1795,
+      "step": 1471
+    },
+    {
+      "epoch": 1.4724141164702573,
+      "grad_norm": 0.17085832357406616,
+      "learning_rate": 0.00010277749412291824,
+      "loss": 0.1888,
+      "step": 1472
+    },
+    {
+      "epoch": 1.473414397799381,
+      "grad_norm": 0.1720651537179947,
+      "learning_rate": 0.00010267270847755048,
+      "loss": 0.2,
+      "step": 1473
+    },
+    {
+      "epoch": 1.4744146791285049,
+      "grad_norm": 0.1681736707687378,
+      "learning_rate": 0.00010256791989535952,
+      "loss": 0.1862,
+      "step": 1474
+    },
+    {
+      "epoch": 1.4754149604576288,
+      "grad_norm": 0.15286403894424438,
+      "learning_rate": 0.00010246312849148899,
+      "loss": 0.1566,
+      "step": 1475
+    },
+    {
+      "epoch": 1.4764152417867527,
+      "grad_norm": 0.17605531215667725,
+      "learning_rate": 0.00010235833438108571,
+      "loss": 0.1831,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4774155231158763,
+      "grad_norm": 0.16123445332050323,
+      "learning_rate": 0.00010225353767929944,
+      "loss": 0.1747,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4784158044450002,
+      "grad_norm": 0.195534810423851,
+      "learning_rate": 0.00010214873850128282,
+      "loss": 0.1859,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4794160857741239,
+      "grad_norm": 0.17487269639968872,
+      "learning_rate": 0.00010204393696219117,
+      "loss": 0.2025,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4804163671032478,
+      "grad_norm": 0.17277052998542786,
+      "learning_rate": 0.00010193913317718244,
+      "loss": 0.1907,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4814166484323716,
+      "grad_norm": 0.17025963962078094,
+      "learning_rate": 0.00010183432726141706,
+      "loss": 0.1724,
+      "step": 1481
+    },
+    {
+      "epoch": 1.4824169297614955,
+      "grad_norm": 0.18107976019382477,
+      "learning_rate": 0.00010172951933005775,
+      "loss": 0.1868,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4834172110906192,
+      "grad_norm": 0.31617456674575806,
+      "learning_rate": 0.00010162470949826948,
+      "loss": 0.1909,
+      "step": 1483
+    },
+    {
+      "epoch": 1.484417492419743,
+      "grad_norm": 0.14407062530517578,
+      "learning_rate": 0.0001015198978812193,
+      "loss": 0.1766,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4854177737488667,
+      "grad_norm": 0.17117851972579956,
+      "learning_rate": 0.00010141508459407623,
+      "loss": 0.2068,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4864180550779906,
+      "grad_norm": 0.15799662470817566,
+      "learning_rate": 0.0001013102697520111,
+      "loss": 0.1831,
+      "step": 1486
+    },
+    {
+      "epoch": 1.4874183364071145,
+      "grad_norm": 0.14950646460056305,
+      "learning_rate": 0.00010120545347019647,
+      "loss": 0.1716,
+      "step": 1487
+    },
+    {
+      "epoch": 1.4884186177362384,
+      "grad_norm": 0.1806739866733551,
+      "learning_rate": 0.00010110063586380646,
+      "loss": 0.1959,
+      "step": 1488
+    },
+    {
+      "epoch": 1.489418899065362,
+      "grad_norm": 0.17430248856544495,
+      "learning_rate": 0.00010099581704801673,
+      "loss": 0.1913,
+      "step": 1489
+    },
+    {
+      "epoch": 1.490419180394486,
+      "grad_norm": 0.19120921194553375,
+      "learning_rate": 0.00010089099713800414,
+      "loss": 0.2127,
+      "step": 1490
+    },
+    {
+      "epoch": 1.4914194617236098,
+      "grad_norm": 0.14513321220874786,
+      "learning_rate": 0.00010078617624894684,
+      "loss": 0.162,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4924197430527335,
+      "grad_norm": 0.1949031502008438,
+      "learning_rate": 0.000100681354496024,
+      "loss": 0.2051,
+      "step": 1492
+    },
+    {
+      "epoch": 1.4934200243818574,
+      "grad_norm": 0.17670153081417084,
+      "learning_rate": 0.00010057653199441581,
+      "loss": 0.1888,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4944203057109813,
+      "grad_norm": 0.19661328196525574,
+      "learning_rate": 0.00010047170885930324,
+      "loss": 0.2033,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4954205870401052,
+      "grad_norm": 0.18825002014636993,
+      "learning_rate": 0.00010036688520586788,
+      "loss": 0.1933,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4964208683692288,
+      "grad_norm": 0.1821403205394745,
+      "learning_rate": 0.00010026206114929209,
+      "loss": 0.2039,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4974211496983527,
+      "grad_norm": 0.18830259144306183,
+      "learning_rate": 0.00010015723680475846,
+      "loss": 0.1915,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4984214310274764,
+      "grad_norm": 0.17046190798282623,
+      "learning_rate": 0.00010005241228745004,
+      "loss": 0.1812,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4994217123566003,
+      "grad_norm": 0.16060322523117065,
+      "learning_rate": 9.994758771254997e-05,
+      "loss": 0.1645,
+      "step": 1499
+    },
+    {
+      "epoch": 1.5004219936857242,
+      "grad_norm": 0.1693597286939621,
+      "learning_rate": 9.984276319524154e-05,
+      "loss": 0.1797,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2997,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2846369280956826e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}