diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,121 +1,5060 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.013908205841446454,
-  "eval_steps": 3,
-  "global_step": 10,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 719,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.0013908205841446453,
-      "grad_norm": 0.8333961367607117,
-      "learning_rate": 2e-05,
-      "loss": 1.3831,
-      "step": 1
-    },
-    {
-      "epoch": 0.0013908205841446453,
-      "eval_loss": 1.014951229095459,
-      "eval_runtime": 4.9385,
-      "eval_samples_per_second": 30.779,
-      "eval_steps_per_second": 30.779,
+      "grad_norm": 0.8009753823280334,
+      "learning_rate": 0.00019999904542231437,
+      "loss": 1.3829,
       "step": 1
     },
     {
       "epoch": 0.0027816411682892906,
-      "grad_norm": 0.9491351246833801,
-      "learning_rate": 4e-05,
-      "loss": 1.1721,
+      "grad_norm": 0.7863552570343018,
+      "learning_rate": 0.00019999618170748185,
+      "loss": 1.1518,
       "step": 2
     },
     {
       "epoch": 0.004172461752433936,
-      "grad_norm": 0.7000308036804199,
-      "learning_rate": 6e-05,
-      "loss": 0.8161,
-      "step": 3
-    },
-    {
-      "epoch": 0.004172461752433936,
-      "eval_loss": 1.0096120834350586,
-      "eval_runtime": 4.9607,
-      "eval_samples_per_second": 30.641,
-      "eval_steps_per_second": 30.641,
+      "grad_norm": 0.6151663661003113,
+      "learning_rate": 0.00019999140891017515,
+      "loss": 0.7947,
       "step": 3
     },
     {
       "epoch": 0.005563282336578581,
-      "grad_norm": 1.0017286539077759,
-      "learning_rate": 8e-05,
-      "loss": 0.8552,
+      "grad_norm": 0.6827961802482605,
+      "learning_rate": 0.00019998472712151445,
+      "loss": 0.7977,
       "step": 4
     },
     {
       "epoch": 0.006954102920723227,
-      "grad_norm": 1.0567864179611206,
-      "learning_rate": 0.0001,
-      "loss": 0.9583,
+      "grad_norm": 0.756430447101593,
+      "learning_rate": 0.00019997613646906544,
+      "loss": 0.8856,
       "step": 5
     },
     {
       "epoch": 0.008344923504867872,
-      "grad_norm": 0.7813672423362732,
-      "learning_rate": 0.00012,
-      "loss": 0.8737,
-      "step": 6
-    },
-    {
-      "epoch": 0.008344923504867872,
-      "eval_loss": 0.9795357584953308,
-      "eval_runtime": 5.1418,
-      "eval_samples_per_second": 29.562,
-      "eval_steps_per_second": 29.562,
+      "grad_norm": 0.5014609098434448,
+      "learning_rate": 0.00019996563711683704,
+      "loss": 0.813,
       "step": 6
     },
     {
       "epoch": 0.009735744089012517,
-      "grad_norm": 0.6960843801498413,
-      "learning_rate": 0.00014,
-      "loss": 0.7209,
+      "grad_norm": 0.5118421912193298,
+      "learning_rate": 0.0001999532292652782,
+      "loss": 0.6667,
       "step": 7
     },
     {
       "epoch": 0.011126564673157162,
-      "grad_norm": 0.6831668615341187,
-      "learning_rate": 0.00016,
-      "loss": 0.6779,
+      "grad_norm": 0.5027555823326111,
+      "learning_rate": 0.0001999389131512741,
+      "loss": 0.6341,
       "step": 8
     },
     {
       "epoch": 0.012517385257301807,
-      "grad_norm": 0.4121267795562744,
-      "learning_rate": 0.00018,
-      "loss": 0.5585,
-      "step": 9
-    },
-    {
-      "epoch": 0.012517385257301807,
-      "eval_loss": 0.9398213624954224,
-      "eval_runtime": 5.0845,
-      "eval_samples_per_second": 29.895,
-      "eval_steps_per_second": 29.895,
+      "grad_norm": 0.41336894035339355,
+      "learning_rate": 0.00019992268904814154,
+      "loss": 0.5495,
       "step": 9
     },
     {
       "epoch": 0.013908205841446454,
-      "grad_norm": 0.8026168346405029,
-      "learning_rate": 0.0002,
-      "loss": 0.8507,
+      "grad_norm": 0.7500741481781006,
+      "learning_rate": 0.0001999045572656239,
+      "loss": 0.837,
       "step": 10
+    },
+    {
+      "epoch": 0.015299026425591099,
+      "grad_norm": 0.6654464602470398,
+      "learning_rate": 0.00019988451814988508,
+      "loss": 0.6861,
+      "step": 11
+    },
+    {
+      "epoch": 0.016689847009735744,
+      "grad_norm": 0.5394367575645447,
+      "learning_rate": 0.00019986257208350297,
+      "loss": 0.7721,
+      "step": 12
+    },
+    {
+      "epoch": 0.01808066759388039,
+      "grad_norm": 0.7445791959762573,
+      "learning_rate": 0.00019983871948546197,
+      "loss": 1.0295,
+      "step": 13
+    },
+    {
+      "epoch": 0.019471488178025034,
+      "grad_norm": 0.43704089522361755,
+      "learning_rate": 0.00019981296081114533,
+      "loss": 0.5809,
+      "step": 14
+    },
+    {
+      "epoch": 0.02086230876216968,
+      "grad_norm": 0.4476829767227173,
+      "learning_rate": 0.00019978529655232614,
+      "loss": 0.5618,
+      "step": 15
+    },
+    {
+      "epoch": 0.022253129346314324,
+      "grad_norm": 0.5836384892463684,
+      "learning_rate": 0.00019975572723715808,
+      "loss": 0.8827,
+      "step": 16
+    },
+    {
+      "epoch": 0.02364394993045897,
+      "grad_norm": 0.4332376718521118,
+      "learning_rate": 0.0001997242534301653,
+      "loss": 0.5225,
+      "step": 17
+    },
+    {
+      "epoch": 0.025034770514603615,
+      "grad_norm": 0.552960216999054,
+      "learning_rate": 0.0001996908757322317,
+      "loss": 0.8696,
+      "step": 18
+    },
+    {
+      "epoch": 0.02642559109874826,
+      "grad_norm": 0.5399945974349976,
+      "learning_rate": 0.00019965559478058936,
+      "loss": 0.6978,
+      "step": 19
+    },
+    {
+      "epoch": 0.027816411682892908,
+      "grad_norm": 0.4308752715587616,
+      "learning_rate": 0.00019961841124880656,
+      "loss": 0.9733,
+      "step": 20
+    },
+    {
+      "epoch": 0.02920723226703755,
+      "grad_norm": 0.6925644874572754,
+      "learning_rate": 0.00019957932584677458,
+      "loss": 1.0623,
+      "step": 21
+    },
+    {
+      "epoch": 0.030598052851182198,
+      "grad_norm": 0.46195390820503235,
+      "learning_rate": 0.0001995383393206945,
+      "loss": 0.4816,
+      "step": 22
+    },
+    {
+      "epoch": 0.031988873435326845,
+      "grad_norm": 0.6869767308235168,
+      "learning_rate": 0.00019949545245306284,
+      "loss": 0.892,
+      "step": 23
+    },
+    {
+      "epoch": 0.03337969401947149,
+      "grad_norm": 0.552448570728302,
+      "learning_rate": 0.00019945066606265644,
+      "loss": 0.4939,
+      "step": 24
+    },
+    {
+      "epoch": 0.03477051460361613,
+      "grad_norm": 0.5400081872940063,
+      "learning_rate": 0.0001994039810045172,
+      "loss": 0.987,
+      "step": 25
+    },
+    {
+      "epoch": 0.03616133518776078,
+      "grad_norm": 0.4983847141265869,
+      "learning_rate": 0.0001993553981699353,
+      "loss": 1.0772,
+      "step": 26
+    },
+    {
+      "epoch": 0.037552155771905425,
+      "grad_norm": 0.5919345617294312,
+      "learning_rate": 0.0001993049184864326,
+      "loss": 0.9741,
+      "step": 27
+    },
+    {
+      "epoch": 0.03894297635605007,
+      "grad_norm": 0.42608675360679626,
+      "learning_rate": 0.00019925254291774467,
+      "loss": 0.6234,
+      "step": 28
+    },
+    {
+      "epoch": 0.04033379694019471,
+      "grad_norm": 0.7832518219947815,
+      "learning_rate": 0.00019919827246380254,
+      "loss": 0.745,
+      "step": 29
+    },
+    {
+      "epoch": 0.04172461752433936,
+      "grad_norm": 0.5608259439468384,
+      "learning_rate": 0.0001991421081607134,
+      "loss": 0.8893,
+      "step": 30
+    },
+    {
+      "epoch": 0.043115438108484005,
+      "grad_norm": 0.40105724334716797,
+      "learning_rate": 0.00019908405108074115,
+      "loss": 0.7164,
+      "step": 31
+    },
+    {
+      "epoch": 0.04450625869262865,
+      "grad_norm": 0.6342111825942993,
+      "learning_rate": 0.0001990241023322856,
+      "loss": 1.0731,
+      "step": 32
+    },
+    {
+      "epoch": 0.0458970792767733,
+      "grad_norm": 0.6254178881645203,
+      "learning_rate": 0.0001989622630598615,
+      "loss": 1.0253,
+      "step": 33
+    },
+    {
+      "epoch": 0.04728789986091794,
+      "grad_norm": 0.7923423647880554,
+      "learning_rate": 0.00019889853444407667,
+      "loss": 1.3147,
+      "step": 34
+    },
+    {
+      "epoch": 0.048678720445062586,
+      "grad_norm": 0.593407392501831,
+      "learning_rate": 0.00019883291770160942,
+      "loss": 1.341,
+      "step": 35
+    },
+    {
+      "epoch": 0.05006954102920723,
+      "grad_norm": 0.638047993183136,
+      "learning_rate": 0.00019876541408518525,
+      "loss": 0.9105,
+      "step": 36
+    },
+    {
+      "epoch": 0.05146036161335188,
+      "grad_norm": 0.6722823977470398,
+      "learning_rate": 0.00019869602488355311,
+      "loss": 1.0405,
+      "step": 37
+    },
+    {
+      "epoch": 0.05285118219749652,
+      "grad_norm": 0.4327874183654785,
+      "learning_rate": 0.00019862475142146067,
+      "loss": 0.7478,
+      "step": 38
+    },
+    {
+      "epoch": 0.054242002781641166,
+      "grad_norm": 0.7023957371711731,
+      "learning_rate": 0.00019855159505962906,
+      "loss": 0.8305,
+      "step": 39
+    },
+    {
+      "epoch": 0.055632823365785816,
+      "grad_norm": 0.3530227541923523,
+      "learning_rate": 0.00019847655719472688,
+      "loss": 0.5053,
+      "step": 40
+    },
+    {
+      "epoch": 0.05702364394993046,
+      "grad_norm": 0.4187379777431488,
+      "learning_rate": 0.0001983996392593436,
+      "loss": 0.7392,
+      "step": 41
+    },
+    {
+      "epoch": 0.0584144645340751,
+      "grad_norm": 0.6277675032615662,
+      "learning_rate": 0.00019832084272196205,
+      "loss": 0.9229,
+      "step": 42
+    },
+    {
+      "epoch": 0.059805285118219746,
+      "grad_norm": 0.46730366349220276,
+      "learning_rate": 0.00019824016908693063,
+      "loss": 0.6589,
+      "step": 43
+    },
+    {
+      "epoch": 0.061196105702364396,
+      "grad_norm": 0.42245203256607056,
+      "learning_rate": 0.00019815761989443428,
+      "loss": 0.586,
+      "step": 44
+    },
+    {
+      "epoch": 0.06258692628650904,
+      "grad_norm": 0.4314742386341095,
+      "learning_rate": 0.00019807319672046546,
+      "loss": 0.9604,
+      "step": 45
+    },
+    {
+      "epoch": 0.06397774687065369,
+      "grad_norm": 0.5638596415519714,
+      "learning_rate": 0.00019798690117679363,
+      "loss": 0.9601,
+      "step": 46
+    },
+    {
+      "epoch": 0.06536856745479833,
+      "grad_norm": 0.4781525135040283,
+      "learning_rate": 0.00019789873491093488,
+      "loss": 0.7406,
+      "step": 47
+    },
+    {
+      "epoch": 0.06675938803894298,
+      "grad_norm": 0.6311485171318054,
+      "learning_rate": 0.0001978086996061201,
+      "loss": 0.918,
+      "step": 48
+    },
+    {
+      "epoch": 0.06815020862308763,
+      "grad_norm": 0.6816067695617676,
+      "learning_rate": 0.00019771679698126325,
+      "loss": 0.9729,
+      "step": 49
+    },
+    {
+      "epoch": 0.06954102920723226,
+      "grad_norm": 0.6415221691131592,
+      "learning_rate": 0.0001976230287909282,
+      "loss": 1.1286,
+      "step": 50
+    },
+    {
+      "epoch": 0.07093184979137691,
+      "grad_norm": 0.5188845992088318,
+      "learning_rate": 0.0001975273968252954,
+      "loss": 0.5219,
+      "step": 51
+    },
+    {
+      "epoch": 0.07232267037552156,
+      "grad_norm": 0.4242594540119171,
+      "learning_rate": 0.00019742990291012763,
+      "loss": 0.6322,
+      "step": 52
+    },
+    {
+      "epoch": 0.0737134909596662,
+      "grad_norm": 0.5624274015426636,
+      "learning_rate": 0.00019733054890673518,
+      "loss": 0.8793,
+      "step": 53
+    },
+    {
+      "epoch": 0.07510431154381085,
+      "grad_norm": 0.7936162352561951,
+      "learning_rate": 0.00019722933671194044,
+      "loss": 1.2471,
+      "step": 54
+    },
+    {
+      "epoch": 0.07649513212795549,
+      "grad_norm": 0.5537488460540771,
+      "learning_rate": 0.0001971262682580414,
+      "loss": 0.9449,
+      "step": 55
+    },
+    {
+      "epoch": 0.07788595271210014,
+      "grad_norm": 0.4708329737186432,
+      "learning_rate": 0.000197021345512775,
+      "loss": 0.7764,
+      "step": 56
+    },
+    {
+      "epoch": 0.07927677329624479,
+      "grad_norm": 0.469587504863739,
+      "learning_rate": 0.00019691457047927942,
+      "loss": 0.6973,
+      "step": 57
+    },
+    {
+      "epoch": 0.08066759388038942,
+      "grad_norm": 0.6117117404937744,
+      "learning_rate": 0.00019680594519605598,
+      "loss": 1.1396,
+      "step": 58
+    },
+    {
+      "epoch": 0.08205841446453407,
+      "grad_norm": 0.356476366519928,
+      "learning_rate": 0.00019669547173693016,
+      "loss": 0.6137,
+      "step": 59
+    },
+    {
+      "epoch": 0.08344923504867872,
+      "grad_norm": 0.604932427406311,
+      "learning_rate": 0.0001965831522110119,
+      "loss": 0.7565,
+      "step": 60
+    },
+    {
+      "epoch": 0.08484005563282336,
+      "grad_norm": 0.6221743226051331,
+      "learning_rate": 0.0001964689887626554,
+      "loss": 0.8141,
+      "step": 61
+    },
+    {
+      "epoch": 0.08623087621696801,
+      "grad_norm": 0.6533489227294922,
+      "learning_rate": 0.00019635298357141836,
+      "loss": 0.942,
+      "step": 62
+    },
+    {
+      "epoch": 0.08762169680111266,
+      "grad_norm": 0.5665097236633301,
+      "learning_rate": 0.00019623513885202009,
+      "loss": 0.715,
+      "step": 63
+    },
+    {
+      "epoch": 0.0890125173852573,
+      "grad_norm": 0.48491233587265015,
+      "learning_rate": 0.00019611545685429936,
+      "loss": 0.654,
+      "step": 64
+    },
+    {
+      "epoch": 0.09040333796940195,
+      "grad_norm": 0.5510562062263489,
+      "learning_rate": 0.00019599393986317147,
+      "loss": 0.8199,
+      "step": 65
+    },
+    {
+      "epoch": 0.0917941585535466,
+      "grad_norm": 0.4846614599227905,
+      "learning_rate": 0.0001958705901985846,
+      "loss": 0.374,
+      "step": 66
+    },
+    {
+      "epoch": 0.09318497913769123,
+      "grad_norm": 0.38200896978378296,
+      "learning_rate": 0.00019574541021547543,
+      "loss": 0.5371,
+      "step": 67
+    },
+    {
+      "epoch": 0.09457579972183588,
+      "grad_norm": 0.5550661683082581,
+      "learning_rate": 0.0001956184023037244,
+      "loss": 0.6162,
+      "step": 68
+    },
+    {
+      "epoch": 0.09596662030598054,
+      "grad_norm": 0.6405926942825317,
+      "learning_rate": 0.0001954895688881099,
+      "loss": 0.7948,
+      "step": 69
+    },
+    {
+      "epoch": 0.09735744089012517,
+      "grad_norm": 0.6494671106338501,
+      "learning_rate": 0.00019535891242826193,
+      "loss": 0.9949,
+      "step": 70
+    },
+    {
+      "epoch": 0.09874826147426982,
+      "grad_norm": 0.7433452010154724,
+      "learning_rate": 0.00019522643541861534,
+      "loss": 0.9698,
+      "step": 71
+    },
+    {
+      "epoch": 0.10013908205841446,
+      "grad_norm": 0.6889805197715759,
+      "learning_rate": 0.0001950921403883621,
+      "loss": 0.972,
+      "step": 72
+    },
+    {
+      "epoch": 0.10152990264255911,
+      "grad_norm": 0.48119157552719116,
+      "learning_rate": 0.00019495602990140295,
+      "loss": 0.6302,
+      "step": 73
+    },
+    {
+      "epoch": 0.10292072322670376,
+      "grad_norm": 0.594725489616394,
+      "learning_rate": 0.00019481810655629862,
+      "loss": 0.7976,
+      "step": 74
+    },
+    {
+      "epoch": 0.1043115438108484,
+      "grad_norm": 0.4150237441062927,
+      "learning_rate": 0.00019467837298622003,
+      "loss": 0.6382,
+      "step": 75
+    },
+    {
+      "epoch": 0.10570236439499305,
+      "grad_norm": 0.4758132994174957,
+      "learning_rate": 0.00019453683185889813,
+      "loss": 0.7724,
+      "step": 76
+    },
+    {
+      "epoch": 0.1070931849791377,
+      "grad_norm": 0.5909102559089661,
+      "learning_rate": 0.00019439348587657292,
+      "loss": 1.0946,
+      "step": 77
+    },
+    {
+      "epoch": 0.10848400556328233,
+      "grad_norm": 0.7560714483261108,
+      "learning_rate": 0.000194248337775942,
+      "loss": 0.8244,
+      "step": 78
+    },
+    {
+      "epoch": 0.10987482614742698,
+      "grad_norm": 0.7283512353897095,
+      "learning_rate": 0.0001941013903281081,
+      "loss": 0.6333,
+      "step": 79
+    },
+    {
+      "epoch": 0.11126564673157163,
+      "grad_norm": 0.7111244797706604,
+      "learning_rate": 0.0001939526463385263,
+      "loss": 0.6636,
+      "step": 80
+    },
+    {
+      "epoch": 0.11265646731571627,
+      "grad_norm": 0.3831712603569031,
+      "learning_rate": 0.00019380210864695046,
+      "loss": 0.573,
+      "step": 81
+    },
+    {
+      "epoch": 0.11404728789986092,
+      "grad_norm": 0.6651076078414917,
+      "learning_rate": 0.000193649780127379,
+      "loss": 1.0275,
+      "step": 82
+    },
+    {
+      "epoch": 0.11543810848400557,
+      "grad_norm": 0.6869786381721497,
+      "learning_rate": 0.00019349566368800008,
+      "loss": 0.9826,
+      "step": 83
+    },
+    {
+      "epoch": 0.1168289290681502,
+      "grad_norm": 0.5350708365440369,
+      "learning_rate": 0.00019333976227113597,
+      "loss": 0.877,
+      "step": 84
+    },
+    {
+      "epoch": 0.11821974965229486,
+      "grad_norm": 0.7623734474182129,
+      "learning_rate": 0.0001931820788531869,
+      "loss": 0.8944,
+      "step": 85
+    },
+    {
+      "epoch": 0.11961057023643949,
+      "grad_norm": 0.5682898759841919,
+      "learning_rate": 0.00019302261644457436,
+      "loss": 0.7397,
+      "step": 86
+    },
+    {
+      "epoch": 0.12100139082058414,
+      "grad_norm": 0.5998944044113159,
+      "learning_rate": 0.00019286137808968343,
+      "loss": 0.9021,
+      "step": 87
+    },
+    {
+      "epoch": 0.12239221140472879,
+      "grad_norm": 0.6008535623550415,
+      "learning_rate": 0.00019269836686680486,
+      "loss": 0.6896,
+      "step": 88
+    },
+    {
+      "epoch": 0.12378303198887343,
+      "grad_norm": 0.6002005338668823,
+      "learning_rate": 0.00019253358588807617,
+      "loss": 0.6922,
+      "step": 89
+    },
+    {
+      "epoch": 0.12517385257301808,
+      "grad_norm": 0.8469592332839966,
+      "learning_rate": 0.00019236703829942232,
+      "loss": 0.9103,
+      "step": 90
+    },
+    {
+      "epoch": 0.12656467315716272,
+      "grad_norm": 0.7402297258377075,
+      "learning_rate": 0.0001921987272804954,
+      "loss": 0.9818,
+      "step": 91
+    },
+    {
+      "epoch": 0.12795549374130738,
+      "grad_norm": 0.6879746317863464,
+      "learning_rate": 0.00019202865604461442,
+      "loss": 0.8245,
+      "step": 92
+    },
+    {
+      "epoch": 0.12934631432545202,
+      "grad_norm": 0.679220974445343,
+      "learning_rate": 0.00019185682783870345,
+      "loss": 0.9057,
+      "step": 93
+    },
+    {
+      "epoch": 0.13073713490959665,
+      "grad_norm": 0.7069530487060547,
+      "learning_rate": 0.0001916832459432299,
+      "loss": 1.1019,
+      "step": 94
+    },
+    {
+      "epoch": 0.13212795549374132,
+      "grad_norm": 0.5007368922233582,
+      "learning_rate": 0.00019150791367214182,
+      "loss": 0.7764,
+      "step": 95
+    },
+    {
+      "epoch": 0.13351877607788595,
+      "grad_norm": 0.5823730230331421,
+      "learning_rate": 0.0001913308343728048,
+      "loss": 0.8476,
+      "step": 96
+    },
+    {
+      "epoch": 0.1349095966620306,
+      "grad_norm": 0.52687668800354,
+      "learning_rate": 0.00019115201142593766,
+      "loss": 0.5274,
+      "step": 97
+    },
+    {
+      "epoch": 0.13630041724617525,
+      "grad_norm": 0.6914880871772766,
+      "learning_rate": 0.00019097144824554838,
+      "loss": 0.8151,
+      "step": 98
+    },
+    {
+      "epoch": 0.1376912378303199,
+      "grad_norm": 0.5643054246902466,
+      "learning_rate": 0.00019078914827886861,
+      "loss": 0.5966,
+      "step": 99
+    },
+    {
+      "epoch": 0.13908205841446453,
+      "grad_norm": 0.4006194472312927,
+      "learning_rate": 0.00019060511500628794,
+      "loss": 0.6825,
+      "step": 100
+    },
+    {
+      "epoch": 0.1404728789986092,
+      "grad_norm": 0.6900478601455688,
+      "learning_rate": 0.00019041935194128744,
+      "loss": 0.9402,
+      "step": 101
+    },
+    {
+      "epoch": 0.14186369958275383,
+      "grad_norm": 0.5653002262115479,
+      "learning_rate": 0.00019023186263037272,
+      "loss": 0.6778,
+      "step": 102
+    },
+    {
+      "epoch": 0.14325452016689846,
+      "grad_norm": 0.7496730089187622,
+      "learning_rate": 0.00019004265065300597,
+      "loss": 1.1462,
+      "step": 103
+    },
+    {
+      "epoch": 0.14464534075104313,
+      "grad_norm": 0.8458881378173828,
+      "learning_rate": 0.00018985171962153784,
+      "loss": 0.8839,
+      "step": 104
+    },
+    {
+      "epoch": 0.14603616133518776,
+      "grad_norm": 0.759869396686554,
+      "learning_rate": 0.00018965907318113838,
+      "loss": 0.7793,
+      "step": 105
+    },
+    {
+      "epoch": 0.1474269819193324,
+      "grad_norm": 0.8109004497528076,
+      "learning_rate": 0.00018946471500972742,
+      "loss": 0.9288,
+      "step": 106
+    },
+    {
+      "epoch": 0.14881780250347706,
+      "grad_norm": 0.5695831179618835,
+      "learning_rate": 0.00018926864881790446,
+      "loss": 0.7099,
+      "step": 107
+    },
+    {
+      "epoch": 0.1502086230876217,
+      "grad_norm": 0.42937278747558594,
+      "learning_rate": 0.00018907087834887777,
+      "loss": 0.7653,
+      "step": 108
+    },
+    {
+      "epoch": 0.15159944367176634,
+      "grad_norm": 0.7026939392089844,
+      "learning_rate": 0.0001888714073783928,
+      "loss": 0.7911,
+      "step": 109
+    },
+    {
+      "epoch": 0.15299026425591097,
+      "grad_norm": 0.6589177846908569,
+      "learning_rate": 0.00018867023971466036,
+      "loss": 0.7075,
+      "step": 110
+    },
+    {
+      "epoch": 0.15438108484005564,
+      "grad_norm": 0.3478275239467621,
+      "learning_rate": 0.00018846737919828366,
+      "loss": 0.6684,
+      "step": 111
+    },
+    {
+      "epoch": 0.15577190542420027,
+      "grad_norm": 0.45631977915763855,
+      "learning_rate": 0.00018826282970218518,
+      "loss": 0.729,
+      "step": 112
+    },
+    {
+      "epoch": 0.1571627260083449,
+      "grad_norm": 0.8417097926139832,
+      "learning_rate": 0.0001880565951315326,
+      "loss": 0.9948,
+      "step": 113
+    },
+    {
+      "epoch": 0.15855354659248957,
+      "grad_norm": 0.47765666246414185,
+      "learning_rate": 0.0001878486794236643,
+      "loss": 0.8803,
+      "step": 114
+    },
+    {
+      "epoch": 0.1599443671766342,
+      "grad_norm": 0.5094887018203735,
+      "learning_rate": 0.00018763908654801422,
+      "loss": 0.7931,
+      "step": 115
+    },
+    {
+      "epoch": 0.16133518776077885,
+      "grad_norm": 0.5795468688011169,
+      "learning_rate": 0.00018742782050603594,
+      "loss": 0.7923,
+      "step": 116
+    },
+    {
+      "epoch": 0.1627260083449235,
+      "grad_norm": 0.4326666593551636,
+      "learning_rate": 0.0001872148853311265,
+      "loss": 0.7003,
+      "step": 117
+    },
+    {
+      "epoch": 0.16411682892906815,
+      "grad_norm": 0.39254435896873474,
+      "learning_rate": 0.0001870002850885492,
+      "loss": 0.4221,
+      "step": 118
+    },
+    {
+      "epoch": 0.16550764951321278,
+      "grad_norm": 0.5720130205154419,
+      "learning_rate": 0.00018678402387535607,
+      "loss": 0.9065,
+      "step": 119
+    },
+    {
+      "epoch": 0.16689847009735745,
+      "grad_norm": 0.4030667543411255,
+      "learning_rate": 0.00018656610582030975,
+      "loss": 0.5853,
+      "step": 120
+    },
+    {
+      "epoch": 0.16828929068150209,
+      "grad_norm": 0.6296719908714294,
+      "learning_rate": 0.00018634653508380446,
+      "loss": 0.8816,
+      "step": 121
+    },
+    {
+      "epoch": 0.16968011126564672,
+      "grad_norm": 0.5008291602134705,
+      "learning_rate": 0.0001861253158577867,
+      "loss": 0.77,
+      "step": 122
+    },
+    {
+      "epoch": 0.17107093184979139,
+      "grad_norm": 0.5873888731002808,
+      "learning_rate": 0.00018590245236567522,
+      "loss": 0.7593,
+      "step": 123
+    },
+    {
+      "epoch": 0.17246175243393602,
+      "grad_norm": 0.554957926273346,
+      "learning_rate": 0.0001856779488622803,
+      "loss": 0.7214,
+      "step": 124
+    },
+    {
+      "epoch": 0.17385257301808066,
+      "grad_norm": 0.833118736743927,
+      "learning_rate": 0.00018545180963372272,
+      "loss": 0.7297,
+      "step": 125
+    },
+    {
+      "epoch": 0.17524339360222532,
+      "grad_norm": 0.46049150824546814,
+      "learning_rate": 0.00018522403899735168,
+      "loss": 0.6834,
+      "step": 126
+    },
+    {
+      "epoch": 0.17663421418636996,
+      "grad_norm": 0.7460084557533264,
+      "learning_rate": 0.00018499464130166247,
+      "loss": 1.0782,
+      "step": 127
+    },
+    {
+      "epoch": 0.1780250347705146,
+      "grad_norm": 0.49190595746040344,
+      "learning_rate": 0.00018476362092621358,
+      "loss": 0.7146,
+      "step": 128
+    },
+    {
+      "epoch": 0.17941585535465926,
+      "grad_norm": 0.638210654258728,
+      "learning_rate": 0.00018453098228154286,
+      "loss": 0.9643,
+      "step": 129
+    },
+    {
+      "epoch": 0.1808066759388039,
+      "grad_norm": 0.3647628426551819,
+      "learning_rate": 0.00018429672980908355,
+      "loss": 0.5027,
+      "step": 130
+    },
+    {
+      "epoch": 0.18219749652294853,
+      "grad_norm": 0.5246829390525818,
+      "learning_rate": 0.00018406086798107926,
+      "loss": 0.6835,
+      "step": 131
+    },
+    {
+      "epoch": 0.1835883171070932,
+      "grad_norm": 0.6160104274749756,
+      "learning_rate": 0.00018382340130049878,
+      "loss": 0.8247,
+      "step": 132
+    },
+    {
+      "epoch": 0.18497913769123783,
+      "grad_norm": 0.7959861755371094,
+      "learning_rate": 0.00018358433430094997,
+      "loss": 0.8313,
+      "step": 133
+    },
+    {
+      "epoch": 0.18636995827538247,
+      "grad_norm": 0.5451456904411316,
+      "learning_rate": 0.00018334367154659333,
+      "loss": 0.5423,
+      "step": 134
+    },
+    {
+      "epoch": 0.18776077885952713,
+      "grad_norm": 0.44990184903144836,
+      "learning_rate": 0.00018310141763205472,
+      "loss": 0.7677,
+      "step": 135
+    },
+    {
+      "epoch": 0.18915159944367177,
+      "grad_norm": 0.5349050164222717,
+      "learning_rate": 0.00018285757718233781,
+      "loss": 0.6931,
+      "step": 136
+    },
+    {
+      "epoch": 0.1905424200278164,
+      "grad_norm": 0.48450711369514465,
+      "learning_rate": 0.00018261215485273558,
+      "loss": 0.7665,
+      "step": 137
+    },
+    {
+      "epoch": 0.19193324061196107,
+      "grad_norm": 0.9346706867218018,
+      "learning_rate": 0.00018236515532874166,
+      "loss": 0.5957,
+      "step": 138
+    },
+    {
+      "epoch": 0.1933240611961057,
+      "grad_norm": 0.5466428995132446,
+      "learning_rate": 0.00018211658332596074,
+      "loss": 0.9407,
+      "step": 139
+    },
+    {
+      "epoch": 0.19471488178025034,
+      "grad_norm": 0.5062084794044495,
+      "learning_rate": 0.0001818664435900185,
+      "loss": 0.705,
+      "step": 140
+    },
+    {
+      "epoch": 0.19610570236439498,
+      "grad_norm": 0.8536409139633179,
+      "learning_rate": 0.00018161474089647124,
+      "loss": 0.8505,
+      "step": 141
+    },
+    {
+      "epoch": 0.19749652294853964,
+      "grad_norm": 0.6437822580337524,
+      "learning_rate": 0.00018136148005071438,
+      "loss": 0.8731,
+      "step": 142
+    },
+    {
+      "epoch": 0.19888734353268428,
+      "grad_norm": 0.5764851570129395,
+      "learning_rate": 0.000181106665887891,
+      "loss": 0.8047,
+      "step": 143
+    },
+    {
+      "epoch": 0.20027816411682892,
+      "grad_norm": 0.7331337332725525,
+      "learning_rate": 0.00018085030327279932,
+      "loss": 1.0024,
+      "step": 144
+    },
+    {
+      "epoch": 0.20166898470097358,
+      "grad_norm": 0.5635994672775269,
+      "learning_rate": 0.00018059239709980002,
+      "loss": 0.7239,
+      "step": 145
+    },
+    {
+      "epoch": 0.20305980528511822,
+      "grad_norm": 0.8839379549026489,
+      "learning_rate": 0.00018033295229272269,
+      "loss": 1.0873,
+      "step": 146
+    },
+    {
+      "epoch": 0.20445062586926285,
+      "grad_norm": 0.6360731720924377,
+      "learning_rate": 0.00018007197380477173,
+      "loss": 0.6065,
+      "step": 147
+    },
+    {
+      "epoch": 0.20584144645340752,
+      "grad_norm": 0.5467824339866638,
+      "learning_rate": 0.000179809466618432,
+      "loss": 1.0145,
+      "step": 148
+    },
+    {
+      "epoch": 0.20723226703755215,
+      "grad_norm": 0.6928727626800537,
+      "learning_rate": 0.00017954543574537354,
+      "loss": 0.7633,
+      "step": 149
+    },
+    {
+      "epoch": 0.2086230876216968,
+      "grad_norm": 0.40504512190818787,
+      "learning_rate": 0.0001792798862263559,
+      "loss": 0.287,
+      "step": 150
+    },
+    {
+      "epoch": 0.21001390820584145,
+      "grad_norm": 0.6809450387954712,
+      "learning_rate": 0.0001790128231311321,
+      "loss": 0.7642,
+      "step": 151
+    },
+    {
+      "epoch": 0.2114047287899861,
+      "grad_norm": 0.940665066242218,
+      "learning_rate": 0.00017874425155835146,
+      "loss": 0.767,
+      "step": 152
+    },
+    {
+      "epoch": 0.21279554937413073,
+      "grad_norm": 0.6397515535354614,
+      "learning_rate": 0.00017847417663546263,
+      "loss": 0.4997,
+      "step": 153
+    },
+    {
+      "epoch": 0.2141863699582754,
+      "grad_norm": 0.4192731976509094,
+      "learning_rate": 0.00017820260351861545,
+      "loss": 0.786,
+      "step": 154
+    },
+    {
+      "epoch": 0.21557719054242003,
+      "grad_norm": 0.6076131463050842,
+      "learning_rate": 0.00017792953739256278,
+      "loss": 0.9718,
+      "step": 155
+    },
+    {
+      "epoch": 0.21696801112656466,
+      "grad_norm": 0.6200540065765381,
+      "learning_rate": 0.00017765498347056116,
+      "loss": 0.7627,
+      "step": 156
+    },
+    {
+      "epoch": 0.21835883171070933,
+      "grad_norm": 0.7067424058914185,
+      "learning_rate": 0.00017737894699427153,
+      "loss": 0.9193,
+      "step": 157
+    },
+    {
+      "epoch": 0.21974965229485396,
+      "grad_norm": 0.6337076425552368,
+      "learning_rate": 0.00017710143323365914,
+      "loss": 0.8581,
+      "step": 158
+    },
+    {
+      "epoch": 0.2211404728789986,
+      "grad_norm": 0.6896895170211792,
+      "learning_rate": 0.00017682244748689282,
+      "loss": 0.6753,
+      "step": 159
+    },
+    {
+      "epoch": 0.22253129346314326,
+      "grad_norm": 0.5917553901672363,
+      "learning_rate": 0.00017654199508024396,
+      "loss": 0.66,
+      "step": 160
+    },
+    {
+      "epoch": 0.2239221140472879,
+      "grad_norm": 0.7202566266059875,
+      "learning_rate": 0.00017626008136798475,
+      "loss": 0.9858,
+      "step": 161
+    },
+    {
+      "epoch": 0.22531293463143254,
+      "grad_norm": 0.46285754442214966,
+      "learning_rate": 0.00017597671173228595,
+      "loss": 0.5095,
+      "step": 162
+    },
+    {
+      "epoch": 0.2267037552155772,
+      "grad_norm": 0.7485176920890808,
+      "learning_rate": 0.00017569189158311423,
+      "loss": 1.0158,
+      "step": 163
+    },
+    {
+      "epoch": 0.22809457579972184,
+      "grad_norm": 0.4538167119026184,
+      "learning_rate": 0.0001754056263581287,
+      "loss": 0.7216,
+      "step": 164
+    },
+    {
+      "epoch": 0.22948539638386647,
+      "grad_norm": 0.5545186400413513,
+      "learning_rate": 0.00017511792152257735,
+      "loss": 0.6962,
+      "step": 165
+    },
+    {
+      "epoch": 0.23087621696801114,
+      "grad_norm": 0.35297563672065735,
+      "learning_rate": 0.00017482878256919242,
+      "loss": 0.8238,
+      "step": 166
+    },
+    {
+      "epoch": 0.23226703755215578,
+      "grad_norm": 0.4776262044906616,
+      "learning_rate": 0.00017453821501808582,
+      "loss": 0.5246,
+      "step": 167
+    },
+    {
+      "epoch": 0.2336578581363004,
+      "grad_norm": 0.5610180497169495,
+      "learning_rate": 0.00017424622441664352,
+      "loss": 0.9382,
+      "step": 168
+    },
+    {
+      "epoch": 0.23504867872044508,
+      "grad_norm": 0.507335364818573,
+      "learning_rate": 0.00017395281633941984,
+      "loss": 0.5895,
+      "step": 169
+    },
+    {
+      "epoch": 0.2364394993045897,
+      "grad_norm": 0.44601017236709595,
+      "learning_rate": 0.0001736579963880308,
+      "loss": 0.5757,
+      "step": 170
+    },
+    {
+      "epoch": 0.23783031988873435,
+      "grad_norm": 0.4246615171432495,
+      "learning_rate": 0.00017336177019104736,
+      "loss": 0.5395,
+      "step": 171
+    },
+    {
+      "epoch": 0.23922114047287898,
+      "grad_norm": 0.43568405508995056,
+      "learning_rate": 0.0001730641434038878,
+      "loss": 0.6583,
+      "step": 172
+    },
+    {
+      "epoch": 0.24061196105702365,
+      "grad_norm": 0.5562520623207092,
+      "learning_rate": 0.00017276512170871002,
+      "loss": 0.6103,
+      "step": 173
+    },
+    {
+      "epoch": 0.24200278164116829,
+      "grad_norm": 0.6876732110977173,
+      "learning_rate": 0.0001724647108143027,
+      "loss": 0.8665,
+      "step": 174
+    },
+    {
+      "epoch": 0.24339360222531292,
+      "grad_norm": 0.45132243633270264,
+      "learning_rate": 0.0001721629164559766,
+      "loss": 0.5818,
+      "step": 175
+    },
+    {
+      "epoch": 0.24478442280945759,
+      "grad_norm": 0.6815333366394043,
+      "learning_rate": 0.00017185974439545493,
+      "loss": 1.1134,
+      "step": 176
+    },
+    {
+      "epoch": 0.24617524339360222,
+      "grad_norm": 0.6711398959159851,
+      "learning_rate": 0.00017155520042076328,
+      "loss": 0.9138,
+      "step": 177
+    },
+    {
+      "epoch": 0.24756606397774686,
+      "grad_norm": 0.715816080570221,
+      "learning_rate": 0.00017124929034611943,
+      "loss": 0.9356,
+      "step": 178
+    },
+    {
+      "epoch": 0.24895688456189152,
+      "grad_norm": 0.5493932366371155,
+      "learning_rate": 0.00017094202001182192,
+      "loss": 0.639,
+      "step": 179
+    },
+    {
+      "epoch": 0.25034770514603616,
+      "grad_norm": 0.6813206672668457,
+      "learning_rate": 0.0001706333952841389,
+      "loss": 0.8165,
+      "step": 180
+    },
+    {
+      "epoch": 0.2517385257301808,
+      "grad_norm": 0.6707322597503662,
+      "learning_rate": 0.00017032342205519583,
+      "loss": 0.8762,
+      "step": 181
+    },
+    {
+      "epoch": 0.25312934631432543,
+      "grad_norm": 0.7183409333229065,
+      "learning_rate": 0.0001700121062428634,
+      "loss": 0.9158,
+      "step": 182
+    },
+    {
+      "epoch": 0.2545201668984701,
+      "grad_norm": 0.6149759292602539,
+      "learning_rate": 0.00016969945379064402,
+      "loss": 0.9132,
+      "step": 183
+    },
+    {
+      "epoch": 0.25591098748261476,
+      "grad_norm": 0.572540283203125,
+      "learning_rate": 0.0001693854706675589,
+      "loss": 0.6699,
+      "step": 184
+    },
+    {
+      "epoch": 0.2573018080667594,
+      "grad_norm": 0.5242774486541748,
+      "learning_rate": 0.00016907016286803363,
+      "loss": 0.7539,
+      "step": 185
+    },
+    {
+      "epoch": 0.25869262865090403,
+      "grad_norm": 0.6336389780044556,
+      "learning_rate": 0.00016875353641178403,
+      "loss": 0.9734,
+      "step": 186
+    },
+    {
+      "epoch": 0.26008344923504867,
+      "grad_norm": 0.5278305411338806,
+      "learning_rate": 0.00016843559734370107,
+      "loss": 0.6424,
+      "step": 187
+    },
+    {
+      "epoch": 0.2614742698191933,
+      "grad_norm": 0.5382114052772522,
+      "learning_rate": 0.00016811635173373554,
+      "loss": 0.8486,
+      "step": 188
+    },
+    {
+      "epoch": 0.26286509040333794,
+      "grad_norm": 0.5917262434959412,
+      "learning_rate": 0.0001677958056767822,
+      "loss": 0.6542,
+      "step": 189
+    },
+    {
+      "epoch": 0.26425591098748263,
+      "grad_norm": 0.4891544282436371,
+      "learning_rate": 0.00016747396529256326,
+      "loss": 0.5733,
+      "step": 190
+    },
+    {
+      "epoch": 0.26564673157162727,
+      "grad_norm": 0.6395061016082764,
+      "learning_rate": 0.0001671508367255117,
+      "loss": 1.0908,
+      "step": 191
+    },
+    {
+      "epoch": 0.2670375521557719,
+      "grad_norm": 0.4592403471469879,
+      "learning_rate": 0.00016682642614465398,
+      "loss": 0.5655,
+      "step": 192
+    },
+    {
+      "epoch": 0.26842837273991654,
+      "grad_norm": 0.5632762312889099,
+      "learning_rate": 0.00016650073974349208,
+      "loss": 0.828,
+      "step": 193
+    },
+    {
+      "epoch": 0.2698191933240612,
+      "grad_norm": 0.5738239884376526,
+      "learning_rate": 0.00016617378373988545,
+      "loss": 0.7961,
+      "step": 194
+    },
+    {
+      "epoch": 0.2712100139082058,
+      "grad_norm": 0.4357210695743561,
+      "learning_rate": 0.00016584556437593213,
+      "loss": 0.7828,
+      "step": 195
+    },
+    {
+      "epoch": 0.2726008344923505,
+      "grad_norm": 0.44077688455581665,
+      "learning_rate": 0.00016551608791784978,
+      "loss": 0.5819,
+      "step": 196
+    },
+    {
+      "epoch": 0.27399165507649514,
+      "grad_norm": 0.43826064467430115,
+      "learning_rate": 0.0001651853606558559,
+      "loss": 0.4607,
+      "step": 197
+    },
+    {
+      "epoch": 0.2753824756606398,
+      "grad_norm": 0.9474967122077942,
+      "learning_rate": 0.0001648533889040478,
+      "loss": 1.0691,
+      "step": 198
+    },
+    {
+      "epoch": 0.2767732962447844,
+      "grad_norm": 0.8134686946868896,
+      "learning_rate": 0.000164520179000282,
+      "loss": 0.9829,
+      "step": 199
+    },
+    {
+      "epoch": 0.27816411682892905,
+      "grad_norm": 0.706295371055603,
+      "learning_rate": 0.00016418573730605322,
+      "loss": 1.012,
+      "step": 200
+    },
+    {
+      "epoch": 0.2795549374130737,
+      "grad_norm": 0.5357800126075745,
+      "learning_rate": 0.00016385007020637308,
+      "loss": 1.1607,
+      "step": 201
+    },
+    {
+      "epoch": 0.2809457579972184,
+      "grad_norm": 0.3396245539188385,
+      "learning_rate": 0.00016351318410964802,
+      "loss": 0.4552,
+      "step": 202
+    },
+    {
+      "epoch": 0.282336578581363,
+      "grad_norm": 0.7053083181381226,
+      "learning_rate": 0.00016317508544755707,
+      "loss": 1.0218,
+      "step": 203
+    },
+    {
+      "epoch": 0.28372739916550765,
+      "grad_norm": 0.6248435378074646,
+      "learning_rate": 0.000162835780674929,
+      "loss": 0.7015,
+      "step": 204
+    },
+    {
+      "epoch": 0.2851182197496523,
+      "grad_norm": 0.675771176815033,
+      "learning_rate": 0.00016249527626961907,
+      "loss": 0.8192,
+      "step": 205
+    },
+    {
+      "epoch": 0.2865090403337969,
+      "grad_norm": 0.5560746192932129,
+      "learning_rate": 0.00016215357873238548,
+      "loss": 0.9967,
+      "step": 206
+    },
+    {
+      "epoch": 0.28789986091794156,
+      "grad_norm": 0.7881655097007751,
+      "learning_rate": 0.00016181069458676505,
+      "loss": 0.7616,
+      "step": 207
+    },
+    {
+      "epoch": 0.28929068150208626,
+      "grad_norm": 0.4778852164745331,
+      "learning_rate": 0.00016146663037894889,
+      "loss": 0.856,
+      "step": 208
+    },
+    {
+      "epoch": 0.2906815020862309,
+      "grad_norm": 0.6938823461532593,
+      "learning_rate": 0.00016112139267765736,
+      "loss": 0.7036,
+      "step": 209
+    },
+    {
+      "epoch": 0.29207232267037553,
+      "grad_norm": 0.42614930868148804,
+      "learning_rate": 0.00016077498807401448,
+      "loss": 0.5915,
+      "step": 210
+    },
+    {
+      "epoch": 0.29346314325452016,
+      "grad_norm": 0.5251280665397644,
+      "learning_rate": 0.0001604274231814224,
+      "loss": 0.6153,
+      "step": 211
+    },
+    {
+      "epoch": 0.2948539638386648,
+      "grad_norm": 0.6218894124031067,
+      "learning_rate": 0.00016007870463543493,
+      "loss": 1.0625,
+      "step": 212
+    },
+    {
+      "epoch": 0.29624478442280944,
+      "grad_norm": 0.655918300151825,
+      "learning_rate": 0.00015972883909363094,
+      "loss": 0.8486,
+      "step": 213
+    },
+    {
+      "epoch": 0.29763560500695413,
+      "grad_norm": 0.3520936667919159,
+      "learning_rate": 0.00015937783323548713,
+      "loss": 0.5004,
+      "step": 214
+    },
+    {
+      "epoch": 0.29902642559109877,
+      "grad_norm": 0.5861802101135254,
+      "learning_rate": 0.00015902569376225083,
+      "loss": 0.7989,
+      "step": 215
+    },
+    {
+      "epoch": 0.3004172461752434,
+      "grad_norm": 0.6868994832038879,
+      "learning_rate": 0.0001586724273968116,
+      "loss": 0.7204,
+      "step": 216
+    },
+    {
+      "epoch": 0.30180806675938804,
+      "grad_norm": 0.5663580298423767,
+      "learning_rate": 0.00015831804088357326,
+      "loss": 0.8766,
+      "step": 217
+    },
+    {
+      "epoch": 0.3031988873435327,
+      "grad_norm": 0.4419863522052765,
+      "learning_rate": 0.00015796254098832498,
+      "loss": 0.6172,
+      "step": 218
+    },
+    {
+      "epoch": 0.3045897079276773,
+      "grad_norm": 0.5941941142082214,
+      "learning_rate": 0.0001576059344981121,
+      "loss": 0.7334,
+      "step": 219
+    },
+    {
+      "epoch": 0.30598052851182195,
+      "grad_norm": 0.4497317373752594,
+      "learning_rate": 0.00015724822822110656,
+      "loss": 0.7373,
+      "step": 220
+    },
+    {
+      "epoch": 0.30737134909596664,
+      "grad_norm": 0.8395616412162781,
+      "learning_rate": 0.000156889428986477,
+      "loss": 1.0541,
+      "step": 221
+    },
+    {
+      "epoch": 0.3087621696801113,
+      "grad_norm": 0.5366500020027161,
+      "learning_rate": 0.00015652954364425822,
+      "loss": 0.9061,
+      "step": 222
+    },
+    {
+      "epoch": 0.3101529902642559,
+      "grad_norm": 0.6337093710899353,
+      "learning_rate": 0.00015616857906522062,
+      "loss": 0.7976,
+      "step": 223
+    },
+    {
+      "epoch": 0.31154381084840055,
+      "grad_norm": 0.46693211793899536,
+      "learning_rate": 0.00015580654214073886,
+      "loss": 0.7367,
+      "step": 224
+    },
+    {
+      "epoch": 0.3129346314325452,
+      "grad_norm": 0.5937201380729675,
+      "learning_rate": 0.00015544343978266025,
+      "loss": 1.0663,
+      "step": 225
+    },
+    {
+      "epoch": 0.3143254520166898,
+      "grad_norm": 0.46815693378448486,
+      "learning_rate": 0.00015507927892317302,
+      "loss": 0.5882,
+      "step": 226
+    },
+    {
+      "epoch": 0.3157162726008345,
+      "grad_norm": 0.5517436861991882,
+      "learning_rate": 0.00015471406651467376,
+      "loss": 0.8323,
+      "step": 227
+    },
+    {
+      "epoch": 0.31710709318497915,
+      "grad_norm": 0.563980758190155,
+      "learning_rate": 0.0001543478095296348,
+      "loss": 0.6292,
+      "step": 228
+    },
+    {
+      "epoch": 0.3184979137691238,
+      "grad_norm": 0.3477098047733307,
+      "learning_rate": 0.00015398051496047106,
+      "loss": 0.3825,
+      "step": 229
+    },
+    {
+      "epoch": 0.3198887343532684,
+      "grad_norm": 0.6715020537376404,
+      "learning_rate": 0.00015361218981940647,
+      "loss": 0.812,
+      "step": 230
+    },
+    {
+      "epoch": 0.32127955493741306,
+      "grad_norm": 0.5200334787368774,
+      "learning_rate": 0.00015324284113834034,
+      "loss": 0.7529,
+      "step": 231
+    },
+    {
+      "epoch": 0.3226703755215577,
+      "grad_norm": 0.7095698714256287,
+      "learning_rate": 0.00015287247596871274,
+      "loss": 0.8984,
+      "step": 232
+    },
+    {
+      "epoch": 0.3240611961057024,
+      "grad_norm": 0.44531530141830444,
+      "learning_rate": 0.00015250110138137027,
+      "loss": 0.6572,
+      "step": 233
+    },
+    {
+      "epoch": 0.325452016689847,
+      "grad_norm": 0.4628504514694214,
+      "learning_rate": 0.0001521287244664308,
+      "loss": 0.5592,
+      "step": 234
+    },
+    {
+      "epoch": 0.32684283727399166,
+      "grad_norm": 0.7793769836425781,
+      "learning_rate": 0.00015175535233314823,
+      "loss": 0.8252,
+      "step": 235
+    },
+    {
+      "epoch": 0.3282336578581363,
+      "grad_norm": 0.37840721011161804,
+      "learning_rate": 0.00015138099210977663,
+      "loss": 0.4771,
+      "step": 236
+    },
+    {
+      "epoch": 0.32962447844228093,
+      "grad_norm": 0.7621236443519592,
+      "learning_rate": 0.00015100565094343436,
+      "loss": 0.8724,
+      "step": 237
+    },
+    {
+      "epoch": 0.33101529902642557,
+      "grad_norm": 0.4652254581451416,
+      "learning_rate": 0.00015062933599996746,
+      "loss": 0.5944,
+      "step": 238
+    },
+    {
+      "epoch": 0.33240611961057026,
+      "grad_norm": 0.4765583574771881,
+      "learning_rate": 0.00015025205446381285,
+      "loss": 0.6176,
+      "step": 239
+    },
+    {
+      "epoch": 0.3337969401947149,
+      "grad_norm": 0.6718812584877014,
+      "learning_rate": 0.0001498738135378613,
+      "loss": 0.9399,
+      "step": 240
+    },
+    {
+      "epoch": 0.33518776077885953,
+      "grad_norm": 0.597186803817749,
+      "learning_rate": 0.00014949462044331973,
+      "loss": 0.8488,
+      "step": 241
+    },
+    {
+      "epoch": 0.33657858136300417,
+      "grad_norm": 0.5403935313224792,
+      "learning_rate": 0.00014911448241957347,
+      "loss": 0.7961,
+      "step": 242
+    },
+    {
+      "epoch": 0.3379694019471488,
+      "grad_norm": 0.7336134314537048,
+      "learning_rate": 0.00014873340672404804,
+      "loss": 1.0693,
+      "step": 243
+    },
+    {
+      "epoch": 0.33936022253129344,
+      "grad_norm": 0.6288007497787476,
+      "learning_rate": 0.0001483514006320705,
+      "loss": 0.6699,
+      "step": 244
+    },
+    {
+      "epoch": 0.34075104311543813,
+      "grad_norm": 0.544596791267395,
+      "learning_rate": 0.0001479684714367307,
+      "loss": 0.6903,
+      "step": 245
+    },
+    {
+      "epoch": 0.34214186369958277,
+      "grad_norm": 0.612636387348175,
+      "learning_rate": 0.00014758462644874195,
+      "loss": 0.969,
+      "step": 246
+    },
+    {
+      "epoch": 0.3435326842837274,
+      "grad_norm": 0.5399485230445862,
+      "learning_rate": 0.00014719987299630148,
+      "loss": 0.9496,
+      "step": 247
+    },
+    {
+      "epoch": 0.34492350486787204,
+      "grad_norm": 0.5721740126609802,
+      "learning_rate": 0.00014681421842495046,
+      "loss": 0.8159,
+      "step": 248
+    },
+    {
+      "epoch": 0.3463143254520167,
+      "grad_norm": 0.3901589810848236,
+      "learning_rate": 0.00014642767009743384,
+      "loss": 0.55,
+      "step": 249
+    },
+    {
+      "epoch": 0.3477051460361613,
+      "grad_norm": 0.4009230136871338,
+      "learning_rate": 0.0001460402353935598,
+      "loss": 0.5454,
+      "step": 250
+    },
+    {
+      "epoch": 0.34909596662030595,
+      "grad_norm": 0.8864453434944153,
+      "learning_rate": 0.00014565192171005881,
+      "loss": 0.6913,
+      "step": 251
+    },
+    {
+      "epoch": 0.35048678720445064,
+      "grad_norm": 0.7382369637489319,
+      "learning_rate": 0.00014526273646044239,
+      "loss": 0.8861,
+      "step": 252
+    },
+    {
+      "epoch": 0.3518776077885953,
+      "grad_norm": 0.5884717702865601,
+      "learning_rate": 0.00014487268707486165,
+      "loss": 1.0722,
+      "step": 253
+    },
+    {
+      "epoch": 0.3532684283727399,
+      "grad_norm": 0.5429056882858276,
+      "learning_rate": 0.0001444817809999654,
+      "loss": 0.9678,
+      "step": 254
+    },
+    {
+      "epoch": 0.35465924895688455,
+      "grad_norm": 0.9008721113204956,
+      "learning_rate": 0.00014409002569875794,
+      "loss": 0.7893,
+      "step": 255
+    },
+    {
+      "epoch": 0.3560500695410292,
+      "grad_norm": 0.5287141799926758,
+      "learning_rate": 0.00014369742865045666,
+      "loss": 0.6202,
+      "step": 256
+    },
+    {
+      "epoch": 0.3574408901251738,
+      "grad_norm": 0.7288508415222168,
+      "learning_rate": 0.00014330399735034915,
+      "loss": 0.8188,
+      "step": 257
+    },
+    {
+      "epoch": 0.3588317107093185,
+      "grad_norm": 0.5704284906387329,
+      "learning_rate": 0.0001429097393096503,
+      "loss": 1.1271,
+      "step": 258
+    },
+    {
+      "epoch": 0.36022253129346316,
+      "grad_norm": 0.6670523881912231,
+      "learning_rate": 0.00014251466205535859,
+      "loss": 0.7093,
+      "step": 259
+    },
+    {
+      "epoch": 0.3616133518776078,
+      "grad_norm": 0.7255128622055054,
+      "learning_rate": 0.0001421187731301127,
+      "loss": 1.0761,
+      "step": 260
+    },
+    {
+      "epoch": 0.36300417246175243,
+      "grad_norm": 0.36424586176872253,
+      "learning_rate": 0.00014172208009204725,
+      "loss": 0.7407,
+      "step": 261
+    },
+    {
+      "epoch": 0.36439499304589706,
+      "grad_norm": 0.47512948513031006,
+      "learning_rate": 0.00014132459051464873,
+      "loss": 0.7557,
+      "step": 262
+    },
+    {
+      "epoch": 0.3657858136300417,
+      "grad_norm": 0.5641518235206604,
+      "learning_rate": 0.00014092631198661074,
+      "loss": 0.6554,
+      "step": 263
+    },
+    {
+      "epoch": 0.3671766342141864,
+      "grad_norm": 0.9931116104125977,
+      "learning_rate": 0.0001405272521116892,
+      "loss": 1.1471,
+      "step": 264
+    },
+    {
+      "epoch": 0.36856745479833103,
+      "grad_norm": 0.6660618185997009,
+      "learning_rate": 0.00014012741850855714,
+      "loss": 0.6843,
+      "step": 265
+    },
+    {
+      "epoch": 0.36995827538247567,
+      "grad_norm": 0.4106440544128418,
+      "learning_rate": 0.00013972681881065926,
+      "loss": 0.696,
+      "step": 266
+    },
+    {
+      "epoch": 0.3713490959666203,
+      "grad_norm": 0.4855511486530304,
+      "learning_rate": 0.0001393254606660662,
+      "loss": 0.5328,
+      "step": 267
+    },
+    {
+      "epoch": 0.37273991655076494,
+      "grad_norm": 0.47108009457588196,
+      "learning_rate": 0.0001389233517373286,
+      "loss": 0.7797,
+      "step": 268
+    },
+    {
+      "epoch": 0.3741307371349096,
+      "grad_norm": 0.49856990575790405,
+      "learning_rate": 0.00013852049970133058,
+      "loss": 0.5435,
+      "step": 269
+    },
+    {
+      "epoch": 0.37552155771905427,
+      "grad_norm": 0.46097061038017273,
+      "learning_rate": 0.00013811691224914347,
+      "loss": 0.552,
+      "step": 270
+    },
+    {
+      "epoch": 0.3769123783031989,
+      "grad_norm": 0.4893166422843933,
+      "learning_rate": 0.0001377125970858788,
+      "loss": 0.7556,
+      "step": 271
+    },
+    {
+      "epoch": 0.37830319888734354,
+      "grad_norm": 0.5556257367134094,
+      "learning_rate": 0.00013730756193054127,
+      "loss": 0.797,
+      "step": 272
+    },
+    {
+      "epoch": 0.3796940194714882,
+      "grad_norm": 0.5033267736434937,
+      "learning_rate": 0.0001369018145158812,
+      "loss": 0.91,
+      "step": 273
+    },
+    {
+      "epoch": 0.3810848400556328,
+      "grad_norm": 0.6278270483016968,
+      "learning_rate": 0.00013649536258824723,
+      "loss": 0.7328,
+      "step": 274
+    },
+    {
+      "epoch": 0.38247566063977745,
+      "grad_norm": 0.4785560369491577,
+      "learning_rate": 0.00013608821390743812,
+      "loss": 0.6043,
+      "step": 275
+    },
+    {
+      "epoch": 0.38386648122392214,
+      "grad_norm": 0.6780014038085938,
+      "learning_rate": 0.00013568037624655477,
+      "loss": 0.9633,
+      "step": 276
+    },
+    {
+      "epoch": 0.3852573018080668,
+      "grad_norm": 0.5871226191520691,
+      "learning_rate": 0.00013527185739185187,
+      "loss": 0.6637,
+      "step": 277
+    },
+    {
+      "epoch": 0.3866481223922114,
+      "grad_norm": 0.7345836162567139,
+      "learning_rate": 0.00013486266514258898,
+      "loss": 0.6675,
+      "step": 278
+    },
+    {
+      "epoch": 0.38803894297635605,
+      "grad_norm": 0.5815680027008057,
+      "learning_rate": 0.00013445280731088198,
+      "loss": 1.1779,
+      "step": 279
+    },
+    {
+      "epoch": 0.3894297635605007,
+      "grad_norm": 0.507443904876709,
+      "learning_rate": 0.00013404229172155364,
+      "loss": 1.1962,
+      "step": 280
+    },
+    {
+      "epoch": 0.3908205841446453,
+      "grad_norm": 0.4102143347263336,
+      "learning_rate": 0.0001336311262119844,
+      "loss": 0.4496,
+      "step": 281
+    },
+    {
+      "epoch": 0.39221140472878996,
+      "grad_norm": 0.7400050759315491,
+      "learning_rate": 0.00013321931863196268,
+      "loss": 0.7913,
+      "step": 282
+    },
+    {
+      "epoch": 0.39360222531293465,
+      "grad_norm": 0.5921124219894409,
+      "learning_rate": 0.00013280687684353494,
+      "loss": 0.8314,
+      "step": 283
+    },
+    {
+      "epoch": 0.3949930458970793,
+      "grad_norm": 0.4468141496181488,
+      "learning_rate": 0.00013239380872085584,
+      "loss": 0.6859,
+      "step": 284
+    },
+    {
+      "epoch": 0.3963838664812239,
+      "grad_norm": 0.5643270611763,
+      "learning_rate": 0.00013198012215003758,
+      "loss": 0.6837,
+      "step": 285
+    },
+    {
+      "epoch": 0.39777468706536856,
+      "grad_norm": 0.4446156620979309,
+      "learning_rate": 0.00013156582502899954,
+      "loss": 0.6196,
+      "step": 286
+    },
+    {
+      "epoch": 0.3991655076495132,
+      "grad_norm": 0.45944854617118835,
+      "learning_rate": 0.00013115092526731748,
+      "loss": 0.7509,
+      "step": 287
+    },
+    {
+      "epoch": 0.40055632823365783,
+      "grad_norm": 0.4472803771495819,
+      "learning_rate": 0.0001307354307860725,
+      "loss": 0.5207,
+      "step": 288
+    },
+    {
+      "epoch": 0.4019471488178025,
+      "grad_norm": 0.603515088558197,
+      "learning_rate": 0.00013031934951769975,
+      "loss": 0.6541,
+      "step": 289
+    },
+    {
+      "epoch": 0.40333796940194716,
+      "grad_norm": 0.43688806891441345,
+      "learning_rate": 0.00012990268940583715,
+      "loss": 0.6992,
+      "step": 290
+    },
+    {
+      "epoch": 0.4047287899860918,
+      "grad_norm": 0.6521260142326355,
+      "learning_rate": 0.0001294854584051736,
+      "loss": 0.9604,
+      "step": 291
+    },
+    {
+      "epoch": 0.40611961057023643,
+      "grad_norm": 0.8296985626220703,
+      "learning_rate": 0.00012906766448129718,
+      "loss": 1.2484,
+      "step": 292
+    },
+    {
+      "epoch": 0.40751043115438107,
+      "grad_norm": 0.680233359336853,
+      "learning_rate": 0.000128649315610543,
+      "loss": 0.8927,
+      "step": 293
+    },
+    {
+      "epoch": 0.4089012517385257,
+      "grad_norm": 0.6780505180358887,
+      "learning_rate": 0.000128230419779841,
+      "loss": 0.9671,
+      "step": 294
+    },
+    {
+      "epoch": 0.4102920723226704,
+      "grad_norm": 0.4301456809043884,
+      "learning_rate": 0.00012781098498656343,
+      "loss": 0.4716,
+      "step": 295
+    },
+    {
+      "epoch": 0.41168289290681503,
+      "grad_norm": 0.6730073094367981,
+      "learning_rate": 0.00012739101923837223,
+      "loss": 0.8887,
+      "step": 296
+    },
+    {
+      "epoch": 0.41307371349095967,
+      "grad_norm": 0.7534409761428833,
+      "learning_rate": 0.00012697053055306596,
+      "loss": 0.8234,
+      "step": 297
+    },
+    {
+      "epoch": 0.4144645340751043,
+      "grad_norm": 0.42848479747772217,
+      "learning_rate": 0.00012654952695842698,
+      "loss": 0.6855,
+      "step": 298
+    },
+    {
+      "epoch": 0.41585535465924894,
+      "grad_norm": 0.643622636795044,
+      "learning_rate": 0.000126128016492068,
+      "loss": 0.7707,
+      "step": 299
+    },
+    {
+      "epoch": 0.4172461752433936,
+      "grad_norm": 0.5563622117042542,
+      "learning_rate": 0.0001257060072012788,
+      "loss": 0.6389,
+      "step": 300
+    },
+    {
+      "epoch": 0.41863699582753827,
+      "grad_norm": 0.558152437210083,
+      "learning_rate": 0.0001252835071428723,
+      "loss": 0.8391,
+      "step": 301
+    },
+    {
+      "epoch": 0.4200278164116829,
+      "grad_norm": 0.49062344431877136,
+      "learning_rate": 0.00012486052438303124,
+      "loss": 0.7015,
+      "step": 302
+    },
+    {
+      "epoch": 0.42141863699582754,
+      "grad_norm": 0.5379894375801086,
+      "learning_rate": 0.00012443706699715354,
+      "loss": 0.7786,
+      "step": 303
+    },
+    {
+      "epoch": 0.4228094575799722,
+      "grad_norm": 0.5868341326713562,
+      "learning_rate": 0.0001240131430696987,
+      "loss": 0.7629,
+      "step": 304
+    },
+    {
+      "epoch": 0.4242002781641168,
+      "grad_norm": 0.41040000319480896,
+      "learning_rate": 0.00012358876069403312,
+      "loss": 0.6226,
+      "step": 305
+    },
+    {
+      "epoch": 0.42559109874826145,
+      "grad_norm": 0.6131793856620789,
+      "learning_rate": 0.0001231639279722757,
+      "loss": 0.9677,
+      "step": 306
+    },
+    {
+      "epoch": 0.42698191933240615,
+      "grad_norm": 0.692093551158905,
+      "learning_rate": 0.0001227386530151432,
+      "loss": 1.0101,
+      "step": 307
+    },
+    {
+      "epoch": 0.4283727399165508,
+      "grad_norm": 0.5966925024986267,
+      "learning_rate": 0.00012231294394179535,
+      "loss": 0.8283,
+      "step": 308
+    },
+    {
+      "epoch": 0.4297635605006954,
+      "grad_norm": 0.6743679046630859,
+      "learning_rate": 0.00012188680887967975,
+      "loss": 0.7774,
+      "step": 309
+    },
+    {
+      "epoch": 0.43115438108484005,
+      "grad_norm": 0.36354535818099976,
+      "learning_rate": 0.0001214602559643768,
+      "loss": 0.4107,
+      "step": 310
+    },
+    {
+      "epoch": 0.4325452016689847,
+      "grad_norm": 0.5071008801460266,
+      "learning_rate": 0.00012103329333944441,
+      "loss": 0.7568,
+      "step": 311
+    },
+    {
+      "epoch": 0.4339360222531293,
+      "grad_norm": 0.5053457617759705,
+      "learning_rate": 0.00012060592915626252,
+      "loss": 0.8495,
+      "step": 312
+    },
+    {
+      "epoch": 0.43532684283727396,
+      "grad_norm": 0.7441285848617554,
+      "learning_rate": 0.0001201781715738773,
+      "loss": 0.9994,
+      "step": 313
+    },
+    {
+      "epoch": 0.43671766342141866,
+      "grad_norm": 0.4609556198120117,
+      "learning_rate": 0.00011975002875884564,
+      "loss": 0.759,
+      "step": 314
+    },
+    {
+      "epoch": 0.4381084840055633,
+      "grad_norm": 0.47555771470069885,
+      "learning_rate": 0.00011932150888507911,
+      "loss": 0.7284,
+      "step": 315
+    },
+    {
+      "epoch": 0.43949930458970793,
+      "grad_norm": 0.5419169664382935,
+      "learning_rate": 0.00011889262013368792,
+      "loss": 0.5593,
+      "step": 316
+    },
+    {
+      "epoch": 0.44089012517385257,
+      "grad_norm": 0.6996181011199951,
+      "learning_rate": 0.00011846337069282463,
+      "loss": 0.8689,
+      "step": 317
+    },
+    {
+      "epoch": 0.4422809457579972,
+      "grad_norm": 0.4238322377204895,
+      "learning_rate": 0.00011803376875752807,
+      "loss": 0.5631,
+      "step": 318
+    },
+    {
+      "epoch": 0.44367176634214184,
+      "grad_norm": 0.8397965431213379,
+      "learning_rate": 0.00011760382252956663,
+      "loss": 0.7258,
+      "step": 319
+    },
+    {
+      "epoch": 0.44506258692628653,
+      "grad_norm": 0.6113640069961548,
+      "learning_rate": 0.0001171735402172818,
+      "loss": 0.5587,
+      "step": 320
+    },
+    {
+      "epoch": 0.44645340751043117,
+      "grad_norm": 0.5756425261497498,
+      "learning_rate": 0.00011674293003543148,
+      "loss": 0.6747,
+      "step": 321
+    },
+    {
+      "epoch": 0.4478442280945758,
+      "grad_norm": 0.7852854132652283,
+      "learning_rate": 0.00011631200020503307,
+      "loss": 1.0206,
+      "step": 322
+    },
+    {
+      "epoch": 0.44923504867872044,
+      "grad_norm": 0.39330264925956726,
+      "learning_rate": 0.00011588075895320663,
+      "loss": 0.5327,
+      "step": 323
+    },
+    {
+      "epoch": 0.4506258692628651,
+      "grad_norm": 0.5356118679046631,
+      "learning_rate": 0.00011544921451301758,
+      "loss": 0.845,
+      "step": 324
+    },
+    {
+      "epoch": 0.4520166898470097,
+      "grad_norm": 0.5333877801895142,
+      "learning_rate": 0.00011501737512331987,
+      "loss": 1.0495,
+      "step": 325
+    },
+    {
+      "epoch": 0.4534075104311544,
+      "grad_norm": 0.5317532420158386,
+      "learning_rate": 0.00011458524902859835,
+      "loss": 0.8038,
+      "step": 326
+    },
+    {
+      "epoch": 0.45479833101529904,
+      "grad_norm": 0.7273275256156921,
+      "learning_rate": 0.00011415284447881159,
+      "loss": 1.0297,
+      "step": 327
+    },
+    {
+      "epoch": 0.4561891515994437,
+      "grad_norm": 0.5057923197746277,
+      "learning_rate": 0.00011372016972923428,
+      "loss": 0.5413,
+      "step": 328
+    },
+    {
+      "epoch": 0.4575799721835883,
+      "grad_norm": 0.8137738108634949,
+      "learning_rate": 0.00011328723304029962,
+      "loss": 0.6639,
+      "step": 329
+    },
+    {
+      "epoch": 0.45897079276773295,
+      "grad_norm": 0.6706041693687439,
+      "learning_rate": 0.00011285404267744171,
+      "loss": 0.6631,
+      "step": 330
+    },
+    {
+      "epoch": 0.4603616133518776,
+      "grad_norm": 0.45623689889907837,
+      "learning_rate": 0.00011242060691093756,
+      "loss": 0.827,
+      "step": 331
+    },
+    {
+      "epoch": 0.4617524339360223,
+      "grad_norm": 0.3871178925037384,
+      "learning_rate": 0.00011198693401574945,
+      "loss": 0.3481,
+      "step": 332
+    },
+    {
+      "epoch": 0.4631432545201669,
+      "grad_norm": 0.5482321977615356,
+      "learning_rate": 0.00011155303227136671,
+      "loss": 0.626,
+      "step": 333
+    },
+    {
+      "epoch": 0.46453407510431155,
+      "grad_norm": 0.5279803276062012,
+      "learning_rate": 0.00011111890996164782,
+      "loss": 0.5369,
+      "step": 334
+    },
+    {
+      "epoch": 0.4659248956884562,
+      "grad_norm": 0.6390705704689026,
+      "learning_rate": 0.0001106845753746622,
+      "loss": 0.8136,
+      "step": 335
+    },
+    {
+      "epoch": 0.4673157162726008,
+      "grad_norm": 0.4990363121032715,
+      "learning_rate": 0.00011025003680253187,
+      "loss": 0.5547,
+      "step": 336
+    },
+    {
+      "epoch": 0.46870653685674546,
+      "grad_norm": 0.6139110326766968,
+      "learning_rate": 0.00010981530254127341,
+      "loss": 0.8627,
+      "step": 337
+    },
+    {
+      "epoch": 0.47009735744089015,
+      "grad_norm": 0.6849618554115295,
+      "learning_rate": 0.00010938038089063922,
+      "loss": 0.8369,
+      "step": 338
+    },
+    {
+      "epoch": 0.4714881780250348,
+      "grad_norm": 0.814551055431366,
+      "learning_rate": 0.00010894528015395946,
+      "loss": 0.7366,
+      "step": 339
+    },
+    {
+      "epoch": 0.4728789986091794,
+      "grad_norm": 0.5513364672660828,
+      "learning_rate": 0.00010851000863798313,
+      "loss": 0.6329,
+      "step": 340
+    },
+    {
+      "epoch": 0.47426981919332406,
+      "grad_norm": 1.1535307168960571,
+      "learning_rate": 0.00010807457465271984,
+      "loss": 0.9299,
+      "step": 341
+    },
+    {
+      "epoch": 0.4756606397774687,
+      "grad_norm": 0.5856449007987976,
+      "learning_rate": 0.00010763898651128084,
+      "loss": 0.7182,
+      "step": 342
+    },
+    {
+      "epoch": 0.47705146036161333,
+      "grad_norm": 0.38905879855155945,
+      "learning_rate": 0.00010720325252972052,
+      "loss": 0.4019,
+      "step": 343
+    },
+    {
+      "epoch": 0.47844228094575797,
+      "grad_norm": 0.7708204388618469,
+      "learning_rate": 0.00010676738102687767,
+      "loss": 1.2936,
+      "step": 344
+    },
+    {
+      "epoch": 0.47983310152990266,
+      "grad_norm": 0.7093687057495117,
+      "learning_rate": 0.00010633138032421638,
+      "loss": 0.6284,
+      "step": 345
+    },
+    {
+      "epoch": 0.4812239221140473,
+      "grad_norm": 0.6811620593070984,
+      "learning_rate": 0.00010589525874566761,
+      "loss": 0.8611,
+      "step": 346
+    },
+    {
+      "epoch": 0.48261474269819193,
+      "grad_norm": 0.48962119221687317,
+      "learning_rate": 0.00010545902461746984,
+      "loss": 0.3705,
+      "step": 347
+    },
+    {
+      "epoch": 0.48400556328233657,
+      "grad_norm": 0.445976585149765,
+      "learning_rate": 0.00010502268626801037,
+      "loss": 0.7893,
+      "step": 348
+    },
+    {
+      "epoch": 0.4853963838664812,
+      "grad_norm": 0.9196891188621521,
+      "learning_rate": 0.00010458625202766621,
+      "loss": 1.0964,
+      "step": 349
+    },
+    {
+      "epoch": 0.48678720445062584,
+      "grad_norm": 0.46996626257896423,
+      "learning_rate": 0.00010414973022864514,
+      "loss": 0.5653,
+      "step": 350
+    },
+    {
+      "epoch": 0.48817802503477054,
+      "grad_norm": 0.44221487641334534,
+      "learning_rate": 0.00010371312920482653,
+      "loss": 0.4808,
+      "step": 351
+    },
+    {
+      "epoch": 0.48956884561891517,
+      "grad_norm": 0.7322631478309631,
+      "learning_rate": 0.00010327645729160223,
+      "loss": 0.7351,
+      "step": 352
+    },
+    {
+      "epoch": 0.4909596662030598,
+      "grad_norm": 0.9222129583358765,
+      "learning_rate": 0.00010283972282571757,
+      "loss": 0.9798,
+      "step": 353
+    },
+    {
+      "epoch": 0.49235048678720444,
+      "grad_norm": 0.43294304609298706,
+      "learning_rate": 0.00010240293414511207,
+      "loss": 0.6612,
+      "step": 354
+    },
+    {
+      "epoch": 0.4937413073713491,
+      "grad_norm": 0.5410692095756531,
+      "learning_rate": 0.00010196609958876027,
+      "loss": 0.855,
+      "step": 355
+    },
+    {
+      "epoch": 0.4951321279554937,
+      "grad_norm": 0.49566787481307983,
+      "learning_rate": 0.00010152922749651252,
+      "loss": 0.5351,
+      "step": 356
+    },
+    {
+      "epoch": 0.4965229485396384,
+      "grad_norm": 0.6739534735679626,
+      "learning_rate": 0.00010109232620893592,
+      "loss": 0.8021,
+      "step": 357
+    },
+    {
+      "epoch": 0.49791376912378305,
+      "grad_norm": 0.5086596012115479,
+      "learning_rate": 0.00010065540406715483,
+      "loss": 0.8633,
+      "step": 358
+    },
+    {
+      "epoch": 0.4993045897079277,
+      "grad_norm": 0.6426077485084534,
+      "learning_rate": 0.00010021846941269179,
+      "loss": 0.7649,
+      "step": 359
+    },
+    {
+      "epoch": 0.5006954102920723,
+      "grad_norm": 0.6010740995407104,
+      "learning_rate": 9.978153058730823e-05,
+      "loss": 0.8781,
+      "step": 360
+    },
+    {
+      "epoch": 0.502086230876217,
+      "grad_norm": 0.3871017396450043,
+      "learning_rate": 9.934459593284517e-05,
+      "loss": 0.5443,
+      "step": 361
+    },
+    {
+      "epoch": 0.5034770514603616,
+      "grad_norm": 0.5949459075927734,
+      "learning_rate": 9.89076737910641e-05,
+      "loss": 1.0484,
+      "step": 362
+    },
+    {
+      "epoch": 0.5048678720445062,
+      "grad_norm": 0.6995226740837097,
+      "learning_rate": 9.847077250348749e-05,
+      "loss": 1.26,
+      "step": 363
+    },
+    {
+      "epoch": 0.5062586926286509,
+      "grad_norm": 0.5390698313713074,
+      "learning_rate": 9.803390041123979e-05,
+      "loss": 0.6685,
+      "step": 364
+    },
+    {
+      "epoch": 0.5076495132127955,
+      "grad_norm": 0.4320105314254761,
+      "learning_rate": 9.759706585488797e-05,
+      "loss": 0.6356,
+      "step": 365
+    },
+    {
+      "epoch": 0.5090403337969402,
+      "grad_norm": 0.46999648213386536,
+      "learning_rate": 9.716027717428243e-05,
+      "loss": 0.5946,
+      "step": 366
+    },
+    {
+      "epoch": 0.5104311543810849,
+      "grad_norm": 0.5290897488594055,
+      "learning_rate": 9.672354270839778e-05,
+      "loss": 0.8421,
+      "step": 367
+    },
+    {
+      "epoch": 0.5118219749652295,
+      "grad_norm": 0.600607693195343,
+      "learning_rate": 9.628687079517348e-05,
+      "loss": 0.7612,
+      "step": 368
+    },
+    {
+      "epoch": 0.5132127955493742,
+      "grad_norm": 0.764788031578064,
+      "learning_rate": 9.585026977135487e-05,
+      "loss": 0.9364,
+      "step": 369
+    },
+    {
+      "epoch": 0.5146036161335188,
+      "grad_norm": 0.5077704191207886,
+      "learning_rate": 9.541374797233381e-05,
+      "loss": 0.5467,
+      "step": 370
+    },
+    {
+      "epoch": 0.5159944367176634,
+      "grad_norm": 0.6263410449028015,
+      "learning_rate": 9.497731373198966e-05,
+      "loss": 0.8766,
+      "step": 371
+    },
+    {
+      "epoch": 0.5173852573018081,
+      "grad_norm": 0.965828001499176,
+      "learning_rate": 9.45409753825302e-05,
+      "loss": 0.8539,
+      "step": 372
+    },
+    {
+      "epoch": 0.5187760778859527,
+      "grad_norm": 0.46888992190361023,
+      "learning_rate": 9.41047412543324e-05,
+      "loss": 0.4867,
+      "step": 373
+    },
+    {
+      "epoch": 0.5201668984700973,
+      "grad_norm": 0.7452260255813599,
+      "learning_rate": 9.366861967578364e-05,
+      "loss": 1.0975,
+      "step": 374
+    },
+    {
+      "epoch": 0.521557719054242,
+      "grad_norm": 0.5537124872207642,
+      "learning_rate": 9.323261897312238e-05,
+      "loss": 1.0182,
+      "step": 375
+    },
+    {
+      "epoch": 0.5229485396383866,
+      "grad_norm": 0.5639439225196838,
+      "learning_rate": 9.279674747027952e-05,
+      "loss": 0.8048,
+      "step": 376
+    },
+    {
+      "epoch": 0.5243393602225312,
+      "grad_norm": 0.39898842573165894,
+      "learning_rate": 9.236101348871921e-05,
+      "loss": 0.5712,
+      "step": 377
+    },
+    {
+      "epoch": 0.5257301808066759,
+      "grad_norm": 0.673101544380188,
+      "learning_rate": 9.192542534728021e-05,
+      "loss": 0.9192,
+      "step": 378
+    },
+    {
+      "epoch": 0.5271210013908206,
+      "grad_norm": 0.6064724326133728,
+      "learning_rate": 9.148999136201686e-05,
+      "loss": 0.851,
+      "step": 379
+    },
+    {
+      "epoch": 0.5285118219749653,
+      "grad_norm": 0.5122058391571045,
+      "learning_rate": 9.105471984604055e-05,
+      "loss": 0.722,
+      "step": 380
+    },
+    {
+      "epoch": 0.5299026425591099,
+      "grad_norm": 0.7270992994308472,
+      "learning_rate": 9.061961910936078e-05,
+      "loss": 0.9341,
+      "step": 381
+    },
+    {
+      "epoch": 0.5312934631432545,
+      "grad_norm": 0.4852277338504791,
+      "learning_rate": 9.018469745872661e-05,
+      "loss": 0.5898,
+      "step": 382
+    },
+    {
+      "epoch": 0.5326842837273992,
+      "grad_norm": 0.39030471444129944,
+      "learning_rate": 8.974996319746814e-05,
+      "loss": 0.5184,
+      "step": 383
+    },
+    {
+      "epoch": 0.5340751043115438,
+      "grad_norm": 0.6964354515075684,
+      "learning_rate": 8.931542462533783e-05,
+      "loss": 0.8125,
+      "step": 384
+    },
+    {
+      "epoch": 0.5354659248956884,
+      "grad_norm": 0.6369216442108154,
+      "learning_rate": 8.88810900383522e-05,
+      "loss": 0.7592,
+      "step": 385
+    },
+    {
+      "epoch": 0.5368567454798331,
+      "grad_norm": 0.6099286675453186,
+      "learning_rate": 8.84469677286333e-05,
+      "loss": 1.0762,
+      "step": 386
+    },
+    {
+      "epoch": 0.5382475660639777,
+      "grad_norm": 0.6085535287857056,
+      "learning_rate": 8.801306598425059e-05,
+      "loss": 1.3508,
+      "step": 387
+    },
+    {
+      "epoch": 0.5396383866481224,
+      "grad_norm": 0.41708487272262573,
+      "learning_rate": 8.757939308906248e-05,
+      "loss": 0.4395,
+      "step": 388
+    },
+    {
+      "epoch": 0.541029207232267,
+      "grad_norm": 0.37615370750427246,
+      "learning_rate": 8.714595732255834e-05,
+      "loss": 0.5712,
+      "step": 389
+    },
+    {
+      "epoch": 0.5424200278164116,
+      "grad_norm": 0.7704113721847534,
+      "learning_rate": 8.671276695970043e-05,
+      "loss": 0.9996,
+      "step": 390
+    },
+    {
+      "epoch": 0.5438108484005564,
+      "grad_norm": 0.49218448996543884,
+      "learning_rate": 8.627983027076573e-05,
+      "loss": 0.562,
+      "step": 391
+    },
+    {
+      "epoch": 0.545201668984701,
+      "grad_norm": 0.5166805386543274,
+      "learning_rate": 8.584715552118842e-05,
+      "loss": 0.7108,
+      "step": 392
+    },
+    {
+      "epoch": 0.5465924895688457,
+      "grad_norm": 0.5317919850349426,
+      "learning_rate": 8.541475097140167e-05,
+      "loss": 0.7533,
+      "step": 393
+    },
+    {
+      "epoch": 0.5479833101529903,
+      "grad_norm": 0.4382265508174896,
+      "learning_rate": 8.498262487668015e-05,
+      "loss": 0.5953,
+      "step": 394
+    },
+    {
+      "epoch": 0.5493741307371349,
+      "grad_norm": 0.5257577300071716,
+      "learning_rate": 8.455078548698243e-05,
+      "loss": 0.7693,
+      "step": 395
+    },
+    {
+      "epoch": 0.5507649513212796,
+      "grad_norm": 0.9228342771530151,
+      "learning_rate": 8.411924104679339e-05,
+      "loss": 0.9168,
+      "step": 396
+    },
+    {
+      "epoch": 0.5521557719054242,
+      "grad_norm": 0.9047560095787048,
+      "learning_rate": 8.368799979496694e-05,
+      "loss": 1.2655,
+      "step": 397
+    },
+    {
+      "epoch": 0.5535465924895688,
+      "grad_norm": 0.5051538944244385,
+      "learning_rate": 8.325706996456854e-05,
+      "loss": 0.4851,
+      "step": 398
+    },
+    {
+      "epoch": 0.5549374130737135,
+      "grad_norm": 0.3420053720474243,
+      "learning_rate": 8.282645978271823e-05,
+      "loss": 0.3682,
+      "step": 399
+    },
+    {
+      "epoch": 0.5563282336578581,
+      "grad_norm": 0.7413134574890137,
+      "learning_rate": 8.23961774704334e-05,
+      "loss": 1.0863,
+      "step": 400
+    },
+    {
+      "epoch": 0.5577190542420027,
+      "grad_norm": 0.6543962359428406,
+      "learning_rate": 8.196623124247197e-05,
+      "loss": 0.9683,
+      "step": 401
+    },
+    {
+      "epoch": 0.5591098748261474,
+      "grad_norm": 0.7017605900764465,
+      "learning_rate": 8.153662930717538e-05,
+      "loss": 1.0335,
+      "step": 402
+    },
+    {
+      "epoch": 0.5605006954102921,
+      "grad_norm": 0.5047004222869873,
+      "learning_rate": 8.110737986631209e-05,
+      "loss": 0.4696,
+      "step": 403
+    },
+    {
+      "epoch": 0.5618915159944368,
+      "grad_norm": 0.4881282150745392,
+      "learning_rate": 8.06784911149209e-05,
+      "loss": 0.5434,
+      "step": 404
+    },
+    {
+      "epoch": 0.5632823365785814,
+      "grad_norm": 0.8078427910804749,
+      "learning_rate": 8.024997124115437e-05,
+      "loss": 0.8981,
+      "step": 405
+    },
+    {
+      "epoch": 0.564673157162726,
+      "grad_norm": 0.5463523864746094,
+      "learning_rate": 7.982182842612274e-05,
+      "loss": 0.6066,
+      "step": 406
+    },
+    {
+      "epoch": 0.5660639777468707,
+      "grad_norm": 0.8765882253646851,
+      "learning_rate": 7.939407084373751e-05,
+      "loss": 0.7475,
+      "step": 407
+    },
+    {
+      "epoch": 0.5674547983310153,
+      "grad_norm": 0.5860880613327026,
+      "learning_rate": 7.89667066605556e-05,
+      "loss": 0.7044,
+      "step": 408
+    },
+    {
+      "epoch": 0.56884561891516,
+      "grad_norm": 0.6440436840057373,
+      "learning_rate": 7.853974403562323e-05,
+      "loss": 0.8198,
+      "step": 409
+    },
+    {
+      "epoch": 0.5702364394993046,
+      "grad_norm": 0.5287847518920898,
+      "learning_rate": 7.811319112032027e-05,
+      "loss": 0.6231,
+      "step": 410
+    },
+    {
+      "epoch": 0.5716272600834492,
+      "grad_norm": 0.5613576769828796,
+      "learning_rate": 7.768705605820466e-05,
+      "loss": 1.1385,
+      "step": 411
+    },
+    {
+      "epoch": 0.5730180806675939,
+      "grad_norm": 0.5475975871086121,
+      "learning_rate": 7.726134698485679e-05,
+      "loss": 0.8026,
+      "step": 412
+    },
+    {
+      "epoch": 0.5744089012517385,
+      "grad_norm": 0.5117793083190918,
+      "learning_rate": 7.683607202772433e-05,
+      "loss": 0.5665,
+      "step": 413
+    },
+    {
+      "epoch": 0.5757997218358831,
+      "grad_norm": 0.5135944485664368,
+      "learning_rate": 7.641123930596693e-05,
+      "loss": 0.6325,
+      "step": 414
+    },
+    {
+      "epoch": 0.5771905424200278,
+      "grad_norm": 0.7722037434577942,
+      "learning_rate": 7.598685693030136e-05,
+      "loss": 1.0663,
+      "step": 415
+    },
+    {
+      "epoch": 0.5785813630041725,
+      "grad_norm": 0.5279682278633118,
+      "learning_rate": 7.556293300284647e-05,
+      "loss": 0.851,
+      "step": 416
+    },
+    {
+      "epoch": 0.5799721835883171,
+      "grad_norm": 0.6835013031959534,
+      "learning_rate": 7.513947561696877e-05,
+      "loss": 0.6762,
+      "step": 417
+    },
+    {
+      "epoch": 0.5813630041724618,
+      "grad_norm": 0.4907541573047638,
+      "learning_rate": 7.471649285712768e-05,
+      "loss": 0.6388,
+      "step": 418
+    },
+    {
+      "epoch": 0.5827538247566064,
+      "grad_norm": 0.5398621559143066,
+      "learning_rate": 7.429399279872122e-05,
+      "loss": 0.937,
+      "step": 419
+    },
+    {
+      "epoch": 0.5841446453407511,
+      "grad_norm": 0.7209218144416809,
+      "learning_rate": 7.387198350793201e-05,
+      "loss": 0.9247,
+      "step": 420
+    },
+    {
+      "epoch": 0.5855354659248957,
+      "grad_norm": 0.687431812286377,
+      "learning_rate": 7.345047304157304e-05,
+      "loss": 0.7067,
+      "step": 421
+    },
+    {
+      "epoch": 0.5869262865090403,
+      "grad_norm": 0.37900230288505554,
+      "learning_rate": 7.302946944693408e-05,
+      "loss": 0.447,
+      "step": 422
+    },
+    {
+      "epoch": 0.588317107093185,
+      "grad_norm": 0.48476675152778625,
+      "learning_rate": 7.260898076162782e-05,
+      "loss": 0.8235,
+      "step": 423
+    },
+    {
+      "epoch": 0.5897079276773296,
+      "grad_norm": 0.4247264266014099,
+      "learning_rate": 7.218901501343658e-05,
+      "loss": 0.5045,
+      "step": 424
+    },
+    {
+      "epoch": 0.5910987482614742,
+      "grad_norm": 0.8774217963218689,
+      "learning_rate": 7.176958022015902e-05,
+      "loss": 0.722,
+      "step": 425
+    },
+    {
+      "epoch": 0.5924895688456189,
+      "grad_norm": 0.7176122665405273,
+      "learning_rate": 7.135068438945704e-05,
+      "loss": 0.7989,
+      "step": 426
+    },
+    {
+      "epoch": 0.5938803894297635,
+      "grad_norm": 0.7195794582366943,
+      "learning_rate": 7.093233551870287e-05,
+      "loss": 0.9021,
+      "step": 427
+    },
+    {
+      "epoch": 0.5952712100139083,
+      "grad_norm": 0.6383934617042542,
+      "learning_rate": 7.05145415948264e-05,
+      "loss": 0.8931,
+      "step": 428
+    },
+    {
+      "epoch": 0.5966620305980529,
+      "grad_norm": 0.5574417114257812,
+      "learning_rate": 7.009731059416287e-05,
+      "loss": 0.6444,
+      "step": 429
+    },
+    {
+      "epoch": 0.5980528511821975,
+      "grad_norm": 0.5440465211868286,
+      "learning_rate": 6.968065048230028e-05,
+      "loss": 0.4592,
+      "step": 430
+    },
+    {
+      "epoch": 0.5994436717663422,
+      "grad_norm": 0.609474241733551,
+      "learning_rate": 6.926456921392753e-05,
+      "loss": 0.8622,
+      "step": 431
+    },
+    {
+      "epoch": 0.6008344923504868,
+      "grad_norm": 0.5016144514083862,
+      "learning_rate": 6.884907473268254e-05,
+      "loss": 0.7558,
+      "step": 432
+    },
+    {
+      "epoch": 0.6022253129346314,
+      "grad_norm": 0.5324974656105042,
+      "learning_rate": 6.843417497100046e-05,
+      "loss": 0.573,
+      "step": 433
+    },
+    {
+      "epoch": 0.6036161335187761,
+      "grad_norm": 0.5987928509712219,
+      "learning_rate": 6.801987784996244e-05,
+      "loss": 0.7625,
+      "step": 434
+    },
+    {
+      "epoch": 0.6050069541029207,
+      "grad_norm": 0.481094628572464,
+      "learning_rate": 6.760619127914417e-05,
+      "loss": 0.6608,
+      "step": 435
+    },
+    {
+      "epoch": 0.6063977746870653,
+      "grad_norm": 0.5764672756195068,
+      "learning_rate": 6.719312315646507e-05,
+      "loss": 0.4863,
+      "step": 436
+    },
+    {
+      "epoch": 0.60778859527121,
+      "grad_norm": 0.3798976540565491,
+      "learning_rate": 6.678068136803736e-05,
+      "loss": 0.5737,
+      "step": 437
+    },
+    {
+      "epoch": 0.6091794158553546,
+      "grad_norm": 0.6764469146728516,
+      "learning_rate": 6.636887378801563e-05,
+      "loss": 0.985,
+      "step": 438
+    },
+    {
+      "epoch": 0.6105702364394993,
+      "grad_norm": 0.567531943321228,
+      "learning_rate": 6.595770827844637e-05,
+      "loss": 0.6424,
+      "step": 439
+    },
+    {
+      "epoch": 0.6119610570236439,
+      "grad_norm": 0.3781545162200928,
+      "learning_rate": 6.554719268911804e-05,
+      "loss": 0.5021,
+      "step": 440
+    },
+    {
+      "epoch": 0.6133518776077886,
+      "grad_norm": 0.3420570492744446,
+      "learning_rate": 6.513733485741102e-05,
+      "loss": 0.6499,
+      "step": 441
+    },
+    {
+      "epoch": 0.6147426981919333,
+      "grad_norm": 0.4896042048931122,
+      "learning_rate": 6.472814260814814e-05,
+      "loss": 0.7439,
+      "step": 442
+    },
+    {
+      "epoch": 0.6161335187760779,
+      "grad_norm": 0.6409856081008911,
+      "learning_rate": 6.431962375344522e-05,
+      "loss": 0.6023,
+      "step": 443
+    },
+    {
+      "epoch": 0.6175243393602226,
+      "grad_norm": 0.7267551422119141,
+      "learning_rate": 6.39117860925619e-05,
+      "loss": 0.8636,
+      "step": 444
+    },
+    {
+      "epoch": 0.6189151599443672,
+      "grad_norm": 0.5234163403511047,
+      "learning_rate": 6.350463741175281e-05,
+      "loss": 0.8441,
+      "step": 445
+    },
+    {
+      "epoch": 0.6203059805285118,
+      "grad_norm": 0.6535494923591614,
+      "learning_rate": 6.309818548411881e-05,
+      "loss": 0.6902,
+      "step": 446
+    },
+    {
+      "epoch": 0.6216968011126565,
+      "grad_norm": 0.6372959613800049,
+      "learning_rate": 6.269243806945878e-05,
+      "loss": 1.0858,
+      "step": 447
+    },
+    {
+      "epoch": 0.6230876216968011,
+      "grad_norm": 0.6454161405563354,
+      "learning_rate": 6.228740291412121e-05,
+      "loss": 0.6181,
+      "step": 448
+    },
+    {
+      "epoch": 0.6244784422809457,
+      "grad_norm": 0.7606287002563477,
+      "learning_rate": 6.188308775085654e-05,
+      "loss": 0.9042,
+      "step": 449
+    },
+    {
+      "epoch": 0.6258692628650904,
+      "grad_norm": 0.46747350692749023,
+      "learning_rate": 6.147950029866946e-05,
+      "loss": 0.5641,
+      "step": 450
+    },
+    {
+      "epoch": 0.627260083449235,
+      "grad_norm": 0.4410058259963989,
+      "learning_rate": 6.107664826267146e-05,
+      "loss": 0.656,
+      "step": 451
+    },
+    {
+      "epoch": 0.6286509040333796,
+      "grad_norm": 0.3773995041847229,
+      "learning_rate": 6.067453933393383e-05,
+      "loss": 0.5439,
+      "step": 452
+    },
+    {
+      "epoch": 0.6300417246175244,
+      "grad_norm": 0.8339559435844421,
+      "learning_rate": 6.027318118934078e-05,
+      "loss": 1.0035,
+      "step": 453
+    },
+    {
+      "epoch": 0.631432545201669,
+      "grad_norm": 0.500758707523346,
+      "learning_rate": 5.987258149144288e-05,
+      "loss": 0.6246,
+      "step": 454
+    },
+    {
+      "epoch": 0.6328233657858137,
+      "grad_norm": 0.6480860114097595,
+      "learning_rate": 5.9472747888310834e-05,
+      "loss": 0.6126,
+      "step": 455
+    },
+    {
+      "epoch": 0.6342141863699583,
+      "grad_norm": 0.5420281291007996,
+      "learning_rate": 5.907368801338927e-05,
+      "loss": 0.569,
+      "step": 456
+    },
+    {
+      "epoch": 0.6356050069541029,
+      "grad_norm": 0.7778435945510864,
+      "learning_rate": 5.86754094853513e-05,
+      "loss": 0.9918,
+      "step": 457
+    },
+    {
+      "epoch": 0.6369958275382476,
+      "grad_norm": 0.6609676480293274,
+      "learning_rate": 5.827791990795278e-05,
+      "loss": 0.6712,
+      "step": 458
+    },
+    {
+      "epoch": 0.6383866481223922,
+      "grad_norm": 0.5104408860206604,
+      "learning_rate": 5.7881226869887314e-05,
+      "loss": 0.6318,
+      "step": 459
+    },
+    {
+      "epoch": 0.6397774687065368,
+      "grad_norm": 0.49637484550476074,
+      "learning_rate": 5.748533794464142e-05,
+      "loss": 0.7334,
+      "step": 460
+    },
+    {
+      "epoch": 0.6411682892906815,
+      "grad_norm": 0.7833724617958069,
+      "learning_rate": 5.709026069034973e-05,
+      "loss": 1.0324,
+      "step": 461
+    },
+    {
+      "epoch": 0.6425591098748261,
+      "grad_norm": 0.481446772813797,
+      "learning_rate": 5.669600264965088e-05,
+      "loss": 0.5633,
+      "step": 462
+    },
+    {
+      "epoch": 0.6439499304589708,
+      "grad_norm": 0.723691999912262,
+      "learning_rate": 5.630257134954338e-05,
+      "loss": 0.9101,
+      "step": 463
+    },
+    {
+      "epoch": 0.6453407510431154,
+      "grad_norm": 0.525189220905304,
+      "learning_rate": 5.59099743012421e-05,
+      "loss": 0.5756,
+      "step": 464
+    },
+    {
+      "epoch": 0.6467315716272601,
+      "grad_norm": 0.5991864204406738,
+      "learning_rate": 5.551821900003461e-05,
+      "loss": 0.8835,
+      "step": 465
+    },
+    {
+      "epoch": 0.6481223922114048,
+      "grad_norm": 0.39758554100990295,
+      "learning_rate": 5.512731292513835e-05,
+      "loss": 0.4273,
+      "step": 466
+    },
+    {
+      "epoch": 0.6495132127955494,
+      "grad_norm": 0.6345305442810059,
+      "learning_rate": 5.473726353955764e-05,
+      "loss": 0.8338,
+      "step": 467
+    },
+    {
+      "epoch": 0.650904033379694,
+      "grad_norm": 0.5823615789413452,
+      "learning_rate": 5.434807828994123e-05,
+      "loss": 1.0666,
+      "step": 468
+    },
+    {
+      "epoch": 0.6522948539638387,
+      "grad_norm": 0.8367906808853149,
+      "learning_rate": 5.39597646064402e-05,
+      "loss": 1.1912,
+      "step": 469
+    },
+    {
+      "epoch": 0.6536856745479833,
+      "grad_norm": 0.46055009961128235,
+      "learning_rate": 5.357232990256618e-05,
+      "loss": 0.5111,
+      "step": 470
+    },
+    {
+      "epoch": 0.655076495132128,
+      "grad_norm": 0.35955506563186646,
+      "learning_rate": 5.318578157504959e-05,
+      "loss": 0.6477,
+      "step": 471
+    },
+    {
+      "epoch": 0.6564673157162726,
+      "grad_norm": 0.6589727401733398,
+      "learning_rate": 5.280012700369856e-05,
+      "loss": 0.8803,
+      "step": 472
+    },
+    {
+      "epoch": 0.6578581363004172,
+      "grad_norm": 0.8205442428588867,
+      "learning_rate": 5.2415373551258054e-05,
+      "loss": 0.955,
+      "step": 473
+    },
+    {
+      "epoch": 0.6592489568845619,
+      "grad_norm": 0.7387497425079346,
+      "learning_rate": 5.203152856326933e-05,
+      "loss": 1.1923,
+      "step": 474
+    },
+    {
+      "epoch": 0.6606397774687065,
+      "grad_norm": 0.47400394082069397,
+      "learning_rate": 5.164859936792955e-05,
+      "loss": 0.49,
+      "step": 475
+    },
+    {
+      "epoch": 0.6620305980528511,
+      "grad_norm": 0.6762047410011292,
+      "learning_rate": 5.1266593275952034e-05,
+      "loss": 1.0296,
+      "step": 476
+    },
+    {
+      "epoch": 0.6634214186369958,
+      "grad_norm": 0.5361552834510803,
+      "learning_rate": 5.088551758042655e-05,
+      "loss": 0.7153,
+      "step": 477
+    },
+    {
+      "epoch": 0.6648122392211405,
+      "grad_norm": 0.6544700264930725,
+      "learning_rate": 5.050537955668027e-05,
+      "loss": 1.1354,
+      "step": 478
+    },
+    {
+      "epoch": 0.6662030598052852,
+      "grad_norm": 0.6211574673652649,
+      "learning_rate": 5.012618646213869e-05,
+      "loss": 1.3023,
+      "step": 479
+    },
+    {
+      "epoch": 0.6675938803894298,
+      "grad_norm": 0.5443877577781677,
+      "learning_rate": 4.9747945536187145e-05,
+      "loss": 0.763,
+      "step": 480
+    },
+    {
+      "epoch": 0.6689847009735744,
+      "grad_norm": 0.5580876469612122,
+      "learning_rate": 4.937066400003256e-05,
+      "loss": 1.1087,
+      "step": 481
+    },
+    {
+      "epoch": 0.6703755215577191,
+      "grad_norm": 0.5902083516120911,
+      "learning_rate": 4.899434905656563e-05,
+      "loss": 0.8052,
+      "step": 482
+    },
+    {
+      "epoch": 0.6717663421418637,
+      "grad_norm": 0.49830299615859985,
+      "learning_rate": 4.861900789022339e-05,
+      "loss": 0.6586,
+      "step": 483
+    },
+    {
+      "epoch": 0.6731571627260083,
+      "grad_norm": 0.5979222059249878,
+      "learning_rate": 4.824464766685182e-05,
+      "loss": 1.0585,
+      "step": 484
+    },
+    {
+      "epoch": 0.674547983310153,
+      "grad_norm": 0.3839762806892395,
+      "learning_rate": 4.7871275533569223e-05,
+      "loss": 0.645,
+      "step": 485
+    },
+    {
+      "epoch": 0.6759388038942976,
+      "grad_norm": 0.6916873455047607,
+      "learning_rate": 4.749889861862977e-05,
+      "loss": 0.7701,
+      "step": 486
+    },
+    {
+      "epoch": 0.6773296244784422,
+      "grad_norm": 0.46500733494758606,
+      "learning_rate": 4.712752403128729e-05,
+      "loss": 0.6347,
+      "step": 487
+    },
+    {
+      "epoch": 0.6787204450625869,
+      "grad_norm": 0.5402698516845703,
+      "learning_rate": 4.6757158861659714e-05,
+      "loss": 1.0546,
+      "step": 488
+    },
+    {
+      "epoch": 0.6801112656467315,
+      "grad_norm": 0.5720083117485046,
+      "learning_rate": 4.638781018059356e-05,
+      "loss": 0.9169,
+      "step": 489
+    },
+    {
+      "epoch": 0.6815020862308763,
+      "grad_norm": 0.5751729011535645,
+      "learning_rate": 4.601948503952896e-05,
+      "loss": 0.5225,
+      "step": 490
+    },
+    {
+      "epoch": 0.6828929068150209,
+      "grad_norm": 0.40184223651885986,
+      "learning_rate": 4.565219047036521e-05,
+      "loss": 0.6651,
+      "step": 491
+    },
+    {
+      "epoch": 0.6842837273991655,
+      "grad_norm": 0.6558845043182373,
+      "learning_rate": 4.5285933485326226e-05,
+      "loss": 1.2829,
+      "step": 492
+    },
+    {
+      "epoch": 0.6856745479833102,
+      "grad_norm": 0.520807683467865,
+      "learning_rate": 4.492072107682699e-05,
+      "loss": 0.8412,
+      "step": 493
+    },
+    {
+      "epoch": 0.6870653685674548,
+      "grad_norm": 0.4416840970516205,
+      "learning_rate": 4.4556560217339774e-05,
+      "loss": 0.5507,
+      "step": 494
+    },
+    {
+      "epoch": 0.6884561891515995,
+      "grad_norm": 0.4645073115825653,
+      "learning_rate": 4.419345785926119e-05,
+      "loss": 0.7746,
+      "step": 495
+    },
+    {
+      "epoch": 0.6898470097357441,
+      "grad_norm": 0.6210498809814453,
+      "learning_rate": 4.383142093477938e-05,
+      "loss": 0.6232,
+      "step": 496
+    },
+    {
+      "epoch": 0.6912378303198887,
+      "grad_norm": 0.46305474638938904,
+      "learning_rate": 4.3470456355741786e-05,
+      "loss": 0.446,
+      "step": 497
+    },
+    {
+      "epoch": 0.6926286509040334,
+      "grad_norm": 0.6622077822685242,
+      "learning_rate": 4.311057101352304e-05,
+      "loss": 1.1716,
+      "step": 498
+    },
+    {
+      "epoch": 0.694019471488178,
+      "grad_norm": 0.5017058253288269,
+      "learning_rate": 4.275177177889347e-05,
+      "loss": 0.7702,
+      "step": 499
+    },
+    {
+      "epoch": 0.6954102920723226,
+      "grad_norm": 0.4082466661930084,
+      "learning_rate": 4.239406550188791e-05,
+      "loss": 0.6416,
+      "step": 500
+    },
+    {
+      "epoch": 0.6968011126564673,
+      "grad_norm": 0.7320986986160278,
+      "learning_rate": 4.203745901167504e-05,
+      "loss": 0.7718,
+      "step": 501
+    },
+    {
+      "epoch": 0.6981919332406119,
+      "grad_norm": 0.5550965070724487,
+      "learning_rate": 4.168195911642677e-05,
+      "loss": 1.1353,
+      "step": 502
+    },
+    {
+      "epoch": 0.6995827538247567,
+      "grad_norm": 0.6742632389068604,
+      "learning_rate": 4.1327572603188415e-05,
+      "loss": 1.2761,
+      "step": 503
+    },
+    {
+      "epoch": 0.7009735744089013,
+      "grad_norm": 0.4826456606388092,
+      "learning_rate": 4.097430623774921e-05,
+      "loss": 0.5789,
+      "step": 504
+    },
+    {
+      "epoch": 0.7023643949930459,
+      "grad_norm": 0.748701810836792,
+      "learning_rate": 4.062216676451285e-05,
+      "loss": 0.9943,
+      "step": 505
+    },
+    {
+      "epoch": 0.7037552155771906,
+      "grad_norm": 0.5128481984138489,
+      "learning_rate": 4.027116090636909e-05,
+      "loss": 0.8659,
+      "step": 506
+    },
+    {
+      "epoch": 0.7051460361613352,
+      "grad_norm": 0.519477128982544,
+      "learning_rate": 3.9921295364565094e-05,
+      "loss": 1.135,
+      "step": 507
+    },
+    {
+      "epoch": 0.7065368567454798,
+      "grad_norm": 0.5690450072288513,
+      "learning_rate": 3.9572576818577637e-05,
+      "loss": 0.7553,
+      "step": 508
+    },
+    {
+      "epoch": 0.7079276773296245,
+      "grad_norm": 0.8124468326568604,
+      "learning_rate": 3.922501192598557e-05,
+      "loss": 1.0765,
+      "step": 509
+    },
+    {
+      "epoch": 0.7093184979137691,
+      "grad_norm": 0.38741016387939453,
+      "learning_rate": 3.8878607322342674e-05,
+      "loss": 0.4048,
+      "step": 510
+    },
+    {
+      "epoch": 0.7107093184979137,
+      "grad_norm": 0.6302335262298584,
+      "learning_rate": 3.853336962105112e-05,
+      "loss": 0.8698,
+      "step": 511
+    },
+    {
+      "epoch": 0.7121001390820584,
+      "grad_norm": 0.6914311647415161,
+      "learning_rate": 3.8189305413234975e-05,
+      "loss": 0.9812,
+      "step": 512
+    },
+    {
+      "epoch": 0.713490959666203,
+      "grad_norm": 0.41893160343170166,
+      "learning_rate": 3.784642126761457e-05,
+      "loss": 0.6756,
+      "step": 513
+    },
+    {
+      "epoch": 0.7148817802503477,
+      "grad_norm": 0.3986198604106903,
+      "learning_rate": 3.750472373038093e-05,
+      "loss": 0.4829,
+      "step": 514
+    },
+    {
+      "epoch": 0.7162726008344924,
+      "grad_norm": 0.4317716658115387,
+      "learning_rate": 3.7164219325070995e-05,
+      "loss": 0.5276,
+      "step": 515
+    },
+    {
+      "epoch": 0.717663421418637,
+      "grad_norm": 0.4723030924797058,
+      "learning_rate": 3.682491455244294e-05,
+      "loss": 0.7898,
+      "step": 516
+    },
+    {
+      "epoch": 0.7190542420027817,
+      "grad_norm": 0.5028923749923706,
+      "learning_rate": 3.6486815890352e-05,
+      "loss": 0.8993,
+      "step": 517
+    },
+    {
+      "epoch": 0.7204450625869263,
+      "grad_norm": 0.5633566975593567,
+      "learning_rate": 3.6149929793626954e-05,
+      "loss": 0.7664,
+      "step": 518
+    },
+    {
+      "epoch": 0.721835883171071,
+      "grad_norm": 0.7815782427787781,
+      "learning_rate": 3.581426269394679e-05,
+      "loss": 0.8137,
+      "step": 519
+    },
+    {
+      "epoch": 0.7232267037552156,
+      "grad_norm": 0.4387079179286957,
+      "learning_rate": 3.5479820999718036e-05,
+      "loss": 0.5675,
+      "step": 520
+    },
+    {
+      "epoch": 0.7246175243393602,
+      "grad_norm": 0.6255477666854858,
+      "learning_rate": 3.514661109595222e-05,
+      "loss": 0.9465,
+      "step": 521
+    },
+    {
+      "epoch": 0.7260083449235049,
+      "grad_norm": 0.6314586997032166,
+      "learning_rate": 3.481463934414412e-05,
+      "loss": 0.9266,
+      "step": 522
+    },
+    {
+      "epoch": 0.7273991655076495,
+      "grad_norm": 0.6725448369979858,
+      "learning_rate": 3.4483912082150225e-05,
+      "loss": 0.7243,
+      "step": 523
+    },
+    {
+      "epoch": 0.7287899860917941,
+      "grad_norm": 0.5544277429580688,
+      "learning_rate": 3.4154435624067904e-05,
+      "loss": 0.7667,
+      "step": 524
+    },
+    {
+      "epoch": 0.7301808066759388,
+      "grad_norm": 0.621879518032074,
+      "learning_rate": 3.3826216260114604e-05,
+      "loss": 0.8856,
+      "step": 525
+    },
+    {
+      "epoch": 0.7315716272600834,
+      "grad_norm": 0.40648820996284485,
+      "learning_rate": 3.349926025650796e-05,
+      "loss": 0.5985,
+      "step": 526
+    },
+    {
+      "epoch": 0.7329624478442281,
+      "grad_norm": 0.7350391745567322,
+      "learning_rate": 3.3173573855346043e-05,
+      "loss": 0.8338,
+      "step": 527
+    },
+    {
+      "epoch": 0.7343532684283728,
+      "grad_norm": 0.5308395028114319,
+      "learning_rate": 3.2849163274488314e-05,
+      "loss": 0.7613,
+      "step": 528
+    },
+    {
+      "epoch": 0.7357440890125174,
+      "grad_norm": 0.4623577296733856,
+      "learning_rate": 3.252603470743676e-05,
+      "loss": 0.7832,
+      "step": 529
+    },
+    {
+      "epoch": 0.7371349095966621,
+      "grad_norm": 0.5233444571495056,
+      "learning_rate": 3.220419432321783e-05,
+      "loss": 0.7505,
+      "step": 530
+    },
+    {
+      "epoch": 0.7385257301808067,
+      "grad_norm": 0.5883181691169739,
+      "learning_rate": 3.188364826626449e-05,
+      "loss": 0.955,
+      "step": 531
+    },
+    {
+      "epoch": 0.7399165507649513,
+      "grad_norm": 0.7370086908340454,
+      "learning_rate": 3.156440265629899e-05,
+      "loss": 0.78,
+      "step": 532
+    },
+    {
+      "epoch": 0.741307371349096,
+      "grad_norm": 0.6598822474479675,
+      "learning_rate": 3.124646358821599e-05,
+      "loss": 0.8851,
+      "step": 533
+    },
+    {
+      "epoch": 0.7426981919332406,
+      "grad_norm": 0.6320440769195557,
+      "learning_rate": 3.092983713196639e-05,
+      "loss": 1.0022,
+      "step": 534
+    },
+    {
+      "epoch": 0.7440890125173852,
+      "grad_norm": 0.4336588978767395,
+      "learning_rate": 3.061452933244112e-05,
+      "loss": 0.655,
+      "step": 535
+    },
+    {
+      "epoch": 0.7454798331015299,
+      "grad_norm": 0.7890181541442871,
+      "learning_rate": 3.0300546209355995e-05,
+      "loss": 1.1719,
+      "step": 536
+    },
+    {
+      "epoch": 0.7468706536856745,
+      "grad_norm": 0.7184107303619385,
+      "learning_rate": 2.998789375713663e-05,
+      "loss": 0.6999,
+      "step": 537
+    },
+    {
+      "epoch": 0.7482614742698191,
+      "grad_norm": 0.35499611496925354,
+      "learning_rate": 2.9676577944804175e-05,
+      "loss": 0.3939,
+      "step": 538
+    },
+    {
+      "epoch": 0.7496522948539638,
+      "grad_norm": 0.40395939350128174,
+      "learning_rate": 2.9366604715861156e-05,
+      "loss": 0.6397,
+      "step": 539
+    },
+    {
+      "epoch": 0.7510431154381085,
+      "grad_norm": 0.6528116464614868,
+      "learning_rate": 2.9057979988178087e-05,
+      "loss": 0.8695,
+      "step": 540
+    },
+    {
+      "epoch": 0.7524339360222532,
+      "grad_norm": 0.45431745052337646,
+      "learning_rate": 2.8750709653880592e-05,
+      "loss": 0.743,
+      "step": 541
+    },
+    {
+      "epoch": 0.7538247566063978,
+      "grad_norm": 0.7888697385787964,
+      "learning_rate": 2.8444799579236704e-05,
+      "loss": 0.8291,
+      "step": 542
+    },
+    {
+      "epoch": 0.7552155771905424,
+      "grad_norm": 0.5523220896720886,
+      "learning_rate": 2.8140255604545106e-05,
+      "loss": 0.8189,
+      "step": 543
+    },
+    {
+      "epoch": 0.7566063977746871,
+      "grad_norm": 0.5537093877792358,
+      "learning_rate": 2.7837083544023422e-05,
+      "loss": 0.8548,
+      "step": 544
+    },
+    {
+      "epoch": 0.7579972183588317,
+      "grad_norm": 0.8726553916931152,
+      "learning_rate": 2.753528918569732e-05,
+      "loss": 1.0436,
+      "step": 545
+    },
+    {
+      "epoch": 0.7593880389429764,
+      "grad_norm": 0.5013949275016785,
+      "learning_rate": 2.723487829129001e-05,
+      "loss": 0.6692,
+      "step": 546
+    },
+    {
+      "epoch": 0.760778859527121,
+      "grad_norm": 0.5562347769737244,
+      "learning_rate": 2.6935856596112197e-05,
+      "loss": 0.5206,
+      "step": 547
+    },
+    {
+      "epoch": 0.7621696801112656,
+      "grad_norm": 0.7989206910133362,
+      "learning_rate": 2.6638229808952685e-05,
+      "loss": 0.8943,
+      "step": 548
+    },
+    {
+      "epoch": 0.7635605006954103,
+      "grad_norm": 0.5031208395957947,
+      "learning_rate": 2.6342003611969223e-05,
+      "loss": 0.623,
+      "step": 549
+    },
+    {
+      "epoch": 0.7649513212795549,
+      "grad_norm": 0.7044405341148376,
+      "learning_rate": 2.60471836605802e-05,
+      "loss": 0.7729,
+      "step": 550
+    },
+    {
+      "epoch": 0.7663421418636995,
+      "grad_norm": 0.6286691427230835,
+      "learning_rate": 2.575377558335649e-05,
+      "loss": 0.919,
+      "step": 551
+    },
+    {
+      "epoch": 0.7677329624478443,
+      "grad_norm": 0.5914317965507507,
+      "learning_rate": 2.5461784981914193e-05,
+      "loss": 1.1554,
+      "step": 552
+    },
+    {
+      "epoch": 0.7691237830319889,
+      "grad_norm": 0.45661044120788574,
+      "learning_rate": 2.5171217430807605e-05,
+      "loss": 0.6233,
+      "step": 553
+    },
+    {
+      "epoch": 0.7705146036161336,
+      "grad_norm": 0.5765746235847473,
+      "learning_rate": 2.4882078477422687e-05,
+      "loss": 0.8168,
+      "step": 554
+    },
+    {
+      "epoch": 0.7719054242002782,
+      "grad_norm": 0.5740229487419128,
+      "learning_rate": 2.4594373641871314e-05,
+      "loss": 0.9991,
+      "step": 555
+    },
+    {
+      "epoch": 0.7732962447844228,
+      "grad_norm": 0.5498442053794861,
+      "learning_rate": 2.4308108416885778e-05,
+      "loss": 0.7764,
+      "step": 556
+    },
+    {
+      "epoch": 0.7746870653685675,
+      "grad_norm": 0.589480459690094,
+      "learning_rate": 2.4023288267714038e-05,
+      "loss": 0.9341,
+      "step": 557
+    },
+    {
+      "epoch": 0.7760778859527121,
+      "grad_norm": 0.5806849002838135,
+      "learning_rate": 2.373991863201527e-05,
+      "loss": 0.8844,
+      "step": 558
+    },
+    {
+      "epoch": 0.7774687065368567,
+      "grad_norm": 0.6247819066047668,
+      "learning_rate": 2.345800491975606e-05,
+      "loss": 0.8373,
+      "step": 559
+    },
+    {
+      "epoch": 0.7788595271210014,
+      "grad_norm": 0.504511833190918,
+      "learning_rate": 2.317755251310719e-05,
+      "loss": 0.9002,
+      "step": 560
+    },
+    {
+      "epoch": 0.780250347705146,
+      "grad_norm": 0.5619590282440186,
+      "learning_rate": 2.289856676634089e-05,
+      "loss": 0.7567,
+      "step": 561
+    },
+    {
+      "epoch": 0.7816411682892906,
+      "grad_norm": 0.3563636243343353,
+      "learning_rate": 2.2621053005728498e-05,
+      "loss": 0.5357,
+      "step": 562
+    },
+    {
+      "epoch": 0.7830319888734353,
+      "grad_norm": 0.48769277334213257,
+      "learning_rate": 2.234501652943888e-05,
+      "loss": 0.7339,
+      "step": 563
+    },
+    {
+      "epoch": 0.7844228094575799,
+      "grad_norm": 0.8310726881027222,
+      "learning_rate": 2.2070462607437258e-05,
+      "loss": 1.3372,
+      "step": 564
+    },
+    {
+      "epoch": 0.7858136300417247,
+      "grad_norm": 0.689653754234314,
+      "learning_rate": 2.1797396481384546e-05,
+      "loss": 0.9748,
+      "step": 565
+    },
+    {
+      "epoch": 0.7872044506258693,
+      "grad_norm": 0.7067202925682068,
+      "learning_rate": 2.152582336453739e-05,
+      "loss": 1.1216,
+      "step": 566
+    },
+    {
+      "epoch": 0.7885952712100139,
+      "grad_norm": 0.5383090972900391,
+      "learning_rate": 2.125574844164856e-05,
+      "loss": 0.5679,
+      "step": 567
+    },
+    {
+      "epoch": 0.7899860917941586,
+      "grad_norm": 0.8883530497550964,
+      "learning_rate": 2.098717686886793e-05,
+      "loss": 0.7698,
+      "step": 568
+    },
+    {
+      "epoch": 0.7913769123783032,
+      "grad_norm": 0.6829560995101929,
+      "learning_rate": 2.0720113773644102e-05,
+      "loss": 1.1163,
+      "step": 569
+    },
+    {
+      "epoch": 0.7927677329624478,
+      "grad_norm": 0.582198441028595,
+      "learning_rate": 2.0454564254626473e-05,
+      "loss": 0.5864,
+      "step": 570
+    },
+    {
+      "epoch": 0.7941585535465925,
+      "grad_norm": 0.6021276712417603,
+      "learning_rate": 2.019053338156801e-05,
+      "loss": 0.8015,
+      "step": 571
+    },
+    {
+      "epoch": 0.7955493741307371,
+      "grad_norm": 0.47162336111068726,
+      "learning_rate": 1.9928026195228288e-05,
+      "loss": 0.6959,
+      "step": 572
+    },
+    {
+      "epoch": 0.7969401947148818,
+      "grad_norm": 0.4599682688713074,
+      "learning_rate": 1.966704770727734e-05,
+      "loss": 0.6825,
+      "step": 573
+    },
+    {
+      "epoch": 0.7983310152990264,
+      "grad_norm": 0.7000210881233215,
+      "learning_rate": 1.940760290019997e-05,
+      "loss": 0.8055,
+      "step": 574
+    },
+    {
+      "epoch": 0.799721835883171,
+      "grad_norm": 0.35347720980644226,
+      "learning_rate": 1.9149696727200695e-05,
+      "loss": 0.6099,
+      "step": 575
+    },
+    {
+      "epoch": 0.8011126564673157,
+      "grad_norm": 0.4522395133972168,
+      "learning_rate": 1.889333411210905e-05,
+      "loss": 0.7279,
+      "step": 576
+    },
+    {
+      "epoch": 0.8025034770514604,
+      "grad_norm": 0.5675965547561646,
+      "learning_rate": 1.8638519949285636e-05,
+      "loss": 0.4292,
+      "step": 577
+    },
+    {
+      "epoch": 0.803894297635605,
+      "grad_norm": 0.3578422963619232,
+      "learning_rate": 1.8385259103528784e-05,
+      "loss": 0.3718,
+      "step": 578
+    },
+    {
+      "epoch": 0.8052851182197497,
+      "grad_norm": 0.629468560218811,
+      "learning_rate": 1.813355640998149e-05,
+      "loss": 0.9091,
+      "step": 579
+    },
+    {
+      "epoch": 0.8066759388038943,
+      "grad_norm": 0.5621572136878967,
+      "learning_rate": 1.7883416674039278e-05,
+      "loss": 0.6464,
+      "step": 580
+    },
+    {
+      "epoch": 0.808066759388039,
+      "grad_norm": 0.4115629196166992,
+      "learning_rate": 1.7634844671258354e-05,
+      "loss": 0.5067,
+      "step": 581
+    },
+    {
+      "epoch": 0.8094575799721836,
+      "grad_norm": 0.5084241628646851,
+      "learning_rate": 1.7387845147264446e-05,
+      "loss": 0.8549,
+      "step": 582
+    },
+    {
+      "epoch": 0.8108484005563282,
+      "grad_norm": 0.35453739762306213,
+      "learning_rate": 1.714242281766224e-05,
+      "loss": 0.439,
+      "step": 583
+    },
+    {
+      "epoch": 0.8122392211404729,
+      "grad_norm": 0.6283367276191711,
+      "learning_rate": 1.6898582367945293e-05,
+      "loss": 0.6407,
+      "step": 584
+    },
+    {
+      "epoch": 0.8136300417246175,
+      "grad_norm": 0.6594211459159851,
+      "learning_rate": 1.665632845340669e-05,
+      "loss": 0.7249,
+      "step": 585
+    },
+    {
+      "epoch": 0.8150208623087621,
+      "grad_norm": 0.6068059206008911,
+      "learning_rate": 1.6415665699050044e-05,
+      "loss": 1.1712,
+      "step": 586
+    },
+    {
+      "epoch": 0.8164116828929068,
+      "grad_norm": 0.7481211423873901,
+      "learning_rate": 1.6176598699501256e-05,
+      "loss": 1.0756,
+      "step": 587
+    },
+    {
+      "epoch": 0.8178025034770514,
+      "grad_norm": 0.44894516468048096,
+      "learning_rate": 1.5939132018920743e-05,
+      "loss": 0.5388,
+      "step": 588
+    },
+    {
+      "epoch": 0.8191933240611962,
+      "grad_norm": 0.39619624614715576,
+      "learning_rate": 1.5703270190916453e-05,
+      "loss": 0.5606,
+      "step": 589
+    },
+    {
+      "epoch": 0.8205841446453408,
+      "grad_norm": 0.5128390192985535,
+      "learning_rate": 1.5469017718457124e-05,
+      "loss": 0.6048,
+      "step": 590
+    },
+    {
+      "epoch": 0.8219749652294854,
+      "grad_norm": 0.46330222487449646,
+      "learning_rate": 1.5236379073786434e-05,
+      "loss": 0.5826,
+      "step": 591
+    },
+    {
+      "epoch": 0.8233657858136301,
+      "grad_norm": 0.5335388779640198,
+      "learning_rate": 1.5005358698337535e-05,
+      "loss": 0.5905,
+      "step": 592
+    },
+    {
+      "epoch": 0.8247566063977747,
+      "grad_norm": 0.7311825156211853,
+      "learning_rate": 1.477596100264832e-05,
+      "loss": 0.8828,
+      "step": 593
+    },
+    {
+      "epoch": 0.8261474269819193,
+      "grad_norm": 0.6245663166046143,
+      "learning_rate": 1.454819036627727e-05,
+      "loss": 0.7556,
+      "step": 594
+    },
+    {
+      "epoch": 0.827538247566064,
+      "grad_norm": 0.5925797820091248,
+      "learning_rate": 1.4322051137719684e-05,
+      "loss": 0.3943,
+      "step": 595
+    },
+    {
+      "epoch": 0.8289290681502086,
+      "grad_norm": 0.5229666829109192,
+      "learning_rate": 1.4097547634324814e-05,
+      "loss": 0.8547,
+      "step": 596
+    },
+    {
+      "epoch": 0.8303198887343533,
+      "grad_norm": 0.49766281247138977,
+      "learning_rate": 1.3874684142213312e-05,
+      "loss": 0.8037,
+      "step": 597
+    },
+    {
+      "epoch": 0.8317107093184979,
+      "grad_norm": 0.5616022944450378,
+      "learning_rate": 1.3653464916195558e-05,
+      "loss": 0.709,
+      "step": 598
+    },
+    {
+      "epoch": 0.8331015299026425,
+      "grad_norm": 0.8873728513717651,
+      "learning_rate": 1.3433894179690265e-05,
+      "loss": 1.0121,
+      "step": 599
+    },
+    {
+      "epoch": 0.8344923504867872,
+      "grad_norm": 0.6128416657447815,
+      "learning_rate": 1.3215976124643947e-05,
+      "loss": 1.1313,
+      "step": 600
+    },
+    {
+      "epoch": 0.8358831710709318,
+      "grad_norm": 0.40921396017074585,
+      "learning_rate": 1.2999714911450833e-05,
+      "loss": 0.5275,
+      "step": 601
+    },
+    {
+      "epoch": 0.8372739916550765,
+      "grad_norm": 0.7264837622642517,
+      "learning_rate": 1.278511466887351e-05,
+      "loss": 0.8739,
+      "step": 602
+    },
+    {
+      "epoch": 0.8386648122392212,
+      "grad_norm": 0.7158416509628296,
+      "learning_rate": 1.2572179493964066e-05,
+      "loss": 0.7647,
+      "step": 603
+    },
+    {
+      "epoch": 0.8400556328233658,
+      "grad_norm": 0.7262928485870361,
+      "learning_rate": 1.2360913451985801e-05,
+      "loss": 0.9295,
+      "step": 604
+    },
+    {
+      "epoch": 0.8414464534075105,
+      "grad_norm": 0.7467795014381409,
+      "learning_rate": 1.2151320576335701e-05,
+      "loss": 0.9113,
+      "step": 605
+    },
+    {
+      "epoch": 0.8428372739916551,
+      "grad_norm": 0.4393318295478821,
+      "learning_rate": 1.1943404868467412e-05,
+      "loss": 0.7946,
+      "step": 606
+    },
+    {
+      "epoch": 0.8442280945757997,
+      "grad_norm": 0.5381770133972168,
+      "learning_rate": 1.1737170297814826e-05,
+      "loss": 0.7992,
+      "step": 607
+    },
+    {
+      "epoch": 0.8456189151599444,
+      "grad_norm": 0.9111826419830322,
+      "learning_rate": 1.1532620801716354e-05,
+      "loss": 0.9123,
+      "step": 608
+    },
+    {
+      "epoch": 0.847009735744089,
+      "grad_norm": 0.6135514378547668,
+      "learning_rate": 1.1329760285339674e-05,
+      "loss": 0.7542,
+      "step": 609
+    },
+    {
+      "epoch": 0.8484005563282336,
+      "grad_norm": 0.6197137236595154,
+      "learning_rate": 1.1128592621607226e-05,
+      "loss": 0.9981,
+      "step": 610
+    },
+    {
+      "epoch": 0.8497913769123783,
+      "grad_norm": 0.4415544271469116,
+      "learning_rate": 1.0929121651122254e-05,
+      "loss": 0.7442,
+      "step": 611
+    },
+    {
+      "epoch": 0.8511821974965229,
+      "grad_norm": 0.5056378245353699,
+      "learning_rate": 1.0731351182095539e-05,
+      "loss": 0.6635,
+      "step": 612
+    },
+    {
+      "epoch": 0.8525730180806675,
+      "grad_norm": 0.4786520004272461,
+      "learning_rate": 1.0535284990272599e-05,
+      "loss": 0.4855,
+      "step": 613
+    },
+    {
+      "epoch": 0.8539638386648123,
+      "grad_norm": 0.5326875448226929,
+      "learning_rate": 1.034092681886164e-05,
+      "loss": 0.8491,
+      "step": 614
+    },
+    {
+      "epoch": 0.8553546592489569,
+      "grad_norm": 0.8259335160255432,
+      "learning_rate": 1.0148280378462182e-05,
+      "loss": 0.6307,
+      "step": 615
+    },
+    {
+      "epoch": 0.8567454798331016,
+      "grad_norm": 0.6941313743591309,
+      "learning_rate": 9.957349346994038e-06,
+      "loss": 0.8014,
+      "step": 616
+    },
+    {
+      "epoch": 0.8581363004172462,
+      "grad_norm": 0.4368307590484619,
+      "learning_rate": 9.768137369627296e-06,
+      "loss": 0.537,
+      "step": 617
+    },
+    {
+      "epoch": 0.8595271210013908,
+      "grad_norm": 0.5466663837432861,
+      "learning_rate": 9.58064805871256e-06,
+      "loss": 0.4442,
+      "step": 618
+    },
+    {
+      "epoch": 0.8609179415855355,
+      "grad_norm": 0.3190799951553345,
+      "learning_rate": 9.394884993712094e-06,
+      "loss": 0.398,
+      "step": 619
+    },
+    {
+      "epoch": 0.8623087621696801,
+      "grad_norm": 0.4227621555328369,
+      "learning_rate": 9.210851721131398e-06,
+      "loss": 0.3173,
+      "step": 620
+    },
+    {
+      "epoch": 0.8636995827538247,
+      "grad_norm": 0.45365777611732483,
+      "learning_rate": 9.028551754451619e-06,
+      "loss": 0.7648,
+      "step": 621
+    },
+    {
+      "epoch": 0.8650904033379694,
+      "grad_norm": 0.6490700840950012,
+      "learning_rate": 8.84798857406235e-06,
+      "loss": 0.6691,
+      "step": 622
+    },
+    {
+      "epoch": 0.866481223922114,
+      "grad_norm": 0.5763516426086426,
+      "learning_rate": 8.669165627195242e-06,
+      "loss": 0.9401,
+      "step": 623
+    },
+    {
+      "epoch": 0.8678720445062587,
+      "grad_norm": 0.7732207775115967,
+      "learning_rate": 8.492086327858185e-06,
+      "loss": 0.7203,
+      "step": 624
+    },
+    {
+      "epoch": 0.8692628650904033,
+      "grad_norm": 0.45827800035476685,
+      "learning_rate": 8.316754056770138e-06,
+      "loss": 0.6771,
+      "step": 625
+    },
+    {
+      "epoch": 0.8706536856745479,
+      "grad_norm": 0.6592226624488831,
+      "learning_rate": 8.143172161296586e-06,
+      "loss": 0.8622,
+      "step": 626
+    },
+    {
+      "epoch": 0.8720445062586927,
+      "grad_norm": 0.49130070209503174,
+      "learning_rate": 7.97134395538559e-06,
+      "loss": 0.9927,
+      "step": 627
+    },
+    {
+      "epoch": 0.8734353268428373,
+      "grad_norm": 0.6076432466506958,
+      "learning_rate": 7.801272719504593e-06,
+      "loss": 0.8569,
+      "step": 628
+    },
+    {
+      "epoch": 0.874826147426982,
+      "grad_norm": 0.6537121534347534,
+      "learning_rate": 7.632961700577723e-06,
+      "loss": 0.8136,
+      "step": 629
+    },
+    {
+      "epoch": 0.8762169680111266,
+      "grad_norm": 0.6857277750968933,
+      "learning_rate": 7.466414111923814e-06,
+      "loss": 0.9315,
+      "step": 630
+    },
+    {
+      "epoch": 0.8776077885952712,
+      "grad_norm": 0.38490551710128784,
+      "learning_rate": 7.301633133195141e-06,
+      "loss": 0.4386,
+      "step": 631
+    },
+    {
+      "epoch": 0.8789986091794159,
+      "grad_norm": 0.5177165865898132,
+      "learning_rate": 7.138621910316589e-06,
+      "loss": 0.585,
+      "step": 632
+    },
+    {
+      "epoch": 0.8803894297635605,
+      "grad_norm": 0.507311999797821,
+      "learning_rate": 6.9773835554256875e-06,
+      "loss": 0.6943,
+      "step": 633
+    },
+    {
+      "epoch": 0.8817802503477051,
+      "grad_norm": 0.6692008972167969,
+      "learning_rate": 6.817921146813111e-06,
+      "loss": 0.9552,
+      "step": 634
+    },
+    {
+      "epoch": 0.8831710709318498,
+      "grad_norm": 0.6843410730361938,
+      "learning_rate": 6.660237728864039e-06,
+      "loss": 0.702,
+      "step": 635
+    },
+    {
+      "epoch": 0.8845618915159944,
+      "grad_norm": 0.45811110734939575,
+      "learning_rate": 6.504336311999915e-06,
+      "loss": 0.4474,
+      "step": 636
+    },
+    {
+      "epoch": 0.885952712100139,
+      "grad_norm": 0.5536378622055054,
+      "learning_rate": 6.3502198726209885e-06,
+      "loss": 0.5747,
+      "step": 637
+    },
+    {
+      "epoch": 0.8873435326842837,
+      "grad_norm": 0.9685459136962891,
+      "learning_rate": 6.197891353049545e-06,
+      "loss": 1.1824,
+      "step": 638
+    },
+    {
+      "epoch": 0.8887343532684284,
+      "grad_norm": 0.5099877715110779,
+      "learning_rate": 6.047353661473698e-06,
+      "loss": 0.6052,
+      "step": 639
+    },
+    {
+      "epoch": 0.8901251738525731,
+      "grad_norm": 0.7409347295761108,
+      "learning_rate": 5.898609671891897e-06,
+      "loss": 1.2359,
+      "step": 640
+    },
+    {
+      "epoch": 0.8915159944367177,
+      "grad_norm": 0.827233612537384,
+      "learning_rate": 5.751662224057997e-06,
+      "loss": 0.8234,
+      "step": 641
+    },
+    {
+      "epoch": 0.8929068150208623,
+      "grad_norm": 0.5103647708892822,
+      "learning_rate": 5.606514123427076e-06,
+      "loss": 0.6315,
+      "step": 642
+    },
+    {
+      "epoch": 0.894297635605007,
+      "grad_norm": 0.507067084312439,
+      "learning_rate": 5.463168141101904e-06,
+      "loss": 1.0608,
+      "step": 643
+    },
+    {
+      "epoch": 0.8956884561891516,
+      "grad_norm": 0.602500319480896,
+      "learning_rate": 5.321627013779984e-06,
+      "loss": 0.5806,
+      "step": 644
+    },
+    {
+      "epoch": 0.8970792767732962,
+      "grad_norm": 0.6329058408737183,
+      "learning_rate": 5.18189344370138e-06,
+      "loss": 1.1104,
+      "step": 645
+    },
+    {
+      "epoch": 0.8984700973574409,
+      "grad_norm": 0.47934696078300476,
+      "learning_rate": 5.04397009859704e-06,
+      "loss": 0.5372,
+      "step": 646
+    },
+    {
+      "epoch": 0.8998609179415855,
+      "grad_norm": 0.647960901260376,
+      "learning_rate": 4.907859611637911e-06,
+      "loss": 0.6997,
+      "step": 647
+    },
+    {
+      "epoch": 0.9012517385257302,
+      "grad_norm": 0.3092882037162781,
+      "learning_rate": 4.773564581384659e-06,
+      "loss": 0.5109,
+      "step": 648
+    },
+    {
+      "epoch": 0.9026425591098748,
+      "grad_norm": 0.41960397362709045,
+      "learning_rate": 4.6410875717380895e-06,
+      "loss": 0.4575,
+      "step": 649
+    },
+    {
+      "epoch": 0.9040333796940194,
+      "grad_norm": 0.5140801668167114,
+      "learning_rate": 4.510431111890134e-06,
+      "loss": 0.8857,
+      "step": 650
+    },
+    {
+      "epoch": 0.9054242002781642,
+      "grad_norm": 0.7521938681602478,
+      "learning_rate": 4.381597696275596e-06,
+      "loss": 0.8543,
+      "step": 651
+    },
+    {
+      "epoch": 0.9068150208623088,
+      "grad_norm": 0.5770370364189148,
+      "learning_rate": 4.254589784524576e-06,
+      "loss": 0.838,
+      "step": 652
+    },
+    {
+      "epoch": 0.9082058414464534,
+      "grad_norm": 0.6081773042678833,
+      "learning_rate": 4.129409801415429e-06,
+      "loss": 1.054,
+      "step": 653
+    },
+    {
+      "epoch": 0.9095966620305981,
+      "grad_norm": 0.5241495966911316,
+      "learning_rate": 4.006060136828549e-06,
+      "loss": 0.7008,
+      "step": 654
+    },
+    {
+      "epoch": 0.9109874826147427,
+      "grad_norm": 0.5054779648780823,
+      "learning_rate": 3.884543145700659e-06,
+      "loss": 0.8989,
+      "step": 655
+    },
+    {
+      "epoch": 0.9123783031988874,
+      "grad_norm": 0.47228187322616577,
+      "learning_rate": 3.7648611479799366e-06,
+      "loss": 0.7675,
+      "step": 656
+    },
+    {
+      "epoch": 0.913769123783032,
+      "grad_norm": 0.6372430324554443,
+      "learning_rate": 3.64701642858164e-06,
+      "loss": 0.838,
+      "step": 657
+    },
+    {
+      "epoch": 0.9151599443671766,
+      "grad_norm": 0.5615782141685486,
+      "learning_rate": 3.5310112373445948e-06,
+      "loss": 0.5772,
+      "step": 658
+    },
+    {
+      "epoch": 0.9165507649513213,
+      "grad_norm": 0.5766297578811646,
+      "learning_rate": 3.416847788988131e-06,
+      "loss": 1.0419,
+      "step": 659
+    },
+    {
+      "epoch": 0.9179415855354659,
+      "grad_norm": 0.45187902450561523,
+      "learning_rate": 3.3045282630698506e-06,
+      "loss": 0.6207,
+      "step": 660
+    },
+    {
+      "epoch": 0.9193324061196105,
+      "grad_norm": 0.46699273586273193,
+      "learning_rate": 3.1940548039440132e-06,
+      "loss": 0.329,
+      "step": 661
+    },
+    {
+      "epoch": 0.9207232267037552,
+      "grad_norm": 0.5114088654518127,
+      "learning_rate": 3.085429520720595e-06,
+      "loss": 0.8997,
+      "step": 662
+    },
+    {
+      "epoch": 0.9221140472878998,
+      "grad_norm": 0.552072286605835,
+      "learning_rate": 2.978654487225041e-06,
+      "loss": 0.804,
+      "step": 663
+    },
+    {
+      "epoch": 0.9235048678720446,
+      "grad_norm": 0.6190208792686462,
+      "learning_rate": 2.8737317419586185e-06,
+      "loss": 0.6113,
+      "step": 664
+    },
+    {
+      "epoch": 0.9248956884561892,
+      "grad_norm": 0.7381587624549866,
+      "learning_rate": 2.7706632880595716e-06,
+      "loss": 0.884,
+      "step": 665
+    },
+    {
+      "epoch": 0.9262865090403338,
+      "grad_norm": 0.6619057059288025,
+      "learning_rate": 2.669451093264819e-06,
+      "loss": 1.1775,
+      "step": 666
+    },
+    {
+      "epoch": 0.9276773296244785,
+      "grad_norm": 0.46355685591697693,
+      "learning_rate": 2.570097089872392e-06,
+      "loss": 0.4978,
+      "step": 667
+    },
+    {
+      "epoch": 0.9290681502086231,
+      "grad_norm": 0.6504789590835571,
+      "learning_rate": 2.472603174704624e-06,
+      "loss": 0.7307,
+      "step": 668
+    },
+    {
+      "epoch": 0.9304589707927677,
+      "grad_norm": 0.5652433633804321,
+      "learning_rate": 2.37697120907181e-06,
+      "loss": 0.6297,
+      "step": 669
+    },
+    {
+      "epoch": 0.9318497913769124,
+      "grad_norm": 0.6597036719322205,
+      "learning_rate": 2.283203018736757e-06,
+      "loss": 0.9534,
+      "step": 670
+    },
+    {
+      "epoch": 0.933240611961057,
+      "grad_norm": 0.5566277503967285,
+      "learning_rate": 2.1913003938798916e-06,
+      "loss": 0.9441,
+      "step": 671
+    },
+    {
+      "epoch": 0.9346314325452016,
+      "grad_norm": 0.7081180214881897,
+      "learning_rate": 2.101265089065152e-06,
+      "loss": 0.9396,
+      "step": 672
+    },
+    {
+      "epoch": 0.9360222531293463,
+      "grad_norm": 0.4790264070034027,
+      "learning_rate": 2.013098823206372e-06,
+      "loss": 0.6906,
+      "step": 673
+    },
+    {
+      "epoch": 0.9374130737134909,
+      "grad_norm": 0.5492000579833984,
+      "learning_rate": 1.9268032795345613e-06,
+      "loss": 0.7632,
+      "step": 674
+    },
+    {
+      "epoch": 0.9388038942976356,
+      "grad_norm": 0.3445267379283905,
+      "learning_rate": 1.842380105565711e-06,
+      "loss": 0.408,
+      "step": 675
+    },
+    {
+      "epoch": 0.9401947148817803,
+      "grad_norm": 0.840276837348938,
+      "learning_rate": 1.7598309130693825e-06,
+      "loss": 0.9225,
+      "step": 676
+    },
+    {
+      "epoch": 0.9415855354659249,
+      "grad_norm": 0.514579176902771,
+      "learning_rate": 1.679157278037946e-06,
+      "loss": 1.023,
+      "step": 677
+    },
+    {
+      "epoch": 0.9429763560500696,
+      "grad_norm": 0.674146294593811,
+      "learning_rate": 1.6003607406564148e-06,
+      "loss": 0.9299,
+      "step": 678
+    },
+    {
+      "epoch": 0.9443671766342142,
+      "grad_norm": 0.6753702759742737,
+      "learning_rate": 1.523442805273123e-06,
+      "loss": 0.8546,
+      "step": 679
+    },
+    {
+      "epoch": 0.9457579972183588,
+      "grad_norm": 0.5497095584869385,
+      "learning_rate": 1.448404940370951e-06,
+      "loss": 0.6086,
+      "step": 680
+    },
+    {
+      "epoch": 0.9471488178025035,
+      "grad_norm": 0.39195868372917175,
+      "learning_rate": 1.3752485785393455e-06,
+      "loss": 0.5186,
+      "step": 681
+    },
+    {
+      "epoch": 0.9485396383866481,
+      "grad_norm": 0.7977584600448608,
+      "learning_rate": 1.3039751164469095e-06,
+      "loss": 1.0067,
+      "step": 682
+    },
+    {
+      "epoch": 0.9499304589707928,
+      "grad_norm": 0.7335418462753296,
+      "learning_rate": 1.2345859148147787e-06,
+      "loss": 1.0071,
+      "step": 683
+    },
+    {
+      "epoch": 0.9513212795549374,
+      "grad_norm": 0.5838838219642639,
+      "learning_rate": 1.1670822983906094e-06,
+      "loss": 0.9712,
+      "step": 684
+    },
+    {
+      "epoch": 0.952712100139082,
+      "grad_norm": 0.49780187010765076,
+      "learning_rate": 1.1014655559233312e-06,
+      "loss": 0.426,
+      "step": 685
+    },
+    {
+      "epoch": 0.9541029207232267,
+      "grad_norm": 0.46391940116882324,
+      "learning_rate": 1.0377369401385007e-06,
+      "loss": 0.5459,
+      "step": 686
+    },
+    {
+      "epoch": 0.9554937413073713,
+      "grad_norm": 0.5011250376701355,
+      "learning_rate": 9.758976677144206e-07,
+      "loss": 0.5285,
+      "step": 687
+    },
+    {
+      "epoch": 0.9568845618915159,
+      "grad_norm": 0.690269947052002,
+      "learning_rate": 9.159489192588688e-07,
+      "loss": 0.9357,
+      "step": 688
+    },
+    {
+      "epoch": 0.9582753824756607,
+      "grad_norm": 0.5723053812980652,
+      "learning_rate": 8.578918392866064e-07,
+      "loss": 0.7386,
+      "step": 689
+    },
+    {
+      "epoch": 0.9596662030598053,
+      "grad_norm": 0.5798659324645996,
+      "learning_rate": 8.01727536197483e-07,
+      "loss": 1.0101,
+      "step": 690
+    },
+    {
+      "epoch": 0.96105702364395,
+      "grad_norm": 0.49990880489349365,
+      "learning_rate": 7.474570822553317e-07,
+      "loss": 0.5511,
+      "step": 691
+    },
+    {
+      "epoch": 0.9624478442280946,
+      "grad_norm": 0.5634463429450989,
+      "learning_rate": 6.950815135674083e-07,
+      "loss": 0.6329,
+      "step": 692
+    },
+    {
+      "epoch": 0.9638386648122392,
+      "grad_norm": 0.48899972438812256,
+      "learning_rate": 6.446018300647172e-07,
+      "loss": 0.8477,
+      "step": 693
+    },
+    {
+      "epoch": 0.9652294853963839,
+      "grad_norm": 0.6336762309074402,
+      "learning_rate": 5.960189954828277e-07,
+      "loss": 0.7969,
+      "step": 694
+    },
+    {
+      "epoch": 0.9666203059805285,
+      "grad_norm": 0.3693409562110901,
+      "learning_rate": 5.493339373435657e-07,
+      "loss": 0.7909,
+      "step": 695
+    },
+    {
+      "epoch": 0.9680111265646731,
+      "grad_norm": 0.4643179476261139,
+      "learning_rate": 5.045475469371952e-07,
+      "loss": 0.4941,
+      "step": 696
+    },
+    {
+      "epoch": 0.9694019471488178,
+      "grad_norm": 0.539460301399231,
+      "learning_rate": 4.6166067930550895e-07,
+      "loss": 0.5685,
+      "step": 697
+    },
+    {
+      "epoch": 0.9707927677329624,
+      "grad_norm": 0.5916941165924072,
+      "learning_rate": 4.206741532254421e-07,
+      "loss": 1.1108,
+      "step": 698
+    },
+    {
+      "epoch": 0.972183588317107,
+      "grad_norm": 0.6084417104721069,
+      "learning_rate": 3.815887511934735e-07,
+      "loss": 1.0283,
+      "step": 699
+    },
+    {
+      "epoch": 0.9735744089012517,
+      "grad_norm": 0.5716700553894043,
+      "learning_rate": 3.444052194106262e-07,
+      "loss": 0.4118,
+      "step": 700
+    },
+    {
+      "epoch": 0.9749652294853964,
+      "grad_norm": 0.5906838774681091,
+      "learning_rate": 3.091242677683126e-07,
+      "loss": 0.8837,
+      "step": 701
+    },
+    {
+      "epoch": 0.9763560500695411,
+      "grad_norm": 0.6931248307228088,
+      "learning_rate": 2.757465698347117e-07,
+      "loss": 0.8323,
+      "step": 702
+    },
+    {
+      "epoch": 0.9777468706536857,
+      "grad_norm": 0.5451267957687378,
+      "learning_rate": 2.44272762841935e-07,
+      "loss": 0.5634,
+      "step": 703
+    },
+    {
+      "epoch": 0.9791376912378303,
+      "grad_norm": 0.4953474700450897,
+      "learning_rate": 2.1470344767386963e-07,
+      "loss": 0.6848,
+      "step": 704
+    },
+    {
+      "epoch": 0.980528511821975,
+      "grad_norm": 0.6026284098625183,
+      "learning_rate": 1.870391888546652e-07,
+      "loss": 0.7091,
+      "step": 705
+    },
+    {
+      "epoch": 0.9819193324061196,
+      "grad_norm": 0.39837387204170227,
+      "learning_rate": 1.612805145380314e-07,
+      "loss": 0.428,
+      "step": 706
+    },
+    {
+      "epoch": 0.9833101529902643,
+      "grad_norm": 0.7257748246192932,
+      "learning_rate": 1.374279164970571e-07,
+      "loss": 0.9248,
+      "step": 707
+    },
+    {
+      "epoch": 0.9847009735744089,
+      "grad_norm": 0.7016385793685913,
+      "learning_rate": 1.1548185011490686e-07,
+      "loss": 0.83,
+      "step": 708
+    },
+    {
+      "epoch": 0.9860917941585535,
+      "grad_norm": 0.4826374650001526,
+      "learning_rate": 9.544273437609441e-08,
+      "loss": 0.6353,
+      "step": 709
+    },
+    {
+      "epoch": 0.9874826147426982,
+      "grad_norm": 0.5491541028022766,
+      "learning_rate": 7.731095185846693e-08,
+      "loss": 0.7307,
+      "step": 710
+    },
+    {
+      "epoch": 0.9888734353268428,
+      "grad_norm": 0.5050036907196045,
+      "learning_rate": 6.108684872592197e-08,
+      "loss": 0.554,
+      "step": 711
+    },
+    {
+      "epoch": 0.9902642559109874,
+      "grad_norm": 0.551762044429779,
+      "learning_rate": 4.67707347218016e-08,
+      "loss": 0.7418,
+      "step": 712
+    },
+    {
+      "epoch": 0.9916550764951322,
+      "grad_norm": 0.5763559937477112,
+      "learning_rate": 3.436288316296388e-08,
+      "loss": 0.7295,
+      "step": 713
+    },
+    {
+      "epoch": 0.9930458970792768,
+      "grad_norm": 0.48820456862449646,
+      "learning_rate": 2.3863530934564727e-08,
+      "loss": 0.6172,
+      "step": 714
+    },
+    {
+      "epoch": 0.9944367176634215,
+      "grad_norm": 0.6964418292045593,
+      "learning_rate": 1.5272878485561582e-08,
+      "loss": 1.0286,
+      "step": 715
+    },
+    {
+      "epoch": 0.9958275382475661,
+      "grad_norm": 0.8657813668251038,
+      "learning_rate": 8.591089824860898e-09,
+      "loss": 1.2155,
+      "step": 716
+    },
+    {
+      "epoch": 0.9972183588317107,
+      "grad_norm": 0.4634774327278137,
+      "learning_rate": 3.818292518176225e-09,
+      "loss": 0.4614,
+      "step": 717
+    },
+    {
+      "epoch": 0.9986091794158554,
+      "grad_norm": 0.4873647391796112,
+      "learning_rate": 9.545776856412225e-10,
+      "loss": 0.7717,
+      "step": 718
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.091496229171753,
+      "learning_rate": 0.0,
+      "loss": 1.321,
+      "step": 719
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.8003649711608887,
+      "eval_runtime": 4.6434,
+      "eval_samples_per_second": 32.735,
+      "eval_steps_per_second": 32.735,
+      "step": 719
     }
   ],
   "logging_steps": 1,
-  "max_steps": 10,
+  "max_steps": 719,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 3,
+  "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -128,7 +5067,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 258586429095936.0,
+  "total_flos": 1.7974763641110528e+16,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null