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"
@@ -10,7003 +10,7011 @@
   "log_history": [
     {
       "epoch": 3.578009553285507e-05,
-      "grad_norm": 1.6233201026916504,
+      "grad_norm": 1.4940814971923828,
       "learning_rate": 0.00019999950652018584,
       "loss": 1.524,
       "step": 1
     },
     {
       "epoch": 7.156019106571014e-05,
-      "grad_norm": 3.6991474628448486,
+      "grad_norm": 3.4283759593963623,
       "learning_rate": 0.0001999980260856137,
-      "loss": 2.3704,
+      "loss": 2.3643,
       "step": 2
     },
     {
       "epoch": 0.00010734028659856522,
-      "grad_norm": 1.373162031173706,
+      "grad_norm": 1.3139270544052124,
       "learning_rate": 0.000199995558710895,
-      "loss": 0.8958,
+      "loss": 0.8935,
       "step": 3
     },
     {
       "epoch": 0.00014312038213142028,
-      "grad_norm": 1.098609447479248,
+      "grad_norm": 1.062879204750061,
       "learning_rate": 0.00019999210442038162,
-      "loss": 0.5749,
+      "loss": 0.5728,
       "step": 4
     },
     {
       "epoch": 0.00017890047766427536,
-      "grad_norm": 1.9097572565078735,
+      "grad_norm": 1.8600218296051025,
       "learning_rate": 0.00019998766324816607,
-      "loss": 0.6967,
+      "loss": 0.6961,
       "step": 5
     },
     {
       "epoch": 0.00021468057319713044,
-      "grad_norm": 1.6201567649841309,
+      "grad_norm": 1.5905139446258545,
       "learning_rate": 0.0001999822352380809,
-      "loss": 0.2754,
+      "loss": 0.2833,
       "step": 6
     },
     {
       "epoch": 0.0002504606687299855,
-      "grad_norm": 7.474282264709473,
+      "grad_norm": 6.56783390045166,
       "learning_rate": 0.00019997582044369843,
-      "loss": 0.2521,
+      "loss": 0.2166,
       "step": 7
     },
     {
       "epoch": 0.00028624076426284056,
-      "grad_norm": 1.5669171810150146,
+      "grad_norm": 1.2253773212432861,
       "learning_rate": 0.00019996841892833,
-      "loss": 0.0411,
+      "loss": 0.0389,
       "step": 8
     },
     {
       "epoch": 0.00032202085979569564,
-      "grad_norm": 0.6482686400413513,
+      "grad_norm": 0.7007305026054382,
       "learning_rate": 0.00019996003076502565,
-      "loss": 0.0148,
+      "loss": 0.0157,
       "step": 9
     },
     {
       "epoch": 0.0003578009553285507,
-      "grad_norm": 1.1002771854400635,
+      "grad_norm": 1.0908080339431763,
       "learning_rate": 0.00019995065603657316,
-      "loss": 0.0767,
+      "loss": 0.0769,
       "step": 10
     },
     {
       "epoch": 0.0003935810508614058,
-      "grad_norm": 0.19515837728977203,
+      "grad_norm": 0.3062255382537842,
       "learning_rate": 0.0001999402948354973,
-      "loss": 0.0043,
+      "loss": 0.0058,
       "step": 11
     },
     {
       "epoch": 0.0004293611463942609,
-      "grad_norm": 0.6988247036933899,
+      "grad_norm": 0.5422049164772034,
       "learning_rate": 0.00019992894726405893,
-      "loss": 0.0111,
+      "loss": 0.0098,
       "step": 12
     },
     {
       "epoch": 0.00046514124192711595,
-      "grad_norm": 1.3810594081878662,
+      "grad_norm": 0.9161967635154724,
       "learning_rate": 0.000199916613434254,
-      "loss": 0.0242,
+      "loss": 0.0162,
       "step": 13
     },
     {
       "epoch": 0.000500921337459971,
-      "grad_norm": 0.7165325880050659,
+      "grad_norm": 0.17783205211162567,
       "learning_rate": 0.0001999032934678125,
-      "loss": 0.0072,
+      "loss": 0.0023,
       "step": 14
     },
     {
       "epoch": 0.000536701432992826,
-      "grad_norm": 1.1972315311431885,
+      "grad_norm": 0.27469125390052795,
       "learning_rate": 0.00019988898749619702,
-      "loss": 0.0106,
+      "loss": 0.0035,
       "step": 15
     },
     {
       "epoch": 0.0005724815285256811,
-      "grad_norm": 3.4662997722625732,
+      "grad_norm": 3.371744394302368,
       "learning_rate": 0.00019987369566060176,
-      "loss": 0.0189,
+      "loss": 0.0233,
       "step": 16
     },
     {
       "epoch": 0.0006082616240585362,
-      "grad_norm": 0.22282695770263672,
+      "grad_norm": 0.6902956366539001,
       "learning_rate": 0.00019985741811195097,
-      "loss": 0.0012,
+      "loss": 0.0026,
       "step": 17
     },
     {
       "epoch": 0.0006440417195913913,
-      "grad_norm": 0.023447470739483833,
+      "grad_norm": 0.013953864574432373,
       "learning_rate": 0.00019984015501089752,
-      "loss": 0.0006,
+      "loss": 0.0004,
       "step": 18
     },
     {
       "epoch": 0.0006798218151242464,
-      "grad_norm": 0.019725991412997246,
+      "grad_norm": 0.005650808569043875,
       "learning_rate": 0.0001998219065278212,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 19
     },
     {
       "epoch": 0.0007156019106571014,
-      "grad_norm": 1.6941359043121338,
+      "grad_norm": 0.1283494383096695,
       "learning_rate": 0.00019980267284282717,
-      "loss": 0.006,
+      "loss": 0.0008,
       "step": 20
     },
     {
       "epoch": 0.0007513820061899565,
-      "grad_norm": 0.7053670287132263,
+      "grad_norm": 0.15634015202522278,
       "learning_rate": 0.00019978245414574417,
-      "loss": 0.0029,
+      "loss": 0.0011,
       "step": 21
     },
     {
       "epoch": 0.0007871621017228116,
-      "grad_norm": 0.07644766569137573,
+      "grad_norm": 0.3101247549057007,
       "learning_rate": 0.00019976125063612252,
-      "loss": 0.0008,
+      "loss": 0.001,
       "step": 22
     },
     {
       "epoch": 0.0008229421972556667,
-      "grad_norm": 1.149553656578064,
+      "grad_norm": 1.2682957649230957,
       "learning_rate": 0.00019973906252323238,
-      "loss": 0.0071,
+      "loss": 0.0052,
       "step": 23
     },
     {
       "epoch": 0.0008587222927885217,
-      "grad_norm": 0.057650115340948105,
+      "grad_norm": 0.329192578792572,
       "learning_rate": 0.0001997158900260614,
-      "loss": 0.0009,
+      "loss": 0.0015,
       "step": 24
     },
     {
       "epoch": 0.0008945023883213768,
-      "grad_norm": 1.8726991415023804,
+      "grad_norm": 0.8752695322036743,
       "learning_rate": 0.0001996917333733128,
-      "loss": 0.0147,
+      "loss": 0.004,
       "step": 25
     },
     {
       "epoch": 0.0009302824838542319,
-      "grad_norm": 0.06484769284725189,
+      "grad_norm": 0.16207607090473175,
       "learning_rate": 0.00019966659280340297,
-      "loss": 0.0008,
+      "loss": 0.0007,
       "step": 26
     },
     {
       "epoch": 0.000966062579387087,
-      "grad_norm": 0.032708559185266495,
+      "grad_norm": 0.002739664865657687,
       "learning_rate": 0.00019964046856445924,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 27
     },
     {
       "epoch": 0.001001842674919942,
-      "grad_norm": 0.06042423099279404,
+      "grad_norm": 0.8487679958343506,
       "learning_rate": 0.00019961336091431727,
-      "loss": 0.001,
+      "loss": 0.0045,
       "step": 28
     },
     {
       "epoch": 0.001037622770452797,
-      "grad_norm": 0.24676547944545746,
+      "grad_norm": 0.7072966694831848,
       "learning_rate": 0.00019958527012051857,
-      "loss": 0.0015,
+      "loss": 0.0016,
       "step": 29
     },
     {
       "epoch": 0.001073402865985652,
-      "grad_norm": 0.051142849028110504,
+      "grad_norm": 0.7657275199890137,
       "learning_rate": 0.00019955619646030802,
-      "loss": 0.0007,
+      "loss": 0.0024,
       "step": 30
     },
     {
       "epoch": 0.0011091829615185072,
-      "grad_norm": 0.0212563369423151,
+      "grad_norm": 0.05178741365671158,
       "learning_rate": 0.00019952614022063084,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 31
     },
     {
       "epoch": 0.0011449630570513623,
-      "grad_norm": 0.03593383729457855,
+      "grad_norm": 0.8519219756126404,
       "learning_rate": 0.00019949510169813003,
-      "loss": 0.0004,
+      "loss": 0.0044,
       "step": 32
     },
     {
       "epoch": 0.0011807431525842173,
-      "grad_norm": 0.770592987537384,
+      "grad_norm": 0.04824286326766014,
       "learning_rate": 0.00019946308119914323,
-      "loss": 0.0039,
+      "loss": 0.0012,
       "step": 33
     },
     {
       "epoch": 0.0012165232481170724,
-      "grad_norm": 0.4447389841079712,
+      "grad_norm": 0.10049160569906235,
       "learning_rate": 0.0001994300790396999,
-      "loss": 0.004,
+      "loss": 0.0013,
       "step": 34
     },
     {
       "epoch": 0.0012523033436499275,
-      "grad_norm": 0.022715669125318527,
+      "grad_norm": 0.021055705845355988,
       "learning_rate": 0.000199396095545518,
-      "loss": 0.0004,
+      "loss": 0.0002,
       "step": 35
     },
     {
       "epoch": 0.0012880834391827826,
-      "grad_norm": 0.3500293493270874,
+      "grad_norm": 0.79359370470047,
       "learning_rate": 0.00019936113105200085,
-      "loss": 0.0018,
+      "loss": 0.0017,
       "step": 36
     },
     {
       "epoch": 0.0013238635347156376,
-      "grad_norm": 0.44238173961639404,
+      "grad_norm": 0.3737424612045288,
       "learning_rate": 0.00019932518590423394,
-      "loss": 0.0067,
+      "loss": 0.0039,
       "step": 37
     },
     {
       "epoch": 0.0013596436302484927,
-      "grad_norm": 0.028947863727808,
+      "grad_norm": 0.0020007130224257708,
       "learning_rate": 0.00019928826045698136,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 38
     },
     {
       "epoch": 0.0013954237257813478,
-      "grad_norm": 0.0033073609229177237,
+      "grad_norm": 0.0023270663805305958,
       "learning_rate": 0.0001992503550746824,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 39
     },
     {
       "epoch": 0.0014312038213142029,
-      "grad_norm": 0.02347722463309765,
+      "grad_norm": 0.02750846929848194,
       "learning_rate": 0.0001992114701314478,
-      "loss": 0.0002,
+      "loss": 0.0004,
       "step": 40
     },
     {
       "epoch": 0.001466983916847058,
-      "grad_norm": 0.0015688989078626037,
+      "grad_norm": 0.015391684137284756,
       "learning_rate": 0.0001991716060110563,
       "loss": 0.0001,
       "step": 41
     },
     {
       "epoch": 0.001502764012379913,
-      "grad_norm": 1.501746416091919,
+      "grad_norm": 2.5219709873199463,
       "learning_rate": 0.00019913076310695068,
-      "loss": 0.0707,
+      "loss": 0.0401,
       "step": 42
     },
     {
       "epoch": 0.0015385441079127681,
-      "grad_norm": 0.007480164989829063,
+      "grad_norm": 0.0812700092792511,
       "learning_rate": 0.00019908894182223388,
-      "loss": 0.0001,
+      "loss": 0.0004,
       "step": 43
     },
     {
       "epoch": 0.0015743242034456232,
-      "grad_norm": 0.00436406908556819,
+      "grad_norm": 0.01355485338717699,
       "learning_rate": 0.00019904614256966512,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 44
     },
     {
       "epoch": 0.0016101042989784783,
-      "grad_norm": 0.001093980041332543,
+      "grad_norm": 0.0012422667350620031,
       "learning_rate": 0.00019900236577165576,
       "loss": 0.0001,
       "step": 45
     },
     {
       "epoch": 0.0016458843945113333,
-      "grad_norm": 0.0010382820619270205,
+      "grad_norm": 0.0012065384071320295,
       "learning_rate": 0.0001989576118602651,
       "loss": 0.0001,
       "step": 46
     },
     {
       "epoch": 0.0016816644900441884,
-      "grad_norm": 1.1161930561065674,
+      "grad_norm": 0.01380125805735588,
       "learning_rate": 0.00019891188127719618,
-      "loss": 0.0047,
+      "loss": 0.0006,
       "step": 47
     },
     {
       "epoch": 0.0017174445855770435,
-      "grad_norm": 0.7130130529403687,
+      "grad_norm": 0.028274118900299072,
       "learning_rate": 0.0001988651744737914,
-      "loss": 0.0052,
+      "loss": 0.0004,
       "step": 48
     },
     {
       "epoch": 0.0017532246811098986,
-      "grad_norm": 0.010638763196766376,
+      "grad_norm": 0.3730854094028473,
       "learning_rate": 0.00019881749191102808,
-      "loss": 0.0002,
+      "loss": 0.0046,
       "step": 49
     },
     {
       "epoch": 0.0017890047766427537,
-      "grad_norm": 0.05554693564772606,
+      "grad_norm": 0.004828768782317638,
       "learning_rate": 0.00019876883405951377,
-      "loss": 0.0005,
+      "loss": 0.0001,
       "step": 50
     },
     {
       "epoch": 0.0018247848721756087,
-      "grad_norm": 1.6455415487289429,
+      "grad_norm": 2.613616943359375,
       "learning_rate": 0.00019871920139948192,
-      "loss": 0.0849,
+      "loss": 0.0984,
       "step": 51
     },
     {
       "epoch": 0.0018605649677084638,
-      "grad_norm": 0.0026788024697452784,
+      "grad_norm": 0.00172997429035604,
       "learning_rate": 0.0001986685944207868,
       "loss": 0.0001,
       "step": 52
     },
     {
       "epoch": 0.0018963450632413189,
-      "grad_norm": 0.01997757889330387,
+      "grad_norm": 0.0117227453738451,
       "learning_rate": 0.0001986170136228989,
-      "loss": 0.0003,
+      "loss": 0.0002,
       "step": 53
     },
     {
       "epoch": 0.001932125158774174,
-      "grad_norm": 0.8625125885009766,
+      "grad_norm": 0.14388525485992432,
       "learning_rate": 0.00019856445951489982,
-      "loss": 0.0061,
+      "loss": 0.0013,
       "step": 54
     },
     {
       "epoch": 0.001967905254307029,
-      "grad_norm": 0.07987687736749649,
+      "grad_norm": 0.10388373583555222,
       "learning_rate": 0.0001985109326154774,
-      "loss": 0.0012,
+      "loss": 0.0008,
       "step": 55
     },
     {
       "epoch": 0.002003685349839884,
-      "grad_norm": 0.044615488499403,
+      "grad_norm": 0.0035523439291864634,
       "learning_rate": 0.00019845643345292054,
-      "loss": 0.0011,
+      "loss": 0.0002,
       "step": 56
     },
     {
       "epoch": 0.002039465445372739,
-      "grad_norm": 0.05630040913820267,
+      "grad_norm": 0.006682439241558313,
       "learning_rate": 0.00019840096256511398,
-      "loss": 0.0018,
+      "loss": 0.0003,
       "step": 57
     },
     {
       "epoch": 0.002075245540905594,
-      "grad_norm": 0.20012424886226654,
+      "grad_norm": 0.013288861140608788,
       "learning_rate": 0.00019834452049953297,
-      "loss": 0.0023,
+      "loss": 0.0004,
       "step": 58
     },
     {
       "epoch": 0.002111025636438449,
-      "grad_norm": 0.799498975276947,
+      "grad_norm": 0.007986829616129398,
       "learning_rate": 0.00019828710781323792,
-      "loss": 0.014,
+      "loss": 0.0003,
       "step": 59
     },
     {
       "epoch": 0.002146805731971304,
-      "grad_norm": 0.03027612529695034,
+      "grad_norm": 0.022813275456428528,
       "learning_rate": 0.0001982287250728689,
-      "loss": 0.0013,
+      "loss": 0.0007,
       "step": 60
     },
     {
       "epoch": 0.0021825858275041593,
-      "grad_norm": 0.023778654634952545,
+      "grad_norm": 0.017735198140144348,
       "learning_rate": 0.0001981693728546399,
-      "loss": 0.0011,
+      "loss": 0.0006,
       "step": 61
     },
     {
       "epoch": 0.0022183659230370144,
-      "grad_norm": 0.038122598081827164,
+      "grad_norm": 0.031412046402692795,
       "learning_rate": 0.0001981090517443334,
-      "loss": 0.0015,
+      "loss": 0.0008,
       "step": 62
     },
     {
       "epoch": 0.0022541460185698694,
-      "grad_norm": 1.425979733467102,
+      "grad_norm": 1.2430920600891113,
       "learning_rate": 0.00019804776233729444,
-      "loss": 0.0272,
+      "loss": 0.015,
       "step": 63
     },
     {
       "epoch": 0.0022899261141027245,
-      "grad_norm": 0.019651517271995544,
+      "grad_norm": 0.008373609744012356,
       "learning_rate": 0.0001979855052384247,
-      "loss": 0.0009,
+      "loss": 0.0004,
       "step": 64
     },
     {
       "epoch": 0.0023257062096355796,
-      "grad_norm": 0.019404835999011993,
+      "grad_norm": 0.012491998262703419,
       "learning_rate": 0.00019792228106217658,
-      "loss": 0.0008,
+      "loss": 0.0005,
       "step": 65
     },
     {
       "epoch": 0.0023614863051684347,
-      "grad_norm": 0.044460229575634,
+      "grad_norm": 2.589193105697632,
       "learning_rate": 0.00019785809043254722,
-      "loss": 0.0013,
+      "loss": 0.0061,
       "step": 66
     },
     {
       "epoch": 0.0023972664007012898,
-      "grad_norm": 0.008166422136127949,
+      "grad_norm": 0.007409463636577129,
       "learning_rate": 0.0001977929339830722,
-      "loss": 0.0005,
+      "loss": 0.0003,
       "step": 67
     },
     {
       "epoch": 0.002433046496234145,
-      "grad_norm": 0.026238959282636642,
+      "grad_norm": 0.1197238564491272,
       "learning_rate": 0.00019772681235681936,
-      "loss": 0.0006,
+      "loss": 0.0008,
       "step": 68
     },
     {
       "epoch": 0.002468826591767,
-      "grad_norm": 0.012247659265995026,
+      "grad_norm": 0.08556798845529556,
       "learning_rate": 0.00019765972620638248,
-      "loss": 0.0006,
+      "loss": 0.001,
       "step": 69
     },
     {
       "epoch": 0.002504606687299855,
-      "grad_norm": 0.029236063361167908,
+      "grad_norm": 0.010016442276537418,
       "learning_rate": 0.00019759167619387476,
-      "loss": 0.0009,
+      "loss": 0.0004,
       "step": 70
     },
     {
       "epoch": 0.00254038678283271,
-      "grad_norm": 0.03290548175573349,
+      "grad_norm": 0.10208962112665176,
       "learning_rate": 0.00019752266299092236,
-      "loss": 0.0007,
+      "loss": 0.0015,
       "step": 71
     },
     {
       "epoch": 0.002576166878365565,
-      "grad_norm": 0.006300569511950016,
+      "grad_norm": 0.00643647788092494,
       "learning_rate": 0.00019745268727865774,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 72
     },
     {
       "epoch": 0.00261194697389842,
-      "grad_norm": 0.005245477892458439,
+      "grad_norm": 0.0046915169805288315,
       "learning_rate": 0.0001973817497477129,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 73
     },
     {
       "epoch": 0.0026477270694312753,
-      "grad_norm": 0.022848384454846382,
+      "grad_norm": 0.25393474102020264,
       "learning_rate": 0.00019730985109821266,
-      "loss": 0.0005,
+      "loss": 0.0027,
       "step": 74
     },
     {
       "epoch": 0.0026835071649641304,
-      "grad_norm": 0.03462355583906174,
+      "grad_norm": 0.007773595862090588,
       "learning_rate": 0.00019723699203976766,
-      "loss": 0.0006,
+      "loss": 0.0004,
       "step": 75
     },
     {
       "epoch": 0.0027192872604969854,
-      "grad_norm": 0.005384132731705904,
+      "grad_norm": 0.004822044633328915,
       "learning_rate": 0.0001971631732914674,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 76
     },
     {
       "epoch": 0.0027550673560298405,
-      "grad_norm": 0.007270114030689001,
+      "grad_norm": 0.003923391457647085,
       "learning_rate": 0.0001970883955818731,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 77
     },
     {
       "epoch": 0.0027908474515626956,
-      "grad_norm": 0.005563640967011452,
+      "grad_norm": 0.016484705731272697,
       "learning_rate": 0.0001970126596490106,
-      "loss": 0.0003,
+      "loss": 0.0006,
       "step": 78
     },
     {
       "epoch": 0.0028266275470955507,
-      "grad_norm": 0.00865593459457159,
+      "grad_norm": 0.046882666647434235,
       "learning_rate": 0.00019693596624036292,
-      "loss": 0.0003,
+      "loss": 0.0007,
       "step": 79
     },
     {
       "epoch": 0.0028624076426284058,
-      "grad_norm": 0.003771064803004265,
+      "grad_norm": 0.0036497802939265966,
       "learning_rate": 0.0001968583161128631,
       "loss": 0.0002,
       "step": 80
     },
     {
       "epoch": 0.002898187738161261,
-      "grad_norm": 0.004731202498078346,
+      "grad_norm": 0.0067147244699299335,
       "learning_rate": 0.00019677971003288655,
       "loss": 0.0003,
       "step": 81
     },
     {
       "epoch": 0.002933967833694116,
-      "grad_norm": 0.004008135758340359,
+      "grad_norm": 0.00465575372800231,
       "learning_rate": 0.00019670014877624353,
       "loss": 0.0002,
       "step": 82
     },
     {
       "epoch": 0.002969747929226971,
-      "grad_norm": 0.06694934517145157,
+      "grad_norm": 0.0572395846247673,
       "learning_rate": 0.00019661963312817148,
-      "loss": 0.0006,
+      "loss": 0.001,
       "step": 83
     },
     {
       "epoch": 0.003005528024759826,
-      "grad_norm": 0.004304195754230022,
+      "grad_norm": 0.006505393423140049,
       "learning_rate": 0.0001965381638833274,
-      "loss": 0.0002,
+      "loss": 0.0003,
       "step": 84
     },
     {
       "epoch": 0.003041308120292681,
-      "grad_norm": 0.003030312480404973,
+      "grad_norm": 0.005373773165047169,
       "learning_rate": 0.00019645574184577982,
       "loss": 0.0002,
       "step": 85
     },
     {
       "epoch": 0.0030770882158255362,
-      "grad_norm": 0.006374573800712824,
+      "grad_norm": 0.028440192341804504,
       "learning_rate": 0.000196372367829001,
-      "loss": 0.0002,
+      "loss": 0.0003,
       "step": 86
     },
     {
       "epoch": 0.0031128683113583913,
-      "grad_norm": 0.006917024031281471,
+      "grad_norm": 0.021315911784768105,
       "learning_rate": 0.00019628804265585877,
-      "loss": 0.0003,
+      "loss": 0.0005,
       "step": 87
     },
     {
       "epoch": 0.0031486484068912464,
-      "grad_norm": 5.839280605316162,
+      "grad_norm": 0.0548127144575119,
       "learning_rate": 0.0001962027671586086,
-      "loss": 0.1345,
+      "loss": 0.0004,
       "step": 88
     },
     {
       "epoch": 0.0031844285024241015,
-      "grad_norm": 0.0044294423423707485,
+      "grad_norm": 0.026664718985557556,
       "learning_rate": 0.0001961165421788852,
-      "loss": 0.0002,
+      "loss": 0.0004,
       "step": 89
     },
     {
       "epoch": 0.0032202085979569565,
-      "grad_norm": 0.0021441711578518152,
+      "grad_norm": 0.0035014627501368523,
       "learning_rate": 0.0001960293685676943,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 90
     },
     {
       "epoch": 0.0032559886934898116,
-      "grad_norm": 2.4447829723358154,
+      "grad_norm": 2.119551181793213,
       "learning_rate": 0.0001959412471854043,
-      "loss": 0.0337,
+      "loss": 0.0222,
       "step": 91
     },
     {
       "epoch": 0.0032917687890226667,
-      "grad_norm": 0.08363230526447296,
+      "grad_norm": 0.10606224089860916,
       "learning_rate": 0.0001958521789017376,
-      "loss": 0.0006,
+      "loss": 0.0009,
       "step": 92
     },
     {
       "epoch": 0.0033275488845555218,
-      "grad_norm": 0.003032706445083022,
+      "grad_norm": 0.0029525908175855875,
       "learning_rate": 0.00019576216459576222,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 93
     },
     {
       "epoch": 0.003363328980088377,
-      "grad_norm": 0.004137410782277584,
+      "grad_norm": 0.003076421096920967,
       "learning_rate": 0.00019567120515588308,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 94
     },
     {
       "epoch": 0.003399109075621232,
-      "grad_norm": 0.059599414467811584,
+      "grad_norm": 0.021403349936008453,
       "learning_rate": 0.00019557930147983302,
-      "loss": 0.0009,
+      "loss": 0.0004,
       "step": 95
     },
     {
       "epoch": 0.003434889171154087,
-      "grad_norm": 0.0038637579418718815,
+      "grad_norm": 0.001615760731510818,
       "learning_rate": 0.00019548645447466431,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 96
     },
     {
       "epoch": 0.003470669266686942,
-      "grad_norm": 0.004651807248592377,
+      "grad_norm": 0.0412115603685379,
       "learning_rate": 0.00019539266505673938,
-      "loss": 0.0003,
+      "loss": 0.0002,
       "step": 97
     },
     {
       "epoch": 0.003506449362219797,
-      "grad_norm": 0.014560009352862835,
+      "grad_norm": 0.606522262096405,
       "learning_rate": 0.00019529793415172192,
-      "loss": 0.0003,
+      "loss": 0.0036,
       "step": 98
     },
     {
       "epoch": 0.0035422294577526522,
-      "grad_norm": 0.005952389445155859,
+      "grad_norm": 7.513233184814453,
       "learning_rate": 0.00019520226269456768,
-      "loss": 0.0003,
+      "loss": 0.0428,
       "step": 99
     },
     {
       "epoch": 0.0035780095532855073,
-      "grad_norm": 0.002610633848235011,
+      "grad_norm": 0.0012378046521916986,
       "learning_rate": 0.00019510565162951537,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 100
     },
     {
       "epoch": 0.0036137896488183624,
-      "grad_norm": 0.004542678594589233,
+      "grad_norm": 0.04065516218543053,
       "learning_rate": 0.00019500810191007718,
-      "loss": 0.0003,
+      "loss": 0.0005,
       "step": 101
     },
     {
       "epoch": 0.0036495697443512175,
-      "grad_norm": 0.10543578863143921,
+      "grad_norm": 0.4646848440170288,
       "learning_rate": 0.00019490961449902946,
-      "loss": 0.0009,
+      "loss": 0.0041,
       "step": 102
     },
     {
       "epoch": 0.0036853498398840725,
-      "grad_norm": 0.004441316705197096,
+      "grad_norm": 0.0042214663699269295,
       "learning_rate": 0.0001948101903684032,
-      "loss": 0.0003,
+      "loss": 0.0002,
       "step": 103
     },
     {
       "epoch": 0.0037211299354169276,
-      "grad_norm": 0.18022280931472778,
+      "grad_norm": 0.3199816346168518,
       "learning_rate": 0.00019470983049947444,
-      "loss": 0.0013,
+      "loss": 0.0014,
       "step": 104
     },
     {
       "epoch": 0.0037569100309497827,
-      "grad_norm": 0.10101974010467529,
+      "grad_norm": 0.0463225431740284,
       "learning_rate": 0.00019460853588275454,
-      "loss": 0.0026,
+      "loss": 0.0009,
       "step": 105
     },
     {
       "epoch": 0.0037926901264826378,
-      "grad_norm": 1.7630839347839355,
+      "grad_norm": 0.31589099764823914,
       "learning_rate": 0.00019450630751798048,
-      "loss": 0.0132,
+      "loss": 0.0017,
       "step": 106
     },
     {
       "epoch": 0.003828470222015493,
-      "grad_norm": 0.003979897126555443,
+      "grad_norm": 0.16430477797985077,
       "learning_rate": 0.000194403146414105,
-      "loss": 0.0002,
+      "loss": 0.0012,
       "step": 107
     },
     {
       "epoch": 0.003864250317548348,
-      "grad_norm": 0.007163205649703741,
+      "grad_norm": 0.8777446746826172,
       "learning_rate": 0.00019429905358928646,
-      "loss": 0.0002,
+      "loss": 0.0065,
       "step": 108
     },
     {
       "epoch": 0.003900030413081203,
-      "grad_norm": 0.07740266621112823,
+      "grad_norm": 0.3933359682559967,
       "learning_rate": 0.00019419403007087907,
-      "loss": 0.0016,
+      "loss": 0.0026,
       "step": 109
     },
     {
       "epoch": 0.003935810508614058,
-      "grad_norm": 0.016551105305552483,
+      "grad_norm": 0.0034663318656384945,
       "learning_rate": 0.00019408807689542257,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 110
     },
     {
       "epoch": 0.003971590604146913,
-      "grad_norm": 0.007171209901571274,
+      "grad_norm": 0.0011069366009905934,
       "learning_rate": 0.00019398119510863197,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 111
     },
     {
       "epoch": 0.004007370699679768,
-      "grad_norm": 0.004873023834079504,
+      "grad_norm": 0.0016587006393820047,
       "learning_rate": 0.00019387338576538744,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 112
     },
     {
       "epoch": 0.004043150795212623,
-      "grad_norm": 0.057489000260829926,
+      "grad_norm": 0.002230650745332241,
       "learning_rate": 0.00019376464992972356,
-      "loss": 0.0004,
+      "loss": 0.0001,
       "step": 113
     },
     {
       "epoch": 0.004078930890745478,
-      "grad_norm": 0.0026130485348403454,
+      "grad_norm": 0.000850948563311249,
       "learning_rate": 0.00019365498867481923,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 114
     },
     {
       "epoch": 0.004114710986278333,
-      "grad_norm": 0.01955575682222843,
+      "grad_norm": 0.0022517703473567963,
       "learning_rate": 0.00019354440308298675,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 115
     },
     {
       "epoch": 0.004150491081811188,
-      "grad_norm": 0.07882912456989288,
+      "grad_norm": 0.5699942708015442,
       "learning_rate": 0.00019343289424566122,
-      "loss": 0.0012,
+      "loss": 0.0051,
       "step": 116
     },
     {
       "epoch": 0.004186271177344043,
-      "grad_norm": 0.011640047654509544,
+      "grad_norm": 0.1059614047408104,
       "learning_rate": 0.00019332046326338986,
-      "loss": 0.0002,
+      "loss": 0.0007,
       "step": 117
     },
     {
       "epoch": 0.004222051272876898,
-      "grad_norm": 0.0029720300808548927,
+      "grad_norm": 0.001039605587720871,
       "learning_rate": 0.0001932071112458211,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 118
     },
     {
       "epoch": 0.004257831368409753,
-      "grad_norm": 0.001956366468220949,
+      "grad_norm": 0.004336313344538212,
       "learning_rate": 0.00019309283931169356,
       "loss": 0.0001,
       "step": 119
     },
     {
       "epoch": 0.004293611463942608,
-      "grad_norm": 0.01011357270181179,
+      "grad_norm": 0.683163583278656,
       "learning_rate": 0.00019297764858882514,
-      "loss": 0.0002,
+      "loss": 0.0033,
       "step": 120
     },
     {
       "epoch": 0.0043293915594754635,
-      "grad_norm": 0.0019783275201916695,
+      "grad_norm": 0.0009113202686421573,
       "learning_rate": 0.00019286154021410173,
       "loss": 0.0001,
       "step": 121
     },
     {
       "epoch": 0.004365171655008319,
-      "grad_norm": 0.002392844995483756,
+      "grad_norm": 0.0011711944825947285,
       "learning_rate": 0.00019274451533346615,
       "loss": 0.0001,
       "step": 122
     },
     {
       "epoch": 0.004400951750541174,
-      "grad_norm": 0.006437376607209444,
+      "grad_norm": 0.0020785066299140453,
       "learning_rate": 0.00019262657510190666,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 123
     },
     {
       "epoch": 0.004436731846074029,
-      "grad_norm": 0.0031083389185369015,
+      "grad_norm": 0.5635930299758911,
       "learning_rate": 0.0001925077206834458,
-      "loss": 0.0001,
+      "loss": 0.003,
       "step": 124
     },
     {
       "epoch": 0.004472511941606884,
-      "grad_norm": 0.0392051637172699,
+      "grad_norm": 0.019610170274972916,
       "learning_rate": 0.0001923879532511287,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 125
     },
     {
       "epoch": 0.004508292037139739,
-      "grad_norm": 0.003594709560275078,
+      "grad_norm": 0.003800542326644063,
       "learning_rate": 0.0001922672739870115,
       "loss": 0.0001,
       "step": 126
     },
     {
       "epoch": 0.004544072132672594,
-      "grad_norm": 0.001110642566345632,
+      "grad_norm": 0.0015734516782686114,
       "learning_rate": 0.00019214568408214985,
       "loss": 0.0001,
       "step": 127
     },
     {
       "epoch": 0.004579852228205449,
-      "grad_norm": 0.4432847499847412,
+      "grad_norm": 0.16732634603977203,
       "learning_rate": 0.00019202318473658705,
-      "loss": 0.0027,
+      "loss": 0.0014,
       "step": 128
     },
     {
       "epoch": 0.004615632323738304,
-      "grad_norm": 0.10707289725542068,
+      "grad_norm": 0.03286750242114067,
       "learning_rate": 0.00019189977715934213,
       "loss": 0.0004,
       "step": 129
     },
     {
       "epoch": 0.004651412419271159,
-      "grad_norm": 0.28691592812538147,
+      "grad_norm": 0.013986490666866302,
       "learning_rate": 0.00019177546256839812,
-      "loss": 0.0018,
+      "loss": 0.0003,
       "step": 130
     },
     {
       "epoch": 0.004687192514804014,
-      "grad_norm": 0.0006904619513079524,
+      "grad_norm": 0.0010463737417012453,
       "learning_rate": 0.0001916502421906898,
       "loss": 0.0001,
       "step": 131
     },
     {
       "epoch": 0.004722972610336869,
-      "grad_norm": 0.0038885704707354307,
+      "grad_norm": 0.05846157670021057,
       "learning_rate": 0.00019152411726209176,
-      "loss": 0.0001,
+      "loss": 0.0004,
       "step": 132
     },
     {
       "epoch": 0.004758752705869724,
-      "grad_norm": 0.0012742049293592572,
+      "grad_norm": 0.0012616278836503625,
       "learning_rate": 0.00019139708902740613,
       "loss": 0.0001,
       "step": 133
     },
     {
       "epoch": 0.0047945328014025795,
-      "grad_norm": 0.0069363294169306755,
+      "grad_norm": 0.002735038986429572,
       "learning_rate": 0.0001912691587403503,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 134
     },
     {
       "epoch": 0.004830312896935435,
-      "grad_norm": 0.14127081632614136,
+      "grad_norm": 0.05025576055049896,
       "learning_rate": 0.00019114032766354453,
-      "loss": 0.0011,
+      "loss": 0.0004,
       "step": 135
     },
     {
       "epoch": 0.00486609299246829,
-      "grad_norm": 0.006388662848621607,
+      "grad_norm": 0.00338102038949728,
       "learning_rate": 0.00019101059706849957,
       "loss": 0.0002,
       "step": 136
     },
     {
       "epoch": 0.004901873088001145,
-      "grad_norm": 0.017692791298031807,
+      "grad_norm": 0.016932694241404533,
       "learning_rate": 0.00019087996823560402,
       "loss": 0.0003,
       "step": 137
     },
     {
       "epoch": 0.004937653183534,
-      "grad_norm": 0.27376821637153625,
+      "grad_norm": 0.02689458057284355,
       "learning_rate": 0.0001907484424541117,
-      "loss": 0.0008,
+      "loss": 0.0004,
       "step": 138
     },
     {
       "epoch": 0.004973433279066855,
-      "grad_norm": 0.002759058028459549,
+      "grad_norm": 0.4155731201171875,
       "learning_rate": 0.00019061602102212898,
-      "loss": 0.0001,
+      "loss": 0.0043,
       "step": 139
     },
     {
       "epoch": 0.00500921337459971,
-      "grad_norm": 0.0010609851451590657,
+      "grad_norm": 0.10929184406995773,
       "learning_rate": 0.00019048270524660196,
-      "loss": 0.0001,
+      "loss": 0.0005,
       "step": 140
     },
     {
       "epoch": 0.005044993470132565,
-      "grad_norm": 0.01087561808526516,
+      "grad_norm": 0.011052286252379417,
       "learning_rate": 0.0001903484964433035,
       "loss": 0.0001,
       "step": 141
     },
     {
       "epoch": 0.00508077356566542,
-      "grad_norm": 0.11954966932535172,
+      "grad_norm": 0.01035118568688631,
       "learning_rate": 0.00019021339593682028,
-      "loss": 0.0007,
+      "loss": 0.0003,
       "step": 142
     },
     {
       "epoch": 0.005116553661198275,
-      "grad_norm": 0.004676848649978638,
+      "grad_norm": 0.019157446920871735,
       "learning_rate": 0.00019007740506053983,
       "loss": 0.0003,
       "step": 143
     },
     {
       "epoch": 0.00515233375673113,
-      "grad_norm": 0.004206379409879446,
+      "grad_norm": 0.0006221545045264065,
       "learning_rate": 0.0001899405251566371,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 144
     },
     {
       "epoch": 0.005188113852263985,
-      "grad_norm": 0.0014629610814154148,
+      "grad_norm": 0.027849700301885605,
       "learning_rate": 0.00018980275757606157,
-      "loss": 0.0001,
+      "loss": 0.0003,
       "step": 145
     },
     {
       "epoch": 0.00522389394779684,
-      "grad_norm": 0.0006727775908075273,
+      "grad_norm": 0.001835379982367158,
       "learning_rate": 0.00018966410367852362,
       "loss": 0.0001,
       "step": 146
     },
     {
       "epoch": 0.0052596740433296955,
-      "grad_norm": 0.023915646597743034,
+      "grad_norm": 1.2824780941009521,
       "learning_rate": 0.00018952456483248119,
-      "loss": 0.0002,
+      "loss": 0.0366,
       "step": 147
     },
     {
       "epoch": 0.005295454138862551,
-      "grad_norm": 0.003377806395292282,
+      "grad_norm": 0.025991423055529594,
       "learning_rate": 0.0001893841424151264,
-      "loss": 0.0002,
+      "loss": 0.0004,
       "step": 148
     },
     {
       "epoch": 0.005331234234395406,
-      "grad_norm": 0.002208555582910776,
+      "grad_norm": 0.0042706448584795,
       "learning_rate": 0.0001892428378123718,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 149
     },
     {
       "epoch": 0.005367014329928261,
-      "grad_norm": 0.000618081830907613,
+      "grad_norm": 0.0003646000404842198,
       "learning_rate": 0.0001891006524188368,
       "loss": 0.0,
       "step": 150
     },
     {
       "epoch": 0.005402794425461116,
-      "grad_norm": 0.0016522255027666688,
+      "grad_norm": 0.002723448909819126,
       "learning_rate": 0.00018895758763783383,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 151
     },
     {
       "epoch": 0.005438574520993971,
-      "grad_norm": 0.0025217332877218723,
+      "grad_norm": 0.0007927878759801388,
       "learning_rate": 0.00018881364488135448,
       "loss": 0.0001,
       "step": 152
     },
     {
       "epoch": 0.005474354616526826,
-      "grad_norm": 0.001731122494675219,
+      "grad_norm": 0.0023497738875448704,
       "learning_rate": 0.00018866882557005567,
       "loss": 0.0001,
       "step": 153
     },
     {
       "epoch": 0.005510134712059681,
-      "grad_norm": 0.0008009941084310412,
+      "grad_norm": 0.00858394242823124,
       "learning_rate": 0.00018852313113324552,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 154
     },
     {
       "epoch": 0.005545914807592536,
-      "grad_norm": 0.0014807975385338068,
+      "grad_norm": 0.02338087372481823,
       "learning_rate": 0.00018837656300886937,
-      "loss": 0.0001,
+      "loss": 0.0003,
       "step": 155
     },
     {
       "epoch": 0.005581694903125391,
-      "grad_norm": 0.14768290519714355,
+      "grad_norm": 0.006292372476309538,
       "learning_rate": 0.00018822912264349534,
-      "loss": 0.0021,
+      "loss": 0.0002,
       "step": 156
     },
     {
       "epoch": 0.005617474998658246,
-      "grad_norm": 0.22848215699195862,
+      "grad_norm": 0.1395585685968399,
       "learning_rate": 0.00018808081149230036,
-      "loss": 0.0029,
+      "loss": 0.002,
       "step": 157
     },
     {
       "epoch": 0.005653255094191101,
-      "grad_norm": 0.00042707924148999155,
+      "grad_norm": 0.06656046211719513,
       "learning_rate": 0.00018793163101905563,
-      "loss": 0.0,
+      "loss": 0.0005,
       "step": 158
     },
     {
       "epoch": 0.0056890351897239564,
-      "grad_norm": 0.02177448756992817,
+      "grad_norm": 0.005269424058496952,
       "learning_rate": 0.00018778158269611218,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 159
     },
     {
       "epoch": 0.0057248152852568115,
-      "grad_norm": 0.0023135808296501637,
+      "grad_norm": 0.005869029089808464,
       "learning_rate": 0.00018763066800438636,
       "loss": 0.0001,
       "step": 160
     },
     {
       "epoch": 0.005760595380789667,
-      "grad_norm": 0.0023537538945674896,
+      "grad_norm": 0.007193957921117544,
       "learning_rate": 0.0001874788884333453,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 161
     },
     {
       "epoch": 0.005796375476322522,
-      "grad_norm": 0.0006432220106944442,
+      "grad_norm": 0.23244933784008026,
       "learning_rate": 0.00018732624548099204,
-      "loss": 0.0001,
+      "loss": 0.0015,
       "step": 162
     },
     {
       "epoch": 0.005832155571855377,
-      "grad_norm": 0.0044472129084169865,
+      "grad_norm": 0.0010494483867660165,
       "learning_rate": 0.0001871727406538509,
       "loss": 0.0001,
       "step": 163
     },
     {
       "epoch": 0.005867935667388232,
-      "grad_norm": 0.03199806436896324,
+      "grad_norm": 0.001451204065233469,
       "learning_rate": 0.0001870183754669526,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 164
     },
     {
       "epoch": 0.005903715762921087,
-      "grad_norm": 0.01239110715687275,
+      "grad_norm": 0.0035805683583021164,
       "learning_rate": 0.00018686315144381913,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 165
     },
     {
       "epoch": 0.005939495858453942,
-      "grad_norm": 0.002915807766839862,
+      "grad_norm": 0.003529100678861141,
       "learning_rate": 0.000186707070116449,
       "loss": 0.0002,
       "step": 166
     },
     {
       "epoch": 0.005975275953986797,
-      "grad_norm": 0.1886855959892273,
+      "grad_norm": 0.0062582106329500675,
       "learning_rate": 0.0001865501330253019,
-      "loss": 0.0005,
+      "loss": 0.0002,
       "step": 167
     },
     {
       "epoch": 0.006011056049519652,
-      "grad_norm": 2.766813039779663,
+      "grad_norm": 2.6403636932373047,
       "learning_rate": 0.00018639234171928353,
-      "loss": 0.1067,
+      "loss": 0.0772,
       "step": 168
     },
     {
       "epoch": 0.006046836145052507,
-      "grad_norm": 0.002398906508460641,
+      "grad_norm": 3.0866150856018066,
       "learning_rate": 0.0001862336977557304,
-      "loss": 0.0001,
+      "loss": 0.0508,
       "step": 169
     },
     {
       "epoch": 0.006082616240585362,
-      "grad_norm": 0.0011222879402339458,
+      "grad_norm": 0.00488316873088479,
       "learning_rate": 0.0001860742027003944,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 170
     },
     {
       "epoch": 0.006118396336118217,
-      "grad_norm": 0.0013398065930232406,
+      "grad_norm": 0.005987066309899092,
       "learning_rate": 0.00018591385812742725,
       "loss": 0.0001,
       "step": 171
     },
     {
       "epoch": 0.0061541764316510724,
-      "grad_norm": 1.195412516593933,
+      "grad_norm": 0.002021647058427334,
       "learning_rate": 0.00018575266561936523,
-      "loss": 0.0694,
+      "loss": 0.0001,
       "step": 172
     },
     {
       "epoch": 0.0061899565271839275,
-      "grad_norm": 0.015319638885557652,
+      "grad_norm": 0.005465734284371138,
       "learning_rate": 0.00018559062676711332,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 173
     },
     {
       "epoch": 0.006225736622716783,
-      "grad_norm": 0.39182984828948975,
+      "grad_norm": 0.04289950057864189,
       "learning_rate": 0.0001854277431699295,
-      "loss": 0.0025,
+      "loss": 0.0003,
       "step": 174
     },
     {
       "epoch": 0.006261516718249638,
-      "grad_norm": 0.008146129548549652,
+      "grad_norm": 0.0013694529188796878,
       "learning_rate": 0.00018526401643540922,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 175
     },
     {
       "epoch": 0.006297296813782493,
-      "grad_norm": 0.05075721815228462,
+      "grad_norm": 0.005164470989257097,
       "learning_rate": 0.00018509944817946922,
-      "loss": 0.0008,
+      "loss": 0.0001,
       "step": 176
     },
     {
       "epoch": 0.006333076909315348,
-      "grad_norm": 0.015434350818395615,
+      "grad_norm": 0.006279651075601578,
       "learning_rate": 0.00018493404002633166,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 177
     },
     {
       "epoch": 0.006368857004848203,
-      "grad_norm": 0.036245349794626236,
+      "grad_norm": 0.01261226087808609,
       "learning_rate": 0.00018476779360850832,
-      "loss": 0.0007,
+      "loss": 0.0002,
       "step": 178
     },
     {
       "epoch": 0.006404637100381058,
-      "grad_norm": 0.005251400172710419,
+      "grad_norm": 0.0049860430881381035,
       "learning_rate": 0.00018460071056678422,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 179
     },
     {
       "epoch": 0.006440417195913913,
-      "grad_norm": 0.2937351167201996,
+      "grad_norm": 0.06854995340108871,
       "learning_rate": 0.00018443279255020152,
-      "loss": 0.0052,
+      "loss": 0.0007,
       "step": 180
     },
     {
       "epoch": 0.006476197291446768,
-      "grad_norm": 0.1004151776432991,
+      "grad_norm": 0.0030735009349882603,
       "learning_rate": 0.00018426404121604323,
-      "loss": 0.0009,
+      "loss": 0.0001,
       "step": 181
     },
     {
       "epoch": 0.006511977386979623,
-      "grad_norm": 0.014370848424732685,
+      "grad_norm": 0.06162787973880768,
       "learning_rate": 0.00018409445822981693,
-      "loss": 0.0006,
+      "loss": 0.0008,
       "step": 182
     },
     {
       "epoch": 0.006547757482512478,
-      "grad_norm": 0.030721843242645264,
+      "grad_norm": 1.4743658304214478,
       "learning_rate": 0.00018392404526523817,
-      "loss": 0.0009,
+      "loss": 0.036,
       "step": 183
     },
     {
       "epoch": 0.006583537578045333,
-      "grad_norm": 0.004438372328877449,
+      "grad_norm": 0.0014406999107450247,
       "learning_rate": 0.0001837528040042142,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 184
     },
     {
       "epoch": 0.0066193176735781884,
-      "grad_norm": 0.00464064022526145,
+      "grad_norm": 0.001721672248095274,
       "learning_rate": 0.00018358073613682706,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 185
     },
     {
       "epoch": 0.0066550977691110435,
-      "grad_norm": 0.27975496649742126,
+      "grad_norm": 0.007810389623045921,
       "learning_rate": 0.00018340784336131713,
-      "loss": 0.0019,
+      "loss": 0.0003,
       "step": 186
     },
     {
       "epoch": 0.006690877864643899,
-      "grad_norm": 0.004852125886827707,
+      "grad_norm": 0.007608585525304079,
       "learning_rate": 0.00018323412738406635,
       "loss": 0.0003,
       "step": 187
     },
     {
       "epoch": 0.006726657960176754,
-      "grad_norm": 0.004563230089843273,
+      "grad_norm": 0.007065481040626764,
       "learning_rate": 0.00018305958991958127,
       "loss": 0.0003,
       "step": 188
     },
     {
       "epoch": 0.006762438055709609,
-      "grad_norm": 0.005834383424371481,
+      "grad_norm": 0.03384295850992203,
       "learning_rate": 0.0001828842326904762,
-      "loss": 0.0004,
+      "loss": 0.0006,
       "step": 189
     },
     {
       "epoch": 0.006798218151242464,
-      "grad_norm": 0.8059268593788147,
+      "grad_norm": 0.3284815847873688,
       "learning_rate": 0.00018270805742745617,
-      "loss": 0.0189,
+      "loss": 0.0059,
       "step": 190
     },
     {
       "epoch": 0.006833998246775319,
-      "grad_norm": 0.012366785667836666,
+      "grad_norm": 0.029190680012106895,
       "learning_rate": 0.00018253106586929997,
-      "loss": 0.0005,
+      "loss": 0.0006,
       "step": 191
     },
     {
       "epoch": 0.006869778342308174,
-      "grad_norm": 0.045971453189849854,
+      "grad_norm": 0.9759858846664429,
       "learning_rate": 0.00018235325976284275,
-      "loss": 0.0008,
+      "loss": 0.147,
       "step": 192
     },
     {
       "epoch": 0.006905558437841029,
-      "grad_norm": 0.007568427361547947,
+      "grad_norm": 0.02130061574280262,
       "learning_rate": 0.00018217464086295904,
-      "loss": 0.0004,
+      "loss": 0.0008,
       "step": 193
     },
     {
       "epoch": 0.006941338533373884,
-      "grad_norm": 0.0033736741170287132,
+      "grad_norm": 0.011916525661945343,
       "learning_rate": 0.00018199521093254523,
-      "loss": 0.0002,
+      "loss": 0.0005,
       "step": 194
     },
     {
       "epoch": 0.006977118628906739,
-      "grad_norm": 0.002627680543810129,
+      "grad_norm": 0.010565850883722305,
       "learning_rate": 0.00018181497174250236,
-      "loss": 0.0002,
+      "loss": 0.0006,
       "step": 195
     },
     {
       "epoch": 0.007012898724439594,
-      "grad_norm": 0.0065056337043643,
+      "grad_norm": 0.022635890170931816,
       "learning_rate": 0.00018163392507171842,
-      "loss": 0.0003,
+      "loss": 0.0013,
       "step": 196
     },
     {
       "epoch": 0.007048678819972449,
-      "grad_norm": 0.005994074046611786,
+      "grad_norm": 1.2287172079086304,
       "learning_rate": 0.00018145207270705096,
-      "loss": 0.0002,
+      "loss": 0.0092,
       "step": 197
     },
     {
       "epoch": 0.0070844589155053045,
-      "grad_norm": 1.088179349899292,
+      "grad_norm": 0.27400410175323486,
       "learning_rate": 0.0001812694164433094,
-      "loss": 0.0661,
+      "loss": 0.0099,
       "step": 198
     },
     {
       "epoch": 0.0071202390110381595,
-      "grad_norm": 0.0036484820302575827,
+      "grad_norm": 0.014746950939297676,
       "learning_rate": 0.00018108595808323736,
-      "loss": 0.0002,
+      "loss": 0.0006,
       "step": 199
     },
     {
       "epoch": 0.007156019106571015,
-      "grad_norm": 1.7597136497497559,
+      "grad_norm": 0.13460226356983185,
       "learning_rate": 0.00018090169943749476,
-      "loss": 0.0502,
+      "loss": 0.0215,
       "step": 200
     },
     {
       "epoch": 0.00719179920210387,
-      "grad_norm": 0.002321377396583557,
+      "grad_norm": 0.0137384794652462,
       "learning_rate": 0.00018071664232464002,
-      "loss": 0.0002,
+      "loss": 0.001,
       "step": 201
     },
     {
       "epoch": 0.007227579297636725,
-      "grad_norm": 0.004088441841304302,
+      "grad_norm": 0.47047391533851624,
       "learning_rate": 0.0001805307885711122,
-      "loss": 0.0003,
+      "loss": 0.0075,
       "step": 202
     },
     {
       "epoch": 0.00726335939316958,
-      "grad_norm": 0.0046237860806286335,
+      "grad_norm": 0.019856106489896774,
       "learning_rate": 0.00018034414001121278,
-      "loss": 0.0003,
+      "loss": 0.0011,
       "step": 203
     },
     {
       "epoch": 0.007299139488702435,
-      "grad_norm": 0.017054418101906776,
+      "grad_norm": 0.01606709137558937,
       "learning_rate": 0.00018015669848708767,
-      "loss": 0.0007,
+      "loss": 0.0012,
       "step": 204
     },
     {
       "epoch": 0.00733491958423529,
-      "grad_norm": 0.13485275208950043,
+      "grad_norm": 0.01651693508028984,
       "learning_rate": 0.00017996846584870908,
-      "loss": 0.0025,
+      "loss": 0.0008,
       "step": 205
     },
     {
       "epoch": 0.007370699679768145,
-      "grad_norm": 0.018125206232070923,
+      "grad_norm": 0.009905004873871803,
       "learning_rate": 0.0001797794439538571,
       "loss": 0.0007,
       "step": 206
     },
     {
       "epoch": 0.007406479775301,
-      "grad_norm": 0.02075115777552128,
+      "grad_norm": 0.013038737699389458,
       "learning_rate": 0.0001795896346681016,
-      "loss": 0.0009,
+      "loss": 0.0008,
       "step": 207
     },
     {
       "epoch": 0.007442259870833855,
-      "grad_norm": 0.015594271942973137,
+      "grad_norm": 0.009600020945072174,
       "learning_rate": 0.00017939903986478355,
-      "loss": 0.0006,
+      "loss": 0.0007,
       "step": 208
     },
     {
       "epoch": 0.00747803996636671,
-      "grad_norm": 0.01755349524319172,
+      "grad_norm": 0.005098757334053516,
       "learning_rate": 0.00017920766142499672,
-      "loss": 0.0006,
+      "loss": 0.0004,
       "step": 209
     },
     {
       "epoch": 0.007513820061899565,
-      "grad_norm": 0.03347839415073395,
+      "grad_norm": 0.005413730163127184,
       "learning_rate": 0.00017901550123756906,
-      "loss": 0.0006,
+      "loss": 0.0004,
       "step": 210
     },
     {
       "epoch": 0.0075496001574324205,
-      "grad_norm": 0.012991059571504593,
+      "grad_norm": 0.005524770822376013,
       "learning_rate": 0.00017882256119904403,
-      "loss": 0.0006,
+      "loss": 0.0004,
       "step": 211
     },
     {
       "epoch": 0.0075853802529652755,
-      "grad_norm": 0.1412995159626007,
+      "grad_norm": 0.02472483552992344,
       "learning_rate": 0.00017862884321366188,
-      "loss": 0.0034,
+      "loss": 0.0009,
       "step": 212
     },
     {
       "epoch": 0.007621160348498131,
-      "grad_norm": 0.02368590049445629,
+      "grad_norm": 0.004835169296711683,
       "learning_rate": 0.000178434349193341,
-      "loss": 0.001,
+      "loss": 0.0004,
       "step": 213
     },
     {
       "epoch": 0.007656940444030986,
-      "grad_norm": 0.008637255989015102,
+      "grad_norm": 0.005667371209710836,
       "learning_rate": 0.0001782390810576588,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 214
     },
     {
       "epoch": 0.007692720539563841,
-      "grad_norm": 0.1594046950340271,
+      "grad_norm": 0.005504743196070194,
       "learning_rate": 0.000178043040733833,
-      "loss": 0.0016,
+      "loss": 0.0003,
       "step": 215
     },
     {
       "epoch": 0.007728500635096696,
-      "grad_norm": 1.6184616088867188,
+      "grad_norm": 0.004330215975642204,
       "learning_rate": 0.00017784623015670238,
-      "loss": 0.0365,
+      "loss": 0.0003,
       "step": 216
     },
     {
       "epoch": 0.007764280730629551,
-      "grad_norm": 0.6205570697784424,
+      "grad_norm": 0.008505130186676979,
       "learning_rate": 0.00017764865126870786,
-      "loss": 0.025,
+      "loss": 0.0006,
       "step": 217
     },
     {
       "epoch": 0.007800060826162406,
-      "grad_norm": 0.00536901643499732,
+      "grad_norm": 0.005010291468352079,
       "learning_rate": 0.00017745030601987337,
-      "loss": 0.0004,
+      "loss": 0.0003,
       "step": 218
     },
     {
       "epoch": 0.00783584092169526,
-      "grad_norm": 0.008372887037694454,
+      "grad_norm": 0.003250179113820195,
       "learning_rate": 0.00017725119636778644,
-      "loss": 0.0006,
+      "loss": 0.0002,
       "step": 219
     },
     {
       "epoch": 0.007871621017228115,
-      "grad_norm": 0.17965705692768097,
+      "grad_norm": 0.00220483704470098,
       "learning_rate": 0.00017705132427757895,
-      "loss": 0.0035,
+      "loss": 0.0002,
       "step": 220
     },
     {
       "epoch": 0.00790740111276097,
-      "grad_norm": 0.020489655435085297,
+      "grad_norm": 0.02156749553978443,
       "learning_rate": 0.00017685069172190766,
-      "loss": 0.0009,
+      "loss": 0.0003,
       "step": 221
     },
     {
       "epoch": 0.007943181208293825,
-      "grad_norm": 0.16530613601207733,
+      "grad_norm": 0.00591958686709404,
       "learning_rate": 0.00017664930068093498,
-      "loss": 0.002,
+      "loss": 0.0002,
       "step": 222
     },
     {
       "epoch": 0.00797896130382668,
-      "grad_norm": 0.14881941676139832,
+      "grad_norm": 0.01073639839887619,
       "learning_rate": 0.00017644715314230918,
-      "loss": 0.0029,
+      "loss": 0.0004,
       "step": 223
     },
     {
       "epoch": 0.008014741399359536,
-      "grad_norm": 0.009059658274054527,
+      "grad_norm": 0.007549067493528128,
       "learning_rate": 0.0001762442511011448,
-      "loss": 0.0006,
+      "loss": 0.0002,
       "step": 224
     },
     {
       "epoch": 0.00805052149489239,
-      "grad_norm": 0.005003109108656645,
+      "grad_norm": 0.0013972257729619741,
       "learning_rate": 0.0001760405965600031,
-      "loss": 0.0004,
+      "loss": 0.0001,
       "step": 225
     },
     {
       "epoch": 0.008086301590425246,
-      "grad_norm": 0.03776596486568451,
+      "grad_norm": 0.06953338533639908,
       "learning_rate": 0.0001758361915288722,
-      "loss": 0.001,
+      "loss": 0.0007,
       "step": 226
     },
     {
       "epoch": 0.0081220816859581,
-      "grad_norm": 0.0055121988989412785,
+      "grad_norm": 0.0011373666347935796,
       "learning_rate": 0.0001756310380251472,
-      "loss": 0.0004,
+      "loss": 0.0001,
       "step": 227
     },
     {
       "epoch": 0.008157861781490956,
-      "grad_norm": 0.06609760224819183,
+      "grad_norm": 0.0281671229749918,
       "learning_rate": 0.00017542513807361037,
-      "loss": 0.0014,
+      "loss": 0.0003,
       "step": 228
     },
     {
       "epoch": 0.008193641877023811,
-      "grad_norm": 0.01043468527495861,
+      "grad_norm": 0.003383536823093891,
       "learning_rate": 0.00017521849370641114,
-      "loss": 0.0005,
+      "loss": 0.0002,
       "step": 229
     },
     {
       "epoch": 0.008229421972556666,
-      "grad_norm": NaN,
-      "learning_rate": 0.00017521849370641114,
-      "loss": 0.0081,
+      "grad_norm": 0.004913358949124813,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.0002,
       "step": 230
     },
     {
       "epoch": 0.008265202068089521,
-      "grad_norm": 0.0070306649431586266,
-      "learning_rate": 0.00017501110696304596,
-      "loss": 0.0005,
+      "grad_norm": 0.0018077875720337033,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 0.0001,
       "step": 231
     },
     {
       "epoch": 0.008300982163622376,
-      "grad_norm": 0.012813176028430462,
-      "learning_rate": 0.00017480297989033825,
-      "loss": 0.0004,
+      "grad_norm": 0.0008154757670126855,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 0.0001,
       "step": 232
     },
     {
       "epoch": 0.008336762259155231,
-      "grad_norm": 2.705077648162842,
-      "learning_rate": 0.00017459411454241822,
-      "loss": 0.1548,
+      "grad_norm": 0.4949972927570343,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 0.0095,
       "step": 233
     },
     {
       "epoch": 0.008372542354688086,
-      "grad_norm": 0.005047104321420193,
-      "learning_rate": 0.00017438451298070252,
-      "loss": 0.0004,
+      "grad_norm": 0.0014726183144375682,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.0001,
       "step": 234
     },
     {
       "epoch": 0.008408322450220941,
-      "grad_norm": 0.16771042346954346,
-      "learning_rate": 0.00017417417727387394,
-      "loss": 0.0016,
+      "grad_norm": 0.000985615304671228,
+      "learning_rate": 0.000173963109497861,
+      "loss": 0.0001,
       "step": 235
     },
     {
       "epoch": 0.008444102545753797,
-      "grad_norm": 0.005447797942906618,
-      "learning_rate": 0.000173963109497861,
-      "loss": 0.0004,
+      "grad_norm": 0.0015382058918476105,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.0001,
       "step": 236
     },
     {
       "epoch": 0.008479882641286652,
-      "grad_norm": 0.0037992142606526613,
-      "learning_rate": 0.0001737513117358174,
-      "loss": 0.0003,
+      "grad_norm": 0.001569434185512364,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 0.0001,
       "step": 237
     },
     {
       "epoch": 0.008515662736819507,
-      "grad_norm": 0.03688381612300873,
-      "learning_rate": 0.0001735387860781016,
-      "loss": 0.0008,
+      "grad_norm": 0.01702137105166912,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 0.0004,
       "step": 238
     },
     {
       "epoch": 0.008551442832352362,
-      "grad_norm": 0.20397980511188507,
-      "learning_rate": 0.00017332553462225602,
-      "loss": 0.0009,
+      "grad_norm": 0.9787658452987671,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 0.03,
       "step": 239
     },
     {
       "epoch": 0.008587222927885217,
-      "grad_norm": 0.005972681101411581,
-      "learning_rate": 0.00017311155947298643,
-      "loss": 0.0004,
+      "grad_norm": 0.0007933751912787557,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.0001,
       "step": 240
     },
     {
       "epoch": 0.008623003023418072,
-      "grad_norm": 0.011358226649463177,
-      "learning_rate": 0.00017289686274214118,
-      "loss": 0.0006,
+      "grad_norm": 0.0029444191604852676,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 0.0002,
       "step": 241
     },
     {
       "epoch": 0.008658783118950927,
-      "grad_norm": 0.0113879619166255,
-      "learning_rate": 0.0001726814465486903,
-      "loss": 0.0008,
+      "grad_norm": 0.0021299768704921007,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 0.0001,
       "step": 242
     },
     {
       "epoch": 0.008694563214483782,
-      "grad_norm": 0.008897796273231506,
-      "learning_rate": 0.0001724653130187047,
-      "loss": 0.0005,
+      "grad_norm": 0.0013644633581861854,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 0.0001,
       "step": 243
     },
     {
       "epoch": 0.008730343310016637,
-      "grad_norm": 0.014523560181260109,
-      "learning_rate": 0.00017224846428533499,
-      "loss": 0.0007,
+      "grad_norm": 0.0023565045557916164,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 0.0001,
       "step": 244
     },
     {
       "epoch": 0.008766123405549492,
-      "grad_norm": 0.014025084674358368,
-      "learning_rate": 0.0001720309024887907,
-      "loss": 0.0007,
+      "grad_norm": 0.001884597004391253,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 0.0001,
       "step": 245
     },
     {
       "epoch": 0.008801903501082347,
-      "grad_norm": 0.07501209527254105,
-      "learning_rate": 0.00017181262977631888,
-      "loss": 0.0017,
+      "grad_norm": 0.08846484124660492,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 0.0008,
       "step": 246
     },
     {
       "epoch": 0.008837683596615202,
-      "grad_norm": 0.008224425837397575,
-      "learning_rate": 0.00017159364830218312,
-      "loss": 0.0005,
+      "grad_norm": 0.06892094761133194,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 0.0006,
       "step": 247
     },
     {
       "epoch": 0.008873463692148057,
-      "grad_norm": 0.010049921460449696,
-      "learning_rate": 0.00017137396022764214,
-      "loss": 0.0005,
+      "grad_norm": 0.002805879106745124,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 0.0002,
       "step": 248
     },
     {
       "epoch": 0.008909243787680913,
-      "grad_norm": 0.048693351447582245,
-      "learning_rate": 0.00017115356772092857,
-      "loss": 0.0011,
+      "grad_norm": 0.016457632184028625,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 0.0003,
       "step": 249
     },
     {
       "epoch": 0.008945023883213768,
-      "grad_norm": 0.04673377797007561,
-      "learning_rate": 0.0001709324729572274,
-      "loss": 0.001,
+      "grad_norm": 2.7799627780914307,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.043,
       "step": 250
     },
     {
       "epoch": 0.008980803978746623,
-      "grad_norm": 1.0758200883865356,
-      "learning_rate": 0.00017071067811865476,
-      "loss": 0.0139,
+      "grad_norm": 0.07246087491512299,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.0016,
       "step": 251
     },
     {
       "epoch": 0.009016584074279478,
-      "grad_norm": 0.009010600857436657,
-      "learning_rate": 0.00017048818539423615,
-      "loss": 0.0004,
+      "grad_norm": 0.003910925704985857,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 0.0002,
       "step": 252
     },
     {
       "epoch": 0.009052364169812333,
-      "grad_norm": 0.0034245457500219345,
-      "learning_rate": 0.00017026499697988493,
-      "loss": 0.0003,
+      "grad_norm": 0.002456547226756811,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 0.0002,
       "step": 253
     },
     {
       "epoch": 0.009088144265345188,
-      "grad_norm": 0.15912485122680664,
-      "learning_rate": 0.00017004111507838064,
-      "loss": 0.0009,
+      "grad_norm": 0.0028989992570132017,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 0.0002,
       "step": 254
     },
     {
       "epoch": 0.009123924360878043,
-      "grad_norm": 0.004494361579418182,
-      "learning_rate": 0.00016981654189934727,
+      "grad_norm": 0.0032434300519526005,
+      "learning_rate": 0.00016959127965923142,
       "loss": 0.0003,
       "step": 255
     },
     {
       "epoch": 0.009159704456410898,
-      "grad_norm": 0.2626131772994995,
-      "learning_rate": 0.00016959127965923142,
-      "loss": 0.0025,
+      "grad_norm": 0.00594690814614296,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 0.0004,
       "step": 256
     },
     {
       "epoch": 0.009195484551943753,
-      "grad_norm": 0.003236124524846673,
-      "learning_rate": 0.0001693653305812805,
-      "loss": 0.0002,
+      "grad_norm": 0.038472067564725876,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 0.0011,
       "step": 257
     },
     {
       "epoch": 0.009231264647476608,
-      "grad_norm": 0.0019491679267957807,
-      "learning_rate": 0.00016913869689552064,
-      "loss": 0.0001,
+      "grad_norm": 0.006935628596693277,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 0.0004,
       "step": 258
     },
     {
       "epoch": 0.009267044743009463,
-      "grad_norm": 0.06664927303791046,
-      "learning_rate": 0.00016891138083873487,
-      "loss": 0.0008,
+      "grad_norm": 0.08398015797138214,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 0.0015,
       "step": 259
     },
     {
       "epoch": 0.009302824838542318,
-      "grad_norm": 0.0012031469959765673,
-      "learning_rate": 0.00016868338465444085,
-      "loss": 0.0001,
+      "grad_norm": 0.04922045022249222,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.0004,
       "step": 260
     },
     {
       "epoch": 0.009338604934075173,
-      "grad_norm": 0.0035160172265022993,
-      "learning_rate": 0.00016845471059286887,
-      "loss": 0.0002,
+      "grad_norm": 0.013643675483763218,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 0.0008,
       "step": 261
     },
     {
       "epoch": 0.009374385029608029,
-      "grad_norm": 0.001314219436608255,
-      "learning_rate": 0.00016822536091093965,
-      "loss": 0.0001,
+      "grad_norm": 0.011866009794175625,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 0.0003,
       "step": 262
     },
     {
       "epoch": 0.009410165125140884,
-      "grad_norm": 0.0014836995396763086,
-      "learning_rate": 0.00016799533787224192,
-      "loss": 0.0001,
+      "grad_norm": 0.00603786064311862,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 0.0003,
       "step": 263
     },
     {
       "epoch": 0.009445945220673739,
-      "grad_norm": 0.005519370548427105,
-      "learning_rate": 0.00016776464374701025,
-      "loss": 0.0002,
+      "grad_norm": 0.015595407225191593,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 0.0007,
       "step": 264
     },
     {
       "epoch": 0.009481725316206594,
-      "grad_norm": 0.001866112113930285,
-      "learning_rate": 0.00016753328081210245,
-      "loss": 0.0001,
+      "grad_norm": 0.005519995000213385,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 0.0004,
       "step": 265
     },
     {
       "epoch": 0.009517505411739449,
-      "grad_norm": 0.03768831863999367,
-      "learning_rate": 0.00016730125135097735,
-      "loss": 0.0002,
+      "grad_norm": 0.013364361599087715,
+      "learning_rate": 0.000167068557653672,
+      "loss": 0.0004,
       "step": 266
     },
     {
       "epoch": 0.009553285507272304,
-      "grad_norm": 0.00720590865239501,
-      "learning_rate": 0.000167068557653672,
-      "loss": 0.0002,
+      "grad_norm": 0.011465894058346748,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 0.0005,
       "step": 267
     },
     {
       "epoch": 0.009589065602805159,
-      "grad_norm": 0.003128788201138377,
-      "learning_rate": 0.0001668352020167793,
-      "loss": 0.0001,
+      "grad_norm": 0.0022764878813177347,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 0.0002,
       "step": 268
     },
     {
       "epoch": 0.009624845698338014,
-      "grad_norm": 0.002723650773987174,
-      "learning_rate": 0.00016660118674342517,
-      "loss": 0.0001,
+      "grad_norm": 0.04595799744129181,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 0.0011,
       "step": 269
     },
     {
       "epoch": 0.00966062579387087,
-      "grad_norm": 0.0018245120299980044,
-      "learning_rate": 0.00016636651414324587,
-      "loss": 0.0002,
+      "grad_norm": 0.008233287371695042,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.0005,
       "step": 270
     },
     {
       "epoch": 0.009696405889403724,
-      "grad_norm": 0.008595665916800499,
-      "learning_rate": 0.00016613118653236518,
-      "loss": 0.0002,
+      "grad_norm": 0.016462009400129318,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 0.0004,
       "step": 271
     },
     {
       "epoch": 0.00973218598493658,
-      "grad_norm": 0.0439569465816021,
-      "learning_rate": 0.0001658952062333717,
-      "loss": 0.0004,
+      "grad_norm": 0.08745335787534714,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 0.0013,
       "step": 272
     },
     {
       "epoch": 0.009767966080469434,
-      "grad_norm": 0.012526117265224457,
-      "learning_rate": 0.00016565857557529566,
-      "loss": 0.0002,
+      "grad_norm": 0.03373585268855095,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 0.0007,
       "step": 273
     },
     {
       "epoch": 0.00980374617600229,
-      "grad_norm": 0.536148726940155,
-      "learning_rate": 0.00016542129689358612,
-      "loss": 0.0066,
+      "grad_norm": 0.5139638185501099,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 0.0155,
       "step": 274
     },
     {
       "epoch": 0.009839526271535145,
-      "grad_norm": 0.0010097100166603923,
-      "learning_rate": 0.0001651833725300879,
-      "loss": 0.0001,
+      "grad_norm": 0.006406674161553383,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.0002,
       "step": 275
     },
     {
       "epoch": 0.009875306367068,
-      "grad_norm": 0.42853671312332153,
-      "learning_rate": 0.00016494480483301836,
-      "loss": 0.0542,
+      "grad_norm": 0.34044501185417175,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 0.0352,
       "step": 276
     },
     {
       "epoch": 0.009911086462600855,
-      "grad_norm": 0.01222367025911808,
-      "learning_rate": 0.00016470559615694446,
-      "loss": 0.0002,
+      "grad_norm": 0.0065045482479035854,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 0.0004,
       "step": 277
     },
     {
       "epoch": 0.00994686655813371,
-      "grad_norm": 0.40589821338653564,
-      "learning_rate": 0.00016446574886275913,
-      "loss": 0.0009,
+      "grad_norm": 0.06801342964172363,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 0.0013,
       "step": 278
     },
     {
       "epoch": 0.009982646653666565,
-      "grad_norm": 0.17128930985927582,
-      "learning_rate": 0.00016422526531765846,
-      "loss": 0.0013,
+      "grad_norm": 0.0377359464764595,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 0.0003,
       "step": 279
     },
     {
       "epoch": 0.01001842674919942,
-      "grad_norm": 0.003709537675604224,
-      "learning_rate": 0.00016398414789511786,
-      "loss": 0.0001,
+      "grad_norm": 0.015254302881658077,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.0007,
       "step": 280
     },
     {
       "epoch": 0.010054206844732275,
-      "grad_norm": 0.2669826149940491,
-      "learning_rate": 0.000163742398974869,
-      "loss": 0.0008,
+      "grad_norm": 0.006927182897925377,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 0.0002,
       "step": 281
     },
     {
       "epoch": 0.01008998694026513,
-      "grad_norm": 0.06789790838956833,
-      "learning_rate": 0.00016350002094287609,
-      "loss": 0.0005,
+      "grad_norm": 0.02803247794508934,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 0.0006,
       "step": 282
     },
     {
       "epoch": 0.010125767035797985,
-      "grad_norm": 0.002458445494994521,
-      "learning_rate": 0.00016325701619131246,
-      "loss": 0.0001,
+      "grad_norm": 0.01434150617569685,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 0.0005,
       "step": 283
     },
     {
       "epoch": 0.01016154713133084,
-      "grad_norm": 0.0014181917067617178,
-      "learning_rate": 0.00016301338711853693,
-      "loss": 0.0001,
+      "grad_norm": 0.001818214193917811,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 0.0002,
       "step": 284
     },
     {
       "epoch": 0.010197327226863695,
-      "grad_norm": 0.0011788622941821814,
-      "learning_rate": 0.00016276913612907007,
+      "grad_norm": 0.0021695042960345745,
+      "learning_rate": 0.00016252426563357055,
       "loss": 0.0001,
       "step": 285
     },
     {
       "epoch": 0.01023310732239655,
-      "grad_norm": 1.1278730630874634,
-      "learning_rate": 0.00016252426563357055,
-      "loss": 0.0025,
+      "grad_norm": 0.001164208515547216,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 0.0001,
       "step": 286
     },
     {
       "epoch": 0.010268887417929405,
-      "grad_norm": 0.001996925100684166,
-      "learning_rate": 0.00016227877804881127,
-      "loss": 0.0001,
+      "grad_norm": 0.002602560678496957,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 0.0002,
       "step": 287
     },
     {
       "epoch": 0.01030466751346226,
-      "grad_norm": 0.0013124500401318073,
-      "learning_rate": 0.00016203267579765563,
+      "grad_norm": 0.002262539230287075,
+      "learning_rate": 0.00016178596130903344,
       "loss": 0.0001,
       "step": 288
     },
     {
       "epoch": 0.010340447608995116,
-      "grad_norm": 0.001296933158300817,
-      "learning_rate": 0.00016178596130903344,
+      "grad_norm": 0.0017323383362963796,
+      "learning_rate": 0.00016153863701791717,
       "loss": 0.0001,
       "step": 289
     },
     {
       "epoch": 0.01037622770452797,
-      "grad_norm": 0.005077495705336332,
-      "learning_rate": 0.00016153863701791717,
-      "loss": 0.0002,
+      "grad_norm": 0.005987819284200668,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.0003,
       "step": 290
     },
     {
       "epoch": 0.010412007800060826,
-      "grad_norm": 0.0018195759039372206,
-      "learning_rate": 0.00016129070536529766,
-      "loss": 0.0001,
+      "grad_norm": 0.002981544006615877,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 0.0002,
       "step": 291
     },
     {
       "epoch": 0.01044778789559368,
-      "grad_norm": 0.002947065280750394,
-      "learning_rate": 0.00016104216879816026,
+      "grad_norm": 0.004128920845687389,
+      "learning_rate": 0.00016079302976946055,
       "loss": 0.0002,
       "step": 292
     },
     {
       "epoch": 0.010483567991126536,
-      "grad_norm": 0.0028106742538511753,
-      "learning_rate": 0.00016079302976946055,
-      "loss": 0.0001,
+      "grad_norm": 0.0421392023563385,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 0.0006,
       "step": 293
     },
     {
       "epoch": 0.010519348086659391,
-      "grad_norm": 0.09077682346105576,
-      "learning_rate": 0.00016054329073810015,
-      "loss": 0.0011,
+      "grad_norm": 0.007170453667640686,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 0.0003,
       "step": 294
     },
     {
       "epoch": 0.010555128182192246,
-      "grad_norm": 0.027377746999263763,
-      "learning_rate": 0.00016029295416890248,
+      "grad_norm": 0.012681632302701473,
+      "learning_rate": 0.00016004202253258842,
       "loss": 0.0003,
       "step": 295
     },
     {
       "epoch": 0.010590908277725101,
-      "grad_norm": 0.049133822321891785,
-      "learning_rate": 0.00016004202253258842,
-      "loss": 0.0004,
+      "grad_norm": 0.00398437911644578,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 0.0002,
       "step": 296
     },
     {
       "epoch": 0.010626688373257956,
-      "grad_norm": 0.014056004583835602,
-      "learning_rate": 0.0001597904983057519,
-      "loss": 0.0004,
+      "grad_norm": 0.02630179561674595,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 0.0005,
       "step": 297
     },
     {
       "epoch": 0.010662468468790811,
-      "grad_norm": 0.03864206746220589,
-      "learning_rate": 0.00015953838397083552,
-      "loss": 0.0004,
+      "grad_norm": 0.007436979562044144,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 0.0003,
       "step": 298
     },
     {
       "epoch": 0.010698248564323666,
-      "grad_norm": 0.0013771051308140159,
-      "learning_rate": 0.00015928568201610595,
+      "grad_norm": 0.0016087141120806336,
+      "learning_rate": 0.00015903239493562948,
       "loss": 0.0001,
       "step": 299
     },
     {
       "epoch": 0.010734028659856521,
-      "grad_norm": 0.06298744678497314,
-      "learning_rate": 0.00015903239493562948,
-      "loss": 0.0003,
+      "grad_norm": 0.0017188222846016288,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.0001,
       "step": 300
     },
     {
       "epoch": 0.010769808755389377,
-      "grad_norm": 0.23744501173496246,
-      "learning_rate": 0.00015877852522924732,
-      "loss": 0.001,
+      "grad_norm": 0.004087434615939856,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.0002,
       "step": 301
     },
     {
       "epoch": 0.010805588850922232,
-      "grad_norm": 0.008492168970406055,
-      "learning_rate": 0.00015852407540255104,
-      "loss": 0.0002,
+      "grad_norm": 0.25743332505226135,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 0.0021,
       "step": 302
     },
     {
       "epoch": 0.010841368946455087,
-      "grad_norm": 0.014776664786040783,
-      "learning_rate": 0.00015826904796685762,
+      "grad_norm": 0.004327516537159681,
+      "learning_rate": 0.00015801344543918495,
       "loss": 0.0002,
       "step": 303
     },
     {
       "epoch": 0.010877149041987942,
-      "grad_norm": 0.001956879161298275,
-      "learning_rate": 0.00015801344543918495,
-      "loss": 0.0001,
+      "grad_norm": 0.00956784002482891,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 0.0002,
       "step": 304
     },
     {
       "epoch": 0.010912929137520797,
-      "grad_norm": 0.00224097166210413,
-      "learning_rate": 0.00015775727034222675,
+      "grad_norm": 0.0017107793828472495,
+      "learning_rate": 0.00015750052520432787,
       "loss": 0.0001,
       "step": 305
     },
     {
       "epoch": 0.010948709233053652,
-      "grad_norm": 0.0012625321978703141,
-      "learning_rate": 0.00015750052520432787,
-      "loss": 0.0001,
+      "grad_norm": 0.011659998446702957,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 0.0002,
       "step": 306
     },
     {
       "epoch": 0.010984489328586507,
-      "grad_norm": 0.00540123600512743,
-      "learning_rate": 0.0001572432125594591,
-      "loss": 0.0003,
+      "grad_norm": 0.007189000491052866,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 0.0004,
       "step": 307
     },
     {
       "epoch": 0.011020269424119362,
-      "grad_norm": 0.0011538398684933782,
-      "learning_rate": 0.00015698533494719238,
-      "loss": 0.0001,
+      "grad_norm": 0.0030206877272576094,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 0.0002,
       "step": 308
     },
     {
       "epoch": 0.011056049519652217,
-      "grad_norm": 0.002988644177094102,
-      "learning_rate": 0.00015672689491267567,
+      "grad_norm": 0.0016745973844081163,
+      "learning_rate": 0.00015646789500660773,
       "loss": 0.0001,
       "step": 309
     },
     {
       "epoch": 0.011091829615185072,
-      "grad_norm": 0.0029471865855157375,
-      "learning_rate": 0.00015646789500660773,
+      "grad_norm": 0.0018627311801537871,
+      "learning_rate": 0.00015620833778521307,
       "loss": 0.0001,
       "step": 310
     },
     {
       "epoch": 0.011127609710717927,
-      "grad_norm": 0.003695828840136528,
-      "learning_rate": 0.00015620833778521307,
+      "grad_norm": 0.0017725079087540507,
+      "learning_rate": 0.0001559482258102167,
       "loss": 0.0001,
       "step": 311
     },
     {
       "epoch": 0.011163389806250782,
-      "grad_norm": 0.0008560797432437539,
-      "learning_rate": 0.0001559482258102167,
+      "grad_norm": 0.0018159925239160657,
+      "learning_rate": 0.00015568756164881882,
       "loss": 0.0001,
       "step": 312
     },
     {
       "epoch": 0.011199169901783637,
-      "grad_norm": 0.14824579656124115,
-      "learning_rate": 0.00015568756164881882,
-      "loss": 0.0005,
+      "grad_norm": 0.0026159253902733326,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 0.0001,
       "step": 313
     },
     {
       "epoch": 0.011234949997316493,
-      "grad_norm": 0.001580694573931396,
-      "learning_rate": 0.00015542634787366942,
-      "loss": 0.0001,
+      "grad_norm": 0.005072129424661398,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 0.0002,
       "step": 314
     },
     {
       "epoch": 0.011270730092849348,
-      "grad_norm": 0.002695061033591628,
-      "learning_rate": 0.00015516458706284303,
+      "grad_norm": 0.009669373743236065,
+      "learning_rate": 0.0001549022817998132,
       "loss": 0.0001,
       "step": 315
     },
     {
       "epoch": 0.011306510188382203,
-      "grad_norm": 0.41342833638191223,
-      "learning_rate": 0.0001549022817998132,
-      "loss": 0.0026,
+      "grad_norm": 0.029667021706700325,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 0.0008,
       "step": 316
     },
     {
       "epoch": 0.011342290283915058,
-      "grad_norm": 0.0013819708256050944,
-      "learning_rate": 0.00015463943467342693,
+      "grad_norm": 0.0014934055507183075,
+      "learning_rate": 0.00015437604827787927,
       "loss": 0.0001,
       "step": 317
     },
     {
       "epoch": 0.011378070379447913,
-      "grad_norm": 0.0010516056790947914,
-      "learning_rate": 0.00015437604827787927,
+      "grad_norm": 0.0008770531858317554,
+      "learning_rate": 0.00015411212521268758,
       "loss": 0.0001,
       "step": 318
     },
     {
       "epoch": 0.011413850474980768,
-      "grad_norm": 0.010458525270223618,
-      "learning_rate": 0.00015411212521268758,
+      "grad_norm": 0.001008510240353644,
+      "learning_rate": 0.00015384766808266602,
       "loss": 0.0001,
       "step": 319
     },
     {
       "epoch": 0.011449630570513623,
-      "grad_norm": 0.0034740527626127005,
-      "learning_rate": 0.00015384766808266602,
-      "loss": 0.0001,
+      "grad_norm": 0.6345011591911316,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.0042,
       "step": 320
     },
     {
       "epoch": 0.011485410666046478,
-      "grad_norm": 0.0020982057321816683,
-      "learning_rate": 0.00015358267949789966,
-      "loss": 0.0001,
+      "grad_norm": 0.005814701318740845,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.0002,
       "step": 321
     },
     {
       "epoch": 0.011521190761579333,
-      "grad_norm": 0.0006513833650387824,
-      "learning_rate": 0.00015331716207371888,
+      "grad_norm": 0.0008782675722613931,
+      "learning_rate": 0.0001530511184306734,
       "loss": 0.0001,
       "step": 322
     },
     {
       "epoch": 0.011556970857112188,
-      "grad_norm": 0.0004633856297004968,
-      "learning_rate": 0.0001530511184306734,
-      "loss": 0.0,
+      "grad_norm": 0.0006713901530019939,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 0.0001,
       "step": 323
     },
     {
       "epoch": 0.011592750952645043,
-      "grad_norm": 0.05308767780661583,
-      "learning_rate": 0.00015278455119450664,
-      "loss": 0.0004,
+      "grad_norm": 0.005472630262374878,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 0.0002,
       "step": 324
     },
     {
       "epoch": 0.011628531048177898,
-      "grad_norm": 0.00755698885768652,
-      "learning_rate": 0.0001525174629961296,
+      "grad_norm": 0.005173255689442158,
+      "learning_rate": 0.0001522498564715949,
       "loss": 0.0001,
       "step": 325
     },
     {
       "epoch": 0.011664311143710753,
-      "grad_norm": 0.003082417882978916,
-      "learning_rate": 0.0001522498564715949,
-      "loss": 0.0001,
+      "grad_norm": 0.003191661089658737,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 0.0002,
       "step": 326
     },
     {
       "epoch": 0.011700091239243609,
-      "grad_norm": 0.0007315215771086514,
-      "learning_rate": 0.00015198173426207094,
+      "grad_norm": 0.0036652470007538795,
+      "learning_rate": 0.00015171309901381572,
       "loss": 0.0001,
       "step": 327
     },
     {
       "epoch": 0.011735871334776464,
-      "grad_norm": 0.002395337913185358,
-      "learning_rate": 0.00015171309901381572,
+      "grad_norm": 0.001312348758801818,
+      "learning_rate": 0.00015144395337815064,
       "loss": 0.0001,
       "step": 328
     },
     {
       "epoch": 0.011771651430309319,
-      "grad_norm": 0.0012004328891634941,
-      "learning_rate": 0.00015144395337815064,
+      "grad_norm": 0.0021110977977514267,
+      "learning_rate": 0.00015117430001143452,
       "loss": 0.0001,
       "step": 329
     },
     {
       "epoch": 0.011807431525842174,
-      "grad_norm": 0.029387248679995537,
-      "learning_rate": 0.00015117430001143452,
-      "loss": 0.0005,
+      "grad_norm": 0.004200182855129242,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 0.0002,
       "step": 330
     },
     {
       "epoch": 0.011843211621375029,
-      "grad_norm": 0.0007639694958925247,
-      "learning_rate": 0.00015090414157503714,
-      "loss": 0.0001,
+      "grad_norm": 0.0005367898265831172,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.0,
       "step": 331
     },
     {
       "epoch": 0.011878991716907884,
-      "grad_norm": 0.0016086171381175518,
-      "learning_rate": 0.00015063348073531324,
-      "loss": 0.0001,
+      "grad_norm": 0.007367909885942936,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 0.0002,
       "step": 332
     },
     {
       "epoch": 0.011914771812440739,
-      "grad_norm": 0.0008903722045943141,
-      "learning_rate": 0.0001503623201635761,
-      "loss": 0.0001,
+      "grad_norm": 0.013929062522947788,
+      "learning_rate": 0.000150090662536071,
+      "loss": 0.0002,
       "step": 333
     },
     {
       "epoch": 0.011950551907973594,
-      "grad_norm": 0.007725098170340061,
-      "learning_rate": 0.000150090662536071,
-      "loss": 0.0002,
+      "grad_norm": 0.009065449237823486,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 0.0001,
       "step": 334
     },
     {
       "epoch": 0.01198633200350645,
-      "grad_norm": 0.0007498833001591265,
-      "learning_rate": 0.0001498185105339491,
+      "grad_norm": 0.00572764640673995,
+      "learning_rate": 0.00014954586684324078,
       "loss": 0.0001,
       "step": 335
     },
     {
       "epoch": 0.012022112099039304,
-      "grad_norm": 0.002171967178583145,
-      "learning_rate": 0.00014954586684324078,
-      "loss": 0.0001,
+      "grad_norm": 0.1300538331270218,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 0.0009,
       "step": 336
     },
     {
       "epoch": 0.01205789219457216,
-      "grad_norm": 0.0014792379224672914,
-      "learning_rate": 0.00014927273415482915,
+      "grad_norm": 0.0021966646891087294,
+      "learning_rate": 0.00014899911516442365,
       "loss": 0.0001,
       "step": 337
     },
     {
       "epoch": 0.012093672290105014,
-      "grad_norm": 0.004606639035046101,
-      "learning_rate": 0.00014899911516442365,
-      "loss": 0.0001,
+      "grad_norm": 0.029224226251244545,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 0.0002,
       "step": 338
     },
     {
       "epoch": 0.01212945238563787,
-      "grad_norm": 0.0008608319913037121,
-      "learning_rate": 0.00014872501257253323,
-      "loss": 0.0001,
+      "grad_norm": 0.0069636073894798756,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 0.0002,
       "step": 339
     },
     {
       "epoch": 0.012165232481170725,
-      "grad_norm": 0.0007085691322572529,
-      "learning_rate": 0.0001484504290844398,
+      "grad_norm": 0.0027267655823379755,
+      "learning_rate": 0.00014817536741017152,
       "loss": 0.0001,
       "step": 340
     },
     {
       "epoch": 0.01220101257670358,
-      "grad_norm": 0.0027327192947268486,
-      "learning_rate": 0.00014817536741017152,
+      "grad_norm": 0.0007624703575856984,
+      "learning_rate": 0.00014789983026447612,
       "loss": 0.0001,
       "step": 341
     },
     {
       "epoch": 0.012236792672236435,
-      "grad_norm": 0.002414976479485631,
-      "learning_rate": 0.00014789983026447612,
+      "grad_norm": 0.000985614606179297,
+      "learning_rate": 0.0001476238203667939,
       "loss": 0.0001,
       "step": 342
     },
     {
       "epoch": 0.01227257276776929,
-      "grad_norm": 1.7449041604995728,
-      "learning_rate": 0.0001476238203667939,
-      "loss": 0.0026,
+      "grad_norm": 0.0015349843306466937,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 0.0001,
       "step": 343
     },
     {
       "epoch": 0.012308352863302145,
-      "grad_norm": 0.093377985060215,
-      "learning_rate": 0.0001473473404412312,
-      "loss": 0.0012,
+      "grad_norm": 0.21670787036418915,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 0.003,
       "step": 344
     },
     {
       "epoch": 0.012344132958835,
-      "grad_norm": 0.004823881201446056,
-      "learning_rate": 0.0001470703932165333,
+      "grad_norm": 0.0010241931304335594,
+      "learning_rate": 0.00014679298142605734,
       "loss": 0.0001,
       "step": 345
     },
     {
       "epoch": 0.012379913054367855,
-      "grad_norm": 0.0015184949152171612,
-      "learning_rate": 0.00014679298142605734,
+      "grad_norm": 0.0006662398809567094,
+      "learning_rate": 0.00014651510780774583,
       "loss": 0.0001,
       "step": 346
     },
     {
       "epoch": 0.01241569314990071,
-      "grad_norm": 0.0010811861138790846,
-      "learning_rate": 0.00014651510780774583,
+      "grad_norm": 0.001260753022506833,
+      "learning_rate": 0.00014623677510409918,
       "loss": 0.0001,
       "step": 347
     },
     {
       "epoch": 0.012451473245433565,
-      "grad_norm": 0.0012885968899354339,
-      "learning_rate": 0.00014623677510409918,
+      "grad_norm": 0.0011193427490070462,
+      "learning_rate": 0.00014595798606214882,
       "loss": 0.0001,
       "step": 348
     },
     {
       "epoch": 0.01248725334096642,
-      "grad_norm": 0.0012518796138465405,
-      "learning_rate": 0.00014595798606214882,
+      "grad_norm": 0.0006154956645332277,
+      "learning_rate": 0.00014567874343342997,
       "loss": 0.0001,
       "step": 349
     },
     {
       "epoch": 0.012523033436499275,
-      "grad_norm": 0.0038816649466753006,
-      "learning_rate": 0.00014567874343342997,
+      "grad_norm": 0.0007212765631265938,
+      "learning_rate": 0.00014539904997395468,
       "loss": 0.0001,
       "step": 350
     },
     {
       "epoch": 0.01255881353203213,
-      "grad_norm": 0.020913008600473404,
-      "learning_rate": 0.00014539904997395468,
+      "grad_norm": 0.010549906641244888,
+      "learning_rate": 0.00014511890844418453,
       "loss": 0.0003,
       "step": 351
     },
     {
       "epoch": 0.012594593627564986,
-      "grad_norm": 0.00418338505551219,
-      "learning_rate": 0.00014511890844418453,
-      "loss": 0.0001,
+      "grad_norm": 0.0005157431005500257,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 0.0,
       "step": 352
     },
     {
       "epoch": 0.01263037372309784,
-      "grad_norm": 1.1656886339187622,
-      "learning_rate": 0.00014483832160900326,
-      "loss": 0.0105,
+      "grad_norm": 0.00042760182986967266,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 0.0001,
       "step": 353
     },
     {
       "epoch": 0.012666153818630696,
-      "grad_norm": 0.9182770252227783,
-      "learning_rate": 0.00014455729223768966,
-      "loss": 0.0038,
+      "grad_norm": 0.004275426734238863,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 0.0001,
       "step": 354
     },
     {
       "epoch": 0.01270193391416355,
-      "grad_norm": 0.0016101684886962175,
-      "learning_rate": 0.0001442758231038902,
+      "grad_norm": 0.00281893671490252,
+      "learning_rate": 0.00014399391698559152,
       "loss": 0.0001,
       "step": 355
     },
     {
       "epoch": 0.012737714009696406,
-      "grad_norm": 0.0013801216846331954,
-      "learning_rate": 0.00014399391698559152,
+      "grad_norm": 0.0008448857697658241,
+      "learning_rate": 0.0001437115766650933,
       "loss": 0.0001,
       "step": 356
     },
     {
       "epoch": 0.012773494105229261,
-      "grad_norm": 0.0037179661449044943,
-      "learning_rate": 0.0001437115766650933,
-      "loss": 0.0001,
+      "grad_norm": 0.3817155361175537,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 0.0024,
       "step": 357
     },
     {
       "epoch": 0.012809274200762116,
-      "grad_norm": 0.0018248152919113636,
-      "learning_rate": 0.00014342880492898048,
-      "loss": 0.0001,
+      "grad_norm": 0.0003617767943069339,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 0.0,
       "step": 358
     },
     {
       "epoch": 0.012845054296294971,
-      "grad_norm": 1.6455618143081665,
-      "learning_rate": 0.0001431456045680959,
-      "loss": 0.0142,
+      "grad_norm": 0.01476916391402483,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 0.0002,
       "step": 359
     },
     {
       "epoch": 0.012880834391827826,
-      "grad_norm": 0.01422668807208538,
-      "learning_rate": 0.00014286197837751286,
-      "loss": 0.0002,
+      "grad_norm": 0.00045559153659269214,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.0,
       "step": 360
     },
     {
       "epoch": 0.012916614487360681,
-      "grad_norm": 0.0013264876324683428,
-      "learning_rate": 0.00014257792915650728,
-      "loss": 0.0001,
+      "grad_norm": 0.0004463440563995391,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.0,
       "step": 361
     },
     {
       "epoch": 0.012952394582893536,
-      "grad_norm": 0.0019532067235559225,
-      "learning_rate": 0.00014229345970853032,
+      "grad_norm": 0.0008974383235909045,
+      "learning_rate": 0.00014200857284118066,
       "loss": 0.0001,
       "step": 362
     },
     {
       "epoch": 0.012988174678426391,
-      "grad_norm": 0.002958630910143256,
-      "learning_rate": 0.00014200857284118066,
+      "grad_norm": 0.0006577485473826528,
+      "learning_rate": 0.00014172327136617656,
       "loss": 0.0001,
       "step": 363
     },
     {
       "epoch": 0.013023954773959246,
-      "grad_norm": 2.0066685676574707,
-      "learning_rate": 0.00014172327136617656,
-      "loss": 0.0083,
+      "grad_norm": 0.00020119256805628538,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 0.0,
       "step": 364
     },
     {
       "epoch": 0.013059734869492102,
-      "grad_norm": 0.002405961509793997,
-      "learning_rate": 0.00014143755809932845,
+      "grad_norm": 0.0012024675961583853,
+      "learning_rate": 0.00014115143586051088,
       "loss": 0.0001,
       "step": 365
     },
     {
       "epoch": 0.013095514965024957,
-      "grad_norm": 0.001647467608563602,
-      "learning_rate": 0.00014115143586051088,
+      "grad_norm": 0.0014580293791368604,
+      "learning_rate": 0.00014086490747363493,
       "loss": 0.0001,
       "step": 366
     },
     {
       "epoch": 0.013131295060557812,
-      "grad_norm": 0.017699865624308586,
-      "learning_rate": 0.00014086490747363493,
-      "loss": 0.0002,
+      "grad_norm": 0.0003630980208981782,
+      "learning_rate": 0.00014057797576662,
+      "loss": 0.0,
       "step": 367
     },
     {
       "epoch": 0.013167075156090667,
-      "grad_norm": 0.0012630497803911567,
-      "learning_rate": 0.00014057797576662,
-      "loss": 0.0001,
+      "grad_norm": 0.0006153484573587775,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 0.0,
       "step": 368
     },
     {
       "epoch": 0.013202855251623522,
-      "grad_norm": 0.0029656332917511463,
-      "learning_rate": 0.00014029064357136628,
-      "loss": 0.0002,
+      "grad_norm": 0.0015171675477176905,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 0.0001,
       "step": 369
     },
     {
       "epoch": 0.013238635347156377,
-      "grad_norm": 0.001950636738911271,
-      "learning_rate": 0.00014000291372372647,
-      "loss": 0.0001,
+      "grad_norm": 0.0003758160746656358,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 0.0,
       "step": 370
     },
     {
       "epoch": 0.013274415442689232,
-      "grad_norm": 0.001674666884355247,
-      "learning_rate": 0.00013971478906347806,
+      "grad_norm": 0.0028044632636010647,
+      "learning_rate": 0.00013942627243429512,
       "loss": 0.0001,
       "step": 371
     },
     {
       "epoch": 0.013310195538222087,
-      "grad_norm": 0.0025913359131664038,
-      "learning_rate": 0.00013942627243429512,
+      "grad_norm": 0.0005744365626014769,
+      "learning_rate": 0.00013913736668372026,
       "loss": 0.0001,
       "step": 372
     },
     {
       "epoch": 0.013345975633754942,
-      "grad_norm": 0.0019228560850024223,
-      "learning_rate": 0.00013913736668372026,
-      "loss": 0.0001,
+      "grad_norm": 0.03431812301278114,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 0.0003,
       "step": 373
     },
     {
       "epoch": 0.013381755729287797,
-      "grad_norm": 0.0024691151920706034,
-      "learning_rate": 0.00013884807466313663,
+      "grad_norm": 0.002126405481249094,
+      "learning_rate": 0.00013855839922773968,
       "loss": 0.0001,
       "step": 374
     },
     {
       "epoch": 0.013417535824820652,
-      "grad_norm": 0.002829766133800149,
-      "learning_rate": 0.00013855839922773968,
-      "loss": 0.0001,
+      "grad_norm": 0.0006087534129619598,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.0,
       "step": 375
     },
     {
       "epoch": 0.013453315920353507,
-      "grad_norm": 0.0016127845738083124,
-      "learning_rate": 0.000138268343236509,
-      "loss": 0.0001,
+      "grad_norm": 0.0005148574709892273,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 0.0,
       "step": 376
     },
     {
       "epoch": 0.013489096015886362,
-      "grad_norm": 0.0034204639960080385,
-      "learning_rate": 0.00013797790955218014,
+      "grad_norm": 0.0026000775396823883,
+      "learning_rate": 0.00013768710104121627,
       "loss": 0.0001,
       "step": 377
     },
     {
       "epoch": 0.013524876111419218,
-      "grad_norm": 0.0250179972499609,
-      "learning_rate": 0.00013768710104121627,
-      "loss": 0.0004,
+      "grad_norm": 0.0016088751144707203,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 0.0001,
       "step": 378
     },
     {
       "epoch": 0.013560656206952073,
-      "grad_norm": 0.0031712076161056757,
-      "learning_rate": 0.00013739592057378003,
-      "loss": 0.0001,
+      "grad_norm": 0.00028388932696543634,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 0.0,
       "step": 379
     },
     {
       "epoch": 0.013596436302484928,
-      "grad_norm": 0.9804368615150452,
-      "learning_rate": 0.0001371043710237051,
-      "loss": 0.006,
+      "grad_norm": 0.0008413681644015014,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 0.0,
       "step": 380
     },
     {
       "epoch": 0.013632216398017783,
-      "grad_norm": 0.0034811091609299183,
-      "learning_rate": 0.00013681245526846783,
-      "loss": 0.0001,
+      "grad_norm": 0.0004796941066160798,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.0,
       "step": 381
     },
     {
       "epoch": 0.013667996493550638,
-      "grad_norm": 0.0016928123077377677,
-      "learning_rate": 0.0001365201761891588,
+      "grad_norm": 0.007888946682214737,
+      "learning_rate": 0.00013622753667045457,
       "loss": 0.0001,
       "step": 382
     },
     {
       "epoch": 0.013703776589083493,
-      "grad_norm": 0.006847256328910589,
-      "learning_rate": 0.00013622753667045457,
-      "loss": 0.0002,
+      "grad_norm": 0.0006594446022063494,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 0.0001,
       "step": 383
     },
     {
       "epoch": 0.013739556684616348,
-      "grad_norm": 0.001284815720282495,
-      "learning_rate": 0.00013593453960058908,
+      "grad_norm": 0.0005899665993638337,
+      "learning_rate": 0.00013564118787132506,
       "loss": 0.0001,
       "step": 384
     },
     {
       "epoch": 0.013775336780149203,
-      "grad_norm": 0.14246053993701935,
-      "learning_rate": 0.00013564118787132506,
-      "loss": 0.002,
+      "grad_norm": 0.03223516419529915,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 0.0003,
       "step": 385
     },
     {
       "epoch": 0.013811116875682058,
-      "grad_norm": 2.7674007415771484,
-      "learning_rate": 0.00013534748437792573,
-      "loss": 0.0364,
+      "grad_norm": 0.0013933213194832206,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 0.0001,
       "step": 386
     },
     {
       "epoch": 0.013846896971214913,
-      "grad_norm": 0.0014613076345995069,
-      "learning_rate": 0.0001350534320191259,
-      "loss": 0.0001,
+      "grad_norm": 0.0004967997083440423,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 0.0,
       "step": 387
     },
     {
       "epoch": 0.013882677066747768,
-      "grad_norm": 0.009488871321082115,
-      "learning_rate": 0.0001347590336971037,
-      "loss": 0.0004,
+      "grad_norm": 0.0020627062767744064,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 0.0001,
       "step": 388
     },
     {
       "epoch": 0.013918457162280623,
-      "grad_norm": 0.00213580671697855,
-      "learning_rate": 0.0001344642923174517,
+      "grad_norm": 0.0016057563479989767,
+      "learning_rate": 0.00013416921078914835,
       "loss": 0.0001,
       "step": 389
     },
     {
       "epoch": 0.013954237257813478,
-      "grad_norm": 0.009764998219907284,
-      "learning_rate": 0.00013416921078914835,
-      "loss": 0.0002,
+      "grad_norm": 0.0008467244333587587,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 0.0,
       "step": 390
     },
     {
       "epoch": 0.013990017353346334,
-      "grad_norm": 0.0015421390999108553,
-      "learning_rate": 0.00013387379202452917,
+      "grad_norm": 0.001441538450308144,
+      "learning_rate": 0.00013357803893925807,
       "loss": 0.0001,
       "step": 391
     },
     {
       "epoch": 0.014025797448879189,
-      "grad_norm": 0.044142596423625946,
-      "learning_rate": 0.00013357803893925807,
-      "loss": 0.0004,
+      "grad_norm": 0.0010898974724113941,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 0.0001,
       "step": 392
     },
     {
       "epoch": 0.014061577544412044,
-      "grad_norm": 0.0050599705427885056,
-      "learning_rate": 0.00013328195445229868,
-      "loss": 0.0002,
+      "grad_norm": 0.0007908979896456003,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 0.0001,
       "step": 393
     },
     {
       "epoch": 0.014097357639944899,
-      "grad_norm": 0.0021553614642471075,
-      "learning_rate": 0.00013298554148588528,
-      "loss": 0.0002,
+      "grad_norm": 0.0010799920419231057,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 0.0001,
       "step": 394
     },
     {
       "epoch": 0.014133137735477754,
-      "grad_norm": 0.001767953741364181,
-      "learning_rate": 0.00013268880296549425,
+      "grad_norm": 0.0009932275861501694,
+      "learning_rate": 0.00013239174181981495,
       "loss": 0.0001,
       "step": 395
     },
     {
       "epoch": 0.014168917831010609,
-      "grad_norm": 0.005165969021618366,
-      "learning_rate": 0.00013239174181981495,
-      "loss": 0.0002,
+      "grad_norm": 0.00047487238771282136,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 0.0,
       "step": 396
     },
     {
       "epoch": 0.014204697926543464,
-      "grad_norm": 0.18705643713474274,
-      "learning_rate": 0.00013209436098072095,
-      "loss": 0.0021,
+      "grad_norm": 1.1907325983047485,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 0.0067,
       "step": 397
     },
     {
       "epoch": 0.014240478022076319,
-      "grad_norm": 0.03174612671136856,
-      "learning_rate": 0.00013179666338324108,
-      "loss": 0.0004,
+      "grad_norm": 0.007387212011963129,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 0.0001,
       "step": 398
     },
     {
       "epoch": 0.014276258117609174,
-      "grad_norm": 0.001790829235687852,
-      "learning_rate": 0.0001314986519655305,
+      "grad_norm": 0.0006041564629413188,
+      "learning_rate": 0.0001312003296688415,
       "loss": 0.0001,
       "step": 399
     },
     {
       "epoch": 0.01431203821314203,
-      "grad_norm": 0.005095229018479586,
-      "learning_rate": 0.0001312003296688415,
-      "loss": 0.0002,
+      "grad_norm": 0.005637326277792454,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.0001,
       "step": 400
     },
     {
       "epoch": 0.014347818308674884,
-      "grad_norm": 0.0028803693130612373,
-      "learning_rate": 0.00013090169943749476,
-      "loss": 0.0002,
+      "grad_norm": 0.0007078536436893046,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.0,
       "step": 401
     },
     {
       "epoch": 0.01438359840420774,
-      "grad_norm": 0.6724929213523865,
-      "learning_rate": 0.0001306027642188501,
-      "loss": 0.0045,
+      "grad_norm": 0.0007561934762634337,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 0.0,
       "step": 402
     },
     {
       "epoch": 0.014419378499740594,
-      "grad_norm": 0.0036115716211497784,
-      "learning_rate": 0.00013030352696327742,
-      "loss": 0.0002,
+      "grad_norm": 0.0013270349008962512,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 0.0,
       "step": 403
     },
     {
       "epoch": 0.01445515859527345,
-      "grad_norm": 0.0012860867427662015,
-      "learning_rate": 0.00013000399062412763,
+      "grad_norm": 0.0005704747745767236,
+      "learning_rate": 0.0001297041581577035,
       "loss": 0.0001,
       "step": 404
     },
     {
       "epoch": 0.014490938690806305,
-      "grad_norm": 0.0015187220415100455,
-      "learning_rate": 0.0001297041581577035,
-      "loss": 0.0001,
+      "grad_norm": 0.0007336271810345352,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 0.0,
       "step": 405
     },
     {
       "epoch": 0.01452671878633916,
-      "grad_norm": 0.002080498496070504,
-      "learning_rate": 0.0001294040325232304,
-      "loss": 0.0002,
+      "grad_norm": 0.0029871645383536816,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 0.0001,
       "step": 406
     },
     {
       "epoch": 0.014562498881872015,
-      "grad_norm": 0.0009239813662134111,
-      "learning_rate": 0.00012910361668282719,
+      "grad_norm": 0.0043711112812161446,
+      "learning_rate": 0.00012880291360147693,
       "loss": 0.0001,
       "step": 407
     },
     {
       "epoch": 0.01459827897740487,
-      "grad_norm": 0.0009995809523388743,
-      "learning_rate": 0.00012880291360147693,
+      "grad_norm": 0.010784786194562912,
+      "learning_rate": 0.0001285019262469976,
       "loss": 0.0001,
       "step": 408
     },
     {
       "epoch": 0.014634059072937725,
-      "grad_norm": 0.0034637765493243933,
-      "learning_rate": 0.0001285019262469976,
+      "grad_norm": 0.006265713833272457,
+      "learning_rate": 0.00012820065759001293,
       "loss": 0.0002,
       "step": 409
     },
     {
       "epoch": 0.01466983916847058,
-      "grad_norm": 0.0018366362201049924,
-      "learning_rate": 0.00012820065759001293,
+      "grad_norm": 0.002279860433191061,
+      "learning_rate": 0.00012789911060392294,
       "loss": 0.0001,
       "step": 410
     },
     {
       "epoch": 0.014705619264003435,
-      "grad_norm": 0.0007655214285477996,
-      "learning_rate": 0.00012789911060392294,
-      "loss": 0.0001,
+      "grad_norm": 0.0015534780686721206,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.0,
       "step": 411
     },
     {
       "epoch": 0.01474139935953629,
-      "grad_norm": 0.008909448981285095,
-      "learning_rate": 0.0001275972882648746,
-      "loss": 0.0002,
+      "grad_norm": 0.0033339038491249084,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 0.0001,
       "step": 412
     },
     {
       "epoch": 0.014777179455069145,
-      "grad_norm": 0.0012444983003661036,
-      "learning_rate": 0.00012729519355173254,
-      "loss": 0.0001,
+      "grad_norm": 0.05229537934064865,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 0.0003,
       "step": 413
     },
     {
       "epoch": 0.014812959550602,
-      "grad_norm": 0.0023903120309114456,
-      "learning_rate": 0.00012699282944604967,
-      "loss": 0.0001,
+      "grad_norm": 0.08266580104827881,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 0.0008,
       "step": 414
     },
     {
       "epoch": 0.014848739646134855,
-      "grad_norm": 0.00184944411739707,
-      "learning_rate": 0.00012669019893203759,
+      "grad_norm": 0.0018661010544747114,
+      "learning_rate": 0.0001263873049965373,
       "loss": 0.0001,
       "step": 415
     },
     {
       "epoch": 0.01488451974166771,
-      "grad_norm": 0.0012689301511272788,
-      "learning_rate": 0.0001263873049965373,
+      "grad_norm": 0.0022214509081095457,
+      "learning_rate": 0.00012608415062898972,
       "loss": 0.0001,
       "step": 416
     },
     {
       "epoch": 0.014920299837200566,
-      "grad_norm": 0.25221753120422363,
-      "learning_rate": 0.00012608415062898972,
-      "loss": 0.007,
+      "grad_norm": 0.05749466270208359,
+      "learning_rate": 0.000125780738821406,
+      "loss": 0.0018,
       "step": 417
     },
     {
       "epoch": 0.01495607993273342,
-      "grad_norm": 0.0011498252861201763,
-      "learning_rate": 0.000125780738821406,
+      "grad_norm": 0.004766398575156927,
+      "learning_rate": 0.00012547707256833823,
       "loss": 0.0001,
       "step": 418
     },
     {
       "epoch": 0.014991860028266276,
-      "grad_norm": 0.0015317858196794987,
-      "learning_rate": 0.00012547707256833823,
-      "loss": 0.0001,
+      "grad_norm": 0.009270568378269672,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 0.0002,
       "step": 419
     },
     {
       "epoch": 0.01502764012379913,
-      "grad_norm": 0.002275897655636072,
-      "learning_rate": 0.00012517315486684972,
+      "grad_norm": 0.006138884928077459,
+      "learning_rate": 0.0001248689887164855,
       "loss": 0.0001,
       "step": 420
     },
     {
       "epoch": 0.015063420219331986,
-      "grad_norm": 0.003881698241457343,
-      "learning_rate": 0.0001248689887164855,
+      "grad_norm": 0.0009792116470634937,
+      "learning_rate": 0.00012456457711924266,
       "loss": 0.0001,
       "step": 421
     },
     {
       "epoch": 0.015099200314864841,
-      "grad_norm": 0.01241608802229166,
-      "learning_rate": 0.00012456457711924266,
-      "loss": 0.0002,
+      "grad_norm": 0.07432135939598083,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 0.0013,
       "step": 422
     },
     {
       "epoch": 0.015134980410397696,
-      "grad_norm": 0.0016596848145127296,
-      "learning_rate": 0.00012425992307954075,
-      "loss": 0.0001,
+      "grad_norm": 0.02271825633943081,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 0.0002,
       "step": 423
     },
     {
       "epoch": 0.015170760505930551,
-      "grad_norm": 0.0013861729530617595,
-      "learning_rate": 0.0001239550296041922,
+      "grad_norm": 0.0027169391978532076,
+      "learning_rate": 0.00012364989970237248,
       "loss": 0.0001,
       "step": 424
     },
     {
       "epoch": 0.015206540601463406,
-      "grad_norm": 0.014387309551239014,
-      "learning_rate": 0.00012364989970237248,
-      "loss": 0.0004,
+      "grad_norm": 0.8292320966720581,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.02,
       "step": 425
     },
     {
       "epoch": 0.015242320696996261,
-      "grad_norm": 0.1316864937543869,
-      "learning_rate": 0.00012334453638559057,
-      "loss": 0.0011,
+      "grad_norm": 0.002853796351701021,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 0.0001,
       "step": 426
     },
     {
       "epoch": 0.015278100792529116,
-      "grad_norm": 0.024847203865647316,
-      "learning_rate": 0.00012303894266765908,
-      "loss": 0.0003,
+      "grad_norm": 0.0020515965297818184,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 0.0001,
       "step": 427
     },
     {
       "epoch": 0.015313880888061971,
-      "grad_norm": 0.0006036808481439948,
-      "learning_rate": 0.00012273312156466464,
+      "grad_norm": 0.0009926969651132822,
+      "learning_rate": 0.00012242707609493814,
       "loss": 0.0001,
       "step": 428
     },
     {
       "epoch": 0.015349660983594826,
-      "grad_norm": 0.005189947318285704,
-      "learning_rate": 0.00012242707609493814,
-      "loss": 0.0002,
+      "grad_norm": 0.015119269490242004,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 0.0003,
       "step": 429
     },
     {
       "epoch": 0.015385441079127682,
-      "grad_norm": 0.0014469908783212304,
-      "learning_rate": 0.00012212080927902474,
+      "grad_norm": 0.003984940703958273,
+      "learning_rate": 0.00012181432413965428,
       "loss": 0.0001,
       "step": 430
     },
     {
       "epoch": 0.015421221174660537,
-      "grad_norm": 0.001749535440467298,
-      "learning_rate": 0.00012181432413965428,
+      "grad_norm": 0.0013063902733847499,
+      "learning_rate": 0.00012150762370171136,
       "loss": 0.0001,
       "step": 431
     },
     {
       "epoch": 0.015457001270193392,
-      "grad_norm": 0.001347336103208363,
-      "learning_rate": 0.00012150762370171136,
-      "loss": 0.0001,
+      "grad_norm": 0.01197239849716425,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 0.0002,
       "step": 432
     },
     {
       "epoch": 0.015492781365726247,
-      "grad_norm": 0.0013699530391022563,
-      "learning_rate": 0.00012120071099220549,
+      "grad_norm": 0.0015485131880268455,
+      "learning_rate": 0.00012089358904024117,
       "loss": 0.0001,
       "step": 433
     },
     {
       "epoch": 0.015528561461259102,
-      "grad_norm": 0.001586375990882516,
-      "learning_rate": 0.00012089358904024117,
+      "grad_norm": 0.0016194679774343967,
+      "learning_rate": 0.00012058626087698814,
       "loss": 0.0001,
       "step": 434
     },
     {
       "epoch": 0.015564341556791957,
-      "grad_norm": 0.0023582857102155685,
-      "learning_rate": 0.00012058626087698814,
-      "loss": 0.0001,
+      "grad_norm": 0.14436820149421692,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 0.0013,
       "step": 435
     },
     {
       "epoch": 0.015600121652324812,
-      "grad_norm": 0.001533446367830038,
-      "learning_rate": 0.00012027872953565125,
-      "loss": 0.0001,
+      "grad_norm": 0.01180162001401186,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 0.0002,
       "step": 436
     },
     {
       "epoch": 0.015635901747857665,
-      "grad_norm": 0.0008434976334683597,
-      "learning_rate": 0.00011997099805144069,
-      "loss": 0.0001,
+      "grad_norm": 0.0005493052303791046,
+      "learning_rate": 0.000119663069461542,
+      "loss": 0.0,
       "step": 437
     },
     {
       "epoch": 0.01567168184339052,
-      "grad_norm": 0.001348627032712102,
-      "learning_rate": 0.000119663069461542,
+      "grad_norm": 0.0012306346325203776,
+      "learning_rate": 0.00011935494680508606,
       "loss": 0.0001,
       "step": 438
     },
     {
       "epoch": 0.015707461938923375,
-      "grad_norm": 0.0014621143927797675,
-      "learning_rate": 0.00011935494680508606,
+      "grad_norm": 0.0011705871438607574,
+      "learning_rate": 0.00011904663312311901,
       "loss": 0.0001,
       "step": 439
     },
     {
       "epoch": 0.01574324203445623,
-      "grad_norm": 0.00228674802929163,
-      "learning_rate": 0.00011904663312311901,
+      "grad_norm": 0.000773779465816915,
+      "learning_rate": 0.00011873813145857249,
       "loss": 0.0001,
       "step": 440
     },
     {
       "epoch": 0.015779022129989086,
-      "grad_norm": 0.0007072591688483953,
-      "learning_rate": 0.00011873813145857249,
+      "grad_norm": 0.0009600446210242808,
+      "learning_rate": 0.00011842944485623335,
       "loss": 0.0001,
       "step": 441
     },
     {
       "epoch": 0.01581480222552194,
-      "grad_norm": 0.0018175409641116858,
-      "learning_rate": 0.00011842944485623335,
+      "grad_norm": 0.0016839904710650444,
+      "learning_rate": 0.00011812057636271374,
       "loss": 0.0001,
       "step": 442
     },
     {
       "epoch": 0.015850582321054796,
-      "grad_norm": 0.003016152884811163,
-      "learning_rate": 0.00011812057636271374,
+      "grad_norm": 0.0010681917192414403,
+      "learning_rate": 0.000117811529026421,
       "loss": 0.0001,
       "step": 443
     },
     {
       "epoch": 0.01588636241658765,
-      "grad_norm": 0.003892874578014016,
-      "learning_rate": 0.000117811529026421,
+      "grad_norm": 0.0053033060394227505,
+      "learning_rate": 0.00011750230589752762,
       "loss": 0.0001,
       "step": 444
     },
     {
       "epoch": 0.015922142512120506,
-      "grad_norm": 0.003585763741284609,
-      "learning_rate": 0.00011750230589752762,
+      "grad_norm": 0.001242525060661137,
+      "learning_rate": 0.00011719291002794096,
       "loss": 0.0001,
       "step": 445
     },
     {
       "epoch": 0.01595792260765336,
-      "grad_norm": 0.002816618187353015,
-      "learning_rate": 0.00011719291002794096,
-      "loss": 0.0001,
+      "grad_norm": 0.8490289449691772,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 0.0311,
       "step": 446
     },
     {
       "epoch": 0.015993702703186216,
-      "grad_norm": 0.0010174558265134692,
-      "learning_rate": 0.00011688334447127338,
+      "grad_norm": 0.0019403910264372826,
+      "learning_rate": 0.00011657361228281199,
       "loss": 0.0001,
       "step": 447
     },
     {
       "epoch": 0.01602948279871907,
-      "grad_norm": 0.0007345742778852582,
-      "learning_rate": 0.00011657361228281199,
-      "loss": 0.0,
-      "step": 448
+      "grad_norm": 0.0009070538799278438,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 0.0001,
+      "step": 448
     },
     {
       "epoch": 0.016065262894251926,
-      "grad_norm": 0.0008107659523375332,
-      "learning_rate": 0.00011626371651948838,
+      "grad_norm": 0.0023522900883108377,
+      "learning_rate": 0.00011595366023984864,
       "loss": 0.0001,
       "step": 449
     },
     {
       "epoch": 0.01610104298978478,
-      "grad_norm": 0.009027695283293724,
-      "learning_rate": 0.00011595366023984864,
+      "grad_norm": 0.005850024987012148,
+      "learning_rate": 0.0001156434465040231,
       "loss": 0.0002,
       "step": 450
     },
     {
       "epoch": 0.016136823085317636,
-      "grad_norm": 0.005168377421796322,
-      "learning_rate": 0.0001156434465040231,
-      "loss": 0.0002,
+      "grad_norm": 0.0020392134319990873,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.0001,
       "step": 451
     },
     {
       "epoch": 0.01617260318085049,
-      "grad_norm": 0.0012759423116222024,
-      "learning_rate": 0.00011533307837369607,
+      "grad_norm": 0.0016762950690463185,
+      "learning_rate": 0.00011502255891207572,
       "loss": 0.0001,
       "step": 452
     },
     {
       "epoch": 0.016208383276383347,
-      "grad_norm": 0.11494802683591843,
-      "learning_rate": 0.00011502255891207572,
-      "loss": 0.0006,
+      "grad_norm": 0.015178056433796883,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 0.0003,
       "step": 453
     },
     {
       "epoch": 0.0162441633719162,
-      "grad_norm": 0.004281184170395136,
-      "learning_rate": 0.00011471189118386375,
-      "loss": 0.0001,
+      "grad_norm": 0.015929946675896645,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 0.0004,
       "step": 454
     },
     {
       "epoch": 0.016279943467449057,
-      "grad_norm": 0.003102266462519765,
-      "learning_rate": 0.00011440107825522521,
-      "loss": 0.0001,
+      "grad_norm": 0.00879756174981594,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 0.0003,
       "step": 455
     },
     {
       "epoch": 0.016315723562981912,
-      "grad_norm": 0.004538760986179113,
-      "learning_rate": 0.00011409012319375827,
+      "grad_norm": 0.0025617824867367744,
+      "learning_rate": 0.0001137790290684638,
       "loss": 0.0001,
       "step": 456
     },
     {
       "epoch": 0.016351503658514767,
-      "grad_norm": 0.010061625391244888,
-      "learning_rate": 0.0001137790290684638,
-      "loss": 0.0001,
+      "grad_norm": 0.004707003477960825,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 0.0002,
       "step": 457
     },
     {
       "epoch": 0.016387283754047622,
-      "grad_norm": 0.0004581700195558369,
-      "learning_rate": 0.00011346779894971527,
-      "loss": 0.0,
+      "grad_norm": 0.002390863373875618,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 0.0001,
       "step": 458
     },
     {
       "epoch": 0.016423063849580477,
-      "grad_norm": 0.0005029060412198305,
-      "learning_rate": 0.00011315643590922827,
-      "loss": 0.0,
+      "grad_norm": 0.00811726227402687,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 0.0002,
       "step": 459
     },
     {
       "epoch": 0.016458843945113332,
-      "grad_norm": 0.0008376644109375775,
-      "learning_rate": 0.0001128449430200303,
-      "loss": 0.0001,
+      "grad_norm": 0.004762257914990187,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 0.0002,
       "step": 460
     },
     {
       "epoch": 0.016494624040646187,
-      "grad_norm": 0.0006250397418625653,
-      "learning_rate": 0.00011253332335643043,
+      "grad_norm": 0.0016065690433606505,
+      "learning_rate": 0.00011222157999398895,
       "loss": 0.0001,
       "step": 461
     },
     {
       "epoch": 0.016530404136179042,
-      "grad_norm": 0.0056080990470945835,
-      "learning_rate": 0.00011222157999398895,
-      "loss": 0.0001,
+      "grad_norm": 0.007567454129457474,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 0.0003,
       "step": 462
     },
     {
       "epoch": 0.016566184231711897,
-      "grad_norm": 0.0009173589642159641,
-      "learning_rate": 0.00011190971600948699,
-      "loss": 0.0001,
+      "grad_norm": 0.019168654456734657,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 0.0005,
       "step": 463
     },
     {
       "epoch": 0.016601964327244752,
-      "grad_norm": 0.0008181494777090847,
-      "learning_rate": 0.00011159773448089614,
+      "grad_norm": 0.0021388723980635405,
+      "learning_rate": 0.00011128563848734816,
       "loss": 0.0001,
       "step": 464
     },
     {
       "epoch": 0.016637744422777608,
-      "grad_norm": 0.00035973888589069247,
-      "learning_rate": 0.00011128563848734816,
-      "loss": 0.0,
+      "grad_norm": 0.00245377398096025,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 0.0001,
       "step": 465
     },
     {
       "epoch": 0.016673524518310463,
-      "grad_norm": 0.0007581185200251639,
-      "learning_rate": 0.00011097343110910452,
-      "loss": 0.0001,
+      "grad_norm": 0.023226840421557426,
+      "learning_rate": 0.000110661115427526,
+      "loss": 0.0003,
       "step": 466
     },
     {
       "epoch": 0.016709304613843318,
-      "grad_norm": 0.010919281281530857,
-      "learning_rate": 0.000110661115427526,
-      "loss": 0.0003,
+      "grad_norm": 0.01760197803378105,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 0.0004,
       "step": 467
     },
     {
       "epoch": 0.016745084709376173,
-      "grad_norm": 0.1749548614025116,
-      "learning_rate": 0.00011034869452504226,
-      "loss": 0.0006,
+      "grad_norm": 0.0035117343068122864,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 0.0001,
       "step": 468
     },
     {
       "epoch": 0.016780864804909028,
-      "grad_norm": 0.002724736463278532,
-      "learning_rate": 0.00011003617148512149,
-      "loss": 0.0001,
+      "grad_norm": 0.010483100078999996,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 0.0004,
       "step": 469
     },
     {
       "epoch": 0.016816644900441883,
-      "grad_norm": 0.00045277512981556356,
-      "learning_rate": 0.00010972354939223996,
-      "loss": 0.0,
+      "grad_norm": 0.019027642905712128,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 0.0004,
       "step": 470
     },
     {
       "epoch": 0.016852424995974738,
-      "grad_norm": 0.000540713022928685,
-      "learning_rate": 0.00010941083133185146,
-      "loss": 0.0,
+      "grad_norm": 0.0012662605149671435,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.0001,
       "step": 471
     },
     {
       "epoch": 0.016888205091507593,
-      "grad_norm": 0.000585521396715194,
-      "learning_rate": 0.00010909802039035701,
-      "loss": 0.0,
+      "grad_norm": 0.0016224164282903075,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 0.0001,
       "step": 472
     },
     {
       "epoch": 0.016923985187040448,
-      "grad_norm": 0.0005423309048637748,
-      "learning_rate": 0.00010878511965507434,
-      "loss": 0.0001,
+      "grad_norm": 0.005407401826232672,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 0.0002,
       "step": 473
     },
     {
       "epoch": 0.016959765282573303,
-      "grad_norm": 0.0038358867168426514,
-      "learning_rate": 0.00010847213221420736,
+      "grad_norm": 0.003952264320105314,
+      "learning_rate": 0.00010815906115681578,
       "loss": 0.0001,
       "step": 474
     },
     {
       "epoch": 0.01699554537810616,
-      "grad_norm": 0.0006008525961078703,
-      "learning_rate": 0.00010815906115681578,
-      "loss": 0.0,
+      "grad_norm": 0.0009798280661925673,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.0001,
       "step": 475
     },
     {
       "epoch": 0.017031325473639013,
-      "grad_norm": 0.00032003739033825696,
-      "learning_rate": 0.0001078459095727845,
-      "loss": 0.0,
+      "grad_norm": 0.0017176421824842691,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 0.0001,
       "step": 476
     },
     {
       "epoch": 0.01706710556917187,
-      "grad_norm": 0.0514453649520874,
-      "learning_rate": 0.00010753268055279329,
-      "loss": 0.001,
+      "grad_norm": 0.11902503669261932,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 0.0015,
       "step": 477
     },
     {
       "epoch": 0.017102885664704724,
-      "grad_norm": 0.0005073443753644824,
-      "learning_rate": 0.0001072193771882861,
+      "grad_norm": 0.0015040796715766191,
+      "learning_rate": 0.00010690600257144061,
       "loss": 0.0001,
       "step": 478
     },
     {
       "epoch": 0.01713866576023758,
-      "grad_norm": 0.0031762297730892897,
-      "learning_rate": 0.00010690600257144061,
-      "loss": 0.0001,
+      "grad_norm": 0.022858543321490288,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 0.0003,
       "step": 479
     },
     {
       "epoch": 0.017174445855770434,
-      "grad_norm": 0.0006731804460287094,
-      "learning_rate": 0.0001065925597951378,
+      "grad_norm": 0.0016762394225224853,
+      "learning_rate": 0.00010627905195293135,
       "loss": 0.0001,
       "step": 480
     },
     {
       "epoch": 0.01721022595130329,
-      "grad_norm": 0.0019265701994299889,
-      "learning_rate": 0.00010627905195293135,
+      "grad_norm": 0.003237987868487835,
+      "learning_rate": 0.00010596548213901708,
       "loss": 0.0001,
       "step": 481
     },
     {
       "epoch": 0.017246006046836144,
-      "grad_norm": 0.0010831786785274744,
-      "learning_rate": 0.00010596548213901708,
+      "grad_norm": 0.0022923192009329796,
+      "learning_rate": 0.00010565185344820247,
       "loss": 0.0001,
       "step": 482
     },
     {
       "epoch": 0.017281786142369,
-      "grad_norm": 0.004170823376625776,
-      "learning_rate": 0.00010565185344820247,
-      "loss": 0.0001,
+      "grad_norm": 0.20378467440605164,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 0.0047,
       "step": 483
     },
     {
       "epoch": 0.017317566237901854,
-      "grad_norm": 0.0005037914961576462,
-      "learning_rate": 0.00010533816897587606,
-      "loss": 0.0,
+      "grad_norm": 0.0037489025853574276,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 0.0001,
       "step": 484
     },
     {
       "epoch": 0.01735334633343471,
-      "grad_norm": 0.0014839319046586752,
-      "learning_rate": 0.00010502443181797697,
+      "grad_norm": 0.001064359094016254,
+      "learning_rate": 0.00010471064507096426,
       "loss": 0.0001,
       "step": 485
     },
     {
       "epoch": 0.017389126428967564,
-      "grad_norm": 0.0003253934846725315,
-      "learning_rate": 0.00010471064507096426,
+      "grad_norm": 0.000837884726934135,
+      "learning_rate": 0.0001043968118317865,
       "loss": 0.0,
       "step": 486
     },
     {
       "epoch": 0.01742490652450042,
-      "grad_norm": 0.0012966105714440346,
-      "learning_rate": 0.0001043968118317865,
-      "loss": 0.0001,
+      "grad_norm": 0.2281397432088852,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 0.0009,
       "step": 487
     },
     {
       "epoch": 0.017460686620033274,
-      "grad_norm": 0.0018304159166291356,
-      "learning_rate": 0.00010408293519785101,
-      "loss": 0.0001,
+      "grad_norm": 0.019389115273952484,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 0.0003,
       "step": 488
     },
     {
       "epoch": 0.01749646671556613,
-      "grad_norm": 0.0005496956291608512,
-      "learning_rate": 0.00010376901826699348,
+      "grad_norm": 0.0031479692552238703,
+      "learning_rate": 0.00010345506413744726,
       "loss": 0.0001,
       "step": 489
     },
     {
       "epoch": 0.017532246811098984,
-      "grad_norm": 0.0008559661218896508,
-      "learning_rate": 0.00010345506413744726,
+      "grad_norm": 0.004489561077207327,
+      "learning_rate": 0.00010314107590781284,
       "loss": 0.0001,
       "step": 490
     },
     {
       "epoch": 0.01756802690663184,
-      "grad_norm": 0.0006337561644613743,
-      "learning_rate": 0.00010314107590781284,
-      "loss": 0.0,
+      "grad_norm": 0.005492003634572029,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.0001,
       "step": 491
     },
     {
       "epoch": 0.017603807002164695,
-      "grad_norm": 0.0025075278244912624,
-      "learning_rate": 0.00010282705667702734,
+      "grad_norm": 0.0008158302516676486,
+      "learning_rate": 0.00010251300954433376,
       "loss": 0.0001,
       "step": 492
     },
     {
       "epoch": 0.01763958709769755,
-      "grad_norm": 0.5806146860122681,
-      "learning_rate": 0.00010251300954433376,
-      "loss": 0.0021,
+      "grad_norm": 0.016751619055867195,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 0.0002,
       "step": 493
     },
     {
       "epoch": 0.017675367193230405,
-      "grad_norm": 0.0006832131184637547,
-      "learning_rate": 0.00010219893760925052,
+      "grad_norm": 0.001021690433844924,
+      "learning_rate": 0.00010188484397154084,
       "loss": 0.0001,
       "step": 494
     },
     {
       "epoch": 0.01771114728876326,
-      "grad_norm": 0.005751829594373703,
-      "learning_rate": 0.00010188484397154084,
-      "loss": 0.0002,
+      "grad_norm": 0.008309063501656055,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 0.0003,
       "step": 495
     },
     {
       "epoch": 0.017746927384296115,
-      "grad_norm": 0.0013581750681623816,
-      "learning_rate": 0.00010157073173118208,
-      "loss": 0.0001,
+      "grad_norm": 3.1554737091064453,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 0.0277,
       "step": 496
     },
     {
       "epoch": 0.01778270747982897,
-      "grad_norm": 0.0002741411735769361,
-      "learning_rate": 0.00010125660398833528,
+      "grad_norm": 0.0003272468165960163,
+      "learning_rate": 0.00010094246384331442,
       "loss": 0.0,
       "step": 497
     },
     {
       "epoch": 0.017818487575361825,
-      "grad_norm": 0.0002076040836982429,
-      "learning_rate": 0.00010094246384331442,
+      "grad_norm": 0.0007006549858488142,
+      "learning_rate": 0.00010062831439655591,
       "loss": 0.0,
       "step": 498
     },
     {
       "epoch": 0.01785426767089468,
-      "grad_norm": 0.0015356852672994137,
-      "learning_rate": 0.00010062831439655591,
+      "grad_norm": 0.0005305926315486431,
+      "learning_rate": 0.00010031415874858797,
       "loss": 0.0001,
       "step": 499
     },
     {
       "epoch": 0.017890047766427535,
-      "grad_norm": 0.0011746383970603347,
-      "learning_rate": 0.00010031415874858797,
+      "grad_norm": 0.0013227920280769467,
+      "learning_rate": 0.0001,
       "loss": 0.0001,
       "step": 500
     },
     {
       "epoch": 0.01792582786196039,
-      "grad_norm": 0.0005788155249319971,
-      "learning_rate": 0.0001,
+      "grad_norm": 0.001107820076867938,
+      "learning_rate": 9.968584125141204e-05,
       "loss": 0.0001,
       "step": 501
     },
     {
       "epoch": 0.017961607957493245,
-      "grad_norm": 0.0005044927820563316,
-      "learning_rate": 9.968584125141204e-05,
+      "grad_norm": 0.001051326747983694,
+      "learning_rate": 9.937168560344412e-05,
       "loss": 0.0001,
       "step": 502
     },
     {
       "epoch": 0.0179973880530261,
-      "grad_norm": 0.00038509120349772274,
-      "learning_rate": 9.937168560344412e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0029785106889903545,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 0.0001,
       "step": 503
     },
     {
       "epoch": 0.018033168148558956,
-      "grad_norm": 0.0004976564086973667,
-      "learning_rate": 9.90575361566856e-05,
+      "grad_norm": 0.0009764096466824412,
+      "learning_rate": 9.874339601166473e-05,
       "loss": 0.0,
       "step": 504
     },
     {
       "epoch": 0.01806894824409181,
-      "grad_norm": 0.06314973533153534,
-      "learning_rate": 9.874339601166473e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0004290563229005784,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 0.0,
       "step": 505
     },
     {
       "epoch": 0.018104728339624666,
-      "grad_norm": 0.0003536037402227521,
-      "learning_rate": 9.842926826881796e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0006004414171911776,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 0.0001,
       "step": 506
     },
     {
       "epoch": 0.01814050843515752,
-      "grad_norm": 0.0007825393113307655,
-      "learning_rate": 9.81151560284592e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0011966415913775563,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 0.0001,
       "step": 507
     },
     {
       "epoch": 0.018176288530690376,
-      "grad_norm": 0.00037882779724895954,
-      "learning_rate": 9.78010623907495e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0027027500327676535,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 0.0001,
       "step": 508
     },
     {
       "epoch": 0.01821206862622323,
-      "grad_norm": 0.00039204652421176434,
-      "learning_rate": 9.748699045566626e-05,
-      "loss": 0.0,
+      "grad_norm": 0.008526821620762348,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 0.0002,
       "step": 509
     },
     {
       "epoch": 0.018247848721756086,
-      "grad_norm": 0.0012885596370324492,
-      "learning_rate": 9.717294332297268e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.009162348695099354,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 0.0003,
       "step": 510
     },
     {
       "epoch": 0.01828362881728894,
-      "grad_norm": 0.0017713923007249832,
-      "learning_rate": 9.685892409218717e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.14034569263458252,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.0003,
       "step": 511
     },
     {
       "epoch": 0.018319408912821796,
-      "grad_norm": 0.0014103096909821033,
-      "learning_rate": 9.654493586255278e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.012582145631313324,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 0.0002,
       "step": 512
     },
     {
       "epoch": 0.01835518900835465,
-      "grad_norm": 0.00025732445647008717,
-      "learning_rate": 9.623098173300654e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0017335236771032214,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 0.0001,
       "step": 513
     },
     {
       "epoch": 0.018390969103887506,
-      "grad_norm": 0.00040602468652650714,
-      "learning_rate": 9.591706480214901e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0018851887434720993,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.0001,
       "step": 514
     },
     {
       "epoch": 0.01842674919942036,
-      "grad_norm": 0.005421795882284641,
-      "learning_rate": 9.560318816821353e-05,
+      "grad_norm": 0.003452875418588519,
+      "learning_rate": 9.528935492903575e-05,
       "loss": 0.0001,
       "step": 515
     },
     {
       "epoch": 0.018462529294953216,
-      "grad_norm": 0.0008320416673086584,
-      "learning_rate": 9.528935492903575e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.005577889736741781,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 0.0002,
       "step": 516
     },
     {
       "epoch": 0.01849830939048607,
-      "grad_norm": 0.018853820860385895,
-      "learning_rate": 9.497556818202306e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.047353558242321014,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 0.0008,
       "step": 517
     },
     {
       "epoch": 0.018534089486018927,
-      "grad_norm": 0.0005207477952353656,
-      "learning_rate": 9.466183102412395e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0012205071980133653,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 0.0001,
       "step": 518
     },
     {
       "epoch": 0.01856986958155178,
-      "grad_norm": 0.0004097981145605445,
-      "learning_rate": 9.434814655179755e-05,
-      "loss": 0.0,
+      "grad_norm": 0.00221634516492486,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 0.0001,
       "step": 519
     },
     {
       "epoch": 0.018605649677084637,
-      "grad_norm": 0.0018601935589686036,
-      "learning_rate": 9.403451786098294e-05,
+      "grad_norm": 0.0012431563809514046,
+      "learning_rate": 9.372094804706867e-05,
       "loss": 0.0001,
       "step": 520
     },
     {
       "epoch": 0.018641429772617492,
-      "grad_norm": 0.0036646395456045866,
-      "learning_rate": 9.372094804706867e-05,
-      "loss": 0.0,
+      "grad_norm": 0.011209934949874878,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.0002,
       "step": 521
     },
     {
       "epoch": 0.018677209868150347,
-      "grad_norm": 0.0011853273026645184,
-      "learning_rate": 9.340744020486222e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0011534786317497492,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 0.0001,
       "step": 522
     },
     {
       "epoch": 0.018712989963683202,
-      "grad_norm": 0.0003728955052793026,
-      "learning_rate": 9.309399742855942e-05,
-      "loss": 0.0,
+      "grad_norm": 0.001877554226666689,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 0.0001,
       "step": 523
     },
     {
       "epoch": 0.018748770059216057,
-      "grad_norm": 0.00022109539713710546,
-      "learning_rate": 9.278062281171393e-05,
+      "grad_norm": 0.0005466327420435846,
+      "learning_rate": 9.246731944720675e-05,
       "loss": 0.0,
       "step": 524
     },
     {
       "epoch": 0.018784550154748912,
-      "grad_norm": 0.0012523216428235173,
-      "learning_rate": 9.246731944720675e-05,
+      "grad_norm": 0.00111495575401932,
+      "learning_rate": 9.215409042721552e-05,
       "loss": 0.0001,
       "step": 525
     },
     {
       "epoch": 0.018820330250281767,
-      "grad_norm": 0.0011476062936708331,
-      "learning_rate": 9.215409042721552e-05,
+      "grad_norm": 0.0009859137935563922,
+      "learning_rate": 9.184093884318425e-05,
       "loss": 0.0001,
       "step": 526
     },
     {
       "epoch": 0.018856110345814622,
-      "grad_norm": 0.0009583460632711649,
-      "learning_rate": 9.184093884318425e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0020015740301460028,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 0.0001,
       "step": 527
     },
     {
       "epoch": 0.018891890441347477,
-      "grad_norm": 0.0008978561381809413,
-      "learning_rate": 9.152786778579267e-05,
-      "loss": 0.0,
+      "grad_norm": 0.14584803581237793,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 0.0011,
       "step": 528
     },
     {
       "epoch": 0.018927670536880332,
-      "grad_norm": 0.005942204967141151,
-      "learning_rate": 9.121488034492569e-05,
-      "loss": 0.0001,
+      "grad_norm": 2.284058094024658,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 0.0192,
       "step": 529
     },
     {
       "epoch": 0.018963450632413188,
-      "grad_norm": 0.00018503722094465047,
-      "learning_rate": 9.090197960964301e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0032946874853223562,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 0.0001,
       "step": 530
     },
     {
       "epoch": 0.018999230727946043,
-      "grad_norm": 0.0003717237268574536,
-      "learning_rate": 9.058916866814858e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0009048773790709674,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.0001,
       "step": 531
     },
     {
       "epoch": 0.019035010823478898,
-      "grad_norm": 0.0034299001563340425,
-      "learning_rate": 9.027645060776006e-05,
+      "grad_norm": 0.0011141895083710551,
+      "learning_rate": 8.99638285148785e-05,
       "loss": 0.0001,
       "step": 532
     },
     {
       "epoch": 0.019070790919011753,
-      "grad_norm": 0.000308646762277931,
-      "learning_rate": 8.99638285148785e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0008086450980044901,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 0.0001,
       "step": 533
     },
     {
       "epoch": 0.019106571014544608,
-      "grad_norm": 0.0022994359023869038,
-      "learning_rate": 8.965130547495776e-05,
+      "grad_norm": 0.0009894418762996793,
+      "learning_rate": 8.933888457247402e-05,
       "loss": 0.0001,
       "step": 534
     },
     {
       "epoch": 0.019142351110077463,
-      "grad_norm": 1.244136929512024,
-      "learning_rate": 8.933888457247402e-05,
-      "loss": 0.0055,
+      "grad_norm": 0.10656411945819855,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 0.0008,
       "step": 535
     },
     {
       "epoch": 0.019178131205610318,
-      "grad_norm": 0.018538277596235275,
-      "learning_rate": 8.902656889089548e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00936040747910738,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 0.0002,
       "step": 536
     },
     {
       "epoch": 0.019213911301143173,
-      "grad_norm": 0.00022319707204587758,
-      "learning_rate": 8.871436151265184e-05,
+      "grad_norm": 0.00042610568925738335,
+      "learning_rate": 8.840226551910387e-05,
       "loss": 0.0,
       "step": 537
     },
     {
       "epoch": 0.019249691396676028,
-      "grad_norm": 0.05897924304008484,
-      "learning_rate": 8.840226551910387e-05,
-      "loss": 0.0008,
+      "grad_norm": 0.1127459779381752,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 0.0015,
       "step": 538
     },
     {
       "epoch": 0.019285471492208883,
-      "grad_norm": 0.0004285780305508524,
-      "learning_rate": 8.809028399051302e-05,
-      "loss": 0.0,
+      "grad_norm": 0.31594935059547424,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 0.0049,
       "step": 539
     },
     {
       "epoch": 0.01932125158774174,
-      "grad_norm": 0.0005444411071948707,
-      "learning_rate": 8.777842000601105e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0016649545868858695,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 0.0001,
       "step": 540
     },
     {
       "epoch": 0.019357031683274593,
-      "grad_norm": 0.3667220175266266,
-      "learning_rate": 8.746667664356956e-05,
-      "loss": 0.0011,
+      "grad_norm": 0.01569303870201111,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.0002,
       "step": 541
     },
     {
       "epoch": 0.01939281177880745,
-      "grad_norm": 0.002283938927575946,
-      "learning_rate": 8.715505697996971e-05,
-      "loss": 0.0,
+      "grad_norm": 0.004535132087767124,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.0001,
       "step": 542
     },
     {
       "epoch": 0.019428591874340304,
-      "grad_norm": 0.007592503447085619,
-      "learning_rate": 8.684356409077176e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.013089085929095745,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 0.0003,
       "step": 543
     },
     {
       "epoch": 0.01946437196987316,
-      "grad_norm": 0.0003751645563170314,
-      "learning_rate": 8.653220105028474e-05,
-      "loss": 0.0,
+      "grad_norm": 0.002690235385671258,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.0001,
       "step": 544
     },
     {
       "epoch": 0.019500152065406014,
-      "grad_norm": 0.007481827866286039,
-      "learning_rate": 8.62209709315362e-05,
+      "grad_norm": 0.00139359082095325,
+      "learning_rate": 8.590987680624174e-05,
       "loss": 0.0001,
       "step": 545
     },
     {
       "epoch": 0.01953593216093887,
-      "grad_norm": 0.4077150523662567,
-      "learning_rate": 8.590987680624174e-05,
-      "loss": 0.006,
+      "grad_norm": 0.3601469099521637,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 0.0063,
       "step": 546
     },
     {
       "epoch": 0.019571712256471724,
-      "grad_norm": 0.6080543398857117,
-      "learning_rate": 8.559892174477479e-05,
-      "loss": 0.0017,
+      "grad_norm": 0.0005300491466186941,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 0.0,
       "step": 547
     },
     {
       "epoch": 0.01960749235200458,
-      "grad_norm": 0.010074864141643047,
-      "learning_rate": 8.528810881613626e-05,
+      "grad_norm": 0.004550674930214882,
+      "learning_rate": 8.497744108792429e-05,
       "loss": 0.0001,
       "step": 548
     },
     {
       "epoch": 0.019643272447537434,
-      "grad_norm": 0.0048607406206429005,
-      "learning_rate": 8.497744108792429e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.21630121767520905,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 0.0014,
       "step": 549
     },
     {
       "epoch": 0.01967905254307029,
-      "grad_norm": 0.0009256759658455849,
-      "learning_rate": 8.466692162630392e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0005254748393781483,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.0001,
       "step": 550
     },
     {
       "epoch": 0.019714832638603144,
-      "grad_norm": 0.0008813950698822737,
-      "learning_rate": 8.435655349597689e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0007254169904626906,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.0001,
       "step": 551
     },
     {
       "epoch": 0.019750612734136,
-      "grad_norm": 0.0013631863985210657,
-      "learning_rate": 8.404633976015134e-05,
-      "loss": 0.0001,
+      "grad_norm": 1.2918907403945923,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 0.005,
       "step": 552
     },
     {
       "epoch": 0.019786392829668854,
-      "grad_norm": 0.00032280024606734514,
-      "learning_rate": 8.373628348051165e-05,
+      "grad_norm": 0.0005183537723496556,
+      "learning_rate": 8.342638771718802e-05,
       "loss": 0.0,
       "step": 553
     },
     {
       "epoch": 0.01982217292520171,
-      "grad_norm": 0.0005799188511446118,
-      "learning_rate": 8.342638771718802e-05,
+      "grad_norm": 0.0008759180782362819,
+      "learning_rate": 8.311665552872662e-05,
       "loss": 0.0001,
       "step": 554
     },
     {
       "epoch": 0.019857953020734564,
-      "grad_norm": 0.0010495373280718923,
-      "learning_rate": 8.311665552872662e-05,
+      "grad_norm": 0.0011825045803561807,
+      "learning_rate": 8.280708997205904e-05,
       "loss": 0.0,
       "step": 555
     },
     {
       "epoch": 0.01989373311626742,
-      "grad_norm": 0.0010906853713095188,
-      "learning_rate": 8.280708997205904e-05,
+      "grad_norm": 0.0011933365603908896,
+      "learning_rate": 8.249769410247239e-05,
       "loss": 0.0001,
       "step": 556
     },
     {
       "epoch": 0.019929513211800275,
-      "grad_norm": 0.015486394055187702,
-      "learning_rate": 8.249769410247239e-05,
+      "grad_norm": 0.0029679578728973866,
+      "learning_rate": 8.218847097357898e-05,
       "loss": 0.0002,
       "step": 557
     },
     {
       "epoch": 0.01996529330733313,
-      "grad_norm": 0.0028266129083931446,
-      "learning_rate": 8.218847097357898e-05,
+      "grad_norm": 0.002634399337694049,
+      "learning_rate": 8.187942363728625e-05,
       "loss": 0.0001,
       "step": 558
     },
     {
       "epoch": 0.020001073402865985,
-      "grad_norm": 0.0004184865683782846,
-      "learning_rate": 8.187942363728625e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0009519772138446569,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 0.0001,
       "step": 559
     },
     {
       "epoch": 0.02003685349839884,
-      "grad_norm": 0.0015919478610157967,
-      "learning_rate": 8.157055514376666e-05,
+      "grad_norm": 0.012963998131453991,
+      "learning_rate": 8.126186854142752e-05,
       "loss": 0.0001,
       "step": 560
     },
     {
       "epoch": 0.020072633593931695,
-      "grad_norm": 0.001232324168086052,
-      "learning_rate": 8.126186854142752e-05,
+      "grad_norm": 0.00807272456586361,
+      "learning_rate": 8.095336687688102e-05,
       "loss": 0.0001,
       "step": 561
     },
     {
       "epoch": 0.02010841368946455,
-      "grad_norm": 0.0008730905828997493,
-      "learning_rate": 8.095336687688102e-05,
-      "loss": 0.0,
+      "grad_norm": 0.001012706314213574,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 0.0001,
       "step": 562
     },
     {
       "epoch": 0.020144193784997405,
-      "grad_norm": 0.0004459127376321703,
-      "learning_rate": 8.064505319491398e-05,
-      "loss": 0.0,
+      "grad_norm": 0.004604503512382507,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 0.0001,
       "step": 563
     },
     {
       "epoch": 0.02017997388053026,
-      "grad_norm": 0.0011548546608537436,
-      "learning_rate": 8.033693053845801e-05,
+      "grad_norm": 0.000794455932918936,
+      "learning_rate": 8.002900194855932e-05,
       "loss": 0.0001,
       "step": 564
     },
     {
       "epoch": 0.020215753976063115,
-      "grad_norm": 0.0012672626180574298,
-      "learning_rate": 8.002900194855932e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0009159276960417628,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 0.0,
       "step": 565
     },
     {
       "epoch": 0.02025153407159597,
-      "grad_norm": 0.20966456830501556,
-      "learning_rate": 7.972127046434878e-05,
-      "loss": 0.001,
+      "grad_norm": 0.004088504705578089,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 0.0002,
       "step": 566
     },
     {
       "epoch": 0.020287314167128825,
-      "grad_norm": 0.00276308786123991,
-      "learning_rate": 7.941373912301189e-05,
+      "grad_norm": 0.0018173141870647669,
+      "learning_rate": 7.910641095975886e-05,
       "loss": 0.0001,
       "step": 567
     },
     {
       "epoch": 0.02032309426266168,
-      "grad_norm": 0.0008582459995523095,
-      "learning_rate": 7.910641095975886e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00044192952918820083,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 0.0,
       "step": 568
     },
     {
       "epoch": 0.020358874358194536,
-      "grad_norm": 0.0016381151508539915,
-      "learning_rate": 7.879928900779456e-05,
+      "grad_norm": 0.0005358615890145302,
+      "learning_rate": 7.849237629828869e-05,
       "loss": 0.0001,
       "step": 569
     },
     {
       "epoch": 0.02039465445372739,
-      "grad_norm": 0.0029155344236642122,
-      "learning_rate": 7.849237629828869e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.024256188422441483,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 0.0002,
       "step": 570
     },
     {
       "epoch": 0.020430434549260246,
-      "grad_norm": 0.12655425071716309,
-      "learning_rate": 7.818567586034577e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.0016058129258453846,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.0001,
       "step": 571
     },
     {
       "epoch": 0.0204662146447931,
-      "grad_norm": 0.0015873904339969158,
-      "learning_rate": 7.787919072097531e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00042367077548988163,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 0.0,
       "step": 572
     },
     {
       "epoch": 0.020501994740325956,
-      "grad_norm": 0.000485215219669044,
-      "learning_rate": 7.75729239050619e-05,
+      "grad_norm": 0.0005126690957695246,
+      "learning_rate": 7.726687843533538e-05,
       "loss": 0.0,
       "step": 573
     },
     {
       "epoch": 0.02053777483585881,
-      "grad_norm": 0.002026630798354745,
-      "learning_rate": 7.726687843533538e-05,
+      "grad_norm": 0.0016349004581570625,
+      "learning_rate": 7.696105733234098e-05,
       "loss": 0.0001,
       "step": 574
     },
     {
       "epoch": 0.020573554931391666,
-      "grad_norm": 0.0015160428592935205,
-      "learning_rate": 7.696105733234098e-05,
+      "grad_norm": 0.0010834066197276115,
+      "learning_rate": 7.66554636144095e-05,
       "loss": 0.0001,
       "step": 575
     },
     {
       "epoch": 0.02060933502692452,
-      "grad_norm": 0.001741651096381247,
-      "learning_rate": 7.66554636144095e-05,
+      "grad_norm": 0.001176648074761033,
+      "learning_rate": 7.635010029762756e-05,
       "loss": 0.0001,
       "step": 576
     },
     {
       "epoch": 0.020645115122457376,
-      "grad_norm": 0.002401455072686076,
-      "learning_rate": 7.635010029762756e-05,
+      "grad_norm": 0.0007962558884173632,
+      "learning_rate": 7.604497039580785e-05,
       "loss": 0.0001,
       "step": 577
     },
     {
       "epoch": 0.02068089521799023,
-      "grad_norm": 0.0009681973024271429,
-      "learning_rate": 7.604497039580785e-05,
+      "grad_norm": 0.0017486009746789932,
+      "learning_rate": 7.574007692045928e-05,
       "loss": 0.0001,
       "step": 578
     },
     {
       "epoch": 0.020716675313523086,
-      "grad_norm": 0.002101348014548421,
-      "learning_rate": 7.574007692045928e-05,
+      "grad_norm": 0.0006156065501272678,
+      "learning_rate": 7.543542288075739e-05,
       "loss": 0.0001,
       "step": 579
     },
     {
       "epoch": 0.02075245540905594,
-      "grad_norm": 0.001094807405024767,
-      "learning_rate": 7.543542288075739e-05,
+      "grad_norm": 0.0005485773435793817,
+      "learning_rate": 7.513101128351454e-05,
       "loss": 0.0001,
       "step": 580
     },
     {
       "epoch": 0.020788235504588796,
-      "grad_norm": 0.02314869873225689,
-      "learning_rate": 7.513101128351454e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.0023702112957835197,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.0001,
       "step": 581
     },
     {
       "epoch": 0.02082401560012165,
-      "grad_norm": 0.023619646206498146,
-      "learning_rate": 7.48268451331503e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.0008253026171587408,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 0.0001,
       "step": 582
     },
     {
       "epoch": 0.020859795695654507,
-      "grad_norm": 0.0786769688129425,
-      "learning_rate": 7.45229274316618e-05,
-      "loss": 0.0007,
+      "grad_norm": 0.0005967675824649632,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 0.0,
       "step": 583
     },
     {
       "epoch": 0.02089557579118736,
-      "grad_norm": 0.0013753515668213367,
-      "learning_rate": 7.421926117859403e-05,
+      "grad_norm": 0.0028061717748641968,
+      "learning_rate": 7.391584937101033e-05,
       "loss": 0.0001,
       "step": 584
     },
     {
       "epoch": 0.020931355886720217,
-      "grad_norm": 0.0017551406053826213,
-      "learning_rate": 7.391584937101033e-05,
+      "grad_norm": 0.00168716034386307,
+      "learning_rate": 7.361269500346274e-05,
       "loss": 0.0001,
       "step": 585
     },
     {
       "epoch": 0.020967135982253072,
-      "grad_norm": 0.0011576368706300855,
-      "learning_rate": 7.361269500346274e-05,
+      "grad_norm": 0.0015768698649480939,
+      "learning_rate": 7.330980106796246e-05,
       "loss": 0.0001,
       "step": 586
     },
     {
       "epoch": 0.021002916077785927,
-      "grad_norm": 0.0665171816945076,
-      "learning_rate": 7.330980106796246e-05,
-      "loss": 0.0005,
+      "grad_norm": 0.001126199378632009,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 0.0001,
       "step": 587
     },
     {
       "epoch": 0.021038696173318782,
-      "grad_norm": 0.0012440182035788894,
-      "learning_rate": 7.300717055395039e-05,
+      "grad_norm": 0.001138276536948979,
+      "learning_rate": 7.270480644826749e-05,
       "loss": 0.0001,
       "step": 588
     },
     {
       "epoch": 0.021074476268851637,
-      "grad_norm": 0.002847005147486925,
-      "learning_rate": 7.270480644826749e-05,
+      "grad_norm": 0.002184976125136018,
+      "learning_rate": 7.240271173512546e-05,
       "loss": 0.0001,
       "step": 589
     },
     {
       "epoch": 0.021110256364384492,
-      "grad_norm": 0.000652329355943948,
-      "learning_rate": 7.240271173512546e-05,
-      "loss": 0.0,
+      "grad_norm": 0.005373840685933828,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 0.0001,
       "step": 590
     },
     {
       "epoch": 0.021146036459917347,
-      "grad_norm": 0.9554212093353271,
-      "learning_rate": 7.210088939607708e-05,
-      "loss": 0.0086,
+      "grad_norm": 0.021403763443231583,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.0003,
       "step": 591
     },
     {
       "epoch": 0.021181816555450202,
-      "grad_norm": 0.0006063065375201404,
-      "learning_rate": 7.179934240998706e-05,
+      "grad_norm": 0.0008738846518099308,
+      "learning_rate": 7.149807375300239e-05,
       "loss": 0.0001,
       "step": 592
     },
     {
       "epoch": 0.021217596650983057,
-      "grad_norm": 0.05093079060316086,
-      "learning_rate": 7.149807375300239e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.01297273300588131,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 0.0001,
       "step": 593
     },
     {
       "epoch": 0.021253376746515912,
-      "grad_norm": 0.0011723345378413796,
-      "learning_rate": 7.119708639852312e-05,
+      "grad_norm": 0.0009190509445033967,
+      "learning_rate": 7.089638331717284e-05,
       "loss": 0.0001,
       "step": 594
     },
     {
       "epoch": 0.021289156842048768,
-      "grad_norm": 0.0012588248355314136,
-      "learning_rate": 7.089638331717284e-05,
+      "grad_norm": 0.0008834196487441659,
+      "learning_rate": 7.059596747676962e-05,
       "loss": 0.0001,
       "step": 595
     },
     {
       "epoch": 0.021324936937581623,
-      "grad_norm": 0.01622002013027668,
-      "learning_rate": 7.059596747676962e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.0009299754165112972,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 0.0001,
       "step": 596
     },
     {
       "epoch": 0.021360717033114478,
-      "grad_norm": 0.001236697193235159,
-      "learning_rate": 7.029584184229653e-05,
+      "grad_norm": 0.0009522477630525827,
+      "learning_rate": 6.999600937587239e-05,
       "loss": 0.0001,
       "step": 597
     },
     {
       "epoch": 0.021396497128647333,
-      "grad_norm": 0.00358847645111382,
-      "learning_rate": 6.999600937587239e-05,
+      "grad_norm": 0.001419812091626227,
+      "learning_rate": 6.969647303672262e-05,
       "loss": 0.0001,
       "step": 598
     },
     {
       "epoch": 0.021432277224180188,
-      "grad_norm": 0.0012909158831462264,
-      "learning_rate": 6.969647303672262e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.014936032705008984,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 0.0003,
       "step": 599
     },
     {
       "epoch": 0.021468057319713043,
-      "grad_norm": 0.000985512277111411,
-      "learning_rate": 6.939723578114993e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0007730189245194197,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.0,
       "step": 600
     },
     {
       "epoch": 0.021503837415245898,
-      "grad_norm": 0.001460637548007071,
-      "learning_rate": 6.909830056250527e-05,
+      "grad_norm": 0.0008386225090362132,
+      "learning_rate": 6.879967033115853e-05,
       "loss": 0.0001,
       "step": 601
     },
     {
       "epoch": 0.021539617510778753,
-      "grad_norm": 0.0011585768079385161,
-      "learning_rate": 6.879967033115853e-05,
+      "grad_norm": 0.0006314528291113675,
+      "learning_rate": 6.850134803446954e-05,
       "loss": 0.0001,
       "step": 602
     },
     {
       "epoch": 0.021575397606311608,
-      "grad_norm": 0.0038794241845607758,
-      "learning_rate": 6.850134803446954e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.004665585700422525,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 0.0002,
       "step": 603
     },
     {
       "epoch": 0.021611177701844463,
-      "grad_norm": 0.0004806022916454822,
-      "learning_rate": 6.820333661675893e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0008167759515345097,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 0.0001,
       "step": 604
     },
     {
       "epoch": 0.02164695779737732,
-      "grad_norm": 0.0005523916915990412,
-      "learning_rate": 6.790563901927907e-05,
+      "grad_norm": 0.0006439369171857834,
+      "learning_rate": 6.760825818018508e-05,
       "loss": 0.0001,
       "step": 605
     },
     {
       "epoch": 0.021682737892910173,
-      "grad_norm": 0.00164729671087116,
-      "learning_rate": 6.760825818018508e-05,
+      "grad_norm": 0.0010109387803822756,
+      "learning_rate": 6.731119703450577e-05,
       "loss": 0.0001,
       "step": 606
     },
     {
       "epoch": 0.02171851798844303,
-      "grad_norm": 0.00039394403574988246,
-      "learning_rate": 6.731119703450577e-05,
+      "grad_norm": 0.00030730641447007656,
+      "learning_rate": 6.701445851411472e-05,
       "loss": 0.0,
       "step": 607
     },
     {
       "epoch": 0.021754298083975884,
-      "grad_norm": 0.00944399181753397,
-      "learning_rate": 6.701445851411472e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.003990877885371447,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 0.0002,
       "step": 608
     },
     {
       "epoch": 0.02179007817950874,
-      "grad_norm": 0.0006950633833184838,
-      "learning_rate": 6.671804554770135e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00034466804936528206,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 0.0,
       "step": 609
     },
     {
       "epoch": 0.021825858275041594,
-      "grad_norm": 0.22722935676574707,
-      "learning_rate": 6.642196106074194e-05,
-      "loss": 0.0022,
+      "grad_norm": 0.001726245740428567,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 0.0001,
       "step": 610
     },
     {
       "epoch": 0.02186163837057445,
-      "grad_norm": 0.003627099795266986,
-      "learning_rate": 6.612620797547087e-05,
+      "grad_norm": 0.001501871389336884,
+      "learning_rate": 6.583078921085167e-05,
       "loss": 0.0001,
       "step": 611
     },
     {
       "epoch": 0.021897418466107304,
-      "grad_norm": 0.0012314224150031805,
-      "learning_rate": 6.583078921085167e-05,
+      "grad_norm": 0.0008260820177383721,
+      "learning_rate": 6.55357076825483e-05,
       "loss": 0.0001,
       "step": 612
     },
     {
       "epoch": 0.02193319856164016,
-      "grad_norm": 0.0005978898843750358,
-      "learning_rate": 6.55357076825483e-05,
-      "loss": 0.0,
+      "grad_norm": 0.01785612665116787,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 0.0002,
       "step": 613
     },
     {
       "epoch": 0.021968978657173014,
-      "grad_norm": 0.016302533447742462,
-      "learning_rate": 6.52409663028963e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.002857221057638526,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 0.0,
       "step": 614
     },
     {
       "epoch": 0.02200475875270587,
-      "grad_norm": 0.2945321202278137,
-      "learning_rate": 6.494656798087412e-05,
-      "loss": 0.0013,
+      "grad_norm": 0.0006424558814615011,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 0.0,
       "step": 615
     },
     {
       "epoch": 0.022040538848238724,
-      "grad_norm": 0.0009686471312306821,
-      "learning_rate": 6.465251562207431e-05,
+      "grad_norm": 0.0006789982435293496,
+      "learning_rate": 6.435881212867493e-05,
       "loss": 0.0001,
       "step": 616
     },
     {
       "epoch": 0.02207631894377158,
-      "grad_norm": 0.0006426291074603796,
-      "learning_rate": 6.435881212867493e-05,
+      "grad_norm": 0.0005097879911772907,
+      "learning_rate": 6.406546039941094e-05,
       "loss": 0.0,
       "step": 617
     },
     {
       "epoch": 0.022112099039304434,
-      "grad_norm": 0.0011405334807932377,
-      "learning_rate": 6.406546039941094e-05,
+      "grad_norm": 0.0009265893022529781,
+      "learning_rate": 6.377246332954544e-05,
       "loss": 0.0001,
       "step": 618
     },
     {
       "epoch": 0.02214787913483729,
-      "grad_norm": 0.0005600455915555358,
-      "learning_rate": 6.377246332954544e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0006062160828150809,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 0.0001,
       "step": 619
     },
     {
       "epoch": 0.022183659230370145,
-      "grad_norm": 0.0030411367770284414,
-      "learning_rate": 6.347982381084123e-05,
+      "grad_norm": 0.0005328463157638907,
+      "learning_rate": 6.318754473153221e-05,
       "loss": 0.0001,
       "step": 620
     },
     {
       "epoch": 0.022219439325903,
-      "grad_norm": 0.0007103204261511564,
-      "learning_rate": 6.318754473153221e-05,
+      "grad_norm": 0.00047331457608379424,
+      "learning_rate": 6.289562897629492e-05,
       "loss": 0.0,
       "step": 621
     },
     {
       "epoch": 0.022255219421435855,
-      "grad_norm": 0.14851249754428864,
-      "learning_rate": 6.289562897629492e-05,
-      "loss": 0.0016,
+      "grad_norm": 0.08468073606491089,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.0005,
       "step": 622
     },
     {
       "epoch": 0.02229099951696871,
-      "grad_norm": 0.00035604022559709847,
-      "learning_rate": 6.260407942621998e-05,
+      "grad_norm": 0.000396040384657681,
+      "learning_rate": 6.231289895878375e-05,
       "loss": 0.0,
       "step": 623
     },
     {
       "epoch": 0.022326779612501565,
-      "grad_norm": 0.000956276897341013,
-      "learning_rate": 6.231289895878375e-05,
+      "grad_norm": 0.000619475613348186,
+      "learning_rate": 6.20220904478199e-05,
       "loss": 0.0001,
       "step": 624
     },
     {
       "epoch": 0.02236255970803442,
-      "grad_norm": 0.0019058961188420653,
-      "learning_rate": 6.20220904478199e-05,
+      "grad_norm": 0.006687821354717016,
+      "learning_rate": 6.173165676349103e-05,
       "loss": 0.0001,
       "step": 625
     },
     {
       "epoch": 0.022398339803567275,
-      "grad_norm": 0.0006445760955102742,
-      "learning_rate": 6.173165676349103e-05,
+      "grad_norm": 0.00044290735968388617,
+      "learning_rate": 6.144160077226036e-05,
       "loss": 0.0001,
       "step": 626
     },
     {
       "epoch": 0.02243411989910013,
-      "grad_norm": 0.0071877241134643555,
-      "learning_rate": 6.144160077226036e-05,
+      "grad_norm": 0.009390274994075298,
+      "learning_rate": 6.11519253368634e-05,
       "loss": 0.0001,
       "step": 627
     },
     {
       "epoch": 0.022469899994632985,
-      "grad_norm": 0.0021069920621812344,
-      "learning_rate": 6.11519253368634e-05,
+      "grad_norm": 0.0016717095859348774,
+      "learning_rate": 6.086263331627976e-05,
       "loss": 0.0001,
       "step": 628
     },
     {
       "epoch": 0.02250568009016584,
-      "grad_norm": 0.0006769267492927611,
-      "learning_rate": 6.086263331627976e-05,
+      "grad_norm": 0.00042551616206765175,
+      "learning_rate": 6.05737275657049e-05,
       "loss": 0.0,
       "step": 629
     },
     {
       "epoch": 0.022541460185698695,
-      "grad_norm": 2.390087366104126,
-      "learning_rate": 6.05737275657049e-05,
-      "loss": 0.0163,
+      "grad_norm": 1.9672138690948486,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 0.0221,
       "step": 630
     },
     {
       "epoch": 0.02257724028123155,
-      "grad_norm": 0.00025465633370913565,
-      "learning_rate": 6.0285210936521955e-05,
+      "grad_norm": 0.0004953712341375649,
+      "learning_rate": 5.999708627627354e-05,
       "loss": 0.0,
       "step": 631
     },
     {
       "epoch": 0.022613020376764405,
-      "grad_norm": 0.0006367940804921091,
-      "learning_rate": 5.999708627627354e-05,
+      "grad_norm": 0.0006604749942198396,
+      "learning_rate": 5.9709356428633746e-05,
       "loss": 0.0001,
       "step": 632
     },
     {
       "epoch": 0.02264880047229726,
-      "grad_norm": 0.000536702573299408,
-      "learning_rate": 5.9709356428633746e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0006702704122290015,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 0.0001,
       "step": 633
     },
     {
       "epoch": 0.022684580567830116,
-      "grad_norm": 0.03368332237005234,
-      "learning_rate": 5.9422024233380013e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0017094590002670884,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 0.0,
       "step": 634
     },
     {
       "epoch": 0.02272036066336297,
-      "grad_norm": 0.0007999651716090739,
-      "learning_rate": 5.913509252636511e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.000579804356675595,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 0.0,
       "step": 635
     },
     {
       "epoch": 0.022756140758895826,
-      "grad_norm": 0.00032151470077224076,
-      "learning_rate": 5.884856413948913e-05,
+      "grad_norm": 0.0003796029486693442,
+      "learning_rate": 5.856244190067159e-05,
       "loss": 0.0,
       "step": 636
     },
     {
       "epoch": 0.02279192085442868,
-      "grad_norm": 0.002231956459581852,
-      "learning_rate": 5.856244190067159e-05,
+      "grad_norm": 0.0005215677083469927,
+      "learning_rate": 5.82767286338235e-05,
       "loss": 0.0,
       "step": 637
     },
     {
       "epoch": 0.022827700949961536,
-      "grad_norm": 0.1039712056517601,
-      "learning_rate": 5.82767286338235e-05,
-      "loss": 0.0016,
+      "grad_norm": 0.041516367346048355,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 0.0006,
       "step": 638
     },
     {
       "epoch": 0.02286348104549439,
-      "grad_norm": 0.000668445834890008,
-      "learning_rate": 5.799142715881938e-05,
+      "grad_norm": 0.0007705031312070787,
+      "learning_rate": 5.770654029146969e-05,
       "loss": 0.0,
       "step": 639
     },
     {
       "epoch": 0.022899261141027246,
-      "grad_norm": 0.0016554845497012138,
-      "learning_rate": 5.770654029146969e-05,
+      "grad_norm": 0.0017117122188210487,
+      "learning_rate": 5.7422070843492734e-05,
       "loss": 0.0001,
       "step": 640
     },
     {
       "epoch": 0.0229350412365601,
-      "grad_norm": 0.0002977857948280871,
-      "learning_rate": 5.7422070843492734e-05,
+      "grad_norm": 0.0010502267396077514,
+      "learning_rate": 5.713802162248718e-05,
       "loss": 0.0,
       "step": 641
     },
     {
       "epoch": 0.022970821332092956,
-      "grad_norm": 0.0005066939629614353,
-      "learning_rate": 5.713802162248718e-05,
+      "grad_norm": 0.000529443146660924,
+      "learning_rate": 5.6854395431904094e-05,
       "loss": 0.0,
       "step": 642
     },
     {
       "epoch": 0.02300660142762581,
-      "grad_norm": 0.0005043231067247689,
-      "learning_rate": 5.6854395431904094e-05,
+      "grad_norm": 0.0006939378217794001,
+      "learning_rate": 5.657119507101954e-05,
       "loss": 0.0,
       "step": 643
     },
     {
       "epoch": 0.023042381523158666,
-      "grad_norm": 0.0005800691433250904,
-      "learning_rate": 5.657119507101954e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00042171860695816576,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 0.0,
       "step": 644
     },
     {
       "epoch": 0.02307816161869152,
-      "grad_norm": 0.0003004905011039227,
-      "learning_rate": 5.6288423334906735e-05,
+      "grad_norm": 0.000523390423040837,
+      "learning_rate": 5.6006083014408484e-05,
       "loss": 0.0,
       "step": 645
     },
     {
       "epoch": 0.023113941714224377,
-      "grad_norm": 0.000490049016661942,
-      "learning_rate": 5.6006083014408484e-05,
+      "grad_norm": 0.000571369775570929,
+      "learning_rate": 5.572417689610987e-05,
       "loss": 0.0,
       "step": 646
     },
     {
       "epoch": 0.02314972180975723,
-      "grad_norm": 0.00039324729004874825,
-      "learning_rate": 5.572417689610987e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0006714938790537417,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 0.0001,
       "step": 647
     },
     {
       "epoch": 0.023185501905290087,
-      "grad_norm": 0.0005152429803274572,
-      "learning_rate": 5.544270776231038e-05,
+      "grad_norm": 0.0005017356015741825,
+      "learning_rate": 5.5161678390996796e-05,
       "loss": 0.0,
       "step": 648
     },
     {
       "epoch": 0.023221282000822942,
-      "grad_norm": 0.00030412263004109263,
-      "learning_rate": 5.5161678390996796e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0011646583443507552,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 0.0001,
       "step": 649
     },
     {
       "epoch": 0.023257062096355797,
-      "grad_norm": 0.000380836077965796,
-      "learning_rate": 5.488109155581549e-05,
+      "grad_norm": 0.00028895260766148567,
+      "learning_rate": 5.4600950026045326e-05,
       "loss": 0.0,
       "step": 650
     },
     {
       "epoch": 0.023292842191888652,
-      "grad_norm": 0.003229961497709155,
-      "learning_rate": 5.4600950026045326e-05,
+      "grad_norm": 0.0008328702533617616,
+      "learning_rate": 5.4321256566570036e-05,
       "loss": 0.0001,
       "step": 651
     },
     {
       "epoch": 0.023328622287421507,
-      "grad_norm": 0.00026713154511526227,
-      "learning_rate": 5.4321256566570036e-05,
+      "grad_norm": 0.00439800601452589,
+      "learning_rate": 5.404201393785122e-05,
       "loss": 0.0,
       "step": 652
     },
     {
       "epoch": 0.023364402382954362,
-      "grad_norm": 0.0008247068617492914,
-      "learning_rate": 5.404201393785122e-05,
-      "loss": 0.0,
+      "grad_norm": 0.002686363412067294,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 0.0001,
       "step": 653
     },
     {
       "epoch": 0.023400182478487217,
-      "grad_norm": 0.006118099670857191,
-      "learning_rate": 5.3763224895900846e-05,
+      "grad_norm": 0.0007387395016849041,
+      "learning_rate": 5.348489219225416e-05,
       "loss": 0.0001,
       "step": 654
     },
     {
       "epoch": 0.023435962574020072,
-      "grad_norm": 0.00907434243708849,
-      "learning_rate": 5.348489219225416e-05,
+      "grad_norm": 0.006214470602571964,
+      "learning_rate": 5.320701857394268e-05,
       "loss": 0.0002,
       "step": 655
     },
     {
       "epoch": 0.023471742669552927,
-      "grad_norm": 0.00029860183713026345,
-      "learning_rate": 5.320701857394268e-05,
+      "grad_norm": 0.000319985905662179,
+      "learning_rate": 5.292960678346675e-05,
       "loss": 0.0,
       "step": 656
     },
     {
       "epoch": 0.023507522765085782,
-      "grad_norm": 0.002056960016489029,
-      "learning_rate": 5.292960678346675e-05,
+      "grad_norm": 0.0005882193217985332,
+      "learning_rate": 5.265265955876879e-05,
       "loss": 0.0001,
       "step": 657
     },
     {
       "epoch": 0.023543302860618637,
-      "grad_norm": 0.0003817214455921203,
-      "learning_rate": 5.265265955876879e-05,
+      "grad_norm": 0.0004243079456500709,
+      "learning_rate": 5.237617963320608e-05,
       "loss": 0.0,
       "step": 658
     },
     {
       "epoch": 0.023579082956151493,
-      "grad_norm": 0.00038575747748836875,
-      "learning_rate": 5.237617963320608e-05,
+      "grad_norm": 0.0003599765768740326,
+      "learning_rate": 5.210016973552391e-05,
       "loss": 0.0,
       "step": 659
     },
     {
       "epoch": 0.023614863051684348,
-      "grad_norm": 0.00042975644464604557,
-      "learning_rate": 5.210016973552391e-05,
+      "grad_norm": 0.0007886773091740906,
+      "learning_rate": 5.182463258982846e-05,
       "loss": 0.0,
       "step": 660
     },
     {
       "epoch": 0.023650643147217203,
-      "grad_norm": 0.00037854621768929064,
-      "learning_rate": 5.182463258982846e-05,
+      "grad_norm": 0.00048393604811280966,
+      "learning_rate": 5.1549570915560206e-05,
       "loss": 0.0,
       "step": 661
     },
     {
       "epoch": 0.023686423242750058,
-      "grad_norm": 0.0002989709610119462,
-      "learning_rate": 5.1549570915560206e-05,
-      "loss": 0.0,
+      "grad_norm": 0.248721644282341,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.0029,
       "step": 662
     },
     {
       "epoch": 0.023722203338282913,
-      "grad_norm": 0.0021638735197484493,
-      "learning_rate": 5.127498742746675e-05,
+      "grad_norm": 0.001305613899603486,
+      "learning_rate": 5.100088483557634e-05,
       "loss": 0.0001,
       "step": 663
     },
     {
       "epoch": 0.023757983433815768,
-      "grad_norm": 0.0031420416198670864,
-      "learning_rate": 5.100088483557634e-05,
+      "grad_norm": 0.0008696125587448478,
+      "learning_rate": 5.072726584517086e-05,
       "loss": 0.0001,
       "step": 664
     },
     {
       "epoch": 0.023793763529348623,
-      "grad_norm": 0.006429425440728664,
-      "learning_rate": 5.072726584517086e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.0011506453156471252,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 0.0001,
       "step": 665
     },
     {
       "epoch": 0.023829543624881478,
-      "grad_norm": 0.0007081036455929279,
-      "learning_rate": 5.045413315675924e-05,
+      "grad_norm": 0.0009921109303832054,
+      "learning_rate": 5.018148946605092e-05,
       "loss": 0.0001,
       "step": 666
     },
     {
       "epoch": 0.023865323720414333,
-      "grad_norm": 0.0005449209129437804,
-      "learning_rate": 5.018148946605092e-05,
+      "grad_norm": 0.0002917166566476226,
+      "learning_rate": 4.990933746392899e-05,
       "loss": 0.0,
       "step": 667
     },
     {
       "epoch": 0.023901103815947188,
-      "grad_norm": 0.06656917184591293,
-      "learning_rate": 4.990933746392899e-05,
-      "loss": 0.0006,
+      "grad_norm": 0.0019778795540332794,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 0.0,
       "step": 668
     },
     {
       "epoch": 0.023936883911480043,
-      "grad_norm": 0.0011547500034794211,
-      "learning_rate": 4.9637679836423924e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.16058020293712616,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 0.0007,
       "step": 669
     },
     {
       "epoch": 0.0239726640070129,
-      "grad_norm": 0.004564153496176004,
-      "learning_rate": 4.9366519264686725e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.000699148396961391,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 0.0,
       "step": 670
     },
     {
       "epoch": 0.024008444102545753,
-      "grad_norm": 0.0009544348577037454,
-      "learning_rate": 4.909585842496287e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.000457806047052145,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 0.0,
       "step": 671
     },
     {
       "epoch": 0.02404422419807861,
-      "grad_norm": 0.0004808309022337198,
-      "learning_rate": 4.8825699988565485e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0014318109024316072,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.0001,
       "step": 672
     },
     {
       "epoch": 0.024080004293611464,
-      "grad_norm": 0.0025187938008457422,
-      "learning_rate": 4.8556046621849346e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0009062907774932683,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 0.0,
       "step": 673
     },
     {
       "epoch": 0.02411578438914432,
-      "grad_norm": 0.0007290815701708198,
-      "learning_rate": 4.828690098618429e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0007847001543268561,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 0.0001,
       "step": 674
     },
     {
       "epoch": 0.024151564484677174,
-      "grad_norm": 0.0007663268479518592,
-      "learning_rate": 4.8018265737929044e-05,
+      "grad_norm": 0.0003682690439745784,
+      "learning_rate": 4.7750143528405126e-05,
       "loss": 0.0,
       "step": 675
     },
     {
       "epoch": 0.02418734458021003,
-      "grad_norm": 0.00024741943343542516,
-      "learning_rate": 4.7750143528405126e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0014180666767060757,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 0.0001,
       "step": 676
     },
     {
       "epoch": 0.024223124675742884,
-      "grad_norm": 0.0006556051666848361,
-      "learning_rate": 4.748253700387042e-05,
+      "grad_norm": 0.0004101350496057421,
+      "learning_rate": 4.721544880549337e-05,
       "loss": 0.0001,
       "step": 677
     },
     {
       "epoch": 0.02425890477127574,
-      "grad_norm": 0.0004872358404099941,
-      "learning_rate": 4.721544880549337e-05,
-      "loss": 0.0,
+      "grad_norm": 0.04580630362033844,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 0.0002,
       "step": 678
     },
     {
       "epoch": 0.024294684866808594,
-      "grad_norm": 0.014162393286824226,
-      "learning_rate": 4.694888156932658e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.0020803629886358976,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 0.0001,
       "step": 679
     },
     {
       "epoch": 0.02433046496234145,
-      "grad_norm": 0.0005920436233282089,
-      "learning_rate": 4.668283792628114e-05,
+      "grad_norm": 0.0007991611491888762,
+      "learning_rate": 4.6417320502100316e-05,
       "loss": 0.0001,
       "step": 680
     },
     {
       "epoch": 0.024366245057874304,
-      "grad_norm": 0.0018332705367356539,
-      "learning_rate": 4.6417320502100316e-05,
+      "grad_norm": 0.0005774202290922403,
+      "learning_rate": 4.615233191733398e-05,
       "loss": 0.0001,
       "step": 681
     },
     {
       "epoch": 0.02440202515340716,
-      "grad_norm": 0.0003595963935367763,
-      "learning_rate": 4.615233191733398e-05,
+      "grad_norm": 0.00040934517164714634,
+      "learning_rate": 4.588787478731242e-05,
       "loss": 0.0,
       "step": 682
     },
     {
       "epoch": 0.024437805248940014,
-      "grad_norm": 0.005255323369055986,
-      "learning_rate": 4.588787478731242e-05,
+      "grad_norm": 0.004855821840465069,
+      "learning_rate": 4.5623951722120736e-05,
       "loss": 0.0001,
       "step": 683
     },
     {
       "epoch": 0.02447358534447287,
-      "grad_norm": 0.00842368509620428,
-      "learning_rate": 4.5623951722120736e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.022048255428671837,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 0.0005,
       "step": 684
     },
     {
       "epoch": 0.024509365440005725,
-      "grad_norm": 0.0006906003109179437,
-      "learning_rate": 4.5360565326573104e-05,
+      "grad_norm": 0.0009525329223833978,
+      "learning_rate": 4.5097718200186814e-05,
       "loss": 0.0001,
       "step": 685
     },
     {
       "epoch": 0.02454514553553858,
-      "grad_norm": 0.0013611942995339632,
-      "learning_rate": 4.5097718200186814e-05,
+      "grad_norm": 0.015187989920377731,
+      "learning_rate": 4.483541293715698e-05,
       "loss": 0.0001,
       "step": 686
     },
     {
       "epoch": 0.024580925631071435,
-      "grad_norm": 0.0024976772256195545,
-      "learning_rate": 4.483541293715698e-05,
+      "grad_norm": 0.0002924858999904245,
+      "learning_rate": 4.457365212633058e-05,
       "loss": 0.0,
       "step": 687
     },
     {
       "epoch": 0.02461670572660429,
-      "grad_norm": 0.0005024508573114872,
-      "learning_rate": 4.457365212633058e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0030987297650426626,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 0.0001,
       "step": 688
     },
     {
       "epoch": 0.024652485822137145,
-      "grad_norm": 0.0006341970292851329,
-      "learning_rate": 4.431243835118124e-05,
-      "loss": 0.0,
+      "grad_norm": 0.07195527851581573,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 0.0004,
       "step": 689
     },
     {
       "epoch": 0.02468826591767,
-      "grad_norm": 0.0003372261708136648,
-      "learning_rate": 4.4051774189783315e-05,
+      "grad_norm": 0.00034498111926950514,
+      "learning_rate": 4.379166221478697e-05,
       "loss": 0.0,
       "step": 690
     },
     {
       "epoch": 0.024724046013202855,
-      "grad_norm": 0.00036810620804317296,
-      "learning_rate": 4.379166221478697e-05,
+      "grad_norm": 0.0004273645463399589,
+      "learning_rate": 4.3532104993392306e-05,
       "loss": 0.0,
       "step": 691
     },
     {
       "epoch": 0.02475982610873571,
-      "grad_norm": 0.0015038405545055866,
-      "learning_rate": 4.3532104993392306e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0006148627726361156,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.0,
       "step": 692
     },
     {
       "epoch": 0.024795606204268565,
-      "grad_norm": 0.0019158747745677829,
-      "learning_rate": 4.327310508732437e-05,
+      "grad_norm": 0.0018783046398311853,
+      "learning_rate": 4.301466505280762e-05,
       "loss": 0.0001,
       "step": 693
     },
     {
       "epoch": 0.02483138629980142,
-      "grad_norm": 0.0023140893317759037,
-      "learning_rate": 4.301466505280762e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0005436090286821127,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 0.0,
       "step": 694
     },
     {
       "epoch": 0.024867166395334275,
-      "grad_norm": 0.000414346344769001,
-      "learning_rate": 4.2756787440540936e-05,
+      "grad_norm": 0.0003239592188037932,
+      "learning_rate": 4.249947479567218e-05,
       "loss": 0.0,
       "step": 695
     },
     {
       "epoch": 0.02490294649086713,
-      "grad_norm": 0.0003080443129874766,
-      "learning_rate": 4.249947479567218e-05,
+      "grad_norm": 0.00047223715228028595,
+      "learning_rate": 4.224272965777326e-05,
       "loss": 0.0,
       "step": 696
     },
     {
       "epoch": 0.024938726586399985,
-      "grad_norm": 0.003244619583711028,
-      "learning_rate": 4.224272965777326e-05,
+      "grad_norm": 0.005179517436772585,
+      "learning_rate": 4.1986554560815096e-05,
       "loss": 0.0001,
       "step": 697
     },
     {
       "epoch": 0.02497450668193284,
-      "grad_norm": 0.004701513797044754,
-      "learning_rate": 4.1986554560815096e-05,
+      "grad_norm": 0.001426901202648878,
+      "learning_rate": 4.173095203314241e-05,
       "loss": 0.0001,
       "step": 698
     },
     {
       "epoch": 0.025010286777465696,
-      "grad_norm": 0.0010928139090538025,
-      "learning_rate": 4.173095203314241e-05,
+      "grad_norm": 0.0039044171571731567,
+      "learning_rate": 4.1475924597449024e-05,
       "loss": 0.0001,
       "step": 699
     },
     {
       "epoch": 0.02504606687299855,
-      "grad_norm": 0.00047846059896983206,
-      "learning_rate": 4.1475924597449024e-05,
+      "grad_norm": 0.000497961591463536,
+      "learning_rate": 4.12214747707527e-05,
       "loss": 0.0,
       "step": 700
     },
     {
       "epoch": 0.025081846968531406,
-      "grad_norm": 0.00044242056901566684,
-      "learning_rate": 4.12214747707527e-05,
+      "grad_norm": 0.0007646341691724956,
+      "learning_rate": 4.096760506437057e-05,
       "loss": 0.0,
       "step": 701
     },
     {
       "epoch": 0.02511762706406426,
-      "grad_norm": 0.0006623343215323985,
-      "learning_rate": 4.096760506437057e-05,
+      "grad_norm": 0.0016085326205939054,
+      "learning_rate": 4.071431798389408e-05,
       "loss": 0.0,
       "step": 702
     },
     {
       "epoch": 0.025153407159597116,
-      "grad_norm": 0.00047430131235159934,
-      "learning_rate": 4.071431798389408e-05,
+      "grad_norm": 0.0005108435289002955,
+      "learning_rate": 4.0461616029164526e-05,
       "loss": 0.0,
       "step": 703
     },
     {
       "epoch": 0.02518918725512997,
-      "grad_norm": 0.0005569083732552826,
-      "learning_rate": 4.0461616029164526e-05,
+      "grad_norm": 0.0007308077183552086,
+      "learning_rate": 4.020950169424815e-05,
       "loss": 0.0,
       "step": 704
     },
     {
       "epoch": 0.025224967350662826,
-      "grad_norm": 0.05185364931821823,
-      "learning_rate": 4.020950169424815e-05,
-      "loss": 0.0005,
+      "grad_norm": 0.017377054318785667,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 0.0001,
       "step": 705
     },
     {
       "epoch": 0.02526074744619568,
-      "grad_norm": 0.0001887798134703189,
-      "learning_rate": 3.9957977467411615e-05,
+      "grad_norm": 0.00024152210971806198,
+      "learning_rate": 3.9707045831097555e-05,
       "loss": 0.0,
       "step": 706
     },
     {
       "epoch": 0.025296527541728536,
-      "grad_norm": 0.04409027472138405,
-      "learning_rate": 3.9707045831097555e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.003006350016221404,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 0.0001,
       "step": 707
     },
     {
       "epoch": 0.02533230763726139,
-      "grad_norm": 0.0002629173977766186,
-      "learning_rate": 3.945670926189987e-05,
+      "grad_norm": 0.00035428308183327317,
+      "learning_rate": 3.920697023053949e-05,
       "loss": 0.0,
       "step": 708
     },
     {
       "epoch": 0.025368087732794246,
-      "grad_norm": 0.0021778519731014967,
-      "learning_rate": 3.920697023053949e-05,
+      "grad_norm": 0.002442484488710761,
+      "learning_rate": 3.895783120183976e-05,
       "loss": 0.0001,
       "step": 709
     },
     {
       "epoch": 0.0254038678283271,
-      "grad_norm": 0.00028650613967329264,
-      "learning_rate": 3.895783120183976e-05,
+      "grad_norm": 0.0002176981361117214,
+      "learning_rate": 3.8709294634702376e-05,
       "loss": 0.0,
       "step": 710
     },
     {
       "epoch": 0.025439647923859957,
-      "grad_norm": 0.0003651605802588165,
-      "learning_rate": 3.8709294634702376e-05,
+      "grad_norm": 0.000730315106920898,
+      "learning_rate": 3.846136298208285e-05,
       "loss": 0.0,
       "step": 711
     },
     {
       "epoch": 0.02547542801939281,
-      "grad_norm": 0.0047093345783650875,
-      "learning_rate": 3.846136298208285e-05,
+      "grad_norm": 0.0019350670045241714,
+      "learning_rate": 3.821403869096658e-05,
       "loss": 0.0001,
       "step": 712
     },
     {
       "epoch": 0.025511208114925667,
-      "grad_norm": 0.023773379623889923,
-      "learning_rate": 3.821403869096658e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.001800208236090839,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 0.0001,
       "step": 713
     },
     {
       "epoch": 0.025546988210458522,
-      "grad_norm": 0.0001875998277682811,
-      "learning_rate": 3.796732420234443e-05,
+      "grad_norm": 0.00022908074606675655,
+      "learning_rate": 3.7721221951188765e-05,
       "loss": 0.0,
       "step": 714
     },
     {
       "epoch": 0.025582768305991377,
-      "grad_norm": 0.001957617700099945,
-      "learning_rate": 3.7721221951188765e-05,
+      "grad_norm": 0.0016731393989175558,
+      "learning_rate": 3.747573436642951e-05,
       "loss": 0.0001,
       "step": 715
     },
     {
       "epoch": 0.025618548401524232,
-      "grad_norm": 0.0013016173616051674,
-      "learning_rate": 3.747573436642951e-05,
+      "grad_norm": 0.00038809413672424853,
+      "learning_rate": 3.7230863870929964e-05,
       "loss": 0.0,
       "step": 716
     },
     {
       "epoch": 0.025654328497057087,
-      "grad_norm": 0.0004720388096757233,
-      "learning_rate": 3.7230863870929964e-05,
-      "loss": 0.0,
+      "grad_norm": 0.003590758191421628,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 0.0001,
       "step": 717
     },
     {
       "epoch": 0.025690108592589942,
-      "grad_norm": 0.00025122895021922886,
-      "learning_rate": 3.698661288146311e-05,
+      "grad_norm": 0.0006934832199476659,
+      "learning_rate": 3.674298380868756e-05,
       "loss": 0.0,
       "step": 718
     },
     {
       "epoch": 0.025725888688122797,
-      "grad_norm": 0.00023847611737437546,
-      "learning_rate": 3.674298380868756e-05,
+      "grad_norm": 0.0002972781367134303,
+      "learning_rate": 3.649997905712396e-05,
       "loss": 0.0,
       "step": 719
     },
     {
       "epoch": 0.025761668783655652,
-      "grad_norm": 0.00513820955529809,
-      "learning_rate": 3.649997905712396e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.00044423117651604116,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.0,
       "step": 720
     },
     {
       "epoch": 0.025797448879188507,
-      "grad_norm": 0.0010151059832423925,
-      "learning_rate": 3.6257601025131026e-05,
+      "grad_norm": 0.0009410815546289086,
+      "learning_rate": 3.601585210488218e-05,
       "loss": 0.0,
       "step": 721
     },
     {
       "epoch": 0.025833228974721362,
-      "grad_norm": 0.0009786454029381275,
-      "learning_rate": 3.601585210488218e-05,
+      "grad_norm": 0.0018991850083693862,
+      "learning_rate": 3.577473468234156e-05,
       "loss": 0.0001,
       "step": 722
     },
     {
       "epoch": 0.025869009070254217,
-      "grad_norm": 0.0005224223714321852,
-      "learning_rate": 3.577473468234156e-05,
+      "grad_norm": 0.0007036121096462011,
+      "learning_rate": 3.553425113724088e-05,
       "loss": 0.0,
       "step": 723
     },
     {
       "epoch": 0.025904789165787073,
-      "grad_norm": 0.0004743629542645067,
-      "learning_rate": 3.553425113724088e-05,
-      "loss": 0.0,
+      "grad_norm": 0.036000337451696396,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 0.0001,
       "step": 724
     },
     {
       "epoch": 0.025940569261319928,
-      "grad_norm": 0.0011396905174478889,
-      "learning_rate": 3.52944038430556e-05,
+      "grad_norm": 0.0010548071004450321,
+      "learning_rate": 3.5055195166981645e-05,
       "loss": 0.0001,
       "step": 725
     },
     {
       "epoch": 0.025976349356852783,
-      "grad_norm": 0.0004826972435694188,
-      "learning_rate": 3.5055195166981645e-05,
+      "grad_norm": 0.0002591910306364298,
+      "learning_rate": 3.481662746991214e-05,
       "loss": 0.0,
       "step": 726
     },
     {
       "epoch": 0.026012129452385638,
-      "grad_norm": 0.000776326865889132,
-      "learning_rate": 3.481662746991214e-05,
-      "loss": 0.0,
+      "grad_norm": 0.014273213222622871,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 0.0002,
       "step": 727
     },
     {
       "epoch": 0.026047909547918493,
-      "grad_norm": 0.0005714445142075419,
-      "learning_rate": 3.4578703106413904e-05,
+      "grad_norm": 0.00023336062440648675,
+      "learning_rate": 3.4341424424704375e-05,
       "loss": 0.0,
       "step": 728
     },
     {
       "epoch": 0.026083689643451348,
-      "grad_norm": 0.0008399290381930768,
-      "learning_rate": 3.4341424424704375e-05,
-      "loss": 0.0,
+      "grad_norm": 0.028344599530100822,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 0.0002,
       "step": 729
     },
     {
       "epoch": 0.026119469738984203,
-      "grad_norm": 0.0006017305422574282,
-      "learning_rate": 3.4104793766628304e-05,
+      "grad_norm": 0.0005131230573169887,
+      "learning_rate": 3.386881346763483e-05,
       "loss": 0.0,
       "step": 730
     },
     {
       "epoch": 0.026155249834517058,
-      "grad_norm": 0.0006686564884148538,
-      "learning_rate": 3.386881346763483e-05,
+      "grad_norm": 0.0010172853944823146,
+      "learning_rate": 3.363348585675414e-05,
       "loss": 0.0001,
       "step": 731
     },
     {
       "epoch": 0.026191029930049913,
-      "grad_norm": 0.00019212601182516664,
-      "learning_rate": 3.363348585675414e-05,
+      "grad_norm": 0.0003787051828112453,
+      "learning_rate": 3.339881325657484e-05,
       "loss": 0.0,
       "step": 732
     },
     {
       "epoch": 0.026226810025582768,
-      "grad_norm": 0.0006051249802112579,
-      "learning_rate": 3.339881325657484e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0009722478571347892,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 0.0001,
       "step": 733
     },
     {
       "epoch": 0.026262590121115623,
-      "grad_norm": 0.00269461702555418,
-      "learning_rate": 3.316479798322072e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.025112910196185112,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 0.0004,
       "step": 734
     },
     {
       "epoch": 0.02629837021664848,
-      "grad_norm": 0.0006747948355041444,
-      "learning_rate": 3.2931442346328004e-05,
-      "loss": 0.0,
+      "grad_norm": 0.002964810933917761,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 0.0001,
       "step": 735
     },
     {
       "epoch": 0.026334150312181333,
-      "grad_norm": 0.0005372610175982118,
-      "learning_rate": 3.269874864902269e-05,
+      "grad_norm": 0.00032617044053040445,
+      "learning_rate": 3.246671918789755e-05,
       "loss": 0.0,
       "step": 736
     },
     {
       "epoch": 0.02636993040771419,
-      "grad_norm": 0.004006940871477127,
-      "learning_rate": 3.246671918789755e-05,
+      "grad_norm": 0.005642542615532875,
+      "learning_rate": 3.223535625298979e-05,
       "loss": 0.0001,
       "step": 737
     },
     {
       "epoch": 0.026405710503247044,
-      "grad_norm": 0.0003320134710520506,
-      "learning_rate": 3.223535625298979e-05,
+      "grad_norm": 0.0002995952090714127,
+      "learning_rate": 3.200466212775808e-05,
       "loss": 0.0,
       "step": 738
     },
     {
       "epoch": 0.0264414905987799,
-      "grad_norm": 0.00025272424682043493,
-      "learning_rate": 3.200466212775808e-05,
+      "grad_norm": 0.00034134986344724894,
+      "learning_rate": 3.1774639089060363e-05,
       "loss": 0.0,
       "step": 739
     },
     {
       "epoch": 0.026477270694312754,
-      "grad_norm": 0.0029009163845330477,
-      "learning_rate": 3.1774639089060363e-05,
-      "loss": 0.0002,
+      "grad_norm": 0.0021360015962272882,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 0.0001,
       "step": 740
     },
     {
       "epoch": 0.02651305078984561,
-      "grad_norm": 0.0003290913300588727,
-      "learning_rate": 3.154528940713113e-05,
+      "grad_norm": 0.0003997268504463136,
+      "learning_rate": 3.1316615345559185e-05,
       "loss": 0.0,
       "step": 741
     },
     {
       "epoch": 0.026548830885378464,
-      "grad_norm": 0.016316860914230347,
-      "learning_rate": 3.1316615345559185e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.7127798795700073,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 0.0027,
       "step": 742
     },
     {
       "epoch": 0.02658461098091132,
-      "grad_norm": 0.0006960472092032433,
-      "learning_rate": 3.108861916126518e-05,
+      "grad_norm": 0.0017480599926784635,
+      "learning_rate": 3.086130310447937e-05,
       "loss": 0.0001,
       "step": 743
     },
     {
       "epoch": 0.026620391076444174,
-      "grad_norm": 0.0006541897892020643,
-      "learning_rate": 3.086130310447937e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0009076651185750961,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 0.0001,
       "step": 744
     },
     {
       "epoch": 0.02665617117197703,
-      "grad_norm": 0.00304150627925992,
-      "learning_rate": 3.063466941871952e-05,
+      "grad_norm": 0.002970295725390315,
+      "learning_rate": 3.0408720340768572e-05,
       "loss": 0.0001,
       "step": 745
     },
     {
       "epoch": 0.026691951267509884,
-      "grad_norm": 0.030295995995402336,
-      "learning_rate": 3.0408720340768572e-05,
-      "loss": 0.0003,
+      "grad_norm": 0.00616664020344615,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 0.0001,
       "step": 746
     },
     {
       "epoch": 0.02672773136304274,
-      "grad_norm": 0.002916352590546012,
-      "learning_rate": 3.018345810065275e-05,
+      "grad_norm": 0.0025385096669197083,
+      "learning_rate": 2.9958884921619367e-05,
       "loss": 0.0001,
       "step": 747
     },
     {
       "epoch": 0.026763511458575594,
-      "grad_norm": 0.0026307753287255764,
-      "learning_rate": 2.9958884921619367e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0010526994010433555,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 0.0,
       "step": 748
     },
     {
       "epoch": 0.02679929155410845,
-      "grad_norm": 0.00039545123581774533,
-      "learning_rate": 2.9735003020115092e-05,
+      "grad_norm": 0.0003839809214696288,
+      "learning_rate": 2.9511814605763855e-05,
       "loss": 0.0,
       "step": 749
     },
     {
       "epoch": 0.026835071649641305,
-      "grad_norm": 0.001498616300523281,
-      "learning_rate": 2.9511814605763855e-05,
+      "grad_norm": 0.001031058025546372,
+      "learning_rate": 2.9289321881345254e-05,
       "loss": 0.0001,
       "step": 750
     },
     {
       "epoch": 0.02687085174517416,
-      "grad_norm": 0.0004944049869664013,
-      "learning_rate": 2.9289321881345254e-05,
+      "grad_norm": 0.000416404043789953,
+      "learning_rate": 2.9067527042772636e-05,
       "loss": 0.0,
       "step": 751
     },
     {
       "epoch": 0.026906631840707015,
-      "grad_norm": 0.0003450750664342195,
-      "learning_rate": 2.9067527042772636e-05,
+      "grad_norm": 0.0006090101669542491,
+      "learning_rate": 2.8846432279071467e-05,
       "loss": 0.0,
       "step": 752
     },
     {
       "epoch": 0.02694241193623987,
-      "grad_norm": 0.0005600374424830079,
-      "learning_rate": 2.8846432279071467e-05,
+      "grad_norm": 0.0008759150514379144,
+      "learning_rate": 2.8626039772357882e-05,
       "loss": 0.0,
       "step": 753
     },
     {
       "epoch": 0.026978192031772725,
-      "grad_norm": 0.0005709484685212374,
-      "learning_rate": 2.8626039772357882e-05,
+      "grad_norm": 0.006808442063629627,
+      "learning_rate": 2.840635169781688e-05,
       "loss": 0.0001,
       "step": 754
     },
     {
       "epoch": 0.02701397212730558,
-      "grad_norm": 0.00028206201386637986,
-      "learning_rate": 2.840635169781688e-05,
+      "grad_norm": 0.0003685827541630715,
+      "learning_rate": 2.8187370223681132e-05,
       "loss": 0.0,
       "step": 755
     },
     {
       "epoch": 0.027049752222838435,
-      "grad_norm": 0.0008898506057448685,
-      "learning_rate": 2.8187370223681132e-05,
-      "loss": 0.0,
+      "grad_norm": 0.001124305883422494,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 0.0001,
       "step": 756
     },
     {
       "epoch": 0.02708553231837129,
-      "grad_norm": 0.0008389144786633551,
-      "learning_rate": 2.7969097511209308e-05,
+      "grad_norm": 0.00040056207217276096,
+      "learning_rate": 2.775153571466502e-05,
       "loss": 0.0,
       "step": 757
     },
     {
       "epoch": 0.027121312413904145,
-      "grad_norm": 0.010121572762727737,
-      "learning_rate": 2.775153571466502e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0006837467662990093,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 0.0,
       "step": 758
     },
     {
       "epoch": 0.027157092509437,
-      "grad_norm": 0.0006029063370078802,
-      "learning_rate": 2.753468698129533e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0013138077920302749,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 0.0001,
       "step": 759
     },
     {
       "epoch": 0.027192872604969855,
-      "grad_norm": 0.000917385274078697,
-      "learning_rate": 2.7318553451309726e-05,
+      "grad_norm": 0.00030627817614004016,
+      "learning_rate": 2.7103137257858868e-05,
       "loss": 0.0,
       "step": 760
     },
     {
       "epoch": 0.02722865270050271,
-      "grad_norm": 0.00039812002796679735,
-      "learning_rate": 2.7103137257858868e-05,
+      "grad_norm": 0.0007263069855980575,
+      "learning_rate": 2.688844052701359e-05,
       "loss": 0.0,
       "step": 761
     },
     {
       "epoch": 0.027264432796035565,
-      "grad_norm": 0.07431972771883011,
-      "learning_rate": 2.688844052701359e-05,
-      "loss": 0.0006,
+      "grad_norm": 0.0025389890652149916,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 0.0001,
       "step": 762
     },
     {
       "epoch": 0.02730021289156842,
-      "grad_norm": 0.0006493106484413147,
-      "learning_rate": 2.6674465377744017e-05,
+      "grad_norm": 0.0011348793050274253,
+      "learning_rate": 2.646121392189841e-05,
       "loss": 0.0,
       "step": 763
     },
     {
       "epoch": 0.027335992987101276,
-      "grad_norm": 0.00030248210532590747,
-      "learning_rate": 2.646121392189841e-05,
+      "grad_norm": 0.0003449412470217794,
+      "learning_rate": 2.624868826418262e-05,
       "loss": 0.0,
       "step": 764
     },
     {
       "epoch": 0.02737177308263413,
-      "grad_norm": 0.00029110562172718346,
-      "learning_rate": 2.624868826418262e-05,
+      "grad_norm": 0.0005563509184867144,
+      "learning_rate": 2.603689050213902e-05,
       "loss": 0.0,
       "step": 765
     },
     {
       "epoch": 0.027407553178166986,
-      "grad_norm": 0.0010446965461596847,
-      "learning_rate": 2.603689050213902e-05,
+      "grad_norm": 0.0008122368017211556,
+      "learning_rate": 2.582582272612609e-05,
       "loss": 0.0001,
       "step": 766
     },
     {
       "epoch": 0.02744333327369984,
-      "grad_norm": 0.003952876664698124,
-      "learning_rate": 2.582582272612609e-05,
+      "grad_norm": 0.000786035496275872,
+      "learning_rate": 2.561548701929749e-05,
       "loss": 0.0001,
       "step": 767
     },
     {
       "epoch": 0.027479113369232696,
-      "grad_norm": 0.00027852607308886945,
-      "learning_rate": 2.561548701929749e-05,
-      "loss": 0.0,
+      "grad_norm": 0.02800973877310753,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 0.0002,
       "step": 768
     },
     {
       "epoch": 0.02751489346476555,
-      "grad_norm": 0.0004770857922267169,
-      "learning_rate": 2.540588545758179e-05,
+      "grad_norm": 0.0003413466038182378,
+      "learning_rate": 2.5197020109661772e-05,
       "loss": 0.0,
       "step": 769
     },
     {
       "epoch": 0.027550673560298406,
-      "grad_norm": 0.00418006582185626,
-      "learning_rate": 2.5197020109661772e-05,
+      "grad_norm": 0.030475104227662086,
+      "learning_rate": 2.4988893036954043e-05,
       "loss": 0.0001,
       "step": 770
     },
     {
       "epoch": 0.02758645365583126,
-      "grad_norm": 0.0006785297300666571,
-      "learning_rate": 2.4988893036954043e-05,
-      "loss": 0.0,
+      "grad_norm": 0.000724347191862762,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 0.0001,
       "step": 771
     },
     {
       "epoch": 0.027622233751364116,
-      "grad_norm": 0.0006333782221190631,
-      "learning_rate": 2.4781506293588873e-05,
+      "grad_norm": 0.00045491516357287765,
+      "learning_rate": 2.4574861926389615e-05,
       "loss": 0.0,
       "step": 772
     },
     {
       "epoch": 0.02765801384689697,
-      "grad_norm": 0.0006415366660803556,
-      "learning_rate": 2.4574861926389615e-05,
-      "loss": 0.0,
+      "grad_norm": 0.0005608052015304565,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 0.0001,
       "step": 773
     },
     {
       "epoch": 0.027693793942429826,
-      "grad_norm": 0.0005563084268942475,
-      "learning_rate": 2.436896197485282e-05,
+      "grad_norm": 0.0009342865669168532,
+      "learning_rate": 2.4163808471127812e-05,
       "loss": 0.0,
       "step": 774
     },
     {
       "epoch": 0.02772957403796268,
-      "grad_norm": 0.0009341281256638467,
-      "learning_rate": 2.4163808471127812e-05,
+      "grad_norm": 0.0005581339355558157,
+      "learning_rate": 2.3959403439996907e-05,
       "loss": 0.0,
       "step": 775
     },
     {
       "epoch": 0.027765354133495537,
-      "grad_norm": 0.0005788183188997209,
-      "learning_rate": 2.3959403439996907e-05,
-      "loss": 0.0,
+      "grad_norm": 0.002823425456881523,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 0.0001,
       "step": 776
     },
     {
       "epoch": 0.02780113422902839,
-      "grad_norm": 0.0016393050318583846,
-      "learning_rate": 2.37557488988552e-05,
+      "grad_norm": 0.0017435902263969183,
+      "learning_rate": 2.3552846857690846e-05,
       "loss": 0.0001,
       "step": 777
     },
     {
       "epoch": 0.027836914324561247,
-      "grad_norm": 0.0009015945252031088,
-      "learning_rate": 2.3552846857690846e-05,
-      "loss": 0.0001,
+      "grad_norm": 1.2892370223999023,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 0.0044,
       "step": 778
     },
     {
       "epoch": 0.027872694420094102,
-      "grad_norm": 0.000347122986568138,
-      "learning_rate": 2.3350699319065026e-05,
+      "grad_norm": 0.00036629923852160573,
+      "learning_rate": 2.3149308278092342e-05,
       "loss": 0.0,
       "step": 779
     },
     {
       "epoch": 0.027908474515626957,
-      "grad_norm": 0.00029451295267790556,
-      "learning_rate": 2.3149308278092342e-05,
+      "grad_norm": 0.0004008855321444571,
+      "learning_rate": 2.2948675722421086e-05,
       "loss": 0.0,
       "step": 780
     },
     {
       "epoch": 0.027944254611159812,
-      "grad_norm": 0.0005590521031990647,
-      "learning_rate": 2.2948675722421086e-05,
+      "grad_norm": 0.0011866980930790305,
+      "learning_rate": 2.2748803632213557e-05,
       "loss": 0.0,
       "step": 781
     },
     {
       "epoch": 0.027980034706692667,
-      "grad_norm": 0.0022998007480055094,
-      "learning_rate": 2.2748803632213557e-05,
+      "grad_norm": 0.0009693397441878915,
+      "learning_rate": 2.254969398012663e-05,
       "loss": 0.0001,
       "step": 782
     },
     {
       "epoch": 0.028015814802225522,
-      "grad_norm": 0.00033281047944910824,
-      "learning_rate": 2.254969398012663e-05,
+      "grad_norm": 0.0003140924090985209,
+      "learning_rate": 2.235134873129213e-05,
       "loss": 0.0,
       "step": 783
     },
     {
       "epoch": 0.028051594897758377,
-      "grad_norm": 0.006088844034820795,
-      "learning_rate": 2.235134873129213e-05,
-      "loss": 0.0001,
+      "grad_norm": 0.0012890478828921914,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 0.0,
       "step": 784
     },
     {
       "epoch": 0.028087374993291232,
-      "grad_norm": 0.00030026666354388,
-      "learning_rate": 2.2153769843297667e-05,
+      "grad_norm": 0.00023359716578852385,
+      "learning_rate": 2.195695926616702e-05,
       "loss": 0.0,
       "step": 785
     },
     {
       "epoch": 0.028123155088824087,
-      "grad_norm": 0.001902211457490921,
+      "grad_norm": NaN,
       "learning_rate": 2.195695926616702e-05,
-      "loss": 0.0001,
+      "loss": 0.0227,
       "step": 786
     },
     {
       "epoch": 0.028158935184356942,
-      "grad_norm": 0.00032737571746110916,
+      "grad_norm": 0.0002614323457237333,
       "learning_rate": 2.1760918942341192e-05,
       "loss": 0.0,
       "step": 787
     },
     {
       "epoch": 0.028194715279889798,
-      "grad_norm": 0.0006129693356342614,
+      "grad_norm": 0.0004320174048189074,
       "learning_rate": 2.1565650806658975e-05,
       "loss": 0.0,
       "step": 788
     },
     {
       "epoch": 0.028230495375422653,
-      "grad_norm": 0.0002605107438284904,
+      "grad_norm": 0.0003752956399694085,
       "learning_rate": 2.137115678633811e-05,
       "loss": 0.0,
       "step": 789
     },
     {
       "epoch": 0.028266275470955508,
-      "grad_norm": 0.0008888985612429678,
+      "grad_norm": 0.000625955464784056,
       "learning_rate": 2.1177438800956007e-05,
       "loss": 0.0,
       "step": 790
     },
     {
       "epoch": 0.028302055566488363,
-      "grad_norm": 0.0029829475097358227,
+      "grad_norm": 0.0027366415597498417,
       "learning_rate": 2.098449876243096e-05,
       "loss": 0.0001,
       "step": 791
     },
     {
       "epoch": 0.028337835662021218,
-      "grad_norm": 0.0012633143924176693,
+      "grad_norm": 0.0010580029338598251,
       "learning_rate": 2.07923385750033e-05,
       "loss": 0.0,
       "step": 792
     },
     {
       "epoch": 0.028373615757554073,
-      "grad_norm": 0.00044179640826769173,
+      "grad_norm": 0.0003477209247648716,
       "learning_rate": 2.0600960135216462e-05,
       "loss": 0.0,
       "step": 793
     },
     {
       "epoch": 0.028409395853086928,
-      "grad_norm": 0.021729029715061188,
+      "grad_norm": 0.0003285256680101156,
       "learning_rate": 2.0410365331898416e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 794
     },
     {
       "epoch": 0.028445175948619783,
-      "grad_norm": 0.0018337086075916886,
+      "grad_norm": 0.0011474394705146551,
       "learning_rate": 2.0220556046142893e-05,
       "loss": 0.0001,
       "step": 795
     },
     {
       "epoch": 0.028480956044152638,
-      "grad_norm": 0.0005180193693377078,
+      "grad_norm": 0.0004369614471215755,
       "learning_rate": 2.0031534151290943e-05,
       "loss": 0.0,
       "step": 796
     },
     {
       "epoch": 0.028516736139685493,
-      "grad_norm": 0.0016789495712146163,
+      "grad_norm": 0.00162404531147331,
       "learning_rate": 1.9843301512912327e-05,
       "loss": 0.0001,
       "step": 797
     },
     {
       "epoch": 0.02855251623521835,
-      "grad_norm": 0.0008929551113396883,
+      "grad_norm": 0.0009096541325561702,
       "learning_rate": 1.965585998878724e-05,
       "loss": 0.0001,
       "step": 798
     },
     {
       "epoch": 0.028588296330751203,
-      "grad_norm": 0.07494878768920898,
+      "grad_norm": 0.006506095640361309,
       "learning_rate": 1.946921142888781e-05,
-      "loss": 0.0015,
+      "loss": 0.0003,
       "step": 799
     },
     {
       "epoch": 0.02862407642628406,
-      "grad_norm": 0.000517186475917697,
+      "grad_norm": 0.0012401292333379388,
       "learning_rate": 1.928335767535997e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 800
     },
     {
       "epoch": 0.028659856521816914,
-      "grad_norm": 0.000289576273644343,
+      "grad_norm": 0.0010858647292479873,
       "learning_rate": 1.9098300562505266e-05,
       "loss": 0.0,
       "step": 801
     },
     {
       "epoch": 0.02869563661734977,
-      "grad_norm": 0.00325192348100245,
+      "grad_norm": 0.0007080929935909808,
       "learning_rate": 1.891404191676265e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 802
     },
     {
       "epoch": 0.028731416712882624,
-      "grad_norm": 0.004330039955675602,
+      "grad_norm": 0.0020980716217309237,
       "learning_rate": 1.8730583556690605e-05,
       "loss": 0.0001,
       "step": 803
     },
     {
       "epoch": 0.02876719680841548,
-      "grad_norm": 0.002672772854566574,
+      "grad_norm": 0.003992127254605293,
       "learning_rate": 1.854792729294905e-05,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 804
     },
     {
       "epoch": 0.028802976903948334,
-      "grad_norm": 0.026112854480743408,
+      "grad_norm": 0.016032757237553596,
       "learning_rate": 1.8366074928281607e-05,
       "loss": 0.0004,
       "step": 805
     },
     {
       "epoch": 0.02883875699948119,
-      "grad_norm": 0.00033803124097175896,
+      "grad_norm": 0.013076834380626678,
       "learning_rate": 1.818502825749764e-05,
-      "loss": 0.0,
+      "loss": 0.0002,
       "step": 806
     },
     {
       "epoch": 0.028874537095014044,
-      "grad_norm": 0.00023536061053164303,
+      "grad_norm": 0.0002816877677105367,
       "learning_rate": 1.8004789067454764e-05,
       "loss": 0.0,
       "step": 807
     },
     {
       "epoch": 0.0289103171905469,
-      "grad_norm": 0.0005821982049383223,
+      "grad_norm": 0.002238611225038767,
       "learning_rate": 1.7825359137040988e-05,
       "loss": 0.0001,
       "step": 808
     },
     {
       "epoch": 0.028946097286079754,
-      "grad_norm": 0.01212245225906372,
+      "grad_norm": 0.000730925879906863,
       "learning_rate": 1.7646740237157256e-05,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 809
     },
     {
       "epoch": 0.02898187738161261,
-      "grad_norm": 0.0006459580617956817,
+      "grad_norm": 0.0032203211449086666,
       "learning_rate": 1.7468934130700044e-05,
       "loss": 0.0001,
       "step": 810
     },
     {
       "epoch": 0.029017657477145464,
-      "grad_norm": 0.001510439207777381,
+      "grad_norm": 0.0004394878342282027,
       "learning_rate": 1.7291942572543807e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 811
     },
     {
       "epoch": 0.02905343757267832,
-      "grad_norm": 0.00034279425744898617,
+      "grad_norm": 0.0012944753980264068,
       "learning_rate": 1.7115767309523812e-05,
       "loss": 0.0,
       "step": 812
     },
     {
       "epoch": 0.029089217668211174,
-      "grad_norm": 0.00032300263410434127,
+      "grad_norm": 0.0003389958874322474,
       "learning_rate": 1.6940410080418723e-05,
       "loss": 0.0,
       "step": 813
     },
     {
       "epoch": 0.02912499776374403,
-      "grad_norm": 0.00041772317490540445,
+      "grad_norm": 0.00032627416658215225,
       "learning_rate": 1.6765872615933677e-05,
       "loss": 0.0,
       "step": 814
     },
     {
       "epoch": 0.029160777859276885,
-      "grad_norm": 0.00318315951153636,
+      "grad_norm": 0.0004266609321348369,
       "learning_rate": 1.6592156638682886e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 815
     },
     {
       "epoch": 0.02919655795480974,
-      "grad_norm": 0.0006948874797672033,
+      "grad_norm": 0.0006152820424176753,
       "learning_rate": 1.6419263863172997e-05,
       "loss": 0.0001,
       "step": 816
     },
     {
       "epoch": 0.029232338050342595,
-      "grad_norm": 0.0005434126942418516,
+      "grad_norm": 0.0014175320975482464,
       "learning_rate": 1.6247195995785837e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 817
     },
     {
       "epoch": 0.02926811814587545,
-      "grad_norm": 0.0004638167447410524,
+      "grad_norm": 0.0005357126356102526,
       "learning_rate": 1.6075954734761845e-05,
       "loss": 0.0,
       "step": 818
     },
     {
       "epoch": 0.029303898241408305,
-      "grad_norm": 0.003182518994435668,
+      "grad_norm": 0.0012012666556984186,
       "learning_rate": 1.5905541770183096e-05,
       "loss": 0.0001,
       "step": 819
     },
     {
       "epoch": 0.02933967833694116,
-      "grad_norm": 0.0003212198498658836,
+      "grad_norm": 0.005087379831820726,
       "learning_rate": 1.5735958783956794e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 820
     },
     {
       "epoch": 0.029375458432474015,
-      "grad_norm": 0.0007989992736838758,
+      "grad_norm": 0.0005693581770174205,
       "learning_rate": 1.5567207449798515e-05,
       "loss": 0.0,
       "step": 821
     },
     {
       "epoch": 0.02941123852800687,
-      "grad_norm": 0.004399982281029224,
+      "grad_norm": 0.005406623240560293,
       "learning_rate": 1.539928943321579e-05,
       "loss": 0.0001,
       "step": 822
     },
     {
       "epoch": 0.029447018623539725,
-      "grad_norm": 0.001447922084480524,
+      "grad_norm": 0.0009374113869853318,
       "learning_rate": 1.5232206391491699e-05,
       "loss": 0.0,
       "step": 823
     },
     {
       "epoch": 0.02948279871907258,
-      "grad_norm": 0.0011744245421141386,
+      "grad_norm": 0.0003775600343942642,
       "learning_rate": 1.5065959973668353e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 824
     },
     {
       "epoch": 0.029518578814605435,
-      "grad_norm": 0.0003306670696474612,
+      "grad_norm": 0.0003483894106466323,
       "learning_rate": 1.4900551820530828e-05,
       "loss": 0.0,
       "step": 825
     },
     {
       "epoch": 0.02955435891013829,
-      "grad_norm": 0.0010293950326740742,
+      "grad_norm": 0.0003740845713764429,
       "learning_rate": 1.4735983564590783e-05,
       "loss": 0.0,
       "step": 826
     },
     {
       "epoch": 0.029590139005671146,
-      "grad_norm": 0.00033167112269438803,
+      "grad_norm": 0.0002811375306919217,
       "learning_rate": 1.4572256830070497e-05,
       "loss": 0.0,
       "step": 827
     },
     {
       "epoch": 0.029625919101204,
-      "grad_norm": 0.0009287113789469004,
+      "grad_norm": 0.0006832360522821546,
       "learning_rate": 1.4409373232886702e-05,
       "loss": 0.0001,
       "step": 828
     },
     {
       "epoch": 0.029661699196736856,
-      "grad_norm": 0.0002370228903600946,
+      "grad_norm": 0.000255306891631335,
       "learning_rate": 1.4247334380634792e-05,
       "loss": 0.0,
       "step": 829
     },
     {
       "epoch": 0.02969747929226971,
-      "grad_norm": 0.00026000026264227927,
+      "grad_norm": 0.00022817605349700898,
       "learning_rate": 1.4086141872572789e-05,
       "loss": 0.0,
       "step": 830
     },
     {
       "epoch": 0.029733259387802566,
-      "grad_norm": 0.0002268932876177132,
+      "grad_norm": 0.00029117820668034256,
       "learning_rate": 1.3925797299605647e-05,
       "loss": 0.0,
       "step": 831
     },
     {
       "epoch": 0.02976903948333542,
-      "grad_norm": 0.0004335989651735872,
+      "grad_norm": 0.000468272075522691,
       "learning_rate": 1.3766302244269624e-05,
       "loss": 0.0,
       "step": 832
     },
     {
       "epoch": 0.029804819578868276,
-      "grad_norm": 0.0037346568424254656,
+      "grad_norm": 0.007703302428126335,
       "learning_rate": 1.3607658280716473e-05,
-      "loss": 0.0001,
+      "loss": 0.0003,
       "step": 833
     },
     {
       "epoch": 0.02984059967440113,
-      "grad_norm": 0.0011007866123691201,
+      "grad_norm": 0.0002647882793098688,
       "learning_rate": 1.3449866974698122e-05,
       "loss": 0.0,
       "step": 834
     },
     {
       "epoch": 0.029876379769933986,
-      "grad_norm": 0.0003231363371014595,
+      "grad_norm": 0.000288227602140978,
       "learning_rate": 1.3292929883550998e-05,
       "loss": 0.0,
       "step": 835
     },
     {
       "epoch": 0.02991215986546684,
-      "grad_norm": 0.0006045972695574164,
+      "grad_norm": 0.0011180542642250657,
       "learning_rate": 1.3136848556180892e-05,
       "loss": 0.0,
       "step": 836
     },
     {
       "epoch": 0.029947939960999696,
-      "grad_norm": 0.000930872920434922,
+      "grad_norm": 0.0015647278632968664,
       "learning_rate": 1.2981624533047432e-05,
       "loss": 0.0001,
       "step": 837
     },
     {
       "epoch": 0.02998372005653255,
-      "grad_norm": 0.004588426556438208,
+      "grad_norm": 0.22069744765758514,
       "learning_rate": 1.2827259346149122e-05,
-      "loss": 0.0001,
+      "loss": 0.0027,
       "step": 838
     },
     {
       "epoch": 0.030019500152065406,
-      "grad_norm": 0.0004836541775148362,
+      "grad_norm": 0.001163410721346736,
       "learning_rate": 1.2673754519008008e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 839
     },
     {
       "epoch": 0.03005528024759826,
-      "grad_norm": 0.00027530387160368264,
+      "grad_norm": 0.00020182439766358584,
       "learning_rate": 1.2521111566654731e-05,
       "loss": 0.0,
       "step": 840
     },
     {
       "epoch": 0.030091060343131117,
-      "grad_norm": 0.001107738702557981,
+      "grad_norm": 0.0008434744086116552,
       "learning_rate": 1.2369331995613665e-05,
       "loss": 0.0001,
       "step": 841
     },
     {
       "epoch": 0.03012684043866397,
-      "grad_norm": 0.0003097807348240167,
+      "grad_norm": 0.0004498157650232315,
       "learning_rate": 1.2218417303887842e-05,
       "loss": 0.0,
       "step": 842
     },
     {
       "epoch": 0.030162620534196827,
-      "grad_norm": 0.000548563024494797,
+      "grad_norm": 0.00038383417995646596,
       "learning_rate": 1.206836898094439e-05,
       "loss": 0.0,
       "step": 843
     },
     {
       "epoch": 0.030198400629729682,
-      "grad_norm": 0.003870975226163864,
+      "grad_norm": 0.0030869089532643557,
       "learning_rate": 1.191918850769964e-05,
       "loss": 0.0001,
       "step": 844
     },
     {
       "epoch": 0.030234180725262537,
-      "grad_norm": 0.0009440371068194509,
+      "grad_norm": 0.0009010889334604144,
       "learning_rate": 1.1770877356504683e-05,
       "loss": 0.0001,
       "step": 845
     },
     {
       "epoch": 0.030269960820795392,
-      "grad_norm": 0.0007617096998728812,
+      "grad_norm": 0.0013517620973289013,
       "learning_rate": 1.1623436991130654e-05,
       "loss": 0.0,
       "step": 846
     },
     {
       "epoch": 0.030305740916328247,
-      "grad_norm": 0.0005047709564678371,
+      "grad_norm": 0.0005510963965207338,
       "learning_rate": 1.1476868866754486e-05,
       "loss": 0.0,
       "step": 847
     },
     {
       "epoch": 0.030341521011861102,
-      "grad_norm": 0.0021626728121191263,
+      "grad_norm": 0.0006830064230598509,
       "learning_rate": 1.1331174429944347e-05,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 848
     },
     {
       "epoch": 0.030377301107393957,
-      "grad_norm": 0.0005678251036442816,
+      "grad_norm": 0.0007867237436585128,
       "learning_rate": 1.1186355118645554e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 849
     },
     {
       "epoch": 0.030413081202926812,
-      "grad_norm": 0.0003562841739039868,
+      "grad_norm": 0.0006200165371410549,
       "learning_rate": 1.1042412362166222e-05,
       "loss": 0.0,
       "step": 850
     },
     {
       "epoch": 0.030448861298459667,
-      "grad_norm": 0.000597378530073911,
+      "grad_norm": 0.001512668328359723,
       "learning_rate": 1.0899347581163221e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 851
     },
     {
       "epoch": 0.030484641393992522,
-      "grad_norm": 0.0012710741721093655,
+      "grad_norm": 0.0007012730347923934,
       "learning_rate": 1.0757162187628222e-05,
       "loss": 0.0,
       "step": 852
     },
     {
       "epoch": 0.030520421489525378,
-      "grad_norm": 0.0005619203438982368,
+      "grad_norm": 0.0004401567275635898,
       "learning_rate": 1.0615857584873623e-05,
       "loss": 0.0001,
       "step": 853
     },
     {
       "epoch": 0.030556201585058233,
-      "grad_norm": 0.0002655864227563143,
+      "grad_norm": 0.00031436511198990047,
       "learning_rate": 1.0475435167518843e-05,
       "loss": 0.0,
       "step": 854
     },
     {
       "epoch": 0.030591981680591088,
-      "grad_norm": 0.00112536724191159,
+      "grad_norm": 0.001058329944498837,
       "learning_rate": 1.0335896321476413e-05,
       "loss": 0.0001,
       "step": 855
     },
     {
       "epoch": 0.030627761776123943,
-      "grad_norm": 0.0006863333401270211,
+      "grad_norm": 0.0007222515996545553,
       "learning_rate": 1.0197242423938446e-05,
       "loss": 0.0,
       "step": 856
     },
     {
       "epoch": 0.030663541871656798,
-      "grad_norm": 0.000631397299002856,
+      "grad_norm": 0.0014382116496562958,
       "learning_rate": 1.0059474843362892e-05,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 857
     },
     {
       "epoch": 0.030699321967189653,
-      "grad_norm": 0.0016255303053185344,
+      "grad_norm": 0.0014747374225407839,
       "learning_rate": 9.922594939460194e-06,
       "loss": 0.0001,
       "step": 858
     },
     {
       "epoch": 0.030735102062722508,
-      "grad_norm": 0.00026934832567349076,
+      "grad_norm": 0.00035679215216077864,
       "learning_rate": 9.786604063179728e-06,
       "loss": 0.0,
       "step": 859
     },
     {
       "epoch": 0.030770882158255363,
-      "grad_norm": 0.00026072084438055754,
+      "grad_norm": 0.00018847925821319222,
       "learning_rate": 9.651503556696516e-06,
       "loss": 0.0,
       "step": 860
     },
     {
       "epoch": 0.030806662253788218,
-      "grad_norm": 0.0004206086159683764,
+      "grad_norm": 0.00047721504233777523,
       "learning_rate": 9.517294753398064e-06,
       "loss": 0.0,
       "step": 861
     },
     {
       "epoch": 0.030842442349321073,
-      "grad_norm": 0.00024543897598050535,
+      "grad_norm": 0.00044450335553847253,
       "learning_rate": 9.383978977871021e-06,
       "loss": 0.0,
       "step": 862
     },
     {
       "epoch": 0.03087822244485393,
-      "grad_norm": 0.0003954702988266945,
+      "grad_norm": 0.000441976822912693,
       "learning_rate": 9.251557545888312e-06,
       "loss": 0.0,
       "step": 863
     },
     {
       "epoch": 0.030914002540386783,
-      "grad_norm": 0.0005821630475111306,
+      "grad_norm": 0.0003807090106420219,
       "learning_rate": 9.120031764395987e-06,
       "loss": 0.0,
       "step": 864
     },
     {
       "epoch": 0.03094978263591964,
-      "grad_norm": 0.00018100690795108676,
+      "grad_norm": 0.00022103596711531281,
       "learning_rate": 8.989402931500434e-06,
       "loss": 0.0,
       "step": 865
     },
     {
       "epoch": 0.030985562731452494,
-      "grad_norm": 0.00047442514915019274,
+      "grad_norm": 0.0011002712417393923,
       "learning_rate": 8.85967233645547e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 866
     },
     {
       "epoch": 0.03102134282698535,
-      "grad_norm": 0.0002546119212638587,
+      "grad_norm": 0.000323577260132879,
       "learning_rate": 8.730841259649725e-06,
       "loss": 0.0,
       "step": 867
     },
     {
       "epoch": 0.031057122922518204,
-      "grad_norm": 0.00017287160153500736,
+      "grad_norm": 0.0002585575566627085,
       "learning_rate": 8.602910972593892e-06,
       "loss": 0.0,
       "step": 868
     },
     {
       "epoch": 0.03109290301805106,
-      "grad_norm": 0.00031131732976064086,
+      "grad_norm": 0.0009426943142898381,
       "learning_rate": 8.475882737908248e-06,
       "loss": 0.0,
       "step": 869
     },
     {
       "epoch": 0.031128683113583914,
-      "grad_norm": 0.00036871631164103746,
+      "grad_norm": 0.0017182575538754463,
       "learning_rate": 8.34975780931021e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 870
     },
     {
       "epoch": 0.03116446320911677,
-      "grad_norm": 0.0036162729375064373,
+      "grad_norm": 0.0020622152369469404,
       "learning_rate": 8.224537431601886e-06,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 871
     },
     {
       "epoch": 0.031200243304649624,
-      "grad_norm": 0.0007622698904015124,
+      "grad_norm": 0.0012766410363838077,
       "learning_rate": 8.100222840657878e-06,
       "loss": 0.0001,
       "step": 872
     },
     {
       "epoch": 0.03123602340018248,
-      "grad_norm": 0.0003165509842801839,
+      "grad_norm": 0.0012319182278588414,
       "learning_rate": 7.976815263412963e-06,
       "loss": 0.0,
       "step": 873
     },
     {
       "epoch": 0.03127180349571533,
-      "grad_norm": 0.00027726407279260457,
+      "grad_norm": 0.05080391466617584,
       "learning_rate": 7.854315917850163e-06,
-      "loss": 0.0,
+      "loss": 0.0002,
       "step": 874
     },
     {
       "epoch": 0.03130758359124819,
-      "grad_norm": 0.001894214772619307,
+      "grad_norm": 0.0002540445129852742,
       "learning_rate": 7.73272601298851e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 875
     },
     {
       "epoch": 0.03134336368678104,
-      "grad_norm": 0.0031109314877539873,
+      "grad_norm": 0.0005007732543163002,
       "learning_rate": 7.612046748871327e-06,
       "loss": 0.0,
       "step": 876
     },
     {
       "epoch": 0.0313791437823139,
-      "grad_norm": 0.00024636011221446097,
+      "grad_norm": 0.00018185537192039192,
       "learning_rate": 7.492279316554207e-06,
       "loss": 0.0,
       "step": 877
     },
     {
       "epoch": 0.03141492387784675,
-      "grad_norm": 0.0014846808044239879,
+      "grad_norm": 0.0014064558781683445,
       "learning_rate": 7.3734248980933395e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 878
     },
     {
       "epoch": 0.03145070397337961,
-      "grad_norm": 0.000900342536624521,
+      "grad_norm": 0.002208331134170294,
       "learning_rate": 7.255484666533874e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 879
     },
     {
       "epoch": 0.03148648406891246,
-      "grad_norm": 0.0005137972184456885,
+      "grad_norm": 0.0004916066536679864,
       "learning_rate": 7.138459785898266e-06,
       "loss": 0.0,
       "step": 880
     },
     {
       "epoch": 0.03152226416444532,
-      "grad_norm": 0.00021104987536091357,
+      "grad_norm": 0.0004903101944364607,
       "learning_rate": 7.022351411174866e-06,
       "loss": 0.0,
       "step": 881
     },
     {
       "epoch": 0.03155804425997817,
-      "grad_norm": 0.0007743556052446365,
+      "grad_norm": 0.0016464211512356997,
       "learning_rate": 6.907160688306425e-06,
       "loss": 0.0001,
       "step": 882
     },
     {
       "epoch": 0.03159382435551103,
-      "grad_norm": 0.00026683160103857517,
+      "grad_norm": 0.00020444215624593198,
       "learning_rate": 6.7928887541789055e-06,
       "loss": 0.0,
       "step": 883
     },
     {
       "epoch": 0.03162960445104388,
-      "grad_norm": 0.00034998750197701156,
+      "grad_norm": 0.039082761853933334,
       "learning_rate": 6.679536736610137e-06,
-      "loss": 0.0,
+      "loss": 0.0003,
       "step": 884
     },
     {
       "epoch": 0.03166538454657674,
-      "grad_norm": 0.0005239667370915413,
+      "grad_norm": 0.0004566194547805935,
       "learning_rate": 6.5671057543387985e-06,
       "loss": 0.0,
       "step": 885
     },
     {
       "epoch": 0.03170116464210959,
-      "grad_norm": 0.00047340718447230756,
+      "grad_norm": 0.00025817123241722584,
       "learning_rate": 6.455596917013273e-06,
       "loss": 0.0,
       "step": 886
     },
     {
       "epoch": 0.03173694473764245,
-      "grad_norm": 0.034872330725193024,
+      "grad_norm": 0.0003827243053819984,
       "learning_rate": 6.345011325180772e-06,
-      "loss": 0.0002,
+      "loss": 0.0,
       "step": 887
     },
     {
       "epoch": 0.0317727248331753,
-      "grad_norm": 0.0004132964531891048,
+      "grad_norm": 0.0004853069840464741,
       "learning_rate": 6.235350070276447e-06,
       "loss": 0.0,
       "step": 888
     },
     {
       "epoch": 0.03180850492870816,
-      "grad_norm": 0.0032511220779269934,
+      "grad_norm": 0.00047525035915896297,
       "learning_rate": 6.126614234612593e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 889
     },
     {
       "epoch": 0.03184428502424101,
-      "grad_norm": 0.0008952480275183916,
+      "grad_norm": 0.0005343626835383475,
       "learning_rate": 6.018804891368035e-06,
       "loss": 0.0,
       "step": 890
     },
     {
       "epoch": 0.03188006511977387,
-      "grad_norm": 0.0003391039790585637,
+      "grad_norm": 0.0003221333317924291,
       "learning_rate": 5.911923104577455e-06,
       "loss": 0.0,
       "step": 891
     },
     {
       "epoch": 0.03191584521530672,
-      "grad_norm": 0.0003913507389370352,
+      "grad_norm": 0.0007317602867260575,
       "learning_rate": 5.805969929120947e-06,
       "loss": 0.0,
       "step": 892
     },
     {
       "epoch": 0.03195162531083958,
-      "grad_norm": 0.047737907618284225,
+      "grad_norm": 0.0011024659033864737,
       "learning_rate": 5.700946410713548e-06,
-      "loss": 0.0005,
+      "loss": 0.0001,
       "step": 893
     },
     {
       "epoch": 0.03198740540637243,
-      "grad_norm": 0.0005478477687574923,
+      "grad_norm": 0.0030077833216637373,
       "learning_rate": 5.5968535858950345e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 894
     },
     {
       "epoch": 0.03202318550190529,
-      "grad_norm": 0.0012483817990869284,
+      "grad_norm": 0.0056787352077662945,
       "learning_rate": 5.49369248201953e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 895
     },
     {
       "epoch": 0.03205896559743814,
-      "grad_norm": 0.001570535241626203,
+      "grad_norm": 0.0005671354592777789,
       "learning_rate": 5.39146411724547e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 896
     },
     {
       "epoch": 0.032094745692971,
-      "grad_norm": 0.0006657909834757447,
+      "grad_norm": 0.0006770672625862062,
       "learning_rate": 5.290169500525577e-06,
       "loss": 0.0,
       "step": 897
     },
     {
       "epoch": 0.03213052578850385,
-      "grad_norm": 0.0005471473559737206,
+      "grad_norm": 0.00038105715066194534,
       "learning_rate": 5.189809631596798e-06,
       "loss": 0.0,
       "step": 898
     },
     {
       "epoch": 0.03216630588403671,
-      "grad_norm": 0.0005784566747024655,
+      "grad_norm": 0.0031658518128097057,
       "learning_rate": 5.0903855009705514e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 899
     },
     {
       "epoch": 0.03220208597956956,
-      "grad_norm": 0.0036765148397535086,
+      "grad_norm": 0.0007510024588555098,
       "learning_rate": 4.991898089922819e-06,
       "loss": 0.0,
       "step": 900
     },
     {
       "epoch": 0.03223786607510242,
-      "grad_norm": 0.00033510179491713643,
+      "grad_norm": 0.00039749740972183645,
       "learning_rate": 4.8943483704846475e-06,
       "loss": 0.0,
       "step": 901
     },
     {
       "epoch": 0.03227364617063527,
-      "grad_norm": 0.0007261609425768256,
+      "grad_norm": 0.0016887275269255042,
       "learning_rate": 4.797737305432337e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 902
     },
     {
       "epoch": 0.03230942626616813,
-      "grad_norm": 0.0009459084249101579,
+      "grad_norm": 0.0017919878009706736,
       "learning_rate": 4.702065848278126e-06,
       "loss": 0.0001,
       "step": 903
     },
     {
       "epoch": 0.03234520636170098,
-      "grad_norm": 0.0004293612437322736,
+      "grad_norm": 0.0026703733019530773,
       "learning_rate": 4.607334943260655e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 904
     },
     {
       "epoch": 0.03238098645723384,
-      "grad_norm": 0.016933247447013855,
+      "grad_norm": 0.015657538548111916,
       "learning_rate": 4.513545525335705e-06,
       "loss": 0.0002,
       "step": 905
     },
     {
       "epoch": 0.03241676655276669,
-      "grad_norm": 0.00017119102994911373,
+      "grad_norm": 0.0002171381638618186,
       "learning_rate": 4.420698520166988e-06,
       "loss": 0.0,
       "step": 906
     },
     {
       "epoch": 0.03245254664829955,
-      "grad_norm": 0.00030199214234016836,
+      "grad_norm": 0.00031557457987219095,
       "learning_rate": 4.328794844116946e-06,
       "loss": 0.0,
       "step": 907
     },
     {
       "epoch": 0.0324883267438324,
-      "grad_norm": 0.0008139758137986064,
+      "grad_norm": 0.0007938373018987477,
       "learning_rate": 4.237835404237778e-06,
       "loss": 0.0,
       "step": 908
     },
     {
       "epoch": 0.03252410683936526,
-      "grad_norm": 0.00030492927180603147,
+      "grad_norm": 0.0008096988312900066,
       "learning_rate": 4.147821098262405e-06,
       "loss": 0.0,
       "step": 909
     },
     {
       "epoch": 0.032559886934898113,
-      "grad_norm": 0.00040922616608440876,
+      "grad_norm": 0.0005311848362907767,
       "learning_rate": 4.0587528145957235e-06,
       "loss": 0.0,
       "step": 910
     },
     {
       "epoch": 0.03259566703043097,
-      "grad_norm": 0.043064460158348083,
+      "grad_norm": 0.0013915644958615303,
       "learning_rate": 3.970631432305694e-06,
-      "loss": 0.0003,
+      "loss": 0.0,
       "step": 911
     },
     {
       "epoch": 0.032631447125963824,
-      "grad_norm": 0.000747754645999521,
+      "grad_norm": 0.0006781916017644107,
       "learning_rate": 3.883457821114811e-06,
       "loss": 0.0001,
       "step": 912
     },
     {
       "epoch": 0.03266722722149668,
-      "grad_norm": 0.0001841056946432218,
+      "grad_norm": 0.0002211649261880666,
       "learning_rate": 3.797232841391407e-06,
       "loss": 0.0,
       "step": 913
     },
     {
       "epoch": 0.032703007317029534,
-      "grad_norm": 0.0017745071090757847,
+      "grad_norm": 0.0013252223143354058,
       "learning_rate": 3.711957344141237e-06,
       "loss": 0.0001,
       "step": 914
     },
     {
       "epoch": 0.03273878741256239,
-      "grad_norm": 0.0002622443134896457,
+      "grad_norm": 0.0002602491294965148,
       "learning_rate": 3.627632170999029e-06,
       "loss": 0.0,
       "step": 915
     },
     {
       "epoch": 0.032774567508095244,
-      "grad_norm": 0.00042836813372559845,
+      "grad_norm": 0.00041321589378640056,
       "learning_rate": 3.5442581542201923e-06,
       "loss": 0.0,
       "step": 916
     },
     {
       "epoch": 0.0328103476036281,
-      "grad_norm": 0.0010627715382725,
+      "grad_norm": 0.0012678432976827025,
       "learning_rate": 3.461836116672612e-06,
       "loss": 0.0001,
       "step": 917
     },
     {
       "epoch": 0.032846127699160954,
-      "grad_norm": 0.00016633706400170922,
+      "grad_norm": 0.0001828453823691234,
       "learning_rate": 3.380366871828522e-06,
       "loss": 0.0,
       "step": 918
     },
     {
       "epoch": 0.03288190779469381,
-      "grad_norm": 0.0018421194981783628,
+      "grad_norm": 0.001912861131131649,
       "learning_rate": 3.2998512237565005e-06,
       "loss": 0.0001,
       "step": 919
     },
     {
       "epoch": 0.032917687890226664,
-      "grad_norm": 0.0004653351497836411,
+      "grad_norm": 0.00035834297887049615,
       "learning_rate": 3.2202899671134546e-06,
       "loss": 0.0,
       "step": 920
     },
     {
       "epoch": 0.03295346798575952,
-      "grad_norm": 0.00026660299045033753,
+      "grad_norm": 0.0003436813422013074,
       "learning_rate": 3.1416838871368924e-06,
       "loss": 0.0,
       "step": 921
     },
     {
       "epoch": 0.032989248081292374,
-      "grad_norm": 0.000764848489779979,
+      "grad_norm": 0.0003717408108059317,
       "learning_rate": 3.064033759637064e-06,
       "loss": 0.0,
       "step": 922
     },
     {
       "epoch": 0.03302502817682523,
-      "grad_norm": 0.0006508257938548923,
+      "grad_norm": 0.0005891156033612788,
       "learning_rate": 2.9873403509894203e-06,
       "loss": 0.0,
       "step": 923
     },
     {
       "epoch": 0.033060808272358085,
-      "grad_norm": 0.0003445699985604733,
+      "grad_norm": 0.00025783927412703633,
       "learning_rate": 2.9116044181269007e-06,
       "loss": 0.0,
       "step": 924
     },
     {
       "epoch": 0.03309658836789094,
-      "grad_norm": 0.011917107738554478,
+      "grad_norm": 0.00029453469323925674,
       "learning_rate": 2.836826708532603e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 925
     },
     {
       "epoch": 0.033132368463423795,
-      "grad_norm": 0.0006622477085329592,
+      "grad_norm": 0.0010198934469372034,
       "learning_rate": 2.7630079602323442e-06,
       "loss": 0.0001,
       "step": 926
     },
     {
       "epoch": 0.03316814855895665,
-      "grad_norm": 0.0005797698395326734,
+      "grad_norm": 0.004502155818045139,
       "learning_rate": 2.690148901787337e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 927
     },
     {
       "epoch": 0.033203928654489505,
-      "grad_norm": 0.00038914880133233964,
+      "grad_norm": 0.00021984580962453038,
       "learning_rate": 2.618250252287113e-06,
       "loss": 0.0,
       "step": 928
     },
     {
       "epoch": 0.03323970875002236,
-      "grad_norm": 0.0016823149053379893,
+      "grad_norm": 0.0007490668795071542,
       "learning_rate": 2.5473127213422763e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 929
     },
     {
       "epoch": 0.033275488845555215,
-      "grad_norm": 0.00021772326726932079,
+      "grad_norm": 0.0006981800543144345,
       "learning_rate": 2.4773370090776626e-06,
       "loss": 0.0,
       "step": 930
     },
     {
       "epoch": 0.033311268941088074,
-      "grad_norm": 0.0009772880002856255,
+      "grad_norm": 0.0020569113548845053,
       "learning_rate": 2.4083238061252567e-06,
       "loss": 0.0001,
       "step": 931
     },
     {
       "epoch": 0.033347049036620925,
-      "grad_norm": 0.0007127180579118431,
+      "grad_norm": 0.0018162206979468465,
       "learning_rate": 2.3402737936175425e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 932
     },
     {
       "epoch": 0.033382829132153784,
-      "grad_norm": 0.00041331606917083263,
+      "grad_norm": 0.00036989006912335753,
       "learning_rate": 2.273187643180652e-06,
       "loss": 0.0001,
       "step": 933
     },
     {
       "epoch": 0.033418609227686635,
-      "grad_norm": 0.0020597451366484165,
+      "grad_norm": 0.0003000934957526624,
       "learning_rate": 2.2070660169278166e-06,
       "loss": 0.0,
       "step": 934
     },
     {
       "epoch": 0.033454389323219494,
-      "grad_norm": 0.0027623341884464025,
+      "grad_norm": 0.001891492516733706,
       "learning_rate": 2.141909567452793e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 935
     },
     {
       "epoch": 0.033490169418752345,
-      "grad_norm": 0.0011368190171197057,
+      "grad_norm": 0.0007120324298739433,
       "learning_rate": 2.0777189378234143e-06,
       "loss": 0.0001,
       "step": 936
     },
     {
       "epoch": 0.033525949514285204,
-      "grad_norm": 0.0004307616618461907,
+      "grad_norm": 0.0004102467792108655,
       "learning_rate": 2.014494761575314e-06,
       "loss": 0.0,
       "step": 937
     },
     {
       "epoch": 0.033561729609818056,
-      "grad_norm": 0.004937837366014719,
+      "grad_norm": 0.00032427874975837767,
       "learning_rate": 1.9522376627055583e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 938
     },
     {
       "epoch": 0.033597509705350914,
-      "grad_norm": 0.0003586305247154087,
+      "grad_norm": 0.0007267544860951602,
       "learning_rate": 1.8909482556666024e-06,
       "loss": 0.0,
       "step": 939
     },
     {
       "epoch": 0.033633289800883766,
-      "grad_norm": 0.0004006979870609939,
+      "grad_norm": 0.000546029070392251,
       "learning_rate": 1.8306271453601199e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 940
     },
     {
       "epoch": 0.033669069896416624,
-      "grad_norm": 0.0473586842417717,
+      "grad_norm": 0.013414498418569565,
       "learning_rate": 1.771274927131139e-06,
-      "loss": 0.0003,
+      "loss": 0.0002,
       "step": 941
     },
     {
       "epoch": 0.033704849991949476,
-      "grad_norm": 0.015022980980575085,
+      "grad_norm": 0.0038646236062049866,
       "learning_rate": 1.712892186762083e-06,
-      "loss": 0.0003,
+      "loss": 0.0001,
       "step": 942
     },
     {
       "epoch": 0.033740630087482334,
-      "grad_norm": 0.0012572897830978036,
+      "grad_norm": 0.0012384445872157812,
       "learning_rate": 1.6554795004670388e-06,
       "loss": 0.0001,
       "step": 943
     },
     {
       "epoch": 0.033776410183015186,
-      "grad_norm": 0.0017297244630753994,
+      "grad_norm": 0.0006567625678144395,
       "learning_rate": 1.5990374348860305e-06,
       "loss": 0.0001,
       "step": 944
     },
     {
       "epoch": 0.033812190278548045,
-      "grad_norm": 0.00041694368701428175,
+      "grad_norm": 0.0008650491945445538,
       "learning_rate": 1.543566547079467e-06,
       "loss": 0.0,
       "step": 945
     },
     {
       "epoch": 0.033847970374080896,
-      "grad_norm": 0.0010576274944469333,
+      "grad_norm": 0.0006459559663198888,
       "learning_rate": 1.4890673845226133e-06,
       "loss": 0.0001,
       "step": 946
     },
     {
       "epoch": 0.033883750469613755,
-      "grad_norm": 0.027320250868797302,
+      "grad_norm": 0.0005734586156904697,
       "learning_rate": 1.4355404851001952e-06,
-      "loss": 0.0002,
+      "loss": 0.0,
       "step": 947
     },
     {
       "epoch": 0.033919530565146606,
-      "grad_norm": 0.003841983387246728,
+      "grad_norm": 0.0004001356428489089,
       "learning_rate": 1.3829863771011253e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 948
     },
     {
       "epoch": 0.033955310660679465,
-      "grad_norm": 0.0002816573833115399,
+      "grad_norm": 0.0002406181301921606,
       "learning_rate": 1.3314055792131964e-06,
       "loss": 0.0,
       "step": 949
     },
     {
       "epoch": 0.03399109075621232,
-      "grad_norm": 0.0001992636389331892,
+      "grad_norm": 0.00018440363055560738,
       "learning_rate": 1.280798600518085e-06,
       "loss": 0.0,
       "step": 950
     },
     {
       "epoch": 0.034026870851745175,
-      "grad_norm": 0.00025782539159990847,
+      "grad_norm": 0.00029384635854512453,
       "learning_rate": 1.231165940486234e-06,
       "loss": 0.0,
       "step": 951
     },
     {
       "epoch": 0.03406265094727803,
-      "grad_norm": 0.0006700262310914695,
+      "grad_norm": 0.0008108361507765949,
       "learning_rate": 1.1825080889719563e-06,
       "loss": 0.0,
       "step": 952
     },
     {
       "epoch": 0.034098431042810885,
-      "grad_norm": 0.003840812249109149,
+      "grad_norm": 0.0003570028638932854,
       "learning_rate": 1.134825526208605e-06,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 953
     },
     {
       "epoch": 0.03413421113834374,
-      "grad_norm": 0.0004403336497489363,
+      "grad_norm": 0.0013423620257526636,
       "learning_rate": 1.0881187228038215e-06,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 954
     },
     {
       "epoch": 0.034169991233876595,
-      "grad_norm": 0.00725098280236125,
+      "grad_norm": 0.005175633821636438,
       "learning_rate": 1.0423881397349068e-06,
       "loss": 0.0002,
       "step": 955
     },
     {
       "epoch": 0.03420577132940945,
-      "grad_norm": 0.0001444380613975227,
+      "grad_norm": 0.000696335278917104,
       "learning_rate": 9.976342283442463e-07,
       "loss": 0.0,
       "step": 956
     },
     {
       "epoch": 0.034241551424942306,
-      "grad_norm": 0.00039639120223000646,
+      "grad_norm": 0.00048268615501001477,
       "learning_rate": 9.538574303348813e-07,
       "loss": 0.0,
       "step": 957
     },
     {
       "epoch": 0.03427733152047516,
-      "grad_norm": 0.00030446768505498767,
+      "grad_norm": 0.0002519102126825601,
       "learning_rate": 9.110581777661331e-07,
       "loss": 0.0,
       "step": 958
     },
     {
       "epoch": 0.034313111616008016,
-      "grad_norm": 0.00035549886524677277,
+      "grad_norm": 0.0014691960532218218,
       "learning_rate": 8.692368930493521e-07,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 959
     },
     {
       "epoch": 0.03434889171154087,
-      "grad_norm": 0.0008194675319828093,
+      "grad_norm": 0.0012190488632768393,
       "learning_rate": 8.283939889437209e-07,
       "loss": 0.0,
       "step": 960
     },
     {
       "epoch": 0.034384671807073726,
-      "grad_norm": 0.00032311270479112864,
+      "grad_norm": 0.00024619302712380886,
       "learning_rate": 7.885298685522235e-07,
       "loss": 0.0,
       "step": 961
     },
     {
       "epoch": 0.03442045190260658,
-      "grad_norm": 0.006602166220545769,
+      "grad_norm": 0.0028114488814026117,
       "learning_rate": 7.496449253176274e-07,
-      "loss": 0.0002,
+      "loss": 0.0001,
       "step": 962
     },
     {
       "epoch": 0.034456231998139436,
-      "grad_norm": 0.0003104876959696412,
+      "grad_norm": 0.0008010416058823466,
       "learning_rate": 7.117395430186414e-07,
       "loss": 0.0,
       "step": 963
     },
     {
       "epoch": 0.03449201209367229,
-      "grad_norm": 0.0028515146113932133,
+      "grad_norm": 0.007256933022290468,
       "learning_rate": 6.748140957660631e-07,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 964
     },
     {
       "epoch": 0.034527792189205146,
-      "grad_norm": 0.0003642117662820965,
+      "grad_norm": 0.0004452927387319505,
       "learning_rate": 6.388689479991605e-07,
       "loss": 0.0,
       "step": 965
     },
     {
       "epoch": 0.034563572284738,
-      "grad_norm": 0.0008696328150108457,
+      "grad_norm": 0.0015947840875014663,
       "learning_rate": 6.039044544820404e-07,
       "loss": 0.0001,
       "step": 966
     },
     {
       "epoch": 0.034599352380270856,
-      "grad_norm": 0.004562847316265106,
+      "grad_norm": 0.006381189450621605,
       "learning_rate": 5.699209603001076e-07,
       "loss": 0.0001,
       "step": 967
     },
     {
       "epoch": 0.03463513247580371,
-      "grad_norm": 0.000312760123051703,
+      "grad_norm": 0.00047914046444930136,
       "learning_rate": 5.369188008567672e-07,
       "loss": 0.0,
       "step": 968
     },
     {
       "epoch": 0.034670912571336567,
-      "grad_norm": 0.0002709199907258153,
+      "grad_norm": 0.01858740672469139,
       "learning_rate": 5.048983018699827e-07,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 969
     },
     {
       "epoch": 0.03470669266686942,
-      "grad_norm": 0.0005034608766436577,
+      "grad_norm": 0.001538052223622799,
       "learning_rate": 4.738597793691679e-07,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 970
     },
     {
       "epoch": 0.03474247276240228,
-      "grad_norm": 0.0017387273255735636,
+      "grad_norm": 0.027744438499212265,
       "learning_rate": 4.438035396920004e-07,
-      "loss": 0.0001,
+      "loss": 0.0002,
       "step": 971
     },
     {
       "epoch": 0.03477825285793513,
-      "grad_norm": 0.001279267598874867,
+      "grad_norm": 0.00037432098179124296,
       "learning_rate": 4.1472987948143473e-07,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 972
     },
     {
       "epoch": 0.03481403295346799,
-      "grad_norm": 0.002777617424726486,
+      "grad_norm": 0.0005791685543954372,
       "learning_rate": 3.866390856827495e-07,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 973
     },
     {
       "epoch": 0.03484981304900084,
-      "grad_norm": 0.00028009427478536963,
+      "grad_norm": 0.0002605381014291197,
       "learning_rate": 3.595314355407609e-07,
       "loss": 0.0,
       "step": 974
     },
     {
       "epoch": 0.0348855931445337,
-      "grad_norm": 0.0006999737815931439,
+      "grad_norm": 0.0015622248174622655,
       "learning_rate": 3.3340719659701313e-07,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 975
     },
     {
       "epoch": 0.03492137324006655,
-      "grad_norm": 0.00026642257580533624,
+      "grad_norm": 0.00022047720267437398,
       "learning_rate": 3.0826662668720364e-07,
       "loss": 0.0,
       "step": 976
     },
     {
       "epoch": 0.03495715333559941,
-      "grad_norm": 0.0005622879252769053,
+      "grad_norm": 0.0006391391507349908,
       "learning_rate": 2.841099739386066e-07,
       "loss": 0.0,
       "step": 977
     },
     {
       "epoch": 0.03499293343113226,
-      "grad_norm": 0.07269036024808884,
+      "grad_norm": 0.2375146448612213,
       "learning_rate": 2.609374767676309e-07,
-      "loss": 0.0015,
+      "loss": 0.0042,
       "step": 978
     },
     {
       "epoch": 0.03502871352666512,
-      "grad_norm": 0.0005635939887724817,
+      "grad_norm": 0.0008213887922465801,
       "learning_rate": 2.387493638774774e-07,
       "loss": 0.0,
       "step": 979
     },
     {
       "epoch": 0.03506449362219797,
-      "grad_norm": 0.0002554512466304004,
+      "grad_norm": 0.0002956699172500521,
       "learning_rate": 2.175458542558517e-07,
       "loss": 0.0,
       "step": 980
     },
     {
       "epoch": 0.03510027371773083,
-      "grad_norm": 0.0005932984640821815,
+      "grad_norm": 0.0015261947410181165,
       "learning_rate": 1.973271571728441e-07,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 981
     },
     {
       "epoch": 0.03513605381326368,
-      "grad_norm": 0.00019584623805712909,
+      "grad_norm": 0.0001553288020659238,
       "learning_rate": 1.7809347217881966e-07,
       "loss": 0.0,
       "step": 982
     },
     {
       "epoch": 0.03517183390879654,
-      "grad_norm": 0.0009659458301030099,
+      "grad_norm": 0.0015854670200496912,
       "learning_rate": 1.598449891024978e-07,
       "loss": 0.0001,
       "step": 983
     },
     {
       "epoch": 0.03520761400432939,
-      "grad_norm": 0.5401099324226379,
+      "grad_norm": 0.010904781520366669,
       "learning_rate": 1.425818880490315e-07,
-      "loss": 0.0088,
+      "loss": 0.0001,
       "step": 984
     },
     {
       "epoch": 0.03524339409986225,
-      "grad_norm": 0.00018495542462915182,
+      "grad_norm": 0.00020986709569115192,
       "learning_rate": 1.2630433939825327e-07,
       "loss": 0.0,
       "step": 985
     },
     {
       "epoch": 0.0352791741953951,
-      "grad_norm": 0.0002920171828009188,
+      "grad_norm": 0.0002844655537046492,
       "learning_rate": 1.1101250380300965e-07,
       "loss": 0.0,
       "step": 986
     },
     {
       "epoch": 0.03531495429092796,
-      "grad_norm": 0.00020017915812786669,
+      "grad_norm": 0.00026205452741123736,
       "learning_rate": 9.670653218752934e-08,
       "loss": 0.0,
       "step": 987
     },
     {
       "epoch": 0.03535073438646081,
-      "grad_norm": 0.0010177441872656345,
+      "grad_norm": 0.0008910206379368901,
       "learning_rate": 8.33865657459909e-08,
       "loss": 0.0001,
       "step": 988
     },
     {
       "epoch": 0.03538651448199367,
-      "grad_norm": 0.0067826565355062485,
+      "grad_norm": 0.0008078790851868689,
       "learning_rate": 7.105273594107953e-08,
-      "loss": 0.0001,
+      "loss": 0.0,
       "step": 989
     },
     {
       "epoch": 0.03542229457752652,
-      "grad_norm": 0.0020364392548799515,
+      "grad_norm": 0.00078550138277933,
       "learning_rate": 5.970516450271025e-08,
       "loss": 0.0001,
       "step": 990
     },
     {
       "epoch": 0.03545807467305938,
-      "grad_norm": 0.001456920406781137,
+      "grad_norm": 0.004978294484317303,
       "learning_rate": 4.934396342684e-08,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 991
     },
     {
       "epoch": 0.03549385476859223,
-      "grad_norm": 0.004736809991300106,
+      "grad_norm": 0.0011256830766797066,
       "learning_rate": 3.996923497434635e-08,
       "loss": 0.0001,
       "step": 992
     },
     {
       "epoch": 0.03552963486412509,
-      "grad_norm": 0.00036006749724037945,
+      "grad_norm": 0.01774189993739128,
       "learning_rate": 3.1581071670006015e-08,
-      "loss": 0.0,
+      "loss": 0.0001,
       "step": 993
     },
     {
       "epoch": 0.03556541495965794,
-      "grad_norm": 0.00035346896038390696,
+      "grad_norm": 0.0002433753979858011,
       "learning_rate": 2.417955630159563e-08,
       "loss": 0.0,
       "step": 994
     },
     {
       "epoch": 0.0356011950551908,
-      "grad_norm": 0.0005142191657796502,
+      "grad_norm": 0.0002669621608220041,
       "learning_rate": 1.7764761919103477e-08,
       "loss": 0.0,
       "step": 995
     },
     {
       "epoch": 0.03563697515072365,
-      "grad_norm": 0.0003540781617630273,
+      "grad_norm": 0.0005420253146439791,
       "learning_rate": 1.2336751833941229e-08,
       "loss": 0.0,
       "step": 996
     },
     {
       "epoch": 0.03567275524625651,
-      "grad_norm": 0.00011586742766667157,
+      "grad_norm": 0.00019939179765060544,
       "learning_rate": 7.895579618388827e-09,
       "loss": 0.0,
       "step": 997
     },
     {
       "epoch": 0.03570853534178936,
-      "grad_norm": 0.0008120646234601736,
+      "grad_norm": 0.0011194644030183554,
       "learning_rate": 4.4412891050171765e-09,
       "loss": 0.0001,
       "step": 998
     },
     {
       "epoch": 0.03574431543732222,
-      "grad_norm": 0.00044350314419716597,
+      "grad_norm": 0.0004404011706355959,
       "learning_rate": 1.973914386288467e-09,
       "loss": 0.0,
       "step": 999
     },
     {
       "epoch": 0.03578009553285507,
-      "grad_norm": 0.0007635752554051578,
+      "grad_norm": 0.000688655418343842,
       "learning_rate": 4.934798141786879e-10,
       "loss": 0.0001,
       "step": 1000
+    },
+    {
+      "epoch": 0.03578009553285507,
+      "eval_loss": 6.354823563015088e-05,
+      "eval_runtime": 859.7993,
+      "eval_samples_per_second": 6.843,
+      "eval_steps_per_second": 6.843,
+      "step": 1000
     }
   ],
   "logging_steps": 1,