diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,9194 +3,2327 @@
   "best_model_checkpoint": null,
   "epoch": 0.9999235532451648,
   "eval_steps": 500,
-  "global_step": 6540,
+  "global_step": 1635,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00015289350967051448,
-      "grad_norm": 0.25844491471013625,
-      "learning_rate": 1.5290519877675841e-06,
-      "loss": 1.5751,
+      "epoch": 0.0006115740386820579,
+      "grad_norm": 0.12052220602859784,
+      "learning_rate": 6.0975609756097564e-06,
+      "loss": 1.6245,
       "step": 1
     },
     {
-      "epoch": 0.0007644675483525725,
-      "grad_norm": 0.23350215723561454,
-      "learning_rate": 7.645259938837921e-06,
-      "loss": 1.6179,
+      "epoch": 0.00305787019341029,
+      "grad_norm": 0.11718955528750491,
+      "learning_rate": 3.048780487804878e-05,
+      "loss": 1.5875,
       "step": 5
     },
     {
-      "epoch": 0.001528935096705145,
-      "grad_norm": 0.23748541894682731,
-      "learning_rate": 1.5290519877675842e-05,
-      "loss": 1.5771,
+      "epoch": 0.00611574038682058,
+      "grad_norm": 0.07463826304526239,
+      "learning_rate": 6.097560975609756e-05,
+      "loss": 1.5734,
       "step": 10
     },
     {
-      "epoch": 0.0022934026450577173,
-      "grad_norm": 0.20165559806581407,
-      "learning_rate": 2.2935779816513765e-05,
-      "loss": 1.5702,
+      "epoch": 0.009173610580230869,
+      "grad_norm": 0.07065399016948878,
+      "learning_rate": 9.146341463414634e-05,
+      "loss": 1.4926,
       "step": 15
     },
     {
-      "epoch": 0.00305787019341029,
-      "grad_norm": 0.15999746610192153,
-      "learning_rate": 3.0581039755351684e-05,
-      "loss": 1.5676,
+      "epoch": 0.01223148077364116,
+      "grad_norm": 0.06072873212977001,
+      "learning_rate": 0.00012195121951219512,
+      "loss": 1.458,
       "step": 20
     },
     {
-      "epoch": 0.0038223377417628624,
-      "grad_norm": 0.15565801558030515,
-      "learning_rate": 3.82262996941896e-05,
-      "loss": 1.5135,
+      "epoch": 0.01528935096705145,
+      "grad_norm": 0.0562127370758177,
+      "learning_rate": 0.0001524390243902439,
+      "loss": 1.4146,
       "step": 25
     },
     {
-      "epoch": 0.0045868052901154345,
-      "grad_norm": 0.1783724098220412,
-      "learning_rate": 4.587155963302753e-05,
-      "loss": 1.5051,
+      "epoch": 0.018347221160461738,
+      "grad_norm": 0.0369940258125185,
+      "learning_rate": 0.00018292682926829268,
+      "loss": 1.3727,
       "step": 30
     },
     {
-      "epoch": 0.005351272838468007,
-      "grad_norm": 0.14330557099797261,
-      "learning_rate": 5.3516819571865445e-05,
-      "loss": 1.4676,
+      "epoch": 0.021405091353872027,
+      "grad_norm": 0.03030856084221372,
+      "learning_rate": 0.00021341463414634146,
+      "loss": 1.3452,
       "step": 35
     },
     {
-      "epoch": 0.00611574038682058,
-      "grad_norm": 0.119715470035023,
-      "learning_rate": 6.116207951070337e-05,
-      "loss": 1.4462,
+      "epoch": 0.02446296154728232,
+      "grad_norm": 0.029052665253947514,
+      "learning_rate": 0.00024390243902439024,
+      "loss": 1.3062,
       "step": 40
     },
     {
-      "epoch": 0.006880207935173152,
-      "grad_norm": 0.11217846902570104,
-      "learning_rate": 6.880733944954129e-05,
-      "loss": 1.3848,
+      "epoch": 0.027520831740692607,
+      "grad_norm": 0.027169249076678257,
+      "learning_rate": 0.00027439024390243905,
+      "loss": 1.3193,
       "step": 45
     },
     {
-      "epoch": 0.007644675483525725,
-      "grad_norm": 0.10423836816079957,
-      "learning_rate": 7.64525993883792e-05,
-      "loss": 1.3812,
+      "epoch": 0.0305787019341029,
+      "grad_norm": 0.026308314290126287,
+      "learning_rate": 0.0003048780487804878,
+      "loss": 1.3059,
       "step": 50
     },
     {
-      "epoch": 0.008409143031878296,
-      "grad_norm": 0.10698713236366712,
-      "learning_rate": 8.409785932721712e-05,
-      "loss": 1.352,
+      "epoch": 0.033636572127513184,
+      "grad_norm": 0.02558312476230526,
+      "learning_rate": 0.0003353658536585366,
+      "loss": 1.2978,
       "step": 55
     },
     {
-      "epoch": 0.009173610580230869,
-      "grad_norm": 0.09868273967248942,
-      "learning_rate": 9.174311926605506e-05,
-      "loss": 1.374,
+      "epoch": 0.036694442320923476,
+      "grad_norm": 0.02366166156410773,
+      "learning_rate": 0.00036585365853658537,
+      "loss": 1.2835,
       "step": 60
     },
     {
-      "epoch": 0.009938078128583442,
-      "grad_norm": 0.08791146321468311,
-      "learning_rate": 9.938837920489297e-05,
-      "loss": 1.375,
+      "epoch": 0.03975231251433377,
+      "grad_norm": 0.023696910212749084,
+      "learning_rate": 0.0003963414634146342,
+      "loss": 1.2801,
       "step": 65
     },
     {
-      "epoch": 0.010702545676936013,
-      "grad_norm": 0.09568154037565704,
-      "learning_rate": 0.00010703363914373089,
-      "loss": 1.3381,
+      "epoch": 0.04281018270774405,
+      "grad_norm": 0.023961941292451862,
+      "learning_rate": 0.0004268292682926829,
+      "loss": 1.2862,
       "step": 70
     },
     {
-      "epoch": 0.011467013225288586,
-      "grad_norm": 0.09637997956354047,
-      "learning_rate": 0.00011467889908256881,
-      "loss": 1.3444,
+      "epoch": 0.045868052901154345,
+      "grad_norm": 0.026300989505884765,
+      "learning_rate": 0.00045731707317073173,
+      "loss": 1.2738,
       "step": 75
     },
     {
-      "epoch": 0.01223148077364116,
-      "grad_norm": 0.09217973928630054,
-      "learning_rate": 0.00012232415902140674,
-      "loss": 1.3413,
+      "epoch": 0.04892592309456464,
+      "grad_norm": 0.024085664746400153,
+      "learning_rate": 0.0004878048780487805,
+      "loss": 1.267,
       "step": 80
     },
     {
-      "epoch": 0.01299594832199373,
-      "grad_norm": 0.08948686274207032,
-      "learning_rate": 0.00012996941896024463,
-      "loss": 1.3062,
+      "epoch": 0.05198379328797492,
+      "grad_norm": 0.02400704254026435,
+      "learning_rate": 0.0005182926829268293,
+      "loss": 1.2687,
       "step": 85
     },
     {
-      "epoch": 0.013760415870346304,
-      "grad_norm": 0.08961531821702039,
-      "learning_rate": 0.00013761467889908258,
-      "loss": 1.3533,
+      "epoch": 0.055041663481385214,
+      "grad_norm": 0.028737409033153787,
+      "learning_rate": 0.0005487804878048781,
+      "loss": 1.248,
       "step": 90
     },
     {
-      "epoch": 0.014524883418698877,
-      "grad_norm": 0.09082536551122133,
-      "learning_rate": 0.0001452599388379205,
-      "loss": 1.3422,
+      "epoch": 0.058099533674795506,
+      "grad_norm": 0.033246208383736264,
+      "learning_rate": 0.0005792682926829268,
+      "loss": 1.2693,
       "step": 95
     },
     {
-      "epoch": 0.01528935096705145,
-      "grad_norm": 0.08401976510617441,
-      "learning_rate": 0.0001529051987767584,
-      "loss": 1.3419,
+      "epoch": 0.0611574038682058,
+      "grad_norm": 0.02684140817882857,
+      "learning_rate": 0.0006097560975609756,
+      "loss": 1.2562,
       "step": 100
     },
     {
-      "epoch": 0.016053818515404022,
-      "grad_norm": 0.13830177296315044,
-      "learning_rate": 0.00016055045871559635,
-      "loss": 1.313,
+      "epoch": 0.06421527406161609,
+      "grad_norm": 0.026557034480265206,
+      "learning_rate": 0.0006402439024390244,
+      "loss": 1.2286,
       "step": 105
     },
     {
-      "epoch": 0.016818286063756592,
-      "grad_norm": 0.08531248094678413,
-      "learning_rate": 0.00016819571865443425,
-      "loss": 1.308,
+      "epoch": 0.06727314425502637,
+      "grad_norm": 0.025820861749653763,
+      "learning_rate": 0.0006707317073170732,
+      "loss": 1.2678,
       "step": 110
     },
     {
-      "epoch": 0.017582753612109165,
-      "grad_norm": 0.09250245345131555,
-      "learning_rate": 0.00017584097859327217,
-      "loss": 1.3213,
+      "epoch": 0.07033101444843666,
+      "grad_norm": 0.029846828731098105,
+      "learning_rate": 0.0007012195121951219,
+      "loss": 1.2335,
       "step": 115
     },
     {
-      "epoch": 0.018347221160461738,
-      "grad_norm": 0.09366340563313799,
-      "learning_rate": 0.00018348623853211012,
-      "loss": 1.3243,
+      "epoch": 0.07338888464184695,
+      "grad_norm": 0.03508112839172254,
+      "learning_rate": 0.0007317073170731707,
+      "loss": 1.2663,
       "step": 120
     },
     {
-      "epoch": 0.01911168870881431,
-      "grad_norm": 0.09609974405919121,
-      "learning_rate": 0.00019113149847094801,
-      "loss": 1.2916,
+      "epoch": 0.07644675483525724,
+      "grad_norm": 0.0352184112282257,
+      "learning_rate": 0.0007621951219512195,
+      "loss": 1.2399,
       "step": 125
     },
     {
-      "epoch": 0.019876156257166884,
-      "grad_norm": 0.09358863492514614,
-      "learning_rate": 0.00019877675840978594,
-      "loss": 1.3238,
+      "epoch": 0.07950462502866754,
+      "grad_norm": 0.03322649965458742,
+      "learning_rate": 0.0007926829268292683,
+      "loss": 1.2503,
       "step": 130
     },
     {
-      "epoch": 0.020640623805519457,
-      "grad_norm": 0.08913000043934122,
-      "learning_rate": 0.00020642201834862386,
-      "loss": 1.2873,
+      "epoch": 0.08256249522207783,
+      "grad_norm": 0.028514272146253784,
+      "learning_rate": 0.000823170731707317,
+      "loss": 1.2468,
       "step": 135
     },
     {
-      "epoch": 0.021405091353872027,
-      "grad_norm": 0.0884733062294589,
-      "learning_rate": 0.00021406727828746178,
-      "loss": 1.2956,
+      "epoch": 0.0856203654154881,
+      "grad_norm": 0.028671295675629906,
+      "learning_rate": 0.0008536585365853659,
+      "loss": 1.23,
       "step": 140
     },
     {
-      "epoch": 0.0221695589022246,
-      "grad_norm": 0.08986088170695355,
-      "learning_rate": 0.0002217125382262997,
-      "loss": 1.2232,
+      "epoch": 0.0886782356088984,
+      "grad_norm": 0.030269321761170552,
+      "learning_rate": 0.0008841463414634147,
+      "loss": 1.2296,
       "step": 145
     },
     {
-      "epoch": 0.022934026450577173,
-      "grad_norm": 0.09329784590851617,
-      "learning_rate": 0.00022935779816513763,
-      "loss": 1.2924,
+      "epoch": 0.09173610580230869,
+      "grad_norm": 0.02846427788729512,
+      "learning_rate": 0.0009146341463414635,
+      "loss": 1.2241,
       "step": 150
     },
     {
-      "epoch": 0.023698493998929746,
-      "grad_norm": 0.09051908574519414,
-      "learning_rate": 0.00023700305810397555,
-      "loss": 1.2969,
+      "epoch": 0.09479397599571898,
+      "grad_norm": 0.03975120343371244,
+      "learning_rate": 0.0009451219512195122,
+      "loss": 1.244,
       "step": 155
     },
     {
-      "epoch": 0.02446296154728232,
-      "grad_norm": 0.08933188925183595,
-      "learning_rate": 0.00024464831804281347,
-      "loss": 1.2705,
+      "epoch": 0.09785184618912927,
+      "grad_norm": 0.03216136084206205,
+      "learning_rate": 0.000975609756097561,
+      "loss": 1.2457,
       "step": 160
     },
     {
-      "epoch": 0.02522742909563489,
-      "grad_norm": 0.09515017865326594,
-      "learning_rate": 0.00025229357798165137,
-      "loss": 1.3025,
+      "epoch": 0.10090971638253957,
+      "grad_norm": 0.03152661590457252,
+      "learning_rate": 0.000999998859712815,
+      "loss": 1.2326,
       "step": 165
     },
     {
-      "epoch": 0.02599189664398746,
-      "grad_norm": 0.0954616739977279,
-      "learning_rate": 0.00025993883792048926,
-      "loss": 1.2614,
+      "epoch": 0.10396758657594984,
+      "grad_norm": 0.031922913335073945,
+      "learning_rate": 0.0009999589502074485,
+      "loss": 1.249,
       "step": 170
     },
     {
-      "epoch": 0.026756364192340034,
-      "grad_norm": 0.08878617974216362,
-      "learning_rate": 0.00026758409785932727,
-      "loss": 1.3293,
+      "epoch": 0.10702545676936014,
+      "grad_norm": 0.04658476708749259,
+      "learning_rate": 0.0009998620315437507,
+      "loss": 1.2447,
       "step": 175
     },
     {
-      "epoch": 0.027520831740692607,
-      "grad_norm": 0.08664411019437332,
-      "learning_rate": 0.00027522935779816516,
-      "loss": 1.2649,
+      "epoch": 0.11008332696277043,
+      "grad_norm": 0.03617623229838695,
+      "learning_rate": 0.000999708114773132,
+      "loss": 1.2419,
       "step": 180
     },
     {
-      "epoch": 0.02828529928904518,
-      "grad_norm": 0.08911333992689248,
-      "learning_rate": 0.00028287461773700306,
-      "loss": 1.3307,
+      "epoch": 0.11314119715618072,
+      "grad_norm": 0.03418014417549059,
+      "learning_rate": 0.0009994972174463639,
+      "loss": 1.2197,
       "step": 185
     },
     {
-      "epoch": 0.029049766837397753,
-      "grad_norm": 0.0919892953815635,
-      "learning_rate": 0.000290519877675841,
-      "loss": 1.2688,
+      "epoch": 0.11619906734959101,
+      "grad_norm": 0.02577313047796619,
+      "learning_rate": 0.0009992293636115793,
+      "loss": 1.2334,
       "step": 190
     },
     {
-      "epoch": 0.029814234385750326,
-      "grad_norm": 0.08982364488062158,
-      "learning_rate": 0.0002981651376146789,
-      "loss": 1.3118,
+      "epoch": 0.1192569375430013,
+      "grad_norm": 0.026546715317612456,
+      "learning_rate": 0.0009989045838115292,
+      "loss": 1.2381,
       "step": 195
     },
     {
-      "epoch": 0.0305787019341029,
-      "grad_norm": 0.10317910874972114,
-      "learning_rate": 0.0003058103975535168,
-      "loss": 1.2273,
+      "epoch": 0.1223148077364116,
+      "grad_norm": 0.0326169458915842,
+      "learning_rate": 0.0009985229150800998,
+      "loss": 1.2143,
       "step": 200
     },
     {
-      "epoch": 0.03134316948245547,
-      "grad_norm": 0.0988002844753528,
-      "learning_rate": 0.00031345565749235475,
-      "loss": 1.282,
+      "epoch": 0.12537267792982187,
+      "grad_norm": 0.03223237682583807,
+      "learning_rate": 0.0009980844009380904,
+      "loss": 1.231,
       "step": 205
     },
     {
-      "epoch": 0.032107637030808045,
-      "grad_norm": 0.09497408054902434,
-      "learning_rate": 0.0003211009174311927,
-      "loss": 1.2672,
+      "epoch": 0.12843054812323218,
+      "grad_norm": 0.029643352282643223,
+      "learning_rate": 0.0009975890913882512,
+      "loss": 1.2469,
       "step": 210
     },
     {
-      "epoch": 0.032872104579160615,
-      "grad_norm": 0.0882881624063108,
-      "learning_rate": 0.0003287461773700306,
-      "loss": 1.2976,
+      "epoch": 0.13148841831664246,
+      "grad_norm": 0.027522481255624117,
+      "learning_rate": 0.0009970370429095794,
+      "loss": 1.2062,
       "step": 215
     },
     {
-      "epoch": 0.033636572127513184,
-      "grad_norm": 0.09864983052632054,
-      "learning_rate": 0.0003363914373088685,
-      "loss": 1.2793,
+      "epoch": 0.13454628851005274,
+      "grad_norm": 0.033075124541469436,
+      "learning_rate": 0.0009964283184508817,
+      "loss": 1.224,
       "step": 220
     },
     {
-      "epoch": 0.03440103967586576,
-      "grad_norm": 0.08805919903970645,
-      "learning_rate": 0.00034403669724770644,
-      "loss": 1.2742,
+      "epoch": 0.13760415870346304,
+      "grad_norm": 0.027377314114451212,
+      "learning_rate": 0.0009957629874235948,
+      "loss": 1.2086,
       "step": 225
     },
     {
-      "epoch": 0.03516550722421833,
-      "grad_norm": 0.09477232508209228,
-      "learning_rate": 0.00035168195718654434,
-      "loss": 1.2745,
+      "epoch": 0.14066202889687332,
+      "grad_norm": 0.03090829597386635,
+      "learning_rate": 0.000995041125693871,
+      "loss": 1.217,
       "step": 230
     },
     {
-      "epoch": 0.035929974772570906,
-      "grad_norm": 0.08374687785846717,
-      "learning_rate": 0.00035932721712538223,
-      "loss": 1.2962,
+      "epoch": 0.14371989909028363,
+      "grad_norm": 0.035975979448955,
+      "learning_rate": 0.0009942628155739275,
+      "loss": 1.2133,
       "step": 235
     },
     {
-      "epoch": 0.036694442320923476,
-      "grad_norm": 0.09194750042724101,
-      "learning_rate": 0.00036697247706422024,
-      "loss": 1.243,
+      "epoch": 0.1467777692836939,
+      "grad_norm": 0.029218388374324843,
+      "learning_rate": 0.00099342814581266,
+      "loss": 1.2056,
       "step": 240
     },
     {
-      "epoch": 0.03745890986927605,
-      "grad_norm": 0.09258202800876715,
-      "learning_rate": 0.00037461773700305813,
-      "loss": 1.293,
+      "epoch": 0.1498356394771042,
+      "grad_norm": 0.028688837478414952,
+      "learning_rate": 0.000992537211585524,
+      "loss": 1.2145,
       "step": 245
     },
     {
-      "epoch": 0.03822337741762862,
-      "grad_norm": 0.107430134301919,
-      "learning_rate": 0.00038226299694189603,
-      "loss": 1.2844,
+      "epoch": 0.1528935096705145,
+      "grad_norm": 0.0302429094788238,
+      "learning_rate": 0.0009915901144836802,
+      "loss": 1.2035,
       "step": 250
     },
     {
-      "epoch": 0.03898784496598119,
-      "grad_norm": 0.08615411635234742,
-      "learning_rate": 0.0003899082568807339,
-      "loss": 1.2506,
+      "epoch": 0.15595137986392477,
+      "grad_norm": 0.024901652208543252,
+      "learning_rate": 0.000990586962502413,
+      "loss": 1.2202,
       "step": 255
     },
     {
-      "epoch": 0.03975231251433377,
-      "grad_norm": 0.09531408169170505,
-      "learning_rate": 0.0003975535168195719,
-      "loss": 1.2637,
+      "epoch": 0.15900925005733507,
+      "grad_norm": 0.027892260766706888,
+      "learning_rate": 0.0009895278700288131,
+      "loss": 1.1807,
       "step": 260
     },
     {
-      "epoch": 0.04051678006268634,
-      "grad_norm": 0.08440291399032539,
-      "learning_rate": 0.00040519877675840977,
-      "loss": 1.2847,
+      "epoch": 0.16206712025074535,
+      "grad_norm": 0.03226968566991327,
+      "learning_rate": 0.000988412957828737,
+      "loss": 1.2072,
       "step": 265
     },
     {
-      "epoch": 0.041281247611038914,
-      "grad_norm": 0.09782115096835144,
-      "learning_rate": 0.0004128440366972477,
-      "loss": 1.2672,
+      "epoch": 0.16512499044415566,
+      "grad_norm": 0.03872574944189164,
+      "learning_rate": 0.000987242353033034,
+      "loss": 1.2037,
       "step": 270
     },
     {
-      "epoch": 0.042045715159391484,
-      "grad_norm": 0.10289516485156369,
-      "learning_rate": 0.00042048929663608567,
-      "loss": 1.3019,
+      "epoch": 0.16818286063756593,
+      "grad_norm": 0.033428446400930414,
+      "learning_rate": 0.000986016189123051,
+      "loss": 1.1912,
       "step": 275
     },
     {
-      "epoch": 0.04281018270774405,
-      "grad_norm": 0.10696722168409797,
-      "learning_rate": 0.00042813455657492356,
-      "loss": 1.2832,
+      "epoch": 0.1712407308309762,
+      "grad_norm": 0.0385131025139653,
+      "learning_rate": 0.0009847346059154124,
+      "loss": 1.2141,
       "step": 280
     },
     {
-      "epoch": 0.04357465025609663,
-      "grad_norm": 0.10607951895487863,
-      "learning_rate": 0.00043577981651376146,
-      "loss": 1.2772,
+      "epoch": 0.17429860102438652,
+      "grad_norm": 0.026232059261105282,
+      "learning_rate": 0.0009833977495460754,
+      "loss": 1.2019,
       "step": 285
     },
     {
-      "epoch": 0.0443391178044492,
-      "grad_norm": 0.08619472519717984,
-      "learning_rate": 0.0004434250764525994,
-      "loss": 1.2527,
+      "epoch": 0.1773564712177968,
+      "grad_norm": 0.024160196147301412,
+      "learning_rate": 0.000982005772453668,
+      "loss": 1.1787,
       "step": 290
     },
     {
-      "epoch": 0.045103585352801775,
-      "grad_norm": 0.09914776526024943,
-      "learning_rate": 0.0004510703363914373,
-      "loss": 1.3135,
+      "epoch": 0.1804143414112071,
+      "grad_norm": 0.022453529587095535,
+      "learning_rate": 0.0009805588333621066,
+      "loss": 1.1845,
       "step": 295
     },
     {
-      "epoch": 0.045868052901154345,
-      "grad_norm": 0.09807796164274683,
-      "learning_rate": 0.00045871559633027525,
-      "loss": 1.2597,
+      "epoch": 0.18347221160461738,
+      "grad_norm": 0.02720482538531544,
+      "learning_rate": 0.0009790570972624964,
+      "loss": 1.1738,
       "step": 300
     },
     {
-      "epoch": 0.04663252044950692,
-      "grad_norm": 0.10283330899312029,
-      "learning_rate": 0.00046636085626911315,
-      "loss": 1.2876,
+      "epoch": 0.18653008179802769,
+      "grad_norm": 0.02771296154155037,
+      "learning_rate": 0.000977500735394318,
+      "loss": 1.203,
       "step": 305
     },
     {
-      "epoch": 0.04739698799785949,
-      "grad_norm": 0.08801929670660691,
-      "learning_rate": 0.0004740061162079511,
-      "loss": 1.2711,
+      "epoch": 0.18958795199143796,
+      "grad_norm": 0.029629116640802273,
+      "learning_rate": 0.0009758899252259019,
+      "loss": 1.1867,
       "step": 310
     },
     {
-      "epoch": 0.04816145554621206,
-      "grad_norm": 0.10424236208657252,
-      "learning_rate": 0.000481651376146789,
-      "loss": 1.238,
+      "epoch": 0.19264582218484824,
+      "grad_norm": 0.024073515796522824,
+      "learning_rate": 0.0009742248504341918,
+      "loss": 1.1892,
       "step": 315
     },
     {
-      "epoch": 0.04892592309456464,
-      "grad_norm": 0.08993328037756107,
-      "learning_rate": 0.0004892966360856269,
-      "loss": 1.2834,
+      "epoch": 0.19570369237825855,
+      "grad_norm": 0.026256282210288567,
+      "learning_rate": 0.0009725057008838005,
+      "loss": 1.1895,
       "step": 320
     },
     {
-      "epoch": 0.04969039064291721,
-      "grad_norm": 0.10465808153748822,
-      "learning_rate": 0.0004969418960244648,
-      "loss": 1.2948,
+      "epoch": 0.19876156257166883,
+      "grad_norm": 0.027127167143907085,
+      "learning_rate": 0.0009707326726053596,
+      "loss": 1.1903,
       "step": 325
     },
     {
-      "epoch": 0.05045485819126978,
-      "grad_norm": 0.08903569990977848,
-      "learning_rate": 0.0005045871559633027,
-      "loss": 1.2672,
+      "epoch": 0.20181943276507913,
+      "grad_norm": 0.024066276808509948,
+      "learning_rate": 0.0009689059677731673,
+      "loss": 1.1732,
       "step": 330
     },
     {
-      "epoch": 0.05121932573962235,
-      "grad_norm": 0.10316518201667073,
-      "learning_rate": 0.0005122324159021407,
-      "loss": 1.2767,
+      "epoch": 0.2048773029584894,
+      "grad_norm": 0.03467571344471047,
+      "learning_rate": 0.0009670257946821346,
+      "loss": 1.1817,
       "step": 335
     },
     {
-      "epoch": 0.05198379328797492,
-      "grad_norm": 0.11171289817500524,
-      "learning_rate": 0.0005198776758409785,
-      "loss": 1.2606,
+      "epoch": 0.2079351731518997,
+      "grad_norm": 0.029277519444581707,
+      "learning_rate": 0.0009650923677240338,
+      "loss": 1.1949,
       "step": 340
     },
     {
-      "epoch": 0.0527482608363275,
-      "grad_norm": 0.09273950516093567,
-      "learning_rate": 0.0005275229357798165,
-      "loss": 1.3147,
+      "epoch": 0.21099304334531,
+      "grad_norm": 0.058400563521677724,
+      "learning_rate": 0.0009631059073630522,
+      "loss": 1.2189,
       "step": 345
     },
     {
-      "epoch": 0.05351272838468007,
-      "grad_norm": 0.09177716088521977,
-      "learning_rate": 0.0005351681957186545,
-      "loss": 1.2496,
+      "epoch": 0.21405091353872027,
+      "grad_norm": 0.035079123932762336,
+      "learning_rate": 0.0009610666401106527,
+      "loss": 1.183,
       "step": 350
     },
     {
-      "epoch": 0.054277195933032645,
-      "grad_norm": 0.11167903832080238,
-      "learning_rate": 0.0005428134556574924,
-      "loss": 1.2371,
+      "epoch": 0.21710878373213058,
+      "grad_norm": 0.027974026832942114,
+      "learning_rate": 0.0009589747984997454,
+      "loss": 1.1889,
       "step": 355
     },
     {
-      "epoch": 0.055041663481385214,
-      "grad_norm": 0.08734633477134537,
-      "learning_rate": 0.0005504587155963303,
-      "loss": 1.2304,
+      "epoch": 0.22016665392554086,
+      "grad_norm": 0.02669021024572584,
+      "learning_rate": 0.0009568306210581728,
+      "loss": 1.1787,
       "step": 360
     },
     {
-      "epoch": 0.05580613102973779,
-      "grad_norm": 0.08372782755538714,
-      "learning_rate": 0.0005581039755351683,
-      "loss": 1.2885,
+      "epoch": 0.22322452411895116,
+      "grad_norm": 0.02179715758089009,
+      "learning_rate": 0.00095463435228151,
+      "loss": 1.2111,
       "step": 365
     },
     {
-      "epoch": 0.05657059857809036,
-      "grad_norm": 0.0978669585781756,
-      "learning_rate": 0.0005657492354740061,
-      "loss": 1.2748,
+      "epoch": 0.22628239431236144,
+      "grad_norm": 0.02384963533856956,
+      "learning_rate": 0.0009523862426051868,
+      "loss": 1.1784,
       "step": 370
     },
     {
-      "epoch": 0.05733506612644293,
-      "grad_norm": 0.09122050324394194,
-      "learning_rate": 0.0005733944954128441,
-      "loss": 1.248,
+      "epoch": 0.22934026450577172,
+      "grad_norm": 0.028747252177852817,
+      "learning_rate": 0.00095008654837593,
+      "loss": 1.1883,
       "step": 375
     },
     {
-      "epoch": 0.058099533674795506,
-      "grad_norm": 0.10143711570915029,
-      "learning_rate": 0.000581039755351682,
-      "loss": 1.3127,
+      "epoch": 0.23239813469918202,
+      "grad_norm": 0.040361089898442656,
+      "learning_rate": 0.0009477355318225334,
+      "loss": 1.1769,
       "step": 380
     },
     {
-      "epoch": 0.058864001223148076,
-      "grad_norm": 0.088650537773865,
-      "learning_rate": 0.0005886850152905199,
-      "loss": 1.2848,
+      "epoch": 0.2354560048925923,
+      "grad_norm": 0.02974537276209321,
+      "learning_rate": 0.000945333461025956,
+      "loss": 1.1765,
       "step": 385
     },
     {
-      "epoch": 0.05962846877150065,
-      "grad_norm": 0.09185160704256218,
-      "learning_rate": 0.0005963302752293578,
-      "loss": 1.3034,
+      "epoch": 0.2385138750860026,
+      "grad_norm": 0.02326511823628993,
+      "learning_rate": 0.0009428806098887542,
+      "loss": 1.1613,
       "step": 390
     },
     {
-      "epoch": 0.06039293631985322,
-      "grad_norm": 0.08093356027719242,
-      "learning_rate": 0.0006039755351681957,
-      "loss": 1.229,
+      "epoch": 0.24157174527941289,
+      "grad_norm": 0.02788935788617548,
+      "learning_rate": 0.0009403772581038481,
+      "loss": 1.1885,
       "step": 395
     },
     {
-      "epoch": 0.0611574038682058,
-      "grad_norm": 0.0860338941976869,
-      "learning_rate": 0.0006116207951070336,
-      "loss": 1.269,
+      "epoch": 0.2446296154728232,
+      "grad_norm": 0.02733650935761564,
+      "learning_rate": 0.0009378236911226298,
+      "loss": 1.1413,
       "step": 400
     },
     {
-      "epoch": 0.06192187141655837,
-      "grad_norm": 0.08948858211939976,
-      "learning_rate": 0.0006192660550458716,
-      "loss": 1.2351,
+      "epoch": 0.24768748566623347,
+      "grad_norm": 0.026735407407419445,
+      "learning_rate": 0.0009352202001224133,
+      "loss": 1.1689,
       "step": 405
     },
     {
-      "epoch": 0.06268633896491094,
-      "grad_norm": 0.09429709970238356,
-      "learning_rate": 0.0006269113149847095,
-      "loss": 1.2485,
+      "epoch": 0.25074535585964375,
+      "grad_norm": 0.02496495083042904,
+      "learning_rate": 0.000932567081973233,
+      "loss": 1.1766,
       "step": 410
     },
     {
-      "epoch": 0.06345080651326351,
-      "grad_norm": 0.08705352394865537,
-      "learning_rate": 0.0006345565749235475,
-      "loss": 1.2537,
+      "epoch": 0.25380322605305405,
+      "grad_norm": 0.022860603493433784,
+      "learning_rate": 0.0009298646392039917,
+      "loss": 1.1601,
       "step": 415
     },
     {
-      "epoch": 0.06421527406161609,
-      "grad_norm": 0.1082260673439644,
-      "learning_rate": 0.0006422018348623854,
-      "loss": 1.2463,
+      "epoch": 0.25686109624646436,
+      "grad_norm": 0.023001099418093644,
+      "learning_rate": 0.0009271131799679638,
+      "loss": 1.17,
       "step": 420
     },
     {
-      "epoch": 0.06497974160996865,
-      "grad_norm": 0.09403045912297836,
-      "learning_rate": 0.0006498470948012232,
-      "loss": 1.304,
+      "epoch": 0.2599189664398746,
+      "grad_norm": 0.027817090057780446,
+      "learning_rate": 0.0009243130180076583,
+      "loss": 1.1614,
       "step": 425
     },
     {
-      "epoch": 0.06574420915832123,
-      "grad_norm": 0.10025637341779624,
-      "learning_rate": 0.0006574923547400612,
-      "loss": 1.2816,
+      "epoch": 0.2629768366332849,
+      "grad_norm": 0.024779365625928928,
+      "learning_rate": 0.000921464472619042,
+      "loss": 1.1744,
       "step": 430
     },
     {
-      "epoch": 0.0665086767066738,
-      "grad_norm": 0.10084432286928596,
-      "learning_rate": 0.0006651376146788991,
-      "loss": 1.3113,
+      "epoch": 0.2660347068266952,
+      "grad_norm": 0.02415651195912707,
+      "learning_rate": 0.0009185678686151322,
+      "loss": 1.1579,
       "step": 435
     },
     {
-      "epoch": 0.06727314425502637,
-      "grad_norm": 0.10873149177853551,
-      "learning_rate": 0.000672782874617737,
-      "loss": 1.264,
+      "epoch": 0.2690925770201055,
+      "grad_norm": 0.022815366652922293,
+      "learning_rate": 0.0009156235362889584,
+      "loss": 1.1745,
       "step": 440
     },
     {
-      "epoch": 0.06803761180337894,
-      "grad_norm": 0.08450706822759692,
-      "learning_rate": 0.0006804281345565749,
-      "loss": 1.2385,
+      "epoch": 0.2721504472135158,
+      "grad_norm": 0.021130481291703748,
+      "learning_rate": 0.0009126318113758999,
+      "loss": 1.1607,
       "step": 445
     },
     {
-      "epoch": 0.06880207935173152,
-      "grad_norm": 0.1055703401455179,
-      "learning_rate": 0.0006880733944954129,
-      "loss": 1.2603,
+      "epoch": 0.2752083174069261,
+      "grad_norm": 0.026167022396722542,
+      "learning_rate": 0.0009095930350154026,
+      "loss": 1.1874,
       "step": 450
     },
     {
-      "epoch": 0.0695665469000841,
-      "grad_norm": 0.09853955754584204,
-      "learning_rate": 0.0006957186544342507,
-      "loss": 1.2655,
+      "epoch": 0.2782661876003364,
+      "grad_norm": 0.03886381769784487,
+      "learning_rate": 0.0009065075537120796,
+      "loss": 1.1896,
       "step": 455
     },
     {
-      "epoch": 0.07033101444843666,
-      "grad_norm": 0.09398452513918377,
-      "learning_rate": 0.0007033639143730887,
-      "loss": 1.2665,
+      "epoch": 0.28132405779374664,
+      "grad_norm": 0.03341109366532242,
+      "learning_rate": 0.0009033757192962004,
+      "loss": 1.1575,
       "step": 460
     },
     {
-      "epoch": 0.07109548199678924,
-      "grad_norm": 0.0963643560310258,
-      "learning_rate": 0.0007110091743119266,
-      "loss": 1.2876,
+      "epoch": 0.28438192798715695,
+      "grad_norm": 0.03306894493562745,
+      "learning_rate": 0.0009001978888835723,
+      "loss": 1.1452,
       "step": 465
     },
     {
-      "epoch": 0.07185994954514181,
-      "grad_norm": 0.0856944074129867,
-      "learning_rate": 0.0007186544342507645,
-      "loss": 1.3032,
+      "epoch": 0.28743979818056725,
+      "grad_norm": 0.02506155876387002,
+      "learning_rate": 0.0008969744248348192,
+      "loss": 1.1589,
       "step": 470
     },
     {
-      "epoch": 0.07262441709349438,
-      "grad_norm": 0.08659127693558222,
-      "learning_rate": 0.0007262996941896025,
-      "loss": 1.3092,
+      "epoch": 0.2904976683739775,
+      "grad_norm": 0.025436594808100562,
+      "learning_rate": 0.0008937056947140625,
+      "loss": 1.188,
       "step": 475
     },
     {
-      "epoch": 0.07338888464184695,
-      "grad_norm": 0.0832981276048925,
-      "learning_rate": 0.0007339449541284405,
-      "loss": 1.2758,
+      "epoch": 0.2935555385673878,
+      "grad_norm": 0.023518916924014855,
+      "learning_rate": 0.0008903920712470089,
+      "loss": 1.1513,
       "step": 480
     },
     {
-      "epoch": 0.07415335219019953,
-      "grad_norm": 0.08230388249939921,
-      "learning_rate": 0.0007415902140672783,
-      "loss": 1.2788,
+      "epoch": 0.2966134087607981,
+      "grad_norm": 0.025275737643640807,
+      "learning_rate": 0.0008870339322784491,
+      "loss": 1.1646,
       "step": 485
     },
     {
-      "epoch": 0.0749178197385521,
-      "grad_norm": 0.10161902796868667,
-      "learning_rate": 0.0007492354740061163,
-      "loss": 1.2714,
+      "epoch": 0.2996712789542084,
+      "grad_norm": 0.02705678175117531,
+      "learning_rate": 0.0008836316607291732,
+      "loss": 1.1468,
       "step": 490
     },
     {
-      "epoch": 0.07568228728690467,
-      "grad_norm": 0.11416418403977804,
-      "learning_rate": 0.0007568807339449541,
-      "loss": 1.2577,
+      "epoch": 0.30272914914761867,
+      "grad_norm": 0.03300749247554839,
+      "learning_rate": 0.0008801856445523064,
+      "loss": 1.1593,
       "step": 495
     },
     {
-      "epoch": 0.07644675483525724,
-      "grad_norm": 0.09847812040362659,
-      "learning_rate": 0.0007645259938837921,
-      "loss": 1.2771,
+      "epoch": 0.305787019341029,
+      "grad_norm": 0.03070807591397227,
+      "learning_rate": 0.0008766962766890733,
+      "loss": 1.1494,
       "step": 500
     },
     {
-      "epoch": 0.07721122238360982,
-      "grad_norm": 0.08845565348706537,
-      "learning_rate": 0.00077217125382263,
-      "loss": 1.3169,
+      "epoch": 0.3088448895344393,
+      "grad_norm": 0.03171884578840513,
+      "learning_rate": 0.0008731639550239895,
+      "loss": 1.1549,
       "step": 505
     },
     {
-      "epoch": 0.07797568993196238,
-      "grad_norm": 0.09498376109239919,
-      "learning_rate": 0.0007798165137614678,
-      "loss": 1.2596,
+      "epoch": 0.31190275972784953,
+      "grad_norm": 0.026683099702115617,
+      "learning_rate": 0.0008695890823394938,
+      "loss": 1.1599,
       "step": 510
     },
     {
-      "epoch": 0.07874015748031496,
-      "grad_norm": 0.08367225972468065,
-      "learning_rate": 0.0007874617737003058,
-      "loss": 1.2634,
+      "epoch": 0.31496062992125984,
+      "grad_norm": 0.02544573139035867,
+      "learning_rate": 0.0008659720662700183,
+      "loss": 1.14,
       "step": 515
     },
     {
-      "epoch": 0.07950462502866754,
-      "grad_norm": 0.07963132772921086,
-      "learning_rate": 0.0007951070336391437,
-      "loss": 1.2872,
+      "epoch": 0.31801850011467014,
+      "grad_norm": 0.022400712579356837,
+      "learning_rate": 0.0008623133192555081,
+      "loss": 1.1631,
       "step": 520
     },
     {
-      "epoch": 0.0802690925770201,
-      "grad_norm": 0.0903050811552875,
-      "learning_rate": 0.0008027522935779816,
-      "loss": 1.2628,
+      "epoch": 0.3210763703080804,
+      "grad_norm": 0.021516644941970493,
+      "learning_rate": 0.0008586132584943902,
+      "loss": 1.1414,
       "step": 525
     },
     {
-      "epoch": 0.08103356012537268,
-      "grad_norm": 0.0835760948608985,
-      "learning_rate": 0.0008103975535168195,
-      "loss": 1.295,
+      "epoch": 0.3241342405014907,
+      "grad_norm": 0.022140987842416558,
+      "learning_rate": 0.0008548723058960033,
+      "loss": 1.1514,
       "step": 530
     },
     {
-      "epoch": 0.08179802767372525,
-      "grad_norm": 0.09170552594575584,
-      "learning_rate": 0.0008180428134556576,
-      "loss": 1.2793,
+      "epoch": 0.327192110694901,
+      "grad_norm": 0.02141121718297312,
+      "learning_rate": 0.0008510908880324864,
+      "loss": 1.1496,
       "step": 535
     },
     {
-      "epoch": 0.08256249522207783,
-      "grad_norm": 0.10994383729222293,
-      "learning_rate": 0.0008256880733944954,
-      "loss": 1.2947,
+      "epoch": 0.3302499808883113,
+      "grad_norm": 0.0230067385747688,
+      "learning_rate": 0.000847269436090139,
+      "loss": 1.1565,
       "step": 540
     },
     {
-      "epoch": 0.08332696277043039,
-      "grad_norm": 0.07270122130569258,
-      "learning_rate": 0.0008333333333333334,
-      "loss": 1.2592,
+      "epoch": 0.33330785108172156,
+      "grad_norm": 0.026857409244505942,
+      "learning_rate": 0.0008434083858202541,
+      "loss": 1.1541,
       "step": 545
     },
     {
-      "epoch": 0.08409143031878297,
-      "grad_norm": 0.09011634909964011,
-      "learning_rate": 0.0008409785932721713,
-      "loss": 1.2491,
+      "epoch": 0.33636572127513187,
+      "grad_norm": 0.031962019558439914,
+      "learning_rate": 0.0008395081774894297,
+      "loss": 1.1369,
       "step": 550
     },
     {
-      "epoch": 0.08485589786713554,
-      "grad_norm": 0.11082174674890546,
-      "learning_rate": 0.0008486238532110092,
-      "loss": 1.2943,
+      "epoch": 0.3394235914685422,
+      "grad_norm": 0.025328863714439345,
+      "learning_rate": 0.0008355692558293664,
+      "loss": 1.1462,
       "step": 555
     },
     {
-      "epoch": 0.0856203654154881,
-      "grad_norm": 0.08500973543905206,
-      "learning_rate": 0.0008562691131498471,
-      "loss": 1.2768,
+      "epoch": 0.3424814616619524,
+      "grad_norm": 0.02322136360750488,
+      "learning_rate": 0.0008315920699861561,
+      "loss": 1.1364,
       "step": 560
     },
     {
-      "epoch": 0.08638483296384068,
-      "grad_norm": 0.08246253797224005,
-      "learning_rate": 0.0008639143730886851,
-      "loss": 1.2644,
+      "epoch": 0.34553933185536273,
+      "grad_norm": 0.023916932106576035,
+      "learning_rate": 0.0008275770734690669,
+      "loss": 1.1376,
       "step": 565
     },
     {
-      "epoch": 0.08714930051219326,
-      "grad_norm": 0.08202189446656684,
-      "learning_rate": 0.0008715596330275229,
-      "loss": 1.2794,
+      "epoch": 0.34859720204877304,
+      "grad_norm": 0.027655600181262414,
+      "learning_rate": 0.0008235247240988294,
+      "loss": 1.1433,
       "step": 570
     },
     {
-      "epoch": 0.08791376806054584,
-      "grad_norm": 0.08492301097879416,
-      "learning_rate": 0.0008792048929663609,
-      "loss": 1.265,
+      "epoch": 0.35165507224218334,
+      "grad_norm": 0.01944294724363515,
+      "learning_rate": 0.0008194354839554344,
+      "loss": 1.1422,
       "step": 575
     },
     {
-      "epoch": 0.0886782356088984,
-      "grad_norm": 0.10488937164616725,
-      "learning_rate": 0.0008868501529051988,
-      "loss": 1.2663,
+      "epoch": 0.3547129424355936,
+      "grad_norm": 0.024586207516548813,
+      "learning_rate": 0.0008153098193254413,
+      "loss": 1.1481,
       "step": 580
     },
     {
-      "epoch": 0.08944270315725097,
-      "grad_norm": 0.09957818602896881,
-      "learning_rate": 0.0008944954128440367,
-      "loss": 1.2554,
+      "epoch": 0.3577708126290039,
+      "grad_norm": 0.02722431056639769,
+      "learning_rate": 0.0008111482006488091,
+      "loss": 1.1479,
       "step": 585
     },
     {
-      "epoch": 0.09020717070560355,
-      "grad_norm": 0.09037700230444715,
-      "learning_rate": 0.0009021406727828746,
-      "loss": 1.2805,
+      "epoch": 0.3608286828224142,
+      "grad_norm": 0.021338457814223475,
+      "learning_rate": 0.0008069511024652537,
+      "loss": 1.1444,
       "step": 590
     },
     {
-      "epoch": 0.09097163825395611,
-      "grad_norm": 0.08940645361248405,
-      "learning_rate": 0.0009097859327217126,
-      "loss": 1.2877,
+      "epoch": 0.36388655301582445,
+      "grad_norm": 0.025185341202656784,
+      "learning_rate": 0.0008027190033601363,
+      "loss": 1.1301,
       "step": 595
     },
     {
-      "epoch": 0.09173610580230869,
-      "grad_norm": 0.08800769193925378,
-      "learning_rate": 0.0009174311926605505,
-      "loss": 1.2453,
+      "epoch": 0.36694442320923476,
+      "grad_norm": 0.020971837960705533,
+      "learning_rate": 0.000798452385909892,
+      "loss": 1.1431,
       "step": 600
     },
     {
-      "epoch": 0.09250057335066127,
-      "grad_norm": 0.09302580222012974,
-      "learning_rate": 0.0009250764525993885,
-      "loss": 1.2957,
+      "epoch": 0.37000229340264507,
+      "grad_norm": 0.02495043534094607,
+      "learning_rate": 0.0007941517366270032,
+      "loss": 1.1356,
       "step": 605
     },
     {
-      "epoch": 0.09326504089901384,
-      "grad_norm": 0.08726126516074638,
-      "learning_rate": 0.0009327217125382263,
-      "loss": 1.2796,
+      "epoch": 0.37306016359605537,
+      "grad_norm": 0.025056486993634036,
+      "learning_rate": 0.0007898175459045224,
+      "loss": 1.1529,
       "step": 610
     },
     {
-      "epoch": 0.0940295084473664,
-      "grad_norm": 0.09201111299602344,
-      "learning_rate": 0.0009403669724770643,
-      "loss": 1.3069,
+      "epoch": 0.3761180337894656,
+      "grad_norm": 0.02276962025168977,
+      "learning_rate": 0.0007854503079601551,
+      "loss": 1.1236,
       "step": 615
     },
     {
-      "epoch": 0.09479397599571898,
-      "grad_norm": 0.10065773151168467,
-      "learning_rate": 0.0009480122324159022,
-      "loss": 1.2825,
+      "epoch": 0.37917590398287593,
+      "grad_norm": 0.021078280161330308,
+      "learning_rate": 0.0007810505207799046,
+      "loss": 1.137,
       "step": 620
     },
     {
-      "epoch": 0.09555844354407156,
-      "grad_norm": 0.07547253089389518,
-      "learning_rate": 0.00095565749235474,
-      "loss": 1.2775,
+      "epoch": 0.38223377417628623,
+      "grad_norm": 0.02117005306879687,
+      "learning_rate": 0.000776618686061288,
+      "loss": 1.1449,
       "step": 625
     },
     {
-      "epoch": 0.09632291109242412,
-      "grad_norm": 0.0795033812217181,
-      "learning_rate": 0.000963302752293578,
-      "loss": 1.3211,
+      "epoch": 0.3852916443696965,
+      "grad_norm": 0.024567794869791014,
+      "learning_rate": 0.0007721553091561288,
+      "loss": 1.139,
       "step": 630
     },
     {
-      "epoch": 0.0970873786407767,
-      "grad_norm": 0.08486183593424045,
-      "learning_rate": 0.0009709480122324159,
-      "loss": 1.2822,
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.019473444654760176,
+      "learning_rate": 0.000767660899012933,
+      "loss": 1.1386,
       "step": 635
     },
     {
-      "epoch": 0.09785184618912927,
-      "grad_norm": 0.09356435073800304,
-      "learning_rate": 0.0009785932721712539,
-      "loss": 1.3051,
+      "epoch": 0.3914073847565171,
+      "grad_norm": 0.022412121303153214,
+      "learning_rate": 0.0007631359681188544,
+      "loss": 1.1368,
       "step": 640
     },
     {
-      "epoch": 0.09861631373748185,
-      "grad_norm": 0.08681809761708813,
-      "learning_rate": 0.0009862385321100918,
-      "loss": 1.2784,
+      "epoch": 0.3944652549499274,
+      "grad_norm": 0.022050962340279187,
+      "learning_rate": 0.000758581032441257,
+      "loss": 1.1223,
       "step": 645
     },
     {
-      "epoch": 0.09938078128583441,
-      "grad_norm": 0.08805484661340418,
-      "learning_rate": 0.0009938837920489296,
-      "loss": 1.2898,
+      "epoch": 0.39752312514333765,
+      "grad_norm": 0.02203032943298021,
+      "learning_rate": 0.0007539966113688812,
+      "loss": 1.1167,
       "step": 650
     },
     {
-      "epoch": 0.10014524883418699,
-      "grad_norm": 0.10779188617692295,
-      "learning_rate": 0.0009999999287804496,
-      "loss": 1.3224,
+      "epoch": 0.40058099533674796,
+      "grad_norm": 0.023078837241982387,
+      "learning_rate": 0.0007493832276526181,
+      "loss": 1.1344,
       "step": 655
     },
     {
-      "epoch": 0.10090971638253957,
-      "grad_norm": 0.09449739996386951,
-      "learning_rate": 0.000999997436098313,
-      "loss": 1.2743,
+      "epoch": 0.40363886553015826,
+      "grad_norm": 0.023966897106968064,
+      "learning_rate": 0.0007447414073459017,
+      "loss": 1.1227,
       "step": 660
     },
     {
-      "epoch": 0.10167418393089213,
-      "grad_norm": 0.0963645222292544,
-      "learning_rate": 0.0009999913824589408,
-      "loss": 1.3092,
+      "epoch": 0.4066967357235685,
+      "grad_norm": 0.02168228889218337,
+      "learning_rate": 0.000740071679744725,
+      "loss": 1.1246,
       "step": 665
     },
     {
-      "epoch": 0.1024386514792447,
-      "grad_norm": 0.11266749758021803,
-      "learning_rate": 0.000999981767905447,
-      "loss": 1.296,
+      "epoch": 0.4097546059169788,
+      "grad_norm": 0.020601206009860356,
+      "learning_rate": 0.0007353745773272844,
+      "loss": 1.1227,
       "step": 670
     },
     {
-      "epoch": 0.10320311902759728,
-      "grad_norm": 0.37487450908071945,
-      "learning_rate": 0.000999968592506306,
-      "loss": 1.3499,
+      "epoch": 0.4128124761103891,
+      "grad_norm": 0.021802890558372243,
+      "learning_rate": 0.0007306506356932631,
+      "loss": 1.1037,
       "step": 675
     },
     {
-      "epoch": 0.10396758657594984,
-      "grad_norm": 0.14259324943638688,
-      "learning_rate": 0.000999951856355352,
-      "loss": 1.3757,
+      "epoch": 0.4158703463037994,
+      "grad_norm": 0.02158104255188381,
+      "learning_rate": 0.0007259003935027581,
+      "loss": 1.1166,
       "step": 680
     },
     {
-      "epoch": 0.10473205412430242,
-      "grad_norm": 0.30655516403662986,
-      "learning_rate": 0.0009999315595717797,
-      "loss": 1.4443,
+      "epoch": 0.4189282164972097,
+      "grad_norm": 0.02059398752407913,
+      "learning_rate": 0.0007211243924148578,
+      "loss": 1.1195,
       "step": 685
     },
     {
-      "epoch": 0.105496521672655,
-      "grad_norm": 1.0200782964851494,
-      "learning_rate": 0.000999907702300141,
-      "loss": 2.5802,
+      "epoch": 0.42198608669062,
+      "grad_norm": 0.022849828674870364,
+      "learning_rate": 0.0007163231770258777,
+      "loss": 1.1358,
       "step": 690
     },
     {
-      "epoch": 0.10626098922100757,
-      "grad_norm": 0.26954529309057906,
-      "learning_rate": 0.0009998802847103467,
-      "loss": 1.5332,
+      "epoch": 0.4250439568840303,
+      "grad_norm": 0.022923693857214923,
+      "learning_rate": 0.0007114972948072613,
+      "loss": 1.1388,
       "step": 695
     },
     {
-      "epoch": 0.10702545676936014,
-      "grad_norm": 0.2069498533750865,
-      "learning_rate": 0.0009998493069976635,
-      "loss": 1.464,
+      "epoch": 0.42810182707744054,
+      "grad_norm": 0.02318319704681315,
+      "learning_rate": 0.0007066472960431541,
+      "loss": 1.1154,
       "step": 700
     },
     {
-      "epoch": 0.10778992431771271,
-      "grad_norm": 0.11686570615912077,
-      "learning_rate": 0.000999814769382713,
-      "loss": 1.4344,
+      "epoch": 0.43115969727085085,
+      "grad_norm": 0.02746835308407499,
+      "learning_rate": 0.0007017737337676546,
+      "loss": 1.1222,
       "step": 705
     },
     {
-      "epoch": 0.10855439186606529,
-      "grad_norm": 0.24123211327141858,
-      "learning_rate": 0.0009997766721114708,
-      "loss": 1.4062,
+      "epoch": 0.43421756746426116,
+      "grad_norm": 0.02776028621219042,
+      "learning_rate": 0.0006968771637017541,
+      "loss": 1.1011,
       "step": 710
     },
     {
-      "epoch": 0.10931885941441785,
-      "grad_norm": 0.11666901013683244,
-      "learning_rate": 0.0009997350154552632,
-      "loss": 1.3302,
+      "epoch": 0.4372754376576714,
+      "grad_norm": 0.024266687223156804,
+      "learning_rate": 0.0006919581441899689,
+      "loss": 1.1246,
       "step": 715
     },
     {
-      "epoch": 0.11008332696277043,
-      "grad_norm": 0.12636047196811126,
-      "learning_rate": 0.000999689799710767,
-      "loss": 1.3883,
+      "epoch": 0.4403333078510817,
+      "grad_norm": 0.02908301149621474,
+      "learning_rate": 0.0006870172361366733,
+      "loss": 1.1225,
       "step": 720
     },
     {
-      "epoch": 0.110847794511123,
-      "grad_norm": 0.12003129463073069,
-      "learning_rate": 0.000999641025200007,
-      "loss": 1.3407,
+      "epoch": 0.443391178044492,
+      "grad_norm": 0.019862559304176748,
+      "learning_rate": 0.0006820550029421413,
+      "loss": 1.1367,
       "step": 725
     },
     {
-      "epoch": 0.11161226205947558,
-      "grad_norm": 0.09823944621605302,
-      "learning_rate": 0.0009995886922703524,
-      "loss": 1.3159,
+      "epoch": 0.4464490482379023,
+      "grad_norm": 0.032405405252682935,
+      "learning_rate": 0.000677072010438303,
+      "loss": 1.1234,
       "step": 730
     },
     {
-      "epoch": 0.11237672960782814,
-      "grad_norm": 0.13329784014265705,
-      "learning_rate": 0.0009995328012945158,
-      "loss": 1.3444,
+      "epoch": 0.4495069184313126,
+      "grad_norm": 0.030753416752446607,
+      "learning_rate": 0.0006720688268242248,
+      "loss": 1.1109,
       "step": 735
     },
     {
-      "epoch": 0.11314119715618072,
-      "grad_norm": 0.0786031784176081,
-      "learning_rate": 0.0009994733526705502,
-      "loss": 1.3057,
+      "epoch": 0.4525647886247229,
+      "grad_norm": 0.022597155712349652,
+      "learning_rate": 0.0006670460226013181,
+      "loss": 1.0944,
       "step": 740
     },
     {
-      "epoch": 0.1139056647045333,
-      "grad_norm": 0.07828669230906563,
-      "learning_rate": 0.0009994103468218456,
-      "loss": 1.3281,
+      "epoch": 0.4556226588181332,
+      "grad_norm": 0.022634198101240257,
+      "learning_rate": 0.0006620041705082871,
+      "loss": 1.1089,
       "step": 745
     },
     {
-      "epoch": 0.11467013225288586,
-      "grad_norm": 0.11852517280818008,
-      "learning_rate": 0.0009993437841971267,
-      "loss": 1.3017,
+      "epoch": 0.45868052901154344,
+      "grad_norm": 0.02271478233831073,
+      "learning_rate": 0.00065694384545582,
+      "loss": 1.1023,
       "step": 750
     },
     {
-      "epoch": 0.11543459980123844,
-      "grad_norm": 0.08558417909039735,
-      "learning_rate": 0.0009992736652704491,
-      "loss": 1.3356,
+      "epoch": 0.46173839920495374,
+      "grad_norm": 0.027216769976956785,
+      "learning_rate": 0.000651865624461034,
+      "loss": 1.1058,
       "step": 755
     },
     {
-      "epoch": 0.11619906734959101,
-      "grad_norm": 0.08594422889881462,
-      "learning_rate": 0.0009991999905411967,
-      "loss": 1.3238,
+      "epoch": 0.46479626939836405,
+      "grad_norm": 0.02071517184981896,
+      "learning_rate": 0.0006467700865816791,
+      "loss": 1.1083,
       "step": 760
     },
     {
-      "epoch": 0.11696353489794359,
-      "grad_norm": 0.19597834457509825,
-      "learning_rate": 0.0009991227605340768,
-      "loss": 1.3349,
+      "epoch": 0.46785413959177435,
+      "grad_norm": 0.02555353317127748,
+      "learning_rate": 0.0006416578128501092,
+      "loss": 1.0981,
       "step": 765
     },
     {
-      "epoch": 0.11772800244629615,
-      "grad_norm": 0.07985410977636408,
-      "learning_rate": 0.0009990419757991181,
-      "loss": 1.3364,
+      "epoch": 0.4709120097851846,
+      "grad_norm": 0.026631260066706193,
+      "learning_rate": 0.0006365293862070285,
+      "loss": 1.1091,
       "step": 770
     },
     {
-      "epoch": 0.11849246999464873,
-      "grad_norm": 0.1703047512929766,
-      "learning_rate": 0.0009989576369116653,
-      "loss": 1.3149,
+      "epoch": 0.4739698799785949,
+      "grad_norm": 0.022198028831592595,
+      "learning_rate": 0.0006313853914350207,
+      "loss": 1.1265,
       "step": 775
     },
     {
-      "epoch": 0.1192569375430013,
-      "grad_norm": 0.10335327963273161,
-      "learning_rate": 0.0009988697444723762,
-      "loss": 1.3157,
+      "epoch": 0.4770277501720052,
+      "grad_norm": 0.020216593671054946,
+      "learning_rate": 0.0006262264150918667,
+      "loss": 1.0885,
       "step": 780
     },
     {
-      "epoch": 0.12002140509135387,
-      "grad_norm": 0.08362146242319474,
-      "learning_rate": 0.000998778299107216,
-      "loss": 1.3042,
+      "epoch": 0.48008562036541547,
+      "grad_norm": 0.02929161675770138,
+      "learning_rate": 0.0006210530454436612,
+      "loss": 1.1049,
       "step": 785
     },
     {
-      "epoch": 0.12078587263970644,
-      "grad_norm": 0.08660098140550453,
-      "learning_rate": 0.0009986833014674543,
-      "loss": 1.3064,
+      "epoch": 0.48314349055882577,
+      "grad_norm": 0.022879526501220927,
+      "learning_rate": 0.0006158658723977341,
+      "loss": 1.1114,
       "step": 790
     },
     {
-      "epoch": 0.12155034018805902,
-      "grad_norm": 0.09078797333385379,
-      "learning_rate": 0.0009985847522296597,
-      "loss": 1.2792,
+      "epoch": 0.4862013607522361,
+      "grad_norm": 0.02011420292019708,
+      "learning_rate": 0.0006106654874353845,
+      "loss": 1.1147,
       "step": 795
     },
     {
-      "epoch": 0.1223148077364116,
-      "grad_norm": 0.09242815151780588,
-      "learning_rate": 0.000998482652095695,
-      "loss": 1.2806,
+      "epoch": 0.4892592309456464,
+      "grad_norm": 0.020765554185496773,
+      "learning_rate": 0.0006054524835444363,
+      "loss": 1.126,
       "step": 800
     },
     {
-      "epoch": 0.12307927528476416,
-      "grad_norm": 0.09723297340419826,
-      "learning_rate": 0.0009983770017927122,
-      "loss": 1.3326,
+      "epoch": 0.49231710113905663,
+      "grad_norm": 0.021368017020777876,
+      "learning_rate": 0.0006002274551516198,
+      "loss": 1.1141,
       "step": 805
     },
     {
-      "epoch": 0.12384374283311674,
-      "grad_norm": 0.09272895680580698,
-      "learning_rate": 0.0009982678020731476,
-      "loss": 1.3065,
+      "epoch": 0.49537497133246694,
+      "grad_norm": 0.023303801824493854,
+      "learning_rate": 0.0005949909980547917,
+      "loss": 1.1056,
       "step": 810
     },
     {
-      "epoch": 0.12460821038146931,
-      "grad_norm": 0.12136097817347706,
-      "learning_rate": 0.0009981550537147166,
-      "loss": 1.2986,
+      "epoch": 0.49843284152587725,
+      "grad_norm": 0.020988017572428172,
+      "learning_rate": 0.0005897437093549972,
+      "loss": 1.1046,
       "step": 815
     },
     {
-      "epoch": 0.12537267792982187,
-      "grad_norm": 0.08788778565772157,
-      "learning_rate": 0.000998038757520407,
-      "loss": 1.2805,
+      "epoch": 0.5014907117192875,
+      "grad_norm": 0.0222701220933236,
+      "learning_rate": 0.0005844861873883838,
+      "loss": 1.0896,
       "step": 820
     },
     {
-      "epoch": 0.12613714547817445,
-      "grad_norm": 0.08105841411135437,
-      "learning_rate": 0.000997918914318475,
-      "loss": 1.3446,
+      "epoch": 0.5045485819126978,
+      "grad_norm": 0.02390246675277257,
+      "learning_rate": 0.0005792190316579754,
+      "loss": 1.1003,
       "step": 825
     },
     {
-      "epoch": 0.12690161302652703,
-      "grad_norm": 0.07619675061909292,
-      "learning_rate": 0.0009977955249624378,
-      "loss": 1.316,
+      "epoch": 0.5076064521061081,
+      "grad_norm": 0.024599305753849318,
+      "learning_rate": 0.0005739428427653107,
+      "loss": 1.1006,
       "step": 830
     },
     {
-      "epoch": 0.1276660805748796,
-      "grad_norm": 0.0780011000474007,
-      "learning_rate": 0.0009976685903310685,
-      "loss": 1.2802,
+      "epoch": 0.5106643222995184,
+      "grad_norm": 0.021909626214719365,
+      "learning_rate": 0.0005686582223419595,
+      "loss": 1.0978,
       "step": 835
     },
     {
-      "epoch": 0.12843054812323218,
-      "grad_norm": 0.1123562503265519,
-      "learning_rate": 0.000997538111328389,
-      "loss": 1.3186,
+      "epoch": 0.5137221924929287,
+      "grad_norm": 0.020358006124918764,
+      "learning_rate": 0.0005633657729809194,
+      "loss": 1.0959,
       "step": 840
     },
     {
-      "epoch": 0.12919501567158473,
-      "grad_norm": 0.09717794248478596,
-      "learning_rate": 0.0009974040888836647,
-      "loss": 1.2766,
+      "epoch": 0.5167800626863389,
+      "grad_norm": 0.0198487307695255,
+      "learning_rate": 0.0005580660981679034,
+      "loss": 1.1,
       "step": 845
     },
     {
-      "epoch": 0.1299594832199373,
-      "grad_norm": 0.13339328046425755,
-      "learning_rate": 0.0009972665239513967,
-      "loss": 1.3051,
+      "epoch": 0.5198379328797492,
+      "grad_norm": 0.020615430577747694,
+      "learning_rate": 0.0005527598022125259,
+      "loss": 1.095,
       "step": 850
     },
     {
-      "epoch": 0.13072395076828988,
-      "grad_norm": 0.09665450497996363,
-      "learning_rate": 0.0009971254175113156,
-      "loss": 1.277,
+      "epoch": 0.5228958030731595,
+      "grad_norm": 0.02292927406023271,
+      "learning_rate": 0.0005474474901793951,
+      "loss": 1.1095,
       "step": 855
     },
     {
-      "epoch": 0.13148841831664246,
-      "grad_norm": 0.10830546508500209,
-      "learning_rate": 0.000996980770568375,
-      "loss": 1.2626,
+      "epoch": 0.5259536732665698,
+      "grad_norm": 0.023765712603430467,
+      "learning_rate": 0.0005421297678191189,
+      "loss": 1.0771,
       "step": 860
     },
     {
-      "epoch": 0.13225288586499503,
-      "grad_norm": 0.0892103153317403,
-      "learning_rate": 0.000996832584152743,
-      "loss": 1.2664,
+      "epoch": 0.5290115434599801,
+      "grad_norm": 0.01983795845597764,
+      "learning_rate": 0.0005368072414992314,
+      "loss": 1.1005,
       "step": 865
     },
     {
-      "epoch": 0.1330173534133476,
-      "grad_norm": 0.08476734354247635,
-      "learning_rate": 0.0009966808593197959,
-      "loss": 1.3118,
+      "epoch": 0.5320694136533904,
+      "grad_norm": 0.026870791253807334,
+      "learning_rate": 0.0005314805181350515,
+      "loss": 1.0942,
       "step": 870
     },
     {
-      "epoch": 0.1337818209617002,
-      "grad_norm": 0.07122127752767866,
-      "learning_rate": 0.0009965255971501107,
-      "loss": 1.2838,
+      "epoch": 0.5351272838468007,
+      "grad_norm": 0.01859295942083269,
+      "learning_rate": 0.0005261502051204772,
+      "loss": 1.0795,
       "step": 875
     },
     {
-      "epoch": 0.13454628851005274,
-      "grad_norm": 0.07151006579443676,
-      "learning_rate": 0.0009963667987494568,
-      "loss": 1.324,
+      "epoch": 0.538185154040211,
+      "grad_norm": 0.024323351242619737,
+      "learning_rate": 0.0005208169102587264,
+      "loss": 1.0888,
       "step": 880
     },
     {
-      "epoch": 0.1353107560584053,
-      "grad_norm": 0.08908137143916416,
-      "learning_rate": 0.000996204465248789,
-      "loss": 1.2627,
+      "epoch": 0.5412430242336213,
+      "grad_norm": 0.02524057653331344,
+      "learning_rate": 0.0005154812416930298,
+      "loss": 1.0908,
       "step": 885
     },
     {
-      "epoch": 0.1360752236067579,
-      "grad_norm": 0.07219573935906125,
-      "learning_rate": 0.0009960385978042383,
-      "loss": 1.2644,
+      "epoch": 0.5443008944270316,
+      "grad_norm": 0.018971386204347403,
+      "learning_rate": 0.0005101438078372858,
+      "loss": 1.0976,
       "step": 890
     },
     {
-      "epoch": 0.13683969115511047,
-      "grad_norm": 0.07819307629570971,
-      "learning_rate": 0.0009958691975971044,
-      "loss": 1.2889,
+      "epoch": 0.5473587646204419,
+      "grad_norm": 0.019908423848410486,
+      "learning_rate": 0.000504805217306685,
+      "loss": 1.093,
       "step": 895
     },
     {
-      "epoch": 0.13760415870346304,
-      "grad_norm": 0.07151962644193259,
-      "learning_rate": 0.0009956962658338472,
-      "loss": 1.2671,
+      "epoch": 0.5504166348138522,
+      "grad_norm": 0.0195002156944646,
+      "learning_rate": 0.0004994660788483113,
+      "loss": 1.0982,
       "step": 900
     },
     {
-      "epoch": 0.13836862625181562,
-      "grad_norm": 0.07987119674459224,
-      "learning_rate": 0.0009955198037460786,
-      "loss": 1.2814,
+      "epoch": 0.5534745050072625,
+      "grad_norm": 0.019088935901819853,
+      "learning_rate": 0.0004941270012717269,
+      "loss": 1.0946,
       "step": 905
     },
     {
-      "epoch": 0.1391330938001682,
-      "grad_norm": 0.07361684559744329,
-      "learning_rate": 0.0009953398125905528,
-      "loss": 1.2827,
+      "epoch": 0.5565323752006728,
+      "grad_norm": 0.020956181295062905,
+      "learning_rate": 0.000488788593379552,
+      "loss": 1.0927,
       "step": 910
     },
     {
-      "epoch": 0.13989756134852074,
-      "grad_norm": 0.06549715454074158,
-      "learning_rate": 0.000995156293649158,
-      "loss": 1.3029,
+      "epoch": 0.559590245394083,
+      "grad_norm": 0.02058468157622052,
+      "learning_rate": 0.00048345146389804425,
+      "loss": 1.0778,
       "step": 915
     },
     {
-      "epoch": 0.14066202889687332,
-      "grad_norm": 0.10854965228474792,
-      "learning_rate": 0.000994969248228907,
-      "loss": 1.252,
+      "epoch": 0.5626481155874933,
+      "grad_norm": 0.023584369471082292,
+      "learning_rate": 0.00047811622140768727,
+      "loss": 1.0821,
       "step": 920
     },
     {
-      "epoch": 0.1414264964452259,
-      "grad_norm": 0.08012814087734642,
-      "learning_rate": 0.000994778677661928,
-      "loss": 1.2825,
+      "epoch": 0.5657059857809036,
+      "grad_norm": 0.019349421074761057,
+      "learning_rate": 0.0004727834742737955,
+      "loss": 1.107,
       "step": 925
     },
     {
-      "epoch": 0.14219096399357847,
-      "grad_norm": 0.07656969405845815,
-      "learning_rate": 0.0009945845833054555,
-      "loss": 1.2584,
+      "epoch": 0.5687638559743139,
+      "grad_norm": 0.020725187288479826,
+      "learning_rate": 0.0004674538305771433,
+      "loss": 1.1124,
       "step": 930
     },
     {
-      "epoch": 0.14295543154193105,
-      "grad_norm": 0.0851834014068238,
-      "learning_rate": 0.00099438696654182,
-      "loss": 1.2469,
+      "epoch": 0.5718217261677242,
+      "grad_norm": 0.02330784277601695,
+      "learning_rate": 0.00046212789804462765,
+      "loss": 1.0776,
       "step": 935
     },
     {
-      "epoch": 0.14371989909028363,
-      "grad_norm": 0.08868829561728953,
-      "learning_rate": 0.0009941858287784383,
-      "loss": 1.3047,
+      "epoch": 0.5748795963611345,
+      "grad_norm": 0.020779970593931846,
+      "learning_rate": 0.0004568062839799696,
+      "loss": 1.0936,
       "step": 940
     },
     {
-      "epoch": 0.1444843666386362,
-      "grad_norm": 0.10386951169055887,
-      "learning_rate": 0.0009939811714478036,
-      "loss": 1.2649,
+      "epoch": 0.5779374665545448,
+      "grad_norm": 0.02310004433856537,
+      "learning_rate": 0.000451489595194466,
+      "loss": 1.0816,
       "step": 945
     },
     {
-      "epoch": 0.14524883418698875,
-      "grad_norm": 0.13301369066388727,
-      "learning_rate": 0.0009937729960074755,
-      "loss": 1.253,
+      "epoch": 0.580995336747955,
+      "grad_norm": 0.020049876509611084,
+      "learning_rate": 0.00044617843793779517,
+      "loss": 1.0776,
       "step": 950
     },
     {
-      "epoch": 0.14601330173534133,
-      "grad_norm": 0.10263670019334172,
-      "learning_rate": 0.000993561303940069,
-      "loss": 1.2709,
+      "epoch": 0.5840532069413653,
+      "grad_norm": 0.019669039054560983,
+      "learning_rate": 0.00044087341782888823,
+      "loss": 1.0813,
       "step": 955
     },
     {
-      "epoch": 0.1467777692836939,
-      "grad_norm": 0.10665302368773437,
-      "learning_rate": 0.0009933460967532454,
-      "loss": 1.2766,
+      "epoch": 0.5871110771347756,
+      "grad_norm": 0.019875654721293167,
+      "learning_rate": 0.0004355751397868713,
+      "loss": 1.0899,
       "step": 960
     },
     {
-      "epoch": 0.14754223683204648,
-      "grad_norm": 0.087814360033473,
-      "learning_rate": 0.0009931273759796989,
-      "loss": 1.285,
+      "epoch": 0.5901689473281859,
+      "grad_norm": 0.020596024583719397,
+      "learning_rate": 0.0004302842079620885,
+      "loss": 1.0753,
       "step": 965
     },
     {
-      "epoch": 0.14830670438039906,
-      "grad_norm": 0.0975006197140381,
-      "learning_rate": 0.0009929051431771486,
-      "loss": 1.2741,
+      "epoch": 0.5932268175215962,
+      "grad_norm": 0.020513561070135005,
+      "learning_rate": 0.0004250012256672117,
+      "loss": 1.0659,
       "step": 970
     },
     {
-      "epoch": 0.14907117192875163,
-      "grad_norm": 0.07858075681067718,
-      "learning_rate": 0.0009926793999283255,
-      "loss": 1.2972,
+      "epoch": 0.5962846877150065,
+      "grad_norm": 0.021144651470144736,
+      "learning_rate": 0.00041972679530844647,
+      "loss": 1.0839,
       "step": 975
     },
     {
-      "epoch": 0.1498356394771042,
-      "grad_norm": 0.0844266755046617,
-      "learning_rate": 0.0009924501478409617,
-      "loss": 1.2449,
+      "epoch": 0.5993425579084168,
+      "grad_norm": 0.021711998053494025,
+      "learning_rate": 0.0004144615183168409,
+      "loss": 1.0849,
       "step": 980
     },
     {
-      "epoch": 0.15060010702545676,
-      "grad_norm": 0.08128550636606187,
-      "learning_rate": 0.0009922173885477798,
-      "loss": 1.2777,
+      "epoch": 0.602400428101827,
+      "grad_norm": 0.022407376804965174,
+      "learning_rate": 0.0004092059950797062,
+      "loss": 1.0609,
       "step": 985
     },
     {
-      "epoch": 0.15136457457380934,
-      "grad_norm": 0.08733018440635835,
-      "learning_rate": 0.0009919811237064798,
-      "loss": 1.256,
+      "epoch": 0.6054582982952373,
+      "grad_norm": 0.02333806103193432,
+      "learning_rate": 0.0004039608248721548,
+      "loss": 1.0886,
       "step": 990
     },
     {
-      "epoch": 0.1521290421221619,
-      "grad_norm": 0.0866321204115484,
-      "learning_rate": 0.0009917413549997284,
-      "loss": 1.3139,
+      "epoch": 0.6085161684886476,
+      "grad_norm": 0.024102796234038874,
+      "learning_rate": 0.00039872660578876785,
+      "loss": 1.0737,
       "step": 995
     },
     {
-      "epoch": 0.1528935096705145,
-      "grad_norm": 0.08070145917396142,
-      "learning_rate": 0.0009914980841351465,
-      "loss": 1.2179,
+      "epoch": 0.611574038682058,
+      "grad_norm": 0.019357290501165046,
+      "learning_rate": 0.00039350393467539515,
+      "loss": 1.08,
       "step": 1000
     },
     {
-      "epoch": 0.15365797721886706,
-      "grad_norm": 0.08759371152417407,
-      "learning_rate": 0.0009912513128452973,
-      "loss": 1.2498,
+      "epoch": 0.6146319088754683,
+      "grad_norm": 0.020142530815836432,
+      "learning_rate": 0.0003882934070610981,
+      "loss": 1.0656,
       "step": 1005
     },
     {
-      "epoch": 0.15442244476721964,
-      "grad_norm": 0.07992027185128928,
-      "learning_rate": 0.000991001042887674,
-      "loss": 1.2898,
+      "epoch": 0.6176897790688786,
+      "grad_norm": 0.018608575204815583,
+      "learning_rate": 0.00038309561709024356,
+      "loss": 1.0637,
       "step": 1010
     },
     {
-      "epoch": 0.15518691231557222,
-      "grad_norm": 0.09915223808905828,
-      "learning_rate": 0.0009907472760446868,
-      "loss": 1.2784,
+      "epoch": 0.6207476492622889,
+      "grad_norm": 0.021214222795490626,
+      "learning_rate": 0.0003779111574547537,
+      "loss": 1.08,
       "step": 1015
     },
     {
-      "epoch": 0.15595137986392477,
-      "grad_norm": 0.07529248014649702,
-      "learning_rate": 0.0009904900141236505,
-      "loss": 1.3012,
+      "epoch": 0.6238055194556991,
+      "grad_norm": 0.022933622407233706,
+      "learning_rate": 0.00037274061932652413,
+      "loss": 1.0664,
       "step": 1020
     },
     {
-      "epoch": 0.15671584741227734,
-      "grad_norm": 0.07116825555952118,
-      "learning_rate": 0.0009902292589567723,
-      "loss": 1.225,
+      "epoch": 0.6268633896491094,
+      "grad_norm": 0.019706448651309442,
+      "learning_rate": 0.000367584592290013,
+      "loss": 1.0806,
       "step": 1025
     },
     {
-      "epoch": 0.15748031496062992,
-      "grad_norm": 0.0721270831702372,
-      "learning_rate": 0.0009899650124011376,
-      "loss": 1.2453,
+      "epoch": 0.6299212598425197,
+      "grad_norm": 0.02379452277464871,
+      "learning_rate": 0.00036244366427501265,
+      "loss": 1.0664,
       "step": 1030
     },
     {
-      "epoch": 0.1582447825089825,
-      "grad_norm": 0.08089190163094111,
-      "learning_rate": 0.0009896972763386973,
-      "loss": 1.2494,
+      "epoch": 0.63297913003593,
+      "grad_norm": 0.022836969922118866,
+      "learning_rate": 0.0003573184214896087,
+      "loss": 1.0702,
       "step": 1035
     },
     {
-      "epoch": 0.15900925005733507,
-      "grad_norm": 0.07205992090968301,
-      "learning_rate": 0.0009894260526762547,
-      "loss": 1.2383,
+      "epoch": 0.6360370002293403,
+      "grad_norm": 0.0220847265087786,
+      "learning_rate": 0.0003522094483533369,
+      "loss": 1.0777,
       "step": 1040
     },
     {
-      "epoch": 0.15977371760568765,
-      "grad_norm": 0.32600838075966987,
-      "learning_rate": 0.0009891513433454512,
-      "loss": 1.2791,
+      "epoch": 0.6390948704227506,
+      "grad_norm": 0.021238471439531986,
+      "learning_rate": 0.000347117327430542,
+      "loss": 1.0827,
       "step": 1045
     },
     {
-      "epoch": 0.1605381851540402,
-      "grad_norm": 0.06955922752630836,
-      "learning_rate": 0.0009888731503027535,
-      "loss": 1.2568,
+      "epoch": 0.6421527406161608,
+      "grad_norm": 0.01967371179793323,
+      "learning_rate": 0.0003420426393639508,
+      "loss": 1.0638,
       "step": 1050
     },
     {
-      "epoch": 0.16130265270239277,
-      "grad_norm": 0.10616255888347151,
-      "learning_rate": 0.0009885914755294386,
-      "loss": 1.2747,
+      "epoch": 0.6452106108095711,
+      "grad_norm": 0.022296186314483487,
+      "learning_rate": 0.00033698596280846125,
+      "loss": 1.0639,
       "step": 1055
     },
     {
-      "epoch": 0.16206712025074535,
-      "grad_norm": 0.11210822875448243,
-      "learning_rate": 0.0009883063210315804,
-      "loss": 1.3125,
+      "epoch": 0.6482684810029814,
+      "grad_norm": 0.02241422770345421,
+      "learning_rate": 0.0003319478743651609,
+      "loss": 1.0851,
       "step": 1060
     },
     {
-      "epoch": 0.16283158779909793,
-      "grad_norm": 0.10366169122798494,
-      "learning_rate": 0.0009880176888400352,
-      "loss": 1.3034,
+      "epoch": 0.6513263511963917,
+      "grad_norm": 0.02523497459965589,
+      "learning_rate": 0.0003269289485155783,
+      "loss": 1.0847,
       "step": 1065
     },
     {
-      "epoch": 0.1635960553474505,
-      "grad_norm": 0.6763547146869965,
-      "learning_rate": 0.0009877255810104275,
-      "loss": 1.2877,
+      "epoch": 0.654384221389802,
+      "grad_norm": 0.025512702330510183,
+      "learning_rate": 0.00032192975755617506,
+      "loss": 1.0662,
       "step": 1070
     },
     {
-      "epoch": 0.16436052289580308,
-      "grad_norm": 0.12323825764543753,
-      "learning_rate": 0.000987429999623135,
-      "loss": 1.2683,
+      "epoch": 0.6574420915832123,
+      "grad_norm": 0.019599747988122102,
+      "learning_rate": 0.00031695087153308976,
+      "loss": 1.0766,
       "step": 1075
     },
     {
-      "epoch": 0.16512499044415566,
-      "grad_norm": 0.14377535741587366,
-      "learning_rate": 0.0009871309467832737,
-      "loss": 1.3187,
+      "epoch": 0.6604999617766226,
+      "grad_norm": 0.01962272505893558,
+      "learning_rate": 0.0003119928581771362,
+      "loss": 1.0625,
       "step": 1080
     },
     {
-      "epoch": 0.1658894579925082,
-      "grad_norm": 0.0913096688213412,
-      "learning_rate": 0.0009868284246206834,
-      "loss": 1.2987,
+      "epoch": 0.6635578319700328,
+      "grad_norm": 0.019771169590040102,
+      "learning_rate": 0.0003070562828390662,
+      "loss": 1.0689,
       "step": 1085
     },
     {
-      "epoch": 0.16665392554086078,
-      "grad_norm": 0.1261260904406275,
-      "learning_rate": 0.0009865224352899118,
-      "loss": 1.2607,
+      "epoch": 0.6666157021634431,
+      "grad_norm": 0.02030493911602512,
+      "learning_rate": 0.00030214170842510435,
+      "loss": 1.0769,
       "step": 1090
     },
     {
-      "epoch": 0.16741839308921336,
-      "grad_norm": 0.07986542658284047,
-      "learning_rate": 0.0009862129809702005,
-      "loss": 1.2815,
+      "epoch": 0.6696735723568534,
+      "grad_norm": 0.024601514965605153,
+      "learning_rate": 0.0002972496953327607,
+      "loss": 1.062,
       "step": 1095
     },
     {
-      "epoch": 0.16818286063756593,
-      "grad_norm": 0.11156681330527404,
-      "learning_rate": 0.0009859000638654674,
-      "loss": 1.2537,
+      "epoch": 0.6727314425502637,
+      "grad_norm": 0.01994828738024262,
+      "learning_rate": 0.00029238080138693025,
+      "loss": 1.0775,
       "step": 1100
     },
     {
-      "epoch": 0.1689473281859185,
-      "grad_norm": 0.07080066680629181,
-      "learning_rate": 0.000985583686204293,
-      "loss": 1.2753,
+      "epoch": 0.675789312743674,
+      "grad_norm": 0.028014612142516805,
+      "learning_rate": 0.0002875355817762856,
+      "loss": 1.0567,
       "step": 1105
     },
     {
-      "epoch": 0.1697117957342711,
-      "grad_norm": 0.08751280576746387,
-      "learning_rate": 0.0009852638502399036,
-      "loss": 1.2832,
+      "epoch": 0.6788471829370843,
+      "grad_norm": 0.02154035563943042,
+      "learning_rate": 0.0002827145889899698,
+      "loss": 1.0451,
       "step": 1110
     },
     {
-      "epoch": 0.17047626328262366,
-      "grad_norm": 0.07513697430712966,
-      "learning_rate": 0.0009849405582501552,
-      "loss": 1.2826,
+      "epoch": 0.6819050531304947,
+      "grad_norm": 0.019908200712777067,
+      "learning_rate": 0.0002779183727545973,
+      "loss": 1.0715,
       "step": 1115
     },
     {
-      "epoch": 0.1712407308309762,
-      "grad_norm": 0.07040068722014306,
-      "learning_rate": 0.0009846138125375174,
-      "loss": 1.3079,
+      "epoch": 0.6849629233239048,
+      "grad_norm": 0.019831506624958018,
+      "learning_rate": 0.0002731474799715701,
+      "loss": 1.0794,
       "step": 1120
     },
     {
-      "epoch": 0.1720051983793288,
-      "grad_norm": 0.07554763974923043,
-      "learning_rate": 0.0009842836154290576,
-      "loss": 1.2688,
+      "epoch": 0.6880207935173152,
+      "grad_norm": 0.020609762986790316,
+      "learning_rate": 0.00026840245465471535,
+      "loss": 1.0648,
       "step": 1125
     },
     {
-      "epoch": 0.17276966592768137,
-      "grad_norm": 0.0756416076694629,
-      "learning_rate": 0.0009839499692764226,
-      "loss": 1.2845,
+      "epoch": 0.6910786637107255,
+      "grad_norm": 0.018373460203891627,
+      "learning_rate": 0.0002636838378682534,
+      "loss": 1.0642,
       "step": 1130
     },
     {
-      "epoch": 0.17353413347603394,
-      "grad_norm": 0.08613603811511952,
-      "learning_rate": 0.0009836128764558248,
-      "loss": 1.2946,
+      "epoch": 0.6941365339041358,
+      "grad_norm": 0.021227917010586402,
+      "learning_rate": 0.00025899216766510114,
+      "loss": 1.0574,
       "step": 1135
     },
     {
-      "epoch": 0.17429860102438652,
-      "grad_norm": 0.09369729151974834,
-      "learning_rate": 0.000983272339368022,
-      "loss": 1.2453,
+      "epoch": 0.6971944040975461,
+      "grad_norm": 0.021984002313747204,
+      "learning_rate": 0.0002543279790255192,
+      "loss": 1.0603,
       "step": 1140
     },
     {
-      "epoch": 0.1750630685727391,
-      "grad_norm": 0.07994080822915721,
-      "learning_rate": 0.0009829283604383034,
-      "loss": 1.2671,
+      "epoch": 0.7002522742909564,
+      "grad_norm": 0.02355426186863231,
+      "learning_rate": 0.00024969180379610934,
+      "loss": 1.0586,
       "step": 1145
     },
     {
-      "epoch": 0.17582753612109167,
-      "grad_norm": 0.10575664278992154,
-      "learning_rate": 0.0009825809421164695,
-      "loss": 1.2497,
+      "epoch": 0.7033101444843667,
+      "grad_norm": 0.024678418156208182,
+      "learning_rate": 0.00024508417062916885,
+      "loss": 1.0661,
       "step": 1150
     },
     {
-      "epoch": 0.17659200366944422,
-      "grad_norm": 0.10195388695968033,
-      "learning_rate": 0.0009822300868768169,
-      "loss": 1.2472,
+      "epoch": 0.7063680146777769,
+      "grad_norm": 0.02082314009145273,
+      "learning_rate": 0.0002405056049224103,
+      "loss": 1.0651,
       "step": 1155
     },
     {
-      "epoch": 0.1773564712177968,
-      "grad_norm": 0.08122506603589001,
-      "learning_rate": 0.000981875797218119,
-      "loss": 1.244,
+      "epoch": 0.7094258848711872,
+      "grad_norm": 0.02162054379835859,
+      "learning_rate": 0.0002359566287590511,
+      "loss": 1.0715,
       "step": 1160
     },
     {
-      "epoch": 0.17812093876614937,
-      "grad_norm": 0.09114356191080156,
-      "learning_rate": 0.00098151807566361,
-      "loss": 1.2631,
+      "epoch": 0.7124837550645975,
+      "grad_norm": 0.021178840967147644,
+      "learning_rate": 0.00023143776084828178,
+      "loss": 1.045,
       "step": 1165
     },
     {
-      "epoch": 0.17888540631450195,
-      "grad_norm": 0.12085241472592927,
-      "learning_rate": 0.0009811569247609646,
-      "loss": 1.2685,
+      "epoch": 0.7155416252580078,
+      "grad_norm": 0.023789035651104056,
+      "learning_rate": 0.00022694951646611884,
+      "loss": 1.0618,
       "step": 1170
     },
     {
-      "epoch": 0.17964987386285453,
-      "grad_norm": 0.10376652091732114,
-      "learning_rate": 0.0009807923470822825,
-      "loss": 1.2457,
+      "epoch": 0.7185994954514181,
+      "grad_norm": 0.019246434577242334,
+      "learning_rate": 0.0002224924073966484,
+      "loss": 1.0638,
       "step": 1175
     },
     {
-      "epoch": 0.1804143414112071,
-      "grad_norm": 0.08582105488308488,
-      "learning_rate": 0.0009804243452240676,
-      "loss": 1.2655,
+      "epoch": 0.7216573656448284,
+      "grad_norm": 0.018166777369856113,
+      "learning_rate": 0.00021806694187366976,
+      "loss": 1.0551,
       "step": 1180
     },
     {
-      "epoch": 0.18117880895955968,
-      "grad_norm": 0.08069706067779515,
-      "learning_rate": 0.0009800529218072112,
-      "loss": 1.242,
+      "epoch": 0.7247152358382387,
+      "grad_norm": 0.01932908428350948,
+      "learning_rate": 0.00021367362452274198,
+      "loss": 1.0559,
       "step": 1185
     },
     {
-      "epoch": 0.18194327650791223,
-      "grad_norm": 0.09899431989367119,
-      "learning_rate": 0.0009796780794769726,
-      "loss": 1.2136,
+      "epoch": 0.7277731060316489,
+      "grad_norm": 0.019018633917993133,
+      "learning_rate": 0.00020931295630364222,
+      "loss": 1.0542,
       "step": 1190
     },
     {
-      "epoch": 0.1827077440562648,
-      "grad_norm": 0.0721358139689406,
-      "learning_rate": 0.0009792998209029607,
-      "loss": 1.2691,
+      "epoch": 0.7308309762250592,
+      "grad_norm": 0.019157449963553598,
+      "learning_rate": 0.00020498543445324346,
+      "loss": 1.0572,
       "step": 1195
     },
     {
-      "epoch": 0.18347221160461738,
-      "grad_norm": 0.1851037362834397,
-      "learning_rate": 0.0009789181487791145,
-      "loss": 1.252,
+      "epoch": 0.7338888464184695,
+      "grad_norm": 0.018701309680460174,
+      "learning_rate": 0.00020069155242881525,
+      "loss": 1.064,
       "step": 1200
     },
     {
-      "epoch": 0.18423667915296996,
-      "grad_norm": 0.07551235120975437,
-      "learning_rate": 0.0009785330658236841,
-      "loss": 1.2771,
+      "epoch": 0.7369467166118798,
+      "grad_norm": 0.020107359392638562,
+      "learning_rate": 0.00019643179985175525,
+      "loss": 1.0427,
       "step": 1205
     },
     {
-      "epoch": 0.18500114670132253,
-      "grad_norm": 0.07538379469221401,
-      "learning_rate": 0.0009781445747792114,
-      "loss": 1.2605,
+      "epoch": 0.7400045868052901,
+      "grad_norm": 0.018162781703447546,
+      "learning_rate": 0.00019220666245176016,
+      "loss": 1.0486,
       "step": 1210
     },
     {
-      "epoch": 0.1857656142496751,
-      "grad_norm": 0.07311552086280827,
-      "learning_rate": 0.0009777526784125105,
-      "loss": 1.2735,
+      "epoch": 0.7430624569987004,
+      "grad_norm": 0.019847164271580075,
+      "learning_rate": 0.00018801662201143815,
+      "loss": 1.0353,
       "step": 1215
     },
     {
-      "epoch": 0.18653008179802769,
-      "grad_norm": 0.06571777631977245,
-      "learning_rate": 0.0009773573795146485,
-      "loss": 1.2642,
+      "epoch": 0.7461203271921107,
+      "grad_norm": 0.020732633406869302,
+      "learning_rate": 0.00018386215631137248,
+      "loss": 1.064,
       "step": 1220
     },
     {
-      "epoch": 0.18729454934638023,
-      "grad_norm": 0.23163418700965988,
-      "learning_rate": 0.0009769586809009244,
-      "loss": 1.2063,
+      "epoch": 0.7491781973855209,
+      "grad_norm": 0.020348214250739332,
+      "learning_rate": 0.00017974373907564167,
+      "loss": 1.052,
       "step": 1225
     },
     {
-      "epoch": 0.1880590168947328,
-      "grad_norm": 0.07247911397915323,
-      "learning_rate": 0.0009765565854108503,
-      "loss": 1.2543,
+      "epoch": 0.7522360675789312,
+      "grad_norm": 0.019382936149259185,
+      "learning_rate": 0.000175661839917801,
+      "loss": 1.0632,
       "step": 1230
     },
     {
-      "epoch": 0.1888234844430854,
-      "grad_norm": 0.07853414468894612,
-      "learning_rate": 0.0009761510959081305,
-      "loss": 1.2966,
+      "epoch": 0.7552939377723416,
+      "grad_norm": 0.020617077209785874,
+      "learning_rate": 0.0001716169242873346,
+      "loss": 1.0603,
       "step": 1235
     },
     {
-      "epoch": 0.18958795199143796,
-      "grad_norm": 0.07106732796502367,
-      "learning_rate": 0.0009757422152806414,
-      "loss": 1.2454,
+      "epoch": 0.7583518079657519,
+      "grad_norm": 0.018721534354599865,
+      "learning_rate": 0.0001676094534165807,
+      "loss": 1.0504,
       "step": 1240
     },
     {
-      "epoch": 0.19035241953979054,
-      "grad_norm": 0.06666091797081679,
-      "learning_rate": 0.0009753299464404109,
-      "loss": 1.252,
+      "epoch": 0.7614096781591622,
+      "grad_norm": 0.019523624424920942,
+      "learning_rate": 0.00016363988426813865,
+      "loss": 1.0605,
       "step": 1245
     },
     {
-      "epoch": 0.19111688708814312,
-      "grad_norm": 0.0809358295124488,
-      "learning_rate": 0.0009749142923235974,
-      "loss": 1.2823,
+      "epoch": 0.7644675483525725,
+      "grad_norm": 0.021325379023544762,
+      "learning_rate": 0.00015970866948276208,
+      "loss": 1.0502,
       "step": 1250
     },
     {
-      "epoch": 0.1918813546364957,
-      "grad_norm": 0.07398963462471533,
-      "learning_rate": 0.0009744952558904692,
-      "loss": 1.2211,
+      "epoch": 0.7675254185459828,
+      "grad_norm": 0.018381651158354734,
+      "learning_rate": 0.00015581625732774562,
+      "loss": 1.0695,
       "step": 1255
     },
     {
-      "epoch": 0.19264582218484824,
-      "grad_norm": 0.06648581697987127,
-      "learning_rate": 0.000974072840125383,
-      "loss": 1.2452,
+      "epoch": 0.770583288739393,
+      "grad_norm": 0.017773314789403843,
+      "learning_rate": 0.00015196309164581046,
+      "loss": 1.0564,
       "step": 1260
     },
     {
-      "epoch": 0.19341028973320082,
-      "grad_norm": 0.08734480773613341,
-      "learning_rate": 0.0009736470480367634,
-      "loss": 1.2983,
+      "epoch": 0.7736411589328033,
+      "grad_norm": 0.018755748204595284,
+      "learning_rate": 0.00014814961180449339,
+      "loss": 1.0372,
       "step": 1265
     },
     {
-      "epoch": 0.1941747572815534,
-      "grad_norm": 0.0814299487892339,
-      "learning_rate": 0.000973217882657081,
-      "loss": 1.2668,
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.025065839963610673,
+      "learning_rate": 0.0001443762526460468,
+      "loss": 1.058,
       "step": 1270
     },
     {
-      "epoch": 0.19493922482990597,
-      "grad_norm": 0.08233314539840832,
-      "learning_rate": 0.00097278534704283,
-      "loss": 1.2196,
+      "epoch": 0.7797568993196239,
+      "grad_norm": 0.022671037009366964,
+      "learning_rate": 0.00014064344443785504,
+      "loss": 1.0486,
       "step": 1275
     },
     {
-      "epoch": 0.19570369237825855,
-      "grad_norm": 0.06781465946994458,
-      "learning_rate": 0.0009723494442745084,
-      "loss": 1.2194,
+      "epoch": 0.7828147695130342,
+      "grad_norm": 0.021001365810248977,
+      "learning_rate": 0.0001369516128233716,
+      "loss": 1.0684,
       "step": 1280
     },
     {
-      "epoch": 0.19646815992661112,
-      "grad_norm": 0.09257477131832359,
-      "learning_rate": 0.0009719101774565941,
-      "loss": 1.228,
+      "epoch": 0.7858726397064445,
+      "grad_norm": 0.02187887791955955,
+      "learning_rate": 0.00013330117877358383,
+      "loss": 1.0488,
       "step": 1285
     },
     {
-      "epoch": 0.1972326274749637,
-      "grad_norm": 0.07505866072283653,
-      "learning_rate": 0.000971467549717524,
-      "loss": 1.2683,
+      "epoch": 0.7889305098998548,
+      "grad_norm": 0.017361526578702963,
+      "learning_rate": 0.00012969255853901142,
+      "loss": 1.0378,
       "step": 1290
     },
     {
-      "epoch": 0.19799709502331625,
-      "grad_norm": 0.06785480940064653,
-      "learning_rate": 0.0009710215642096711,
-      "loss": 1.249,
+      "epoch": 0.791988380093265,
+      "grad_norm": 0.01766771162161045,
+      "learning_rate": 0.00012612616360224138,
+      "loss": 1.0369,
       "step": 1295
     },
     {
-      "epoch": 0.19876156257166883,
-      "grad_norm": 0.06775548552563401,
-      "learning_rate": 0.0009705722241093222,
-      "loss": 1.2575,
+      "epoch": 0.7950462502866753,
+      "grad_norm": 0.021322450785295783,
+      "learning_rate": 0.00012260240063100846,
+      "loss": 1.0639,
       "step": 1300
     },
     {
-      "epoch": 0.1995260301200214,
-      "grad_norm": 0.06486741211895572,
-      "learning_rate": 0.0009701195326166555,
-      "loss": 1.2114,
+      "epoch": 0.7981041204800856,
+      "grad_norm": 0.020380850502793722,
+      "learning_rate": 0.00011912167143182279,
+      "loss": 1.0493,
       "step": 1305
     },
     {
-      "epoch": 0.20029049766837398,
-      "grad_norm": 0.07620009922689974,
-      "learning_rate": 0.0009696634929557178,
-      "loss": 1.2423,
+      "epoch": 0.8011619906734959,
+      "grad_norm": 0.018397926034012132,
+      "learning_rate": 0.00011568437290415395,
+      "loss": 1.0541,
       "step": 1310
     },
     {
-      "epoch": 0.20105496521672656,
-      "grad_norm": 0.07533128179658055,
-      "learning_rate": 0.0009692041083744011,
-      "loss": 1.1992,
+      "epoch": 0.8042198608669062,
+      "grad_norm": 0.017133257784555725,
+      "learning_rate": 0.00011229089699517242,
+      "loss": 1.0523,
       "step": 1315
     },
     {
-      "epoch": 0.20181943276507913,
-      "grad_norm": 0.08244330003493801,
-      "learning_rate": 0.0009687413821444199,
-      "loss": 1.2613,
+      "epoch": 0.8072777310603165,
+      "grad_norm": 0.01819525737040671,
+      "learning_rate": 0.00010894163065505753,
+      "loss": 1.0468,
       "step": 1320
     },
     {
-      "epoch": 0.2025839003134317,
-      "grad_norm": 0.06717152873434033,
-      "learning_rate": 0.0009682753175612876,
-      "loss": 1.2424,
+      "epoch": 0.8103356012537268,
+      "grad_norm": 0.02028545883645919,
+      "learning_rate": 0.00010563695579287402,
+      "loss": 1.0629,
       "step": 1325
     },
     {
-      "epoch": 0.20334836786178426,
-      "grad_norm": 0.10045766887445413,
-      "learning_rate": 0.0009678059179442935,
-      "loss": 1.2244,
+      "epoch": 0.813393471447137,
+      "grad_norm": 0.01787173264557234,
+      "learning_rate": 0.00010237724923302406,
+      "loss": 1.0351,
       "step": 1330
     },
     {
-      "epoch": 0.20411283541013683,
-      "grad_norm": 0.06751762043917765,
-      "learning_rate": 0.0009673331866364786,
-      "loss": 1.2375,
+      "epoch": 0.8164513416405473,
+      "grad_norm": 0.018194742191756992,
+      "learning_rate": 9.916288267227824e-05,
+      "loss": 1.0598,
       "step": 1335
     },
     {
-      "epoch": 0.2048773029584894,
-      "grad_norm": 0.07281491279798423,
-      "learning_rate": 0.0009668571270046121,
-      "loss": 1.2563,
+      "epoch": 0.8195092118339576,
+      "grad_norm": 0.019378519353161313,
+      "learning_rate": 9.599422263739289e-05,
+      "loss": 1.0608,
       "step": 1340
     },
     {
-      "epoch": 0.205641770506842,
-      "grad_norm": 0.07449446576944249,
-      "learning_rate": 0.0009663777424391675,
-      "loss": 1.2537,
+      "epoch": 0.822567082027368,
+      "grad_norm": 0.01902789612845712,
+      "learning_rate": 9.287163044331498e-05,
+      "loss": 1.0501,
       "step": 1345
     },
     {
-      "epoch": 0.20640623805519456,
-      "grad_norm": 0.06486106585475247,
-      "learning_rate": 0.0009658950363542979,
-      "loss": 1.2626,
+      "epoch": 0.8256249522207783,
+      "grad_norm": 0.018581047910824787,
+      "learning_rate": 8.979546215198265e-05,
+      "loss": 1.0458,
       "step": 1350
     },
     {
-      "epoch": 0.20717070560354714,
-      "grad_norm": 0.068273879782427,
-      "learning_rate": 0.0009654090121878125,
-      "loss": 1.2601,
+      "epoch": 0.8286828224141886,
+      "grad_norm": 0.01799606775719105,
+      "learning_rate": 8.676606853172426e-05,
+      "loss": 1.0403,
       "step": 1355
     },
     {
-      "epoch": 0.2079351731518997,
-      "grad_norm": 0.07549308029701757,
-      "learning_rate": 0.0009649196734011519,
-      "loss": 1.2414,
+      "epoch": 0.8317406926075988,
+      "grad_norm": 0.01874997273444636,
+      "learning_rate": 8.378379501726096e-05,
+      "loss": 1.0672,
       "step": 1360
     },
     {
-      "epoch": 0.20869964070025226,
-      "grad_norm": 0.07060653649003629,
-      "learning_rate": 0.0009644270234793623,
-      "loss": 1.2503,
+      "epoch": 0.8347985628010091,
+      "grad_norm": 0.01729209570770116,
+      "learning_rate": 8.084898167031746e-05,
+      "loss": 1.0519,
       "step": 1365
     },
     {
-      "epoch": 0.20946410824860484,
-      "grad_norm": 0.5778206445656846,
-      "learning_rate": 0.0009639310659310729,
-      "loss": 1.2925,
+      "epoch": 0.8378564329944194,
+      "grad_norm": 0.020489335055326883,
+      "learning_rate": 7.796196314084614e-05,
+      "loss": 1.0332,
       "step": 1370
     },
     {
-      "epoch": 0.21022857579695742,
-      "grad_norm": 0.086771304690071,
-      "learning_rate": 0.0009634318042884689,
-      "loss": 1.286,
+      "epoch": 0.8409143031878297,
+      "grad_norm": 0.018824889864362003,
+      "learning_rate": 7.512306862886681e-05,
+      "loss": 1.0349,
       "step": 1375
     },
     {
-      "epoch": 0.21099304334531,
-      "grad_norm": 0.07780731029224541,
-      "learning_rate": 0.000962929242107267,
-      "loss": 1.2886,
+      "epoch": 0.84397217338124,
+      "grad_norm": 0.018493904385427234,
+      "learning_rate": 7.233262184692974e-05,
+      "loss": 1.0458,
       "step": 1380
     },
     {
-      "epoch": 0.21175751089366257,
-      "grad_norm": 0.3625052155879589,
-      "learning_rate": 0.0009624233829666907,
-      "loss": 1.208,
+      "epoch": 0.8470300435746503,
+      "grad_norm": 0.021431180850593753,
+      "learning_rate": 6.959094098320295e-05,
+      "loss": 1.042,
       "step": 1385
     },
     {
-      "epoch": 0.21252197844201515,
-      "grad_norm": 0.06645909024594576,
-      "learning_rate": 0.0009619142304694439,
-      "loss": 1.2426,
+      "epoch": 0.8500879137680606,
+      "grad_norm": 0.01774287881523113,
+      "learning_rate": 6.68983386651898e-05,
+      "loss": 1.0463,
       "step": 1390
     },
     {
-      "epoch": 0.2132864459903677,
-      "grad_norm": 0.10577040811890297,
-      "learning_rate": 0.0009614017882416859,
-      "loss": 1.2787,
+      "epoch": 0.8531457839614708,
+      "grad_norm": 0.019191338242596227,
+      "learning_rate": 6.425512192408112e-05,
+      "loss": 1.0428,
       "step": 1395
     },
     {
-      "epoch": 0.21405091353872027,
-      "grad_norm": 0.09769123070834913,
-      "learning_rate": 0.0009608860599330048,
-      "loss": 1.2174,
+      "epoch": 0.8562036541548811,
+      "grad_norm": 0.01894785926804795,
+      "learning_rate": 6.166159215974521e-05,
+      "loss": 1.0435,
       "step": 1400
     },
     {
-      "epoch": 0.21481538108707285,
-      "grad_norm": 0.08553593141915698,
-      "learning_rate": 0.0009603670492163926,
-      "loss": 1.2382,
+      "epoch": 0.8592615243482914,
+      "grad_norm": 0.016825213446883102,
+      "learning_rate": 5.911804510635965e-05,
+      "loss": 1.0447,
       "step": 1405
     },
     {
-      "epoch": 0.21557984863542543,
-      "grad_norm": 0.08406132520975527,
-      "learning_rate": 0.0009598447597882181,
-      "loss": 1.2541,
+      "epoch": 0.8623193945417017,
+      "grad_norm": 0.016963551769271477,
+      "learning_rate": 5.6624770798689665e-05,
+      "loss": 1.0412,
       "step": 1410
     },
     {
-      "epoch": 0.216344316183778,
-      "grad_norm": 0.09071815773985022,
-      "learning_rate": 0.000959319195368201,
-      "loss": 1.2277,
+      "epoch": 0.865377264735112,
+      "grad_norm": 0.016723668733144115,
+      "learning_rate": 5.418205353901562e-05,
+      "loss": 1.0684,
       "step": 1415
     },
     {
-      "epoch": 0.21710878373213058,
-      "grad_norm": 0.073298406707831,
-      "learning_rate": 0.0009587903596993855,
-      "loss": 1.2373,
+      "epoch": 0.8684351349285223,
+      "grad_norm": 0.019698338107007437,
+      "learning_rate": 5.179017186471557e-05,
+      "loss": 1.0379,
       "step": 1420
     },
     {
-      "epoch": 0.21787325128048315,
-      "grad_norm": 0.08328106823130824,
-      "learning_rate": 0.000958258256548113,
-      "loss": 1.2093,
+      "epoch": 0.8714930051219326,
+      "grad_norm": 0.017265614379559306,
+      "learning_rate": 4.94493985165036e-05,
+      "loss": 1.0454,
       "step": 1425
     },
     {
-      "epoch": 0.2186377188288357,
-      "grad_norm": 0.076889746663869,
-      "learning_rate": 0.0009577228897039963,
-      "loss": 1.2324,
+      "epoch": 0.8745508753153428,
+      "grad_norm": 0.01778176463464062,
+      "learning_rate": 4.7160000407329765e-05,
+      "loss": 1.0431,
       "step": 1430
     },
     {
-      "epoch": 0.21940218637718828,
-      "grad_norm": 0.10417983319043374,
-      "learning_rate": 0.0009571842629798917,
-      "loss": 1.2314,
+      "epoch": 0.8776087455087531,
+      "grad_norm": 0.019057294957546256,
+      "learning_rate": 4.492223859194533e-05,
+      "loss": 1.0442,
       "step": 1435
     },
     {
-      "epoch": 0.22016665392554086,
-      "grad_norm": 0.07670976249694544,
-      "learning_rate": 0.0009566423802118725,
-      "loss": 1.2418,
+      "epoch": 0.8806666157021634,
+      "grad_norm": 0.018453421610507247,
+      "learning_rate": 4.273636823713467e-05,
+      "loss": 1.0437,
       "step": 1440
     },
     {
-      "epoch": 0.22093112147389343,
-      "grad_norm": 0.06138227389521626,
-      "learning_rate": 0.0009560972452592009,
-      "loss": 1.2728,
+      "epoch": 0.8837244858955737,
+      "grad_norm": 0.01765155626929793,
+      "learning_rate": 4.0602638592619257e-05,
+      "loss": 1.0415,
       "step": 1445
     },
     {
-      "epoch": 0.221695589022246,
-      "grad_norm": 0.06168948726191564,
-      "learning_rate": 0.0009555488620043018,
-      "loss": 1.2739,
+      "epoch": 0.886782356088984,
+      "grad_norm": 0.017862026611963264,
+      "learning_rate": 3.8521292962636625e-05,
+      "loss": 1.0436,
       "step": 1450
     },
     {
-      "epoch": 0.22246005657059859,
-      "grad_norm": 0.07139543551614756,
-      "learning_rate": 0.0009549972343527336,
-      "loss": 1.2635,
+      "epoch": 0.8898402262823943,
+      "grad_norm": 0.017606797127937533,
+      "learning_rate": 3.649256867819667e-05,
+      "loss": 1.04,
       "step": 1455
     },
     {
-      "epoch": 0.22322452411895116,
-      "grad_norm": 0.08125497986153993,
-      "learning_rate": 0.0009544423662331612,
-      "loss": 1.2419,
+      "epoch": 0.8928980964758046,
+      "grad_norm": 0.017151387009890048,
+      "learning_rate": 3.451669707001942e-05,
+      "loss": 1.0578,
       "step": 1460
     },
     {
-      "epoch": 0.2239889916673037,
-      "grad_norm": 0.06174524531157182,
-      "learning_rate": 0.0009538842615973284,
-      "loss": 1.2286,
+      "epoch": 0.8959559666692148,
+      "grad_norm": 0.017195927947563305,
+      "learning_rate": 3.259390344215707e-05,
+      "loss": 1.0462,
       "step": 1465
     },
     {
-      "epoch": 0.2247534592156563,
-      "grad_norm": 0.08712335845180429,
-      "learning_rate": 0.0009533229244200292,
-      "loss": 1.2095,
+      "epoch": 0.8990138368626251,
+      "grad_norm": 0.01845679676674692,
+      "learning_rate": 3.072440704630258e-05,
+      "loss": 1.0555,
       "step": 1470
     },
     {
-      "epoch": 0.22551792676400886,
-      "grad_norm": 0.07266991262135022,
-      "learning_rate": 0.0009527583586990791,
-      "loss": 1.2463,
+      "epoch": 0.9020717070560355,
+      "grad_norm": 0.016944392213100467,
+      "learning_rate": 2.8908421056789703e-05,
+      "loss": 1.033,
       "step": 1475
     },
     {
-      "epoch": 0.22628239431236144,
-      "grad_norm": 0.07327264508555179,
-      "learning_rate": 0.0009521905684552877,
-      "loss": 1.2243,
+      "epoch": 0.9051295772494458,
+      "grad_norm": 0.018432906406152354,
+      "learning_rate": 2.7146152546284663e-05,
+      "loss": 1.0508,
       "step": 1480
     },
     {
-      "epoch": 0.22704686186071402,
-      "grad_norm": 0.09100386022395632,
-      "learning_rate": 0.0009516195577324294,
-      "loss": 1.2602,
+      "epoch": 0.9081874474428561,
+      "grad_norm": 0.017706430050248204,
+      "learning_rate": 2.5437802462174144e-05,
+      "loss": 1.0407,
       "step": 1485
     },
     {
-      "epoch": 0.2278113294090666,
-      "grad_norm": 0.06911064137719194,
-      "learning_rate": 0.0009510453305972142,
-      "loss": 1.2512,
+      "epoch": 0.9112453176362664,
+      "grad_norm": 0.01825937227605355,
+      "learning_rate": 2.3783565603652024e-05,
+      "loss": 1.0191,
       "step": 1490
     },
     {
-      "epoch": 0.22857579695741917,
-      "grad_norm": 0.10444912049605627,
-      "learning_rate": 0.0009504678911392601,
-      "loss": 1.2157,
+      "epoch": 0.9143031878296767,
+      "grad_norm": 0.017170141434300536,
+      "learning_rate": 2.218363059950651e-05,
+      "loss": 1.0381,
       "step": 1495
     },
     {
-      "epoch": 0.22934026450577172,
-      "grad_norm": 0.09169840988364335,
-      "learning_rate": 0.0009498872434710623,
-      "loss": 1.2266,
+      "epoch": 0.9173610580230869,
+      "grad_norm": 0.021018198836607237,
+      "learning_rate": 2.0638179886611542e-05,
+      "loss": 1.0342,
       "step": 1500
     },
     {
-      "epoch": 0.2301047320541243,
-      "grad_norm": 0.06835850906504734,
-      "learning_rate": 0.000949303391727965,
-      "loss": 1.1967,
+      "epoch": 0.9204189282164972,
+      "grad_norm": 0.018670756565939955,
+      "learning_rate": 1.9147389689123884e-05,
+      "loss": 1.038,
       "step": 1505
     },
     {
-      "epoch": 0.23086919960247687,
-      "grad_norm": 0.08072558468127804,
-      "learning_rate": 0.000948716340068132,
-      "loss": 1.2446,
+      "epoch": 0.9234767984099075,
+      "grad_norm": 0.01840746015098489,
+      "learning_rate": 1.771142999838854e-05,
+      "loss": 1.0576,
       "step": 1510
     },
     {
-      "epoch": 0.23163366715082945,
-      "grad_norm": 0.09703619217717847,
-      "learning_rate": 0.0009481260926725161,
-      "loss": 1.2335,
+      "epoch": 0.9265346686033178,
+      "grad_norm": 0.01791437481490544,
+      "learning_rate": 1.633046455355519e-05,
+      "loss": 1.0547,
       "step": 1515
     },
     {
-      "epoch": 0.23239813469918202,
-      "grad_norm": 0.07534829838458987,
-      "learning_rate": 0.0009475326537448307,
-      "loss": 1.2297,
+      "epoch": 0.9295925387967281,
+      "grad_norm": 0.016561566008186503,
+      "learning_rate": 1.5004650822907306e-05,
+      "loss": 1.0429,
       "step": 1520
     },
     {
-      "epoch": 0.2331626022475346,
-      "grad_norm": 0.08322927425588308,
-      "learning_rate": 0.0009469360275115185,
-      "loss": 1.2265,
+      "epoch": 0.9326504089901384,
+      "grad_norm": 0.017211669368685612,
+      "learning_rate": 1.373413998590639e-05,
+      "loss": 1.0412,
       "step": 1525
     },
     {
-      "epoch": 0.23392706979588718,
-      "grad_norm": 0.06820300897660575,
-      "learning_rate": 0.0009463362182217223,
-      "loss": 1.2338,
+      "epoch": 0.9357082791835487,
+      "grad_norm": 0.01917485719407648,
+      "learning_rate": 1.2519076915953443e-05,
+      "loss": 1.0329,
       "step": 1530
     },
     {
-      "epoch": 0.23469153734423973,
-      "grad_norm": 0.07576589358564349,
-      "learning_rate": 0.0009457332301472545,
-      "loss": 1.204,
+      "epoch": 0.9387661493769589,
+      "grad_norm": 0.019235379531641828,
+      "learning_rate": 1.1359600163869276e-05,
+      "loss": 1.0509,
       "step": 1535
     },
     {
-      "epoch": 0.2354560048925923,
-      "grad_norm": 0.07033059584476081,
-      "learning_rate": 0.0009451270675825665,
-      "loss": 1.2368,
+      "epoch": 0.9418240195703692,
+      "grad_norm": 0.017132686744869362,
+      "learning_rate": 1.0255841942096056e-05,
+      "loss": 1.0402,
       "step": 1540
     },
     {
-      "epoch": 0.23622047244094488,
-      "grad_norm": 0.07998529784318079,
-      "learning_rate": 0.0009445177348447187,
-      "loss": 1.1867,
+      "epoch": 0.9448818897637795,
+      "grad_norm": 0.018593609192373937,
+      "learning_rate": 9.207928109621122e-06,
+      "loss": 1.0409,
       "step": 1545
     },
     {
-      "epoch": 0.23698493998929746,
-      "grad_norm": 0.06330372086398396,
-      "learning_rate": 0.0009439052362733486,
-      "loss": 1.2153,
+      "epoch": 0.9479397599571898,
+      "grad_norm": 0.01776794448761931,
+      "learning_rate": 8.21597815762587e-06,
+      "loss": 1.0242,
       "step": 1550
     },
-    {
-      "epoch": 0.23774940753765003,
-      "grad_norm": 0.0825543412216852,
-      "learning_rate": 0.0009432895762306415,
-      "loss": 1.2114,
-      "step": 1555
-    },
-    {
-      "epoch": 0.2385138750860026,
-      "grad_norm": 0.07266151001303742,
-      "learning_rate": 0.0009426707591012975,
-      "loss": 1.2227,
-      "step": 1560
-    },
-    {
-      "epoch": 0.23927834263435518,
-      "grad_norm": 0.08994379252698968,
-      "learning_rate": 0.0009420487892925023,
-      "loss": 1.2153,
-      "step": 1565
-    },
-    {
-      "epoch": 0.24004281018270773,
-      "grad_norm": 0.06997215800326952,
-      "learning_rate": 0.0009414236712338941,
-      "loss": 1.2456,
-      "step": 1570
-    },
-    {
-      "epoch": 0.2408072777310603,
-      "grad_norm": 0.07195887577227815,
-      "learning_rate": 0.0009407954093775333,
-      "loss": 1.2606,
-      "step": 1575
-    },
-    {
-      "epoch": 0.24157174527941289,
-      "grad_norm": 0.0669981044756401,
-      "learning_rate": 0.00094016400819787,
-      "loss": 1.2293,
-      "step": 1580
-    },
-    {
-      "epoch": 0.24233621282776546,
-      "grad_norm": 0.08395936334147859,
-      "learning_rate": 0.0009395294721917122,
-      "loss": 1.1822,
-      "step": 1585
-    },
-    {
-      "epoch": 0.24310068037611804,
-      "grad_norm": 0.08734808358295865,
-      "learning_rate": 0.0009388918058781945,
-      "loss": 1.2026,
-      "step": 1590
-    },
-    {
-      "epoch": 0.24386514792447062,
-      "grad_norm": 0.06832693115838025,
-      "learning_rate": 0.0009382510137987452,
-      "loss": 1.2111,
-      "step": 1595
-    },
-    {
-      "epoch": 0.2446296154728232,
-      "grad_norm": 0.0706613941441332,
-      "learning_rate": 0.0009376071005170539,
-      "loss": 1.1989,
-      "step": 1600
-    },
-    {
-      "epoch": 0.24539408302117574,
-      "grad_norm": 0.06738770199309628,
-      "learning_rate": 0.0009369600706190395,
-      "loss": 1.2405,
-      "step": 1605
-    },
-    {
-      "epoch": 0.24615855056952832,
-      "grad_norm": 0.0815414234541493,
-      "learning_rate": 0.0009363099287128173,
-      "loss": 1.2008,
-      "step": 1610
-    },
-    {
-      "epoch": 0.2469230181178809,
-      "grad_norm": 0.07295828491472306,
-      "learning_rate": 0.000935656679428666,
-      "loss": 1.2355,
-      "step": 1615
-    },
-    {
-      "epoch": 0.24768748566623347,
-      "grad_norm": 0.06339989075543628,
-      "learning_rate": 0.0009350003274189949,
-      "loss": 1.1979,
-      "step": 1620
-    },
-    {
-      "epoch": 0.24845195321458605,
-      "grad_norm": 0.06230503242046638,
-      "learning_rate": 0.000934340877358311,
-      "loss": 1.2229,
-      "step": 1625
-    },
-    {
-      "epoch": 0.24921642076293862,
-      "grad_norm": 0.07324576424114503,
-      "learning_rate": 0.0009336783339431851,
-      "loss": 1.2468,
-      "step": 1630
-    },
-    {
-      "epoch": 0.2499808883112912,
-      "grad_norm": 0.06752028292519564,
-      "learning_rate": 0.0009330127018922195,
-      "loss": 1.2213,
-      "step": 1635
-    },
-    {
-      "epoch": 0.25074535585964375,
-      "grad_norm": 0.07188634308822926,
-      "learning_rate": 0.0009323439859460122,
-      "loss": 1.2177,
-      "step": 1640
-    },
-    {
-      "epoch": 0.2515098234079963,
-      "grad_norm": 0.06474900340430122,
-      "learning_rate": 0.000931672190867126,
-      "loss": 1.2038,
-      "step": 1645
-    },
-    {
-      "epoch": 0.2522742909563489,
-      "grad_norm": 0.06803113030721539,
-      "learning_rate": 0.0009309973214400524,
-      "loss": 1.2429,
-      "step": 1650
-    },
-    {
-      "epoch": 0.2530387585047015,
-      "grad_norm": 0.0696030870632886,
-      "learning_rate": 0.0009303193824711783,
-      "loss": 1.1881,
-      "step": 1655
-    },
-    {
-      "epoch": 0.25380322605305405,
-      "grad_norm": 0.06427278412655327,
-      "learning_rate": 0.0009296383787887519,
-      "loss": 1.1999,
-      "step": 1660
-    },
-    {
-      "epoch": 0.25456769360140663,
-      "grad_norm": 0.06151334873890928,
-      "learning_rate": 0.000928954315242848,
-      "loss": 1.2104,
-      "step": 1665
-    },
-    {
-      "epoch": 0.2553321611497592,
-      "grad_norm": 0.07147857753540077,
-      "learning_rate": 0.0009282671967053335,
-      "loss": 1.1903,
-      "step": 1670
-    },
-    {
-      "epoch": 0.2560966286981118,
-      "grad_norm": 0.06167324666150717,
-      "learning_rate": 0.0009275770280698332,
-      "loss": 1.2496,
-      "step": 1675
-    },
-    {
-      "epoch": 0.25686109624646436,
-      "grad_norm": 0.06467673259045754,
-      "learning_rate": 0.0009268838142516944,
-      "loss": 1.2307,
-      "step": 1680
-    },
-    {
-      "epoch": 0.25762556379481694,
-      "grad_norm": 0.06363717052606994,
-      "learning_rate": 0.0009261875601879514,
-      "loss": 1.2183,
-      "step": 1685
-    },
-    {
-      "epoch": 0.25839003134316946,
-      "grad_norm": 0.08631057056855553,
-      "learning_rate": 0.0009254882708372922,
-      "loss": 1.2058,
-      "step": 1690
-    },
-    {
-      "epoch": 0.25915449889152203,
-      "grad_norm": 0.059947021170670405,
-      "learning_rate": 0.0009247859511800207,
-      "loss": 1.2075,
-      "step": 1695
-    },
-    {
-      "epoch": 0.2599189664398746,
-      "grad_norm": 0.06378174138594353,
-      "learning_rate": 0.0009240806062180233,
-      "loss": 1.2141,
-      "step": 1700
-    },
-    {
-      "epoch": 0.2606834339882272,
-      "grad_norm": 0.06373001825884964,
-      "learning_rate": 0.0009233722409747325,
-      "loss": 1.1995,
-      "step": 1705
-    },
-    {
-      "epoch": 0.26144790153657976,
-      "grad_norm": 0.07986652766185562,
-      "learning_rate": 0.0009226608604950905,
-      "loss": 1.2691,
-      "step": 1710
-    },
-    {
-      "epoch": 0.26221236908493234,
-      "grad_norm": 0.06532339660591398,
-      "learning_rate": 0.0009219464698455143,
-      "loss": 1.2049,
-      "step": 1715
-    },
-    {
-      "epoch": 0.2629768366332849,
-      "grad_norm": 0.07511908390592772,
-      "learning_rate": 0.0009212290741138591,
-      "loss": 1.2109,
-      "step": 1720
-    },
-    {
-      "epoch": 0.2637413041816375,
-      "grad_norm": 0.06805686624245322,
-      "learning_rate": 0.0009205086784093823,
-      "loss": 1.1938,
-      "step": 1725
-    },
-    {
-      "epoch": 0.26450577172999007,
-      "grad_norm": 0.07608315580128151,
-      "learning_rate": 0.0009197852878627064,
-      "loss": 1.2247,
-      "step": 1730
-    },
-    {
-      "epoch": 0.26527023927834265,
-      "grad_norm": 0.07460173165874026,
-      "learning_rate": 0.000919058907625784,
-      "loss": 1.1986,
-      "step": 1735
-    },
-    {
-      "epoch": 0.2660347068266952,
-      "grad_norm": 0.0653540774845259,
-      "learning_rate": 0.0009183295428718591,
-      "loss": 1.2015,
-      "step": 1740
-    },
-    {
-      "epoch": 0.2667991743750478,
-      "grad_norm": 0.06088986122123603,
-      "learning_rate": 0.0009175971987954319,
-      "loss": 1.2347,
-      "step": 1745
-    },
-    {
-      "epoch": 0.2675636419234004,
-      "grad_norm": 0.06182727669114958,
-      "learning_rate": 0.0009168618806122209,
-      "loss": 1.2168,
-      "step": 1750
-    },
-    {
-      "epoch": 0.26832810947175295,
-      "grad_norm": 0.06945925182347323,
-      "learning_rate": 0.0009161235935591264,
-      "loss": 1.2107,
-      "step": 1755
-    },
-    {
-      "epoch": 0.2690925770201055,
-      "grad_norm": 0.07350549175394744,
-      "learning_rate": 0.0009153823428941923,
-      "loss": 1.2252,
-      "step": 1760
-    },
-    {
-      "epoch": 0.26985704456845805,
-      "grad_norm": 0.06552770776828734,
-      "learning_rate": 0.0009146381338965698,
-      "loss": 1.2268,
-      "step": 1765
-    },
-    {
-      "epoch": 0.2706215121168106,
-      "grad_norm": 0.0866708613189302,
-      "learning_rate": 0.0009138909718664788,
-      "loss": 1.1928,
-      "step": 1770
-    },
-    {
-      "epoch": 0.2713859796651632,
-      "grad_norm": 0.07542325154916106,
-      "learning_rate": 0.0009131408621251704,
-      "loss": 1.1992,
-      "step": 1775
-    },
-    {
-      "epoch": 0.2721504472135158,
-      "grad_norm": 0.08169121697992505,
-      "learning_rate": 0.0009123878100148899,
-      "loss": 1.2173,
-      "step": 1780
-    },
-    {
-      "epoch": 0.27291491476186835,
-      "grad_norm": 0.07641997824597095,
-      "learning_rate": 0.0009116318208988372,
-      "loss": 1.2592,
-      "step": 1785
-    },
-    {
-      "epoch": 0.27367938231022093,
-      "grad_norm": 0.08606275877785033,
-      "learning_rate": 0.0009108729001611297,
-      "loss": 1.1776,
-      "step": 1790
-    },
-    {
-      "epoch": 0.2744438498585735,
-      "grad_norm": 0.07432766120012625,
-      "learning_rate": 0.000910111053206764,
-      "loss": 1.2192,
-      "step": 1795
-    },
-    {
-      "epoch": 0.2752083174069261,
-      "grad_norm": 0.08764002357094997,
-      "learning_rate": 0.0009093462854615766,
-      "loss": 1.28,
-      "step": 1800
-    },
-    {
-      "epoch": 0.27597278495527866,
-      "grad_norm": 0.06475582520577117,
-      "learning_rate": 0.0009085786023722058,
-      "loss": 1.2357,
-      "step": 1805
-    },
-    {
-      "epoch": 0.27673725250363124,
-      "grad_norm": 0.07215552635866873,
-      "learning_rate": 0.0009078080094060529,
-      "loss": 1.3359,
-      "step": 1810
-    },
-    {
-      "epoch": 0.2775017200519838,
-      "grad_norm": 0.07686482863798352,
-      "learning_rate": 0.0009070345120512436,
-      "loss": 1.202,
-      "step": 1815
-    },
-    {
-      "epoch": 0.2782661876003364,
-      "grad_norm": 0.07764075676679777,
-      "learning_rate": 0.0009062581158165877,
-      "loss": 1.2308,
-      "step": 1820
-    },
-    {
-      "epoch": 0.2790306551486889,
-      "grad_norm": 0.06490582289798909,
-      "learning_rate": 0.0009054788262315414,
-      "loss": 1.1967,
-      "step": 1825
-    },
-    {
-      "epoch": 0.2797951226970415,
-      "grad_norm": 0.064554320995236,
-      "learning_rate": 0.000904696648846167,
-      "loss": 1.2148,
-      "step": 1830
-    },
-    {
-      "epoch": 0.28055959024539406,
-      "grad_norm": 0.09976478815425889,
-      "learning_rate": 0.0009039115892310931,
-      "loss": 1.2129,
-      "step": 1835
-    },
-    {
-      "epoch": 0.28132405779374664,
-      "grad_norm": 0.08662063449022793,
-      "learning_rate": 0.0009031236529774765,
-      "loss": 1.1943,
-      "step": 1840
-    },
-    {
-      "epoch": 0.2820885253420992,
-      "grad_norm": 0.06867780089471395,
-      "learning_rate": 0.0009023328456969598,
-      "loss": 1.1957,
-      "step": 1845
-    },
-    {
-      "epoch": 0.2828529928904518,
-      "grad_norm": 0.08054420978971974,
-      "learning_rate": 0.0009015391730216343,
-      "loss": 1.1775,
-      "step": 1850
-    },
-    {
-      "epoch": 0.28361746043880437,
-      "grad_norm": 0.06748059150550935,
-      "learning_rate": 0.0009007426406039975,
-      "loss": 1.197,
-      "step": 1855
-    },
-    {
-      "epoch": 0.28438192798715695,
-      "grad_norm": 0.061636919853196595,
-      "learning_rate": 0.0008999432541169144,
-      "loss": 1.1895,
-      "step": 1860
-    },
-    {
-      "epoch": 0.2851463955355095,
-      "grad_norm": 0.06329361104758795,
-      "learning_rate": 0.0008991410192535765,
-      "loss": 1.2115,
-      "step": 1865
-    },
-    {
-      "epoch": 0.2859108630838621,
-      "grad_norm": 0.06647515121483408,
-      "learning_rate": 0.000898335941727461,
-      "loss": 1.17,
-      "step": 1870
-    },
-    {
-      "epoch": 0.2866753306322147,
-      "grad_norm": 0.08264733362665477,
-      "learning_rate": 0.0008975280272722906,
-      "loss": 1.2323,
-      "step": 1875
-    },
-    {
-      "epoch": 0.28743979818056725,
-      "grad_norm": 0.06871644318138265,
-      "learning_rate": 0.0008967172816419926,
-      "loss": 1.2038,
-      "step": 1880
-    },
-    {
-      "epoch": 0.28820426572891983,
-      "grad_norm": 0.06608040766728891,
-      "learning_rate": 0.0008959037106106575,
-      "loss": 1.2133,
-      "step": 1885
-    },
-    {
-      "epoch": 0.2889687332772724,
-      "grad_norm": 0.0843249746514495,
-      "learning_rate": 0.0008950873199724985,
-      "loss": 1.2476,
-      "step": 1890
-    },
-    {
-      "epoch": 0.2897332008256249,
-      "grad_norm": 0.06391405872287036,
-      "learning_rate": 0.0008942681155418093,
-      "loss": 1.2173,
-      "step": 1895
-    },
-    {
-      "epoch": 0.2904976683739775,
-      "grad_norm": 0.07858223937418263,
-      "learning_rate": 0.0008934461031529241,
-      "loss": 1.2367,
-      "step": 1900
-    },
-    {
-      "epoch": 0.2912621359223301,
-      "grad_norm": 0.09167454894661381,
-      "learning_rate": 0.0008926212886601749,
-      "loss": 1.2095,
-      "step": 1905
-    },
-    {
-      "epoch": 0.29202660347068266,
-      "grad_norm": 0.0652987710283515,
-      "learning_rate": 0.0008917936779378497,
-      "loss": 1.1657,
-      "step": 1910
-    },
-    {
-      "epoch": 0.29279107101903523,
-      "grad_norm": 0.08532622110352704,
-      "learning_rate": 0.0008909632768801519,
-      "loss": 1.2182,
-      "step": 1915
-    },
-    {
-      "epoch": 0.2935555385673878,
-      "grad_norm": 0.13889204437606176,
-      "learning_rate": 0.0008901300914011569,
-      "loss": 1.1919,
-      "step": 1920
-    },
-    {
-      "epoch": 0.2943200061157404,
-      "grad_norm": 0.06762914304742695,
-      "learning_rate": 0.0008892941274347707,
-      "loss": 1.2174,
-      "step": 1925
-    },
-    {
-      "epoch": 0.29508447366409296,
-      "grad_norm": 0.07275455107428999,
-      "learning_rate": 0.0008884553909346882,
-      "loss": 1.2023,
-      "step": 1930
-    },
-    {
-      "epoch": 0.29584894121244554,
-      "grad_norm": 0.0671901981946541,
-      "learning_rate": 0.0008876138878743493,
-      "loss": 1.1984,
-      "step": 1935
-    },
-    {
-      "epoch": 0.2966134087607981,
-      "grad_norm": 0.06303430763742675,
-      "learning_rate": 0.0008867696242468976,
-      "loss": 1.2203,
-      "step": 1940
-    },
-    {
-      "epoch": 0.2973778763091507,
-      "grad_norm": 0.06625636101894779,
-      "learning_rate": 0.0008859226060651373,
-      "loss": 1.1984,
-      "step": 1945
-    },
-    {
-      "epoch": 0.29814234385750327,
-      "grad_norm": 0.06254460803911173,
-      "learning_rate": 0.0008850728393614902,
-      "loss": 1.149,
-      "step": 1950
-    },
-    {
-      "epoch": 0.29890681140585584,
-      "grad_norm": 0.06693727583070225,
-      "learning_rate": 0.0008842203301879535,
-      "loss": 1.1963,
-      "step": 1955
-    },
-    {
-      "epoch": 0.2996712789542084,
-      "grad_norm": 0.06974909946676806,
-      "learning_rate": 0.0008833650846160555,
-      "loss": 1.2125,
-      "step": 1960
-    },
-    {
-      "epoch": 0.30043574650256094,
-      "grad_norm": 5.2895162289649615,
-      "learning_rate": 0.0008825071087368131,
-      "loss": 1.2391,
-      "step": 1965
-    },
-    {
-      "epoch": 0.3012002140509135,
-      "grad_norm": 0.09460329161529786,
-      "learning_rate": 0.0008816464086606888,
-      "loss": 1.2074,
-      "step": 1970
-    },
-    {
-      "epoch": 0.3019646815992661,
-      "grad_norm": 0.1507554888140455,
-      "learning_rate": 0.0008807829905175462,
-      "loss": 1.2141,
-      "step": 1975
-    },
-    {
-      "epoch": 0.30272914914761867,
-      "grad_norm": 0.07284165587078718,
-      "learning_rate": 0.000879916860456607,
-      "loss": 1.1938,
-      "step": 1980
-    },
-    {
-      "epoch": 0.30349361669597125,
-      "grad_norm": 0.08317773836093713,
-      "learning_rate": 0.0008790480246464071,
-      "loss": 1.2209,
-      "step": 1985
-    },
-    {
-      "epoch": 0.3042580842443238,
-      "grad_norm": 0.11170869912072523,
-      "learning_rate": 0.0008781764892747525,
-      "loss": 1.1916,
-      "step": 1990
-    },
-    {
-      "epoch": 0.3050225517926764,
-      "grad_norm": 0.06299927829436214,
-      "learning_rate": 0.0008773022605486754,
-      "loss": 1.2109,
-      "step": 1995
-    },
-    {
-      "epoch": 0.305787019341029,
-      "grad_norm": 0.08391531668184286,
-      "learning_rate": 0.0008764253446943899,
-      "loss": 1.1906,
-      "step": 2000
-    },
-    {
-      "epoch": 0.30655148688938155,
-      "grad_norm": 0.0698777375605405,
-      "learning_rate": 0.0008755457479572478,
-      "loss": 1.1927,
-      "step": 2005
-    },
-    {
-      "epoch": 0.30731595443773413,
-      "grad_norm": 0.0764854535360263,
-      "learning_rate": 0.0008746634766016941,
-      "loss": 1.1993,
-      "step": 2010
-    },
-    {
-      "epoch": 0.3080804219860867,
-      "grad_norm": 0.07915862196006107,
-      "learning_rate": 0.0008737785369112219,
-      "loss": 1.2232,
-      "step": 2015
-    },
-    {
-      "epoch": 0.3088448895344393,
-      "grad_norm": 0.06725444808758432,
-      "learning_rate": 0.0008728909351883283,
-      "loss": 1.2007,
-      "step": 2020
-    },
-    {
-      "epoch": 0.30960935708279186,
-      "grad_norm": 0.07910670587364323,
-      "learning_rate": 0.0008720006777544696,
-      "loss": 1.2094,
-      "step": 2025
-    },
-    {
-      "epoch": 0.31037382463114443,
-      "grad_norm": 0.07747525026031941,
-      "learning_rate": 0.0008711077709500153,
-      "loss": 1.2051,
-      "step": 2030
-    },
-    {
-      "epoch": 0.31113829217949696,
-      "grad_norm": 0.06741852376910797,
-      "learning_rate": 0.0008702122211342039,
-      "loss": 1.2331,
-      "step": 2035
-    },
-    {
-      "epoch": 0.31190275972784953,
-      "grad_norm": 0.06866481465327473,
-      "learning_rate": 0.0008693140346850975,
-      "loss": 1.183,
-      "step": 2040
-    },
-    {
-      "epoch": 0.3126672272762021,
-      "grad_norm": 0.0978479410648741,
-      "learning_rate": 0.0008684132179995359,
-      "loss": 1.1953,
-      "step": 2045
-    },
-    {
-      "epoch": 0.3134316948245547,
-      "grad_norm": 0.06892340870931946,
-      "learning_rate": 0.0008675097774930912,
-      "loss": 1.1691,
-      "step": 2050
-    },
-    {
-      "epoch": 0.31419616237290726,
-      "grad_norm": 0.059001834743335244,
-      "learning_rate": 0.0008666037196000227,
-      "loss": 1.1996,
-      "step": 2055
-    },
-    {
-      "epoch": 0.31496062992125984,
-      "grad_norm": 0.06755590313805718,
-      "learning_rate": 0.0008656950507732302,
-      "loss": 1.1778,
-      "step": 2060
-    },
-    {
-      "epoch": 0.3157250974696124,
-      "grad_norm": 0.06507056486580198,
-      "learning_rate": 0.0008647837774842085,
-      "loss": 1.2057,
-      "step": 2065
-    },
-    {
-      "epoch": 0.316489565017965,
-      "grad_norm": 0.06785984666263632,
-      "learning_rate": 0.0008638699062230011,
-      "loss": 1.2116,
-      "step": 2070
-    },
-    {
-      "epoch": 0.31725403256631757,
-      "grad_norm": 0.08655861530553224,
-      "learning_rate": 0.0008629534434981547,
-      "loss": 1.1711,
-      "step": 2075
-    },
-    {
-      "epoch": 0.31801850011467014,
-      "grad_norm": 0.09502942427023862,
-      "learning_rate": 0.0008620343958366718,
-      "loss": 1.2589,
-      "step": 2080
-    },
-    {
-      "epoch": 0.3187829676630227,
-      "grad_norm": 0.07389015223749948,
-      "learning_rate": 0.0008611127697839648,
-      "loss": 1.1662,
-      "step": 2085
-    },
-    {
-      "epoch": 0.3195474352113753,
-      "grad_norm": 0.06284898980071424,
-      "learning_rate": 0.0008601885719038092,
-      "loss": 1.2171,
-      "step": 2090
-    },
-    {
-      "epoch": 0.3203119027597279,
-      "grad_norm": 0.06925613572330289,
-      "learning_rate": 0.0008592618087782971,
-      "loss": 1.1941,
-      "step": 2095
-    },
-    {
-      "epoch": 0.3210763703080804,
-      "grad_norm": 0.06565728652078977,
-      "learning_rate": 0.00085833248700779,
-      "loss": 1.1722,
-      "step": 2100
-    },
-    {
-      "epoch": 0.32184083785643297,
-      "grad_norm": 0.06497325207608327,
-      "learning_rate": 0.0008574006132108721,
-      "loss": 1.2005,
-      "step": 2105
-    },
-    {
-      "epoch": 0.32260530540478555,
-      "grad_norm": 0.0830860825359688,
-      "learning_rate": 0.0008564661940243027,
-      "loss": 1.1849,
-      "step": 2110
-    },
-    {
-      "epoch": 0.3233697729531381,
-      "grad_norm": 0.06766203461784877,
-      "learning_rate": 0.0008555292361029696,
-      "loss": 1.1927,
-      "step": 2115
-    },
-    {
-      "epoch": 0.3241342405014907,
-      "grad_norm": 0.07153735721703837,
-      "learning_rate": 0.0008545897461198413,
-      "loss": 1.2189,
-      "step": 2120
-    },
-    {
-      "epoch": 0.3248987080498433,
-      "grad_norm": 0.07558372140579875,
-      "learning_rate": 0.0008536477307659192,
-      "loss": 1.1919,
-      "step": 2125
-    },
-    {
-      "epoch": 0.32566317559819585,
-      "grad_norm": 0.08212311724687331,
-      "learning_rate": 0.0008527031967501906,
-      "loss": 1.1633,
-      "step": 2130
-    },
-    {
-      "epoch": 0.32642764314654843,
-      "grad_norm": 0.06765162215684241,
-      "learning_rate": 0.0008517561507995805,
-      "loss": 1.2469,
-      "step": 2135
-    },
-    {
-      "epoch": 0.327192110694901,
-      "grad_norm": 0.06925030760862601,
-      "learning_rate": 0.0008508065996589037,
-      "loss": 1.1679,
-      "step": 2140
-    },
-    {
-      "epoch": 0.3279565782432536,
-      "grad_norm": 0.07384137246492231,
-      "learning_rate": 0.0008498545500908168,
-      "loss": 1.1988,
-      "step": 2145
-    },
-    {
-      "epoch": 0.32872104579160616,
-      "grad_norm": 0.06472585249971541,
-      "learning_rate": 0.0008489000088757703,
-      "loss": 1.1971,
-      "step": 2150
-    },
-    {
-      "epoch": 0.32948551333995874,
-      "grad_norm": 0.08266273525218829,
-      "learning_rate": 0.0008479429828119598,
-      "loss": 1.1802,
-      "step": 2155
-    },
-    {
-      "epoch": 0.3302499808883113,
-      "grad_norm": 0.06160840327052949,
-      "learning_rate": 0.0008469834787152783,
-      "loss": 1.2191,
-      "step": 2160
-    },
-    {
-      "epoch": 0.3310144484366639,
-      "grad_norm": 0.06108701849003786,
-      "learning_rate": 0.0008460215034192667,
-      "loss": 1.1599,
-      "step": 2165
-    },
-    {
-      "epoch": 0.3317789159850164,
-      "grad_norm": 0.0675003939511652,
-      "learning_rate": 0.000845057063775066,
-      "loss": 1.2091,
-      "step": 2170
-    },
-    {
-      "epoch": 0.332543383533369,
-      "grad_norm": 0.06491128889515788,
-      "learning_rate": 0.0008440901666513681,
-      "loss": 1.1932,
-      "step": 2175
-    },
-    {
-      "epoch": 0.33330785108172156,
-      "grad_norm": 0.06739173874329502,
-      "learning_rate": 0.0008431208189343669,
-      "loss": 1.2222,
-      "step": 2180
-    },
-    {
-      "epoch": 0.33407231863007414,
-      "grad_norm": 0.06837629340068371,
-      "learning_rate": 0.0008421490275277093,
-      "loss": 1.1827,
-      "step": 2185
-    },
-    {
-      "epoch": 0.3348367861784267,
-      "grad_norm": 0.07301400162009927,
-      "learning_rate": 0.000841174799352446,
-      "loss": 1.1898,
-      "step": 2190
-    },
-    {
-      "epoch": 0.3356012537267793,
-      "grad_norm": 0.07696744621983317,
-      "learning_rate": 0.0008401981413469826,
-      "loss": 1.1807,
-      "step": 2195
-    },
-    {
-      "epoch": 0.33636572127513187,
-      "grad_norm": 0.06350723808650868,
-      "learning_rate": 0.0008392190604670293,
-      "loss": 1.1597,
-      "step": 2200
-    },
-    {
-      "epoch": 0.33713018882348444,
-      "grad_norm": 0.06598685397103475,
-      "learning_rate": 0.0008382375636855522,
-      "loss": 1.1897,
-      "step": 2205
-    },
-    {
-      "epoch": 0.337894656371837,
-      "grad_norm": 0.0666673189819934,
-      "learning_rate": 0.0008372536579927233,
-      "loss": 1.1718,
-      "step": 2210
-    },
-    {
-      "epoch": 0.3386591239201896,
-      "grad_norm": 0.062110253414163125,
-      "learning_rate": 0.0008362673503958707,
-      "loss": 1.1791,
-      "step": 2215
-    },
-    {
-      "epoch": 0.3394235914685422,
-      "grad_norm": 0.06769528641288612,
-      "learning_rate": 0.0008352786479194288,
-      "loss": 1.2181,
-      "step": 2220
-    },
-    {
-      "epoch": 0.34018805901689475,
-      "grad_norm": 0.08008302678521025,
-      "learning_rate": 0.000834287557604888,
-      "loss": 1.1546,
-      "step": 2225
-    },
-    {
-      "epoch": 0.3409525265652473,
-      "grad_norm": 0.0764574913479765,
-      "learning_rate": 0.0008332940865107452,
-      "loss": 1.1988,
-      "step": 2230
-    },
-    {
-      "epoch": 0.3417169941135999,
-      "grad_norm": 0.07425606538696802,
-      "learning_rate": 0.000832298241712453,
-      "loss": 1.19,
-      "step": 2235
-    },
-    {
-      "epoch": 0.3424814616619524,
-      "grad_norm": 0.06744819104094633,
-      "learning_rate": 0.0008313000303023688,
-      "loss": 1.1711,
-      "step": 2240
-    },
-    {
-      "epoch": 0.343245929210305,
-      "grad_norm": 0.061154631516393325,
-      "learning_rate": 0.0008302994593897055,
-      "loss": 1.1914,
-      "step": 2245
-    },
-    {
-      "epoch": 0.3440103967586576,
-      "grad_norm": 0.07404312882255028,
-      "learning_rate": 0.0008292965361004801,
-      "loss": 1.1682,
-      "step": 2250
-    },
-    {
-      "epoch": 0.34477486430701015,
-      "grad_norm": 0.06742969913673322,
-      "learning_rate": 0.0008282912675774632,
-      "loss": 1.2123,
-      "step": 2255
-    },
-    {
-      "epoch": 0.34553933185536273,
-      "grad_norm": 0.06384028111357246,
-      "learning_rate": 0.000827283660980128,
-      "loss": 1.1466,
-      "step": 2260
-    },
-    {
-      "epoch": 0.3463037994037153,
-      "grad_norm": 0.0807612774368799,
-      "learning_rate": 0.0008262737234845991,
-      "loss": 1.1709,
-      "step": 2265
-    },
-    {
-      "epoch": 0.3470682669520679,
-      "grad_norm": 0.06299523953324049,
-      "learning_rate": 0.0008252614622836021,
-      "loss": 1.1561,
-      "step": 2270
-    },
-    {
-      "epoch": 0.34783273450042046,
-      "grad_norm": 0.07042573805382132,
-      "learning_rate": 0.0008242468845864115,
-      "loss": 1.2122,
-      "step": 2275
-    },
-    {
-      "epoch": 0.34859720204877304,
-      "grad_norm": 0.07326150426716001,
-      "learning_rate": 0.0008232299976187999,
-      "loss": 1.191,
-      "step": 2280
-    },
-    {
-      "epoch": 0.3493616695971256,
-      "grad_norm": 0.07733654467734337,
-      "learning_rate": 0.0008222108086229865,
-      "loss": 1.1814,
-      "step": 2285
-    },
-    {
-      "epoch": 0.3501261371454782,
-      "grad_norm": 0.06653554190977337,
-      "learning_rate": 0.0008211893248575854,
-      "loss": 1.1706,
-      "step": 2290
-    },
-    {
-      "epoch": 0.35089060469383077,
-      "grad_norm": 0.07586818669104102,
-      "learning_rate": 0.0008201655535975541,
-      "loss": 1.1816,
-      "step": 2295
-    },
-    {
-      "epoch": 0.35165507224218334,
-      "grad_norm": 0.0722200099530391,
-      "learning_rate": 0.0008191395021341408,
-      "loss": 1.2203,
-      "step": 2300
-    },
-    {
-      "epoch": 0.3524195397905359,
-      "grad_norm": 0.07729811512497685,
-      "learning_rate": 0.0008181111777748342,
-      "loss": 1.2101,
-      "step": 2305
-    },
-    {
-      "epoch": 0.35318400733888844,
-      "grad_norm": 0.06548329762746453,
-      "learning_rate": 0.0008170805878433099,
-      "loss": 1.2126,
-      "step": 2310
-    },
-    {
-      "epoch": 0.353948474887241,
-      "grad_norm": 0.16878411753529185,
-      "learning_rate": 0.0008160477396793788,
-      "loss": 1.1993,
-      "step": 2315
-    },
-    {
-      "epoch": 0.3547129424355936,
-      "grad_norm": 0.08450927348552127,
-      "learning_rate": 0.0008150126406389351,
-      "loss": 1.1903,
-      "step": 2320
-    },
-    {
-      "epoch": 0.35547740998394617,
-      "grad_norm": 0.07393765853159336,
-      "learning_rate": 0.0008139752980939033,
-      "loss": 1.1896,
-      "step": 2325
-    },
-    {
-      "epoch": 0.35624187753229875,
-      "grad_norm": 0.0645007516624614,
-      "learning_rate": 0.000812935719432186,
-      "loss": 1.203,
-      "step": 2330
-    },
-    {
-      "epoch": 0.3570063450806513,
-      "grad_norm": 0.07480048796944509,
-      "learning_rate": 0.0008118939120576117,
-      "loss": 1.2232,
-      "step": 2335
-    },
-    {
-      "epoch": 0.3577708126290039,
-      "grad_norm": 0.06387184157320323,
-      "learning_rate": 0.0008108498833898814,
-      "loss": 1.1844,
-      "step": 2340
-    },
-    {
-      "epoch": 0.3585352801773565,
-      "grad_norm": 0.0706157487905302,
-      "learning_rate": 0.0008098036408645161,
-      "loss": 1.1545,
-      "step": 2345
-    },
-    {
-      "epoch": 0.35929974772570905,
-      "grad_norm": 0.06346402094887067,
-      "learning_rate": 0.0008087551919328038,
-      "loss": 1.1998,
-      "step": 2350
-    },
-    {
-      "epoch": 0.3600642152740616,
-      "grad_norm": 0.07219725356620699,
-      "learning_rate": 0.0008077045440617464,
-      "loss": 1.2208,
-      "step": 2355
-    },
-    {
-      "epoch": 0.3608286828224142,
-      "grad_norm": 0.06980068763494783,
-      "learning_rate": 0.0008066517047340065,
-      "loss": 1.1953,
-      "step": 2360
-    },
-    {
-      "epoch": 0.3615931503707668,
-      "grad_norm": 0.06836413221657871,
-      "learning_rate": 0.0008055966814478543,
-      "loss": 1.1998,
-      "step": 2365
-    },
-    {
-      "epoch": 0.36235761791911936,
-      "grad_norm": 0.06020785995804865,
-      "learning_rate": 0.0008045394817171138,
-      "loss": 1.1495,
-      "step": 2370
-    },
-    {
-      "epoch": 0.36312208546747193,
-      "grad_norm": 0.06584082195664415,
-      "learning_rate": 0.0008034801130711101,
-      "loss": 1.1687,
-      "step": 2375
-    },
-    {
-      "epoch": 0.36388655301582445,
-      "grad_norm": 0.07621192168498155,
-      "learning_rate": 0.000802418583054614,
-      "loss": 1.185,
-      "step": 2380
-    },
-    {
-      "epoch": 0.36465102056417703,
-      "grad_norm": 0.0858231914116967,
-      "learning_rate": 0.0008013548992277909,
-      "loss": 1.1875,
-      "step": 2385
-    },
-    {
-      "epoch": 0.3654154881125296,
-      "grad_norm": 0.06562775419380161,
-      "learning_rate": 0.0008002890691661446,
-      "loss": 1.213,
-      "step": 2390
-    },
-    {
-      "epoch": 0.3661799556608822,
-      "grad_norm": 0.07523928065889657,
-      "learning_rate": 0.0007992211004604646,
-      "loss": 1.1824,
-      "step": 2395
-    },
-    {
-      "epoch": 0.36694442320923476,
-      "grad_norm": 0.08499994679950856,
-      "learning_rate": 0.0007981510007167718,
-      "loss": 1.1871,
-      "step": 2400
-    },
-    {
-      "epoch": 0.36770889075758734,
-      "grad_norm": 0.06903788845491694,
-      "learning_rate": 0.0007970787775562641,
-      "loss": 1.1822,
-      "step": 2405
-    },
-    {
-      "epoch": 0.3684733583059399,
-      "grad_norm": 0.06597575725856182,
-      "learning_rate": 0.000796004438615262,
-      "loss": 1.1925,
-      "step": 2410
-    },
-    {
-      "epoch": 0.3692378258542925,
-      "grad_norm": 0.07344580250828273,
-      "learning_rate": 0.0007949279915451553,
-      "loss": 1.1823,
-      "step": 2415
-    },
-    {
-      "epoch": 0.37000229340264507,
-      "grad_norm": 0.060426789873964576,
-      "learning_rate": 0.0007938494440123468,
-      "loss": 1.1681,
-      "step": 2420
-    },
-    {
-      "epoch": 0.37076676095099764,
-      "grad_norm": 0.07015273475512843,
-      "learning_rate": 0.0007927688036981994,
-      "loss": 1.2012,
-      "step": 2425
-    },
-    {
-      "epoch": 0.3715312284993502,
-      "grad_norm": 0.07711911062361188,
-      "learning_rate": 0.0007916860782989806,
-      "loss": 1.1532,
-      "step": 2430
-    },
-    {
-      "epoch": 0.3722956960477028,
-      "grad_norm": 0.06654784595079923,
-      "learning_rate": 0.0007906012755258072,
-      "loss": 1.2284,
-      "step": 2435
-    },
-    {
-      "epoch": 0.37306016359605537,
-      "grad_norm": 0.062029588756078474,
-      "learning_rate": 0.0007895144031045918,
-      "loss": 1.2247,
-      "step": 2440
-    },
-    {
-      "epoch": 0.3738246311444079,
-      "grad_norm": 0.0877602095214219,
-      "learning_rate": 0.0007884254687759862,
-      "loss": 1.1617,
-      "step": 2445
-    },
-    {
-      "epoch": 0.37458909869276047,
-      "grad_norm": 0.06287647828722921,
-      "learning_rate": 0.0007873344802953277,
-      "loss": 1.1672,
-      "step": 2450
-    },
-    {
-      "epoch": 0.37535356624111305,
-      "grad_norm": 0.06438896028122171,
-      "learning_rate": 0.0007862414454325826,
-      "loss": 1.1483,
-      "step": 2455
-    },
-    {
-      "epoch": 0.3761180337894656,
-      "grad_norm": 0.0593997017120742,
-      "learning_rate": 0.0007851463719722913,
-      "loss": 1.1945,
-      "step": 2460
-    },
-    {
-      "epoch": 0.3768825013378182,
-      "grad_norm": 0.09414055116242716,
-      "learning_rate": 0.0007840492677135139,
-      "loss": 1.2039,
-      "step": 2465
-    },
-    {
-      "epoch": 0.3776469688861708,
-      "grad_norm": 0.07238337604705601,
-      "learning_rate": 0.0007829501404697728,
-      "loss": 1.1972,
-      "step": 2470
-    },
-    {
-      "epoch": 0.37841143643452335,
-      "grad_norm": 0.05784642990402551,
-      "learning_rate": 0.0007818489980689985,
-      "loss": 1.1969,
-      "step": 2475
-    },
-    {
-      "epoch": 0.37917590398287593,
-      "grad_norm": 0.06420946957423379,
-      "learning_rate": 0.0007807458483534731,
-      "loss": 1.1344,
-      "step": 2480
-    },
-    {
-      "epoch": 0.3799403715312285,
-      "grad_norm": 0.0622062481944738,
-      "learning_rate": 0.0007796406991797748,
-      "loss": 1.1632,
-      "step": 2485
-    },
-    {
-      "epoch": 0.3807048390795811,
-      "grad_norm": 0.06698914216587033,
-      "learning_rate": 0.0007785335584187218,
-      "loss": 1.2193,
-      "step": 2490
-    },
-    {
-      "epoch": 0.38146930662793366,
-      "grad_norm": 0.06501159827802083,
-      "learning_rate": 0.0007774244339553168,
-      "loss": 1.1949,
-      "step": 2495
-    },
-    {
-      "epoch": 0.38223377417628623,
-      "grad_norm": 0.06352771347650159,
-      "learning_rate": 0.0007763133336886891,
-      "loss": 1.1694,
-      "step": 2500
-    },
-    {
-      "epoch": 0.3829982417246388,
-      "grad_norm": 0.062043540050934584,
-      "learning_rate": 0.0007752002655320411,
-      "loss": 1.1754,
-      "step": 2505
-    },
-    {
-      "epoch": 0.3837627092729914,
-      "grad_norm": 0.05993356867245502,
-      "learning_rate": 0.0007740852374125892,
-      "loss": 1.1586,
-      "step": 2510
-    },
-    {
-      "epoch": 0.3845271768213439,
-      "grad_norm": 0.08359997915488934,
-      "learning_rate": 0.0007729682572715092,
-      "loss": 1.2118,
-      "step": 2515
-    },
-    {
-      "epoch": 0.3852916443696965,
-      "grad_norm": 0.07542827308050223,
-      "learning_rate": 0.0007718493330638788,
-      "loss": 1.1713,
-      "step": 2520
-    },
-    {
-      "epoch": 0.38605611191804906,
-      "grad_norm": 0.06013940005700568,
-      "learning_rate": 0.0007707284727586216,
-      "loss": 1.2009,
-      "step": 2525
-    },
-    {
-      "epoch": 0.38682057946640164,
-      "grad_norm": 0.05981452616943288,
-      "learning_rate": 0.0007696056843384493,
-      "loss": 1.1794,
-      "step": 2530
-    },
-    {
-      "epoch": 0.3875850470147542,
-      "grad_norm": 0.06567705884727439,
-      "learning_rate": 0.0007684809757998066,
-      "loss": 1.155,
-      "step": 2535
-    },
-    {
-      "epoch": 0.3883495145631068,
-      "grad_norm": 0.07472356348523346,
-      "learning_rate": 0.0007673543551528121,
-      "loss": 1.1884,
-      "step": 2540
-    },
-    {
-      "epoch": 0.38911398211145937,
-      "grad_norm": 0.06125765376781909,
-      "learning_rate": 0.0007662258304212033,
-      "loss": 1.163,
-      "step": 2545
-    },
-    {
-      "epoch": 0.38987844965981194,
-      "grad_norm": 0.0660322307444375,
-      "learning_rate": 0.0007650954096422775,
-      "loss": 1.1842,
-      "step": 2550
-    },
-    {
-      "epoch": 0.3906429172081645,
-      "grad_norm": 0.07730780676951493,
-      "learning_rate": 0.0007639631008668364,
-      "loss": 1.1628,
-      "step": 2555
-    },
-    {
-      "epoch": 0.3914073847565171,
-      "grad_norm": 0.08716048156438365,
-      "learning_rate": 0.0007628289121591276,
-      "loss": 1.2012,
-      "step": 2560
-    },
-    {
-      "epoch": 0.3921718523048697,
-      "grad_norm": 0.05864132359328279,
-      "learning_rate": 0.0007616928515967875,
-      "loss": 1.1638,
-      "step": 2565
-    },
-    {
-      "epoch": 0.39293631985322225,
-      "grad_norm": 0.06514781329333548,
-      "learning_rate": 0.0007605549272707835,
-      "loss": 1.1472,
-      "step": 2570
-    },
-    {
-      "epoch": 0.3937007874015748,
-      "grad_norm": 0.07818283105138371,
-      "learning_rate": 0.0007594151472853572,
-      "loss": 1.1711,
-      "step": 2575
-    },
-    {
-      "epoch": 0.3944652549499274,
-      "grad_norm": 0.07142906060531046,
-      "learning_rate": 0.0007582735197579657,
-      "loss": 1.1789,
-      "step": 2580
-    },
-    {
-      "epoch": 0.3952297224982799,
-      "grad_norm": 0.06460911519595092,
-      "learning_rate": 0.0007571300528192242,
-      "loss": 1.1553,
-      "step": 2585
-    },
-    {
-      "epoch": 0.3959941900466325,
-      "grad_norm": 0.06458967247713916,
-      "learning_rate": 0.0007559847546128482,
-      "loss": 1.1672,
-      "step": 2590
-    },
-    {
-      "epoch": 0.3967586575949851,
-      "grad_norm": 0.06555873403698737,
-      "learning_rate": 0.0007548376332955956,
-      "loss": 1.1389,
-      "step": 2595
-    },
-    {
-      "epoch": 0.39752312514333765,
-      "grad_norm": 0.06320995634242295,
-      "learning_rate": 0.0007536886970372079,
-      "loss": 1.172,
-      "step": 2600
-    },
-    {
-      "epoch": 0.39828759269169023,
-      "grad_norm": 0.0692346965255682,
-      "learning_rate": 0.0007525379540203532,
-      "loss": 1.1449,
-      "step": 2605
-    },
-    {
-      "epoch": 0.3990520602400428,
-      "grad_norm": 0.06669313758415195,
-      "learning_rate": 0.0007513854124405664,
-      "loss": 1.1875,
-      "step": 2610
-    },
-    {
-      "epoch": 0.3998165277883954,
-      "grad_norm": 0.06168539003076038,
-      "learning_rate": 0.0007502310805061927,
-      "loss": 1.1822,
-      "step": 2615
-    },
-    {
-      "epoch": 0.40058099533674796,
-      "grad_norm": 0.07577936750925265,
-      "learning_rate": 0.0007490749664383271,
-      "loss": 1.1878,
-      "step": 2620
-    },
-    {
-      "epoch": 0.40134546288510053,
-      "grad_norm": 0.07513830844488023,
-      "learning_rate": 0.0007479170784707574,
-      "loss": 1.1538,
-      "step": 2625
-    },
-    {
-      "epoch": 0.4021099304334531,
-      "grad_norm": 0.0704004134116212,
-      "learning_rate": 0.0007467574248499052,
-      "loss": 1.159,
-      "step": 2630
-    },
-    {
-      "epoch": 0.4028743979818057,
-      "grad_norm": 0.06393476454512485,
-      "learning_rate": 0.0007455960138347665,
-      "loss": 1.1533,
-      "step": 2635
-    },
-    {
-      "epoch": 0.40363886553015826,
-      "grad_norm": 0.060988648131048324,
-      "learning_rate": 0.0007444328536968537,
-      "loss": 1.1878,
-      "step": 2640
-    },
-    {
-      "epoch": 0.40440333307851084,
-      "grad_norm": 0.07642776920542559,
-      "learning_rate": 0.0007432679527201364,
-      "loss": 1.21,
-      "step": 2645
-    },
-    {
-      "epoch": 0.4051678006268634,
-      "grad_norm": 0.07143064451662551,
-      "learning_rate": 0.0007421013192009823,
-      "loss": 1.146,
-      "step": 2650
-    },
-    {
-      "epoch": 0.40593226817521594,
-      "grad_norm": 0.06608214029577218,
-      "learning_rate": 0.000740932961448098,
-      "loss": 1.142,
-      "step": 2655
-    },
-    {
-      "epoch": 0.4066967357235685,
-      "grad_norm": 0.2372583228137919,
-      "learning_rate": 0.0007397628877824702,
-      "loss": 1.1696,
-      "step": 2660
-    },
-    {
-      "epoch": 0.4074612032719211,
-      "grad_norm": 0.06401105720414174,
-      "learning_rate": 0.0007385911065373061,
-      "loss": 1.1434,
-      "step": 2665
-    },
-    {
-      "epoch": 0.40822567082027367,
-      "grad_norm": 0.08634710294909215,
-      "learning_rate": 0.0007374176260579745,
-      "loss": 1.1806,
-      "step": 2670
-    },
-    {
-      "epoch": 0.40899013836862624,
-      "grad_norm": 0.0686184603428492,
-      "learning_rate": 0.0007362424547019457,
-      "loss": 1.1754,
-      "step": 2675
-    },
-    {
-      "epoch": 0.4097546059169788,
-      "grad_norm": 0.07213216317785283,
-      "learning_rate": 0.0007350656008387326,
-      "loss": 1.1699,
-      "step": 2680
-    },
-    {
-      "epoch": 0.4105190734653314,
-      "grad_norm": 0.0739348063942258,
-      "learning_rate": 0.0007338870728498308,
-      "loss": 1.1276,
-      "step": 2685
-    },
-    {
-      "epoch": 0.411283541013684,
-      "grad_norm": 0.07052527755070558,
-      "learning_rate": 0.0007327068791286586,
-      "loss": 1.1166,
-      "step": 2690
-    },
-    {
-      "epoch": 0.41204800856203655,
-      "grad_norm": 0.061734383183528134,
-      "learning_rate": 0.0007315250280804982,
-      "loss": 1.1694,
-      "step": 2695
-    },
-    {
-      "epoch": 0.4128124761103891,
-      "grad_norm": 0.06851752144246956,
-      "learning_rate": 0.0007303415281224346,
-      "loss": 1.1696,
-      "step": 2700
-    },
-    {
-      "epoch": 0.4135769436587417,
-      "grad_norm": 0.06924429941802722,
-      "learning_rate": 0.0007291563876832966,
-      "loss": 1.134,
-      "step": 2705
-    },
-    {
-      "epoch": 0.4143414112070943,
-      "grad_norm": 0.062236735706859295,
-      "learning_rate": 0.0007279696152035963,
-      "loss": 1.1756,
-      "step": 2710
-    },
-    {
-      "epoch": 0.41510587875544686,
-      "grad_norm": 0.06605748129048904,
-      "learning_rate": 0.000726781219135469,
-      "loss": 1.1616,
-      "step": 2715
-    },
-    {
-      "epoch": 0.4158703463037994,
-      "grad_norm": 0.05979864244024164,
-      "learning_rate": 0.0007255912079426136,
-      "loss": 1.1586,
-      "step": 2720
-    },
-    {
-      "epoch": 0.41663481385215195,
-      "grad_norm": 0.05911691466595612,
-      "learning_rate": 0.0007243995901002312,
-      "loss": 1.203,
-      "step": 2725
-    },
-    {
-      "epoch": 0.41739928140050453,
-      "grad_norm": 0.07120579590463852,
-      "learning_rate": 0.0007232063740949656,
-      "loss": 1.1739,
-      "step": 2730
-    },
-    {
-      "epoch": 0.4181637489488571,
-      "grad_norm": 0.07696734291297799,
-      "learning_rate": 0.0007220115684248429,
-      "loss": 1.0993,
-      "step": 2735
-    },
-    {
-      "epoch": 0.4189282164972097,
-      "grad_norm": 0.08907278539208283,
-      "learning_rate": 0.0007208151815992107,
-      "loss": 1.1619,
-      "step": 2740
-    },
-    {
-      "epoch": 0.41969268404556226,
-      "grad_norm": 0.062313137263563724,
-      "learning_rate": 0.000719617222138677,
-      "loss": 1.1466,
-      "step": 2745
-    },
-    {
-      "epoch": 0.42045715159391484,
-      "grad_norm": 0.07514342527228114,
-      "learning_rate": 0.0007184176985750506,
-      "loss": 1.215,
-      "step": 2750
-    },
-    {
-      "epoch": 0.4212216191422674,
-      "grad_norm": 0.07162533739741002,
-      "learning_rate": 0.0007172166194512792,
-      "loss": 1.1552,
-      "step": 2755
-    },
-    {
-      "epoch": 0.42198608669062,
-      "grad_norm": 0.06920738404553707,
-      "learning_rate": 0.0007160139933213898,
-      "loss": 1.1844,
-      "step": 2760
-    },
-    {
-      "epoch": 0.42275055423897256,
-      "grad_norm": 0.06239605059074051,
-      "learning_rate": 0.0007148098287504266,
-      "loss": 1.1899,
-      "step": 2765
-    },
-    {
-      "epoch": 0.42351502178732514,
-      "grad_norm": 0.07204617294710451,
-      "learning_rate": 0.0007136041343143908,
-      "loss": 1.1943,
-      "step": 2770
-    },
-    {
-      "epoch": 0.4242794893356777,
-      "grad_norm": 0.07656583544762036,
-      "learning_rate": 0.0007123969186001791,
-      "loss": 1.1758,
-      "step": 2775
-    },
-    {
-      "epoch": 0.4250439568840303,
-      "grad_norm": 0.06931848694667721,
-      "learning_rate": 0.0007111881902055223,
-      "loss": 1.1583,
-      "step": 2780
-    },
-    {
-      "epoch": 0.42580842443238287,
-      "grad_norm": 0.07497458994101201,
-      "learning_rate": 0.000709977957738925,
-      "loss": 1.1558,
-      "step": 2785
-    },
-    {
-      "epoch": 0.4265728919807354,
-      "grad_norm": 0.0798030234936708,
-      "learning_rate": 0.0007087662298196037,
-      "loss": 1.1589,
-      "step": 2790
-    },
-    {
-      "epoch": 0.42733735952908797,
-      "grad_norm": 0.07255931989483083,
-      "learning_rate": 0.000707553015077425,
-      "loss": 1.1175,
-      "step": 2795
-    },
-    {
-      "epoch": 0.42810182707744054,
-      "grad_norm": 0.07101246431967022,
-      "learning_rate": 0.000706338322152845,
-      "loss": 1.181,
-      "step": 2800
-    },
-    {
-      "epoch": 0.4288662946257931,
-      "grad_norm": 0.05920290047242307,
-      "learning_rate": 0.000705122159696847,
-      "loss": 1.1552,
-      "step": 2805
-    },
-    {
-      "epoch": 0.4296307621741457,
-      "grad_norm": 0.07048728251583447,
-      "learning_rate": 0.0007039045363708807,
-      "loss": 1.1649,
-      "step": 2810
-    },
-    {
-      "epoch": 0.4303952297224983,
-      "grad_norm": 0.057337298176797065,
-      "learning_rate": 0.0007026854608467994,
-      "loss": 1.1646,
-      "step": 2815
-    },
-    {
-      "epoch": 0.43115969727085085,
-      "grad_norm": 0.06921836651212089,
-      "learning_rate": 0.0007014649418067994,
-      "loss": 1.1547,
-      "step": 2820
-    },
-    {
-      "epoch": 0.4319241648192034,
-      "grad_norm": 0.06676730022332167,
-      "learning_rate": 0.0007002429879433577,
-      "loss": 1.1512,
-      "step": 2825
-    },
-    {
-      "epoch": 0.432688632367556,
-      "grad_norm": 0.0630615394746687,
-      "learning_rate": 0.0006990196079591694,
-      "loss": 1.1271,
-      "step": 2830
-    },
-    {
-      "epoch": 0.4334530999159086,
-      "grad_norm": 0.06891965405826847,
-      "learning_rate": 0.000697794810567087,
-      "loss": 1.1506,
-      "step": 2835
-    },
-    {
-      "epoch": 0.43421756746426116,
-      "grad_norm": 0.0658246172183584,
-      "learning_rate": 0.0006965686044900577,
-      "loss": 1.1165,
-      "step": 2840
-    },
-    {
-      "epoch": 0.43498203501261373,
-      "grad_norm": 0.08178737314912204,
-      "learning_rate": 0.0006953409984610607,
-      "loss": 1.1679,
-      "step": 2845
-    },
-    {
-      "epoch": 0.4357465025609663,
-      "grad_norm": 0.0658367549932726,
-      "learning_rate": 0.0006941120012230463,
-      "loss": 1.1854,
-      "step": 2850
-    },
-    {
-      "epoch": 0.4365109701093189,
-      "grad_norm": 0.06579939923600828,
-      "learning_rate": 0.0006928816215288722,
-      "loss": 1.1589,
-      "step": 2855
-    },
-    {
-      "epoch": 0.4372754376576714,
-      "grad_norm": 0.06821461913343178,
-      "learning_rate": 0.0006916498681412429,
-      "loss": 1.1409,
-      "step": 2860
-    },
-    {
-      "epoch": 0.438039905206024,
-      "grad_norm": 0.06302865374733113,
-      "learning_rate": 0.0006904167498326451,
-      "loss": 1.1557,
-      "step": 2865
-    },
-    {
-      "epoch": 0.43880437275437656,
-      "grad_norm": 0.05918815408942707,
-      "learning_rate": 0.0006891822753852874,
-      "loss": 1.1565,
-      "step": 2870
-    },
-    {
-      "epoch": 0.43956884030272914,
-      "grad_norm": 0.06542754378140148,
-      "learning_rate": 0.0006879464535910358,
-      "loss": 1.1577,
-      "step": 2875
-    },
-    {
-      "epoch": 0.4403333078510817,
-      "grad_norm": 0.0647221155275831,
-      "learning_rate": 0.000686709293251353,
-      "loss": 1.1602,
-      "step": 2880
-    },
-    {
-      "epoch": 0.4410977753994343,
-      "grad_norm": 0.0579063394917837,
-      "learning_rate": 0.0006854708031772341,
-      "loss": 1.1827,
-      "step": 2885
-    },
-    {
-      "epoch": 0.44186224294778687,
-      "grad_norm": 0.06365215547735684,
-      "learning_rate": 0.0006842309921891447,
-      "loss": 1.164,
-      "step": 2890
-    },
-    {
-      "epoch": 0.44262671049613944,
-      "grad_norm": 0.06261024715809171,
-      "learning_rate": 0.000682989869116958,
-      "loss": 1.191,
-      "step": 2895
-    },
-    {
-      "epoch": 0.443391178044492,
-      "grad_norm": 0.07234668451704347,
-      "learning_rate": 0.0006817474427998916,
-      "loss": 1.1505,
-      "step": 2900
-    },
-    {
-      "epoch": 0.4441556455928446,
-      "grad_norm": 0.07197614974444562,
-      "learning_rate": 0.0006805037220864449,
-      "loss": 1.1858,
-      "step": 2905
-    },
-    {
-      "epoch": 0.44492011314119717,
-      "grad_norm": 0.06560613456796847,
-      "learning_rate": 0.000679258715834336,
-      "loss": 1.1559,
-      "step": 2910
-    },
-    {
-      "epoch": 0.44568458068954975,
-      "grad_norm": 0.06198035141165582,
-      "learning_rate": 0.000678012432910438,
-      "loss": 1.1545,
-      "step": 2915
-    },
-    {
-      "epoch": 0.4464490482379023,
-      "grad_norm": 0.06444936717504604,
-      "learning_rate": 0.0006767648821907172,
-      "loss": 1.1336,
-      "step": 2920
-    },
-    {
-      "epoch": 0.4472135157862549,
-      "grad_norm": 0.0645392975838134,
-      "learning_rate": 0.0006755160725601685,
-      "loss": 1.1391,
-      "step": 2925
-    },
-    {
-      "epoch": 0.4479779833346074,
-      "grad_norm": 0.06065312532733533,
-      "learning_rate": 0.0006742660129127529,
-      "loss": 1.1231,
-      "step": 2930
-    },
-    {
-      "epoch": 0.44874245088296,
-      "grad_norm": 0.05770027644463755,
-      "learning_rate": 0.000673014712151334,
-      "loss": 1.1665,
-      "step": 2935
-    },
-    {
-      "epoch": 0.4495069184313126,
-      "grad_norm": 0.059687456397664214,
-      "learning_rate": 0.0006717621791876146,
-      "loss": 1.1478,
-      "step": 2940
-    },
-    {
-      "epoch": 0.45027138597966515,
-      "grad_norm": 0.0709234344203115,
-      "learning_rate": 0.0006705084229420729,
-      "loss": 1.1418,
-      "step": 2945
-    },
-    {
-      "epoch": 0.4510358535280177,
-      "grad_norm": 0.08126674149433259,
-      "learning_rate": 0.0006692534523438993,
-      "loss": 1.079,
-      "step": 2950
-    },
-    {
-      "epoch": 0.4518003210763703,
-      "grad_norm": 0.07012397401994788,
-      "learning_rate": 0.0006679972763309333,
-      "loss": 1.1583,
-      "step": 2955
-    },
-    {
-      "epoch": 0.4525647886247229,
-      "grad_norm": 0.060697021218761534,
-      "learning_rate": 0.0006667399038495986,
-      "loss": 1.1223,
-      "step": 2960
-    },
-    {
-      "epoch": 0.45332925617307546,
-      "grad_norm": 0.07452330710843885,
-      "learning_rate": 0.0006654813438548404,
-      "loss": 1.1536,
-      "step": 2965
-    },
-    {
-      "epoch": 0.45409372372142803,
-      "grad_norm": 0.06928736586445865,
-      "learning_rate": 0.0006642216053100616,
-      "loss": 1.1137,
-      "step": 2970
-    },
-    {
-      "epoch": 0.4548581912697806,
-      "grad_norm": 0.06759852135632291,
-      "learning_rate": 0.000662960697187058,
-      "loss": 1.1315,
-      "step": 2975
-    },
-    {
-      "epoch": 0.4556226588181332,
-      "grad_norm": 0.06959880066718058,
-      "learning_rate": 0.0006616986284659557,
-      "loss": 1.165,
-      "step": 2980
-    },
-    {
-      "epoch": 0.45638712636648576,
-      "grad_norm": 0.06392136917578246,
-      "learning_rate": 0.000660435408135146,
-      "loss": 1.1531,
-      "step": 2985
-    },
-    {
-      "epoch": 0.45715159391483834,
-      "grad_norm": 0.06630940031172995,
-      "learning_rate": 0.0006591710451912225,
-      "loss": 1.1314,
-      "step": 2990
-    },
-    {
-      "epoch": 0.45791606146319086,
-      "grad_norm": 0.10697784394354062,
-      "learning_rate": 0.000657905548638916,
-      "loss": 1.1522,
-      "step": 2995
-    },
-    {
-      "epoch": 0.45868052901154344,
-      "grad_norm": 0.060670680576808696,
-      "learning_rate": 0.0006566389274910309,
-      "loss": 1.0984,
-      "step": 3000
-    },
-    {
-      "epoch": 0.459444996559896,
-      "grad_norm": 0.06606537631594284,
-      "learning_rate": 0.000655371190768381,
-      "loss": 1.1592,
-      "step": 3005
-    },
-    {
-      "epoch": 0.4602094641082486,
-      "grad_norm": 0.06083196601012183,
-      "learning_rate": 0.000654102347499725,
-      "loss": 1.1495,
-      "step": 3010
-    },
-    {
-      "epoch": 0.46097393165660117,
-      "grad_norm": 0.05898668376641134,
-      "learning_rate": 0.0006528324067217025,
-      "loss": 1.1208,
-      "step": 3015
-    },
-    {
-      "epoch": 0.46173839920495374,
-      "grad_norm": 0.05423508964655473,
-      "learning_rate": 0.0006515613774787698,
-      "loss": 1.1107,
-      "step": 3020
-    },
-    {
-      "epoch": 0.4625028667533063,
-      "grad_norm": 0.06527274534130718,
-      "learning_rate": 0.0006502892688231342,
-      "loss": 1.1523,
-      "step": 3025
-    },
-    {
-      "epoch": 0.4632673343016589,
-      "grad_norm": 0.06470168742423997,
-      "learning_rate": 0.0006490160898146918,
-      "loss": 1.1557,
-      "step": 3030
-    },
-    {
-      "epoch": 0.46403180185001147,
-      "grad_norm": 0.0651661791184414,
-      "learning_rate": 0.000647741849520961,
-      "loss": 1.1149,
-      "step": 3035
-    },
-    {
-      "epoch": 0.46479626939836405,
-      "grad_norm": 0.06518913789788636,
-      "learning_rate": 0.0006464665570170185,
-      "loss": 1.1251,
-      "step": 3040
-    },
-    {
-      "epoch": 0.4655607369467166,
-      "grad_norm": 0.06041300256236077,
-      "learning_rate": 0.0006451902213854352,
-      "loss": 1.1574,
-      "step": 3045
-    },
-    {
-      "epoch": 0.4663252044950692,
-      "grad_norm": 0.07182863661248022,
-      "learning_rate": 0.0006439128517162109,
-      "loss": 1.113,
-      "step": 3050
-    },
-    {
-      "epoch": 0.4670896720434218,
-      "grad_norm": 0.07484708764880833,
-      "learning_rate": 0.0006426344571067096,
-      "loss": 1.1027,
-      "step": 3055
-    },
-    {
-      "epoch": 0.46785413959177435,
-      "grad_norm": 0.06402028003954885,
-      "learning_rate": 0.0006413550466615952,
-      "loss": 1.1398,
-      "step": 3060
-    },
-    {
-      "epoch": 0.4686186071401269,
-      "grad_norm": 0.05903548215228208,
-      "learning_rate": 0.0006400746294927662,
-      "loss": 1.1552,
-      "step": 3065
-    },
-    {
-      "epoch": 0.46938307468847945,
-      "grad_norm": 0.0701643196016173,
-      "learning_rate": 0.000638793214719291,
-      "loss": 1.1333,
-      "step": 3070
-    },
-    {
-      "epoch": 0.47014754223683203,
-      "grad_norm": 0.08476933332297756,
-      "learning_rate": 0.0006375108114673424,
-      "loss": 1.0994,
-      "step": 3075
-    },
-    {
-      "epoch": 0.4709120097851846,
-      "grad_norm": 0.06475816517225963,
-      "learning_rate": 0.0006362274288701342,
-      "loss": 1.1759,
-      "step": 3080
-    },
-    {
-      "epoch": 0.4716764773335372,
-      "grad_norm": 0.06895976508220301,
-      "learning_rate": 0.0006349430760678538,
-      "loss": 1.1462,
-      "step": 3085
-    },
-    {
-      "epoch": 0.47244094488188976,
-      "grad_norm": 0.05934511939754207,
-      "learning_rate": 0.0006336577622075992,
-      "loss": 1.1841,
-      "step": 3090
-    },
-    {
-      "epoch": 0.47320541243024233,
-      "grad_norm": 0.060053855049644576,
-      "learning_rate": 0.0006323714964433126,
-      "loss": 1.1397,
-      "step": 3095
-    },
-    {
-      "epoch": 0.4739698799785949,
-      "grad_norm": 0.0727525002225497,
-      "learning_rate": 0.0006310842879357157,
-      "loss": 1.1602,
-      "step": 3100
-    },
-    {
-      "epoch": 0.4747343475269475,
-      "grad_norm": 0.06560733009273371,
-      "learning_rate": 0.0006297961458522444,
-      "loss": 1.145,
-      "step": 3105
-    },
-    {
-      "epoch": 0.47549881507530006,
-      "grad_norm": 0.06296234629565342,
-      "learning_rate": 0.0006285070793669836,
-      "loss": 1.0818,
-      "step": 3110
-    },
-    {
-      "epoch": 0.47626328262365264,
-      "grad_norm": 0.07048483200966957,
-      "learning_rate": 0.0006272170976606016,
-      "loss": 1.1443,
-      "step": 3115
-    },
-    {
-      "epoch": 0.4770277501720052,
-      "grad_norm": 0.060550118582415664,
-      "learning_rate": 0.0006259262099202849,
-      "loss": 1.1041,
-      "step": 3120
-    },
-    {
-      "epoch": 0.4777922177203578,
-      "grad_norm": 0.05970124171786194,
-      "learning_rate": 0.0006246344253396727,
-      "loss": 1.1376,
-      "step": 3125
-    },
-    {
-      "epoch": 0.47855668526871037,
-      "grad_norm": 0.06257524696187984,
-      "learning_rate": 0.0006233417531187914,
-      "loss": 1.1472,
-      "step": 3130
-    },
-    {
-      "epoch": 0.4793211528170629,
-      "grad_norm": 0.0588918841092477,
-      "learning_rate": 0.0006220482024639893,
-      "loss": 1.1095,
-      "step": 3135
-    },
-    {
-      "epoch": 0.48008562036541547,
-      "grad_norm": 0.07088042128141556,
-      "learning_rate": 0.0006207537825878707,
-      "loss": 1.1452,
-      "step": 3140
-    },
-    {
-      "epoch": 0.48085008791376804,
-      "grad_norm": 0.05691526198340547,
-      "learning_rate": 0.0006194585027092306,
-      "loss": 1.141,
-      "step": 3145
-    },
-    {
-      "epoch": 0.4816145554621206,
-      "grad_norm": 0.062009323742541686,
-      "learning_rate": 0.0006181623720529888,
-      "loss": 1.1483,
-      "step": 3150
-    },
-    {
-      "epoch": 0.4823790230104732,
-      "grad_norm": 0.07737737480019008,
-      "learning_rate": 0.0006168653998501242,
-      "loss": 1.1398,
-      "step": 3155
-    },
-    {
-      "epoch": 0.48314349055882577,
-      "grad_norm": 0.0727754305441971,
-      "learning_rate": 0.0006155675953376094,
-      "loss": 1.1404,
-      "step": 3160
-    },
-    {
-      "epoch": 0.48390795810717835,
-      "grad_norm": 0.057391906142971814,
-      "learning_rate": 0.0006142689677583447,
-      "loss": 1.1316,
-      "step": 3165
-    },
-    {
-      "epoch": 0.4846724256555309,
-      "grad_norm": 0.06229610749138481,
-      "learning_rate": 0.0006129695263610915,
-      "loss": 1.1871,
-      "step": 3170
-    },
-    {
-      "epoch": 0.4854368932038835,
-      "grad_norm": 0.09096557645409899,
-      "learning_rate": 0.0006116692804004083,
-      "loss": 1.1198,
-      "step": 3175
-    },
-    {
-      "epoch": 0.4862013607522361,
-      "grad_norm": 0.06325095081093489,
-      "learning_rate": 0.0006103682391365828,
-      "loss": 1.1363,
-      "step": 3180
-    },
-    {
-      "epoch": 0.48696582830058865,
-      "grad_norm": 0.09162697104182367,
-      "learning_rate": 0.0006090664118355673,
-      "loss": 1.1527,
-      "step": 3185
-    },
-    {
-      "epoch": 0.48773029584894123,
-      "grad_norm": 0.06262487814739563,
-      "learning_rate": 0.000607763807768912,
-      "loss": 1.198,
-      "step": 3190
-    },
-    {
-      "epoch": 0.4884947633972938,
-      "grad_norm": 0.06150141308221176,
-      "learning_rate": 0.0006064604362136991,
-      "loss": 1.1479,
-      "step": 3195
-    },
-    {
-      "epoch": 0.4892592309456464,
-      "grad_norm": 0.07871059881475885,
-      "learning_rate": 0.000605156306452477,
-      "loss": 1.1168,
-      "step": 3200
-    },
-    {
-      "epoch": 0.4900236984939989,
-      "grad_norm": 0.06617353704318331,
-      "learning_rate": 0.0006038514277731941,
-      "loss": 1.1265,
-      "step": 3205
-    },
-    {
-      "epoch": 0.4907881660423515,
-      "grad_norm": 0.06533259508339928,
-      "learning_rate": 0.0006025458094691323,
-      "loss": 1.1142,
-      "step": 3210
-    },
-    {
-      "epoch": 0.49155263359070406,
-      "grad_norm": 0.06562349068949161,
-      "learning_rate": 0.0006012394608388411,
-      "loss": 1.1551,
-      "step": 3215
-    },
-    {
-      "epoch": 0.49231710113905663,
-      "grad_norm": 0.07913282984614999,
-      "learning_rate": 0.0005999323911860712,
-      "loss": 1.1805,
-      "step": 3220
-    },
-    {
-      "epoch": 0.4930815686874092,
-      "grad_norm": 0.06329112653333181,
-      "learning_rate": 0.0005986246098197093,
-      "loss": 1.1808,
-      "step": 3225
-    },
-    {
-      "epoch": 0.4938460362357618,
-      "grad_norm": 0.05983365597317458,
-      "learning_rate": 0.0005973161260537095,
-      "loss": 1.1359,
-      "step": 3230
-    },
-    {
-      "epoch": 0.49461050378411436,
-      "grad_norm": 0.06404699549747081,
-      "learning_rate": 0.0005960069492070294,
-      "loss": 1.109,
-      "step": 3235
-    },
-    {
-      "epoch": 0.49537497133246694,
-      "grad_norm": 0.06392702550097112,
-      "learning_rate": 0.0005946970886035625,
-      "loss": 1.1195,
-      "step": 3240
-    },
-    {
-      "epoch": 0.4961394388808195,
-      "grad_norm": 0.06887946962999059,
-      "learning_rate": 0.0005933865535720714,
-      "loss": 1.1338,
-      "step": 3245
-    },
-    {
-      "epoch": 0.4969039064291721,
-      "grad_norm": 0.06321912983691443,
-      "learning_rate": 0.0005920753534461225,
-      "loss": 1.1628,
-      "step": 3250
-    },
-    {
-      "epoch": 0.49766837397752467,
-      "grad_norm": 0.05892992472619733,
-      "learning_rate": 0.0005907634975640191,
-      "loss": 1.1343,
-      "step": 3255
-    },
-    {
-      "epoch": 0.49843284152587725,
-      "grad_norm": 0.08626283038221394,
-      "learning_rate": 0.000589450995268734,
-      "loss": 1.1082,
-      "step": 3260
-    },
-    {
-      "epoch": 0.4991973090742298,
-      "grad_norm": 0.07608683156808083,
-      "learning_rate": 0.0005881378559078448,
-      "loss": 1.0954,
-      "step": 3265
-    },
-    {
-      "epoch": 0.4999617766225824,
-      "grad_norm": 0.06202931308381165,
-      "learning_rate": 0.0005868240888334653,
-      "loss": 1.1253,
-      "step": 3270
-    },
-    {
-      "epoch": 0.5007262441709349,
-      "grad_norm": 0.0693162463919691,
-      "learning_rate": 0.0005855097034021803,
-      "loss": 1.1302,
-      "step": 3275
-    },
-    {
-      "epoch": 0.5014907117192875,
-      "grad_norm": 0.06857586212270887,
-      "learning_rate": 0.0005841947089749782,
-      "loss": 1.1145,
-      "step": 3280
-    },
-    {
-      "epoch": 0.5022551792676401,
-      "grad_norm": 0.056543922942401514,
-      "learning_rate": 0.0005828791149171855,
-      "loss": 1.166,
-      "step": 3285
-    },
-    {
-      "epoch": 0.5030196468159926,
-      "grad_norm": 0.061216299482396534,
-      "learning_rate": 0.0005815629305983979,
-      "loss": 1.1258,
-      "step": 3290
-    },
-    {
-      "epoch": 0.5037841143643452,
-      "grad_norm": 0.08922514630551934,
-      "learning_rate": 0.000580246165392416,
-      "loss": 1.1201,
-      "step": 3295
-    },
-    {
-      "epoch": 0.5045485819126978,
-      "grad_norm": 0.06366534833418323,
-      "learning_rate": 0.000578928828677177,
-      "loss": 1.0926,
-      "step": 3300
-    },
-    {
-      "epoch": 0.5053130494610504,
-      "grad_norm": 0.05868222940320121,
-      "learning_rate": 0.0005776109298346885,
-      "loss": 1.123,
-      "step": 3305
-    },
-    {
-      "epoch": 0.506077517009403,
-      "grad_norm": 0.055139863246591446,
-      "learning_rate": 0.0005762924782509612,
-      "loss": 1.1111,
-      "step": 3310
-    },
-    {
-      "epoch": 0.5068419845577555,
-      "grad_norm": 0.06573454791906676,
-      "learning_rate": 0.0005749734833159429,
-      "loss": 1.113,
-      "step": 3315
-    },
-    {
-      "epoch": 0.5076064521061081,
-      "grad_norm": 0.08427239378689733,
-      "learning_rate": 0.0005736539544234508,
-      "loss": 1.1569,
-      "step": 3320
-    },
-    {
-      "epoch": 0.5083709196544607,
-      "grad_norm": 0.07558080456630767,
-      "learning_rate": 0.0005723339009711051,
-      "loss": 1.1376,
-      "step": 3325
-    },
-    {
-      "epoch": 0.5091353872028133,
-      "grad_norm": 0.05898916837152998,
-      "learning_rate": 0.0005710133323602616,
-      "loss": 1.1007,
-      "step": 3330
-    },
-    {
-      "epoch": 0.5098998547511658,
-      "grad_norm": 0.06278710607065763,
-      "learning_rate": 0.0005696922579959455,
-      "loss": 1.1288,
-      "step": 3335
-    },
-    {
-      "epoch": 0.5106643222995184,
-      "grad_norm": 0.06074976725732031,
-      "learning_rate": 0.0005683706872867833,
-      "loss": 1.131,
-      "step": 3340
-    },
-    {
-      "epoch": 0.511428789847871,
-      "grad_norm": 0.06500965731679485,
-      "learning_rate": 0.0005670486296449372,
-      "loss": 1.1073,
-      "step": 3345
-    },
-    {
-      "epoch": 0.5121932573962236,
-      "grad_norm": 0.059725258913741155,
-      "learning_rate": 0.0005657260944860367,
-      "loss": 1.122,
-      "step": 3350
-    },
-    {
-      "epoch": 0.5129577249445761,
-      "grad_norm": 0.06081393034395375,
-      "learning_rate": 0.0005644030912291125,
-      "loss": 1.1397,
-      "step": 3355
-    },
-    {
-      "epoch": 0.5137221924929287,
-      "grad_norm": 0.06500953558187635,
-      "learning_rate": 0.0005630796292965288,
-      "loss": 1.128,
-      "step": 3360
-    },
-    {
-      "epoch": 0.5144866600412813,
-      "grad_norm": 0.05991814710809413,
-      "learning_rate": 0.0005617557181139169,
-      "loss": 1.115,
-      "step": 3365
-    },
-    {
-      "epoch": 0.5152511275896339,
-      "grad_norm": 0.06774178943497386,
-      "learning_rate": 0.000560431367110107,
-      "loss": 1.1321,
-      "step": 3370
-    },
-    {
-      "epoch": 0.5160155951379863,
-      "grad_norm": 0.06508939987253937,
-      "learning_rate": 0.0005591065857170623,
-      "loss": 1.1377,
-      "step": 3375
-    },
-    {
-      "epoch": 0.5167800626863389,
-      "grad_norm": 0.06522977314137465,
-      "learning_rate": 0.000557781383369811,
-      "loss": 1.117,
-      "step": 3380
-    },
-    {
-      "epoch": 0.5175445302346915,
-      "grad_norm": 0.0636581730317922,
-      "learning_rate": 0.000556455769506379,
-      "loss": 1.0904,
-      "step": 3385
-    },
-    {
-      "epoch": 0.5183089977830441,
-      "grad_norm": 0.05994794288590292,
-      "learning_rate": 0.0005551297535677235,
-      "loss": 1.1361,
-      "step": 3390
-    },
-    {
-      "epoch": 0.5190734653313966,
-      "grad_norm": 0.07212580912491869,
-      "learning_rate": 0.000553803344997665,
-      "loss": 1.1197,
-      "step": 3395
-    },
-    {
-      "epoch": 0.5198379328797492,
-      "grad_norm": 0.06315090246868217,
-      "learning_rate": 0.0005524765532428204,
-      "loss": 1.1205,
-      "step": 3400
-    },
-    {
-      "epoch": 0.5206024004281018,
-      "grad_norm": 0.07345186841532578,
-      "learning_rate": 0.0005511493877525352,
-      "loss": 1.121,
-      "step": 3405
-    },
-    {
-      "epoch": 0.5213668679764544,
-      "grad_norm": 0.05751935815321791,
-      "learning_rate": 0.0005498218579788173,
-      "loss": 1.1279,
-      "step": 3410
-    },
-    {
-      "epoch": 0.522131335524807,
-      "grad_norm": 0.06091507104043262,
-      "learning_rate": 0.0005484939733762687,
-      "loss": 1.1366,
-      "step": 3415
-    },
-    {
-      "epoch": 0.5228958030731595,
-      "grad_norm": 0.059482594137463464,
-      "learning_rate": 0.0005471657434020182,
-      "loss": 1.1536,
-      "step": 3420
-    },
-    {
-      "epoch": 0.5236602706215121,
-      "grad_norm": 0.06510428118556476,
-      "learning_rate": 0.0005458371775156548,
-      "loss": 1.0995,
-      "step": 3425
-    },
-    {
-      "epoch": 0.5244247381698647,
-      "grad_norm": 0.06702378445812333,
-      "learning_rate": 0.0005445082851791597,
-      "loss": 1.0935,
-      "step": 3430
-    },
-    {
-      "epoch": 0.5251892057182173,
-      "grad_norm": 0.06267153627313525,
-      "learning_rate": 0.0005431790758568388,
-      "loss": 1.0971,
-      "step": 3435
-    },
-    {
-      "epoch": 0.5259536732665698,
-      "grad_norm": 0.06690110297685588,
-      "learning_rate": 0.0005418495590152557,
-      "loss": 1.1132,
-      "step": 3440
-    },
-    {
-      "epoch": 0.5267181408149224,
-      "grad_norm": 0.06857945644968291,
-      "learning_rate": 0.0005405197441231645,
-      "loss": 1.129,
-      "step": 3445
-    },
-    {
-      "epoch": 0.527482608363275,
-      "grad_norm": 0.0679067326661273,
-      "learning_rate": 0.0005391896406514414,
-      "loss": 1.1197,
-      "step": 3450
-    },
-    {
-      "epoch": 0.5282470759116276,
-      "grad_norm": 0.06495452970096306,
-      "learning_rate": 0.0005378592580730182,
-      "loss": 1.1191,
-      "step": 3455
-    },
-    {
-      "epoch": 0.5290115434599801,
-      "grad_norm": 0.07011699388524864,
-      "learning_rate": 0.0005365286058628144,
-      "loss": 1.1244,
-      "step": 3460
-    },
-    {
-      "epoch": 0.5297760110083327,
-      "grad_norm": 0.08305004226196393,
-      "learning_rate": 0.0005351976934976702,
-      "loss": 1.1122,
-      "step": 3465
-    },
-    {
-      "epoch": 0.5305404785566853,
-      "grad_norm": 0.05964304908716454,
-      "learning_rate": 0.0005338665304562776,
-      "loss": 1.0962,
-      "step": 3470
-    },
-    {
-      "epoch": 0.5313049461050379,
-      "grad_norm": 0.06608417735470593,
-      "learning_rate": 0.0005325351262191149,
-      "loss": 1.1091,
-      "step": 3475
-    },
-    {
-      "epoch": 0.5320694136533904,
-      "grad_norm": 0.06157287568970666,
-      "learning_rate": 0.0005312034902683779,
-      "loss": 1.1544,
-      "step": 3480
-    },
-    {
-      "epoch": 0.532833881201743,
-      "grad_norm": 0.0604884131717689,
-      "learning_rate": 0.0005298716320879124,
-      "loss": 1.1189,
-      "step": 3485
-    },
-    {
-      "epoch": 0.5335983487500956,
-      "grad_norm": 0.07023563410010202,
-      "learning_rate": 0.000528539561163147,
-      "loss": 1.0744,
-      "step": 3490
-    },
-    {
-      "epoch": 0.5343628162984482,
-      "grad_norm": 0.06000529959619642,
-      "learning_rate": 0.000527207286981026,
-      "loss": 1.1147,
-      "step": 3495
-    },
-    {
-      "epoch": 0.5351272838468007,
-      "grad_norm": 0.06234895133269599,
-      "learning_rate": 0.0005258748190299404,
-      "loss": 1.0933,
-      "step": 3500
-    },
-    {
-      "epoch": 0.5358917513951533,
-      "grad_norm": 0.0631654264689642,
-      "learning_rate": 0.0005245421667996618,
-      "loss": 1.151,
-      "step": 3505
-    },
-    {
-      "epoch": 0.5366562189435059,
-      "grad_norm": 0.06627341764797444,
-      "learning_rate": 0.000523209339781274,
-      "loss": 1.0995,
-      "step": 3510
-    },
-    {
-      "epoch": 0.5374206864918584,
-      "grad_norm": 0.07201555609957591,
-      "learning_rate": 0.0005218763474671058,
-      "loss": 1.0883,
-      "step": 3515
-    },
-    {
-      "epoch": 0.538185154040211,
-      "grad_norm": 0.0713275731303358,
-      "learning_rate": 0.000520543199350663,
-      "loss": 1.105,
-      "step": 3520
-    },
-    {
-      "epoch": 0.5389496215885635,
-      "grad_norm": 0.06821865094908078,
-      "learning_rate": 0.0005192099049265613,
-      "loss": 1.0769,
-      "step": 3525
-    },
-    {
-      "epoch": 0.5397140891369161,
-      "grad_norm": 0.06557378936774873,
-      "learning_rate": 0.0005178764736904582,
-      "loss": 1.1454,
-      "step": 3530
-    },
-    {
-      "epoch": 0.5404785566852687,
-      "grad_norm": 0.06045084371867397,
-      "learning_rate": 0.0005165429151389856,
-      "loss": 1.105,
-      "step": 3535
-    },
-    {
-      "epoch": 0.5412430242336213,
-      "grad_norm": 0.06242346673470062,
-      "learning_rate": 0.0005152092387696821,
-      "loss": 1.1108,
-      "step": 3540
-    },
-    {
-      "epoch": 0.5420074917819738,
-      "grad_norm": 0.06172204231908025,
-      "learning_rate": 0.0005138754540809253,
-      "loss": 1.1299,
-      "step": 3545
-    },
-    {
-      "epoch": 0.5427719593303264,
-      "grad_norm": 0.05825804942006388,
-      "learning_rate": 0.0005125415705718646,
-      "loss": 1.1023,
-      "step": 3550
-    },
-    {
-      "epoch": 0.543536426878679,
-      "grad_norm": 0.06796344310929125,
-      "learning_rate": 0.0005112075977423532,
-      "loss": 1.1216,
-      "step": 3555
-    },
-    {
-      "epoch": 0.5443008944270316,
-      "grad_norm": 0.058632756425142235,
-      "learning_rate": 0.0005098735450928799,
-      "loss": 1.1226,
-      "step": 3560
-    },
-    {
-      "epoch": 0.5450653619753841,
-      "grad_norm": 0.06265232814361309,
-      "learning_rate": 0.0005085394221245021,
-      "loss": 1.1166,
-      "step": 3565
-    },
-    {
-      "epoch": 0.5458298295237367,
-      "grad_norm": 0.06577998911234535,
-      "learning_rate": 0.0005072052383387787,
-      "loss": 1.0894,
-      "step": 3570
-    },
-    {
-      "epoch": 0.5465942970720893,
-      "grad_norm": 0.07545067157181294,
-      "learning_rate": 0.0005058710032377008,
-      "loss": 1.1264,
-      "step": 3575
-    },
-    {
-      "epoch": 0.5473587646204419,
-      "grad_norm": 0.07311060599396883,
-      "learning_rate": 0.0005045367263236257,
-      "loss": 1.1257,
-      "step": 3580
-    },
-    {
-      "epoch": 0.5481232321687944,
-      "grad_norm": 0.06148576987581805,
-      "learning_rate": 0.000503202417099208,
-      "loss": 1.1184,
-      "step": 3585
-    },
-    {
-      "epoch": 0.548887699717147,
-      "grad_norm": 0.06467766525296764,
-      "learning_rate": 0.0005018680850673327,
-      "loss": 1.1132,
-      "step": 3590
-    },
-    {
-      "epoch": 0.5496521672654996,
-      "grad_norm": 0.06393600821432976,
-      "learning_rate": 0.0005005337397310469,
-      "loss": 1.1094,
-      "step": 3595
-    },
-    {
-      "epoch": 0.5504166348138522,
-      "grad_norm": 0.0621691916365975,
-      "learning_rate": 0.0004991993905934931,
-      "loss": 1.1361,
-      "step": 3600
-    },
-    {
-      "epoch": 0.5511811023622047,
-      "grad_norm": 0.06724351675549163,
-      "learning_rate": 0.0004978650471578402,
-      "loss": 1.1147,
-      "step": 3605
-    },
-    {
-      "epoch": 0.5519455699105573,
-      "grad_norm": 0.05679252783616519,
-      "learning_rate": 0.000496530718927217,
-      "loss": 1.1256,
-      "step": 3610
-    },
-    {
-      "epoch": 0.5527100374589099,
-      "grad_norm": 0.057881098463542474,
-      "learning_rate": 0.0004951964154046432,
-      "loss": 1.1106,
-      "step": 3615
-    },
-    {
-      "epoch": 0.5534745050072625,
-      "grad_norm": 0.06292887711056025,
-      "learning_rate": 0.0004938621460929639,
-      "loss": 1.1134,
-      "step": 3620
-    },
-    {
-      "epoch": 0.554238972555615,
-      "grad_norm": 0.06590848453173402,
-      "learning_rate": 0.0004925279204947789,
-      "loss": 1.1275,
-      "step": 3625
-    },
-    {
-      "epoch": 0.5550034401039676,
-      "grad_norm": 0.06325873950215483,
-      "learning_rate": 0.0004911937481123783,
-      "loss": 1.09,
-      "step": 3630
-    },
-    {
-      "epoch": 0.5557679076523202,
-      "grad_norm": 0.05608984482155,
-      "learning_rate": 0.0004898596384476718,
-      "loss": 1.1412,
-      "step": 3635
-    },
-    {
-      "epoch": 0.5565323752006728,
-      "grad_norm": 0.07138390520772304,
-      "learning_rate": 0.0004885256010021232,
-      "loss": 1.087,
-      "step": 3640
-    },
-    {
-      "epoch": 0.5572968427490254,
-      "grad_norm": 0.06847884401062614,
-      "learning_rate": 0.0004871916452766822,
-      "loss": 1.0987,
-      "step": 3645
-    },
-    {
-      "epoch": 0.5580613102973778,
-      "grad_norm": 0.06778957445525216,
-      "learning_rate": 0.00048585778077171586,
-      "loss": 1.0987,
-      "step": 3650
-    },
-    {
-      "epoch": 0.5588257778457304,
-      "grad_norm": 0.062458614692809604,
-      "learning_rate": 0.00048452401698694154,
-      "loss": 1.0803,
-      "step": 3655
-    },
-    {
-      "epoch": 0.559590245394083,
-      "grad_norm": 0.06609876186419025,
-      "learning_rate": 0.00048319036342135983,
-      "loss": 1.1123,
-      "step": 3660
-    },
-    {
-      "epoch": 0.5603547129424356,
-      "grad_norm": 0.06691503571502525,
-      "learning_rate": 0.00048185682957318604,
-      "loss": 1.0857,
-      "step": 3665
-    },
-    {
-      "epoch": 0.5611191804907881,
-      "grad_norm": 0.06633733246653169,
-      "learning_rate": 0.00048052342493978275,
-      "loss": 1.1082,
-      "step": 3670
-    },
-    {
-      "epoch": 0.5618836480391407,
-      "grad_norm": 0.06504840824047256,
-      "learning_rate": 0.0004791901590175926,
-      "loss": 1.1154,
-      "step": 3675
-    },
-    {
-      "epoch": 0.5626481155874933,
-      "grad_norm": 0.0603761920675834,
-      "learning_rate": 0.0004778570413020702,
-      "loss": 1.0871,
-      "step": 3680
-    },
-    {
-      "epoch": 0.5634125831358459,
-      "grad_norm": 0.058914355257856044,
-      "learning_rate": 0.0004765240812876141,
-      "loss": 1.1103,
-      "step": 3685
-    },
-    {
-      "epoch": 0.5641770506841984,
-      "grad_norm": 0.06502988982340185,
-      "learning_rate": 0.00047519128846750083,
-      "loss": 1.15,
-      "step": 3690
-    },
-    {
-      "epoch": 0.564941518232551,
-      "grad_norm": 0.07280984153816127,
-      "learning_rate": 0.00047385867233381516,
-      "loss": 1.1121,
-      "step": 3695
-    },
-    {
-      "epoch": 0.5657059857809036,
-      "grad_norm": 0.06994408916357317,
-      "learning_rate": 0.0004725262423773838,
-      "loss": 1.1293,
-      "step": 3700
-    },
-    {
-      "epoch": 0.5664704533292562,
-      "grad_norm": 0.0621673770156963,
-      "learning_rate": 0.0004711940080877079,
-      "loss": 1.0852,
-      "step": 3705
-    },
-    {
-      "epoch": 0.5672349208776087,
-      "grad_norm": 0.05817651318162582,
-      "learning_rate": 0.0004698619789528948,
-      "loss": 1.1301,
-      "step": 3710
-    },
-    {
-      "epoch": 0.5679993884259613,
-      "grad_norm": 0.06081671877206097,
-      "learning_rate": 0.00046853016445959014,
-      "loss": 1.1398,
-      "step": 3715
-    },
-    {
-      "epoch": 0.5687638559743139,
-      "grad_norm": 0.06188571604259829,
-      "learning_rate": 0.00046719857409291226,
-      "loss": 1.1731,
-      "step": 3720
-    },
-    {
-      "epoch": 0.5695283235226665,
-      "grad_norm": 0.06243919731382698,
-      "learning_rate": 0.00046586721733638204,
-      "loss": 1.0922,
-      "step": 3725
-    },
-    {
-      "epoch": 0.570292791071019,
-      "grad_norm": 0.06337869980709998,
-      "learning_rate": 0.00046453610367185695,
-      "loss": 1.1325,
-      "step": 3730
-    },
-    {
-      "epoch": 0.5710572586193716,
-      "grad_norm": 0.06302625565447471,
-      "learning_rate": 0.00046320524257946363,
-      "loss": 1.0613,
-      "step": 3735
-    },
-    {
-      "epoch": 0.5718217261677242,
-      "grad_norm": 0.06598219655258462,
-      "learning_rate": 0.00046187464353752945,
-      "loss": 1.0962,
-      "step": 3740
-    },
-    {
-      "epoch": 0.5725861937160768,
-      "grad_norm": 0.06698025179482835,
-      "learning_rate": 0.0004605443160225152,
-      "loss": 1.1107,
-      "step": 3745
-    },
-    {
-      "epoch": 0.5733506612644294,
-      "grad_norm": 0.06756310322497935,
-      "learning_rate": 0.00045921426950894885,
-      "loss": 1.1511,
-      "step": 3750
-    },
-    {
-      "epoch": 0.5741151288127819,
-      "grad_norm": 0.06396486586356562,
-      "learning_rate": 0.00045788451346935605,
-      "loss": 1.0729,
-      "step": 3755
-    },
-    {
-      "epoch": 0.5748795963611345,
-      "grad_norm": 0.0732526041311009,
-      "learning_rate": 0.0004565550573741942,
-      "loss": 1.1073,
-      "step": 3760
-    },
-    {
-      "epoch": 0.5756440639094871,
-      "grad_norm": 0.06998612377117433,
-      "learning_rate": 0.0004552259106917846,
-      "loss": 1.128,
-      "step": 3765
-    },
-    {
-      "epoch": 0.5764085314578397,
-      "grad_norm": 0.06616565417315289,
-      "learning_rate": 0.0004538970828882447,
-      "loss": 1.0956,
-      "step": 3770
-    },
-    {
-      "epoch": 0.5771729990061922,
-      "grad_norm": 0.06108422217657452,
-      "learning_rate": 0.00045256858342742094,
-      "loss": 1.0987,
-      "step": 3775
-    },
-    {
-      "epoch": 0.5779374665545448,
-      "grad_norm": 0.05824837749243198,
-      "learning_rate": 0.0004512404217708217,
-      "loss": 1.0806,
-      "step": 3780
-    },
-    {
-      "epoch": 0.5787019341028974,
-      "grad_norm": 0.06266413521523766,
-      "learning_rate": 0.00044991260737754877,
-      "loss": 1.102,
-      "step": 3785
-    },
-    {
-      "epoch": 0.5794664016512499,
-      "grad_norm": 0.060441791106991334,
-      "learning_rate": 0.0004485851497042312,
-      "loss": 1.0884,
-      "step": 3790
-    },
-    {
-      "epoch": 0.5802308691996024,
-      "grad_norm": 0.060797439754043565,
-      "learning_rate": 0.0004472580582049578,
-      "loss": 1.098,
-      "step": 3795
-    },
-    {
-      "epoch": 0.580995336747955,
-      "grad_norm": 0.062023942489464336,
-      "learning_rate": 0.000445931342331209,
-      "loss": 1.0919,
-      "step": 3800
-    },
-    {
-      "epoch": 0.5817598042963076,
-      "grad_norm": 0.06418430339347835,
-      "learning_rate": 0.00044460501153179016,
-      "loss": 1.1017,
-      "step": 3805
-    },
-    {
-      "epoch": 0.5825242718446602,
-      "grad_norm": 0.060619213747711925,
-      "learning_rate": 0.00044327907525276456,
-      "loss": 1.0849,
-      "step": 3810
-    },
-    {
-      "epoch": 0.5832887393930127,
-      "grad_norm": 0.057326188671437316,
-      "learning_rate": 0.0004419535429373848,
-      "loss": 1.1172,
-      "step": 3815
-    },
-    {
-      "epoch": 0.5840532069413653,
-      "grad_norm": 0.06607758733161419,
-      "learning_rate": 0.00044062842402602776,
-      "loss": 1.0928,
-      "step": 3820
-    },
-    {
-      "epoch": 0.5848176744897179,
-      "grad_norm": 0.07029009256072538,
-      "learning_rate": 0.000439303727956125,
-      "loss": 1.095,
-      "step": 3825
-    },
-    {
-      "epoch": 0.5855821420380705,
-      "grad_norm": 0.0710751111387438,
-      "learning_rate": 0.0004379794641620969,
-      "loss": 1.0881,
-      "step": 3830
-    },
-    {
-      "epoch": 0.586346609586423,
-      "grad_norm": 0.06321187821581618,
-      "learning_rate": 0.00043665564207528556,
-      "loss": 1.1249,
-      "step": 3835
-    },
-    {
-      "epoch": 0.5871110771347756,
-      "grad_norm": 0.06167451377736427,
-      "learning_rate": 0.0004353322711238869,
-      "loss": 1.1137,
-      "step": 3840
-    },
-    {
-      "epoch": 0.5878755446831282,
-      "grad_norm": 0.060565668504396965,
-      "learning_rate": 0.0004340093607328839,
-      "loss": 1.1036,
-      "step": 3845
-    },
-    {
-      "epoch": 0.5886400122314808,
-      "grad_norm": 0.0622976248859124,
-      "learning_rate": 0.00043268692032397984,
-      "loss": 1.0913,
-      "step": 3850
-    },
-    {
-      "epoch": 0.5894044797798333,
-      "grad_norm": 0.06391396857013981,
-      "learning_rate": 0.0004313649593155301,
-      "loss": 1.0847,
-      "step": 3855
-    },
-    {
-      "epoch": 0.5901689473281859,
-      "grad_norm": 0.07423916257372261,
-      "learning_rate": 0.0004300434871224763,
-      "loss": 1.0901,
-      "step": 3860
-    },
-    {
-      "epoch": 0.5909334148765385,
-      "grad_norm": 0.062338256577621556,
-      "learning_rate": 0.00042872251315627884,
-      "loss": 1.0926,
-      "step": 3865
-    },
-    {
-      "epoch": 0.5916978824248911,
-      "grad_norm": 0.06178660389841415,
-      "learning_rate": 0.0004274020468248494,
-      "loss": 1.0754,
-      "step": 3870
-    },
-    {
-      "epoch": 0.5924623499732437,
-      "grad_norm": 0.058847696827950484,
-      "learning_rate": 0.0004260820975324844,
-      "loss": 1.0788,
-      "step": 3875
-    },
-    {
-      "epoch": 0.5932268175215962,
-      "grad_norm": 0.061443746079662805,
-      "learning_rate": 0.00042476267467979827,
-      "loss": 1.0796,
-      "step": 3880
-    },
-    {
-      "epoch": 0.5939912850699488,
-      "grad_norm": 0.06810138628824276,
-      "learning_rate": 0.0004234437876636557,
-      "loss": 1.0977,
-      "step": 3885
-    },
-    {
-      "epoch": 0.5947557526183014,
-      "grad_norm": 0.07494921338482857,
-      "learning_rate": 0.00042212544587710534,
-      "loss": 1.0967,
-      "step": 3890
-    },
-    {
-      "epoch": 0.595520220166654,
-      "grad_norm": 0.06867273849143388,
-      "learning_rate": 0.00042080765870931294,
-      "loss": 1.1117,
-      "step": 3895
-    },
-    {
-      "epoch": 0.5962846877150065,
-      "grad_norm": 0.06708271782909088,
-      "learning_rate": 0.00041949043554549405,
-      "loss": 1.0985,
-      "step": 3900
-    },
-    {
-      "epoch": 0.5970491552633591,
-      "grad_norm": 0.05385745159990605,
-      "learning_rate": 0.00041817378576684746,
-      "loss": 1.1013,
-      "step": 3905
-    },
-    {
-      "epoch": 0.5978136228117117,
-      "grad_norm": 0.059032475231314455,
-      "learning_rate": 0.0004168577187504884,
-      "loss": 1.0984,
-      "step": 3910
-    },
-    {
-      "epoch": 0.5985780903600643,
-      "grad_norm": 0.0527291048836883,
-      "learning_rate": 0.00041554224386938186,
-      "loss": 1.0873,
-      "step": 3915
-    },
-    {
-      "epoch": 0.5993425579084168,
-      "grad_norm": 0.06282472550713358,
-      "learning_rate": 0.00041422737049227496,
-      "loss": 1.1133,
-      "step": 3920
-    },
-    {
-      "epoch": 0.6001070254567693,
-      "grad_norm": 0.06637056841657221,
-      "learning_rate": 0.000412913107983632,
-      "loss": 1.0733,
-      "step": 3925
-    },
-    {
-      "epoch": 0.6008714930051219,
-      "grad_norm": 0.07134078790379195,
-      "learning_rate": 0.00041159946570356584,
-      "loss": 1.0589,
-      "step": 3930
-    },
-    {
-      "epoch": 0.6016359605534745,
-      "grad_norm": 0.059395620562726396,
-      "learning_rate": 0.0004102864530077724,
-      "loss": 1.0624,
-      "step": 3935
-    },
-    {
-      "epoch": 0.602400428101827,
-      "grad_norm": 0.059597386222636405,
-      "learning_rate": 0.000408974079247464,
-      "loss": 1.1008,
-      "step": 3940
-    },
-    {
-      "epoch": 0.6031648956501796,
-      "grad_norm": 0.057839152335023306,
-      "learning_rate": 0.00040766235376930214,
-      "loss": 1.1103,
-      "step": 3945
-    },
-    {
-      "epoch": 0.6039293631985322,
-      "grad_norm": 0.061849685617478466,
-      "learning_rate": 0.0004063512859153311,
-      "loss": 1.0816,
-      "step": 3950
-    },
-    {
-      "epoch": 0.6046938307468848,
-      "grad_norm": 0.06508877768701075,
-      "learning_rate": 0.00040504088502291234,
-      "loss": 1.1081,
-      "step": 3955
-    },
-    {
-      "epoch": 0.6054582982952373,
-      "grad_norm": 0.07683797410703767,
-      "learning_rate": 0.00040373116042465647,
-      "loss": 1.1111,
-      "step": 3960
-    },
-    {
-      "epoch": 0.6062227658435899,
-      "grad_norm": 0.06757768507720088,
-      "learning_rate": 0.0004024221214483579,
-      "loss": 1.0925,
-      "step": 3965
-    },
-    {
-      "epoch": 0.6069872333919425,
-      "grad_norm": 0.06593559024726058,
-      "learning_rate": 0.00040111377741692834,
-      "loss": 1.0891,
-      "step": 3970
-    },
-    {
-      "epoch": 0.6077517009402951,
-      "grad_norm": 0.05952734135732054,
-      "learning_rate": 0.00039980613764832975,
-      "loss": 1.0705,
-      "step": 3975
-    },
-    {
-      "epoch": 0.6085161684886476,
-      "grad_norm": 0.06174361114814803,
-      "learning_rate": 0.00039849921145550807,
-      "loss": 1.1021,
-      "step": 3980
-    },
-    {
-      "epoch": 0.6092806360370002,
-      "grad_norm": 0.07045898974414644,
-      "learning_rate": 0.00039719300814632823,
-      "loss": 1.1107,
-      "step": 3985
-    },
-    {
-      "epoch": 0.6100451035853528,
-      "grad_norm": 0.059203133295731424,
-      "learning_rate": 0.00039588753702350584,
-      "loss": 1.1137,
-      "step": 3990
-    },
-    {
-      "epoch": 0.6108095711337054,
-      "grad_norm": 0.07049531573486653,
-      "learning_rate": 0.00039458280738454213,
-      "loss": 1.0764,
-      "step": 3995
-    },
-    {
-      "epoch": 0.611574038682058,
-      "grad_norm": 0.06407896548037507,
-      "learning_rate": 0.0003932788285216579,
-      "loss": 1.0802,
-      "step": 4000
-    },
-    {
-      "epoch": 0.6123385062304105,
-      "grad_norm": 0.062705551728603,
-      "learning_rate": 0.0003919756097217266,
-      "loss": 1.0549,
-      "step": 4005
-    },
-    {
-      "epoch": 0.6131029737787631,
-      "grad_norm": 0.06559872784175859,
-      "learning_rate": 0.0003906731602662087,
-      "loss": 1.1088,
-      "step": 4010
-    },
-    {
-      "epoch": 0.6138674413271157,
-      "grad_norm": 0.06427225726298613,
-      "learning_rate": 0.0003893714894310855,
-      "loss": 1.0621,
-      "step": 4015
-    },
-    {
-      "epoch": 0.6146319088754683,
-      "grad_norm": 0.06406941756324243,
-      "learning_rate": 0.0003880706064867926,
-      "loss": 1.0855,
-      "step": 4020
-    },
-    {
-      "epoch": 0.6153963764238208,
-      "grad_norm": 0.06576877266118943,
-      "learning_rate": 0.00038677052069815477,
-      "loss": 1.0793,
-      "step": 4025
-    },
-    {
-      "epoch": 0.6161608439721734,
-      "grad_norm": 0.06459197784453244,
-      "learning_rate": 0.0003854712413243192,
-      "loss": 1.091,
-      "step": 4030
-    },
-    {
-      "epoch": 0.616925311520526,
-      "grad_norm": 0.05811327949041547,
-      "learning_rate": 0.0003841727776186899,
-      "loss": 1.08,
-      "step": 4035
-    },
-    {
-      "epoch": 0.6176897790688786,
-      "grad_norm": 0.07215085049566354,
-      "learning_rate": 0.00038287513882886195,
-      "loss": 1.0611,
-      "step": 4040
-    },
-    {
-      "epoch": 0.6184542466172311,
-      "grad_norm": 0.06252405374750149,
-      "learning_rate": 0.00038157833419655507,
-      "loss": 1.088,
-      "step": 4045
-    },
-    {
-      "epoch": 0.6192187141655837,
-      "grad_norm": 0.06008502958042993,
-      "learning_rate": 0.00038028237295754805,
-      "loss": 1.1198,
-      "step": 4050
-    },
-    {
-      "epoch": 0.6199831817139363,
-      "grad_norm": 0.06330961932652555,
-      "learning_rate": 0.0003789872643416138,
-      "loss": 1.0562,
-      "step": 4055
-    },
-    {
-      "epoch": 0.6207476492622889,
-      "grad_norm": 0.058986400518205764,
-      "learning_rate": 0.00037769301757245214,
-      "loss": 1.1104,
-      "step": 4060
-    },
-    {
-      "epoch": 0.6215121168106413,
-      "grad_norm": 0.05855844822094068,
-      "learning_rate": 0.00037639964186762505,
-      "loss": 1.0794,
-      "step": 4065
-    },
-    {
-      "epoch": 0.6222765843589939,
-      "grad_norm": 0.06760118473112167,
-      "learning_rate": 0.00037510714643849107,
-      "loss": 1.0467,
-      "step": 4070
-    },
-    {
-      "epoch": 0.6230410519073465,
-      "grad_norm": 0.05867534449074983,
-      "learning_rate": 0.00037381554049013946,
-      "loss": 1.0938,
-      "step": 4075
-    },
-    {
-      "epoch": 0.6238055194556991,
-      "grad_norm": 0.07682190561405723,
-      "learning_rate": 0.00037252483322132386,
-      "loss": 1.0884,
-      "step": 4080
-    },
-    {
-      "epoch": 0.6245699870040516,
-      "grad_norm": 0.07229420016537297,
-      "learning_rate": 0.00037123503382439894,
-      "loss": 1.0984,
-      "step": 4085
-    },
-    {
-      "epoch": 0.6253344545524042,
-      "grad_norm": 0.06570366560388953,
-      "learning_rate": 0.00036994615148525236,
-      "loss": 1.0887,
-      "step": 4090
-    },
-    {
-      "epoch": 0.6260989221007568,
-      "grad_norm": 0.061864508617144216,
-      "learning_rate": 0.00036865819538324095,
-      "loss": 1.1112,
-      "step": 4095
-    },
-    {
-      "epoch": 0.6268633896491094,
-      "grad_norm": 0.0656439941017718,
-      "learning_rate": 0.0003673711746911252,
-      "loss": 1.0741,
-      "step": 4100
-    },
-    {
-      "epoch": 0.627627857197462,
-      "grad_norm": 0.055905760438808745,
-      "learning_rate": 0.00036608509857500335,
-      "loss": 1.0459,
-      "step": 4105
-    },
-    {
-      "epoch": 0.6283923247458145,
-      "grad_norm": 0.06674862047297604,
-      "learning_rate": 0.00036479997619424606,
-      "loss": 1.0918,
-      "step": 4110
-    },
-    {
-      "epoch": 0.6291567922941671,
-      "grad_norm": 0.060859055527957194,
-      "learning_rate": 0.0003635158167014326,
-      "loss": 1.0804,
-      "step": 4115
-    },
-    {
-      "epoch": 0.6299212598425197,
-      "grad_norm": 0.06537100475249158,
-      "learning_rate": 0.0003622326292422834,
-      "loss": 1.0847,
-      "step": 4120
-    },
-    {
-      "epoch": 0.6306857273908723,
-      "grad_norm": 0.06227343013649914,
-      "learning_rate": 0.0003609504229555969,
-      "loss": 1.0853,
-      "step": 4125
-    },
-    {
-      "epoch": 0.6314501949392248,
-      "grad_norm": 0.06841807299363888,
-      "learning_rate": 0.0003596692069731835,
-      "loss": 1.0578,
-      "step": 4130
-    },
-    {
-      "epoch": 0.6322146624875774,
-      "grad_norm": 0.0736451804635189,
-      "learning_rate": 0.00035838899041980055,
-      "loss": 1.0817,
-      "step": 4135
-    },
-    {
-      "epoch": 0.63297913003593,
-      "grad_norm": 0.05488620850993582,
-      "learning_rate": 0.0003571097824130873,
-      "loss": 1.0921,
-      "step": 4140
-    },
-    {
-      "epoch": 0.6337435975842826,
-      "grad_norm": 0.06282145805968094,
-      "learning_rate": 0.0003558315920635011,
-      "loss": 1.0378,
-      "step": 4145
-    },
-    {
-      "epoch": 0.6345080651326351,
-      "grad_norm": 0.06445119265430188,
-      "learning_rate": 0.0003545544284742506,
-      "loss": 1.0929,
-      "step": 4150
-    },
-    {
-      "epoch": 0.6352725326809877,
-      "grad_norm": 0.056909601872114204,
-      "learning_rate": 0.0003532783007412321,
-      "loss": 1.1056,
-      "step": 4155
-    },
-    {
-      "epoch": 0.6360370002293403,
-      "grad_norm": 0.060124580046915954,
-      "learning_rate": 0.0003520032179529652,
-      "loss": 1.1178,
-      "step": 4160
-    },
-    {
-      "epoch": 0.6368014677776929,
-      "grad_norm": 0.06314252699342483,
-      "learning_rate": 0.00035072918919052683,
-      "loss": 1.0837,
-      "step": 4165
-    },
-    {
-      "epoch": 0.6375659353260454,
-      "grad_norm": 0.06686385252802637,
-      "learning_rate": 0.0003494562235274873,
-      "loss": 1.0949,
-      "step": 4170
-    },
-    {
-      "epoch": 0.638330402874398,
-      "grad_norm": 0.06101338870198361,
-      "learning_rate": 0.0003481843300298459,
-      "loss": 1.0786,
-      "step": 4175
-    },
-    {
-      "epoch": 0.6390948704227506,
-      "grad_norm": 0.060474410756929865,
-      "learning_rate": 0.0003469135177559656,
-      "loss": 1.1157,
-      "step": 4180
-    },
-    {
-      "epoch": 0.6398593379711032,
-      "grad_norm": 0.06235109250261373,
-      "learning_rate": 0.00034564379575650916,
-      "loss": 1.0814,
-      "step": 4185
-    },
-    {
-      "epoch": 0.6406238055194557,
-      "grad_norm": 0.07102337030993396,
-      "learning_rate": 0.0003443751730743745,
-      "loss": 1.0545,
-      "step": 4190
-    },
-    {
-      "epoch": 0.6413882730678083,
-      "grad_norm": 0.06922307618811946,
-      "learning_rate": 0.00034310765874463007,
-      "loss": 1.0813,
-      "step": 4195
-    },
-    {
-      "epoch": 0.6421527406161608,
-      "grad_norm": 0.06734659315916373,
-      "learning_rate": 0.000341841261794451,
-      "loss": 1.0821,
-      "step": 4200
-    },
-    {
-      "epoch": 0.6429172081645134,
-      "grad_norm": 0.07173326298860931,
-      "learning_rate": 0.0003405759912430543,
-      "loss": 1.0853,
-      "step": 4205
-    },
-    {
-      "epoch": 0.6436816757128659,
-      "grad_norm": 0.060997631416030235,
-      "learning_rate": 0.00033931185610163437,
-      "loss": 1.0501,
-      "step": 4210
-    },
-    {
-      "epoch": 0.6444461432612185,
-      "grad_norm": 0.0678413835009176,
-      "learning_rate": 0.0003380488653733004,
-      "loss": 1.0596,
-      "step": 4215
-    },
-    {
-      "epoch": 0.6452106108095711,
-      "grad_norm": 0.06764818450679011,
-      "learning_rate": 0.00033678702805301007,
-      "loss": 1.1031,
-      "step": 4220
-    },
-    {
-      "epoch": 0.6459750783579237,
-      "grad_norm": 0.06064356181998237,
-      "learning_rate": 0.0003355263531275069,
-      "loss": 1.0926,
-      "step": 4225
-    },
-    {
-      "epoch": 0.6467395459062762,
-      "grad_norm": 0.07643917264349942,
-      "learning_rate": 0.0003342668495752561,
-      "loss": 1.102,
-      "step": 4230
-    },
-    {
-      "epoch": 0.6475040134546288,
-      "grad_norm": 0.06014169982393995,
-      "learning_rate": 0.00033300852636638026,
-      "loss": 1.097,
-      "step": 4235
-    },
-    {
-      "epoch": 0.6482684810029814,
-      "grad_norm": 0.07020185439225457,
-      "learning_rate": 0.00033175139246259533,
-      "loss": 1.081,
-      "step": 4240
-    },
-    {
-      "epoch": 0.649032948551334,
-      "grad_norm": 0.061835432421758194,
-      "learning_rate": 0.0003304954568171478,
-      "loss": 1.151,
-      "step": 4245
-    },
-    {
-      "epoch": 0.6497974160996866,
-      "grad_norm": 0.05936020939229027,
-      "learning_rate": 0.00032924072837474936,
-      "loss": 1.06,
-      "step": 4250
-    },
-    {
-      "epoch": 0.6505618836480391,
-      "grad_norm": 0.06073312626281398,
-      "learning_rate": 0.0003279872160715146,
-      "loss": 1.0694,
-      "step": 4255
-    },
-    {
-      "epoch": 0.6513263511963917,
-      "grad_norm": 0.05551866637743777,
-      "learning_rate": 0.00032673492883489693,
-      "loss": 1.0963,
-      "step": 4260
-    },
-    {
-      "epoch": 0.6520908187447443,
-      "grad_norm": 0.057095759169672695,
-      "learning_rate": 0.0003254838755836245,
-      "loss": 1.0736,
-      "step": 4265
-    },
-    {
-      "epoch": 0.6528552862930969,
-      "grad_norm": 0.05824058789550292,
-      "learning_rate": 0.0003242340652276371,
-      "loss": 1.0632,
-      "step": 4270
-    },
-    {
-      "epoch": 0.6536197538414494,
-      "grad_norm": 0.06447238459426781,
-      "learning_rate": 0.00032298550666802315,
-      "loss": 1.0641,
-      "step": 4275
-    },
-    {
-      "epoch": 0.654384221389802,
-      "grad_norm": 0.07348007672974653,
-      "learning_rate": 0.0003217382087969554,
-      "loss": 1.0919,
-      "step": 4280
-    },
-    {
-      "epoch": 0.6551486889381546,
-      "grad_norm": 0.06202495691783003,
-      "learning_rate": 0.000320492180497628,
-      "loss": 1.1166,
-      "step": 4285
-    },
-    {
-      "epoch": 0.6559131564865072,
-      "grad_norm": 0.05877415292293241,
-      "learning_rate": 0.0003192474306441936,
-      "loss": 1.045,
-      "step": 4290
-    },
-    {
-      "epoch": 0.6566776240348597,
-      "grad_norm": 0.05983915724082635,
-      "learning_rate": 0.00031800396810169963,
-      "loss": 1.0807,
-      "step": 4295
-    },
-    {
-      "epoch": 0.6574420915832123,
-      "grad_norm": 0.06659604980593338,
-      "learning_rate": 0.00031676180172602523,
-      "loss": 1.101,
-      "step": 4300
-    },
-    {
-      "epoch": 0.6582065591315649,
-      "grad_norm": 0.06638853925039498,
-      "learning_rate": 0.00031552094036381873,
-      "loss": 1.0828,
-      "step": 4305
-    },
-    {
-      "epoch": 0.6589710266799175,
-      "grad_norm": 0.06147261804251727,
-      "learning_rate": 0.0003142813928524338,
-      "loss": 1.0734,
-      "step": 4310
-    },
-    {
-      "epoch": 0.65973549422827,
-      "grad_norm": 0.06859661731494095,
-      "learning_rate": 0.000313043168019867,
-      "loss": 1.0989,
-      "step": 4315
-    },
-    {
-      "epoch": 0.6604999617766226,
-      "grad_norm": 0.06412265796659967,
-      "learning_rate": 0.00031180627468469494,
-      "loss": 1.0289,
-      "step": 4320
-    },
-    {
-      "epoch": 0.6612644293249752,
-      "grad_norm": 0.06624448181748696,
-      "learning_rate": 0.0003105707216560114,
-      "loss": 1.0555,
-      "step": 4325
-    },
-    {
-      "epoch": 0.6620288968733278,
-      "grad_norm": 0.05973287720890624,
-      "learning_rate": 0.0003093365177333644,
-      "loss": 1.0933,
-      "step": 4330
-    },
-    {
-      "epoch": 0.6627933644216804,
-      "grad_norm": 0.06311393586816254,
-      "learning_rate": 0.0003081036717066938,
-      "loss": 1.0869,
-      "step": 4335
-    },
-    {
-      "epoch": 0.6635578319700328,
-      "grad_norm": 0.06408202960534053,
-      "learning_rate": 0.0003068721923562688,
-      "loss": 1.0702,
-      "step": 4340
-    },
-    {
-      "epoch": 0.6643222995183854,
-      "grad_norm": 0.05725889171250293,
-      "learning_rate": 0.0003056420884526244,
-      "loss": 1.0778,
-      "step": 4345
-    },
-    {
-      "epoch": 0.665086767066738,
-      "grad_norm": 0.05862869882873112,
-      "learning_rate": 0.00030441336875650115,
-      "loss": 1.0706,
-      "step": 4350
-    },
-    {
-      "epoch": 0.6658512346150905,
-      "grad_norm": 0.061265898924405884,
-      "learning_rate": 0.00030318604201877994,
-      "loss": 1.0869,
-      "step": 4355
-    },
-    {
-      "epoch": 0.6666157021634431,
-      "grad_norm": 0.06660471122968116,
-      "learning_rate": 0.0003019601169804216,
-      "loss": 1.096,
-      "step": 4360
-    },
-    {
-      "epoch": 0.6673801697117957,
-      "grad_norm": 0.06883224741443655,
-      "learning_rate": 0.0003007356023724045,
-      "loss": 1.0725,
-      "step": 4365
-    },
-    {
-      "epoch": 0.6681446372601483,
-      "grad_norm": 0.06613070267430747,
-      "learning_rate": 0.00029951250691566156,
-      "loss": 1.0636,
-      "step": 4370
-    },
-    {
-      "epoch": 0.6689091048085009,
-      "grad_norm": 0.05514526491107288,
-      "learning_rate": 0.00029829083932101844,
-      "loss": 1.0718,
-      "step": 4375
-    },
-    {
-      "epoch": 0.6696735723568534,
-      "grad_norm": 0.06035585077122042,
-      "learning_rate": 0.00029707060828913225,
-      "loss": 1.0601,
-      "step": 4380
-    },
-    {
-      "epoch": 0.670438039905206,
-      "grad_norm": 0.06401962624174869,
-      "learning_rate": 0.0002958518225104283,
-      "loss": 1.1028,
-      "step": 4385
-    },
-    {
-      "epoch": 0.6712025074535586,
-      "grad_norm": 0.06137496192489166,
-      "learning_rate": 0.00029463449066503936,
-      "loss": 1.0915,
-      "step": 4390
-    },
-    {
-      "epoch": 0.6719669750019112,
-      "grad_norm": 0.061351092117864735,
-      "learning_rate": 0.00029341862142274306,
-      "loss": 1.1007,
-      "step": 4395
-    },
-    {
-      "epoch": 0.6727314425502637,
-      "grad_norm": 0.06025596628948086,
-      "learning_rate": 0.00029220422344290053,
-      "loss": 1.0421,
-      "step": 4400
-    },
-    {
-      "epoch": 0.6734959100986163,
-      "grad_norm": 0.057708276936275546,
-      "learning_rate": 0.0002909913053743949,
-      "loss": 1.073,
-      "step": 4405
-    },
-    {
-      "epoch": 0.6742603776469689,
-      "grad_norm": 0.05934715222749201,
-      "learning_rate": 0.000289779875855569,
-      "loss": 1.0912,
-      "step": 4410
-    },
-    {
-      "epoch": 0.6750248451953215,
-      "grad_norm": 0.05766840269365502,
-      "learning_rate": 0.00028856994351416433,
-      "loss": 1.0883,
-      "step": 4415
-    },
-    {
-      "epoch": 0.675789312743674,
-      "grad_norm": 0.05304655893207912,
-      "learning_rate": 0.0002873615169672601,
-      "loss": 1.0014,
-      "step": 4420
-    },
-    {
-      "epoch": 0.6765537802920266,
-      "grad_norm": 0.06298752379204364,
-      "learning_rate": 0.00028615460482121096,
-      "loss": 1.0319,
-      "step": 4425
-    },
-    {
-      "epoch": 0.6773182478403792,
-      "grad_norm": 0.06572666933184156,
-      "learning_rate": 0.00028494921567158557,
-      "loss": 1.0627,
-      "step": 4430
-    },
-    {
-      "epoch": 0.6780827153887318,
-      "grad_norm": 0.05564925869133632,
-      "learning_rate": 0.0002837453581031068,
-      "loss": 1.0647,
-      "step": 4435
-    },
-    {
-      "epoch": 0.6788471829370843,
-      "grad_norm": 0.08038005236660152,
-      "learning_rate": 0.00028254304068958924,
-      "loss": 1.0436,
-      "step": 4440
-    },
-    {
-      "epoch": 0.6796116504854369,
-      "grad_norm": 0.07054627355167405,
-      "learning_rate": 0.0002813422719938784,
-      "loss": 1.0817,
-      "step": 4445
-    },
-    {
-      "epoch": 0.6803761180337895,
-      "grad_norm": 0.06023512683419835,
-      "learning_rate": 0.00028014306056779026,
-      "loss": 1.0844,
-      "step": 4450
-    },
-    {
-      "epoch": 0.6811405855821421,
-      "grad_norm": 0.06141926824245988,
-      "learning_rate": 0.0002789454149520497,
-      "loss": 1.0579,
-      "step": 4455
-    },
-    {
-      "epoch": 0.6819050531304947,
-      "grad_norm": 0.05608369336806669,
-      "learning_rate": 0.00027774934367622993,
-      "loss": 1.0871,
-      "step": 4460
-    },
-    {
-      "epoch": 0.6826695206788472,
-      "grad_norm": 0.06081877401068415,
-      "learning_rate": 0.0002765548552586926,
-      "loss": 1.0833,
-      "step": 4465
-    },
-    {
-      "epoch": 0.6834339882271998,
-      "grad_norm": 0.058671855702705646,
-      "learning_rate": 0.00027536195820652504,
-      "loss": 1.0955,
-      "step": 4470
-    },
-    {
-      "epoch": 0.6841984557755524,
-      "grad_norm": 0.06570010670273176,
-      "learning_rate": 0.0002741706610154816,
-      "loss": 1.0776,
-      "step": 4475
-    },
-    {
-      "epoch": 0.6849629233239048,
-      "grad_norm": 0.0621757019848029,
-      "learning_rate": 0.00027298097216992283,
-      "loss": 1.0787,
-      "step": 4480
-    },
-    {
-      "epoch": 0.6857273908722574,
-      "grad_norm": 0.05545832112022665,
-      "learning_rate": 0.00027179290014275447,
-      "loss": 1.0497,
-      "step": 4485
-    },
-    {
-      "epoch": 0.68649185842061,
-      "grad_norm": 0.06138638133688741,
-      "learning_rate": 0.0002706064533953667,
-      "loss": 1.0874,
-      "step": 4490
-    },
-    {
-      "epoch": 0.6872563259689626,
-      "grad_norm": 0.06238495792553165,
-      "learning_rate": 0.00026942164037757567,
-      "loss": 1.0474,
-      "step": 4495
-    },
-    {
-      "epoch": 0.6880207935173152,
-      "grad_norm": 0.06011230852606334,
-      "learning_rate": 0.00026823846952756127,
-      "loss": 1.0881,
-      "step": 4500
-    },
-    {
-      "epoch": 0.6887852610656677,
-      "grad_norm": 0.059382096793669986,
-      "learning_rate": 0.00026705694927180856,
-      "loss": 1.0691,
-      "step": 4505
-    },
-    {
-      "epoch": 0.6895497286140203,
-      "grad_norm": 0.06765967195924834,
-      "learning_rate": 0.00026587708802504675,
-      "loss": 1.0529,
-      "step": 4510
-    },
-    {
-      "epoch": 0.6903141961623729,
-      "grad_norm": 0.06898816086964697,
-      "learning_rate": 0.00026469889419018983,
-      "loss": 1.0509,
-      "step": 4515
-    },
-    {
-      "epoch": 0.6910786637107255,
-      "grad_norm": 0.06243515555844469,
-      "learning_rate": 0.0002635223761582763,
-      "loss": 1.0923,
-      "step": 4520
-    },
-    {
-      "epoch": 0.691843131259078,
-      "grad_norm": 0.06953610098318412,
-      "learning_rate": 0.00026234754230841076,
-      "loss": 1.0357,
-      "step": 4525
-    },
-    {
-      "epoch": 0.6926075988074306,
-      "grad_norm": 0.057115769917451306,
-      "learning_rate": 0.0002611744010077014,
-      "loss": 1.0823,
-      "step": 4530
-    },
-    {
-      "epoch": 0.6933720663557832,
-      "grad_norm": 0.05928833387518554,
-      "learning_rate": 0.0002600029606112033,
-      "loss": 1.0685,
-      "step": 4535
-    },
-    {
-      "epoch": 0.6941365339041358,
-      "grad_norm": 0.06514424505474277,
-      "learning_rate": 0.00025883322946185775,
-      "loss": 1.0522,
-      "step": 4540
-    },
-    {
-      "epoch": 0.6949010014524883,
-      "grad_norm": 0.06345812694465712,
-      "learning_rate": 0.00025766521589043267,
-      "loss": 1.063,
-      "step": 4545
-    },
-    {
-      "epoch": 0.6956654690008409,
-      "grad_norm": 0.06710277327176996,
-      "learning_rate": 0.00025649892821546334,
-      "loss": 1.102,
-      "step": 4550
-    },
-    {
-      "epoch": 0.6964299365491935,
-      "grad_norm": 0.0601077328567939,
-      "learning_rate": 0.0002553343747431934,
-      "loss": 1.0402,
-      "step": 4555
-    },
-    {
-      "epoch": 0.6971944040975461,
-      "grad_norm": 0.06375489248294168,
-      "learning_rate": 0.0002541715637675156,
-      "loss": 1.048,
-      "step": 4560
-    },
-    {
-      "epoch": 0.6979588716458986,
-      "grad_norm": 0.061468942934610936,
-      "learning_rate": 0.0002530105035699131,
-      "loss": 1.0518,
-      "step": 4565
-    },
-    {
-      "epoch": 0.6987233391942512,
-      "grad_norm": 0.05855568239256093,
-      "learning_rate": 0.0002518512024193993,
-      "loss": 1.0779,
-      "step": 4570
-    },
-    {
-      "epoch": 0.6994878067426038,
-      "grad_norm": 0.06351375706901624,
-      "learning_rate": 0.00025069366857246,
-      "loss": 1.0445,
-      "step": 4575
-    },
-    {
-      "epoch": 0.7002522742909564,
-      "grad_norm": 0.06505940038931915,
-      "learning_rate": 0.00024953791027299503,
-      "loss": 1.0701,
-      "step": 4580
-    },
-    {
-      "epoch": 0.701016741839309,
-      "grad_norm": 0.054690941924002506,
-      "learning_rate": 0.0002483839357522582,
-      "loss": 1.0496,
-      "step": 4585
-    },
-    {
-      "epoch": 0.7017812093876615,
-      "grad_norm": 0.0596197660604312,
-      "learning_rate": 0.0002472317532287994,
-      "loss": 1.0715,
-      "step": 4590
-    },
-    {
-      "epoch": 0.7025456769360141,
-      "grad_norm": 0.06143144522893759,
-      "learning_rate": 0.00024608137090840614,
-      "loss": 1.0836,
-      "step": 4595
-    },
-    {
-      "epoch": 0.7033101444843667,
-      "grad_norm": 0.06261739468488044,
-      "learning_rate": 0.0002449327969840449,
-      "loss": 1.0627,
-      "step": 4600
-    },
-    {
-      "epoch": 0.7040746120327193,
-      "grad_norm": 0.06305255828688207,
-      "learning_rate": 0.00024378603963580293,
-      "loss": 1.058,
-      "step": 4605
-    },
-    {
-      "epoch": 0.7048390795810718,
-      "grad_norm": 0.06250340445854399,
-      "learning_rate": 0.00024264110703082982,
-      "loss": 1.0866,
-      "step": 4610
-    },
-    {
-      "epoch": 0.7056035471294243,
-      "grad_norm": 0.06906200872639721,
-      "learning_rate": 0.00024149800732327942,
-      "loss": 1.0653,
-      "step": 4615
-    },
-    {
-      "epoch": 0.7063680146777769,
-      "grad_norm": 0.06481175868184377,
-      "learning_rate": 0.00024035674865425177,
-      "loss": 1.0565,
-      "step": 4620
-    },
-    {
-      "epoch": 0.7071324822261295,
-      "grad_norm": 0.0624819873899462,
-      "learning_rate": 0.00023921733915173565,
-      "loss": 1.069,
-      "step": 4625
-    },
-    {
-      "epoch": 0.707896949774482,
-      "grad_norm": 0.0701032532427006,
-      "learning_rate": 0.0002380797869305491,
-      "loss": 1.1104,
-      "step": 4630
-    },
-    {
-      "epoch": 0.7086614173228346,
-      "grad_norm": 0.06214752273965523,
-      "learning_rate": 0.0002369441000922834,
-      "loss": 1.0545,
-      "step": 4635
-    },
-    {
-      "epoch": 0.7094258848711872,
-      "grad_norm": 0.06024114616320267,
-      "learning_rate": 0.00023581028672524485,
-      "loss": 1.0555,
-      "step": 4640
-    },
-    {
-      "epoch": 0.7101903524195398,
-      "grad_norm": 0.05406841159348691,
-      "learning_rate": 0.00023467835490439647,
-      "loss": 1.0367,
-      "step": 4645
-    },
-    {
-      "epoch": 0.7109548199678923,
-      "grad_norm": 0.059729814605345245,
-      "learning_rate": 0.00023354831269130133,
-      "loss": 1.0416,
-      "step": 4650
-    },
-    {
-      "epoch": 0.7117192875162449,
-      "grad_norm": 0.0636881199097813,
-      "learning_rate": 0.0002324201681340646,
-      "loss": 1.0688,
-      "step": 4655
-    },
-    {
-      "epoch": 0.7124837550645975,
-      "grad_norm": 0.06458838613385967,
-      "learning_rate": 0.0002312939292672765,
-      "loss": 1.0565,
-      "step": 4660
-    },
-    {
-      "epoch": 0.7132482226129501,
-      "grad_norm": 0.055775934963736715,
-      "learning_rate": 0.00023016960411195508,
-      "loss": 1.0496,
-      "step": 4665
-    },
-    {
-      "epoch": 0.7140126901613026,
-      "grad_norm": 0.06470166002706611,
-      "learning_rate": 0.00022904720067548907,
-      "loss": 1.073,
-      "step": 4670
-    },
-    {
-      "epoch": 0.7147771577096552,
-      "grad_norm": 0.06899482171245135,
-      "learning_rate": 0.00022792672695158057,
-      "loss": 1.0684,
-      "step": 4675
-    },
-    {
-      "epoch": 0.7155416252580078,
-      "grad_norm": 0.054926399175138495,
-      "learning_rate": 0.00022680819092018851,
-      "loss": 1.0778,
-      "step": 4680
-    },
-    {
-      "epoch": 0.7163060928063604,
-      "grad_norm": 0.06076409853777293,
-      "learning_rate": 0.00022569160054747195,
-      "loss": 1.0726,
-      "step": 4685
-    },
-    {
-      "epoch": 0.717070560354713,
-      "grad_norm": 0.05595718010117591,
-      "learning_rate": 0.00022457696378573285,
-      "loss": 1.0407,
-      "step": 4690
-    },
-    {
-      "epoch": 0.7178350279030655,
-      "grad_norm": 0.05621732756049289,
-      "learning_rate": 0.00022346428857335904,
-      "loss": 1.0532,
-      "step": 4695
-    },
-    {
-      "epoch": 0.7185994954514181,
-      "grad_norm": 0.06019576108707039,
-      "learning_rate": 0.00022235358283476936,
-      "loss": 1.0965,
-      "step": 4700
-    },
-    {
-      "epoch": 0.7193639629997707,
-      "grad_norm": 0.08373809700675898,
-      "learning_rate": 0.00022124485448035553,
-      "loss": 1.0519,
-      "step": 4705
-    },
-    {
-      "epoch": 0.7201284305481233,
-      "grad_norm": 0.07459730823974277,
-      "learning_rate": 0.00022013811140642652,
-      "loss": 1.0614,
-      "step": 4710
-    },
-    {
-      "epoch": 0.7208928980964758,
-      "grad_norm": 0.0571730569494354,
-      "learning_rate": 0.00021903336149515224,
-      "loss": 1.0561,
-      "step": 4715
-    },
-    {
-      "epoch": 0.7216573656448284,
-      "grad_norm": 0.05546776042814926,
-      "learning_rate": 0.00021793061261450748,
-      "loss": 1.0567,
-      "step": 4720
-    },
-    {
-      "epoch": 0.722421833193181,
-      "grad_norm": 0.06438680600968821,
-      "learning_rate": 0.00021682987261821546,
-      "loss": 1.0629,
-      "step": 4725
-    },
-    {
-      "epoch": 0.7231863007415336,
-      "grad_norm": 0.0743342279298406,
-      "learning_rate": 0.0002157311493456931,
-      "loss": 1.0397,
-      "step": 4730
-    },
-    {
-      "epoch": 0.7239507682898861,
-      "grad_norm": 0.06983357774921824,
-      "learning_rate": 0.0002146344506219931,
-      "loss": 1.0536,
-      "step": 4735
-    },
-    {
-      "epoch": 0.7247152358382387,
-      "grad_norm": 0.059060993062700295,
-      "learning_rate": 0.00021353978425775005,
-      "loss": 1.0728,
-      "step": 4740
-    },
-    {
-      "epoch": 0.7254797033865913,
-      "grad_norm": 0.0579147373891631,
-      "learning_rate": 0.0002124471580491245,
-      "loss": 1.0557,
-      "step": 4745
-    },
-    {
-      "epoch": 0.7262441709349439,
-      "grad_norm": 0.0621748767051743,
-      "learning_rate": 0.00021135657977774665,
-      "loss": 1.0392,
-      "step": 4750
-    },
-    {
-      "epoch": 0.7270086384832963,
-      "grad_norm": 0.05605518237597559,
-      "learning_rate": 0.00021026805721066134,
-      "loss": 1.0644,
-      "step": 4755
-    },
-    {
-      "epoch": 0.7277731060316489,
-      "grad_norm": 0.058785481188473786,
-      "learning_rate": 0.0002091815981002731,
-      "loss": 1.0571,
-      "step": 4760
-    },
-    {
-      "epoch": 0.7285375735800015,
-      "grad_norm": 0.06048464507849506,
-      "learning_rate": 0.0002080972101842904,
-      "loss": 1.0716,
-      "step": 4765
-    },
-    {
-      "epoch": 0.7293020411283541,
-      "grad_norm": 0.06504524837348831,
-      "learning_rate": 0.00020701490118567079,
-      "loss": 1.0539,
-      "step": 4770
-    },
-    {
-      "epoch": 0.7300665086767066,
-      "grad_norm": 0.058155612162554556,
-      "learning_rate": 0.000205934678812566,
-      "loss": 1.0634,
-      "step": 4775
-    },
-    {
-      "epoch": 0.7308309762250592,
-      "grad_norm": 0.08474842750770611,
-      "learning_rate": 0.00020485655075826663,
-      "loss": 1.0494,
-      "step": 4780
-    },
-    {
-      "epoch": 0.7315954437734118,
-      "grad_norm": 0.06240607837640539,
-      "learning_rate": 0.00020378052470114823,
-      "loss": 1.0964,
-      "step": 4785
-    },
-    {
-      "epoch": 0.7323599113217644,
-      "grad_norm": 0.05804984432313353,
-      "learning_rate": 0.0002027066083046155,
-      "loss": 1.0112,
-      "step": 4790
-    },
-    {
-      "epoch": 0.7331243788701169,
-      "grad_norm": 0.05649066467343644,
-      "learning_rate": 0.000201634809217048,
-      "loss": 1.0986,
-      "step": 4795
-    },
-    {
-      "epoch": 0.7338888464184695,
-      "grad_norm": 0.05637504209092696,
-      "learning_rate": 0.00020056513507174685,
-      "loss": 1.0532,
-      "step": 4800
-    },
-    {
-      "epoch": 0.7346533139668221,
-      "grad_norm": 0.06298093925215675,
-      "learning_rate": 0.00019949759348687857,
-      "loss": 1.0139,
-      "step": 4805
-    },
-    {
-      "epoch": 0.7354177815151747,
-      "grad_norm": 0.06517336083169355,
-      "learning_rate": 0.00019843219206542212,
-      "loss": 1.0883,
-      "step": 4810
-    },
-    {
-      "epoch": 0.7361822490635272,
-      "grad_norm": 0.05724115630119351,
-      "learning_rate": 0.00019736893839511423,
-      "loss": 1.0196,
-      "step": 4815
-    },
-    {
-      "epoch": 0.7369467166118798,
-      "grad_norm": 0.06914879934829934,
-      "learning_rate": 0.0001963078400483953,
-      "loss": 1.048,
-      "step": 4820
-    },
-    {
-      "epoch": 0.7377111841602324,
-      "grad_norm": 0.07095787700138627,
-      "learning_rate": 0.00019524890458235566,
-      "loss": 1.0561,
-      "step": 4825
-    },
-    {
-      "epoch": 0.738475651708585,
-      "grad_norm": 0.06227294786877535,
-      "learning_rate": 0.00019419213953868235,
-      "loss": 0.9981,
-      "step": 4830
-    },
-    {
-      "epoch": 0.7392401192569376,
-      "grad_norm": 0.05688377294396278,
-      "learning_rate": 0.0001931375524436037,
-      "loss": 1.0683,
-      "step": 4835
-    },
-    {
-      "epoch": 0.7400045868052901,
-      "grad_norm": 0.06299988853768748,
-      "learning_rate": 0.00019208515080783723,
-      "loss": 1.0661,
-      "step": 4840
-    },
-    {
-      "epoch": 0.7407690543536427,
-      "grad_norm": 0.05374647681244872,
-      "learning_rate": 0.0001910349421265362,
-      "loss": 1.0292,
-      "step": 4845
-    },
-    {
-      "epoch": 0.7415335219019953,
-      "grad_norm": 0.05994041675068698,
-      "learning_rate": 0.00018998693387923542,
-      "loss": 1.0409,
-      "step": 4850
-    },
-    {
-      "epoch": 0.7422979894503479,
-      "grad_norm": 0.06160205659605142,
-      "learning_rate": 0.00018894113352979774,
-      "loss": 1.0132,
-      "step": 4855
-    },
-    {
-      "epoch": 0.7430624569987004,
-      "grad_norm": 0.06087875817228139,
-      "learning_rate": 0.00018789754852636243,
-      "loss": 1.0508,
-      "step": 4860
-    },
-    {
-      "epoch": 0.743826924547053,
-      "grad_norm": 0.05831613296562593,
-      "learning_rate": 0.00018685618630129064,
-      "loss": 1.0738,
-      "step": 4865
-    },
-    {
-      "epoch": 0.7445913920954056,
-      "grad_norm": 0.05710988576473131,
-      "learning_rate": 0.00018581705427111306,
-      "loss": 1.0502,
-      "step": 4870
-    },
-    {
-      "epoch": 0.7453558596437582,
-      "grad_norm": 0.055938286857235675,
-      "learning_rate": 0.00018478015983647717,
-      "loss": 1.0545,
-      "step": 4875
-    },
-    {
-      "epoch": 0.7461203271921107,
-      "grad_norm": 0.06130999146989328,
-      "learning_rate": 0.0001837455103820942,
-      "loss": 1.0672,
-      "step": 4880
-    },
-    {
-      "epoch": 0.7468847947404633,
-      "grad_norm": 0.0700486673269101,
-      "learning_rate": 0.0001827131132766867,
-      "loss": 1.0598,
-      "step": 4885
-    },
-    {
-      "epoch": 0.7476492622888158,
-      "grad_norm": 0.061673007556164776,
-      "learning_rate": 0.0001816829758729368,
-      "loss": 1.0591,
-      "step": 4890
-    },
-    {
-      "epoch": 0.7484137298371684,
-      "grad_norm": 0.05301433905384343,
-      "learning_rate": 0.00018065510550743213,
-      "loss": 1.032,
-      "step": 4895
-    },
-    {
-      "epoch": 0.7491781973855209,
-      "grad_norm": 0.05637326883587409,
-      "learning_rate": 0.000179629509500615,
-      "loss": 1.0485,
-      "step": 4900
-    },
-    {
-      "epoch": 0.7499426649338735,
-      "grad_norm": 0.06567737463780914,
-      "learning_rate": 0.0001786061951567303,
-      "loss": 1.0517,
-      "step": 4905
-    },
-    {
-      "epoch": 0.7507071324822261,
-      "grad_norm": 0.06049809996712097,
-      "learning_rate": 0.00017758516976377247,
-      "loss": 1.0563,
-      "step": 4910
-    },
-    {
-      "epoch": 0.7514716000305787,
-      "grad_norm": 0.0576348511640816,
-      "learning_rate": 0.00017656644059343423,
-      "loss": 1.0455,
-      "step": 4915
-    },
-    {
-      "epoch": 0.7522360675789312,
-      "grad_norm": 0.05822968830584341,
-      "learning_rate": 0.00017555001490105488,
-      "loss": 1.0848,
-      "step": 4920
-    },
-    {
-      "epoch": 0.7530005351272838,
-      "grad_norm": 0.06417289661323497,
-      "learning_rate": 0.00017453589992556833,
-      "loss": 1.0481,
-      "step": 4925
-    },
-    {
-      "epoch": 0.7537650026756364,
-      "grad_norm": 0.059833408626991574,
-      "learning_rate": 0.0001735241028894518,
-      "loss": 1.0682,
-      "step": 4930
-    },
-    {
-      "epoch": 0.754529470223989,
-      "grad_norm": 0.061813707081001805,
-      "learning_rate": 0.00017251463099867415,
-      "loss": 1.0578,
-      "step": 4935
-    },
-    {
-      "epoch": 0.7552939377723416,
-      "grad_norm": 0.06410222905138455,
-      "learning_rate": 0.0001715074914426446,
-      "loss": 1.0629,
-      "step": 4940
-    },
-    {
-      "epoch": 0.7560584053206941,
-      "grad_norm": 0.062252515513016034,
-      "learning_rate": 0.00017050269139416198,
-      "loss": 1.0241,
-      "step": 4945
-    },
-    {
-      "epoch": 0.7568228728690467,
-      "grad_norm": 0.06040469252714623,
-      "learning_rate": 0.00016950023800936298,
-      "loss": 1.0177,
-      "step": 4950
-    },
-    {
-      "epoch": 0.7575873404173993,
-      "grad_norm": 0.05861174886953368,
-      "learning_rate": 0.00016850013842767102,
-      "loss": 1.0575,
-      "step": 4955
-    },
-    {
-      "epoch": 0.7583518079657519,
-      "grad_norm": 0.05545703925778717,
-      "learning_rate": 0.0001675023997717468,
-      "loss": 1.0852,
-      "step": 4960
-    },
-    {
-      "epoch": 0.7591162755141044,
-      "grad_norm": 0.061973289366438845,
-      "learning_rate": 0.000166507029147436,
-      "loss": 1.0398,
-      "step": 4965
-    },
-    {
-      "epoch": 0.759880743062457,
-      "grad_norm": 0.05581024876428727,
-      "learning_rate": 0.00016551403364371936,
-      "loss": 1.0494,
-      "step": 4970
-    },
-    {
-      "epoch": 0.7606452106108096,
-      "grad_norm": 0.05851335959521461,
-      "learning_rate": 0.0001645234203326622,
-      "loss": 1.0588,
-      "step": 4975
-    },
-    {
-      "epoch": 0.7614096781591622,
-      "grad_norm": 0.06189891483918274,
-      "learning_rate": 0.00016353519626936396,
-      "loss": 1.0855,
-      "step": 4980
-    },
-    {
-      "epoch": 0.7621741457075147,
-      "grad_norm": 0.06925049708184426,
-      "learning_rate": 0.00016254936849190795,
-      "loss": 1.0521,
-      "step": 4985
-    },
-    {
-      "epoch": 0.7629386132558673,
-      "grad_norm": 0.05519111923786063,
-      "learning_rate": 0.00016156594402131158,
-      "loss": 1.0116,
-      "step": 4990
-    },
-    {
-      "epoch": 0.7637030808042199,
-      "grad_norm": 0.0661038161121552,
-      "learning_rate": 0.00016058492986147538,
-      "loss": 1.0273,
-      "step": 4995
-    },
-    {
-      "epoch": 0.7644675483525725,
-      "grad_norm": 0.062383945489791016,
-      "learning_rate": 0.00015960633299913406,
-      "loss": 1.0873,
-      "step": 5000
-    },
-    {
-      "epoch": 0.765232015900925,
-      "grad_norm": 0.06897279137476997,
-      "learning_rate": 0.00015863016040380685,
-      "loss": 1.0594,
-      "step": 5005
-    },
-    {
-      "epoch": 0.7659964834492776,
-      "grad_norm": 0.056082673963443117,
-      "learning_rate": 0.000157656419027747,
-      "loss": 1.0708,
-      "step": 5010
-    },
-    {
-      "epoch": 0.7667609509976302,
-      "grad_norm": 0.058420523194276514,
-      "learning_rate": 0.00015668511580589284,
-      "loss": 1.0628,
-      "step": 5015
-    },
-    {
-      "epoch": 0.7675254185459828,
-      "grad_norm": 0.06310313725455063,
-      "learning_rate": 0.0001557162576558183,
-      "loss": 1.0722,
-      "step": 5020
-    },
-    {
-      "epoch": 0.7682898860943354,
-      "grad_norm": 0.06596435296148517,
-      "learning_rate": 0.00015474985147768367,
-      "loss": 1.0406,
-      "step": 5025
-    },
-    {
-      "epoch": 0.7690543536426878,
-      "grad_norm": 0.06971977148141095,
-      "learning_rate": 0.00015378590415418637,
-      "loss": 1.0339,
-      "step": 5030
-    },
-    {
-      "epoch": 0.7698188211910404,
-      "grad_norm": 0.05474565393634749,
-      "learning_rate": 0.00015282442255051205,
-      "loss": 1.0795,
-      "step": 5035
-    },
-    {
-      "epoch": 0.770583288739393,
-      "grad_norm": 0.06320387891306647,
-      "learning_rate": 0.00015186541351428545,
-      "loss": 1.0629,
-      "step": 5040
-    },
-    {
-      "epoch": 0.7713477562877455,
-      "grad_norm": 0.06290589598630533,
-      "learning_rate": 0.00015090888387552187,
-      "loss": 1.045,
-      "step": 5045
-    },
-    {
-      "epoch": 0.7721122238360981,
-      "grad_norm": 0.06282443758043982,
-      "learning_rate": 0.0001499548404465788,
-      "loss": 1.0133,
-      "step": 5050
-    },
-    {
-      "epoch": 0.7728766913844507,
-      "grad_norm": 0.061142142507514305,
-      "learning_rate": 0.00014900329002210682,
-      "loss": 1.0481,
-      "step": 5055
-    },
-    {
-      "epoch": 0.7736411589328033,
-      "grad_norm": 0.056267952874468424,
-      "learning_rate": 0.00014805423937900086,
-      "loss": 1.0295,
-      "step": 5060
-    },
-    {
-      "epoch": 0.7744056264811559,
-      "grad_norm": 0.056082796704510815,
-      "learning_rate": 0.00014710769527635372,
-      "loss": 1.0509,
-      "step": 5065
-    },
-    {
-      "epoch": 0.7751700940295084,
-      "grad_norm": 0.06610725348457465,
-      "learning_rate": 0.00014616366445540575,
-      "loss": 1.0911,
-      "step": 5070
-    },
-    {
-      "epoch": 0.775934561577861,
-      "grad_norm": 0.06318782224901359,
-      "learning_rate": 0.0001452221536394983,
-      "loss": 1.0382,
-      "step": 5075
-    },
-    {
-      "epoch": 0.7766990291262136,
-      "grad_norm": 0.08384451404877709,
-      "learning_rate": 0.00014428316953402526,
-      "loss": 1.0365,
-      "step": 5080
-    },
-    {
-      "epoch": 0.7774634966745662,
-      "grad_norm": 0.0671394576521587,
-      "learning_rate": 0.00014334671882638538,
-      "loss": 1.0328,
-      "step": 5085
-    },
-    {
-      "epoch": 0.7782279642229187,
-      "grad_norm": 0.055317503030393636,
-      "learning_rate": 0.00014241280818593472,
-      "loss": 1.0427,
-      "step": 5090
-    },
-    {
-      "epoch": 0.7789924317712713,
-      "grad_norm": 0.06850422084885686,
-      "learning_rate": 0.0001414814442639391,
-      "loss": 1.0653,
-      "step": 5095
-    },
-    {
-      "epoch": 0.7797568993196239,
-      "grad_norm": 0.06250617863046679,
-      "learning_rate": 0.00014055263369352672,
-      "loss": 1.0339,
-      "step": 5100
-    },
-    {
-      "epoch": 0.7805213668679765,
-      "grad_norm": 0.06251159274247055,
-      "learning_rate": 0.00013962638308964083,
-      "loss": 1.0615,
-      "step": 5105
-    },
-    {
-      "epoch": 0.781285834416329,
-      "grad_norm": 0.06035781872130478,
-      "learning_rate": 0.00013870269904899302,
-      "loss": 1.0662,
-      "step": 5110
-    },
-    {
-      "epoch": 0.7820503019646816,
-      "grad_norm": 0.05488178375326448,
-      "learning_rate": 0.0001377815881500159,
-      "loss": 1.0777,
-      "step": 5115
-    },
-    {
-      "epoch": 0.7828147695130342,
-      "grad_norm": 0.056280611763081405,
-      "learning_rate": 0.0001368630569528156,
-      "loss": 1.0451,
-      "step": 5120
-    },
-    {
-      "epoch": 0.7835792370613868,
-      "grad_norm": 0.059108351931859225,
-      "learning_rate": 0.0001359471119991268,
-      "loss": 1.0645,
-      "step": 5125
-    },
-    {
-      "epoch": 0.7843437046097393,
-      "grad_norm": 0.05661993372727011,
-      "learning_rate": 0.0001350337598122644,
-      "loss": 1.0362,
-      "step": 5130
-    },
-    {
-      "epoch": 0.7851081721580919,
-      "grad_norm": 0.05986614836392417,
-      "learning_rate": 0.0001341230068970779,
-      "loss": 1.0175,
-      "step": 5135
-    },
-    {
-      "epoch": 0.7858726397064445,
-      "grad_norm": 0.06809753874365711,
-      "learning_rate": 0.00013321485973990493,
-      "loss": 1.0583,
-      "step": 5140
-    },
-    {
-      "epoch": 0.7866371072547971,
-      "grad_norm": 0.06323544711504882,
-      "learning_rate": 0.00013230932480852485,
-      "loss": 1.036,
-      "step": 5145
-    },
-    {
-      "epoch": 0.7874015748031497,
-      "grad_norm": 0.06793389651951727,
-      "learning_rate": 0.00013140640855211345,
-      "loss": 1.0233,
-      "step": 5150
-    },
-    {
-      "epoch": 0.7881660423515022,
-      "grad_norm": 0.06817907964044204,
-      "learning_rate": 0.000130506117401196,
-      "loss": 1.0297,
-      "step": 5155
-    },
-    {
-      "epoch": 0.7889305098998548,
-      "grad_norm": 0.0603564707452043,
-      "learning_rate": 0.00012960845776760154,
-      "loss": 1.0359,
-      "step": 5160
-    },
-    {
-      "epoch": 0.7896949774482073,
-      "grad_norm": 0.05755456407134064,
-      "learning_rate": 0.00012871343604441837,
-      "loss": 1.0425,
-      "step": 5165
-    },
-    {
-      "epoch": 0.7904594449965598,
-      "grad_norm": 0.06924984234647509,
-      "learning_rate": 0.00012782105860594724,
-      "loss": 1.0133,
-      "step": 5170
-    },
-    {
-      "epoch": 0.7912239125449124,
-      "grad_norm": 0.06525653855106175,
-      "learning_rate": 0.00012693133180765654,
-      "loss": 1.0236,
-      "step": 5175
-    },
-    {
-      "epoch": 0.791988380093265,
-      "grad_norm": 0.058531934314296985,
-      "learning_rate": 0.0001260442619861369,
-      "loss": 1.0413,
-      "step": 5180
-    },
-    {
-      "epoch": 0.7927528476416176,
-      "grad_norm": 0.06357634178206084,
-      "learning_rate": 0.00012515985545905606,
-      "loss": 1.0336,
-      "step": 5185
-    },
-    {
-      "epoch": 0.7935173151899702,
-      "grad_norm": 0.06604250681255697,
-      "learning_rate": 0.00012427811852511395,
-      "loss": 1.1083,
-      "step": 5190
-    },
-    {
-      "epoch": 0.7942817827383227,
-      "grad_norm": 0.055731873100012756,
-      "learning_rate": 0.0001233990574639981,
-      "loss": 1.0716,
-      "step": 5195
-    },
-    {
-      "epoch": 0.7950462502866753,
-      "grad_norm": 0.05731005090974728,
-      "learning_rate": 0.00012252267853633797,
-      "loss": 1.0179,
-      "step": 5200
-    },
-    {
-      "epoch": 0.7958107178350279,
-      "grad_norm": 0.05717571227226772,
-      "learning_rate": 0.00012164898798366126,
-      "loss": 1.0229,
-      "step": 5205
-    },
-    {
-      "epoch": 0.7965751853833805,
-      "grad_norm": 0.06450794458214343,
-      "learning_rate": 0.0001207779920283496,
-      "loss": 1.0382,
-      "step": 5210
-    },
-    {
-      "epoch": 0.797339652931733,
-      "grad_norm": 0.06110455802138656,
-      "learning_rate": 0.00011990969687359349,
-      "loss": 1.072,
-      "step": 5215
-    },
-    {
-      "epoch": 0.7981041204800856,
-      "grad_norm": 0.06285806039928286,
-      "learning_rate": 0.00011904410870334803,
-      "loss": 1.038,
-      "step": 5220
-    },
-    {
-      "epoch": 0.7988685880284382,
-      "grad_norm": 0.05602690523165866,
-      "learning_rate": 0.00011818123368229022,
-      "loss": 1.0422,
-      "step": 5225
-    },
-    {
-      "epoch": 0.7996330555767908,
-      "grad_norm": 0.05976622450742704,
-      "learning_rate": 0.00011732107795577345,
-      "loss": 1.051,
-      "step": 5230
-    },
-    {
-      "epoch": 0.8003975231251433,
-      "grad_norm": 0.06267233414280757,
-      "learning_rate": 0.00011646364764978467,
-      "loss": 1.0601,
-      "step": 5235
-    },
-    {
-      "epoch": 0.8011619906734959,
-      "grad_norm": 0.07549985414221129,
-      "learning_rate": 0.00011560894887090051,
-      "loss": 1.0348,
-      "step": 5240
-    },
-    {
-      "epoch": 0.8019264582218485,
-      "grad_norm": 0.059342180679561214,
-      "learning_rate": 0.00011475698770624382,
-      "loss": 1.0652,
-      "step": 5245
-    },
-    {
-      "epoch": 0.8026909257702011,
-      "grad_norm": 0.06346928729977709,
-      "learning_rate": 0.00011390777022344006,
-      "loss": 1.0449,
-      "step": 5250
-    },
-    {
-      "epoch": 0.8034553933185536,
-      "grad_norm": 0.054691568332852346,
-      "learning_rate": 0.000113061302470575,
-      "loss": 1.0218,
-      "step": 5255
-    },
-    {
-      "epoch": 0.8042198608669062,
-      "grad_norm": 0.055409675325726666,
-      "learning_rate": 0.00011221759047615004,
-      "loss": 1.0436,
-      "step": 5260
-    },
-    {
-      "epoch": 0.8049843284152588,
-      "grad_norm": 0.06398528486658486,
-      "learning_rate": 0.00011137664024904065,
-      "loss": 1.0385,
-      "step": 5265
-    },
-    {
-      "epoch": 0.8057487959636114,
-      "grad_norm": 0.06396202261992998,
-      "learning_rate": 0.0001105384577784534,
-      "loss": 1.0752,
-      "step": 5270
-    },
-    {
-      "epoch": 0.806513263511964,
-      "grad_norm": 0.05484742690897295,
-      "learning_rate": 0.00010970304903388267,
-      "loss": 1.0219,
-      "step": 5275
-    },
-    {
-      "epoch": 0.8072777310603165,
-      "grad_norm": 0.062189301119376504,
-      "learning_rate": 0.00010887041996506857,
-      "loss": 1.0211,
-      "step": 5280
-    },
-    {
-      "epoch": 0.8080421986086691,
-      "grad_norm": 0.062008186236770825,
-      "learning_rate": 0.00010804057650195448,
-      "loss": 1.069,
-      "step": 5285
-    },
-    {
-      "epoch": 0.8088066661570217,
-      "grad_norm": 0.06107669674272538,
-      "learning_rate": 0.000107213524554645,
-      "loss": 1.1088,
-      "step": 5290
-    },
-    {
-      "epoch": 0.8095711337053743,
-      "grad_norm": 0.05868772060077093,
-      "learning_rate": 0.00010638927001336341,
-      "loss": 1.007,
-      "step": 5295
-    },
-    {
-      "epoch": 0.8103356012537268,
-      "grad_norm": 0.06532420618997664,
-      "learning_rate": 0.00010556781874841026,
-      "loss": 1.0363,
-      "step": 5300
-    },
-    {
-      "epoch": 0.8111000688020793,
-      "grad_norm": 0.05416277654812083,
-      "learning_rate": 0.00010474917661012118,
-      "loss": 1.0184,
-      "step": 5305
-    },
-    {
-      "epoch": 0.8118645363504319,
-      "grad_norm": 0.05527840984915126,
-      "learning_rate": 0.00010393334942882521,
-      "loss": 1.024,
-      "step": 5310
-    },
-    {
-      "epoch": 0.8126290038987845,
-      "grad_norm": 0.06124101383734983,
-      "learning_rate": 0.000103120343014804,
-      "loss": 1.0251,
-      "step": 5315
-    },
-    {
-      "epoch": 0.813393471447137,
-      "grad_norm": 0.06459260135386527,
-      "learning_rate": 0.00010231016315824875,
-      "loss": 1.0474,
-      "step": 5320
-    },
-    {
-      "epoch": 0.8141579389954896,
-      "grad_norm": 0.05584252284948542,
-      "learning_rate": 0.0001015028156292212,
-      "loss": 1.0966,
-      "step": 5325
-    },
-    {
-      "epoch": 0.8149224065438422,
-      "grad_norm": 0.06552357487141015,
-      "learning_rate": 0.00010069830617761067,
-      "loss": 1.0496,
-      "step": 5330
-    },
-    {
-      "epoch": 0.8156868740921948,
-      "grad_norm": 0.05524468981155268,
-      "learning_rate": 9.98966405330941e-05,
-      "loss": 1.0519,
-      "step": 5335
-    },
-    {
-      "epoch": 0.8164513416405473,
-      "grad_norm": 0.05208274069273543,
-      "learning_rate": 9.909782440509491e-05,
-      "loss": 1.0089,
-      "step": 5340
-    },
-    {
-      "epoch": 0.8172158091888999,
-      "grad_norm": 0.05988429182702099,
-      "learning_rate": 9.830186348274239e-05,
-      "loss": 1.0319,
-      "step": 5345
-    },
-    {
-      "epoch": 0.8179802767372525,
-      "grad_norm": 0.06029492738382687,
-      "learning_rate": 9.750876343483112e-05,
-      "loss": 1.0574,
-      "step": 5350
-    },
-    {
-      "epoch": 0.8187447442856051,
-      "grad_norm": 0.05562867844505153,
-      "learning_rate": 9.671852990978109e-05,
-      "loss": 1.0655,
-      "step": 5355
-    },
-    {
-      "epoch": 0.8195092118339576,
-      "grad_norm": 0.055724731096407874,
-      "learning_rate": 9.593116853559646e-05,
-      "loss": 1.0604,
-      "step": 5360
-    },
-    {
-      "epoch": 0.8202736793823102,
-      "grad_norm": 0.05764217796697168,
-      "learning_rate": 9.514668491982631e-05,
-      "loss": 1.0664,
-      "step": 5365
-    },
-    {
-      "epoch": 0.8210381469306628,
-      "grad_norm": 0.05752272726794207,
-      "learning_rate": 9.43650846495247e-05,
-      "loss": 1.0412,
-      "step": 5370
-    },
-    {
-      "epoch": 0.8218026144790154,
-      "grad_norm": 0.06966191316852606,
-      "learning_rate": 9.35863732912104e-05,
-      "loss": 1.0369,
-      "step": 5375
-    },
-    {
-      "epoch": 0.822567082027368,
-      "grad_norm": 0.0683006602529788,
-      "learning_rate": 9.281055639082747e-05,
-      "loss": 1.023,
-      "step": 5380
-    },
-    {
-      "epoch": 0.8233315495757205,
-      "grad_norm": 0.06139287047512268,
-      "learning_rate": 9.203763947370591e-05,
-      "loss": 1.0404,
-      "step": 5385
-    },
-    {
-      "epoch": 0.8240960171240731,
-      "grad_norm": 0.057208454420155645,
-      "learning_rate": 9.12676280445221e-05,
-      "loss": 1.0402,
-      "step": 5390
-    },
-    {
-      "epoch": 0.8248604846724257,
-      "grad_norm": 0.060773388016504705,
-      "learning_rate": 9.050052758725975e-05,
-      "loss": 1.053,
-      "step": 5395
-    },
-    {
-      "epoch": 0.8256249522207783,
-      "grad_norm": 0.06256383640892278,
-      "learning_rate": 8.973634356517063e-05,
-      "loss": 1.0165,
-      "step": 5400
-    },
-    {
-      "epoch": 0.8263894197691308,
-      "grad_norm": 0.05964678294589494,
-      "learning_rate": 8.89750814207359e-05,
-      "loss": 1.0403,
-      "step": 5405
-    },
-    {
-      "epoch": 0.8271538873174834,
-      "grad_norm": 0.05381486961488011,
-      "learning_rate": 8.821674657562723e-05,
-      "loss": 1.0269,
-      "step": 5410
-    },
-    {
-      "epoch": 0.827918354865836,
-      "grad_norm": 0.06186356736730895,
-      "learning_rate": 8.746134443066839e-05,
-      "loss": 1.0329,
-      "step": 5415
-    },
-    {
-      "epoch": 0.8286828224141886,
-      "grad_norm": 0.0762563854805148,
-      "learning_rate": 8.670888036579639e-05,
-      "loss": 1.026,
-      "step": 5420
-    },
-    {
-      "epoch": 0.8294472899625411,
-      "grad_norm": 0.06234950390412603,
-      "learning_rate": 8.595935974002317e-05,
-      "loss": 1.074,
-      "step": 5425
-    },
-    {
-      "epoch": 0.8302117575108937,
-      "grad_norm": 0.06064159367934313,
-      "learning_rate": 8.521278789139813e-05,
-      "loss": 1.0455,
-      "step": 5430
-    },
-    {
-      "epoch": 0.8309762250592463,
-      "grad_norm": 0.05867647677565349,
-      "learning_rate": 8.446917013696937e-05,
-      "loss": 1.0575,
-      "step": 5435
-    },
-    {
-      "epoch": 0.8317406926075988,
-      "grad_norm": 0.06005112304825767,
-      "learning_rate": 8.372851177274604e-05,
-      "loss": 1.0533,
-      "step": 5440
-    },
-    {
-      "epoch": 0.8325051601559513,
-      "grad_norm": 0.056801197118919115,
-      "learning_rate": 8.299081807366076e-05,
-      "loss": 1.0202,
-      "step": 5445
-    },
-    {
-      "epoch": 0.8332696277043039,
-      "grad_norm": 0.06433320913665841,
-      "learning_rate": 8.225609429353187e-05,
-      "loss": 1.0427,
-      "step": 5450
-    },
-    {
-      "epoch": 0.8340340952526565,
-      "grad_norm": 0.056930526224854554,
-      "learning_rate": 8.152434566502609e-05,
-      "loss": 1.0571,
-      "step": 5455
-    },
-    {
-      "epoch": 0.8347985628010091,
-      "grad_norm": 0.06701689056833854,
-      "learning_rate": 8.07955773996213e-05,
-      "loss": 1.056,
-      "step": 5460
-    },
-    {
-      "epoch": 0.8355630303493616,
-      "grad_norm": 0.0619431235582593,
-      "learning_rate": 8.006979468756942e-05,
-      "loss": 1.0356,
-      "step": 5465
-    },
-    {
-      "epoch": 0.8363274978977142,
-      "grad_norm": 0.07144070886689532,
-      "learning_rate": 7.934700269785921e-05,
-      "loss": 1.0338,
-      "step": 5470
-    },
-    {
-      "epoch": 0.8370919654460668,
-      "grad_norm": 0.05933587591040727,
-      "learning_rate": 7.862720657818001e-05,
-      "loss": 0.9977,
-      "step": 5475
-    },
-    {
-      "epoch": 0.8378564329944194,
-      "grad_norm": 0.060067430203855944,
-      "learning_rate": 7.791041145488453e-05,
-      "loss": 1.0338,
-      "step": 5480
-    },
-    {
-      "epoch": 0.8386209005427719,
-      "grad_norm": 0.05476141839391333,
-      "learning_rate": 7.719662243295217e-05,
-      "loss": 1.0267,
-      "step": 5485
-    },
-    {
-      "epoch": 0.8393853680911245,
-      "grad_norm": 0.060883085756603644,
-      "learning_rate": 7.64858445959537e-05,
-      "loss": 1.0211,
-      "step": 5490
-    },
-    {
-      "epoch": 0.8401498356394771,
-      "grad_norm": 0.06166847844911407,
-      "learning_rate": 7.57780830060139e-05,
-      "loss": 0.9904,
-      "step": 5495
-    },
-    {
-      "epoch": 0.8409143031878297,
-      "grad_norm": 0.056351980585675226,
-      "learning_rate": 7.50733427037762e-05,
-      "loss": 1.0611,
-      "step": 5500
-    },
-    {
-      "epoch": 0.8416787707361822,
-      "grad_norm": 0.06572816953871387,
-      "learning_rate": 7.437162870836638e-05,
-      "loss": 1.0406,
-      "step": 5505
-    },
-    {
-      "epoch": 0.8424432382845348,
-      "grad_norm": 0.05899387419863725,
-      "learning_rate": 7.367294601735708e-05,
-      "loss": 1.0148,
-      "step": 5510
-    },
-    {
-      "epoch": 0.8432077058328874,
-      "grad_norm": 0.07468273941671154,
-      "learning_rate": 7.297729960673244e-05,
-      "loss": 1.063,
-      "step": 5515
-    },
-    {
-      "epoch": 0.84397217338124,
-      "grad_norm": 0.05921366189210043,
-      "learning_rate": 7.228469443085206e-05,
-      "loss": 1.0227,
-      "step": 5520
-    },
-    {
-      "epoch": 0.8447366409295926,
-      "grad_norm": 0.0582034769285556,
-      "learning_rate": 7.15951354224157e-05,
-      "loss": 1.0282,
-      "step": 5525
-    },
-    {
-      "epoch": 0.8455011084779451,
-      "grad_norm": 0.06412673335503208,
-      "learning_rate": 7.090862749242921e-05,
-      "loss": 1.0661,
-      "step": 5530
-    },
-    {
-      "epoch": 0.8462655760262977,
-      "grad_norm": 0.06369336136081105,
-      "learning_rate": 7.022517553016827e-05,
-      "loss": 1.0106,
-      "step": 5535
-    },
-    {
-      "epoch": 0.8470300435746503,
-      "grad_norm": 0.058369722400650574,
-      "learning_rate": 6.954478440314427e-05,
-      "loss": 1.0188,
-      "step": 5540
-    },
-    {
-      "epoch": 0.8477945111230029,
-      "grad_norm": 0.05858354379438574,
-      "learning_rate": 6.886745895706947e-05,
-      "loss": 1.0354,
-      "step": 5545
-    },
-    {
-      "epoch": 0.8485589786713554,
-      "grad_norm": 0.058696989044599954,
-      "learning_rate": 6.819320401582258e-05,
-      "loss": 1.0528,
-      "step": 5550
-    },
-    {
-      "epoch": 0.849323446219708,
-      "grad_norm": 0.053612155113326045,
-      "learning_rate": 6.752202438141402e-05,
-      "loss": 1.0318,
-      "step": 5555
-    },
-    {
-      "epoch": 0.8500879137680606,
-      "grad_norm": 0.06426653835618115,
-      "learning_rate": 6.685392483395259e-05,
-      "loss": 1.0242,
-      "step": 5560
-    },
-    {
-      "epoch": 0.8508523813164132,
-      "grad_norm": 0.0635526144399548,
-      "learning_rate": 6.618891013161026e-05,
-      "loss": 1.0592,
-      "step": 5565
-    },
-    {
-      "epoch": 0.8516168488647657,
-      "grad_norm": 0.06555734215101829,
-      "learning_rate": 6.552698501058918e-05,
-      "loss": 1.0422,
-      "step": 5570
-    },
-    {
-      "epoch": 0.8523813164131183,
-      "grad_norm": 0.055809657555843446,
-      "learning_rate": 6.486815418508774e-05,
-      "loss": 0.9978,
-      "step": 5575
-    },
-    {
-      "epoch": 0.8531457839614708,
-      "grad_norm": 0.05874038849008624,
-      "learning_rate": 6.421242234726682e-05,
-      "loss": 1.0349,
-      "step": 5580
-    },
-    {
-      "epoch": 0.8539102515098234,
-      "grad_norm": 0.06670573325924009,
-      "learning_rate": 6.355979416721614e-05,
-      "loss": 1.0444,
-      "step": 5585
-    },
-    {
-      "epoch": 0.8546747190581759,
-      "grad_norm": 0.05580177701866413,
-      "learning_rate": 6.29102742929219e-05,
-      "loss": 0.9998,
-      "step": 5590
-    },
-    {
-      "epoch": 0.8554391866065285,
-      "grad_norm": 0.06253027310359607,
-      "learning_rate": 6.22638673502327e-05,
-      "loss": 1.0142,
-      "step": 5595
-    },
-    {
-      "epoch": 0.8562036541548811,
-      "grad_norm": 0.06370909036637372,
-      "learning_rate": 6.162057794282716e-05,
-      "loss": 1.0731,
-      "step": 5600
-    },
-    {
-      "epoch": 0.8569681217032337,
-      "grad_norm": 0.06829313690074416,
-      "learning_rate": 6.098041065218091e-05,
-      "loss": 1.0409,
-      "step": 5605
-    },
-    {
-      "epoch": 0.8577325892515862,
-      "grad_norm": 0.9460628196157908,
-      "learning_rate": 6.034337003753393e-05,
-      "loss": 1.0322,
-      "step": 5610
-    },
-    {
-      "epoch": 0.8584970567999388,
-      "grad_norm": 0.05588947262819443,
-      "learning_rate": 5.970946063585825e-05,
-      "loss": 1.0452,
-      "step": 5615
-    },
-    {
-      "epoch": 0.8592615243482914,
-      "grad_norm": 0.057798441761553625,
-      "learning_rate": 5.9078686961825836e-05,
-      "loss": 1.0114,
-      "step": 5620
-    },
-    {
-      "epoch": 0.860025991896644,
-      "grad_norm": 0.05559558436221299,
-      "learning_rate": 5.845105350777552e-05,
-      "loss": 1.0428,
-      "step": 5625
-    },
-    {
-      "epoch": 0.8607904594449965,
-      "grad_norm": 0.0605007497185841,
-      "learning_rate": 5.782656474368209e-05,
-      "loss": 1.033,
-      "step": 5630
-    },
-    {
-      "epoch": 0.8615549269933491,
-      "grad_norm": 0.05966863081807496,
-      "learning_rate": 5.720522511712406e-05,
-      "loss": 1.0186,
-      "step": 5635
-    },
-    {
-      "epoch": 0.8623193945417017,
-      "grad_norm": 0.06499735693928305,
-      "learning_rate": 5.6587039053251856e-05,
-      "loss": 1.0291,
-      "step": 5640
-    },
-    {
-      "epoch": 0.8630838620900543,
-      "grad_norm": 0.0625076833633183,
-      "learning_rate": 5.5972010954756015e-05,
-      "loss": 1.0514,
-      "step": 5645
-    },
-    {
-      "epoch": 0.8638483296384069,
-      "grad_norm": 0.05810931772236003,
-      "learning_rate": 5.5360145201836745e-05,
-      "loss": 1.0511,
-      "step": 5650
-    },
-    {
-      "epoch": 0.8646127971867594,
-      "grad_norm": 0.059530519769012705,
-      "learning_rate": 5.4751446152171866e-05,
-      "loss": 1.0486,
-      "step": 5655
-    },
-    {
-      "epoch": 0.865377264735112,
-      "grad_norm": 0.06283023161053312,
-      "learning_rate": 5.414591814088626e-05,
-      "loss": 1.0857,
-      "step": 5660
-    },
-    {
-      "epoch": 0.8661417322834646,
-      "grad_norm": 0.05570338995935867,
-      "learning_rate": 5.35435654805207e-05,
-      "loss": 1.0054,
-      "step": 5665
-    },
-    {
-      "epoch": 0.8669061998318172,
-      "grad_norm": 0.05444953831071218,
-      "learning_rate": 5.29443924610013e-05,
-      "loss": 1.0478,
-      "step": 5670
-    },
-    {
-      "epoch": 0.8676706673801697,
-      "grad_norm": 0.061787584612685256,
-      "learning_rate": 5.2348403349608834e-05,
-      "loss": 1.0321,
-      "step": 5675
-    },
-    {
-      "epoch": 0.8684351349285223,
-      "grad_norm": 0.05892829060325815,
-      "learning_rate": 5.17556023909489e-05,
-      "loss": 1.0268,
-      "step": 5680
-    },
-    {
-      "epoch": 0.8691996024768749,
-      "grad_norm": 0.06444520050715744,
-      "learning_rate": 5.1165993806920494e-05,
-      "loss": 1.0257,
-      "step": 5685
-    },
-    {
-      "epoch": 0.8699640700252275,
-      "grad_norm": 0.058670167066170946,
-      "learning_rate": 5.057958179668709e-05,
-      "loss": 1.0266,
-      "step": 5690
-    },
-    {
-      "epoch": 0.87072853757358,
-      "grad_norm": 0.058886717538286906,
-      "learning_rate": 4.9996370536646464e-05,
-      "loss": 1.0171,
-      "step": 5695
-    },
-    {
-      "epoch": 0.8714930051219326,
-      "grad_norm": 0.05872445786329025,
-      "learning_rate": 4.9416364180400574e-05,
-      "loss": 1.0689,
-      "step": 5700
-    },
-    {
-      "epoch": 0.8722574726702852,
-      "grad_norm": 0.06068126926877506,
-      "learning_rate": 4.883956685872626e-05,
-      "loss": 1.0528,
-      "step": 5705
-    },
-    {
-      "epoch": 0.8730219402186378,
-      "grad_norm": 0.06024646060572423,
-      "learning_rate": 4.826598267954574e-05,
-      "loss": 1.0091,
-      "step": 5710
-    },
-    {
-      "epoch": 0.8737864077669902,
-      "grad_norm": 0.056061378995899455,
-      "learning_rate": 4.7695615727897376e-05,
-      "loss": 1.0327,
-      "step": 5715
-    },
-    {
-      "epoch": 0.8745508753153428,
-      "grad_norm": 0.05642355016792108,
-      "learning_rate": 4.712847006590693e-05,
-      "loss": 1.0372,
-      "step": 5720
-    },
-    {
-      "epoch": 0.8753153428636954,
-      "grad_norm": 0.063941135432836,
-      "learning_rate": 4.6564549732757644e-05,
-      "loss": 1.0343,
-      "step": 5725
-    },
-    {
-      "epoch": 0.876079810412048,
-      "grad_norm": 0.06401648968226159,
-      "learning_rate": 4.6003858744662565e-05,
-      "loss": 1.0077,
-      "step": 5730
-    },
-    {
-      "epoch": 0.8768442779604005,
-      "grad_norm": 0.05861465222047972,
-      "learning_rate": 4.544640109483561e-05,
-      "loss": 1.0679,
-      "step": 5735
-    },
-    {
-      "epoch": 0.8776087455087531,
-      "grad_norm": 0.06161017849684759,
-      "learning_rate": 4.4892180753462744e-05,
-      "loss": 1.0213,
-      "step": 5740
-    },
-    {
-      "epoch": 0.8783732130571057,
-      "grad_norm": 0.06386075268338762,
-      "learning_rate": 4.434120166767408e-05,
-      "loss": 1.0436,
-      "step": 5745
-    },
-    {
-      "epoch": 0.8791376806054583,
-      "grad_norm": 0.06479505346064894,
-      "learning_rate": 4.37934677615156e-05,
-      "loss": 1.0518,
-      "step": 5750
-    },
-    {
-      "epoch": 0.8799021481538108,
-      "grad_norm": 0.06233866178912154,
-      "learning_rate": 4.32489829359215e-05,
-      "loss": 1.0067,
-      "step": 5755
-    },
-    {
-      "epoch": 0.8806666157021634,
-      "grad_norm": 0.0630500847286222,
-      "learning_rate": 4.270775106868585e-05,
-      "loss": 1.0303,
-      "step": 5760
-    },
-    {
-      "epoch": 0.881431083250516,
-      "grad_norm": 0.056777284717789075,
-      "learning_rate": 4.2169776014435554e-05,
-      "loss": 1.0409,
-      "step": 5765
-    },
-    {
-      "epoch": 0.8821955507988686,
-      "grad_norm": 0.05842993078979994,
-      "learning_rate": 4.163506160460262e-05,
-      "loss": 1.0421,
-      "step": 5770
-    },
-    {
-      "epoch": 0.8829600183472212,
-      "grad_norm": 0.054652077273351,
-      "learning_rate": 4.1103611647396735e-05,
-      "loss": 1.0526,
-      "step": 5775
-    },
-    {
-      "epoch": 0.8837244858955737,
-      "grad_norm": 0.05789228866767471,
-      "learning_rate": 4.0575429927778684e-05,
-      "loss": 0.9877,
-      "step": 5780
-    },
-    {
-      "epoch": 0.8844889534439263,
-      "grad_norm": 0.06435394005015357,
-      "learning_rate": 4.005052020743261e-05,
-      "loss": 1.0197,
-      "step": 5785
-    },
-    {
-      "epoch": 0.8852534209922789,
-      "grad_norm": 0.0668458189028938,
-      "learning_rate": 3.952888622473977e-05,
-      "loss": 1.0316,
-      "step": 5790
-    },
-    {
-      "epoch": 0.8860178885406315,
-      "grad_norm": 0.058224692824791244,
-      "learning_rate": 3.901053169475194e-05,
-      "loss": 1.0214,
-      "step": 5795
-    },
-    {
-      "epoch": 0.886782356088984,
-      "grad_norm": 0.062009025031709276,
-      "learning_rate": 3.849546030916473e-05,
-      "loss": 1.0556,
-      "step": 5800
-    },
-    {
-      "epoch": 0.8875468236373366,
-      "grad_norm": 0.05649898959762725,
-      "learning_rate": 3.798367573629113e-05,
-      "loss": 1.0078,
-      "step": 5805
-    },
-    {
-      "epoch": 0.8883112911856892,
-      "grad_norm": 0.05872883539211935,
-      "learning_rate": 3.747518162103597e-05,
-      "loss": 1.0756,
-      "step": 5810
-    },
-    {
-      "epoch": 0.8890757587340418,
-      "grad_norm": 0.059954064410168935,
-      "learning_rate": 3.696998158486925e-05,
-      "loss": 1.039,
-      "step": 5815
-    },
-    {
-      "epoch": 0.8898402262823943,
-      "grad_norm": 0.06002820321062443,
-      "learning_rate": 3.646807922580098e-05,
-      "loss": 0.9925,
-      "step": 5820
-    },
-    {
-      "epoch": 0.8906046938307469,
-      "grad_norm": 0.06355876238149161,
-      "learning_rate": 3.5969478118355156e-05,
-      "loss": 1.0486,
-      "step": 5825
-    },
-    {
-      "epoch": 0.8913691613790995,
-      "grad_norm": 0.05532906568084624,
-      "learning_rate": 3.5474181813544335e-05,
-      "loss": 1.0117,
-      "step": 5830
-    },
-    {
-      "epoch": 0.8921336289274521,
-      "grad_norm": 0.06045611131725339,
-      "learning_rate": 3.4982193838844544e-05,
-      "loss": 1.0701,
-      "step": 5835
-    },
-    {
-      "epoch": 0.8928980964758046,
-      "grad_norm": 0.0609764116986175,
-      "learning_rate": 3.4493517698170164e-05,
-      "loss": 1.0551,
-      "step": 5840
-    },
-    {
-      "epoch": 0.8936625640241572,
-      "grad_norm": 0.05428762758654894,
-      "learning_rate": 3.400815687184872e-05,
-      "loss": 1.0155,
-      "step": 5845
-    },
-    {
-      "epoch": 0.8944270315725098,
-      "grad_norm": 0.05965962849250994,
-      "learning_rate": 3.352611481659595e-05,
-      "loss": 1.0511,
-      "step": 5850
-    },
-    {
-      "epoch": 0.8951914991208623,
-      "grad_norm": 0.056939343112244536,
-      "learning_rate": 3.304739496549203e-05,
-      "loss": 1.0378,
-      "step": 5855
-    },
-    {
-      "epoch": 0.8959559666692148,
-      "grad_norm": 0.055743517095070166,
-      "learning_rate": 3.257200072795619e-05,
-      "loss": 1.0267,
-      "step": 5860
-    },
-    {
-      "epoch": 0.8967204342175674,
-      "grad_norm": 0.06219829393402533,
-      "learning_rate": 3.2099935489722896e-05,
-      "loss": 1.0407,
-      "step": 5865
-    },
-    {
-      "epoch": 0.89748490176592,
-      "grad_norm": 0.060687226345486525,
-      "learning_rate": 3.16312026128176e-05,
-      "loss": 1.0483,
-      "step": 5870
-    },
-    {
-      "epoch": 0.8982493693142726,
-      "grad_norm": 0.05981997075175033,
-      "learning_rate": 3.1165805435532936e-05,
-      "loss": 1.0382,
-      "step": 5875
-    },
-    {
-      "epoch": 0.8990138368626251,
-      "grad_norm": 0.057428646353623006,
-      "learning_rate": 3.070374727240466e-05,
-      "loss": 1.0479,
-      "step": 5880
-    },
-    {
-      "epoch": 0.8997783044109777,
-      "grad_norm": 0.06072172144354419,
-      "learning_rate": 3.0245031414188663e-05,
-      "loss": 0.9941,
-      "step": 5885
-    },
-    {
-      "epoch": 0.9005427719593303,
-      "grad_norm": 0.056470945217116215,
-      "learning_rate": 2.9789661127836432e-05,
-      "loss": 1.0396,
-      "step": 5890
-    },
-    {
-      "epoch": 0.9013072395076829,
-      "grad_norm": 0.05688971005581688,
-      "learning_rate": 2.933763965647307e-05,
-      "loss": 1.0013,
-      "step": 5895
-    },
-    {
-      "epoch": 0.9020717070560355,
-      "grad_norm": 0.06009286163449355,
-      "learning_rate": 2.8888970219373312e-05,
-      "loss": 1.0559,
-      "step": 5900
-    },
-    {
-      "epoch": 0.902836174604388,
-      "grad_norm": 0.07379788163258001,
-      "learning_rate": 2.8443656011938878e-05,
-      "loss": 0.9891,
-      "step": 5905
-    },
-    {
-      "epoch": 0.9036006421527406,
-      "grad_norm": 0.056877317710044964,
-      "learning_rate": 2.8001700205675663e-05,
-      "loss": 1.0763,
-      "step": 5910
-    },
-    {
-      "epoch": 0.9043651097010932,
-      "grad_norm": 0.05890117332064067,
-      "learning_rate": 2.7563105948171352e-05,
-      "loss": 1.0734,
-      "step": 5915
-    },
-    {
-      "epoch": 0.9051295772494458,
-      "grad_norm": 0.05414882138341149,
-      "learning_rate": 2.7127876363072734e-05,
-      "loss": 1.015,
-      "step": 5920
-    },
-    {
-      "epoch": 0.9058940447977983,
-      "grad_norm": 0.055796033504062224,
-      "learning_rate": 2.669601455006354e-05,
-      "loss": 1.0413,
-      "step": 5925
-    },
-    {
-      "epoch": 0.9066585123461509,
-      "grad_norm": 0.07287607712954004,
-      "learning_rate": 2.626752358484247e-05,
-      "loss": 1.0328,
-      "step": 5930
-    },
-    {
-      "epoch": 0.9074229798945035,
-      "grad_norm": 0.0544793234349126,
-      "learning_rate": 2.5842406519101135e-05,
-      "loss": 1.0159,
-      "step": 5935
-    },
-    {
-      "epoch": 0.9081874474428561,
-      "grad_norm": 0.061462858436265366,
-      "learning_rate": 2.54206663805025e-05,
-      "loss": 1.0265,
-      "step": 5940
-    },
-    {
-      "epoch": 0.9089519149912086,
-      "grad_norm": 0.05724469979574169,
-      "learning_rate": 2.500230617265925e-05,
-      "loss": 0.9856,
-      "step": 5945
-    },
-    {
-      "epoch": 0.9097163825395612,
-      "grad_norm": 0.0544995228115746,
-      "learning_rate": 2.4587328875112002e-05,
-      "loss": 0.9981,
-      "step": 5950
-    },
-    {
-      "epoch": 0.9104808500879138,
-      "grad_norm": 0.053865460569715765,
-      "learning_rate": 2.4175737443308975e-05,
-      "loss": 1.0465,
-      "step": 5955
-    },
-    {
-      "epoch": 0.9112453176362664,
-      "grad_norm": 0.055490991260585244,
-      "learning_rate": 2.3767534808584125e-05,
-      "loss": 0.9892,
-      "step": 5960
-    },
-    {
-      "epoch": 0.912009785184619,
-      "grad_norm": 0.058919461159295805,
-      "learning_rate": 2.3362723878136592e-05,
-      "loss": 1.0345,
-      "step": 5965
-    },
-    {
-      "epoch": 0.9127742527329715,
-      "grad_norm": 0.05703229353352163,
-      "learning_rate": 2.2961307535009966e-05,
-      "loss": 1.0403,
-      "step": 5970
-    },
-    {
-      "epoch": 0.9135387202813241,
-      "grad_norm": 0.06454466876602472,
-      "learning_rate": 2.2563288638071778e-05,
-      "loss": 0.9969,
-      "step": 5975
-    },
-    {
-      "epoch": 0.9143031878296767,
-      "grad_norm": 0.06269175867935511,
-      "learning_rate": 2.2168670021993076e-05,
-      "loss": 1.0319,
-      "step": 5980
-    },
-    {
-      "epoch": 0.9150676553780293,
-      "grad_norm": 0.054325191582656224,
-      "learning_rate": 2.1777454497228456e-05,
-      "loss": 1.0105,
-      "step": 5985
-    },
-    {
-      "epoch": 0.9158321229263817,
-      "grad_norm": 0.05741162522426686,
-      "learning_rate": 2.1389644849995506e-05,
-      "loss": 1.0486,
-      "step": 5990
-    },
-    {
-      "epoch": 0.9165965904747343,
-      "grad_norm": 0.059760672805936554,
-      "learning_rate": 2.100524384225555e-05,
-      "loss": 1.0196,
-      "step": 5995
-    },
-    {
-      "epoch": 0.9173610580230869,
-      "grad_norm": 0.05529093705054985,
-      "learning_rate": 2.0624254211693892e-05,
-      "loss": 1.0128,
-      "step": 6000
-    },
-    {
-      "epoch": 0.9181255255714394,
-      "grad_norm": 0.0560885235521163,
-      "learning_rate": 2.0246678671699936e-05,
-      "loss": 1.0228,
-      "step": 6005
-    },
-    {
-      "epoch": 0.918889993119792,
-      "grad_norm": 0.06287594156329433,
-      "learning_rate": 1.987251991134803e-05,
-      "loss": 1.0095,
-      "step": 6010
-    },
-    {
-      "epoch": 0.9196544606681446,
-      "grad_norm": 0.0571702237792438,
-      "learning_rate": 1.9501780595378715e-05,
-      "loss": 1.0274,
-      "step": 6015
-    },
-    {
-      "epoch": 0.9204189282164972,
-      "grad_norm": 0.06161643032987793,
-      "learning_rate": 1.9134463364179176e-05,
-      "loss": 1.0419,
-      "step": 6020
-    },
-    {
-      "epoch": 0.9211833957648498,
-      "grad_norm": 0.062217905827859084,
-      "learning_rate": 1.8770570833764712e-05,
-      "loss": 1.0231,
-      "step": 6025
-    },
-    {
-      "epoch": 0.9219478633132023,
-      "grad_norm": 0.056345001582428215,
-      "learning_rate": 1.8410105595760063e-05,
-      "loss": 1.0567,
-      "step": 6030
-    },
-    {
-      "epoch": 0.9227123308615549,
-      "grad_norm": 0.06041690355549645,
-      "learning_rate": 1.8053070217381007e-05,
-      "loss": 1.0428,
-      "step": 6035
-    },
-    {
-      "epoch": 0.9234767984099075,
-      "grad_norm": 0.059578321030006376,
-      "learning_rate": 1.7699467241416023e-05,
-      "loss": 1.0511,
-      "step": 6040
-    },
-    {
-      "epoch": 0.9242412659582601,
-      "grad_norm": 0.053569365750345925,
-      "learning_rate": 1.7349299186208258e-05,
-      "loss": 1.0511,
-      "step": 6045
-    },
-    {
-      "epoch": 0.9250057335066126,
-      "grad_norm": 0.055706266872641214,
-      "learning_rate": 1.7002568545637308e-05,
-      "loss": 1.0615,
-      "step": 6050
-    },
-    {
-      "epoch": 0.9257702010549652,
-      "grad_norm": 0.06210337329106453,
-      "learning_rate": 1.6659277789101757e-05,
-      "loss": 1.04,
-      "step": 6055
-    },
-    {
-      "epoch": 0.9265346686033178,
-      "grad_norm": 0.05720829659281351,
-      "learning_rate": 1.6319429361501713e-05,
-      "loss": 1.0181,
-      "step": 6060
-    },
-    {
-      "epoch": 0.9272991361516704,
-      "grad_norm": 0.05466393755249225,
-      "learning_rate": 1.5983025683220964e-05,
-      "loss": 1.024,
-      "step": 6065
-    },
-    {
-      "epoch": 0.9280636037000229,
-      "grad_norm": 0.06279571773285912,
-      "learning_rate": 1.5650069150110023e-05,
-      "loss": 1.0058,
-      "step": 6070
-    },
-    {
-      "epoch": 0.9288280712483755,
-      "grad_norm": 0.06240450137014157,
-      "learning_rate": 1.5320562133468997e-05,
-      "loss": 1.0495,
-      "step": 6075
-    },
-    {
-      "epoch": 0.9295925387967281,
-      "grad_norm": 0.0647440463087532,
-      "learning_rate": 1.4994506980030576e-05,
-      "loss": 1.0403,
-      "step": 6080
-    },
-    {
-      "epoch": 0.9303570063450807,
-      "grad_norm": 0.05313875552378329,
-      "learning_rate": 1.4671906011943848e-05,
-      "loss": 1.0331,
-      "step": 6085
-    },
-    {
-      "epoch": 0.9311214738934332,
-      "grad_norm": 0.05521374207034005,
-      "learning_rate": 1.4352761526756907e-05,
-      "loss": 1.0075,
-      "step": 6090
-    },
-    {
-      "epoch": 0.9318859414417858,
-      "grad_norm": 0.05982237948906845,
-      "learning_rate": 1.4037075797401155e-05,
-      "loss": 1.0629,
-      "step": 6095
-    },
-    {
-      "epoch": 0.9326504089901384,
-      "grad_norm": 0.06261396303304399,
-      "learning_rate": 1.3724851072174916e-05,
-      "loss": 1.0106,
-      "step": 6100
-    },
-    {
-      "epoch": 0.933414876538491,
-      "grad_norm": 0.05700687805485836,
-      "learning_rate": 1.3416089574727396e-05,
-      "loss": 1.0337,
-      "step": 6105
-    },
-    {
-      "epoch": 0.9341793440868436,
-      "grad_norm": 0.057915780300408407,
-      "learning_rate": 1.3110793504042873e-05,
-      "loss": 1.0059,
-      "step": 6110
-    },
-    {
-      "epoch": 0.9349438116351961,
-      "grad_norm": 0.05380257804448096,
-      "learning_rate": 1.2808965034424913e-05,
-      "loss": 1.0018,
-      "step": 6115
-    },
-    {
-      "epoch": 0.9357082791835487,
-      "grad_norm": 0.060812954266141986,
-      "learning_rate": 1.2510606315481121e-05,
-      "loss": 1.0407,
-      "step": 6120
-    },
-    {
-      "epoch": 0.9364727467319013,
-      "grad_norm": 0.059692939386243565,
-      "learning_rate": 1.22157194721077e-05,
-      "loss": 1.0539,
-      "step": 6125
-    },
-    {
-      "epoch": 0.9372372142802537,
-      "grad_norm": 0.06410611806486131,
-      "learning_rate": 1.1924306604474245e-05,
-      "loss": 1.0402,
-      "step": 6130
-    },
-    {
-      "epoch": 0.9380016818286063,
-      "grad_norm": 0.05657159861756608,
-      "learning_rate": 1.1636369788008972e-05,
-      "loss": 1.0383,
-      "step": 6135
-    },
-    {
-      "epoch": 0.9387661493769589,
-      "grad_norm": 0.06587261877884344,
-      "learning_rate": 1.1351911073383681e-05,
-      "loss": 1.0233,
-      "step": 6140
-    },
-    {
-      "epoch": 0.9395306169253115,
-      "grad_norm": 0.06201339266734815,
-      "learning_rate": 1.1070932486499542e-05,
-      "loss": 1.0289,
-      "step": 6145
-    },
-    {
-      "epoch": 0.9402950844736641,
-      "grad_norm": 0.07085274861735513,
-      "learning_rate": 1.0793436028472214e-05,
-      "loss": 1.0377,
-      "step": 6150
-    },
-    {
-      "epoch": 0.9410595520220166,
-      "grad_norm": 0.056399221307596455,
-      "learning_rate": 1.0519423675617811e-05,
-      "loss": 1.0336,
-      "step": 6155
-    },
-    {
-      "epoch": 0.9418240195703692,
-      "grad_norm": 0.05838303840904757,
-      "learning_rate": 1.0248897379438905e-05,
-      "loss": 1.0088,
-      "step": 6160
-    },
-    {
-      "epoch": 0.9425884871187218,
-      "grad_norm": 0.05511937754131694,
-      "learning_rate": 9.981859066610643e-06,
-      "loss": 1.0325,
-      "step": 6165
-    },
-    {
-      "epoch": 0.9433529546670744,
-      "grad_norm": 0.05862068288977634,
-      "learning_rate": 9.718310638966609e-06,
-      "loss": 1.0566,
-      "step": 6170
-    },
-    {
-      "epoch": 0.9441174222154269,
-      "grad_norm": 0.06229665145510315,
-      "learning_rate": 9.45825397348593e-06,
-      "loss": 1.0129,
-      "step": 6175
-    },
-    {
-      "epoch": 0.9448818897637795,
-      "grad_norm": 0.058461286059371405,
-      "learning_rate": 9.201690922279405e-06,
-      "loss": 1.0108,
-      "step": 6180
-    },
-    {
-      "epoch": 0.9456463573121321,
-      "grad_norm": 0.060236921179087834,
-      "learning_rate": 8.948623312576454e-06,
-      "loss": 0.9994,
-      "step": 6185
-    },
-    {
-      "epoch": 0.9464108248604847,
-      "grad_norm": 0.061654492292078825,
-      "learning_rate": 8.699052946712248e-06,
-      "loss": 1.0435,
-      "step": 6190
-    },
-    {
-      "epoch": 0.9471752924088372,
-      "grad_norm": 0.05414771023362534,
-      "learning_rate": 8.452981602114717e-06,
-      "loss": 1.0159,
-      "step": 6195
-    },
-    {
-      "epoch": 0.9479397599571898,
-      "grad_norm": 0.05913688312421346,
-      "learning_rate": 8.210411031291776e-06,
-      "loss": 0.9874,
-      "step": 6200
-    },
-    {
-      "epoch": 0.9487042275055424,
-      "grad_norm": 0.05691389065492228,
-      "learning_rate": 7.97134296181934e-06,
-      "loss": 1.0112,
-      "step": 6205
-    },
-    {
-      "epoch": 0.949468695053895,
-      "grad_norm": 0.055429285880686574,
-      "learning_rate": 7.735779096328389e-06,
-      "loss": 1.017,
-      "step": 6210
-    },
-    {
-      "epoch": 0.9502331626022475,
-      "grad_norm": 0.06097722521863537,
-      "learning_rate": 7.5037211124932e-06,
-      "loss": 1.0011,
-      "step": 6215
-    },
     {
       "epoch": 0.9509976301506001,
-      "grad_norm": 0.05532466685476817,
-      "learning_rate": 7.275170663019415e-06,
-      "loss": 1.0269,
-      "step": 6220
-    },
-    {
-      "epoch": 0.9517620976989527,
-      "grad_norm": 0.06042708468611293,
-      "learning_rate": 7.0501293756320975e-06,
-      "loss": 1.0602,
-      "step": 6225
-    },
-    {
-      "epoch": 0.9525265652473053,
-      "grad_norm": 0.05543873760281028,
-      "learning_rate": 6.828598853064249e-06,
-      "loss": 1.0048,
-      "step": 6230
-    },
-    {
-      "epoch": 0.9532910327956579,
-      "grad_norm": 0.0674868225950688,
-      "learning_rate": 6.610580673045485e-06,
-      "loss": 1.0069,
-      "step": 6235
+      "grad_norm": 0.019644926869063512,
+      "learning_rate": 7.280105195860254e-06,
+      "loss": 1.0306,
+      "step": 1555
     },
     {
       "epoch": 0.9540555003440104,
-      "grad_norm": 0.05549130127970468,
-      "learning_rate": 6.396076388290484e-06,
-      "loss": 0.9908,
-      "step": 6240
-    },
-    {
-      "epoch": 0.954819967892363,
-      "grad_norm": 0.055652396632378566,
-      "learning_rate": 6.185087526488331e-06,
-      "loss": 1.0001,
-      "step": 6245
-    },
-    {
-      "epoch": 0.9555844354407156,
-      "grad_norm": 0.05267551540657268,
-      "learning_rate": 5.977615590291363e-06,
-      "loss": 1.0544,
-      "step": 6250
-    },
-    {
-      "epoch": 0.9563489029890682,
-      "grad_norm": 0.05592750269276289,
-      "learning_rate": 5.773662057304452e-06,
-      "loss": 1.0415,
-      "step": 6255
+      "grad_norm": 0.016710041757227387,
+      "learning_rate": 6.400415939745275e-06,
+      "loss": 1.0281,
+      "step": 1560
     },
     {
       "epoch": 0.9571133705374207,
-      "grad_norm": 0.05646475680673014,
-      "learning_rate": 5.573228380074735e-06,
-      "loss": 1.0531,
-      "step": 6260
-    },
-    {
-      "epoch": 0.9578778380857733,
-      "grad_norm": 0.06321256859481447,
-      "learning_rate": 5.376315986081071e-06,
-      "loss": 0.9959,
-      "step": 6265
-    },
-    {
-      "epoch": 0.9586423056341258,
-      "grad_norm": 0.06453658248744666,
-      "learning_rate": 5.1829262777238205e-06,
-      "loss": 1.0466,
-      "step": 6270
-    },
-    {
-      "epoch": 0.9594067731824784,
-      "grad_norm": 0.05935378265190903,
-      "learning_rate": 4.993060632314972e-06,
-      "loss": 1.0193,
-      "step": 6275
+      "grad_norm": 0.018527383142328113,
+      "learning_rate": 5.577010698204488e-06,
+      "loss": 1.0493,
+      "step": 1565
     },
     {
       "epoch": 0.9601712407308309,
-      "grad_norm": 0.056930907655133194,
-      "learning_rate": 4.806720402068476e-06,
-      "loss": 1.0026,
-      "step": 6280
-    },
-    {
-      "epoch": 0.9609357082791835,
-      "grad_norm": 0.05573893520025273,
-      "learning_rate": 4.623906914090203e-06,
-      "loss": 0.9942,
-      "step": 6285
-    },
-    {
-      "epoch": 0.9617001758275361,
-      "grad_norm": 0.059293037343727206,
-      "learning_rate": 4.444621470368893e-06,
-      "loss": 1.0574,
-      "step": 6290
-    },
-    {
-      "epoch": 0.9624646433758887,
-      "grad_norm": 0.057319030664173376,
-      "learning_rate": 4.268865347766715e-06,
-      "loss": 1.047,
-      "step": 6295
+      "grad_norm": 0.01587941593451175,
+      "learning_rate": 4.809983362225878e-06,
+      "loss": 1.0287,
+      "step": 1570
     },
     {
       "epoch": 0.9632291109242412,
-      "grad_norm": 0.05986032444262332,
-      "learning_rate": 4.09663979801006e-06,
-      "loss": 1.0445,
-      "step": 6300
-    },
-    {
-      "epoch": 0.9639935784725938,
-      "grad_norm": 0.06772012126843469,
-      "learning_rate": 3.927946047680986e-06,
-      "loss": 1.0249,
-      "step": 6305
-    },
-    {
-      "epoch": 0.9647580460209464,
-      "grad_norm": 0.052915625380592655,
-      "learning_rate": 3.7627852982081156e-06,
-      "loss": 1.0115,
-      "step": 6310
-    },
-    {
-      "epoch": 0.965522513569299,
-      "grad_norm": 0.05719483772507953,
-      "learning_rate": 3.601158725858034e-06,
-      "loss": 1.0089,
-      "step": 6315
+      "grad_norm": 0.01784011524370286,
+      "learning_rate": 4.0994213941557e-06,
+      "loss": 1.0494,
+      "step": 1575
     },
     {
       "epoch": 0.9662869811176515,
-      "grad_norm": 0.06742198460058717,
-      "learning_rate": 3.4430674817274575e-06,
-      "loss": 0.9958,
-      "step": 6320
-    },
-    {
-      "epoch": 0.9670514486660041,
-      "grad_norm": 0.05934089002940797,
-      "learning_rate": 3.288512691734413e-06,
-      "loss": 1.0607,
-      "step": 6325
-    },
-    {
-      "epoch": 0.9678159162143567,
-      "grad_norm": 0.05426755164335042,
-      "learning_rate": 3.1374954566105173e-06,
-      "loss": 1.0188,
-      "step": 6330
-    },
-    {
-      "epoch": 0.9685803837627093,
-      "grad_norm": 0.05770507635439541,
-      "learning_rate": 2.9900168518931513e-06,
-      "loss": 1.0476,
-      "step": 6335
+      "grad_norm": 0.022229608533606754,
+      "learning_rate": 3.4454058177253998e-06,
+      "loss": 1.0245,
+      "step": 1580
     },
     {
       "epoch": 0.9693448513110619,
-      "grad_norm": 0.061035126427298775,
-      "learning_rate": 2.8460779279176897e-06,
-      "loss": 1.0285,
-      "step": 6340
-    },
-    {
-      "epoch": 0.9701093188594144,
-      "grad_norm": 0.05994621621761424,
-      "learning_rate": 2.705679709810116e-06,
-      "loss": 1.0273,
-      "step": 6345
-    },
-    {
-      "epoch": 0.970873786407767,
-      "grad_norm": 0.05622211095511261,
-      "learning_rate": 2.5688231974796418e-06,
-      "loss": 1.0306,
-      "step": 6350
-    },
-    {
-      "epoch": 0.9716382539561196,
-      "grad_norm": 0.05672371489636327,
-      "learning_rate": 2.4355093656116546e-06,
-      "loss": 1.0404,
-      "step": 6355
+      "grad_norm": 0.01771675468251995,
+      "learning_rate": 2.8480112088128396e-06,
+      "loss": 1.0539,
+      "step": 1585
     },
     {
       "epoch": 0.9724027215044722,
-      "grad_norm": 0.05819240760299045,
-      "learning_rate": 2.3057391636606697e-06,
-      "loss": 0.9942,
-      "step": 6360
-    },
-    {
-      "epoch": 0.9731671890528247,
-      "grad_norm": 0.053968015068043766,
-      "learning_rate": 2.1795135158438338e-06,
-      "loss": 1.0023,
-      "step": 6365
-    },
-    {
-      "epoch": 0.9739316566011773,
-      "grad_norm": 0.05692302172149137,
-      "learning_rate": 2.0568333211338774e-06,
-      "loss": 1.0448,
-      "step": 6370
-    },
-    {
-      "epoch": 0.9746961241495299,
-      "grad_norm": 0.056053764675146125,
-      "learning_rate": 1.937699453253228e-06,
-      "loss": 1.0052,
-      "step": 6375
+      "grad_norm": 0.019318982313098143,
+      "learning_rate": 2.307305686938266e-06,
+      "loss": 1.0364,
+      "step": 1590
     },
     {
       "epoch": 0.9754605916978825,
-      "grad_norm": 0.05688214554755461,
-      "learning_rate": 1.8221127606674603e-06,
-      "loss": 1.0529,
-      "step": 6380
-    },
-    {
-      "epoch": 0.976225059246235,
-      "grad_norm": 0.059143411541020596,
-      "learning_rate": 1.710074066579248e-06,
-      "loss": 1.0184,
-      "step": 6385
-    },
-    {
-      "epoch": 0.9769895267945876,
-      "grad_norm": 0.05396274376529374,
-      "learning_rate": 1.6015841689227539e-06,
-      "loss": 1.0269,
-      "step": 6390
-    },
-    {
-      "epoch": 0.9777539943429402,
-      "grad_norm": 0.06599763151002173,
-      "learning_rate": 1.4966438403577475e-06,
-      "loss": 1.0267,
-      "step": 6395
+      "grad_norm": 0.01763551427180363,
+      "learning_rate": 1.8233509074969124e-06,
+      "loss": 1.0387,
+      "step": 1595
     },
     {
       "epoch": 0.9785184618912928,
-      "grad_norm": 0.054046587912488314,
-      "learning_rate": 1.3952538282639982e-06,
-      "loss": 1.0196,
-      "step": 6400
-    },
-    {
-      "epoch": 0.9792829294396452,
-      "grad_norm": 0.05563950879198749,
-      "learning_rate": 1.297414854736223e-06,
-      "loss": 1.0423,
-      "step": 6405
-    },
-    {
-      "epoch": 0.9800473969879978,
-      "grad_norm": 0.05499476733743179,
-      "learning_rate": 1.2031276165789252e-06,
-      "loss": 1.0444,
-      "step": 6410
-    },
-    {
-      "epoch": 0.9808118645363504,
-      "grad_norm": 0.05728822187639338,
-      "learning_rate": 1.1123927853010085e-06,
-      "loss": 1.008,
-      "step": 6415
+      "grad_norm": 0.020265580264344555,
+      "learning_rate": 1.3962020547287345e-06,
+      "loss": 1.0375,
+      "step": 1600
     },
     {
       "epoch": 0.981576332084703,
-      "grad_norm": 0.05951022694716498,
-      "learning_rate": 1.025211007111615e-06,
-      "loss": 1.0416,
-      "step": 6420
-    },
-    {
-      "epoch": 0.9823407996330555,
-      "grad_norm": 0.05918659466424548,
-      "learning_rate": 9.415829029150724e-07,
-      "loss": 1.0156,
-      "step": 6425
-    },
-    {
-      "epoch": 0.9831052671814081,
-      "grad_norm": 0.056548103687334934,
-      "learning_rate": 8.615090683066206e-07,
-      "loss": 1.0567,
-      "step": 6430
-    },
-    {
-      "epoch": 0.9838697347297607,
-      "grad_norm": 0.06106341663875317,
-      "learning_rate": 7.84990073568248e-07,
-      "loss": 1.0436,
-      "step": 6435
+      "grad_norm": 0.017705249951308937,
+      "learning_rate": 1.0259078354257212e-06,
+      "loss": 1.0463,
+      "step": 1605
     },
     {
       "epoch": 0.9846342022781133,
-      "grad_norm": 0.05745985936443068,
-      "learning_rate": 7.120264636643614e-07,
-      "loss": 1.0382,
-      "step": 6440
-    },
-    {
-      "epoch": 0.9853986698264658,
-      "grad_norm": 0.06113618554750734,
-      "learning_rate": 6.426187582383447e-07,
-      "loss": 0.9907,
-      "step": 6445
-    },
-    {
-      "epoch": 0.9861631373748184,
-      "grad_norm": 0.060054142652089436,
-      "learning_rate": 5.767674516083954e-07,
-      "loss": 1.0316,
-      "step": 6450
-    },
-    {
-      "epoch": 0.986927604923171,
-      "grad_norm": 0.05945875123547078,
-      "learning_rate": 5.144730127643605e-07,
-      "loss": 1.0362,
-      "step": 6455
+      "grad_norm": 0.01899318507681047,
+      "learning_rate": 7.12510473377892e-07,
+      "loss": 1.0499,
+      "step": 1610
     },
     {
       "epoch": 0.9876920724715236,
-      "grad_norm": 0.06106867404334743,
-      "learning_rate": 4.5573588536407253e-07,
-      "loss": 1.0046,
-      "step": 6460
-    },
-    {
-      "epoch": 0.9884565400198762,
-      "grad_norm": 0.058622274212216934,
-      "learning_rate": 4.005564877305745e-07,
-      "loss": 1.0808,
-      "step": 6465
-    },
-    {
-      "epoch": 0.9892210075682287,
-      "grad_norm": 0.055370471737170614,
-      "learning_rate": 3.4893521284878884e-07,
-      "loss": 1.0308,
-      "step": 6470
-    },
-    {
-      "epoch": 0.9899854751165813,
-      "grad_norm": 0.058651345695096446,
-      "learning_rate": 3.0087242836285275e-07,
-      "loss": 1.0341,
-      "step": 6475
+      "grad_norm": 0.20527421150604633,
+      "learning_rate": 4.5604570455887175e-07,
+      "loss": 1.0321,
+      "step": 1615
     },
     {
       "epoch": 0.9907499426649339,
-      "grad_norm": 0.05881488906932957,
-      "learning_rate": 2.5636847657367624e-07,
-      "loss": 1.0514,
-      "step": 6480
-    },
-    {
-      "epoch": 0.9915144102132865,
-      "grad_norm": 0.05718945757039933,
-      "learning_rate": 2.1542367443616595e-07,
-      "loss": 1.0102,
-      "step": 6485
-    },
-    {
-      "epoch": 0.992278877761639,
-      "grad_norm": 0.05591730925908283,
-      "learning_rate": 1.780383135571717e-07,
-      "loss": 1.0249,
-      "step": 6490
-    },
-    {
-      "epoch": 0.9930433453099916,
-      "grad_norm": 0.06348320627570112,
-      "learning_rate": 1.4421266019348787e-07,
-      "loss": 1.057,
-      "step": 6495
+      "grad_norm": 0.01969626150373465,
+      "learning_rate": 2.565427730508163e-07,
+      "loss": 1.0622,
+      "step": 1620
     },
     {
       "epoch": 0.9938078128583442,
-      "grad_norm": 0.06390840223331544,
-      "learning_rate": 1.1394695524963305e-07,
-      "loss": 1.0271,
-      "step": 6500
-    },
-    {
-      "epoch": 0.9945722804066968,
-      "grad_norm": 0.1081874050336723,
-      "learning_rate": 8.724141427657317e-08,
-      "loss": 1.0346,
-      "step": 6505
-    },
-    {
-      "epoch": 0.9953367479550493,
-      "grad_norm": 0.05829569266484604,
-      "learning_rate": 6.409622746977872e-08,
-      "loss": 1.0286,
-      "step": 6510
-    },
-    {
-      "epoch": 0.9961012155034019,
-      "grad_norm": 0.0578107659452591,
-      "learning_rate": 4.4511559668225507e-08,
-      "loss": 1.0164,
-      "step": 6515
+      "grad_norm": 0.01846687245559012,
+      "learning_rate": 1.1402442770985788e-07,
+      "loss": 1.0443,
+      "step": 1625
     },
     {
       "epoch": 0.9968656830517545,
-      "grad_norm": 0.06124806496876297,
-      "learning_rate": 2.8487550352951362e-08,
-      "loss": 1.0361,
-      "step": 6520
-    },
-    {
-      "epoch": 0.9976301506001071,
-      "grad_norm": 0.0593758021573862,
-      "learning_rate": 1.6024313646056942e-08,
-      "loss": 1.0053,
-      "step": 6525
-    },
-    {
-      "epoch": 0.9983946181484596,
-      "grad_norm": 0.06060901602693129,
-      "learning_rate": 7.121938310261644e-09,
-      "loss": 1.0459,
-      "step": 6530
-    },
-    {
-      "epoch": 0.9991590856968122,
-      "grad_norm": 0.05833116627694336,
-      "learning_rate": 1.7804877476823578e-09,
-      "loss": 1.0245,
-      "step": 6535
+      "grad_norm": 0.01754718213075641,
+      "learning_rate": 2.8506919571902325e-08,
+      "loss": 1.0413,
+      "step": 1630
     },
     {
       "epoch": 0.9999235532451648,
-      "grad_norm": 0.055856565745703204,
+      "grad_norm": 0.01654015476154834,
       "learning_rate": 0.0,
-      "loss": 0.9952,
-      "step": 6540
+      "loss": 1.031,
+      "step": 1635
     },
     {
       "epoch": 0.9999235532451648,
-      "eval_loss": 1.4338810443878174,
-      "eval_runtime": 608.009,
-      "eval_samples_per_second": 34.46,
-      "eval_steps_per_second": 8.615,
-      "step": 6540
+      "eval_loss": 1.415837049484253,
+      "eval_runtime": 621.6735,
+      "eval_samples_per_second": 33.703,
+      "eval_steps_per_second": 8.426,
+      "step": 1635
     },
     {
       "epoch": 0.9999235532451648,
-      "step": 6540,
-      "total_flos": 1.024299027881001e+17,
-      "train_loss": 1.1466178722155569,
-      "train_runtime": 11966.5261,
-      "train_samples_per_second": 8.745,
-      "train_steps_per_second": 0.547
+      "step": 1635,
+      "total_flos": 1.1949402256048128e+17,
+      "train_loss": 1.1248935338554031,
+      "train_runtime": 10312.3192,
+      "train_samples_per_second": 10.148,
+      "train_steps_per_second": 0.159
     }
   ],
   "logging_steps": 5,
-  "max_steps": 6540,
+  "max_steps": 1635,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
@@ -9206,8 +2339,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.024299027881001e+17,
-  "train_batch_size": 4,
+  "total_flos": 1.1949402256048128e+17,
+  "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null
 }