diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.7987685418415897,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 10000,
+  "global_step": 10719,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -70008,6 +70008,5039 @@
       "learning_rate": 1.2284763283065159e-06,
       "loss": 0.4598,
       "step": 10000
+    },
+    {
+      "epoch": 2.799048418695774,
+      "grad_norm": 0.2623289816675507,
+      "learning_rate": 1.225080257965755e-06,
+      "loss": 0.4304,
+      "step": 10001
+    },
+    {
+      "epoch": 2.799328295549958,
+      "grad_norm": 0.25573096800671996,
+      "learning_rate": 1.2216888300491559e-06,
+      "loss": 0.4428,
+      "step": 10002
+    },
+    {
+      "epoch": 2.799608172404142,
+      "grad_norm": 0.2668555762116913,
+      "learning_rate": 1.2183020448795212e-06,
+      "loss": 0.4618,
+      "step": 10003
+    },
+    {
+      "epoch": 2.7998880492583265,
+      "grad_norm": 0.26229813824137216,
+      "learning_rate": 1.2149199027791992e-06,
+      "loss": 0.4494,
+      "step": 10004
+    },
+    {
+      "epoch": 2.8001679261125103,
+      "grad_norm": 0.25880574709037374,
+      "learning_rate": 1.211542404070104e-06,
+      "loss": 0.4527,
+      "step": 10005
+    },
+    {
+      "epoch": 2.8004478029666946,
+      "grad_norm": 0.24930135750690222,
+      "learning_rate": 1.2081695490737178e-06,
+      "loss": 0.4423,
+      "step": 10006
+    },
+    {
+      "epoch": 2.800727679820879,
+      "grad_norm": 0.25562693305388845,
+      "learning_rate": 1.2048013381110611e-06,
+      "loss": 0.4355,
+      "step": 10007
+    },
+    {
+      "epoch": 2.8010075566750627,
+      "grad_norm": 0.2573744662139402,
+      "learning_rate": 1.2014377715027225e-06,
+      "loss": 0.4669,
+      "step": 10008
+    },
+    {
+      "epoch": 2.801287433529247,
+      "grad_norm": 0.2521840692937912,
+      "learning_rate": 1.198078849568851e-06,
+      "loss": 0.4362,
+      "step": 10009
+    },
+    {
+      "epoch": 2.8015673103834313,
+      "grad_norm": 0.25530231186014696,
+      "learning_rate": 1.1947245726291523e-06,
+      "loss": 0.418,
+      "step": 10010
+    },
+    {
+      "epoch": 2.8018471872376156,
+      "grad_norm": 0.25835305322672464,
+      "learning_rate": 1.1913749410028873e-06,
+      "loss": 0.4524,
+      "step": 10011
+    },
+    {
+      "epoch": 2.8021270640917995,
+      "grad_norm": 0.25387526679496275,
+      "learning_rate": 1.188029955008868e-06,
+      "loss": 0.4566,
+      "step": 10012
+    },
+    {
+      "epoch": 2.8024069409459837,
+      "grad_norm": 0.25688080755971887,
+      "learning_rate": 1.1846896149654785e-06,
+      "loss": 0.4442,
+      "step": 10013
+    },
+    {
+      "epoch": 2.802686817800168,
+      "grad_norm": 0.2509635911430038,
+      "learning_rate": 1.1813539211906587e-06,
+      "loss": 0.4494,
+      "step": 10014
+    },
+    {
+      "epoch": 2.8029666946543523,
+      "grad_norm": 0.2552682125098111,
+      "learning_rate": 1.1780228740018994e-06,
+      "loss": 0.4557,
+      "step": 10015
+    },
+    {
+      "epoch": 2.803246571508536,
+      "grad_norm": 0.26217435155468116,
+      "learning_rate": 1.174696473716247e-06,
+      "loss": 0.4307,
+      "step": 10016
+    },
+    {
+      "epoch": 2.8035264483627205,
+      "grad_norm": 0.2523713093152466,
+      "learning_rate": 1.1713747206503145e-06,
+      "loss": 0.4498,
+      "step": 10017
+    },
+    {
+      "epoch": 2.8038063252169048,
+      "grad_norm": 0.25678219498147636,
+      "learning_rate": 1.1680576151202717e-06,
+      "loss": 0.4608,
+      "step": 10018
+    },
+    {
+      "epoch": 2.8040862020710886,
+      "grad_norm": 0.2531984694325335,
+      "learning_rate": 1.1647451574418321e-06,
+      "loss": 0.4462,
+      "step": 10019
+    },
+    {
+      "epoch": 2.804366078925273,
+      "grad_norm": 0.2519919251645383,
+      "learning_rate": 1.1614373479302832e-06,
+      "loss": 0.4429,
+      "step": 10020
+    },
+    {
+      "epoch": 2.804645955779457,
+      "grad_norm": 0.2537678668261987,
+      "learning_rate": 1.1581341869004613e-06,
+      "loss": 0.4659,
+      "step": 10021
+    },
+    {
+      "epoch": 2.804925832633641,
+      "grad_norm": 0.25675874919280917,
+      "learning_rate": 1.1548356746667655e-06,
+      "loss": 0.435,
+      "step": 10022
+    },
+    {
+      "epoch": 2.8052057094878253,
+      "grad_norm": 0.2579020260027916,
+      "learning_rate": 1.1515418115431553e-06,
+      "loss": 0.4521,
+      "step": 10023
+    },
+    {
+      "epoch": 2.8054855863420096,
+      "grad_norm": 0.2618712561245846,
+      "learning_rate": 1.1482525978431357e-06,
+      "loss": 0.4553,
+      "step": 10024
+    },
+    {
+      "epoch": 2.8057654631961935,
+      "grad_norm": 0.26263634255425594,
+      "learning_rate": 1.1449680338797786e-06,
+      "loss": 0.4383,
+      "step": 10025
+    },
+    {
+      "epoch": 2.8060453400503778,
+      "grad_norm": 0.25277594804249975,
+      "learning_rate": 1.1416881199657059e-06,
+      "loss": 0.4436,
+      "step": 10026
+    },
+    {
+      "epoch": 2.806325216904562,
+      "grad_norm": 0.24431848116031987,
+      "learning_rate": 1.1384128564131069e-06,
+      "loss": 0.4566,
+      "step": 10027
+    },
+    {
+      "epoch": 2.806605093758746,
+      "grad_norm": 0.2609154288450035,
+      "learning_rate": 1.1351422435337211e-06,
+      "loss": 0.4437,
+      "step": 10028
+    },
+    {
+      "epoch": 2.80688497061293,
+      "grad_norm": 0.25695753681581046,
+      "learning_rate": 1.131876281638844e-06,
+      "loss": 0.432,
+      "step": 10029
+    },
+    {
+      "epoch": 2.8071648474671145,
+      "grad_norm": 0.2538585151636701,
+      "learning_rate": 1.1286149710393323e-06,
+      "loss": 0.4597,
+      "step": 10030
+    },
+    {
+      "epoch": 2.8074447243212988,
+      "grad_norm": 0.23442293755687932,
+      "learning_rate": 1.1253583120456102e-06,
+      "loss": 0.4393,
+      "step": 10031
+    },
+    {
+      "epoch": 2.807724601175483,
+      "grad_norm": 0.24571792553483443,
+      "learning_rate": 1.122106304967635e-06,
+      "loss": 0.4572,
+      "step": 10032
+    },
+    {
+      "epoch": 2.808004478029667,
+      "grad_norm": 0.24975815194917134,
+      "learning_rate": 1.1188589501149427e-06,
+      "loss": 0.4537,
+      "step": 10033
+    },
+    {
+      "epoch": 2.808284354883851,
+      "grad_norm": 0.2522156074258436,
+      "learning_rate": 1.1156162477966137e-06,
+      "loss": 0.4228,
+      "step": 10034
+    },
+    {
+      "epoch": 2.8085642317380355,
+      "grad_norm": 0.25234913236963324,
+      "learning_rate": 1.1123781983212955e-06,
+      "loss": 0.4632,
+      "step": 10035
+    },
+    {
+      "epoch": 2.8088441085922193,
+      "grad_norm": 0.2617953931586556,
+      "learning_rate": 1.1091448019971806e-06,
+      "loss": 0.4499,
+      "step": 10036
+    },
+    {
+      "epoch": 2.8091239854464036,
+      "grad_norm": 0.268164279012252,
+      "learning_rate": 1.1059160591320338e-06,
+      "loss": 0.4774,
+      "step": 10037
+    },
+    {
+      "epoch": 2.809403862300588,
+      "grad_norm": 0.2532397829516345,
+      "learning_rate": 1.1026919700331596e-06,
+      "loss": 0.4791,
+      "step": 10038
+    },
+    {
+      "epoch": 2.8096837391547718,
+      "grad_norm": 0.25454581027538686,
+      "learning_rate": 1.0994725350074342e-06,
+      "loss": 0.4517,
+      "step": 10039
+    },
+    {
+      "epoch": 2.809963616008956,
+      "grad_norm": 0.2641984689771092,
+      "learning_rate": 1.0962577543612795e-06,
+      "loss": 0.4472,
+      "step": 10040
+    },
+    {
+      "epoch": 2.8102434928631403,
+      "grad_norm": 0.36658894526412195,
+      "learning_rate": 1.0930476284006897e-06,
+      "loss": 0.4572,
+      "step": 10041
+    },
+    {
+      "epoch": 2.810523369717324,
+      "grad_norm": 0.25492350997962254,
+      "learning_rate": 1.0898421574312034e-06,
+      "loss": 0.4352,
+      "step": 10042
+    },
+    {
+      "epoch": 2.8108032465715085,
+      "grad_norm": 0.2533002255898336,
+      "learning_rate": 1.0866413417579103e-06,
+      "loss": 0.4584,
+      "step": 10043
+    },
+    {
+      "epoch": 2.8110831234256928,
+      "grad_norm": 0.25750211312832716,
+      "learning_rate": 1.0834451816854775e-06,
+      "loss": 0.444,
+      "step": 10044
+    },
+    {
+      "epoch": 2.8113630002798766,
+      "grad_norm": 0.25995239658813213,
+      "learning_rate": 1.080253677518106e-06,
+      "loss": 0.4479,
+      "step": 10045
+    },
+    {
+      "epoch": 2.811642877134061,
+      "grad_norm": 0.25958867059504953,
+      "learning_rate": 1.0770668295595755e-06,
+      "loss": 0.4433,
+      "step": 10046
+    },
+    {
+      "epoch": 2.811922753988245,
+      "grad_norm": 0.26457879133391693,
+      "learning_rate": 1.0738846381131983e-06,
+      "loss": 0.4362,
+      "step": 10047
+    },
+    {
+      "epoch": 2.8122026308424295,
+      "grad_norm": 0.2597805161367181,
+      "learning_rate": 1.0707071034818773e-06,
+      "loss": 0.4655,
+      "step": 10048
+    },
+    {
+      "epoch": 2.8124825076966133,
+      "grad_norm": 0.2573105787220753,
+      "learning_rate": 1.0675342259680366e-06,
+      "loss": 0.4514,
+      "step": 10049
+    },
+    {
+      "epoch": 2.8127623845507976,
+      "grad_norm": 0.2626867445065791,
+      "learning_rate": 1.0643660058736793e-06,
+      "loss": 0.4468,
+      "step": 10050
+    },
+    {
+      "epoch": 2.813042261404982,
+      "grad_norm": 0.25000800106284804,
+      "learning_rate": 1.0612024435003531e-06,
+      "loss": 0.4472,
+      "step": 10051
+    },
+    {
+      "epoch": 2.813322138259166,
+      "grad_norm": 0.2588721745046232,
+      "learning_rate": 1.0580435391491728e-06,
+      "loss": 0.4349,
+      "step": 10052
+    },
+    {
+      "epoch": 2.81360201511335,
+      "grad_norm": 0.25125222756847393,
+      "learning_rate": 1.0548892931208033e-06,
+      "loss": 0.4439,
+      "step": 10053
+    },
+    {
+      "epoch": 2.8138818919675344,
+      "grad_norm": 0.24978774064115405,
+      "learning_rate": 1.0517397057154654e-06,
+      "loss": 0.4489,
+      "step": 10054
+    },
+    {
+      "epoch": 2.8141617688217186,
+      "grad_norm": 0.25557688299734493,
+      "learning_rate": 1.048594777232942e-06,
+      "loss": 0.4524,
+      "step": 10055
+    },
+    {
+      "epoch": 2.8144416456759025,
+      "grad_norm": 0.2553222781900778,
+      "learning_rate": 1.0454545079725654e-06,
+      "loss": 0.442,
+      "step": 10056
+    },
+    {
+      "epoch": 2.814721522530087,
+      "grad_norm": 0.2625996339854685,
+      "learning_rate": 1.04231889823323e-06,
+      "loss": 0.4616,
+      "step": 10057
+    },
+    {
+      "epoch": 2.815001399384271,
+      "grad_norm": 0.26536237523411943,
+      "learning_rate": 1.0391879483133804e-06,
+      "loss": 0.4553,
+      "step": 10058
+    },
+    {
+      "epoch": 2.815281276238455,
+      "grad_norm": 0.2518798493314616,
+      "learning_rate": 1.0360616585110338e-06,
+      "loss": 0.444,
+      "step": 10059
+    },
+    {
+      "epoch": 2.815561153092639,
+      "grad_norm": 0.2674145425133259,
+      "learning_rate": 1.0329400291237413e-06,
+      "loss": 0.4639,
+      "step": 10060
+    },
+    {
+      "epoch": 2.8158410299468235,
+      "grad_norm": 0.25435428497609625,
+      "learning_rate": 1.0298230604486258e-06,
+      "loss": 0.449,
+      "step": 10061
+    },
+    {
+      "epoch": 2.8161209068010074,
+      "grad_norm": 0.2590284413753409,
+      "learning_rate": 1.0267107527823617e-06,
+      "loss": 0.4425,
+      "step": 10062
+    },
+    {
+      "epoch": 2.8164007836551916,
+      "grad_norm": 0.25201643767313303,
+      "learning_rate": 1.023603106421178e-06,
+      "loss": 0.4552,
+      "step": 10063
+    },
+    {
+      "epoch": 2.816680660509376,
+      "grad_norm": 0.26165099284783544,
+      "learning_rate": 1.0205001216608612e-06,
+      "loss": 0.4538,
+      "step": 10064
+    },
+    {
+      "epoch": 2.81696053736356,
+      "grad_norm": 0.27776039496702404,
+      "learning_rate": 1.0174017987967577e-06,
+      "loss": 0.4715,
+      "step": 10065
+    },
+    {
+      "epoch": 2.817240414217744,
+      "grad_norm": 0.25777329208975347,
+      "learning_rate": 1.0143081381237706e-06,
+      "loss": 0.4371,
+      "step": 10066
+    },
+    {
+      "epoch": 2.8175202910719284,
+      "grad_norm": 0.2690546753062122,
+      "learning_rate": 1.0112191399363534e-06,
+      "loss": 0.4494,
+      "step": 10067
+    },
+    {
+      "epoch": 2.8178001679261127,
+      "grad_norm": 0.25587850210427726,
+      "learning_rate": 1.008134804528521e-06,
+      "loss": 0.4486,
+      "step": 10068
+    },
+    {
+      "epoch": 2.818080044780297,
+      "grad_norm": 0.25385824906574944,
+      "learning_rate": 1.0050551321938384e-06,
+      "loss": 0.4499,
+      "step": 10069
+    },
+    {
+      "epoch": 2.818359921634481,
+      "grad_norm": 0.2529017673823629,
+      "learning_rate": 1.0019801232254322e-06,
+      "loss": 0.4263,
+      "step": 10070
+    },
+    {
+      "epoch": 2.818639798488665,
+      "grad_norm": 0.2634481033538776,
+      "learning_rate": 9.989097779159796e-07,
+      "loss": 0.4462,
+      "step": 10071
+    },
+    {
+      "epoch": 2.8189196753428494,
+      "grad_norm": 0.2487106793995691,
+      "learning_rate": 9.958440965577243e-07,
+      "loss": 0.4597,
+      "step": 10072
+    },
+    {
+      "epoch": 2.819199552197033,
+      "grad_norm": 0.2479782160319531,
+      "learning_rate": 9.927830794424553e-07,
+      "loss": 0.4496,
+      "step": 10073
+    },
+    {
+      "epoch": 2.8194794290512175,
+      "grad_norm": 0.2533961878193845,
+      "learning_rate": 9.897267268615284e-07,
+      "loss": 0.4307,
+      "step": 10074
+    },
+    {
+      "epoch": 2.819759305905402,
+      "grad_norm": 0.2466751349058953,
+      "learning_rate": 9.866750391058389e-07,
+      "loss": 0.4368,
+      "step": 10075
+    },
+    {
+      "epoch": 2.8200391827595856,
+      "grad_norm": 0.24374236508360309,
+      "learning_rate": 9.83628016465854e-07,
+      "loss": 0.4352,
+      "step": 10076
+    },
+    {
+      "epoch": 2.82031905961377,
+      "grad_norm": 0.2543213559113662,
+      "learning_rate": 9.80585659231592e-07,
+      "loss": 0.4259,
+      "step": 10077
+    },
+    {
+      "epoch": 2.8205989364679542,
+      "grad_norm": 0.25099459130558754,
+      "learning_rate": 9.775479676926269e-07,
+      "loss": 0.4377,
+      "step": 10078
+    },
+    {
+      "epoch": 2.820878813322138,
+      "grad_norm": 0.2572368803838283,
+      "learning_rate": 9.745149421380828e-07,
+      "loss": 0.4492,
+      "step": 10079
+    },
+    {
+      "epoch": 2.8211586901763224,
+      "grad_norm": 0.2601264172895387,
+      "learning_rate": 9.714865828566455e-07,
+      "loss": 0.4614,
+      "step": 10080
+    },
+    {
+      "epoch": 2.8214385670305067,
+      "grad_norm": 0.24809535239113745,
+      "learning_rate": 9.684628901365567e-07,
+      "loss": 0.4686,
+      "step": 10081
+    },
+    {
+      "epoch": 2.8217184438846905,
+      "grad_norm": 0.2583757231300886,
+      "learning_rate": 9.654438642656083e-07,
+      "loss": 0.4374,
+      "step": 10082
+    },
+    {
+      "epoch": 2.821998320738875,
+      "grad_norm": 0.25979970122250534,
+      "learning_rate": 9.624295055311704e-07,
+      "loss": 0.4878,
+      "step": 10083
+    },
+    {
+      "epoch": 2.822278197593059,
+      "grad_norm": 0.2627535858238464,
+      "learning_rate": 9.594198142201305e-07,
+      "loss": 0.4486,
+      "step": 10084
+    },
+    {
+      "epoch": 2.822558074447243,
+      "grad_norm": 0.24754921991327522,
+      "learning_rate": 9.564147906189703e-07,
+      "loss": 0.4519,
+      "step": 10085
+    },
+    {
+      "epoch": 2.8228379513014272,
+      "grad_norm": 0.27070248197086716,
+      "learning_rate": 9.534144350136942e-07,
+      "loss": 0.4693,
+      "step": 10086
+    },
+    {
+      "epoch": 2.8231178281556115,
+      "grad_norm": 0.24472943267326952,
+      "learning_rate": 9.504187476898907e-07,
+      "loss": 0.4334,
+      "step": 10087
+    },
+    {
+      "epoch": 2.823397705009796,
+      "grad_norm": 0.24744547999958846,
+      "learning_rate": 9.474277289326817e-07,
+      "loss": 0.4531,
+      "step": 10088
+    },
+    {
+      "epoch": 2.82367758186398,
+      "grad_norm": 0.25962567214023186,
+      "learning_rate": 9.444413790267558e-07,
+      "loss": 0.4222,
+      "step": 10089
+    },
+    {
+      "epoch": 2.823957458718164,
+      "grad_norm": 0.25505431645769044,
+      "learning_rate": 9.41459698256364e-07,
+      "loss": 0.456,
+      "step": 10090
+    },
+    {
+      "epoch": 2.8242373355723482,
+      "grad_norm": 0.26849471916892415,
+      "learning_rate": 9.384826869052898e-07,
+      "loss": 0.4498,
+      "step": 10091
+    },
+    {
+      "epoch": 2.8245172124265325,
+      "grad_norm": 0.2554869188266186,
+      "learning_rate": 9.355103452568958e-07,
+      "loss": 0.4382,
+      "step": 10092
+    },
+    {
+      "epoch": 2.8247970892807164,
+      "grad_norm": 0.253352558640311,
+      "learning_rate": 9.325426735940946e-07,
+      "loss": 0.4439,
+      "step": 10093
+    },
+    {
+      "epoch": 2.8250769661349007,
+      "grad_norm": 0.24457205420612252,
+      "learning_rate": 9.295796721993433e-07,
+      "loss": 0.4317,
+      "step": 10094
+    },
+    {
+      "epoch": 2.825356842989085,
+      "grad_norm": 0.2611003726754562,
+      "learning_rate": 9.266213413546609e-07,
+      "loss": 0.4707,
+      "step": 10095
+    },
+    {
+      "epoch": 2.825636719843269,
+      "grad_norm": 0.2572010642600289,
+      "learning_rate": 9.236676813416334e-07,
+      "loss": 0.4458,
+      "step": 10096
+    },
+    {
+      "epoch": 2.825916596697453,
+      "grad_norm": 0.23567654616912856,
+      "learning_rate": 9.207186924413858e-07,
+      "loss": 0.4161,
+      "step": 10097
+    },
+    {
+      "epoch": 2.8261964735516374,
+      "grad_norm": 0.2524806826636959,
+      "learning_rate": 9.17774374934599e-07,
+      "loss": 0.4612,
+      "step": 10098
+    },
+    {
+      "epoch": 2.8264763504058212,
+      "grad_norm": 0.2583721915497328,
+      "learning_rate": 9.148347291015158e-07,
+      "loss": 0.4534,
+      "step": 10099
+    },
+    {
+      "epoch": 2.8267562272600055,
+      "grad_norm": 0.2443844209188402,
+      "learning_rate": 9.1189975522194e-07,
+      "loss": 0.4351,
+      "step": 10100
+    },
+    {
+      "epoch": 2.82703610411419,
+      "grad_norm": 0.2592075075953771,
+      "learning_rate": 9.089694535752257e-07,
+      "loss": 0.4483,
+      "step": 10101
+    },
+    {
+      "epoch": 2.8273159809683737,
+      "grad_norm": 0.2627400968198566,
+      "learning_rate": 9.060438244402725e-07,
+      "loss": 0.4623,
+      "step": 10102
+    },
+    {
+      "epoch": 2.827595857822558,
+      "grad_norm": 0.24670093257172748,
+      "learning_rate": 9.031228680955517e-07,
+      "loss": 0.4376,
+      "step": 10103
+    },
+    {
+      "epoch": 2.8278757346767422,
+      "grad_norm": 0.25509140273022224,
+      "learning_rate": 9.002065848190744e-07,
+      "loss": 0.4488,
+      "step": 10104
+    },
+    {
+      "epoch": 2.8281556115309265,
+      "grad_norm": 0.2513643430278304,
+      "learning_rate": 8.97294974888413e-07,
+      "loss": 0.4434,
+      "step": 10105
+    },
+    {
+      "epoch": 2.8284354883851104,
+      "grad_norm": 0.25221197470587714,
+      "learning_rate": 8.943880385807068e-07,
+      "loss": 0.4452,
+      "step": 10106
+    },
+    {
+      "epoch": 2.8287153652392947,
+      "grad_norm": 0.25070011355054483,
+      "learning_rate": 8.914857761726292e-07,
+      "loss": 0.434,
+      "step": 10107
+    },
+    {
+      "epoch": 2.828995242093479,
+      "grad_norm": 0.2582137023259711,
+      "learning_rate": 8.885881879404201e-07,
+      "loss": 0.4528,
+      "step": 10108
+    },
+    {
+      "epoch": 2.8292751189476633,
+      "grad_norm": 0.2592773347128377,
+      "learning_rate": 8.856952741598756e-07,
+      "loss": 0.4697,
+      "step": 10109
+    },
+    {
+      "epoch": 2.829554995801847,
+      "grad_norm": 0.2629416907627084,
+      "learning_rate": 8.828070351063533e-07,
+      "loss": 0.4501,
+      "step": 10110
+    },
+    {
+      "epoch": 2.8298348726560314,
+      "grad_norm": 0.2631384755840109,
+      "learning_rate": 8.799234710547444e-07,
+      "loss": 0.45,
+      "step": 10111
+    },
+    {
+      "epoch": 2.8301147495102157,
+      "grad_norm": 0.24555445924895378,
+      "learning_rate": 8.770445822795126e-07,
+      "loss": 0.4541,
+      "step": 10112
+    },
+    {
+      "epoch": 2.8303946263643995,
+      "grad_norm": 0.25707171887557456,
+      "learning_rate": 8.741703690546721e-07,
+      "loss": 0.4402,
+      "step": 10113
+    },
+    {
+      "epoch": 2.830674503218584,
+      "grad_norm": 0.2556283650182902,
+      "learning_rate": 8.713008316537929e-07,
+      "loss": 0.4572,
+      "step": 10114
+    },
+    {
+      "epoch": 2.830954380072768,
+      "grad_norm": 0.2600383651816295,
+      "learning_rate": 8.684359703500067e-07,
+      "loss": 0.4621,
+      "step": 10115
+    },
+    {
+      "epoch": 2.831234256926952,
+      "grad_norm": 0.267705736677189,
+      "learning_rate": 8.65575785415973e-07,
+      "loss": 0.4598,
+      "step": 10116
+    },
+    {
+      "epoch": 2.8315141337811363,
+      "grad_norm": 0.24332224527834756,
+      "learning_rate": 8.627202771239462e-07,
+      "loss": 0.4593,
+      "step": 10117
+    },
+    {
+      "epoch": 2.8317940106353205,
+      "grad_norm": 0.2585644170136799,
+      "learning_rate": 8.598694457457035e-07,
+      "loss": 0.4578,
+      "step": 10118
+    },
+    {
+      "epoch": 2.8320738874895044,
+      "grad_norm": 0.2528586860303206,
+      "learning_rate": 8.570232915525888e-07,
+      "loss": 0.4631,
+      "step": 10119
+    },
+    {
+      "epoch": 2.8323537643436887,
+      "grad_norm": 0.26658090330048756,
+      "learning_rate": 8.541818148155079e-07,
+      "loss": 0.4456,
+      "step": 10120
+    },
+    {
+      "epoch": 2.832633641197873,
+      "grad_norm": 0.25805663437445653,
+      "learning_rate": 8.513450158049108e-07,
+      "loss": 0.4587,
+      "step": 10121
+    },
+    {
+      "epoch": 2.832913518052057,
+      "grad_norm": 0.26390238260937576,
+      "learning_rate": 8.485128947908039e-07,
+      "loss": 0.4603,
+      "step": 10122
+    },
+    {
+      "epoch": 2.833193394906241,
+      "grad_norm": 0.2628755722632489,
+      "learning_rate": 8.456854520427493e-07,
+      "loss": 0.437,
+      "step": 10123
+    },
+    {
+      "epoch": 2.8334732717604254,
+      "grad_norm": 0.26083288223354306,
+      "learning_rate": 8.428626878298707e-07,
+      "loss": 0.4331,
+      "step": 10124
+    },
+    {
+      "epoch": 2.8337531486146097,
+      "grad_norm": 0.25799908741638145,
+      "learning_rate": 8.400446024208309e-07,
+      "loss": 0.4564,
+      "step": 10125
+    },
+    {
+      "epoch": 2.834033025468794,
+      "grad_norm": 0.24081806697538238,
+      "learning_rate": 8.372311960838652e-07,
+      "loss": 0.4382,
+      "step": 10126
+    },
+    {
+      "epoch": 2.834312902322978,
+      "grad_norm": 0.2574413268314077,
+      "learning_rate": 8.344224690867485e-07,
+      "loss": 0.4434,
+      "step": 10127
+    },
+    {
+      "epoch": 2.834592779177162,
+      "grad_norm": 0.29761285533028925,
+      "learning_rate": 8.316184216968226e-07,
+      "loss": 0.453,
+      "step": 10128
+    },
+    {
+      "epoch": 2.8348726560313464,
+      "grad_norm": 0.26541255095918426,
+      "learning_rate": 8.288190541809738e-07,
+      "loss": 0.446,
+      "step": 10129
+    },
+    {
+      "epoch": 2.8351525328855303,
+      "grad_norm": 0.2551603304358025,
+      "learning_rate": 8.260243668056555e-07,
+      "loss": 0.4409,
+      "step": 10130
+    },
+    {
+      "epoch": 2.8354324097397146,
+      "grad_norm": 0.2681181178604902,
+      "learning_rate": 8.232343598368552e-07,
+      "loss": 0.454,
+      "step": 10131
+    },
+    {
+      "epoch": 2.835712286593899,
+      "grad_norm": 0.267916569740636,
+      "learning_rate": 8.204490335401382e-07,
+      "loss": 0.4453,
+      "step": 10132
+    },
+    {
+      "epoch": 2.8359921634480827,
+      "grad_norm": 0.25766534832339966,
+      "learning_rate": 8.17668388180609e-07,
+      "loss": 0.444,
+      "step": 10133
+    },
+    {
+      "epoch": 2.836272040302267,
+      "grad_norm": 0.25295408765533794,
+      "learning_rate": 8.148924240229283e-07,
+      "loss": 0.4301,
+      "step": 10134
+    },
+    {
+      "epoch": 2.8365519171564513,
+      "grad_norm": 0.2590464448281375,
+      "learning_rate": 8.121211413313178e-07,
+      "loss": 0.4527,
+      "step": 10135
+    },
+    {
+      "epoch": 2.836831794010635,
+      "grad_norm": 0.27162404825376435,
+      "learning_rate": 8.093545403695502e-07,
+      "loss": 0.4445,
+      "step": 10136
+    },
+    {
+      "epoch": 2.8371116708648194,
+      "grad_norm": 0.26622854836678916,
+      "learning_rate": 8.065926214009479e-07,
+      "loss": 0.4467,
+      "step": 10137
+    },
+    {
+      "epoch": 2.8373915477190037,
+      "grad_norm": 0.2540515828910898,
+      "learning_rate": 8.038353846884006e-07,
+      "loss": 0.4431,
+      "step": 10138
+    },
+    {
+      "epoch": 2.8376714245731876,
+      "grad_norm": 0.25453815423627246,
+      "learning_rate": 8.010828304943318e-07,
+      "loss": 0.4441,
+      "step": 10139
+    },
+    {
+      "epoch": 2.837951301427372,
+      "grad_norm": 0.2685794575098386,
+      "learning_rate": 7.983349590807376e-07,
+      "loss": 0.4476,
+      "step": 10140
+    },
+    {
+      "epoch": 2.838231178281556,
+      "grad_norm": 0.26090741048944566,
+      "learning_rate": 7.955917707091642e-07,
+      "loss": 0.444,
+      "step": 10141
+    },
+    {
+      "epoch": 2.8385110551357404,
+      "grad_norm": 0.24961955398553431,
+      "learning_rate": 7.928532656407029e-07,
+      "loss": 0.4578,
+      "step": 10142
+    },
+    {
+      "epoch": 2.8387909319899243,
+      "grad_norm": 0.261480653970001,
+      "learning_rate": 7.901194441360116e-07,
+      "loss": 0.4365,
+      "step": 10143
+    },
+    {
+      "epoch": 2.8390708088441086,
+      "grad_norm": 0.26065977212184543,
+      "learning_rate": 7.87390306455299e-07,
+      "loss": 0.4538,
+      "step": 10144
+    },
+    {
+      "epoch": 2.839350685698293,
+      "grad_norm": 0.24750876625727614,
+      "learning_rate": 7.846658528583184e-07,
+      "loss": 0.4522,
+      "step": 10145
+    },
+    {
+      "epoch": 2.839630562552477,
+      "grad_norm": 0.2553102416355813,
+      "learning_rate": 7.819460836043957e-07,
+      "loss": 0.4559,
+      "step": 10146
+    },
+    {
+      "epoch": 2.839910439406661,
+      "grad_norm": 0.24769185971067034,
+      "learning_rate": 7.79230998952385e-07,
+      "loss": 0.4552,
+      "step": 10147
+    },
+    {
+      "epoch": 2.8401903162608453,
+      "grad_norm": 0.2672748874110806,
+      "learning_rate": 7.765205991607238e-07,
+      "loss": 0.4659,
+      "step": 10148
+    },
+    {
+      "epoch": 2.8404701931150296,
+      "grad_norm": 0.2604953786766096,
+      "learning_rate": 7.738148844873783e-07,
+      "loss": 0.4744,
+      "step": 10149
+    },
+    {
+      "epoch": 2.8407500699692134,
+      "grad_norm": 0.2583480179782524,
+      "learning_rate": 7.711138551898867e-07,
+      "loss": 0.4479,
+      "step": 10150
+    },
+    {
+      "epoch": 2.8410299468233977,
+      "grad_norm": 0.2546325371505603,
+      "learning_rate": 7.684175115253378e-07,
+      "loss": 0.439,
+      "step": 10151
+    },
+    {
+      "epoch": 2.841309823677582,
+      "grad_norm": 0.26613969797264936,
+      "learning_rate": 7.657258537503598e-07,
+      "loss": 0.461,
+      "step": 10152
+    },
+    {
+      "epoch": 2.841589700531766,
+      "grad_norm": 0.24737645835701677,
+      "learning_rate": 7.630388821211587e-07,
+      "loss": 0.4282,
+      "step": 10153
+    },
+    {
+      "epoch": 2.84186957738595,
+      "grad_norm": 0.2624667092229066,
+      "learning_rate": 7.603565968934689e-07,
+      "loss": 0.4708,
+      "step": 10154
+    },
+    {
+      "epoch": 2.8421494542401344,
+      "grad_norm": 0.2420140631051068,
+      "learning_rate": 7.576789983226029e-07,
+      "loss": 0.4299,
+      "step": 10155
+    },
+    {
+      "epoch": 2.8424293310943183,
+      "grad_norm": 0.25496191355468883,
+      "learning_rate": 7.550060866634123e-07,
+      "loss": 0.4377,
+      "step": 10156
+    },
+    {
+      "epoch": 2.8427092079485026,
+      "grad_norm": 0.24663098820574658,
+      "learning_rate": 7.523378621703048e-07,
+      "loss": 0.4334,
+      "step": 10157
+    },
+    {
+      "epoch": 2.842989084802687,
+      "grad_norm": 0.25869772024576165,
+      "learning_rate": 7.496743250972494e-07,
+      "loss": 0.4677,
+      "step": 10158
+    },
+    {
+      "epoch": 2.8432689616568707,
+      "grad_norm": 0.259342332895863,
+      "learning_rate": 7.470154756977543e-07,
+      "loss": 0.4556,
+      "step": 10159
+    },
+    {
+      "epoch": 2.843548838511055,
+      "grad_norm": 0.25720323808780676,
+      "learning_rate": 7.443613142248951e-07,
+      "loss": 0.447,
+      "step": 10160
+    },
+    {
+      "epoch": 2.8438287153652393,
+      "grad_norm": 0.26139628238782625,
+      "learning_rate": 7.417118409312918e-07,
+      "loss": 0.4468,
+      "step": 10161
+    },
+    {
+      "epoch": 2.8441085922194236,
+      "grad_norm": 0.2630166216465488,
+      "learning_rate": 7.390670560691315e-07,
+      "loss": 0.455,
+      "step": 10162
+    },
+    {
+      "epoch": 2.844388469073608,
+      "grad_norm": 0.2659934121702161,
+      "learning_rate": 7.364269598901408e-07,
+      "loss": 0.4334,
+      "step": 10163
+    },
+    {
+      "epoch": 2.8446683459277917,
+      "grad_norm": 0.25686988035527164,
+      "learning_rate": 7.337915526456019e-07,
+      "loss": 0.4514,
+      "step": 10164
+    },
+    {
+      "epoch": 2.844948222781976,
+      "grad_norm": 0.251348655283329,
+      "learning_rate": 7.311608345863641e-07,
+      "loss": 0.4539,
+      "step": 10165
+    },
+    {
+      "epoch": 2.8452280996361603,
+      "grad_norm": 0.25654330476923476,
+      "learning_rate": 7.285348059628105e-07,
+      "loss": 0.4593,
+      "step": 10166
+    },
+    {
+      "epoch": 2.845507976490344,
+      "grad_norm": 0.24776546205526637,
+      "learning_rate": 7.259134670248968e-07,
+      "loss": 0.4672,
+      "step": 10167
+    },
+    {
+      "epoch": 2.8457878533445284,
+      "grad_norm": 0.2556660013031803,
+      "learning_rate": 7.232968180221122e-07,
+      "loss": 0.4325,
+      "step": 10168
+    },
+    {
+      "epoch": 2.8460677301987127,
+      "grad_norm": 0.2448147767428879,
+      "learning_rate": 7.206848592035242e-07,
+      "loss": 0.4377,
+      "step": 10169
+    },
+    {
+      "epoch": 2.8463476070528966,
+      "grad_norm": 0.25705771035811886,
+      "learning_rate": 7.180775908177339e-07,
+      "loss": 0.443,
+      "step": 10170
+    },
+    {
+      "epoch": 2.846627483907081,
+      "grad_norm": 0.24945762219879117,
+      "learning_rate": 7.154750131128984e-07,
+      "loss": 0.4569,
+      "step": 10171
+    },
+    {
+      "epoch": 2.846907360761265,
+      "grad_norm": 0.2540273156999752,
+      "learning_rate": 7.128771263367418e-07,
+      "loss": 0.4303,
+      "step": 10172
+    },
+    {
+      "epoch": 2.847187237615449,
+      "grad_norm": 0.2580889678828135,
+      "learning_rate": 7.102839307365272e-07,
+      "loss": 0.4505,
+      "step": 10173
+    },
+    {
+      "epoch": 2.8474671144696333,
+      "grad_norm": 0.2603424115721654,
+      "learning_rate": 7.076954265590741e-07,
+      "loss": 0.467,
+      "step": 10174
+    },
+    {
+      "epoch": 2.8477469913238176,
+      "grad_norm": 0.2530692836062023,
+      "learning_rate": 7.05111614050763e-07,
+      "loss": 0.4487,
+      "step": 10175
+    },
+    {
+      "epoch": 2.8480268681780014,
+      "grad_norm": 0.2559129157286122,
+      "learning_rate": 7.025324934575139e-07,
+      "loss": 0.4416,
+      "step": 10176
+    },
+    {
+      "epoch": 2.8483067450321857,
+      "grad_norm": 0.24773205322018566,
+      "learning_rate": 6.999580650248194e-07,
+      "loss": 0.4329,
+      "step": 10177
+    },
+    {
+      "epoch": 2.84858662188637,
+      "grad_norm": 0.25521126431205277,
+      "learning_rate": 6.973883289977112e-07,
+      "loss": 0.4543,
+      "step": 10178
+    },
+    {
+      "epoch": 2.8488664987405543,
+      "grad_norm": 0.2595084443543985,
+      "learning_rate": 6.948232856207771e-07,
+      "loss": 0.4529,
+      "step": 10179
+    },
+    {
+      "epoch": 2.849146375594738,
+      "grad_norm": 0.2606741603953674,
+      "learning_rate": 6.922629351381604e-07,
+      "loss": 0.4353,
+      "step": 10180
+    },
+    {
+      "epoch": 2.8494262524489224,
+      "grad_norm": 0.2557838497999989,
+      "learning_rate": 6.897072777935609e-07,
+      "loss": 0.4424,
+      "step": 10181
+    },
+    {
+      "epoch": 2.8497061293031067,
+      "grad_norm": 0.24571771064985215,
+      "learning_rate": 6.871563138302173e-07,
+      "loss": 0.4505,
+      "step": 10182
+    },
+    {
+      "epoch": 2.849986006157291,
+      "grad_norm": 0.26667683305745005,
+      "learning_rate": 6.846100434909353e-07,
+      "loss": 0.4542,
+      "step": 10183
+    },
+    {
+      "epoch": 2.850265883011475,
+      "grad_norm": 0.2600320263030447,
+      "learning_rate": 6.820684670180766e-07,
+      "loss": 0.435,
+      "step": 10184
+    },
+    {
+      "epoch": 2.850545759865659,
+      "grad_norm": 0.2537212753909362,
+      "learning_rate": 6.795315846535422e-07,
+      "loss": 0.4445,
+      "step": 10185
+    },
+    {
+      "epoch": 2.8508256367198435,
+      "grad_norm": 0.26412510291968333,
+      "learning_rate": 6.769993966387999e-07,
+      "loss": 0.4496,
+      "step": 10186
+    },
+    {
+      "epoch": 2.8511055135740273,
+      "grad_norm": 0.25422280946962045,
+      "learning_rate": 6.744719032148627e-07,
+      "loss": 0.45,
+      "step": 10187
+    },
+    {
+      "epoch": 2.8513853904282116,
+      "grad_norm": 0.2533516046710153,
+      "learning_rate": 6.719491046222992e-07,
+      "loss": 0.428,
+      "step": 10188
+    },
+    {
+      "epoch": 2.851665267282396,
+      "grad_norm": 0.24407088865412283,
+      "learning_rate": 6.694310011012284e-07,
+      "loss": 0.4572,
+      "step": 10189
+    },
+    {
+      "epoch": 2.8519451441365797,
+      "grad_norm": 0.24943438860892966,
+      "learning_rate": 6.669175928913251e-07,
+      "loss": 0.4165,
+      "step": 10190
+    },
+    {
+      "epoch": 2.852225020990764,
+      "grad_norm": 0.25713363585411575,
+      "learning_rate": 6.644088802318205e-07,
+      "loss": 0.4453,
+      "step": 10191
+    },
+    {
+      "epoch": 2.8525048978449483,
+      "grad_norm": 0.26567928250076694,
+      "learning_rate": 6.619048633614955e-07,
+      "loss": 0.4505,
+      "step": 10192
+    },
+    {
+      "epoch": 2.852784774699132,
+      "grad_norm": 0.25818673891369537,
+      "learning_rate": 6.594055425186763e-07,
+      "loss": 0.449,
+      "step": 10193
+    },
+    {
+      "epoch": 2.8530646515533165,
+      "grad_norm": 0.2540198476376862,
+      "learning_rate": 6.569109179412558e-07,
+      "loss": 0.4309,
+      "step": 10194
+    },
+    {
+      "epoch": 2.8533445284075007,
+      "grad_norm": 0.2545216361991795,
+      "learning_rate": 6.54420989866672e-07,
+      "loss": 0.4281,
+      "step": 10195
+    },
+    {
+      "epoch": 2.8536244052616846,
+      "grad_norm": 0.25968360621510583,
+      "learning_rate": 6.519357585319242e-07,
+      "loss": 0.468,
+      "step": 10196
+    },
+    {
+      "epoch": 2.853904282115869,
+      "grad_norm": 0.24687151281696984,
+      "learning_rate": 6.494552241735452e-07,
+      "loss": 0.4264,
+      "step": 10197
+    },
+    {
+      "epoch": 2.854184158970053,
+      "grad_norm": 0.2583821078491018,
+      "learning_rate": 6.469793870276464e-07,
+      "loss": 0.4752,
+      "step": 10198
+    },
+    {
+      "epoch": 2.8544640358242375,
+      "grad_norm": 0.2459137265381298,
+      "learning_rate": 6.445082473298669e-07,
+      "loss": 0.4725,
+      "step": 10199
+    },
+    {
+      "epoch": 2.8547439126784218,
+      "grad_norm": 0.24742088586054187,
+      "learning_rate": 6.420418053154243e-07,
+      "loss": 0.4407,
+      "step": 10200
+    },
+    {
+      "epoch": 2.8550237895326056,
+      "grad_norm": 0.25517969356297493,
+      "learning_rate": 6.395800612190639e-07,
+      "loss": 0.4355,
+      "step": 10201
+    },
+    {
+      "epoch": 2.85530366638679,
+      "grad_norm": 0.2598944885613344,
+      "learning_rate": 6.37123015275104e-07,
+      "loss": 0.4621,
+      "step": 10202
+    },
+    {
+      "epoch": 2.855583543240974,
+      "grad_norm": 0.25983576281556636,
+      "learning_rate": 6.346706677174075e-07,
+      "loss": 0.448,
+      "step": 10203
+    },
+    {
+      "epoch": 2.855863420095158,
+      "grad_norm": 0.2442368834802085,
+      "learning_rate": 6.322230187793876e-07,
+      "loss": 0.4305,
+      "step": 10204
+    },
+    {
+      "epoch": 2.8561432969493423,
+      "grad_norm": 0.25932837759420585,
+      "learning_rate": 6.297800686940081e-07,
+      "loss": 0.4534,
+      "step": 10205
+    },
+    {
+      "epoch": 2.8564231738035266,
+      "grad_norm": 0.26187189125601396,
+      "learning_rate": 6.273418176937995e-07,
+      "loss": 0.4604,
+      "step": 10206
+    },
+    {
+      "epoch": 2.8567030506577105,
+      "grad_norm": 0.2503901417519331,
+      "learning_rate": 6.249082660108318e-07,
+      "loss": 0.4152,
+      "step": 10207
+    },
+    {
+      "epoch": 2.8569829275118948,
+      "grad_norm": 0.2557843101499479,
+      "learning_rate": 6.224794138767309e-07,
+      "loss": 0.4565,
+      "step": 10208
+    },
+    {
+      "epoch": 2.857262804366079,
+      "grad_norm": 0.25384466886807133,
+      "learning_rate": 6.200552615226784e-07,
+      "loss": 0.4452,
+      "step": 10209
+    },
+    {
+      "epoch": 2.857542681220263,
+      "grad_norm": 0.2548841014459113,
+      "learning_rate": 6.176358091794011e-07,
+      "loss": 0.4368,
+      "step": 10210
+    },
+    {
+      "epoch": 2.857822558074447,
+      "grad_norm": 0.25757124576997104,
+      "learning_rate": 6.152210570771921e-07,
+      "loss": 0.4439,
+      "step": 10211
+    },
+    {
+      "epoch": 2.8581024349286315,
+      "grad_norm": 0.2569331306938241,
+      "learning_rate": 6.128110054458847e-07,
+      "loss": 0.4612,
+      "step": 10212
+    },
+    {
+      "epoch": 2.8583823117828153,
+      "grad_norm": 0.25282275686772765,
+      "learning_rate": 6.104056545148673e-07,
+      "loss": 0.4577,
+      "step": 10213
+    },
+    {
+      "epoch": 2.8586621886369996,
+      "grad_norm": 0.25896247226299324,
+      "learning_rate": 6.080050045130847e-07,
+      "loss": 0.4509,
+      "step": 10214
+    },
+    {
+      "epoch": 2.858942065491184,
+      "grad_norm": 0.2576693397622,
+      "learning_rate": 6.056090556690319e-07,
+      "loss": 0.4411,
+      "step": 10215
+    },
+    {
+      "epoch": 2.859221942345368,
+      "grad_norm": 0.2636597329257613,
+      "learning_rate": 6.032178082107542e-07,
+      "loss": 0.4533,
+      "step": 10216
+    },
+    {
+      "epoch": 2.859501819199552,
+      "grad_norm": 0.24697365713431213,
+      "learning_rate": 6.008312623658586e-07,
+      "loss": 0.4389,
+      "step": 10217
+    },
+    {
+      "epoch": 2.8597816960537363,
+      "grad_norm": 0.2500930445428214,
+      "learning_rate": 5.984494183614909e-07,
+      "loss": 0.4566,
+      "step": 10218
+    },
+    {
+      "epoch": 2.8600615729079206,
+      "grad_norm": 0.26216006480801274,
+      "learning_rate": 5.960722764243587e-07,
+      "loss": 0.467,
+      "step": 10219
+    },
+    {
+      "epoch": 2.860341449762105,
+      "grad_norm": 0.24926684599386656,
+      "learning_rate": 5.936998367807201e-07,
+      "loss": 0.4326,
+      "step": 10220
+    },
+    {
+      "epoch": 2.8606213266162888,
+      "grad_norm": 0.2562319805043352,
+      "learning_rate": 5.913320996563831e-07,
+      "loss": 0.4631,
+      "step": 10221
+    },
+    {
+      "epoch": 2.860901203470473,
+      "grad_norm": 0.2457629153440099,
+      "learning_rate": 5.889690652767177e-07,
+      "loss": 0.4501,
+      "step": 10222
+    },
+    {
+      "epoch": 2.8611810803246573,
+      "grad_norm": 0.26272183713329245,
+      "learning_rate": 5.866107338666271e-07,
+      "loss": 0.4568,
+      "step": 10223
+    },
+    {
+      "epoch": 2.861460957178841,
+      "grad_norm": 0.25666367549779606,
+      "learning_rate": 5.842571056505875e-07,
+      "loss": 0.4489,
+      "step": 10224
+    },
+    {
+      "epoch": 2.8617408340330255,
+      "grad_norm": 0.2705683877873388,
+      "learning_rate": 5.819081808526139e-07,
+      "loss": 0.4643,
+      "step": 10225
+    },
+    {
+      "epoch": 2.8620207108872098,
+      "grad_norm": 0.2543820024194871,
+      "learning_rate": 5.795639596962832e-07,
+      "loss": 0.4503,
+      "step": 10226
+    },
+    {
+      "epoch": 2.8623005877413936,
+      "grad_norm": 0.25693972439495794,
+      "learning_rate": 5.772244424047169e-07,
+      "loss": 0.4444,
+      "step": 10227
+    },
+    {
+      "epoch": 2.862580464595578,
+      "grad_norm": 0.25490621766923677,
+      "learning_rate": 5.748896292005868e-07,
+      "loss": 0.4486,
+      "step": 10228
+    },
+    {
+      "epoch": 2.862860341449762,
+      "grad_norm": 0.26485194854054805,
+      "learning_rate": 5.725595203061318e-07,
+      "loss": 0.4659,
+      "step": 10229
+    },
+    {
+      "epoch": 2.863140218303946,
+      "grad_norm": 0.24588014051314325,
+      "learning_rate": 5.702341159431246e-07,
+      "loss": 0.4458,
+      "step": 10230
+    },
+    {
+      "epoch": 2.8634200951581303,
+      "grad_norm": 0.25924629499595886,
+      "learning_rate": 5.679134163328992e-07,
+      "loss": 0.4423,
+      "step": 10231
+    },
+    {
+      "epoch": 2.8636999720123146,
+      "grad_norm": 0.2553099022686705,
+      "learning_rate": 5.655974216963456e-07,
+      "loss": 0.447,
+      "step": 10232
+    },
+    {
+      "epoch": 2.8639798488664985,
+      "grad_norm": 0.259680346674402,
+      "learning_rate": 5.632861322538985e-07,
+      "loss": 0.4425,
+      "step": 10233
+    },
+    {
+      "epoch": 2.8642597257206828,
+      "grad_norm": 0.26740786597677946,
+      "learning_rate": 5.609795482255486e-07,
+      "loss": 0.4423,
+      "step": 10234
+    },
+    {
+      "epoch": 2.864539602574867,
+      "grad_norm": 0.2680068141821788,
+      "learning_rate": 5.586776698308372e-07,
+      "loss": 0.4648,
+      "step": 10235
+    },
+    {
+      "epoch": 2.8648194794290514,
+      "grad_norm": 0.2563372948009287,
+      "learning_rate": 5.563804972888609e-07,
+      "loss": 0.4618,
+      "step": 10236
+    },
+    {
+      "epoch": 2.8650993562832356,
+      "grad_norm": 0.24611658743221432,
+      "learning_rate": 5.540880308182617e-07,
+      "loss": 0.4475,
+      "step": 10237
+    },
+    {
+      "epoch": 2.8653792331374195,
+      "grad_norm": 0.24907400613042632,
+      "learning_rate": 5.518002706372372e-07,
+      "loss": 0.4479,
+      "step": 10238
+    },
+    {
+      "epoch": 2.865659109991604,
+      "grad_norm": 0.2481189654103723,
+      "learning_rate": 5.495172169635409e-07,
+      "loss": 0.4535,
+      "step": 10239
+    },
+    {
+      "epoch": 2.865938986845788,
+      "grad_norm": 0.24972684085267743,
+      "learning_rate": 5.472388700144771e-07,
+      "loss": 0.4486,
+      "step": 10240
+    },
+    {
+      "epoch": 2.866218863699972,
+      "grad_norm": 0.2493928388889635,
+      "learning_rate": 5.449652300068997e-07,
+      "loss": 0.442,
+      "step": 10241
+    },
+    {
+      "epoch": 2.866498740554156,
+      "grad_norm": 0.25502178405778514,
+      "learning_rate": 5.426962971572081e-07,
+      "loss": 0.4567,
+      "step": 10242
+    },
+    {
+      "epoch": 2.8667786174083405,
+      "grad_norm": 0.25717676914814624,
+      "learning_rate": 5.404320716813683e-07,
+      "loss": 0.4245,
+      "step": 10243
+    },
+    {
+      "epoch": 2.8670584942625243,
+      "grad_norm": 0.26027840464450835,
+      "learning_rate": 5.381725537948856e-07,
+      "loss": 0.4528,
+      "step": 10244
+    },
+    {
+      "epoch": 2.8673383711167086,
+      "grad_norm": 0.258521387544155,
+      "learning_rate": 5.35917743712827e-07,
+      "loss": 0.4327,
+      "step": 10245
+    },
+    {
+      "epoch": 2.867618247970893,
+      "grad_norm": 0.25270403591986484,
+      "learning_rate": 5.336676416498043e-07,
+      "loss": 0.4459,
+      "step": 10246
+    },
+    {
+      "epoch": 2.8678981248250768,
+      "grad_norm": 0.2552451258449106,
+      "learning_rate": 5.314222478199793e-07,
+      "loss": 0.4565,
+      "step": 10247
+    },
+    {
+      "epoch": 2.868178001679261,
+      "grad_norm": 0.25016155667035667,
+      "learning_rate": 5.291815624370755e-07,
+      "loss": 0.4381,
+      "step": 10248
+    },
+    {
+      "epoch": 2.8684578785334454,
+      "grad_norm": 0.25687106582087943,
+      "learning_rate": 5.269455857143613e-07,
+      "loss": 0.4437,
+      "step": 10249
+    },
+    {
+      "epoch": 2.868737755387629,
+      "grad_norm": 0.25635311037631414,
+      "learning_rate": 5.247143178646552e-07,
+      "loss": 0.4523,
+      "step": 10250
+    },
+    {
+      "epoch": 2.8690176322418135,
+      "grad_norm": 0.2604944605147867,
+      "learning_rate": 5.224877591003374e-07,
+      "loss": 0.4581,
+      "step": 10251
+    },
+    {
+      "epoch": 2.869297509095998,
+      "grad_norm": 0.25014607563839375,
+      "learning_rate": 5.202659096333218e-07,
+      "loss": 0.4462,
+      "step": 10252
+    },
+    {
+      "epoch": 2.869577385950182,
+      "grad_norm": 0.26158734638337017,
+      "learning_rate": 5.180487696750946e-07,
+      "loss": 0.4418,
+      "step": 10253
+    },
+    {
+      "epoch": 2.869857262804366,
+      "grad_norm": 0.2628142041741432,
+      "learning_rate": 5.158363394366816e-07,
+      "loss": 0.4513,
+      "step": 10254
+    },
+    {
+      "epoch": 2.87013713965855,
+      "grad_norm": 0.25316528938359184,
+      "learning_rate": 5.136286191286643e-07,
+      "loss": 0.4393,
+      "step": 10255
+    },
+    {
+      "epoch": 2.8704170165127345,
+      "grad_norm": 0.2495107263010776,
+      "learning_rate": 5.114256089611747e-07,
+      "loss": 0.445,
+      "step": 10256
+    },
+    {
+      "epoch": 2.870696893366919,
+      "grad_norm": 0.25869314266387045,
+      "learning_rate": 5.092273091438948e-07,
+      "loss": 0.4538,
+      "step": 10257
+    },
+    {
+      "epoch": 2.8709767702211026,
+      "grad_norm": 0.2554676058510657,
+      "learning_rate": 5.070337198860631e-07,
+      "loss": 0.4623,
+      "step": 10258
+    },
+    {
+      "epoch": 2.871256647075287,
+      "grad_norm": 0.24537075832570931,
+      "learning_rate": 5.048448413964624e-07,
+      "loss": 0.4546,
+      "step": 10259
+    },
+    {
+      "epoch": 2.8715365239294712,
+      "grad_norm": 0.2672389569675875,
+      "learning_rate": 5.026606738834317e-07,
+      "loss": 0.4557,
+      "step": 10260
+    },
+    {
+      "epoch": 2.871816400783655,
+      "grad_norm": 0.24970304197948354,
+      "learning_rate": 5.004812175548656e-07,
+      "loss": 0.4586,
+      "step": 10261
+    },
+    {
+      "epoch": 2.8720962776378394,
+      "grad_norm": 0.26349661025087207,
+      "learning_rate": 4.983064726181986e-07,
+      "loss": 0.4649,
+      "step": 10262
+    },
+    {
+      "epoch": 2.8723761544920237,
+      "grad_norm": 0.24778771885993947,
+      "learning_rate": 4.961364392804313e-07,
+      "loss": 0.4433,
+      "step": 10263
+    },
+    {
+      "epoch": 2.8726560313462075,
+      "grad_norm": 0.26672455734691347,
+      "learning_rate": 4.939711177481099e-07,
+      "loss": 0.4429,
+      "step": 10264
+    },
+    {
+      "epoch": 2.872935908200392,
+      "grad_norm": 0.2548923418425755,
+      "learning_rate": 4.91810508227325e-07,
+      "loss": 0.4397,
+      "step": 10265
+    },
+    {
+      "epoch": 2.873215785054576,
+      "grad_norm": 0.25746339083089437,
+      "learning_rate": 4.896546109237232e-07,
+      "loss": 0.4322,
+      "step": 10266
+    },
+    {
+      "epoch": 2.87349566190876,
+      "grad_norm": 0.2571032628499338,
+      "learning_rate": 4.875034260425126e-07,
+      "loss": 0.4665,
+      "step": 10267
+    },
+    {
+      "epoch": 2.8737755387629442,
+      "grad_norm": 0.2611386216444229,
+      "learning_rate": 4.853569537884406e-07,
+      "loss": 0.4496,
+      "step": 10268
+    },
+    {
+      "epoch": 2.8740554156171285,
+      "grad_norm": 0.2554305595408042,
+      "learning_rate": 4.832151943658048e-07,
+      "loss": 0.4374,
+      "step": 10269
+    },
+    {
+      "epoch": 2.8743352924713124,
+      "grad_norm": 0.26483360105662895,
+      "learning_rate": 4.810781479784588e-07,
+      "loss": 0.455,
+      "step": 10270
+    },
+    {
+      "epoch": 2.8746151693254967,
+      "grad_norm": 0.252441584593939,
+      "learning_rate": 4.789458148298176e-07,
+      "loss": 0.4432,
+      "step": 10271
+    },
+    {
+      "epoch": 2.874895046179681,
+      "grad_norm": 0.2603269128030735,
+      "learning_rate": 4.768181951228301e-07,
+      "loss": 0.4651,
+      "step": 10272
+    },
+    {
+      "epoch": 2.8751749230338652,
+      "grad_norm": 0.24957648626480253,
+      "learning_rate": 4.746952890600065e-07,
+      "loss": 0.455,
+      "step": 10273
+    },
+    {
+      "epoch": 2.8754547998880495,
+      "grad_norm": 0.26391183669801865,
+      "learning_rate": 4.725770968434018e-07,
+      "loss": 0.4541,
+      "step": 10274
+    },
+    {
+      "epoch": 2.8757346767422334,
+      "grad_norm": 0.25716393251231623,
+      "learning_rate": 4.7046361867463807e-07,
+      "loss": 0.4608,
+      "step": 10275
+    },
+    {
+      "epoch": 2.8760145535964177,
+      "grad_norm": 0.26668582501615473,
+      "learning_rate": 4.683548547548655e-07,
+      "loss": 0.4512,
+      "step": 10276
+    },
+    {
+      "epoch": 2.876294430450602,
+      "grad_norm": 0.2455419391449943,
+      "learning_rate": 4.6625080528480137e-07,
+      "loss": 0.4527,
+      "step": 10277
+    },
+    {
+      "epoch": 2.876574307304786,
+      "grad_norm": 0.2570455687678307,
+      "learning_rate": 4.641514704647132e-07,
+      "loss": 0.457,
+      "step": 10278
+    },
+    {
+      "epoch": 2.87685418415897,
+      "grad_norm": 0.24449720654527113,
+      "learning_rate": 4.620568504944134e-07,
+      "loss": 0.4511,
+      "step": 10279
+    },
+    {
+      "epoch": 2.8771340610131544,
+      "grad_norm": 0.24351161087254086,
+      "learning_rate": 4.5996694557327025e-07,
+      "loss": 0.439,
+      "step": 10280
+    },
+    {
+      "epoch": 2.8774139378673382,
+      "grad_norm": 0.26099511112058876,
+      "learning_rate": 4.5788175590020246e-07,
+      "loss": 0.469,
+      "step": 10281
+    },
+    {
+      "epoch": 2.8776938147215225,
+      "grad_norm": 0.2582471683662392,
+      "learning_rate": 4.5580128167367895e-07,
+      "loss": 0.4309,
+      "step": 10282
+    },
+    {
+      "epoch": 2.877973691575707,
+      "grad_norm": 0.24795866638590033,
+      "learning_rate": 4.5372552309171925e-07,
+      "loss": 0.4668,
+      "step": 10283
+    },
+    {
+      "epoch": 2.8782535684298907,
+      "grad_norm": 0.25154649095883064,
+      "learning_rate": 4.516544803518985e-07,
+      "loss": 0.4477,
+      "step": 10284
+    },
+    {
+      "epoch": 2.878533445284075,
+      "grad_norm": 0.2505105586047212,
+      "learning_rate": 4.495881536513369e-07,
+      "loss": 0.419,
+      "step": 10285
+    },
+    {
+      "epoch": 2.8788133221382592,
+      "grad_norm": 0.25352983127315026,
+      "learning_rate": 4.4752654318670485e-07,
+      "loss": 0.4531,
+      "step": 10286
+    },
+    {
+      "epoch": 2.879093198992443,
+      "grad_norm": 0.25948613905647183,
+      "learning_rate": 4.4546964915423986e-07,
+      "loss": 0.4618,
+      "step": 10287
+    },
+    {
+      "epoch": 2.8793730758466274,
+      "grad_norm": 0.2582180835960804,
+      "learning_rate": 4.434174717497075e-07,
+      "loss": 0.4542,
+      "step": 10288
+    },
+    {
+      "epoch": 2.8796529527008117,
+      "grad_norm": 0.257995531526806,
+      "learning_rate": 4.413700111684349e-07,
+      "loss": 0.4611,
+      "step": 10289
+    },
+    {
+      "epoch": 2.879932829554996,
+      "grad_norm": 0.2536050936538131,
+      "learning_rate": 4.3932726760531064e-07,
+      "loss": 0.4331,
+      "step": 10290
+    },
+    {
+      "epoch": 2.88021270640918,
+      "grad_norm": 0.26348346728174327,
+      "learning_rate": 4.3728924125475137e-07,
+      "loss": 0.4652,
+      "step": 10291
+    },
+    {
+      "epoch": 2.880492583263364,
+      "grad_norm": 0.25005631660677363,
+      "learning_rate": 4.3525593231074633e-07,
+      "loss": 0.4544,
+      "step": 10292
+    },
+    {
+      "epoch": 2.8807724601175484,
+      "grad_norm": 0.2555331171203246,
+      "learning_rate": 4.3322734096682417e-07,
+      "loss": 0.457,
+      "step": 10293
+    },
+    {
+      "epoch": 2.8810523369717327,
+      "grad_norm": 0.2545421363229972,
+      "learning_rate": 4.3120346741606367e-07,
+      "loss": 0.4686,
+      "step": 10294
+    },
+    {
+      "epoch": 2.8813322138259165,
+      "grad_norm": 0.2617427853569952,
+      "learning_rate": 4.2918431185110517e-07,
+      "loss": 0.4435,
+      "step": 10295
+    },
+    {
+      "epoch": 2.881612090680101,
+      "grad_norm": 0.2611070268049796,
+      "learning_rate": 4.271698744641339e-07,
+      "loss": 0.4822,
+      "step": 10296
+    },
+    {
+      "epoch": 2.881891967534285,
+      "grad_norm": 0.2651306945435216,
+      "learning_rate": 4.2516015544687426e-07,
+      "loss": 0.4603,
+      "step": 10297
+    },
+    {
+      "epoch": 2.882171844388469,
+      "grad_norm": 0.3124262786349235,
+      "learning_rate": 4.2315515499062317e-07,
+      "loss": 0.4459,
+      "step": 10298
+    },
+    {
+      "epoch": 2.8824517212426533,
+      "grad_norm": 0.2685649925102902,
+      "learning_rate": 4.211548732862114e-07,
+      "loss": 0.4392,
+      "step": 10299
+    },
+    {
+      "epoch": 2.8827315980968375,
+      "grad_norm": 0.2532537109973419,
+      "learning_rate": 4.19159310524031e-07,
+      "loss": 0.442,
+      "step": 10300
+    },
+    {
+      "epoch": 2.8830114749510214,
+      "grad_norm": 0.25293340380592855,
+      "learning_rate": 4.17168466894019e-07,
+      "loss": 0.4272,
+      "step": 10301
+    },
+    {
+      "epoch": 2.8832913518052057,
+      "grad_norm": 0.2626769114687486,
+      "learning_rate": 4.1518234258566824e-07,
+      "loss": 0.4514,
+      "step": 10302
+    },
+    {
+      "epoch": 2.88357122865939,
+      "grad_norm": 0.2548871588665236,
+      "learning_rate": 4.132009377880108e-07,
+      "loss": 0.4563,
+      "step": 10303
+    },
+    {
+      "epoch": 2.883851105513574,
+      "grad_norm": 0.2499079155484336,
+      "learning_rate": 4.1122425268964593e-07,
+      "loss": 0.4488,
+      "step": 10304
+    },
+    {
+      "epoch": 2.884130982367758,
+      "grad_norm": 0.2774741682164875,
+      "learning_rate": 4.092522874787119e-07,
+      "loss": 0.4595,
+      "step": 10305
+    },
+    {
+      "epoch": 2.8844108592219424,
+      "grad_norm": 0.2620330159986628,
+      "learning_rate": 4.072850423428975e-07,
+      "loss": 0.4553,
+      "step": 10306
+    },
+    {
+      "epoch": 2.8846907360761262,
+      "grad_norm": 0.24761164989931458,
+      "learning_rate": 4.053225174694586e-07,
+      "loss": 0.4532,
+      "step": 10307
+    },
+    {
+      "epoch": 2.8849706129303105,
+      "grad_norm": 0.24479380194816994,
+      "learning_rate": 4.0336471304517897e-07,
+      "loss": 0.4172,
+      "step": 10308
+    },
+    {
+      "epoch": 2.885250489784495,
+      "grad_norm": 0.2514815231592224,
+      "learning_rate": 4.014116292564041e-07,
+      "loss": 0.4459,
+      "step": 10309
+    },
+    {
+      "epoch": 2.885530366638679,
+      "grad_norm": 0.2923214184067987,
+      "learning_rate": 3.9946326628903516e-07,
+      "loss": 0.4461,
+      "step": 10310
+    },
+    {
+      "epoch": 2.8858102434928634,
+      "grad_norm": 0.24988649212900807,
+      "learning_rate": 3.975196243285129e-07,
+      "loss": 0.4624,
+      "step": 10311
+    },
+    {
+      "epoch": 2.8860901203470473,
+      "grad_norm": 0.2483774134025153,
+      "learning_rate": 3.9558070355983357e-07,
+      "loss": 0.447,
+      "step": 10312
+    },
+    {
+      "epoch": 2.8863699972012316,
+      "grad_norm": 0.2575682884161871,
+      "learning_rate": 3.9364650416755525e-07,
+      "loss": 0.4319,
+      "step": 10313
+    },
+    {
+      "epoch": 2.886649874055416,
+      "grad_norm": 0.2563758815757559,
+      "learning_rate": 3.9171702633576945e-07,
+      "loss": 0.423,
+      "step": 10314
+    },
+    {
+      "epoch": 2.8869297509095997,
+      "grad_norm": 0.25206218781065665,
+      "learning_rate": 3.8979227024811826e-07,
+      "loss": 0.4592,
+      "step": 10315
+    },
+    {
+      "epoch": 2.887209627763784,
+      "grad_norm": 0.25445140057837873,
+      "learning_rate": 3.878722360878051e-07,
+      "loss": 0.4538,
+      "step": 10316
+    },
+    {
+      "epoch": 2.8874895046179683,
+      "grad_norm": 0.25892850164373754,
+      "learning_rate": 3.8595692403758376e-07,
+      "loss": 0.4565,
+      "step": 10317
+    },
+    {
+      "epoch": 2.887769381472152,
+      "grad_norm": 0.2571744291136916,
+      "learning_rate": 3.8404633427975846e-07,
+      "loss": 0.4458,
+      "step": 10318
+    },
+    {
+      "epoch": 2.8880492583263364,
+      "grad_norm": 0.25961005904007983,
+      "learning_rate": 3.82140466996167e-07,
+      "loss": 0.4681,
+      "step": 10319
+    },
+    {
+      "epoch": 2.8883291351805207,
+      "grad_norm": 0.2626162006435372,
+      "learning_rate": 3.802393223682199e-07,
+      "loss": 0.4348,
+      "step": 10320
+    },
+    {
+      "epoch": 2.8886090120347045,
+      "grad_norm": 0.26403517659005993,
+      "learning_rate": 3.7834290057686684e-07,
+      "loss": 0.4463,
+      "step": 10321
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.2640114446640457,
+      "learning_rate": 3.764512018026134e-07,
+      "loss": 0.4491,
+      "step": 10322
+    },
+    {
+      "epoch": 2.889168765743073,
+      "grad_norm": 0.2567861703073457,
+      "learning_rate": 3.7456422622551e-07,
+      "loss": 0.4629,
+      "step": 10323
+    },
+    {
+      "epoch": 2.889448642597257,
+      "grad_norm": 0.26970481734620666,
+      "learning_rate": 3.726819740251575e-07,
+      "loss": 0.4425,
+      "step": 10324
+    },
+    {
+      "epoch": 2.8897285194514413,
+      "grad_norm": 0.2573265949733196,
+      "learning_rate": 3.70804445380718e-07,
+      "loss": 0.4339,
+      "step": 10325
+    },
+    {
+      "epoch": 2.8900083963056256,
+      "grad_norm": 0.2457024445347563,
+      "learning_rate": 3.68931640470882e-07,
+      "loss": 0.4542,
+      "step": 10326
+    },
+    {
+      "epoch": 2.8902882731598094,
+      "grad_norm": 0.2518158589030567,
+      "learning_rate": 3.67063559473918e-07,
+      "loss": 0.4413,
+      "step": 10327
+    },
+    {
+      "epoch": 2.8905681500139937,
+      "grad_norm": 0.2541691212113756,
+      "learning_rate": 3.6520020256762267e-07,
+      "loss": 0.4354,
+      "step": 10328
+    },
+    {
+      "epoch": 2.890848026868178,
+      "grad_norm": 0.26148669070858493,
+      "learning_rate": 3.6334156992935406e-07,
+      "loss": 0.4758,
+      "step": 10329
+    },
+    {
+      "epoch": 2.8911279037223623,
+      "grad_norm": 0.2578137534929587,
+      "learning_rate": 3.614876617360152e-07,
+      "loss": 0.4694,
+      "step": 10330
+    },
+    {
+      "epoch": 2.8914077805765466,
+      "grad_norm": 0.25462776023270506,
+      "learning_rate": 3.596384781640705e-07,
+      "loss": 0.4343,
+      "step": 10331
+    },
+    {
+      "epoch": 2.8916876574307304,
+      "grad_norm": 0.2574136099033175,
+      "learning_rate": 3.5779401938951794e-07,
+      "loss": 0.4517,
+      "step": 10332
+    },
+    {
+      "epoch": 2.8919675342849147,
+      "grad_norm": 0.2528803426108313,
+      "learning_rate": 3.559542855879172e-07,
+      "loss": 0.4427,
+      "step": 10333
+    },
+    {
+      "epoch": 2.892247411139099,
+      "grad_norm": 0.2428749181134538,
+      "learning_rate": 3.5411927693437265e-07,
+      "loss": 0.4381,
+      "step": 10334
+    },
+    {
+      "epoch": 2.892527287993283,
+      "grad_norm": 0.2563684456973611,
+      "learning_rate": 3.522889936035445e-07,
+      "loss": 0.4615,
+      "step": 10335
+    },
+    {
+      "epoch": 2.892807164847467,
+      "grad_norm": 0.2603322799709365,
+      "learning_rate": 3.504634357696379e-07,
+      "loss": 0.469,
+      "step": 10336
+    },
+    {
+      "epoch": 2.8930870417016514,
+      "grad_norm": 0.24520091679492342,
+      "learning_rate": 3.4864260360641386e-07,
+      "loss": 0.4374,
+      "step": 10337
+    },
+    {
+      "epoch": 2.8933669185558353,
+      "grad_norm": 0.265476476947951,
+      "learning_rate": 3.468264972871782e-07,
+      "loss": 0.463,
+      "step": 10338
+    },
+    {
+      "epoch": 2.8936467954100196,
+      "grad_norm": 0.2716861089064996,
+      "learning_rate": 3.4501511698478705e-07,
+      "loss": 0.4538,
+      "step": 10339
+    },
+    {
+      "epoch": 2.893926672264204,
+      "grad_norm": 0.24986600046345175,
+      "learning_rate": 3.43208462871647e-07,
+      "loss": 0.4434,
+      "step": 10340
+    },
+    {
+      "epoch": 2.8942065491183877,
+      "grad_norm": 0.25661580131767164,
+      "learning_rate": 3.41406535119726e-07,
+      "loss": 0.4358,
+      "step": 10341
+    },
+    {
+      "epoch": 2.894486425972572,
+      "grad_norm": 0.26288992965346053,
+      "learning_rate": 3.396093339005202e-07,
+      "loss": 0.4532,
+      "step": 10342
+    },
+    {
+      "epoch": 2.8947663028267563,
+      "grad_norm": 0.2598594542669984,
+      "learning_rate": 3.378168593850983e-07,
+      "loss": 0.4471,
+      "step": 10343
+    },
+    {
+      "epoch": 2.89504617968094,
+      "grad_norm": 0.2620213552419905,
+      "learning_rate": 3.3602911174406283e-07,
+      "loss": 0.4515,
+      "step": 10344
+    },
+    {
+      "epoch": 2.8953260565351244,
+      "grad_norm": 0.33377074231948806,
+      "learning_rate": 3.34246091147572e-07,
+      "loss": 0.4552,
+      "step": 10345
+    },
+    {
+      "epoch": 2.8956059333893087,
+      "grad_norm": 0.2577495637966244,
+      "learning_rate": 3.324677977653401e-07,
+      "loss": 0.4417,
+      "step": 10346
+    },
+    {
+      "epoch": 2.895885810243493,
+      "grad_norm": 0.275292397605857,
+      "learning_rate": 3.306942317666206e-07,
+      "loss": 0.462,
+      "step": 10347
+    },
+    {
+      "epoch": 2.8961656870976773,
+      "grad_norm": 0.24961360541812772,
+      "learning_rate": 3.289253933202285e-07,
+      "loss": 0.4533,
+      "step": 10348
+    },
+    {
+      "epoch": 2.896445563951861,
+      "grad_norm": 0.2661825316055105,
+      "learning_rate": 3.2716128259451804e-07,
+      "loss": 0.4429,
+      "step": 10349
+    },
+    {
+      "epoch": 2.8967254408060454,
+      "grad_norm": 0.25305679025577094,
+      "learning_rate": 3.254018997574049e-07,
+      "loss": 0.45,
+      "step": 10350
+    },
+    {
+      "epoch": 2.8970053176602297,
+      "grad_norm": 0.25119578479911997,
+      "learning_rate": 3.236472449763384e-07,
+      "loss": 0.4816,
+      "step": 10351
+    },
+    {
+      "epoch": 2.8972851945144136,
+      "grad_norm": 0.25395181283716584,
+      "learning_rate": 3.218973184183349e-07,
+      "loss": 0.4181,
+      "step": 10352
+    },
+    {
+      "epoch": 2.897565071368598,
+      "grad_norm": 0.25487831155720125,
+      "learning_rate": 3.2015212024995024e-07,
+      "loss": 0.4494,
+      "step": 10353
+    },
+    {
+      "epoch": 2.897844948222782,
+      "grad_norm": 0.2599533908619156,
+      "learning_rate": 3.184116506372903e-07,
+      "loss": 0.4703,
+      "step": 10354
+    },
+    {
+      "epoch": 2.898124825076966,
+      "grad_norm": 0.25358339034142985,
+      "learning_rate": 3.1667590974602256e-07,
+      "loss": 0.4428,
+      "step": 10355
+    },
+    {
+      "epoch": 2.8984047019311503,
+      "grad_norm": 0.26625769348057865,
+      "learning_rate": 3.1494489774134273e-07,
+      "loss": 0.4752,
+      "step": 10356
+    },
+    {
+      "epoch": 2.8986845787853346,
+      "grad_norm": 0.2534972553995035,
+      "learning_rate": 3.132186147880245e-07,
+      "loss": 0.4547,
+      "step": 10357
+    },
+    {
+      "epoch": 2.8989644556395184,
+      "grad_norm": 0.24764399053567493,
+      "learning_rate": 3.114970610503698e-07,
+      "loss": 0.4588,
+      "step": 10358
+    },
+    {
+      "epoch": 2.8992443324937027,
+      "grad_norm": 0.2565092323150165,
+      "learning_rate": 3.0978023669223087e-07,
+      "loss": 0.4616,
+      "step": 10359
+    },
+    {
+      "epoch": 2.899524209347887,
+      "grad_norm": 0.2525542577336461,
+      "learning_rate": 3.0806814187702703e-07,
+      "loss": 0.4575,
+      "step": 10360
+    },
+    {
+      "epoch": 2.899804086202071,
+      "grad_norm": 0.26140960295508986,
+      "learning_rate": 3.0636077676770567e-07,
+      "loss": 0.4497,
+      "step": 10361
+    },
+    {
+      "epoch": 2.900083963056255,
+      "grad_norm": 0.2679370880273387,
+      "learning_rate": 3.0465814152678685e-07,
+      "loss": 0.4545,
+      "step": 10362
+    },
+    {
+      "epoch": 2.9003638399104394,
+      "grad_norm": 0.2699083430806211,
+      "learning_rate": 3.0296023631631865e-07,
+      "loss": 0.4643,
+      "step": 10363
+    },
+    {
+      "epoch": 2.9006437167646233,
+      "grad_norm": 0.26170266192893454,
+      "learning_rate": 3.0126706129791073e-07,
+      "loss": 0.4561,
+      "step": 10364
+    },
+    {
+      "epoch": 2.9009235936188076,
+      "grad_norm": 0.24858769529730645,
+      "learning_rate": 2.99578616632723e-07,
+      "loss": 0.453,
+      "step": 10365
+    },
+    {
+      "epoch": 2.901203470472992,
+      "grad_norm": 0.2574791076494076,
+      "learning_rate": 2.9789490248146033e-07,
+      "loss": 0.4434,
+      "step": 10366
+    },
+    {
+      "epoch": 2.901483347327176,
+      "grad_norm": 0.26240149623407355,
+      "learning_rate": 2.9621591900437784e-07,
+      "loss": 0.4402,
+      "step": 10367
+    },
+    {
+      "epoch": 2.9017632241813605,
+      "grad_norm": 0.2507582670062776,
+      "learning_rate": 2.945416663612921e-07,
+      "loss": 0.4458,
+      "step": 10368
+    },
+    {
+      "epoch": 2.9020431010355443,
+      "grad_norm": 0.2471406035226927,
+      "learning_rate": 2.928721447115479e-07,
+      "loss": 0.4346,
+      "step": 10369
+    },
+    {
+      "epoch": 2.9023229778897286,
+      "grad_norm": 0.25689886731465,
+      "learning_rate": 2.91207354214057e-07,
+      "loss": 0.4527,
+      "step": 10370
+    },
+    {
+      "epoch": 2.902602854743913,
+      "grad_norm": 0.25295420038947664,
+      "learning_rate": 2.895472950272704e-07,
+      "loss": 0.4379,
+      "step": 10371
+    },
+    {
+      "epoch": 2.9028827315980967,
+      "grad_norm": 0.2560257149664785,
+      "learning_rate": 2.878919673091951e-07,
+      "loss": 0.4282,
+      "step": 10372
+    },
+    {
+      "epoch": 2.903162608452281,
+      "grad_norm": 0.2696114038415163,
+      "learning_rate": 2.8624137121738836e-07,
+      "loss": 0.4706,
+      "step": 10373
+    },
+    {
+      "epoch": 2.9034424853064653,
+      "grad_norm": 0.2500674589002211,
+      "learning_rate": 2.845955069089579e-07,
+      "loss": 0.4488,
+      "step": 10374
+    },
+    {
+      "epoch": 2.903722362160649,
+      "grad_norm": 0.2554546188397463,
+      "learning_rate": 2.829543745405505e-07,
+      "loss": 0.4372,
+      "step": 10375
+    },
+    {
+      "epoch": 2.9040022390148335,
+      "grad_norm": 0.24931022196936325,
+      "learning_rate": 2.813179742683747e-07,
+      "loss": 0.4526,
+      "step": 10376
+    },
+    {
+      "epoch": 2.9042821158690177,
+      "grad_norm": 0.2928468198012615,
+      "learning_rate": 2.796863062481836e-07,
+      "loss": 0.4511,
+      "step": 10377
+    },
+    {
+      "epoch": 2.9045619927232016,
+      "grad_norm": 0.26388584571093515,
+      "learning_rate": 2.780593706352752e-07,
+      "loss": 0.4539,
+      "step": 10378
+    },
+    {
+      "epoch": 2.904841869577386,
+      "grad_norm": 0.2498804927398296,
+      "learning_rate": 2.7643716758451457e-07,
+      "loss": 0.4373,
+      "step": 10379
+    },
+    {
+      "epoch": 2.90512174643157,
+      "grad_norm": 0.26863275546009896,
+      "learning_rate": 2.748196972502892e-07,
+      "loss": 0.4611,
+      "step": 10380
+    },
+    {
+      "epoch": 2.905401623285754,
+      "grad_norm": 0.2517534306363316,
+      "learning_rate": 2.7320695978655943e-07,
+      "loss": 0.4529,
+      "step": 10381
+    },
+    {
+      "epoch": 2.9056815001399383,
+      "grad_norm": 0.2560660861816635,
+      "learning_rate": 2.71598955346819e-07,
+      "loss": 0.463,
+      "step": 10382
+    },
+    {
+      "epoch": 2.9059613769941226,
+      "grad_norm": 0.2605364040440226,
+      "learning_rate": 2.6999568408413443e-07,
+      "loss": 0.4483,
+      "step": 10383
+    },
+    {
+      "epoch": 2.906241253848307,
+      "grad_norm": 0.24860598895959993,
+      "learning_rate": 2.6839714615108926e-07,
+      "loss": 0.4366,
+      "step": 10384
+    },
+    {
+      "epoch": 2.9065211307024907,
+      "grad_norm": 0.2540350770123497,
+      "learning_rate": 2.6680334169983946e-07,
+      "loss": 0.4506,
+      "step": 10385
+    },
+    {
+      "epoch": 2.906801007556675,
+      "grad_norm": 0.26270021658466275,
+      "learning_rate": 2.652142708820915e-07,
+      "loss": 0.4541,
+      "step": 10386
+    },
+    {
+      "epoch": 2.9070808844108593,
+      "grad_norm": 0.2660916319559597,
+      "learning_rate": 2.6362993384907997e-07,
+      "loss": 0.4614,
+      "step": 10387
+    },
+    {
+      "epoch": 2.9073607612650436,
+      "grad_norm": 0.25039301873484276,
+      "learning_rate": 2.62050330751612e-07,
+      "loss": 0.4466,
+      "step": 10388
+    },
+    {
+      "epoch": 2.9076406381192275,
+      "grad_norm": 0.24631662305707414,
+      "learning_rate": 2.60475461740034e-07,
+      "loss": 0.4223,
+      "step": 10389
+    },
+    {
+      "epoch": 2.9079205149734118,
+      "grad_norm": 0.24299943031929547,
+      "learning_rate": 2.5890532696424274e-07,
+      "loss": 0.4428,
+      "step": 10390
+    },
+    {
+      "epoch": 2.908200391827596,
+      "grad_norm": 0.25736270614928747,
+      "learning_rate": 2.573399265736798e-07,
+      "loss": 0.4457,
+      "step": 10391
+    },
+    {
+      "epoch": 2.90848026868178,
+      "grad_norm": 0.25310670202243307,
+      "learning_rate": 2.557792607173481e-07,
+      "loss": 0.4447,
+      "step": 10392
+    },
+    {
+      "epoch": 2.908760145535964,
+      "grad_norm": 0.2601763089070998,
+      "learning_rate": 2.5422332954379014e-07,
+      "loss": 0.4497,
+      "step": 10393
+    },
+    {
+      "epoch": 2.9090400223901485,
+      "grad_norm": 0.25673423214608204,
+      "learning_rate": 2.5267213320109283e-07,
+      "loss": 0.4462,
+      "step": 10394
+    },
+    {
+      "epoch": 2.9093198992443323,
+      "grad_norm": 0.2732253329291374,
+      "learning_rate": 2.5112567183691595e-07,
+      "loss": 0.4579,
+      "step": 10395
+    },
+    {
+      "epoch": 2.9095997760985166,
+      "grad_norm": 0.24489089427254132,
+      "learning_rate": 2.4958394559843614e-07,
+      "loss": 0.4637,
+      "step": 10396
+    },
+    {
+      "epoch": 2.909879652952701,
+      "grad_norm": 0.2597770859891086,
+      "learning_rate": 2.4804695463240826e-07,
+      "loss": 0.4524,
+      "step": 10397
+    },
+    {
+      "epoch": 2.9101595298068847,
+      "grad_norm": 0.25115464762045403,
+      "learning_rate": 2.4651469908511527e-07,
+      "loss": 0.4323,
+      "step": 10398
+    },
+    {
+      "epoch": 2.910439406661069,
+      "grad_norm": 0.25904281835353893,
+      "learning_rate": 2.4498717910240163e-07,
+      "loss": 0.4532,
+      "step": 10399
+    },
+    {
+      "epoch": 2.9107192835152533,
+      "grad_norm": 0.26533887036075104,
+      "learning_rate": 2.4346439482965664e-07,
+      "loss": 0.4473,
+      "step": 10400
+    },
+    {
+      "epoch": 2.910999160369437,
+      "grad_norm": 0.25950334993597235,
+      "learning_rate": 2.419463464118199e-07,
+      "loss": 0.4503,
+      "step": 10401
+    },
+    {
+      "epoch": 2.9112790372236215,
+      "grad_norm": 0.2592923155127461,
+      "learning_rate": 2.404330339933869e-07,
+      "loss": 0.4539,
+      "step": 10402
+    },
+    {
+      "epoch": 2.9115589140778058,
+      "grad_norm": 0.24565959064147494,
+      "learning_rate": 2.3892445771838134e-07,
+      "loss": 0.4426,
+      "step": 10403
+    },
+    {
+      "epoch": 2.91183879093199,
+      "grad_norm": 0.2564125328004262,
+      "learning_rate": 2.37420617730405e-07,
+      "loss": 0.4456,
+      "step": 10404
+    },
+    {
+      "epoch": 2.9121186677861743,
+      "grad_norm": 0.25657770161499827,
+      "learning_rate": 2.359215141725879e-07,
+      "loss": 0.4584,
+      "step": 10405
+    },
+    {
+      "epoch": 2.912398544640358,
+      "grad_norm": 0.2671894407257729,
+      "learning_rate": 2.3442714718761027e-07,
+      "loss": 0.4621,
+      "step": 10406
+    },
+    {
+      "epoch": 2.9126784214945425,
+      "grad_norm": 0.2676332770972902,
+      "learning_rate": 2.3293751691771394e-07,
+      "loss": 0.4649,
+      "step": 10407
+    },
+    {
+      "epoch": 2.9129582983487268,
+      "grad_norm": 0.2594919036123832,
+      "learning_rate": 2.314526235046799e-07,
+      "loss": 0.4604,
+      "step": 10408
+    },
+    {
+      "epoch": 2.9132381752029106,
+      "grad_norm": 0.2593525779189398,
+      "learning_rate": 2.299724670898451e-07,
+      "loss": 0.4349,
+      "step": 10409
+    },
+    {
+      "epoch": 2.913518052057095,
+      "grad_norm": 0.25144203995667563,
+      "learning_rate": 2.2849704781408577e-07,
+      "loss": 0.4491,
+      "step": 10410
+    },
+    {
+      "epoch": 2.913797928911279,
+      "grad_norm": 0.2635956450162497,
+      "learning_rate": 2.2702636581783399e-07,
+      "loss": 0.4605,
+      "step": 10411
+    },
+    {
+      "epoch": 2.914077805765463,
+      "grad_norm": 0.2600994568333365,
+      "learning_rate": 2.2556042124107223e-07,
+      "loss": 0.4368,
+      "step": 10412
+    },
+    {
+      "epoch": 2.9143576826196473,
+      "grad_norm": 0.2564418348360325,
+      "learning_rate": 2.2409921422333335e-07,
+      "loss": 0.4497,
+      "step": 10413
+    },
+    {
+      "epoch": 2.9146375594738316,
+      "grad_norm": 0.249296973255,
+      "learning_rate": 2.226427449036894e-07,
+      "loss": 0.4578,
+      "step": 10414
+    },
+    {
+      "epoch": 2.9149174363280155,
+      "grad_norm": 0.2624082979505958,
+      "learning_rate": 2.2119101342076841e-07,
+      "loss": 0.4557,
+      "step": 10415
+    },
+    {
+      "epoch": 2.9151973131821998,
+      "grad_norm": 0.2482312610438998,
+      "learning_rate": 2.1974401991274873e-07,
+      "loss": 0.4452,
+      "step": 10416
+    },
+    {
+      "epoch": 2.915477190036384,
+      "grad_norm": 0.28259425383700054,
+      "learning_rate": 2.1830176451735905e-07,
+      "loss": 0.4542,
+      "step": 10417
+    },
+    {
+      "epoch": 2.915757066890568,
+      "grad_norm": 0.25328344799056274,
+      "learning_rate": 2.1686424737187295e-07,
+      "loss": 0.4447,
+      "step": 10418
+    },
+    {
+      "epoch": 2.916036943744752,
+      "grad_norm": 0.2536325406702005,
+      "learning_rate": 2.1543146861311426e-07,
+      "loss": 0.4425,
+      "step": 10419
+    },
+    {
+      "epoch": 2.9163168205989365,
+      "grad_norm": 0.24890910384382822,
+      "learning_rate": 2.1400342837745168e-07,
+      "loss": 0.4497,
+      "step": 10420
+    },
+    {
+      "epoch": 2.916596697453121,
+      "grad_norm": 0.2544036580834295,
+      "learning_rate": 2.125801268008043e-07,
+      "loss": 0.4449,
+      "step": 10421
+    },
+    {
+      "epoch": 2.9168765743073046,
+      "grad_norm": 0.25742946778671977,
+      "learning_rate": 2.1116156401865815e-07,
+      "loss": 0.4379,
+      "step": 10422
+    },
+    {
+      "epoch": 2.917156451161489,
+      "grad_norm": 0.24188272688832368,
+      "learning_rate": 2.0974774016601638e-07,
+      "loss": 0.4396,
+      "step": 10423
+    },
+    {
+      "epoch": 2.917436328015673,
+      "grad_norm": 0.25419956448380476,
+      "learning_rate": 2.0833865537744912e-07,
+      "loss": 0.4445,
+      "step": 10424
+    },
+    {
+      "epoch": 2.9177162048698575,
+      "grad_norm": 0.2541023006339829,
+      "learning_rate": 2.0693430978708795e-07,
+      "loss": 0.4506,
+      "step": 10425
+    },
+    {
+      "epoch": 2.9179960817240413,
+      "grad_norm": 0.24882416696091192,
+      "learning_rate": 2.0553470352858152e-07,
+      "loss": 0.4491,
+      "step": 10426
+    },
+    {
+      "epoch": 2.9182759585782256,
+      "grad_norm": 0.2613166798429044,
+      "learning_rate": 2.0413983673515659e-07,
+      "loss": 0.4515,
+      "step": 10427
+    },
+    {
+      "epoch": 2.91855583543241,
+      "grad_norm": 0.2532369966568343,
+      "learning_rate": 2.027497095395736e-07,
+      "loss": 0.4486,
+      "step": 10428
+    },
+    {
+      "epoch": 2.9188357122865938,
+      "grad_norm": 0.26685547084956296,
+      "learning_rate": 2.0136432207414347e-07,
+      "loss": 0.4506,
+      "step": 10429
+    },
+    {
+      "epoch": 2.919115589140778,
+      "grad_norm": 0.2614000751207966,
+      "learning_rate": 1.9998367447073285e-07,
+      "loss": 0.4389,
+      "step": 10430
+    },
+    {
+      "epoch": 2.9193954659949624,
+      "grad_norm": 0.26655927508353255,
+      "learning_rate": 1.9860776686075332e-07,
+      "loss": 0.4532,
+      "step": 10431
+    },
+    {
+      "epoch": 2.919675342849146,
+      "grad_norm": 0.2578419302241896,
+      "learning_rate": 1.9723659937515572e-07,
+      "loss": 0.4331,
+      "step": 10432
+    },
+    {
+      "epoch": 2.9199552197033305,
+      "grad_norm": 0.26151256479973034,
+      "learning_rate": 1.958701721444578e-07,
+      "loss": 0.4619,
+      "step": 10433
+    },
+    {
+      "epoch": 2.920235096557515,
+      "grad_norm": 0.2533984994980884,
+      "learning_rate": 1.9450848529871114e-07,
+      "loss": 0.4559,
+      "step": 10434
+    },
+    {
+      "epoch": 2.9205149734116986,
+      "grad_norm": 0.26181913630805637,
+      "learning_rate": 1.9315153896752868e-07,
+      "loss": 0.4638,
+      "step": 10435
+    },
+    {
+      "epoch": 2.920794850265883,
+      "grad_norm": 0.25441984983057697,
+      "learning_rate": 1.9179933328005718e-07,
+      "loss": 0.4444,
+      "step": 10436
+    },
+    {
+      "epoch": 2.921074727120067,
+      "grad_norm": 0.2568916127723896,
+      "learning_rate": 1.9045186836500472e-07,
+      "loss": 0.4368,
+      "step": 10437
+    },
+    {
+      "epoch": 2.921354603974251,
+      "grad_norm": 0.2611152004511161,
+      "learning_rate": 1.8910914435062433e-07,
+      "loss": 0.4619,
+      "step": 10438
+    },
+    {
+      "epoch": 2.9216344808284354,
+      "grad_norm": 0.25755190571414255,
+      "learning_rate": 1.8777116136471372e-07,
+      "loss": 0.4443,
+      "step": 10439
+    },
+    {
+      "epoch": 2.9219143576826196,
+      "grad_norm": 0.25147781885391063,
+      "learning_rate": 1.8643791953462664e-07,
+      "loss": 0.4446,
+      "step": 10440
+    },
+    {
+      "epoch": 2.922194234536804,
+      "grad_norm": 0.24820029427422324,
+      "learning_rate": 1.8510941898726153e-07,
+      "loss": 0.4537,
+      "step": 10441
+    },
+    {
+      "epoch": 2.9224741113909882,
+      "grad_norm": 0.2638186979540325,
+      "learning_rate": 1.837856598490617e-07,
+      "loss": 0.4554,
+      "step": 10442
+    },
+    {
+      "epoch": 2.922753988245172,
+      "grad_norm": 0.25462418346204635,
+      "learning_rate": 1.8246664224603193e-07,
+      "loss": 0.4428,
+      "step": 10443
+    },
+    {
+      "epoch": 2.9230338650993564,
+      "grad_norm": 0.2594138620877634,
+      "learning_rate": 1.8115236630370513e-07,
+      "loss": 0.4358,
+      "step": 10444
+    },
+    {
+      "epoch": 2.9233137419535407,
+      "grad_norm": 0.25762916254734597,
+      "learning_rate": 1.7984283214718122e-07,
+      "loss": 0.4703,
+      "step": 10445
+    },
+    {
+      "epoch": 2.9235936188077245,
+      "grad_norm": 0.24532429258654168,
+      "learning_rate": 1.7853803990110495e-07,
+      "loss": 0.462,
+      "step": 10446
+    },
+    {
+      "epoch": 2.923873495661909,
+      "grad_norm": 0.2562022449504887,
+      "learning_rate": 1.7723798968966032e-07,
+      "loss": 0.4544,
+      "step": 10447
+    },
+    {
+      "epoch": 2.924153372516093,
+      "grad_norm": 0.25893836124805875,
+      "learning_rate": 1.7594268163659278e-07,
+      "loss": 0.4451,
+      "step": 10448
+    },
+    {
+      "epoch": 2.924433249370277,
+      "grad_norm": 0.24266450193199873,
+      "learning_rate": 1.7465211586519258e-07,
+      "loss": 0.4507,
+      "step": 10449
+    },
+    {
+      "epoch": 2.9247131262244612,
+      "grad_norm": 0.2524898468017339,
+      "learning_rate": 1.7336629249828372e-07,
+      "loss": 0.4432,
+      "step": 10450
+    },
+    {
+      "epoch": 2.9249930030786455,
+      "grad_norm": 0.25327670454061246,
+      "learning_rate": 1.7208521165826829e-07,
+      "loss": 0.4703,
+      "step": 10451
+    },
+    {
+      "epoch": 2.9252728799328294,
+      "grad_norm": 0.2652988133563846,
+      "learning_rate": 1.7080887346707098e-07,
+      "loss": 0.4556,
+      "step": 10452
+    },
+    {
+      "epoch": 2.9255527567870137,
+      "grad_norm": 0.25212927409896774,
+      "learning_rate": 1.6953727804617237e-07,
+      "loss": 0.4334,
+      "step": 10453
+    },
+    {
+      "epoch": 2.925832633641198,
+      "grad_norm": 0.257078013870885,
+      "learning_rate": 1.6827042551660898e-07,
+      "loss": 0.4199,
+      "step": 10454
+    },
+    {
+      "epoch": 2.926112510495382,
+      "grad_norm": 0.26032378169304704,
+      "learning_rate": 1.670083159989566e-07,
+      "loss": 0.4716,
+      "step": 10455
+    },
+    {
+      "epoch": 2.926392387349566,
+      "grad_norm": 0.24862097369804104,
+      "learning_rate": 1.6575094961335248e-07,
+      "loss": 0.4319,
+      "step": 10456
+    },
+    {
+      "epoch": 2.9266722642037504,
+      "grad_norm": 0.24885533344996946,
+      "learning_rate": 1.644983264794564e-07,
+      "loss": 0.4413,
+      "step": 10457
+    },
+    {
+      "epoch": 2.9269521410579347,
+      "grad_norm": 0.261188700334077,
+      "learning_rate": 1.632504467165119e-07,
+      "loss": 0.4465,
+      "step": 10458
+    },
+    {
+      "epoch": 2.9272320179121185,
+      "grad_norm": 0.24969608494229537,
+      "learning_rate": 1.620073104432851e-07,
+      "loss": 0.4313,
+      "step": 10459
+    },
+    {
+      "epoch": 2.927511894766303,
+      "grad_norm": 0.24858008443318744,
+      "learning_rate": 1.6076891777809245e-07,
+      "loss": 0.4744,
+      "step": 10460
+    },
+    {
+      "epoch": 2.927791771620487,
+      "grad_norm": 0.25462668422505136,
+      "learning_rate": 1.5953526883881188e-07,
+      "loss": 0.4702,
+      "step": 10461
+    },
+    {
+      "epoch": 2.9280716484746714,
+      "grad_norm": 0.2523815260372512,
+      "learning_rate": 1.5830636374286056e-07,
+      "loss": 0.4711,
+      "step": 10462
+    },
+    {
+      "epoch": 2.9283515253288552,
+      "grad_norm": 0.25767637274775135,
+      "learning_rate": 1.5708220260721163e-07,
+      "loss": 0.4455,
+      "step": 10463
+    },
+    {
+      "epoch": 2.9286314021830395,
+      "grad_norm": 0.2574606985716124,
+      "learning_rate": 1.5586278554837187e-07,
+      "loss": 0.4387,
+      "step": 10464
+    },
+    {
+      "epoch": 2.928911279037224,
+      "grad_norm": 0.2585395013878928,
+      "learning_rate": 1.546481126824151e-07,
+      "loss": 0.4426,
+      "step": 10465
+    },
+    {
+      "epoch": 2.9291911558914077,
+      "grad_norm": 0.2567128032532917,
+      "learning_rate": 1.534381841249488e-07,
+      "loss": 0.4581,
+      "step": 10466
+    },
+    {
+      "epoch": 2.929471032745592,
+      "grad_norm": 0.2551522452384083,
+      "learning_rate": 1.5223299999113094e-07,
+      "loss": 0.4508,
+      "step": 10467
+    },
+    {
+      "epoch": 2.9297509095997762,
+      "grad_norm": 0.26152786331414646,
+      "learning_rate": 1.5103256039568635e-07,
+      "loss": 0.4462,
+      "step": 10468
+    },
+    {
+      "epoch": 2.93003078645396,
+      "grad_norm": 0.2531210010499005,
+      "learning_rate": 1.4983686545285701e-07,
+      "loss": 0.4444,
+      "step": 10469
+    },
+    {
+      "epoch": 2.9303106633081444,
+      "grad_norm": 0.24830215591602056,
+      "learning_rate": 1.4864591527646298e-07,
+      "loss": 0.4222,
+      "step": 10470
+    },
+    {
+      "epoch": 2.9305905401623287,
+      "grad_norm": 0.2636931535798214,
+      "learning_rate": 1.4745970997985248e-07,
+      "loss": 0.456,
+      "step": 10471
+    },
+    {
+      "epoch": 2.9308704170165125,
+      "grad_norm": 0.2524364263885123,
+      "learning_rate": 1.4627824967592963e-07,
+      "loss": 0.4496,
+      "step": 10472
+    },
+    {
+      "epoch": 2.931150293870697,
+      "grad_norm": 0.2616100064256884,
+      "learning_rate": 1.451015344771489e-07,
+      "loss": 0.4612,
+      "step": 10473
+    },
+    {
+      "epoch": 2.931430170724881,
+      "grad_norm": 0.25763637866034367,
+      "learning_rate": 1.439295644955041e-07,
+      "loss": 0.448,
+      "step": 10474
+    },
+    {
+      "epoch": 2.931710047579065,
+      "grad_norm": 0.2584985683151651,
+      "learning_rate": 1.4276233984255593e-07,
+      "loss": 0.4615,
+      "step": 10475
+    },
+    {
+      "epoch": 2.9319899244332492,
+      "grad_norm": 0.2511241667910256,
+      "learning_rate": 1.415998606293878e-07,
+      "loss": 0.4303,
+      "step": 10476
+    },
+    {
+      "epoch": 2.9322698012874335,
+      "grad_norm": 0.2545468157726461,
+      "learning_rate": 1.404421269666556e-07,
+      "loss": 0.4323,
+      "step": 10477
+    },
+    {
+      "epoch": 2.932549678141618,
+      "grad_norm": 0.24978718053471582,
+      "learning_rate": 1.3928913896454897e-07,
+      "loss": 0.4596,
+      "step": 10478
+    },
+    {
+      "epoch": 2.932829554995802,
+      "grad_norm": 0.24891041253653481,
+      "learning_rate": 1.381408967328135e-07,
+      "loss": 0.4408,
+      "step": 10479
+    },
+    {
+      "epoch": 2.933109431849986,
+      "grad_norm": 0.25436583588653244,
+      "learning_rate": 1.3699740038073394e-07,
+      "loss": 0.4401,
+      "step": 10480
+    },
+    {
+      "epoch": 2.9333893087041703,
+      "grad_norm": 0.25555326435865844,
+      "learning_rate": 1.3585865001715104e-07,
+      "loss": 0.4463,
+      "step": 10481
+    },
+    {
+      "epoch": 2.9336691855583545,
+      "grad_norm": 0.25808759275433096,
+      "learning_rate": 1.347246457504503e-07,
+      "loss": 0.4591,
+      "step": 10482
+    },
+    {
+      "epoch": 2.9339490624125384,
+      "grad_norm": 0.26026257449568424,
+      "learning_rate": 1.3359538768856762e-07,
+      "loss": 0.4479,
+      "step": 10483
+    },
+    {
+      "epoch": 2.9342289392667227,
+      "grad_norm": 0.262707843378836,
+      "learning_rate": 1.3247087593898922e-07,
+      "loss": 0.451,
+      "step": 10484
+    },
+    {
+      "epoch": 2.934508816120907,
+      "grad_norm": 0.2564618605525408,
+      "learning_rate": 1.3135111060874617e-07,
+      "loss": 0.4358,
+      "step": 10485
+    },
+    {
+      "epoch": 2.934788692975091,
+      "grad_norm": 0.2568705199636224,
+      "learning_rate": 1.3023609180441431e-07,
+      "loss": 0.4473,
+      "step": 10486
+    },
+    {
+      "epoch": 2.935068569829275,
+      "grad_norm": 0.25888006691203863,
+      "learning_rate": 1.2912581963212543e-07,
+      "loss": 0.4572,
+      "step": 10487
+    },
+    {
+      "epoch": 2.9353484466834594,
+      "grad_norm": 0.26257361864420226,
+      "learning_rate": 1.2802029419755613e-07,
+      "loss": 0.4479,
+      "step": 10488
+    },
+    {
+      "epoch": 2.9356283235376432,
+      "grad_norm": 0.25256274125923245,
+      "learning_rate": 1.269195156059333e-07,
+      "loss": 0.4529,
+      "step": 10489
+    },
+    {
+      "epoch": 2.9359082003918275,
+      "grad_norm": 0.26555130805791666,
+      "learning_rate": 1.2582348396202316e-07,
+      "loss": 0.4576,
+      "step": 10490
+    },
+    {
+      "epoch": 2.936188077246012,
+      "grad_norm": 0.2657930704720208,
+      "learning_rate": 1.247321993701478e-07,
+      "loss": 0.459,
+      "step": 10491
+    },
+    {
+      "epoch": 2.9364679541001957,
+      "grad_norm": 0.25168542573681696,
+      "learning_rate": 1.236456619341797e-07,
+      "loss": 0.4326,
+      "step": 10492
+    },
+    {
+      "epoch": 2.93674783095438,
+      "grad_norm": 0.2560893374026327,
+      "learning_rate": 1.225638717575306e-07,
+      "loss": 0.4643,
+      "step": 10493
+    },
+    {
+      "epoch": 2.9370277078085643,
+      "grad_norm": 0.24755384682714268,
+      "learning_rate": 1.214868289431792e-07,
+      "loss": 0.4479,
+      "step": 10494
+    },
+    {
+      "epoch": 2.9373075846627485,
+      "grad_norm": 0.26603348984649644,
+      "learning_rate": 1.2041453359362132e-07,
+      "loss": 0.4439,
+      "step": 10495
+    },
+    {
+      "epoch": 2.9375874615169324,
+      "grad_norm": 0.2509154217666826,
+      "learning_rate": 1.1934698581093086e-07,
+      "loss": 0.4408,
+      "step": 10496
+    },
+    {
+      "epoch": 2.9378673383711167,
+      "grad_norm": 0.25680508562855797,
+      "learning_rate": 1.1828418569670985e-07,
+      "loss": 0.4351,
+      "step": 10497
+    },
+    {
+      "epoch": 2.938147215225301,
+      "grad_norm": 0.2585395777558349,
+      "learning_rate": 1.1722613335212185e-07,
+      "loss": 0.4556,
+      "step": 10498
+    },
+    {
+      "epoch": 2.9384270920794853,
+      "grad_norm": 0.24513710160148572,
+      "learning_rate": 1.1617282887787518e-07,
+      "loss": 0.4251,
+      "step": 10499
+    },
+    {
+      "epoch": 2.938706968933669,
+      "grad_norm": 0.2549706151869714,
+      "learning_rate": 1.1512427237421741e-07,
+      "loss": 0.4419,
+      "step": 10500
+    },
+    {
+      "epoch": 2.9389868457878534,
+      "grad_norm": 0.2394454558323718,
+      "learning_rate": 1.1408046394095206e-07,
+      "loss": 0.4508,
+      "step": 10501
+    },
+    {
+      "epoch": 2.9392667226420377,
+      "grad_norm": 0.24841273548598636,
+      "learning_rate": 1.1304140367742744e-07,
+      "loss": 0.4425,
+      "step": 10502
+    },
+    {
+      "epoch": 2.9395465994962215,
+      "grad_norm": 0.26824065946500614,
+      "learning_rate": 1.1200709168254774e-07,
+      "loss": 0.4657,
+      "step": 10503
+    },
+    {
+      "epoch": 2.939826476350406,
+      "grad_norm": 0.25734686141639923,
+      "learning_rate": 1.1097752805475647e-07,
+      "loss": 0.4425,
+      "step": 10504
+    },
+    {
+      "epoch": 2.94010635320459,
+      "grad_norm": 0.25906967232550493,
+      "learning_rate": 1.0995271289204745e-07,
+      "loss": 0.4325,
+      "step": 10505
+    },
+    {
+      "epoch": 2.940386230058774,
+      "grad_norm": 0.25985462367295037,
+      "learning_rate": 1.0893264629197042e-07,
+      "loss": 0.4374,
+      "step": 10506
+    },
+    {
+      "epoch": 2.9406661069129583,
+      "grad_norm": 0.2424408673662371,
+      "learning_rate": 1.0791732835160328e-07,
+      "loss": 0.4225,
+      "step": 10507
+    },
+    {
+      "epoch": 2.9409459837671426,
+      "grad_norm": 0.2572039680041926,
+      "learning_rate": 1.0690675916759097e-07,
+      "loss": 0.4438,
+      "step": 10508
+    },
+    {
+      "epoch": 2.9412258606213264,
+      "grad_norm": 0.25194635813849986,
+      "learning_rate": 1.0590093883611762e-07,
+      "loss": 0.4151,
+      "step": 10509
+    },
+    {
+      "epoch": 2.9415057374755107,
+      "grad_norm": 0.26942250750658214,
+      "learning_rate": 1.0489986745292335e-07,
+      "loss": 0.4737,
+      "step": 10510
+    },
+    {
+      "epoch": 2.941785614329695,
+      "grad_norm": 0.249986867792104,
+      "learning_rate": 1.0390354511328749e-07,
+      "loss": 0.454,
+      "step": 10511
+    },
+    {
+      "epoch": 2.942065491183879,
+      "grad_norm": 0.24610171411008774,
+      "learning_rate": 1.0291197191203971e-07,
+      "loss": 0.4396,
+      "step": 10512
+    },
+    {
+      "epoch": 2.942345368038063,
+      "grad_norm": 0.255544816937748,
+      "learning_rate": 1.0192514794356012e-07,
+      "loss": 0.4401,
+      "step": 10513
+    },
+    {
+      "epoch": 2.9426252448922474,
+      "grad_norm": 0.2455925774087127,
+      "learning_rate": 1.0094307330177355e-07,
+      "loss": 0.4536,
+      "step": 10514
+    },
+    {
+      "epoch": 2.9429051217464317,
+      "grad_norm": 0.2583339543631677,
+      "learning_rate": 9.99657480801497e-08,
+      "loss": 0.4508,
+      "step": 10515
+    },
+    {
+      "epoch": 2.943184998600616,
+      "grad_norm": 0.25668429968223977,
+      "learning_rate": 9.899317237172523e-08,
+      "loss": 0.419,
+      "step": 10516
+    },
+    {
+      "epoch": 2.9434648754548,
+      "grad_norm": 0.24562189192693365,
+      "learning_rate": 9.80253462690539e-08,
+      "loss": 0.4555,
+      "step": 10517
+    },
+    {
+      "epoch": 2.943744752308984,
+      "grad_norm": 0.2622676135147595,
+      "learning_rate": 9.706226986426203e-08,
+      "loss": 0.4386,
+      "step": 10518
+    },
+    {
+      "epoch": 2.9440246291631684,
+      "grad_norm": 0.26528886042279054,
+      "learning_rate": 9.610394324902073e-08,
+      "loss": 0.4399,
+      "step": 10519
+    },
+    {
+      "epoch": 2.9443045060173523,
+      "grad_norm": 0.2687492946457659,
+      "learning_rate": 9.515036651453479e-08,
+      "loss": 0.4565,
+      "step": 10520
+    },
+    {
+      "epoch": 2.9445843828715366,
+      "grad_norm": 0.25029425954088713,
+      "learning_rate": 9.420153975157053e-08,
+      "loss": 0.45,
+      "step": 10521
+    },
+    {
+      "epoch": 2.944864259725721,
+      "grad_norm": 0.2530798242372889,
+      "learning_rate": 9.325746305043348e-08,
+      "loss": 0.4337,
+      "step": 10522
+    },
+    {
+      "epoch": 2.9451441365799047,
+      "grad_norm": 0.25875219638515645,
+      "learning_rate": 9.231813650099064e-08,
+      "loss": 0.4519,
+      "step": 10523
+    },
+    {
+      "epoch": 2.945424013434089,
+      "grad_norm": 0.25682815207388776,
+      "learning_rate": 9.138356019264271e-08,
+      "loss": 0.4389,
+      "step": 10524
+    },
+    {
+      "epoch": 2.9457038902882733,
+      "grad_norm": 0.25104282666120115,
+      "learning_rate": 9.045373421433523e-08,
+      "loss": 0.425,
+      "step": 10525
+    },
+    {
+      "epoch": 2.945983767142457,
+      "grad_norm": 0.26342494863030874,
+      "learning_rate": 8.952865865458626e-08,
+      "loss": 0.4703,
+      "step": 10526
+    },
+    {
+      "epoch": 2.9462636439966414,
+      "grad_norm": 0.26385994849672095,
+      "learning_rate": 8.860833360143095e-08,
+      "loss": 0.4356,
+      "step": 10527
+    },
+    {
+      "epoch": 2.9465435208508257,
+      "grad_norm": 0.251010870200779,
+      "learning_rate": 8.769275914247143e-08,
+      "loss": 0.4662,
+      "step": 10528
+    },
+    {
+      "epoch": 2.9468233977050096,
+      "grad_norm": 0.2471674470410353,
+      "learning_rate": 8.678193536484914e-08,
+      "loss": 0.4394,
+      "step": 10529
+    },
+    {
+      "epoch": 2.947103274559194,
+      "grad_norm": 0.24808625130165926,
+      "learning_rate": 8.58758623552669e-08,
+      "loss": 0.4527,
+      "step": 10530
+    },
+    {
+      "epoch": 2.947383151413378,
+      "grad_norm": 0.2553979940589703,
+      "learning_rate": 8.497454019995022e-08,
+      "loss": 0.4581,
+      "step": 10531
+    },
+    {
+      "epoch": 2.9476630282675624,
+      "grad_norm": 0.27036146065247973,
+      "learning_rate": 8.407796898470266e-08,
+      "loss": 0.4543,
+      "step": 10532
+    },
+    {
+      "epoch": 2.9479429051217463,
+      "grad_norm": 0.26083548862095113,
+      "learning_rate": 8.318614879485043e-08,
+      "loss": 0.4528,
+      "step": 10533
+    },
+    {
+      "epoch": 2.9482227819759306,
+      "grad_norm": 0.25369304134129755,
+      "learning_rate": 8.229907971528116e-08,
+      "loss": 0.4301,
+      "step": 10534
+    },
+    {
+      "epoch": 2.948502658830115,
+      "grad_norm": 0.2587089753348896,
+      "learning_rate": 8.14167618304218e-08,
+      "loss": 0.4547,
+      "step": 10535
+    },
+    {
+      "epoch": 2.948782535684299,
+      "grad_norm": 0.2561463390198463,
+      "learning_rate": 8.053919522425513e-08,
+      "loss": 0.4417,
+      "step": 10536
+    },
+    {
+      "epoch": 2.949062412538483,
+      "grad_norm": 0.2804159977991239,
+      "learning_rate": 7.966637998031434e-08,
+      "loss": 0.4435,
+      "step": 10537
+    },
+    {
+      "epoch": 2.9493422893926673,
+      "grad_norm": 0.2636994221235027,
+      "learning_rate": 7.879831618166633e-08,
+      "loss": 0.4542,
+      "step": 10538
+    },
+    {
+      "epoch": 2.9496221662468516,
+      "grad_norm": 0.25934639410112037,
+      "learning_rate": 7.793500391093944e-08,
+      "loss": 0.445,
+      "step": 10539
+    },
+    {
+      "epoch": 2.9499020431010354,
+      "grad_norm": 0.28139162564843395,
+      "learning_rate": 7.707644325029572e-08,
+      "loss": 0.4317,
+      "step": 10540
+    },
+    {
+      "epoch": 2.9501819199552197,
+      "grad_norm": 0.26420461175765253,
+      "learning_rate": 7.622263428146426e-08,
+      "loss": 0.4356,
+      "step": 10541
+    },
+    {
+      "epoch": 2.950461796809404,
+      "grad_norm": 0.25441472194653697,
+      "learning_rate": 7.537357708570225e-08,
+      "loss": 0.446,
+      "step": 10542
+    },
+    {
+      "epoch": 2.950741673663588,
+      "grad_norm": 0.2671800826870824,
+      "learning_rate": 7.452927174383396e-08,
+      "loss": 0.4549,
+      "step": 10543
+    },
+    {
+      "epoch": 2.951021550517772,
+      "grad_norm": 0.25564475592140923,
+      "learning_rate": 7.368971833620619e-08,
+      "loss": 0.4551,
+      "step": 10544
+    },
+    {
+      "epoch": 2.9513014273719564,
+      "grad_norm": 0.25828170287220853,
+      "learning_rate": 7.285491694273838e-08,
+      "loss": 0.445,
+      "step": 10545
+    },
+    {
+      "epoch": 2.9515813042261403,
+      "grad_norm": 0.24080567738555247,
+      "learning_rate": 7.202486764288918e-08,
+      "loss": 0.4379,
+      "step": 10546
+    },
+    {
+      "epoch": 2.9518611810803246,
+      "grad_norm": 0.25363021744294506,
+      "learning_rate": 7.119957051565651e-08,
+      "loss": 0.4627,
+      "step": 10547
+    },
+    {
+      "epoch": 2.952141057934509,
+      "grad_norm": 0.25326867966327166,
+      "learning_rate": 7.03790256395942e-08,
+      "loss": 0.4262,
+      "step": 10548
+    },
+    {
+      "epoch": 2.9524209347886927,
+      "grad_norm": 0.26917960408980013,
+      "learning_rate": 6.956323309280089e-08,
+      "loss": 0.4603,
+      "step": 10549
+    },
+    {
+      "epoch": 2.952700811642877,
+      "grad_norm": 0.26027908392323174,
+      "learning_rate": 6.875219295293111e-08,
+      "loss": 0.461,
+      "step": 10550
+    },
+    {
+      "epoch": 2.9529806884970613,
+      "grad_norm": 0.2591632379456325,
+      "learning_rate": 6.794590529717315e-08,
+      "loss": 0.4536,
+      "step": 10551
+    },
+    {
+      "epoch": 2.9532605653512456,
+      "grad_norm": 0.2510652354734244,
+      "learning_rate": 6.714437020227115e-08,
+      "loss": 0.4235,
+      "step": 10552
+    },
+    {
+      "epoch": 2.95354044220543,
+      "grad_norm": 0.2526340838303442,
+      "learning_rate": 6.634758774451966e-08,
+      "loss": 0.4352,
+      "step": 10553
+    },
+    {
+      "epoch": 2.9538203190596137,
+      "grad_norm": 0.2575515578872242,
+      "learning_rate": 6.555555799974689e-08,
+      "loss": 0.4536,
+      "step": 10554
+    },
+    {
+      "epoch": 2.954100195913798,
+      "grad_norm": 0.2603208898237062,
+      "learning_rate": 6.476828104335364e-08,
+      "loss": 0.4133,
+      "step": 10555
+    },
+    {
+      "epoch": 2.9543800727679823,
+      "grad_norm": 0.24904628086688163,
+      "learning_rate": 6.398575695026332e-08,
+      "loss": 0.431,
+      "step": 10556
+    },
+    {
+      "epoch": 2.954659949622166,
+      "grad_norm": 0.26696013383377293,
+      "learning_rate": 6.320798579495524e-08,
+      "loss": 0.4594,
+      "step": 10557
+    },
+    {
+      "epoch": 2.9549398264763505,
+      "grad_norm": 0.25246866520967143,
+      "learning_rate": 6.243496765146461e-08,
+      "loss": 0.4408,
+      "step": 10558
+    },
+    {
+      "epoch": 2.9552197033305347,
+      "grad_norm": 0.254437865474292,
+      "learning_rate": 6.166670259336594e-08,
+      "loss": 0.4652,
+      "step": 10559
+    },
+    {
+      "epoch": 2.9554995801847186,
+      "grad_norm": 0.26299613136944583,
+      "learning_rate": 6.090319069377848e-08,
+      "loss": 0.4597,
+      "step": 10560
+    },
+    {
+      "epoch": 2.955779457038903,
+      "grad_norm": 0.2615353362609834,
+      "learning_rate": 6.0144432025383e-08,
+      "loss": 0.4629,
+      "step": 10561
+    },
+    {
+      "epoch": 2.956059333893087,
+      "grad_norm": 0.2512308203604363,
+      "learning_rate": 5.939042666038841e-08,
+      "loss": 0.4444,
+      "step": 10562
+    },
+    {
+      "epoch": 2.956339210747271,
+      "grad_norm": 0.2377900486143696,
+      "learning_rate": 5.864117467057062e-08,
+      "loss": 0.4592,
+      "step": 10563
+    },
+    {
+      "epoch": 2.9566190876014553,
+      "grad_norm": 0.25875069490141756,
+      "learning_rate": 5.789667612723371e-08,
+      "loss": 0.4335,
+      "step": 10564
+    },
+    {
+      "epoch": 2.9568989644556396,
+      "grad_norm": 0.24683365444532673,
+      "learning_rate": 5.7156931101248755e-08,
+      "loss": 0.4351,
+      "step": 10565
+    },
+    {
+      "epoch": 2.9571788413098234,
+      "grad_norm": 0.26012406806987626,
+      "learning_rate": 5.642193966302056e-08,
+      "loss": 0.4473,
+      "step": 10566
+    },
+    {
+      "epoch": 2.9574587181640077,
+      "grad_norm": 0.2542922961412427,
+      "learning_rate": 5.569170188250983e-08,
+      "loss": 0.4543,
+      "step": 10567
+    },
+    {
+      "epoch": 2.957738595018192,
+      "grad_norm": 0.25715451992754895,
+      "learning_rate": 5.496621782921097e-08,
+      "loss": 0.443,
+      "step": 10568
+    },
+    {
+      "epoch": 2.9580184718723763,
+      "grad_norm": 0.2590939919758361,
+      "learning_rate": 5.4245487572190946e-08,
+      "loss": 0.4282,
+      "step": 10569
+    },
+    {
+      "epoch": 2.95829834872656,
+      "grad_norm": 0.24454462290233103,
+      "learning_rate": 5.35295111800338e-08,
+      "loss": 0.4258,
+      "step": 10570
+    },
+    {
+      "epoch": 2.9585782255807445,
+      "grad_norm": 0.24034675492651944,
+      "learning_rate": 5.281828872089611e-08,
+      "loss": 0.4318,
+      "step": 10571
+    },
+    {
+      "epoch": 2.9588581024349287,
+      "grad_norm": 0.2668304845642728,
+      "learning_rate": 5.2111820262473745e-08,
+      "loss": 0.4531,
+      "step": 10572
+    },
+    {
+      "epoch": 2.959137979289113,
+      "grad_norm": 0.24610859132094123,
+      "learning_rate": 5.141010587200179e-08,
+      "loss": 0.4518,
+      "step": 10573
+    },
+    {
+      "epoch": 2.959417856143297,
+      "grad_norm": 0.2525784737926144,
+      "learning_rate": 5.071314561627682e-08,
+      "loss": 0.4499,
+      "step": 10574
+    },
+    {
+      "epoch": 2.959697732997481,
+      "grad_norm": 0.2589959315475041,
+      "learning_rate": 5.002093956162912e-08,
+      "loss": 0.4473,
+      "step": 10575
+    },
+    {
+      "epoch": 2.9599776098516655,
+      "grad_norm": 0.24563499364177488,
+      "learning_rate": 4.933348777395042e-08,
+      "loss": 0.4311,
+      "step": 10576
+    },
+    {
+      "epoch": 2.9602574867058493,
+      "grad_norm": 0.24229642325126785,
+      "learning_rate": 4.865079031866615e-08,
+      "loss": 0.4356,
+      "step": 10577
+    },
+    {
+      "epoch": 2.9605373635600336,
+      "grad_norm": 0.2569134940403133,
+      "learning_rate": 4.797284726075768e-08,
+      "loss": 0.4439,
+      "step": 10578
+    },
+    {
+      "epoch": 2.960817240414218,
+      "grad_norm": 0.2544678366133651,
+      "learning_rate": 4.729965866475117e-08,
+      "loss": 0.4644,
+      "step": 10579
+    },
+    {
+      "epoch": 2.9610971172684017,
+      "grad_norm": 0.25061094843996773,
+      "learning_rate": 4.663122459472869e-08,
+      "loss": 0.4305,
+      "step": 10580
+    },
+    {
+      "epoch": 2.961376994122586,
+      "grad_norm": 0.25322228356666415,
+      "learning_rate": 4.5967545114306007e-08,
+      "loss": 0.4518,
+      "step": 10581
+    },
+    {
+      "epoch": 2.9616568709767703,
+      "grad_norm": 0.24853741746919858,
+      "learning_rate": 4.530862028664928e-08,
+      "loss": 0.4444,
+      "step": 10582
+    },
+    {
+      "epoch": 2.961936747830954,
+      "grad_norm": 0.25579650896961803,
+      "learning_rate": 4.465445017448056e-08,
+      "loss": 0.4376,
+      "step": 10583
+    },
+    {
+      "epoch": 2.9622166246851385,
+      "grad_norm": 0.25032094678732375,
+      "learning_rate": 4.400503484006113e-08,
+      "loss": 0.4392,
+      "step": 10584
+    },
+    {
+      "epoch": 2.9624965015393228,
+      "grad_norm": 0.25834259594052383,
+      "learning_rate": 4.336037434520823e-08,
+      "loss": 0.456,
+      "step": 10585
+    },
+    {
+      "epoch": 2.9627763783935066,
+      "grad_norm": 0.247699967081848,
+      "learning_rate": 4.272046875127278e-08,
+      "loss": 0.4352,
+      "step": 10586
+    },
+    {
+      "epoch": 2.963056255247691,
+      "grad_norm": 0.255500336642797,
+      "learning_rate": 4.208531811916716e-08,
+      "loss": 0.4507,
+      "step": 10587
+    },
+    {
+      "epoch": 2.963336132101875,
+      "grad_norm": 0.2481995662102386,
+      "learning_rate": 4.1454922509337466e-08,
+      "loss": 0.4505,
+      "step": 10588
+    },
+    {
+      "epoch": 2.9636160089560595,
+      "grad_norm": 0.25844158185485583,
+      "learning_rate": 4.082928198179681e-08,
+      "loss": 0.4587,
+      "step": 10589
+    },
+    {
+      "epoch": 2.9638958858102438,
+      "grad_norm": 0.25711496564190917,
+      "learning_rate": 4.020839659609199e-08,
+      "loss": 0.4555,
+      "step": 10590
+    },
+    {
+      "epoch": 2.9641757626644276,
+      "grad_norm": 0.2588625423650436,
+      "learning_rate": 3.959226641130909e-08,
+      "loss": 0.4533,
+      "step": 10591
+    },
+    {
+      "epoch": 2.964455639518612,
+      "grad_norm": 0.24528164675911024,
+      "learning_rate": 3.8980891486101176e-08,
+      "loss": 0.4379,
+      "step": 10592
+    },
+    {
+      "epoch": 2.964735516372796,
+      "grad_norm": 0.25861177214476944,
+      "learning_rate": 3.837427187866061e-08,
+      "loss": 0.4488,
+      "step": 10593
+    },
+    {
+      "epoch": 2.96501539322698,
+      "grad_norm": 0.2520859346888829,
+      "learning_rate": 3.777240764671342e-08,
+      "loss": 0.4513,
+      "step": 10594
+    },
+    {
+      "epoch": 2.9652952700811643,
+      "grad_norm": 0.2604468848180648,
+      "learning_rate": 3.7175298847558216e-08,
+      "loss": 0.4585,
+      "step": 10595
+    },
+    {
+      "epoch": 2.9655751469353486,
+      "grad_norm": 0.2582001394915078,
+      "learning_rate": 3.6582945538027324e-08,
+      "loss": 0.4321,
+      "step": 10596
+    },
+    {
+      "epoch": 2.9658550237895325,
+      "grad_norm": 0.25229843018884607,
+      "learning_rate": 3.599534777449232e-08,
+      "loss": 0.463,
+      "step": 10597
+    },
+    {
+      "epoch": 2.9661349006437168,
+      "grad_norm": 0.25236312467740846,
+      "learning_rate": 3.5412505612886225e-08,
+      "loss": 0.4519,
+      "step": 10598
+    },
+    {
+      "epoch": 2.966414777497901,
+      "grad_norm": 0.25724033603089713,
+      "learning_rate": 3.4834419108681346e-08,
+      "loss": 0.4533,
+      "step": 10599
+    },
+    {
+      "epoch": 2.966694654352085,
+      "grad_norm": 0.2581127353937177,
+      "learning_rate": 3.426108831691144e-08,
+      "loss": 0.4595,
+      "step": 10600
+    },
+    {
+      "epoch": 2.966974531206269,
+      "grad_norm": 0.24611693632209142,
+      "learning_rate": 3.369251329213285e-08,
+      "loss": 0.415,
+      "step": 10601
+    },
+    {
+      "epoch": 2.9672544080604535,
+      "grad_norm": 0.25310950920344844,
+      "learning_rate": 3.312869408846897e-08,
+      "loss": 0.429,
+      "step": 10602
+    },
+    {
+      "epoch": 2.9675342849146373,
+      "grad_norm": 0.24968979714091358,
+      "learning_rate": 3.2569630759582415e-08,
+      "loss": 0.4566,
+      "step": 10603
+    },
+    {
+      "epoch": 2.9678141617688216,
+      "grad_norm": 0.2609776394190351,
+      "learning_rate": 3.201532335868618e-08,
+      "loss": 0.4458,
+      "step": 10604
+    },
+    {
+      "epoch": 2.968094038623006,
+      "grad_norm": 0.25969592415796383,
+      "learning_rate": 3.146577193854361e-08,
+      "loss": 0.4498,
+      "step": 10605
+    },
+    {
+      "epoch": 2.9683739154771898,
+      "grad_norm": 0.24716778897711522,
+      "learning_rate": 3.092097655145176e-08,
+      "loss": 0.4332,
+      "step": 10606
+    },
+    {
+      "epoch": 2.968653792331374,
+      "grad_norm": 0.27328902605297223,
+      "learning_rate": 3.038093724927471e-08,
+      "loss": 0.4511,
+      "step": 10607
+    },
+    {
+      "epoch": 2.9689336691855583,
+      "grad_norm": 0.2507518753167543,
+      "learning_rate": 2.984565408341022e-08,
+      "loss": 0.4494,
+      "step": 10608
+    },
+    {
+      "epoch": 2.9692135460397426,
+      "grad_norm": 0.2609750291364599,
+      "learning_rate": 2.9315127104800887e-08,
+      "loss": 0.4491,
+      "step": 10609
+    },
+    {
+      "epoch": 2.969493422893927,
+      "grad_norm": 0.25655339861321874,
+      "learning_rate": 2.878935636395075e-08,
+      "loss": 0.4316,
+      "step": 10610
+    },
+    {
+      "epoch": 2.9697732997481108,
+      "grad_norm": 0.2530639883196548,
+      "learning_rate": 2.8268341910903108e-08,
+      "loss": 0.4525,
+      "step": 10611
+    },
+    {
+      "epoch": 2.970053176602295,
+      "grad_norm": 0.23225610378688546,
+      "learning_rate": 2.7752083795240525e-08,
+      "loss": 0.4282,
+      "step": 10612
+    },
+    {
+      "epoch": 2.9703330534564794,
+      "grad_norm": 0.2543375083380285,
+      "learning_rate": 2.7240582066107022e-08,
+      "loss": 0.4498,
+      "step": 10613
+    },
+    {
+      "epoch": 2.970612930310663,
+      "grad_norm": 0.2527960226970521,
+      "learning_rate": 2.6733836772185884e-08,
+      "loss": 0.462,
+      "step": 10614
+    },
+    {
+      "epoch": 2.9708928071648475,
+      "grad_norm": 0.2554175886404961,
+      "learning_rate": 2.623184796170519e-08,
+      "loss": 0.4319,
+      "step": 10615
+    },
+    {
+      "epoch": 2.971172684019032,
+      "grad_norm": 0.25718644400633023,
+      "learning_rate": 2.57346156824545e-08,
+      "loss": 0.4517,
+      "step": 10616
+    },
+    {
+      "epoch": 2.9714525608732156,
+      "grad_norm": 0.302200558832404,
+      "learning_rate": 2.5242139981751513e-08,
+      "loss": 0.4469,
+      "step": 10617
+    },
+    {
+      "epoch": 2.9717324377274,
+      "grad_norm": 0.2549185842864372,
+      "learning_rate": 2.4754420906475396e-08,
+      "loss": 0.4441,
+      "step": 10618
+    },
+    {
+      "epoch": 2.972012314581584,
+      "grad_norm": 0.2552989797038095,
+      "learning_rate": 2.4271458503044576e-08,
+      "loss": 0.4555,
+      "step": 10619
+    },
+    {
+      "epoch": 2.972292191435768,
+      "grad_norm": 0.25867881848070834,
+      "learning_rate": 2.3793252817427836e-08,
+      "loss": 0.4586,
+      "step": 10620
+    },
+    {
+      "epoch": 2.9725720682899524,
+      "grad_norm": 0.25767245445537734,
+      "learning_rate": 2.3319803895144322e-08,
+      "loss": 0.437,
+      "step": 10621
+    },
+    {
+      "epoch": 2.9728519451441366,
+      "grad_norm": 0.261765852972024,
+      "learning_rate": 2.2851111781257983e-08,
+      "loss": 0.4512,
+      "step": 10622
+    },
+    {
+      "epoch": 2.9731318219983205,
+      "grad_norm": 0.25522924545223363,
+      "learning_rate": 2.238717652037203e-08,
+      "loss": 0.449,
+      "step": 10623
+    },
+    {
+      "epoch": 2.973411698852505,
+      "grad_norm": 0.24927821271923611,
+      "learning_rate": 2.1927998156651142e-08,
+      "loss": 0.4506,
+      "step": 10624
+    },
+    {
+      "epoch": 2.973691575706689,
+      "grad_norm": 0.25831566762208635,
+      "learning_rate": 2.1473576733793686e-08,
+      "loss": 0.4281,
+      "step": 10625
+    },
+    {
+      "epoch": 2.9739714525608734,
+      "grad_norm": 0.25375287107557304,
+      "learning_rate": 2.1023912295059512e-08,
+      "loss": 0.4466,
+      "step": 10626
+    },
+    {
+      "epoch": 2.974251329415057,
+      "grad_norm": 0.25946073380754525,
+      "learning_rate": 2.0579004883236608e-08,
+      "loss": 0.4414,
+      "step": 10627
+    },
+    {
+      "epoch": 2.9745312062692415,
+      "grad_norm": 0.25171033719930647,
+      "learning_rate": 2.0138854540685538e-08,
+      "loss": 0.435,
+      "step": 10628
+    },
+    {
+      "epoch": 2.974811083123426,
+      "grad_norm": 0.26115332869155145,
+      "learning_rate": 1.9703461309295013e-08,
+      "loss": 0.4663,
+      "step": 10629
+    },
+    {
+      "epoch": 2.97509095997761,
+      "grad_norm": 0.25790911409843703,
+      "learning_rate": 1.927282523049856e-08,
+      "loss": 0.4663,
+      "step": 10630
+    },
+    {
+      "epoch": 2.975370836831794,
+      "grad_norm": 0.2607999674754798,
+      "learning_rate": 1.884694634529116e-08,
+      "loss": 0.4595,
+      "step": 10631
+    },
+    {
+      "epoch": 2.975650713685978,
+      "grad_norm": 0.2534597997717571,
+      "learning_rate": 1.842582469420706e-08,
+      "loss": 0.4221,
+      "step": 10632
+    },
+    {
+      "epoch": 2.9759305905401625,
+      "grad_norm": 0.2503832993380337,
+      "learning_rate": 1.8009460317330862e-08,
+      "loss": 0.4242,
+      "step": 10633
+    },
+    {
+      "epoch": 2.9762104673943464,
+      "grad_norm": 0.25792042031621715,
+      "learning_rate": 1.7597853254291972e-08,
+      "loss": 0.4446,
+      "step": 10634
+    },
+    {
+      "epoch": 2.9764903442485307,
+      "grad_norm": 0.24700530197851792,
+      "learning_rate": 1.7191003544259064e-08,
+      "loss": 0.4664,
+      "step": 10635
+    },
+    {
+      "epoch": 2.976770221102715,
+      "grad_norm": 0.26387028208771174,
+      "learning_rate": 1.6788911225967817e-08,
+      "loss": 0.438,
+      "step": 10636
+    },
+    {
+      "epoch": 2.977050097956899,
+      "grad_norm": 0.2465718699073755,
+      "learning_rate": 1.639157633768762e-08,
+      "loss": 0.4598,
+      "step": 10637
+    },
+    {
+      "epoch": 2.977329974811083,
+      "grad_norm": 0.2632178783866457,
+      "learning_rate": 1.5998998917227116e-08,
+      "loss": 0.4506,
+      "step": 10638
+    },
+    {
+      "epoch": 2.9776098516652674,
+      "grad_norm": 0.25198103145232476,
+      "learning_rate": 1.5611179001967513e-08,
+      "loss": 0.4497,
+      "step": 10639
+    },
+    {
+      "epoch": 2.977889728519451,
+      "grad_norm": 0.2615169372245782,
+      "learning_rate": 1.5228116628807078e-08,
+      "loss": 0.4634,
+      "step": 10640
+    },
+    {
+      "epoch": 2.9781696053736355,
+      "grad_norm": 0.24759420280801528,
+      "learning_rate": 1.484981183421108e-08,
+      "loss": 0.4492,
+      "step": 10641
+    },
+    {
+      "epoch": 2.97844948222782,
+      "grad_norm": 0.2575222446333656,
+      "learning_rate": 1.447626465419516e-08,
+      "loss": 0.4414,
+      "step": 10642
+    },
+    {
+      "epoch": 2.9787293590820036,
+      "grad_norm": 0.2578216050855136,
+      "learning_rate": 1.4107475124297553e-08,
+      "loss": 0.4615,
+      "step": 10643
+    },
+    {
+      "epoch": 2.979009235936188,
+      "grad_norm": 0.24764862867426704,
+      "learning_rate": 1.3743443279634617e-08,
+      "loss": 0.4396,
+      "step": 10644
+    },
+    {
+      "epoch": 2.9792891127903722,
+      "grad_norm": 0.2790608325780094,
+      "learning_rate": 1.3384169154850856e-08,
+      "loss": 0.4493,
+      "step": 10645
+    },
+    {
+      "epoch": 2.9795689896445565,
+      "grad_norm": 0.24885669381579884,
+      "learning_rate": 1.3029652784135592e-08,
+      "loss": 0.4471,
+      "step": 10646
+    },
+    {
+      "epoch": 2.979848866498741,
+      "grad_norm": 0.2516985946364248,
+      "learning_rate": 1.2679894201239606e-08,
+      "loss": 0.4282,
+      "step": 10647
+    },
+    {
+      "epoch": 2.9801287433529247,
+      "grad_norm": 0.2546918838471997,
+      "learning_rate": 1.2334893439447382e-08,
+      "loss": 0.4247,
+      "step": 10648
+    },
+    {
+      "epoch": 2.980408620207109,
+      "grad_norm": 0.26154533642566097,
+      "learning_rate": 1.1994650531604868e-08,
+      "loss": 0.4514,
+      "step": 10649
+    },
+    {
+      "epoch": 2.9806884970612932,
+      "grad_norm": 0.2541071458160839,
+      "learning_rate": 1.1659165510086167e-08,
+      "loss": 0.4453,
+      "step": 10650
+    },
+    {
+      "epoch": 2.980968373915477,
+      "grad_norm": 0.2557609039926244,
+      "learning_rate": 1.1328438406826847e-08,
+      "loss": 0.4518,
+      "step": 10651
+    },
+    {
+      "epoch": 2.9812482507696614,
+      "grad_norm": 0.2603397966746705,
+      "learning_rate": 1.100246925331283e-08,
+      "loss": 0.4518,
+      "step": 10652
+    },
+    {
+      "epoch": 2.9815281276238457,
+      "grad_norm": 0.25378766274204506,
+      "learning_rate": 1.0681258080558198e-08,
+      "loss": 0.4632,
+      "step": 10653
+    },
+    {
+      "epoch": 2.9818080044780295,
+      "grad_norm": 0.2577461103752013,
+      "learning_rate": 1.0364804919144044e-08,
+      "loss": 0.4321,
+      "step": 10654
+    },
+    {
+      "epoch": 2.982087881332214,
+      "grad_norm": 0.2548663096401936,
+      "learning_rate": 1.0053109799190719e-08,
+      "loss": 0.4583,
+      "step": 10655
+    },
+    {
+      "epoch": 2.982367758186398,
+      "grad_norm": 0.2678503092268285,
+      "learning_rate": 9.74617275035783e-09,
+      "loss": 0.4446,
+      "step": 10656
+    },
+    {
+      "epoch": 2.982647635040582,
+      "grad_norm": 0.25644673755161296,
+      "learning_rate": 9.443993801866447e-09,
+      "loss": 0.4487,
+      "step": 10657
+    },
+    {
+      "epoch": 2.9829275118947662,
+      "grad_norm": 0.26481506867740406,
+      "learning_rate": 9.146572982476897e-09,
+      "loss": 0.4555,
+      "step": 10658
+    },
+    {
+      "epoch": 2.9832073887489505,
+      "grad_norm": 0.26361043389181155,
+      "learning_rate": 8.85391032049987e-09,
+      "loss": 0.4423,
+      "step": 10659
+    },
+    {
+      "epoch": 2.9834872656031344,
+      "grad_norm": 0.25615383736953834,
+      "learning_rate": 8.566005843790858e-09,
+      "loss": 0.4583,
+      "step": 10660
+    },
+    {
+      "epoch": 2.9837671424573187,
+      "grad_norm": 0.24828969986997024,
+      "learning_rate": 8.282859579744617e-09,
+      "loss": 0.4262,
+      "step": 10661
+    },
+    {
+      "epoch": 2.984047019311503,
+      "grad_norm": 0.25552999391652786,
+      "learning_rate": 8.004471555322913e-09,
+      "loss": 0.4646,
+      "step": 10662
+    },
+    {
+      "epoch": 2.9843268961656872,
+      "grad_norm": 0.2636652872115676,
+      "learning_rate": 7.730841797010113e-09,
+      "loss": 0.4618,
+      "step": 10663
+    },
+    {
+      "epoch": 2.984606773019871,
+      "grad_norm": 0.2583926307618832,
+      "learning_rate": 7.461970330863156e-09,
+      "loss": 0.4385,
+      "step": 10664
+    },
+    {
+      "epoch": 2.9848866498740554,
+      "grad_norm": 0.2518994518625086,
+      "learning_rate": 7.197857182467127e-09,
+      "loss": 0.438,
+      "step": 10665
+    },
+    {
+      "epoch": 2.9851665267282397,
+      "grad_norm": 0.2549280585315688,
+      "learning_rate": 6.938502376963029e-09,
+      "loss": 0.431,
+      "step": 10666
+    },
+    {
+      "epoch": 2.985446403582424,
+      "grad_norm": 0.26188810220058695,
+      "learning_rate": 6.683905939031121e-09,
+      "loss": 0.4571,
+      "step": 10667
+    },
+    {
+      "epoch": 2.985726280436608,
+      "grad_norm": 0.2629607222117437,
+      "learning_rate": 6.434067892907569e-09,
+      "loss": 0.4575,
+      "step": 10668
+    },
+    {
+      "epoch": 2.986006157290792,
+      "grad_norm": 0.2497045188475996,
+      "learning_rate": 6.188988262373352e-09,
+      "loss": 0.4428,
+      "step": 10669
+    },
+    {
+      "epoch": 2.9862860341449764,
+      "grad_norm": 0.26736706499842067,
+      "learning_rate": 5.948667070754255e-09,
+      "loss": 0.4459,
+      "step": 10670
+    },
+    {
+      "epoch": 2.9865659109991602,
+      "grad_norm": 0.24828775818061782,
+      "learning_rate": 5.713104340926423e-09,
+      "loss": 0.4284,
+      "step": 10671
+    },
+    {
+      "epoch": 2.9868457878533445,
+      "grad_norm": 0.26353052160753476,
+      "learning_rate": 5.482300095305259e-09,
+      "loss": 0.4721,
+      "step": 10672
+    },
+    {
+      "epoch": 2.987125664707529,
+      "grad_norm": 0.25205314276037316,
+      "learning_rate": 5.256254355862078e-09,
+      "loss": 0.4565,
+      "step": 10673
+    },
+    {
+      "epoch": 2.9874055415617127,
+      "grad_norm": 0.2559405262948841,
+      "learning_rate": 5.034967144113001e-09,
+      "loss": 0.4575,
+      "step": 10674
+    },
+    {
+      "epoch": 2.987685418415897,
+      "grad_norm": 0.24907589377603614,
+      "learning_rate": 4.818438481118959e-09,
+      "loss": 0.4411,
+      "step": 10675
+    },
+    {
+      "epoch": 2.9879652952700813,
+      "grad_norm": 0.2549880059005624,
+      "learning_rate": 4.606668387491242e-09,
+      "loss": 0.4556,
+      "step": 10676
+    },
+    {
+      "epoch": 2.988245172124265,
+      "grad_norm": 0.2570026338796731,
+      "learning_rate": 4.399656883380398e-09,
+      "loss": 0.4777,
+      "step": 10677
+    },
+    {
+      "epoch": 2.9885250489784494,
+      "grad_norm": 0.24982827713037273,
+      "learning_rate": 4.197403988492887e-09,
+      "loss": 0.4447,
+      "step": 10678
+    },
+    {
+      "epoch": 2.9888049258326337,
+      "grad_norm": 0.2588687368315132,
+      "learning_rate": 3.999909722085527e-09,
+      "loss": 0.4514,
+      "step": 10679
+    },
+    {
+      "epoch": 2.9890848026868175,
+      "grad_norm": 0.2582412441405769,
+      "learning_rate": 3.807174102948841e-09,
+      "loss": 0.4494,
+      "step": 10680
+    },
+    {
+      "epoch": 2.989364679541002,
+      "grad_norm": 0.2508054374346863,
+      "learning_rate": 3.6191971494292652e-09,
+      "loss": 0.4488,
+      "step": 10681
+    },
+    {
+      "epoch": 2.989644556395186,
+      "grad_norm": 0.24647912747459155,
+      "learning_rate": 3.435978879418045e-09,
+      "loss": 0.4673,
+      "step": 10682
+    },
+    {
+      "epoch": 2.9899244332493704,
+      "grad_norm": 0.2600041063000193,
+      "learning_rate": 3.2575193103567826e-09,
+      "loss": 0.4533,
+      "step": 10683
+    },
+    {
+      "epoch": 2.9902043101035547,
+      "grad_norm": 0.25480134712200003,
+      "learning_rate": 3.0838184592263396e-09,
+      "loss": 0.4655,
+      "step": 10684
+    },
+    {
+      "epoch": 2.9904841869577385,
+      "grad_norm": 0.25467713191116054,
+      "learning_rate": 2.9148763425634886e-09,
+      "loss": 0.4577,
+      "step": 10685
+    },
+    {
+      "epoch": 2.990764063811923,
+      "grad_norm": 0.2425309705348517,
+      "learning_rate": 2.750692976444258e-09,
+      "loss": 0.4377,
+      "step": 10686
+    },
+    {
+      "epoch": 2.991043940666107,
+      "grad_norm": 0.26213250626211815,
+      "learning_rate": 2.5912683765061398e-09,
+      "loss": 0.4705,
+      "step": 10687
+    },
+    {
+      "epoch": 2.991323817520291,
+      "grad_norm": 0.2612357437354752,
+      "learning_rate": 2.436602557909229e-09,
+      "loss": 0.4615,
+      "step": 10688
+    },
+    {
+      "epoch": 2.9916036943744753,
+      "grad_norm": 0.260821699009736,
+      "learning_rate": 2.286695535386185e-09,
+      "loss": 0.4519,
+      "step": 10689
+    },
+    {
+      "epoch": 2.9918835712286596,
+      "grad_norm": 0.2515545516960629,
+      "learning_rate": 2.1415473231978236e-09,
+      "loss": 0.4532,
+      "step": 10690
+    },
+    {
+      "epoch": 2.9921634480828434,
+      "grad_norm": 0.2640350546718974,
+      "learning_rate": 2.001157935160869e-09,
+      "loss": 0.4628,
+      "step": 10691
+    },
+    {
+      "epoch": 2.9924433249370277,
+      "grad_norm": 0.2552477225519158,
+      "learning_rate": 1.865527384642407e-09,
+      "loss": 0.443,
+      "step": 10692
+    },
+    {
+      "epoch": 2.992723201791212,
+      "grad_norm": 0.28604816511846115,
+      "learning_rate": 1.7346556845432294e-09,
+      "loss": 0.4417,
+      "step": 10693
+    },
+    {
+      "epoch": 2.993003078645396,
+      "grad_norm": 0.2607651957497237,
+      "learning_rate": 1.6085428473311404e-09,
+      "loss": 0.4655,
+      "step": 10694
+    },
+    {
+      "epoch": 2.99328295549958,
+      "grad_norm": 0.2537250574074743,
+      "learning_rate": 1.4871888849965488e-09,
+      "loss": 0.4519,
+      "step": 10695
+    },
+    {
+      "epoch": 2.9935628323537644,
+      "grad_norm": 0.2496928430714104,
+      "learning_rate": 1.3705938091024273e-09,
+      "loss": 0.4401,
+      "step": 10696
+    },
+    {
+      "epoch": 2.9938427092079483,
+      "grad_norm": 0.2458011649695888,
+      "learning_rate": 1.2587576307343528e-09,
+      "loss": 0.4475,
+      "step": 10697
+    },
+    {
+      "epoch": 2.9941225860621326,
+      "grad_norm": 0.24453034008079477,
+      "learning_rate": 1.1516803605504666e-09,
+      "loss": 0.4365,
+      "step": 10698
+    },
+    {
+      "epoch": 2.994402462916317,
+      "grad_norm": 0.26694280093260125,
+      "learning_rate": 1.0493620087315136e-09,
+      "loss": 0.4531,
+      "step": 10699
+    },
+    {
+      "epoch": 2.994682339770501,
+      "grad_norm": 0.25832226564127514,
+      "learning_rate": 9.51802585019701e-10,
+      "loss": 0.4457,
+      "step": 10700
+    },
+    {
+      "epoch": 2.994962216624685,
+      "grad_norm": 0.25240393231643893,
+      "learning_rate": 8.590020987020442e-10,
+      "loss": 0.4708,
+      "step": 10701
+    },
+    {
+      "epoch": 2.9952420934788693,
+      "grad_norm": 0.253125299590793,
+      "learning_rate": 7.709605586103674e-10,
+      "loss": 0.4476,
+      "step": 10702
+    },
+    {
+      "epoch": 2.9955219703330536,
+      "grad_norm": 0.2445135905145077,
+      "learning_rate": 6.876779731213035e-10,
+      "loss": 0.4462,
+      "step": 10703
+    },
+    {
+      "epoch": 2.995801847187238,
+      "grad_norm": 0.25008011484227777,
+      "learning_rate": 6.091543501673958e-10,
+      "loss": 0.445,
+      "step": 10704
+    },
+    {
+      "epoch": 2.9960817240414217,
+      "grad_norm": 0.25643300402049357,
+      "learning_rate": 5.353896972204453e-10,
+      "loss": 0.4406,
+      "step": 10705
+    },
+    {
+      "epoch": 2.996361600895606,
+      "grad_norm": 0.24847707003519692,
+      "learning_rate": 4.663840213026127e-10,
+      "loss": 0.4677,
+      "step": 10706
+    },
+    {
+      "epoch": 2.9966414777497903,
+      "grad_norm": 0.2520890475535469,
+      "learning_rate": 4.0213732897531607e-10,
+      "loss": 0.449,
+      "step": 10707
+    },
+    {
+      "epoch": 2.996921354603974,
+      "grad_norm": 0.24197743329090915,
+      "learning_rate": 3.4264962636143536e-10,
+      "loss": 0.4382,
+      "step": 10708
+    },
+    {
+      "epoch": 2.9972012314581584,
+      "grad_norm": 0.2663302724507473,
+      "learning_rate": 2.8792091912310804e-10,
+      "loss": 0.4549,
+      "step": 10709
+    },
+    {
+      "epoch": 2.9974811083123427,
+      "grad_norm": 0.27210717025743686,
+      "learning_rate": 2.379512124617289e-10,
+      "loss": 0.4738,
+      "step": 10710
+    },
+    {
+      "epoch": 2.9977609851665266,
+      "grad_norm": 0.25045671556430316,
+      "learning_rate": 1.9274051114015478e-10,
+      "loss": 0.4365,
+      "step": 10711
+    },
+    {
+      "epoch": 2.998040862020711,
+      "grad_norm": 0.2669187587751103,
+      "learning_rate": 1.522888194604999e-10,
+      "loss": 0.4566,
+      "step": 10712
+    },
+    {
+      "epoch": 2.998320738874895,
+      "grad_norm": 0.2600536100682567,
+      "learning_rate": 1.16596141269687e-10,
+      "loss": 0.4347,
+      "step": 10713
+    },
+    {
+      "epoch": 2.998600615729079,
+      "grad_norm": 0.2541319276216827,
+      "learning_rate": 8.566247997054966e-11,
+      "loss": 0.4663,
+      "step": 10714
+    },
+    {
+      "epoch": 2.9988804925832633,
+      "grad_norm": 0.2576446088985775,
+      "learning_rate": 5.948783850517892e-11,
+      "loss": 0.4636,
+      "step": 10715
+    },
+    {
+      "epoch": 2.9991603694374476,
+      "grad_norm": 0.2536248818180751,
+      "learning_rate": 3.807221936047434e-11,
+      "loss": 0.435,
+      "step": 10716
+    },
+    {
+      "epoch": 2.9994402462916314,
+      "grad_norm": 0.2545279038011797,
+      "learning_rate": 2.1415624579246284e-11,
+      "loss": 0.4368,
+      "step": 10717
+    },
+    {
+      "epoch": 2.9997201231458157,
+      "grad_norm": 0.2604701422540892,
+      "learning_rate": 9.51805574356257e-12,
+      "loss": 0.4551,
+      "step": 10718
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2550703778343758,
+      "learning_rate": 2.3795139914017937e-12,
+      "loss": 0.4401,
+      "step": 10719
     }
   ],
   "logging_steps": 1,
@@ -70022,12 +75055,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 2349444853596160.0,
+  "total_flos": 2518172089335808.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null