diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.539942252165544,
+  "epoch": 1.9249278152069298,
   "eval_steps": 500,
-  "global_step": 40000,
+  "global_step": 50000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -56007,6 +56007,14006 @@
       "learning_rate": 2.5063532569201454e-05,
       "loss": 0.765,
       "step": 40000
+    },
+    {
+      "epoch": 1.5401347449470646,
+      "grad_norm": 1.959747076034546,
+      "learning_rate": 2.5043512613218933e-05,
+      "loss": 0.8532,
+      "step": 40005
+    },
+    {
+      "epoch": 1.5403272377285853,
+      "grad_norm": 1.0300486087799072,
+      "learning_rate": 2.502349951151064e-05,
+      "loss": 0.7775,
+      "step": 40010
+    },
+    {
+      "epoch": 1.540519730510106,
+      "grad_norm": 1.6674004793167114,
+      "learning_rate": 2.5003493265906664e-05,
+      "loss": 0.9549,
+      "step": 40015
+    },
+    {
+      "epoch": 1.5407122232916266,
+      "grad_norm": 1.486302137374878,
+      "learning_rate": 2.4983493878236374e-05,
+      "loss": 0.7536,
+      "step": 40020
+    },
+    {
+      "epoch": 1.5409047160731473,
+      "grad_norm": 0.8989318013191223,
+      "learning_rate": 2.496350135032869e-05,
+      "loss": 0.805,
+      "step": 40025
+    },
+    {
+      "epoch": 1.541097208854668,
+      "grad_norm": 1.692155361175537,
+      "learning_rate": 2.49435156840117e-05,
+      "loss": 0.8884,
+      "step": 40030
+    },
+    {
+      "epoch": 1.5412897016361886,
+      "grad_norm": 0.7467209100723267,
+      "learning_rate": 2.4923536881112997e-05,
+      "loss": 0.7268,
+      "step": 40035
+    },
+    {
+      "epoch": 1.5414821944177093,
+      "grad_norm": 1.0695582628250122,
+      "learning_rate": 2.490356494345951e-05,
+      "loss": 0.6906,
+      "step": 40040
+    },
+    {
+      "epoch": 1.54167468719923,
+      "grad_norm": 1.119673728942871,
+      "learning_rate": 2.4883599872877583e-05,
+      "loss": 0.8238,
+      "step": 40045
+    },
+    {
+      "epoch": 1.5418671799807506,
+      "grad_norm": 1.4105931520462036,
+      "learning_rate": 2.4863641671192806e-05,
+      "loss": 0.7352,
+      "step": 40050
+    },
+    {
+      "epoch": 1.5420596727622713,
+      "grad_norm": 0.8062414526939392,
+      "learning_rate": 2.4843690340230265e-05,
+      "loss": 0.8423,
+      "step": 40055
+    },
+    {
+      "epoch": 1.542252165543792,
+      "grad_norm": 1.100274682044983,
+      "learning_rate": 2.4823745881814374e-05,
+      "loss": 0.7665,
+      "step": 40060
+    },
+    {
+      "epoch": 1.5424446583253129,
+      "grad_norm": 1.5957090854644775,
+      "learning_rate": 2.4803808297768937e-05,
+      "loss": 0.7665,
+      "step": 40065
+    },
+    {
+      "epoch": 1.5426371511068335,
+      "grad_norm": 1.4533089399337769,
+      "learning_rate": 2.4783877589917125e-05,
+      "loss": 0.7874,
+      "step": 40070
+    },
+    {
+      "epoch": 1.5428296438883542,
+      "grad_norm": 1.0134419202804565,
+      "learning_rate": 2.4763953760081414e-05,
+      "loss": 0.8836,
+      "step": 40075
+    },
+    {
+      "epoch": 1.5430221366698749,
+      "grad_norm": 1.2606679201126099,
+      "learning_rate": 2.4744036810083738e-05,
+      "loss": 0.8233,
+      "step": 40080
+    },
+    {
+      "epoch": 1.5432146294513955,
+      "grad_norm": 1.0272793769836426,
+      "learning_rate": 2.472412674174538e-05,
+      "loss": 0.8782,
+      "step": 40085
+    },
+    {
+      "epoch": 1.5434071222329164,
+      "grad_norm": 0.853097140789032,
+      "learning_rate": 2.4704223556886998e-05,
+      "loss": 0.7858,
+      "step": 40090
+    },
+    {
+      "epoch": 1.543599615014437,
+      "grad_norm": 1.5345498323440552,
+      "learning_rate": 2.4684327257328522e-05,
+      "loss": 0.7227,
+      "step": 40095
+    },
+    {
+      "epoch": 1.5437921077959578,
+      "grad_norm": 1.565811276435852,
+      "learning_rate": 2.4664437844889454e-05,
+      "loss": 0.8012,
+      "step": 40100
+    },
+    {
+      "epoch": 1.5439846005774784,
+      "grad_norm": 1.7416857481002808,
+      "learning_rate": 2.4644555321388462e-05,
+      "loss": 0.7459,
+      "step": 40105
+    },
+    {
+      "epoch": 1.544177093358999,
+      "grad_norm": 1.2711718082427979,
+      "learning_rate": 2.4624679688643716e-05,
+      "loss": 0.7688,
+      "step": 40110
+    },
+    {
+      "epoch": 1.5443695861405198,
+      "grad_norm": 1.38042414188385,
+      "learning_rate": 2.4604810948472677e-05,
+      "loss": 0.8042,
+      "step": 40115
+    },
+    {
+      "epoch": 1.5445620789220404,
+      "grad_norm": 1.0315542221069336,
+      "learning_rate": 2.458494910269228e-05,
+      "loss": 0.8793,
+      "step": 40120
+    },
+    {
+      "epoch": 1.5447545717035611,
+      "grad_norm": 1.4990910291671753,
+      "learning_rate": 2.456509415311864e-05,
+      "loss": 0.766,
+      "step": 40125
+    },
+    {
+      "epoch": 1.5449470644850818,
+      "grad_norm": 1.0002154111862183,
+      "learning_rate": 2.4545246101567476e-05,
+      "loss": 0.8452,
+      "step": 40130
+    },
+    {
+      "epoch": 1.5451395572666025,
+      "grad_norm": 1.0912269353866577,
+      "learning_rate": 2.4525404949853702e-05,
+      "loss": 0.71,
+      "step": 40135
+    },
+    {
+      "epoch": 1.5453320500481231,
+      "grad_norm": 1.2968130111694336,
+      "learning_rate": 2.4505570699791668e-05,
+      "loss": 0.6927,
+      "step": 40140
+    },
+    {
+      "epoch": 1.5455245428296438,
+      "grad_norm": 0.9810501933097839,
+      "learning_rate": 2.448574335319508e-05,
+      "loss": 0.821,
+      "step": 40145
+    },
+    {
+      "epoch": 1.5457170356111645,
+      "grad_norm": 1.6552939414978027,
+      "learning_rate": 2.446592291187706e-05,
+      "loss": 0.8898,
+      "step": 40150
+    },
+    {
+      "epoch": 1.5459095283926851,
+      "grad_norm": 2.9171223640441895,
+      "learning_rate": 2.4446109377649996e-05,
+      "loss": 0.8895,
+      "step": 40155
+    },
+    {
+      "epoch": 1.5461020211742058,
+      "grad_norm": 1.7031809091567993,
+      "learning_rate": 2.4426302752325735e-05,
+      "loss": 0.7551,
+      "step": 40160
+    },
+    {
+      "epoch": 1.5462945139557267,
+      "grad_norm": 0.709255039691925,
+      "learning_rate": 2.4406503037715445e-05,
+      "loss": 0.7308,
+      "step": 40165
+    },
+    {
+      "epoch": 1.5464870067372474,
+      "grad_norm": 1.2134255170822144,
+      "learning_rate": 2.4386710235629708e-05,
+      "loss": 0.8311,
+      "step": 40170
+    },
+    {
+      "epoch": 1.546679499518768,
+      "grad_norm": 0.7918189764022827,
+      "learning_rate": 2.436692434787844e-05,
+      "loss": 0.7764,
+      "step": 40175
+    },
+    {
+      "epoch": 1.5468719923002887,
+      "grad_norm": 1.3109943866729736,
+      "learning_rate": 2.4347145376270896e-05,
+      "loss": 0.8472,
+      "step": 40180
+    },
+    {
+      "epoch": 1.5470644850818094,
+      "grad_norm": 1.0307000875473022,
+      "learning_rate": 2.4327373322615754e-05,
+      "loss": 0.8618,
+      "step": 40185
+    },
+    {
+      "epoch": 1.5472569778633303,
+      "grad_norm": 0.8745335936546326,
+      "learning_rate": 2.430760818872103e-05,
+      "loss": 0.7465,
+      "step": 40190
+    },
+    {
+      "epoch": 1.547449470644851,
+      "grad_norm": 1.271261215209961,
+      "learning_rate": 2.428784997639415e-05,
+      "loss": 0.6754,
+      "step": 40195
+    },
+    {
+      "epoch": 1.5476419634263716,
+      "grad_norm": 0.9863643646240234,
+      "learning_rate": 2.42680986874418e-05,
+      "loss": 0.7717,
+      "step": 40200
+    },
+    {
+      "epoch": 1.5478344562078923,
+      "grad_norm": 1.5643993616104126,
+      "learning_rate": 2.4248354323670185e-05,
+      "loss": 0.8209,
+      "step": 40205
+    },
+    {
+      "epoch": 1.548026948989413,
+      "grad_norm": 1.2394016981124878,
+      "learning_rate": 2.4228616886884713e-05,
+      "loss": 0.8473,
+      "step": 40210
+    },
+    {
+      "epoch": 1.5482194417709336,
+      "grad_norm": 1.1060245037078857,
+      "learning_rate": 2.420888637889034e-05,
+      "loss": 0.6498,
+      "step": 40215
+    },
+    {
+      "epoch": 1.5484119345524543,
+      "grad_norm": 1.6194653511047363,
+      "learning_rate": 2.4189162801491206e-05,
+      "loss": 0.839,
+      "step": 40220
+    },
+    {
+      "epoch": 1.548604427333975,
+      "grad_norm": 0.9702569842338562,
+      "learning_rate": 2.4169446156490938e-05,
+      "loss": 0.8054,
+      "step": 40225
+    },
+    {
+      "epoch": 1.5487969201154956,
+      "grad_norm": 0.964542031288147,
+      "learning_rate": 2.4149736445692483e-05,
+      "loss": 0.7966,
+      "step": 40230
+    },
+    {
+      "epoch": 1.5489894128970163,
+      "grad_norm": 1.230900526046753,
+      "learning_rate": 2.413003367089821e-05,
+      "loss": 0.8332,
+      "step": 40235
+    },
+    {
+      "epoch": 1.549181905678537,
+      "grad_norm": 0.8581804633140564,
+      "learning_rate": 2.411033783390969e-05,
+      "loss": 0.8305,
+      "step": 40240
+    },
+    {
+      "epoch": 1.5493743984600576,
+      "grad_norm": 1.232013463973999,
+      "learning_rate": 2.4090648936528125e-05,
+      "loss": 0.7663,
+      "step": 40245
+    },
+    {
+      "epoch": 1.5495668912415783,
+      "grad_norm": 1.6917791366577148,
+      "learning_rate": 2.407096698055382e-05,
+      "loss": 0.8289,
+      "step": 40250
+    },
+    {
+      "epoch": 1.549759384023099,
+      "grad_norm": 1.128301978111267,
+      "learning_rate": 2.4051291967786605e-05,
+      "loss": 0.8291,
+      "step": 40255
+    },
+    {
+      "epoch": 1.5499518768046199,
+      "grad_norm": 1.8056789636611938,
+      "learning_rate": 2.4031623900025624e-05,
+      "loss": 0.7495,
+      "step": 40260
+    },
+    {
+      "epoch": 1.5501443695861405,
+      "grad_norm": 1.2374507188796997,
+      "learning_rate": 2.4011962779069432e-05,
+      "loss": 0.7374,
+      "step": 40265
+    },
+    {
+      "epoch": 1.5503368623676612,
+      "grad_norm": 1.8159735202789307,
+      "learning_rate": 2.3992308606715828e-05,
+      "loss": 0.7937,
+      "step": 40270
+    },
+    {
+      "epoch": 1.5505293551491819,
+      "grad_norm": 1.4417248964309692,
+      "learning_rate": 2.3972661384762096e-05,
+      "loss": 0.8844,
+      "step": 40275
+    },
+    {
+      "epoch": 1.5507218479307026,
+      "grad_norm": 1.61116623878479,
+      "learning_rate": 2.3953021115004858e-05,
+      "loss": 0.8133,
+      "step": 40280
+    },
+    {
+      "epoch": 1.5509143407122234,
+      "grad_norm": 1.6744754314422607,
+      "learning_rate": 2.393338779924006e-05,
+      "loss": 0.7884,
+      "step": 40285
+    },
+    {
+      "epoch": 1.5511068334937441,
+      "grad_norm": 1.0945570468902588,
+      "learning_rate": 2.3913761439263095e-05,
+      "loss": 0.9085,
+      "step": 40290
+    },
+    {
+      "epoch": 1.5512993262752648,
+      "grad_norm": 1.8907279968261719,
+      "learning_rate": 2.3894142036868583e-05,
+      "loss": 0.9115,
+      "step": 40295
+    },
+    {
+      "epoch": 1.5514918190567855,
+      "grad_norm": 1.0637871026992798,
+      "learning_rate": 2.3874529593850624e-05,
+      "loss": 0.7011,
+      "step": 40300
+    },
+    {
+      "epoch": 1.5516843118383061,
+      "grad_norm": 1.661790132522583,
+      "learning_rate": 2.3854924112002665e-05,
+      "loss": 0.7207,
+      "step": 40305
+    },
+    {
+      "epoch": 1.5518768046198268,
+      "grad_norm": 1.138431429862976,
+      "learning_rate": 2.3835325593117498e-05,
+      "loss": 0.8999,
+      "step": 40310
+    },
+    {
+      "epoch": 1.5520692974013475,
+      "grad_norm": 1.0955361127853394,
+      "learning_rate": 2.381573403898719e-05,
+      "loss": 0.8033,
+      "step": 40315
+    },
+    {
+      "epoch": 1.5522617901828681,
+      "grad_norm": 0.4663851857185364,
+      "learning_rate": 2.3796149451403405e-05,
+      "loss": 0.8223,
+      "step": 40320
+    },
+    {
+      "epoch": 1.5524542829643888,
+      "grad_norm": 0.9485166072845459,
+      "learning_rate": 2.3776571832156914e-05,
+      "loss": 0.8846,
+      "step": 40325
+    },
+    {
+      "epoch": 1.5526467757459095,
+      "grad_norm": 0.8865009546279907,
+      "learning_rate": 2.3757001183037998e-05,
+      "loss": 0.7741,
+      "step": 40330
+    },
+    {
+      "epoch": 1.5528392685274301,
+      "grad_norm": 1.5837032794952393,
+      "learning_rate": 2.3737437505836257e-05,
+      "loss": 0.8743,
+      "step": 40335
+    },
+    {
+      "epoch": 1.5530317613089508,
+      "grad_norm": 1.4707670211791992,
+      "learning_rate": 2.3717880802340698e-05,
+      "loss": 0.8894,
+      "step": 40340
+    },
+    {
+      "epoch": 1.5532242540904715,
+      "grad_norm": 1.321226954460144,
+      "learning_rate": 2.3698331074339553e-05,
+      "loss": 0.9362,
+      "step": 40345
+    },
+    {
+      "epoch": 1.5534167468719922,
+      "grad_norm": 1.60404372215271,
+      "learning_rate": 2.3678788323620638e-05,
+      "loss": 0.7989,
+      "step": 40350
+    },
+    {
+      "epoch": 1.5536092396535128,
+      "grad_norm": 1.4875679016113281,
+      "learning_rate": 2.3659252551970922e-05,
+      "loss": 0.8456,
+      "step": 40355
+    },
+    {
+      "epoch": 1.5538017324350337,
+      "grad_norm": 1.7115007638931274,
+      "learning_rate": 2.3639723761176847e-05,
+      "loss": 0.8346,
+      "step": 40360
+    },
+    {
+      "epoch": 1.5539942252165544,
+      "grad_norm": 1.4487700462341309,
+      "learning_rate": 2.362020195302419e-05,
+      "loss": 0.7359,
+      "step": 40365
+    },
+    {
+      "epoch": 1.554186717998075,
+      "grad_norm": 1.6501834392547607,
+      "learning_rate": 2.3600687129298126e-05,
+      "loss": 0.769,
+      "step": 40370
+    },
+    {
+      "epoch": 1.5543792107795957,
+      "grad_norm": 1.4506645202636719,
+      "learning_rate": 2.3581179291783094e-05,
+      "loss": 0.9698,
+      "step": 40375
+    },
+    {
+      "epoch": 1.5545717035611166,
+      "grad_norm": 1.098969578742981,
+      "learning_rate": 2.356167844226299e-05,
+      "loss": 0.8277,
+      "step": 40380
+    },
+    {
+      "epoch": 1.5547641963426373,
+      "grad_norm": 0.9754076600074768,
+      "learning_rate": 2.3542184582521034e-05,
+      "loss": 0.754,
+      "step": 40385
+    },
+    {
+      "epoch": 1.554956689124158,
+      "grad_norm": 1.5836515426635742,
+      "learning_rate": 2.3522697714339814e-05,
+      "loss": 0.8996,
+      "step": 40390
+    },
+    {
+      "epoch": 1.5551491819056786,
+      "grad_norm": 1.1244460344314575,
+      "learning_rate": 2.3503217839501302e-05,
+      "loss": 0.721,
+      "step": 40395
+    },
+    {
+      "epoch": 1.5553416746871993,
+      "grad_norm": 0.96147620677948,
+      "learning_rate": 2.3483744959786735e-05,
+      "loss": 0.7998,
+      "step": 40400
+    },
+    {
+      "epoch": 1.55553416746872,
+      "grad_norm": 0.9615457653999329,
+      "learning_rate": 2.3464279076976837e-05,
+      "loss": 0.9238,
+      "step": 40405
+    },
+    {
+      "epoch": 1.5557266602502406,
+      "grad_norm": 2.3235690593719482,
+      "learning_rate": 2.344482019285159e-05,
+      "loss": 0.9334,
+      "step": 40410
+    },
+    {
+      "epoch": 1.5559191530317613,
+      "grad_norm": 1.226501703262329,
+      "learning_rate": 2.3425368309190455e-05,
+      "loss": 0.8288,
+      "step": 40415
+    },
+    {
+      "epoch": 1.556111645813282,
+      "grad_norm": 1.4257621765136719,
+      "learning_rate": 2.3405923427772057e-05,
+      "loss": 0.5693,
+      "step": 40420
+    },
+    {
+      "epoch": 1.5563041385948027,
+      "grad_norm": 1.0261151790618896,
+      "learning_rate": 2.3386485550374636e-05,
+      "loss": 0.9288,
+      "step": 40425
+    },
+    {
+      "epoch": 1.5564966313763233,
+      "grad_norm": 1.0929063558578491,
+      "learning_rate": 2.3367054678775557e-05,
+      "loss": 0.7563,
+      "step": 40430
+    },
+    {
+      "epoch": 1.556689124157844,
+      "grad_norm": 0.8762004375457764,
+      "learning_rate": 2.3347630814751687e-05,
+      "loss": 0.7824,
+      "step": 40435
+    },
+    {
+      "epoch": 1.5568816169393647,
+      "grad_norm": 1.4772645235061646,
+      "learning_rate": 2.3328213960079205e-05,
+      "loss": 0.8197,
+      "step": 40440
+    },
+    {
+      "epoch": 1.5570741097208853,
+      "grad_norm": 1.4557557106018066,
+      "learning_rate": 2.330880411653368e-05,
+      "loss": 0.8008,
+      "step": 40445
+    },
+    {
+      "epoch": 1.557266602502406,
+      "grad_norm": 1.2183635234832764,
+      "learning_rate": 2.3289401285889934e-05,
+      "loss": 0.769,
+      "step": 40450
+    },
+    {
+      "epoch": 1.557459095283927,
+      "grad_norm": 0.9675846695899963,
+      "learning_rate": 2.327000546992233e-05,
+      "loss": 0.8019,
+      "step": 40455
+    },
+    {
+      "epoch": 1.5576515880654476,
+      "grad_norm": 1.5185277462005615,
+      "learning_rate": 2.32506166704044e-05,
+      "loss": 0.6456,
+      "step": 40460
+    },
+    {
+      "epoch": 1.5578440808469682,
+      "grad_norm": 1.6056936979293823,
+      "learning_rate": 2.3231234889109165e-05,
+      "loss": 0.8468,
+      "step": 40465
+    },
+    {
+      "epoch": 1.558036573628489,
+      "grad_norm": 1.182332992553711,
+      "learning_rate": 2.3211860127808948e-05,
+      "loss": 0.991,
+      "step": 40470
+    },
+    {
+      "epoch": 1.5582290664100096,
+      "grad_norm": 1.1876064538955688,
+      "learning_rate": 2.3192492388275454e-05,
+      "loss": 0.7254,
+      "step": 40475
+    },
+    {
+      "epoch": 1.5584215591915305,
+      "grad_norm": 1.7529163360595703,
+      "learning_rate": 2.3173131672279725e-05,
+      "loss": 0.9351,
+      "step": 40480
+    },
+    {
+      "epoch": 1.5586140519730511,
+      "grad_norm": 1.0285496711730957,
+      "learning_rate": 2.3153777981592207e-05,
+      "loss": 0.835,
+      "step": 40485
+    },
+    {
+      "epoch": 1.5588065447545718,
+      "grad_norm": 1.4586035013198853,
+      "learning_rate": 2.3134431317982597e-05,
+      "loss": 0.7694,
+      "step": 40490
+    },
+    {
+      "epoch": 1.5589990375360925,
+      "grad_norm": 1.656645655632019,
+      "learning_rate": 2.311509168322006e-05,
+      "loss": 0.7945,
+      "step": 40495
+    },
+    {
+      "epoch": 1.5591915303176132,
+      "grad_norm": 1.66303551197052,
+      "learning_rate": 2.30957590790731e-05,
+      "loss": 0.7964,
+      "step": 40500
+    },
+    {
+      "epoch": 1.5593840230991338,
+      "grad_norm": 1.375352382659912,
+      "learning_rate": 2.307643350730947e-05,
+      "loss": 0.9546,
+      "step": 40505
+    },
+    {
+      "epoch": 1.5595765158806545,
+      "grad_norm": 1.1943459510803223,
+      "learning_rate": 2.305711496969648e-05,
+      "loss": 0.6616,
+      "step": 40510
+    },
+    {
+      "epoch": 1.5597690086621752,
+      "grad_norm": 1.4341318607330322,
+      "learning_rate": 2.3037803468000597e-05,
+      "loss": 0.8164,
+      "step": 40515
+    },
+    {
+      "epoch": 1.5599615014436958,
+      "grad_norm": 1.3383629322052002,
+      "learning_rate": 2.301849900398776e-05,
+      "loss": 0.8711,
+      "step": 40520
+    },
+    {
+      "epoch": 1.5601539942252165,
+      "grad_norm": 1.8261584043502808,
+      "learning_rate": 2.2999201579423236e-05,
+      "loss": 0.8195,
+      "step": 40525
+    },
+    {
+      "epoch": 1.5603464870067372,
+      "grad_norm": 2.052516222000122,
+      "learning_rate": 2.2979911196071668e-05,
+      "loss": 0.9955,
+      "step": 40530
+    },
+    {
+      "epoch": 1.5605389797882578,
+      "grad_norm": 1.2053110599517822,
+      "learning_rate": 2.2960627855696958e-05,
+      "loss": 0.8021,
+      "step": 40535
+    },
+    {
+      "epoch": 1.5607314725697785,
+      "grad_norm": 1.301857590675354,
+      "learning_rate": 2.294135156006255e-05,
+      "loss": 0.8293,
+      "step": 40540
+    },
+    {
+      "epoch": 1.5609239653512992,
+      "grad_norm": 1.1416163444519043,
+      "learning_rate": 2.2922082310931036e-05,
+      "loss": 0.8486,
+      "step": 40545
+    },
+    {
+      "epoch": 1.56111645813282,
+      "grad_norm": 1.3656333684921265,
+      "learning_rate": 2.2902820110064503e-05,
+      "loss": 0.7598,
+      "step": 40550
+    },
+    {
+      "epoch": 1.5613089509143407,
+      "grad_norm": 2.20866060256958,
+      "learning_rate": 2.288356495922436e-05,
+      "loss": 0.7617,
+      "step": 40555
+    },
+    {
+      "epoch": 1.5615014436958614,
+      "grad_norm": 1.5742281675338745,
+      "learning_rate": 2.2864316860171375e-05,
+      "loss": 0.9051,
+      "step": 40560
+    },
+    {
+      "epoch": 1.561693936477382,
+      "grad_norm": 1.7272670269012451,
+      "learning_rate": 2.2845075814665573e-05,
+      "loss": 0.8531,
+      "step": 40565
+    },
+    {
+      "epoch": 1.5618864292589028,
+      "grad_norm": 1.8699864149093628,
+      "learning_rate": 2.2825841824466543e-05,
+      "loss": 0.7353,
+      "step": 40570
+    },
+    {
+      "epoch": 1.5620789220404236,
+      "grad_norm": 1.5581755638122559,
+      "learning_rate": 2.2806614891333022e-05,
+      "loss": 0.7466,
+      "step": 40575
+    },
+    {
+      "epoch": 1.5622714148219443,
+      "grad_norm": 1.340266227722168,
+      "learning_rate": 2.2787395017023205e-05,
+      "loss": 0.7033,
+      "step": 40580
+    },
+    {
+      "epoch": 1.562463907603465,
+      "grad_norm": 2.0326457023620605,
+      "learning_rate": 2.276818220329463e-05,
+      "loss": 0.8551,
+      "step": 40585
+    },
+    {
+      "epoch": 1.5626564003849857,
+      "grad_norm": 1.2885262966156006,
+      "learning_rate": 2.2748976451904203e-05,
+      "loss": 0.7386,
+      "step": 40590
+    },
+    {
+      "epoch": 1.5628488931665063,
+      "grad_norm": 0.5644479393959045,
+      "learning_rate": 2.2729777764608108e-05,
+      "loss": 0.618,
+      "step": 40595
+    },
+    {
+      "epoch": 1.563041385948027,
+      "grad_norm": 1.572131633758545,
+      "learning_rate": 2.2710586143161972e-05,
+      "loss": 0.7215,
+      "step": 40600
+    },
+    {
+      "epoch": 1.5632338787295477,
+      "grad_norm": 1.1839364767074585,
+      "learning_rate": 2.2691401589320737e-05,
+      "loss": 0.7607,
+      "step": 40605
+    },
+    {
+      "epoch": 1.5634263715110683,
+      "grad_norm": 1.3653178215026855,
+      "learning_rate": 2.2672224104838713e-05,
+      "loss": 0.6762,
+      "step": 40610
+    },
+    {
+      "epoch": 1.563618864292589,
+      "grad_norm": 1.8901535272598267,
+      "learning_rate": 2.2653053691469563e-05,
+      "loss": 0.8515,
+      "step": 40615
+    },
+    {
+      "epoch": 1.5638113570741097,
+      "grad_norm": 1.8246430158615112,
+      "learning_rate": 2.2633890350966248e-05,
+      "loss": 0.85,
+      "step": 40620
+    },
+    {
+      "epoch": 1.5640038498556303,
+      "grad_norm": 1.524446725845337,
+      "learning_rate": 2.2614734085081158e-05,
+      "loss": 0.7622,
+      "step": 40625
+    },
+    {
+      "epoch": 1.564196342637151,
+      "grad_norm": 1.142157793045044,
+      "learning_rate": 2.2595584895566e-05,
+      "loss": 0.7196,
+      "step": 40630
+    },
+    {
+      "epoch": 1.5643888354186717,
+      "grad_norm": 1.0782500505447388,
+      "learning_rate": 2.2576442784171892e-05,
+      "loss": 0.7491,
+      "step": 40635
+    },
+    {
+      "epoch": 1.5645813282001924,
+      "grad_norm": 1.034097671508789,
+      "learning_rate": 2.2557307752649137e-05,
+      "loss": 0.851,
+      "step": 40640
+    },
+    {
+      "epoch": 1.564773820981713,
+      "grad_norm": 1.894054889678955,
+      "learning_rate": 2.253817980274764e-05,
+      "loss": 0.6389,
+      "step": 40645
+    },
+    {
+      "epoch": 1.564966313763234,
+      "grad_norm": 0.6015737056732178,
+      "learning_rate": 2.251905893621642e-05,
+      "loss": 0.749,
+      "step": 40650
+    },
+    {
+      "epoch": 1.5651588065447546,
+      "grad_norm": 0.9904623627662659,
+      "learning_rate": 2.2499945154804013e-05,
+      "loss": 0.836,
+      "step": 40655
+    },
+    {
+      "epoch": 1.5653512993262753,
+      "grad_norm": 1.3582662343978882,
+      "learning_rate": 2.2480838460258226e-05,
+      "loss": 0.8239,
+      "step": 40660
+    },
+    {
+      "epoch": 1.565543792107796,
+      "grad_norm": 1.4158751964569092,
+      "learning_rate": 2.2461738854326263e-05,
+      "loss": 0.8528,
+      "step": 40665
+    },
+    {
+      "epoch": 1.5657362848893166,
+      "grad_norm": 1.0276498794555664,
+      "learning_rate": 2.244264633875459e-05,
+      "loss": 0.7761,
+      "step": 40670
+    },
+    {
+      "epoch": 1.5659287776708375,
+      "grad_norm": 1.4849311113357544,
+      "learning_rate": 2.24235609152892e-05,
+      "loss": 0.934,
+      "step": 40675
+    },
+    {
+      "epoch": 1.5661212704523582,
+      "grad_norm": 1.024477481842041,
+      "learning_rate": 2.2404482585675225e-05,
+      "loss": 0.675,
+      "step": 40680
+    },
+    {
+      "epoch": 1.5663137632338788,
+      "grad_norm": 1.2070591449737549,
+      "learning_rate": 2.2385411351657303e-05,
+      "loss": 0.7506,
+      "step": 40685
+    },
+    {
+      "epoch": 1.5665062560153995,
+      "grad_norm": 1.2458285093307495,
+      "learning_rate": 2.2366347214979366e-05,
+      "loss": 0.6916,
+      "step": 40690
+    },
+    {
+      "epoch": 1.5666987487969202,
+      "grad_norm": 1.7189850807189941,
+      "learning_rate": 2.2347290177384726e-05,
+      "loss": 0.7032,
+      "step": 40695
+    },
+    {
+      "epoch": 1.5668912415784408,
+      "grad_norm": 1.3378642797470093,
+      "learning_rate": 2.2328240240615972e-05,
+      "loss": 0.6995,
+      "step": 40700
+    },
+    {
+      "epoch": 1.5670837343599615,
+      "grad_norm": 1.4483795166015625,
+      "learning_rate": 2.2309197406415117e-05,
+      "loss": 0.7532,
+      "step": 40705
+    },
+    {
+      "epoch": 1.5672762271414822,
+      "grad_norm": 1.1107079982757568,
+      "learning_rate": 2.2290161676523503e-05,
+      "loss": 0.7787,
+      "step": 40710
+    },
+    {
+      "epoch": 1.5674687199230029,
+      "grad_norm": 1.3404992818832397,
+      "learning_rate": 2.2271133052681825e-05,
+      "loss": 0.7247,
+      "step": 40715
+    },
+    {
+      "epoch": 1.5676612127045235,
+      "grad_norm": 1.0901662111282349,
+      "learning_rate": 2.2252111536630148e-05,
+      "loss": 0.8229,
+      "step": 40720
+    },
+    {
+      "epoch": 1.5678537054860442,
+      "grad_norm": 1.3521157503128052,
+      "learning_rate": 2.2233097130107782e-05,
+      "loss": 0.6774,
+      "step": 40725
+    },
+    {
+      "epoch": 1.5680461982675649,
+      "grad_norm": 1.1512130498886108,
+      "learning_rate": 2.221408983485358e-05,
+      "loss": 0.8416,
+      "step": 40730
+    },
+    {
+      "epoch": 1.5682386910490855,
+      "grad_norm": 1.2217384576797485,
+      "learning_rate": 2.219508965260555e-05,
+      "loss": 0.8276,
+      "step": 40735
+    },
+    {
+      "epoch": 1.5684311838306062,
+      "grad_norm": 1.0364052057266235,
+      "learning_rate": 2.217609658510117e-05,
+      "loss": 0.7916,
+      "step": 40740
+    },
+    {
+      "epoch": 1.568623676612127,
+      "grad_norm": 1.143720269203186,
+      "learning_rate": 2.2157110634077215e-05,
+      "loss": 0.98,
+      "step": 40745
+    },
+    {
+      "epoch": 1.5688161693936478,
+      "grad_norm": 1.7398204803466797,
+      "learning_rate": 2.2138131801269857e-05,
+      "loss": 0.8718,
+      "step": 40750
+    },
+    {
+      "epoch": 1.5690086621751684,
+      "grad_norm": 1.4039459228515625,
+      "learning_rate": 2.2119160088414502e-05,
+      "loss": 1.0399,
+      "step": 40755
+    },
+    {
+      "epoch": 1.569201154956689,
+      "grad_norm": 1.1382685899734497,
+      "learning_rate": 2.2100195497246103e-05,
+      "loss": 0.8066,
+      "step": 40760
+    },
+    {
+      "epoch": 1.5693936477382098,
+      "grad_norm": 1.2807059288024902,
+      "learning_rate": 2.208123802949875e-05,
+      "loss": 0.8231,
+      "step": 40765
+    },
+    {
+      "epoch": 1.5695861405197307,
+      "grad_norm": 1.0869920253753662,
+      "learning_rate": 2.2062287686906026e-05,
+      "loss": 0.7107,
+      "step": 40770
+    },
+    {
+      "epoch": 1.5697786333012513,
+      "grad_norm": 1.4672589302062988,
+      "learning_rate": 2.20433444712008e-05,
+      "loss": 0.6363,
+      "step": 40775
+    },
+    {
+      "epoch": 1.569971126082772,
+      "grad_norm": 1.185530185699463,
+      "learning_rate": 2.2024408384115337e-05,
+      "loss": 0.7595,
+      "step": 40780
+    },
+    {
+      "epoch": 1.5701636188642927,
+      "grad_norm": 1.210688829421997,
+      "learning_rate": 2.2005479427381126e-05,
+      "loss": 0.7845,
+      "step": 40785
+    },
+    {
+      "epoch": 1.5703561116458133,
+      "grad_norm": 0.9549643397331238,
+      "learning_rate": 2.1986557602729207e-05,
+      "loss": 0.7966,
+      "step": 40790
+    },
+    {
+      "epoch": 1.570548604427334,
+      "grad_norm": 1.4014770984649658,
+      "learning_rate": 2.1967642911889787e-05,
+      "loss": 0.7121,
+      "step": 40795
+    },
+    {
+      "epoch": 1.5707410972088547,
+      "grad_norm": 1.2537885904312134,
+      "learning_rate": 2.1948735356592497e-05,
+      "loss": 0.8161,
+      "step": 40800
+    },
+    {
+      "epoch": 1.5709335899903754,
+      "grad_norm": 1.4344686269760132,
+      "learning_rate": 2.1929834938566317e-05,
+      "loss": 0.9185,
+      "step": 40805
+    },
+    {
+      "epoch": 1.571126082771896,
+      "grad_norm": 1.5772898197174072,
+      "learning_rate": 2.1910941659539585e-05,
+      "loss": 0.7617,
+      "step": 40810
+    },
+    {
+      "epoch": 1.5713185755534167,
+      "grad_norm": 0.9699328541755676,
+      "learning_rate": 2.189205552123993e-05,
+      "loss": 0.7618,
+      "step": 40815
+    },
+    {
+      "epoch": 1.5715110683349374,
+      "grad_norm": 1.8218029737472534,
+      "learning_rate": 2.1873176525394378e-05,
+      "loss": 0.8044,
+      "step": 40820
+    },
+    {
+      "epoch": 1.571703561116458,
+      "grad_norm": 1.0945191383361816,
+      "learning_rate": 2.1854304673729287e-05,
+      "loss": 0.8599,
+      "step": 40825
+    },
+    {
+      "epoch": 1.5718960538979787,
+      "grad_norm": 1.4021522998809814,
+      "learning_rate": 2.183543996797036e-05,
+      "loss": 0.8188,
+      "step": 40830
+    },
+    {
+      "epoch": 1.5720885466794994,
+      "grad_norm": 0.9821120500564575,
+      "learning_rate": 2.181658240984269e-05,
+      "loss": 0.8562,
+      "step": 40835
+    },
+    {
+      "epoch": 1.57228103946102,
+      "grad_norm": 1.1095430850982666,
+      "learning_rate": 2.1797732001070613e-05,
+      "loss": 0.7987,
+      "step": 40840
+    },
+    {
+      "epoch": 1.572473532242541,
+      "grad_norm": 1.1971663236618042,
+      "learning_rate": 2.17788887433779e-05,
+      "loss": 0.942,
+      "step": 40845
+    },
+    {
+      "epoch": 1.5726660250240616,
+      "grad_norm": 1.4306426048278809,
+      "learning_rate": 2.176005263848765e-05,
+      "loss": 0.922,
+      "step": 40850
+    },
+    {
+      "epoch": 1.5728585178055823,
+      "grad_norm": 1.3114017248153687,
+      "learning_rate": 2.1741223688122313e-05,
+      "loss": 0.76,
+      "step": 40855
+    },
+    {
+      "epoch": 1.573051010587103,
+      "grad_norm": 1.3243577480316162,
+      "learning_rate": 2.1722401894003608e-05,
+      "loss": 0.7896,
+      "step": 40860
+    },
+    {
+      "epoch": 1.5732435033686238,
+      "grad_norm": 1.723612666130066,
+      "learning_rate": 2.1703587257852755e-05,
+      "loss": 0.9003,
+      "step": 40865
+    },
+    {
+      "epoch": 1.5734359961501445,
+      "grad_norm": 1.3303959369659424,
+      "learning_rate": 2.1684779781390152e-05,
+      "loss": 0.8007,
+      "step": 40870
+    },
+    {
+      "epoch": 1.5736284889316652,
+      "grad_norm": 1.080396056175232,
+      "learning_rate": 2.166597946633565e-05,
+      "loss": 0.7348,
+      "step": 40875
+    },
+    {
+      "epoch": 1.5738209817131859,
+      "grad_norm": 0.7364131808280945,
+      "learning_rate": 2.1647186314408408e-05,
+      "loss": 0.8049,
+      "step": 40880
+    },
+    {
+      "epoch": 1.5740134744947065,
+      "grad_norm": 0.581088662147522,
+      "learning_rate": 2.162840032732697e-05,
+      "loss": 0.7795,
+      "step": 40885
+    },
+    {
+      "epoch": 1.5742059672762272,
+      "grad_norm": 1.5997105836868286,
+      "learning_rate": 2.1609621506809097e-05,
+      "loss": 0.7561,
+      "step": 40890
+    },
+    {
+      "epoch": 1.5743984600577479,
+      "grad_norm": 1.5176271200180054,
+      "learning_rate": 2.1590849854572114e-05,
+      "loss": 0.6649,
+      "step": 40895
+    },
+    {
+      "epoch": 1.5745909528392685,
+      "grad_norm": 2.400883197784424,
+      "learning_rate": 2.1572085372332463e-05,
+      "loss": 0.8173,
+      "step": 40900
+    },
+    {
+      "epoch": 1.5747834456207892,
+      "grad_norm": 1.7215204238891602,
+      "learning_rate": 2.1553328061806065e-05,
+      "loss": 0.8483,
+      "step": 40905
+    },
+    {
+      "epoch": 1.5749759384023099,
+      "grad_norm": 1.2602962255477905,
+      "learning_rate": 2.1534577924708155e-05,
+      "loss": 0.7841,
+      "step": 40910
+    },
+    {
+      "epoch": 1.5751684311838305,
+      "grad_norm": 2.122758388519287,
+      "learning_rate": 2.1515834962753346e-05,
+      "loss": 0.7834,
+      "step": 40915
+    },
+    {
+      "epoch": 1.5753609239653512,
+      "grad_norm": 1.6246029138565063,
+      "learning_rate": 2.1497099177655476e-05,
+      "loss": 0.7015,
+      "step": 40920
+    },
+    {
+      "epoch": 1.5755534167468719,
+      "grad_norm": 1.3783003091812134,
+      "learning_rate": 2.147837057112786e-05,
+      "loss": 0.9372,
+      "step": 40925
+    },
+    {
+      "epoch": 1.5757459095283926,
+      "grad_norm": 1.430508017539978,
+      "learning_rate": 2.14596491448831e-05,
+      "loss": 0.8844,
+      "step": 40930
+    },
+    {
+      "epoch": 1.5759384023099132,
+      "grad_norm": 1.1913694143295288,
+      "learning_rate": 2.1440934900633148e-05,
+      "loss": 0.8196,
+      "step": 40935
+    },
+    {
+      "epoch": 1.5761308950914341,
+      "grad_norm": 0.9103003740310669,
+      "learning_rate": 2.1422227840089303e-05,
+      "loss": 0.8644,
+      "step": 40940
+    },
+    {
+      "epoch": 1.5763233878729548,
+      "grad_norm": 1.2136942148208618,
+      "learning_rate": 2.1403527964962176e-05,
+      "loss": 0.8063,
+      "step": 40945
+    },
+    {
+      "epoch": 1.5765158806544755,
+      "grad_norm": 1.0555609464645386,
+      "learning_rate": 2.1384835276961767e-05,
+      "loss": 0.7176,
+      "step": 40950
+    },
+    {
+      "epoch": 1.5767083734359961,
+      "grad_norm": 1.4605985879898071,
+      "learning_rate": 2.1369886302441366e-05,
+      "loss": 0.8852,
+      "step": 40955
+    },
+    {
+      "epoch": 1.5769008662175168,
+      "grad_norm": 1.9542622566223145,
+      "learning_rate": 2.1351206555576065e-05,
+      "loss": 0.8881,
+      "step": 40960
+    },
+    {
+      "epoch": 1.5770933589990377,
+      "grad_norm": 0.9330739974975586,
+      "learning_rate": 2.1332534000621972e-05,
+      "loss": 0.701,
+      "step": 40965
+    },
+    {
+      "epoch": 1.5772858517805584,
+      "grad_norm": 1.3597526550292969,
+      "learning_rate": 2.1313868639286494e-05,
+      "loss": 0.928,
+      "step": 40970
+    },
+    {
+      "epoch": 1.577478344562079,
+      "grad_norm": 1.327373743057251,
+      "learning_rate": 2.1295210473276484e-05,
+      "loss": 0.8114,
+      "step": 40975
+    },
+    {
+      "epoch": 1.5776708373435997,
+      "grad_norm": 1.1426680088043213,
+      "learning_rate": 2.12765595042981e-05,
+      "loss": 0.8321,
+      "step": 40980
+    },
+    {
+      "epoch": 1.5778633301251204,
+      "grad_norm": 1.3431953191757202,
+      "learning_rate": 2.1257915734056854e-05,
+      "loss": 0.7583,
+      "step": 40985
+    },
+    {
+      "epoch": 1.578055822906641,
+      "grad_norm": 1.6750779151916504,
+      "learning_rate": 2.123927916425763e-05,
+      "loss": 0.7295,
+      "step": 40990
+    },
+    {
+      "epoch": 1.5782483156881617,
+      "grad_norm": 1.191678762435913,
+      "learning_rate": 2.122064979660455e-05,
+      "loss": 0.731,
+      "step": 40995
+    },
+    {
+      "epoch": 1.5784408084696824,
+      "grad_norm": 1.6331219673156738,
+      "learning_rate": 2.120202763280118e-05,
+      "loss": 0.7803,
+      "step": 41000
+    },
+    {
+      "epoch": 1.578633301251203,
+      "grad_norm": 1.1647868156433105,
+      "learning_rate": 2.1183412674550396e-05,
+      "loss": 0.8047,
+      "step": 41005
+    },
+    {
+      "epoch": 1.5788257940327237,
+      "grad_norm": 1.082160472869873,
+      "learning_rate": 2.1164804923554438e-05,
+      "loss": 0.7405,
+      "step": 41010
+    },
+    {
+      "epoch": 1.5790182868142444,
+      "grad_norm": 2.297400951385498,
+      "learning_rate": 2.1146204381514788e-05,
+      "loss": 1.0183,
+      "step": 41015
+    },
+    {
+      "epoch": 1.579210779595765,
+      "grad_norm": 1.0821828842163086,
+      "learning_rate": 2.1127611050132435e-05,
+      "loss": 0.9013,
+      "step": 41020
+    },
+    {
+      "epoch": 1.5794032723772857,
+      "grad_norm": 0.7487876415252686,
+      "learning_rate": 2.1109024931107547e-05,
+      "loss": 0.7388,
+      "step": 41025
+    },
+    {
+      "epoch": 1.5795957651588064,
+      "grad_norm": 1.8875432014465332,
+      "learning_rate": 2.109044602613971e-05,
+      "loss": 0.9369,
+      "step": 41030
+    },
+    {
+      "epoch": 1.5797882579403273,
+      "grad_norm": 1.49079167842865,
+      "learning_rate": 2.1071874336927876e-05,
+      "loss": 0.8084,
+      "step": 41035
+    },
+    {
+      "epoch": 1.579980750721848,
+      "grad_norm": 1.307343602180481,
+      "learning_rate": 2.10533098651703e-05,
+      "loss": 0.8797,
+      "step": 41040
+    },
+    {
+      "epoch": 1.5801732435033686,
+      "grad_norm": 1.5751981735229492,
+      "learning_rate": 2.1034752612564503e-05,
+      "loss": 0.6406,
+      "step": 41045
+    },
+    {
+      "epoch": 1.5803657362848893,
+      "grad_norm": 1.0212968587875366,
+      "learning_rate": 2.1016202580807544e-05,
+      "loss": 0.8681,
+      "step": 41050
+    },
+    {
+      "epoch": 1.58055822906641,
+      "grad_norm": 0.9244958758354187,
+      "learning_rate": 2.0997659771595612e-05,
+      "loss": 0.76,
+      "step": 41055
+    },
+    {
+      "epoch": 1.5807507218479309,
+      "grad_norm": 2.024540662765503,
+      "learning_rate": 2.0979124186624356e-05,
+      "loss": 0.8869,
+      "step": 41060
+    },
+    {
+      "epoch": 1.5809432146294515,
+      "grad_norm": 1.2309986352920532,
+      "learning_rate": 2.0960595827588713e-05,
+      "loss": 0.8485,
+      "step": 41065
+    },
+    {
+      "epoch": 1.5811357074109722,
+      "grad_norm": 1.1824150085449219,
+      "learning_rate": 2.0942074696183033e-05,
+      "loss": 0.8635,
+      "step": 41070
+    },
+    {
+      "epoch": 1.5813282001924929,
+      "grad_norm": 1.0491746664047241,
+      "learning_rate": 2.092356079410086e-05,
+      "loss": 0.8331,
+      "step": 41075
+    },
+    {
+      "epoch": 1.5815206929740135,
+      "grad_norm": 1.1368030309677124,
+      "learning_rate": 2.090505412303526e-05,
+      "loss": 0.6937,
+      "step": 41080
+    },
+    {
+      "epoch": 1.5817131857555342,
+      "grad_norm": 1.2097387313842773,
+      "learning_rate": 2.0886554684678485e-05,
+      "loss": 0.7318,
+      "step": 41085
+    },
+    {
+      "epoch": 1.5819056785370549,
+      "grad_norm": 2.0431954860687256,
+      "learning_rate": 2.0868062480722206e-05,
+      "loss": 0.8352,
+      "step": 41090
+    },
+    {
+      "epoch": 1.5820981713185756,
+      "grad_norm": 1.763728141784668,
+      "learning_rate": 2.084957751285741e-05,
+      "loss": 0.7479,
+      "step": 41095
+    },
+    {
+      "epoch": 1.5822906641000962,
+      "grad_norm": 1.1527445316314697,
+      "learning_rate": 2.0831099782774455e-05,
+      "loss": 0.8316,
+      "step": 41100
+    },
+    {
+      "epoch": 1.582483156881617,
+      "grad_norm": 1.493467092514038,
+      "learning_rate": 2.0812629292162955e-05,
+      "loss": 0.772,
+      "step": 41105
+    },
+    {
+      "epoch": 1.5826756496631376,
+      "grad_norm": 1.3418041467666626,
+      "learning_rate": 2.0794166042711936e-05,
+      "loss": 0.7486,
+      "step": 41110
+    },
+    {
+      "epoch": 1.5828681424446582,
+      "grad_norm": 0.8369565606117249,
+      "learning_rate": 2.0775710036109762e-05,
+      "loss": 0.7444,
+      "step": 41115
+    },
+    {
+      "epoch": 1.583060635226179,
+      "grad_norm": 1.1210306882858276,
+      "learning_rate": 2.0757261274044048e-05,
+      "loss": 0.7875,
+      "step": 41120
+    },
+    {
+      "epoch": 1.5832531280076996,
+      "grad_norm": 1.1858896017074585,
+      "learning_rate": 2.07388197582019e-05,
+      "loss": 0.7042,
+      "step": 41125
+    },
+    {
+      "epoch": 1.5834456207892202,
+      "grad_norm": 0.9110993146896362,
+      "learning_rate": 2.072038549026961e-05,
+      "loss": 0.7668,
+      "step": 41130
+    },
+    {
+      "epoch": 1.5836381135707411,
+      "grad_norm": 1.5564924478530884,
+      "learning_rate": 2.070195847193288e-05,
+      "loss": 0.7743,
+      "step": 41135
+    },
+    {
+      "epoch": 1.5838306063522618,
+      "grad_norm": 1.6773813962936401,
+      "learning_rate": 2.068353870487675e-05,
+      "loss": 0.7443,
+      "step": 41140
+    },
+    {
+      "epoch": 1.5840230991337825,
+      "grad_norm": 1.0506762266159058,
+      "learning_rate": 2.066512619078561e-05,
+      "loss": 0.681,
+      "step": 41145
+    },
+    {
+      "epoch": 1.5842155919153031,
+      "grad_norm": 1.599478006362915,
+      "learning_rate": 2.0646720931343078e-05,
+      "loss": 0.7801,
+      "step": 41150
+    },
+    {
+      "epoch": 1.5844080846968238,
+      "grad_norm": 0.9615776538848877,
+      "learning_rate": 2.06283229282323e-05,
+      "loss": 0.6929,
+      "step": 41155
+    },
+    {
+      "epoch": 1.5846005774783447,
+      "grad_norm": 1.4779179096221924,
+      "learning_rate": 2.0609932183135582e-05,
+      "loss": 0.8112,
+      "step": 41160
+    },
+    {
+      "epoch": 1.5847930702598654,
+      "grad_norm": 1.1211433410644531,
+      "learning_rate": 2.059154869773465e-05,
+      "loss": 0.7598,
+      "step": 41165
+    },
+    {
+      "epoch": 1.584985563041386,
+      "grad_norm": 1.0643943548202515,
+      "learning_rate": 2.0573172473710567e-05,
+      "loss": 0.8607,
+      "step": 41170
+    },
+    {
+      "epoch": 1.5851780558229067,
+      "grad_norm": 2.5210201740264893,
+      "learning_rate": 2.0554803512743724e-05,
+      "loss": 0.9052,
+      "step": 41175
+    },
+    {
+      "epoch": 1.5853705486044274,
+      "grad_norm": 1.6734479665756226,
+      "learning_rate": 2.053644181651376e-05,
+      "loss": 0.8745,
+      "step": 41180
+    },
+    {
+      "epoch": 1.585563041385948,
+      "grad_norm": 1.5669529438018799,
+      "learning_rate": 2.051808738669987e-05,
+      "loss": 0.7996,
+      "step": 41185
+    },
+    {
+      "epoch": 1.5857555341674687,
+      "grad_norm": 1.2366806268692017,
+      "learning_rate": 2.0499740224980325e-05,
+      "loss": 0.719,
+      "step": 41190
+    },
+    {
+      "epoch": 1.5859480269489894,
+      "grad_norm": 1.0737662315368652,
+      "learning_rate": 2.0481400333032897e-05,
+      "loss": 0.6766,
+      "step": 41195
+    },
+    {
+      "epoch": 1.58614051973051,
+      "grad_norm": 2.486666679382324,
+      "learning_rate": 2.0463067712534656e-05,
+      "loss": 0.9854,
+      "step": 41200
+    },
+    {
+      "epoch": 1.5863330125120307,
+      "grad_norm": 1.0239930152893066,
+      "learning_rate": 2.0444742365162005e-05,
+      "loss": 0.8435,
+      "step": 41205
+    },
+    {
+      "epoch": 1.5865255052935514,
+      "grad_norm": 1.0103557109832764,
+      "learning_rate": 2.042642429259064e-05,
+      "loss": 0.8264,
+      "step": 41210
+    },
+    {
+      "epoch": 1.586717998075072,
+      "grad_norm": 1.7183647155761719,
+      "learning_rate": 2.040811349649564e-05,
+      "loss": 0.781,
+      "step": 41215
+    },
+    {
+      "epoch": 1.5869104908565927,
+      "grad_norm": 1.1959103345870972,
+      "learning_rate": 2.038980997855142e-05,
+      "loss": 0.7821,
+      "step": 41220
+    },
+    {
+      "epoch": 1.5871029836381134,
+      "grad_norm": 1.4438652992248535,
+      "learning_rate": 2.03715137404317e-05,
+      "loss": 0.8455,
+      "step": 41225
+    },
+    {
+      "epoch": 1.5872954764196343,
+      "grad_norm": 1.0840461254119873,
+      "learning_rate": 2.0353224783809587e-05,
+      "loss": 0.7846,
+      "step": 41230
+    },
+    {
+      "epoch": 1.587487969201155,
+      "grad_norm": 1.6811954975128174,
+      "learning_rate": 2.033494311035744e-05,
+      "loss": 0.8039,
+      "step": 41235
+    },
+    {
+      "epoch": 1.5876804619826757,
+      "grad_norm": 1.2145577669143677,
+      "learning_rate": 2.0316668721747e-05,
+      "loss": 0.8147,
+      "step": 41240
+    },
+    {
+      "epoch": 1.5878729547641963,
+      "grad_norm": 1.1736549139022827,
+      "learning_rate": 2.0298401619649353e-05,
+      "loss": 1.0046,
+      "step": 41245
+    },
+    {
+      "epoch": 1.588065447545717,
+      "grad_norm": 1.1645379066467285,
+      "learning_rate": 2.028014180573491e-05,
+      "loss": 0.8703,
+      "step": 41250
+    },
+    {
+      "epoch": 1.5882579403272379,
+      "grad_norm": 1.2913013696670532,
+      "learning_rate": 2.0261889281673397e-05,
+      "loss": 0.8041,
+      "step": 41255
+    },
+    {
+      "epoch": 1.5884504331087586,
+      "grad_norm": 1.1715826988220215,
+      "learning_rate": 2.0243644049133926e-05,
+      "loss": 0.8894,
+      "step": 41260
+    },
+    {
+      "epoch": 1.5886429258902792,
+      "grad_norm": 1.0594674348831177,
+      "learning_rate": 2.0225406109784805e-05,
+      "loss": 0.8548,
+      "step": 41265
+    },
+    {
+      "epoch": 1.5888354186718,
+      "grad_norm": 2.4181666374206543,
+      "learning_rate": 2.0207175465293904e-05,
+      "loss": 0.8165,
+      "step": 41270
+    },
+    {
+      "epoch": 1.5890279114533206,
+      "grad_norm": 0.9365816712379456,
+      "learning_rate": 2.0188952117328186e-05,
+      "loss": 0.8741,
+      "step": 41275
+    },
+    {
+      "epoch": 1.5892204042348412,
+      "grad_norm": 2.1108558177948,
+      "learning_rate": 2.0170736067554108e-05,
+      "loss": 0.8202,
+      "step": 41280
+    },
+    {
+      "epoch": 1.589412897016362,
+      "grad_norm": 1.5129166841506958,
+      "learning_rate": 2.0152527317637394e-05,
+      "loss": 0.7616,
+      "step": 41285
+    },
+    {
+      "epoch": 1.5896053897978826,
+      "grad_norm": 1.32972252368927,
+      "learning_rate": 2.0134325869243144e-05,
+      "loss": 0.905,
+      "step": 41290
+    },
+    {
+      "epoch": 1.5897978825794032,
+      "grad_norm": 0.950168788433075,
+      "learning_rate": 2.011613172403567e-05,
+      "loss": 0.6316,
+      "step": 41295
+    },
+    {
+      "epoch": 1.589990375360924,
+      "grad_norm": 0.9195626378059387,
+      "learning_rate": 2.009794488367882e-05,
+      "loss": 0.8385,
+      "step": 41300
+    },
+    {
+      "epoch": 1.5901828681424446,
+      "grad_norm": 1.176831603050232,
+      "learning_rate": 2.007976534983559e-05,
+      "loss": 0.8977,
+      "step": 41305
+    },
+    {
+      "epoch": 1.5903753609239653,
+      "grad_norm": 1.0982027053833008,
+      "learning_rate": 2.0061593124168398e-05,
+      "loss": 0.8981,
+      "step": 41310
+    },
+    {
+      "epoch": 1.590567853705486,
+      "grad_norm": 1.0368854999542236,
+      "learning_rate": 2.0043428208338987e-05,
+      "loss": 0.6795,
+      "step": 41315
+    },
+    {
+      "epoch": 1.5907603464870066,
+      "grad_norm": 0.7842391729354858,
+      "learning_rate": 2.0025270604008384e-05,
+      "loss": 0.6644,
+      "step": 41320
+    },
+    {
+      "epoch": 1.5909528392685275,
+      "grad_norm": 1.4070217609405518,
+      "learning_rate": 2.0007120312837e-05,
+      "loss": 0.7587,
+      "step": 41325
+    },
+    {
+      "epoch": 1.5911453320500482,
+      "grad_norm": 1.3950802087783813,
+      "learning_rate": 1.998897733648456e-05,
+      "loss": 0.8808,
+      "step": 41330
+    },
+    {
+      "epoch": 1.5913378248315688,
+      "grad_norm": 0.873643696308136,
+      "learning_rate": 1.9970841676610143e-05,
+      "loss": 0.8078,
+      "step": 41335
+    },
+    {
+      "epoch": 1.5915303176130895,
+      "grad_norm": 1.5144951343536377,
+      "learning_rate": 1.995271333487205e-05,
+      "loss": 0.7164,
+      "step": 41340
+    },
+    {
+      "epoch": 1.5917228103946102,
+      "grad_norm": 1.031936526298523,
+      "learning_rate": 1.993459231292811e-05,
+      "loss": 0.6438,
+      "step": 41345
+    },
+    {
+      "epoch": 1.591915303176131,
+      "grad_norm": 1.1276479959487915,
+      "learning_rate": 1.991647861243531e-05,
+      "loss": 0.8856,
+      "step": 41350
+    },
+    {
+      "epoch": 1.5921077959576517,
+      "grad_norm": 0.9697341322898865,
+      "learning_rate": 1.9898372235050022e-05,
+      "loss": 0.788,
+      "step": 41355
+    },
+    {
+      "epoch": 1.5923002887391724,
+      "grad_norm": 1.0358269214630127,
+      "learning_rate": 1.9880273182427965e-05,
+      "loss": 0.7337,
+      "step": 41360
+    },
+    {
+      "epoch": 1.592492781520693,
+      "grad_norm": 1.1483445167541504,
+      "learning_rate": 1.9862181456224216e-05,
+      "loss": 0.8329,
+      "step": 41365
+    },
+    {
+      "epoch": 1.5926852743022137,
+      "grad_norm": 0.7864820957183838,
+      "learning_rate": 1.9844097058093047e-05,
+      "loss": 0.664,
+      "step": 41370
+    },
+    {
+      "epoch": 1.5928777670837344,
+      "grad_norm": 0.8901979327201843,
+      "learning_rate": 1.9826019989688283e-05,
+      "loss": 0.7352,
+      "step": 41375
+    },
+    {
+      "epoch": 1.593070259865255,
+      "grad_norm": 1.5492757558822632,
+      "learning_rate": 1.9807950252662854e-05,
+      "loss": 0.8297,
+      "step": 41380
+    },
+    {
+      "epoch": 1.5932627526467757,
+      "grad_norm": 1.5439116954803467,
+      "learning_rate": 1.9789887848669143e-05,
+      "loss": 0.7502,
+      "step": 41385
+    },
+    {
+      "epoch": 1.5934552454282964,
+      "grad_norm": 1.6057932376861572,
+      "learning_rate": 1.9771832779358857e-05,
+      "loss": 0.8005,
+      "step": 41390
+    },
+    {
+      "epoch": 1.593647738209817,
+      "grad_norm": 1.489683985710144,
+      "learning_rate": 1.9753785046383022e-05,
+      "loss": 0.8509,
+      "step": 41395
+    },
+    {
+      "epoch": 1.5938402309913378,
+      "grad_norm": 1.2398594617843628,
+      "learning_rate": 1.9735744651391906e-05,
+      "loss": 0.839,
+      "step": 41400
+    },
+    {
+      "epoch": 1.5940327237728584,
+      "grad_norm": 0.5610536932945251,
+      "learning_rate": 1.9717711596035292e-05,
+      "loss": 0.686,
+      "step": 41405
+    },
+    {
+      "epoch": 1.594225216554379,
+      "grad_norm": 1.1076955795288086,
+      "learning_rate": 1.9699685881962115e-05,
+      "loss": 0.7216,
+      "step": 41410
+    },
+    {
+      "epoch": 1.5944177093358998,
+      "grad_norm": 0.8716223835945129,
+      "learning_rate": 1.9681667510820713e-05,
+      "loss": 0.8062,
+      "step": 41415
+    },
+    {
+      "epoch": 1.5946102021174204,
+      "grad_norm": 0.9298526048660278,
+      "learning_rate": 1.9663656484258764e-05,
+      "loss": 0.8884,
+      "step": 41420
+    },
+    {
+      "epoch": 1.5948026948989413,
+      "grad_norm": 1.1228723526000977,
+      "learning_rate": 1.9645652803923266e-05,
+      "loss": 0.7249,
+      "step": 41425
+    },
+    {
+      "epoch": 1.594995187680462,
+      "grad_norm": 1.0350871086120605,
+      "learning_rate": 1.9627656471460498e-05,
+      "loss": 0.7171,
+      "step": 41430
+    },
+    {
+      "epoch": 1.5951876804619827,
+      "grad_norm": 1.5256117582321167,
+      "learning_rate": 1.9609667488516138e-05,
+      "loss": 1.0133,
+      "step": 41435
+    },
+    {
+      "epoch": 1.5953801732435033,
+      "grad_norm": 1.4316469430923462,
+      "learning_rate": 1.9591685856735144e-05,
+      "loss": 0.7591,
+      "step": 41440
+    },
+    {
+      "epoch": 1.595572666025024,
+      "grad_norm": 1.8641667366027832,
+      "learning_rate": 1.9573711577761812e-05,
+      "loss": 0.8817,
+      "step": 41445
+    },
+    {
+      "epoch": 1.595765158806545,
+      "grad_norm": 1.348741054534912,
+      "learning_rate": 1.9555744653239815e-05,
+      "loss": 0.9214,
+      "step": 41450
+    },
+    {
+      "epoch": 1.5959576515880656,
+      "grad_norm": 1.2496349811553955,
+      "learning_rate": 1.9537785084812044e-05,
+      "loss": 0.7907,
+      "step": 41455
+    },
+    {
+      "epoch": 1.5961501443695862,
+      "grad_norm": 1.1491204500198364,
+      "learning_rate": 1.9519832874120824e-05,
+      "loss": 0.8712,
+      "step": 41460
+    },
+    {
+      "epoch": 1.596342637151107,
+      "grad_norm": 0.7588009238243103,
+      "learning_rate": 1.9501888022807745e-05,
+      "loss": 0.7473,
+      "step": 41465
+    },
+    {
+      "epoch": 1.5965351299326276,
+      "grad_norm": 1.0782458782196045,
+      "learning_rate": 1.9483950532513783e-05,
+      "loss": 0.7309,
+      "step": 41470
+    },
+    {
+      "epoch": 1.5967276227141483,
+      "grad_norm": 1.52632474899292,
+      "learning_rate": 1.9466020404879127e-05,
+      "loss": 0.8698,
+      "step": 41475
+    },
+    {
+      "epoch": 1.596920115495669,
+      "grad_norm": 1.286037564277649,
+      "learning_rate": 1.9448097641543462e-05,
+      "loss": 0.8768,
+      "step": 41480
+    },
+    {
+      "epoch": 1.5971126082771896,
+      "grad_norm": 0.8476093411445618,
+      "learning_rate": 1.9430182244145646e-05,
+      "loss": 0.9017,
+      "step": 41485
+    },
+    {
+      "epoch": 1.5973051010587103,
+      "grad_norm": 1.0609407424926758,
+      "learning_rate": 1.9412274214323923e-05,
+      "loss": 0.8112,
+      "step": 41490
+    },
+    {
+      "epoch": 1.597497593840231,
+      "grad_norm": 1.7310614585876465,
+      "learning_rate": 1.9394373553715885e-05,
+      "loss": 1.0244,
+      "step": 41495
+    },
+    {
+      "epoch": 1.5976900866217516,
+      "grad_norm": 1.7635456323623657,
+      "learning_rate": 1.9376480263958453e-05,
+      "loss": 0.8217,
+      "step": 41500
+    },
+    {
+      "epoch": 1.5978825794032723,
+      "grad_norm": 0.9183743000030518,
+      "learning_rate": 1.9358594346687765e-05,
+      "loss": 0.8512,
+      "step": 41505
+    },
+    {
+      "epoch": 1.598075072184793,
+      "grad_norm": 1.0998347997665405,
+      "learning_rate": 1.9340715803539466e-05,
+      "loss": 0.7804,
+      "step": 41510
+    },
+    {
+      "epoch": 1.5982675649663136,
+      "grad_norm": 1.388878345489502,
+      "learning_rate": 1.9322844636148375e-05,
+      "loss": 0.8285,
+      "step": 41515
+    },
+    {
+      "epoch": 1.5984600577478345,
+      "grad_norm": 1.1632678508758545,
+      "learning_rate": 1.930498084614869e-05,
+      "loss": 0.7176,
+      "step": 41520
+    },
+    {
+      "epoch": 1.5986525505293552,
+      "grad_norm": 0.8704983592033386,
+      "learning_rate": 1.9287124435173964e-05,
+      "loss": 0.7018,
+      "step": 41525
+    },
+    {
+      "epoch": 1.5988450433108758,
+      "grad_norm": 1.2163587808609009,
+      "learning_rate": 1.9269275404857022e-05,
+      "loss": 0.6076,
+      "step": 41530
+    },
+    {
+      "epoch": 1.5990375360923965,
+      "grad_norm": 0.8491012454032898,
+      "learning_rate": 1.9251433756830095e-05,
+      "loss": 0.7527,
+      "step": 41535
+    },
+    {
+      "epoch": 1.5992300288739172,
+      "grad_norm": 0.996997058391571,
+      "learning_rate": 1.9233599492724607e-05,
+      "loss": 0.6848,
+      "step": 41540
+    },
+    {
+      "epoch": 1.599422521655438,
+      "grad_norm": 1.3385977745056152,
+      "learning_rate": 1.9215772614171413e-05,
+      "loss": 0.8753,
+      "step": 41545
+    },
+    {
+      "epoch": 1.5996150144369587,
+      "grad_norm": 0.9963300228118896,
+      "learning_rate": 1.919795312280067e-05,
+      "loss": 0.8053,
+      "step": 41550
+    },
+    {
+      "epoch": 1.5998075072184794,
+      "grad_norm": 0.8091634511947632,
+      "learning_rate": 1.918014102024187e-05,
+      "loss": 0.8183,
+      "step": 41555
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6417735815048218,
+      "learning_rate": 1.916233630812374e-05,
+      "loss": 0.7796,
+      "step": 41560
+    },
+    {
+      "epoch": 1.6001924927815208,
+      "grad_norm": 0.8539758324623108,
+      "learning_rate": 1.914453898807451e-05,
+      "loss": 0.7709,
+      "step": 41565
+    },
+    {
+      "epoch": 1.6003849855630414,
+      "grad_norm": 1.14285409450531,
+      "learning_rate": 1.912674906172155e-05,
+      "loss": 0.762,
+      "step": 41570
+    },
+    {
+      "epoch": 1.600577478344562,
+      "grad_norm": 1.2252355813980103,
+      "learning_rate": 1.910896653069165e-05,
+      "loss": 0.9033,
+      "step": 41575
+    },
+    {
+      "epoch": 1.6007699711260828,
+      "grad_norm": 1.3306511640548706,
+      "learning_rate": 1.9091191396610895e-05,
+      "loss": 0.8857,
+      "step": 41580
+    },
+    {
+      "epoch": 1.6009624639076034,
+      "grad_norm": 0.9147464632987976,
+      "learning_rate": 1.9073423661104762e-05,
+      "loss": 0.7447,
+      "step": 41585
+    },
+    {
+      "epoch": 1.601154956689124,
+      "grad_norm": 1.4393258094787598,
+      "learning_rate": 1.9055663325797877e-05,
+      "loss": 0.7681,
+      "step": 41590
+    },
+    {
+      "epoch": 1.6013474494706448,
+      "grad_norm": 1.5441241264343262,
+      "learning_rate": 1.903791039231443e-05,
+      "loss": 0.9034,
+      "step": 41595
+    },
+    {
+      "epoch": 1.6015399422521654,
+      "grad_norm": 1.5993388891220093,
+      "learning_rate": 1.9020164862277724e-05,
+      "loss": 0.7531,
+      "step": 41600
+    },
+    {
+      "epoch": 1.6017324350336861,
+      "grad_norm": 1.3545401096343994,
+      "learning_rate": 1.900242673731051e-05,
+      "loss": 0.7417,
+      "step": 41605
+    },
+    {
+      "epoch": 1.6019249278152068,
+      "grad_norm": 2.0168793201446533,
+      "learning_rate": 1.89846960190348e-05,
+      "loss": 0.83,
+      "step": 41610
+    },
+    {
+      "epoch": 1.6021174205967275,
+      "grad_norm": 0.8939989805221558,
+      "learning_rate": 1.8966972709071985e-05,
+      "loss": 0.7973,
+      "step": 41615
+    },
+    {
+      "epoch": 1.6023099133782484,
+      "grad_norm": 1.6125478744506836,
+      "learning_rate": 1.894925680904268e-05,
+      "loss": 0.8518,
+      "step": 41620
+    },
+    {
+      "epoch": 1.602502406159769,
+      "grad_norm": 1.4295742511749268,
+      "learning_rate": 1.8931548320566972e-05,
+      "loss": 0.782,
+      "step": 41625
+    },
+    {
+      "epoch": 1.6026948989412897,
+      "grad_norm": 0.9275886416435242,
+      "learning_rate": 1.8913847245264116e-05,
+      "loss": 0.8498,
+      "step": 41630
+    },
+    {
+      "epoch": 1.6028873917228104,
+      "grad_norm": 1.2459815740585327,
+      "learning_rate": 1.8896153584752785e-05,
+      "loss": 0.8219,
+      "step": 41635
+    },
+    {
+      "epoch": 1.6030798845043313,
+      "grad_norm": 1.0280636548995972,
+      "learning_rate": 1.887846734065094e-05,
+      "loss": 0.7993,
+      "step": 41640
+    },
+    {
+      "epoch": 1.603272377285852,
+      "grad_norm": 0.9304599761962891,
+      "learning_rate": 1.886078851457591e-05,
+      "loss": 0.8449,
+      "step": 41645
+    },
+    {
+      "epoch": 1.6034648700673726,
+      "grad_norm": 1.3438136577606201,
+      "learning_rate": 1.884311710814425e-05,
+      "loss": 0.706,
+      "step": 41650
+    },
+    {
+      "epoch": 1.6036573628488933,
+      "grad_norm": 1.3000657558441162,
+      "learning_rate": 1.8825453122971904e-05,
+      "loss": 0.8801,
+      "step": 41655
+    },
+    {
+      "epoch": 1.603849855630414,
+      "grad_norm": 0.9097469449043274,
+      "learning_rate": 1.880779656067414e-05,
+      "loss": 0.7417,
+      "step": 41660
+    },
+    {
+      "epoch": 1.6040423484119346,
+      "grad_norm": 1.5576119422912598,
+      "learning_rate": 1.8790147422865532e-05,
+      "loss": 0.8143,
+      "step": 41665
+    },
+    {
+      "epoch": 1.6042348411934553,
+      "grad_norm": 1.0649263858795166,
+      "learning_rate": 1.8772505711160003e-05,
+      "loss": 0.6215,
+      "step": 41670
+    },
+    {
+      "epoch": 1.604427333974976,
+      "grad_norm": 1.2119357585906982,
+      "learning_rate": 1.8754871427170716e-05,
+      "loss": 0.7959,
+      "step": 41675
+    },
+    {
+      "epoch": 1.6046198267564966,
+      "grad_norm": 1.5147475004196167,
+      "learning_rate": 1.8737244572510238e-05,
+      "loss": 0.6919,
+      "step": 41680
+    },
+    {
+      "epoch": 1.6048123195380173,
+      "grad_norm": 1.540199637413025,
+      "learning_rate": 1.8719625148790432e-05,
+      "loss": 0.9114,
+      "step": 41685
+    },
+    {
+      "epoch": 1.605004812319538,
+      "grad_norm": 1.3708090782165527,
+      "learning_rate": 1.8702013157622488e-05,
+      "loss": 0.7575,
+      "step": 41690
+    },
+    {
+      "epoch": 1.6051973051010586,
+      "grad_norm": 0.8762415647506714,
+      "learning_rate": 1.8684408600616855e-05,
+      "loss": 0.7135,
+      "step": 41695
+    },
+    {
+      "epoch": 1.6053897978825793,
+      "grad_norm": 1.0719411373138428,
+      "learning_rate": 1.866681147938343e-05,
+      "loss": 0.7409,
+      "step": 41700
+    },
+    {
+      "epoch": 1.6055822906641,
+      "grad_norm": 1.3425824642181396,
+      "learning_rate": 1.864922179553128e-05,
+      "loss": 1.0134,
+      "step": 41705
+    },
+    {
+      "epoch": 1.6057747834456206,
+      "grad_norm": 1.2449208498001099,
+      "learning_rate": 1.8631639550668912e-05,
+      "loss": 0.8516,
+      "step": 41710
+    },
+    {
+      "epoch": 1.6059672762271415,
+      "grad_norm": 0.9909605979919434,
+      "learning_rate": 1.861406474640408e-05,
+      "loss": 0.8218,
+      "step": 41715
+    },
+    {
+      "epoch": 1.6061597690086622,
+      "grad_norm": 1.1478825807571411,
+      "learning_rate": 1.8596497384343926e-05,
+      "loss": 0.9639,
+      "step": 41720
+    },
+    {
+      "epoch": 1.6063522617901829,
+      "grad_norm": 1.305647611618042,
+      "learning_rate": 1.857893746609478e-05,
+      "loss": 0.8432,
+      "step": 41725
+    },
+    {
+      "epoch": 1.6065447545717035,
+      "grad_norm": 1.028073787689209,
+      "learning_rate": 1.8561384993262497e-05,
+      "loss": 0.7462,
+      "step": 41730
+    },
+    {
+      "epoch": 1.6067372473532242,
+      "grad_norm": 1.058986783027649,
+      "learning_rate": 1.8543839967452047e-05,
+      "loss": 0.8153,
+      "step": 41735
+    },
+    {
+      "epoch": 1.606929740134745,
+      "grad_norm": 1.5460129976272583,
+      "learning_rate": 1.8526302390267836e-05,
+      "loss": 0.7346,
+      "step": 41740
+    },
+    {
+      "epoch": 1.6071222329162658,
+      "grad_norm": 1.0355900526046753,
+      "learning_rate": 1.8508772263313556e-05,
+      "loss": 0.8171,
+      "step": 41745
+    },
+    {
+      "epoch": 1.6073147256977864,
+      "grad_norm": 1.0379014015197754,
+      "learning_rate": 1.849124958819224e-05,
+      "loss": 0.7432,
+      "step": 41750
+    },
+    {
+      "epoch": 1.607507218479307,
+      "grad_norm": 1.5380711555480957,
+      "learning_rate": 1.847373436650619e-05,
+      "loss": 0.9295,
+      "step": 41755
+    },
+    {
+      "epoch": 1.6076997112608278,
+      "grad_norm": 1.3490667343139648,
+      "learning_rate": 1.8456226599857064e-05,
+      "loss": 0.9094,
+      "step": 41760
+    },
+    {
+      "epoch": 1.6078922040423484,
+      "grad_norm": 1.5740834474563599,
+      "learning_rate": 1.8438726289845833e-05,
+      "loss": 0.7674,
+      "step": 41765
+    },
+    {
+      "epoch": 1.6080846968238691,
+      "grad_norm": 0.9285767674446106,
+      "learning_rate": 1.8421233438072795e-05,
+      "loss": 0.7368,
+      "step": 41770
+    },
+    {
+      "epoch": 1.6082771896053898,
+      "grad_norm": 1.3774783611297607,
+      "learning_rate": 1.840374804613757e-05,
+      "loss": 0.7023,
+      "step": 41775
+    },
+    {
+      "epoch": 1.6084696823869105,
+      "grad_norm": 1.00751793384552,
+      "learning_rate": 1.8386270115639013e-05,
+      "loss": 0.7488,
+      "step": 41780
+    },
+    {
+      "epoch": 1.6086621751684311,
+      "grad_norm": 1.7420424222946167,
+      "learning_rate": 1.836879964817546e-05,
+      "loss": 0.9531,
+      "step": 41785
+    },
+    {
+      "epoch": 1.6088546679499518,
+      "grad_norm": 1.4065346717834473,
+      "learning_rate": 1.8351336645344408e-05,
+      "loss": 0.8083,
+      "step": 41790
+    },
+    {
+      "epoch": 1.6090471607314725,
+      "grad_norm": 1.2782877683639526,
+      "learning_rate": 1.8333881108742736e-05,
+      "loss": 0.9427,
+      "step": 41795
+    },
+    {
+      "epoch": 1.6092396535129931,
+      "grad_norm": 1.675103783607483,
+      "learning_rate": 1.8316433039966653e-05,
+      "loss": 0.9521,
+      "step": 41800
+    },
+    {
+      "epoch": 1.6094321462945138,
+      "grad_norm": 1.5068132877349854,
+      "learning_rate": 1.8298992440611686e-05,
+      "loss": 0.8201,
+      "step": 41805
+    },
+    {
+      "epoch": 1.6096246390760347,
+      "grad_norm": 1.5667356252670288,
+      "learning_rate": 1.828155931227259e-05,
+      "loss": 0.882,
+      "step": 41810
+    },
+    {
+      "epoch": 1.6098171318575554,
+      "grad_norm": 1.710042119026184,
+      "learning_rate": 1.8264133656543613e-05,
+      "loss": 0.867,
+      "step": 41815
+    },
+    {
+      "epoch": 1.610009624639076,
+      "grad_norm": 1.5348536968231201,
+      "learning_rate": 1.824671547501814e-05,
+      "loss": 0.8304,
+      "step": 41820
+    },
+    {
+      "epoch": 1.6102021174205967,
+      "grad_norm": 1.3562618494033813,
+      "learning_rate": 1.8229304769288956e-05,
+      "loss": 0.8852,
+      "step": 41825
+    },
+    {
+      "epoch": 1.6103946102021174,
+      "grad_norm": 1.1753439903259277,
+      "learning_rate": 1.8211901540948183e-05,
+      "loss": 0.8705,
+      "step": 41830
+    },
+    {
+      "epoch": 1.6105871029836383,
+      "grad_norm": 0.7911993265151978,
+      "learning_rate": 1.8194505791587245e-05,
+      "loss": 0.642,
+      "step": 41835
+    },
+    {
+      "epoch": 1.610779595765159,
+      "grad_norm": 1.2742297649383545,
+      "learning_rate": 1.8177117522796784e-05,
+      "loss": 0.7778,
+      "step": 41840
+    },
+    {
+      "epoch": 1.6109720885466796,
+      "grad_norm": 0.5650262236595154,
+      "learning_rate": 1.8159736736166943e-05,
+      "loss": 0.6833,
+      "step": 41845
+    },
+    {
+      "epoch": 1.6111645813282003,
+      "grad_norm": 0.8346420526504517,
+      "learning_rate": 1.8142363433287026e-05,
+      "loss": 0.7115,
+      "step": 41850
+    },
+    {
+      "epoch": 1.611357074109721,
+      "grad_norm": 2.0539052486419678,
+      "learning_rate": 1.812499761574571e-05,
+      "loss": 0.9054,
+      "step": 41855
+    },
+    {
+      "epoch": 1.6115495668912416,
+      "grad_norm": 1.3676276206970215,
+      "learning_rate": 1.8107639285131005e-05,
+      "loss": 0.7522,
+      "step": 41860
+    },
+    {
+      "epoch": 1.6117420596727623,
+      "grad_norm": 0.8344974517822266,
+      "learning_rate": 1.809028844303018e-05,
+      "loss": 0.7912,
+      "step": 41865
+    },
+    {
+      "epoch": 1.611934552454283,
+      "grad_norm": 2.75577449798584,
+      "learning_rate": 1.807294509102988e-05,
+      "loss": 0.7293,
+      "step": 41870
+    },
+    {
+      "epoch": 1.6121270452358036,
+      "grad_norm": 1.7650033235549927,
+      "learning_rate": 1.8055609230716032e-05,
+      "loss": 0.7737,
+      "step": 41875
+    },
+    {
+      "epoch": 1.6123195380173243,
+      "grad_norm": 1.5389347076416016,
+      "learning_rate": 1.8038280863673907e-05,
+      "loss": 0.7563,
+      "step": 41880
+    },
+    {
+      "epoch": 1.612512030798845,
+      "grad_norm": 1.5038697719573975,
+      "learning_rate": 1.8020959991488006e-05,
+      "loss": 0.7828,
+      "step": 41885
+    },
+    {
+      "epoch": 1.6127045235803656,
+      "grad_norm": 1.3098371028900146,
+      "learning_rate": 1.8003646615742308e-05,
+      "loss": 0.7512,
+      "step": 41890
+    },
+    {
+      "epoch": 1.6128970163618863,
+      "grad_norm": 1.2375686168670654,
+      "learning_rate": 1.7986340738019912e-05,
+      "loss": 0.8669,
+      "step": 41895
+    },
+    {
+      "epoch": 1.613089509143407,
+      "grad_norm": 1.7652937173843384,
+      "learning_rate": 1.7969042359903376e-05,
+      "loss": 0.6875,
+      "step": 41900
+    },
+    {
+      "epoch": 1.6132820019249277,
+      "grad_norm": 1.404327630996704,
+      "learning_rate": 1.795175148297451e-05,
+      "loss": 0.786,
+      "step": 41905
+    },
+    {
+      "epoch": 1.6134744947064485,
+      "grad_norm": 1.2358356714248657,
+      "learning_rate": 1.7934468108814472e-05,
+      "loss": 0.7852,
+      "step": 41910
+    },
+    {
+      "epoch": 1.6136669874879692,
+      "grad_norm": 1.1059602499008179,
+      "learning_rate": 1.7917192239003644e-05,
+      "loss": 0.7523,
+      "step": 41915
+    },
+    {
+      "epoch": 1.6138594802694899,
+      "grad_norm": 1.4840894937515259,
+      "learning_rate": 1.7899923875121882e-05,
+      "loss": 0.7958,
+      "step": 41920
+    },
+    {
+      "epoch": 1.6140519730510106,
+      "grad_norm": 1.5999846458435059,
+      "learning_rate": 1.7882663018748193e-05,
+      "loss": 0.8017,
+      "step": 41925
+    },
+    {
+      "epoch": 1.6142444658325312,
+      "grad_norm": 1.9015750885009766,
+      "learning_rate": 1.7865409671460996e-05,
+      "loss": 0.8296,
+      "step": 41930
+    },
+    {
+      "epoch": 1.6144369586140521,
+      "grad_norm": 1.0894391536712646,
+      "learning_rate": 1.7848163834837995e-05,
+      "loss": 0.8116,
+      "step": 41935
+    },
+    {
+      "epoch": 1.6146294513955728,
+      "grad_norm": 1.131320834159851,
+      "learning_rate": 1.783092551045623e-05,
+      "loss": 0.7442,
+      "step": 41940
+    },
+    {
+      "epoch": 1.6148219441770935,
+      "grad_norm": 2.2992284297943115,
+      "learning_rate": 1.781369469989196e-05,
+      "loss": 0.8746,
+      "step": 41945
+    },
+    {
+      "epoch": 1.6150144369586141,
+      "grad_norm": 1.1256681680679321,
+      "learning_rate": 1.7796471404720916e-05,
+      "loss": 0.721,
+      "step": 41950
+    },
+    {
+      "epoch": 1.6152069297401348,
+      "grad_norm": 1.115007758140564,
+      "learning_rate": 1.7779255626518e-05,
+      "loss": 0.7711,
+      "step": 41955
+    },
+    {
+      "epoch": 1.6153994225216555,
+      "grad_norm": 1.9384006261825562,
+      "learning_rate": 1.7762047366857483e-05,
+      "loss": 1.0144,
+      "step": 41960
+    },
+    {
+      "epoch": 1.6155919153031761,
+      "grad_norm": 1.0422606468200684,
+      "learning_rate": 1.7744846627312962e-05,
+      "loss": 0.8137,
+      "step": 41965
+    },
+    {
+      "epoch": 1.6157844080846968,
+      "grad_norm": 0.9894946217536926,
+      "learning_rate": 1.7727653409457358e-05,
+      "loss": 0.7198,
+      "step": 41970
+    },
+    {
+      "epoch": 1.6159769008662175,
+      "grad_norm": 1.2634601593017578,
+      "learning_rate": 1.771046771486281e-05,
+      "loss": 0.9481,
+      "step": 41975
+    },
+    {
+      "epoch": 1.6161693936477382,
+      "grad_norm": 1.1557183265686035,
+      "learning_rate": 1.7693289545100876e-05,
+      "loss": 0.8676,
+      "step": 41980
+    },
+    {
+      "epoch": 1.6163618864292588,
+      "grad_norm": 1.1935524940490723,
+      "learning_rate": 1.767611890174238e-05,
+      "loss": 0.8162,
+      "step": 41985
+    },
+    {
+      "epoch": 1.6165543792107795,
+      "grad_norm": 1.1359078884124756,
+      "learning_rate": 1.7658955786357455e-05,
+      "loss": 0.8553,
+      "step": 41990
+    },
+    {
+      "epoch": 1.6167468719923002,
+      "grad_norm": 1.311218500137329,
+      "learning_rate": 1.76418002005156e-05,
+      "loss": 0.745,
+      "step": 41995
+    },
+    {
+      "epoch": 1.6169393647738208,
+      "grad_norm": 1.595218300819397,
+      "learning_rate": 1.7624652145785523e-05,
+      "loss": 0.6425,
+      "step": 42000
+    },
+    {
+      "epoch": 1.6171318575553417,
+      "grad_norm": 1.5510958433151245,
+      "learning_rate": 1.7607511623735317e-05,
+      "loss": 0.7513,
+      "step": 42005
+    },
+    {
+      "epoch": 1.6173243503368624,
+      "grad_norm": 1.0332719087600708,
+      "learning_rate": 1.759037863593237e-05,
+      "loss": 0.7869,
+      "step": 42010
+    },
+    {
+      "epoch": 1.617516843118383,
+      "grad_norm": 1.556477427482605,
+      "learning_rate": 1.7573253183943404e-05,
+      "loss": 0.6109,
+      "step": 42015
+    },
+    {
+      "epoch": 1.6177093358999037,
+      "grad_norm": 3.525594472885132,
+      "learning_rate": 1.755613526933436e-05,
+      "loss": 0.9172,
+      "step": 42020
+    },
+    {
+      "epoch": 1.6179018286814244,
+      "grad_norm": 2.301964044570923,
+      "learning_rate": 1.7539024893670664e-05,
+      "loss": 1.1278,
+      "step": 42025
+    },
+    {
+      "epoch": 1.6180943214629453,
+      "grad_norm": 0.9457361698150635,
+      "learning_rate": 1.7521922058516827e-05,
+      "loss": 0.6383,
+      "step": 42030
+    },
+    {
+      "epoch": 1.618286814244466,
+      "grad_norm": 2.5374155044555664,
+      "learning_rate": 1.7504826765436898e-05,
+      "loss": 0.74,
+      "step": 42035
+    },
+    {
+      "epoch": 1.6184793070259866,
+      "grad_norm": 1.9584330320358276,
+      "learning_rate": 1.7487739015994064e-05,
+      "loss": 0.693,
+      "step": 42040
+    },
+    {
+      "epoch": 1.6186717998075073,
+      "grad_norm": 1.7019546031951904,
+      "learning_rate": 1.7470658811750905e-05,
+      "loss": 0.8639,
+      "step": 42045
+    },
+    {
+      "epoch": 1.618864292589028,
+      "grad_norm": 0.9637229442596436,
+      "learning_rate": 1.7453586154269287e-05,
+      "loss": 0.7905,
+      "step": 42050
+    },
+    {
+      "epoch": 1.6190567853705486,
+      "grad_norm": 1.0330854654312134,
+      "learning_rate": 1.7436521045110422e-05,
+      "loss": 0.6918,
+      "step": 42055
+    },
+    {
+      "epoch": 1.6192492781520693,
+      "grad_norm": 1.4338631629943848,
+      "learning_rate": 1.741946348583474e-05,
+      "loss": 0.7246,
+      "step": 42060
+    },
+    {
+      "epoch": 1.61944177093359,
+      "grad_norm": 1.057032823562622,
+      "learning_rate": 1.740241347800209e-05,
+      "loss": 0.8211,
+      "step": 42065
+    },
+    {
+      "epoch": 1.6196342637151107,
+      "grad_norm": 1.3909273147583008,
+      "learning_rate": 1.738537102317156e-05,
+      "loss": 0.8421,
+      "step": 42070
+    },
+    {
+      "epoch": 1.6198267564966313,
+      "grad_norm": 1.2499020099639893,
+      "learning_rate": 1.7368336122901573e-05,
+      "loss": 0.7806,
+      "step": 42075
+    },
+    {
+      "epoch": 1.620019249278152,
+      "grad_norm": 0.8923901319503784,
+      "learning_rate": 1.7351308778749897e-05,
+      "loss": 0.6891,
+      "step": 42080
+    },
+    {
+      "epoch": 1.6202117420596727,
+      "grad_norm": 1.582690954208374,
+      "learning_rate": 1.7334288992273505e-05,
+      "loss": 0.8041,
+      "step": 42085
+    },
+    {
+      "epoch": 1.6204042348411933,
+      "grad_norm": 2.02708101272583,
+      "learning_rate": 1.731727676502878e-05,
+      "loss": 0.9111,
+      "step": 42090
+    },
+    {
+      "epoch": 1.620596727622714,
+      "grad_norm": 0.7452544569969177,
+      "learning_rate": 1.730027209857137e-05,
+      "loss": 0.7931,
+      "step": 42095
+    },
+    {
+      "epoch": 1.6207892204042347,
+      "grad_norm": 1.4981223344802856,
+      "learning_rate": 1.7283274994456267e-05,
+      "loss": 0.8637,
+      "step": 42100
+    },
+    {
+      "epoch": 1.6209817131857556,
+      "grad_norm": 1.1747374534606934,
+      "learning_rate": 1.7266285454237664e-05,
+      "loss": 0.8657,
+      "step": 42105
+    },
+    {
+      "epoch": 1.6211742059672762,
+      "grad_norm": 1.1444958448410034,
+      "learning_rate": 1.7249303479469247e-05,
+      "loss": 0.8245,
+      "step": 42110
+    },
+    {
+      "epoch": 1.621366698748797,
+      "grad_norm": 1.2287684679031372,
+      "learning_rate": 1.7232329071703833e-05,
+      "loss": 0.8287,
+      "step": 42115
+    },
+    {
+      "epoch": 1.6215591915303176,
+      "grad_norm": 2.1734228134155273,
+      "learning_rate": 1.7215362232493638e-05,
+      "loss": 0.8472,
+      "step": 42120
+    },
+    {
+      "epoch": 1.6217516843118385,
+      "grad_norm": 1.0564271211624146,
+      "learning_rate": 1.719840296339017e-05,
+      "loss": 0.7898,
+      "step": 42125
+    },
+    {
+      "epoch": 1.6219441770933591,
+      "grad_norm": 1.1211154460906982,
+      "learning_rate": 1.718145126594426e-05,
+      "loss": 0.8268,
+      "step": 42130
+    },
+    {
+      "epoch": 1.6221366698748798,
+      "grad_norm": 1.5351982116699219,
+      "learning_rate": 1.7164507141705967e-05,
+      "loss": 0.8677,
+      "step": 42135
+    },
+    {
+      "epoch": 1.6223291626564005,
+      "grad_norm": 1.2579121589660645,
+      "learning_rate": 1.7147570592224803e-05,
+      "loss": 0.9406,
+      "step": 42140
+    },
+    {
+      "epoch": 1.6225216554379212,
+      "grad_norm": 1.0373250246047974,
+      "learning_rate": 1.7130641619049436e-05,
+      "loss": 0.7835,
+      "step": 42145
+    },
+    {
+      "epoch": 1.6227141482194418,
+      "grad_norm": 1.5844688415527344,
+      "learning_rate": 1.7113720223727937e-05,
+      "loss": 0.8265,
+      "step": 42150
+    },
+    {
+      "epoch": 1.6229066410009625,
+      "grad_norm": 1.3220007419586182,
+      "learning_rate": 1.7096806407807653e-05,
+      "loss": 0.8124,
+      "step": 42155
+    },
+    {
+      "epoch": 1.6230991337824832,
+      "grad_norm": 1.0909702777862549,
+      "learning_rate": 1.7079900172835263e-05,
+      "loss": 0.7539,
+      "step": 42160
+    },
+    {
+      "epoch": 1.6232916265640038,
+      "grad_norm": 1.1731175184249878,
+      "learning_rate": 1.7063001520356658e-05,
+      "loss": 0.8448,
+      "step": 42165
+    },
+    {
+      "epoch": 1.6234841193455245,
+      "grad_norm": 1.3320856094360352,
+      "learning_rate": 1.7046110451917207e-05,
+      "loss": 0.81,
+      "step": 42170
+    },
+    {
+      "epoch": 1.6236766121270452,
+      "grad_norm": 1.6085408926010132,
+      "learning_rate": 1.7029226969061407e-05,
+      "loss": 0.8147,
+      "step": 42175
+    },
+    {
+      "epoch": 1.6238691049085658,
+      "grad_norm": 1.2450767755508423,
+      "learning_rate": 1.7012351073333168e-05,
+      "loss": 0.7533,
+      "step": 42180
+    },
+    {
+      "epoch": 1.6240615976900865,
+      "grad_norm": 0.9313931465148926,
+      "learning_rate": 1.6995482766275682e-05,
+      "loss": 0.8205,
+      "step": 42185
+    },
+    {
+      "epoch": 1.6242540904716072,
+      "grad_norm": 2.4100301265716553,
+      "learning_rate": 1.697862204943148e-05,
+      "loss": 0.8043,
+      "step": 42190
+    },
+    {
+      "epoch": 1.6244465832531279,
+      "grad_norm": 1.371580958366394,
+      "learning_rate": 1.6965138941945725e-05,
+      "loss": 1.6104,
+      "step": 42195
+    },
+    {
+      "epoch": 1.6246390760346487,
+      "grad_norm": 0.8591129779815674,
+      "learning_rate": 1.6948291891370227e-05,
+      "loss": 0.7858,
+      "step": 42200
+    },
+    {
+      "epoch": 1.6248315688161694,
+      "grad_norm": 1.4864912033081055,
+      "learning_rate": 1.6931452435323226e-05,
+      "loss": 0.8206,
+      "step": 42205
+    },
+    {
+      "epoch": 1.62502406159769,
+      "grad_norm": 1.010176420211792,
+      "learning_rate": 1.6914620575344663e-05,
+      "loss": 0.766,
+      "step": 42210
+    },
+    {
+      "epoch": 1.6252165543792108,
+      "grad_norm": 1.7786433696746826,
+      "learning_rate": 1.6897796312973634e-05,
+      "loss": 0.8346,
+      "step": 42215
+    },
+    {
+      "epoch": 1.6254090471607314,
+      "grad_norm": 0.9494479894638062,
+      "learning_rate": 1.688097964974863e-05,
+      "loss": 0.8556,
+      "step": 42220
+    },
+    {
+      "epoch": 1.6256015399422523,
+      "grad_norm": 1.153810739517212,
+      "learning_rate": 1.686417058720743e-05,
+      "loss": 0.7756,
+      "step": 42225
+    },
+    {
+      "epoch": 1.625794032723773,
+      "grad_norm": 1.232348918914795,
+      "learning_rate": 1.6847369126887124e-05,
+      "loss": 0.8343,
+      "step": 42230
+    },
+    {
+      "epoch": 1.6259865255052937,
+      "grad_norm": 1.010253667831421,
+      "learning_rate": 1.6830575270324022e-05,
+      "loss": 0.9215,
+      "step": 42235
+    },
+    {
+      "epoch": 1.6261790182868143,
+      "grad_norm": 2.2589335441589355,
+      "learning_rate": 1.6813789019053926e-05,
+      "loss": 0.7625,
+      "step": 42240
+    },
+    {
+      "epoch": 1.626371511068335,
+      "grad_norm": 2.0202784538269043,
+      "learning_rate": 1.679701037461173e-05,
+      "loss": 0.8108,
+      "step": 42245
+    },
+    {
+      "epoch": 1.6265640038498557,
+      "grad_norm": 1.7455151081085205,
+      "learning_rate": 1.6780239338531777e-05,
+      "loss": 0.946,
+      "step": 42250
+    },
+    {
+      "epoch": 1.6267564966313763,
+      "grad_norm": 1.2754290103912354,
+      "learning_rate": 1.676347591234765e-05,
+      "loss": 0.9888,
+      "step": 42255
+    },
+    {
+      "epoch": 1.626948989412897,
+      "grad_norm": 1.6412559747695923,
+      "learning_rate": 1.6746720097592285e-05,
+      "loss": 0.8801,
+      "step": 42260
+    },
+    {
+      "epoch": 1.6271414821944177,
+      "grad_norm": 1.385617733001709,
+      "learning_rate": 1.6729971895797835e-05,
+      "loss": 0.8472,
+      "step": 42265
+    },
+    {
+      "epoch": 1.6273339749759383,
+      "grad_norm": 1.6804052591323853,
+      "learning_rate": 1.6713231308495846e-05,
+      "loss": 0.8365,
+      "step": 42270
+    },
+    {
+      "epoch": 1.627526467757459,
+      "grad_norm": 1.2974660396575928,
+      "learning_rate": 1.6696498337217125e-05,
+      "loss": 0.7316,
+      "step": 42275
+    },
+    {
+      "epoch": 1.6277189605389797,
+      "grad_norm": 1.354999303817749,
+      "learning_rate": 1.6679772983491804e-05,
+      "loss": 0.9026,
+      "step": 42280
+    },
+    {
+      "epoch": 1.6279114533205004,
+      "grad_norm": 1.667365312576294,
+      "learning_rate": 1.666305524884931e-05,
+      "loss": 0.8071,
+      "step": 42285
+    },
+    {
+      "epoch": 1.628103946102021,
+      "grad_norm": 0.764616847038269,
+      "learning_rate": 1.664634513481832e-05,
+      "loss": 0.6431,
+      "step": 42290
+    },
+    {
+      "epoch": 1.628296438883542,
+      "grad_norm": 1.2956533432006836,
+      "learning_rate": 1.6629642642926947e-05,
+      "loss": 0.8066,
+      "step": 42295
+    },
+    {
+      "epoch": 1.6284889316650626,
+      "grad_norm": 1.1321804523468018,
+      "learning_rate": 1.661294777470245e-05,
+      "loss": 0.7923,
+      "step": 42300
+    },
+    {
+      "epoch": 1.6286814244465833,
+      "grad_norm": 1.4168885946273804,
+      "learning_rate": 1.659626053167149e-05,
+      "loss": 0.6876,
+      "step": 42305
+    },
+    {
+      "epoch": 1.628873917228104,
+      "grad_norm": 1.1318153142929077,
+      "learning_rate": 1.6579580915360003e-05,
+      "loss": 1.0224,
+      "step": 42310
+    },
+    {
+      "epoch": 1.6290664100096246,
+      "grad_norm": 1.2682236433029175,
+      "learning_rate": 1.656290892729325e-05,
+      "loss": 0.6783,
+      "step": 42315
+    },
+    {
+      "epoch": 1.6292589027911455,
+      "grad_norm": 1.0369542837142944,
+      "learning_rate": 1.654624456899572e-05,
+      "loss": 0.746,
+      "step": 42320
+    },
+    {
+      "epoch": 1.6294513955726662,
+      "grad_norm": 1.284934639930725,
+      "learning_rate": 1.6529587841991336e-05,
+      "loss": 0.7735,
+      "step": 42325
+    },
+    {
+      "epoch": 1.6296438883541868,
+      "grad_norm": 0.9802597761154175,
+      "learning_rate": 1.6512938747803186e-05,
+      "loss": 0.7332,
+      "step": 42330
+    },
+    {
+      "epoch": 1.6298363811357075,
+      "grad_norm": 1.6991342306137085,
+      "learning_rate": 1.649629728795372e-05,
+      "loss": 0.8675,
+      "step": 42335
+    },
+    {
+      "epoch": 1.6300288739172282,
+      "grad_norm": 1.371645212173462,
+      "learning_rate": 1.6479663463964722e-05,
+      "loss": 0.7313,
+      "step": 42340
+    },
+    {
+      "epoch": 1.6302213666987488,
+      "grad_norm": 1.8463810682296753,
+      "learning_rate": 1.646303727735724e-05,
+      "loss": 0.9124,
+      "step": 42345
+    },
+    {
+      "epoch": 1.6304138594802695,
+      "grad_norm": 2.088982582092285,
+      "learning_rate": 1.6446418729651604e-05,
+      "loss": 0.8852,
+      "step": 42350
+    },
+    {
+      "epoch": 1.6306063522617902,
+      "grad_norm": 1.3795658349990845,
+      "learning_rate": 1.6429807822367482e-05,
+      "loss": 0.838,
+      "step": 42355
+    },
+    {
+      "epoch": 1.6307988450433109,
+      "grad_norm": 1.5469810962677002,
+      "learning_rate": 1.641320455702383e-05,
+      "loss": 0.9927,
+      "step": 42360
+    },
+    {
+      "epoch": 1.6309913378248315,
+      "grad_norm": 1.1996501684188843,
+      "learning_rate": 1.6396608935138902e-05,
+      "loss": 0.7755,
+      "step": 42365
+    },
+    {
+      "epoch": 1.6311838306063522,
+      "grad_norm": 0.9938015937805176,
+      "learning_rate": 1.63800209582303e-05,
+      "loss": 0.7619,
+      "step": 42370
+    },
+    {
+      "epoch": 1.6313763233878729,
+      "grad_norm": 1.932384729385376,
+      "learning_rate": 1.636344062781482e-05,
+      "loss": 0.8774,
+      "step": 42375
+    },
+    {
+      "epoch": 1.6315688161693935,
+      "grad_norm": 0.921647310256958,
+      "learning_rate": 1.6346867945408662e-05,
+      "loss": 0.7319,
+      "step": 42380
+    },
+    {
+      "epoch": 1.6317613089509142,
+      "grad_norm": 1.4289840459823608,
+      "learning_rate": 1.6330302912527263e-05,
+      "loss": 0.7474,
+      "step": 42385
+    },
+    {
+      "epoch": 1.6319538017324349,
+      "grad_norm": 2.0691914558410645,
+      "learning_rate": 1.6313745530685443e-05,
+      "loss": 0.9581,
+      "step": 42390
+    },
+    {
+      "epoch": 1.6321462945139558,
+      "grad_norm": 1.3220711946487427,
+      "learning_rate": 1.6297195801397157e-05,
+      "loss": 0.6706,
+      "step": 42395
+    },
+    {
+      "epoch": 1.6323387872954764,
+      "grad_norm": 1.0470640659332275,
+      "learning_rate": 1.6280653726175897e-05,
+      "loss": 0.6795,
+      "step": 42400
+    },
+    {
+      "epoch": 1.632531280076997,
+      "grad_norm": 1.722261667251587,
+      "learning_rate": 1.626411930653423e-05,
+      "loss": 0.8447,
+      "step": 42405
+    },
+    {
+      "epoch": 1.6327237728585178,
+      "grad_norm": 2.425245761871338,
+      "learning_rate": 1.624759254398417e-05,
+      "loss": 0.9633,
+      "step": 42410
+    },
+    {
+      "epoch": 1.6329162656400384,
+      "grad_norm": 1.6226695775985718,
+      "learning_rate": 1.6231073440036947e-05,
+      "loss": 0.8426,
+      "step": 42415
+    },
+    {
+      "epoch": 1.6331087584215593,
+      "grad_norm": 2.568894147872925,
+      "learning_rate": 1.621456199620317e-05,
+      "loss": 0.729,
+      "step": 42420
+    },
+    {
+      "epoch": 1.63330125120308,
+      "grad_norm": 1.7268556356430054,
+      "learning_rate": 1.6198058213992617e-05,
+      "loss": 0.7958,
+      "step": 42425
+    },
+    {
+      "epoch": 1.6334937439846007,
+      "grad_norm": 1.1411241292953491,
+      "learning_rate": 1.618156209491456e-05,
+      "loss": 0.8699,
+      "step": 42430
+    },
+    {
+      "epoch": 1.6336862367661213,
+      "grad_norm": 1.1613513231277466,
+      "learning_rate": 1.6165073640477368e-05,
+      "loss": 0.8642,
+      "step": 42435
+    },
+    {
+      "epoch": 1.633878729547642,
+      "grad_norm": 1.3415052890777588,
+      "learning_rate": 1.6148592852188838e-05,
+      "loss": 0.9454,
+      "step": 42440
+    },
+    {
+      "epoch": 1.6340712223291627,
+      "grad_norm": 0.5342742204666138,
+      "learning_rate": 1.613211973155604e-05,
+      "loss": 0.8002,
+      "step": 42445
+    },
+    {
+      "epoch": 1.6342637151106834,
+      "grad_norm": 1.4904274940490723,
+      "learning_rate": 1.6115654280085335e-05,
+      "loss": 0.8969,
+      "step": 42450
+    },
+    {
+      "epoch": 1.634456207892204,
+      "grad_norm": 0.9641575813293457,
+      "learning_rate": 1.609919649928231e-05,
+      "loss": 0.7795,
+      "step": 42455
+    },
+    {
+      "epoch": 1.6346487006737247,
+      "grad_norm": 1.267978310585022,
+      "learning_rate": 1.6082746390652026e-05,
+      "loss": 0.7523,
+      "step": 42460
+    },
+    {
+      "epoch": 1.6348411934552454,
+      "grad_norm": 1.0590537786483765,
+      "learning_rate": 1.606630395569866e-05,
+      "loss": 0.7176,
+      "step": 42465
+    },
+    {
+      "epoch": 1.635033686236766,
+      "grad_norm": 0.7965215444564819,
+      "learning_rate": 1.604986919592578e-05,
+      "loss": 0.8919,
+      "step": 42470
+    },
+    {
+      "epoch": 1.6352261790182867,
+      "grad_norm": 0.9418613314628601,
+      "learning_rate": 1.603344211283625e-05,
+      "loss": 0.7038,
+      "step": 42475
+    },
+    {
+      "epoch": 1.6354186717998074,
+      "grad_norm": 1.7966370582580566,
+      "learning_rate": 1.6017022707932237e-05,
+      "loss": 0.7553,
+      "step": 42480
+    },
+    {
+      "epoch": 1.635611164581328,
+      "grad_norm": 0.9546367526054382,
+      "learning_rate": 1.6000610982715135e-05,
+      "loss": 0.7796,
+      "step": 42485
+    },
+    {
+      "epoch": 1.635803657362849,
+      "grad_norm": 1.3808022737503052,
+      "learning_rate": 1.598420693868571e-05,
+      "loss": 0.6758,
+      "step": 42490
+    },
+    {
+      "epoch": 1.6359961501443696,
+      "grad_norm": 0.980531632900238,
+      "learning_rate": 1.5967810577344034e-05,
+      "loss": 0.8068,
+      "step": 42495
+    },
+    {
+      "epoch": 1.6361886429258903,
+      "grad_norm": 1.2579573392868042,
+      "learning_rate": 1.5951421900189366e-05,
+      "loss": 0.9325,
+      "step": 42500
+    },
+    {
+      "epoch": 1.636381135707411,
+      "grad_norm": 1.0082838535308838,
+      "learning_rate": 1.5935040908720455e-05,
+      "loss": 0.8162,
+      "step": 42505
+    },
+    {
+      "epoch": 1.6365736284889316,
+      "grad_norm": 1.9807016849517822,
+      "learning_rate": 1.5918667604435132e-05,
+      "loss": 0.7665,
+      "step": 42510
+    },
+    {
+      "epoch": 1.6367661212704525,
+      "grad_norm": 1.1579536199569702,
+      "learning_rate": 1.5902301988830682e-05,
+      "loss": 0.856,
+      "step": 42515
+    },
+    {
+      "epoch": 1.6369586140519732,
+      "grad_norm": 1.3920001983642578,
+      "learning_rate": 1.588594406340361e-05,
+      "loss": 0.8391,
+      "step": 42520
+    },
+    {
+      "epoch": 1.6371511068334939,
+      "grad_norm": 1.4542222023010254,
+      "learning_rate": 1.5869593829649787e-05,
+      "loss": 0.81,
+      "step": 42525
+    },
+    {
+      "epoch": 1.6373435996150145,
+      "grad_norm": 1.2081284523010254,
+      "learning_rate": 1.5853251289064242e-05,
+      "loss": 0.8316,
+      "step": 42530
+    },
+    {
+      "epoch": 1.6375360923965352,
+      "grad_norm": 3.1090595722198486,
+      "learning_rate": 1.583691644314148e-05,
+      "loss": 0.9793,
+      "step": 42535
+    },
+    {
+      "epoch": 1.6377285851780559,
+      "grad_norm": 1.5080655813217163,
+      "learning_rate": 1.5820589293375142e-05,
+      "loss": 0.9199,
+      "step": 42540
+    },
+    {
+      "epoch": 1.6379210779595765,
+      "grad_norm": 1.2855561971664429,
+      "learning_rate": 1.5804269841258323e-05,
+      "loss": 0.7934,
+      "step": 42545
+    },
+    {
+      "epoch": 1.6381135707410972,
+      "grad_norm": 1.6619151830673218,
+      "learning_rate": 1.578795808828326e-05,
+      "loss": 0.8615,
+      "step": 42550
+    },
+    {
+      "epoch": 1.6383060635226179,
+      "grad_norm": 1.0249853134155273,
+      "learning_rate": 1.5771654035941574e-05,
+      "loss": 0.8015,
+      "step": 42555
+    },
+    {
+      "epoch": 1.6384985563041385,
+      "grad_norm": 1.0512340068817139,
+      "learning_rate": 1.575535768572416e-05,
+      "loss": 0.7564,
+      "step": 42560
+    },
+    {
+      "epoch": 1.6386910490856592,
+      "grad_norm": 1.487186074256897,
+      "learning_rate": 1.5739069039121245e-05,
+      "loss": 0.8638,
+      "step": 42565
+    },
+    {
+      "epoch": 1.6388835418671799,
+      "grad_norm": 1.0381027460098267,
+      "learning_rate": 1.5722788097622267e-05,
+      "loss": 0.917,
+      "step": 42570
+    },
+    {
+      "epoch": 1.6390760346487006,
+      "grad_norm": 1.8956249952316284,
+      "learning_rate": 1.5706514862716028e-05,
+      "loss": 0.8835,
+      "step": 42575
+    },
+    {
+      "epoch": 1.6392685274302212,
+      "grad_norm": 0.9786503314971924,
+      "learning_rate": 1.5690249335890605e-05,
+      "loss": 0.7697,
+      "step": 42580
+    },
+    {
+      "epoch": 1.6394610202117421,
+      "grad_norm": 1.1818338632583618,
+      "learning_rate": 1.567399151863339e-05,
+      "loss": 0.8998,
+      "step": 42585
+    },
+    {
+      "epoch": 1.6396535129932628,
+      "grad_norm": 1.3527082204818726,
+      "learning_rate": 1.565774141243106e-05,
+      "loss": 0.7461,
+      "step": 42590
+    },
+    {
+      "epoch": 1.6398460057747835,
+      "grad_norm": 1.5790470838546753,
+      "learning_rate": 1.5641499018769545e-05,
+      "loss": 0.8563,
+      "step": 42595
+    },
+    {
+      "epoch": 1.6400384985563041,
+      "grad_norm": 0.7987781763076782,
+      "learning_rate": 1.5625264339134115e-05,
+      "loss": 0.7464,
+      "step": 42600
+    },
+    {
+      "epoch": 1.6402309913378248,
+      "grad_norm": 1.0493252277374268,
+      "learning_rate": 1.5609037375009326e-05,
+      "loss": 0.6797,
+      "step": 42605
+    },
+    {
+      "epoch": 1.6404234841193457,
+      "grad_norm": 1.0951621532440186,
+      "learning_rate": 1.559281812787906e-05,
+      "loss": 0.7957,
+      "step": 42610
+    },
+    {
+      "epoch": 1.6406159769008664,
+      "grad_norm": 1.529111623764038,
+      "learning_rate": 1.5576606599226383e-05,
+      "loss": 0.894,
+      "step": 42615
+    },
+    {
+      "epoch": 1.640808469682387,
+      "grad_norm": 0.8548210263252258,
+      "learning_rate": 1.5560402790533823e-05,
+      "loss": 0.7031,
+      "step": 42620
+    },
+    {
+      "epoch": 1.6410009624639077,
+      "grad_norm": 1.763548493385315,
+      "learning_rate": 1.554420670328305e-05,
+      "loss": 0.7808,
+      "step": 42625
+    },
+    {
+      "epoch": 1.6411934552454284,
+      "grad_norm": 1.7161123752593994,
+      "learning_rate": 1.5528018338955098e-05,
+      "loss": 0.686,
+      "step": 42630
+    },
+    {
+      "epoch": 1.641385948026949,
+      "grad_norm": 2.433823347091675,
+      "learning_rate": 1.5511837699030295e-05,
+      "loss": 0.788,
+      "step": 42635
+    },
+    {
+      "epoch": 1.6415784408084697,
+      "grad_norm": 1.3295754194259644,
+      "learning_rate": 1.549566478498827e-05,
+      "loss": 0.765,
+      "step": 42640
+    },
+    {
+      "epoch": 1.6417709335899904,
+      "grad_norm": 1.3078267574310303,
+      "learning_rate": 1.547949959830787e-05,
+      "loss": 0.7929,
+      "step": 42645
+    },
+    {
+      "epoch": 1.641963426371511,
+      "grad_norm": 2.7279410362243652,
+      "learning_rate": 1.5463342140467373e-05,
+      "loss": 0.9705,
+      "step": 42650
+    },
+    {
+      "epoch": 1.6421559191530317,
+      "grad_norm": 1.5727055072784424,
+      "learning_rate": 1.5447192412944223e-05,
+      "loss": 0.8038,
+      "step": 42655
+    },
+    {
+      "epoch": 1.6423484119345524,
+      "grad_norm": 1.5005487203598022,
+      "learning_rate": 1.5431050417215208e-05,
+      "loss": 0.8304,
+      "step": 42660
+    },
+    {
+      "epoch": 1.642540904716073,
+      "grad_norm": 1.3737221956253052,
+      "learning_rate": 1.541491615475642e-05,
+      "loss": 0.9049,
+      "step": 42665
+    },
+    {
+      "epoch": 1.6427333974975937,
+      "grad_norm": 0.9405394196510315,
+      "learning_rate": 1.5398789627043243e-05,
+      "loss": 0.7638,
+      "step": 42670
+    },
+    {
+      "epoch": 1.6429258902791144,
+      "grad_norm": 1.3767449855804443,
+      "learning_rate": 1.5382670835550293e-05,
+      "loss": 0.791,
+      "step": 42675
+    },
+    {
+      "epoch": 1.643118383060635,
+      "grad_norm": 0.9052772521972656,
+      "learning_rate": 1.5366559781751566e-05,
+      "loss": 0.8165,
+      "step": 42680
+    },
+    {
+      "epoch": 1.643310875842156,
+      "grad_norm": 1.9760810136795044,
+      "learning_rate": 1.53504564671203e-05,
+      "loss": 1.0684,
+      "step": 42685
+    },
+    {
+      "epoch": 1.6435033686236766,
+      "grad_norm": 0.9905810952186584,
+      "learning_rate": 1.533436089312904e-05,
+      "loss": 0.7722,
+      "step": 42690
+    },
+    {
+      "epoch": 1.6436958614051973,
+      "grad_norm": 1.814131259918213,
+      "learning_rate": 1.531827306124963e-05,
+      "loss": 0.7637,
+      "step": 42695
+    },
+    {
+      "epoch": 1.643888354186718,
+      "grad_norm": 1.4231481552124023,
+      "learning_rate": 1.530219297295318e-05,
+      "loss": 0.8173,
+      "step": 42700
+    },
+    {
+      "epoch": 1.6440808469682386,
+      "grad_norm": 1.3890655040740967,
+      "learning_rate": 1.5286120629710098e-05,
+      "loss": 0.7146,
+      "step": 42705
+    },
+    {
+      "epoch": 1.6442733397497595,
+      "grad_norm": 1.4815847873687744,
+      "learning_rate": 1.527005603299011e-05,
+      "loss": 0.8625,
+      "step": 42710
+    },
+    {
+      "epoch": 1.6444658325312802,
+      "grad_norm": 1.5338302850723267,
+      "learning_rate": 1.5253999184262235e-05,
+      "loss": 0.6705,
+      "step": 42715
+    },
+    {
+      "epoch": 1.6446583253128009,
+      "grad_norm": 1.2656974792480469,
+      "learning_rate": 1.523795008499469e-05,
+      "loss": 0.9487,
+      "step": 42720
+    },
+    {
+      "epoch": 1.6448508180943215,
+      "grad_norm": 1.0494598150253296,
+      "learning_rate": 1.5221908736655167e-05,
+      "loss": 0.7775,
+      "step": 42725
+    },
+    {
+      "epoch": 1.6450433108758422,
+      "grad_norm": 1.1466418504714966,
+      "learning_rate": 1.5205875140710458e-05,
+      "loss": 0.9629,
+      "step": 42730
+    },
+    {
+      "epoch": 1.6452358036573629,
+      "grad_norm": 1.3125535249710083,
+      "learning_rate": 1.5189849298626769e-05,
+      "loss": 0.9284,
+      "step": 42735
+    },
+    {
+      "epoch": 1.6454282964388836,
+      "grad_norm": 1.2060669660568237,
+      "learning_rate": 1.5173831211869539e-05,
+      "loss": 0.8197,
+      "step": 42740
+    },
+    {
+      "epoch": 1.6456207892204042,
+      "grad_norm": 1.393114447593689,
+      "learning_rate": 1.5157820881903539e-05,
+      "loss": 0.8467,
+      "step": 42745
+    },
+    {
+      "epoch": 1.645813282001925,
+      "grad_norm": 2.256540060043335,
+      "learning_rate": 1.5141818310192756e-05,
+      "loss": 0.9494,
+      "step": 42750
+    },
+    {
+      "epoch": 1.6460057747834456,
+      "grad_norm": 1.1511859893798828,
+      "learning_rate": 1.5125823498200598e-05,
+      "loss": 0.8653,
+      "step": 42755
+    },
+    {
+      "epoch": 1.6461982675649662,
+      "grad_norm": 1.7524446249008179,
+      "learning_rate": 1.5109836447389613e-05,
+      "loss": 0.7703,
+      "step": 42760
+    },
+    {
+      "epoch": 1.646390760346487,
+      "grad_norm": 1.9260464906692505,
+      "learning_rate": 1.5093857159221747e-05,
+      "loss": 0.8431,
+      "step": 42765
+    },
+    {
+      "epoch": 1.6465832531280076,
+      "grad_norm": 1.6984754800796509,
+      "learning_rate": 1.5077885635158185e-05,
+      "loss": 0.7436,
+      "step": 42770
+    },
+    {
+      "epoch": 1.6467757459095282,
+      "grad_norm": 1.1173107624053955,
+      "learning_rate": 1.5061921876659446e-05,
+      "loss": 0.7743,
+      "step": 42775
+    },
+    {
+      "epoch": 1.6469682386910491,
+      "grad_norm": 1.5459610223770142,
+      "learning_rate": 1.5045965885185253e-05,
+      "loss": 0.7828,
+      "step": 42780
+    },
+    {
+      "epoch": 1.6471607314725698,
+      "grad_norm": 1.2092872858047485,
+      "learning_rate": 1.503001766219475e-05,
+      "loss": 0.8356,
+      "step": 42785
+    },
+    {
+      "epoch": 1.6473532242540905,
+      "grad_norm": 1.4177119731903076,
+      "learning_rate": 1.5014077209146227e-05,
+      "loss": 0.6802,
+      "step": 42790
+    },
+    {
+      "epoch": 1.6475457170356111,
+      "grad_norm": 1.1195024251937866,
+      "learning_rate": 1.499814452749737e-05,
+      "loss": 0.8973,
+      "step": 42795
+    },
+    {
+      "epoch": 1.6477382098171318,
+      "grad_norm": 1.1029400825500488,
+      "learning_rate": 1.4982219618705119e-05,
+      "loss": 0.7615,
+      "step": 42800
+    },
+    {
+      "epoch": 1.6479307025986527,
+      "grad_norm": 1.2463111877441406,
+      "learning_rate": 1.4966302484225681e-05,
+      "loss": 0.8533,
+      "step": 42805
+    },
+    {
+      "epoch": 1.6481231953801734,
+      "grad_norm": 1.6013331413269043,
+      "learning_rate": 1.4950393125514605e-05,
+      "loss": 0.7185,
+      "step": 42810
+    },
+    {
+      "epoch": 1.648315688161694,
+      "grad_norm": 1.4675661325454712,
+      "learning_rate": 1.4934491544026663e-05,
+      "loss": 0.6616,
+      "step": 42815
+    },
+    {
+      "epoch": 1.6485081809432147,
+      "grad_norm": 1.2098175287246704,
+      "learning_rate": 1.4918597741215957e-05,
+      "loss": 0.7673,
+      "step": 42820
+    },
+    {
+      "epoch": 1.6487006737247354,
+      "grad_norm": 0.9863401651382446,
+      "learning_rate": 1.4902711718535866e-05,
+      "loss": 0.7465,
+      "step": 42825
+    },
+    {
+      "epoch": 1.648893166506256,
+      "grad_norm": 1.357666254043579,
+      "learning_rate": 1.4886833477439099e-05,
+      "loss": 0.8188,
+      "step": 42830
+    },
+    {
+      "epoch": 1.6490856592877767,
+      "grad_norm": 2.8012890815734863,
+      "learning_rate": 1.4870963019377548e-05,
+      "loss": 0.8675,
+      "step": 42835
+    },
+    {
+      "epoch": 1.6492781520692974,
+      "grad_norm": 2.035005807876587,
+      "learning_rate": 1.4855100345802542e-05,
+      "loss": 0.9065,
+      "step": 42840
+    },
+    {
+      "epoch": 1.649470644850818,
+      "grad_norm": 1.256974458694458,
+      "learning_rate": 1.4839245458164553e-05,
+      "loss": 0.8464,
+      "step": 42845
+    },
+    {
+      "epoch": 1.6496631376323387,
+      "grad_norm": 1.1360011100769043,
+      "learning_rate": 1.4823398357913432e-05,
+      "loss": 0.8485,
+      "step": 42850
+    },
+    {
+      "epoch": 1.6498556304138594,
+      "grad_norm": 1.2088253498077393,
+      "learning_rate": 1.4807559046498287e-05,
+      "loss": 0.8786,
+      "step": 42855
+    },
+    {
+      "epoch": 1.65004812319538,
+      "grad_norm": 1.3639460802078247,
+      "learning_rate": 1.4791727525367539e-05,
+      "loss": 0.9163,
+      "step": 42860
+    },
+    {
+      "epoch": 1.6502406159769007,
+      "grad_norm": 0.8863676190376282,
+      "learning_rate": 1.4775903795968804e-05,
+      "loss": 0.7723,
+      "step": 42865
+    },
+    {
+      "epoch": 1.6504331087584214,
+      "grad_norm": 1.2062976360321045,
+      "learning_rate": 1.476008785974916e-05,
+      "loss": 0.7728,
+      "step": 42870
+    },
+    {
+      "epoch": 1.650625601539942,
+      "grad_norm": 2.016284227371216,
+      "learning_rate": 1.4744279718154797e-05,
+      "loss": 0.8003,
+      "step": 42875
+    },
+    {
+      "epoch": 1.650818094321463,
+      "grad_norm": 2.4403231143951416,
+      "learning_rate": 1.4728479372631287e-05,
+      "loss": 0.8282,
+      "step": 42880
+    },
+    {
+      "epoch": 1.6510105871029837,
+      "grad_norm": 1.3682914972305298,
+      "learning_rate": 1.4712686824623466e-05,
+      "loss": 0.8877,
+      "step": 42885
+    },
+    {
+      "epoch": 1.6512030798845043,
+      "grad_norm": 1.542219877243042,
+      "learning_rate": 1.469690207557548e-05,
+      "loss": 0.8553,
+      "step": 42890
+    },
+    {
+      "epoch": 1.651395572666025,
+      "grad_norm": 1.1081209182739258,
+      "learning_rate": 1.4681125126930695e-05,
+      "loss": 0.8886,
+      "step": 42895
+    },
+    {
+      "epoch": 1.6515880654475459,
+      "grad_norm": 2.007857084274292,
+      "learning_rate": 1.4665355980131834e-05,
+      "loss": 0.8171,
+      "step": 42900
+    },
+    {
+      "epoch": 1.6517805582290666,
+      "grad_norm": 1.1662068367004395,
+      "learning_rate": 1.4649594636620878e-05,
+      "loss": 0.8906,
+      "step": 42905
+    },
+    {
+      "epoch": 1.6519730510105872,
+      "grad_norm": 1.0259616374969482,
+      "learning_rate": 1.4633841097839096e-05,
+      "loss": 0.6473,
+      "step": 42910
+    },
+    {
+      "epoch": 1.652165543792108,
+      "grad_norm": 0.8987205624580383,
+      "learning_rate": 1.4618095365227069e-05,
+      "loss": 0.7654,
+      "step": 42915
+    },
+    {
+      "epoch": 1.6523580365736286,
+      "grad_norm": 1.9955406188964844,
+      "learning_rate": 1.460235744022459e-05,
+      "loss": 0.882,
+      "step": 42920
+    },
+    {
+      "epoch": 1.6525505293551492,
+      "grad_norm": 1.3027300834655762,
+      "learning_rate": 1.458662732427083e-05,
+      "loss": 0.9002,
+      "step": 42925
+    },
+    {
+      "epoch": 1.65274302213667,
+      "grad_norm": 0.7334115505218506,
+      "learning_rate": 1.457090501880417e-05,
+      "loss": 0.8804,
+      "step": 42930
+    },
+    {
+      "epoch": 1.6529355149181906,
+      "grad_norm": 1.0303765535354614,
+      "learning_rate": 1.4555190525262363e-05,
+      "loss": 0.7637,
+      "step": 42935
+    },
+    {
+      "epoch": 1.6531280076997112,
+      "grad_norm": 0.9630566239356995,
+      "learning_rate": 1.4539483845082324e-05,
+      "loss": 0.8624,
+      "step": 42940
+    },
+    {
+      "epoch": 1.653320500481232,
+      "grad_norm": 1.3009661436080933,
+      "learning_rate": 1.4523784979700395e-05,
+      "loss": 0.799,
+      "step": 42945
+    },
+    {
+      "epoch": 1.6535129932627526,
+      "grad_norm": 1.6627331972122192,
+      "learning_rate": 1.4508093930552092e-05,
+      "loss": 0.9133,
+      "step": 42950
+    },
+    {
+      "epoch": 1.6537054860442733,
+      "grad_norm": 1.2457365989685059,
+      "learning_rate": 1.4492410699072256e-05,
+      "loss": 0.8653,
+      "step": 42955
+    },
+    {
+      "epoch": 1.653897978825794,
+      "grad_norm": 2.1877925395965576,
+      "learning_rate": 1.447673528669503e-05,
+      "loss": 0.9323,
+      "step": 42960
+    },
+    {
+      "epoch": 1.6540904716073146,
+      "grad_norm": 1.2046828269958496,
+      "learning_rate": 1.4461067694853847e-05,
+      "loss": 0.7993,
+      "step": 42965
+    },
+    {
+      "epoch": 1.6542829643888353,
+      "grad_norm": 1.9960542917251587,
+      "learning_rate": 1.4445407924981325e-05,
+      "loss": 0.7632,
+      "step": 42970
+    },
+    {
+      "epoch": 1.6544754571703562,
+      "grad_norm": 1.4354302883148193,
+      "learning_rate": 1.4429755978509551e-05,
+      "loss": 0.7538,
+      "step": 42975
+    },
+    {
+      "epoch": 1.6546679499518768,
+      "grad_norm": 1.3290390968322754,
+      "learning_rate": 1.4414111856869727e-05,
+      "loss": 0.7158,
+      "step": 42980
+    },
+    {
+      "epoch": 1.6548604427333975,
+      "grad_norm": 0.9498152136802673,
+      "learning_rate": 1.4398475561492409e-05,
+      "loss": 0.7065,
+      "step": 42985
+    },
+    {
+      "epoch": 1.6550529355149182,
+      "grad_norm": 1.9382153749465942,
+      "learning_rate": 1.438284709380745e-05,
+      "loss": 0.6704,
+      "step": 42990
+    },
+    {
+      "epoch": 1.6552454282964388,
+      "grad_norm": 1.109632968902588,
+      "learning_rate": 1.4367226455243988e-05,
+      "loss": 0.8012,
+      "step": 42995
+    },
+    {
+      "epoch": 1.6554379210779597,
+      "grad_norm": 1.8244105577468872,
+      "learning_rate": 1.4351613647230344e-05,
+      "loss": 0.7368,
+      "step": 43000
+    },
+    {
+      "epoch": 1.6556304138594804,
+      "grad_norm": 1.2546757459640503,
+      "learning_rate": 1.4336008671194311e-05,
+      "loss": 0.7095,
+      "step": 43005
+    },
+    {
+      "epoch": 1.655822906641001,
+      "grad_norm": 1.7501899003982544,
+      "learning_rate": 1.4320411528562806e-05,
+      "loss": 0.6884,
+      "step": 43010
+    },
+    {
+      "epoch": 1.6560153994225217,
+      "grad_norm": 1.3198050260543823,
+      "learning_rate": 1.430482222076207e-05,
+      "loss": 0.679,
+      "step": 43015
+    },
+    {
+      "epoch": 1.6562078922040424,
+      "grad_norm": 1.1397435665130615,
+      "learning_rate": 1.428924074921768e-05,
+      "loss": 0.7621,
+      "step": 43020
+    },
+    {
+      "epoch": 1.656400384985563,
+      "grad_norm": 1.1400476694107056,
+      "learning_rate": 1.427366711535445e-05,
+      "loss": 0.7215,
+      "step": 43025
+    },
+    {
+      "epoch": 1.6565928777670837,
+      "grad_norm": 1.151114821434021,
+      "learning_rate": 1.4258101320596462e-05,
+      "loss": 0.856,
+      "step": 43030
+    },
+    {
+      "epoch": 1.6567853705486044,
+      "grad_norm": 1.7196004390716553,
+      "learning_rate": 1.4242543366367122e-05,
+      "loss": 0.8217,
+      "step": 43035
+    },
+    {
+      "epoch": 1.656977863330125,
+      "grad_norm": 1.8382436037063599,
+      "learning_rate": 1.4226993254089127e-05,
+      "loss": 0.8671,
+      "step": 43040
+    },
+    {
+      "epoch": 1.6571703561116458,
+      "grad_norm": 0.8144475221633911,
+      "learning_rate": 1.4211450985184349e-05,
+      "loss": 0.9136,
+      "step": 43045
+    },
+    {
+      "epoch": 1.6573628488931664,
+      "grad_norm": 1.0420215129852295,
+      "learning_rate": 1.4195916561074129e-05,
+      "loss": 0.8111,
+      "step": 43050
+    },
+    {
+      "epoch": 1.657555341674687,
+      "grad_norm": 1.1034629344940186,
+      "learning_rate": 1.41803899831789e-05,
+      "loss": 0.7063,
+      "step": 43055
+    },
+    {
+      "epoch": 1.6577478344562078,
+      "grad_norm": 1.1261588335037231,
+      "learning_rate": 1.4164871252918544e-05,
+      "loss": 0.9457,
+      "step": 43060
+    },
+    {
+      "epoch": 1.6579403272377284,
+      "grad_norm": 1.1537283658981323,
+      "learning_rate": 1.4149360371712084e-05,
+      "loss": 1.0019,
+      "step": 43065
+    },
+    {
+      "epoch": 1.6581328200192493,
+      "grad_norm": 1.0885237455368042,
+      "learning_rate": 1.4133857340977908e-05,
+      "loss": 0.8041,
+      "step": 43070
+    },
+    {
+      "epoch": 1.65832531280077,
+      "grad_norm": 1.2921056747436523,
+      "learning_rate": 1.411836216213367e-05,
+      "loss": 0.8121,
+      "step": 43075
+    },
+    {
+      "epoch": 1.6585178055822907,
+      "grad_norm": 1.4436019659042358,
+      "learning_rate": 1.4102874836596325e-05,
+      "loss": 0.7213,
+      "step": 43080
+    },
+    {
+      "epoch": 1.6587102983638113,
+      "grad_norm": 1.3824903964996338,
+      "learning_rate": 1.4087395365782008e-05,
+      "loss": 0.8929,
+      "step": 43085
+    },
+    {
+      "epoch": 1.658902791145332,
+      "grad_norm": 1.0888464450836182,
+      "learning_rate": 1.407192375110632e-05,
+      "loss": 0.828,
+      "step": 43090
+    },
+    {
+      "epoch": 1.659095283926853,
+      "grad_norm": 1.3013052940368652,
+      "learning_rate": 1.405645999398395e-05,
+      "loss": 0.7308,
+      "step": 43095
+    },
+    {
+      "epoch": 1.6592877767083736,
+      "grad_norm": 1.4749294519424438,
+      "learning_rate": 1.4041004095828992e-05,
+      "loss": 0.8799,
+      "step": 43100
+    },
+    {
+      "epoch": 1.6594802694898942,
+      "grad_norm": 1.4504752159118652,
+      "learning_rate": 1.4025556058054789e-05,
+      "loss": 0.7252,
+      "step": 43105
+    },
+    {
+      "epoch": 1.659672762271415,
+      "grad_norm": 1.064914584159851,
+      "learning_rate": 1.4010115882073971e-05,
+      "loss": 0.8309,
+      "step": 43110
+    },
+    {
+      "epoch": 1.6598652550529356,
+      "grad_norm": 1.1333822011947632,
+      "learning_rate": 1.3994683569298406e-05,
+      "loss": 0.8019,
+      "step": 43115
+    },
+    {
+      "epoch": 1.6600577478344563,
+      "grad_norm": 1.1221379041671753,
+      "learning_rate": 1.39792591211393e-05,
+      "loss": 0.9147,
+      "step": 43120
+    },
+    {
+      "epoch": 1.660250240615977,
+      "grad_norm": 1.3781883716583252,
+      "learning_rate": 1.3963842539007098e-05,
+      "loss": 0.7619,
+      "step": 43125
+    },
+    {
+      "epoch": 1.6604427333974976,
+      "grad_norm": 0.8834213614463806,
+      "learning_rate": 1.394843382431158e-05,
+      "loss": 0.9033,
+      "step": 43130
+    },
+    {
+      "epoch": 1.6606352261790183,
+      "grad_norm": 1.049788236618042,
+      "learning_rate": 1.3933032978461757e-05,
+      "loss": 0.6857,
+      "step": 43135
+    },
+    {
+      "epoch": 1.660827718960539,
+      "grad_norm": 1.520080804824829,
+      "learning_rate": 1.3917640002865905e-05,
+      "loss": 0.8151,
+      "step": 43140
+    },
+    {
+      "epoch": 1.6610202117420596,
+      "grad_norm": 1.1372543573379517,
+      "learning_rate": 1.3902254898931633e-05,
+      "loss": 0.7423,
+      "step": 43145
+    },
+    {
+      "epoch": 1.6612127045235803,
+      "grad_norm": 0.8767010569572449,
+      "learning_rate": 1.3886877668065802e-05,
+      "loss": 0.874,
+      "step": 43150
+    },
+    {
+      "epoch": 1.661405197305101,
+      "grad_norm": 1.2810707092285156,
+      "learning_rate": 1.3871508311674587e-05,
+      "loss": 0.7602,
+      "step": 43155
+    },
+    {
+      "epoch": 1.6615976900866216,
+      "grad_norm": 1.0464438199996948,
+      "learning_rate": 1.3856146831163341e-05,
+      "loss": 0.6807,
+      "step": 43160
+    },
+    {
+      "epoch": 1.6617901828681423,
+      "grad_norm": 2.142082452774048,
+      "learning_rate": 1.3840793227936867e-05,
+      "loss": 0.7548,
+      "step": 43165
+    },
+    {
+      "epoch": 1.6619826756496632,
+      "grad_norm": 0.9313360452651978,
+      "learning_rate": 1.382544750339907e-05,
+      "loss": 0.804,
+      "step": 43170
+    },
+    {
+      "epoch": 1.6621751684311838,
+      "grad_norm": 0.9089387059211731,
+      "learning_rate": 1.3810109658953252e-05,
+      "loss": 0.6871,
+      "step": 43175
+    },
+    {
+      "epoch": 1.6623676612127045,
+      "grad_norm": 1.2716861963272095,
+      "learning_rate": 1.3794779696001948e-05,
+      "loss": 0.914,
+      "step": 43180
+    },
+    {
+      "epoch": 1.6625601539942252,
+      "grad_norm": 1.5578489303588867,
+      "learning_rate": 1.3779457615947e-05,
+      "loss": 0.7148,
+      "step": 43185
+    },
+    {
+      "epoch": 1.6627526467757459,
+      "grad_norm": 0.8890004754066467,
+      "learning_rate": 1.3764143420189457e-05,
+      "loss": 0.7453,
+      "step": 43190
+    },
+    {
+      "epoch": 1.6629451395572667,
+      "grad_norm": 1.7845152616500854,
+      "learning_rate": 1.3748837110129774e-05,
+      "loss": 0.8449,
+      "step": 43195
+    },
+    {
+      "epoch": 1.6631376323387874,
+      "grad_norm": 1.1087902784347534,
+      "learning_rate": 1.3733538687167558e-05,
+      "loss": 0.8144,
+      "step": 43200
+    },
+    {
+      "epoch": 1.663330125120308,
+      "grad_norm": 1.7164037227630615,
+      "learning_rate": 1.3718248152701773e-05,
+      "loss": 0.8841,
+      "step": 43205
+    },
+    {
+      "epoch": 1.6635226179018288,
+      "grad_norm": 1.1538746356964111,
+      "learning_rate": 1.3702965508130616e-05,
+      "loss": 0.9051,
+      "step": 43210
+    },
+    {
+      "epoch": 1.6637151106833494,
+      "grad_norm": 1.328781247138977,
+      "learning_rate": 1.3687690754851634e-05,
+      "loss": 0.7957,
+      "step": 43215
+    },
+    {
+      "epoch": 1.66390760346487,
+      "grad_norm": 1.2953364849090576,
+      "learning_rate": 1.367242389426151e-05,
+      "loss": 0.8322,
+      "step": 43220
+    },
+    {
+      "epoch": 1.6641000962463908,
+      "grad_norm": 1.3127518892288208,
+      "learning_rate": 1.3657164927756405e-05,
+      "loss": 0.7086,
+      "step": 43225
+    },
+    {
+      "epoch": 1.6642925890279114,
+      "grad_norm": 1.5213780403137207,
+      "learning_rate": 1.3641913856731569e-05,
+      "loss": 0.8327,
+      "step": 43230
+    },
+    {
+      "epoch": 1.664485081809432,
+      "grad_norm": 1.8218622207641602,
+      "learning_rate": 1.3626670682581655e-05,
+      "loss": 0.9758,
+      "step": 43235
+    },
+    {
+      "epoch": 1.6646775745909528,
+      "grad_norm": 1.4787392616271973,
+      "learning_rate": 1.3611435406700546e-05,
+      "loss": 0.6978,
+      "step": 43240
+    },
+    {
+      "epoch": 1.6648700673724735,
+      "grad_norm": 1.1849435567855835,
+      "learning_rate": 1.3596208030481372e-05,
+      "loss": 0.9329,
+      "step": 43245
+    },
+    {
+      "epoch": 1.6650625601539941,
+      "grad_norm": 0.34040671586990356,
+      "learning_rate": 1.358098855531661e-05,
+      "loss": 0.67,
+      "step": 43250
+    },
+    {
+      "epoch": 1.6652550529355148,
+      "grad_norm": 1.1814277172088623,
+      "learning_rate": 1.3565776982597966e-05,
+      "loss": 0.8672,
+      "step": 43255
+    },
+    {
+      "epoch": 1.6654475457170355,
+      "grad_norm": 1.2392293214797974,
+      "learning_rate": 1.3550573313716463e-05,
+      "loss": 0.9071,
+      "step": 43260
+    },
+    {
+      "epoch": 1.6656400384985564,
+      "grad_norm": 1.1878242492675781,
+      "learning_rate": 1.3535377550062323e-05,
+      "loss": 0.7641,
+      "step": 43265
+    },
+    {
+      "epoch": 1.665832531280077,
+      "grad_norm": 0.885995090007782,
+      "learning_rate": 1.3520189693025164e-05,
+      "loss": 0.7289,
+      "step": 43270
+    },
+    {
+      "epoch": 1.6660250240615977,
+      "grad_norm": 1.223071575164795,
+      "learning_rate": 1.3505009743993757e-05,
+      "loss": 0.6685,
+      "step": 43275
+    },
+    {
+      "epoch": 1.6662175168431184,
+      "grad_norm": 1.0630251169204712,
+      "learning_rate": 1.3489837704356235e-05,
+      "loss": 0.8338,
+      "step": 43280
+    },
+    {
+      "epoch": 1.666410009624639,
+      "grad_norm": 1.187340497970581,
+      "learning_rate": 1.3474673575499986e-05,
+      "loss": 0.6967,
+      "step": 43285
+    },
+    {
+      "epoch": 1.66660250240616,
+      "grad_norm": 2.237581729888916,
+      "learning_rate": 1.3459517358811668e-05,
+      "loss": 0.8451,
+      "step": 43290
+    },
+    {
+      "epoch": 1.6667949951876806,
+      "grad_norm": 1.1633507013320923,
+      "learning_rate": 1.3444369055677175e-05,
+      "loss": 0.7811,
+      "step": 43295
+    },
+    {
+      "epoch": 1.6669874879692013,
+      "grad_norm": 1.9150440692901611,
+      "learning_rate": 1.3429228667481797e-05,
+      "loss": 0.8467,
+      "step": 43300
+    },
+    {
+      "epoch": 1.667179980750722,
+      "grad_norm": 0.9145646691322327,
+      "learning_rate": 1.341409619560996e-05,
+      "loss": 0.8669,
+      "step": 43305
+    },
+    {
+      "epoch": 1.6673724735322426,
+      "grad_norm": 1.4156630039215088,
+      "learning_rate": 1.3398971641445434e-05,
+      "loss": 0.7688,
+      "step": 43310
+    },
+    {
+      "epoch": 1.6675649663137633,
+      "grad_norm": 1.122641682624817,
+      "learning_rate": 1.3383855006371281e-05,
+      "loss": 0.9059,
+      "step": 43315
+    },
+    {
+      "epoch": 1.667757459095284,
+      "grad_norm": 1.1816550493240356,
+      "learning_rate": 1.3368746291769806e-05,
+      "loss": 0.6822,
+      "step": 43320
+    },
+    {
+      "epoch": 1.6679499518768046,
+      "grad_norm": 1.2884527444839478,
+      "learning_rate": 1.3353645499022605e-05,
+      "loss": 0.989,
+      "step": 43325
+    },
+    {
+      "epoch": 1.6681424446583253,
+      "grad_norm": 2.128404140472412,
+      "learning_rate": 1.333855262951056e-05,
+      "loss": 0.835,
+      "step": 43330
+    },
+    {
+      "epoch": 1.668334937439846,
+      "grad_norm": 1.3156408071517944,
+      "learning_rate": 1.3323467684613789e-05,
+      "loss": 0.7899,
+      "step": 43335
+    },
+    {
+      "epoch": 1.6685274302213666,
+      "grad_norm": 2.0468761920928955,
+      "learning_rate": 1.3308390665711701e-05,
+      "loss": 0.7012,
+      "step": 43340
+    },
+    {
+      "epoch": 1.6687199230028873,
+      "grad_norm": 1.0322657823562622,
+      "learning_rate": 1.3293321574183016e-05,
+      "loss": 0.6967,
+      "step": 43345
+    },
+    {
+      "epoch": 1.668912415784408,
+      "grad_norm": 0.9102632999420166,
+      "learning_rate": 1.3278260411405697e-05,
+      "loss": 0.7604,
+      "step": 43350
+    },
+    {
+      "epoch": 1.6691049085659286,
+      "grad_norm": 1.8231315612792969,
+      "learning_rate": 1.3263207178756997e-05,
+      "loss": 0.6853,
+      "step": 43355
+    },
+    {
+      "epoch": 1.6692974013474493,
+      "grad_norm": 1.3293402194976807,
+      "learning_rate": 1.3248161877613408e-05,
+      "loss": 0.9419,
+      "step": 43360
+    },
+    {
+      "epoch": 1.6694898941289702,
+      "grad_norm": 1.4297016859054565,
+      "learning_rate": 1.3233124509350736e-05,
+      "loss": 0.9883,
+      "step": 43365
+    },
+    {
+      "epoch": 1.6696823869104909,
+      "grad_norm": 1.4253541231155396,
+      "learning_rate": 1.3218095075344051e-05,
+      "loss": 0.8432,
+      "step": 43370
+    },
+    {
+      "epoch": 1.6698748796920115,
+      "grad_norm": 1.1191109418869019,
+      "learning_rate": 1.3203073576967717e-05,
+      "loss": 0.87,
+      "step": 43375
+    },
+    {
+      "epoch": 1.6700673724735322,
+      "grad_norm": 1.2235162258148193,
+      "learning_rate": 1.3188060015595271e-05,
+      "loss": 0.943,
+      "step": 43380
+    },
+    {
+      "epoch": 1.670259865255053,
+      "grad_norm": 0.8941404223442078,
+      "learning_rate": 1.3173054392599715e-05,
+      "loss": 0.7541,
+      "step": 43385
+    },
+    {
+      "epoch": 1.6704523580365738,
+      "grad_norm": 1.3058192729949951,
+      "learning_rate": 1.3158056709353139e-05,
+      "loss": 0.7509,
+      "step": 43390
+    },
+    {
+      "epoch": 1.6706448508180944,
+      "grad_norm": 1.0127733945846558,
+      "learning_rate": 1.3143066967226992e-05,
+      "loss": 0.6995,
+      "step": 43395
+    },
+    {
+      "epoch": 1.670837343599615,
+      "grad_norm": 1.4045414924621582,
+      "learning_rate": 1.3128085167592007e-05,
+      "loss": 0.751,
+      "step": 43400
+    },
+    {
+      "epoch": 1.6710298363811358,
+      "grad_norm": 1.8028631210327148,
+      "learning_rate": 1.3113111311818171e-05,
+      "loss": 0.8459,
+      "step": 43405
+    },
+    {
+      "epoch": 1.6712223291626565,
+      "grad_norm": 2.1013052463531494,
+      "learning_rate": 1.3098145401274697e-05,
+      "loss": 0.8916,
+      "step": 43410
+    },
+    {
+      "epoch": 1.6714148219441771,
+      "grad_norm": 1.4260973930358887,
+      "learning_rate": 1.3083187437330192e-05,
+      "loss": 0.6563,
+      "step": 43415
+    },
+    {
+      "epoch": 1.6716073147256978,
+      "grad_norm": 1.321631908416748,
+      "learning_rate": 1.3068237421352414e-05,
+      "loss": 0.7359,
+      "step": 43420
+    },
+    {
+      "epoch": 1.6717998075072185,
+      "grad_norm": 1.628443717956543,
+      "learning_rate": 1.3053295354708439e-05,
+      "loss": 0.7925,
+      "step": 43425
+    },
+    {
+      "epoch": 1.6719923002887391,
+      "grad_norm": 1.8474700450897217,
+      "learning_rate": 1.3038361238764641e-05,
+      "loss": 0.8156,
+      "step": 43430
+    },
+    {
+      "epoch": 1.6721847930702598,
+      "grad_norm": 1.3773387670516968,
+      "learning_rate": 1.3023435074886658e-05,
+      "loss": 0.7901,
+      "step": 43435
+    },
+    {
+      "epoch": 1.6723772858517805,
+      "grad_norm": 0.8466984033584595,
+      "learning_rate": 1.3008516864439357e-05,
+      "loss": 0.8321,
+      "step": 43440
+    },
+    {
+      "epoch": 1.6725697786333011,
+      "grad_norm": 1.169967770576477,
+      "learning_rate": 1.2993606608786913e-05,
+      "loss": 0.7029,
+      "step": 43445
+    },
+    {
+      "epoch": 1.6727622714148218,
+      "grad_norm": 1.8645579814910889,
+      "learning_rate": 1.2978704309292789e-05,
+      "loss": 0.8151,
+      "step": 43450
+    },
+    {
+      "epoch": 1.6729547641963425,
+      "grad_norm": 0.84092777967453,
+      "learning_rate": 1.2963809967319685e-05,
+      "loss": 0.8796,
+      "step": 43455
+    },
+    {
+      "epoch": 1.6731472569778634,
+      "grad_norm": 1.5797427892684937,
+      "learning_rate": 1.2948923584229622e-05,
+      "loss": 0.8069,
+      "step": 43460
+    },
+    {
+      "epoch": 1.673339749759384,
+      "grad_norm": 1.3241761922836304,
+      "learning_rate": 1.2934045161383824e-05,
+      "loss": 0.9362,
+      "step": 43465
+    },
+    {
+      "epoch": 1.6735322425409047,
+      "grad_norm": 1.459402322769165,
+      "learning_rate": 1.2919174700142822e-05,
+      "loss": 0.8827,
+      "step": 43470
+    },
+    {
+      "epoch": 1.6737247353224254,
+      "grad_norm": 1.3563963174819946,
+      "learning_rate": 1.2904312201866443e-05,
+      "loss": 0.7035,
+      "step": 43475
+    },
+    {
+      "epoch": 1.673917228103946,
+      "grad_norm": 1.3299168348312378,
+      "learning_rate": 1.2889457667913785e-05,
+      "loss": 0.8429,
+      "step": 43480
+    },
+    {
+      "epoch": 1.674109720885467,
+      "grad_norm": 1.492050051689148,
+      "learning_rate": 1.2874611099643108e-05,
+      "loss": 0.9821,
+      "step": 43485
+    },
+    {
+      "epoch": 1.6743022136669876,
+      "grad_norm": 0.9983692169189453,
+      "learning_rate": 1.2859772498412149e-05,
+      "loss": 0.6268,
+      "step": 43490
+    },
+    {
+      "epoch": 1.6744947064485083,
+      "grad_norm": 1.0508897304534912,
+      "learning_rate": 1.2844941865577719e-05,
+      "loss": 0.8499,
+      "step": 43495
+    },
+    {
+      "epoch": 1.674687199230029,
+      "grad_norm": 1.9004470109939575,
+      "learning_rate": 1.2830119202496016e-05,
+      "loss": 0.8397,
+      "step": 43500
+    },
+    {
+      "epoch": 1.6748796920115496,
+      "grad_norm": 1.5322777032852173,
+      "learning_rate": 1.2815304510522453e-05,
+      "loss": 0.9092,
+      "step": 43505
+    },
+    {
+      "epoch": 1.6750721847930703,
+      "grad_norm": 1.5270432233810425,
+      "learning_rate": 1.2800497791011768e-05,
+      "loss": 0.739,
+      "step": 43510
+    },
+    {
+      "epoch": 1.675264677574591,
+      "grad_norm": 1.1705281734466553,
+      "learning_rate": 1.2785699045317878e-05,
+      "loss": 0.7841,
+      "step": 43515
+    },
+    {
+      "epoch": 1.6754571703561116,
+      "grad_norm": 1.2318379878997803,
+      "learning_rate": 1.2770908274794102e-05,
+      "loss": 0.7705,
+      "step": 43520
+    },
+    {
+      "epoch": 1.6756496631376323,
+      "grad_norm": 1.481648325920105,
+      "learning_rate": 1.2756125480792912e-05,
+      "loss": 0.92,
+      "step": 43525
+    },
+    {
+      "epoch": 1.675842155919153,
+      "grad_norm": 1.4268490076065063,
+      "learning_rate": 1.2741350664666108e-05,
+      "loss": 0.8993,
+      "step": 43530
+    },
+    {
+      "epoch": 1.6760346487006736,
+      "grad_norm": 1.973768711090088,
+      "learning_rate": 1.2726583827764748e-05,
+      "loss": 0.7195,
+      "step": 43535
+    },
+    {
+      "epoch": 1.6762271414821943,
+      "grad_norm": 1.2458196878433228,
+      "learning_rate": 1.271182497143919e-05,
+      "loss": 0.7451,
+      "step": 43540
+    },
+    {
+      "epoch": 1.676419634263715,
+      "grad_norm": 1.4529706239700317,
+      "learning_rate": 1.2697074097038964e-05,
+      "loss": 0.726,
+      "step": 43545
+    },
+    {
+      "epoch": 1.6766121270452357,
+      "grad_norm": 2.0508511066436768,
+      "learning_rate": 1.2682331205913012e-05,
+      "loss": 0.8247,
+      "step": 43550
+    },
+    {
+      "epoch": 1.6768046198267565,
+      "grad_norm": 1.2535037994384766,
+      "learning_rate": 1.2667596299409434e-05,
+      "loss": 0.8559,
+      "step": 43555
+    },
+    {
+      "epoch": 1.6769971126082772,
+      "grad_norm": 1.0279282331466675,
+      "learning_rate": 1.2652869378875654e-05,
+      "loss": 0.7823,
+      "step": 43560
+    },
+    {
+      "epoch": 1.6771896053897979,
+      "grad_norm": 0.9795730113983154,
+      "learning_rate": 1.2638150445658337e-05,
+      "loss": 0.6945,
+      "step": 43565
+    },
+    {
+      "epoch": 1.6773820981713186,
+      "grad_norm": 1.2508692741394043,
+      "learning_rate": 1.2623439501103452e-05,
+      "loss": 0.6393,
+      "step": 43570
+    },
+    {
+      "epoch": 1.6775745909528392,
+      "grad_norm": 1.5805447101593018,
+      "learning_rate": 1.260873654655622e-05,
+      "loss": 0.7677,
+      "step": 43575
+    },
+    {
+      "epoch": 1.6777670837343601,
+      "grad_norm": 1.477809190750122,
+      "learning_rate": 1.2594041583361105e-05,
+      "loss": 0.8169,
+      "step": 43580
+    },
+    {
+      "epoch": 1.6779595765158808,
+      "grad_norm": 1.8884755373001099,
+      "learning_rate": 1.257935461286187e-05,
+      "loss": 0.7033,
+      "step": 43585
+    },
+    {
+      "epoch": 1.6781520692974015,
+      "grad_norm": 1.1278042793273926,
+      "learning_rate": 1.2564675636401557e-05,
+      "loss": 0.6912,
+      "step": 43590
+    },
+    {
+      "epoch": 1.6783445620789221,
+      "grad_norm": 1.1749690771102905,
+      "learning_rate": 1.2550004655322457e-05,
+      "loss": 0.8726,
+      "step": 43595
+    },
+    {
+      "epoch": 1.6785370548604428,
+      "grad_norm": 2.014747142791748,
+      "learning_rate": 1.2535341670966094e-05,
+      "loss": 0.835,
+      "step": 43600
+    },
+    {
+      "epoch": 1.6787295476419635,
+      "grad_norm": 1.0013171434402466,
+      "learning_rate": 1.2520686684673377e-05,
+      "loss": 0.7554,
+      "step": 43605
+    },
+    {
+      "epoch": 1.6789220404234841,
+      "grad_norm": 2.000584602355957,
+      "learning_rate": 1.2506039697784345e-05,
+      "loss": 0.9073,
+      "step": 43610
+    },
+    {
+      "epoch": 1.6791145332050048,
+      "grad_norm": 1.03412926197052,
+      "learning_rate": 1.2491400711638378e-05,
+      "loss": 0.9339,
+      "step": 43615
+    },
+    {
+      "epoch": 1.6793070259865255,
+      "grad_norm": 1.905938744544983,
+      "learning_rate": 1.2476769727574133e-05,
+      "loss": 0.7793,
+      "step": 43620
+    },
+    {
+      "epoch": 1.6794995187680462,
+      "grad_norm": 1.2626748085021973,
+      "learning_rate": 1.2462146746929538e-05,
+      "loss": 0.6789,
+      "step": 43625
+    },
+    {
+      "epoch": 1.6796920115495668,
+      "grad_norm": 1.5359333753585815,
+      "learning_rate": 1.2447531771041677e-05,
+      "loss": 0.7657,
+      "step": 43630
+    },
+    {
+      "epoch": 1.6798845043310875,
+      "grad_norm": 0.68184894323349,
+      "learning_rate": 1.2432924801247115e-05,
+      "loss": 1.0114,
+      "step": 43635
+    },
+    {
+      "epoch": 1.6800769971126082,
+      "grad_norm": 2.0016860961914062,
+      "learning_rate": 1.2418325838881462e-05,
+      "loss": 0.7352,
+      "step": 43640
+    },
+    {
+      "epoch": 1.6802694898941288,
+      "grad_norm": 1.29106605052948,
+      "learning_rate": 1.240373488527975e-05,
+      "loss": 0.743,
+      "step": 43645
+    },
+    {
+      "epoch": 1.6804619826756495,
+      "grad_norm": 1.8269158601760864,
+      "learning_rate": 1.23891519417762e-05,
+      "loss": 0.8502,
+      "step": 43650
+    },
+    {
+      "epoch": 1.6806544754571704,
+      "grad_norm": 0.9514836072921753,
+      "learning_rate": 1.2374577009704357e-05,
+      "loss": 0.8288,
+      "step": 43655
+    },
+    {
+      "epoch": 1.680846968238691,
+      "grad_norm": 1.3545150756835938,
+      "learning_rate": 1.2360010090396968e-05,
+      "loss": 0.9582,
+      "step": 43660
+    },
+    {
+      "epoch": 1.6810394610202117,
+      "grad_norm": 1.4287010431289673,
+      "learning_rate": 1.2345451185186097e-05,
+      "loss": 0.8761,
+      "step": 43665
+    },
+    {
+      "epoch": 1.6812319538017324,
+      "grad_norm": 0.9901794195175171,
+      "learning_rate": 1.2330900295403048e-05,
+      "loss": 0.878,
+      "step": 43670
+    },
+    {
+      "epoch": 1.681424446583253,
+      "grad_norm": 1.1503381729125977,
+      "learning_rate": 1.231635742237841e-05,
+      "loss": 0.747,
+      "step": 43675
+    },
+    {
+      "epoch": 1.681616939364774,
+      "grad_norm": 0.8054229617118835,
+      "learning_rate": 1.2301822567442067e-05,
+      "loss": 0.724,
+      "step": 43680
+    },
+    {
+      "epoch": 1.6818094321462946,
+      "grad_norm": 2.6240620613098145,
+      "learning_rate": 1.2287295731923077e-05,
+      "loss": 0.7472,
+      "step": 43685
+    },
+    {
+      "epoch": 1.6820019249278153,
+      "grad_norm": 0.9073877930641174,
+      "learning_rate": 1.2272776917149841e-05,
+      "loss": 0.7837,
+      "step": 43690
+    },
+    {
+      "epoch": 1.682194417709336,
+      "grad_norm": 0.9025770425796509,
+      "learning_rate": 1.2258266124450024e-05,
+      "loss": 0.8009,
+      "step": 43695
+    },
+    {
+      "epoch": 1.6823869104908566,
+      "grad_norm": 0.9586546421051025,
+      "learning_rate": 1.224376335515055e-05,
+      "loss": 0.8342,
+      "step": 43700
+    },
+    {
+      "epoch": 1.6825794032723773,
+      "grad_norm": 1.0272797346115112,
+      "learning_rate": 1.222926861057755e-05,
+      "loss": 0.8561,
+      "step": 43705
+    },
+    {
+      "epoch": 1.682771896053898,
+      "grad_norm": 1.4825763702392578,
+      "learning_rate": 1.2214781892056548e-05,
+      "loss": 0.938,
+      "step": 43710
+    },
+    {
+      "epoch": 1.6829643888354187,
+      "grad_norm": 1.5734376907348633,
+      "learning_rate": 1.2200303200912199e-05,
+      "loss": 0.7255,
+      "step": 43715
+    },
+    {
+      "epoch": 1.6831568816169393,
+      "grad_norm": 1.2651904821395874,
+      "learning_rate": 1.21858325384685e-05,
+      "loss": 0.762,
+      "step": 43720
+    },
+    {
+      "epoch": 1.68334937439846,
+      "grad_norm": 1.516546368598938,
+      "learning_rate": 1.2171369906048703e-05,
+      "loss": 0.7557,
+      "step": 43725
+    },
+    {
+      "epoch": 1.6835418671799807,
+      "grad_norm": 1.8338733911514282,
+      "learning_rate": 1.2156915304975325e-05,
+      "loss": 0.7487,
+      "step": 43730
+    },
+    {
+      "epoch": 1.6837343599615013,
+      "grad_norm": 2.001396417617798,
+      "learning_rate": 1.21424687365701e-05,
+      "loss": 0.8106,
+      "step": 43735
+    },
+    {
+      "epoch": 1.683926852743022,
+      "grad_norm": 1.8002668619155884,
+      "learning_rate": 1.212803020215415e-05,
+      "loss": 0.6761,
+      "step": 43740
+    },
+    {
+      "epoch": 1.6841193455245427,
+      "grad_norm": 1.0602316856384277,
+      "learning_rate": 1.2113599703047728e-05,
+      "loss": 0.7588,
+      "step": 43745
+    },
+    {
+      "epoch": 1.6843118383060636,
+      "grad_norm": 1.10280442237854,
+      "learning_rate": 1.2099177240570403e-05,
+      "loss": 0.8926,
+      "step": 43750
+    },
+    {
+      "epoch": 1.6845043310875842,
+      "grad_norm": 1.1837788820266724,
+      "learning_rate": 1.208476281604104e-05,
+      "loss": 0.8216,
+      "step": 43755
+    },
+    {
+      "epoch": 1.684696823869105,
+      "grad_norm": 1.3022428750991821,
+      "learning_rate": 1.2070356430777752e-05,
+      "loss": 0.6659,
+      "step": 43760
+    },
+    {
+      "epoch": 1.6848893166506256,
+      "grad_norm": 1.8487025499343872,
+      "learning_rate": 1.205595808609784e-05,
+      "loss": 0.7536,
+      "step": 43765
+    },
+    {
+      "epoch": 1.6850818094321462,
+      "grad_norm": 1.2362842559814453,
+      "learning_rate": 1.2041567783318031e-05,
+      "loss": 0.8764,
+      "step": 43770
+    },
+    {
+      "epoch": 1.6852743022136671,
+      "grad_norm": 1.0950088500976562,
+      "learning_rate": 1.2027185523754159e-05,
+      "loss": 0.9292,
+      "step": 43775
+    },
+    {
+      "epoch": 1.6854667949951878,
+      "grad_norm": 1.8685953617095947,
+      "learning_rate": 1.2012811308721395e-05,
+      "loss": 0.6901,
+      "step": 43780
+    },
+    {
+      "epoch": 1.6856592877767085,
+      "grad_norm": 1.1124467849731445,
+      "learning_rate": 1.1998445139534209e-05,
+      "loss": 0.835,
+      "step": 43785
+    },
+    {
+      "epoch": 1.6858517805582292,
+      "grad_norm": 1.07558012008667,
+      "learning_rate": 1.1984087017506228e-05,
+      "loss": 0.8145,
+      "step": 43790
+    },
+    {
+      "epoch": 1.6860442733397498,
+      "grad_norm": 1.631927490234375,
+      "learning_rate": 1.1969736943950439e-05,
+      "loss": 0.9775,
+      "step": 43795
+    },
+    {
+      "epoch": 1.6862367661212705,
+      "grad_norm": 1.3678085803985596,
+      "learning_rate": 1.1955394920179053e-05,
+      "loss": 0.7821,
+      "step": 43800
+    },
+    {
+      "epoch": 1.6864292589027912,
+      "grad_norm": 1.4719831943511963,
+      "learning_rate": 1.1941060947503591e-05,
+      "loss": 0.7762,
+      "step": 43805
+    },
+    {
+      "epoch": 1.6866217516843118,
+      "grad_norm": 1.012670636177063,
+      "learning_rate": 1.1926735027234726e-05,
+      "loss": 0.8692,
+      "step": 43810
+    },
+    {
+      "epoch": 1.6868142444658325,
+      "grad_norm": 1.3539128303527832,
+      "learning_rate": 1.1912417160682543e-05,
+      "loss": 0.6571,
+      "step": 43815
+    },
+    {
+      "epoch": 1.6870067372473532,
+      "grad_norm": 1.2712576389312744,
+      "learning_rate": 1.1898107349156274e-05,
+      "loss": 0.7417,
+      "step": 43820
+    },
+    {
+      "epoch": 1.6871992300288738,
+      "grad_norm": 2.003711462020874,
+      "learning_rate": 1.188380559396446e-05,
+      "loss": 0.8174,
+      "step": 43825
+    },
+    {
+      "epoch": 1.6873917228103945,
+      "grad_norm": 2.019716739654541,
+      "learning_rate": 1.186951189641491e-05,
+      "loss": 0.6646,
+      "step": 43830
+    },
+    {
+      "epoch": 1.6875842155919152,
+      "grad_norm": 1.0991610288619995,
+      "learning_rate": 1.1855226257814688e-05,
+      "loss": 0.7247,
+      "step": 43835
+    },
+    {
+      "epoch": 1.6877767083734359,
+      "grad_norm": 1.6221376657485962,
+      "learning_rate": 1.184094867947011e-05,
+      "loss": 0.8466,
+      "step": 43840
+    },
+    {
+      "epoch": 1.6879692011549567,
+      "grad_norm": 1.4020202159881592,
+      "learning_rate": 1.1826679162686805e-05,
+      "loss": 0.8511,
+      "step": 43845
+    },
+    {
+      "epoch": 1.6881616939364774,
+      "grad_norm": 1.3694734573364258,
+      "learning_rate": 1.1812417708769552e-05,
+      "loss": 0.8935,
+      "step": 43850
+    },
+    {
+      "epoch": 1.688354186717998,
+      "grad_norm": 1.3666578531265259,
+      "learning_rate": 1.1798164319022554e-05,
+      "loss": 0.8224,
+      "step": 43855
+    },
+    {
+      "epoch": 1.6885466794995188,
+      "grad_norm": 1.5298011302947998,
+      "learning_rate": 1.1783918994749122e-05,
+      "loss": 0.834,
+      "step": 43860
+    },
+    {
+      "epoch": 1.6887391722810394,
+      "grad_norm": 1.589240550994873,
+      "learning_rate": 1.1769681737251914e-05,
+      "loss": 0.7738,
+      "step": 43865
+    },
+    {
+      "epoch": 1.6889316650625603,
+      "grad_norm": 1.132411241531372,
+      "learning_rate": 1.1755452547832846e-05,
+      "loss": 0.8,
+      "step": 43870
+    },
+    {
+      "epoch": 1.689124157844081,
+      "grad_norm": 0.9924208521842957,
+      "learning_rate": 1.1741231427793097e-05,
+      "loss": 0.7401,
+      "step": 43875
+    },
+    {
+      "epoch": 1.6893166506256017,
+      "grad_norm": 1.3757164478302002,
+      "learning_rate": 1.172701837843304e-05,
+      "loss": 0.9371,
+      "step": 43880
+    },
+    {
+      "epoch": 1.6895091434071223,
+      "grad_norm": 1.4748291969299316,
+      "learning_rate": 1.1712813401052414e-05,
+      "loss": 0.7713,
+      "step": 43885
+    },
+    {
+      "epoch": 1.689701636188643,
+      "grad_norm": 2.083977460861206,
+      "learning_rate": 1.1698616496950143e-05,
+      "loss": 0.7284,
+      "step": 43890
+    },
+    {
+      "epoch": 1.6898941289701637,
+      "grad_norm": 1.163346529006958,
+      "learning_rate": 1.1684427667424458e-05,
+      "loss": 0.7961,
+      "step": 43895
+    },
+    {
+      "epoch": 1.6900866217516843,
+      "grad_norm": 1.0689622163772583,
+      "learning_rate": 1.1670246913772841e-05,
+      "loss": 0.8038,
+      "step": 43900
+    },
+    {
+      "epoch": 1.690279114533205,
+      "grad_norm": 1.596536636352539,
+      "learning_rate": 1.1656074237291991e-05,
+      "loss": 0.852,
+      "step": 43905
+    },
+    {
+      "epoch": 1.6904716073147257,
+      "grad_norm": 2.0381526947021484,
+      "learning_rate": 1.1641909639277936e-05,
+      "loss": 0.8369,
+      "step": 43910
+    },
+    {
+      "epoch": 1.6906641000962463,
+      "grad_norm": 0.8638001084327698,
+      "learning_rate": 1.162775312102592e-05,
+      "loss": 0.5935,
+      "step": 43915
+    },
+    {
+      "epoch": 1.690856592877767,
+      "grad_norm": 1.7793620824813843,
+      "learning_rate": 1.161360468383048e-05,
+      "loss": 0.8603,
+      "step": 43920
+    },
+    {
+      "epoch": 1.6910490856592877,
+      "grad_norm": 1.2274523973464966,
+      "learning_rate": 1.1599464328985355e-05,
+      "loss": 0.8225,
+      "step": 43925
+    },
+    {
+      "epoch": 1.6912415784408084,
+      "grad_norm": 1.3787938356399536,
+      "learning_rate": 1.158533205778366e-05,
+      "loss": 0.9415,
+      "step": 43930
+    },
+    {
+      "epoch": 1.691434071222329,
+      "grad_norm": 1.7021337747573853,
+      "learning_rate": 1.157120787151763e-05,
+      "loss": 0.9423,
+      "step": 43935
+    },
+    {
+      "epoch": 1.6916265640038497,
+      "grad_norm": 1.1812869310379028,
+      "learning_rate": 1.1557091771478855e-05,
+      "loss": 0.8007,
+      "step": 43940
+    },
+    {
+      "epoch": 1.6918190567853706,
+      "grad_norm": 1.0627899169921875,
+      "learning_rate": 1.1542983758958148e-05,
+      "loss": 0.8179,
+      "step": 43945
+    },
+    {
+      "epoch": 1.6920115495668913,
+      "grad_norm": 1.115240216255188,
+      "learning_rate": 1.152888383524563e-05,
+      "loss": 0.811,
+      "step": 43950
+    },
+    {
+      "epoch": 1.692204042348412,
+      "grad_norm": 0.7895986437797546,
+      "learning_rate": 1.151479200163058e-05,
+      "loss": 0.8463,
+      "step": 43955
+    },
+    {
+      "epoch": 1.6923965351299326,
+      "grad_norm": 1.0171902179718018,
+      "learning_rate": 1.1500708259401682e-05,
+      "loss": 0.7925,
+      "step": 43960
+    },
+    {
+      "epoch": 1.6925890279114533,
+      "grad_norm": 0.9279621243476868,
+      "learning_rate": 1.1486632609846726e-05,
+      "loss": 0.7073,
+      "step": 43965
+    },
+    {
+      "epoch": 1.6927815206929742,
+      "grad_norm": 1.056172490119934,
+      "learning_rate": 1.1472565054252882e-05,
+      "loss": 0.7971,
+      "step": 43970
+    },
+    {
+      "epoch": 1.6929740134744948,
+      "grad_norm": 1.2332504987716675,
+      "learning_rate": 1.1458505593906522e-05,
+      "loss": 0.8648,
+      "step": 43975
+    },
+    {
+      "epoch": 1.6931665062560155,
+      "grad_norm": 1.2039694786071777,
+      "learning_rate": 1.1444454230093315e-05,
+      "loss": 0.7207,
+      "step": 43980
+    },
+    {
+      "epoch": 1.6933589990375362,
+      "grad_norm": 1.3070398569107056,
+      "learning_rate": 1.1430410964098115e-05,
+      "loss": 0.6398,
+      "step": 43985
+    },
+    {
+      "epoch": 1.6935514918190568,
+      "grad_norm": 1.1500900983810425,
+      "learning_rate": 1.1416375797205114e-05,
+      "loss": 0.7082,
+      "step": 43990
+    },
+    {
+      "epoch": 1.6937439846005775,
+      "grad_norm": 2.088250160217285,
+      "learning_rate": 1.1402348730697731e-05,
+      "loss": 0.7245,
+      "step": 43995
+    },
+    {
+      "epoch": 1.6939364773820982,
+      "grad_norm": 1.7187893390655518,
+      "learning_rate": 1.1388329765858651e-05,
+      "loss": 0.7806,
+      "step": 44000
+    },
+    {
+      "epoch": 1.6941289701636189,
+      "grad_norm": 1.2111575603485107,
+      "learning_rate": 1.137431890396985e-05,
+      "loss": 0.6781,
+      "step": 44005
+    },
+    {
+      "epoch": 1.6943214629451395,
+      "grad_norm": 1.4132143259048462,
+      "learning_rate": 1.1360316146312455e-05,
+      "loss": 1.0329,
+      "step": 44010
+    },
+    {
+      "epoch": 1.6945139557266602,
+      "grad_norm": 1.3241591453552246,
+      "learning_rate": 1.1346321494166978e-05,
+      "loss": 0.8164,
+      "step": 44015
+    },
+    {
+      "epoch": 1.6947064485081809,
+      "grad_norm": 2.9741406440734863,
+      "learning_rate": 1.1332334948813117e-05,
+      "loss": 0.8342,
+      "step": 44020
+    },
+    {
+      "epoch": 1.6948989412897015,
+      "grad_norm": 1.5332659482955933,
+      "learning_rate": 1.1318356511529871e-05,
+      "loss": 0.6292,
+      "step": 44025
+    },
+    {
+      "epoch": 1.6950914340712222,
+      "grad_norm": 1.2463070154190063,
+      "learning_rate": 1.1304386183595428e-05,
+      "loss": 0.7732,
+      "step": 44030
+    },
+    {
+      "epoch": 1.6952839268527429,
+      "grad_norm": 1.6358565092086792,
+      "learning_rate": 1.1290423966287345e-05,
+      "loss": 0.7089,
+      "step": 44035
+    },
+    {
+      "epoch": 1.6954764196342638,
+      "grad_norm": 1.1293914318084717,
+      "learning_rate": 1.1276469860882332e-05,
+      "loss": 0.8117,
+      "step": 44040
+    },
+    {
+      "epoch": 1.6956689124157844,
+      "grad_norm": 1.2490708827972412,
+      "learning_rate": 1.1262523868656405e-05,
+      "loss": 0.8166,
+      "step": 44045
+    },
+    {
+      "epoch": 1.695861405197305,
+      "grad_norm": 1.3659740686416626,
+      "learning_rate": 1.124858599088484e-05,
+      "loss": 0.7455,
+      "step": 44050
+    },
+    {
+      "epoch": 1.6960538979788258,
+      "grad_norm": 1.2188682556152344,
+      "learning_rate": 1.1234656228842177e-05,
+      "loss": 0.8226,
+      "step": 44055
+    },
+    {
+      "epoch": 1.6962463907603464,
+      "grad_norm": 1.0454777479171753,
+      "learning_rate": 1.122073458380215e-05,
+      "loss": 0.8888,
+      "step": 44060
+    },
+    {
+      "epoch": 1.6964388835418673,
+      "grad_norm": 2.2984609603881836,
+      "learning_rate": 1.1206821057037886e-05,
+      "loss": 0.8094,
+      "step": 44065
+    },
+    {
+      "epoch": 1.696631376323388,
+      "grad_norm": 1.107490062713623,
+      "learning_rate": 1.11929156498216e-05,
+      "loss": 0.8469,
+      "step": 44070
+    },
+    {
+      "epoch": 1.6968238691049087,
+      "grad_norm": 0.9354246854782104,
+      "learning_rate": 1.1179018363424899e-05,
+      "loss": 0.6261,
+      "step": 44075
+    },
+    {
+      "epoch": 1.6970163618864293,
+      "grad_norm": 1.164993405342102,
+      "learning_rate": 1.1165129199118574e-05,
+      "loss": 0.61,
+      "step": 44080
+    },
+    {
+      "epoch": 1.69720885466795,
+      "grad_norm": 0.9452062845230103,
+      "learning_rate": 1.1151248158172722e-05,
+      "loss": 0.7721,
+      "step": 44085
+    },
+    {
+      "epoch": 1.6974013474494707,
+      "grad_norm": 0.8486310839653015,
+      "learning_rate": 1.1137375241856619e-05,
+      "loss": 0.9994,
+      "step": 44090
+    },
+    {
+      "epoch": 1.6975938402309914,
+      "grad_norm": 1.195550560951233,
+      "learning_rate": 1.1123510451438934e-05,
+      "loss": 0.812,
+      "step": 44095
+    },
+    {
+      "epoch": 1.697786333012512,
+      "grad_norm": 1.6935982704162598,
+      "learning_rate": 1.1109653788187447e-05,
+      "loss": 0.7816,
+      "step": 44100
+    },
+    {
+      "epoch": 1.6979788257940327,
+      "grad_norm": 1.238194465637207,
+      "learning_rate": 1.1095805253369274e-05,
+      "loss": 0.8888,
+      "step": 44105
+    },
+    {
+      "epoch": 1.6981713185755534,
+      "grad_norm": 1.1349773406982422,
+      "learning_rate": 1.108196484825077e-05,
+      "loss": 0.6549,
+      "step": 44110
+    },
+    {
+      "epoch": 1.698363811357074,
+      "grad_norm": 1.0032782554626465,
+      "learning_rate": 1.1068132574097557e-05,
+      "loss": 0.7298,
+      "step": 44115
+    },
+    {
+      "epoch": 1.6985563041385947,
+      "grad_norm": 2.31643009185791,
+      "learning_rate": 1.1054308432174521e-05,
+      "loss": 0.871,
+      "step": 44120
+    },
+    {
+      "epoch": 1.6987487969201154,
+      "grad_norm": 1.0429223775863647,
+      "learning_rate": 1.1040492423745752e-05,
+      "loss": 0.7892,
+      "step": 44125
+    },
+    {
+      "epoch": 1.698941289701636,
+      "grad_norm": 1.4634356498718262,
+      "learning_rate": 1.102668455007464e-05,
+      "loss": 0.679,
+      "step": 44130
+    },
+    {
+      "epoch": 1.6991337824831567,
+      "grad_norm": 1.5524640083312988,
+      "learning_rate": 1.1012884812423829e-05,
+      "loss": 0.9072,
+      "step": 44135
+    },
+    {
+      "epoch": 1.6993262752646776,
+      "grad_norm": 0.9184067845344543,
+      "learning_rate": 1.0999093212055244e-05,
+      "loss": 0.6238,
+      "step": 44140
+    },
+    {
+      "epoch": 1.6995187680461983,
+      "grad_norm": 1.144820213317871,
+      "learning_rate": 1.0985309750229966e-05,
+      "loss": 0.7566,
+      "step": 44145
+    },
+    {
+      "epoch": 1.699711260827719,
+      "grad_norm": 1.1043323278427124,
+      "learning_rate": 1.0971534428208485e-05,
+      "loss": 0.7834,
+      "step": 44150
+    },
+    {
+      "epoch": 1.6999037536092396,
+      "grad_norm": 0.9373242259025574,
+      "learning_rate": 1.0957767247250395e-05,
+      "loss": 0.7216,
+      "step": 44155
+    },
+    {
+      "epoch": 1.7000962463907605,
+      "grad_norm": 0.977449357509613,
+      "learning_rate": 1.0944008208614643e-05,
+      "loss": 0.7585,
+      "step": 44160
+    },
+    {
+      "epoch": 1.7002887391722812,
+      "grad_norm": 1.5112518072128296,
+      "learning_rate": 1.09302573135594e-05,
+      "loss": 0.7718,
+      "step": 44165
+    },
+    {
+      "epoch": 1.7004812319538019,
+      "grad_norm": 1.095468282699585,
+      "learning_rate": 1.0916514563342106e-05,
+      "loss": 0.9487,
+      "step": 44170
+    },
+    {
+      "epoch": 1.7006737247353225,
+      "grad_norm": 1.084921956062317,
+      "learning_rate": 1.0902779959219401e-05,
+      "loss": 0.8464,
+      "step": 44175
+    },
+    {
+      "epoch": 1.7008662175168432,
+      "grad_norm": 2.3119983673095703,
+      "learning_rate": 1.0889053502447278e-05,
+      "loss": 0.7722,
+      "step": 44180
+    },
+    {
+      "epoch": 1.7010587102983639,
+      "grad_norm": 0.5821973085403442,
+      "learning_rate": 1.08753351942809e-05,
+      "loss": 0.715,
+      "step": 44185
+    },
+    {
+      "epoch": 1.7012512030798845,
+      "grad_norm": 1.1047406196594238,
+      "learning_rate": 1.086162503597472e-05,
+      "loss": 0.7629,
+      "step": 44190
+    },
+    {
+      "epoch": 1.7014436958614052,
+      "grad_norm": 1.3281866312026978,
+      "learning_rate": 1.0847923028782437e-05,
+      "loss": 0.7576,
+      "step": 44195
+    },
+    {
+      "epoch": 1.7016361886429259,
+      "grad_norm": 1.2818950414657593,
+      "learning_rate": 1.0834229173957045e-05,
+      "loss": 0.722,
+      "step": 44200
+    },
+    {
+      "epoch": 1.7018286814244465,
+      "grad_norm": 0.4923422336578369,
+      "learning_rate": 1.08205434727507e-05,
+      "loss": 0.7202,
+      "step": 44205
+    },
+    {
+      "epoch": 1.7020211742059672,
+      "grad_norm": 1.5802128314971924,
+      "learning_rate": 1.0806865926414889e-05,
+      "loss": 0.7272,
+      "step": 44210
+    },
+    {
+      "epoch": 1.7022136669874879,
+      "grad_norm": 2.5249600410461426,
+      "learning_rate": 1.0793196536200346e-05,
+      "loss": 0.8254,
+      "step": 44215
+    },
+    {
+      "epoch": 1.7024061597690086,
+      "grad_norm": 1.4376665353775024,
+      "learning_rate": 1.0779535303357035e-05,
+      "loss": 0.8114,
+      "step": 44220
+    },
+    {
+      "epoch": 1.7025986525505292,
+      "grad_norm": 1.4501975774765015,
+      "learning_rate": 1.0765882229134205e-05,
+      "loss": 0.953,
+      "step": 44225
+    },
+    {
+      "epoch": 1.70279114533205,
+      "grad_norm": 1.0345442295074463,
+      "learning_rate": 1.0752237314780311e-05,
+      "loss": 0.7878,
+      "step": 44230
+    },
+    {
+      "epoch": 1.7029836381135708,
+      "grad_norm": 0.9220666289329529,
+      "learning_rate": 1.07386005615431e-05,
+      "loss": 0.8297,
+      "step": 44235
+    },
+    {
+      "epoch": 1.7031761308950915,
+      "grad_norm": 1.8300734758377075,
+      "learning_rate": 1.0724971970669561e-05,
+      "loss": 0.7797,
+      "step": 44240
+    },
+    {
+      "epoch": 1.7033686236766121,
+      "grad_norm": 1.5990431308746338,
+      "learning_rate": 1.0711351543405967e-05,
+      "loss": 0.79,
+      "step": 44245
+    },
+    {
+      "epoch": 1.7035611164581328,
+      "grad_norm": 1.114249587059021,
+      "learning_rate": 1.0697739280997753e-05,
+      "loss": 0.7867,
+      "step": 44250
+    },
+    {
+      "epoch": 1.7037536092396535,
+      "grad_norm": 1.217193603515625,
+      "learning_rate": 1.0684135184689748e-05,
+      "loss": 0.756,
+      "step": 44255
+    },
+    {
+      "epoch": 1.7039461020211744,
+      "grad_norm": 1.1771777868270874,
+      "learning_rate": 1.0670539255725886e-05,
+      "loss": 0.7378,
+      "step": 44260
+    },
+    {
+      "epoch": 1.704138594802695,
+      "grad_norm": 1.1569862365722656,
+      "learning_rate": 1.0656951495349466e-05,
+      "loss": 0.8281,
+      "step": 44265
+    },
+    {
+      "epoch": 1.7043310875842157,
+      "grad_norm": 1.5429112911224365,
+      "learning_rate": 1.064337190480299e-05,
+      "loss": 0.7918,
+      "step": 44270
+    },
+    {
+      "epoch": 1.7045235803657364,
+      "grad_norm": 1.192523717880249,
+      "learning_rate": 1.0629800485328235e-05,
+      "loss": 0.8646,
+      "step": 44275
+    },
+    {
+      "epoch": 1.704716073147257,
+      "grad_norm": 2.4051706790924072,
+      "learning_rate": 1.061623723816616e-05,
+      "loss": 0.7809,
+      "step": 44280
+    },
+    {
+      "epoch": 1.7049085659287777,
+      "grad_norm": 1.8021824359893799,
+      "learning_rate": 1.0602682164557121e-05,
+      "loss": 0.8048,
+      "step": 44285
+    },
+    {
+      "epoch": 1.7051010587102984,
+      "grad_norm": 0.9928067922592163,
+      "learning_rate": 1.0589135265740569e-05,
+      "loss": 0.8615,
+      "step": 44290
+    },
+    {
+      "epoch": 1.705293551491819,
+      "grad_norm": 0.8101016879081726,
+      "learning_rate": 1.0575596542955312e-05,
+      "loss": 0.6574,
+      "step": 44295
+    },
+    {
+      "epoch": 1.7054860442733397,
+      "grad_norm": 1.1880675554275513,
+      "learning_rate": 1.0562065997439364e-05,
+      "loss": 0.7205,
+      "step": 44300
+    },
+    {
+      "epoch": 1.7056785370548604,
+      "grad_norm": 1.8503644466400146,
+      "learning_rate": 1.054854363043003e-05,
+      "loss": 0.8311,
+      "step": 44305
+    },
+    {
+      "epoch": 1.705871029836381,
+      "grad_norm": 1.2988847494125366,
+      "learning_rate": 1.053502944316378e-05,
+      "loss": 0.8,
+      "step": 44310
+    },
+    {
+      "epoch": 1.7060635226179017,
+      "grad_norm": 1.75754714012146,
+      "learning_rate": 1.0521523436876479e-05,
+      "loss": 0.7782,
+      "step": 44315
+    },
+    {
+      "epoch": 1.7062560153994224,
+      "grad_norm": 1.1873159408569336,
+      "learning_rate": 1.0508025612803096e-05,
+      "loss": 0.8162,
+      "step": 44320
+    },
+    {
+      "epoch": 1.706448508180943,
+      "grad_norm": 1.8230105638504028,
+      "learning_rate": 1.0494535972177932e-05,
+      "loss": 0.7841,
+      "step": 44325
+    },
+    {
+      "epoch": 1.706641000962464,
+      "grad_norm": 1.1108636856079102,
+      "learning_rate": 1.0481054516234546e-05,
+      "loss": 0.7711,
+      "step": 44330
+    },
+    {
+      "epoch": 1.7068334937439846,
+      "grad_norm": 1.4503856897354126,
+      "learning_rate": 1.0467581246205726e-05,
+      "loss": 0.8515,
+      "step": 44335
+    },
+    {
+      "epoch": 1.7070259865255053,
+      "grad_norm": 1.6712833642959595,
+      "learning_rate": 1.0454116163323491e-05,
+      "loss": 0.7857,
+      "step": 44340
+    },
+    {
+      "epoch": 1.707218479307026,
+      "grad_norm": 2.288127899169922,
+      "learning_rate": 1.0440659268819143e-05,
+      "loss": 0.6903,
+      "step": 44345
+    },
+    {
+      "epoch": 1.7074109720885466,
+      "grad_norm": 1.7935099601745605,
+      "learning_rate": 1.0427210563923228e-05,
+      "loss": 0.7954,
+      "step": 44350
+    },
+    {
+      "epoch": 1.7076034648700675,
+      "grad_norm": 1.5855534076690674,
+      "learning_rate": 1.0413770049865546e-05,
+      "loss": 0.8879,
+      "step": 44355
+    },
+    {
+      "epoch": 1.7077959576515882,
+      "grad_norm": 0.9473410844802856,
+      "learning_rate": 1.0400337727875153e-05,
+      "loss": 1.0308,
+      "step": 44360
+    },
+    {
+      "epoch": 1.7079884504331089,
+      "grad_norm": 1.1841708421707153,
+      "learning_rate": 1.0386913599180293e-05,
+      "loss": 0.9523,
+      "step": 44365
+    },
+    {
+      "epoch": 1.7081809432146295,
+      "grad_norm": 1.8651456832885742,
+      "learning_rate": 1.037349766500859e-05,
+      "loss": 0.8538,
+      "step": 44370
+    },
+    {
+      "epoch": 1.7083734359961502,
+      "grad_norm": 1.2280970811843872,
+      "learning_rate": 1.036008992658679e-05,
+      "loss": 0.7538,
+      "step": 44375
+    },
+    {
+      "epoch": 1.7085659287776709,
+      "grad_norm": 1.300763726234436,
+      "learning_rate": 1.0346690385140956e-05,
+      "loss": 0.9075,
+      "step": 44380
+    },
+    {
+      "epoch": 1.7087584215591916,
+      "grad_norm": 1.2974258661270142,
+      "learning_rate": 1.0333299041896383e-05,
+      "loss": 0.8188,
+      "step": 44385
+    },
+    {
+      "epoch": 1.7089509143407122,
+      "grad_norm": 1.9830495119094849,
+      "learning_rate": 1.0319915898077648e-05,
+      "loss": 0.7887,
+      "step": 44390
+    },
+    {
+      "epoch": 1.709143407122233,
+      "grad_norm": 1.0923196077346802,
+      "learning_rate": 1.0306540954908483e-05,
+      "loss": 0.8392,
+      "step": 44395
+    },
+    {
+      "epoch": 1.7093358999037536,
+      "grad_norm": 1.537200689315796,
+      "learning_rate": 1.0293174213612023e-05,
+      "loss": 0.8336,
+      "step": 44400
+    },
+    {
+      "epoch": 1.7095283926852742,
+      "grad_norm": 1.2134389877319336,
+      "learning_rate": 1.027981567541051e-05,
+      "loss": 0.7743,
+      "step": 44405
+    },
+    {
+      "epoch": 1.709720885466795,
+      "grad_norm": 0.6608363389968872,
+      "learning_rate": 1.026646534152551e-05,
+      "loss": 0.6123,
+      "step": 44410
+    },
+    {
+      "epoch": 1.7099133782483156,
+      "grad_norm": 1.0624284744262695,
+      "learning_rate": 1.0253123213177828e-05,
+      "loss": 0.6536,
+      "step": 44415
+    },
+    {
+      "epoch": 1.7101058710298362,
+      "grad_norm": 1.9946283102035522,
+      "learning_rate": 1.0239789291587531e-05,
+      "loss": 0.8706,
+      "step": 44420
+    },
+    {
+      "epoch": 1.710298363811357,
+      "grad_norm": 1.8715145587921143,
+      "learning_rate": 1.0226463577973877e-05,
+      "loss": 0.8116,
+      "step": 44425
+    },
+    {
+      "epoch": 1.7104908565928778,
+      "grad_norm": 1.597334384918213,
+      "learning_rate": 1.0213146073555424e-05,
+      "loss": 0.6751,
+      "step": 44430
+    },
+    {
+      "epoch": 1.7106833493743985,
+      "grad_norm": 1.811631202697754,
+      "learning_rate": 1.0199836779549987e-05,
+      "loss": 0.779,
+      "step": 44435
+    },
+    {
+      "epoch": 1.7108758421559191,
+      "grad_norm": 1.2666516304016113,
+      "learning_rate": 1.0186535697174603e-05,
+      "loss": 0.8132,
+      "step": 44440
+    },
+    {
+      "epoch": 1.7110683349374398,
+      "grad_norm": 1.3092554807662964,
+      "learning_rate": 1.017324282764559e-05,
+      "loss": 0.6895,
+      "step": 44445
+    },
+    {
+      "epoch": 1.7112608277189605,
+      "grad_norm": 1.3765099048614502,
+      "learning_rate": 1.0159958172178452e-05,
+      "loss": 0.7155,
+      "step": 44450
+    },
+    {
+      "epoch": 1.7114533205004814,
+      "grad_norm": 1.301037311553955,
+      "learning_rate": 1.0146681731988006e-05,
+      "loss": 0.827,
+      "step": 44455
+    },
+    {
+      "epoch": 1.711645813282002,
+      "grad_norm": 1.229777455329895,
+      "learning_rate": 1.0133413508288292e-05,
+      "loss": 0.7375,
+      "step": 44460
+    },
+    {
+      "epoch": 1.7118383060635227,
+      "grad_norm": 1.2582491636276245,
+      "learning_rate": 1.0120153502292618e-05,
+      "loss": 0.9079,
+      "step": 44465
+    },
+    {
+      "epoch": 1.7120307988450434,
+      "grad_norm": 1.4134953022003174,
+      "learning_rate": 1.0106901715213468e-05,
+      "loss": 0.7504,
+      "step": 44470
+    },
+    {
+      "epoch": 1.712223291626564,
+      "grad_norm": 1.3563551902770996,
+      "learning_rate": 1.0093658148262709e-05,
+      "loss": 0.7363,
+      "step": 44475
+    },
+    {
+      "epoch": 1.7124157844080847,
+      "grad_norm": 1.642959713935852,
+      "learning_rate": 1.0080422802651312e-05,
+      "loss": 0.7629,
+      "step": 44480
+    },
+    {
+      "epoch": 1.7126082771896054,
+      "grad_norm": 1.9216701984405518,
+      "learning_rate": 1.0067195679589591e-05,
+      "loss": 0.8328,
+      "step": 44485
+    },
+    {
+      "epoch": 1.712800769971126,
+      "grad_norm": 0.9144423007965088,
+      "learning_rate": 1.0053976780287078e-05,
+      "loss": 0.7265,
+      "step": 44490
+    },
+    {
+      "epoch": 1.7129932627526467,
+      "grad_norm": 1.0702438354492188,
+      "learning_rate": 1.0040766105952559e-05,
+      "loss": 0.8567,
+      "step": 44495
+    },
+    {
+      "epoch": 1.7131857555341674,
+      "grad_norm": 1.9012739658355713,
+      "learning_rate": 1.0027563657794026e-05,
+      "loss": 0.8425,
+      "step": 44500
+    },
+    {
+      "epoch": 1.713378248315688,
+      "grad_norm": 1.2741745710372925,
+      "learning_rate": 1.0014369437018823e-05,
+      "loss": 0.8694,
+      "step": 44505
+    },
+    {
+      "epoch": 1.7135707410972087,
+      "grad_norm": 2.203899383544922,
+      "learning_rate": 1.0001183444833417e-05,
+      "loss": 0.7839,
+      "step": 44510
+    },
+    {
+      "epoch": 1.7137632338787294,
+      "grad_norm": 1.059478998184204,
+      "learning_rate": 9.98800568244359e-06,
+      "loss": 0.7702,
+      "step": 44515
+    },
+    {
+      "epoch": 1.71395572666025,
+      "grad_norm": 2.4111287593841553,
+      "learning_rate": 9.974836151054367e-06,
+      "loss": 0.7798,
+      "step": 44520
+    },
+    {
+      "epoch": 1.714148219441771,
+      "grad_norm": 1.2681207656860352,
+      "learning_rate": 9.96167485187004e-06,
+      "loss": 0.818,
+      "step": 44525
+    },
+    {
+      "epoch": 1.7143407122232917,
+      "grad_norm": 1.0277059078216553,
+      "learning_rate": 9.948521786094079e-06,
+      "loss": 0.6552,
+      "step": 44530
+    },
+    {
+      "epoch": 1.7145332050048123,
+      "grad_norm": 0.9331089854240417,
+      "learning_rate": 9.935376954929265e-06,
+      "loss": 0.8268,
+      "step": 44535
+    },
+    {
+      "epoch": 1.714725697786333,
+      "grad_norm": 1.3255749940872192,
+      "learning_rate": 9.922240359577606e-06,
+      "loss": 0.8625,
+      "step": 44540
+    },
+    {
+      "epoch": 1.7149181905678537,
+      "grad_norm": 1.65230393409729,
+      "learning_rate": 9.90911200124035e-06,
+      "loss": 0.7516,
+      "step": 44545
+    },
+    {
+      "epoch": 1.7151106833493746,
+      "grad_norm": 1.164333462715149,
+      "learning_rate": 9.895991881118028e-06,
+      "loss": 0.8635,
+      "step": 44550
+    },
+    {
+      "epoch": 1.7153031761308952,
+      "grad_norm": 1.0087164640426636,
+      "learning_rate": 9.882880000410344e-06,
+      "loss": 0.703,
+      "step": 44555
+    },
+    {
+      "epoch": 1.715495668912416,
+      "grad_norm": 1.272603988647461,
+      "learning_rate": 9.869776360316307e-06,
+      "loss": 0.6907,
+      "step": 44560
+    },
+    {
+      "epoch": 1.7156881616939366,
+      "grad_norm": 0.8092613816261292,
+      "learning_rate": 9.85668096203417e-06,
+      "loss": 0.7481,
+      "step": 44565
+    },
+    {
+      "epoch": 1.7158806544754572,
+      "grad_norm": 2.1582627296447754,
+      "learning_rate": 9.84359380676143e-06,
+      "loss": 0.9793,
+      "step": 44570
+    },
+    {
+      "epoch": 1.716073147256978,
+      "grad_norm": 2.2021024227142334,
+      "learning_rate": 9.830514895694775e-06,
+      "loss": 0.7774,
+      "step": 44575
+    },
+    {
+      "epoch": 1.7162656400384986,
+      "grad_norm": 0.9477373361587524,
+      "learning_rate": 9.817444230030247e-06,
+      "loss": 0.7769,
+      "step": 44580
+    },
+    {
+      "epoch": 1.7164581328200192,
+      "grad_norm": 1.2492976188659668,
+      "learning_rate": 9.804381810963015e-06,
+      "loss": 0.7954,
+      "step": 44585
+    },
+    {
+      "epoch": 1.71665062560154,
+      "grad_norm": 1.2578282356262207,
+      "learning_rate": 9.791327639687587e-06,
+      "loss": 0.9006,
+      "step": 44590
+    },
+    {
+      "epoch": 1.7168431183830606,
+      "grad_norm": 1.1162382364273071,
+      "learning_rate": 9.778281717397652e-06,
+      "loss": 0.755,
+      "step": 44595
+    },
+    {
+      "epoch": 1.7170356111645813,
+      "grad_norm": 1.123497724533081,
+      "learning_rate": 9.765244045286227e-06,
+      "loss": 0.711,
+      "step": 44600
+    },
+    {
+      "epoch": 1.717228103946102,
+      "grad_norm": 1.5501790046691895,
+      "learning_rate": 9.75221462454543e-06,
+      "loss": 0.8798,
+      "step": 44605
+    },
+    {
+      "epoch": 1.7174205967276226,
+      "grad_norm": 1.510923147201538,
+      "learning_rate": 9.739193456366813e-06,
+      "loss": 0.8363,
+      "step": 44610
+    },
+    {
+      "epoch": 1.7176130895091433,
+      "grad_norm": 1.6598632335662842,
+      "learning_rate": 9.726180541941e-06,
+      "loss": 0.8614,
+      "step": 44615
+    },
+    {
+      "epoch": 1.717805582290664,
+      "grad_norm": 1.1419432163238525,
+      "learning_rate": 9.713175882458003e-06,
+      "loss": 0.8118,
+      "step": 44620
+    },
+    {
+      "epoch": 1.7179980750721848,
+      "grad_norm": 1.0711899995803833,
+      "learning_rate": 9.700179479106953e-06,
+      "loss": 0.7971,
+      "step": 44625
+    },
+    {
+      "epoch": 1.7181905678537055,
+      "grad_norm": 1.0233666896820068,
+      "learning_rate": 9.687191333076306e-06,
+      "loss": 0.7606,
+      "step": 44630
+    },
+    {
+      "epoch": 1.7183830606352262,
+      "grad_norm": 2.493830919265747,
+      "learning_rate": 9.674211445553738e-06,
+      "loss": 0.9428,
+      "step": 44635
+    },
+    {
+      "epoch": 1.7185755534167468,
+      "grad_norm": 1.0951639413833618,
+      "learning_rate": 9.661239817726209e-06,
+      "loss": 0.9051,
+      "step": 44640
+    },
+    {
+      "epoch": 1.7187680461982677,
+      "grad_norm": 1.3658279180526733,
+      "learning_rate": 9.648276450779836e-06,
+      "loss": 0.5982,
+      "step": 44645
+    },
+    {
+      "epoch": 1.7189605389797884,
+      "grad_norm": 1.604722499847412,
+      "learning_rate": 9.635321345900061e-06,
+      "loss": 0.8538,
+      "step": 44650
+    },
+    {
+      "epoch": 1.719153031761309,
+      "grad_norm": 1.2309184074401855,
+      "learning_rate": 9.622374504271536e-06,
+      "loss": 0.7353,
+      "step": 44655
+    },
+    {
+      "epoch": 1.7193455245428297,
+      "grad_norm": 1.8740421533584595,
+      "learning_rate": 9.60943592707817e-06,
+      "loss": 0.8622,
+      "step": 44660
+    },
+    {
+      "epoch": 1.7195380173243504,
+      "grad_norm": 1.5069630146026611,
+      "learning_rate": 9.596505615503116e-06,
+      "loss": 0.7259,
+      "step": 44665
+    },
+    {
+      "epoch": 1.719730510105871,
+      "grad_norm": 0.8005820512771606,
+      "learning_rate": 9.583583570728738e-06,
+      "loss": 0.876,
+      "step": 44670
+    },
+    {
+      "epoch": 1.7199230028873917,
+      "grad_norm": 1.9598007202148438,
+      "learning_rate": 9.570669793936694e-06,
+      "loss": 0.799,
+      "step": 44675
+    },
+    {
+      "epoch": 1.7201154956689124,
+      "grad_norm": 1.1604948043823242,
+      "learning_rate": 9.557764286307846e-06,
+      "loss": 0.7464,
+      "step": 44680
+    },
+    {
+      "epoch": 1.720307988450433,
+      "grad_norm": 0.9628298282623291,
+      "learning_rate": 9.544867049022354e-06,
+      "loss": 0.7404,
+      "step": 44685
+    },
+    {
+      "epoch": 1.7205004812319538,
+      "grad_norm": 1.5390831232070923,
+      "learning_rate": 9.53197808325953e-06,
+      "loss": 0.8366,
+      "step": 44690
+    },
+    {
+      "epoch": 1.7206929740134744,
+      "grad_norm": 0.8520711660385132,
+      "learning_rate": 9.51909739019804e-06,
+      "loss": 0.7651,
+      "step": 44695
+    },
+    {
+      "epoch": 1.720885466794995,
+      "grad_norm": 1.1799733638763428,
+      "learning_rate": 9.506224971015709e-06,
+      "loss": 0.7417,
+      "step": 44700
+    },
+    {
+      "epoch": 1.7210779595765158,
+      "grad_norm": 1.2342482805252075,
+      "learning_rate": 9.49336082688962e-06,
+      "loss": 0.829,
+      "step": 44705
+    },
+    {
+      "epoch": 1.7212704523580364,
+      "grad_norm": 1.1500221490859985,
+      "learning_rate": 9.48050495899615e-06,
+      "loss": 0.8516,
+      "step": 44710
+    },
+    {
+      "epoch": 1.721462945139557,
+      "grad_norm": 1.6604115962982178,
+      "learning_rate": 9.467657368510874e-06,
+      "loss": 0.8381,
+      "step": 44715
+    },
+    {
+      "epoch": 1.721655437921078,
+      "grad_norm": 1.3249857425689697,
+      "learning_rate": 9.454818056608573e-06,
+      "loss": 0.7726,
+      "step": 44720
+    },
+    {
+      "epoch": 1.7218479307025987,
+      "grad_norm": 1.1568611860275269,
+      "learning_rate": 9.441987024463384e-06,
+      "loss": 0.7704,
+      "step": 44725
+    },
+    {
+      "epoch": 1.7220404234841193,
+      "grad_norm": 1.5718170404434204,
+      "learning_rate": 9.429164273248581e-06,
+      "loss": 0.7952,
+      "step": 44730
+    },
+    {
+      "epoch": 1.72223291626564,
+      "grad_norm": 1.3416763544082642,
+      "learning_rate": 9.416349804136726e-06,
+      "loss": 0.77,
+      "step": 44735
+    },
+    {
+      "epoch": 1.7224254090471607,
+      "grad_norm": 1.5594481229782104,
+      "learning_rate": 9.403543618299614e-06,
+      "loss": 0.7476,
+      "step": 44740
+    },
+    {
+      "epoch": 1.7226179018286816,
+      "grad_norm": 1.517449975013733,
+      "learning_rate": 9.390745716908312e-06,
+      "loss": 0.6838,
+      "step": 44745
+    },
+    {
+      "epoch": 1.7228103946102022,
+      "grad_norm": 0.9632403254508972,
+      "learning_rate": 9.377956101133068e-06,
+      "loss": 0.7085,
+      "step": 44750
+    },
+    {
+      "epoch": 1.723002887391723,
+      "grad_norm": 1.5700565576553345,
+      "learning_rate": 9.365174772143426e-06,
+      "loss": 0.8404,
+      "step": 44755
+    },
+    {
+      "epoch": 1.7231953801732436,
+      "grad_norm": 1.4765150547027588,
+      "learning_rate": 9.35240173110814e-06,
+      "loss": 0.811,
+      "step": 44760
+    },
+    {
+      "epoch": 1.7233878729547643,
+      "grad_norm": 1.0645157098770142,
+      "learning_rate": 9.339636979195244e-06,
+      "loss": 0.7075,
+      "step": 44765
+    },
+    {
+      "epoch": 1.723580365736285,
+      "grad_norm": 1.7466483116149902,
+      "learning_rate": 9.326880517571978e-06,
+      "loss": 0.8613,
+      "step": 44770
+    },
+    {
+      "epoch": 1.7237728585178056,
+      "grad_norm": 1.016494870185852,
+      "learning_rate": 9.314132347404824e-06,
+      "loss": 0.697,
+      "step": 44775
+    },
+    {
+      "epoch": 1.7239653512993263,
+      "grad_norm": 1.2336843013763428,
+      "learning_rate": 9.301392469859527e-06,
+      "loss": 0.8933,
+      "step": 44780
+    },
+    {
+      "epoch": 1.724157844080847,
+      "grad_norm": 1.7032203674316406,
+      "learning_rate": 9.288660886101075e-06,
+      "loss": 0.7887,
+      "step": 44785
+    },
+    {
+      "epoch": 1.7243503368623676,
+      "grad_norm": 0.8340054154396057,
+      "learning_rate": 9.275937597293682e-06,
+      "loss": 0.8859,
+      "step": 44790
+    },
+    {
+      "epoch": 1.7245428296438883,
+      "grad_norm": 1.1701608896255493,
+      "learning_rate": 9.263222604600774e-06,
+      "loss": 0.7429,
+      "step": 44795
+    },
+    {
+      "epoch": 1.724735322425409,
+      "grad_norm": 0.891636073589325,
+      "learning_rate": 9.25051590918512e-06,
+      "loss": 0.7672,
+      "step": 44800
+    },
+    {
+      "epoch": 1.7249278152069296,
+      "grad_norm": 1.2118922472000122,
+      "learning_rate": 9.237817512208602e-06,
+      "loss": 0.8563,
+      "step": 44805
+    },
+    {
+      "epoch": 1.7251203079884503,
+      "grad_norm": 1.1670253276824951,
+      "learning_rate": 9.225127414832436e-06,
+      "loss": 0.8383,
+      "step": 44810
+    },
+    {
+      "epoch": 1.7253128007699712,
+      "grad_norm": 1.722845435142517,
+      "learning_rate": 9.212445618217035e-06,
+      "loss": 0.8429,
+      "step": 44815
+    },
+    {
+      "epoch": 1.7255052935514918,
+      "grad_norm": 1.2573529481887817,
+      "learning_rate": 9.199772123522088e-06,
+      "loss": 0.9031,
+      "step": 44820
+    },
+    {
+      "epoch": 1.7256977863330125,
+      "grad_norm": 1.844647765159607,
+      "learning_rate": 9.187106931906442e-06,
+      "loss": 0.8546,
+      "step": 44825
+    },
+    {
+      "epoch": 1.7258902791145332,
+      "grad_norm": 2.288663864135742,
+      "learning_rate": 9.174450044528327e-06,
+      "loss": 0.7559,
+      "step": 44830
+    },
+    {
+      "epoch": 1.7260827718960539,
+      "grad_norm": 1.2201417684555054,
+      "learning_rate": 9.161801462545084e-06,
+      "loss": 0.7306,
+      "step": 44835
+    },
+    {
+      "epoch": 1.7262752646775748,
+      "grad_norm": 1.2007523775100708,
+      "learning_rate": 9.14916118711333e-06,
+      "loss": 0.7666,
+      "step": 44840
+    },
+    {
+      "epoch": 1.7264677574590954,
+      "grad_norm": 0.9047454595565796,
+      "learning_rate": 9.136529219388968e-06,
+      "loss": 0.7994,
+      "step": 44845
+    },
+    {
+      "epoch": 1.726660250240616,
+      "grad_norm": 0.8384250402450562,
+      "learning_rate": 9.123905560527102e-06,
+      "loss": 0.6711,
+      "step": 44850
+    },
+    {
+      "epoch": 1.7268527430221368,
+      "grad_norm": 1.0188740491867065,
+      "learning_rate": 9.111290211682044e-06,
+      "loss": 0.7907,
+      "step": 44855
+    },
+    {
+      "epoch": 1.7270452358036574,
+      "grad_norm": 1.1332343816757202,
+      "learning_rate": 9.09868317400745e-06,
+      "loss": 0.7105,
+      "step": 44860
+    },
+    {
+      "epoch": 1.727237728585178,
+      "grad_norm": 0.6927266716957092,
+      "learning_rate": 9.086084448656095e-06,
+      "loss": 0.9879,
+      "step": 44865
+    },
+    {
+      "epoch": 1.7274302213666988,
+      "grad_norm": 1.233681559562683,
+      "learning_rate": 9.073494036780062e-06,
+      "loss": 0.7542,
+      "step": 44870
+    },
+    {
+      "epoch": 1.7276227141482194,
+      "grad_norm": 1.4534273147583008,
+      "learning_rate": 9.06091193953067e-06,
+      "loss": 0.7775,
+      "step": 44875
+    },
+    {
+      "epoch": 1.72781520692974,
+      "grad_norm": 1.0983202457427979,
+      "learning_rate": 9.048338158058467e-06,
+      "loss": 0.9082,
+      "step": 44880
+    },
+    {
+      "epoch": 1.7280076997112608,
+      "grad_norm": 0.7827137112617493,
+      "learning_rate": 9.035772693513256e-06,
+      "loss": 0.7013,
+      "step": 44885
+    },
+    {
+      "epoch": 1.7282001924927815,
+      "grad_norm": 1.1753028631210327,
+      "learning_rate": 9.023215547044028e-06,
+      "loss": 0.6837,
+      "step": 44890
+    },
+    {
+      "epoch": 1.7283926852743021,
+      "grad_norm": 1.4475585222244263,
+      "learning_rate": 9.010666719799077e-06,
+      "loss": 0.7162,
+      "step": 44895
+    },
+    {
+      "epoch": 1.7285851780558228,
+      "grad_norm": 1.2361037731170654,
+      "learning_rate": 8.998126212925906e-06,
+      "loss": 0.7424,
+      "step": 44900
+    },
+    {
+      "epoch": 1.7287776708373435,
+      "grad_norm": 1.0601775646209717,
+      "learning_rate": 8.985594027571276e-06,
+      "loss": 0.6877,
+      "step": 44905
+    },
+    {
+      "epoch": 1.7289701636188641,
+      "grad_norm": 1.1087630987167358,
+      "learning_rate": 8.973070164881126e-06,
+      "loss": 0.7211,
+      "step": 44910
+    },
+    {
+      "epoch": 1.729162656400385,
+      "grad_norm": 1.3524638414382935,
+      "learning_rate": 8.96055462600074e-06,
+      "loss": 0.7259,
+      "step": 44915
+    },
+    {
+      "epoch": 1.7293551491819057,
+      "grad_norm": 1.305503487586975,
+      "learning_rate": 8.948047412074534e-06,
+      "loss": 0.8636,
+      "step": 44920
+    },
+    {
+      "epoch": 1.7295476419634264,
+      "grad_norm": 1.1058146953582764,
+      "learning_rate": 8.935548524246229e-06,
+      "loss": 0.8791,
+      "step": 44925
+    },
+    {
+      "epoch": 1.729740134744947,
+      "grad_norm": 1.93928062915802,
+      "learning_rate": 8.923057963658766e-06,
+      "loss": 0.7892,
+      "step": 44930
+    },
+    {
+      "epoch": 1.7299326275264677,
+      "grad_norm": 0.9320961236953735,
+      "learning_rate": 8.91057573145434e-06,
+      "loss": 0.8925,
+      "step": 44935
+    },
+    {
+      "epoch": 1.7301251203079886,
+      "grad_norm": 1.0383516550064087,
+      "learning_rate": 8.898101828774318e-06,
+      "loss": 0.7483,
+      "step": 44940
+    },
+    {
+      "epoch": 1.7303176130895093,
+      "grad_norm": 1.3876311779022217,
+      "learning_rate": 8.885636256759422e-06,
+      "loss": 0.8479,
+      "step": 44945
+    },
+    {
+      "epoch": 1.73051010587103,
+      "grad_norm": 2.2601609230041504,
+      "learning_rate": 8.873179016549505e-06,
+      "loss": 0.9106,
+      "step": 44950
+    },
+    {
+      "epoch": 1.7307025986525506,
+      "grad_norm": 0.9025186896324158,
+      "learning_rate": 8.8607301092837e-06,
+      "loss": 0.7625,
+      "step": 44955
+    },
+    {
+      "epoch": 1.7308950914340713,
+      "grad_norm": 2.425801992416382,
+      "learning_rate": 8.848289536100374e-06,
+      "loss": 0.9814,
+      "step": 44960
+    },
+    {
+      "epoch": 1.731087584215592,
+      "grad_norm": 1.1960424184799194,
+      "learning_rate": 8.835857298137173e-06,
+      "loss": 0.7639,
+      "step": 44965
+    },
+    {
+      "epoch": 1.7312800769971126,
+      "grad_norm": 1.3602185249328613,
+      "learning_rate": 8.8234333965309e-06,
+      "loss": 0.7965,
+      "step": 44970
+    },
+    {
+      "epoch": 1.7314725697786333,
+      "grad_norm": 1.1339243650436401,
+      "learning_rate": 8.811017832417645e-06,
+      "loss": 0.9131,
+      "step": 44975
+    },
+    {
+      "epoch": 1.731665062560154,
+      "grad_norm": 1.8668417930603027,
+      "learning_rate": 8.798610606932744e-06,
+      "loss": 0.8395,
+      "step": 44980
+    },
+    {
+      "epoch": 1.7318575553416746,
+      "grad_norm": 2.117316246032715,
+      "learning_rate": 8.786211721210747e-06,
+      "loss": 0.6666,
+      "step": 44985
+    },
+    {
+      "epoch": 1.7320500481231953,
+      "grad_norm": 0.8929471373558044,
+      "learning_rate": 8.773821176385466e-06,
+      "loss": 0.829,
+      "step": 44990
+    },
+    {
+      "epoch": 1.732242540904716,
+      "grad_norm": 2.004077196121216,
+      "learning_rate": 8.761438973589908e-06,
+      "loss": 0.8636,
+      "step": 44995
+    },
+    {
+      "epoch": 1.7324350336862366,
+      "grad_norm": 0.8660832643508911,
+      "learning_rate": 8.749065113956357e-06,
+      "loss": 0.7696,
+      "step": 45000
+    },
+    {
+      "epoch": 1.7326275264677573,
+      "grad_norm": 1.3923887014389038,
+      "learning_rate": 8.736699598616305e-06,
+      "loss": 0.8972,
+      "step": 45005
+    },
+    {
+      "epoch": 1.7328200192492782,
+      "grad_norm": 0.835381805896759,
+      "learning_rate": 8.72434242870055e-06,
+      "loss": 0.7019,
+      "step": 45010
+    },
+    {
+      "epoch": 1.7330125120307989,
+      "grad_norm": 0.8896954655647278,
+      "learning_rate": 8.711993605338985e-06,
+      "loss": 0.8025,
+      "step": 45015
+    },
+    {
+      "epoch": 1.7332050048123195,
+      "grad_norm": 1.4621267318725586,
+      "learning_rate": 8.699653129660912e-06,
+      "loss": 0.7648,
+      "step": 45020
+    },
+    {
+      "epoch": 1.7333974975938402,
+      "grad_norm": 1.1056898832321167,
+      "learning_rate": 8.687321002794734e-06,
+      "loss": 0.7427,
+      "step": 45025
+    },
+    {
+      "epoch": 1.7335899903753609,
+      "grad_norm": 1.546595573425293,
+      "learning_rate": 8.674997225868165e-06,
+      "loss": 0.8088,
+      "step": 45030
+    },
+    {
+      "epoch": 1.7337824831568818,
+      "grad_norm": 0.7969051599502563,
+      "learning_rate": 8.662681800008121e-06,
+      "loss": 0.7576,
+      "step": 45035
+    },
+    {
+      "epoch": 1.7339749759384024,
+      "grad_norm": 0.9694302678108215,
+      "learning_rate": 8.650374726340793e-06,
+      "loss": 0.9295,
+      "step": 45040
+    },
+    {
+      "epoch": 1.734167468719923,
+      "grad_norm": 0.5235655903816223,
+      "learning_rate": 8.638076005991524e-06,
+      "loss": 0.7925,
+      "step": 45045
+    },
+    {
+      "epoch": 1.7343599615014438,
+      "grad_norm": 1.4736652374267578,
+      "learning_rate": 8.625785640085026e-06,
+      "loss": 0.8665,
+      "step": 45050
+    },
+    {
+      "epoch": 1.7345524542829645,
+      "grad_norm": 1.5209852457046509,
+      "learning_rate": 8.613503629745113e-06,
+      "loss": 0.8142,
+      "step": 45055
+    },
+    {
+      "epoch": 1.7347449470644851,
+      "grad_norm": 1.1320667266845703,
+      "learning_rate": 8.601229976094904e-06,
+      "loss": 0.7985,
+      "step": 45060
+    },
+    {
+      "epoch": 1.7349374398460058,
+      "grad_norm": 1.1330043077468872,
+      "learning_rate": 8.588964680256761e-06,
+      "loss": 0.8037,
+      "step": 45065
+    },
+    {
+      "epoch": 1.7351299326275265,
+      "grad_norm": 1.6776443719863892,
+      "learning_rate": 8.576707743352275e-06,
+      "loss": 0.8429,
+      "step": 45070
+    },
+    {
+      "epoch": 1.7353224254090471,
+      "grad_norm": 0.8801172971725464,
+      "learning_rate": 8.564459166502204e-06,
+      "loss": 0.8425,
+      "step": 45075
+    },
+    {
+      "epoch": 1.7355149181905678,
+      "grad_norm": 0.9202898740768433,
+      "learning_rate": 8.552218950826662e-06,
+      "loss": 0.7558,
+      "step": 45080
+    },
+    {
+      "epoch": 1.7357074109720885,
+      "grad_norm": 0.8619052767753601,
+      "learning_rate": 8.539987097444912e-06,
+      "loss": 0.7801,
+      "step": 45085
+    },
+    {
+      "epoch": 1.7358999037536091,
+      "grad_norm": 1.3250690698623657,
+      "learning_rate": 8.527763607475459e-06,
+      "loss": 0.8215,
+      "step": 45090
+    },
+    {
+      "epoch": 1.7360923965351298,
+      "grad_norm": 1.4232019186019897,
+      "learning_rate": 8.515548482036106e-06,
+      "loss": 0.9854,
+      "step": 45095
+    },
+    {
+      "epoch": 1.7362848893166505,
+      "grad_norm": 0.8504436016082764,
+      "learning_rate": 8.503341722243785e-06,
+      "loss": 0.7121,
+      "step": 45100
+    },
+    {
+      "epoch": 1.7364773820981714,
+      "grad_norm": 1.5012342929840088,
+      "learning_rate": 8.491143329214768e-06,
+      "loss": 0.8461,
+      "step": 45105
+    },
+    {
+      "epoch": 1.736669874879692,
+      "grad_norm": 1.4541362524032593,
+      "learning_rate": 8.478953304064485e-06,
+      "loss": 0.7671,
+      "step": 45110
+    },
+    {
+      "epoch": 1.7368623676612127,
+      "grad_norm": 1.0640056133270264,
+      "learning_rate": 8.466771647907679e-06,
+      "loss": 0.7623,
+      "step": 45115
+    },
+    {
+      "epoch": 1.7370548604427334,
+      "grad_norm": 1.897068738937378,
+      "learning_rate": 8.454598361858223e-06,
+      "loss": 0.8152,
+      "step": 45120
+    },
+    {
+      "epoch": 1.737247353224254,
+      "grad_norm": 1.7493306398391724,
+      "learning_rate": 8.44243344702934e-06,
+      "loss": 0.8049,
+      "step": 45125
+    },
+    {
+      "epoch": 1.737439846005775,
+      "grad_norm": 1.0308282375335693,
+      "learning_rate": 8.430276904533375e-06,
+      "loss": 0.9274,
+      "step": 45130
+    },
+    {
+      "epoch": 1.7376323387872956,
+      "grad_norm": 1.5391184091567993,
+      "learning_rate": 8.418128735482033e-06,
+      "loss": 0.8391,
+      "step": 45135
+    },
+    {
+      "epoch": 1.7378248315688163,
+      "grad_norm": 1.622266173362732,
+      "learning_rate": 8.40598894098612e-06,
+      "loss": 0.7576,
+      "step": 45140
+    },
+    {
+      "epoch": 1.738017324350337,
+      "grad_norm": 1.7522730827331543,
+      "learning_rate": 8.393857522155758e-06,
+      "loss": 0.6953,
+      "step": 45145
+    },
+    {
+      "epoch": 1.7382098171318576,
+      "grad_norm": 1.5102694034576416,
+      "learning_rate": 8.3817344801003e-06,
+      "loss": 0.7033,
+      "step": 45150
+    },
+    {
+      "epoch": 1.7384023099133783,
+      "grad_norm": 1.0262799263000488,
+      "learning_rate": 8.369619815928321e-06,
+      "loss": 0.8477,
+      "step": 45155
+    },
+    {
+      "epoch": 1.738594802694899,
+      "grad_norm": 1.1589692831039429,
+      "learning_rate": 8.357513530747585e-06,
+      "loss": 0.6228,
+      "step": 45160
+    },
+    {
+      "epoch": 1.7387872954764196,
+      "grad_norm": 1.038752555847168,
+      "learning_rate": 8.345415625665187e-06,
+      "loss": 0.8333,
+      "step": 45165
+    },
+    {
+      "epoch": 1.7389797882579403,
+      "grad_norm": 1.9438576698303223,
+      "learning_rate": 8.333326101787365e-06,
+      "loss": 0.8507,
+      "step": 45170
+    },
+    {
+      "epoch": 1.739172281039461,
+      "grad_norm": 1.3162816762924194,
+      "learning_rate": 8.321244960219632e-06,
+      "loss": 0.8264,
+      "step": 45175
+    },
+    {
+      "epoch": 1.7393647738209816,
+      "grad_norm": 1.4058345556259155,
+      "learning_rate": 8.309172202066728e-06,
+      "loss": 0.8086,
+      "step": 45180
+    },
+    {
+      "epoch": 1.7395572666025023,
+      "grad_norm": 1.2361959218978882,
+      "learning_rate": 8.297107828432649e-06,
+      "loss": 0.6796,
+      "step": 45185
+    },
+    {
+      "epoch": 1.739749759384023,
+      "grad_norm": 1.3547639846801758,
+      "learning_rate": 8.285051840420565e-06,
+      "loss": 0.8473,
+      "step": 45190
+    },
+    {
+      "epoch": 1.7399422521655437,
+      "grad_norm": 1.3733632564544678,
+      "learning_rate": 8.273004239132932e-06,
+      "loss": 0.7467,
+      "step": 45195
+    },
+    {
+      "epoch": 1.7401347449470643,
+      "grad_norm": 1.2235511541366577,
+      "learning_rate": 8.26096502567143e-06,
+      "loss": 0.8072,
+      "step": 45200
+    },
+    {
+      "epoch": 1.7403272377285852,
+      "grad_norm": 1.30812406539917,
+      "learning_rate": 8.24893420113696e-06,
+      "loss": 0.8308,
+      "step": 45205
+    },
+    {
+      "epoch": 1.740519730510106,
+      "grad_norm": 1.1475272178649902,
+      "learning_rate": 8.236911766629674e-06,
+      "loss": 0.7597,
+      "step": 45210
+    },
+    {
+      "epoch": 1.7407122232916266,
+      "grad_norm": 1.2318791151046753,
+      "learning_rate": 8.224897723248926e-06,
+      "loss": 0.7935,
+      "step": 45215
+    },
+    {
+      "epoch": 1.7409047160731472,
+      "grad_norm": 0.8889015913009644,
+      "learning_rate": 8.212892072093314e-06,
+      "loss": 0.7292,
+      "step": 45220
+    },
+    {
+      "epoch": 1.741097208854668,
+      "grad_norm": 2.2487030029296875,
+      "learning_rate": 8.200894814260695e-06,
+      "loss": 0.8981,
+      "step": 45225
+    },
+    {
+      "epoch": 1.7412897016361888,
+      "grad_norm": 2.454103469848633,
+      "learning_rate": 8.188905950848157e-06,
+      "loss": 0.9552,
+      "step": 45230
+    },
+    {
+      "epoch": 1.7414821944177095,
+      "grad_norm": 1.027070164680481,
+      "learning_rate": 8.176925482951925e-06,
+      "loss": 0.8115,
+      "step": 45235
+    },
+    {
+      "epoch": 1.7416746871992301,
+      "grad_norm": 1.2418460845947266,
+      "learning_rate": 8.16495341166763e-06,
+      "loss": 0.7263,
+      "step": 45240
+    },
+    {
+      "epoch": 1.7418671799807508,
+      "grad_norm": 1.0683330297470093,
+      "learning_rate": 8.152989738089978e-06,
+      "loss": 0.8064,
+      "step": 45245
+    },
+    {
+      "epoch": 1.7420596727622715,
+      "grad_norm": 0.9566577672958374,
+      "learning_rate": 8.14103446331298e-06,
+      "loss": 0.7102,
+      "step": 45250
+    },
+    {
+      "epoch": 1.7422521655437921,
+      "grad_norm": 0.8572260737419128,
+      "learning_rate": 8.129087588429873e-06,
+      "loss": 0.8575,
+      "step": 45255
+    },
+    {
+      "epoch": 1.7424446583253128,
+      "grad_norm": 0.9579412341117859,
+      "learning_rate": 8.11714911453314e-06,
+      "loss": 0.8609,
+      "step": 45260
+    },
+    {
+      "epoch": 1.7426371511068335,
+      "grad_norm": 1.0199538469314575,
+      "learning_rate": 8.105219042714406e-06,
+      "loss": 0.7884,
+      "step": 45265
+    },
+    {
+      "epoch": 1.7428296438883542,
+      "grad_norm": 1.9061774015426636,
+      "learning_rate": 8.093297374064679e-06,
+      "loss": 0.7509,
+      "step": 45270
+    },
+    {
+      "epoch": 1.7430221366698748,
+      "grad_norm": 1.1046141386032104,
+      "learning_rate": 8.081384109674073e-06,
+      "loss": 0.861,
+      "step": 45275
+    },
+    {
+      "epoch": 1.7432146294513955,
+      "grad_norm": 0.9744033217430115,
+      "learning_rate": 8.069479250631972e-06,
+      "loss": 0.8627,
+      "step": 45280
+    },
+    {
+      "epoch": 1.7434071222329162,
+      "grad_norm": 1.1140841245651245,
+      "learning_rate": 8.057582798027019e-06,
+      "loss": 0.8167,
+      "step": 45285
+    },
+    {
+      "epoch": 1.7435996150144368,
+      "grad_norm": 0.7895367741584778,
+      "learning_rate": 8.045694752947076e-06,
+      "loss": 0.7456,
+      "step": 45290
+    },
+    {
+      "epoch": 1.7437921077959575,
+      "grad_norm": 1.113579511642456,
+      "learning_rate": 8.033815116479182e-06,
+      "loss": 0.7822,
+      "step": 45295
+    },
+    {
+      "epoch": 1.7439846005774784,
+      "grad_norm": 1.654111385345459,
+      "learning_rate": 8.021943889709682e-06,
+      "loss": 0.8276,
+      "step": 45300
+    },
+    {
+      "epoch": 1.744177093358999,
+      "grad_norm": 1.1563928127288818,
+      "learning_rate": 8.010081073724107e-06,
+      "loss": 0.7169,
+      "step": 45305
+    },
+    {
+      "epoch": 1.7443695861405197,
+      "grad_norm": 1.4437462091445923,
+      "learning_rate": 7.998226669607245e-06,
+      "loss": 0.759,
+      "step": 45310
+    },
+    {
+      "epoch": 1.7445620789220404,
+      "grad_norm": 1.2180962562561035,
+      "learning_rate": 7.986380678443117e-06,
+      "loss": 0.8452,
+      "step": 45315
+    },
+    {
+      "epoch": 1.744754571703561,
+      "grad_norm": 1.4761182069778442,
+      "learning_rate": 7.974543101314912e-06,
+      "loss": 0.8095,
+      "step": 45320
+    },
+    {
+      "epoch": 1.744947064485082,
+      "grad_norm": 0.9857079982757568,
+      "learning_rate": 7.962713939305143e-06,
+      "loss": 0.7167,
+      "step": 45325
+    },
+    {
+      "epoch": 1.7451395572666026,
+      "grad_norm": 1.562577724456787,
+      "learning_rate": 7.95089319349549e-06,
+      "loss": 0.829,
+      "step": 45330
+    },
+    {
+      "epoch": 1.7453320500481233,
+      "grad_norm": 1.0676238536834717,
+      "learning_rate": 7.939080864966897e-06,
+      "loss": 0.7187,
+      "step": 45335
+    },
+    {
+      "epoch": 1.745524542829644,
+      "grad_norm": 1.6058528423309326,
+      "learning_rate": 7.92727695479948e-06,
+      "loss": 0.7018,
+      "step": 45340
+    },
+    {
+      "epoch": 1.7457170356111646,
+      "grad_norm": 1.4000190496444702,
+      "learning_rate": 7.915481464072694e-06,
+      "loss": 0.7528,
+      "step": 45345
+    },
+    {
+      "epoch": 1.7459095283926853,
+      "grad_norm": 1.476828694343567,
+      "learning_rate": 7.903694393865102e-06,
+      "loss": 0.6115,
+      "step": 45350
+    },
+    {
+      "epoch": 1.746102021174206,
+      "grad_norm": 0.9761121273040771,
+      "learning_rate": 7.891915745254574e-06,
+      "loss": 0.775,
+      "step": 45355
+    },
+    {
+      "epoch": 1.7462945139557267,
+      "grad_norm": 1.4030824899673462,
+      "learning_rate": 7.880145519318205e-06,
+      "loss": 0.7931,
+      "step": 45360
+    },
+    {
+      "epoch": 1.7464870067372473,
+      "grad_norm": 2.230025291442871,
+      "learning_rate": 7.868383717132299e-06,
+      "loss": 0.9373,
+      "step": 45365
+    },
+    {
+      "epoch": 1.746679499518768,
+      "grad_norm": 1.5030139684677124,
+      "learning_rate": 7.856630339772341e-06,
+      "loss": 0.8815,
+      "step": 45370
+    },
+    {
+      "epoch": 1.7468719923002887,
+      "grad_norm": 1.0745930671691895,
+      "learning_rate": 7.844885388313194e-06,
+      "loss": 0.7682,
+      "step": 45375
+    },
+    {
+      "epoch": 1.7470644850818093,
+      "grad_norm": 1.3051297664642334,
+      "learning_rate": 7.833148863828766e-06,
+      "loss": 0.7516,
+      "step": 45380
+    },
+    {
+      "epoch": 1.74725697786333,
+      "grad_norm": 1.2218395471572876,
+      "learning_rate": 7.821420767392362e-06,
+      "loss": 0.7248,
+      "step": 45385
+    },
+    {
+      "epoch": 1.7474494706448507,
+      "grad_norm": 1.7142714262008667,
+      "learning_rate": 7.809701100076384e-06,
+      "loss": 0.953,
+      "step": 45390
+    },
+    {
+      "epoch": 1.7476419634263713,
+      "grad_norm": 1.6819877624511719,
+      "learning_rate": 7.797989862952525e-06,
+      "loss": 0.8154,
+      "step": 45395
+    },
+    {
+      "epoch": 1.7478344562078922,
+      "grad_norm": 2.180309295654297,
+      "learning_rate": 7.786287057091723e-06,
+      "loss": 0.7794,
+      "step": 45400
+    },
+    {
+      "epoch": 1.748026948989413,
+      "grad_norm": 1.5509096384048462,
+      "learning_rate": 7.77459268356413e-06,
+      "loss": 0.9624,
+      "step": 45405
+    },
+    {
+      "epoch": 1.7482194417709336,
+      "grad_norm": 1.4458515644073486,
+      "learning_rate": 7.762906743439069e-06,
+      "loss": 0.836,
+      "step": 45410
+    },
+    {
+      "epoch": 1.7484119345524542,
+      "grad_norm": 1.2456862926483154,
+      "learning_rate": 7.751229237785173e-06,
+      "loss": 0.863,
+      "step": 45415
+    },
+    {
+      "epoch": 1.7486044273339751,
+      "grad_norm": 1.0490598678588867,
+      "learning_rate": 7.739560167670279e-06,
+      "loss": 0.8947,
+      "step": 45420
+    },
+    {
+      "epoch": 1.7487969201154958,
+      "grad_norm": 1.548740029335022,
+      "learning_rate": 7.727899534161431e-06,
+      "loss": 0.8212,
+      "step": 45425
+    },
+    {
+      "epoch": 1.7489894128970165,
+      "grad_norm": 1.1737068891525269,
+      "learning_rate": 7.716247338324945e-06,
+      "loss": 0.6706,
+      "step": 45430
+    },
+    {
+      "epoch": 1.7491819056785372,
+      "grad_norm": 1.5018718242645264,
+      "learning_rate": 7.7046035812263e-06,
+      "loss": 0.7855,
+      "step": 45435
+    },
+    {
+      "epoch": 1.7493743984600578,
+      "grad_norm": 1.2772295475006104,
+      "learning_rate": 7.692968263930255e-06,
+      "loss": 0.7608,
+      "step": 45440
+    },
+    {
+      "epoch": 1.7495668912415785,
+      "grad_norm": 1.0926804542541504,
+      "learning_rate": 7.681341387500784e-06,
+      "loss": 0.7587,
+      "step": 45445
+    },
+    {
+      "epoch": 1.7497593840230992,
+      "grad_norm": 1.2796525955200195,
+      "learning_rate": 7.669722953001113e-06,
+      "loss": 0.8821,
+      "step": 45450
+    },
+    {
+      "epoch": 1.7499518768046198,
+      "grad_norm": 0.9413089752197266,
+      "learning_rate": 7.658112961493602e-06,
+      "loss": 0.7892,
+      "step": 45455
+    },
+    {
+      "epoch": 1.7501443695861405,
+      "grad_norm": 1.1582038402557373,
+      "learning_rate": 7.646511414039981e-06,
+      "loss": 0.759,
+      "step": 45460
+    },
+    {
+      "epoch": 1.7503368623676612,
+      "grad_norm": 1.9622395038604736,
+      "learning_rate": 7.6349183117011e-06,
+      "loss": 0.7499,
+      "step": 45465
+    },
+    {
+      "epoch": 1.7505293551491818,
+      "grad_norm": 1.9417619705200195,
+      "learning_rate": 7.623333655537068e-06,
+      "loss": 0.8312,
+      "step": 45470
+    },
+    {
+      "epoch": 1.7507218479307025,
+      "grad_norm": 1.0554540157318115,
+      "learning_rate": 7.611757446607238e-06,
+      "loss": 0.8,
+      "step": 45475
+    },
+    {
+      "epoch": 1.7509143407122232,
+      "grad_norm": 0.9951033592224121,
+      "learning_rate": 7.6001896859701935e-06,
+      "loss": 0.7422,
+      "step": 45480
+    },
+    {
+      "epoch": 1.7511068334937439,
+      "grad_norm": 1.446395993232727,
+      "learning_rate": 7.58863037468367e-06,
+      "loss": 0.8357,
+      "step": 45485
+    },
+    {
+      "epoch": 1.7512993262752645,
+      "grad_norm": 1.0408679246902466,
+      "learning_rate": 7.577079513804764e-06,
+      "loss": 0.8824,
+      "step": 45490
+    },
+    {
+      "epoch": 1.7514918190567854,
+      "grad_norm": 2.0275330543518066,
+      "learning_rate": 7.565537104389675e-06,
+      "loss": 0.865,
+      "step": 45495
+    },
+    {
+      "epoch": 1.751684311838306,
+      "grad_norm": 0.5688580870628357,
+      "learning_rate": 7.5540031474938936e-06,
+      "loss": 0.7585,
+      "step": 45500
+    },
+    {
+      "epoch": 1.7518768046198268,
+      "grad_norm": 0.9278910160064697,
+      "learning_rate": 7.54247764417213e-06,
+      "loss": 0.8563,
+      "step": 45505
+    },
+    {
+      "epoch": 1.7520692974013474,
+      "grad_norm": 1.2069793939590454,
+      "learning_rate": 7.53096059547832e-06,
+      "loss": 0.8896,
+      "step": 45510
+    },
+    {
+      "epoch": 1.752261790182868,
+      "grad_norm": 1.3929320573806763,
+      "learning_rate": 7.51945200246561e-06,
+      "loss": 0.6949,
+      "step": 45515
+    },
+    {
+      "epoch": 1.752454282964389,
+      "grad_norm": 1.403387427330017,
+      "learning_rate": 7.507951866186369e-06,
+      "loss": 0.8077,
+      "step": 45520
+    },
+    {
+      "epoch": 1.7526467757459097,
+      "grad_norm": 1.3101508617401123,
+      "learning_rate": 7.496460187692233e-06,
+      "loss": 0.8342,
+      "step": 45525
+    },
+    {
+      "epoch": 1.7528392685274303,
+      "grad_norm": 1.3768970966339111,
+      "learning_rate": 7.484976968034041e-06,
+      "loss": 0.7386,
+      "step": 45530
+    },
+    {
+      "epoch": 1.753031761308951,
+      "grad_norm": 1.012230396270752,
+      "learning_rate": 7.473502208261862e-06,
+      "loss": 0.7605,
+      "step": 45535
+    },
+    {
+      "epoch": 1.7532242540904717,
+      "grad_norm": 0.9598119258880615,
+      "learning_rate": 7.462035909424947e-06,
+      "loss": 0.7581,
+      "step": 45540
+    },
+    {
+      "epoch": 1.7534167468719923,
+      "grad_norm": 1.7553858757019043,
+      "learning_rate": 7.450578072571857e-06,
+      "loss": 0.6762,
+      "step": 45545
+    },
+    {
+      "epoch": 1.753609239653513,
+      "grad_norm": 1.4688769578933716,
+      "learning_rate": 7.439128698750309e-06,
+      "loss": 0.661,
+      "step": 45550
+    },
+    {
+      "epoch": 1.7538017324350337,
+      "grad_norm": 1.9736151695251465,
+      "learning_rate": 7.427687789007299e-06,
+      "loss": 0.8019,
+      "step": 45555
+    },
+    {
+      "epoch": 1.7539942252165543,
+      "grad_norm": 0.9081621170043945,
+      "learning_rate": 7.416255344388967e-06,
+      "loss": 0.8455,
+      "step": 45560
+    },
+    {
+      "epoch": 1.754186717998075,
+      "grad_norm": 0.812998354434967,
+      "learning_rate": 7.404831365940812e-06,
+      "loss": 0.8729,
+      "step": 45565
+    },
+    {
+      "epoch": 1.7543792107795957,
+      "grad_norm": 1.4121613502502441,
+      "learning_rate": 7.393415854707419e-06,
+      "loss": 0.7995,
+      "step": 45570
+    },
+    {
+      "epoch": 1.7545717035611164,
+      "grad_norm": 1.363328456878662,
+      "learning_rate": 7.382008811732688e-06,
+      "loss": 0.7346,
+      "step": 45575
+    },
+    {
+      "epoch": 1.754764196342637,
+      "grad_norm": 1.5376365184783936,
+      "learning_rate": 7.370610238059716e-06,
+      "loss": 0.703,
+      "step": 45580
+    },
+    {
+      "epoch": 1.7549566891241577,
+      "grad_norm": 1.3588625192642212,
+      "learning_rate": 7.359220134730838e-06,
+      "loss": 0.8671,
+      "step": 45585
+    },
+    {
+      "epoch": 1.7551491819056786,
+      "grad_norm": 1.2630447149276733,
+      "learning_rate": 7.347838502787563e-06,
+      "loss": 0.8949,
+      "step": 45590
+    },
+    {
+      "epoch": 1.7553416746871993,
+      "grad_norm": 1.2255010604858398,
+      "learning_rate": 7.336465343270715e-06,
+      "loss": 0.8899,
+      "step": 45595
+    },
+    {
+      "epoch": 1.75553416746872,
+      "grad_norm": 0.93909752368927,
+      "learning_rate": 7.325100657220263e-06,
+      "loss": 0.6891,
+      "step": 45600
+    },
+    {
+      "epoch": 1.7557266602502406,
+      "grad_norm": 1.0947200059890747,
+      "learning_rate": 7.31374444567543e-06,
+      "loss": 0.7669,
+      "step": 45605
+    },
+    {
+      "epoch": 1.7559191530317613,
+      "grad_norm": 1.0405867099761963,
+      "learning_rate": 7.302396709674686e-06,
+      "loss": 0.9153,
+      "step": 45610
+    },
+    {
+      "epoch": 1.7561116458132822,
+      "grad_norm": 1.1569056510925293,
+      "learning_rate": 7.291057450255712e-06,
+      "loss": 0.7694,
+      "step": 45615
+    },
+    {
+      "epoch": 1.7563041385948028,
+      "grad_norm": 2.7271223068237305,
+      "learning_rate": 7.279726668455366e-06,
+      "loss": 0.9885,
+      "step": 45620
+    },
+    {
+      "epoch": 1.7564966313763235,
+      "grad_norm": 1.421797752380371,
+      "learning_rate": 7.268404365309822e-06,
+      "loss": 0.9199,
+      "step": 45625
+    },
+    {
+      "epoch": 1.7566891241578442,
+      "grad_norm": 1.7297556400299072,
+      "learning_rate": 7.257090541854405e-06,
+      "loss": 0.8034,
+      "step": 45630
+    },
+    {
+      "epoch": 1.7568816169393648,
+      "grad_norm": 1.0657942295074463,
+      "learning_rate": 7.245785199123689e-06,
+      "loss": 0.7211,
+      "step": 45635
+    },
+    {
+      "epoch": 1.7570741097208855,
+      "grad_norm": 1.2450711727142334,
+      "learning_rate": 7.23448833815148e-06,
+      "loss": 0.9191,
+      "step": 45640
+    },
+    {
+      "epoch": 1.7572666025024062,
+      "grad_norm": 1.1879651546478271,
+      "learning_rate": 7.223199959970784e-06,
+      "loss": 0.7244,
+      "step": 45645
+    },
+    {
+      "epoch": 1.7574590952839269,
+      "grad_norm": 1.0603852272033691,
+      "learning_rate": 7.211920065613875e-06,
+      "loss": 0.757,
+      "step": 45650
+    },
+    {
+      "epoch": 1.7576515880654475,
+      "grad_norm": 1.9512847661972046,
+      "learning_rate": 7.200648656112208e-06,
+      "loss": 0.6762,
+      "step": 45655
+    },
+    {
+      "epoch": 1.7578440808469682,
+      "grad_norm": 1.7029409408569336,
+      "learning_rate": 7.189385732496479e-06,
+      "loss": 0.7148,
+      "step": 45660
+    },
+    {
+      "epoch": 1.7580365736284889,
+      "grad_norm": 1.464708685874939,
+      "learning_rate": 7.178131295796609e-06,
+      "loss": 0.7853,
+      "step": 45665
+    },
+    {
+      "epoch": 1.7582290664100095,
+      "grad_norm": 1.0993046760559082,
+      "learning_rate": 7.166885347041763e-06,
+      "loss": 0.8098,
+      "step": 45670
+    },
+    {
+      "epoch": 1.7584215591915302,
+      "grad_norm": 1.787937045097351,
+      "learning_rate": 7.155647887260253e-06,
+      "loss": 0.9171,
+      "step": 45675
+    },
+    {
+      "epoch": 1.7586140519730509,
+      "grad_norm": 1.478559136390686,
+      "learning_rate": 7.144418917479734e-06,
+      "loss": 0.7519,
+      "step": 45680
+    },
+    {
+      "epoch": 1.7588065447545715,
+      "grad_norm": 1.5054091215133667,
+      "learning_rate": 7.133198438726985e-06,
+      "loss": 0.6784,
+      "step": 45685
+    },
+    {
+      "epoch": 1.7589990375360924,
+      "grad_norm": 1.190794587135315,
+      "learning_rate": 7.121986452028051e-06,
+      "loss": 0.7847,
+      "step": 45690
+    },
+    {
+      "epoch": 1.759191530317613,
+      "grad_norm": 1.474478006362915,
+      "learning_rate": 7.110782958408191e-06,
+      "loss": 0.7256,
+      "step": 45695
+    },
+    {
+      "epoch": 1.7593840230991338,
+      "grad_norm": 0.9803133010864258,
+      "learning_rate": 7.0995879588919176e-06,
+      "loss": 0.7769,
+      "step": 45700
+    },
+    {
+      "epoch": 1.7595765158806544,
+      "grad_norm": 1.1921695470809937,
+      "learning_rate": 7.08840145450288e-06,
+      "loss": 0.7835,
+      "step": 45705
+    },
+    {
+      "epoch": 1.7597690086621751,
+      "grad_norm": 0.8569211363792419,
+      "learning_rate": 7.077223446264081e-06,
+      "loss": 0.7431,
+      "step": 45710
+    },
+    {
+      "epoch": 1.759961501443696,
+      "grad_norm": 1.798243522644043,
+      "learning_rate": 7.066053935197625e-06,
+      "loss": 0.9124,
+      "step": 45715
+    },
+    {
+      "epoch": 1.7601539942252167,
+      "grad_norm": 1.0873985290527344,
+      "learning_rate": 7.054892922324896e-06,
+      "loss": 0.7808,
+      "step": 45720
+    },
+    {
+      "epoch": 1.7603464870067373,
+      "grad_norm": 0.899448573589325,
+      "learning_rate": 7.043740408666511e-06,
+      "loss": 0.6514,
+      "step": 45725
+    },
+    {
+      "epoch": 1.760538979788258,
+      "grad_norm": 0.8346114754676819,
+      "learning_rate": 7.032596395242308e-06,
+      "loss": 0.7472,
+      "step": 45730
+    },
+    {
+      "epoch": 1.7607314725697787,
+      "grad_norm": 1.0554609298706055,
+      "learning_rate": 7.021460883071296e-06,
+      "loss": 0.7476,
+      "step": 45735
+    },
+    {
+      "epoch": 1.7609239653512994,
+      "grad_norm": 2.069948196411133,
+      "learning_rate": 7.010333873171749e-06,
+      "loss": 0.8979,
+      "step": 45740
+    },
+    {
+      "epoch": 1.76111645813282,
+      "grad_norm": 1.2894495725631714,
+      "learning_rate": 6.999215366561174e-06,
+      "loss": 0.8933,
+      "step": 45745
+    },
+    {
+      "epoch": 1.7613089509143407,
+      "grad_norm": 0.8491725325584412,
+      "learning_rate": 6.988105364256281e-06,
+      "loss": 0.7834,
+      "step": 45750
+    },
+    {
+      "epoch": 1.7615014436958614,
+      "grad_norm": 1.3714752197265625,
+      "learning_rate": 6.977003867273024e-06,
+      "loss": 0.8987,
+      "step": 45755
+    },
+    {
+      "epoch": 1.761693936477382,
+      "grad_norm": 1.4427109956741333,
+      "learning_rate": 6.965910876626525e-06,
+      "loss": 0.7007,
+      "step": 45760
+    },
+    {
+      "epoch": 1.7618864292589027,
+      "grad_norm": 2.3159167766571045,
+      "learning_rate": 6.954826393331182e-06,
+      "loss": 0.7829,
+      "step": 45765
+    },
+    {
+      "epoch": 1.7620789220404234,
+      "grad_norm": 1.3657716512680054,
+      "learning_rate": 6.9437504184005964e-06,
+      "loss": 0.8195,
+      "step": 45770
+    },
+    {
+      "epoch": 1.762271414821944,
+      "grad_norm": 1.9438031911849976,
+      "learning_rate": 6.932682952847602e-06,
+      "loss": 0.851,
+      "step": 45775
+    },
+    {
+      "epoch": 1.7624639076034647,
+      "grad_norm": 1.6237000226974487,
+      "learning_rate": 6.9216239976842125e-06,
+      "loss": 0.7393,
+      "step": 45780
+    },
+    {
+      "epoch": 1.7626564003849856,
+      "grad_norm": 1.7042378187179565,
+      "learning_rate": 6.910573553921762e-06,
+      "loss": 0.7954,
+      "step": 45785
+    },
+    {
+      "epoch": 1.7628488931665063,
+      "grad_norm": 1.3898320198059082,
+      "learning_rate": 6.899531622570665e-06,
+      "loss": 0.7336,
+      "step": 45790
+    },
+    {
+      "epoch": 1.763041385948027,
+      "grad_norm": 0.9973713159561157,
+      "learning_rate": 6.888498204640681e-06,
+      "loss": 0.7636,
+      "step": 45795
+    },
+    {
+      "epoch": 1.7632338787295476,
+      "grad_norm": 1.350479006767273,
+      "learning_rate": 6.877473301140725e-06,
+      "loss": 0.7871,
+      "step": 45800
+    },
+    {
+      "epoch": 1.7634263715110683,
+      "grad_norm": 0.9988759756088257,
+      "learning_rate": 6.866456913078967e-06,
+      "loss": 0.7535,
+      "step": 45805
+    },
+    {
+      "epoch": 1.7636188642925892,
+      "grad_norm": 1.9600189924240112,
+      "learning_rate": 6.855449041462736e-06,
+      "loss": 0.8776,
+      "step": 45810
+    },
+    {
+      "epoch": 1.7638113570741099,
+      "grad_norm": 1.4199215173721313,
+      "learning_rate": 6.844449687298704e-06,
+      "loss": 0.8597,
+      "step": 45815
+    },
+    {
+      "epoch": 1.7640038498556305,
+      "grad_norm": 0.9643407464027405,
+      "learning_rate": 6.833458851592633e-06,
+      "loss": 0.8721,
+      "step": 45820
+    },
+    {
+      "epoch": 1.7641963426371512,
+      "grad_norm": 1.3059306144714355,
+      "learning_rate": 6.822476535349576e-06,
+      "loss": 0.8427,
+      "step": 45825
+    },
+    {
+      "epoch": 1.7643888354186719,
+      "grad_norm": 1.1498661041259766,
+      "learning_rate": 6.811502739573794e-06,
+      "loss": 0.8962,
+      "step": 45830
+    },
+    {
+      "epoch": 1.7645813282001925,
+      "grad_norm": 1.7601451873779297,
+      "learning_rate": 6.800537465268786e-06,
+      "loss": 0.6432,
+      "step": 45835
+    },
+    {
+      "epoch": 1.7647738209817132,
+      "grad_norm": 1.7763153314590454,
+      "learning_rate": 6.7895807134372265e-06,
+      "loss": 0.7724,
+      "step": 45840
+    },
+    {
+      "epoch": 1.7649663137632339,
+      "grad_norm": 0.8977097868919373,
+      "learning_rate": 6.778632485081038e-06,
+      "loss": 0.9375,
+      "step": 45845
+    },
+    {
+      "epoch": 1.7651588065447545,
+      "grad_norm": 2.03190541267395,
+      "learning_rate": 6.767692781201385e-06,
+      "loss": 0.8175,
+      "step": 45850
+    },
+    {
+      "epoch": 1.7653512993262752,
+      "grad_norm": 1.9310617446899414,
+      "learning_rate": 6.756761602798611e-06,
+      "loss": 0.7512,
+      "step": 45855
+    },
+    {
+      "epoch": 1.7655437921077959,
+      "grad_norm": 1.507844090461731,
+      "learning_rate": 6.7458389508723295e-06,
+      "loss": 0.989,
+      "step": 45860
+    },
+    {
+      "epoch": 1.7657362848893166,
+      "grad_norm": 1.0742762088775635,
+      "learning_rate": 6.734924826421296e-06,
+      "loss": 0.7757,
+      "step": 45865
+    },
+    {
+      "epoch": 1.7659287776708372,
+      "grad_norm": 3.061866283416748,
+      "learning_rate": 6.724019230443579e-06,
+      "loss": 0.8364,
+      "step": 45870
+    },
+    {
+      "epoch": 1.766121270452358,
+      "grad_norm": 1.1504889726638794,
+      "learning_rate": 6.713122163936392e-06,
+      "loss": 0.7237,
+      "step": 45875
+    },
+    {
+      "epoch": 1.7663137632338786,
+      "grad_norm": 0.9810017347335815,
+      "learning_rate": 6.702233627896237e-06,
+      "loss": 0.8952,
+      "step": 45880
+    },
+    {
+      "epoch": 1.7665062560153995,
+      "grad_norm": 1.719298005104065,
+      "learning_rate": 6.691353623318752e-06,
+      "loss": 0.9511,
+      "step": 45885
+    },
+    {
+      "epoch": 1.7666987487969201,
+      "grad_norm": 1.0837730169296265,
+      "learning_rate": 6.680482151198886e-06,
+      "loss": 1.0137,
+      "step": 45890
+    },
+    {
+      "epoch": 1.7668912415784408,
+      "grad_norm": 1.030633807182312,
+      "learning_rate": 6.669619212530709e-06,
+      "loss": 0.8197,
+      "step": 45895
+    },
+    {
+      "epoch": 1.7670837343599615,
+      "grad_norm": 1.0519740581512451,
+      "learning_rate": 6.658764808307638e-06,
+      "loss": 0.7061,
+      "step": 45900
+    },
+    {
+      "epoch": 1.7672762271414824,
+      "grad_norm": 2.0290040969848633,
+      "learning_rate": 6.647918939522168e-06,
+      "loss": 0.8013,
+      "step": 45905
+    },
+    {
+      "epoch": 1.767468719923003,
+      "grad_norm": 1.4919228553771973,
+      "learning_rate": 6.637081607166129e-06,
+      "loss": 0.7577,
+      "step": 45910
+    },
+    {
+      "epoch": 1.7676612127045237,
+      "grad_norm": 1.4189355373382568,
+      "learning_rate": 6.626252812230494e-06,
+      "loss": 0.6824,
+      "step": 45915
+    },
+    {
+      "epoch": 1.7678537054860444,
+      "grad_norm": 1.654653787612915,
+      "learning_rate": 6.615432555705503e-06,
+      "loss": 0.7618,
+      "step": 45920
+    },
+    {
+      "epoch": 1.768046198267565,
+      "grad_norm": 1.591550350189209,
+      "learning_rate": 6.604620838580566e-06,
+      "loss": 0.8461,
+      "step": 45925
+    },
+    {
+      "epoch": 1.7682386910490857,
+      "grad_norm": 1.0825369358062744,
+      "learning_rate": 6.593817661844393e-06,
+      "loss": 0.8319,
+      "step": 45930
+    },
+    {
+      "epoch": 1.7684311838306064,
+      "grad_norm": 1.893295407295227,
+      "learning_rate": 6.5830230264848245e-06,
+      "loss": 0.8209,
+      "step": 45935
+    },
+    {
+      "epoch": 1.768623676612127,
+      "grad_norm": 0.8829330205917358,
+      "learning_rate": 6.572236933488962e-06,
+      "loss": 0.7583,
+      "step": 45940
+    },
+    {
+      "epoch": 1.7688161693936477,
+      "grad_norm": 1.44454026222229,
+      "learning_rate": 6.561459383843138e-06,
+      "loss": 0.8129,
+      "step": 45945
+    },
+    {
+      "epoch": 1.7690086621751684,
+      "grad_norm": 1.0190069675445557,
+      "learning_rate": 6.5506903785328865e-06,
+      "loss": 0.7915,
+      "step": 45950
+    },
+    {
+      "epoch": 1.769201154956689,
+      "grad_norm": 1.2443015575408936,
+      "learning_rate": 6.539929918542953e-06,
+      "loss": 0.6906,
+      "step": 45955
+    },
+    {
+      "epoch": 1.7693936477382097,
+      "grad_norm": 1.2522984743118286,
+      "learning_rate": 6.529178004857295e-06,
+      "loss": 0.7546,
+      "step": 45960
+    },
+    {
+      "epoch": 1.7695861405197304,
+      "grad_norm": 0.4453613758087158,
+      "learning_rate": 6.5184346384591365e-06,
+      "loss": 0.7618,
+      "step": 45965
+    },
+    {
+      "epoch": 1.769778633301251,
+      "grad_norm": 1.1776552200317383,
+      "learning_rate": 6.507699820330859e-06,
+      "loss": 0.7245,
+      "step": 45970
+    },
+    {
+      "epoch": 1.7699711260827717,
+      "grad_norm": 0.9290987253189087,
+      "learning_rate": 6.4969735514541216e-06,
+      "loss": 0.8716,
+      "step": 45975
+    },
+    {
+      "epoch": 1.7701636188642926,
+      "grad_norm": 1.227397084236145,
+      "learning_rate": 6.48625583280974e-06,
+      "loss": 0.9719,
+      "step": 45980
+    },
+    {
+      "epoch": 1.7703561116458133,
+      "grad_norm": 1.5179232358932495,
+      "learning_rate": 6.4755466653777965e-06,
+      "loss": 0.7412,
+      "step": 45985
+    },
+    {
+      "epoch": 1.770548604427334,
+      "grad_norm": 1.2842813730239868,
+      "learning_rate": 6.464846050137552e-06,
+      "loss": 0.8631,
+      "step": 45990
+    },
+    {
+      "epoch": 1.7707410972088546,
+      "grad_norm": 1.102016806602478,
+      "learning_rate": 6.454153988067557e-06,
+      "loss": 0.7781,
+      "step": 45995
+    },
+    {
+      "epoch": 1.7709335899903753,
+      "grad_norm": 1.2582801580429077,
+      "learning_rate": 6.443470480145452e-06,
+      "loss": 0.8891,
+      "step": 46000
+    },
+    {
+      "epoch": 1.7711260827718962,
+      "grad_norm": 1.6584640741348267,
+      "learning_rate": 6.432795527348246e-06,
+      "loss": 0.845,
+      "step": 46005
+    },
+    {
+      "epoch": 1.7713185755534169,
+      "grad_norm": 0.9719191789627075,
+      "learning_rate": 6.422129130652055e-06,
+      "loss": 0.7511,
+      "step": 46010
+    },
+    {
+      "epoch": 1.7715110683349375,
+      "grad_norm": 0.968356192111969,
+      "learning_rate": 6.411471291032245e-06,
+      "loss": 0.8289,
+      "step": 46015
+    },
+    {
+      "epoch": 1.7717035611164582,
+      "grad_norm": 1.0656431913375854,
+      "learning_rate": 6.400822009463437e-06,
+      "loss": 0.8338,
+      "step": 46020
+    },
+    {
+      "epoch": 1.7718960538979789,
+      "grad_norm": 0.6920823454856873,
+      "learning_rate": 6.3901812869194165e-06,
+      "loss": 0.6625,
+      "step": 46025
+    },
+    {
+      "epoch": 1.7720885466794996,
+      "grad_norm": 1.6432719230651855,
+      "learning_rate": 6.379549124373185e-06,
+      "loss": 0.8689,
+      "step": 46030
+    },
+    {
+      "epoch": 1.7722810394610202,
+      "grad_norm": 1.784363865852356,
+      "learning_rate": 6.368925522797042e-06,
+      "loss": 0.8466,
+      "step": 46035
+    },
+    {
+      "epoch": 1.772473532242541,
+      "grad_norm": 0.9978721141815186,
+      "learning_rate": 6.358310483162389e-06,
+      "loss": 0.6407,
+      "step": 46040
+    },
+    {
+      "epoch": 1.7726660250240616,
+      "grad_norm": 1.793863296508789,
+      "learning_rate": 6.347704006439936e-06,
+      "loss": 0.7827,
+      "step": 46045
+    },
+    {
+      "epoch": 1.7728585178055822,
+      "grad_norm": 0.9259586930274963,
+      "learning_rate": 6.337106093599543e-06,
+      "loss": 0.8073,
+      "step": 46050
+    },
+    {
+      "epoch": 1.773051010587103,
+      "grad_norm": 1.971684455871582,
+      "learning_rate": 6.326516745610367e-06,
+      "loss": 0.7771,
+      "step": 46055
+    },
+    {
+      "epoch": 1.7732435033686236,
+      "grad_norm": 1.9362154006958008,
+      "learning_rate": 6.315935963440689e-06,
+      "loss": 0.9013,
+      "step": 46060
+    },
+    {
+      "epoch": 1.7734359961501442,
+      "grad_norm": 1.4866061210632324,
+      "learning_rate": 6.305363748058057e-06,
+      "loss": 0.9351,
+      "step": 46065
+    },
+    {
+      "epoch": 1.773628488931665,
+      "grad_norm": 0.8735952973365784,
+      "learning_rate": 6.2948001004292435e-06,
+      "loss": 0.7682,
+      "step": 46070
+    },
+    {
+      "epoch": 1.7738209817131858,
+      "grad_norm": 1.655439019203186,
+      "learning_rate": 6.2842450215202295e-06,
+      "loss": 0.9094,
+      "step": 46075
+    },
+    {
+      "epoch": 1.7740134744947065,
+      "grad_norm": 1.2115354537963867,
+      "learning_rate": 6.2736985122962e-06,
+      "loss": 0.8336,
+      "step": 46080
+    },
+    {
+      "epoch": 1.7742059672762271,
+      "grad_norm": 1.3702316284179688,
+      "learning_rate": 6.263160573721561e-06,
+      "loss": 0.7794,
+      "step": 46085
+    },
+    {
+      "epoch": 1.7743984600577478,
+      "grad_norm": 0.8656373023986816,
+      "learning_rate": 6.252631206759929e-06,
+      "loss": 0.6995,
+      "step": 46090
+    },
+    {
+      "epoch": 1.7745909528392685,
+      "grad_norm": 1.3100272417068481,
+      "learning_rate": 6.242110412374158e-06,
+      "loss": 1.0182,
+      "step": 46095
+    },
+    {
+      "epoch": 1.7747834456207894,
+      "grad_norm": 0.7958429455757141,
+      "learning_rate": 6.231598191526311e-06,
+      "loss": 0.7499,
+      "step": 46100
+    },
+    {
+      "epoch": 1.77497593840231,
+      "grad_norm": 1.7569608688354492,
+      "learning_rate": 6.221094545177619e-06,
+      "loss": 0.6442,
+      "step": 46105
+    },
+    {
+      "epoch": 1.7751684311838307,
+      "grad_norm": 1.1067934036254883,
+      "learning_rate": 6.2105994742886365e-06,
+      "loss": 0.7889,
+      "step": 46110
+    },
+    {
+      "epoch": 1.7753609239653514,
+      "grad_norm": 1.4211006164550781,
+      "learning_rate": 6.200112979819017e-06,
+      "loss": 0.7786,
+      "step": 46115
+    },
+    {
+      "epoch": 1.775553416746872,
+      "grad_norm": 0.9069014191627502,
+      "learning_rate": 6.189635062727695e-06,
+      "loss": 0.8214,
+      "step": 46120
+    },
+    {
+      "epoch": 1.7757459095283927,
+      "grad_norm": 1.2809123992919922,
+      "learning_rate": 6.179165723972824e-06,
+      "loss": 0.8923,
+      "step": 46125
+    },
+    {
+      "epoch": 1.7759384023099134,
+      "grad_norm": 1.3749951124191284,
+      "learning_rate": 6.16870496451174e-06,
+      "loss": 0.7566,
+      "step": 46130
+    },
+    {
+      "epoch": 1.776130895091434,
+      "grad_norm": 1.8189775943756104,
+      "learning_rate": 6.158252785300988e-06,
+      "loss": 0.9776,
+      "step": 46135
+    },
+    {
+      "epoch": 1.7763233878729547,
+      "grad_norm": 1.2725715637207031,
+      "learning_rate": 6.147809187296405e-06,
+      "loss": 0.7726,
+      "step": 46140
+    },
+    {
+      "epoch": 1.7765158806544754,
+      "grad_norm": 1.479418396949768,
+      "learning_rate": 6.137374171452948e-06,
+      "loss": 0.8255,
+      "step": 46145
+    },
+    {
+      "epoch": 1.776708373435996,
+      "grad_norm": 0.9075431227684021,
+      "learning_rate": 6.1269477387248306e-06,
+      "loss": 0.9074,
+      "step": 46150
+    },
+    {
+      "epoch": 1.7769008662175168,
+      "grad_norm": 1.1781271696090698,
+      "learning_rate": 6.116529890065492e-06,
+      "loss": 0.7702,
+      "step": 46155
+    },
+    {
+      "epoch": 1.7770933589990374,
+      "grad_norm": 1.5959038734436035,
+      "learning_rate": 6.1061206264275805e-06,
+      "loss": 0.7611,
+      "step": 46160
+    },
+    {
+      "epoch": 1.777285851780558,
+      "grad_norm": 1.5484418869018555,
+      "learning_rate": 6.095719948762934e-06,
+      "loss": 0.8188,
+      "step": 46165
+    },
+    {
+      "epoch": 1.7774783445620788,
+      "grad_norm": 1.0949671268463135,
+      "learning_rate": 6.085327858022672e-06,
+      "loss": 0.7638,
+      "step": 46170
+    },
+    {
+      "epoch": 1.7776708373435997,
+      "grad_norm": 1.3612221479415894,
+      "learning_rate": 6.07494435515702e-06,
+      "loss": 0.8475,
+      "step": 46175
+    },
+    {
+      "epoch": 1.7778633301251203,
+      "grad_norm": 1.385720133781433,
+      "learning_rate": 6.06456944111552e-06,
+      "loss": 0.7755,
+      "step": 46180
+    },
+    {
+      "epoch": 1.778055822906641,
+      "grad_norm": 1.639599323272705,
+      "learning_rate": 6.05420311684689e-06,
+      "loss": 0.7855,
+      "step": 46185
+    },
+    {
+      "epoch": 1.7782483156881617,
+      "grad_norm": 1.1654934883117676,
+      "learning_rate": 6.043845383299018e-06,
+      "loss": 0.9169,
+      "step": 46190
+    },
+    {
+      "epoch": 1.7784408084696823,
+      "grad_norm": 0.8145331740379333,
+      "learning_rate": 6.033496241419112e-06,
+      "loss": 0.7727,
+      "step": 46195
+    },
+    {
+      "epoch": 1.7786333012512032,
+      "grad_norm": 1.7755072116851807,
+      "learning_rate": 6.0231556921534925e-06,
+      "loss": 0.9423,
+      "step": 46200
+    },
+    {
+      "epoch": 1.778825794032724,
+      "grad_norm": 0.9639081358909607,
+      "learning_rate": 6.012823736447748e-06,
+      "loss": 0.7255,
+      "step": 46205
+    },
+    {
+      "epoch": 1.7790182868142446,
+      "grad_norm": 1.6836614608764648,
+      "learning_rate": 6.002500375246667e-06,
+      "loss": 0.7516,
+      "step": 46210
+    },
+    {
+      "epoch": 1.7792107795957652,
+      "grad_norm": 0.9299353957176208,
+      "learning_rate": 5.9921856094942604e-06,
+      "loss": 0.7697,
+      "step": 46215
+    },
+    {
+      "epoch": 1.779403272377286,
+      "grad_norm": 1.0631067752838135,
+      "learning_rate": 5.981879440133709e-06,
+      "loss": 0.732,
+      "step": 46220
+    },
+    {
+      "epoch": 1.7795957651588066,
+      "grad_norm": 1.403578519821167,
+      "learning_rate": 5.97158186810749e-06,
+      "loss": 0.7957,
+      "step": 46225
+    },
+    {
+      "epoch": 1.7797882579403272,
+      "grad_norm": 2.0105881690979004,
+      "learning_rate": 5.961292894357217e-06,
+      "loss": 0.681,
+      "step": 46230
+    },
+    {
+      "epoch": 1.779980750721848,
+      "grad_norm": 1.8827532529830933,
+      "learning_rate": 5.951012519823762e-06,
+      "loss": 0.9873,
+      "step": 46235
+    },
+    {
+      "epoch": 1.7801732435033686,
+      "grad_norm": 0.4478518068790436,
+      "learning_rate": 5.940740745447193e-06,
+      "loss": 0.7578,
+      "step": 46240
+    },
+    {
+      "epoch": 1.7803657362848893,
+      "grad_norm": 1.727774977684021,
+      "learning_rate": 5.930477572166815e-06,
+      "loss": 0.7855,
+      "step": 46245
+    },
+    {
+      "epoch": 1.78055822906641,
+      "grad_norm": 0.8507458567619324,
+      "learning_rate": 5.920223000921077e-06,
+      "loss": 0.7533,
+      "step": 46250
+    },
+    {
+      "epoch": 1.7807507218479306,
+      "grad_norm": 1.1110725402832031,
+      "learning_rate": 5.909977032647762e-06,
+      "loss": 0.9928,
+      "step": 46255
+    },
+    {
+      "epoch": 1.7809432146294513,
+      "grad_norm": 1.1344168186187744,
+      "learning_rate": 5.899739668283732e-06,
+      "loss": 0.7474,
+      "step": 46260
+    },
+    {
+      "epoch": 1.781135707410972,
+      "grad_norm": 1.6794506311416626,
+      "learning_rate": 5.8895109087651616e-06,
+      "loss": 0.7804,
+      "step": 46265
+    },
+    {
+      "epoch": 1.7813282001924928,
+      "grad_norm": 1.0263005495071411,
+      "learning_rate": 5.879290755027378e-06,
+      "loss": 0.8814,
+      "step": 46270
+    },
+    {
+      "epoch": 1.7815206929740135,
+      "grad_norm": 1.7495372295379639,
+      "learning_rate": 5.869079208004991e-06,
+      "loss": 0.9627,
+      "step": 46275
+    },
+    {
+      "epoch": 1.7817131857555342,
+      "grad_norm": 1.242300271987915,
+      "learning_rate": 5.85887626863173e-06,
+      "loss": 0.8365,
+      "step": 46280
+    },
+    {
+      "epoch": 1.7819056785370548,
+      "grad_norm": 2.084526777267456,
+      "learning_rate": 5.848681937840605e-06,
+      "loss": 0.8406,
+      "step": 46285
+    },
+    {
+      "epoch": 1.7820981713185755,
+      "grad_norm": 0.9705522656440735,
+      "learning_rate": 5.8384962165638265e-06,
+      "loss": 0.7655,
+      "step": 46290
+    },
+    {
+      "epoch": 1.7822906641000964,
+      "grad_norm": 1.0004565715789795,
+      "learning_rate": 5.8283191057328045e-06,
+      "loss": 0.7498,
+      "step": 46295
+    },
+    {
+      "epoch": 1.782483156881617,
+      "grad_norm": 1.0826503038406372,
+      "learning_rate": 5.818150606278183e-06,
+      "loss": 0.8392,
+      "step": 46300
+    },
+    {
+      "epoch": 1.7826756496631377,
+      "grad_norm": 2.7211012840270996,
+      "learning_rate": 5.807990719129786e-06,
+      "loss": 0.8496,
+      "step": 46305
+    },
+    {
+      "epoch": 1.7828681424446584,
+      "grad_norm": 2.3229408264160156,
+      "learning_rate": 5.797839445216657e-06,
+      "loss": 0.7426,
+      "step": 46310
+    },
+    {
+      "epoch": 1.783060635226179,
+      "grad_norm": 2.3727757930755615,
+      "learning_rate": 5.78769678546709e-06,
+      "loss": 0.7056,
+      "step": 46315
+    },
+    {
+      "epoch": 1.7832531280076998,
+      "grad_norm": 1.557446837425232,
+      "learning_rate": 5.777562740808562e-06,
+      "loss": 0.8112,
+      "step": 46320
+    },
+    {
+      "epoch": 1.7834456207892204,
+      "grad_norm": 1.3999435901641846,
+      "learning_rate": 5.7674373121677226e-06,
+      "loss": 0.7352,
+      "step": 46325
+    },
+    {
+      "epoch": 1.783638113570741,
+      "grad_norm": 1.531625509262085,
+      "learning_rate": 5.757320500470542e-06,
+      "loss": 0.6754,
+      "step": 46330
+    },
+    {
+      "epoch": 1.7838306063522618,
+      "grad_norm": 1.059308648109436,
+      "learning_rate": 5.74721230664208e-06,
+      "loss": 0.7609,
+      "step": 46335
+    },
+    {
+      "epoch": 1.7840230991337824,
+      "grad_norm": 1.5941221714019775,
+      "learning_rate": 5.737112731606698e-06,
+      "loss": 0.8795,
+      "step": 46340
+    },
+    {
+      "epoch": 1.784215591915303,
+      "grad_norm": 1.1354868412017822,
+      "learning_rate": 5.7270217762879105e-06,
+      "loss": 0.7511,
+      "step": 46345
+    },
+    {
+      "epoch": 1.7844080846968238,
+      "grad_norm": 1.1096879243850708,
+      "learning_rate": 5.716939441608504e-06,
+      "loss": 0.6635,
+      "step": 46350
+    },
+    {
+      "epoch": 1.7846005774783444,
+      "grad_norm": 2.472426176071167,
+      "learning_rate": 5.706865728490373e-06,
+      "loss": 0.805,
+      "step": 46355
+    },
+    {
+      "epoch": 1.784793070259865,
+      "grad_norm": 1.060925841331482,
+      "learning_rate": 5.696800637854782e-06,
+      "loss": 0.8253,
+      "step": 46360
+    },
+    {
+      "epoch": 1.784985563041386,
+      "grad_norm": 1.1766963005065918,
+      "learning_rate": 5.686744170622049e-06,
+      "loss": 0.6249,
+      "step": 46365
+    },
+    {
+      "epoch": 1.7851780558229067,
+      "grad_norm": 1.531556248664856,
+      "learning_rate": 5.676696327711795e-06,
+      "loss": 0.7294,
+      "step": 46370
+    },
+    {
+      "epoch": 1.7853705486044273,
+      "grad_norm": 1.9163962602615356,
+      "learning_rate": 5.666657110042828e-06,
+      "loss": 0.7667,
+      "step": 46375
+    },
+    {
+      "epoch": 1.785563041385948,
+      "grad_norm": 1.7495805025100708,
+      "learning_rate": 5.65662651853317e-06,
+      "loss": 0.8198,
+      "step": 46380
+    },
+    {
+      "epoch": 1.7857555341674687,
+      "grad_norm": 0.7979353666305542,
+      "learning_rate": 5.6466045541000546e-06,
+      "loss": 0.7781,
+      "step": 46385
+    },
+    {
+      "epoch": 1.7859480269489896,
+      "grad_norm": 1.199220895767212,
+      "learning_rate": 5.636591217659903e-06,
+      "loss": 0.8034,
+      "step": 46390
+    },
+    {
+      "epoch": 1.7861405197305102,
+      "grad_norm": 1.2027233839035034,
+      "learning_rate": 5.626586510128384e-06,
+      "loss": 0.7879,
+      "step": 46395
+    },
+    {
+      "epoch": 1.786333012512031,
+      "grad_norm": 0.7976201772689819,
+      "learning_rate": 5.616590432420376e-06,
+      "loss": 0.6871,
+      "step": 46400
+    },
+    {
+      "epoch": 1.7865255052935516,
+      "grad_norm": 1.0275331735610962,
+      "learning_rate": 5.606602985449949e-06,
+      "loss": 0.7664,
+      "step": 46405
+    },
+    {
+      "epoch": 1.7867179980750723,
+      "grad_norm": 2.2883615493774414,
+      "learning_rate": 5.59662417013036e-06,
+      "loss": 0.957,
+      "step": 46410
+    },
+    {
+      "epoch": 1.786910490856593,
+      "grad_norm": 0.8585185408592224,
+      "learning_rate": 5.586653987374125e-06,
+      "loss": 0.6444,
+      "step": 46415
+    },
+    {
+      "epoch": 1.7871029836381136,
+      "grad_norm": 2.087956666946411,
+      "learning_rate": 5.576692438092956e-06,
+      "loss": 0.8373,
+      "step": 46420
+    },
+    {
+      "epoch": 1.7872954764196343,
+      "grad_norm": 1.8517224788665771,
+      "learning_rate": 5.566739523197762e-06,
+      "loss": 0.8641,
+      "step": 46425
+    },
+    {
+      "epoch": 1.787487969201155,
+      "grad_norm": 1.3989286422729492,
+      "learning_rate": 5.556795243598678e-06,
+      "loss": 0.8464,
+      "step": 46430
+    },
+    {
+      "epoch": 1.7876804619826756,
+      "grad_norm": 1.8883378505706787,
+      "learning_rate": 5.546859600205057e-06,
+      "loss": 0.8401,
+      "step": 46435
+    },
+    {
+      "epoch": 1.7878729547641963,
+      "grad_norm": 1.1194407939910889,
+      "learning_rate": 5.536932593925403e-06,
+      "loss": 0.8559,
+      "step": 46440
+    },
+    {
+      "epoch": 1.788065447545717,
+      "grad_norm": 1.501118779182434,
+      "learning_rate": 5.527014225667526e-06,
+      "loss": 0.7472,
+      "step": 46445
+    },
+    {
+      "epoch": 1.7882579403272376,
+      "grad_norm": 1.1292697191238403,
+      "learning_rate": 5.517104496338365e-06,
+      "loss": 0.8151,
+      "step": 46450
+    },
+    {
+      "epoch": 1.7884504331087583,
+      "grad_norm": 1.5353429317474365,
+      "learning_rate": 5.507203406844097e-06,
+      "loss": 0.693,
+      "step": 46455
+    },
+    {
+      "epoch": 1.788642925890279,
+      "grad_norm": 1.1052043437957764,
+      "learning_rate": 5.497310958090129e-06,
+      "loss": 0.74,
+      "step": 46460
+    },
+    {
+      "epoch": 1.7888354186717998,
+      "grad_norm": 1.2323806285858154,
+      "learning_rate": 5.487427150981061e-06,
+      "loss": 0.7457,
+      "step": 46465
+    },
+    {
+      "epoch": 1.7890279114533205,
+      "grad_norm": 1.003144383430481,
+      "learning_rate": 5.477551986420659e-06,
+      "loss": 0.6708,
+      "step": 46470
+    },
+    {
+      "epoch": 1.7892204042348412,
+      "grad_norm": 1.9035112857818604,
+      "learning_rate": 5.467685465312e-06,
+      "loss": 0.7999,
+      "step": 46475
+    },
+    {
+      "epoch": 1.7894128970163619,
+      "grad_norm": 2.7132394313812256,
+      "learning_rate": 5.457827588557285e-06,
+      "loss": 0.7343,
+      "step": 46480
+    },
+    {
+      "epoch": 1.7896053897978825,
+      "grad_norm": 2.29404878616333,
+      "learning_rate": 5.4479783570579366e-06,
+      "loss": 0.8014,
+      "step": 46485
+    },
+    {
+      "epoch": 1.7897978825794034,
+      "grad_norm": 1.4308083057403564,
+      "learning_rate": 5.438137771714635e-06,
+      "loss": 0.888,
+      "step": 46490
+    },
+    {
+      "epoch": 1.789990375360924,
+      "grad_norm": 1.0539064407348633,
+      "learning_rate": 5.428305833427216e-06,
+      "loss": 0.8976,
+      "step": 46495
+    },
+    {
+      "epoch": 1.7901828681424448,
+      "grad_norm": 1.6149170398712158,
+      "learning_rate": 5.41848254309475e-06,
+      "loss": 0.8853,
+      "step": 46500
+    },
+    {
+      "epoch": 1.7903753609239654,
+      "grad_norm": 0.8334898948669434,
+      "learning_rate": 5.408667901615494e-06,
+      "loss": 0.6841,
+      "step": 46505
+    },
+    {
+      "epoch": 1.790567853705486,
+      "grad_norm": 1.9171321392059326,
+      "learning_rate": 5.398861909886965e-06,
+      "loss": 0.6301,
+      "step": 46510
+    },
+    {
+      "epoch": 1.7907603464870068,
+      "grad_norm": 1.2018779516220093,
+      "learning_rate": 5.391023344927259e-06,
+      "loss": 1.3233,
+      "step": 46515
+    },
+    {
+      "epoch": 1.7909528392685274,
+      "grad_norm": 1.1932569742202759,
+      "learning_rate": 5.381232925009128e-06,
+      "loss": 0.7556,
+      "step": 46520
+    },
+    {
+      "epoch": 1.7911453320500481,
+      "grad_norm": 0.987067699432373,
+      "learning_rate": 5.37145115735046e-06,
+      "loss": 0.6867,
+      "step": 46525
+    },
+    {
+      "epoch": 1.7913378248315688,
+      "grad_norm": 1.0372260808944702,
+      "learning_rate": 5.361678042845731e-06,
+      "loss": 0.7445,
+      "step": 46530
+    },
+    {
+      "epoch": 1.7915303176130895,
+      "grad_norm": 0.9253237843513489,
+      "learning_rate": 5.351913582388635e-06,
+      "loss": 0.7732,
+      "step": 46535
+    },
+    {
+      "epoch": 1.7917228103946101,
+      "grad_norm": 0.9296650886535645,
+      "learning_rate": 5.342157776872025e-06,
+      "loss": 0.7397,
+      "step": 46540
+    },
+    {
+      "epoch": 1.7919153031761308,
+      "grad_norm": 2.767267942428589,
+      "learning_rate": 5.332410627188067e-06,
+      "loss": 0.7794,
+      "step": 46545
+    },
+    {
+      "epoch": 1.7921077959576515,
+      "grad_norm": 1.5188485383987427,
+      "learning_rate": 5.322672134228024e-06,
+      "loss": 0.7275,
+      "step": 46550
+    },
+    {
+      "epoch": 1.7923002887391721,
+      "grad_norm": 1.1152452230453491,
+      "learning_rate": 5.312942298882439e-06,
+      "loss": 0.7639,
+      "step": 46555
+    },
+    {
+      "epoch": 1.792492781520693,
+      "grad_norm": 1.1599907875061035,
+      "learning_rate": 5.303221122041036e-06,
+      "loss": 0.7895,
+      "step": 46560
+    },
+    {
+      "epoch": 1.7926852743022137,
+      "grad_norm": 1.0555464029312134,
+      "learning_rate": 5.293508604592768e-06,
+      "loss": 0.8518,
+      "step": 46565
+    },
+    {
+      "epoch": 1.7928777670837344,
+      "grad_norm": 1.0384188890457153,
+      "learning_rate": 5.283804747425747e-06,
+      "loss": 0.7783,
+      "step": 46570
+    },
+    {
+      "epoch": 1.793070259865255,
+      "grad_norm": 1.148019790649414,
+      "learning_rate": 5.274109551427342e-06,
+      "loss": 0.8909,
+      "step": 46575
+    },
+    {
+      "epoch": 1.7932627526467757,
+      "grad_norm": 2.0219950675964355,
+      "learning_rate": 5.264423017484122e-06,
+      "loss": 0.7428,
+      "step": 46580
+    },
+    {
+      "epoch": 1.7934552454282966,
+      "grad_norm": 1.3461076021194458,
+      "learning_rate": 5.254745146481843e-06,
+      "loss": 0.675,
+      "step": 46585
+    },
+    {
+      "epoch": 1.7936477382098173,
+      "grad_norm": 1.6037160158157349,
+      "learning_rate": 5.2450759393055104e-06,
+      "loss": 0.8687,
+      "step": 46590
+    },
+    {
+      "epoch": 1.793840230991338,
+      "grad_norm": 0.929437518119812,
+      "learning_rate": 5.235415396839283e-06,
+      "loss": 0.9081,
+      "step": 46595
+    },
+    {
+      "epoch": 1.7940327237728586,
+      "grad_norm": 0.9498231410980225,
+      "learning_rate": 5.225763519966542e-06,
+      "loss": 0.7378,
+      "step": 46600
+    },
+    {
+      "epoch": 1.7942252165543793,
+      "grad_norm": 1.3678570985794067,
+      "learning_rate": 5.216120309569917e-06,
+      "loss": 1.001,
+      "step": 46605
+    },
+    {
+      "epoch": 1.7944177093359,
+      "grad_norm": 1.7113560438156128,
+      "learning_rate": 5.206485766531222e-06,
+      "loss": 0.8138,
+      "step": 46610
+    },
+    {
+      "epoch": 1.7946102021174206,
+      "grad_norm": 1.4284687042236328,
+      "learning_rate": 5.19685989173142e-06,
+      "loss": 0.7692,
+      "step": 46615
+    },
+    {
+      "epoch": 1.7948026948989413,
+      "grad_norm": 1.2268130779266357,
+      "learning_rate": 5.1872426860507975e-06,
+      "loss": 0.7408,
+      "step": 46620
+    },
+    {
+      "epoch": 1.794995187680462,
+      "grad_norm": 1.4930412769317627,
+      "learning_rate": 5.177634150368738e-06,
+      "loss": 0.9012,
+      "step": 46625
+    },
+    {
+      "epoch": 1.7951876804619826,
+      "grad_norm": 1.3684518337249756,
+      "learning_rate": 5.1680342855638945e-06,
+      "loss": 0.927,
+      "step": 46630
+    },
+    {
+      "epoch": 1.7953801732435033,
+      "grad_norm": 1.5305689573287964,
+      "learning_rate": 5.158443092514109e-06,
+      "loss": 0.8608,
+      "step": 46635
+    },
+    {
+      "epoch": 1.795572666025024,
+      "grad_norm": 1.6696339845657349,
+      "learning_rate": 5.148860572096459e-06,
+      "loss": 0.8078,
+      "step": 46640
+    },
+    {
+      "epoch": 1.7957651588065446,
+      "grad_norm": 1.3109546899795532,
+      "learning_rate": 5.139286725187143e-06,
+      "loss": 0.7532,
+      "step": 46645
+    },
+    {
+      "epoch": 1.7959576515880653,
+      "grad_norm": 1.0849995613098145,
+      "learning_rate": 5.129721552661681e-06,
+      "loss": 0.9083,
+      "step": 46650
+    },
+    {
+      "epoch": 1.796150144369586,
+      "grad_norm": 2.3269553184509277,
+      "learning_rate": 5.120165055394721e-06,
+      "loss": 0.7511,
+      "step": 46655
+    },
+    {
+      "epoch": 1.7963426371511069,
+      "grad_norm": 1.75252103805542,
+      "learning_rate": 5.110617234260151e-06,
+      "loss": 0.7646,
+      "step": 46660
+    },
+    {
+      "epoch": 1.7965351299326275,
+      "grad_norm": 1.0314127206802368,
+      "learning_rate": 5.1010780901310395e-06,
+      "loss": 0.6695,
+      "step": 46665
+    },
+    {
+      "epoch": 1.7967276227141482,
+      "grad_norm": 1.0665658712387085,
+      "learning_rate": 5.091547623879711e-06,
+      "loss": 0.8726,
+      "step": 46670
+    },
+    {
+      "epoch": 1.7969201154956689,
+      "grad_norm": 1.8659719228744507,
+      "learning_rate": 5.082025836377624e-06,
+      "loss": 0.8587,
+      "step": 46675
+    },
+    {
+      "epoch": 1.7971126082771898,
+      "grad_norm": 1.0363880395889282,
+      "learning_rate": 5.072512728495493e-06,
+      "loss": 0.7135,
+      "step": 46680
+    },
+    {
+      "epoch": 1.7973051010587104,
+      "grad_norm": 1.0198173522949219,
+      "learning_rate": 5.063008301103245e-06,
+      "loss": 0.8571,
+      "step": 46685
+    },
+    {
+      "epoch": 1.7974975938402311,
+      "grad_norm": 1.313674807548523,
+      "learning_rate": 5.0535125550699834e-06,
+      "loss": 0.8776,
+      "step": 46690
+    },
+    {
+      "epoch": 1.7976900866217518,
+      "grad_norm": 0.9799014925956726,
+      "learning_rate": 5.044025491264049e-06,
+      "loss": 0.8095,
+      "step": 46695
+    },
+    {
+      "epoch": 1.7978825794032725,
+      "grad_norm": 0.9720066785812378,
+      "learning_rate": 5.0345471105529345e-06,
+      "loss": 0.7772,
+      "step": 46700
+    },
+    {
+      "epoch": 1.7980750721847931,
+      "grad_norm": 1.2402182817459106,
+      "learning_rate": 5.025077413803425e-06,
+      "loss": 0.7378,
+      "step": 46705
+    },
+    {
+      "epoch": 1.7982675649663138,
+      "grad_norm": 1.4397085905075073,
+      "learning_rate": 5.015616401881418e-06,
+      "loss": 0.8355,
+      "step": 46710
+    },
+    {
+      "epoch": 1.7984600577478345,
+      "grad_norm": 1.2105292081832886,
+      "learning_rate": 5.0061640756520754e-06,
+      "loss": 0.8271,
+      "step": 46715
+    },
+    {
+      "epoch": 1.7986525505293551,
+      "grad_norm": 1.2763471603393555,
+      "learning_rate": 4.996720435979763e-06,
+      "loss": 0.7776,
+      "step": 46720
+    },
+    {
+      "epoch": 1.7988450433108758,
+      "grad_norm": 1.6655856370925903,
+      "learning_rate": 4.987285483728033e-06,
+      "loss": 0.7638,
+      "step": 46725
+    },
+    {
+      "epoch": 1.7990375360923965,
+      "grad_norm": 1.3366689682006836,
+      "learning_rate": 4.977859219759617e-06,
+      "loss": 0.6957,
+      "step": 46730
+    },
+    {
+      "epoch": 1.7992300288739171,
+      "grad_norm": 0.9900131225585938,
+      "learning_rate": 4.968441644936539e-06,
+      "loss": 0.7898,
+      "step": 46735
+    },
+    {
+      "epoch": 1.7994225216554378,
+      "grad_norm": 1.191714882850647,
+      "learning_rate": 4.95903276011993e-06,
+      "loss": 0.8838,
+      "step": 46740
+    },
+    {
+      "epoch": 1.7996150144369585,
+      "grad_norm": 1.0521550178527832,
+      "learning_rate": 4.949632566170181e-06,
+      "loss": 0.7767,
+      "step": 46745
+    },
+    {
+      "epoch": 1.7998075072184792,
+      "grad_norm": 1.6969459056854248,
+      "learning_rate": 4.940241063946893e-06,
+      "loss": 0.8962,
+      "step": 46750
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4018580913543701,
+      "learning_rate": 4.930858254308856e-06,
+      "loss": 0.8324,
+      "step": 46755
+    },
+    {
+      "epoch": 1.8001924927815207,
+      "grad_norm": 1.4890011548995972,
+      "learning_rate": 4.921484138114029e-06,
+      "loss": 0.6562,
+      "step": 46760
+    },
+    {
+      "epoch": 1.8003849855630414,
+      "grad_norm": 1.6725586652755737,
+      "learning_rate": 4.912118716219672e-06,
+      "loss": 0.7311,
+      "step": 46765
+    },
+    {
+      "epoch": 1.800577478344562,
+      "grad_norm": 1.5450921058654785,
+      "learning_rate": 4.9027619894821404e-06,
+      "loss": 0.8046,
+      "step": 46770
+    },
+    {
+      "epoch": 1.8007699711260827,
+      "grad_norm": 1.5963780879974365,
+      "learning_rate": 4.893413958757065e-06,
+      "loss": 0.7562,
+      "step": 46775
+    },
+    {
+      "epoch": 1.8009624639076036,
+      "grad_norm": 2.59673810005188,
+      "learning_rate": 4.88407462489926e-06,
+      "loss": 0.7906,
+      "step": 46780
+    },
+    {
+      "epoch": 1.8011549566891243,
+      "grad_norm": 1.0897094011306763,
+      "learning_rate": 4.874743988762753e-06,
+      "loss": 0.6483,
+      "step": 46785
+    },
+    {
+      "epoch": 1.801347449470645,
+      "grad_norm": 1.5644094944000244,
+      "learning_rate": 4.865422051200752e-06,
+      "loss": 0.668,
+      "step": 46790
+    },
+    {
+      "epoch": 1.8015399422521656,
+      "grad_norm": 1.7178467512130737,
+      "learning_rate": 4.856108813065696e-06,
+      "loss": 0.7188,
+      "step": 46795
+    },
+    {
+      "epoch": 1.8017324350336863,
+      "grad_norm": 1.8276166915893555,
+      "learning_rate": 4.846804275209216e-06,
+      "loss": 0.7897,
+      "step": 46800
+    },
+    {
+      "epoch": 1.801924927815207,
+      "grad_norm": 1.107142448425293,
+      "learning_rate": 4.8375084384821635e-06,
+      "loss": 0.6513,
+      "step": 46805
+    },
+    {
+      "epoch": 1.8021174205967276,
+      "grad_norm": 1.1322238445281982,
+      "learning_rate": 4.82822130373457e-06,
+      "loss": 0.8737,
+      "step": 46810
+    },
+    {
+      "epoch": 1.8023099133782483,
+      "grad_norm": 1.5215245485305786,
+      "learning_rate": 4.818942871815679e-06,
+      "loss": 0.7255,
+      "step": 46815
+    },
+    {
+      "epoch": 1.802502406159769,
+      "grad_norm": 1.0915231704711914,
+      "learning_rate": 4.809673143573934e-06,
+      "loss": 0.6991,
+      "step": 46820
+    },
+    {
+      "epoch": 1.8026948989412896,
+      "grad_norm": 1.1544078588485718,
+      "learning_rate": 4.800412119857012e-06,
+      "loss": 0.8788,
+      "step": 46825
+    },
+    {
+      "epoch": 1.8028873917228103,
+      "grad_norm": 1.521331787109375,
+      "learning_rate": 4.791159801511769e-06,
+      "loss": 0.8318,
+      "step": 46830
+    },
+    {
+      "epoch": 1.803079884504331,
+      "grad_norm": 1.9487007856369019,
+      "learning_rate": 4.781916189384239e-06,
+      "loss": 0.7977,
+      "step": 46835
+    },
+    {
+      "epoch": 1.8032723772858517,
+      "grad_norm": 0.9875978827476501,
+      "learning_rate": 4.772681284319736e-06,
+      "loss": 0.7773,
+      "step": 46840
+    },
+    {
+      "epoch": 1.8034648700673723,
+      "grad_norm": 0.9986359477043152,
+      "learning_rate": 4.7634550871626935e-06,
+      "loss": 0.7418,
+      "step": 46845
+    },
+    {
+      "epoch": 1.8036573628488932,
+      "grad_norm": 0.9883870482444763,
+      "learning_rate": 4.754237598756806e-06,
+      "loss": 0.6378,
+      "step": 46850
+    },
+    {
+      "epoch": 1.803849855630414,
+      "grad_norm": 1.2948468923568726,
+      "learning_rate": 4.745028819944941e-06,
+      "loss": 0.6721,
+      "step": 46855
+    },
+    {
+      "epoch": 1.8040423484119346,
+      "grad_norm": 1.422836422920227,
+      "learning_rate": 4.735828751569194e-06,
+      "loss": 0.77,
+      "step": 46860
+    },
+    {
+      "epoch": 1.8042348411934552,
+      "grad_norm": 0.956969678401947,
+      "learning_rate": 4.726637394470812e-06,
+      "loss": 0.7077,
+      "step": 46865
+    },
+    {
+      "epoch": 1.804427333974976,
+      "grad_norm": 1.3407295942306519,
+      "learning_rate": 4.717454749490336e-06,
+      "loss": 0.7287,
+      "step": 46870
+    },
+    {
+      "epoch": 1.8046198267564968,
+      "grad_norm": 1.0364344120025635,
+      "learning_rate": 4.7082808174674255e-06,
+      "loss": 0.8322,
+      "step": 46875
+    },
+    {
+      "epoch": 1.8048123195380175,
+      "grad_norm": 1.2396444082260132,
+      "learning_rate": 4.6991155992409885e-06,
+      "loss": 0.6264,
+      "step": 46880
+    },
+    {
+      "epoch": 1.8050048123195381,
+      "grad_norm": 1.7426707744598389,
+      "learning_rate": 4.6899590956491105e-06,
+      "loss": 0.6881,
+      "step": 46885
+    },
+    {
+      "epoch": 1.8051973051010588,
+      "grad_norm": 1.3045077323913574,
+      "learning_rate": 4.68081130752912e-06,
+      "loss": 0.7758,
+      "step": 46890
+    },
+    {
+      "epoch": 1.8053897978825795,
+      "grad_norm": 1.462531566619873,
+      "learning_rate": 4.671672235717494e-06,
+      "loss": 0.9664,
+      "step": 46895
+    },
+    {
+      "epoch": 1.8055822906641001,
+      "grad_norm": 1.063500165939331,
+      "learning_rate": 4.662541881049942e-06,
+      "loss": 0.8255,
+      "step": 46900
+    },
+    {
+      "epoch": 1.8057747834456208,
+      "grad_norm": 1.0657175779342651,
+      "learning_rate": 4.653420244361395e-06,
+      "loss": 0.8206,
+      "step": 46905
+    },
+    {
+      "epoch": 1.8059672762271415,
+      "grad_norm": 1.0056036710739136,
+      "learning_rate": 4.644307326485941e-06,
+      "loss": 0.8588,
+      "step": 46910
+    },
+    {
+      "epoch": 1.8061597690086622,
+      "grad_norm": 1.1666234731674194,
+      "learning_rate": 4.635203128256927e-06,
+      "loss": 0.5983,
+      "step": 46915
+    },
+    {
+      "epoch": 1.8063522617901828,
+      "grad_norm": 1.6504344940185547,
+      "learning_rate": 4.626107650506839e-06,
+      "loss": 0.7481,
+      "step": 46920
+    },
+    {
+      "epoch": 1.8065447545717035,
+      "grad_norm": 1.4112893342971802,
+      "learning_rate": 4.617020894067403e-06,
+      "loss": 0.9829,
+      "step": 46925
+    },
+    {
+      "epoch": 1.8067372473532242,
+      "grad_norm": 1.058632493019104,
+      "learning_rate": 4.607942859769565e-06,
+      "loss": 0.7069,
+      "step": 46930
+    },
+    {
+      "epoch": 1.8069297401347448,
+      "grad_norm": 1.1201543807983398,
+      "learning_rate": 4.598873548443427e-06,
+      "loss": 0.8061,
+      "step": 46935
+    },
+    {
+      "epoch": 1.8071222329162655,
+      "grad_norm": 0.8915508389472961,
+      "learning_rate": 4.589812960918338e-06,
+      "loss": 0.82,
+      "step": 46940
+    },
+    {
+      "epoch": 1.8073147256977862,
+      "grad_norm": 2.371730327606201,
+      "learning_rate": 4.580761098022835e-06,
+      "loss": 0.8294,
+      "step": 46945
+    },
+    {
+      "epoch": 1.807507218479307,
+      "grad_norm": 1.3790137767791748,
+      "learning_rate": 4.571717960584598e-06,
+      "loss": 0.7928,
+      "step": 46950
+    },
+    {
+      "epoch": 1.8076997112608277,
+      "grad_norm": 1.2915151119232178,
+      "learning_rate": 4.562683549430624e-06,
+      "loss": 0.7858,
+      "step": 46955
+    },
+    {
+      "epoch": 1.8078922040423484,
+      "grad_norm": 1.653926968574524,
+      "learning_rate": 4.553657865387018e-06,
+      "loss": 0.8588,
+      "step": 46960
+    },
+    {
+      "epoch": 1.808084696823869,
+      "grad_norm": 1.2625577449798584,
+      "learning_rate": 4.54464090927913e-06,
+      "loss": 0.8673,
+      "step": 46965
+    },
+    {
+      "epoch": 1.8082771896053897,
+      "grad_norm": 0.9362401962280273,
+      "learning_rate": 4.5356326819314894e-06,
+      "loss": 0.7839,
+      "step": 46970
+    },
+    {
+      "epoch": 1.8084696823869106,
+      "grad_norm": 1.078352451324463,
+      "learning_rate": 4.526633184167861e-06,
+      "loss": 0.8862,
+      "step": 46975
+    },
+    {
+      "epoch": 1.8086621751684313,
+      "grad_norm": 1.3022172451019287,
+      "learning_rate": 4.517642416811152e-06,
+      "loss": 0.8553,
+      "step": 46980
+    },
+    {
+      "epoch": 1.808854667949952,
+      "grad_norm": 1.2438225746154785,
+      "learning_rate": 4.50866038068356e-06,
+      "loss": 0.7513,
+      "step": 46985
+    },
+    {
+      "epoch": 1.8090471607314726,
+      "grad_norm": 0.9540335536003113,
+      "learning_rate": 4.4996870766063845e-06,
+      "loss": 0.7287,
+      "step": 46990
+    },
+    {
+      "epoch": 1.8092396535129933,
+      "grad_norm": 1.5209485292434692,
+      "learning_rate": 4.490722505400191e-06,
+      "loss": 0.8749,
+      "step": 46995
+    },
+    {
+      "epoch": 1.809432146294514,
+      "grad_norm": 1.1515183448791504,
+      "learning_rate": 4.481766667884757e-06,
+      "loss": 0.8147,
+      "step": 47000
+    },
+    {
+      "epoch": 1.8096246390760347,
+      "grad_norm": 1.2854846715927124,
+      "learning_rate": 4.472819564878994e-06,
+      "loss": 0.8026,
+      "step": 47005
+    },
+    {
+      "epoch": 1.8098171318575553,
+      "grad_norm": 1.7539061307907104,
+      "learning_rate": 4.4638811972010696e-06,
+      "loss": 0.8365,
+      "step": 47010
+    },
+    {
+      "epoch": 1.810009624639076,
+      "grad_norm": 1.8201215267181396,
+      "learning_rate": 4.454951565668341e-06,
+      "loss": 0.9819,
+      "step": 47015
+    },
+    {
+      "epoch": 1.8102021174205967,
+      "grad_norm": 1.3828089237213135,
+      "learning_rate": 4.4460306710973786e-06,
+      "loss": 0.7404,
+      "step": 47020
+    },
+    {
+      "epoch": 1.8103946102021173,
+      "grad_norm": 2.6402223110198975,
+      "learning_rate": 4.437118514303895e-06,
+      "loss": 0.887,
+      "step": 47025
+    },
+    {
+      "epoch": 1.810587102983638,
+      "grad_norm": 1.3061926364898682,
+      "learning_rate": 4.428215096102906e-06,
+      "loss": 0.8308,
+      "step": 47030
+    },
+    {
+      "epoch": 1.8107795957651587,
+      "grad_norm": 1.3701341152191162,
+      "learning_rate": 4.419320417308526e-06,
+      "loss": 0.7879,
+      "step": 47035
+    },
+    {
+      "epoch": 1.8109720885466793,
+      "grad_norm": 1.355481505393982,
+      "learning_rate": 4.410434478734127e-06,
+      "loss": 0.8011,
+      "step": 47040
+    },
+    {
+      "epoch": 1.8111645813282002,
+      "grad_norm": 1.3704595565795898,
+      "learning_rate": 4.401557281192281e-06,
+      "loss": 0.8246,
+      "step": 47045
+    },
+    {
+      "epoch": 1.811357074109721,
+      "grad_norm": 1.3024649620056152,
+      "learning_rate": 4.39268882549474e-06,
+      "loss": 0.8316,
+      "step": 47050
+    },
+    {
+      "epoch": 1.8115495668912416,
+      "grad_norm": 1.1199791431427002,
+      "learning_rate": 4.383829112452454e-06,
+      "loss": 0.8883,
+      "step": 47055
+    },
+    {
+      "epoch": 1.8117420596727623,
+      "grad_norm": 1.2826857566833496,
+      "learning_rate": 4.374978142875608e-06,
+      "loss": 0.7167,
+      "step": 47060
+    },
+    {
+      "epoch": 1.811934552454283,
+      "grad_norm": 0.9372978806495667,
+      "learning_rate": 4.366135917573555e-06,
+      "loss": 0.7517,
+      "step": 47065
+    },
+    {
+      "epoch": 1.8121270452358038,
+      "grad_norm": 1.2647088766098022,
+      "learning_rate": 4.357302437354848e-06,
+      "loss": 0.6803,
+      "step": 47070
+    },
+    {
+      "epoch": 1.8123195380173245,
+      "grad_norm": 0.9146122932434082,
+      "learning_rate": 4.348477703027254e-06,
+      "loss": 0.8121,
+      "step": 47075
+    },
+    {
+      "epoch": 1.8125120307988452,
+      "grad_norm": 1.3470911979675293,
+      "learning_rate": 4.3396617153977585e-06,
+      "loss": 0.7432,
+      "step": 47080
+    },
+    {
+      "epoch": 1.8127045235803658,
+      "grad_norm": 1.119855284690857,
+      "learning_rate": 4.330854475272483e-06,
+      "loss": 0.694,
+      "step": 47085
+    },
+    {
+      "epoch": 1.8128970163618865,
+      "grad_norm": 0.8852443099021912,
+      "learning_rate": 4.322055983456841e-06,
+      "loss": 0.8118,
+      "step": 47090
+    },
+    {
+      "epoch": 1.8130895091434072,
+      "grad_norm": 1.6684072017669678,
+      "learning_rate": 4.313266240755354e-06,
+      "loss": 0.9023,
+      "step": 47095
+    },
+    {
+      "epoch": 1.8132820019249278,
+      "grad_norm": 1.1562302112579346,
+      "learning_rate": 4.304485247971812e-06,
+      "loss": 0.7498,
+      "step": 47100
+    },
+    {
+      "epoch": 1.8134744947064485,
+      "grad_norm": 1.1458892822265625,
+      "learning_rate": 4.2957130059091635e-06,
+      "loss": 0.7468,
+      "step": 47105
+    },
+    {
+      "epoch": 1.8136669874879692,
+      "grad_norm": 1.3492836952209473,
+      "learning_rate": 4.286949515369587e-06,
+      "loss": 0.7202,
+      "step": 47110
+    },
+    {
+      "epoch": 1.8138594802694898,
+      "grad_norm": 1.3468396663665771,
+      "learning_rate": 4.278194777154432e-06,
+      "loss": 0.9034,
+      "step": 47115
+    },
+    {
+      "epoch": 1.8140519730510105,
+      "grad_norm": 1.1316415071487427,
+      "learning_rate": 4.269448792064257e-06,
+      "loss": 0.8748,
+      "step": 47120
+    },
+    {
+      "epoch": 1.8142444658325312,
+      "grad_norm": 2.3300271034240723,
+      "learning_rate": 4.260711560898833e-06,
+      "loss": 0.6797,
+      "step": 47125
+    },
+    {
+      "epoch": 1.8144369586140519,
+      "grad_norm": 1.528640866279602,
+      "learning_rate": 4.251983084457134e-06,
+      "loss": 0.8478,
+      "step": 47130
+    },
+    {
+      "epoch": 1.8146294513955725,
+      "grad_norm": 1.0944819450378418,
+      "learning_rate": 4.2432633635373084e-06,
+      "loss": 0.7424,
+      "step": 47135
+    },
+    {
+      "epoch": 1.8148219441770932,
+      "grad_norm": 1.4346030950546265,
+      "learning_rate": 4.234552398936709e-06,
+      "loss": 0.7892,
+      "step": 47140
+    },
+    {
+      "epoch": 1.815014436958614,
+      "grad_norm": 1.987533450126648,
+      "learning_rate": 4.225850191451908e-06,
+      "loss": 0.728,
+      "step": 47145
+    },
+    {
+      "epoch": 1.8152069297401348,
+      "grad_norm": 1.4819614887237549,
+      "learning_rate": 4.217156741878658e-06,
+      "loss": 0.8437,
+      "step": 47150
+    },
+    {
+      "epoch": 1.8153994225216554,
+      "grad_norm": 1.6680911779403687,
+      "learning_rate": 4.2084720510119355e-06,
+      "loss": 0.9606,
+      "step": 47155
+    },
+    {
+      "epoch": 1.815591915303176,
+      "grad_norm": 1.2909523248672485,
+      "learning_rate": 4.19979611964586e-06,
+      "loss": 0.8903,
+      "step": 47160
+    },
+    {
+      "epoch": 1.815784408084697,
+      "grad_norm": 1.6894687414169312,
+      "learning_rate": 4.19112894857383e-06,
+      "loss": 0.8166,
+      "step": 47165
+    },
+    {
+      "epoch": 1.8159769008662177,
+      "grad_norm": 1.0353437662124634,
+      "learning_rate": 4.1824705385883786e-06,
+      "loss": 0.6971,
+      "step": 47170
+    },
+    {
+      "epoch": 1.8161693936477383,
+      "grad_norm": 1.8149842023849487,
+      "learning_rate": 4.17382089048125e-06,
+      "loss": 0.8447,
+      "step": 47175
+    },
+    {
+      "epoch": 1.816361886429259,
+      "grad_norm": 1.26813805103302,
+      "learning_rate": 4.165180005043423e-06,
+      "loss": 0.7669,
+      "step": 47180
+    },
+    {
+      "epoch": 1.8165543792107797,
+      "grad_norm": 1.110489010810852,
+      "learning_rate": 4.156547883065043e-06,
+      "loss": 0.7015,
+      "step": 47185
+    },
+    {
+      "epoch": 1.8167468719923003,
+      "grad_norm": 1.940280795097351,
+      "learning_rate": 4.147924525335434e-06,
+      "loss": 0.7819,
+      "step": 47190
+    },
+    {
+      "epoch": 1.816939364773821,
+      "grad_norm": 0.8619933724403381,
+      "learning_rate": 4.139309932643187e-06,
+      "loss": 0.9221,
+      "step": 47195
+    },
+    {
+      "epoch": 1.8171318575553417,
+      "grad_norm": 1.2719132900238037,
+      "learning_rate": 4.130704105776018e-06,
+      "loss": 0.7197,
+      "step": 47200
+    },
+    {
+      "epoch": 1.8173243503368623,
+      "grad_norm": 1.1760082244873047,
+      "learning_rate": 4.1221070455208735e-06,
+      "loss": 0.8961,
+      "step": 47205
+    },
+    {
+      "epoch": 1.817516843118383,
+      "grad_norm": 1.3913214206695557,
+      "learning_rate": 4.113518752663914e-06,
+      "loss": 0.787,
+      "step": 47210
+    },
+    {
+      "epoch": 1.8177093358999037,
+      "grad_norm": 1.941649079322815,
+      "learning_rate": 4.104939227990478e-06,
+      "loss": 0.918,
+      "step": 47215
+    },
+    {
+      "epoch": 1.8179018286814244,
+      "grad_norm": 0.8367112278938293,
+      "learning_rate": 4.096368472285106e-06,
+      "loss": 0.636,
+      "step": 47220
+    },
+    {
+      "epoch": 1.818094321462945,
+      "grad_norm": 1.0153110027313232,
+      "learning_rate": 4.0878064863315356e-06,
+      "loss": 0.8449,
+      "step": 47225
+    },
+    {
+      "epoch": 1.8182868142444657,
+      "grad_norm": 1.1490223407745361,
+      "learning_rate": 4.079253270912687e-06,
+      "loss": 0.7839,
+      "step": 47230
+    },
+    {
+      "epoch": 1.8184793070259864,
+      "grad_norm": 1.2465404272079468,
+      "learning_rate": 4.070708826810721e-06,
+      "loss": 0.6734,
+      "step": 47235
+    },
+    {
+      "epoch": 1.8186717998075073,
+      "grad_norm": 1.5955170392990112,
+      "learning_rate": 4.06217315480697e-06,
+      "loss": 0.868,
+      "step": 47240
+    },
+    {
+      "epoch": 1.818864292589028,
+      "grad_norm": 2.5519890785217285,
+      "learning_rate": 4.05364625568192e-06,
+      "loss": 0.6769,
+      "step": 47245
+    },
+    {
+      "epoch": 1.8190567853705486,
+      "grad_norm": 0.968748152256012,
+      "learning_rate": 4.04512813021537e-06,
+      "loss": 0.7728,
+      "step": 47250
+    },
+    {
+      "epoch": 1.8192492781520693,
+      "grad_norm": 1.381900668144226,
+      "learning_rate": 4.0366187791861855e-06,
+      "loss": 0.6947,
+      "step": 47255
+    },
+    {
+      "epoch": 1.81944177093359,
+      "grad_norm": 0.9397858381271362,
+      "learning_rate": 4.028118203372521e-06,
+      "loss": 0.7413,
+      "step": 47260
+    },
+    {
+      "epoch": 1.8196342637151108,
+      "grad_norm": 1.2155348062515259,
+      "learning_rate": 4.019626403551691e-06,
+      "loss": 0.8101,
+      "step": 47265
+    },
+    {
+      "epoch": 1.8198267564966315,
+      "grad_norm": 1.4671454429626465,
+      "learning_rate": 4.011143380500226e-06,
+      "loss": 0.6367,
+      "step": 47270
+    },
+    {
+      "epoch": 1.8200192492781522,
+      "grad_norm": 1.2641961574554443,
+      "learning_rate": 4.002669134993808e-06,
+      "loss": 0.8871,
+      "step": 47275
+    },
+    {
+      "epoch": 1.8202117420596728,
+      "grad_norm": 1.482218861579895,
+      "learning_rate": 3.994203667807395e-06,
+      "loss": 0.7126,
+      "step": 47280
+    },
+    {
+      "epoch": 1.8204042348411935,
+      "grad_norm": 1.0232921838760376,
+      "learning_rate": 3.985746979715066e-06,
+      "loss": 0.7237,
+      "step": 47285
+    },
+    {
+      "epoch": 1.8205967276227142,
+      "grad_norm": 2.7515757083892822,
+      "learning_rate": 3.977299071490148e-06,
+      "loss": 0.7837,
+      "step": 47290
+    },
+    {
+      "epoch": 1.8207892204042349,
+      "grad_norm": 1.110560417175293,
+      "learning_rate": 3.968859943905124e-06,
+      "loss": 0.7886,
+      "step": 47295
+    },
+    {
+      "epoch": 1.8209817131857555,
+      "grad_norm": 1.8675193786621094,
+      "learning_rate": 3.96042959773173e-06,
+      "loss": 0.6506,
+      "step": 47300
+    },
+    {
+      "epoch": 1.8211742059672762,
+      "grad_norm": 1.233778953552246,
+      "learning_rate": 3.952008033740817e-06,
+      "loss": 0.7465,
+      "step": 47305
+    },
+    {
+      "epoch": 1.8213666987487969,
+      "grad_norm": 1.6183221340179443,
+      "learning_rate": 3.943595252702537e-06,
+      "loss": 0.7255,
+      "step": 47310
+    },
+    {
+      "epoch": 1.8215591915303175,
+      "grad_norm": 1.736594796180725,
+      "learning_rate": 3.935191255386139e-06,
+      "loss": 0.6987,
+      "step": 47315
+    },
+    {
+      "epoch": 1.8217516843118382,
+      "grad_norm": 1.2381728887557983,
+      "learning_rate": 3.926796042560133e-06,
+      "loss": 0.9369,
+      "step": 47320
+    },
+    {
+      "epoch": 1.8219441770933589,
+      "grad_norm": 0.9554552435874939,
+      "learning_rate": 3.9184096149922025e-06,
+      "loss": 0.8767,
+      "step": 47325
+    },
+    {
+      "epoch": 1.8221366698748795,
+      "grad_norm": 1.7278703451156616,
+      "learning_rate": 3.9100319734492465e-06,
+      "loss": 0.8719,
+      "step": 47330
+    },
+    {
+      "epoch": 1.8223291626564004,
+      "grad_norm": 0.6862698197364807,
+      "learning_rate": 3.901663118697308e-06,
+      "loss": 0.8551,
+      "step": 47335
+    },
+    {
+      "epoch": 1.822521655437921,
+      "grad_norm": 1.7460417747497559,
+      "learning_rate": 3.893303051501685e-06,
+      "loss": 0.7529,
+      "step": 47340
+    },
+    {
+      "epoch": 1.8227141482194418,
+      "grad_norm": 1.0978227853775024,
+      "learning_rate": 3.884951772626854e-06,
+      "loss": 0.6831,
+      "step": 47345
+    },
+    {
+      "epoch": 1.8229066410009624,
+      "grad_norm": 1.4301860332489014,
+      "learning_rate": 3.8766092828364856e-06,
+      "loss": 0.7509,
+      "step": 47350
+    },
+    {
+      "epoch": 1.8230991337824831,
+      "grad_norm": 1.623776912689209,
+      "learning_rate": 3.868275582893444e-06,
+      "loss": 0.891,
+      "step": 47355
+    },
+    {
+      "epoch": 1.823291626564004,
+      "grad_norm": 0.9064698219299316,
+      "learning_rate": 3.859950673559765e-06,
+      "loss": 0.923,
+      "step": 47360
+    },
+    {
+      "epoch": 1.8234841193455247,
+      "grad_norm": 1.351654052734375,
+      "learning_rate": 3.8516345555967396e-06,
+      "loss": 0.8076,
+      "step": 47365
+    },
+    {
+      "epoch": 1.8236766121270453,
+      "grad_norm": 1.3697307109832764,
+      "learning_rate": 3.8433272297648035e-06,
+      "loss": 0.8001,
+      "step": 47370
+    },
+    {
+      "epoch": 1.823869104908566,
+      "grad_norm": 1.3666245937347412,
+      "learning_rate": 3.835028696823628e-06,
+      "loss": 0.7855,
+      "step": 47375
+    },
+    {
+      "epoch": 1.8240615976900867,
+      "grad_norm": 1.0923709869384766,
+      "learning_rate": 3.826738957532017e-06,
+      "loss": 0.8134,
+      "step": 47380
+    },
+    {
+      "epoch": 1.8242540904716074,
+      "grad_norm": 1.4689079523086548,
+      "learning_rate": 3.818458012648063e-06,
+      "loss": 0.8976,
+      "step": 47385
+    },
+    {
+      "epoch": 1.824446583253128,
+      "grad_norm": 0.9157153367996216,
+      "learning_rate": 3.810185862928972e-06,
+      "loss": 0.7001,
+      "step": 47390
+    },
+    {
+      "epoch": 1.8246390760346487,
+      "grad_norm": 1.6523871421813965,
+      "learning_rate": 3.801922509131184e-06,
+      "loss": 0.7818,
+      "step": 47395
+    },
+    {
+      "epoch": 1.8248315688161694,
+      "grad_norm": 1.4929468631744385,
+      "learning_rate": 3.793667952010327e-06,
+      "loss": 0.7585,
+      "step": 47400
+    },
+    {
+      "epoch": 1.82502406159769,
+      "grad_norm": 0.9369055032730103,
+      "learning_rate": 3.7854221923212307e-06,
+      "loss": 0.6929,
+      "step": 47405
+    },
+    {
+      "epoch": 1.8252165543792107,
+      "grad_norm": 2.0922114849090576,
+      "learning_rate": 3.7771852308178925e-06,
+      "loss": 0.8816,
+      "step": 47410
+    },
+    {
+      "epoch": 1.8254090471607314,
+      "grad_norm": 1.3968013525009155,
+      "learning_rate": 3.7689570682535758e-06,
+      "loss": 0.8213,
+      "step": 47415
+    },
+    {
+      "epoch": 1.825601539942252,
+      "grad_norm": 1.271930456161499,
+      "learning_rate": 3.7607377053806567e-06,
+      "loss": 0.7507,
+      "step": 47420
+    },
+    {
+      "epoch": 1.8257940327237727,
+      "grad_norm": 1.283036231994629,
+      "learning_rate": 3.752527142950735e-06,
+      "loss": 0.7738,
+      "step": 47425
+    },
+    {
+      "epoch": 1.8259865255052934,
+      "grad_norm": 1.001495122909546,
+      "learning_rate": 3.7443253817146306e-06,
+      "loss": 0.8228,
+      "step": 47430
+    },
+    {
+      "epoch": 1.8261790182868143,
+      "grad_norm": 1.5130256414413452,
+      "learning_rate": 3.736132422422345e-06,
+      "loss": 0.8288,
+      "step": 47435
+    },
+    {
+      "epoch": 1.826371511068335,
+      "grad_norm": 0.9629761576652527,
+      "learning_rate": 3.7279482658230445e-06,
+      "loss": 0.819,
+      "step": 47440
+    },
+    {
+      "epoch": 1.8265640038498556,
+      "grad_norm": 1.0789672136306763,
+      "learning_rate": 3.7197729126651317e-06,
+      "loss": 0.8386,
+      "step": 47445
+    },
+    {
+      "epoch": 1.8267564966313763,
+      "grad_norm": 1.1328613758087158,
+      "learning_rate": 3.7116063636961964e-06,
+      "loss": 0.916,
+      "step": 47450
+    },
+    {
+      "epoch": 1.826948989412897,
+      "grad_norm": 0.8800186514854431,
+      "learning_rate": 3.703448619662997e-06,
+      "loss": 0.7408,
+      "step": 47455
+    },
+    {
+      "epoch": 1.8271414821944179,
+      "grad_norm": 1.6841456890106201,
+      "learning_rate": 3.695299681311537e-06,
+      "loss": 0.8142,
+      "step": 47460
+    },
+    {
+      "epoch": 1.8273339749759385,
+      "grad_norm": 1.3652926683425903,
+      "learning_rate": 3.6871595493869316e-06,
+      "loss": 0.8027,
+      "step": 47465
+    },
+    {
+      "epoch": 1.8275264677574592,
+      "grad_norm": 1.3112471103668213,
+      "learning_rate": 3.679028224633596e-06,
+      "loss": 0.9375,
+      "step": 47470
+    },
+    {
+      "epoch": 1.8277189605389799,
+      "grad_norm": 1.2834017276763916,
+      "learning_rate": 3.670905707795047e-06,
+      "loss": 0.8115,
+      "step": 47475
+    },
+    {
+      "epoch": 1.8279114533205005,
+      "grad_norm": 1.075247049331665,
+      "learning_rate": 3.6627919996140457e-06,
+      "loss": 0.8848,
+      "step": 47480
+    },
+    {
+      "epoch": 1.8281039461020212,
+      "grad_norm": 1.0291975736618042,
+      "learning_rate": 3.6546871008325433e-06,
+      "loss": 0.9769,
+      "step": 47485
+    },
+    {
+      "epoch": 1.8282964388835419,
+      "grad_norm": 1.429732084274292,
+      "learning_rate": 3.646591012191691e-06,
+      "loss": 0.6725,
+      "step": 47490
+    },
+    {
+      "epoch": 1.8284889316650625,
+      "grad_norm": 1.1338505744934082,
+      "learning_rate": 3.6385037344317862e-06,
+      "loss": 0.8288,
+      "step": 47495
+    },
+    {
+      "epoch": 1.8286814244465832,
+      "grad_norm": 1.1439357995986938,
+      "learning_rate": 3.6304252682924036e-06,
+      "loss": 0.8643,
+      "step": 47500
+    },
+    {
+      "epoch": 1.8288739172281039,
+      "grad_norm": 2.053274631500244,
+      "learning_rate": 3.6223556145122186e-06,
+      "loss": 0.9137,
+      "step": 47505
+    },
+    {
+      "epoch": 1.8290664100096246,
+      "grad_norm": 0.9340258836746216,
+      "learning_rate": 3.6142947738291744e-06,
+      "loss": 0.702,
+      "step": 47510
+    },
+    {
+      "epoch": 1.8292589027911452,
+      "grad_norm": 2.1745176315307617,
+      "learning_rate": 3.6062427469803705e-06,
+      "loss": 0.8425,
+      "step": 47515
+    },
+    {
+      "epoch": 1.829451395572666,
+      "grad_norm": 1.1325429677963257,
+      "learning_rate": 3.5981995347021403e-06,
+      "loss": 0.8262,
+      "step": 47520
+    },
+    {
+      "epoch": 1.8296438883541866,
+      "grad_norm": 1.0720462799072266,
+      "learning_rate": 3.5901651377299285e-06,
+      "loss": 0.8925,
+      "step": 47525
+    },
+    {
+      "epoch": 1.8298363811357075,
+      "grad_norm": 0.8257774114608765,
+      "learning_rate": 3.5821395567984805e-06,
+      "loss": 0.7577,
+      "step": 47530
+    },
+    {
+      "epoch": 1.8300288739172281,
+      "grad_norm": 1.3935045003890991,
+      "learning_rate": 3.5741227926416545e-06,
+      "loss": 0.7907,
+      "step": 47535
+    },
+    {
+      "epoch": 1.8302213666987488,
+      "grad_norm": 1.0141493082046509,
+      "learning_rate": 3.56611484599253e-06,
+      "loss": 0.6785,
+      "step": 47540
+    },
+    {
+      "epoch": 1.8304138594802695,
+      "grad_norm": 1.2392221689224243,
+      "learning_rate": 3.558115717583388e-06,
+      "loss": 0.9234,
+      "step": 47545
+    },
+    {
+      "epoch": 1.8306063522617901,
+      "grad_norm": 1.5585402250289917,
+      "learning_rate": 3.5501254081457104e-06,
+      "loss": 0.9621,
+      "step": 47550
+    },
+    {
+      "epoch": 1.830798845043311,
+      "grad_norm": 1.6715527772903442,
+      "learning_rate": 3.5421439184101234e-06,
+      "loss": 0.869,
+      "step": 47555
+    },
+    {
+      "epoch": 1.8309913378248317,
+      "grad_norm": 0.8488251566886902,
+      "learning_rate": 3.5341712491065103e-06,
+      "loss": 0.8197,
+      "step": 47560
+    },
+    {
+      "epoch": 1.8311838306063524,
+      "grad_norm": 2.0500576496124268,
+      "learning_rate": 3.52620740096391e-06,
+      "loss": 0.9029,
+      "step": 47565
+    },
+    {
+      "epoch": 1.831376323387873,
+      "grad_norm": 2.1723897457122803,
+      "learning_rate": 3.518252374710551e-06,
+      "loss": 0.7979,
+      "step": 47570
+    },
+    {
+      "epoch": 1.8315688161693937,
+      "grad_norm": 0.6698965430259705,
+      "learning_rate": 3.510306171073896e-06,
+      "loss": 0.8421,
+      "step": 47575
+    },
+    {
+      "epoch": 1.8317613089509144,
+      "grad_norm": 1.0468194484710693,
+      "learning_rate": 3.5023687907805415e-06,
+      "loss": 0.7916,
+      "step": 47580
+    },
+    {
+      "epoch": 1.831953801732435,
+      "grad_norm": 1.206809163093567,
+      "learning_rate": 3.4944402345563177e-06,
+      "loss": 0.7774,
+      "step": 47585
+    },
+    {
+      "epoch": 1.8321462945139557,
+      "grad_norm": 1.2247503995895386,
+      "learning_rate": 3.486520503126256e-06,
+      "loss": 0.8677,
+      "step": 47590
+    },
+    {
+      "epoch": 1.8323387872954764,
+      "grad_norm": 1.510204553604126,
+      "learning_rate": 3.4786095972145547e-06,
+      "loss": 0.8027,
+      "step": 47595
+    },
+    {
+      "epoch": 1.832531280076997,
+      "grad_norm": 0.9858173131942749,
+      "learning_rate": 3.4707075175445915e-06,
+      "loss": 0.7336,
+      "step": 47600
+    },
+    {
+      "epoch": 1.8327237728585177,
+      "grad_norm": 1.1785179376602173,
+      "learning_rate": 3.462814264838998e-06,
+      "loss": 0.7863,
+      "step": 47605
+    },
+    {
+      "epoch": 1.8329162656400384,
+      "grad_norm": 1.1377897262573242,
+      "learning_rate": 3.4549298398195316e-06,
+      "loss": 0.7096,
+      "step": 47610
+    },
+    {
+      "epoch": 1.833108758421559,
+      "grad_norm": 1.5963387489318848,
+      "learning_rate": 3.4470542432071704e-06,
+      "loss": 0.8752,
+      "step": 47615
+    },
+    {
+      "epoch": 1.8333012512030797,
+      "grad_norm": 1.3239431381225586,
+      "learning_rate": 3.4391874757221054e-06,
+      "loss": 0.7784,
+      "step": 47620
+    },
+    {
+      "epoch": 1.8334937439846006,
+      "grad_norm": 1.5421568155288696,
+      "learning_rate": 3.431329538083694e-06,
+      "loss": 0.7401,
+      "step": 47625
+    },
+    {
+      "epoch": 1.8336862367661213,
+      "grad_norm": 1.7763359546661377,
+      "learning_rate": 3.423480431010462e-06,
+      "loss": 0.739,
+      "step": 47630
+    },
+    {
+      "epoch": 1.833878729547642,
+      "grad_norm": 1.8040746450424194,
+      "learning_rate": 3.415640155220212e-06,
+      "loss": 0.848,
+      "step": 47635
+    },
+    {
+      "epoch": 1.8340712223291626,
+      "grad_norm": 1.0281445980072021,
+      "learning_rate": 3.4078087114298495e-06,
+      "loss": 0.7374,
+      "step": 47640
+    },
+    {
+      "epoch": 1.8342637151106833,
+      "grad_norm": 1.061551570892334,
+      "learning_rate": 3.399986100355501e-06,
+      "loss": 0.7567,
+      "step": 47645
+    },
+    {
+      "epoch": 1.8344562078922042,
+      "grad_norm": 0.8794161677360535,
+      "learning_rate": 3.392172322712517e-06,
+      "loss": 0.7726,
+      "step": 47650
+    },
+    {
+      "epoch": 1.8346487006737249,
+      "grad_norm": 2.0824060440063477,
+      "learning_rate": 3.384367379215425e-06,
+      "loss": 0.9414,
+      "step": 47655
+    },
+    {
+      "epoch": 1.8348411934552455,
+      "grad_norm": 1.3683054447174072,
+      "learning_rate": 3.3765712705778884e-06,
+      "loss": 0.7369,
+      "step": 47660
+    },
+    {
+      "epoch": 1.8350336862367662,
+      "grad_norm": 1.2808324098587036,
+      "learning_rate": 3.3687839975128477e-06,
+      "loss": 0.7476,
+      "step": 47665
+    },
+    {
+      "epoch": 1.8352261790182869,
+      "grad_norm": 1.3409979343414307,
+      "learning_rate": 3.3610055607323887e-06,
+      "loss": 0.7112,
+      "step": 47670
+    },
+    {
+      "epoch": 1.8354186717998076,
+      "grad_norm": 1.0510163307189941,
+      "learning_rate": 3.353235960947787e-06,
+      "loss": 0.7489,
+      "step": 47675
+    },
+    {
+      "epoch": 1.8356111645813282,
+      "grad_norm": 0.9734552502632141,
+      "learning_rate": 3.345475198869552e-06,
+      "loss": 0.6948,
+      "step": 47680
+    },
+    {
+      "epoch": 1.835803657362849,
+      "grad_norm": 0.9429619908332825,
+      "learning_rate": 3.337723275207316e-06,
+      "loss": 0.5987,
+      "step": 47685
+    },
+    {
+      "epoch": 1.8359961501443696,
+      "grad_norm": 1.4983386993408203,
+      "learning_rate": 3.3299801906699567e-06,
+      "loss": 0.809,
+      "step": 47690
+    },
+    {
+      "epoch": 1.8361886429258902,
+      "grad_norm": 2.2176480293273926,
+      "learning_rate": 3.3222459459655297e-06,
+      "loss": 0.8457,
+      "step": 47695
+    },
+    {
+      "epoch": 1.836381135707411,
+      "grad_norm": 0.8564404249191284,
+      "learning_rate": 3.3145205418012915e-06,
+      "loss": 0.8729,
+      "step": 47700
+    },
+    {
+      "epoch": 1.8365736284889316,
+      "grad_norm": 1.057340383529663,
+      "learning_rate": 3.3068039788836435e-06,
+      "loss": 0.7122,
+      "step": 47705
+    },
+    {
+      "epoch": 1.8367661212704522,
+      "grad_norm": 1.1386967897415161,
+      "learning_rate": 3.299096257918255e-06,
+      "loss": 0.7144,
+      "step": 47710
+    },
+    {
+      "epoch": 1.836958614051973,
+      "grad_norm": 1.4860408306121826,
+      "learning_rate": 3.2913973796099174e-06,
+      "loss": 0.6879,
+      "step": 47715
+    },
+    {
+      "epoch": 1.8371511068334936,
+      "grad_norm": 1.474790096282959,
+      "learning_rate": 3.2837073446626677e-06,
+      "loss": 0.9557,
+      "step": 47720
+    },
+    {
+      "epoch": 1.8373435996150145,
+      "grad_norm": 1.0030909776687622,
+      "learning_rate": 3.276026153779688e-06,
+      "loss": 0.7611,
+      "step": 47725
+    },
+    {
+      "epoch": 1.8375360923965351,
+      "grad_norm": 1.4009109735488892,
+      "learning_rate": 3.2683538076633714e-06,
+      "loss": 0.7646,
+      "step": 47730
+    },
+    {
+      "epoch": 1.8377285851780558,
+      "grad_norm": 1.6726889610290527,
+      "learning_rate": 3.2606903070153127e-06,
+      "loss": 0.8261,
+      "step": 47735
+    },
+    {
+      "epoch": 1.8379210779595765,
+      "grad_norm": 1.5140515565872192,
+      "learning_rate": 3.253035652536307e-06,
+      "loss": 0.8136,
+      "step": 47740
+    },
+    {
+      "epoch": 1.8381135707410972,
+      "grad_norm": 1.0216679573059082,
+      "learning_rate": 3.2453898449262834e-06,
+      "loss": 0.8699,
+      "step": 47745
+    },
+    {
+      "epoch": 1.838306063522618,
+      "grad_norm": 0.9776056408882141,
+      "learning_rate": 3.2377528848844154e-06,
+      "loss": 0.6758,
+      "step": 47750
+    },
+    {
+      "epoch": 1.8384985563041387,
+      "grad_norm": 1.2438172101974487,
+      "learning_rate": 3.2301247731090557e-06,
+      "loss": 0.848,
+      "step": 47755
+    },
+    {
+      "epoch": 1.8386910490856594,
+      "grad_norm": 1.5708495378494263,
+      "learning_rate": 3.2225055102977464e-06,
+      "loss": 0.7504,
+      "step": 47760
+    },
+    {
+      "epoch": 1.83888354186718,
+      "grad_norm": 1.5115718841552734,
+      "learning_rate": 3.2148950971472302e-06,
+      "loss": 0.8011,
+      "step": 47765
+    },
+    {
+      "epoch": 1.8390760346487007,
+      "grad_norm": 1.2615517377853394,
+      "learning_rate": 3.207293534353395e-06,
+      "loss": 0.8359,
+      "step": 47770
+    },
+    {
+      "epoch": 1.8392685274302214,
+      "grad_norm": 1.3792976140975952,
+      "learning_rate": 3.1997008226113734e-06,
+      "loss": 0.8824,
+      "step": 47775
+    },
+    {
+      "epoch": 1.839461020211742,
+      "grad_norm": 0.8459932804107666,
+      "learning_rate": 3.192116962615477e-06,
+      "loss": 0.7326,
+      "step": 47780
+    },
+    {
+      "epoch": 1.8396535129932627,
+      "grad_norm": 1.6874253749847412,
+      "learning_rate": 3.184541955059195e-06,
+      "loss": 0.9437,
+      "step": 47785
+    },
+    {
+      "epoch": 1.8398460057747834,
+      "grad_norm": 1.0827080011367798,
+      "learning_rate": 3.1769758006351846e-06,
+      "loss": 0.7969,
+      "step": 47790
+    },
+    {
+      "epoch": 1.840038498556304,
+      "grad_norm": 1.4325987100601196,
+      "learning_rate": 3.1694185000353703e-06,
+      "loss": 0.6621,
+      "step": 47795
+    },
+    {
+      "epoch": 1.8402309913378248,
+      "grad_norm": 0.8930256366729736,
+      "learning_rate": 3.1618700539507774e-06,
+      "loss": 0.8178,
+      "step": 47800
+    },
+    {
+      "epoch": 1.8404234841193454,
+      "grad_norm": 1.8248862028121948,
+      "learning_rate": 3.154330463071675e-06,
+      "loss": 0.6502,
+      "step": 47805
+    },
+    {
+      "epoch": 1.840615976900866,
+      "grad_norm": 1.318647027015686,
+      "learning_rate": 3.146799728087513e-06,
+      "loss": 0.7669,
+      "step": 47810
+    },
+    {
+      "epoch": 1.8408084696823868,
+      "grad_norm": 1.4266585111618042,
+      "learning_rate": 3.139277849686928e-06,
+      "loss": 0.8331,
+      "step": 47815
+    },
+    {
+      "epoch": 1.8410009624639077,
+      "grad_norm": 1.6008580923080444,
+      "learning_rate": 3.131764828557715e-06,
+      "loss": 0.7556,
+      "step": 47820
+    },
+    {
+      "epoch": 1.8411934552454283,
+      "grad_norm": 1.1239802837371826,
+      "learning_rate": 3.1242606653869355e-06,
+      "loss": 0.7312,
+      "step": 47825
+    },
+    {
+      "epoch": 1.841385948026949,
+      "grad_norm": 1.142177939414978,
+      "learning_rate": 3.116765360860774e-06,
+      "loss": 0.7444,
+      "step": 47830
+    },
+    {
+      "epoch": 1.8415784408084697,
+      "grad_norm": 1.0553315877914429,
+      "learning_rate": 3.109278915664615e-06,
+      "loss": 0.8287,
+      "step": 47835
+    },
+    {
+      "epoch": 1.8417709335899903,
+      "grad_norm": 1.4490617513656616,
+      "learning_rate": 3.101801330483067e-06,
+      "loss": 0.9347,
+      "step": 47840
+    },
+    {
+      "epoch": 1.8419634263715112,
+      "grad_norm": 2.255176305770874,
+      "learning_rate": 3.0943326059999056e-06,
+      "loss": 0.8066,
+      "step": 47845
+    },
+    {
+      "epoch": 1.842155919153032,
+      "grad_norm": 1.7092933654785156,
+      "learning_rate": 3.0868727428980617e-06,
+      "loss": 0.9151,
+      "step": 47850
+    },
+    {
+      "epoch": 1.8423484119345526,
+      "grad_norm": 1.135002613067627,
+      "learning_rate": 3.079421741859734e-06,
+      "loss": 0.873,
+      "step": 47855
+    },
+    {
+      "epoch": 1.8425409047160732,
+      "grad_norm": 1.5216232538223267,
+      "learning_rate": 3.071979603566233e-06,
+      "loss": 0.9477,
+      "step": 47860
+    },
+    {
+      "epoch": 1.842733397497594,
+      "grad_norm": 1.8921546936035156,
+      "learning_rate": 3.0645463286981148e-06,
+      "loss": 0.7256,
+      "step": 47865
+    },
+    {
+      "epoch": 1.8429258902791146,
+      "grad_norm": 1.7546651363372803,
+      "learning_rate": 3.0571219179351016e-06,
+      "loss": 0.7983,
+      "step": 47870
+    },
+    {
+      "epoch": 1.8431183830606352,
+      "grad_norm": 1.1996418237686157,
+      "learning_rate": 3.0497063719561068e-06,
+      "loss": 0.7587,
+      "step": 47875
+    },
+    {
+      "epoch": 1.843310875842156,
+      "grad_norm": 1.1013127565383911,
+      "learning_rate": 3.0422996914392098e-06,
+      "loss": 0.7332,
+      "step": 47880
+    },
+    {
+      "epoch": 1.8435033686236766,
+      "grad_norm": 1.216601014137268,
+      "learning_rate": 3.0349018770617354e-06,
+      "loss": 0.8414,
+      "step": 47885
+    },
+    {
+      "epoch": 1.8436958614051973,
+      "grad_norm": 0.8273448944091797,
+      "learning_rate": 3.0275129295001315e-06,
+      "loss": 0.7199,
+      "step": 47890
+    },
+    {
+      "epoch": 1.843888354186718,
+      "grad_norm": 1.1838254928588867,
+      "learning_rate": 3.020132849430102e-06,
+      "loss": 0.7555,
+      "step": 47895
+    },
+    {
+      "epoch": 1.8440808469682386,
+      "grad_norm": 0.9097992181777954,
+      "learning_rate": 3.0127616375264956e-06,
+      "loss": 0.7774,
+      "step": 47900
+    },
+    {
+      "epoch": 1.8442733397497593,
+      "grad_norm": 3.2226083278656006,
+      "learning_rate": 3.0053992944633404e-06,
+      "loss": 0.8008,
+      "step": 47905
+    },
+    {
+      "epoch": 1.84446583253128,
+      "grad_norm": 1.555508017539978,
+      "learning_rate": 2.998045820913886e-06,
+      "loss": 0.814,
+      "step": 47910
+    },
+    {
+      "epoch": 1.8446583253128006,
+      "grad_norm": 1.5878740549087524,
+      "learning_rate": 2.990701217550573e-06,
+      "loss": 0.7585,
+      "step": 47915
+    },
+    {
+      "epoch": 1.8448508180943215,
+      "grad_norm": 1.0371932983398438,
+      "learning_rate": 2.983365485045009e-06,
+      "loss": 0.8261,
+      "step": 47920
+    },
+    {
+      "epoch": 1.8450433108758422,
+      "grad_norm": 0.9782282114028931,
+      "learning_rate": 2.976038624067978e-06,
+      "loss": 0.6692,
+      "step": 47925
+    },
+    {
+      "epoch": 1.8452358036573628,
+      "grad_norm": 1.5847156047821045,
+      "learning_rate": 2.9687206352895125e-06,
+      "loss": 0.846,
+      "step": 47930
+    },
+    {
+      "epoch": 1.8454282964388835,
+      "grad_norm": 0.9324734807014465,
+      "learning_rate": 2.961411519378754e-06,
+      "loss": 0.8455,
+      "step": 47935
+    },
+    {
+      "epoch": 1.8456207892204044,
+      "grad_norm": 1.0669317245483398,
+      "learning_rate": 2.9541112770041013e-06,
+      "loss": 0.944,
+      "step": 47940
+    },
+    {
+      "epoch": 1.845813282001925,
+      "grad_norm": 1.0065456628799438,
+      "learning_rate": 2.9468199088330985e-06,
+      "loss": 0.7346,
+      "step": 47945
+    },
+    {
+      "epoch": 1.8460057747834457,
+      "grad_norm": 1.4775437116622925,
+      "learning_rate": 2.9395374155325007e-06,
+      "loss": 0.6656,
+      "step": 47950
+    },
+    {
+      "epoch": 1.8461982675649664,
+      "grad_norm": 0.8717533349990845,
+      "learning_rate": 2.93226379776822e-06,
+      "loss": 0.696,
+      "step": 47955
+    },
+    {
+      "epoch": 1.846390760346487,
+      "grad_norm": 0.9198839664459229,
+      "learning_rate": 2.924999056205424e-06,
+      "loss": 0.8188,
+      "step": 47960
+    },
+    {
+      "epoch": 1.8465832531280078,
+      "grad_norm": 1.2768006324768066,
+      "learning_rate": 2.917743191508393e-06,
+      "loss": 0.8166,
+      "step": 47965
+    },
+    {
+      "epoch": 1.8467757459095284,
+      "grad_norm": 1.1566869020462036,
+      "learning_rate": 2.9104962043406293e-06,
+      "loss": 0.7985,
+      "step": 47970
+    },
+    {
+      "epoch": 1.846968238691049,
+      "grad_norm": 1.2734941244125366,
+      "learning_rate": 2.9032580953648357e-06,
+      "loss": 0.8511,
+      "step": 47975
+    },
+    {
+      "epoch": 1.8471607314725698,
+      "grad_norm": 0.7918460369110107,
+      "learning_rate": 2.8960288652428726e-06,
+      "loss": 0.7173,
+      "step": 47980
+    },
+    {
+      "epoch": 1.8473532242540904,
+      "grad_norm": 1.3652775287628174,
+      "learning_rate": 2.8888085146358324e-06,
+      "loss": 0.7498,
+      "step": 47985
+    },
+    {
+      "epoch": 1.847545717035611,
+      "grad_norm": 1.2592366933822632,
+      "learning_rate": 2.881597044203943e-06,
+      "loss": 0.8583,
+      "step": 47990
+    },
+    {
+      "epoch": 1.8477382098171318,
+      "grad_norm": 2.0500288009643555,
+      "learning_rate": 2.8743944546066437e-06,
+      "loss": 0.854,
+      "step": 47995
+    },
+    {
+      "epoch": 1.8479307025986524,
+      "grad_norm": 1.140693187713623,
+      "learning_rate": 2.867200746502585e-06,
+      "loss": 0.6055,
+      "step": 48000
+    },
+    {
+      "epoch": 1.8481231953801731,
+      "grad_norm": 1.287409782409668,
+      "learning_rate": 2.8600159205495748e-06,
+      "loss": 0.8655,
+      "step": 48005
+    },
+    {
+      "epoch": 1.8483156881616938,
+      "grad_norm": 1.279222846031189,
+      "learning_rate": 2.8528399774045977e-06,
+      "loss": 0.7086,
+      "step": 48010
+    },
+    {
+      "epoch": 1.8485081809432147,
+      "grad_norm": 1.2571754455566406,
+      "learning_rate": 2.845672917723885e-06,
+      "loss": 0.8281,
+      "step": 48015
+    },
+    {
+      "epoch": 1.8487006737247353,
+      "grad_norm": 1.407504916191101,
+      "learning_rate": 2.838514742162779e-06,
+      "loss": 0.8478,
+      "step": 48020
+    },
+    {
+      "epoch": 1.848893166506256,
+      "grad_norm": 1.116768479347229,
+      "learning_rate": 2.831365451375867e-06,
+      "loss": 0.7778,
+      "step": 48025
+    },
+    {
+      "epoch": 1.8490856592877767,
+      "grad_norm": 1.208387017250061,
+      "learning_rate": 2.824225046016904e-06,
+      "loss": 0.7465,
+      "step": 48030
+    },
+    {
+      "epoch": 1.8492781520692974,
+      "grad_norm": 1.3610413074493408,
+      "learning_rate": 2.8170935267388343e-06,
+      "loss": 0.7879,
+      "step": 48035
+    },
+    {
+      "epoch": 1.8494706448508182,
+      "grad_norm": 0.9448495507240295,
+      "learning_rate": 2.8099708941937697e-06,
+      "loss": 0.7607,
+      "step": 48040
+    },
+    {
+      "epoch": 1.849663137632339,
+      "grad_norm": 1.0731843709945679,
+      "learning_rate": 2.8028571490330556e-06,
+      "loss": 0.7323,
+      "step": 48045
+    },
+    {
+      "epoch": 1.8498556304138596,
+      "grad_norm": 1.2336262464523315,
+      "learning_rate": 2.795752291907183e-06,
+      "loss": 0.827,
+      "step": 48050
+    },
+    {
+      "epoch": 1.8500481231953803,
+      "grad_norm": 1.0452556610107422,
+      "learning_rate": 2.7886563234658327e-06,
+      "loss": 0.7965,
+      "step": 48055
+    },
+    {
+      "epoch": 1.850240615976901,
+      "grad_norm": 1.5202009677886963,
+      "learning_rate": 2.7815692443579066e-06,
+      "loss": 0.7943,
+      "step": 48060
+    },
+    {
+      "epoch": 1.8504331087584216,
+      "grad_norm": 1.1691776514053345,
+      "learning_rate": 2.774491055231465e-06,
+      "loss": 0.7282,
+      "step": 48065
+    },
+    {
+      "epoch": 1.8506256015399423,
+      "grad_norm": 2.0975561141967773,
+      "learning_rate": 2.7674217567337348e-06,
+      "loss": 0.8538,
+      "step": 48070
+    },
+    {
+      "epoch": 1.850818094321463,
+      "grad_norm": 1.0445414781570435,
+      "learning_rate": 2.760361349511198e-06,
+      "loss": 0.9576,
+      "step": 48075
+    },
+    {
+      "epoch": 1.8510105871029836,
+      "grad_norm": 1.9726266860961914,
+      "learning_rate": 2.75330983420945e-06,
+      "loss": 0.8439,
+      "step": 48080
+    },
+    {
+      "epoch": 1.8512030798845043,
+      "grad_norm": 1.502845048904419,
+      "learning_rate": 2.746267211473319e-06,
+      "loss": 0.9805,
+      "step": 48085
+    },
+    {
+      "epoch": 1.851395572666025,
+      "grad_norm": 1.754075050354004,
+      "learning_rate": 2.7392334819468123e-06,
+      "loss": 0.9215,
+      "step": 48090
+    },
+    {
+      "epoch": 1.8515880654475456,
+      "grad_norm": 0.9861617088317871,
+      "learning_rate": 2.7322086462731157e-06,
+      "loss": 0.9203,
+      "step": 48095
+    },
+    {
+      "epoch": 1.8517805582290663,
+      "grad_norm": 1.416224718093872,
+      "learning_rate": 2.7251927050945813e-06,
+      "loss": 0.834,
+      "step": 48100
+    },
+    {
+      "epoch": 1.851973051010587,
+      "grad_norm": 1.3885353803634644,
+      "learning_rate": 2.7181856590527967e-06,
+      "loss": 0.8002,
+      "step": 48105
+    },
+    {
+      "epoch": 1.8521655437921078,
+      "grad_norm": 1.1199766397476196,
+      "learning_rate": 2.7111875087885042e-06,
+      "loss": 0.8423,
+      "step": 48110
+    },
+    {
+      "epoch": 1.8523580365736285,
+      "grad_norm": 0.9421118497848511,
+      "learning_rate": 2.7041982549416144e-06,
+      "loss": 0.7188,
+      "step": 48115
+    },
+    {
+      "epoch": 1.8525505293551492,
+      "grad_norm": 2.011782646179199,
+      "learning_rate": 2.697217898151294e-06,
+      "loss": 0.8969,
+      "step": 48120
+    },
+    {
+      "epoch": 1.8527430221366699,
+      "grad_norm": 1.3184289932250977,
+      "learning_rate": 2.6902464390558103e-06,
+      "loss": 0.6961,
+      "step": 48125
+    },
+    {
+      "epoch": 1.8529355149181905,
+      "grad_norm": 1.5041390657424927,
+      "learning_rate": 2.683283878292675e-06,
+      "loss": 0.7917,
+      "step": 48130
+    },
+    {
+      "epoch": 1.8531280076997114,
+      "grad_norm": 1.7774921655654907,
+      "learning_rate": 2.6763302164985573e-06,
+      "loss": 1.0091,
+      "step": 48135
+    },
+    {
+      "epoch": 1.853320500481232,
+      "grad_norm": 1.642179250717163,
+      "learning_rate": 2.6693854543093476e-06,
+      "loss": 0.7543,
+      "step": 48140
+    },
+    {
+      "epoch": 1.8535129932627528,
+      "grad_norm": 1.1686973571777344,
+      "learning_rate": 2.662449592360061e-06,
+      "loss": 0.7695,
+      "step": 48145
+    },
+    {
+      "epoch": 1.8537054860442734,
+      "grad_norm": 0.9292591214179993,
+      "learning_rate": 2.655522631284979e-06,
+      "loss": 0.7245,
+      "step": 48150
+    },
+    {
+      "epoch": 1.853897978825794,
+      "grad_norm": 2.467902660369873,
+      "learning_rate": 2.6486045717174836e-06,
+      "loss": 0.8604,
+      "step": 48155
+    },
+    {
+      "epoch": 1.8540904716073148,
+      "grad_norm": 1.306139349937439,
+      "learning_rate": 2.641695414290224e-06,
+      "loss": 0.7734,
+      "step": 48160
+    },
+    {
+      "epoch": 1.8542829643888354,
+      "grad_norm": 1.376107096672058,
+      "learning_rate": 2.634795159634962e-06,
+      "loss": 0.6813,
+      "step": 48165
+    },
+    {
+      "epoch": 1.8544754571703561,
+      "grad_norm": 1.3315315246582031,
+      "learning_rate": 2.6279038083827146e-06,
+      "loss": 0.7047,
+      "step": 48170
+    },
+    {
+      "epoch": 1.8546679499518768,
+      "grad_norm": 1.155645489692688,
+      "learning_rate": 2.6210213611636115e-06,
+      "loss": 0.7626,
+      "step": 48175
+    },
+    {
+      "epoch": 1.8548604427333975,
+      "grad_norm": 1.1847624778747559,
+      "learning_rate": 2.6141478186070487e-06,
+      "loss": 0.8735,
+      "step": 48180
+    },
+    {
+      "epoch": 1.8550529355149181,
+      "grad_norm": 1.0727204084396362,
+      "learning_rate": 2.6072831813415354e-06,
+      "loss": 0.7699,
+      "step": 48185
+    },
+    {
+      "epoch": 1.8552454282964388,
+      "grad_norm": 1.2296266555786133,
+      "learning_rate": 2.600427449994813e-06,
+      "loss": 0.5567,
+      "step": 48190
+    },
+    {
+      "epoch": 1.8554379210779595,
+      "grad_norm": 1.0693732500076294,
+      "learning_rate": 2.593580625193781e-06,
+      "loss": 0.6918,
+      "step": 48195
+    },
+    {
+      "epoch": 1.8556304138594801,
+      "grad_norm": 1.5725680589675903,
+      "learning_rate": 2.58674270756456e-06,
+      "loss": 0.835,
+      "step": 48200
+    },
+    {
+      "epoch": 1.8558229066410008,
+      "grad_norm": 1.0935879945755005,
+      "learning_rate": 2.5799136977323948e-06,
+      "loss": 0.822,
+      "step": 48205
+    },
+    {
+      "epoch": 1.8560153994225217,
+      "grad_norm": 0.9157683849334717,
+      "learning_rate": 2.573093596321774e-06,
+      "loss": 0.8057,
+      "step": 48210
+    },
+    {
+      "epoch": 1.8562078922040424,
+      "grad_norm": 1.6122345924377441,
+      "learning_rate": 2.566282403956355e-06,
+      "loss": 0.9039,
+      "step": 48215
+    },
+    {
+      "epoch": 1.856400384985563,
+      "grad_norm": 1.6490685939788818,
+      "learning_rate": 2.5594801212589613e-06,
+      "loss": 0.9311,
+      "step": 48220
+    },
+    {
+      "epoch": 1.8565928777670837,
+      "grad_norm": 1.5705087184906006,
+      "learning_rate": 2.5526867488516513e-06,
+      "loss": 0.6325,
+      "step": 48225
+    },
+    {
+      "epoch": 1.8567853705486044,
+      "grad_norm": 1.2620983123779297,
+      "learning_rate": 2.5459022873555726e-06,
+      "loss": 0.7455,
+      "step": 48230
+    },
+    {
+      "epoch": 1.8569778633301253,
+      "grad_norm": 1.1116305589675903,
+      "learning_rate": 2.5391267373911842e-06,
+      "loss": 0.65,
+      "step": 48235
+    },
+    {
+      "epoch": 1.857170356111646,
+      "grad_norm": 1.5795679092407227,
+      "learning_rate": 2.532360099578024e-06,
+      "loss": 0.8738,
+      "step": 48240
+    },
+    {
+      "epoch": 1.8573628488931666,
+      "grad_norm": 1.2260464429855347,
+      "learning_rate": 2.5256023745348746e-06,
+      "loss": 0.9088,
+      "step": 48245
+    },
+    {
+      "epoch": 1.8575553416746873,
+      "grad_norm": 1.582663893699646,
+      "learning_rate": 2.518853562879675e-06,
+      "loss": 0.7772,
+      "step": 48250
+    },
+    {
+      "epoch": 1.857747834456208,
+      "grad_norm": 1.497031807899475,
+      "learning_rate": 2.5121136652295764e-06,
+      "loss": 0.8043,
+      "step": 48255
+    },
+    {
+      "epoch": 1.8579403272377286,
+      "grad_norm": 1.0930918455123901,
+      "learning_rate": 2.505382682200863e-06,
+      "loss": 0.7846,
+      "step": 48260
+    },
+    {
+      "epoch": 1.8581328200192493,
+      "grad_norm": 1.2151498794555664,
+      "learning_rate": 2.4986606144090762e-06,
+      "loss": 0.7133,
+      "step": 48265
+    },
+    {
+      "epoch": 1.85832531280077,
+      "grad_norm": 1.189026951789856,
+      "learning_rate": 2.49194746246888e-06,
+      "loss": 0.7892,
+      "step": 48270
+    },
+    {
+      "epoch": 1.8585178055822906,
+      "grad_norm": 1.6892609596252441,
+      "learning_rate": 2.4852432269941607e-06,
+      "loss": 0.8151,
+      "step": 48275
+    },
+    {
+      "epoch": 1.8587102983638113,
+      "grad_norm": 1.126964807510376,
+      "learning_rate": 2.4785479085979724e-06,
+      "loss": 1.0332,
+      "step": 48280
+    },
+    {
+      "epoch": 1.858902791145332,
+      "grad_norm": 1.3323931694030762,
+      "learning_rate": 2.471861507892559e-06,
+      "loss": 0.7541,
+      "step": 48285
+    },
+    {
+      "epoch": 1.8590952839268526,
+      "grad_norm": 1.3149107694625854,
+      "learning_rate": 2.465184025489331e-06,
+      "loss": 0.7269,
+      "step": 48290
+    },
+    {
+      "epoch": 1.8592877767083733,
+      "grad_norm": 1.9822214841842651,
+      "learning_rate": 2.458515461998945e-06,
+      "loss": 0.7885,
+      "step": 48295
+    },
+    {
+      "epoch": 1.859480269489894,
+      "grad_norm": 1.3917293548583984,
+      "learning_rate": 2.4518558180311456e-06,
+      "loss": 0.7677,
+      "step": 48300
+    },
+    {
+      "epoch": 1.8596727622714149,
+      "grad_norm": 0.9052494764328003,
+      "learning_rate": 2.4452050941949357e-06,
+      "loss": 0.7665,
+      "step": 48305
+    },
+    {
+      "epoch": 1.8598652550529355,
+      "grad_norm": 1.0716874599456787,
+      "learning_rate": 2.4385632910984834e-06,
+      "loss": 0.6622,
+      "step": 48310
+    },
+    {
+      "epoch": 1.8600577478344562,
+      "grad_norm": 1.5617578029632568,
+      "learning_rate": 2.431930409349137e-06,
+      "loss": 0.7643,
+      "step": 48315
+    },
+    {
+      "epoch": 1.8602502406159769,
+      "grad_norm": 1.0250791311264038,
+      "learning_rate": 2.4253064495534106e-06,
+      "loss": 0.8006,
+      "step": 48320
+    },
+    {
+      "epoch": 1.8604427333974976,
+      "grad_norm": 1.2734397649765015,
+      "learning_rate": 2.4186914123170423e-06,
+      "loss": 0.6411,
+      "step": 48325
+    },
+    {
+      "epoch": 1.8606352261790184,
+      "grad_norm": 0.9825404286384583,
+      "learning_rate": 2.412085298244937e-06,
+      "loss": 0.7262,
+      "step": 48330
+    },
+    {
+      "epoch": 1.8608277189605391,
+      "grad_norm": 1.2708152532577515,
+      "learning_rate": 2.4054881079411564e-06,
+      "loss": 0.7606,
+      "step": 48335
+    },
+    {
+      "epoch": 1.8610202117420598,
+      "grad_norm": 1.1390200853347778,
+      "learning_rate": 2.3988998420089947e-06,
+      "loss": 0.734,
+      "step": 48340
+    },
+    {
+      "epoch": 1.8612127045235805,
+      "grad_norm": 1.1848487854003906,
+      "learning_rate": 2.3923205010508932e-06,
+      "loss": 0.742,
+      "step": 48345
+    },
+    {
+      "epoch": 1.8614051973051011,
+      "grad_norm": 1.2743182182312012,
+      "learning_rate": 2.385750085668481e-06,
+      "loss": 0.7798,
+      "step": 48350
+    },
+    {
+      "epoch": 1.8615976900866218,
+      "grad_norm": 0.963897168636322,
+      "learning_rate": 2.3791885964625894e-06,
+      "loss": 0.8682,
+      "step": 48355
+    },
+    {
+      "epoch": 1.8617901828681425,
+      "grad_norm": 1.480070948600769,
+      "learning_rate": 2.3726360340332376e-06,
+      "loss": 0.7644,
+      "step": 48360
+    },
+    {
+      "epoch": 1.8619826756496631,
+      "grad_norm": 0.9536148905754089,
+      "learning_rate": 2.3660923989795803e-06,
+      "loss": 0.8048,
+      "step": 48365
+    },
+    {
+      "epoch": 1.8621751684311838,
+      "grad_norm": 1.9504202604293823,
+      "learning_rate": 2.3595576919000163e-06,
+      "loss": 0.8524,
+      "step": 48370
+    },
+    {
+      "epoch": 1.8623676612127045,
+      "grad_norm": 0.9313097596168518,
+      "learning_rate": 2.3530319133920896e-06,
+      "loss": 0.8765,
+      "step": 48375
+    },
+    {
+      "epoch": 1.8625601539942251,
+      "grad_norm": 1.1792408227920532,
+      "learning_rate": 2.3465150640525456e-06,
+      "loss": 0.7123,
+      "step": 48380
+    },
+    {
+      "epoch": 1.8627526467757458,
+      "grad_norm": 0.8612557649612427,
+      "learning_rate": 2.3400071444772964e-06,
+      "loss": 0.6412,
+      "step": 48385
+    },
+    {
+      "epoch": 1.8629451395572665,
+      "grad_norm": 1.3216156959533691,
+      "learning_rate": 2.3335081552614768e-06,
+      "loss": 0.7946,
+      "step": 48390
+    },
+    {
+      "epoch": 1.8631376323387872,
+      "grad_norm": 0.8052036166191101,
+      "learning_rate": 2.3270180969993226e-06,
+      "loss": 0.8188,
+      "step": 48395
+    },
+    {
+      "epoch": 1.8633301251203078,
+      "grad_norm": 1.8680044412612915,
+      "learning_rate": 2.3205369702843703e-06,
+      "loss": 0.767,
+      "step": 48400
+    },
+    {
+      "epoch": 1.8635226179018287,
+      "grad_norm": 0.8687813878059387,
+      "learning_rate": 2.314064775709224e-06,
+      "loss": 0.751,
+      "step": 48405
+    },
+    {
+      "epoch": 1.8637151106833494,
+      "grad_norm": 1.280096173286438,
+      "learning_rate": 2.3076015138657537e-06,
+      "loss": 0.8564,
+      "step": 48410
+    },
+    {
+      "epoch": 1.86390760346487,
+      "grad_norm": 0.4581018090248108,
+      "learning_rate": 2.301147185344965e-06,
+      "loss": 0.6742,
+      "step": 48415
+    },
+    {
+      "epoch": 1.8641000962463907,
+      "grad_norm": 1.88551926612854,
+      "learning_rate": 2.294701790737086e-06,
+      "loss": 0.7482,
+      "step": 48420
+    },
+    {
+      "epoch": 1.8642925890279116,
+      "grad_norm": 1.5057963132858276,
+      "learning_rate": 2.2882653306314673e-06,
+      "loss": 0.8031,
+      "step": 48425
+    },
+    {
+      "epoch": 1.8644850818094323,
+      "grad_norm": 1.2725811004638672,
+      "learning_rate": 2.2818378056167155e-06,
+      "loss": 0.7703,
+      "step": 48430
+    },
+    {
+      "epoch": 1.864677574590953,
+      "grad_norm": 1.0294675827026367,
+      "learning_rate": 2.275419216280572e-06,
+      "loss": 0.8374,
+      "step": 48435
+    },
+    {
+      "epoch": 1.8648700673724736,
+      "grad_norm": 1.523382306098938,
+      "learning_rate": 2.2690095632099785e-06,
+      "loss": 0.9034,
+      "step": 48440
+    },
+    {
+      "epoch": 1.8650625601539943,
+      "grad_norm": 1.2364825010299683,
+      "learning_rate": 2.2626088469910547e-06,
+      "loss": 0.6834,
+      "step": 48445
+    },
+    {
+      "epoch": 1.865255052935515,
+      "grad_norm": 1.2433017492294312,
+      "learning_rate": 2.2562170682090877e-06,
+      "loss": 0.7689,
+      "step": 48450
+    },
+    {
+      "epoch": 1.8654475457170356,
+      "grad_norm": 1.3385852575302124,
+      "learning_rate": 2.2498342274485774e-06,
+      "loss": 0.8025,
+      "step": 48455
+    },
+    {
+      "epoch": 1.8656400384985563,
+      "grad_norm": 2.2688310146331787,
+      "learning_rate": 2.2434603252932006e-06,
+      "loss": 0.8375,
+      "step": 48460
+    },
+    {
+      "epoch": 1.865832531280077,
+      "grad_norm": 1.4485090970993042,
+      "learning_rate": 2.237095362325803e-06,
+      "loss": 0.9097,
+      "step": 48465
+    },
+    {
+      "epoch": 1.8660250240615976,
+      "grad_norm": 1.457838773727417,
+      "learning_rate": 2.230739339128396e-06,
+      "loss": 0.7724,
+      "step": 48470
+    },
+    {
+      "epoch": 1.8662175168431183,
+      "grad_norm": 1.2256953716278076,
+      "learning_rate": 2.2243922562822374e-06,
+      "loss": 0.769,
+      "step": 48475
+    },
+    {
+      "epoch": 1.866410009624639,
+      "grad_norm": 1.5888760089874268,
+      "learning_rate": 2.218054114367685e-06,
+      "loss": 0.7524,
+      "step": 48480
+    },
+    {
+      "epoch": 1.8666025024061597,
+      "grad_norm": 1.6216797828674316,
+      "learning_rate": 2.2117249139643415e-06,
+      "loss": 0.8443,
+      "step": 48485
+    },
+    {
+      "epoch": 1.8667949951876803,
+      "grad_norm": 0.9898841381072998,
+      "learning_rate": 2.2054046556509666e-06,
+      "loss": 0.7376,
+      "step": 48490
+    },
+    {
+      "epoch": 1.866987487969201,
+      "grad_norm": 1.0620770454406738,
+      "learning_rate": 2.1990933400055093e-06,
+      "loss": 0.8391,
+      "step": 48495
+    },
+    {
+      "epoch": 1.867179980750722,
+      "grad_norm": 0.9419720768928528,
+      "learning_rate": 2.192790967605085e-06,
+      "loss": 0.7434,
+      "step": 48500
+    },
+    {
+      "epoch": 1.8673724735322426,
+      "grad_norm": 2.1842215061187744,
+      "learning_rate": 2.1864975390260334e-06,
+      "loss": 0.9107,
+      "step": 48505
+    },
+    {
+      "epoch": 1.8675649663137632,
+      "grad_norm": 2.277456283569336,
+      "learning_rate": 2.180213054843816e-06,
+      "loss": 0.8018,
+      "step": 48510
+    },
+    {
+      "epoch": 1.867757459095284,
+      "grad_norm": 1.2959650754928589,
+      "learning_rate": 2.1739375156331176e-06,
+      "loss": 0.6821,
+      "step": 48515
+    },
+    {
+      "epoch": 1.8679499518768046,
+      "grad_norm": 1.474094271659851,
+      "learning_rate": 2.1676709219677905e-06,
+      "loss": 0.7479,
+      "step": 48520
+    },
+    {
+      "epoch": 1.8681424446583255,
+      "grad_norm": 3.1948869228363037,
+      "learning_rate": 2.161413274420876e-06,
+      "loss": 1.0174,
+      "step": 48525
+    },
+    {
+      "epoch": 1.8683349374398461,
+      "grad_norm": 1.1724430322647095,
+      "learning_rate": 2.1551645735646053e-06,
+      "loss": 0.8481,
+      "step": 48530
+    },
+    {
+      "epoch": 1.8685274302213668,
+      "grad_norm": 1.5295108556747437,
+      "learning_rate": 2.1489248199703773e-06,
+      "loss": 0.7603,
+      "step": 48535
+    },
+    {
+      "epoch": 1.8687199230028875,
+      "grad_norm": 1.8773775100708008,
+      "learning_rate": 2.142694014208757e-06,
+      "loss": 0.8019,
+      "step": 48540
+    },
+    {
+      "epoch": 1.8689124157844081,
+      "grad_norm": 1.0877622365951538,
+      "learning_rate": 2.136472156849523e-06,
+      "loss": 0.8444,
+      "step": 48545
+    },
+    {
+      "epoch": 1.8691049085659288,
+      "grad_norm": 1.389959692955017,
+      "learning_rate": 2.130259248461641e-06,
+      "loss": 0.9192,
+      "step": 48550
+    },
+    {
+      "epoch": 1.8692974013474495,
+      "grad_norm": 0.84064120054245,
+      "learning_rate": 2.1240552896131918e-06,
+      "loss": 0.7529,
+      "step": 48555
+    },
+    {
+      "epoch": 1.8694898941289702,
+      "grad_norm": 0.8932579159736633,
+      "learning_rate": 2.117860280871542e-06,
+      "loss": 0.7907,
+      "step": 48560
+    },
+    {
+      "epoch": 1.8696823869104908,
+      "grad_norm": 1.7808706760406494,
+      "learning_rate": 2.1116742228031616e-06,
+      "loss": 0.8738,
+      "step": 48565
+    },
+    {
+      "epoch": 1.8698748796920115,
+      "grad_norm": 1.6530510187149048,
+      "learning_rate": 2.105497115973709e-06,
+      "loss": 0.8155,
+      "step": 48570
+    },
+    {
+      "epoch": 1.8700673724735322,
+      "grad_norm": 1.3979318141937256,
+      "learning_rate": 2.0993289609480547e-06,
+      "loss": 0.8067,
+      "step": 48575
+    },
+    {
+      "epoch": 1.8702598652550528,
+      "grad_norm": 1.452545404434204,
+      "learning_rate": 2.0931697582902476e-06,
+      "loss": 0.8966,
+      "step": 48580
+    },
+    {
+      "epoch": 1.8704523580365735,
+      "grad_norm": 1.182625651359558,
+      "learning_rate": 2.087019508563481e-06,
+      "loss": 0.7396,
+      "step": 48585
+    },
+    {
+      "epoch": 1.8706448508180942,
+      "grad_norm": 1.7071011066436768,
+      "learning_rate": 2.080878212330173e-06,
+      "loss": 0.773,
+      "step": 48590
+    },
+    {
+      "epoch": 1.870837343599615,
+      "grad_norm": 0.9710677862167358,
+      "learning_rate": 2.074745870151895e-06,
+      "loss": 0.762,
+      "step": 48595
+    },
+    {
+      "epoch": 1.8710298363811357,
+      "grad_norm": 1.9075227975845337,
+      "learning_rate": 2.068622482589411e-06,
+      "loss": 0.7325,
+      "step": 48600
+    },
+    {
+      "epoch": 1.8712223291626564,
+      "grad_norm": 1.70927894115448,
+      "learning_rate": 2.0625080502026606e-06,
+      "loss": 0.7394,
+      "step": 48605
+    },
+    {
+      "epoch": 1.871414821944177,
+      "grad_norm": 1.2238922119140625,
+      "learning_rate": 2.0564025735507864e-06,
+      "loss": 0.8871,
+      "step": 48610
+    },
+    {
+      "epoch": 1.8716073147256977,
+      "grad_norm": 1.3306856155395508,
+      "learning_rate": 2.0503060531920528e-06,
+      "loss": 0.7314,
+      "step": 48615
+    },
+    {
+      "epoch": 1.8717998075072186,
+      "grad_norm": 1.1816201210021973,
+      "learning_rate": 2.0442184896840023e-06,
+      "loss": 0.7342,
+      "step": 48620
+    },
+    {
+      "epoch": 1.8719923002887393,
+      "grad_norm": 1.3129255771636963,
+      "learning_rate": 2.0381398835832567e-06,
+      "loss": 0.6599,
+      "step": 48625
+    },
+    {
+      "epoch": 1.87218479307026,
+      "grad_norm": 2.268023729324341,
+      "learning_rate": 2.032070235445682e-06,
+      "loss": 0.9488,
+      "step": 48630
+    },
+    {
+      "epoch": 1.8723772858517806,
+      "grad_norm": 1.5048736333847046,
+      "learning_rate": 2.0260095458263128e-06,
+      "loss": 0.7311,
+      "step": 48635
+    },
+    {
+      "epoch": 1.8725697786333013,
+      "grad_norm": 2.065556287765503,
+      "learning_rate": 2.019957815279361e-06,
+      "loss": 0.9459,
+      "step": 48640
+    },
+    {
+      "epoch": 1.872762271414822,
+      "grad_norm": 1.1562378406524658,
+      "learning_rate": 2.0139150443581944e-06,
+      "loss": 0.8719,
+      "step": 48645
+    },
+    {
+      "epoch": 1.8729547641963427,
+      "grad_norm": 1.01041579246521,
+      "learning_rate": 2.007881233615394e-06,
+      "loss": 0.8623,
+      "step": 48650
+    },
+    {
+      "epoch": 1.8731472569778633,
+      "grad_norm": 1.1639820337295532,
+      "learning_rate": 2.0018563836027293e-06,
+      "loss": 0.7525,
+      "step": 48655
+    },
+    {
+      "epoch": 1.873339749759384,
+      "grad_norm": 1.1873496770858765,
+      "learning_rate": 1.995840494871104e-06,
+      "loss": 0.7482,
+      "step": 48660
+    },
+    {
+      "epoch": 1.8735322425409047,
+      "grad_norm": 1.128056526184082,
+      "learning_rate": 1.989833567970667e-06,
+      "loss": 0.8526,
+      "step": 48665
+    },
+    {
+      "epoch": 1.8737247353224253,
+      "grad_norm": 2.4238860607147217,
+      "learning_rate": 1.98383560345069e-06,
+      "loss": 0.848,
+      "step": 48670
+    },
+    {
+      "epoch": 1.873917228103946,
+      "grad_norm": 2.036635160446167,
+      "learning_rate": 1.977846601859634e-06,
+      "loss": 0.8757,
+      "step": 48675
+    },
+    {
+      "epoch": 1.8741097208854667,
+      "grad_norm": 1.680403709411621,
+      "learning_rate": 1.9718665637451726e-06,
+      "loss": 0.9098,
+      "step": 48680
+    },
+    {
+      "epoch": 1.8743022136669873,
+      "grad_norm": 0.7588726878166199,
+      "learning_rate": 1.965895489654157e-06,
+      "loss": 0.7572,
+      "step": 48685
+    },
+    {
+      "epoch": 1.874494706448508,
+      "grad_norm": 1.095284104347229,
+      "learning_rate": 1.9599333801325505e-06,
+      "loss": 0.8009,
+      "step": 48690
+    },
+    {
+      "epoch": 1.874687199230029,
+      "grad_norm": 1.7757399082183838,
+      "learning_rate": 1.9539802357256055e-06,
+      "loss": 0.7788,
+      "step": 48695
+    },
+    {
+      "epoch": 1.8748796920115496,
+      "grad_norm": 1.0243725776672363,
+      "learning_rate": 1.9480360569776647e-06,
+      "loss": 0.8988,
+      "step": 48700
+    },
+    {
+      "epoch": 1.8750721847930703,
+      "grad_norm": 1.7366422414779663,
+      "learning_rate": 1.942100844432293e-06,
+      "loss": 0.7865,
+      "step": 48705
+    },
+    {
+      "epoch": 1.875264677574591,
+      "grad_norm": 1.935095191001892,
+      "learning_rate": 1.936174598632212e-06,
+      "loss": 0.9178,
+      "step": 48710
+    },
+    {
+      "epoch": 1.8754571703561116,
+      "grad_norm": 0.9639614224433899,
+      "learning_rate": 1.9302573201193776e-06,
+      "loss": 0.6945,
+      "step": 48715
+    },
+    {
+      "epoch": 1.8756496631376325,
+      "grad_norm": 1.3461066484451294,
+      "learning_rate": 1.924349009434834e-06,
+      "loss": 0.7487,
+      "step": 48720
+    },
+    {
+      "epoch": 1.8758421559191532,
+      "grad_norm": 1.1384443044662476,
+      "learning_rate": 1.9184496671188933e-06,
+      "loss": 0.7095,
+      "step": 48725
+    },
+    {
+      "epoch": 1.8760346487006738,
+      "grad_norm": 0.8311898112297058,
+      "learning_rate": 1.9125592937109916e-06,
+      "loss": 0.7814,
+      "step": 48730
+    },
+    {
+      "epoch": 1.8762271414821945,
+      "grad_norm": 1.4966212511062622,
+      "learning_rate": 1.906677889749775e-06,
+      "loss": 0.9245,
+      "step": 48735
+    },
+    {
+      "epoch": 1.8764196342637152,
+      "grad_norm": 1.414469599723816,
+      "learning_rate": 1.900805455773058e-06,
+      "loss": 0.8376,
+      "step": 48740
+    },
+    {
+      "epoch": 1.8766121270452358,
+      "grad_norm": 1.3187638521194458,
+      "learning_rate": 1.8949419923178336e-06,
+      "loss": 0.7,
+      "step": 48745
+    },
+    {
+      "epoch": 1.8768046198267565,
+      "grad_norm": 1.3405956029891968,
+      "learning_rate": 1.8890874999202946e-06,
+      "loss": 0.9212,
+      "step": 48750
+    },
+    {
+      "epoch": 1.8769971126082772,
+      "grad_norm": 1.8978711366653442,
+      "learning_rate": 1.8832419791157574e-06,
+      "loss": 0.8214,
+      "step": 48755
+    },
+    {
+      "epoch": 1.8771896053897978,
+      "grad_norm": 1.8074771165847778,
+      "learning_rate": 1.8774054304387834e-06,
+      "loss": 0.8833,
+      "step": 48760
+    },
+    {
+      "epoch": 1.8773820981713185,
+      "grad_norm": 1.301705241203308,
+      "learning_rate": 1.871577854423079e-06,
+      "loss": 0.7134,
+      "step": 48765
+    },
+    {
+      "epoch": 1.8775745909528392,
+      "grad_norm": 0.8135727047920227,
+      "learning_rate": 1.8657592516015398e-06,
+      "loss": 0.6364,
+      "step": 48770
+    },
+    {
+      "epoch": 1.8777670837343599,
+      "grad_norm": 0.7817474007606506,
+      "learning_rate": 1.8599496225062296e-06,
+      "loss": 0.6671,
+      "step": 48775
+    },
+    {
+      "epoch": 1.8779595765158805,
+      "grad_norm": 1.2370343208312988,
+      "learning_rate": 1.8541489676684232e-06,
+      "loss": 0.8199,
+      "step": 48780
+    },
+    {
+      "epoch": 1.8781520692974012,
+      "grad_norm": 1.585356593132019,
+      "learning_rate": 1.8483572876185296e-06,
+      "loss": 0.9021,
+      "step": 48785
+    },
+    {
+      "epoch": 1.878344562078922,
+      "grad_norm": 1.4099061489105225,
+      "learning_rate": 1.8425745828861585e-06,
+      "loss": 0.7381,
+      "step": 48790
+    },
+    {
+      "epoch": 1.8785370548604428,
+      "grad_norm": 2.376617670059204,
+      "learning_rate": 1.8368008540001203e-06,
+      "loss": 0.7274,
+      "step": 48795
+    },
+    {
+      "epoch": 1.8787295476419634,
+      "grad_norm": 2.1324985027313232,
+      "learning_rate": 1.8310361014883703e-06,
+      "loss": 0.953,
+      "step": 48800
+    },
+    {
+      "epoch": 1.878922040423484,
+      "grad_norm": 2.0138044357299805,
+      "learning_rate": 1.8252803258780538e-06,
+      "loss": 0.8029,
+      "step": 48805
+    },
+    {
+      "epoch": 1.8791145332050048,
+      "grad_norm": 1.857802152633667,
+      "learning_rate": 1.8195335276955162e-06,
+      "loss": 0.8503,
+      "step": 48810
+    },
+    {
+      "epoch": 1.8793070259865257,
+      "grad_norm": 1.9087456464767456,
+      "learning_rate": 1.813795707466237e-06,
+      "loss": 0.8598,
+      "step": 48815
+    },
+    {
+      "epoch": 1.8794995187680463,
+      "grad_norm": 1.3068690299987793,
+      "learning_rate": 1.8080668657149192e-06,
+      "loss": 0.6669,
+      "step": 48820
+    },
+    {
+      "epoch": 1.879692011549567,
+      "grad_norm": 1.2544077634811401,
+      "learning_rate": 1.802347002965421e-06,
+      "loss": 0.851,
+      "step": 48825
+    },
+    {
+      "epoch": 1.8798845043310877,
+      "grad_norm": 1.29012930393219,
+      "learning_rate": 1.7966361197408022e-06,
+      "loss": 0.8332,
+      "step": 48830
+    },
+    {
+      "epoch": 1.8800769971126083,
+      "grad_norm": 1.0903306007385254,
+      "learning_rate": 1.7909342165632558e-06,
+      "loss": 0.6718,
+      "step": 48835
+    },
+    {
+      "epoch": 1.880269489894129,
+      "grad_norm": 1.1692110300064087,
+      "learning_rate": 1.7852412939542208e-06,
+      "loss": 0.8295,
+      "step": 48840
+    },
+    {
+      "epoch": 1.8804619826756497,
+      "grad_norm": 1.5184317827224731,
+      "learning_rate": 1.7795573524342356e-06,
+      "loss": 0.7913,
+      "step": 48845
+    },
+    {
+      "epoch": 1.8806544754571703,
+      "grad_norm": 1.50426185131073,
+      "learning_rate": 1.7738823925230964e-06,
+      "loss": 0.8514,
+      "step": 48850
+    },
+    {
+      "epoch": 1.880846968238691,
+      "grad_norm": 1.971408486366272,
+      "learning_rate": 1.7682164147397208e-06,
+      "loss": 0.8384,
+      "step": 48855
+    },
+    {
+      "epoch": 1.8810394610202117,
+      "grad_norm": 0.9322729706764221,
+      "learning_rate": 1.7625594196022166e-06,
+      "loss": 0.7074,
+      "step": 48860
+    },
+    {
+      "epoch": 1.8812319538017324,
+      "grad_norm": 1.4018663167953491,
+      "learning_rate": 1.7569114076278924e-06,
+      "loss": 0.7446,
+      "step": 48865
+    },
+    {
+      "epoch": 1.881424446583253,
+      "grad_norm": 1.2637794017791748,
+      "learning_rate": 1.7512723793332242e-06,
+      "loss": 0.7912,
+      "step": 48870
+    },
+    {
+      "epoch": 1.8816169393647737,
+      "grad_norm": 0.7229753732681274,
+      "learning_rate": 1.7456423352338658e-06,
+      "loss": 0.7339,
+      "step": 48875
+    },
+    {
+      "epoch": 1.8818094321462944,
+      "grad_norm": 0.9690874814987183,
+      "learning_rate": 1.7400212758446276e-06,
+      "loss": 0.727,
+      "step": 48880
+    },
+    {
+      "epoch": 1.8820019249278153,
+      "grad_norm": 1.093895673751831,
+      "learning_rate": 1.7344092016795433e-06,
+      "loss": 0.7939,
+      "step": 48885
+    },
+    {
+      "epoch": 1.882194417709336,
+      "grad_norm": 1.0549845695495605,
+      "learning_rate": 1.7288061132517686e-06,
+      "loss": 0.8103,
+      "step": 48890
+    },
+    {
+      "epoch": 1.8823869104908566,
+      "grad_norm": 1.6850600242614746,
+      "learning_rate": 1.723212011073705e-06,
+      "loss": 0.7877,
+      "step": 48895
+    },
+    {
+      "epoch": 1.8825794032723773,
+      "grad_norm": 1.4307023286819458,
+      "learning_rate": 1.7176268956568653e-06,
+      "loss": 0.7448,
+      "step": 48900
+    },
+    {
+      "epoch": 1.882771896053898,
+      "grad_norm": 1.4105719327926636,
+      "learning_rate": 1.7120507675120078e-06,
+      "loss": 0.9085,
+      "step": 48905
+    },
+    {
+      "epoch": 1.8829643888354188,
+      "grad_norm": 1.330667495727539,
+      "learning_rate": 1.7064836271489803e-06,
+      "loss": 0.8304,
+      "step": 48910
+    },
+    {
+      "epoch": 1.8831568816169395,
+      "grad_norm": 1.1804107427597046,
+      "learning_rate": 1.7009254750769088e-06,
+      "loss": 0.8184,
+      "step": 48915
+    },
+    {
+      "epoch": 1.8833493743984602,
+      "grad_norm": 0.7987260818481445,
+      "learning_rate": 1.695376311804031e-06,
+      "loss": 0.6531,
+      "step": 48920
+    },
+    {
+      "epoch": 1.8835418671799808,
+      "grad_norm": 0.7400739789009094,
+      "learning_rate": 1.6898361378377747e-06,
+      "loss": 0.7269,
+      "step": 48925
+    },
+    {
+      "epoch": 1.8837343599615015,
+      "grad_norm": 1.043596863746643,
+      "learning_rate": 1.6843049536847567e-06,
+      "loss": 0.7467,
+      "step": 48930
+    },
+    {
+      "epoch": 1.8839268527430222,
+      "grad_norm": 1.6760101318359375,
+      "learning_rate": 1.6787827598507721e-06,
+      "loss": 0.6619,
+      "step": 48935
+    },
+    {
+      "epoch": 1.8841193455245429,
+      "grad_norm": 1.3724150657653809,
+      "learning_rate": 1.6732695568407842e-06,
+      "loss": 0.7495,
+      "step": 48940
+    },
+    {
+      "epoch": 1.8843118383060635,
+      "grad_norm": 0.6713613867759705,
+      "learning_rate": 1.6677653451589448e-06,
+      "loss": 0.7356,
+      "step": 48945
+    },
+    {
+      "epoch": 1.8845043310875842,
+      "grad_norm": 1.0061167478561401,
+      "learning_rate": 1.6622701253085626e-06,
+      "loss": 0.8709,
+      "step": 48950
+    },
+    {
+      "epoch": 1.8846968238691049,
+      "grad_norm": 1.6838310956954956,
+      "learning_rate": 1.656783897792158e-06,
+      "loss": 0.8505,
+      "step": 48955
+    },
+    {
+      "epoch": 1.8848893166506255,
+      "grad_norm": 1.262903094291687,
+      "learning_rate": 1.651306663111396e-06,
+      "loss": 0.8333,
+      "step": 48960
+    },
+    {
+      "epoch": 1.8850818094321462,
+      "grad_norm": 0.9866868257522583,
+      "learning_rate": 1.645838421767154e-06,
+      "loss": 0.8363,
+      "step": 48965
+    },
+    {
+      "epoch": 1.8852743022136669,
+      "grad_norm": 1.7241532802581787,
+      "learning_rate": 1.6403791742594433e-06,
+      "loss": 0.8477,
+      "step": 48970
+    },
+    {
+      "epoch": 1.8854667949951875,
+      "grad_norm": 1.6335369348526,
+      "learning_rate": 1.6349289210874752e-06,
+      "loss": 0.7239,
+      "step": 48975
+    },
+    {
+      "epoch": 1.8856592877767082,
+      "grad_norm": 2.1081769466400146,
+      "learning_rate": 1.6294876627496624e-06,
+      "loss": 0.6491,
+      "step": 48980
+    },
+    {
+      "epoch": 1.885851780558229,
+      "grad_norm": 1.3898777961730957,
+      "learning_rate": 1.6240553997435403e-06,
+      "loss": 0.7792,
+      "step": 48985
+    },
+    {
+      "epoch": 1.8860442733397498,
+      "grad_norm": 0.9727959632873535,
+      "learning_rate": 1.6186321325658893e-06,
+      "loss": 0.7878,
+      "step": 48990
+    },
+    {
+      "epoch": 1.8862367661212704,
+      "grad_norm": 1.1837464570999146,
+      "learning_rate": 1.6132178617126016e-06,
+      "loss": 1.0175,
+      "step": 48995
+    },
+    {
+      "epoch": 1.8864292589027911,
+      "grad_norm": 2.5681560039520264,
+      "learning_rate": 1.607812587678792e-06,
+      "loss": 0.9884,
+      "step": 49000
+    },
+    {
+      "epoch": 1.8866217516843118,
+      "grad_norm": 0.915593683719635,
+      "learning_rate": 1.602416310958732e-06,
+      "loss": 0.677,
+      "step": 49005
+    },
+    {
+      "epoch": 1.8868142444658327,
+      "grad_norm": 1.0929181575775146,
+      "learning_rate": 1.5970290320458715e-06,
+      "loss": 0.7596,
+      "step": 49010
+    },
+    {
+      "epoch": 1.8870067372473533,
+      "grad_norm": 1.0402964353561401,
+      "learning_rate": 1.5916507514328494e-06,
+      "loss": 0.6634,
+      "step": 49015
+    },
+    {
+      "epoch": 1.887199230028874,
+      "grad_norm": 1.7485926151275635,
+      "learning_rate": 1.5862814696114836e-06,
+      "loss": 0.7347,
+      "step": 49020
+    },
+    {
+      "epoch": 1.8873917228103947,
+      "grad_norm": 2.60465669631958,
+      "learning_rate": 1.5809211870727259e-06,
+      "loss": 0.9895,
+      "step": 49025
+    },
+    {
+      "epoch": 1.8875842155919154,
+      "grad_norm": 1.0043087005615234,
+      "learning_rate": 1.5755699043067728e-06,
+      "loss": 0.7079,
+      "step": 49030
+    },
+    {
+      "epoch": 1.887776708373436,
+      "grad_norm": 0.876469075679779,
+      "learning_rate": 1.5702276218029444e-06,
+      "loss": 0.6647,
+      "step": 49035
+    },
+    {
+      "epoch": 1.8879692011549567,
+      "grad_norm": 1.5398415327072144,
+      "learning_rate": 1.5648943400497495e-06,
+      "loss": 0.8207,
+      "step": 49040
+    },
+    {
+      "epoch": 1.8881616939364774,
+      "grad_norm": 0.8192068934440613,
+      "learning_rate": 1.5595700595349093e-06,
+      "loss": 0.8108,
+      "step": 49045
+    },
+    {
+      "epoch": 1.888354186717998,
+      "grad_norm": 2.2673983573913574,
+      "learning_rate": 1.554254780745279e-06,
+      "loss": 0.8681,
+      "step": 49050
+    },
+    {
+      "epoch": 1.8885466794995187,
+      "grad_norm": 1.8705366849899292,
+      "learning_rate": 1.5489485041669026e-06,
+      "loss": 0.7649,
+      "step": 49055
+    },
+    {
+      "epoch": 1.8887391722810394,
+      "grad_norm": 1.0904033184051514,
+      "learning_rate": 1.5436512302850148e-06,
+      "loss": 0.7311,
+      "step": 49060
+    },
+    {
+      "epoch": 1.88893166506256,
+      "grad_norm": 1.5426052808761597,
+      "learning_rate": 1.5383629595839944e-06,
+      "loss": 0.8157,
+      "step": 49065
+    },
+    {
+      "epoch": 1.8891241578440807,
+      "grad_norm": 1.1552188396453857,
+      "learning_rate": 1.5330836925474434e-06,
+      "loss": 0.8184,
+      "step": 49070
+    },
+    {
+      "epoch": 1.8893166506256014,
+      "grad_norm": 1.0506844520568848,
+      "learning_rate": 1.5278134296580981e-06,
+      "loss": 0.7162,
+      "step": 49075
+    },
+    {
+      "epoch": 1.8895091434071223,
+      "grad_norm": 1.1629083156585693,
+      "learning_rate": 1.5225521713979063e-06,
+      "loss": 0.8053,
+      "step": 49080
+    },
+    {
+      "epoch": 1.889701636188643,
+      "grad_norm": 1.4255765676498413,
+      "learning_rate": 1.5172999182479496e-06,
+      "loss": 0.7563,
+      "step": 49085
+    },
+    {
+      "epoch": 1.8898941289701636,
+      "grad_norm": 1.1012247800827026,
+      "learning_rate": 1.512056670688533e-06,
+      "loss": 0.7282,
+      "step": 49090
+    },
+    {
+      "epoch": 1.8900866217516843,
+      "grad_norm": 1.0495306253433228,
+      "learning_rate": 1.5068224291991174e-06,
+      "loss": 0.9035,
+      "step": 49095
+    },
+    {
+      "epoch": 1.890279114533205,
+      "grad_norm": 1.2971009016036987,
+      "learning_rate": 1.5015971942583196e-06,
+      "loss": 0.7806,
+      "step": 49100
+    },
+    {
+      "epoch": 1.8904716073147259,
+      "grad_norm": 1.4070019721984863,
+      "learning_rate": 1.4963809663439908e-06,
+      "loss": 0.7514,
+      "step": 49105
+    },
+    {
+      "epoch": 1.8906641000962465,
+      "grad_norm": 1.4686723947525024,
+      "learning_rate": 1.4911737459330722e-06,
+      "loss": 0.8591,
+      "step": 49110
+    },
+    {
+      "epoch": 1.8908565928777672,
+      "grad_norm": 1.885849952697754,
+      "learning_rate": 1.48597553350176e-06,
+      "loss": 0.771,
+      "step": 49115
+    },
+    {
+      "epoch": 1.8910490856592879,
+      "grad_norm": 1.6949317455291748,
+      "learning_rate": 1.4807863295253965e-06,
+      "loss": 0.9033,
+      "step": 49120
+    },
+    {
+      "epoch": 1.8912415784408085,
+      "grad_norm": 1.8095852136611938,
+      "learning_rate": 1.4756061344784912e-06,
+      "loss": 0.7384,
+      "step": 49125
+    },
+    {
+      "epoch": 1.8914340712223292,
+      "grad_norm": 1.4184739589691162,
+      "learning_rate": 1.4704349488347313e-06,
+      "loss": 0.8234,
+      "step": 49130
+    },
+    {
+      "epoch": 1.8916265640038499,
+      "grad_norm": 1.2031188011169434,
+      "learning_rate": 1.4652727730670058e-06,
+      "loss": 0.6538,
+      "step": 49135
+    },
+    {
+      "epoch": 1.8918190567853705,
+      "grad_norm": 0.8883029222488403,
+      "learning_rate": 1.4601196076473478e-06,
+      "loss": 0.778,
+      "step": 49140
+    },
+    {
+      "epoch": 1.8920115495668912,
+      "grad_norm": 1.1456712484359741,
+      "learning_rate": 1.4549754530469805e-06,
+      "loss": 0.8069,
+      "step": 49145
+    },
+    {
+      "epoch": 1.8922040423484119,
+      "grad_norm": 1.2302826642990112,
+      "learning_rate": 1.4498403097363167e-06,
+      "loss": 0.6927,
+      "step": 49150
+    },
+    {
+      "epoch": 1.8923965351299326,
+      "grad_norm": 1.2416324615478516,
+      "learning_rate": 1.4457386835319498e-06,
+      "loss": 0.8545,
+      "step": 49155
+    },
+    {
+      "epoch": 1.8925890279114532,
+      "grad_norm": 1.3639918565750122,
+      "learning_rate": 1.4406197617255034e-06,
+      "loss": 0.9177,
+      "step": 49160
+    },
+    {
+      "epoch": 1.892781520692974,
+      "grad_norm": 1.4873005151748657,
+      "learning_rate": 1.4355098525214883e-06,
+      "loss": 0.7928,
+      "step": 49165
+    },
+    {
+      "epoch": 1.8929740134744946,
+      "grad_norm": 0.8439772725105286,
+      "learning_rate": 1.4304089563871525e-06,
+      "loss": 0.9517,
+      "step": 49170
+    },
+    {
+      "epoch": 1.8931665062560152,
+      "grad_norm": 1.5179928541183472,
+      "learning_rate": 1.4253170737889787e-06,
+      "loss": 0.8324,
+      "step": 49175
+    },
+    {
+      "epoch": 1.8933589990375361,
+      "grad_norm": 2.050947904586792,
+      "learning_rate": 1.4202342051925387e-06,
+      "loss": 0.8277,
+      "step": 49180
+    },
+    {
+      "epoch": 1.8935514918190568,
+      "grad_norm": 1.4222488403320312,
+      "learning_rate": 1.415160351062661e-06,
+      "loss": 0.8468,
+      "step": 49185
+    },
+    {
+      "epoch": 1.8937439846005775,
+      "grad_norm": 1.9091302156448364,
+      "learning_rate": 1.4100955118632964e-06,
+      "loss": 0.8484,
+      "step": 49190
+    },
+    {
+      "epoch": 1.8939364773820981,
+      "grad_norm": 1.2363388538360596,
+      "learning_rate": 1.4050396880576189e-06,
+      "loss": 0.7329,
+      "step": 49195
+    },
+    {
+      "epoch": 1.894128970163619,
+      "grad_norm": 1.5010732412338257,
+      "learning_rate": 1.3999928801079033e-06,
+      "loss": 0.7513,
+      "step": 49200
+    },
+    {
+      "epoch": 1.8943214629451397,
+      "grad_norm": 0.9450666904449463,
+      "learning_rate": 1.3949550884756913e-06,
+      "loss": 0.7275,
+      "step": 49205
+    },
+    {
+      "epoch": 1.8945139557266604,
+      "grad_norm": 1.6268372535705566,
+      "learning_rate": 1.3899263136216367e-06,
+      "loss": 0.773,
+      "step": 49210
+    },
+    {
+      "epoch": 1.894706448508181,
+      "grad_norm": 1.2723169326782227,
+      "learning_rate": 1.384906556005583e-06,
+      "loss": 0.7464,
+      "step": 49215
+    },
+    {
+      "epoch": 1.8948989412897017,
+      "grad_norm": 1.1981759071350098,
+      "learning_rate": 1.3798958160865517e-06,
+      "loss": 0.8826,
+      "step": 49220
+    },
+    {
+      "epoch": 1.8950914340712224,
+      "grad_norm": 1.8858412504196167,
+      "learning_rate": 1.374894094322765e-06,
+      "loss": 0.7508,
+      "step": 49225
+    },
+    {
+      "epoch": 1.895283926852743,
+      "grad_norm": 1.544263482093811,
+      "learning_rate": 1.3699013911715685e-06,
+      "loss": 0.7827,
+      "step": 49230
+    },
+    {
+      "epoch": 1.8954764196342637,
+      "grad_norm": 1.369147777557373,
+      "learning_rate": 1.3649177070895414e-06,
+      "loss": 0.8175,
+      "step": 49235
+    },
+    {
+      "epoch": 1.8956689124157844,
+      "grad_norm": 1.682911992073059,
+      "learning_rate": 1.3599430425323856e-06,
+      "loss": 0.6453,
+      "step": 49240
+    },
+    {
+      "epoch": 1.895861405197305,
+      "grad_norm": 1.029921293258667,
+      "learning_rate": 1.354977397955004e-06,
+      "loss": 0.7134,
+      "step": 49245
+    },
+    {
+      "epoch": 1.8960538979788257,
+      "grad_norm": 1.0889908075332642,
+      "learning_rate": 1.3500207738114777e-06,
+      "loss": 0.7324,
+      "step": 49250
+    },
+    {
+      "epoch": 1.8962463907603464,
+      "grad_norm": 1.1377679109573364,
+      "learning_rate": 1.3450731705550557e-06,
+      "loss": 0.7995,
+      "step": 49255
+    },
+    {
+      "epoch": 1.896438883541867,
+      "grad_norm": 1.6833134889602661,
+      "learning_rate": 1.3401345886381644e-06,
+      "loss": 0.6787,
+      "step": 49260
+    },
+    {
+      "epoch": 1.8966313763233877,
+      "grad_norm": 1.080384373664856,
+      "learning_rate": 1.3352050285123986e-06,
+      "loss": 0.7251,
+      "step": 49265
+    },
+    {
+      "epoch": 1.8968238691049084,
+      "grad_norm": 1.5266867876052856,
+      "learning_rate": 1.3302844906285417e-06,
+      "loss": 0.8563,
+      "step": 49270
+    },
+    {
+      "epoch": 1.8970163618864293,
+      "grad_norm": 1.8439712524414062,
+      "learning_rate": 1.325372975436545e-06,
+      "loss": 0.7046,
+      "step": 49275
+    },
+    {
+      "epoch": 1.89720885466795,
+      "grad_norm": 1.3293750286102295,
+      "learning_rate": 1.3204704833855275e-06,
+      "loss": 0.8011,
+      "step": 49280
+    },
+    {
+      "epoch": 1.8974013474494706,
+      "grad_norm": 0.9964110255241394,
+      "learning_rate": 1.3155770149237856e-06,
+      "loss": 0.7107,
+      "step": 49285
+    },
+    {
+      "epoch": 1.8975938402309913,
+      "grad_norm": 1.0830775499343872,
+      "learning_rate": 1.310692570498806e-06,
+      "loss": 0.9186,
+      "step": 49290
+    },
+    {
+      "epoch": 1.897786333012512,
+      "grad_norm": 2.34660267829895,
+      "learning_rate": 1.3058171505572424e-06,
+      "loss": 0.8195,
+      "step": 49295
+    },
+    {
+      "epoch": 1.8979788257940329,
+      "grad_norm": 1.1595919132232666,
+      "learning_rate": 1.300950755544894e-06,
+      "loss": 0.8621,
+      "step": 49300
+    },
+    {
+      "epoch": 1.8981713185755535,
+      "grad_norm": 1.1527278423309326,
+      "learning_rate": 1.2960933859067937e-06,
+      "loss": 0.7031,
+      "step": 49305
+    },
+    {
+      "epoch": 1.8983638113570742,
+      "grad_norm": 1.7028868198394775,
+      "learning_rate": 1.291245042087097e-06,
+      "loss": 0.8185,
+      "step": 49310
+    },
+    {
+      "epoch": 1.8985563041385949,
+      "grad_norm": 1.3725398778915405,
+      "learning_rate": 1.2864057245291384e-06,
+      "loss": 0.7387,
+      "step": 49315
+    },
+    {
+      "epoch": 1.8987487969201156,
+      "grad_norm": 0.7600288987159729,
+      "learning_rate": 1.2815754336754748e-06,
+      "loss": 0.7169,
+      "step": 49320
+    },
+    {
+      "epoch": 1.8989412897016362,
+      "grad_norm": 1.943352460861206,
+      "learning_rate": 1.2767541699677865e-06,
+      "loss": 0.8994,
+      "step": 49325
+    },
+    {
+      "epoch": 1.899133782483157,
+      "grad_norm": 1.6857982873916626,
+      "learning_rate": 1.271941933846943e-06,
+      "loss": 0.8606,
+      "step": 49330
+    },
+    {
+      "epoch": 1.8993262752646776,
+      "grad_norm": 1.2063673734664917,
+      "learning_rate": 1.2671387257530033e-06,
+      "loss": 0.8712,
+      "step": 49335
+    },
+    {
+      "epoch": 1.8995187680461982,
+      "grad_norm": 1.1214938163757324,
+      "learning_rate": 1.2623445461251826e-06,
+      "loss": 0.8094,
+      "step": 49340
+    },
+    {
+      "epoch": 1.899711260827719,
+      "grad_norm": 1.7402042150497437,
+      "learning_rate": 1.257559395401875e-06,
+      "loss": 0.7574,
+      "step": 49345
+    },
+    {
+      "epoch": 1.8999037536092396,
+      "grad_norm": 2.6430463790893555,
+      "learning_rate": 1.2527832740206413e-06,
+      "loss": 0.89,
+      "step": 49350
+    },
+    {
+      "epoch": 1.9000962463907602,
+      "grad_norm": 1.625260829925537,
+      "learning_rate": 1.2480161824182435e-06,
+      "loss": 0.8766,
+      "step": 49355
+    },
+    {
+      "epoch": 1.900288739172281,
+      "grad_norm": 1.6484109163284302,
+      "learning_rate": 1.2432581210305883e-06,
+      "loss": 0.6991,
+      "step": 49360
+    },
+    {
+      "epoch": 1.9004812319538016,
+      "grad_norm": 1.4141255617141724,
+      "learning_rate": 1.2385090902927943e-06,
+      "loss": 0.8387,
+      "step": 49365
+    },
+    {
+      "epoch": 1.9006737247353225,
+      "grad_norm": 2.3338122367858887,
+      "learning_rate": 1.233769090639092e-06,
+      "loss": 0.9757,
+      "step": 49370
+    },
+    {
+      "epoch": 1.9008662175168431,
+      "grad_norm": 1.6034566164016724,
+      "learning_rate": 1.2290381225029345e-06,
+      "loss": 0.8093,
+      "step": 49375
+    },
+    {
+      "epoch": 1.9010587102983638,
+      "grad_norm": 1.138555645942688,
+      "learning_rate": 1.224316186316954e-06,
+      "loss": 0.7738,
+      "step": 49380
+    },
+    {
+      "epoch": 1.9012512030798845,
+      "grad_norm": 1.7570353746414185,
+      "learning_rate": 1.2196032825129377e-06,
+      "loss": 0.8496,
+      "step": 49385
+    },
+    {
+      "epoch": 1.9014436958614052,
+      "grad_norm": 0.8414403796195984,
+      "learning_rate": 1.2148994115218194e-06,
+      "loss": 0.8599,
+      "step": 49390
+    },
+    {
+      "epoch": 1.901636188642926,
+      "grad_norm": 1.4490456581115723,
+      "learning_rate": 1.2102045737737655e-06,
+      "loss": 0.7314,
+      "step": 49395
+    },
+    {
+      "epoch": 1.9018286814244467,
+      "grad_norm": 1.2178351879119873,
+      "learning_rate": 1.2055187696980885e-06,
+      "loss": 0.8655,
+      "step": 49400
+    },
+    {
+      "epoch": 1.9020211742059674,
+      "grad_norm": 1.1484450101852417,
+      "learning_rate": 1.2008419997232567e-06,
+      "loss": 0.7352,
+      "step": 49405
+    },
+    {
+      "epoch": 1.902213666987488,
+      "grad_norm": 2.422426462173462,
+      "learning_rate": 1.1961742642769502e-06,
+      "loss": 0.8654,
+      "step": 49410
+    },
+    {
+      "epoch": 1.9024061597690087,
+      "grad_norm": 1.6864680051803589,
+      "learning_rate": 1.1915155637859942e-06,
+      "loss": 0.7319,
+      "step": 49415
+    },
+    {
+      "epoch": 1.9025986525505294,
+      "grad_norm": 1.0465269088745117,
+      "learning_rate": 1.1868658986763704e-06,
+      "loss": 0.6556,
+      "step": 49420
+    },
+    {
+      "epoch": 1.90279114533205,
+      "grad_norm": 1.5697258710861206,
+      "learning_rate": 1.182225269373305e-06,
+      "loss": 0.8959,
+      "step": 49425
+    },
+    {
+      "epoch": 1.9029836381135707,
+      "grad_norm": 1.0091673135757446,
+      "learning_rate": 1.1775936763011252e-06,
+      "loss": 0.8843,
+      "step": 49430
+    },
+    {
+      "epoch": 1.9031761308950914,
+      "grad_norm": 1.5295612812042236,
+      "learning_rate": 1.1729711198833592e-06,
+      "loss": 0.7747,
+      "step": 49435
+    },
+    {
+      "epoch": 1.903368623676612,
+      "grad_norm": 1.655747413635254,
+      "learning_rate": 1.1683576005427243e-06,
+      "loss": 0.7618,
+      "step": 49440
+    },
+    {
+      "epoch": 1.9035611164581328,
+      "grad_norm": 1.3846567869186401,
+      "learning_rate": 1.163753118701083e-06,
+      "loss": 0.7765,
+      "step": 49445
+    },
+    {
+      "epoch": 1.9037536092396534,
+      "grad_norm": 1.6685348749160767,
+      "learning_rate": 1.1591576747794874e-06,
+      "loss": 0.7638,
+      "step": 49450
+    },
+    {
+      "epoch": 1.903946102021174,
+      "grad_norm": 1.5088547468185425,
+      "learning_rate": 1.154571269198168e-06,
+      "loss": 0.7608,
+      "step": 49455
+    },
+    {
+      "epoch": 1.9041385948026948,
+      "grad_norm": 2.2280006408691406,
+      "learning_rate": 1.1499939023765116e-06,
+      "loss": 0.8825,
+      "step": 49460
+    },
+    {
+      "epoch": 1.9043310875842154,
+      "grad_norm": 1.4215261936187744,
+      "learning_rate": 1.1454255747330834e-06,
+      "loss": 0.953,
+      "step": 49465
+    },
+    {
+      "epoch": 1.9045235803657363,
+      "grad_norm": 1.827169418334961,
+      "learning_rate": 1.1408662866856379e-06,
+      "loss": 0.7853,
+      "step": 49470
+    },
+    {
+      "epoch": 1.904716073147257,
+      "grad_norm": 1.3283016681671143,
+      "learning_rate": 1.1363160386510975e-06,
+      "loss": 0.6103,
+      "step": 49475
+    },
+    {
+      "epoch": 1.9049085659287777,
+      "grad_norm": 1.2778635025024414,
+      "learning_rate": 1.13177483104554e-06,
+      "loss": 0.7057,
+      "step": 49480
+    },
+    {
+      "epoch": 1.9051010587102983,
+      "grad_norm": 1.315544605255127,
+      "learning_rate": 1.1272426642842337e-06,
+      "loss": 0.7023,
+      "step": 49485
+    },
+    {
+      "epoch": 1.905293551491819,
+      "grad_norm": 1.4004102945327759,
+      "learning_rate": 1.1227195387816136e-06,
+      "loss": 0.6889,
+      "step": 49490
+    },
+    {
+      "epoch": 1.90548604427334,
+      "grad_norm": 2.1009254455566406,
+      "learning_rate": 1.118205454951271e-06,
+      "loss": 0.8759,
+      "step": 49495
+    },
+    {
+      "epoch": 1.9056785370548606,
+      "grad_norm": 1.2553610801696777,
+      "learning_rate": 1.1137004132060314e-06,
+      "loss": 0.7435,
+      "step": 49500
+    },
+    {
+      "epoch": 1.9058710298363812,
+      "grad_norm": 0.8491818308830261,
+      "learning_rate": 1.1092044139578205e-06,
+      "loss": 0.6341,
+      "step": 49505
+    },
+    {
+      "epoch": 1.906063522617902,
+      "grad_norm": 1.680105447769165,
+      "learning_rate": 1.1047174576177654e-06,
+      "loss": 0.8919,
+      "step": 49510
+    },
+    {
+      "epoch": 1.9062560153994226,
+      "grad_norm": 1.8832811117172241,
+      "learning_rate": 1.1002395445961932e-06,
+      "loss": 0.7662,
+      "step": 49515
+    },
+    {
+      "epoch": 1.9064485081809432,
+      "grad_norm": 1.2031275033950806,
+      "learning_rate": 1.0957706753025432e-06,
+      "loss": 0.7988,
+      "step": 49520
+    },
+    {
+      "epoch": 1.906641000962464,
+      "grad_norm": 2.169948101043701,
+      "learning_rate": 1.0913108501454993e-06,
+      "loss": 0.8608,
+      "step": 49525
+    },
+    {
+      "epoch": 1.9068334937439846,
+      "grad_norm": 1.2307795286178589,
+      "learning_rate": 1.086860069532869e-06,
+      "loss": 0.7195,
+      "step": 49530
+    },
+    {
+      "epoch": 1.9070259865255053,
+      "grad_norm": 1.0904724597930908,
+      "learning_rate": 1.0824183338716377e-06,
+      "loss": 0.7586,
+      "step": 49535
+    },
+    {
+      "epoch": 1.907218479307026,
+      "grad_norm": 2.3190550804138184,
+      "learning_rate": 1.0779856435679913e-06,
+      "loss": 0.9954,
+      "step": 49540
+    },
+    {
+      "epoch": 1.9074109720885466,
+      "grad_norm": 1.5595754384994507,
+      "learning_rate": 1.0735619990272506e-06,
+      "loss": 0.7631,
+      "step": 49545
+    },
+    {
+      "epoch": 1.9076034648700673,
+      "grad_norm": 1.5422192811965942,
+      "learning_rate": 1.069147400653936e-06,
+      "loss": 0.7709,
+      "step": 49550
+    },
+    {
+      "epoch": 1.907795957651588,
+      "grad_norm": 1.085741639137268,
+      "learning_rate": 1.0647418488517358e-06,
+      "loss": 0.6323,
+      "step": 49555
+    },
+    {
+      "epoch": 1.9079884504331086,
+      "grad_norm": 1.0943596363067627,
+      "learning_rate": 1.0603453440235168e-06,
+      "loss": 0.7902,
+      "step": 49560
+    },
+    {
+      "epoch": 1.9081809432146295,
+      "grad_norm": 0.9542067646980286,
+      "learning_rate": 1.0559578865712905e-06,
+      "loss": 0.8609,
+      "step": 49565
+    },
+    {
+      "epoch": 1.9083734359961502,
+      "grad_norm": 2.1979501247406006,
+      "learning_rate": 1.0515794768962806e-06,
+      "loss": 0.7887,
+      "step": 49570
+    },
+    {
+      "epoch": 1.9085659287776708,
+      "grad_norm": 1.0749517679214478,
+      "learning_rate": 1.0472101153988446e-06,
+      "loss": 0.8218,
+      "step": 49575
+    },
+    {
+      "epoch": 1.9087584215591915,
+      "grad_norm": 1.4718987941741943,
+      "learning_rate": 1.0428498024785404e-06,
+      "loss": 0.8026,
+      "step": 49580
+    },
+    {
+      "epoch": 1.9089509143407122,
+      "grad_norm": 1.4935468435287476,
+      "learning_rate": 1.038498538534105e-06,
+      "loss": 0.7906,
+      "step": 49585
+    },
+    {
+      "epoch": 1.909143407122233,
+      "grad_norm": 1.2869309186935425,
+      "learning_rate": 1.034156323963409e-06,
+      "loss": 0.7009,
+      "step": 49590
+    },
+    {
+      "epoch": 1.9093358999037537,
+      "grad_norm": 1.7238849401474,
+      "learning_rate": 1.0298231591635232e-06,
+      "loss": 0.6669,
+      "step": 49595
+    },
+    {
+      "epoch": 1.9095283926852744,
+      "grad_norm": 1.0271257162094116,
+      "learning_rate": 1.0254990445306978e-06,
+      "loss": 0.893,
+      "step": 49600
+    },
+    {
+      "epoch": 1.909720885466795,
+      "grad_norm": 1.4019641876220703,
+      "learning_rate": 1.0211839804603385e-06,
+      "loss": 0.7958,
+      "step": 49605
+    },
+    {
+      "epoch": 1.9099133782483158,
+      "grad_norm": 2.0241641998291016,
+      "learning_rate": 1.0168779673470296e-06,
+      "loss": 0.9105,
+      "step": 49610
+    },
+    {
+      "epoch": 1.9101058710298364,
+      "grad_norm": 1.333714485168457,
+      "learning_rate": 1.0125810055845231e-06,
+      "loss": 0.7325,
+      "step": 49615
+    },
+    {
+      "epoch": 1.910298363811357,
+      "grad_norm": 1.3105881214141846,
+      "learning_rate": 1.00829309556576e-06,
+      "loss": 0.8457,
+      "step": 49620
+    },
+    {
+      "epoch": 1.9104908565928778,
+      "grad_norm": 1.3533077239990234,
+      "learning_rate": 1.0040142376828266e-06,
+      "loss": 0.7181,
+      "step": 49625
+    },
+    {
+      "epoch": 1.9106833493743984,
+      "grad_norm": 1.5712995529174805,
+      "learning_rate": 9.9974443232701e-07,
+      "loss": 0.9438,
+      "step": 49630
+    },
+    {
+      "epoch": 1.910875842155919,
+      "grad_norm": 1.0359176397323608,
+      "learning_rate": 9.954836798887424e-07,
+      "loss": 0.7692,
+      "step": 49635
+    },
+    {
+      "epoch": 1.9110683349374398,
+      "grad_norm": 1.355884313583374,
+      "learning_rate": 9.912319807576452e-07,
+      "loss": 0.856,
+      "step": 49640
+    },
+    {
+      "epoch": 1.9112608277189604,
+      "grad_norm": 1.1907180547714233,
+      "learning_rate": 9.86989335322519e-07,
+      "loss": 0.7598,
+      "step": 49645
+    },
+    {
+      "epoch": 1.9114533205004811,
+      "grad_norm": 1.5118666887283325,
+      "learning_rate": 9.827557439713086e-07,
+      "loss": 0.9634,
+      "step": 49650
+    },
+    {
+      "epoch": 1.9116458132820018,
+      "grad_norm": 1.2580753564834595,
+      "learning_rate": 9.7853120709116e-07,
+      "loss": 0.7613,
+      "step": 49655
+    },
+    {
+      "epoch": 1.9118383060635225,
+      "grad_norm": 1.043454885482788,
+      "learning_rate": 9.743157250683644e-07,
+      "loss": 0.7669,
+      "step": 49660
+    },
+    {
+      "epoch": 1.9120307988450433,
+      "grad_norm": 1.603769302368164,
+      "learning_rate": 9.701092982884351e-07,
+      "loss": 0.8249,
+      "step": 49665
+    },
+    {
+      "epoch": 1.912223291626564,
+      "grad_norm": 1.0182993412017822,
+      "learning_rate": 9.659119271359762e-07,
+      "loss": 0.7097,
+      "step": 49670
+    },
+    {
+      "epoch": 1.9124157844080847,
+      "grad_norm": 1.6166123151779175,
+      "learning_rate": 9.617236119948358e-07,
+      "loss": 0.6695,
+      "step": 49675
+    },
+    {
+      "epoch": 1.9126082771896054,
+      "grad_norm": 1.0179489850997925,
+      "learning_rate": 9.575443532480076e-07,
+      "loss": 0.7893,
+      "step": 49680
+    },
+    {
+      "epoch": 1.9128007699711262,
+      "grad_norm": 1.4863917827606201,
+      "learning_rate": 9.533741512776417e-07,
+      "loss": 0.7844,
+      "step": 49685
+    },
+    {
+      "epoch": 1.912993262752647,
+      "grad_norm": 2.095845937728882,
+      "learning_rate": 9.492130064650995e-07,
+      "loss": 0.7687,
+      "step": 49690
+    },
+    {
+      "epoch": 1.9131857555341676,
+      "grad_norm": 0.9106748104095459,
+      "learning_rate": 9.450609191908766e-07,
+      "loss": 0.8065,
+      "step": 49695
+    },
+    {
+      "epoch": 1.9133782483156883,
+      "grad_norm": 0.5192959308624268,
+      "learning_rate": 9.40917889834636e-07,
+      "loss": 0.8384,
+      "step": 49700
+    },
+    {
+      "epoch": 1.913570741097209,
+      "grad_norm": 0.8835710287094116,
+      "learning_rate": 9.367839187752636e-07,
+      "loss": 0.8304,
+      "step": 49705
+    },
+    {
+      "epoch": 1.9137632338787296,
+      "grad_norm": 1.3475233316421509,
+      "learning_rate": 9.326590063907681e-07,
+      "loss": 0.8051,
+      "step": 49710
+    },
+    {
+      "epoch": 1.9139557266602503,
+      "grad_norm": 1.0911017656326294,
+      "learning_rate": 9.285431530583366e-07,
+      "loss": 0.6483,
+      "step": 49715
+    },
+    {
+      "epoch": 1.914148219441771,
+      "grad_norm": 1.183143973350525,
+      "learning_rate": 9.244363591543459e-07,
+      "loss": 0.9217,
+      "step": 49720
+    },
+    {
+      "epoch": 1.9143407122232916,
+      "grad_norm": 1.7121336460113525,
+      "learning_rate": 9.2033862505434e-07,
+      "loss": 0.7959,
+      "step": 49725
+    },
+    {
+      "epoch": 1.9145332050048123,
+      "grad_norm": 1.2456363439559937,
+      "learning_rate": 9.162499511330192e-07,
+      "loss": 0.7527,
+      "step": 49730
+    },
+    {
+      "epoch": 1.914725697786333,
+      "grad_norm": 2.138845205307007,
+      "learning_rate": 9.121703377642732e-07,
+      "loss": 0.9166,
+      "step": 49735
+    },
+    {
+      "epoch": 1.9149181905678536,
+      "grad_norm": 1.7342472076416016,
+      "learning_rate": 9.080997853211592e-07,
+      "loss": 0.7697,
+      "step": 49740
+    },
+    {
+      "epoch": 1.9151106833493743,
+      "grad_norm": 1.9589825868606567,
+      "learning_rate": 9.040382941758907e-07,
+      "loss": 0.86,
+      "step": 49745
+    },
+    {
+      "epoch": 1.915303176130895,
+      "grad_norm": 0.9614568948745728,
+      "learning_rate": 8.999858646998704e-07,
+      "loss": 0.8351,
+      "step": 49750
+    },
+    {
+      "epoch": 1.9154956689124156,
+      "grad_norm": 2.190661668777466,
+      "learning_rate": 8.959424972636576e-07,
+      "loss": 0.8423,
+      "step": 49755
+    },
+    {
+      "epoch": 1.9156881616939365,
+      "grad_norm": 1.1599905490875244,
+      "learning_rate": 8.91908192237012e-07,
+      "loss": 0.9698,
+      "step": 49760
+    },
+    {
+      "epoch": 1.9158806544754572,
+      "grad_norm": 0.9257279634475708,
+      "learning_rate": 8.878829499888164e-07,
+      "loss": 0.7918,
+      "step": 49765
+    },
+    {
+      "epoch": 1.9160731472569779,
+      "grad_norm": 1.8219150304794312,
+      "learning_rate": 8.838667708871873e-07,
+      "loss": 0.7945,
+      "step": 49770
+    },
+    {
+      "epoch": 1.9162656400384985,
+      "grad_norm": 1.6939747333526611,
+      "learning_rate": 8.798596552993421e-07,
+      "loss": 0.8348,
+      "step": 49775
+    },
+    {
+      "epoch": 1.9164581328200192,
+      "grad_norm": 1.0439796447753906,
+      "learning_rate": 8.758616035917211e-07,
+      "loss": 0.8166,
+      "step": 49780
+    },
+    {
+      "epoch": 1.91665062560154,
+      "grad_norm": 1.6584985256195068,
+      "learning_rate": 8.718726161299206e-07,
+      "loss": 0.7026,
+      "step": 49785
+    },
+    {
+      "epoch": 1.9168431183830608,
+      "grad_norm": 1.5818768739700317,
+      "learning_rate": 8.678926932787157e-07,
+      "loss": 0.7577,
+      "step": 49790
+    },
+    {
+      "epoch": 1.9170356111645814,
+      "grad_norm": 1.8825992345809937,
+      "learning_rate": 8.639218354020151e-07,
+      "loss": 0.6868,
+      "step": 49795
+    },
+    {
+      "epoch": 1.917228103946102,
+      "grad_norm": 1.7052052021026611,
+      "learning_rate": 8.599600428629617e-07,
+      "loss": 0.705,
+      "step": 49800
+    },
+    {
+      "epoch": 1.9174205967276228,
+      "grad_norm": 1.2492605447769165,
+      "learning_rate": 8.560073160238213e-07,
+      "loss": 0.7604,
+      "step": 49805
+    },
+    {
+      "epoch": 1.9176130895091434,
+      "grad_norm": 1.3016527891159058,
+      "learning_rate": 8.520636552460381e-07,
+      "loss": 0.681,
+      "step": 49810
+    },
+    {
+      "epoch": 1.9178055822906641,
+      "grad_norm": 1.7560960054397583,
+      "learning_rate": 8.481290608902348e-07,
+      "loss": 0.8405,
+      "step": 49815
+    },
+    {
+      "epoch": 1.9179980750721848,
+      "grad_norm": 1.0738979578018188,
+      "learning_rate": 8.442035333162013e-07,
+      "loss": 0.9153,
+      "step": 49820
+    },
+    {
+      "epoch": 1.9181905678537055,
+      "grad_norm": 2.123007297515869,
+      "learning_rate": 8.402870728829282e-07,
+      "loss": 0.8122,
+      "step": 49825
+    },
+    {
+      "epoch": 1.9183830606352261,
+      "grad_norm": 1.9177762269973755,
+      "learning_rate": 8.363796799485069e-07,
+      "loss": 0.9144,
+      "step": 49830
+    },
+    {
+      "epoch": 1.9185755534167468,
+      "grad_norm": 1.4648776054382324,
+      "learning_rate": 8.324813548702847e-07,
+      "loss": 0.8814,
+      "step": 49835
+    },
+    {
+      "epoch": 1.9187680461982675,
+      "grad_norm": 1.2725647687911987,
+      "learning_rate": 8.285920980047102e-07,
+      "loss": 0.8659,
+      "step": 49840
+    },
+    {
+      "epoch": 1.9189605389797881,
+      "grad_norm": 0.6527199149131775,
+      "learning_rate": 8.247119097074319e-07,
+      "loss": 0.7217,
+      "step": 49845
+    },
+    {
+      "epoch": 1.9191530317613088,
+      "grad_norm": 1.6857417821884155,
+      "learning_rate": 8.208407903332771e-07,
+      "loss": 0.8141,
+      "step": 49850
+    },
+    {
+      "epoch": 1.9193455245428297,
+      "grad_norm": 0.8140450716018677,
+      "learning_rate": 8.169787402362406e-07,
+      "loss": 0.7364,
+      "step": 49855
+    },
+    {
+      "epoch": 1.9195380173243504,
+      "grad_norm": 1.138654112815857,
+      "learning_rate": 8.131257597694508e-07,
+      "loss": 0.9553,
+      "step": 49860
+    },
+    {
+      "epoch": 1.919730510105871,
+      "grad_norm": 1.369598150253296,
+      "learning_rate": 8.092818492852705e-07,
+      "loss": 0.7387,
+      "step": 49865
+    },
+    {
+      "epoch": 1.9199230028873917,
+      "grad_norm": 1.1333906650543213,
+      "learning_rate": 8.054470091351851e-07,
+      "loss": 0.7376,
+      "step": 49870
+    },
+    {
+      "epoch": 1.9201154956689124,
+      "grad_norm": 1.7450077533721924,
+      "learning_rate": 8.016212396698697e-07,
+      "loss": 0.9635,
+      "step": 49875
+    },
+    {
+      "epoch": 1.9203079884504333,
+      "grad_norm": 0.9076559543609619,
+      "learning_rate": 7.978045412391555e-07,
+      "loss": 0.7919,
+      "step": 49880
+    },
+    {
+      "epoch": 1.920500481231954,
+      "grad_norm": 1.402205228805542,
+      "learning_rate": 7.939969141920744e-07,
+      "loss": 0.8091,
+      "step": 49885
+    },
+    {
+      "epoch": 1.9206929740134746,
+      "grad_norm": 1.2806485891342163,
+      "learning_rate": 7.901983588767814e-07,
+      "loss": 0.7798,
+      "step": 49890
+    },
+    {
+      "epoch": 1.9208854667949953,
+      "grad_norm": 2.2308027744293213,
+      "learning_rate": 7.864088756406429e-07,
+      "loss": 0.829,
+      "step": 49895
+    },
+    {
+      "epoch": 1.921077959576516,
+      "grad_norm": 1.4024001359939575,
+      "learning_rate": 7.826284648301929e-07,
+      "loss": 0.7619,
+      "step": 49900
+    },
+    {
+      "epoch": 1.9212704523580366,
+      "grad_norm": 1.3569003343582153,
+      "learning_rate": 7.788571267911104e-07,
+      "loss": 0.7127,
+      "step": 49905
+    },
+    {
+      "epoch": 1.9214629451395573,
+      "grad_norm": 2.439872980117798,
+      "learning_rate": 7.75094861868264e-07,
+      "loss": 0.9349,
+      "step": 49910
+    },
+    {
+      "epoch": 1.921655437921078,
+      "grad_norm": 1.2316006422042847,
+      "learning_rate": 7.713416704056897e-07,
+      "loss": 0.8283,
+      "step": 49915
+    },
+    {
+      "epoch": 1.9218479307025986,
+      "grad_norm": 1.0408631563186646,
+      "learning_rate": 7.675975527465906e-07,
+      "loss": 0.765,
+      "step": 49920
+    },
+    {
+      "epoch": 1.9220404234841193,
+      "grad_norm": 1.4583711624145508,
+      "learning_rate": 7.638625092333373e-07,
+      "loss": 0.8184,
+      "step": 49925
+    },
+    {
+      "epoch": 1.92223291626564,
+      "grad_norm": 2.3234689235687256,
+      "learning_rate": 7.601365402074789e-07,
+      "loss": 0.8176,
+      "step": 49930
+    },
+    {
+      "epoch": 1.9224254090471606,
+      "grad_norm": 1.4219454526901245,
+      "learning_rate": 7.564196460097316e-07,
+      "loss": 0.6931,
+      "step": 49935
+    },
+    {
+      "epoch": 1.9226179018286813,
+      "grad_norm": 1.5232850313186646,
+      "learning_rate": 7.527118269799793e-07,
+      "loss": 0.8985,
+      "step": 49940
+    },
+    {
+      "epoch": 1.922810394610202,
+      "grad_norm": 1.4048689603805542,
+      "learning_rate": 7.490130834572728e-07,
+      "loss": 0.9178,
+      "step": 49945
+    },
+    {
+      "epoch": 1.9230028873917226,
+      "grad_norm": 0.891716718673706,
+      "learning_rate": 7.453234157798416e-07,
+      "loss": 0.9839,
+      "step": 49950
+    },
+    {
+      "epoch": 1.9231953801732435,
+      "grad_norm": 1.1254078149795532,
+      "learning_rate": 7.416428242850937e-07,
+      "loss": 0.928,
+      "step": 49955
+    },
+    {
+      "epoch": 1.9233878729547642,
+      "grad_norm": 1.3273284435272217,
+      "learning_rate": 7.379713093095708e-07,
+      "loss": 0.7957,
+      "step": 49960
+    },
+    {
+      "epoch": 1.9235803657362849,
+      "grad_norm": 1.1084377765655518,
+      "learning_rate": 7.343088711890267e-07,
+      "loss": 0.9596,
+      "step": 49965
+    },
+    {
+      "epoch": 1.9237728585178056,
+      "grad_norm": 1.641728162765503,
+      "learning_rate": 7.306555102583601e-07,
+      "loss": 0.8442,
+      "step": 49970
+    },
+    {
+      "epoch": 1.9239653512993262,
+      "grad_norm": 1.5067929029464722,
+      "learning_rate": 7.270112268516483e-07,
+      "loss": 0.7893,
+      "step": 49975
+    },
+    {
+      "epoch": 1.9241578440808471,
+      "grad_norm": 0.8102709054946899,
+      "learning_rate": 7.233760213021357e-07,
+      "loss": 0.7961,
+      "step": 49980
+    },
+    {
+      "epoch": 1.9243503368623678,
+      "grad_norm": 2.6088552474975586,
+      "learning_rate": 7.197498939422343e-07,
+      "loss": 0.8979,
+      "step": 49985
+    },
+    {
+      "epoch": 1.9245428296438885,
+      "grad_norm": 1.865166187286377,
+      "learning_rate": 7.161328451035454e-07,
+      "loss": 0.8434,
+      "step": 49990
+    },
+    {
+      "epoch": 1.9247353224254091,
+      "grad_norm": 1.1561760902404785,
+      "learning_rate": 7.125248751167934e-07,
+      "loss": 0.7265,
+      "step": 49995
+    },
+    {
+      "epoch": 1.9249278152069298,
+      "grad_norm": 1.1306687593460083,
+      "learning_rate": 7.089259843119478e-07,
+      "loss": 0.9502,
+      "step": 50000
     }
   ],
   "logging_steps": 5,
@@ -56014,7 +70014,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 10000,
-  "total_flos": 1.2478892695491133e+18,
+  "total_flos": 1.5612551596802458e+18,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null