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": 0.8505146128020254,
+  "epoch": 1.0006054268259124,
   "eval_steps": 500,
-  "global_step": 15453,
+  "global_step": 18180,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -108178,6 +108178,19095 @@
       "learning_rate": 6.18751132978102e-06,
       "loss": 0.7619,
       "step": 15453
+    },
+    {
+      "epoch": 0.8505696516043811,
+      "grad_norm": 0.7532587647438049,
+      "learning_rate": 6.1870902614890384e-06,
+      "loss": 0.6929,
+      "step": 15454
+    },
+    {
+      "epoch": 0.8506246904067367,
+      "grad_norm": 0.5978666543960571,
+      "learning_rate": 6.186669184275204e-06,
+      "loss": 0.6246,
+      "step": 15455
+    },
+    {
+      "epoch": 0.8506797292090924,
+      "grad_norm": 0.8646023869514465,
+      "learning_rate": 6.186248098142681e-06,
+      "loss": 0.8215,
+      "step": 15456
+    },
+    {
+      "epoch": 0.8507347680114481,
+      "grad_norm": 0.635597825050354,
+      "learning_rate": 6.1858270030946355e-06,
+      "loss": 0.7683,
+      "step": 15457
+    },
+    {
+      "epoch": 0.8507898068138038,
+      "grad_norm": 0.7014510631561279,
+      "learning_rate": 6.185405899134231e-06,
+      "loss": 0.708,
+      "step": 15458
+    },
+    {
+      "epoch": 0.8508448456161594,
+      "grad_norm": 1.6040544509887695,
+      "learning_rate": 6.184984786264633e-06,
+      "loss": 0.893,
+      "step": 15459
+    },
+    {
+      "epoch": 0.850899884418515,
+      "grad_norm": 0.8098211288452148,
+      "learning_rate": 6.184563664489007e-06,
+      "loss": 0.7704,
+      "step": 15460
+    },
+    {
+      "epoch": 0.8509549232208707,
+      "grad_norm": 0.8390217423439026,
+      "learning_rate": 6.184142533810518e-06,
+      "loss": 0.7161,
+      "step": 15461
+    },
+    {
+      "epoch": 0.8510099620232264,
+      "grad_norm": 0.792433500289917,
+      "learning_rate": 6.183721394232329e-06,
+      "loss": 0.7247,
+      "step": 15462
+    },
+    {
+      "epoch": 0.851065000825582,
+      "grad_norm": 0.6644124984741211,
+      "learning_rate": 6.183300245757609e-06,
+      "loss": 0.7996,
+      "step": 15463
+    },
+    {
+      "epoch": 0.8511200396279377,
+      "grad_norm": 0.7543407082557678,
+      "learning_rate": 6.182879088389521e-06,
+      "loss": 0.7912,
+      "step": 15464
+    },
+    {
+      "epoch": 0.8511750784302934,
+      "grad_norm": 0.7752966284751892,
+      "learning_rate": 6.18245792213123e-06,
+      "loss": 0.7222,
+      "step": 15465
+    },
+    {
+      "epoch": 0.851230117232649,
+      "grad_norm": 0.6863895654678345,
+      "learning_rate": 6.182036746985901e-06,
+      "loss": 0.7375,
+      "step": 15466
+    },
+    {
+      "epoch": 0.8512851560350047,
+      "grad_norm": 0.6404759287834167,
+      "learning_rate": 6.1816155629567006e-06,
+      "loss": 0.7666,
+      "step": 15467
+    },
+    {
+      "epoch": 0.8513401948373603,
+      "grad_norm": 0.6879389882087708,
+      "learning_rate": 6.181194370046795e-06,
+      "loss": 0.697,
+      "step": 15468
+    },
+    {
+      "epoch": 0.851395233639716,
+      "grad_norm": 0.6798561215400696,
+      "learning_rate": 6.180773168259347e-06,
+      "loss": 0.7401,
+      "step": 15469
+    },
+    {
+      "epoch": 0.8514502724420716,
+      "grad_norm": 0.6846516728401184,
+      "learning_rate": 6.180351957597524e-06,
+      "loss": 0.7512,
+      "step": 15470
+    },
+    {
+      "epoch": 0.8515053112444273,
+      "grad_norm": 0.6618537902832031,
+      "learning_rate": 6.1799307380644925e-06,
+      "loss": 0.7577,
+      "step": 15471
+    },
+    {
+      "epoch": 0.851560350046783,
+      "grad_norm": 0.7780229449272156,
+      "learning_rate": 6.179509509663417e-06,
+      "loss": 0.7145,
+      "step": 15472
+    },
+    {
+      "epoch": 0.8516153888491387,
+      "grad_norm": 0.8041914701461792,
+      "learning_rate": 6.179088272397464e-06,
+      "loss": 0.7662,
+      "step": 15473
+    },
+    {
+      "epoch": 0.8516704276514943,
+      "grad_norm": 0.6719852685928345,
+      "learning_rate": 6.178667026269799e-06,
+      "loss": 0.7269,
+      "step": 15474
+    },
+    {
+      "epoch": 0.8517254664538499,
+      "grad_norm": 0.6677427291870117,
+      "learning_rate": 6.178245771283589e-06,
+      "loss": 0.6949,
+      "step": 15475
+    },
+    {
+      "epoch": 0.8517805052562056,
+      "grad_norm": 0.7873098850250244,
+      "learning_rate": 6.177824507441998e-06,
+      "loss": 0.7435,
+      "step": 15476
+    },
+    {
+      "epoch": 0.8518355440585613,
+      "grad_norm": 0.726902186870575,
+      "learning_rate": 6.1774032347481935e-06,
+      "loss": 0.8712,
+      "step": 15477
+    },
+    {
+      "epoch": 0.8518905828609169,
+      "grad_norm": 0.6984882950782776,
+      "learning_rate": 6.176981953205342e-06,
+      "loss": 0.8705,
+      "step": 15478
+    },
+    {
+      "epoch": 0.8519456216632726,
+      "grad_norm": 0.9757348895072937,
+      "learning_rate": 6.176560662816609e-06,
+      "loss": 0.8172,
+      "step": 15479
+    },
+    {
+      "epoch": 0.8520006604656283,
+      "grad_norm": 0.8358891606330872,
+      "learning_rate": 6.1761393635851615e-06,
+      "loss": 0.7576,
+      "step": 15480
+    },
+    {
+      "epoch": 0.852055699267984,
+      "grad_norm": 0.7324516177177429,
+      "learning_rate": 6.175718055514165e-06,
+      "loss": 0.6785,
+      "step": 15481
+    },
+    {
+      "epoch": 0.8521107380703395,
+      "grad_norm": 0.7352122664451599,
+      "learning_rate": 6.175296738606785e-06,
+      "loss": 0.6998,
+      "step": 15482
+    },
+    {
+      "epoch": 0.8521657768726952,
+      "grad_norm": 0.7331148982048035,
+      "learning_rate": 6.17487541286619e-06,
+      "loss": 0.7318,
+      "step": 15483
+    },
+    {
+      "epoch": 0.8522208156750509,
+      "grad_norm": 0.7264272570610046,
+      "learning_rate": 6.174454078295547e-06,
+      "loss": 0.8328,
+      "step": 15484
+    },
+    {
+      "epoch": 0.8522758544774066,
+      "grad_norm": 0.6304524540901184,
+      "learning_rate": 6.174032734898021e-06,
+      "loss": 0.6982,
+      "step": 15485
+    },
+    {
+      "epoch": 0.8523308932797622,
+      "grad_norm": 0.709864616394043,
+      "learning_rate": 6.173611382676778e-06,
+      "loss": 0.784,
+      "step": 15486
+    },
+    {
+      "epoch": 0.8523859320821179,
+      "grad_norm": 0.689153254032135,
+      "learning_rate": 6.173190021634987e-06,
+      "loss": 0.596,
+      "step": 15487
+    },
+    {
+      "epoch": 0.8524409708844736,
+      "grad_norm": 0.6987396478652954,
+      "learning_rate": 6.172768651775815e-06,
+      "loss": 0.776,
+      "step": 15488
+    },
+    {
+      "epoch": 0.8524960096868293,
+      "grad_norm": 0.6473208665847778,
+      "learning_rate": 6.172347273102427e-06,
+      "loss": 0.7084,
+      "step": 15489
+    },
+    {
+      "epoch": 0.8525510484891848,
+      "grad_norm": 0.7659436464309692,
+      "learning_rate": 6.17192588561799e-06,
+      "loss": 0.6916,
+      "step": 15490
+    },
+    {
+      "epoch": 0.8526060872915405,
+      "grad_norm": 0.6404485702514648,
+      "learning_rate": 6.171504489325673e-06,
+      "loss": 0.6761,
+      "step": 15491
+    },
+    {
+      "epoch": 0.8526611260938962,
+      "grad_norm": 0.7309281826019287,
+      "learning_rate": 6.171083084228641e-06,
+      "loss": 0.7993,
+      "step": 15492
+    },
+    {
+      "epoch": 0.8527161648962519,
+      "grad_norm": 0.6718485355377197,
+      "learning_rate": 6.170661670330062e-06,
+      "loss": 0.719,
+      "step": 15493
+    },
+    {
+      "epoch": 0.8527712036986075,
+      "grad_norm": 0.7705026865005493,
+      "learning_rate": 6.170240247633101e-06,
+      "loss": 0.7148,
+      "step": 15494
+    },
+    {
+      "epoch": 0.8528262425009632,
+      "grad_norm": 0.9131157398223877,
+      "learning_rate": 6.169818816140931e-06,
+      "loss": 0.7439,
+      "step": 15495
+    },
+    {
+      "epoch": 0.8528812813033189,
+      "grad_norm": 0.7396848201751709,
+      "learning_rate": 6.169397375856715e-06,
+      "loss": 0.7745,
+      "step": 15496
+    },
+    {
+      "epoch": 0.8529363201056746,
+      "grad_norm": 0.613767147064209,
+      "learning_rate": 6.168975926783621e-06,
+      "loss": 0.6724,
+      "step": 15497
+    },
+    {
+      "epoch": 0.8529913589080301,
+      "grad_norm": 0.8665019869804382,
+      "learning_rate": 6.168554468924815e-06,
+      "loss": 0.6896,
+      "step": 15498
+    },
+    {
+      "epoch": 0.8530463977103858,
+      "grad_norm": 0.7870811223983765,
+      "learning_rate": 6.168133002283469e-06,
+      "loss": 0.7177,
+      "step": 15499
+    },
+    {
+      "epoch": 0.8531014365127415,
+      "grad_norm": 0.5830097794532776,
+      "learning_rate": 6.167711526862747e-06,
+      "loss": 0.6465,
+      "step": 15500
+    },
+    {
+      "epoch": 0.8531564753150972,
+      "grad_norm": 0.6497567892074585,
+      "learning_rate": 6.167290042665819e-06,
+      "loss": 0.6842,
+      "step": 15501
+    },
+    {
+      "epoch": 0.8532115141174528,
+      "grad_norm": 0.6574105620384216,
+      "learning_rate": 6.1668685496958515e-06,
+      "loss": 0.7197,
+      "step": 15502
+    },
+    {
+      "epoch": 0.8532665529198085,
+      "grad_norm": 0.7069656252861023,
+      "learning_rate": 6.166447047956011e-06,
+      "loss": 0.8031,
+      "step": 15503
+    },
+    {
+      "epoch": 0.8533215917221642,
+      "grad_norm": 0.700334370136261,
+      "learning_rate": 6.166025537449467e-06,
+      "loss": 0.7731,
+      "step": 15504
+    },
+    {
+      "epoch": 0.8533766305245198,
+      "grad_norm": 0.7227431535720825,
+      "learning_rate": 6.165604018179388e-06,
+      "loss": 0.7741,
+      "step": 15505
+    },
+    {
+      "epoch": 0.8534316693268754,
+      "grad_norm": 0.6752750277519226,
+      "learning_rate": 6.16518249014894e-06,
+      "loss": 0.7176,
+      "step": 15506
+    },
+    {
+      "epoch": 0.8534867081292311,
+      "grad_norm": 0.6595750451087952,
+      "learning_rate": 6.1647609533612925e-06,
+      "loss": 0.7758,
+      "step": 15507
+    },
+    {
+      "epoch": 0.8535417469315868,
+      "grad_norm": 0.7232886552810669,
+      "learning_rate": 6.1643394078196136e-06,
+      "loss": 0.667,
+      "step": 15508
+    },
+    {
+      "epoch": 0.8535967857339424,
+      "grad_norm": 0.7297642827033997,
+      "learning_rate": 6.163917853527072e-06,
+      "loss": 0.6951,
+      "step": 15509
+    },
+    {
+      "epoch": 0.8536518245362981,
+      "grad_norm": 0.6766324043273926,
+      "learning_rate": 6.163496290486834e-06,
+      "loss": 0.7429,
+      "step": 15510
+    },
+    {
+      "epoch": 0.8537068633386538,
+      "grad_norm": 0.6861003041267395,
+      "learning_rate": 6.16307471870207e-06,
+      "loss": 0.8159,
+      "step": 15511
+    },
+    {
+      "epoch": 0.8537619021410094,
+      "grad_norm": 0.687471330165863,
+      "learning_rate": 6.1626531381759494e-06,
+      "loss": 0.7001,
+      "step": 15512
+    },
+    {
+      "epoch": 0.853816940943365,
+      "grad_norm": 0.8295855522155762,
+      "learning_rate": 6.162231548911637e-06,
+      "loss": 0.824,
+      "step": 15513
+    },
+    {
+      "epoch": 0.8538719797457207,
+      "grad_norm": 0.7513163685798645,
+      "learning_rate": 6.161809950912304e-06,
+      "loss": 0.7563,
+      "step": 15514
+    },
+    {
+      "epoch": 0.8539270185480764,
+      "grad_norm": 0.824586033821106,
+      "learning_rate": 6.161388344181119e-06,
+      "loss": 0.9607,
+      "step": 15515
+    },
+    {
+      "epoch": 0.8539820573504321,
+      "grad_norm": 0.9423270225524902,
+      "learning_rate": 6.160966728721249e-06,
+      "loss": 0.7527,
+      "step": 15516
+    },
+    {
+      "epoch": 0.8540370961527877,
+      "grad_norm": 0.6734938621520996,
+      "learning_rate": 6.160545104535866e-06,
+      "loss": 0.7741,
+      "step": 15517
+    },
+    {
+      "epoch": 0.8540921349551434,
+      "grad_norm": 0.6953195929527283,
+      "learning_rate": 6.160123471628133e-06,
+      "loss": 0.7844,
+      "step": 15518
+    },
+    {
+      "epoch": 0.854147173757499,
+      "grad_norm": 0.8023058772087097,
+      "learning_rate": 6.1597018300012245e-06,
+      "loss": 0.815,
+      "step": 15519
+    },
+    {
+      "epoch": 0.8542022125598547,
+      "grad_norm": 1.1232868432998657,
+      "learning_rate": 6.159280179658308e-06,
+      "loss": 0.7702,
+      "step": 15520
+    },
+    {
+      "epoch": 0.8542572513622103,
+      "grad_norm": 0.6074864268302917,
+      "learning_rate": 6.158858520602552e-06,
+      "loss": 0.6025,
+      "step": 15521
+    },
+    {
+      "epoch": 0.854312290164566,
+      "grad_norm": 0.6659427285194397,
+      "learning_rate": 6.158436852837124e-06,
+      "loss": 0.7019,
+      "step": 15522
+    },
+    {
+      "epoch": 0.8543673289669217,
+      "grad_norm": 0.6226561665534973,
+      "learning_rate": 6.158015176365197e-06,
+      "loss": 0.6748,
+      "step": 15523
+    },
+    {
+      "epoch": 0.8544223677692774,
+      "grad_norm": 0.6821898818016052,
+      "learning_rate": 6.157593491189936e-06,
+      "loss": 0.7617,
+      "step": 15524
+    },
+    {
+      "epoch": 0.854477406571633,
+      "grad_norm": 0.6755489110946655,
+      "learning_rate": 6.157171797314513e-06,
+      "loss": 0.6664,
+      "step": 15525
+    },
+    {
+      "epoch": 0.8545324453739886,
+      "grad_norm": 0.5896545052528381,
+      "learning_rate": 6.156750094742096e-06,
+      "loss": 0.7147,
+      "step": 15526
+    },
+    {
+      "epoch": 0.8545874841763443,
+      "grad_norm": 0.7024844288825989,
+      "learning_rate": 6.1563283834758555e-06,
+      "loss": 0.7884,
+      "step": 15527
+    },
+    {
+      "epoch": 0.8546425229787,
+      "grad_norm": 0.8020251393318176,
+      "learning_rate": 6.15590666351896e-06,
+      "loss": 0.8014,
+      "step": 15528
+    },
+    {
+      "epoch": 0.8546975617810556,
+      "grad_norm": 0.7213570475578308,
+      "learning_rate": 6.1554849348745805e-06,
+      "loss": 0.7289,
+      "step": 15529
+    },
+    {
+      "epoch": 0.8547526005834113,
+      "grad_norm": 0.618235170841217,
+      "learning_rate": 6.155063197545884e-06,
+      "loss": 0.7241,
+      "step": 15530
+    },
+    {
+      "epoch": 0.854807639385767,
+      "grad_norm": 0.7157233953475952,
+      "learning_rate": 6.154641451536042e-06,
+      "loss": 0.7591,
+      "step": 15531
+    },
+    {
+      "epoch": 0.8548626781881227,
+      "grad_norm": 0.7156147360801697,
+      "learning_rate": 6.1542196968482245e-06,
+      "loss": 0.7337,
+      "step": 15532
+    },
+    {
+      "epoch": 0.8549177169904782,
+      "grad_norm": 0.7722104787826538,
+      "learning_rate": 6.153797933485601e-06,
+      "loss": 0.7287,
+      "step": 15533
+    },
+    {
+      "epoch": 0.8549727557928339,
+      "grad_norm": 0.8379881381988525,
+      "learning_rate": 6.1533761614513394e-06,
+      "loss": 0.7898,
+      "step": 15534
+    },
+    {
+      "epoch": 0.8550277945951896,
+      "grad_norm": 0.6861830949783325,
+      "learning_rate": 6.152954380748614e-06,
+      "loss": 0.7838,
+      "step": 15535
+    },
+    {
+      "epoch": 0.8550828333975453,
+      "grad_norm": 0.6730731725692749,
+      "learning_rate": 6.15253259138059e-06,
+      "loss": 0.7606,
+      "step": 15536
+    },
+    {
+      "epoch": 0.8551378721999009,
+      "grad_norm": 0.6933832168579102,
+      "learning_rate": 6.152110793350441e-06,
+      "loss": 0.7637,
+      "step": 15537
+    },
+    {
+      "epoch": 0.8551929110022566,
+      "grad_norm": 0.9518343806266785,
+      "learning_rate": 6.151688986661335e-06,
+      "loss": 0.7715,
+      "step": 15538
+    },
+    {
+      "epoch": 0.8552479498046123,
+      "grad_norm": 0.7800498008728027,
+      "learning_rate": 6.151267171316442e-06,
+      "loss": 0.7933,
+      "step": 15539
+    },
+    {
+      "epoch": 0.855302988606968,
+      "grad_norm": 0.8873908519744873,
+      "learning_rate": 6.150845347318934e-06,
+      "loss": 0.8349,
+      "step": 15540
+    },
+    {
+      "epoch": 0.8553580274093235,
+      "grad_norm": 0.6778621673583984,
+      "learning_rate": 6.15042351467198e-06,
+      "loss": 0.7855,
+      "step": 15541
+    },
+    {
+      "epoch": 0.8554130662116792,
+      "grad_norm": 0.6535203456878662,
+      "learning_rate": 6.150001673378751e-06,
+      "loss": 0.7288,
+      "step": 15542
+    },
+    {
+      "epoch": 0.8554681050140349,
+      "grad_norm": 0.7087036967277527,
+      "learning_rate": 6.149579823442418e-06,
+      "loss": 0.8542,
+      "step": 15543
+    },
+    {
+      "epoch": 0.8555231438163906,
+      "grad_norm": 0.8136983513832092,
+      "learning_rate": 6.1491579648661495e-06,
+      "loss": 0.8618,
+      "step": 15544
+    },
+    {
+      "epoch": 0.8555781826187462,
+      "grad_norm": 0.7439128756523132,
+      "learning_rate": 6.148736097653118e-06,
+      "loss": 0.7257,
+      "step": 15545
+    },
+    {
+      "epoch": 0.8556332214211019,
+      "grad_norm": 0.853769838809967,
+      "learning_rate": 6.148314221806493e-06,
+      "loss": 0.8759,
+      "step": 15546
+    },
+    {
+      "epoch": 0.8556882602234576,
+      "grad_norm": 0.6681458950042725,
+      "learning_rate": 6.147892337329446e-06,
+      "loss": 0.6993,
+      "step": 15547
+    },
+    {
+      "epoch": 0.8557432990258133,
+      "grad_norm": 0.6452274918556213,
+      "learning_rate": 6.147470444225147e-06,
+      "loss": 0.6838,
+      "step": 15548
+    },
+    {
+      "epoch": 0.8557983378281688,
+      "grad_norm": 0.7074391841888428,
+      "learning_rate": 6.147048542496769e-06,
+      "loss": 0.7978,
+      "step": 15549
+    },
+    {
+      "epoch": 0.8558533766305245,
+      "grad_norm": 0.634824275970459,
+      "learning_rate": 6.14662663214748e-06,
+      "loss": 0.6602,
+      "step": 15550
+    },
+    {
+      "epoch": 0.8559084154328802,
+      "grad_norm": 0.7253528237342834,
+      "learning_rate": 6.146204713180453e-06,
+      "loss": 0.8662,
+      "step": 15551
+    },
+    {
+      "epoch": 0.8559634542352358,
+      "grad_norm": 0.6583418846130371,
+      "learning_rate": 6.145782785598858e-06,
+      "loss": 0.7502,
+      "step": 15552
+    },
+    {
+      "epoch": 0.8560184930375915,
+      "grad_norm": 0.7276360392570496,
+      "learning_rate": 6.1453608494058645e-06,
+      "loss": 0.7066,
+      "step": 15553
+    },
+    {
+      "epoch": 0.8560735318399472,
+      "grad_norm": 0.6554223299026489,
+      "learning_rate": 6.144938904604646e-06,
+      "loss": 0.7468,
+      "step": 15554
+    },
+    {
+      "epoch": 0.8561285706423029,
+      "grad_norm": 0.6767130494117737,
+      "learning_rate": 6.144516951198374e-06,
+      "loss": 0.724,
+      "step": 15555
+    },
+    {
+      "epoch": 0.8561836094446584,
+      "grad_norm": 0.7025824785232544,
+      "learning_rate": 6.144094989190219e-06,
+      "loss": 0.7658,
+      "step": 15556
+    },
+    {
+      "epoch": 0.8562386482470141,
+      "grad_norm": 0.6780791282653809,
+      "learning_rate": 6.143673018583353e-06,
+      "loss": 0.7273,
+      "step": 15557
+    },
+    {
+      "epoch": 0.8562936870493698,
+      "grad_norm": 0.6552621722221375,
+      "learning_rate": 6.143251039380944e-06,
+      "loss": 0.7374,
+      "step": 15558
+    },
+    {
+      "epoch": 0.8563487258517255,
+      "grad_norm": 0.7451765537261963,
+      "learning_rate": 6.142829051586169e-06,
+      "loss": 0.7918,
+      "step": 15559
+    },
+    {
+      "epoch": 0.8564037646540811,
+      "grad_norm": 0.7365521788597107,
+      "learning_rate": 6.142407055202195e-06,
+      "loss": 0.7142,
+      "step": 15560
+    },
+    {
+      "epoch": 0.8564588034564368,
+      "grad_norm": 0.6708245277404785,
+      "learning_rate": 6.1419850502321976e-06,
+      "loss": 0.773,
+      "step": 15561
+    },
+    {
+      "epoch": 0.8565138422587925,
+      "grad_norm": 0.8878017067909241,
+      "learning_rate": 6.141563036679344e-06,
+      "loss": 0.7704,
+      "step": 15562
+    },
+    {
+      "epoch": 0.8565688810611481,
+      "grad_norm": 0.8903444409370422,
+      "learning_rate": 6.14114101454681e-06,
+      "loss": 0.8422,
+      "step": 15563
+    },
+    {
+      "epoch": 0.8566239198635037,
+      "grad_norm": 0.7124255895614624,
+      "learning_rate": 6.140718983837764e-06,
+      "loss": 0.74,
+      "step": 15564
+    },
+    {
+      "epoch": 0.8566789586658594,
+      "grad_norm": 0.7107509970664978,
+      "learning_rate": 6.14029694455538e-06,
+      "loss": 0.7052,
+      "step": 15565
+    },
+    {
+      "epoch": 0.8567339974682151,
+      "grad_norm": 0.6864815950393677,
+      "learning_rate": 6.13987489670283e-06,
+      "loss": 0.7541,
+      "step": 15566
+    },
+    {
+      "epoch": 0.8567890362705708,
+      "grad_norm": 0.696169912815094,
+      "learning_rate": 6.1394528402832845e-06,
+      "loss": 0.7829,
+      "step": 15567
+    },
+    {
+      "epoch": 0.8568440750729264,
+      "grad_norm": 0.5650855302810669,
+      "learning_rate": 6.139030775299917e-06,
+      "loss": 0.6268,
+      "step": 15568
+    },
+    {
+      "epoch": 0.8568991138752821,
+      "grad_norm": 0.8753485083580017,
+      "learning_rate": 6.138608701755899e-06,
+      "loss": 0.7658,
+      "step": 15569
+    },
+    {
+      "epoch": 0.8569541526776377,
+      "grad_norm": 0.6950936317443848,
+      "learning_rate": 6.138186619654401e-06,
+      "loss": 0.8013,
+      "step": 15570
+    },
+    {
+      "epoch": 0.8570091914799934,
+      "grad_norm": 0.6608526110649109,
+      "learning_rate": 6.1377645289986e-06,
+      "loss": 0.8095,
+      "step": 15571
+    },
+    {
+      "epoch": 0.857064230282349,
+      "grad_norm": 0.6566348075866699,
+      "learning_rate": 6.137342429791664e-06,
+      "loss": 0.787,
+      "step": 15572
+    },
+    {
+      "epoch": 0.8571192690847047,
+      "grad_norm": 0.6536870002746582,
+      "learning_rate": 6.136920322036768e-06,
+      "loss": 0.8754,
+      "step": 15573
+    },
+    {
+      "epoch": 0.8571743078870604,
+      "grad_norm": 0.6461239457130432,
+      "learning_rate": 6.136498205737081e-06,
+      "loss": 0.6948,
+      "step": 15574
+    },
+    {
+      "epoch": 0.8572293466894161,
+      "grad_norm": 0.6441778540611267,
+      "learning_rate": 6.13607608089578e-06,
+      "loss": 0.7097,
+      "step": 15575
+    },
+    {
+      "epoch": 0.8572843854917717,
+      "grad_norm": 0.6770008206367493,
+      "learning_rate": 6.135653947516034e-06,
+      "loss": 0.7497,
+      "step": 15576
+    },
+    {
+      "epoch": 0.8573394242941274,
+      "grad_norm": 0.6479504704475403,
+      "learning_rate": 6.1352318056010175e-06,
+      "loss": 0.6975,
+      "step": 15577
+    },
+    {
+      "epoch": 0.857394463096483,
+      "grad_norm": 0.6586747765541077,
+      "learning_rate": 6.134809655153901e-06,
+      "loss": 0.7977,
+      "step": 15578
+    },
+    {
+      "epoch": 0.8574495018988387,
+      "grad_norm": 0.6888973116874695,
+      "learning_rate": 6.1343874961778604e-06,
+      "loss": 0.7399,
+      "step": 15579
+    },
+    {
+      "epoch": 0.8575045407011943,
+      "grad_norm": 0.6897402405738831,
+      "learning_rate": 6.133965328676066e-06,
+      "loss": 0.7507,
+      "step": 15580
+    },
+    {
+      "epoch": 0.85755957950355,
+      "grad_norm": 0.6857936382293701,
+      "learning_rate": 6.133543152651693e-06,
+      "loss": 0.76,
+      "step": 15581
+    },
+    {
+      "epoch": 0.8576146183059057,
+      "grad_norm": 0.8104296922683716,
+      "learning_rate": 6.133120968107912e-06,
+      "loss": 0.711,
+      "step": 15582
+    },
+    {
+      "epoch": 0.8576696571082614,
+      "grad_norm": 0.786551296710968,
+      "learning_rate": 6.132698775047897e-06,
+      "loss": 0.7603,
+      "step": 15583
+    },
+    {
+      "epoch": 0.857724695910617,
+      "grad_norm": 0.6685918569564819,
+      "learning_rate": 6.132276573474822e-06,
+      "loss": 0.6986,
+      "step": 15584
+    },
+    {
+      "epoch": 0.8577797347129726,
+      "grad_norm": 0.8557218909263611,
+      "learning_rate": 6.131854363391859e-06,
+      "loss": 0.795,
+      "step": 15585
+    },
+    {
+      "epoch": 0.8578347735153283,
+      "grad_norm": 0.6823254823684692,
+      "learning_rate": 6.1314321448021825e-06,
+      "loss": 0.7349,
+      "step": 15586
+    },
+    {
+      "epoch": 0.857889812317684,
+      "grad_norm": 0.772792637348175,
+      "learning_rate": 6.131009917708965e-06,
+      "loss": 0.7547,
+      "step": 15587
+    },
+    {
+      "epoch": 0.8579448511200396,
+      "grad_norm": 0.7231488227844238,
+      "learning_rate": 6.130587682115379e-06,
+      "loss": 0.7997,
+      "step": 15588
+    },
+    {
+      "epoch": 0.8579998899223953,
+      "grad_norm": 0.6683667898178101,
+      "learning_rate": 6.130165438024598e-06,
+      "loss": 0.7863,
+      "step": 15589
+    },
+    {
+      "epoch": 0.858054928724751,
+      "grad_norm": 0.6588496565818787,
+      "learning_rate": 6.129743185439796e-06,
+      "loss": 0.7859,
+      "step": 15590
+    },
+    {
+      "epoch": 0.8581099675271067,
+      "grad_norm": 0.6130164861679077,
+      "learning_rate": 6.129320924364147e-06,
+      "loss": 0.7042,
+      "step": 15591
+    },
+    {
+      "epoch": 0.8581650063294622,
+      "grad_norm": 0.610054612159729,
+      "learning_rate": 6.128898654800824e-06,
+      "loss": 0.6645,
+      "step": 15592
+    },
+    {
+      "epoch": 0.8582200451318179,
+      "grad_norm": 0.6974982023239136,
+      "learning_rate": 6.128476376753002e-06,
+      "loss": 0.8146,
+      "step": 15593
+    },
+    {
+      "epoch": 0.8582750839341736,
+      "grad_norm": 0.7313922047615051,
+      "learning_rate": 6.128054090223853e-06,
+      "loss": 0.7055,
+      "step": 15594
+    },
+    {
+      "epoch": 0.8583301227365292,
+      "grad_norm": 0.7004476189613342,
+      "learning_rate": 6.12763179521655e-06,
+      "loss": 0.7848,
+      "step": 15595
+    },
+    {
+      "epoch": 0.8583851615388849,
+      "grad_norm": 0.6916295289993286,
+      "learning_rate": 6.127209491734269e-06,
+      "loss": 0.7711,
+      "step": 15596
+    },
+    {
+      "epoch": 0.8584402003412406,
+      "grad_norm": 0.648551881313324,
+      "learning_rate": 6.126787179780185e-06,
+      "loss": 0.7098,
+      "step": 15597
+    },
+    {
+      "epoch": 0.8584952391435963,
+      "grad_norm": 0.6482384204864502,
+      "learning_rate": 6.126364859357469e-06,
+      "loss": 0.7596,
+      "step": 15598
+    },
+    {
+      "epoch": 0.8585502779459518,
+      "grad_norm": 0.7109531164169312,
+      "learning_rate": 6.125942530469297e-06,
+      "loss": 0.7539,
+      "step": 15599
+    },
+    {
+      "epoch": 0.8586053167483075,
+      "grad_norm": 0.6109207272529602,
+      "learning_rate": 6.125520193118841e-06,
+      "loss": 0.6764,
+      "step": 15600
+    },
+    {
+      "epoch": 0.8586603555506632,
+      "grad_norm": 0.7050053477287292,
+      "learning_rate": 6.125097847309277e-06,
+      "loss": 0.7304,
+      "step": 15601
+    },
+    {
+      "epoch": 0.8587153943530189,
+      "grad_norm": 0.653078019618988,
+      "learning_rate": 6.124675493043779e-06,
+      "loss": 0.6985,
+      "step": 15602
+    },
+    {
+      "epoch": 0.8587704331553745,
+      "grad_norm": 0.8391665816307068,
+      "learning_rate": 6.124253130325521e-06,
+      "loss": 0.8011,
+      "step": 15603
+    },
+    {
+      "epoch": 0.8588254719577302,
+      "grad_norm": 0.6978835463523865,
+      "learning_rate": 6.123830759157676e-06,
+      "loss": 0.6783,
+      "step": 15604
+    },
+    {
+      "epoch": 0.8588805107600859,
+      "grad_norm": 0.7796862125396729,
+      "learning_rate": 6.123408379543422e-06,
+      "loss": 0.7237,
+      "step": 15605
+    },
+    {
+      "epoch": 0.8589355495624416,
+      "grad_norm": 0.7162224054336548,
+      "learning_rate": 6.12298599148593e-06,
+      "loss": 0.8095,
+      "step": 15606
+    },
+    {
+      "epoch": 0.8589905883647971,
+      "grad_norm": 0.7654495239257812,
+      "learning_rate": 6.122563594988375e-06,
+      "loss": 0.7149,
+      "step": 15607
+    },
+    {
+      "epoch": 0.8590456271671528,
+      "grad_norm": 0.6186618804931641,
+      "learning_rate": 6.122141190053935e-06,
+      "loss": 0.6687,
+      "step": 15608
+    },
+    {
+      "epoch": 0.8591006659695085,
+      "grad_norm": 0.6669701337814331,
+      "learning_rate": 6.121718776685781e-06,
+      "loss": 0.7281,
+      "step": 15609
+    },
+    {
+      "epoch": 0.8591557047718642,
+      "grad_norm": 0.6581971645355225,
+      "learning_rate": 6.121296354887089e-06,
+      "loss": 0.7158,
+      "step": 15610
+    },
+    {
+      "epoch": 0.8592107435742198,
+      "grad_norm": 0.698243260383606,
+      "learning_rate": 6.120873924661034e-06,
+      "loss": 0.7894,
+      "step": 15611
+    },
+    {
+      "epoch": 0.8592657823765755,
+      "grad_norm": 0.6746723651885986,
+      "learning_rate": 6.120451486010791e-06,
+      "loss": 0.7993,
+      "step": 15612
+    },
+    {
+      "epoch": 0.8593208211789312,
+      "grad_norm": 0.727219820022583,
+      "learning_rate": 6.1200290389395335e-06,
+      "loss": 0.8446,
+      "step": 15613
+    },
+    {
+      "epoch": 0.8593758599812868,
+      "grad_norm": 0.7818809151649475,
+      "learning_rate": 6.119606583450438e-06,
+      "loss": 0.7167,
+      "step": 15614
+    },
+    {
+      "epoch": 0.8594308987836424,
+      "grad_norm": 0.692720890045166,
+      "learning_rate": 6.119184119546679e-06,
+      "loss": 0.705,
+      "step": 15615
+    },
+    {
+      "epoch": 0.8594859375859981,
+      "grad_norm": 0.6671997308731079,
+      "learning_rate": 6.1187616472314315e-06,
+      "loss": 0.8383,
+      "step": 15616
+    },
+    {
+      "epoch": 0.8595409763883538,
+      "grad_norm": 0.8043667674064636,
+      "learning_rate": 6.118339166507872e-06,
+      "loss": 0.6775,
+      "step": 15617
+    },
+    {
+      "epoch": 0.8595960151907095,
+      "grad_norm": 0.6313692927360535,
+      "learning_rate": 6.117916677379173e-06,
+      "loss": 0.6327,
+      "step": 15618
+    },
+    {
+      "epoch": 0.8596510539930651,
+      "grad_norm": 0.6770568490028381,
+      "learning_rate": 6.117494179848512e-06,
+      "loss": 0.741,
+      "step": 15619
+    },
+    {
+      "epoch": 0.8597060927954208,
+      "grad_norm": 0.6715630292892456,
+      "learning_rate": 6.117071673919064e-06,
+      "loss": 0.7105,
+      "step": 15620
+    },
+    {
+      "epoch": 0.8597611315977765,
+      "grad_norm": 0.618145763874054,
+      "learning_rate": 6.116649159594006e-06,
+      "loss": 0.7316,
+      "step": 15621
+    },
+    {
+      "epoch": 0.8598161704001321,
+      "grad_norm": 0.7127259969711304,
+      "learning_rate": 6.11622663687651e-06,
+      "loss": 0.7736,
+      "step": 15622
+    },
+    {
+      "epoch": 0.8598712092024877,
+      "grad_norm": 0.6675243377685547,
+      "learning_rate": 6.115804105769754e-06,
+      "loss": 0.6747,
+      "step": 15623
+    },
+    {
+      "epoch": 0.8599262480048434,
+      "grad_norm": 0.7965354323387146,
+      "learning_rate": 6.115381566276912e-06,
+      "loss": 0.7524,
+      "step": 15624
+    },
+    {
+      "epoch": 0.8599812868071991,
+      "grad_norm": 0.5921181440353394,
+      "learning_rate": 6.114959018401163e-06,
+      "loss": 0.679,
+      "step": 15625
+    },
+    {
+      "epoch": 0.8600363256095548,
+      "grad_norm": 0.635802149772644,
+      "learning_rate": 6.1145364621456795e-06,
+      "loss": 0.699,
+      "step": 15626
+    },
+    {
+      "epoch": 0.8600913644119104,
+      "grad_norm": 0.7159842252731323,
+      "learning_rate": 6.114113897513636e-06,
+      "loss": 0.7112,
+      "step": 15627
+    },
+    {
+      "epoch": 0.860146403214266,
+      "grad_norm": 0.7100176215171814,
+      "learning_rate": 6.113691324508213e-06,
+      "loss": 0.7459,
+      "step": 15628
+    },
+    {
+      "epoch": 0.8602014420166217,
+      "grad_norm": 0.6484093070030212,
+      "learning_rate": 6.113268743132583e-06,
+      "loss": 0.6779,
+      "step": 15629
+    },
+    {
+      "epoch": 0.8602564808189774,
+      "grad_norm": 0.6825945377349854,
+      "learning_rate": 6.112846153389924e-06,
+      "loss": 0.7607,
+      "step": 15630
+    },
+    {
+      "epoch": 0.860311519621333,
+      "grad_norm": 0.7553657293319702,
+      "learning_rate": 6.112423555283411e-06,
+      "loss": 0.6945,
+      "step": 15631
+    },
+    {
+      "epoch": 0.8603665584236887,
+      "grad_norm": 0.7892605662345886,
+      "learning_rate": 6.11200094881622e-06,
+      "loss": 0.767,
+      "step": 15632
+    },
+    {
+      "epoch": 0.8604215972260444,
+      "grad_norm": 0.6485433578491211,
+      "learning_rate": 6.111578333991528e-06,
+      "loss": 0.7302,
+      "step": 15633
+    },
+    {
+      "epoch": 0.8604766360284001,
+      "grad_norm": 0.6713895201683044,
+      "learning_rate": 6.111155710812511e-06,
+      "loss": 0.774,
+      "step": 15634
+    },
+    {
+      "epoch": 0.8605316748307557,
+      "grad_norm": 0.9890132546424866,
+      "learning_rate": 6.110733079282345e-06,
+      "loss": 0.7549,
+      "step": 15635
+    },
+    {
+      "epoch": 0.8605867136331113,
+      "grad_norm": 0.6421818137168884,
+      "learning_rate": 6.110310439404206e-06,
+      "loss": 0.7004,
+      "step": 15636
+    },
+    {
+      "epoch": 0.860641752435467,
+      "grad_norm": 0.6384093165397644,
+      "learning_rate": 6.109887791181272e-06,
+      "loss": 0.7465,
+      "step": 15637
+    },
+    {
+      "epoch": 0.8606967912378226,
+      "grad_norm": 0.7991462349891663,
+      "learning_rate": 6.109465134616717e-06,
+      "loss": 0.8041,
+      "step": 15638
+    },
+    {
+      "epoch": 0.8607518300401783,
+      "grad_norm": 0.661189615726471,
+      "learning_rate": 6.1090424697137185e-06,
+      "loss": 0.7717,
+      "step": 15639
+    },
+    {
+      "epoch": 0.860806868842534,
+      "grad_norm": 0.6952805519104004,
+      "learning_rate": 6.108619796475455e-06,
+      "loss": 0.7149,
+      "step": 15640
+    },
+    {
+      "epoch": 0.8608619076448897,
+      "grad_norm": 0.7330671548843384,
+      "learning_rate": 6.108197114905102e-06,
+      "loss": 0.7229,
+      "step": 15641
+    },
+    {
+      "epoch": 0.8609169464472453,
+      "grad_norm": 0.6831181049346924,
+      "learning_rate": 6.107774425005836e-06,
+      "loss": 0.6937,
+      "step": 15642
+    },
+    {
+      "epoch": 0.860971985249601,
+      "grad_norm": 0.7261425852775574,
+      "learning_rate": 6.107351726780833e-06,
+      "loss": 0.7963,
+      "step": 15643
+    },
+    {
+      "epoch": 0.8610270240519566,
+      "grad_norm": 0.6796271800994873,
+      "learning_rate": 6.106929020233272e-06,
+      "loss": 0.7785,
+      "step": 15644
+    },
+    {
+      "epoch": 0.8610820628543123,
+      "grad_norm": 0.6772015690803528,
+      "learning_rate": 6.106506305366328e-06,
+      "loss": 0.7732,
+      "step": 15645
+    },
+    {
+      "epoch": 0.8611371016566679,
+      "grad_norm": 0.6153992414474487,
+      "learning_rate": 6.10608358218318e-06,
+      "loss": 0.6361,
+      "step": 15646
+    },
+    {
+      "epoch": 0.8611921404590236,
+      "grad_norm": 0.9580141305923462,
+      "learning_rate": 6.105660850687003e-06,
+      "loss": 0.7178,
+      "step": 15647
+    },
+    {
+      "epoch": 0.8612471792613793,
+      "grad_norm": 0.8536281585693359,
+      "learning_rate": 6.105238110880975e-06,
+      "loss": 0.6689,
+      "step": 15648
+    },
+    {
+      "epoch": 0.861302218063735,
+      "grad_norm": 0.6578275561332703,
+      "learning_rate": 6.104815362768274e-06,
+      "loss": 0.7474,
+      "step": 15649
+    },
+    {
+      "epoch": 0.8613572568660905,
+      "grad_norm": 0.7298864126205444,
+      "learning_rate": 6.104392606352075e-06,
+      "loss": 0.7134,
+      "step": 15650
+    },
+    {
+      "epoch": 0.8614122956684462,
+      "grad_norm": 1.0637938976287842,
+      "learning_rate": 6.103969841635557e-06,
+      "loss": 0.8614,
+      "step": 15651
+    },
+    {
+      "epoch": 0.8614673344708019,
+      "grad_norm": 0.6678902506828308,
+      "learning_rate": 6.103547068621898e-06,
+      "loss": 0.6736,
+      "step": 15652
+    },
+    {
+      "epoch": 0.8615223732731576,
+      "grad_norm": 0.9442873001098633,
+      "learning_rate": 6.103124287314275e-06,
+      "loss": 0.8002,
+      "step": 15653
+    },
+    {
+      "epoch": 0.8615774120755132,
+      "grad_norm": 0.7156786322593689,
+      "learning_rate": 6.102701497715864e-06,
+      "loss": 0.8764,
+      "step": 15654
+    },
+    {
+      "epoch": 0.8616324508778689,
+      "grad_norm": 0.7954290509223938,
+      "learning_rate": 6.102278699829843e-06,
+      "loss": 0.7308,
+      "step": 15655
+    },
+    {
+      "epoch": 0.8616874896802246,
+      "grad_norm": 0.7544524073600769,
+      "learning_rate": 6.101855893659392e-06,
+      "loss": 0.7064,
+      "step": 15656
+    },
+    {
+      "epoch": 0.8617425284825803,
+      "grad_norm": 0.652656078338623,
+      "learning_rate": 6.101433079207687e-06,
+      "loss": 0.7264,
+      "step": 15657
+    },
+    {
+      "epoch": 0.8617975672849358,
+      "grad_norm": 0.6478716135025024,
+      "learning_rate": 6.101010256477906e-06,
+      "loss": 0.6265,
+      "step": 15658
+    },
+    {
+      "epoch": 0.8618526060872915,
+      "grad_norm": 0.5916007161140442,
+      "learning_rate": 6.1005874254732256e-06,
+      "loss": 0.6485,
+      "step": 15659
+    },
+    {
+      "epoch": 0.8619076448896472,
+      "grad_norm": 0.7353591322898865,
+      "learning_rate": 6.1001645861968264e-06,
+      "loss": 0.759,
+      "step": 15660
+    },
+    {
+      "epoch": 0.8619626836920029,
+      "grad_norm": 0.7352280020713806,
+      "learning_rate": 6.099741738651883e-06,
+      "loss": 0.7392,
+      "step": 15661
+    },
+    {
+      "epoch": 0.8620177224943585,
+      "grad_norm": 0.6027048230171204,
+      "learning_rate": 6.099318882841576e-06,
+      "loss": 0.7039,
+      "step": 15662
+    },
+    {
+      "epoch": 0.8620727612967142,
+      "grad_norm": 0.6907329559326172,
+      "learning_rate": 6.09889601876908e-06,
+      "loss": 0.8164,
+      "step": 15663
+    },
+    {
+      "epoch": 0.8621278000990699,
+      "grad_norm": 0.7133687138557434,
+      "learning_rate": 6.098473146437579e-06,
+      "loss": 0.791,
+      "step": 15664
+    },
+    {
+      "epoch": 0.8621828389014256,
+      "grad_norm": 0.7229570746421814,
+      "learning_rate": 6.098050265850246e-06,
+      "loss": 0.7839,
+      "step": 15665
+    },
+    {
+      "epoch": 0.8622378777037811,
+      "grad_norm": 0.7066009640693665,
+      "learning_rate": 6.097627377010262e-06,
+      "loss": 0.6816,
+      "step": 15666
+    },
+    {
+      "epoch": 0.8622929165061368,
+      "grad_norm": 0.7801152467727661,
+      "learning_rate": 6.097204479920804e-06,
+      "loss": 0.7402,
+      "step": 15667
+    },
+    {
+      "epoch": 0.8623479553084925,
+      "grad_norm": 0.6149227023124695,
+      "learning_rate": 6.096781574585051e-06,
+      "loss": 0.6964,
+      "step": 15668
+    },
+    {
+      "epoch": 0.8624029941108482,
+      "grad_norm": 0.6978023648262024,
+      "learning_rate": 6.096358661006181e-06,
+      "loss": 0.6982,
+      "step": 15669
+    },
+    {
+      "epoch": 0.8624580329132038,
+      "grad_norm": 0.6561335325241089,
+      "learning_rate": 6.095935739187373e-06,
+      "loss": 0.74,
+      "step": 15670
+    },
+    {
+      "epoch": 0.8625130717155595,
+      "grad_norm": 0.7627743482589722,
+      "learning_rate": 6.0955128091318065e-06,
+      "loss": 0.6886,
+      "step": 15671
+    },
+    {
+      "epoch": 0.8625681105179152,
+      "grad_norm": 0.919551432132721,
+      "learning_rate": 6.095089870842657e-06,
+      "loss": 0.7453,
+      "step": 15672
+    },
+    {
+      "epoch": 0.8626231493202708,
+      "grad_norm": 0.732641875743866,
+      "learning_rate": 6.094666924323107e-06,
+      "loss": 0.7502,
+      "step": 15673
+    },
+    {
+      "epoch": 0.8626781881226264,
+      "grad_norm": 0.7035035490989685,
+      "learning_rate": 6.094243969576332e-06,
+      "loss": 0.7825,
+      "step": 15674
+    },
+    {
+      "epoch": 0.8627332269249821,
+      "grad_norm": 0.648766279220581,
+      "learning_rate": 6.093821006605513e-06,
+      "loss": 0.6771,
+      "step": 15675
+    },
+    {
+      "epoch": 0.8627882657273378,
+      "grad_norm": 0.6031193137168884,
+      "learning_rate": 6.093398035413828e-06,
+      "loss": 0.6246,
+      "step": 15676
+    },
+    {
+      "epoch": 0.8628433045296935,
+      "grad_norm": 0.76170414686203,
+      "learning_rate": 6.0929750560044555e-06,
+      "loss": 0.713,
+      "step": 15677
+    },
+    {
+      "epoch": 0.8628983433320491,
+      "grad_norm": 1.023805022239685,
+      "learning_rate": 6.092552068380575e-06,
+      "loss": 0.6824,
+      "step": 15678
+    },
+    {
+      "epoch": 0.8629533821344048,
+      "grad_norm": 0.7333651185035706,
+      "learning_rate": 6.092129072545366e-06,
+      "loss": 0.7213,
+      "step": 15679
+    },
+    {
+      "epoch": 0.8630084209367604,
+      "grad_norm": 0.6620833873748779,
+      "learning_rate": 6.091706068502007e-06,
+      "loss": 0.7436,
+      "step": 15680
+    },
+    {
+      "epoch": 0.863063459739116,
+      "grad_norm": 0.5971367359161377,
+      "learning_rate": 6.091283056253679e-06,
+      "loss": 0.6774,
+      "step": 15681
+    },
+    {
+      "epoch": 0.8631184985414717,
+      "grad_norm": 0.6435208320617676,
+      "learning_rate": 6.090860035803558e-06,
+      "loss": 0.7556,
+      "step": 15682
+    },
+    {
+      "epoch": 0.8631735373438274,
+      "grad_norm": 0.6666582822799683,
+      "learning_rate": 6.090437007154824e-06,
+      "loss": 0.7533,
+      "step": 15683
+    },
+    {
+      "epoch": 0.8632285761461831,
+      "grad_norm": 0.665928840637207,
+      "learning_rate": 6.09001397031066e-06,
+      "loss": 0.7325,
+      "step": 15684
+    },
+    {
+      "epoch": 0.8632836149485387,
+      "grad_norm": 0.6638591885566711,
+      "learning_rate": 6.0895909252742414e-06,
+      "loss": 0.8256,
+      "step": 15685
+    },
+    {
+      "epoch": 0.8633386537508944,
+      "grad_norm": 0.6556721925735474,
+      "learning_rate": 6.089167872048749e-06,
+      "loss": 0.728,
+      "step": 15686
+    },
+    {
+      "epoch": 0.86339369255325,
+      "grad_norm": 0.6327305436134338,
+      "learning_rate": 6.088744810637361e-06,
+      "loss": 0.7584,
+      "step": 15687
+    },
+    {
+      "epoch": 0.8634487313556057,
+      "grad_norm": 0.676216185092926,
+      "learning_rate": 6.088321741043262e-06,
+      "loss": 0.7868,
+      "step": 15688
+    },
+    {
+      "epoch": 0.8635037701579613,
+      "grad_norm": 0.646700918674469,
+      "learning_rate": 6.0878986632696255e-06,
+      "loss": 0.7248,
+      "step": 15689
+    },
+    {
+      "epoch": 0.863558808960317,
+      "grad_norm": 0.6748735308647156,
+      "learning_rate": 6.087475577319635e-06,
+      "loss": 0.7657,
+      "step": 15690
+    },
+    {
+      "epoch": 0.8636138477626727,
+      "grad_norm": 0.6363335251808167,
+      "learning_rate": 6.087052483196467e-06,
+      "loss": 0.7273,
+      "step": 15691
+    },
+    {
+      "epoch": 0.8636688865650284,
+      "grad_norm": 0.6166467666625977,
+      "learning_rate": 6.086629380903305e-06,
+      "loss": 0.6642,
+      "step": 15692
+    },
+    {
+      "epoch": 0.863723925367384,
+      "grad_norm": 1.3258485794067383,
+      "learning_rate": 6.086206270443328e-06,
+      "loss": 0.7227,
+      "step": 15693
+    },
+    {
+      "epoch": 0.8637789641697396,
+      "grad_norm": 0.8923795223236084,
+      "learning_rate": 6.085783151819716e-06,
+      "loss": 0.7513,
+      "step": 15694
+    },
+    {
+      "epoch": 0.8638340029720953,
+      "grad_norm": 0.7227154970169067,
+      "learning_rate": 6.085360025035647e-06,
+      "loss": 0.7078,
+      "step": 15695
+    },
+    {
+      "epoch": 0.863889041774451,
+      "grad_norm": 0.6465400457382202,
+      "learning_rate": 6.084936890094303e-06,
+      "loss": 0.7541,
+      "step": 15696
+    },
+    {
+      "epoch": 0.8639440805768066,
+      "grad_norm": 0.6628104448318481,
+      "learning_rate": 6.084513746998865e-06,
+      "loss": 0.7121,
+      "step": 15697
+    },
+    {
+      "epoch": 0.8639991193791623,
+      "grad_norm": 0.6723392605781555,
+      "learning_rate": 6.08409059575251e-06,
+      "loss": 0.7086,
+      "step": 15698
+    },
+    {
+      "epoch": 0.864054158181518,
+      "grad_norm": 0.7443264126777649,
+      "learning_rate": 6.08366743635842e-06,
+      "loss": 0.7098,
+      "step": 15699
+    },
+    {
+      "epoch": 0.8641091969838737,
+      "grad_norm": 0.7792028188705444,
+      "learning_rate": 6.083244268819777e-06,
+      "loss": 0.8472,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8641642357862293,
+      "grad_norm": 0.7211549878120422,
+      "learning_rate": 6.08282109313976e-06,
+      "loss": 0.7917,
+      "step": 15701
+    },
+    {
+      "epoch": 0.8642192745885849,
+      "grad_norm": 0.6670592427253723,
+      "learning_rate": 6.082397909321549e-06,
+      "loss": 0.7758,
+      "step": 15702
+    },
+    {
+      "epoch": 0.8642743133909406,
+      "grad_norm": 0.8279144167900085,
+      "learning_rate": 6.0819747173683255e-06,
+      "loss": 0.7355,
+      "step": 15703
+    },
+    {
+      "epoch": 0.8643293521932963,
+      "grad_norm": 0.7409362196922302,
+      "learning_rate": 6.081551517283269e-06,
+      "loss": 0.7283,
+      "step": 15704
+    },
+    {
+      "epoch": 0.8643843909956519,
+      "grad_norm": 0.6700742840766907,
+      "learning_rate": 6.081128309069562e-06,
+      "loss": 0.6555,
+      "step": 15705
+    },
+    {
+      "epoch": 0.8644394297980076,
+      "grad_norm": 0.7364388108253479,
+      "learning_rate": 6.080705092730383e-06,
+      "loss": 0.6652,
+      "step": 15706
+    },
+    {
+      "epoch": 0.8644944686003633,
+      "grad_norm": 0.778404176235199,
+      "learning_rate": 6.080281868268913e-06,
+      "loss": 0.7774,
+      "step": 15707
+    },
+    {
+      "epoch": 0.864549507402719,
+      "grad_norm": 0.6663825511932373,
+      "learning_rate": 6.079858635688336e-06,
+      "loss": 0.7665,
+      "step": 15708
+    },
+    {
+      "epoch": 0.8646045462050745,
+      "grad_norm": 0.7061408758163452,
+      "learning_rate": 6.079435394991829e-06,
+      "loss": 0.7824,
+      "step": 15709
+    },
+    {
+      "epoch": 0.8646595850074302,
+      "grad_norm": 0.6537507176399231,
+      "learning_rate": 6.079012146182576e-06,
+      "loss": 0.7313,
+      "step": 15710
+    },
+    {
+      "epoch": 0.8647146238097859,
+      "grad_norm": 0.6154575943946838,
+      "learning_rate": 6.078588889263754e-06,
+      "loss": 0.7066,
+      "step": 15711
+    },
+    {
+      "epoch": 0.8647696626121416,
+      "grad_norm": 0.659093976020813,
+      "learning_rate": 6.078165624238548e-06,
+      "loss": 0.7495,
+      "step": 15712
+    },
+    {
+      "epoch": 0.8648247014144972,
+      "grad_norm": 0.677669107913971,
+      "learning_rate": 6.077742351110138e-06,
+      "loss": 0.7072,
+      "step": 15713
+    },
+    {
+      "epoch": 0.8648797402168529,
+      "grad_norm": 0.7204097509384155,
+      "learning_rate": 6.077319069881705e-06,
+      "loss": 0.7181,
+      "step": 15714
+    },
+    {
+      "epoch": 0.8649347790192086,
+      "grad_norm": 0.6903330683708191,
+      "learning_rate": 6.076895780556429e-06,
+      "loss": 0.8565,
+      "step": 15715
+    },
+    {
+      "epoch": 0.8649898178215643,
+      "grad_norm": 0.8147342205047607,
+      "learning_rate": 6.076472483137493e-06,
+      "loss": 0.6916,
+      "step": 15716
+    },
+    {
+      "epoch": 0.8650448566239198,
+      "grad_norm": 0.7021569013595581,
+      "learning_rate": 6.076049177628079e-06,
+      "loss": 0.6893,
+      "step": 15717
+    },
+    {
+      "epoch": 0.8650998954262755,
+      "grad_norm": 0.6534682512283325,
+      "learning_rate": 6.075625864031368e-06,
+      "loss": 0.6313,
+      "step": 15718
+    },
+    {
+      "epoch": 0.8651549342286312,
+      "grad_norm": 0.7883698344230652,
+      "learning_rate": 6.07520254235054e-06,
+      "loss": 0.7204,
+      "step": 15719
+    },
+    {
+      "epoch": 0.8652099730309869,
+      "grad_norm": 0.6255857944488525,
+      "learning_rate": 6.074779212588777e-06,
+      "loss": 0.7137,
+      "step": 15720
+    },
+    {
+      "epoch": 0.8652650118333425,
+      "grad_norm": 0.7278919816017151,
+      "learning_rate": 6.074355874749261e-06,
+      "loss": 0.8003,
+      "step": 15721
+    },
+    {
+      "epoch": 0.8653200506356982,
+      "grad_norm": 0.7809221744537354,
+      "learning_rate": 6.073932528835176e-06,
+      "loss": 0.8652,
+      "step": 15722
+    },
+    {
+      "epoch": 0.8653750894380539,
+      "grad_norm": 0.6781452894210815,
+      "learning_rate": 6.0735091748496985e-06,
+      "loss": 0.6111,
+      "step": 15723
+    },
+    {
+      "epoch": 0.8654301282404094,
+      "grad_norm": 0.6400741934776306,
+      "learning_rate": 6.073085812796015e-06,
+      "loss": 0.7126,
+      "step": 15724
+    },
+    {
+      "epoch": 0.8654851670427651,
+      "grad_norm": 0.6753132343292236,
+      "learning_rate": 6.072662442677305e-06,
+      "loss": 0.7389,
+      "step": 15725
+    },
+    {
+      "epoch": 0.8655402058451208,
+      "grad_norm": 0.6688135266304016,
+      "learning_rate": 6.072239064496752e-06,
+      "loss": 0.6356,
+      "step": 15726
+    },
+    {
+      "epoch": 0.8655952446474765,
+      "grad_norm": 0.664271354675293,
+      "learning_rate": 6.0718156782575365e-06,
+      "loss": 0.6949,
+      "step": 15727
+    },
+    {
+      "epoch": 0.8656502834498321,
+      "grad_norm": 0.6760862469673157,
+      "learning_rate": 6.071392283962843e-06,
+      "loss": 0.7279,
+      "step": 15728
+    },
+    {
+      "epoch": 0.8657053222521878,
+      "grad_norm": 0.6911706924438477,
+      "learning_rate": 6.07096888161585e-06,
+      "loss": 0.8132,
+      "step": 15729
+    },
+    {
+      "epoch": 0.8657603610545435,
+      "grad_norm": 0.7274359464645386,
+      "learning_rate": 6.070545471219743e-06,
+      "loss": 0.7894,
+      "step": 15730
+    },
+    {
+      "epoch": 0.8658153998568991,
+      "grad_norm": 0.7742472290992737,
+      "learning_rate": 6.070122052777703e-06,
+      "loss": 0.8057,
+      "step": 15731
+    },
+    {
+      "epoch": 0.8658704386592547,
+      "grad_norm": 0.8446773290634155,
+      "learning_rate": 6.06969862629291e-06,
+      "loss": 0.7816,
+      "step": 15732
+    },
+    {
+      "epoch": 0.8659254774616104,
+      "grad_norm": 0.669518232345581,
+      "learning_rate": 6.06927519176855e-06,
+      "loss": 0.696,
+      "step": 15733
+    },
+    {
+      "epoch": 0.8659805162639661,
+      "grad_norm": 0.6845564842224121,
+      "learning_rate": 6.068851749207803e-06,
+      "loss": 0.7486,
+      "step": 15734
+    },
+    {
+      "epoch": 0.8660355550663218,
+      "grad_norm": 0.6650436520576477,
+      "learning_rate": 6.068428298613853e-06,
+      "loss": 0.7215,
+      "step": 15735
+    },
+    {
+      "epoch": 0.8660905938686774,
+      "grad_norm": 0.67397540807724,
+      "learning_rate": 6.068004839989881e-06,
+      "loss": 0.7458,
+      "step": 15736
+    },
+    {
+      "epoch": 0.8661456326710331,
+      "grad_norm": 0.7140672206878662,
+      "learning_rate": 6.067581373339072e-06,
+      "loss": 0.8213,
+      "step": 15737
+    },
+    {
+      "epoch": 0.8662006714733888,
+      "grad_norm": 0.8632931113243103,
+      "learning_rate": 6.067157898664606e-06,
+      "loss": 0.8109,
+      "step": 15738
+    },
+    {
+      "epoch": 0.8662557102757444,
+      "grad_norm": 0.6106804013252258,
+      "learning_rate": 6.066734415969669e-06,
+      "loss": 0.7183,
+      "step": 15739
+    },
+    {
+      "epoch": 0.8663107490781,
+      "grad_norm": 0.8055095672607422,
+      "learning_rate": 6.066310925257438e-06,
+      "loss": 0.7871,
+      "step": 15740
+    },
+    {
+      "epoch": 0.8663657878804557,
+      "grad_norm": 0.6310189366340637,
+      "learning_rate": 6.065887426531102e-06,
+      "loss": 0.5873,
+      "step": 15741
+    },
+    {
+      "epoch": 0.8664208266828114,
+      "grad_norm": 0.6704412698745728,
+      "learning_rate": 6.065463919793842e-06,
+      "loss": 0.6838,
+      "step": 15742
+    },
+    {
+      "epoch": 0.8664758654851671,
+      "grad_norm": 0.6292148232460022,
+      "learning_rate": 6.06504040504884e-06,
+      "loss": 0.6886,
+      "step": 15743
+    },
+    {
+      "epoch": 0.8665309042875227,
+      "grad_norm": 0.8556584715843201,
+      "learning_rate": 6.064616882299277e-06,
+      "loss": 0.8967,
+      "step": 15744
+    },
+    {
+      "epoch": 0.8665859430898784,
+      "grad_norm": 0.6956119537353516,
+      "learning_rate": 6.064193351548341e-06,
+      "loss": 0.7444,
+      "step": 15745
+    },
+    {
+      "epoch": 0.866640981892234,
+      "grad_norm": 1.01414954662323,
+      "learning_rate": 6.063769812799212e-06,
+      "loss": 0.9216,
+      "step": 15746
+    },
+    {
+      "epoch": 0.8666960206945897,
+      "grad_norm": 0.6685424447059631,
+      "learning_rate": 6.063346266055073e-06,
+      "loss": 0.6795,
+      "step": 15747
+    },
+    {
+      "epoch": 0.8667510594969453,
+      "grad_norm": 0.6735886335372925,
+      "learning_rate": 6.062922711319108e-06,
+      "loss": 0.6805,
+      "step": 15748
+    },
+    {
+      "epoch": 0.866806098299301,
+      "grad_norm": 0.6536576747894287,
+      "learning_rate": 6.062499148594502e-06,
+      "loss": 0.6575,
+      "step": 15749
+    },
+    {
+      "epoch": 0.8668611371016567,
+      "grad_norm": 0.6739212870597839,
+      "learning_rate": 6.062075577884437e-06,
+      "loss": 0.6704,
+      "step": 15750
+    },
+    {
+      "epoch": 0.8669161759040124,
+      "grad_norm": 0.73397296667099,
+      "learning_rate": 6.061651999192094e-06,
+      "loss": 0.7892,
+      "step": 15751
+    },
+    {
+      "epoch": 0.866971214706368,
+      "grad_norm": 0.7974724769592285,
+      "learning_rate": 6.06122841252066e-06,
+      "loss": 0.7133,
+      "step": 15752
+    },
+    {
+      "epoch": 0.8670262535087236,
+      "grad_norm": 0.6199150681495667,
+      "learning_rate": 6.060804817873317e-06,
+      "loss": 0.765,
+      "step": 15753
+    },
+    {
+      "epoch": 0.8670812923110793,
+      "grad_norm": 0.709783673286438,
+      "learning_rate": 6.060381215253251e-06,
+      "loss": 0.7332,
+      "step": 15754
+    },
+    {
+      "epoch": 0.867136331113435,
+      "grad_norm": 0.6947084069252014,
+      "learning_rate": 6.059957604663642e-06,
+      "loss": 0.8224,
+      "step": 15755
+    },
+    {
+      "epoch": 0.8671913699157906,
+      "grad_norm": 0.9439684152603149,
+      "learning_rate": 6.059533986107674e-06,
+      "loss": 0.8347,
+      "step": 15756
+    },
+    {
+      "epoch": 0.8672464087181463,
+      "grad_norm": 0.806992769241333,
+      "learning_rate": 6.059110359588534e-06,
+      "loss": 0.8055,
+      "step": 15757
+    },
+    {
+      "epoch": 0.867301447520502,
+      "grad_norm": 0.659092128276825,
+      "learning_rate": 6.058686725109404e-06,
+      "loss": 0.6972,
+      "step": 15758
+    },
+    {
+      "epoch": 0.8673564863228577,
+      "grad_norm": 0.7345813512802124,
+      "learning_rate": 6.058263082673468e-06,
+      "loss": 0.8044,
+      "step": 15759
+    },
+    {
+      "epoch": 0.8674115251252132,
+      "grad_norm": 0.7216777205467224,
+      "learning_rate": 6.057839432283908e-06,
+      "loss": 0.7816,
+      "step": 15760
+    },
+    {
+      "epoch": 0.8674665639275689,
+      "grad_norm": 0.6828186511993408,
+      "learning_rate": 6.0574157739439125e-06,
+      "loss": 0.7534,
+      "step": 15761
+    },
+    {
+      "epoch": 0.8675216027299246,
+      "grad_norm": 0.7324418425559998,
+      "learning_rate": 6.0569921076566615e-06,
+      "loss": 0.7476,
+      "step": 15762
+    },
+    {
+      "epoch": 0.8675766415322803,
+      "grad_norm": 0.5894229412078857,
+      "learning_rate": 6.056568433425342e-06,
+      "loss": 0.6667,
+      "step": 15763
+    },
+    {
+      "epoch": 0.8676316803346359,
+      "grad_norm": 0.6743035912513733,
+      "learning_rate": 6.056144751253135e-06,
+      "loss": 0.6765,
+      "step": 15764
+    },
+    {
+      "epoch": 0.8676867191369916,
+      "grad_norm": 0.6885803937911987,
+      "learning_rate": 6.055721061143229e-06,
+      "loss": 0.6954,
+      "step": 15765
+    },
+    {
+      "epoch": 0.8677417579393473,
+      "grad_norm": 0.6543543338775635,
+      "learning_rate": 6.055297363098806e-06,
+      "loss": 0.6277,
+      "step": 15766
+    },
+    {
+      "epoch": 0.8677967967417028,
+      "grad_norm": 0.7671917080879211,
+      "learning_rate": 6.054873657123049e-06,
+      "loss": 0.7575,
+      "step": 15767
+    },
+    {
+      "epoch": 0.8678518355440585,
+      "grad_norm": 0.7491669654846191,
+      "learning_rate": 6.054449943219144e-06,
+      "loss": 0.727,
+      "step": 15768
+    },
+    {
+      "epoch": 0.8679068743464142,
+      "grad_norm": 0.7161419987678528,
+      "learning_rate": 6.0540262213902765e-06,
+      "loss": 0.7381,
+      "step": 15769
+    },
+    {
+      "epoch": 0.8679619131487699,
+      "grad_norm": 0.7061475515365601,
+      "learning_rate": 6.05360249163963e-06,
+      "loss": 0.7831,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8680169519511255,
+      "grad_norm": 0.7481213212013245,
+      "learning_rate": 6.053178753970389e-06,
+      "loss": 0.7235,
+      "step": 15771
+    },
+    {
+      "epoch": 0.8680719907534812,
+      "grad_norm": 0.6475214958190918,
+      "learning_rate": 6.052755008385736e-06,
+      "loss": 0.6864,
+      "step": 15772
+    },
+    {
+      "epoch": 0.8681270295558369,
+      "grad_norm": 0.7365770936012268,
+      "learning_rate": 6.052331254888862e-06,
+      "loss": 0.7746,
+      "step": 15773
+    },
+    {
+      "epoch": 0.8681820683581926,
+      "grad_norm": 0.6339132785797119,
+      "learning_rate": 6.0519074934829456e-06,
+      "loss": 0.7102,
+      "step": 15774
+    },
+    {
+      "epoch": 0.8682371071605481,
+      "grad_norm": 0.691531240940094,
+      "learning_rate": 6.0514837241711754e-06,
+      "loss": 0.7896,
+      "step": 15775
+    },
+    {
+      "epoch": 0.8682921459629038,
+      "grad_norm": 0.6793948411941528,
+      "learning_rate": 6.051059946956734e-06,
+      "loss": 0.6514,
+      "step": 15776
+    },
+    {
+      "epoch": 0.8683471847652595,
+      "grad_norm": 0.6301077008247375,
+      "learning_rate": 6.050636161842809e-06,
+      "loss": 0.6831,
+      "step": 15777
+    },
+    {
+      "epoch": 0.8684022235676152,
+      "grad_norm": 0.7680420875549316,
+      "learning_rate": 6.0502123688325835e-06,
+      "loss": 0.7504,
+      "step": 15778
+    },
+    {
+      "epoch": 0.8684572623699708,
+      "grad_norm": 0.6260972619056702,
+      "learning_rate": 6.0497885679292415e-06,
+      "loss": 0.7066,
+      "step": 15779
+    },
+    {
+      "epoch": 0.8685123011723265,
+      "grad_norm": 0.663060188293457,
+      "learning_rate": 6.04936475913597e-06,
+      "loss": 0.6634,
+      "step": 15780
+    },
+    {
+      "epoch": 0.8685673399746822,
+      "grad_norm": 0.6798335313796997,
+      "learning_rate": 6.048940942455954e-06,
+      "loss": 0.7055,
+      "step": 15781
+    },
+    {
+      "epoch": 0.8686223787770379,
+      "grad_norm": 0.7080284953117371,
+      "learning_rate": 6.048517117892379e-06,
+      "loss": 0.7606,
+      "step": 15782
+    },
+    {
+      "epoch": 0.8686774175793934,
+      "grad_norm": 0.67658931016922,
+      "learning_rate": 6.04809328544843e-06,
+      "loss": 0.7656,
+      "step": 15783
+    },
+    {
+      "epoch": 0.8687324563817491,
+      "grad_norm": 0.6667472720146179,
+      "learning_rate": 6.047669445127291e-06,
+      "loss": 0.7275,
+      "step": 15784
+    },
+    {
+      "epoch": 0.8687874951841048,
+      "grad_norm": 0.782096266746521,
+      "learning_rate": 6.04724559693215e-06,
+      "loss": 0.7524,
+      "step": 15785
+    },
+    {
+      "epoch": 0.8688425339864605,
+      "grad_norm": 0.7733443379402161,
+      "learning_rate": 6.046821740866192e-06,
+      "loss": 0.7022,
+      "step": 15786
+    },
+    {
+      "epoch": 0.8688975727888161,
+      "grad_norm": 0.6487871408462524,
+      "learning_rate": 6.046397876932602e-06,
+      "loss": 0.7077,
+      "step": 15787
+    },
+    {
+      "epoch": 0.8689526115911718,
+      "grad_norm": 0.6294482350349426,
+      "learning_rate": 6.045974005134564e-06,
+      "loss": 0.6974,
+      "step": 15788
+    },
+    {
+      "epoch": 0.8690076503935275,
+      "grad_norm": 0.6573933362960815,
+      "learning_rate": 6.045550125475268e-06,
+      "loss": 0.735,
+      "step": 15789
+    },
+    {
+      "epoch": 0.8690626891958831,
+      "grad_norm": 0.6794875264167786,
+      "learning_rate": 6.045126237957895e-06,
+      "loss": 0.7511,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8691177279982387,
+      "grad_norm": 0.687599778175354,
+      "learning_rate": 6.0447023425856345e-06,
+      "loss": 0.7164,
+      "step": 15791
+    },
+    {
+      "epoch": 0.8691727668005944,
+      "grad_norm": 0.6593008637428284,
+      "learning_rate": 6.04427843936167e-06,
+      "loss": 0.688,
+      "step": 15792
+    },
+    {
+      "epoch": 0.8692278056029501,
+      "grad_norm": 0.7226807475090027,
+      "learning_rate": 6.043854528289188e-06,
+      "loss": 0.7364,
+      "step": 15793
+    },
+    {
+      "epoch": 0.8692828444053058,
+      "grad_norm": 0.603318452835083,
+      "learning_rate": 6.043430609371375e-06,
+      "loss": 0.6933,
+      "step": 15794
+    },
+    {
+      "epoch": 0.8693378832076614,
+      "grad_norm": 0.8227141499519348,
+      "learning_rate": 6.043006682611416e-06,
+      "loss": 0.7039,
+      "step": 15795
+    },
+    {
+      "epoch": 0.869392922010017,
+      "grad_norm": 0.729284405708313,
+      "learning_rate": 6.042582748012499e-06,
+      "loss": 0.7288,
+      "step": 15796
+    },
+    {
+      "epoch": 0.8694479608123727,
+      "grad_norm": 0.8269371390342712,
+      "learning_rate": 6.042158805577809e-06,
+      "loss": 0.7419,
+      "step": 15797
+    },
+    {
+      "epoch": 0.8695029996147284,
+      "grad_norm": 0.6699450016021729,
+      "learning_rate": 6.0417348553105325e-06,
+      "loss": 0.7893,
+      "step": 15798
+    },
+    {
+      "epoch": 0.869558038417084,
+      "grad_norm": 0.7747042775154114,
+      "learning_rate": 6.041310897213856e-06,
+      "loss": 0.791,
+      "step": 15799
+    },
+    {
+      "epoch": 0.8696130772194397,
+      "grad_norm": 0.7503781318664551,
+      "learning_rate": 6.0408869312909645e-06,
+      "loss": 0.7204,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8696681160217954,
+      "grad_norm": 0.6733731627464294,
+      "learning_rate": 6.0404629575450464e-06,
+      "loss": 0.815,
+      "step": 15801
+    },
+    {
+      "epoch": 0.8697231548241511,
+      "grad_norm": 0.6925041079521179,
+      "learning_rate": 6.040038975979288e-06,
+      "loss": 0.8096,
+      "step": 15802
+    },
+    {
+      "epoch": 0.8697781936265067,
+      "grad_norm": 0.7510724067687988,
+      "learning_rate": 6.039614986596873e-06,
+      "loss": 0.7957,
+      "step": 15803
+    },
+    {
+      "epoch": 0.8698332324288623,
+      "grad_norm": 0.9631650447845459,
+      "learning_rate": 6.039190989400991e-06,
+      "loss": 0.7574,
+      "step": 15804
+    },
+    {
+      "epoch": 0.869888271231218,
+      "grad_norm": 0.7080852389335632,
+      "learning_rate": 6.0387669843948285e-06,
+      "loss": 0.7037,
+      "step": 15805
+    },
+    {
+      "epoch": 0.8699433100335737,
+      "grad_norm": 0.723419725894928,
+      "learning_rate": 6.03834297158157e-06,
+      "loss": 0.7424,
+      "step": 15806
+    },
+    {
+      "epoch": 0.8699983488359293,
+      "grad_norm": 0.6093000173568726,
+      "learning_rate": 6.037918950964404e-06,
+      "loss": 0.6754,
+      "step": 15807
+    },
+    {
+      "epoch": 0.870053387638285,
+      "grad_norm": 0.7614741921424866,
+      "learning_rate": 6.037494922546518e-06,
+      "loss": 0.6856,
+      "step": 15808
+    },
+    {
+      "epoch": 0.8701084264406407,
+      "grad_norm": 0.6535844802856445,
+      "learning_rate": 6.0370708863310965e-06,
+      "loss": 0.8201,
+      "step": 15809
+    },
+    {
+      "epoch": 0.8701634652429963,
+      "grad_norm": 0.724897027015686,
+      "learning_rate": 6.036646842321329e-06,
+      "loss": 0.7399,
+      "step": 15810
+    },
+    {
+      "epoch": 0.870218504045352,
+      "grad_norm": 0.7602331638336182,
+      "learning_rate": 6.036222790520401e-06,
+      "loss": 0.8233,
+      "step": 15811
+    },
+    {
+      "epoch": 0.8702735428477076,
+      "grad_norm": 0.7890536189079285,
+      "learning_rate": 6.035798730931498e-06,
+      "loss": 0.8473,
+      "step": 15812
+    },
+    {
+      "epoch": 0.8703285816500633,
+      "grad_norm": 0.7241165637969971,
+      "learning_rate": 6.035374663557813e-06,
+      "loss": 0.7298,
+      "step": 15813
+    },
+    {
+      "epoch": 0.8703836204524189,
+      "grad_norm": 0.6661847829818726,
+      "learning_rate": 6.034950588402526e-06,
+      "loss": 0.7461,
+      "step": 15814
+    },
+    {
+      "epoch": 0.8704386592547746,
+      "grad_norm": 0.6431320309638977,
+      "learning_rate": 6.034526505468829e-06,
+      "loss": 0.7436,
+      "step": 15815
+    },
+    {
+      "epoch": 0.8704936980571303,
+      "grad_norm": 1.122704267501831,
+      "learning_rate": 6.0341024147599055e-06,
+      "loss": 0.7378,
+      "step": 15816
+    },
+    {
+      "epoch": 0.870548736859486,
+      "grad_norm": 0.6391544938087463,
+      "learning_rate": 6.033678316278947e-06,
+      "loss": 0.7517,
+      "step": 15817
+    },
+    {
+      "epoch": 0.8706037756618416,
+      "grad_norm": 0.6522098183631897,
+      "learning_rate": 6.033254210029139e-06,
+      "loss": 0.7188,
+      "step": 15818
+    },
+    {
+      "epoch": 0.8706588144641972,
+      "grad_norm": 0.7638733386993408,
+      "learning_rate": 6.0328300960136686e-06,
+      "loss": 0.8032,
+      "step": 15819
+    },
+    {
+      "epoch": 0.8707138532665529,
+      "grad_norm": 0.6374132633209229,
+      "learning_rate": 6.032405974235722e-06,
+      "loss": 0.7292,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8707688920689086,
+      "grad_norm": 0.7061800360679626,
+      "learning_rate": 6.03198184469849e-06,
+      "loss": 0.7304,
+      "step": 15821
+    },
+    {
+      "epoch": 0.8708239308712642,
+      "grad_norm": 0.646089494228363,
+      "learning_rate": 6.031557707405159e-06,
+      "loss": 0.6762,
+      "step": 15822
+    },
+    {
+      "epoch": 0.8708789696736199,
+      "grad_norm": 0.8142202496528625,
+      "learning_rate": 6.031133562358916e-06,
+      "loss": 0.7789,
+      "step": 15823
+    },
+    {
+      "epoch": 0.8709340084759756,
+      "grad_norm": 0.6444084644317627,
+      "learning_rate": 6.030709409562949e-06,
+      "loss": 0.7383,
+      "step": 15824
+    },
+    {
+      "epoch": 0.8709890472783313,
+      "grad_norm": 0.8917344808578491,
+      "learning_rate": 6.030285249020448e-06,
+      "loss": 0.7527,
+      "step": 15825
+    },
+    {
+      "epoch": 0.8710440860806868,
+      "grad_norm": 0.6395692825317383,
+      "learning_rate": 6.029861080734597e-06,
+      "loss": 0.6923,
+      "step": 15826
+    },
+    {
+      "epoch": 0.8710991248830425,
+      "grad_norm": 0.6475933790206909,
+      "learning_rate": 6.029436904708586e-06,
+      "loss": 0.7495,
+      "step": 15827
+    },
+    {
+      "epoch": 0.8711541636853982,
+      "grad_norm": 0.7310789823532104,
+      "learning_rate": 6.029012720945602e-06,
+      "loss": 0.7541,
+      "step": 15828
+    },
+    {
+      "epoch": 0.8712092024877539,
+      "grad_norm": 0.8475071787834167,
+      "learning_rate": 6.028588529448835e-06,
+      "loss": 0.7397,
+      "step": 15829
+    },
+    {
+      "epoch": 0.8712642412901095,
+      "grad_norm": 0.6214048266410828,
+      "learning_rate": 6.028164330221471e-06,
+      "loss": 0.7365,
+      "step": 15830
+    },
+    {
+      "epoch": 0.8713192800924652,
+      "grad_norm": 0.6558026671409607,
+      "learning_rate": 6.0277401232667e-06,
+      "loss": 0.79,
+      "step": 15831
+    },
+    {
+      "epoch": 0.8713743188948209,
+      "grad_norm": 0.6652923226356506,
+      "learning_rate": 6.0273159085877074e-06,
+      "loss": 0.7539,
+      "step": 15832
+    },
+    {
+      "epoch": 0.8714293576971766,
+      "grad_norm": 0.7908313870429993,
+      "learning_rate": 6.026891686187686e-06,
+      "loss": 0.6776,
+      "step": 15833
+    },
+    {
+      "epoch": 0.8714843964995321,
+      "grad_norm": 0.6947218775749207,
+      "learning_rate": 6.02646745606982e-06,
+      "loss": 0.7375,
+      "step": 15834
+    },
+    {
+      "epoch": 0.8715394353018878,
+      "grad_norm": 0.7137001156806946,
+      "learning_rate": 6.0260432182373e-06,
+      "loss": 0.7213,
+      "step": 15835
+    },
+    {
+      "epoch": 0.8715944741042435,
+      "grad_norm": 0.6175974011421204,
+      "learning_rate": 6.025618972693314e-06,
+      "loss": 0.6468,
+      "step": 15836
+    },
+    {
+      "epoch": 0.8716495129065992,
+      "grad_norm": 0.6631742119789124,
+      "learning_rate": 6.0251947194410496e-06,
+      "loss": 0.7116,
+      "step": 15837
+    },
+    {
+      "epoch": 0.8717045517089548,
+      "grad_norm": 0.7667781710624695,
+      "learning_rate": 6.024770458483698e-06,
+      "loss": 0.8836,
+      "step": 15838
+    },
+    {
+      "epoch": 0.8717595905113105,
+      "grad_norm": 0.664364218711853,
+      "learning_rate": 6.024346189824444e-06,
+      "loss": 0.7719,
+      "step": 15839
+    },
+    {
+      "epoch": 0.8718146293136662,
+      "grad_norm": 0.7073011994361877,
+      "learning_rate": 6.023921913466477e-06,
+      "loss": 0.7117,
+      "step": 15840
+    },
+    {
+      "epoch": 0.8718696681160218,
+      "grad_norm": 0.7126373052597046,
+      "learning_rate": 6.02349762941299e-06,
+      "loss": 0.7693,
+      "step": 15841
+    },
+    {
+      "epoch": 0.8719247069183774,
+      "grad_norm": 0.7864155173301697,
+      "learning_rate": 6.0230733376671665e-06,
+      "loss": 0.8195,
+      "step": 15842
+    },
+    {
+      "epoch": 0.8719797457207331,
+      "grad_norm": 0.7260663509368896,
+      "learning_rate": 6.0226490382322e-06,
+      "loss": 0.7739,
+      "step": 15843
+    },
+    {
+      "epoch": 0.8720347845230888,
+      "grad_norm": 0.7656667232513428,
+      "learning_rate": 6.0222247311112745e-06,
+      "loss": 0.6552,
+      "step": 15844
+    },
+    {
+      "epoch": 0.8720898233254445,
+      "grad_norm": 0.7063844799995422,
+      "learning_rate": 6.0218004163075826e-06,
+      "loss": 0.7506,
+      "step": 15845
+    },
+    {
+      "epoch": 0.8721448621278001,
+      "grad_norm": 0.6452813744544983,
+      "learning_rate": 6.021376093824313e-06,
+      "loss": 0.6854,
+      "step": 15846
+    },
+    {
+      "epoch": 0.8721999009301558,
+      "grad_norm": 0.6507169008255005,
+      "learning_rate": 6.020951763664653e-06,
+      "loss": 0.7289,
+      "step": 15847
+    },
+    {
+      "epoch": 0.8722549397325114,
+      "grad_norm": 0.6529967784881592,
+      "learning_rate": 6.020527425831793e-06,
+      "loss": 0.7196,
+      "step": 15848
+    },
+    {
+      "epoch": 0.8723099785348671,
+      "grad_norm": 0.8070194125175476,
+      "learning_rate": 6.020103080328924e-06,
+      "loss": 0.7848,
+      "step": 15849
+    },
+    {
+      "epoch": 0.8723650173372227,
+      "grad_norm": 0.7091495394706726,
+      "learning_rate": 6.019678727159232e-06,
+      "loss": 0.7948,
+      "step": 15850
+    },
+    {
+      "epoch": 0.8724200561395784,
+      "grad_norm": 0.8268260955810547,
+      "learning_rate": 6.019254366325907e-06,
+      "loss": 0.7446,
+      "step": 15851
+    },
+    {
+      "epoch": 0.8724750949419341,
+      "grad_norm": 0.7777679562568665,
+      "learning_rate": 6.018829997832139e-06,
+      "loss": 0.8307,
+      "step": 15852
+    },
+    {
+      "epoch": 0.8725301337442897,
+      "grad_norm": 0.6404305696487427,
+      "learning_rate": 6.018405621681117e-06,
+      "loss": 0.6952,
+      "step": 15853
+    },
+    {
+      "epoch": 0.8725851725466454,
+      "grad_norm": 0.8895840644836426,
+      "learning_rate": 6.017981237876033e-06,
+      "loss": 0.7554,
+      "step": 15854
+    },
+    {
+      "epoch": 0.872640211349001,
+      "grad_norm": 0.6717105507850647,
+      "learning_rate": 6.017556846420073e-06,
+      "loss": 0.674,
+      "step": 15855
+    },
+    {
+      "epoch": 0.8726952501513567,
+      "grad_norm": 0.6096089482307434,
+      "learning_rate": 6.017132447316427e-06,
+      "loss": 0.7508,
+      "step": 15856
+    },
+    {
+      "epoch": 0.8727502889537123,
+      "grad_norm": 0.7513056397438049,
+      "learning_rate": 6.016708040568288e-06,
+      "loss": 0.717,
+      "step": 15857
+    },
+    {
+      "epoch": 0.872805327756068,
+      "grad_norm": 0.6977408528327942,
+      "learning_rate": 6.0162836261788425e-06,
+      "loss": 0.7002,
+      "step": 15858
+    },
+    {
+      "epoch": 0.8728603665584237,
+      "grad_norm": 0.6753636598587036,
+      "learning_rate": 6.015859204151282e-06,
+      "loss": 0.7414,
+      "step": 15859
+    },
+    {
+      "epoch": 0.8729154053607794,
+      "grad_norm": 0.7120729684829712,
+      "learning_rate": 6.015434774488795e-06,
+      "loss": 0.6774,
+      "step": 15860
+    },
+    {
+      "epoch": 0.872970444163135,
+      "grad_norm": 0.7560111880302429,
+      "learning_rate": 6.015010337194573e-06,
+      "loss": 0.6887,
+      "step": 15861
+    },
+    {
+      "epoch": 0.8730254829654907,
+      "grad_norm": 0.652497410774231,
+      "learning_rate": 6.0145858922718044e-06,
+      "loss": 0.76,
+      "step": 15862
+    },
+    {
+      "epoch": 0.8730805217678463,
+      "grad_norm": 0.7120025753974915,
+      "learning_rate": 6.01416143972368e-06,
+      "loss": 0.7008,
+      "step": 15863
+    },
+    {
+      "epoch": 0.873135560570202,
+      "grad_norm": 0.7517643570899963,
+      "learning_rate": 6.013736979553389e-06,
+      "loss": 0.8944,
+      "step": 15864
+    },
+    {
+      "epoch": 0.8731905993725576,
+      "grad_norm": 0.6225923299789429,
+      "learning_rate": 6.013312511764122e-06,
+      "loss": 0.6217,
+      "step": 15865
+    },
+    {
+      "epoch": 0.8732456381749133,
+      "grad_norm": 0.8815253376960754,
+      "learning_rate": 6.012888036359071e-06,
+      "loss": 0.8121,
+      "step": 15866
+    },
+    {
+      "epoch": 0.873300676977269,
+      "grad_norm": 0.676211953163147,
+      "learning_rate": 6.012463553341424e-06,
+      "loss": 0.7233,
+      "step": 15867
+    },
+    {
+      "epoch": 0.8733557157796247,
+      "grad_norm": 0.6566252708435059,
+      "learning_rate": 6.012039062714371e-06,
+      "loss": 0.8099,
+      "step": 15868
+    },
+    {
+      "epoch": 0.8734107545819803,
+      "grad_norm": 0.7964142560958862,
+      "learning_rate": 6.011614564481103e-06,
+      "loss": 0.758,
+      "step": 15869
+    },
+    {
+      "epoch": 0.8734657933843359,
+      "grad_norm": 0.6923096776008606,
+      "learning_rate": 6.011190058644811e-06,
+      "loss": 0.6997,
+      "step": 15870
+    },
+    {
+      "epoch": 0.8735208321866916,
+      "grad_norm": 0.6507520079612732,
+      "learning_rate": 6.010765545208687e-06,
+      "loss": 0.7046,
+      "step": 15871
+    },
+    {
+      "epoch": 0.8735758709890473,
+      "grad_norm": 0.8206372857093811,
+      "learning_rate": 6.010341024175918e-06,
+      "loss": 0.8568,
+      "step": 15872
+    },
+    {
+      "epoch": 0.8736309097914029,
+      "grad_norm": 0.6379685997962952,
+      "learning_rate": 6.0099164955496965e-06,
+      "loss": 0.7537,
+      "step": 15873
+    },
+    {
+      "epoch": 0.8736859485937586,
+      "grad_norm": 0.7258248925209045,
+      "learning_rate": 6.009491959333214e-06,
+      "loss": 0.6946,
+      "step": 15874
+    },
+    {
+      "epoch": 0.8737409873961143,
+      "grad_norm": 0.6882272362709045,
+      "learning_rate": 6.0090674155296606e-06,
+      "loss": 0.6508,
+      "step": 15875
+    },
+    {
+      "epoch": 0.87379602619847,
+      "grad_norm": 0.646864652633667,
+      "learning_rate": 6.0086428641422245e-06,
+      "loss": 0.7061,
+      "step": 15876
+    },
+    {
+      "epoch": 0.8738510650008255,
+      "grad_norm": 0.772055983543396,
+      "learning_rate": 6.008218305174099e-06,
+      "loss": 0.7435,
+      "step": 15877
+    },
+    {
+      "epoch": 0.8739061038031812,
+      "grad_norm": 0.660976767539978,
+      "learning_rate": 6.007793738628476e-06,
+      "loss": 0.6834,
+      "step": 15878
+    },
+    {
+      "epoch": 0.8739611426055369,
+      "grad_norm": 0.6279324293136597,
+      "learning_rate": 6.007369164508544e-06,
+      "loss": 0.6903,
+      "step": 15879
+    },
+    {
+      "epoch": 0.8740161814078926,
+      "grad_norm": 0.7111205458641052,
+      "learning_rate": 6.006944582817495e-06,
+      "loss": 0.7338,
+      "step": 15880
+    },
+    {
+      "epoch": 0.8740712202102482,
+      "grad_norm": 0.6149270534515381,
+      "learning_rate": 6.006519993558519e-06,
+      "loss": 0.6639,
+      "step": 15881
+    },
+    {
+      "epoch": 0.8741262590126039,
+      "grad_norm": 0.7477333545684814,
+      "learning_rate": 6.00609539673481e-06,
+      "loss": 0.8032,
+      "step": 15882
+    },
+    {
+      "epoch": 0.8741812978149596,
+      "grad_norm": 0.8613518476486206,
+      "learning_rate": 6.005670792349557e-06,
+      "loss": 0.7508,
+      "step": 15883
+    },
+    {
+      "epoch": 0.8742363366173153,
+      "grad_norm": 0.6627817153930664,
+      "learning_rate": 6.0052461804059515e-06,
+      "loss": 0.7898,
+      "step": 15884
+    },
+    {
+      "epoch": 0.8742913754196708,
+      "grad_norm": 0.6863798499107361,
+      "learning_rate": 6.004821560907185e-06,
+      "loss": 0.7674,
+      "step": 15885
+    },
+    {
+      "epoch": 0.8743464142220265,
+      "grad_norm": 0.6809577941894531,
+      "learning_rate": 6.004396933856449e-06,
+      "loss": 0.8094,
+      "step": 15886
+    },
+    {
+      "epoch": 0.8744014530243822,
+      "grad_norm": 0.6340956687927246,
+      "learning_rate": 6.003972299256934e-06,
+      "loss": 0.6508,
+      "step": 15887
+    },
+    {
+      "epoch": 0.8744564918267379,
+      "grad_norm": 0.6261658072471619,
+      "learning_rate": 6.003547657111831e-06,
+      "loss": 0.7375,
+      "step": 15888
+    },
+    {
+      "epoch": 0.8745115306290935,
+      "grad_norm": 0.7042009830474854,
+      "learning_rate": 6.003123007424332e-06,
+      "loss": 0.6817,
+      "step": 15889
+    },
+    {
+      "epoch": 0.8745665694314492,
+      "grad_norm": 0.719497561454773,
+      "learning_rate": 6.002698350197631e-06,
+      "loss": 0.7689,
+      "step": 15890
+    },
+    {
+      "epoch": 0.8746216082338049,
+      "grad_norm": 0.7034541964530945,
+      "learning_rate": 6.002273685434916e-06,
+      "loss": 0.7956,
+      "step": 15891
+    },
+    {
+      "epoch": 0.8746766470361605,
+      "grad_norm": 0.6404724717140198,
+      "learning_rate": 6.0018490131393815e-06,
+      "loss": 0.672,
+      "step": 15892
+    },
+    {
+      "epoch": 0.8747316858385161,
+      "grad_norm": 0.6812208294868469,
+      "learning_rate": 6.001424333314216e-06,
+      "loss": 0.6911,
+      "step": 15893
+    },
+    {
+      "epoch": 0.8747867246408718,
+      "grad_norm": 0.5907782912254333,
+      "learning_rate": 6.000999645962615e-06,
+      "loss": 0.6476,
+      "step": 15894
+    },
+    {
+      "epoch": 0.8748417634432275,
+      "grad_norm": 1.2116328477859497,
+      "learning_rate": 6.000574951087769e-06,
+      "loss": 0.733,
+      "step": 15895
+    },
+    {
+      "epoch": 0.8748968022455831,
+      "grad_norm": 0.6581991314888,
+      "learning_rate": 6.000150248692868e-06,
+      "loss": 0.7441,
+      "step": 15896
+    },
+    {
+      "epoch": 0.8749518410479388,
+      "grad_norm": 0.7342226505279541,
+      "learning_rate": 5.999725538781107e-06,
+      "loss": 0.6592,
+      "step": 15897
+    },
+    {
+      "epoch": 0.8750068798502945,
+      "grad_norm": 0.6864113211631775,
+      "learning_rate": 5.9993008213556766e-06,
+      "loss": 0.7652,
+      "step": 15898
+    },
+    {
+      "epoch": 0.8750619186526502,
+      "grad_norm": 0.6845645904541016,
+      "learning_rate": 5.9988760964197675e-06,
+      "loss": 0.7471,
+      "step": 15899
+    },
+    {
+      "epoch": 0.8751169574550057,
+      "grad_norm": 0.663165271282196,
+      "learning_rate": 5.998451363976574e-06,
+      "loss": 0.7667,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8751719962573614,
+      "grad_norm": 0.6032472252845764,
+      "learning_rate": 5.998026624029286e-06,
+      "loss": 0.6309,
+      "step": 15901
+    },
+    {
+      "epoch": 0.8752270350597171,
+      "grad_norm": 0.6466236710548401,
+      "learning_rate": 5.997601876581098e-06,
+      "loss": 0.7418,
+      "step": 15902
+    },
+    {
+      "epoch": 0.8752820738620728,
+      "grad_norm": 0.6456779837608337,
+      "learning_rate": 5.997177121635201e-06,
+      "loss": 0.6598,
+      "step": 15903
+    },
+    {
+      "epoch": 0.8753371126644284,
+      "grad_norm": 0.7854783535003662,
+      "learning_rate": 5.996752359194788e-06,
+      "loss": 0.7545,
+      "step": 15904
+    },
+    {
+      "epoch": 0.8753921514667841,
+      "grad_norm": 0.7146682143211365,
+      "learning_rate": 5.99632758926305e-06,
+      "loss": 0.6856,
+      "step": 15905
+    },
+    {
+      "epoch": 0.8754471902691398,
+      "grad_norm": 0.7379660606384277,
+      "learning_rate": 5.995902811843181e-06,
+      "loss": 0.8601,
+      "step": 15906
+    },
+    {
+      "epoch": 0.8755022290714954,
+      "grad_norm": 0.6879674196243286,
+      "learning_rate": 5.995478026938375e-06,
+      "loss": 0.8147,
+      "step": 15907
+    },
+    {
+      "epoch": 0.875557267873851,
+      "grad_norm": 0.6188415884971619,
+      "learning_rate": 5.995053234551821e-06,
+      "loss": 0.6646,
+      "step": 15908
+    },
+    {
+      "epoch": 0.8756123066762067,
+      "grad_norm": 1.0765966176986694,
+      "learning_rate": 5.994628434686713e-06,
+      "loss": 0.6624,
+      "step": 15909
+    },
+    {
+      "epoch": 0.8756673454785624,
+      "grad_norm": 0.6391757726669312,
+      "learning_rate": 5.994203627346245e-06,
+      "loss": 0.7087,
+      "step": 15910
+    },
+    {
+      "epoch": 0.8757223842809181,
+      "grad_norm": 0.7664490938186646,
+      "learning_rate": 5.993778812533609e-06,
+      "loss": 0.6969,
+      "step": 15911
+    },
+    {
+      "epoch": 0.8757774230832737,
+      "grad_norm": 0.6901882290840149,
+      "learning_rate": 5.993353990251995e-06,
+      "loss": 0.7879,
+      "step": 15912
+    },
+    {
+      "epoch": 0.8758324618856294,
+      "grad_norm": 0.6871299147605896,
+      "learning_rate": 5.992929160504599e-06,
+      "loss": 0.6981,
+      "step": 15913
+    },
+    {
+      "epoch": 0.875887500687985,
+      "grad_norm": 0.754436731338501,
+      "learning_rate": 5.9925043232946145e-06,
+      "loss": 0.7549,
+      "step": 15914
+    },
+    {
+      "epoch": 0.8759425394903407,
+      "grad_norm": 0.7250627875328064,
+      "learning_rate": 5.992079478625232e-06,
+      "loss": 0.8134,
+      "step": 15915
+    },
+    {
+      "epoch": 0.8759975782926963,
+      "grad_norm": 0.7294771671295166,
+      "learning_rate": 5.991654626499647e-06,
+      "loss": 0.8551,
+      "step": 15916
+    },
+    {
+      "epoch": 0.876052617095052,
+      "grad_norm": 0.6388968229293823,
+      "learning_rate": 5.991229766921049e-06,
+      "loss": 0.7562,
+      "step": 15917
+    },
+    {
+      "epoch": 0.8761076558974077,
+      "grad_norm": 0.7206701636314392,
+      "learning_rate": 5.990804899892636e-06,
+      "loss": 0.6818,
+      "step": 15918
+    },
+    {
+      "epoch": 0.8761626946997634,
+      "grad_norm": 0.6607910394668579,
+      "learning_rate": 5.990380025417597e-06,
+      "loss": 0.6956,
+      "step": 15919
+    },
+    {
+      "epoch": 0.876217733502119,
+      "grad_norm": 0.6806843280792236,
+      "learning_rate": 5.9899551434991276e-06,
+      "loss": 0.7261,
+      "step": 15920
+    },
+    {
+      "epoch": 0.8762727723044746,
+      "grad_norm": 0.6994869709014893,
+      "learning_rate": 5.989530254140421e-06,
+      "loss": 0.7709,
+      "step": 15921
+    },
+    {
+      "epoch": 0.8763278111068303,
+      "grad_norm": 0.6707572937011719,
+      "learning_rate": 5.9891053573446685e-06,
+      "loss": 0.6962,
+      "step": 15922
+    },
+    {
+      "epoch": 0.876382849909186,
+      "grad_norm": 0.8244118690490723,
+      "learning_rate": 5.988680453115065e-06,
+      "loss": 0.7173,
+      "step": 15923
+    },
+    {
+      "epoch": 0.8764378887115416,
+      "grad_norm": 0.7745859026908875,
+      "learning_rate": 5.988255541454806e-06,
+      "loss": 0.7176,
+      "step": 15924
+    },
+    {
+      "epoch": 0.8764929275138973,
+      "grad_norm": 0.6572975516319275,
+      "learning_rate": 5.98783062236708e-06,
+      "loss": 0.6994,
+      "step": 15925
+    },
+    {
+      "epoch": 0.876547966316253,
+      "grad_norm": 0.7363128066062927,
+      "learning_rate": 5.9874056958550845e-06,
+      "loss": 0.7392,
+      "step": 15926
+    },
+    {
+      "epoch": 0.8766030051186087,
+      "grad_norm": 0.6638974547386169,
+      "learning_rate": 5.986980761922012e-06,
+      "loss": 0.693,
+      "step": 15927
+    },
+    {
+      "epoch": 0.8766580439209642,
+      "grad_norm": 0.6946521997451782,
+      "learning_rate": 5.9865558205710576e-06,
+      "loss": 0.7727,
+      "step": 15928
+    },
+    {
+      "epoch": 0.8767130827233199,
+      "grad_norm": 0.7787197232246399,
+      "learning_rate": 5.9861308718054115e-06,
+      "loss": 0.7203,
+      "step": 15929
+    },
+    {
+      "epoch": 0.8767681215256756,
+      "grad_norm": 0.6275830864906311,
+      "learning_rate": 5.985705915628271e-06,
+      "loss": 0.7613,
+      "step": 15930
+    },
+    {
+      "epoch": 0.8768231603280313,
+      "grad_norm": 0.5722871422767639,
+      "learning_rate": 5.985280952042829e-06,
+      "loss": 0.6635,
+      "step": 15931
+    },
+    {
+      "epoch": 0.8768781991303869,
+      "grad_norm": 0.6991726160049438,
+      "learning_rate": 5.984855981052278e-06,
+      "loss": 0.7965,
+      "step": 15932
+    },
+    {
+      "epoch": 0.8769332379327426,
+      "grad_norm": 0.6550602912902832,
+      "learning_rate": 5.984431002659815e-06,
+      "loss": 0.7591,
+      "step": 15933
+    },
+    {
+      "epoch": 0.8769882767350983,
+      "grad_norm": 0.703481137752533,
+      "learning_rate": 5.984006016868631e-06,
+      "loss": 0.7623,
+      "step": 15934
+    },
+    {
+      "epoch": 0.877043315537454,
+      "grad_norm": 0.5606309175491333,
+      "learning_rate": 5.983581023681922e-06,
+      "loss": 0.5957,
+      "step": 15935
+    },
+    {
+      "epoch": 0.8770983543398095,
+      "grad_norm": 0.6857683062553406,
+      "learning_rate": 5.98315602310288e-06,
+      "loss": 0.7629,
+      "step": 15936
+    },
+    {
+      "epoch": 0.8771533931421652,
+      "grad_norm": 0.6976568698883057,
+      "learning_rate": 5.982731015134699e-06,
+      "loss": 0.7259,
+      "step": 15937
+    },
+    {
+      "epoch": 0.8772084319445209,
+      "grad_norm": 0.7161057591438293,
+      "learning_rate": 5.982305999780578e-06,
+      "loss": 0.8572,
+      "step": 15938
+    },
+    {
+      "epoch": 0.8772634707468765,
+      "grad_norm": 0.7051637768745422,
+      "learning_rate": 5.981880977043706e-06,
+      "loss": 0.8251,
+      "step": 15939
+    },
+    {
+      "epoch": 0.8773185095492322,
+      "grad_norm": 0.6403392553329468,
+      "learning_rate": 5.98145594692728e-06,
+      "loss": 0.7021,
+      "step": 15940
+    },
+    {
+      "epoch": 0.8773735483515879,
+      "grad_norm": 0.8370338678359985,
+      "learning_rate": 5.981030909434493e-06,
+      "loss": 0.6376,
+      "step": 15941
+    },
+    {
+      "epoch": 0.8774285871539436,
+      "grad_norm": 0.7969813346862793,
+      "learning_rate": 5.980605864568541e-06,
+      "loss": 0.7935,
+      "step": 15942
+    },
+    {
+      "epoch": 0.8774836259562991,
+      "grad_norm": 0.7235258221626282,
+      "learning_rate": 5.980180812332619e-06,
+      "loss": 0.8024,
+      "step": 15943
+    },
+    {
+      "epoch": 0.8775386647586548,
+      "grad_norm": 0.6627998352050781,
+      "learning_rate": 5.97975575272992e-06,
+      "loss": 0.8184,
+      "step": 15944
+    },
+    {
+      "epoch": 0.8775937035610105,
+      "grad_norm": 1.123727798461914,
+      "learning_rate": 5.979330685763638e-06,
+      "loss": 0.9015,
+      "step": 15945
+    },
+    {
+      "epoch": 0.8776487423633662,
+      "grad_norm": 0.8116182088851929,
+      "learning_rate": 5.97890561143697e-06,
+      "loss": 0.7327,
+      "step": 15946
+    },
+    {
+      "epoch": 0.8777037811657218,
+      "grad_norm": 0.6537826657295227,
+      "learning_rate": 5.978480529753108e-06,
+      "loss": 0.7215,
+      "step": 15947
+    },
+    {
+      "epoch": 0.8777588199680775,
+      "grad_norm": 0.746971845626831,
+      "learning_rate": 5.978055440715249e-06,
+      "loss": 0.8549,
+      "step": 15948
+    },
+    {
+      "epoch": 0.8778138587704332,
+      "grad_norm": 0.7417864799499512,
+      "learning_rate": 5.9776303443265855e-06,
+      "loss": 0.7806,
+      "step": 15949
+    },
+    {
+      "epoch": 0.8778688975727889,
+      "grad_norm": 0.6726604700088501,
+      "learning_rate": 5.977205240590317e-06,
+      "loss": 0.7745,
+      "step": 15950
+    },
+    {
+      "epoch": 0.8779239363751444,
+      "grad_norm": 0.9483097791671753,
+      "learning_rate": 5.976780129509634e-06,
+      "loss": 0.81,
+      "step": 15951
+    },
+    {
+      "epoch": 0.8779789751775001,
+      "grad_norm": 0.7212807536125183,
+      "learning_rate": 5.976355011087734e-06,
+      "loss": 0.6534,
+      "step": 15952
+    },
+    {
+      "epoch": 0.8780340139798558,
+      "grad_norm": 0.725513756275177,
+      "learning_rate": 5.975929885327808e-06,
+      "loss": 0.6902,
+      "step": 15953
+    },
+    {
+      "epoch": 0.8780890527822115,
+      "grad_norm": 0.6132134795188904,
+      "learning_rate": 5.975504752233057e-06,
+      "loss": 0.7397,
+      "step": 15954
+    },
+    {
+      "epoch": 0.8781440915845671,
+      "grad_norm": 0.7595381736755371,
+      "learning_rate": 5.975079611806672e-06,
+      "loss": 0.5893,
+      "step": 15955
+    },
+    {
+      "epoch": 0.8781991303869228,
+      "grad_norm": 0.7069967985153198,
+      "learning_rate": 5.974654464051851e-06,
+      "loss": 0.7702,
+      "step": 15956
+    },
+    {
+      "epoch": 0.8782541691892785,
+      "grad_norm": 0.650039553642273,
+      "learning_rate": 5.974229308971788e-06,
+      "loss": 0.7708,
+      "step": 15957
+    },
+    {
+      "epoch": 0.8783092079916341,
+      "grad_norm": 0.6799747347831726,
+      "learning_rate": 5.973804146569677e-06,
+      "loss": 0.7094,
+      "step": 15958
+    },
+    {
+      "epoch": 0.8783642467939897,
+      "grad_norm": 0.653275728225708,
+      "learning_rate": 5.973378976848716e-06,
+      "loss": 0.623,
+      "step": 15959
+    },
+    {
+      "epoch": 0.8784192855963454,
+      "grad_norm": 0.6734615564346313,
+      "learning_rate": 5.972953799812098e-06,
+      "loss": 0.7071,
+      "step": 15960
+    },
+    {
+      "epoch": 0.8784743243987011,
+      "grad_norm": 0.6319865584373474,
+      "learning_rate": 5.9725286154630205e-06,
+      "loss": 0.7239,
+      "step": 15961
+    },
+    {
+      "epoch": 0.8785293632010568,
+      "grad_norm": 0.6933672428131104,
+      "learning_rate": 5.972103423804677e-06,
+      "loss": 0.7138,
+      "step": 15962
+    },
+    {
+      "epoch": 0.8785844020034124,
+      "grad_norm": 0.7323144674301147,
+      "learning_rate": 5.971678224840266e-06,
+      "loss": 0.7534,
+      "step": 15963
+    },
+    {
+      "epoch": 0.8786394408057681,
+      "grad_norm": 0.6736310720443726,
+      "learning_rate": 5.971253018572981e-06,
+      "loss": 0.7644,
+      "step": 15964
+    },
+    {
+      "epoch": 0.8786944796081237,
+      "grad_norm": 0.6524562835693359,
+      "learning_rate": 5.970827805006016e-06,
+      "loss": 0.7727,
+      "step": 15965
+    },
+    {
+      "epoch": 0.8787495184104794,
+      "grad_norm": 0.7736978530883789,
+      "learning_rate": 5.970402584142573e-06,
+      "loss": 0.7641,
+      "step": 15966
+    },
+    {
+      "epoch": 0.878804557212835,
+      "grad_norm": 0.7099476456642151,
+      "learning_rate": 5.969977355985842e-06,
+      "loss": 0.7064,
+      "step": 15967
+    },
+    {
+      "epoch": 0.8788595960151907,
+      "grad_norm": 0.6340349912643433,
+      "learning_rate": 5.969552120539021e-06,
+      "loss": 0.7135,
+      "step": 15968
+    },
+    {
+      "epoch": 0.8789146348175464,
+      "grad_norm": 0.649853527545929,
+      "learning_rate": 5.969126877805306e-06,
+      "loss": 0.7068,
+      "step": 15969
+    },
+    {
+      "epoch": 0.8789696736199021,
+      "grad_norm": 0.589920699596405,
+      "learning_rate": 5.9687016277878925e-06,
+      "loss": 0.6891,
+      "step": 15970
+    },
+    {
+      "epoch": 0.8790247124222577,
+      "grad_norm": 0.7485616207122803,
+      "learning_rate": 5.968276370489977e-06,
+      "loss": 0.7305,
+      "step": 15971
+    },
+    {
+      "epoch": 0.8790797512246133,
+      "grad_norm": 0.6765890121459961,
+      "learning_rate": 5.967851105914756e-06,
+      "loss": 0.6862,
+      "step": 15972
+    },
+    {
+      "epoch": 0.879134790026969,
+      "grad_norm": 0.7127717733383179,
+      "learning_rate": 5.967425834065423e-06,
+      "loss": 0.6255,
+      "step": 15973
+    },
+    {
+      "epoch": 0.8791898288293247,
+      "grad_norm": 1.1564292907714844,
+      "learning_rate": 5.967000554945179e-06,
+      "loss": 0.7351,
+      "step": 15974
+    },
+    {
+      "epoch": 0.8792448676316803,
+      "grad_norm": 0.7343658804893494,
+      "learning_rate": 5.966575268557217e-06,
+      "loss": 0.7605,
+      "step": 15975
+    },
+    {
+      "epoch": 0.879299906434036,
+      "grad_norm": 0.695280134677887,
+      "learning_rate": 5.966149974904733e-06,
+      "loss": 0.7425,
+      "step": 15976
+    },
+    {
+      "epoch": 0.8793549452363917,
+      "grad_norm": 0.7075724005699158,
+      "learning_rate": 5.965724673990925e-06,
+      "loss": 0.6501,
+      "step": 15977
+    },
+    {
+      "epoch": 0.8794099840387474,
+      "grad_norm": 0.6764316558837891,
+      "learning_rate": 5.96529936581899e-06,
+      "loss": 0.6871,
+      "step": 15978
+    },
+    {
+      "epoch": 0.879465022841103,
+      "grad_norm": 0.6782918572425842,
+      "learning_rate": 5.964874050392122e-06,
+      "loss": 0.7298,
+      "step": 15979
+    },
+    {
+      "epoch": 0.8795200616434586,
+      "grad_norm": 0.7106739282608032,
+      "learning_rate": 5.964448727713519e-06,
+      "loss": 0.755,
+      "step": 15980
+    },
+    {
+      "epoch": 0.8795751004458143,
+      "grad_norm": 0.7591151595115662,
+      "learning_rate": 5.964023397786378e-06,
+      "loss": 0.7603,
+      "step": 15981
+    },
+    {
+      "epoch": 0.8796301392481699,
+      "grad_norm": 0.6050585508346558,
+      "learning_rate": 5.963598060613896e-06,
+      "loss": 0.6895,
+      "step": 15982
+    },
+    {
+      "epoch": 0.8796851780505256,
+      "grad_norm": 0.7037108540534973,
+      "learning_rate": 5.963172716199267e-06,
+      "loss": 0.7475,
+      "step": 15983
+    },
+    {
+      "epoch": 0.8797402168528813,
+      "grad_norm": 0.6476989984512329,
+      "learning_rate": 5.962747364545692e-06,
+      "loss": 0.7925,
+      "step": 15984
+    },
+    {
+      "epoch": 0.879795255655237,
+      "grad_norm": 0.6409072875976562,
+      "learning_rate": 5.962322005656362e-06,
+      "loss": 0.8134,
+      "step": 15985
+    },
+    {
+      "epoch": 0.8798502944575926,
+      "grad_norm": 0.8730958700180054,
+      "learning_rate": 5.96189663953448e-06,
+      "loss": 0.8283,
+      "step": 15986
+    },
+    {
+      "epoch": 0.8799053332599482,
+      "grad_norm": 0.73405522108078,
+      "learning_rate": 5.96147126618324e-06,
+      "loss": 0.7351,
+      "step": 15987
+    },
+    {
+      "epoch": 0.8799603720623039,
+      "grad_norm": 0.6587926745414734,
+      "learning_rate": 5.961045885605839e-06,
+      "loss": 0.7713,
+      "step": 15988
+    },
+    {
+      "epoch": 0.8800154108646596,
+      "grad_norm": 0.8479684591293335,
+      "learning_rate": 5.9606204978054736e-06,
+      "loss": 0.7537,
+      "step": 15989
+    },
+    {
+      "epoch": 0.8800704496670152,
+      "grad_norm": 0.731315553188324,
+      "learning_rate": 5.960195102785343e-06,
+      "loss": 0.7363,
+      "step": 15990
+    },
+    {
+      "epoch": 0.8801254884693709,
+      "grad_norm": 0.7163324952125549,
+      "learning_rate": 5.9597697005486434e-06,
+      "loss": 0.718,
+      "step": 15991
+    },
+    {
+      "epoch": 0.8801805272717266,
+      "grad_norm": 0.6090518832206726,
+      "learning_rate": 5.9593442910985714e-06,
+      "loss": 0.6002,
+      "step": 15992
+    },
+    {
+      "epoch": 0.8802355660740823,
+      "grad_norm": 0.633756160736084,
+      "learning_rate": 5.958918874438324e-06,
+      "loss": 0.5981,
+      "step": 15993
+    },
+    {
+      "epoch": 0.8802906048764378,
+      "grad_norm": 0.728961169719696,
+      "learning_rate": 5.958493450571099e-06,
+      "loss": 0.7964,
+      "step": 15994
+    },
+    {
+      "epoch": 0.8803456436787935,
+      "grad_norm": 0.8194692134857178,
+      "learning_rate": 5.958068019500094e-06,
+      "loss": 0.748,
+      "step": 15995
+    },
+    {
+      "epoch": 0.8804006824811492,
+      "grad_norm": 0.9324885010719299,
+      "learning_rate": 5.957642581228506e-06,
+      "loss": 0.8263,
+      "step": 15996
+    },
+    {
+      "epoch": 0.8804557212835049,
+      "grad_norm": 0.7585923075675964,
+      "learning_rate": 5.957217135759532e-06,
+      "loss": 0.8208,
+      "step": 15997
+    },
+    {
+      "epoch": 0.8805107600858605,
+      "grad_norm": 0.726158618927002,
+      "learning_rate": 5.956791683096371e-06,
+      "loss": 0.7664,
+      "step": 15998
+    },
+    {
+      "epoch": 0.8805657988882162,
+      "grad_norm": 0.7648130059242249,
+      "learning_rate": 5.95636622324222e-06,
+      "loss": 0.8102,
+      "step": 15999
+    },
+    {
+      "epoch": 0.8806208376905719,
+      "grad_norm": 0.7053543925285339,
+      "learning_rate": 5.955940756200277e-06,
+      "loss": 0.8446,
+      "step": 16000
+    },
+    {
+      "epoch": 0.8806758764929276,
+      "grad_norm": 1.0602153539657593,
+      "learning_rate": 5.955515281973737e-06,
+      "loss": 0.8199,
+      "step": 16001
+    },
+    {
+      "epoch": 0.8807309152952831,
+      "grad_norm": 0.7527370452880859,
+      "learning_rate": 5.955089800565802e-06,
+      "loss": 0.7895,
+      "step": 16002
+    },
+    {
+      "epoch": 0.8807859540976388,
+      "grad_norm": 0.7191178202629089,
+      "learning_rate": 5.954664311979666e-06,
+      "loss": 0.8836,
+      "step": 16003
+    },
+    {
+      "epoch": 0.8808409928999945,
+      "grad_norm": 0.6993798017501831,
+      "learning_rate": 5.95423881621853e-06,
+      "loss": 0.7765,
+      "step": 16004
+    },
+    {
+      "epoch": 0.8808960317023502,
+      "grad_norm": 0.6863248944282532,
+      "learning_rate": 5.9538133132855915e-06,
+      "loss": 0.6294,
+      "step": 16005
+    },
+    {
+      "epoch": 0.8809510705047058,
+      "grad_norm": 0.7736430168151855,
+      "learning_rate": 5.953387803184046e-06,
+      "loss": 0.9003,
+      "step": 16006
+    },
+    {
+      "epoch": 0.8810061093070615,
+      "grad_norm": 0.6101526618003845,
+      "learning_rate": 5.9529622859170935e-06,
+      "loss": 0.7125,
+      "step": 16007
+    },
+    {
+      "epoch": 0.8810611481094172,
+      "grad_norm": 0.7688401341438293,
+      "learning_rate": 5.952536761487932e-06,
+      "loss": 0.7667,
+      "step": 16008
+    },
+    {
+      "epoch": 0.8811161869117728,
+      "grad_norm": 0.6438688039779663,
+      "learning_rate": 5.9521112298997575e-06,
+      "loss": 0.7327,
+      "step": 16009
+    },
+    {
+      "epoch": 0.8811712257141284,
+      "grad_norm": 0.7732130885124207,
+      "learning_rate": 5.951685691155769e-06,
+      "loss": 0.8466,
+      "step": 16010
+    },
+    {
+      "epoch": 0.8812262645164841,
+      "grad_norm": 0.755892813205719,
+      "learning_rate": 5.951260145259168e-06,
+      "loss": 0.8042,
+      "step": 16011
+    },
+    {
+      "epoch": 0.8812813033188398,
+      "grad_norm": 0.7132954001426697,
+      "learning_rate": 5.950834592213151e-06,
+      "loss": 0.7801,
+      "step": 16012
+    },
+    {
+      "epoch": 0.8813363421211955,
+      "grad_norm": 0.702319324016571,
+      "learning_rate": 5.950409032020914e-06,
+      "loss": 0.7487,
+      "step": 16013
+    },
+    {
+      "epoch": 0.8813913809235511,
+      "grad_norm": 0.6477691531181335,
+      "learning_rate": 5.949983464685656e-06,
+      "loss": 0.6942,
+      "step": 16014
+    },
+    {
+      "epoch": 0.8814464197259068,
+      "grad_norm": 0.6817807555198669,
+      "learning_rate": 5.949557890210578e-06,
+      "loss": 0.676,
+      "step": 16015
+    },
+    {
+      "epoch": 0.8815014585282624,
+      "grad_norm": 0.6980645060539246,
+      "learning_rate": 5.949132308598877e-06,
+      "loss": 0.7837,
+      "step": 16016
+    },
+    {
+      "epoch": 0.8815564973306181,
+      "grad_norm": 0.9056459665298462,
+      "learning_rate": 5.948706719853753e-06,
+      "loss": 0.7005,
+      "step": 16017
+    },
+    {
+      "epoch": 0.8816115361329737,
+      "grad_norm": 0.8172656297683716,
+      "learning_rate": 5.948281123978402e-06,
+      "loss": 0.7784,
+      "step": 16018
+    },
+    {
+      "epoch": 0.8816665749353294,
+      "grad_norm": 0.6387753486633301,
+      "learning_rate": 5.947855520976025e-06,
+      "loss": 0.8171,
+      "step": 16019
+    },
+    {
+      "epoch": 0.8817216137376851,
+      "grad_norm": 0.6150957345962524,
+      "learning_rate": 5.947429910849818e-06,
+      "loss": 0.6448,
+      "step": 16020
+    },
+    {
+      "epoch": 0.8817766525400408,
+      "grad_norm": 0.7051831483840942,
+      "learning_rate": 5.947004293602982e-06,
+      "loss": 0.7585,
+      "step": 16021
+    },
+    {
+      "epoch": 0.8818316913423964,
+      "grad_norm": 0.7967584729194641,
+      "learning_rate": 5.946578669238714e-06,
+      "loss": 0.8122,
+      "step": 16022
+    },
+    {
+      "epoch": 0.881886730144752,
+      "grad_norm": 0.6126663088798523,
+      "learning_rate": 5.946153037760216e-06,
+      "loss": 0.6149,
+      "step": 16023
+    },
+    {
+      "epoch": 0.8819417689471077,
+      "grad_norm": 0.6940233111381531,
+      "learning_rate": 5.945727399170684e-06,
+      "loss": 0.7373,
+      "step": 16024
+    },
+    {
+      "epoch": 0.8819968077494633,
+      "grad_norm": 0.652776837348938,
+      "learning_rate": 5.945301753473318e-06,
+      "loss": 0.7029,
+      "step": 16025
+    },
+    {
+      "epoch": 0.882051846551819,
+      "grad_norm": 0.7182415723800659,
+      "learning_rate": 5.944876100671317e-06,
+      "loss": 0.7343,
+      "step": 16026
+    },
+    {
+      "epoch": 0.8821068853541747,
+      "grad_norm": 0.6714525818824768,
+      "learning_rate": 5.944450440767881e-06,
+      "loss": 0.7402,
+      "step": 16027
+    },
+    {
+      "epoch": 0.8821619241565304,
+      "grad_norm": 0.7144107818603516,
+      "learning_rate": 5.944024773766208e-06,
+      "loss": 0.6775,
+      "step": 16028
+    },
+    {
+      "epoch": 0.882216962958886,
+      "grad_norm": 0.6483643054962158,
+      "learning_rate": 5.943599099669497e-06,
+      "loss": 0.698,
+      "step": 16029
+    },
+    {
+      "epoch": 0.8822720017612417,
+      "grad_norm": 0.6388065218925476,
+      "learning_rate": 5.943173418480949e-06,
+      "loss": 0.7814,
+      "step": 16030
+    },
+    {
+      "epoch": 0.8823270405635973,
+      "grad_norm": 0.6891177892684937,
+      "learning_rate": 5.942747730203761e-06,
+      "loss": 0.7423,
+      "step": 16031
+    },
+    {
+      "epoch": 0.882382079365953,
+      "grad_norm": 0.6425840258598328,
+      "learning_rate": 5.942322034841133e-06,
+      "loss": 0.7178,
+      "step": 16032
+    },
+    {
+      "epoch": 0.8824371181683086,
+      "grad_norm": 0.6308293342590332,
+      "learning_rate": 5.941896332396266e-06,
+      "loss": 0.6624,
+      "step": 16033
+    },
+    {
+      "epoch": 0.8824921569706643,
+      "grad_norm": 0.6310557126998901,
+      "learning_rate": 5.941470622872358e-06,
+      "loss": 0.7365,
+      "step": 16034
+    },
+    {
+      "epoch": 0.88254719577302,
+      "grad_norm": 0.7731123566627502,
+      "learning_rate": 5.941044906272609e-06,
+      "loss": 0.6385,
+      "step": 16035
+    },
+    {
+      "epoch": 0.8826022345753757,
+      "grad_norm": 0.7063789963722229,
+      "learning_rate": 5.940619182600217e-06,
+      "loss": 0.7606,
+      "step": 16036
+    },
+    {
+      "epoch": 0.8826572733777313,
+      "grad_norm": 0.6370593309402466,
+      "learning_rate": 5.940193451858384e-06,
+      "loss": 0.7332,
+      "step": 16037
+    },
+    {
+      "epoch": 0.8827123121800869,
+      "grad_norm": 0.7685242891311646,
+      "learning_rate": 5.9397677140503085e-06,
+      "loss": 0.7809,
+      "step": 16038
+    },
+    {
+      "epoch": 0.8827673509824426,
+      "grad_norm": 0.8439769148826599,
+      "learning_rate": 5.93934196917919e-06,
+      "loss": 0.7419,
+      "step": 16039
+    },
+    {
+      "epoch": 0.8828223897847983,
+      "grad_norm": 0.873909056186676,
+      "learning_rate": 5.93891621724823e-06,
+      "loss": 0.6559,
+      "step": 16040
+    },
+    {
+      "epoch": 0.8828774285871539,
+      "grad_norm": 0.711383581161499,
+      "learning_rate": 5.938490458260626e-06,
+      "loss": 0.718,
+      "step": 16041
+    },
+    {
+      "epoch": 0.8829324673895096,
+      "grad_norm": 0.6775739192962646,
+      "learning_rate": 5.938064692219579e-06,
+      "loss": 0.7062,
+      "step": 16042
+    },
+    {
+      "epoch": 0.8829875061918653,
+      "grad_norm": 0.687095582485199,
+      "learning_rate": 5.93763891912829e-06,
+      "loss": 0.7773,
+      "step": 16043
+    },
+    {
+      "epoch": 0.883042544994221,
+      "grad_norm": 0.6648910641670227,
+      "learning_rate": 5.937213138989957e-06,
+      "loss": 0.7267,
+      "step": 16044
+    },
+    {
+      "epoch": 0.8830975837965765,
+      "grad_norm": 0.6296299695968628,
+      "learning_rate": 5.9367873518077815e-06,
+      "loss": 0.7561,
+      "step": 16045
+    },
+    {
+      "epoch": 0.8831526225989322,
+      "grad_norm": 1.1233999729156494,
+      "learning_rate": 5.936361557584961e-06,
+      "loss": 0.7401,
+      "step": 16046
+    },
+    {
+      "epoch": 0.8832076614012879,
+      "grad_norm": 0.7307866811752319,
+      "learning_rate": 5.935935756324699e-06,
+      "loss": 0.7945,
+      "step": 16047
+    },
+    {
+      "epoch": 0.8832627002036436,
+      "grad_norm": 0.627402663230896,
+      "learning_rate": 5.9355099480301944e-06,
+      "loss": 0.7005,
+      "step": 16048
+    },
+    {
+      "epoch": 0.8833177390059992,
+      "grad_norm": 0.6698537468910217,
+      "learning_rate": 5.935084132704648e-06,
+      "loss": 0.7349,
+      "step": 16049
+    },
+    {
+      "epoch": 0.8833727778083549,
+      "grad_norm": 0.7348290681838989,
+      "learning_rate": 5.934658310351258e-06,
+      "loss": 0.8033,
+      "step": 16050
+    },
+    {
+      "epoch": 0.8834278166107106,
+      "grad_norm": 0.6543971300125122,
+      "learning_rate": 5.934232480973228e-06,
+      "loss": 0.706,
+      "step": 16051
+    },
+    {
+      "epoch": 0.8834828554130663,
+      "grad_norm": 0.7279872894287109,
+      "learning_rate": 5.933806644573756e-06,
+      "loss": 0.8142,
+      "step": 16052
+    },
+    {
+      "epoch": 0.8835378942154218,
+      "grad_norm": 0.6433993577957153,
+      "learning_rate": 5.933380801156044e-06,
+      "loss": 0.7092,
+      "step": 16053
+    },
+    {
+      "epoch": 0.8835929330177775,
+      "grad_norm": 0.7375844717025757,
+      "learning_rate": 5.932954950723291e-06,
+      "loss": 0.7632,
+      "step": 16054
+    },
+    {
+      "epoch": 0.8836479718201332,
+      "grad_norm": 0.6837477087974548,
+      "learning_rate": 5.932529093278698e-06,
+      "loss": 0.735,
+      "step": 16055
+    },
+    {
+      "epoch": 0.8837030106224889,
+      "grad_norm": 0.5978528261184692,
+      "learning_rate": 5.932103228825467e-06,
+      "loss": 0.6745,
+      "step": 16056
+    },
+    {
+      "epoch": 0.8837580494248445,
+      "grad_norm": 0.6475256085395813,
+      "learning_rate": 5.931677357366798e-06,
+      "loss": 0.7469,
+      "step": 16057
+    },
+    {
+      "epoch": 0.8838130882272002,
+      "grad_norm": 0.685108482837677,
+      "learning_rate": 5.931251478905888e-06,
+      "loss": 0.7528,
+      "step": 16058
+    },
+    {
+      "epoch": 0.8838681270295559,
+      "grad_norm": 0.7046063542366028,
+      "learning_rate": 5.930825593445945e-06,
+      "loss": 0.8157,
+      "step": 16059
+    },
+    {
+      "epoch": 0.8839231658319116,
+      "grad_norm": 0.6626511216163635,
+      "learning_rate": 5.930399700990165e-06,
+      "loss": 0.791,
+      "step": 16060
+    },
+    {
+      "epoch": 0.8839782046342671,
+      "grad_norm": 0.7249611020088196,
+      "learning_rate": 5.929973801541749e-06,
+      "loss": 0.8428,
+      "step": 16061
+    },
+    {
+      "epoch": 0.8840332434366228,
+      "grad_norm": 0.5953019261360168,
+      "learning_rate": 5.929547895103899e-06,
+      "loss": 0.6589,
+      "step": 16062
+    },
+    {
+      "epoch": 0.8840882822389785,
+      "grad_norm": 0.6847557425498962,
+      "learning_rate": 5.9291219816798165e-06,
+      "loss": 0.7976,
+      "step": 16063
+    },
+    {
+      "epoch": 0.8841433210413342,
+      "grad_norm": 0.7806814312934875,
+      "learning_rate": 5.928696061272701e-06,
+      "loss": 0.7879,
+      "step": 16064
+    },
+    {
+      "epoch": 0.8841983598436898,
+      "grad_norm": 1.028922200202942,
+      "learning_rate": 5.9282701338857555e-06,
+      "loss": 0.732,
+      "step": 16065
+    },
+    {
+      "epoch": 0.8842533986460455,
+      "grad_norm": 0.6143102645874023,
+      "learning_rate": 5.927844199522179e-06,
+      "loss": 0.7216,
+      "step": 16066
+    },
+    {
+      "epoch": 0.8843084374484012,
+      "grad_norm": 0.6739519834518433,
+      "learning_rate": 5.927418258185176e-06,
+      "loss": 0.7134,
+      "step": 16067
+    },
+    {
+      "epoch": 0.8843634762507567,
+      "grad_norm": 0.7888758778572083,
+      "learning_rate": 5.926992309877944e-06,
+      "loss": 0.7396,
+      "step": 16068
+    },
+    {
+      "epoch": 0.8844185150531124,
+      "grad_norm": 0.6926425695419312,
+      "learning_rate": 5.926566354603687e-06,
+      "loss": 0.7629,
+      "step": 16069
+    },
+    {
+      "epoch": 0.8844735538554681,
+      "grad_norm": 0.7800819277763367,
+      "learning_rate": 5.926140392365602e-06,
+      "loss": 0.834,
+      "step": 16070
+    },
+    {
+      "epoch": 0.8845285926578238,
+      "grad_norm": 0.711067259311676,
+      "learning_rate": 5.925714423166897e-06,
+      "loss": 0.7401,
+      "step": 16071
+    },
+    {
+      "epoch": 0.8845836314601794,
+      "grad_norm": 0.645727276802063,
+      "learning_rate": 5.92528844701077e-06,
+      "loss": 0.719,
+      "step": 16072
+    },
+    {
+      "epoch": 0.8846386702625351,
+      "grad_norm": 0.7098503112792969,
+      "learning_rate": 5.924862463900421e-06,
+      "loss": 0.7838,
+      "step": 16073
+    },
+    {
+      "epoch": 0.8846937090648908,
+      "grad_norm": 1.0021764039993286,
+      "learning_rate": 5.924436473839055e-06,
+      "loss": 0.9824,
+      "step": 16074
+    },
+    {
+      "epoch": 0.8847487478672464,
+      "grad_norm": 0.657049298286438,
+      "learning_rate": 5.924010476829871e-06,
+      "loss": 0.6797,
+      "step": 16075
+    },
+    {
+      "epoch": 0.884803786669602,
+      "grad_norm": 0.6779371500015259,
+      "learning_rate": 5.923584472876072e-06,
+      "loss": 0.6697,
+      "step": 16076
+    },
+    {
+      "epoch": 0.8848588254719577,
+      "grad_norm": 0.6699591279029846,
+      "learning_rate": 5.923158461980859e-06,
+      "loss": 0.6779,
+      "step": 16077
+    },
+    {
+      "epoch": 0.8849138642743134,
+      "grad_norm": 0.6137605905532837,
+      "learning_rate": 5.922732444147434e-06,
+      "loss": 0.7195,
+      "step": 16078
+    },
+    {
+      "epoch": 0.8849689030766691,
+      "grad_norm": 0.6648411750793457,
+      "learning_rate": 5.922306419379e-06,
+      "loss": 0.7027,
+      "step": 16079
+    },
+    {
+      "epoch": 0.8850239418790247,
+      "grad_norm": 0.6827279329299927,
+      "learning_rate": 5.921880387678758e-06,
+      "loss": 0.7582,
+      "step": 16080
+    },
+    {
+      "epoch": 0.8850789806813804,
+      "grad_norm": 0.6747342944145203,
+      "learning_rate": 5.921454349049909e-06,
+      "loss": 0.7174,
+      "step": 16081
+    },
+    {
+      "epoch": 0.885134019483736,
+      "grad_norm": 0.7580771446228027,
+      "learning_rate": 5.921028303495654e-06,
+      "loss": 0.7457,
+      "step": 16082
+    },
+    {
+      "epoch": 0.8851890582860917,
+      "grad_norm": 0.8015843033790588,
+      "learning_rate": 5.920602251019198e-06,
+      "loss": 0.7084,
+      "step": 16083
+    },
+    {
+      "epoch": 0.8852440970884473,
+      "grad_norm": 0.7056819796562195,
+      "learning_rate": 5.9201761916237434e-06,
+      "loss": 0.7363,
+      "step": 16084
+    },
+    {
+      "epoch": 0.885299135890803,
+      "grad_norm": 0.6790309548377991,
+      "learning_rate": 5.91975012531249e-06,
+      "loss": 0.7966,
+      "step": 16085
+    },
+    {
+      "epoch": 0.8853541746931587,
+      "grad_norm": 0.6311555504798889,
+      "learning_rate": 5.91932405208864e-06,
+      "loss": 0.7099,
+      "step": 16086
+    },
+    {
+      "epoch": 0.8854092134955144,
+      "grad_norm": 0.8329381942749023,
+      "learning_rate": 5.918897971955397e-06,
+      "loss": 0.8127,
+      "step": 16087
+    },
+    {
+      "epoch": 0.88546425229787,
+      "grad_norm": 0.6810656785964966,
+      "learning_rate": 5.918471884915964e-06,
+      "loss": 0.8052,
+      "step": 16088
+    },
+    {
+      "epoch": 0.8855192911002256,
+      "grad_norm": 0.7434407472610474,
+      "learning_rate": 5.918045790973541e-06,
+      "loss": 0.7936,
+      "step": 16089
+    },
+    {
+      "epoch": 0.8855743299025813,
+      "grad_norm": 0.6702361106872559,
+      "learning_rate": 5.917619690131332e-06,
+      "loss": 0.744,
+      "step": 16090
+    },
+    {
+      "epoch": 0.885629368704937,
+      "grad_norm": 0.6321039199829102,
+      "learning_rate": 5.9171935823925384e-06,
+      "loss": 0.6949,
+      "step": 16091
+    },
+    {
+      "epoch": 0.8856844075072926,
+      "grad_norm": 0.7133139371871948,
+      "learning_rate": 5.916767467760365e-06,
+      "loss": 0.7113,
+      "step": 16092
+    },
+    {
+      "epoch": 0.8857394463096483,
+      "grad_norm": 0.7414994239807129,
+      "learning_rate": 5.916341346238011e-06,
+      "loss": 0.8203,
+      "step": 16093
+    },
+    {
+      "epoch": 0.885794485112004,
+      "grad_norm": 0.6744404435157776,
+      "learning_rate": 5.91591521782868e-06,
+      "loss": 0.7869,
+      "step": 16094
+    },
+    {
+      "epoch": 0.8858495239143597,
+      "grad_norm": 0.7183727025985718,
+      "learning_rate": 5.915489082535577e-06,
+      "loss": 0.7375,
+      "step": 16095
+    },
+    {
+      "epoch": 0.8859045627167152,
+      "grad_norm": 0.740496814250946,
+      "learning_rate": 5.9150629403619035e-06,
+      "loss": 0.8134,
+      "step": 16096
+    },
+    {
+      "epoch": 0.8859596015190709,
+      "grad_norm": 0.696391224861145,
+      "learning_rate": 5.9146367913108605e-06,
+      "loss": 0.7137,
+      "step": 16097
+    },
+    {
+      "epoch": 0.8860146403214266,
+      "grad_norm": 0.6438629031181335,
+      "learning_rate": 5.914210635385652e-06,
+      "loss": 0.698,
+      "step": 16098
+    },
+    {
+      "epoch": 0.8860696791237823,
+      "grad_norm": 0.7644562125205994,
+      "learning_rate": 5.913784472589482e-06,
+      "loss": 0.771,
+      "step": 16099
+    },
+    {
+      "epoch": 0.8861247179261379,
+      "grad_norm": 0.7281080484390259,
+      "learning_rate": 5.913358302925553e-06,
+      "loss": 0.7281,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8861797567284936,
+      "grad_norm": 0.7768884301185608,
+      "learning_rate": 5.912932126397067e-06,
+      "loss": 0.7859,
+      "step": 16101
+    },
+    {
+      "epoch": 0.8862347955308493,
+      "grad_norm": 0.6960753202438354,
+      "learning_rate": 5.9125059430072275e-06,
+      "loss": 0.702,
+      "step": 16102
+    },
+    {
+      "epoch": 0.886289834333205,
+      "grad_norm": 0.6299503445625305,
+      "learning_rate": 5.912079752759238e-06,
+      "loss": 0.6558,
+      "step": 16103
+    },
+    {
+      "epoch": 0.8863448731355605,
+      "grad_norm": 0.7048517465591431,
+      "learning_rate": 5.9116535556563005e-06,
+      "loss": 0.6915,
+      "step": 16104
+    },
+    {
+      "epoch": 0.8863999119379162,
+      "grad_norm": 1.0701110363006592,
+      "learning_rate": 5.9112273517016195e-06,
+      "loss": 0.7721,
+      "step": 16105
+    },
+    {
+      "epoch": 0.8864549507402719,
+      "grad_norm": 0.6834803223609924,
+      "learning_rate": 5.910801140898396e-06,
+      "loss": 0.7474,
+      "step": 16106
+    },
+    {
+      "epoch": 0.8865099895426276,
+      "grad_norm": 0.6799558401107788,
+      "learning_rate": 5.9103749232498366e-06,
+      "loss": 0.655,
+      "step": 16107
+    },
+    {
+      "epoch": 0.8865650283449832,
+      "grad_norm": 0.9704173803329468,
+      "learning_rate": 5.9099486987591425e-06,
+      "loss": 0.911,
+      "step": 16108
+    },
+    {
+      "epoch": 0.8866200671473389,
+      "grad_norm": 0.7304208278656006,
+      "learning_rate": 5.909522467429518e-06,
+      "loss": 0.7315,
+      "step": 16109
+    },
+    {
+      "epoch": 0.8866751059496946,
+      "grad_norm": 0.6966742277145386,
+      "learning_rate": 5.909096229264164e-06,
+      "loss": 0.8658,
+      "step": 16110
+    },
+    {
+      "epoch": 0.8867301447520501,
+      "grad_norm": 0.667934238910675,
+      "learning_rate": 5.908669984266289e-06,
+      "loss": 0.6654,
+      "step": 16111
+    },
+    {
+      "epoch": 0.8867851835544058,
+      "grad_norm": 0.6689571142196655,
+      "learning_rate": 5.908243732439092e-06,
+      "loss": 0.8669,
+      "step": 16112
+    },
+    {
+      "epoch": 0.8868402223567615,
+      "grad_norm": 0.6054841876029968,
+      "learning_rate": 5.9078174737857795e-06,
+      "loss": 0.6063,
+      "step": 16113
+    },
+    {
+      "epoch": 0.8868952611591172,
+      "grad_norm": 0.6113643050193787,
+      "learning_rate": 5.907391208309553e-06,
+      "loss": 0.5609,
+      "step": 16114
+    },
+    {
+      "epoch": 0.8869502999614728,
+      "grad_norm": 0.6858495473861694,
+      "learning_rate": 5.906964936013617e-06,
+      "loss": 0.7651,
+      "step": 16115
+    },
+    {
+      "epoch": 0.8870053387638285,
+      "grad_norm": 0.6587123870849609,
+      "learning_rate": 5.906538656901175e-06,
+      "loss": 0.6834,
+      "step": 16116
+    },
+    {
+      "epoch": 0.8870603775661842,
+      "grad_norm": 0.7558240294456482,
+      "learning_rate": 5.906112370975432e-06,
+      "loss": 0.7911,
+      "step": 16117
+    },
+    {
+      "epoch": 0.8871154163685399,
+      "grad_norm": 0.7324747443199158,
+      "learning_rate": 5.90568607823959e-06,
+      "loss": 0.7132,
+      "step": 16118
+    },
+    {
+      "epoch": 0.8871704551708954,
+      "grad_norm": 0.6696536540985107,
+      "learning_rate": 5.9052597786968545e-06,
+      "loss": 0.6761,
+      "step": 16119
+    },
+    {
+      "epoch": 0.8872254939732511,
+      "grad_norm": 0.6724241375923157,
+      "learning_rate": 5.904833472350429e-06,
+      "loss": 0.7367,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8872805327756068,
+      "grad_norm": 0.8458320498466492,
+      "learning_rate": 5.904407159203517e-06,
+      "loss": 0.8542,
+      "step": 16121
+    },
+    {
+      "epoch": 0.8873355715779625,
+      "grad_norm": 0.6740517020225525,
+      "learning_rate": 5.903980839259323e-06,
+      "loss": 0.711,
+      "step": 16122
+    },
+    {
+      "epoch": 0.8873906103803181,
+      "grad_norm": 0.7465891242027283,
+      "learning_rate": 5.9035545125210505e-06,
+      "loss": 0.8501,
+      "step": 16123
+    },
+    {
+      "epoch": 0.8874456491826738,
+      "grad_norm": 0.9160736203193665,
+      "learning_rate": 5.903128178991905e-06,
+      "loss": 0.8055,
+      "step": 16124
+    },
+    {
+      "epoch": 0.8875006879850295,
+      "grad_norm": 0.8358868956565857,
+      "learning_rate": 5.902701838675089e-06,
+      "loss": 0.7946,
+      "step": 16125
+    },
+    {
+      "epoch": 0.8875557267873851,
+      "grad_norm": 0.760776162147522,
+      "learning_rate": 5.902275491573808e-06,
+      "loss": 0.7944,
+      "step": 16126
+    },
+    {
+      "epoch": 0.8876107655897407,
+      "grad_norm": 0.605964720249176,
+      "learning_rate": 5.901849137691267e-06,
+      "loss": 0.6512,
+      "step": 16127
+    },
+    {
+      "epoch": 0.8876658043920964,
+      "grad_norm": 1.4000526666641235,
+      "learning_rate": 5.9014227770306676e-06,
+      "loss": 0.8047,
+      "step": 16128
+    },
+    {
+      "epoch": 0.8877208431944521,
+      "grad_norm": 0.7314043045043945,
+      "learning_rate": 5.900996409595217e-06,
+      "loss": 0.7674,
+      "step": 16129
+    },
+    {
+      "epoch": 0.8877758819968078,
+      "grad_norm": 1.3130903244018555,
+      "learning_rate": 5.900570035388117e-06,
+      "loss": 0.6922,
+      "step": 16130
+    },
+    {
+      "epoch": 0.8878309207991634,
+      "grad_norm": 0.6799461841583252,
+      "learning_rate": 5.900143654412576e-06,
+      "loss": 0.6921,
+      "step": 16131
+    },
+    {
+      "epoch": 0.8878859596015191,
+      "grad_norm": 0.6657615900039673,
+      "learning_rate": 5.899717266671794e-06,
+      "loss": 0.642,
+      "step": 16132
+    },
+    {
+      "epoch": 0.8879409984038747,
+      "grad_norm": 0.6838696599006653,
+      "learning_rate": 5.899290872168979e-06,
+      "loss": 0.7077,
+      "step": 16133
+    },
+    {
+      "epoch": 0.8879960372062304,
+      "grad_norm": 0.7769932150840759,
+      "learning_rate": 5.898864470907334e-06,
+      "loss": 0.8155,
+      "step": 16134
+    },
+    {
+      "epoch": 0.888051076008586,
+      "grad_norm": 0.6874750852584839,
+      "learning_rate": 5.898438062890065e-06,
+      "loss": 0.7597,
+      "step": 16135
+    },
+    {
+      "epoch": 0.8881061148109417,
+      "grad_norm": 0.7016799449920654,
+      "learning_rate": 5.898011648120375e-06,
+      "loss": 0.7973,
+      "step": 16136
+    },
+    {
+      "epoch": 0.8881611536132974,
+      "grad_norm": 0.743046760559082,
+      "learning_rate": 5.897585226601471e-06,
+      "loss": 0.7217,
+      "step": 16137
+    },
+    {
+      "epoch": 0.8882161924156531,
+      "grad_norm": 0.6889417767524719,
+      "learning_rate": 5.8971587983365566e-06,
+      "loss": 0.7067,
+      "step": 16138
+    },
+    {
+      "epoch": 0.8882712312180087,
+      "grad_norm": 0.631155788898468,
+      "learning_rate": 5.896732363328836e-06,
+      "loss": 0.8113,
+      "step": 16139
+    },
+    {
+      "epoch": 0.8883262700203644,
+      "grad_norm": 0.64445960521698,
+      "learning_rate": 5.8963059215815165e-06,
+      "loss": 0.7801,
+      "step": 16140
+    },
+    {
+      "epoch": 0.88838130882272,
+      "grad_norm": 1.6496944427490234,
+      "learning_rate": 5.895879473097801e-06,
+      "loss": 0.7997,
+      "step": 16141
+    },
+    {
+      "epoch": 0.8884363476250757,
+      "grad_norm": 0.8304264545440674,
+      "learning_rate": 5.895453017880893e-06,
+      "loss": 0.7333,
+      "step": 16142
+    },
+    {
+      "epoch": 0.8884913864274313,
+      "grad_norm": 0.659909725189209,
+      "learning_rate": 5.895026555934002e-06,
+      "loss": 0.7924,
+      "step": 16143
+    },
+    {
+      "epoch": 0.888546425229787,
+      "grad_norm": 0.7013087272644043,
+      "learning_rate": 5.894600087260332e-06,
+      "loss": 0.7704,
+      "step": 16144
+    },
+    {
+      "epoch": 0.8886014640321427,
+      "grad_norm": 0.6914981603622437,
+      "learning_rate": 5.894173611863085e-06,
+      "loss": 0.7377,
+      "step": 16145
+    },
+    {
+      "epoch": 0.8886565028344984,
+      "grad_norm": 0.8310953378677368,
+      "learning_rate": 5.89374712974547e-06,
+      "loss": 0.7893,
+      "step": 16146
+    },
+    {
+      "epoch": 0.888711541636854,
+      "grad_norm": 0.6522740721702576,
+      "learning_rate": 5.8933206409106895e-06,
+      "loss": 0.6915,
+      "step": 16147
+    },
+    {
+      "epoch": 0.8887665804392096,
+      "grad_norm": 0.6072065234184265,
+      "learning_rate": 5.89289414536195e-06,
+      "loss": 0.5744,
+      "step": 16148
+    },
+    {
+      "epoch": 0.8888216192415653,
+      "grad_norm": 0.5975275635719299,
+      "learning_rate": 5.892467643102458e-06,
+      "loss": 0.6281,
+      "step": 16149
+    },
+    {
+      "epoch": 0.888876658043921,
+      "grad_norm": 0.9194470643997192,
+      "learning_rate": 5.892041134135418e-06,
+      "loss": 0.6909,
+      "step": 16150
+    },
+    {
+      "epoch": 0.8889316968462766,
+      "grad_norm": 0.5815016031265259,
+      "learning_rate": 5.891614618464037e-06,
+      "loss": 0.6342,
+      "step": 16151
+    },
+    {
+      "epoch": 0.8889867356486323,
+      "grad_norm": 0.666912853717804,
+      "learning_rate": 5.891188096091517e-06,
+      "loss": 0.8043,
+      "step": 16152
+    },
+    {
+      "epoch": 0.889041774450988,
+      "grad_norm": 0.7708194851875305,
+      "learning_rate": 5.890761567021067e-06,
+      "loss": 0.811,
+      "step": 16153
+    },
+    {
+      "epoch": 0.8890968132533436,
+      "grad_norm": 0.7158086895942688,
+      "learning_rate": 5.890335031255892e-06,
+      "loss": 0.7615,
+      "step": 16154
+    },
+    {
+      "epoch": 0.8891518520556992,
+      "grad_norm": 0.7432296872138977,
+      "learning_rate": 5.889908488799194e-06,
+      "loss": 0.8236,
+      "step": 16155
+    },
+    {
+      "epoch": 0.8892068908580549,
+      "grad_norm": 0.7223588228225708,
+      "learning_rate": 5.889481939654185e-06,
+      "loss": 0.7022,
+      "step": 16156
+    },
+    {
+      "epoch": 0.8892619296604106,
+      "grad_norm": 0.7680726647377014,
+      "learning_rate": 5.889055383824067e-06,
+      "loss": 0.7329,
+      "step": 16157
+    },
+    {
+      "epoch": 0.8893169684627662,
+      "grad_norm": 0.679315984249115,
+      "learning_rate": 5.888628821312048e-06,
+      "loss": 0.7213,
+      "step": 16158
+    },
+    {
+      "epoch": 0.8893720072651219,
+      "grad_norm": 0.9369942545890808,
+      "learning_rate": 5.88820225212133e-06,
+      "loss": 0.7661,
+      "step": 16159
+    },
+    {
+      "epoch": 0.8894270460674776,
+      "grad_norm": 0.710561990737915,
+      "learning_rate": 5.887775676255123e-06,
+      "loss": 0.7869,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8894820848698333,
+      "grad_norm": 0.6641749143600464,
+      "learning_rate": 5.887349093716632e-06,
+      "loss": 0.748,
+      "step": 16161
+    },
+    {
+      "epoch": 0.8895371236721888,
+      "grad_norm": 0.6491042971611023,
+      "learning_rate": 5.886922504509062e-06,
+      "loss": 0.7208,
+      "step": 16162
+    },
+    {
+      "epoch": 0.8895921624745445,
+      "grad_norm": 0.706950843334198,
+      "learning_rate": 5.886495908635622e-06,
+      "loss": 0.7579,
+      "step": 16163
+    },
+    {
+      "epoch": 0.8896472012769002,
+      "grad_norm": 0.7884653806686401,
+      "learning_rate": 5.886069306099514e-06,
+      "loss": 0.7289,
+      "step": 16164
+    },
+    {
+      "epoch": 0.8897022400792559,
+      "grad_norm": 0.7089719176292419,
+      "learning_rate": 5.885642696903948e-06,
+      "loss": 0.7796,
+      "step": 16165
+    },
+    {
+      "epoch": 0.8897572788816115,
+      "grad_norm": 0.7245141267776489,
+      "learning_rate": 5.8852160810521275e-06,
+      "loss": 0.7357,
+      "step": 16166
+    },
+    {
+      "epoch": 0.8898123176839672,
+      "grad_norm": 0.74881511926651,
+      "learning_rate": 5.884789458547258e-06,
+      "loss": 0.707,
+      "step": 16167
+    },
+    {
+      "epoch": 0.8898673564863229,
+      "grad_norm": 0.623418390750885,
+      "learning_rate": 5.88436282939255e-06,
+      "loss": 0.5891,
+      "step": 16168
+    },
+    {
+      "epoch": 0.8899223952886786,
+      "grad_norm": 0.8884579539299011,
+      "learning_rate": 5.883936193591208e-06,
+      "loss": 0.9109,
+      "step": 16169
+    },
+    {
+      "epoch": 0.8899774340910341,
+      "grad_norm": 0.7089982628822327,
+      "learning_rate": 5.883509551146437e-06,
+      "loss": 0.8435,
+      "step": 16170
+    },
+    {
+      "epoch": 0.8900324728933898,
+      "grad_norm": 0.6861062049865723,
+      "learning_rate": 5.883082902061444e-06,
+      "loss": 0.5662,
+      "step": 16171
+    },
+    {
+      "epoch": 0.8900875116957455,
+      "grad_norm": 0.7688663005828857,
+      "learning_rate": 5.882656246339438e-06,
+      "loss": 0.7483,
+      "step": 16172
+    },
+    {
+      "epoch": 0.8901425504981012,
+      "grad_norm": 0.6451166868209839,
+      "learning_rate": 5.882229583983623e-06,
+      "loss": 0.8061,
+      "step": 16173
+    },
+    {
+      "epoch": 0.8901975893004568,
+      "grad_norm": 0.668999195098877,
+      "learning_rate": 5.881802914997208e-06,
+      "loss": 0.698,
+      "step": 16174
+    },
+    {
+      "epoch": 0.8902526281028125,
+      "grad_norm": 0.5772761702537537,
+      "learning_rate": 5.881376239383398e-06,
+      "loss": 0.6718,
+      "step": 16175
+    },
+    {
+      "epoch": 0.8903076669051682,
+      "grad_norm": 0.6677992343902588,
+      "learning_rate": 5.880949557145399e-06,
+      "loss": 0.7835,
+      "step": 16176
+    },
+    {
+      "epoch": 0.8903627057075238,
+      "grad_norm": 0.7227941751480103,
+      "learning_rate": 5.880522868286419e-06,
+      "loss": 0.7166,
+      "step": 16177
+    },
+    {
+      "epoch": 0.8904177445098794,
+      "grad_norm": 0.7365387082099915,
+      "learning_rate": 5.880096172809665e-06,
+      "loss": 0.7743,
+      "step": 16178
+    },
+    {
+      "epoch": 0.8904727833122351,
+      "grad_norm": 0.7826401591300964,
+      "learning_rate": 5.8796694707183435e-06,
+      "loss": 0.7963,
+      "step": 16179
+    },
+    {
+      "epoch": 0.8905278221145908,
+      "grad_norm": 0.6749493479728699,
+      "learning_rate": 5.879242762015662e-06,
+      "loss": 0.7023,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8905828609169465,
+      "grad_norm": 0.7109015583992004,
+      "learning_rate": 5.8788160467048275e-06,
+      "loss": 0.8432,
+      "step": 16181
+    },
+    {
+      "epoch": 0.8906378997193021,
+      "grad_norm": 0.737983226776123,
+      "learning_rate": 5.878389324789047e-06,
+      "loss": 0.807,
+      "step": 16182
+    },
+    {
+      "epoch": 0.8906929385216578,
+      "grad_norm": 0.676296055316925,
+      "learning_rate": 5.877962596271526e-06,
+      "loss": 0.7894,
+      "step": 16183
+    },
+    {
+      "epoch": 0.8907479773240135,
+      "grad_norm": 0.6367083191871643,
+      "learning_rate": 5.877535861155474e-06,
+      "loss": 0.6995,
+      "step": 16184
+    },
+    {
+      "epoch": 0.8908030161263691,
+      "grad_norm": 0.7221261262893677,
+      "learning_rate": 5.877109119444099e-06,
+      "loss": 0.8032,
+      "step": 16185
+    },
+    {
+      "epoch": 0.8908580549287247,
+      "grad_norm": 0.9108307957649231,
+      "learning_rate": 5.8766823711406055e-06,
+      "loss": 0.6949,
+      "step": 16186
+    },
+    {
+      "epoch": 0.8909130937310804,
+      "grad_norm": 0.5985114574432373,
+      "learning_rate": 5.876255616248201e-06,
+      "loss": 0.6981,
+      "step": 16187
+    },
+    {
+      "epoch": 0.8909681325334361,
+      "grad_norm": 0.6146743297576904,
+      "learning_rate": 5.875828854770096e-06,
+      "loss": 0.6869,
+      "step": 16188
+    },
+    {
+      "epoch": 0.8910231713357918,
+      "grad_norm": 1.2942423820495605,
+      "learning_rate": 5.875402086709494e-06,
+      "loss": 0.8142,
+      "step": 16189
+    },
+    {
+      "epoch": 0.8910782101381474,
+      "grad_norm": 0.6676996350288391,
+      "learning_rate": 5.874975312069605e-06,
+      "loss": 0.7476,
+      "step": 16190
+    },
+    {
+      "epoch": 0.891133248940503,
+      "grad_norm": 0.6210917234420776,
+      "learning_rate": 5.874548530853635e-06,
+      "loss": 0.7248,
+      "step": 16191
+    },
+    {
+      "epoch": 0.8911882877428587,
+      "grad_norm": 0.7242050766944885,
+      "learning_rate": 5.874121743064792e-06,
+      "loss": 0.8378,
+      "step": 16192
+    },
+    {
+      "epoch": 0.8912433265452144,
+      "grad_norm": 0.7029538750648499,
+      "learning_rate": 5.873694948706286e-06,
+      "loss": 0.7487,
+      "step": 16193
+    },
+    {
+      "epoch": 0.89129836534757,
+      "grad_norm": 0.7620413899421692,
+      "learning_rate": 5.87326814778132e-06,
+      "loss": 0.7102,
+      "step": 16194
+    },
+    {
+      "epoch": 0.8913534041499257,
+      "grad_norm": 0.7075870633125305,
+      "learning_rate": 5.872841340293105e-06,
+      "loss": 0.7771,
+      "step": 16195
+    },
+    {
+      "epoch": 0.8914084429522814,
+      "grad_norm": 0.706533670425415,
+      "learning_rate": 5.8724145262448495e-06,
+      "loss": 0.8173,
+      "step": 16196
+    },
+    {
+      "epoch": 0.891463481754637,
+      "grad_norm": 0.6712881326675415,
+      "learning_rate": 5.871987705639759e-06,
+      "loss": 0.7933,
+      "step": 16197
+    },
+    {
+      "epoch": 0.8915185205569927,
+      "grad_norm": 0.6531795859336853,
+      "learning_rate": 5.871560878481043e-06,
+      "loss": 0.8013,
+      "step": 16198
+    },
+    {
+      "epoch": 0.8915735593593483,
+      "grad_norm": 0.7291449308395386,
+      "learning_rate": 5.8711340447719086e-06,
+      "loss": 0.7379,
+      "step": 16199
+    },
+    {
+      "epoch": 0.891628598161704,
+      "grad_norm": 0.7187185883522034,
+      "learning_rate": 5.870707204515564e-06,
+      "loss": 0.7627,
+      "step": 16200
+    },
+    {
+      "epoch": 0.8916836369640596,
+      "grad_norm": 0.6900884509086609,
+      "learning_rate": 5.870280357715217e-06,
+      "loss": 0.7779,
+      "step": 16201
+    },
+    {
+      "epoch": 0.8917386757664153,
+      "grad_norm": 0.647745668888092,
+      "learning_rate": 5.869853504374075e-06,
+      "loss": 0.7616,
+      "step": 16202
+    },
+    {
+      "epoch": 0.891793714568771,
+      "grad_norm": 0.6717308759689331,
+      "learning_rate": 5.869426644495347e-06,
+      "loss": 0.6673,
+      "step": 16203
+    },
+    {
+      "epoch": 0.8918487533711267,
+      "grad_norm": 0.7069498300552368,
+      "learning_rate": 5.868999778082242e-06,
+      "loss": 0.7349,
+      "step": 16204
+    },
+    {
+      "epoch": 0.8919037921734823,
+      "grad_norm": 0.72287917137146,
+      "learning_rate": 5.868572905137967e-06,
+      "loss": 0.7272,
+      "step": 16205
+    },
+    {
+      "epoch": 0.891958830975838,
+      "grad_norm": 0.6938319802284241,
+      "learning_rate": 5.868146025665731e-06,
+      "loss": 0.8285,
+      "step": 16206
+    },
+    {
+      "epoch": 0.8920138697781936,
+      "grad_norm": 1.5806810855865479,
+      "learning_rate": 5.867719139668739e-06,
+      "loss": 0.6881,
+      "step": 16207
+    },
+    {
+      "epoch": 0.8920689085805493,
+      "grad_norm": 0.7156746983528137,
+      "learning_rate": 5.867292247150206e-06,
+      "loss": 0.8212,
+      "step": 16208
+    },
+    {
+      "epoch": 0.8921239473829049,
+      "grad_norm": 0.6833271980285645,
+      "learning_rate": 5.866865348113335e-06,
+      "loss": 0.7741,
+      "step": 16209
+    },
+    {
+      "epoch": 0.8921789861852606,
+      "grad_norm": 0.6972640156745911,
+      "learning_rate": 5.866438442561336e-06,
+      "loss": 0.9058,
+      "step": 16210
+    },
+    {
+      "epoch": 0.8922340249876163,
+      "grad_norm": 0.697632372379303,
+      "learning_rate": 5.866011530497419e-06,
+      "loss": 0.8319,
+      "step": 16211
+    },
+    {
+      "epoch": 0.892289063789972,
+      "grad_norm": 0.7249447703361511,
+      "learning_rate": 5.865584611924789e-06,
+      "loss": 0.7491,
+      "step": 16212
+    },
+    {
+      "epoch": 0.8923441025923275,
+      "grad_norm": 0.7094838619232178,
+      "learning_rate": 5.865157686846659e-06,
+      "loss": 0.7371,
+      "step": 16213
+    },
+    {
+      "epoch": 0.8923991413946832,
+      "grad_norm": 0.7066075205802917,
+      "learning_rate": 5.864730755266233e-06,
+      "loss": 0.8273,
+      "step": 16214
+    },
+    {
+      "epoch": 0.8924541801970389,
+      "grad_norm": 0.7090823650360107,
+      "learning_rate": 5.864303817186723e-06,
+      "loss": 0.7642,
+      "step": 16215
+    },
+    {
+      "epoch": 0.8925092189993946,
+      "grad_norm": 0.7501302361488342,
+      "learning_rate": 5.863876872611337e-06,
+      "loss": 0.716,
+      "step": 16216
+    },
+    {
+      "epoch": 0.8925642578017502,
+      "grad_norm": 0.7354205250740051,
+      "learning_rate": 5.863449921543284e-06,
+      "loss": 0.7011,
+      "step": 16217
+    },
+    {
+      "epoch": 0.8926192966041059,
+      "grad_norm": 0.9364498853683472,
+      "learning_rate": 5.863022963985773e-06,
+      "loss": 0.7843,
+      "step": 16218
+    },
+    {
+      "epoch": 0.8926743354064616,
+      "grad_norm": 0.6501762270927429,
+      "learning_rate": 5.86259599994201e-06,
+      "loss": 0.7215,
+      "step": 16219
+    },
+    {
+      "epoch": 0.8927293742088173,
+      "grad_norm": 0.6883421540260315,
+      "learning_rate": 5.862169029415208e-06,
+      "loss": 0.8631,
+      "step": 16220
+    },
+    {
+      "epoch": 0.8927844130111728,
+      "grad_norm": 0.7614920735359192,
+      "learning_rate": 5.861742052408575e-06,
+      "loss": 0.8031,
+      "step": 16221
+    },
+    {
+      "epoch": 0.8928394518135285,
+      "grad_norm": 0.7668151259422302,
+      "learning_rate": 5.861315068925319e-06,
+      "loss": 0.8024,
+      "step": 16222
+    },
+    {
+      "epoch": 0.8928944906158842,
+      "grad_norm": 0.6772485971450806,
+      "learning_rate": 5.860888078968649e-06,
+      "loss": 0.72,
+      "step": 16223
+    },
+    {
+      "epoch": 0.8929495294182399,
+      "grad_norm": 0.742821216583252,
+      "learning_rate": 5.860461082541775e-06,
+      "loss": 0.7432,
+      "step": 16224
+    },
+    {
+      "epoch": 0.8930045682205955,
+      "grad_norm": 0.7056832909584045,
+      "learning_rate": 5.860034079647907e-06,
+      "loss": 0.7089,
+      "step": 16225
+    },
+    {
+      "epoch": 0.8930596070229512,
+      "grad_norm": 0.6898871660232544,
+      "learning_rate": 5.859607070290252e-06,
+      "loss": 0.7505,
+      "step": 16226
+    },
+    {
+      "epoch": 0.8931146458253069,
+      "grad_norm": 0.6888724565505981,
+      "learning_rate": 5.859180054472019e-06,
+      "loss": 0.7638,
+      "step": 16227
+    },
+    {
+      "epoch": 0.8931696846276626,
+      "grad_norm": 0.9010199308395386,
+      "learning_rate": 5.858753032196421e-06,
+      "loss": 0.775,
+      "step": 16228
+    },
+    {
+      "epoch": 0.8932247234300181,
+      "grad_norm": 0.6443523168563843,
+      "learning_rate": 5.858326003466663e-06,
+      "loss": 0.6702,
+      "step": 16229
+    },
+    {
+      "epoch": 0.8932797622323738,
+      "grad_norm": 0.6245587468147278,
+      "learning_rate": 5.857898968285957e-06,
+      "loss": 0.6907,
+      "step": 16230
+    },
+    {
+      "epoch": 0.8933348010347295,
+      "grad_norm": 0.6724962592124939,
+      "learning_rate": 5.857471926657512e-06,
+      "loss": 0.7381,
+      "step": 16231
+    },
+    {
+      "epoch": 0.8933898398370852,
+      "grad_norm": 1.0391525030136108,
+      "learning_rate": 5.857044878584539e-06,
+      "loss": 0.7919,
+      "step": 16232
+    },
+    {
+      "epoch": 0.8934448786394408,
+      "grad_norm": 0.8852080702781677,
+      "learning_rate": 5.8566178240702455e-06,
+      "loss": 0.8572,
+      "step": 16233
+    },
+    {
+      "epoch": 0.8934999174417965,
+      "grad_norm": 0.7087608575820923,
+      "learning_rate": 5.856190763117843e-06,
+      "loss": 0.7739,
+      "step": 16234
+    },
+    {
+      "epoch": 0.8935549562441522,
+      "grad_norm": 0.6688494086265564,
+      "learning_rate": 5.855763695730536e-06,
+      "loss": 0.6978,
+      "step": 16235
+    },
+    {
+      "epoch": 0.8936099950465078,
+      "grad_norm": 0.6174076795578003,
+      "learning_rate": 5.8553366219115415e-06,
+      "loss": 0.7838,
+      "step": 16236
+    },
+    {
+      "epoch": 0.8936650338488634,
+      "grad_norm": 0.6558929681777954,
+      "learning_rate": 5.854909541664065e-06,
+      "loss": 0.6071,
+      "step": 16237
+    },
+    {
+      "epoch": 0.8937200726512191,
+      "grad_norm": 0.678820013999939,
+      "learning_rate": 5.854482454991317e-06,
+      "loss": 0.792,
+      "step": 16238
+    },
+    {
+      "epoch": 0.8937751114535748,
+      "grad_norm": 0.6893227696418762,
+      "learning_rate": 5.854055361896507e-06,
+      "loss": 0.7182,
+      "step": 16239
+    },
+    {
+      "epoch": 0.8938301502559304,
+      "grad_norm": 0.6799605488777161,
+      "learning_rate": 5.853628262382847e-06,
+      "loss": 0.7011,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8938851890582861,
+      "grad_norm": 0.7625865340232849,
+      "learning_rate": 5.853201156453544e-06,
+      "loss": 0.9217,
+      "step": 16241
+    },
+    {
+      "epoch": 0.8939402278606418,
+      "grad_norm": 0.6532776355743408,
+      "learning_rate": 5.8527740441118104e-06,
+      "loss": 0.7527,
+      "step": 16242
+    },
+    {
+      "epoch": 0.8939952666629974,
+      "grad_norm": 0.7904673218727112,
+      "learning_rate": 5.852346925360854e-06,
+      "loss": 0.8359,
+      "step": 16243
+    },
+    {
+      "epoch": 0.894050305465353,
+      "grad_norm": 0.7274239659309387,
+      "learning_rate": 5.851919800203888e-06,
+      "loss": 0.7335,
+      "step": 16244
+    },
+    {
+      "epoch": 0.8941053442677087,
+      "grad_norm": 0.843180239200592,
+      "learning_rate": 5.85149266864412e-06,
+      "loss": 0.7913,
+      "step": 16245
+    },
+    {
+      "epoch": 0.8941603830700644,
+      "grad_norm": 0.7756116390228271,
+      "learning_rate": 5.851065530684763e-06,
+      "loss": 0.7508,
+      "step": 16246
+    },
+    {
+      "epoch": 0.8942154218724201,
+      "grad_norm": 0.7086586952209473,
+      "learning_rate": 5.850638386329022e-06,
+      "loss": 0.7754,
+      "step": 16247
+    },
+    {
+      "epoch": 0.8942704606747757,
+      "grad_norm": 0.9373844265937805,
+      "learning_rate": 5.850211235580112e-06,
+      "loss": 0.7391,
+      "step": 16248
+    },
+    {
+      "epoch": 0.8943254994771314,
+      "grad_norm": 0.6847782135009766,
+      "learning_rate": 5.849784078441243e-06,
+      "loss": 0.6655,
+      "step": 16249
+    },
+    {
+      "epoch": 0.894380538279487,
+      "grad_norm": 0.6071921586990356,
+      "learning_rate": 5.849356914915624e-06,
+      "loss": 0.6933,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8944355770818427,
+      "grad_norm": 0.712497889995575,
+      "learning_rate": 5.848929745006464e-06,
+      "loss": 0.8025,
+      "step": 16251
+    },
+    {
+      "epoch": 0.8944906158841983,
+      "grad_norm": 0.5942297577857971,
+      "learning_rate": 5.848502568716976e-06,
+      "loss": 0.621,
+      "step": 16252
+    },
+    {
+      "epoch": 0.894545654686554,
+      "grad_norm": 0.6706910729408264,
+      "learning_rate": 5.84807538605037e-06,
+      "loss": 0.7664,
+      "step": 16253
+    },
+    {
+      "epoch": 0.8946006934889097,
+      "grad_norm": 0.7494041919708252,
+      "learning_rate": 5.847648197009858e-06,
+      "loss": 0.7418,
+      "step": 16254
+    },
+    {
+      "epoch": 0.8946557322912654,
+      "grad_norm": 0.7373181581497192,
+      "learning_rate": 5.847221001598646e-06,
+      "loss": 0.7133,
+      "step": 16255
+    },
+    {
+      "epoch": 0.894710771093621,
+      "grad_norm": 0.8178310394287109,
+      "learning_rate": 5.84679379981995e-06,
+      "loss": 0.8562,
+      "step": 16256
+    },
+    {
+      "epoch": 0.8947658098959766,
+      "grad_norm": 0.6232174634933472,
+      "learning_rate": 5.8463665916769785e-06,
+      "loss": 0.7723,
+      "step": 16257
+    },
+    {
+      "epoch": 0.8948208486983323,
+      "grad_norm": 0.6817423701286316,
+      "learning_rate": 5.845939377172942e-06,
+      "loss": 0.6706,
+      "step": 16258
+    },
+    {
+      "epoch": 0.894875887500688,
+      "grad_norm": 0.8091211318969727,
+      "learning_rate": 5.845512156311051e-06,
+      "loss": 0.7453,
+      "step": 16259
+    },
+    {
+      "epoch": 0.8949309263030436,
+      "grad_norm": 0.7013124227523804,
+      "learning_rate": 5.845084929094518e-06,
+      "loss": 0.7104,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8949859651053993,
+      "grad_norm": 0.8373565077781677,
+      "learning_rate": 5.844657695526552e-06,
+      "loss": 0.8181,
+      "step": 16261
+    },
+    {
+      "epoch": 0.895041003907755,
+      "grad_norm": 0.7339033484458923,
+      "learning_rate": 5.844230455610364e-06,
+      "loss": 0.7465,
+      "step": 16262
+    },
+    {
+      "epoch": 0.8950960427101107,
+      "grad_norm": 0.7680185437202454,
+      "learning_rate": 5.843803209349167e-06,
+      "loss": 0.7962,
+      "step": 16263
+    },
+    {
+      "epoch": 0.8951510815124663,
+      "grad_norm": 0.7889038920402527,
+      "learning_rate": 5.843375956746171e-06,
+      "loss": 0.7846,
+      "step": 16264
+    },
+    {
+      "epoch": 0.8952061203148219,
+      "grad_norm": 0.7361034154891968,
+      "learning_rate": 5.842948697804587e-06,
+      "loss": 0.7077,
+      "step": 16265
+    },
+    {
+      "epoch": 0.8952611591171776,
+      "grad_norm": 0.7543736100196838,
+      "learning_rate": 5.8425214325276255e-06,
+      "loss": 0.748,
+      "step": 16266
+    },
+    {
+      "epoch": 0.8953161979195333,
+      "grad_norm": 0.6194653511047363,
+      "learning_rate": 5.842094160918499e-06,
+      "loss": 0.8032,
+      "step": 16267
+    },
+    {
+      "epoch": 0.8953712367218889,
+      "grad_norm": 0.789439857006073,
+      "learning_rate": 5.841666882980418e-06,
+      "loss": 0.745,
+      "step": 16268
+    },
+    {
+      "epoch": 0.8954262755242446,
+      "grad_norm": 0.6813651919364929,
+      "learning_rate": 5.841239598716595e-06,
+      "loss": 0.7169,
+      "step": 16269
+    },
+    {
+      "epoch": 0.8954813143266003,
+      "grad_norm": 0.6128388047218323,
+      "learning_rate": 5.84081230813024e-06,
+      "loss": 0.7109,
+      "step": 16270
+    },
+    {
+      "epoch": 0.895536353128956,
+      "grad_norm": 0.7170562148094177,
+      "learning_rate": 5.8403850112245645e-06,
+      "loss": 0.7461,
+      "step": 16271
+    },
+    {
+      "epoch": 0.8955913919313115,
+      "grad_norm": 0.7415544986724854,
+      "learning_rate": 5.83995770800278e-06,
+      "loss": 0.7786,
+      "step": 16272
+    },
+    {
+      "epoch": 0.8956464307336672,
+      "grad_norm": 0.5996596813201904,
+      "learning_rate": 5.8395303984680985e-06,
+      "loss": 0.6926,
+      "step": 16273
+    },
+    {
+      "epoch": 0.8957014695360229,
+      "grad_norm": 0.7399057149887085,
+      "learning_rate": 5.839103082623732e-06,
+      "loss": 0.6953,
+      "step": 16274
+    },
+    {
+      "epoch": 0.8957565083383786,
+      "grad_norm": 0.7090675830841064,
+      "learning_rate": 5.838675760472888e-06,
+      "loss": 0.7638,
+      "step": 16275
+    },
+    {
+      "epoch": 0.8958115471407342,
+      "grad_norm": 0.6865240931510925,
+      "learning_rate": 5.838248432018785e-06,
+      "loss": 0.7023,
+      "step": 16276
+    },
+    {
+      "epoch": 0.8958665859430899,
+      "grad_norm": 0.604603111743927,
+      "learning_rate": 5.83782109726463e-06,
+      "loss": 0.6845,
+      "step": 16277
+    },
+    {
+      "epoch": 0.8959216247454456,
+      "grad_norm": 0.6722466349601746,
+      "learning_rate": 5.837393756213636e-06,
+      "loss": 0.7794,
+      "step": 16278
+    },
+    {
+      "epoch": 0.8959766635478013,
+      "grad_norm": 0.683106541633606,
+      "learning_rate": 5.836966408869014e-06,
+      "loss": 0.7988,
+      "step": 16279
+    },
+    {
+      "epoch": 0.8960317023501568,
+      "grad_norm": 0.7195246815681458,
+      "learning_rate": 5.8365390552339774e-06,
+      "loss": 0.7532,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8960867411525125,
+      "grad_norm": 0.6945170760154724,
+      "learning_rate": 5.836111695311737e-06,
+      "loss": 0.7515,
+      "step": 16281
+    },
+    {
+      "epoch": 0.8961417799548682,
+      "grad_norm": 0.7424500584602356,
+      "learning_rate": 5.8356843291055065e-06,
+      "loss": 0.6813,
+      "step": 16282
+    },
+    {
+      "epoch": 0.8961968187572238,
+      "grad_norm": 0.673574686050415,
+      "learning_rate": 5.835256956618495e-06,
+      "loss": 0.7234,
+      "step": 16283
+    },
+    {
+      "epoch": 0.8962518575595795,
+      "grad_norm": 0.6816020011901855,
+      "learning_rate": 5.834829577853913e-06,
+      "loss": 0.7935,
+      "step": 16284
+    },
+    {
+      "epoch": 0.8963068963619352,
+      "grad_norm": 0.7598507404327393,
+      "learning_rate": 5.834402192814979e-06,
+      "loss": 0.8141,
+      "step": 16285
+    },
+    {
+      "epoch": 0.8963619351642909,
+      "grad_norm": 0.7720094323158264,
+      "learning_rate": 5.8339748015049e-06,
+      "loss": 0.8241,
+      "step": 16286
+    },
+    {
+      "epoch": 0.8964169739666464,
+      "grad_norm": 0.7409939169883728,
+      "learning_rate": 5.833547403926891e-06,
+      "loss": 0.7196,
+      "step": 16287
+    },
+    {
+      "epoch": 0.8964720127690021,
+      "grad_norm": 0.670557975769043,
+      "learning_rate": 5.83312000008416e-06,
+      "loss": 0.7931,
+      "step": 16288
+    },
+    {
+      "epoch": 0.8965270515713578,
+      "grad_norm": 0.6361322999000549,
+      "learning_rate": 5.832692589979925e-06,
+      "loss": 0.7292,
+      "step": 16289
+    },
+    {
+      "epoch": 0.8965820903737135,
+      "grad_norm": 0.6359429359436035,
+      "learning_rate": 5.832265173617393e-06,
+      "loss": 0.7705,
+      "step": 16290
+    },
+    {
+      "epoch": 0.8966371291760691,
+      "grad_norm": 0.7249873876571655,
+      "learning_rate": 5.831837750999781e-06,
+      "loss": 0.7741,
+      "step": 16291
+    },
+    {
+      "epoch": 0.8966921679784248,
+      "grad_norm": 0.6784750819206238,
+      "learning_rate": 5.831410322130296e-06,
+      "loss": 0.8434,
+      "step": 16292
+    },
+    {
+      "epoch": 0.8967472067807805,
+      "grad_norm": 0.7696726322174072,
+      "learning_rate": 5.830982887012157e-06,
+      "loss": 0.8218,
+      "step": 16293
+    },
+    {
+      "epoch": 0.8968022455831361,
+      "grad_norm": 0.5974952578544617,
+      "learning_rate": 5.830555445648572e-06,
+      "loss": 0.5864,
+      "step": 16294
+    },
+    {
+      "epoch": 0.8968572843854917,
+      "grad_norm": 1.4088029861450195,
+      "learning_rate": 5.830127998042755e-06,
+      "loss": 0.705,
+      "step": 16295
+    },
+    {
+      "epoch": 0.8969123231878474,
+      "grad_norm": 0.621288001537323,
+      "learning_rate": 5.8297005441979174e-06,
+      "loss": 0.6799,
+      "step": 16296
+    },
+    {
+      "epoch": 0.8969673619902031,
+      "grad_norm": 0.7229657173156738,
+      "learning_rate": 5.829273084117272e-06,
+      "loss": 0.7222,
+      "step": 16297
+    },
+    {
+      "epoch": 0.8970224007925588,
+      "grad_norm": 0.7076373100280762,
+      "learning_rate": 5.828845617804033e-06,
+      "loss": 0.7813,
+      "step": 16298
+    },
+    {
+      "epoch": 0.8970774395949144,
+      "grad_norm": 0.6931923627853394,
+      "learning_rate": 5.828418145261412e-06,
+      "loss": 0.6936,
+      "step": 16299
+    },
+    {
+      "epoch": 0.8971324783972701,
+      "grad_norm": 0.6719018220901489,
+      "learning_rate": 5.827990666492621e-06,
+      "loss": 0.747,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8971875171996258,
+      "grad_norm": 0.6288262605667114,
+      "learning_rate": 5.827563181500875e-06,
+      "loss": 0.7731,
+      "step": 16301
+    },
+    {
+      "epoch": 0.8972425560019814,
+      "grad_norm": 0.6359015703201294,
+      "learning_rate": 5.8271356902893864e-06,
+      "loss": 0.6408,
+      "step": 16302
+    },
+    {
+      "epoch": 0.897297594804337,
+      "grad_norm": 0.6485893726348877,
+      "learning_rate": 5.826708192861365e-06,
+      "loss": 0.7739,
+      "step": 16303
+    },
+    {
+      "epoch": 0.8973526336066927,
+      "grad_norm": 0.7622523903846741,
+      "learning_rate": 5.826280689220027e-06,
+      "loss": 0.6087,
+      "step": 16304
+    },
+    {
+      "epoch": 0.8974076724090484,
+      "grad_norm": 0.650451123714447,
+      "learning_rate": 5.825853179368586e-06,
+      "loss": 0.6953,
+      "step": 16305
+    },
+    {
+      "epoch": 0.8974627112114041,
+      "grad_norm": 0.7266152501106262,
+      "learning_rate": 5.8254256633102535e-06,
+      "loss": 0.7762,
+      "step": 16306
+    },
+    {
+      "epoch": 0.8975177500137597,
+      "grad_norm": 0.6428011059761047,
+      "learning_rate": 5.824998141048241e-06,
+      "loss": 0.67,
+      "step": 16307
+    },
+    {
+      "epoch": 0.8975727888161154,
+      "grad_norm": 0.6991005539894104,
+      "learning_rate": 5.824570612585764e-06,
+      "loss": 0.79,
+      "step": 16308
+    },
+    {
+      "epoch": 0.897627827618471,
+      "grad_norm": 0.6385177969932556,
+      "learning_rate": 5.824143077926034e-06,
+      "loss": 0.6993,
+      "step": 16309
+    },
+    {
+      "epoch": 0.8976828664208267,
+      "grad_norm": 0.6891354322433472,
+      "learning_rate": 5.823715537072268e-06,
+      "loss": 0.7155,
+      "step": 16310
+    },
+    {
+      "epoch": 0.8977379052231823,
+      "grad_norm": 0.7448866367340088,
+      "learning_rate": 5.823287990027674e-06,
+      "loss": 0.674,
+      "step": 16311
+    },
+    {
+      "epoch": 0.897792944025538,
+      "grad_norm": 0.6892699599266052,
+      "learning_rate": 5.822860436795468e-06,
+      "loss": 0.7719,
+      "step": 16312
+    },
+    {
+      "epoch": 0.8978479828278937,
+      "grad_norm": 1.2621982097625732,
+      "learning_rate": 5.822432877378864e-06,
+      "loss": 0.6985,
+      "step": 16313
+    },
+    {
+      "epoch": 0.8979030216302494,
+      "grad_norm": 0.635137677192688,
+      "learning_rate": 5.822005311781075e-06,
+      "loss": 0.748,
+      "step": 16314
+    },
+    {
+      "epoch": 0.897958060432605,
+      "grad_norm": 0.8765038847923279,
+      "learning_rate": 5.821577740005313e-06,
+      "loss": 0.798,
+      "step": 16315
+    },
+    {
+      "epoch": 0.8980130992349606,
+      "grad_norm": 0.734259843826294,
+      "learning_rate": 5.8211501620547926e-06,
+      "loss": 0.7462,
+      "step": 16316
+    },
+    {
+      "epoch": 0.8980681380373163,
+      "grad_norm": 0.7057023048400879,
+      "learning_rate": 5.820722577932729e-06,
+      "loss": 0.8057,
+      "step": 16317
+    },
+    {
+      "epoch": 0.898123176839672,
+      "grad_norm": 0.7444988489151001,
+      "learning_rate": 5.8202949876423344e-06,
+      "loss": 0.7023,
+      "step": 16318
+    },
+    {
+      "epoch": 0.8981782156420276,
+      "grad_norm": 0.7205658555030823,
+      "learning_rate": 5.819867391186821e-06,
+      "loss": 0.6949,
+      "step": 16319
+    },
+    {
+      "epoch": 0.8982332544443833,
+      "grad_norm": 0.7285442352294922,
+      "learning_rate": 5.819439788569403e-06,
+      "loss": 0.6966,
+      "step": 16320
+    },
+    {
+      "epoch": 0.898288293246739,
+      "grad_norm": 0.7485014796257019,
+      "learning_rate": 5.819012179793295e-06,
+      "loss": 0.7676,
+      "step": 16321
+    },
+    {
+      "epoch": 0.8983433320490947,
+      "grad_norm": 0.7867493629455566,
+      "learning_rate": 5.818584564861712e-06,
+      "loss": 0.7601,
+      "step": 16322
+    },
+    {
+      "epoch": 0.8983983708514502,
+      "grad_norm": 0.6732510328292847,
+      "learning_rate": 5.818156943777867e-06,
+      "loss": 0.7068,
+      "step": 16323
+    },
+    {
+      "epoch": 0.8984534096538059,
+      "grad_norm": 0.648333728313446,
+      "learning_rate": 5.817729316544971e-06,
+      "loss": 0.762,
+      "step": 16324
+    },
+    {
+      "epoch": 0.8985084484561616,
+      "grad_norm": 0.9556308388710022,
+      "learning_rate": 5.817301683166241e-06,
+      "loss": 0.7177,
+      "step": 16325
+    },
+    {
+      "epoch": 0.8985634872585172,
+      "grad_norm": 0.7043321132659912,
+      "learning_rate": 5.816874043644891e-06,
+      "loss": 0.7206,
+      "step": 16326
+    },
+    {
+      "epoch": 0.8986185260608729,
+      "grad_norm": 0.6318387985229492,
+      "learning_rate": 5.816446397984136e-06,
+      "loss": 0.7306,
+      "step": 16327
+    },
+    {
+      "epoch": 0.8986735648632286,
+      "grad_norm": 0.7083125114440918,
+      "learning_rate": 5.816018746187186e-06,
+      "loss": 0.7204,
+      "step": 16328
+    },
+    {
+      "epoch": 0.8987286036655843,
+      "grad_norm": 0.6810079216957092,
+      "learning_rate": 5.815591088257259e-06,
+      "loss": 0.8575,
+      "step": 16329
+    },
+    {
+      "epoch": 0.8987836424679398,
+      "grad_norm": 0.7081509232521057,
+      "learning_rate": 5.815163424197567e-06,
+      "loss": 0.7198,
+      "step": 16330
+    },
+    {
+      "epoch": 0.8988386812702955,
+      "grad_norm": 1.1525241136550903,
+      "learning_rate": 5.814735754011325e-06,
+      "loss": 0.6222,
+      "step": 16331
+    },
+    {
+      "epoch": 0.8988937200726512,
+      "grad_norm": 0.724651575088501,
+      "learning_rate": 5.8143080777017456e-06,
+      "loss": 0.6937,
+      "step": 16332
+    },
+    {
+      "epoch": 0.8989487588750069,
+      "grad_norm": 0.7607846260070801,
+      "learning_rate": 5.813880395272047e-06,
+      "loss": 0.7552,
+      "step": 16333
+    },
+    {
+      "epoch": 0.8990037976773625,
+      "grad_norm": 0.6370975375175476,
+      "learning_rate": 5.813452706725441e-06,
+      "loss": 0.753,
+      "step": 16334
+    },
+    {
+      "epoch": 0.8990588364797182,
+      "grad_norm": 0.7018759250640869,
+      "learning_rate": 5.813025012065141e-06,
+      "loss": 0.8026,
+      "step": 16335
+    },
+    {
+      "epoch": 0.8991138752820739,
+      "grad_norm": 0.7106475234031677,
+      "learning_rate": 5.812597311294363e-06,
+      "loss": 0.6558,
+      "step": 16336
+    },
+    {
+      "epoch": 0.8991689140844296,
+      "grad_norm": 0.715859591960907,
+      "learning_rate": 5.812169604416321e-06,
+      "loss": 0.6581,
+      "step": 16337
+    },
+    {
+      "epoch": 0.8992239528867851,
+      "grad_norm": 1.1907461881637573,
+      "learning_rate": 5.811741891434231e-06,
+      "loss": 0.7618,
+      "step": 16338
+    },
+    {
+      "epoch": 0.8992789916891408,
+      "grad_norm": 0.6529675722122192,
+      "learning_rate": 5.811314172351304e-06,
+      "loss": 0.7317,
+      "step": 16339
+    },
+    {
+      "epoch": 0.8993340304914965,
+      "grad_norm": 0.6876475811004639,
+      "learning_rate": 5.810886447170758e-06,
+      "loss": 0.724,
+      "step": 16340
+    },
+    {
+      "epoch": 0.8993890692938522,
+      "grad_norm": 0.6798568964004517,
+      "learning_rate": 5.8104587158958084e-06,
+      "loss": 0.7173,
+      "step": 16341
+    },
+    {
+      "epoch": 0.8994441080962078,
+      "grad_norm": 1.0269527435302734,
+      "learning_rate": 5.8100309785296664e-06,
+      "loss": 0.737,
+      "step": 16342
+    },
+    {
+      "epoch": 0.8994991468985635,
+      "grad_norm": 0.7578931450843811,
+      "learning_rate": 5.809603235075547e-06,
+      "loss": 0.8119,
+      "step": 16343
+    },
+    {
+      "epoch": 0.8995541857009192,
+      "grad_norm": 0.7005903124809265,
+      "learning_rate": 5.8091754855366675e-06,
+      "loss": 0.7548,
+      "step": 16344
+    },
+    {
+      "epoch": 0.8996092245032749,
+      "grad_norm": 0.6554011702537537,
+      "learning_rate": 5.808747729916242e-06,
+      "loss": 0.6705,
+      "step": 16345
+    },
+    {
+      "epoch": 0.8996642633056304,
+      "grad_norm": 0.6242305040359497,
+      "learning_rate": 5.808319968217485e-06,
+      "loss": 0.7278,
+      "step": 16346
+    },
+    {
+      "epoch": 0.8997193021079861,
+      "grad_norm": 0.9426488876342773,
+      "learning_rate": 5.807892200443611e-06,
+      "loss": 0.7129,
+      "step": 16347
+    },
+    {
+      "epoch": 0.8997743409103418,
+      "grad_norm": 0.671768307685852,
+      "learning_rate": 5.807464426597835e-06,
+      "loss": 0.7688,
+      "step": 16348
+    },
+    {
+      "epoch": 0.8998293797126975,
+      "grad_norm": 0.672828197479248,
+      "learning_rate": 5.807036646683374e-06,
+      "loss": 0.8089,
+      "step": 16349
+    },
+    {
+      "epoch": 0.8998844185150531,
+      "grad_norm": 0.6689574718475342,
+      "learning_rate": 5.806608860703441e-06,
+      "loss": 0.7523,
+      "step": 16350
+    },
+    {
+      "epoch": 0.8999394573174088,
+      "grad_norm": 0.7440112233161926,
+      "learning_rate": 5.8061810686612514e-06,
+      "loss": 0.7802,
+      "step": 16351
+    },
+    {
+      "epoch": 0.8999944961197645,
+      "grad_norm": 0.6520549654960632,
+      "learning_rate": 5.8057532705600206e-06,
+      "loss": 0.6582,
+      "step": 16352
+    },
+    {
+      "epoch": 0.9000495349221201,
+      "grad_norm": 0.73201984167099,
+      "learning_rate": 5.805325466402965e-06,
+      "loss": 0.685,
+      "step": 16353
+    },
+    {
+      "epoch": 0.9001045737244757,
+      "grad_norm": 0.6377021074295044,
+      "learning_rate": 5.804897656193298e-06,
+      "loss": 0.802,
+      "step": 16354
+    },
+    {
+      "epoch": 0.9001596125268314,
+      "grad_norm": 0.6761305928230286,
+      "learning_rate": 5.8044698399342355e-06,
+      "loss": 0.7005,
+      "step": 16355
+    },
+    {
+      "epoch": 0.9002146513291871,
+      "grad_norm": 0.6993832588195801,
+      "learning_rate": 5.804042017628992e-06,
+      "loss": 0.848,
+      "step": 16356
+    },
+    {
+      "epoch": 0.9002696901315428,
+      "grad_norm": 0.6453230381011963,
+      "learning_rate": 5.803614189280786e-06,
+      "loss": 0.7525,
+      "step": 16357
+    },
+    {
+      "epoch": 0.9003247289338984,
+      "grad_norm": 0.620090663433075,
+      "learning_rate": 5.80318635489283e-06,
+      "loss": 0.668,
+      "step": 16358
+    },
+    {
+      "epoch": 0.900379767736254,
+      "grad_norm": 0.7285178899765015,
+      "learning_rate": 5.80275851446834e-06,
+      "loss": 0.7936,
+      "step": 16359
+    },
+    {
+      "epoch": 0.9004348065386097,
+      "grad_norm": 0.7555778622627258,
+      "learning_rate": 5.802330668010532e-06,
+      "loss": 0.7493,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9004898453409654,
+      "grad_norm": 0.8527930974960327,
+      "learning_rate": 5.801902815522622e-06,
+      "loss": 0.7609,
+      "step": 16361
+    },
+    {
+      "epoch": 0.900544884143321,
+      "grad_norm": 0.9000833034515381,
+      "learning_rate": 5.801474957007824e-06,
+      "loss": 0.7719,
+      "step": 16362
+    },
+    {
+      "epoch": 0.9005999229456767,
+      "grad_norm": 0.7018047571182251,
+      "learning_rate": 5.801047092469356e-06,
+      "loss": 0.9116,
+      "step": 16363
+    },
+    {
+      "epoch": 0.9006549617480324,
+      "grad_norm": 0.7304424047470093,
+      "learning_rate": 5.800619221910432e-06,
+      "loss": 0.7339,
+      "step": 16364
+    },
+    {
+      "epoch": 0.900710000550388,
+      "grad_norm": 0.6775519251823425,
+      "learning_rate": 5.80019134533427e-06,
+      "loss": 0.6962,
+      "step": 16365
+    },
+    {
+      "epoch": 0.9007650393527437,
+      "grad_norm": 0.717262864112854,
+      "learning_rate": 5.799763462744081e-06,
+      "loss": 0.7518,
+      "step": 16366
+    },
+    {
+      "epoch": 0.9008200781550993,
+      "grad_norm": 0.9107728004455566,
+      "learning_rate": 5.799335574143086e-06,
+      "loss": 0.7778,
+      "step": 16367
+    },
+    {
+      "epoch": 0.900875116957455,
+      "grad_norm": 0.760864794254303,
+      "learning_rate": 5.798907679534498e-06,
+      "loss": 0.8022,
+      "step": 16368
+    },
+    {
+      "epoch": 0.9009301557598106,
+      "grad_norm": 0.6746829152107239,
+      "learning_rate": 5.798479778921533e-06,
+      "loss": 0.7105,
+      "step": 16369
+    },
+    {
+      "epoch": 0.9009851945621663,
+      "grad_norm": 0.6654032468795776,
+      "learning_rate": 5.798051872307409e-06,
+      "loss": 0.7946,
+      "step": 16370
+    },
+    {
+      "epoch": 0.901040233364522,
+      "grad_norm": 0.6802057027816772,
+      "learning_rate": 5.79762395969534e-06,
+      "loss": 0.7332,
+      "step": 16371
+    },
+    {
+      "epoch": 0.9010952721668777,
+      "grad_norm": 0.6586911678314209,
+      "learning_rate": 5.797196041088542e-06,
+      "loss": 0.7477,
+      "step": 16372
+    },
+    {
+      "epoch": 0.9011503109692333,
+      "grad_norm": 0.7057276964187622,
+      "learning_rate": 5.796768116490233e-06,
+      "loss": 0.7468,
+      "step": 16373
+    },
+    {
+      "epoch": 0.901205349771589,
+      "grad_norm": 0.6421266794204712,
+      "learning_rate": 5.796340185903629e-06,
+      "loss": 0.7073,
+      "step": 16374
+    },
+    {
+      "epoch": 0.9012603885739446,
+      "grad_norm": 0.752708911895752,
+      "learning_rate": 5.795912249331944e-06,
+      "loss": 0.723,
+      "step": 16375
+    },
+    {
+      "epoch": 0.9013154273763003,
+      "grad_norm": 0.6225395798683167,
+      "learning_rate": 5.795484306778395e-06,
+      "loss": 0.6317,
+      "step": 16376
+    },
+    {
+      "epoch": 0.9013704661786559,
+      "grad_norm": 0.6598436236381531,
+      "learning_rate": 5.795056358246201e-06,
+      "loss": 0.7429,
+      "step": 16377
+    },
+    {
+      "epoch": 0.9014255049810116,
+      "grad_norm": 0.6590927243232727,
+      "learning_rate": 5.794628403738576e-06,
+      "loss": 0.8009,
+      "step": 16378
+    },
+    {
+      "epoch": 0.9014805437833673,
+      "grad_norm": 0.7180964350700378,
+      "learning_rate": 5.794200443258736e-06,
+      "loss": 0.7438,
+      "step": 16379
+    },
+    {
+      "epoch": 0.901535582585723,
+      "grad_norm": 0.7282911539077759,
+      "learning_rate": 5.7937724768098965e-06,
+      "loss": 0.8067,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9015906213880786,
+      "grad_norm": 0.7015960812568665,
+      "learning_rate": 5.793344504395278e-06,
+      "loss": 0.7094,
+      "step": 16381
+    },
+    {
+      "epoch": 0.9016456601904342,
+      "grad_norm": 0.7785707116127014,
+      "learning_rate": 5.792916526018093e-06,
+      "loss": 0.6816,
+      "step": 16382
+    },
+    {
+      "epoch": 0.9017006989927899,
+      "grad_norm": 0.5919010639190674,
+      "learning_rate": 5.79248854168156e-06,
+      "loss": 0.6592,
+      "step": 16383
+    },
+    {
+      "epoch": 0.9017557377951456,
+      "grad_norm": 0.8561486005783081,
+      "learning_rate": 5.792060551388894e-06,
+      "loss": 0.8305,
+      "step": 16384
+    },
+    {
+      "epoch": 0.9018107765975012,
+      "grad_norm": 0.6228078603744507,
+      "learning_rate": 5.791632555143315e-06,
+      "loss": 0.6535,
+      "step": 16385
+    },
+    {
+      "epoch": 0.9018658153998569,
+      "grad_norm": 0.7751365303993225,
+      "learning_rate": 5.791204552948037e-06,
+      "loss": 0.7738,
+      "step": 16386
+    },
+    {
+      "epoch": 0.9019208542022126,
+      "grad_norm": 0.6248418092727661,
+      "learning_rate": 5.790776544806277e-06,
+      "loss": 0.7464,
+      "step": 16387
+    },
+    {
+      "epoch": 0.9019758930045683,
+      "grad_norm": 0.6654400825500488,
+      "learning_rate": 5.790348530721253e-06,
+      "loss": 0.7345,
+      "step": 16388
+    },
+    {
+      "epoch": 0.9020309318069238,
+      "grad_norm": 0.7654403448104858,
+      "learning_rate": 5.789920510696181e-06,
+      "loss": 0.8147,
+      "step": 16389
+    },
+    {
+      "epoch": 0.9020859706092795,
+      "grad_norm": 0.630255937576294,
+      "learning_rate": 5.789492484734277e-06,
+      "loss": 0.7007,
+      "step": 16390
+    },
+    {
+      "epoch": 0.9021410094116352,
+      "grad_norm": 0.7767404317855835,
+      "learning_rate": 5.7890644528387595e-06,
+      "loss": 0.7547,
+      "step": 16391
+    },
+    {
+      "epoch": 0.9021960482139909,
+      "grad_norm": 0.8416337966918945,
+      "learning_rate": 5.788636415012842e-06,
+      "loss": 0.8771,
+      "step": 16392
+    },
+    {
+      "epoch": 0.9022510870163465,
+      "grad_norm": 0.6467620134353638,
+      "learning_rate": 5.7882083712597485e-06,
+      "loss": 0.651,
+      "step": 16393
+    },
+    {
+      "epoch": 0.9023061258187022,
+      "grad_norm": 0.8612825870513916,
+      "learning_rate": 5.78778032158269e-06,
+      "loss": 0.8127,
+      "step": 16394
+    },
+    {
+      "epoch": 0.9023611646210579,
+      "grad_norm": 0.7267431020736694,
+      "learning_rate": 5.787352265984886e-06,
+      "loss": 0.7616,
+      "step": 16395
+    },
+    {
+      "epoch": 0.9024162034234136,
+      "grad_norm": 0.736853837966919,
+      "learning_rate": 5.786924204469551e-06,
+      "loss": 0.7907,
+      "step": 16396
+    },
+    {
+      "epoch": 0.9024712422257691,
+      "grad_norm": 0.6155600547790527,
+      "learning_rate": 5.7864961370399065e-06,
+      "loss": 0.6326,
+      "step": 16397
+    },
+    {
+      "epoch": 0.9025262810281248,
+      "grad_norm": 0.7266953587532043,
+      "learning_rate": 5.7860680636991675e-06,
+      "loss": 0.817,
+      "step": 16398
+    },
+    {
+      "epoch": 0.9025813198304805,
+      "grad_norm": 0.7350000143051147,
+      "learning_rate": 5.785639984450551e-06,
+      "loss": 0.6602,
+      "step": 16399
+    },
+    {
+      "epoch": 0.9026363586328362,
+      "grad_norm": 1.1884604692459106,
+      "learning_rate": 5.785211899297274e-06,
+      "loss": 0.8311,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9026913974351918,
+      "grad_norm": 0.7405989170074463,
+      "learning_rate": 5.784783808242555e-06,
+      "loss": 0.8444,
+      "step": 16401
+    },
+    {
+      "epoch": 0.9027464362375475,
+      "grad_norm": 0.9071193933486938,
+      "learning_rate": 5.7843557112896124e-06,
+      "loss": 0.8302,
+      "step": 16402
+    },
+    {
+      "epoch": 0.9028014750399032,
+      "grad_norm": 0.6949760317802429,
+      "learning_rate": 5.783927608441661e-06,
+      "loss": 0.8151,
+      "step": 16403
+    },
+    {
+      "epoch": 0.9028565138422588,
+      "grad_norm": 0.7762002348899841,
+      "learning_rate": 5.783499499701919e-06,
+      "loss": 0.8362,
+      "step": 16404
+    },
+    {
+      "epoch": 0.9029115526446144,
+      "grad_norm": 0.6971734762191772,
+      "learning_rate": 5.783071385073604e-06,
+      "loss": 0.7524,
+      "step": 16405
+    },
+    {
+      "epoch": 0.9029665914469701,
+      "grad_norm": 0.6840386390686035,
+      "learning_rate": 5.782643264559936e-06,
+      "loss": 0.7548,
+      "step": 16406
+    },
+    {
+      "epoch": 0.9030216302493258,
+      "grad_norm": 0.7312326431274414,
+      "learning_rate": 5.78221513816413e-06,
+      "loss": 0.7998,
+      "step": 16407
+    },
+    {
+      "epoch": 0.9030766690516814,
+      "grad_norm": 0.6572339534759521,
+      "learning_rate": 5.781787005889403e-06,
+      "loss": 0.7044,
+      "step": 16408
+    },
+    {
+      "epoch": 0.9031317078540371,
+      "grad_norm": 0.6981173157691956,
+      "learning_rate": 5.7813588677389755e-06,
+      "loss": 0.8207,
+      "step": 16409
+    },
+    {
+      "epoch": 0.9031867466563928,
+      "grad_norm": 0.6956539750099182,
+      "learning_rate": 5.780930723716064e-06,
+      "loss": 0.7379,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9032417854587484,
+      "grad_norm": 0.6781721115112305,
+      "learning_rate": 5.780502573823886e-06,
+      "loss": 0.7051,
+      "step": 16411
+    },
+    {
+      "epoch": 0.903296824261104,
+      "grad_norm": 0.8202298283576965,
+      "learning_rate": 5.7800744180656596e-06,
+      "loss": 0.8002,
+      "step": 16412
+    },
+    {
+      "epoch": 0.9033518630634597,
+      "grad_norm": 0.6392808556556702,
+      "learning_rate": 5.779646256444603e-06,
+      "loss": 0.7204,
+      "step": 16413
+    },
+    {
+      "epoch": 0.9034069018658154,
+      "grad_norm": 0.6195122003555298,
+      "learning_rate": 5.779218088963934e-06,
+      "loss": 0.7297,
+      "step": 16414
+    },
+    {
+      "epoch": 0.9034619406681711,
+      "grad_norm": 0.6970216631889343,
+      "learning_rate": 5.77878991562687e-06,
+      "loss": 0.7572,
+      "step": 16415
+    },
+    {
+      "epoch": 0.9035169794705267,
+      "grad_norm": 0.6491253972053528,
+      "learning_rate": 5.778361736436628e-06,
+      "loss": 0.609,
+      "step": 16416
+    },
+    {
+      "epoch": 0.9035720182728824,
+      "grad_norm": 0.745171070098877,
+      "learning_rate": 5.777933551396429e-06,
+      "loss": 0.7413,
+      "step": 16417
+    },
+    {
+      "epoch": 0.903627057075238,
+      "grad_norm": 0.7239416241645813,
+      "learning_rate": 5.77750536050949e-06,
+      "loss": 0.7539,
+      "step": 16418
+    },
+    {
+      "epoch": 0.9036820958775937,
+      "grad_norm": 0.7006568312644958,
+      "learning_rate": 5.777077163779028e-06,
+      "loss": 0.7071,
+      "step": 16419
+    },
+    {
+      "epoch": 0.9037371346799493,
+      "grad_norm": 0.6700214147567749,
+      "learning_rate": 5.7766489612082625e-06,
+      "loss": 0.7743,
+      "step": 16420
+    },
+    {
+      "epoch": 0.903792173482305,
+      "grad_norm": 0.6496555805206299,
+      "learning_rate": 5.776220752800411e-06,
+      "loss": 0.7058,
+      "step": 16421
+    },
+    {
+      "epoch": 0.9038472122846607,
+      "grad_norm": 0.7568738460540771,
+      "learning_rate": 5.775792538558692e-06,
+      "loss": 0.7937,
+      "step": 16422
+    },
+    {
+      "epoch": 0.9039022510870164,
+      "grad_norm": 0.8599572777748108,
+      "learning_rate": 5.775364318486325e-06,
+      "loss": 0.6478,
+      "step": 16423
+    },
+    {
+      "epoch": 0.903957289889372,
+      "grad_norm": 0.80698162317276,
+      "learning_rate": 5.774936092586526e-06,
+      "loss": 0.7847,
+      "step": 16424
+    },
+    {
+      "epoch": 0.9040123286917277,
+      "grad_norm": 0.9321784973144531,
+      "learning_rate": 5.774507860862516e-06,
+      "loss": 0.8223,
+      "step": 16425
+    },
+    {
+      "epoch": 0.9040673674940833,
+      "grad_norm": 0.6712853312492371,
+      "learning_rate": 5.774079623317513e-06,
+      "loss": 0.6958,
+      "step": 16426
+    },
+    {
+      "epoch": 0.904122406296439,
+      "grad_norm": 0.6142114400863647,
+      "learning_rate": 5.7736513799547326e-06,
+      "loss": 0.6696,
+      "step": 16427
+    },
+    {
+      "epoch": 0.9041774450987946,
+      "grad_norm": 0.7185747027397156,
+      "learning_rate": 5.773223130777396e-06,
+      "loss": 0.7161,
+      "step": 16428
+    },
+    {
+      "epoch": 0.9042324839011503,
+      "grad_norm": 0.6980574727058411,
+      "learning_rate": 5.772794875788721e-06,
+      "loss": 0.8255,
+      "step": 16429
+    },
+    {
+      "epoch": 0.904287522703506,
+      "grad_norm": 0.6682069301605225,
+      "learning_rate": 5.772366614991927e-06,
+      "loss": 0.7574,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9043425615058617,
+      "grad_norm": 0.6451525688171387,
+      "learning_rate": 5.771938348390233e-06,
+      "loss": 0.8147,
+      "step": 16431
+    },
+    {
+      "epoch": 0.9043976003082173,
+      "grad_norm": 0.7066215872764587,
+      "learning_rate": 5.771510075986857e-06,
+      "loss": 0.6447,
+      "step": 16432
+    },
+    {
+      "epoch": 0.9044526391105729,
+      "grad_norm": 0.591195285320282,
+      "learning_rate": 5.771081797785015e-06,
+      "loss": 0.6144,
+      "step": 16433
+    },
+    {
+      "epoch": 0.9045076779129286,
+      "grad_norm": 0.7004978656768799,
+      "learning_rate": 5.770653513787932e-06,
+      "loss": 0.6351,
+      "step": 16434
+    },
+    {
+      "epoch": 0.9045627167152843,
+      "grad_norm": 0.765402615070343,
+      "learning_rate": 5.770225223998822e-06,
+      "loss": 0.8448,
+      "step": 16435
+    },
+    {
+      "epoch": 0.9046177555176399,
+      "grad_norm": 0.7902443408966064,
+      "learning_rate": 5.769796928420907e-06,
+      "loss": 0.736,
+      "step": 16436
+    },
+    {
+      "epoch": 0.9046727943199956,
+      "grad_norm": 0.7513584494590759,
+      "learning_rate": 5.7693686270574026e-06,
+      "loss": 0.7939,
+      "step": 16437
+    },
+    {
+      "epoch": 0.9047278331223513,
+      "grad_norm": 0.7552775144577026,
+      "learning_rate": 5.7689403199115305e-06,
+      "loss": 0.7057,
+      "step": 16438
+    },
+    {
+      "epoch": 0.904782871924707,
+      "grad_norm": 0.6609054207801819,
+      "learning_rate": 5.768512006986508e-06,
+      "loss": 0.7442,
+      "step": 16439
+    },
+    {
+      "epoch": 0.9048379107270625,
+      "grad_norm": 0.7378336787223816,
+      "learning_rate": 5.768083688285555e-06,
+      "loss": 0.8066,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9048929495294182,
+      "grad_norm": 0.6223191022872925,
+      "learning_rate": 5.76765536381189e-06,
+      "loss": 0.6754,
+      "step": 16441
+    },
+    {
+      "epoch": 0.9049479883317739,
+      "grad_norm": 0.707188069820404,
+      "learning_rate": 5.767227033568734e-06,
+      "loss": 0.8307,
+      "step": 16442
+    },
+    {
+      "epoch": 0.9050030271341296,
+      "grad_norm": 0.680635929107666,
+      "learning_rate": 5.7667986975593046e-06,
+      "loss": 0.7659,
+      "step": 16443
+    },
+    {
+      "epoch": 0.9050580659364852,
+      "grad_norm": 0.7325000166893005,
+      "learning_rate": 5.766370355786822e-06,
+      "loss": 0.8449,
+      "step": 16444
+    },
+    {
+      "epoch": 0.9051131047388409,
+      "grad_norm": 0.6698771715164185,
+      "learning_rate": 5.765942008254504e-06,
+      "loss": 0.7689,
+      "step": 16445
+    },
+    {
+      "epoch": 0.9051681435411966,
+      "grad_norm": 0.6456333994865417,
+      "learning_rate": 5.765513654965571e-06,
+      "loss": 0.8054,
+      "step": 16446
+    },
+    {
+      "epoch": 0.9052231823435523,
+      "grad_norm": 0.6837481260299683,
+      "learning_rate": 5.765085295923243e-06,
+      "loss": 0.7327,
+      "step": 16447
+    },
+    {
+      "epoch": 0.9052782211459078,
+      "grad_norm": 0.6525166034698486,
+      "learning_rate": 5.7646569311307385e-06,
+      "loss": 0.6838,
+      "step": 16448
+    },
+    {
+      "epoch": 0.9053332599482635,
+      "grad_norm": 0.6613319516181946,
+      "learning_rate": 5.764228560591277e-06,
+      "loss": 0.765,
+      "step": 16449
+    },
+    {
+      "epoch": 0.9053882987506192,
+      "grad_norm": 0.732072651386261,
+      "learning_rate": 5.763800184308078e-06,
+      "loss": 0.7969,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9054433375529748,
+      "grad_norm": 0.6654138565063477,
+      "learning_rate": 5.763371802284362e-06,
+      "loss": 0.744,
+      "step": 16451
+    },
+    {
+      "epoch": 0.9054983763553305,
+      "grad_norm": 0.685528576374054,
+      "learning_rate": 5.762943414523347e-06,
+      "loss": 0.793,
+      "step": 16452
+    },
+    {
+      "epoch": 0.9055534151576862,
+      "grad_norm": 0.7957682013511658,
+      "learning_rate": 5.762515021028254e-06,
+      "loss": 0.7792,
+      "step": 16453
+    },
+    {
+      "epoch": 0.9056084539600419,
+      "grad_norm": 0.7029872536659241,
+      "learning_rate": 5.762086621802302e-06,
+      "loss": 0.7014,
+      "step": 16454
+    },
+    {
+      "epoch": 0.9056634927623974,
+      "grad_norm": 0.6274085640907288,
+      "learning_rate": 5.761658216848711e-06,
+      "loss": 0.7039,
+      "step": 16455
+    },
+    {
+      "epoch": 0.9057185315647531,
+      "grad_norm": 0.7196398377418518,
+      "learning_rate": 5.761229806170702e-06,
+      "loss": 0.7091,
+      "step": 16456
+    },
+    {
+      "epoch": 0.9057735703671088,
+      "grad_norm": 0.8225924372673035,
+      "learning_rate": 5.760801389771491e-06,
+      "loss": 0.8,
+      "step": 16457
+    },
+    {
+      "epoch": 0.9058286091694645,
+      "grad_norm": 0.659428596496582,
+      "learning_rate": 5.760372967654302e-06,
+      "loss": 0.7364,
+      "step": 16458
+    },
+    {
+      "epoch": 0.9058836479718201,
+      "grad_norm": 0.667123556137085,
+      "learning_rate": 5.759944539822353e-06,
+      "loss": 0.7754,
+      "step": 16459
+    },
+    {
+      "epoch": 0.9059386867741758,
+      "grad_norm": 0.6794472336769104,
+      "learning_rate": 5.759516106278864e-06,
+      "loss": 0.7747,
+      "step": 16460
+    },
+    {
+      "epoch": 0.9059937255765315,
+      "grad_norm": 0.652911365032196,
+      "learning_rate": 5.759087667027056e-06,
+      "loss": 0.7144,
+      "step": 16461
+    },
+    {
+      "epoch": 0.9060487643788872,
+      "grad_norm": 0.7240120768547058,
+      "learning_rate": 5.758659222070149e-06,
+      "loss": 0.7643,
+      "step": 16462
+    },
+    {
+      "epoch": 0.9061038031812427,
+      "grad_norm": 0.7705693244934082,
+      "learning_rate": 5.758230771411361e-06,
+      "loss": 0.6911,
+      "step": 16463
+    },
+    {
+      "epoch": 0.9061588419835984,
+      "grad_norm": 0.6969810128211975,
+      "learning_rate": 5.757802315053913e-06,
+      "loss": 0.7588,
+      "step": 16464
+    },
+    {
+      "epoch": 0.9062138807859541,
+      "grad_norm": 0.6787244081497192,
+      "learning_rate": 5.757373853001026e-06,
+      "loss": 0.7772,
+      "step": 16465
+    },
+    {
+      "epoch": 0.9062689195883098,
+      "grad_norm": 0.7181483507156372,
+      "learning_rate": 5.75694538525592e-06,
+      "loss": 0.7586,
+      "step": 16466
+    },
+    {
+      "epoch": 0.9063239583906654,
+      "grad_norm": 0.6730976700782776,
+      "learning_rate": 5.756516911821816e-06,
+      "loss": 0.758,
+      "step": 16467
+    },
+    {
+      "epoch": 0.9063789971930211,
+      "grad_norm": 0.752193808555603,
+      "learning_rate": 5.756088432701933e-06,
+      "loss": 0.7701,
+      "step": 16468
+    },
+    {
+      "epoch": 0.9064340359953768,
+      "grad_norm": 0.7424039244651794,
+      "learning_rate": 5.75565994789949e-06,
+      "loss": 0.777,
+      "step": 16469
+    },
+    {
+      "epoch": 0.9064890747977324,
+      "grad_norm": 0.7168293595314026,
+      "learning_rate": 5.755231457417711e-06,
+      "loss": 0.752,
+      "step": 16470
+    },
+    {
+      "epoch": 0.906544113600088,
+      "grad_norm": 0.566977858543396,
+      "learning_rate": 5.754802961259814e-06,
+      "loss": 0.6182,
+      "step": 16471
+    },
+    {
+      "epoch": 0.9065991524024437,
+      "grad_norm": 0.6081774234771729,
+      "learning_rate": 5.754374459429021e-06,
+      "loss": 0.7449,
+      "step": 16472
+    },
+    {
+      "epoch": 0.9066541912047994,
+      "grad_norm": 0.8636649250984192,
+      "learning_rate": 5.75394595192855e-06,
+      "loss": 0.7754,
+      "step": 16473
+    },
+    {
+      "epoch": 0.9067092300071551,
+      "grad_norm": 0.705915629863739,
+      "learning_rate": 5.753517438761624e-06,
+      "loss": 0.7838,
+      "step": 16474
+    },
+    {
+      "epoch": 0.9067642688095107,
+      "grad_norm": 0.6954970955848694,
+      "learning_rate": 5.753088919931462e-06,
+      "loss": 0.6729,
+      "step": 16475
+    },
+    {
+      "epoch": 0.9068193076118664,
+      "grad_norm": 0.6736747622489929,
+      "learning_rate": 5.752660395441286e-06,
+      "loss": 0.7344,
+      "step": 16476
+    },
+    {
+      "epoch": 0.906874346414222,
+      "grad_norm": 0.7642985582351685,
+      "learning_rate": 5.752231865294314e-06,
+      "loss": 0.7716,
+      "step": 16477
+    },
+    {
+      "epoch": 0.9069293852165777,
+      "grad_norm": 0.6003983616828918,
+      "learning_rate": 5.7518033294937705e-06,
+      "loss": 0.7094,
+      "step": 16478
+    },
+    {
+      "epoch": 0.9069844240189333,
+      "grad_norm": 0.6081627607345581,
+      "learning_rate": 5.751374788042875e-06,
+      "loss": 0.6534,
+      "step": 16479
+    },
+    {
+      "epoch": 0.907039462821289,
+      "grad_norm": 0.7221750617027283,
+      "learning_rate": 5.750946240944847e-06,
+      "loss": 0.7032,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9070945016236447,
+      "grad_norm": 0.6791507601737976,
+      "learning_rate": 5.750517688202907e-06,
+      "loss": 0.7944,
+      "step": 16481
+    },
+    {
+      "epoch": 0.9071495404260004,
+      "grad_norm": 0.6808193922042847,
+      "learning_rate": 5.750089129820278e-06,
+      "loss": 0.7536,
+      "step": 16482
+    },
+    {
+      "epoch": 0.907204579228356,
+      "grad_norm": 0.6561774611473083,
+      "learning_rate": 5.7496605658001795e-06,
+      "loss": 0.755,
+      "step": 16483
+    },
+    {
+      "epoch": 0.9072596180307116,
+      "grad_norm": 0.6241487264633179,
+      "learning_rate": 5.749231996145833e-06,
+      "loss": 0.7474,
+      "step": 16484
+    },
+    {
+      "epoch": 0.9073146568330673,
+      "grad_norm": 0.676726758480072,
+      "learning_rate": 5.748803420860459e-06,
+      "loss": 0.7416,
+      "step": 16485
+    },
+    {
+      "epoch": 0.907369695635423,
+      "grad_norm": 0.6571675539016724,
+      "learning_rate": 5.748374839947279e-06,
+      "loss": 0.7428,
+      "step": 16486
+    },
+    {
+      "epoch": 0.9074247344377786,
+      "grad_norm": 0.6442016959190369,
+      "learning_rate": 5.7479462534095155e-06,
+      "loss": 0.7265,
+      "step": 16487
+    },
+    {
+      "epoch": 0.9074797732401343,
+      "grad_norm": 0.6728110909461975,
+      "learning_rate": 5.747517661250387e-06,
+      "loss": 0.6492,
+      "step": 16488
+    },
+    {
+      "epoch": 0.90753481204249,
+      "grad_norm": 1.114125370979309,
+      "learning_rate": 5.747089063473115e-06,
+      "loss": 0.7399,
+      "step": 16489
+    },
+    {
+      "epoch": 0.9075898508448457,
+      "grad_norm": 0.6066282391548157,
+      "learning_rate": 5.746660460080923e-06,
+      "loss": 0.5736,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9076448896472012,
+      "grad_norm": 0.6982638835906982,
+      "learning_rate": 5.74623185107703e-06,
+      "loss": 0.765,
+      "step": 16491
+    },
+    {
+      "epoch": 0.9076999284495569,
+      "grad_norm": 0.6360225677490234,
+      "learning_rate": 5.745803236464659e-06,
+      "loss": 0.7082,
+      "step": 16492
+    },
+    {
+      "epoch": 0.9077549672519126,
+      "grad_norm": 0.6267019510269165,
+      "learning_rate": 5.745374616247029e-06,
+      "loss": 0.7498,
+      "step": 16493
+    },
+    {
+      "epoch": 0.9078100060542682,
+      "grad_norm": 0.6618008613586426,
+      "learning_rate": 5.744945990427364e-06,
+      "loss": 0.7073,
+      "step": 16494
+    },
+    {
+      "epoch": 0.9078650448566239,
+      "grad_norm": 0.7970854043960571,
+      "learning_rate": 5.744517359008884e-06,
+      "loss": 0.6968,
+      "step": 16495
+    },
+    {
+      "epoch": 0.9079200836589796,
+      "grad_norm": 0.6457371711730957,
+      "learning_rate": 5.744088721994811e-06,
+      "loss": 0.7102,
+      "step": 16496
+    },
+    {
+      "epoch": 0.9079751224613353,
+      "grad_norm": 0.6841073036193848,
+      "learning_rate": 5.743660079388367e-06,
+      "loss": 0.7448,
+      "step": 16497
+    },
+    {
+      "epoch": 0.9080301612636908,
+      "grad_norm": 0.6577959656715393,
+      "learning_rate": 5.7432314311927725e-06,
+      "loss": 0.6853,
+      "step": 16498
+    },
+    {
+      "epoch": 0.9080852000660465,
+      "grad_norm": 0.9146704077720642,
+      "learning_rate": 5.7428027774112494e-06,
+      "loss": 0.7052,
+      "step": 16499
+    },
+    {
+      "epoch": 0.9081402388684022,
+      "grad_norm": 0.7161448001861572,
+      "learning_rate": 5.7423741180470194e-06,
+      "loss": 0.775,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9081952776707579,
+      "grad_norm": 0.6443139314651489,
+      "learning_rate": 5.741945453103304e-06,
+      "loss": 0.7169,
+      "step": 16501
+    },
+    {
+      "epoch": 0.9082503164731135,
+      "grad_norm": 0.6783748269081116,
+      "learning_rate": 5.7415167825833254e-06,
+      "loss": 0.7082,
+      "step": 16502
+    },
+    {
+      "epoch": 0.9083053552754692,
+      "grad_norm": 0.7961641550064087,
+      "learning_rate": 5.741088106490305e-06,
+      "loss": 0.769,
+      "step": 16503
+    },
+    {
+      "epoch": 0.9083603940778249,
+      "grad_norm": 0.6952986121177673,
+      "learning_rate": 5.740659424827465e-06,
+      "loss": 0.7978,
+      "step": 16504
+    },
+    {
+      "epoch": 0.9084154328801806,
+      "grad_norm": 0.6124480366706848,
+      "learning_rate": 5.740230737598026e-06,
+      "loss": 0.6824,
+      "step": 16505
+    },
+    {
+      "epoch": 0.9084704716825361,
+      "grad_norm": 0.6469553709030151,
+      "learning_rate": 5.739802044805211e-06,
+      "loss": 0.6558,
+      "step": 16506
+    },
+    {
+      "epoch": 0.9085255104848918,
+      "grad_norm": 0.6296014189720154,
+      "learning_rate": 5.739373346452244e-06,
+      "loss": 0.7079,
+      "step": 16507
+    },
+    {
+      "epoch": 0.9085805492872475,
+      "grad_norm": 0.6862651705741882,
+      "learning_rate": 5.7389446425423435e-06,
+      "loss": 0.7787,
+      "step": 16508
+    },
+    {
+      "epoch": 0.9086355880896032,
+      "grad_norm": 0.7167361974716187,
+      "learning_rate": 5.738515933078733e-06,
+      "loss": 0.8556,
+      "step": 16509
+    },
+    {
+      "epoch": 0.9086906268919588,
+      "grad_norm": 0.6557625532150269,
+      "learning_rate": 5.7380872180646356e-06,
+      "loss": 0.6326,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9087456656943145,
+      "grad_norm": 0.6629947423934937,
+      "learning_rate": 5.737658497503271e-06,
+      "loss": 0.6692,
+      "step": 16511
+    },
+    {
+      "epoch": 0.9088007044966702,
+      "grad_norm": 0.723196804523468,
+      "learning_rate": 5.737229771397862e-06,
+      "loss": 0.837,
+      "step": 16512
+    },
+    {
+      "epoch": 0.9088557432990259,
+      "grad_norm": 0.6491259336471558,
+      "learning_rate": 5.7368010397516315e-06,
+      "loss": 0.7818,
+      "step": 16513
+    },
+    {
+      "epoch": 0.9089107821013814,
+      "grad_norm": 0.6776933073997498,
+      "learning_rate": 5.736372302567802e-06,
+      "loss": 0.668,
+      "step": 16514
+    },
+    {
+      "epoch": 0.9089658209037371,
+      "grad_norm": 0.5955449938774109,
+      "learning_rate": 5.735943559849596e-06,
+      "loss": 0.6875,
+      "step": 16515
+    },
+    {
+      "epoch": 0.9090208597060928,
+      "grad_norm": 0.6763395667076111,
+      "learning_rate": 5.735514811600235e-06,
+      "loss": 0.6882,
+      "step": 16516
+    },
+    {
+      "epoch": 0.9090758985084485,
+      "grad_norm": 0.794796884059906,
+      "learning_rate": 5.735086057822941e-06,
+      "loss": 0.8279,
+      "step": 16517
+    },
+    {
+      "epoch": 0.9091309373108041,
+      "grad_norm": 0.6420493721961975,
+      "learning_rate": 5.734657298520936e-06,
+      "loss": 0.6755,
+      "step": 16518
+    },
+    {
+      "epoch": 0.9091859761131598,
+      "grad_norm": 0.7267552614212036,
+      "learning_rate": 5.734228533697445e-06,
+      "loss": 0.7816,
+      "step": 16519
+    },
+    {
+      "epoch": 0.9092410149155155,
+      "grad_norm": 0.7506794929504395,
+      "learning_rate": 5.733799763355689e-06,
+      "loss": 0.6968,
+      "step": 16520
+    },
+    {
+      "epoch": 0.9092960537178711,
+      "grad_norm": 0.6953657269477844,
+      "learning_rate": 5.73337098749889e-06,
+      "loss": 0.7866,
+      "step": 16521
+    },
+    {
+      "epoch": 0.9093510925202267,
+      "grad_norm": 0.8068533539772034,
+      "learning_rate": 5.732942206130271e-06,
+      "loss": 0.7726,
+      "step": 16522
+    },
+    {
+      "epoch": 0.9094061313225824,
+      "grad_norm": 0.8853289484977722,
+      "learning_rate": 5.732513419253054e-06,
+      "loss": 0.7886,
+      "step": 16523
+    },
+    {
+      "epoch": 0.9094611701249381,
+      "grad_norm": 0.6110638380050659,
+      "learning_rate": 5.732084626870463e-06,
+      "loss": 0.6793,
+      "step": 16524
+    },
+    {
+      "epoch": 0.9095162089272938,
+      "grad_norm": 0.6836052536964417,
+      "learning_rate": 5.731655828985719e-06,
+      "loss": 0.7245,
+      "step": 16525
+    },
+    {
+      "epoch": 0.9095712477296494,
+      "grad_norm": 0.6699627041816711,
+      "learning_rate": 5.731227025602046e-06,
+      "loss": 0.7642,
+      "step": 16526
+    },
+    {
+      "epoch": 0.9096262865320051,
+      "grad_norm": 0.6030815243721008,
+      "learning_rate": 5.730798216722667e-06,
+      "loss": 0.7817,
+      "step": 16527
+    },
+    {
+      "epoch": 0.9096813253343607,
+      "grad_norm": 0.9264968633651733,
+      "learning_rate": 5.730369402350804e-06,
+      "loss": 0.8899,
+      "step": 16528
+    },
+    {
+      "epoch": 0.9097363641367164,
+      "grad_norm": 0.6469599604606628,
+      "learning_rate": 5.729940582489679e-06,
+      "loss": 0.7672,
+      "step": 16529
+    },
+    {
+      "epoch": 0.909791402939072,
+      "grad_norm": 0.6422780752182007,
+      "learning_rate": 5.729511757142518e-06,
+      "loss": 0.7308,
+      "step": 16530
+    },
+    {
+      "epoch": 0.9098464417414277,
+      "grad_norm": 0.7130196690559387,
+      "learning_rate": 5.72908292631254e-06,
+      "loss": 0.7283,
+      "step": 16531
+    },
+    {
+      "epoch": 0.9099014805437834,
+      "grad_norm": 0.6389671564102173,
+      "learning_rate": 5.7286540900029705e-06,
+      "loss": 0.6865,
+      "step": 16532
+    },
+    {
+      "epoch": 0.9099565193461391,
+      "grad_norm": 0.6116911768913269,
+      "learning_rate": 5.728225248217033e-06,
+      "loss": 0.7162,
+      "step": 16533
+    },
+    {
+      "epoch": 0.9100115581484947,
+      "grad_norm": 0.6901746392250061,
+      "learning_rate": 5.727796400957949e-06,
+      "loss": 0.7936,
+      "step": 16534
+    },
+    {
+      "epoch": 0.9100665969508503,
+      "grad_norm": 0.6178259253501892,
+      "learning_rate": 5.727367548228941e-06,
+      "loss": 0.7345,
+      "step": 16535
+    },
+    {
+      "epoch": 0.910121635753206,
+      "grad_norm": 0.6640061140060425,
+      "learning_rate": 5.726938690033235e-06,
+      "loss": 0.6847,
+      "step": 16536
+    },
+    {
+      "epoch": 0.9101766745555616,
+      "grad_norm": 0.5669469833374023,
+      "learning_rate": 5.726509826374049e-06,
+      "loss": 0.6054,
+      "step": 16537
+    },
+    {
+      "epoch": 0.9102317133579173,
+      "grad_norm": 0.6553897857666016,
+      "learning_rate": 5.7260809572546126e-06,
+      "loss": 0.7299,
+      "step": 16538
+    },
+    {
+      "epoch": 0.910286752160273,
+      "grad_norm": 0.6646978259086609,
+      "learning_rate": 5.725652082678146e-06,
+      "loss": 0.7952,
+      "step": 16539
+    },
+    {
+      "epoch": 0.9103417909626287,
+      "grad_norm": 0.7033318281173706,
+      "learning_rate": 5.725223202647872e-06,
+      "loss": 0.7477,
+      "step": 16540
+    },
+    {
+      "epoch": 0.9103968297649843,
+      "grad_norm": 0.804336667060852,
+      "learning_rate": 5.724794317167013e-06,
+      "loss": 0.7088,
+      "step": 16541
+    },
+    {
+      "epoch": 0.91045186856734,
+      "grad_norm": 0.6683815717697144,
+      "learning_rate": 5.724365426238796e-06,
+      "loss": 0.7497,
+      "step": 16542
+    },
+    {
+      "epoch": 0.9105069073696956,
+      "grad_norm": 0.7227790951728821,
+      "learning_rate": 5.723936529866442e-06,
+      "loss": 0.7673,
+      "step": 16543
+    },
+    {
+      "epoch": 0.9105619461720513,
+      "grad_norm": 0.7867640852928162,
+      "learning_rate": 5.723507628053175e-06,
+      "loss": 0.8351,
+      "step": 16544
+    },
+    {
+      "epoch": 0.9106169849744069,
+      "grad_norm": 0.6680448055267334,
+      "learning_rate": 5.7230787208022175e-06,
+      "loss": 0.7068,
+      "step": 16545
+    },
+    {
+      "epoch": 0.9106720237767626,
+      "grad_norm": 0.7167722582817078,
+      "learning_rate": 5.722649808116794e-06,
+      "loss": 0.771,
+      "step": 16546
+    },
+    {
+      "epoch": 0.9107270625791183,
+      "grad_norm": 0.6377316117286682,
+      "learning_rate": 5.72222089000013e-06,
+      "loss": 0.7535,
+      "step": 16547
+    },
+    {
+      "epoch": 0.910782101381474,
+      "grad_norm": 0.6644514799118042,
+      "learning_rate": 5.721791966455444e-06,
+      "loss": 0.6642,
+      "step": 16548
+    },
+    {
+      "epoch": 0.9108371401838296,
+      "grad_norm": 0.6652604341506958,
+      "learning_rate": 5.7213630374859644e-06,
+      "loss": 0.7192,
+      "step": 16549
+    },
+    {
+      "epoch": 0.9108921789861852,
+      "grad_norm": 0.7685999274253845,
+      "learning_rate": 5.720934103094913e-06,
+      "loss": 0.7497,
+      "step": 16550
+    },
+    {
+      "epoch": 0.9109472177885409,
+      "grad_norm": 0.6837451457977295,
+      "learning_rate": 5.7205051632855135e-06,
+      "loss": 0.8337,
+      "step": 16551
+    },
+    {
+      "epoch": 0.9110022565908966,
+      "grad_norm": 0.6328313946723938,
+      "learning_rate": 5.720076218060991e-06,
+      "loss": 0.7106,
+      "step": 16552
+    },
+    {
+      "epoch": 0.9110572953932522,
+      "grad_norm": 0.6779924631118774,
+      "learning_rate": 5.7196472674245675e-06,
+      "loss": 0.6518,
+      "step": 16553
+    },
+    {
+      "epoch": 0.9111123341956079,
+      "grad_norm": 0.7673786282539368,
+      "learning_rate": 5.719218311379468e-06,
+      "loss": 0.687,
+      "step": 16554
+    },
+    {
+      "epoch": 0.9111673729979636,
+      "grad_norm": 0.9567201137542725,
+      "learning_rate": 5.718789349928917e-06,
+      "loss": 0.8435,
+      "step": 16555
+    },
+    {
+      "epoch": 0.9112224118003193,
+      "grad_norm": 0.6433333158493042,
+      "learning_rate": 5.7183603830761365e-06,
+      "loss": 0.6911,
+      "step": 16556
+    },
+    {
+      "epoch": 0.9112774506026748,
+      "grad_norm": 0.7093275189399719,
+      "learning_rate": 5.717931410824352e-06,
+      "loss": 0.7928,
+      "step": 16557
+    },
+    {
+      "epoch": 0.9113324894050305,
+      "grad_norm": 0.6292399764060974,
+      "learning_rate": 5.717502433176788e-06,
+      "loss": 0.7971,
+      "step": 16558
+    },
+    {
+      "epoch": 0.9113875282073862,
+      "grad_norm": 0.7010529637336731,
+      "learning_rate": 5.717073450136668e-06,
+      "loss": 0.7363,
+      "step": 16559
+    },
+    {
+      "epoch": 0.9114425670097419,
+      "grad_norm": 0.6663999557495117,
+      "learning_rate": 5.716644461707215e-06,
+      "loss": 0.6988,
+      "step": 16560
+    },
+    {
+      "epoch": 0.9114976058120975,
+      "grad_norm": 0.6320468187332153,
+      "learning_rate": 5.716215467891654e-06,
+      "loss": 0.6654,
+      "step": 16561
+    },
+    {
+      "epoch": 0.9115526446144532,
+      "grad_norm": 0.6906135678291321,
+      "learning_rate": 5.715786468693208e-06,
+      "loss": 0.7749,
+      "step": 16562
+    },
+    {
+      "epoch": 0.9116076834168089,
+      "grad_norm": 0.7253930568695068,
+      "learning_rate": 5.715357464115104e-06,
+      "loss": 0.7325,
+      "step": 16563
+    },
+    {
+      "epoch": 0.9116627222191646,
+      "grad_norm": 0.6567997336387634,
+      "learning_rate": 5.7149284541605655e-06,
+      "loss": 0.7028,
+      "step": 16564
+    },
+    {
+      "epoch": 0.9117177610215201,
+      "grad_norm": 0.7034043669700623,
+      "learning_rate": 5.7144994388328155e-06,
+      "loss": 0.7328,
+      "step": 16565
+    },
+    {
+      "epoch": 0.9117727998238758,
+      "grad_norm": 0.6639497876167297,
+      "learning_rate": 5.714070418135078e-06,
+      "loss": 0.7289,
+      "step": 16566
+    },
+    {
+      "epoch": 0.9118278386262315,
+      "grad_norm": 0.6601856350898743,
+      "learning_rate": 5.713641392070579e-06,
+      "loss": 0.6857,
+      "step": 16567
+    },
+    {
+      "epoch": 0.9118828774285872,
+      "grad_norm": 1.0736916065216064,
+      "learning_rate": 5.713212360642543e-06,
+      "loss": 0.6254,
+      "step": 16568
+    },
+    {
+      "epoch": 0.9119379162309428,
+      "grad_norm": 0.6571956872940063,
+      "learning_rate": 5.712783323854193e-06,
+      "loss": 0.7931,
+      "step": 16569
+    },
+    {
+      "epoch": 0.9119929550332985,
+      "grad_norm": 0.7038640975952148,
+      "learning_rate": 5.712354281708754e-06,
+      "loss": 0.7065,
+      "step": 16570
+    },
+    {
+      "epoch": 0.9120479938356542,
+      "grad_norm": 0.6827693581581116,
+      "learning_rate": 5.711925234209452e-06,
+      "loss": 0.7971,
+      "step": 16571
+    },
+    {
+      "epoch": 0.9121030326380098,
+      "grad_norm": 0.6412268877029419,
+      "learning_rate": 5.71149618135951e-06,
+      "loss": 0.7899,
+      "step": 16572
+    },
+    {
+      "epoch": 0.9121580714403654,
+      "grad_norm": 0.753730833530426,
+      "learning_rate": 5.711067123162154e-06,
+      "loss": 0.6762,
+      "step": 16573
+    },
+    {
+      "epoch": 0.9122131102427211,
+      "grad_norm": 0.62330162525177,
+      "learning_rate": 5.710638059620606e-06,
+      "loss": 0.6661,
+      "step": 16574
+    },
+    {
+      "epoch": 0.9122681490450768,
+      "grad_norm": 0.7718592882156372,
+      "learning_rate": 5.710208990738093e-06,
+      "loss": 0.8146,
+      "step": 16575
+    },
+    {
+      "epoch": 0.9123231878474325,
+      "grad_norm": 0.7234812378883362,
+      "learning_rate": 5.709779916517842e-06,
+      "loss": 0.7633,
+      "step": 16576
+    },
+    {
+      "epoch": 0.9123782266497881,
+      "grad_norm": 0.7249013781547546,
+      "learning_rate": 5.709350836963072e-06,
+      "loss": 0.7339,
+      "step": 16577
+    },
+    {
+      "epoch": 0.9124332654521438,
+      "grad_norm": 0.643532395362854,
+      "learning_rate": 5.708921752077011e-06,
+      "loss": 0.6899,
+      "step": 16578
+    },
+    {
+      "epoch": 0.9124883042544994,
+      "grad_norm": 0.7240098118782043,
+      "learning_rate": 5.708492661862885e-06,
+      "loss": 0.6788,
+      "step": 16579
+    },
+    {
+      "epoch": 0.912543343056855,
+      "grad_norm": 0.6944229602813721,
+      "learning_rate": 5.708063566323918e-06,
+      "loss": 0.8284,
+      "step": 16580
+    },
+    {
+      "epoch": 0.9125983818592107,
+      "grad_norm": 0.7374451160430908,
+      "learning_rate": 5.707634465463334e-06,
+      "loss": 0.7414,
+      "step": 16581
+    },
+    {
+      "epoch": 0.9126534206615664,
+      "grad_norm": 0.6843305230140686,
+      "learning_rate": 5.70720535928436e-06,
+      "loss": 0.7765,
+      "step": 16582
+    },
+    {
+      "epoch": 0.9127084594639221,
+      "grad_norm": 0.6846521496772766,
+      "learning_rate": 5.70677624779022e-06,
+      "loss": 0.7526,
+      "step": 16583
+    },
+    {
+      "epoch": 0.9127634982662777,
+      "grad_norm": 0.782206654548645,
+      "learning_rate": 5.7063471309841375e-06,
+      "loss": 0.8384,
+      "step": 16584
+    },
+    {
+      "epoch": 0.9128185370686334,
+      "grad_norm": 0.590929388999939,
+      "learning_rate": 5.70591800886934e-06,
+      "loss": 0.5523,
+      "step": 16585
+    },
+    {
+      "epoch": 0.912873575870989,
+      "grad_norm": 0.7062276005744934,
+      "learning_rate": 5.705488881449051e-06,
+      "loss": 0.78,
+      "step": 16586
+    },
+    {
+      "epoch": 0.9129286146733447,
+      "grad_norm": 0.6782402992248535,
+      "learning_rate": 5.705059748726496e-06,
+      "loss": 0.7376,
+      "step": 16587
+    },
+    {
+      "epoch": 0.9129836534757003,
+      "grad_norm": 0.6320528984069824,
+      "learning_rate": 5.704630610704902e-06,
+      "loss": 0.6088,
+      "step": 16588
+    },
+    {
+      "epoch": 0.913038692278056,
+      "grad_norm": 0.62249356508255,
+      "learning_rate": 5.704201467387492e-06,
+      "loss": 0.648,
+      "step": 16589
+    },
+    {
+      "epoch": 0.9130937310804117,
+      "grad_norm": 0.6249415278434753,
+      "learning_rate": 5.703772318777492e-06,
+      "loss": 0.6843,
+      "step": 16590
+    },
+    {
+      "epoch": 0.9131487698827674,
+      "grad_norm": 0.7930389046669006,
+      "learning_rate": 5.703343164878128e-06,
+      "loss": 0.8037,
+      "step": 16591
+    },
+    {
+      "epoch": 0.913203808685123,
+      "grad_norm": 0.6056682467460632,
+      "learning_rate": 5.7029140056926256e-06,
+      "loss": 0.726,
+      "step": 16592
+    },
+    {
+      "epoch": 0.9132588474874787,
+      "grad_norm": 0.7311093807220459,
+      "learning_rate": 5.702484841224209e-06,
+      "loss": 0.757,
+      "step": 16593
+    },
+    {
+      "epoch": 0.9133138862898343,
+      "grad_norm": 0.6924651861190796,
+      "learning_rate": 5.702055671476105e-06,
+      "loss": 0.6812,
+      "step": 16594
+    },
+    {
+      "epoch": 0.91336892509219,
+      "grad_norm": 0.6961154937744141,
+      "learning_rate": 5.701626496451539e-06,
+      "loss": 0.7758,
+      "step": 16595
+    },
+    {
+      "epoch": 0.9134239638945456,
+      "grad_norm": 0.7138234972953796,
+      "learning_rate": 5.7011973161537345e-06,
+      "loss": 0.7691,
+      "step": 16596
+    },
+    {
+      "epoch": 0.9134790026969013,
+      "grad_norm": 0.6445374488830566,
+      "learning_rate": 5.70076813058592e-06,
+      "loss": 0.6223,
+      "step": 16597
+    },
+    {
+      "epoch": 0.913534041499257,
+      "grad_norm": 0.6478644013404846,
+      "learning_rate": 5.700338939751318e-06,
+      "loss": 0.6874,
+      "step": 16598
+    },
+    {
+      "epoch": 0.9135890803016127,
+      "grad_norm": 0.7351845502853394,
+      "learning_rate": 5.699909743653158e-06,
+      "loss": 0.751,
+      "step": 16599
+    },
+    {
+      "epoch": 0.9136441191039683,
+      "grad_norm": 0.675804078578949,
+      "learning_rate": 5.6994805422946634e-06,
+      "loss": 0.7637,
+      "step": 16600
+    },
+    {
+      "epoch": 0.9136991579063239,
+      "grad_norm": 0.7249841690063477,
+      "learning_rate": 5.69905133567906e-06,
+      "loss": 0.8253,
+      "step": 16601
+    },
+    {
+      "epoch": 0.9137541967086796,
+      "grad_norm": 0.7853312492370605,
+      "learning_rate": 5.698622123809573e-06,
+      "loss": 0.7469,
+      "step": 16602
+    },
+    {
+      "epoch": 0.9138092355110353,
+      "grad_norm": 0.719670295715332,
+      "learning_rate": 5.698192906689429e-06,
+      "loss": 0.7138,
+      "step": 16603
+    },
+    {
+      "epoch": 0.9138642743133909,
+      "grad_norm": 0.6555393934249878,
+      "learning_rate": 5.6977636843218555e-06,
+      "loss": 0.7134,
+      "step": 16604
+    },
+    {
+      "epoch": 0.9139193131157466,
+      "grad_norm": 0.60044926404953,
+      "learning_rate": 5.697334456710076e-06,
+      "loss": 0.6102,
+      "step": 16605
+    },
+    {
+      "epoch": 0.9139743519181023,
+      "grad_norm": 0.6154021620750427,
+      "learning_rate": 5.696905223857319e-06,
+      "loss": 0.7117,
+      "step": 16606
+    },
+    {
+      "epoch": 0.914029390720458,
+      "grad_norm": 0.7099533081054688,
+      "learning_rate": 5.696475985766807e-06,
+      "loss": 0.7042,
+      "step": 16607
+    },
+    {
+      "epoch": 0.9140844295228135,
+      "grad_norm": 0.6776680946350098,
+      "learning_rate": 5.696046742441769e-06,
+      "loss": 0.785,
+      "step": 16608
+    },
+    {
+      "epoch": 0.9141394683251692,
+      "grad_norm": 0.6876663565635681,
+      "learning_rate": 5.695617493885429e-06,
+      "loss": 0.7235,
+      "step": 16609
+    },
+    {
+      "epoch": 0.9141945071275249,
+      "grad_norm": 0.6305785775184631,
+      "learning_rate": 5.6951882401010136e-06,
+      "loss": 0.7578,
+      "step": 16610
+    },
+    {
+      "epoch": 0.9142495459298806,
+      "grad_norm": 0.5949695110321045,
+      "learning_rate": 5.694758981091752e-06,
+      "loss": 0.6303,
+      "step": 16611
+    },
+    {
+      "epoch": 0.9143045847322362,
+      "grad_norm": 0.6790008544921875,
+      "learning_rate": 5.6943297168608655e-06,
+      "loss": 0.7193,
+      "step": 16612
+    },
+    {
+      "epoch": 0.9143596235345919,
+      "grad_norm": 0.57261061668396,
+      "learning_rate": 5.693900447411585e-06,
+      "loss": 0.6036,
+      "step": 16613
+    },
+    {
+      "epoch": 0.9144146623369476,
+      "grad_norm": 0.7408223152160645,
+      "learning_rate": 5.693471172747131e-06,
+      "loss": 0.7957,
+      "step": 16614
+    },
+    {
+      "epoch": 0.9144697011393033,
+      "grad_norm": 0.6669179201126099,
+      "learning_rate": 5.693041892870736e-06,
+      "loss": 0.7269,
+      "step": 16615
+    },
+    {
+      "epoch": 0.9145247399416588,
+      "grad_norm": 0.9949795603752136,
+      "learning_rate": 5.692612607785622e-06,
+      "loss": 0.8756,
+      "step": 16616
+    },
+    {
+      "epoch": 0.9145797787440145,
+      "grad_norm": 0.7075115442276001,
+      "learning_rate": 5.6921833174950194e-06,
+      "loss": 0.7451,
+      "step": 16617
+    },
+    {
+      "epoch": 0.9146348175463702,
+      "grad_norm": 0.7605398297309875,
+      "learning_rate": 5.69175402200215e-06,
+      "loss": 0.8082,
+      "step": 16618
+    },
+    {
+      "epoch": 0.9146898563487259,
+      "grad_norm": 0.8290950655937195,
+      "learning_rate": 5.691324721310244e-06,
+      "loss": 0.8054,
+      "step": 16619
+    },
+    {
+      "epoch": 0.9147448951510815,
+      "grad_norm": 0.644469678401947,
+      "learning_rate": 5.690895415422526e-06,
+      "loss": 0.7695,
+      "step": 16620
+    },
+    {
+      "epoch": 0.9147999339534372,
+      "grad_norm": 0.6535455584526062,
+      "learning_rate": 5.690466104342224e-06,
+      "loss": 0.7285,
+      "step": 16621
+    },
+    {
+      "epoch": 0.9148549727557929,
+      "grad_norm": 0.6709266304969788,
+      "learning_rate": 5.690036788072562e-06,
+      "loss": 0.7393,
+      "step": 16622
+    },
+    {
+      "epoch": 0.9149100115581484,
+      "grad_norm": 0.6400435566902161,
+      "learning_rate": 5.689607466616767e-06,
+      "loss": 0.6981,
+      "step": 16623
+    },
+    {
+      "epoch": 0.9149650503605041,
+      "grad_norm": 0.6391820311546326,
+      "learning_rate": 5.68917813997807e-06,
+      "loss": 0.8183,
+      "step": 16624
+    },
+    {
+      "epoch": 0.9150200891628598,
+      "grad_norm": 0.6608666181564331,
+      "learning_rate": 5.688748808159693e-06,
+      "loss": 0.768,
+      "step": 16625
+    },
+    {
+      "epoch": 0.9150751279652155,
+      "grad_norm": 0.6189754605293274,
+      "learning_rate": 5.688319471164863e-06,
+      "loss": 0.6962,
+      "step": 16626
+    },
+    {
+      "epoch": 0.9151301667675711,
+      "grad_norm": 0.6954981684684753,
+      "learning_rate": 5.68789012899681e-06,
+      "loss": 0.7455,
+      "step": 16627
+    },
+    {
+      "epoch": 0.9151852055699268,
+      "grad_norm": 0.7653296589851379,
+      "learning_rate": 5.687460781658759e-06,
+      "loss": 0.7123,
+      "step": 16628
+    },
+    {
+      "epoch": 0.9152402443722825,
+      "grad_norm": 0.6047289371490479,
+      "learning_rate": 5.687031429153937e-06,
+      "loss": 0.6735,
+      "step": 16629
+    },
+    {
+      "epoch": 0.9152952831746382,
+      "grad_norm": 0.6627909541130066,
+      "learning_rate": 5.68660207148557e-06,
+      "loss": 0.6597,
+      "step": 16630
+    },
+    {
+      "epoch": 0.9153503219769937,
+      "grad_norm": 0.8544076085090637,
+      "learning_rate": 5.6861727086568855e-06,
+      "loss": 0.8044,
+      "step": 16631
+    },
+    {
+      "epoch": 0.9154053607793494,
+      "grad_norm": 0.7739614248275757,
+      "learning_rate": 5.685743340671111e-06,
+      "loss": 0.7395,
+      "step": 16632
+    },
+    {
+      "epoch": 0.9154603995817051,
+      "grad_norm": 0.7752935290336609,
+      "learning_rate": 5.6853139675314725e-06,
+      "loss": 0.8354,
+      "step": 16633
+    },
+    {
+      "epoch": 0.9155154383840608,
+      "grad_norm": 0.6166974306106567,
+      "learning_rate": 5.684884589241197e-06,
+      "loss": 0.7338,
+      "step": 16634
+    },
+    {
+      "epoch": 0.9155704771864164,
+      "grad_norm": 0.606935977935791,
+      "learning_rate": 5.684455205803513e-06,
+      "loss": 0.7381,
+      "step": 16635
+    },
+    {
+      "epoch": 0.9156255159887721,
+      "grad_norm": 0.9468479752540588,
+      "learning_rate": 5.684025817221647e-06,
+      "loss": 0.7959,
+      "step": 16636
+    },
+    {
+      "epoch": 0.9156805547911278,
+      "grad_norm": 0.6966477036476135,
+      "learning_rate": 5.683596423498827e-06,
+      "loss": 0.7251,
+      "step": 16637
+    },
+    {
+      "epoch": 0.9157355935934834,
+      "grad_norm": 0.630775511264801,
+      "learning_rate": 5.683167024638277e-06,
+      "loss": 0.7181,
+      "step": 16638
+    },
+    {
+      "epoch": 0.915790632395839,
+      "grad_norm": 0.6859539151191711,
+      "learning_rate": 5.682737620643229e-06,
+      "loss": 0.8418,
+      "step": 16639
+    },
+    {
+      "epoch": 0.9158456711981947,
+      "grad_norm": 0.782072901725769,
+      "learning_rate": 5.682308211516907e-06,
+      "loss": 0.687,
+      "step": 16640
+    },
+    {
+      "epoch": 0.9159007100005504,
+      "grad_norm": 0.6942490339279175,
+      "learning_rate": 5.6818787972625385e-06,
+      "loss": 0.771,
+      "step": 16641
+    },
+    {
+      "epoch": 0.9159557488029061,
+      "grad_norm": 0.7138404846191406,
+      "learning_rate": 5.681449377883352e-06,
+      "loss": 0.8126,
+      "step": 16642
+    },
+    {
+      "epoch": 0.9160107876052617,
+      "grad_norm": 0.7067154049873352,
+      "learning_rate": 5.681019953382575e-06,
+      "loss": 0.7199,
+      "step": 16643
+    },
+    {
+      "epoch": 0.9160658264076174,
+      "grad_norm": 0.7254672646522522,
+      "learning_rate": 5.680590523763433e-06,
+      "loss": 0.7619,
+      "step": 16644
+    },
+    {
+      "epoch": 0.916120865209973,
+      "grad_norm": 0.6963558197021484,
+      "learning_rate": 5.680161089029156e-06,
+      "loss": 0.8141,
+      "step": 16645
+    },
+    {
+      "epoch": 0.9161759040123287,
+      "grad_norm": 0.7634423971176147,
+      "learning_rate": 5.679731649182969e-06,
+      "loss": 0.6432,
+      "step": 16646
+    },
+    {
+      "epoch": 0.9162309428146843,
+      "grad_norm": 0.6261885166168213,
+      "learning_rate": 5.679302204228101e-06,
+      "loss": 0.6637,
+      "step": 16647
+    },
+    {
+      "epoch": 0.91628598161704,
+      "grad_norm": 0.666869580745697,
+      "learning_rate": 5.67887275416778e-06,
+      "loss": 0.7428,
+      "step": 16648
+    },
+    {
+      "epoch": 0.9163410204193957,
+      "grad_norm": 0.698590099811554,
+      "learning_rate": 5.678443299005234e-06,
+      "loss": 0.7404,
+      "step": 16649
+    },
+    {
+      "epoch": 0.9163960592217514,
+      "grad_norm": 0.7000448703765869,
+      "learning_rate": 5.678013838743687e-06,
+      "loss": 0.8314,
+      "step": 16650
+    },
+    {
+      "epoch": 0.916451098024107,
+      "grad_norm": 0.7588967680931091,
+      "learning_rate": 5.677584373386372e-06,
+      "loss": 0.795,
+      "step": 16651
+    },
+    {
+      "epoch": 0.9165061368264626,
+      "grad_norm": 1.0171421766281128,
+      "learning_rate": 5.677154902936515e-06,
+      "loss": 0.8069,
+      "step": 16652
+    },
+    {
+      "epoch": 0.9165611756288183,
+      "grad_norm": 0.699393093585968,
+      "learning_rate": 5.676725427397342e-06,
+      "loss": 0.8354,
+      "step": 16653
+    },
+    {
+      "epoch": 0.916616214431174,
+      "grad_norm": 0.7349410653114319,
+      "learning_rate": 5.676295946772081e-06,
+      "loss": 0.8218,
+      "step": 16654
+    },
+    {
+      "epoch": 0.9166712532335296,
+      "grad_norm": 0.6317818760871887,
+      "learning_rate": 5.6758664610639615e-06,
+      "loss": 0.6601,
+      "step": 16655
+    },
+    {
+      "epoch": 0.9167262920358853,
+      "grad_norm": 0.6990076899528503,
+      "learning_rate": 5.67543697027621e-06,
+      "loss": 0.7225,
+      "step": 16656
+    },
+    {
+      "epoch": 0.916781330838241,
+      "grad_norm": 0.6295570135116577,
+      "learning_rate": 5.675007474412055e-06,
+      "loss": 0.6699,
+      "step": 16657
+    },
+    {
+      "epoch": 0.9168363696405967,
+      "grad_norm": 0.6749002933502197,
+      "learning_rate": 5.674577973474724e-06,
+      "loss": 0.7686,
+      "step": 16658
+    },
+    {
+      "epoch": 0.9168914084429522,
+      "grad_norm": 0.7126505374908447,
+      "learning_rate": 5.674148467467446e-06,
+      "loss": 0.7407,
+      "step": 16659
+    },
+    {
+      "epoch": 0.9169464472453079,
+      "grad_norm": 0.7344494462013245,
+      "learning_rate": 5.67371895639345e-06,
+      "loss": 0.8363,
+      "step": 16660
+    },
+    {
+      "epoch": 0.9170014860476636,
+      "grad_norm": 1.137642741203308,
+      "learning_rate": 5.673289440255962e-06,
+      "loss": 0.7756,
+      "step": 16661
+    },
+    {
+      "epoch": 0.9170565248500193,
+      "grad_norm": 0.7077656388282776,
+      "learning_rate": 5.672859919058209e-06,
+      "loss": 0.7276,
+      "step": 16662
+    },
+    {
+      "epoch": 0.9171115636523749,
+      "grad_norm": 0.6685035824775696,
+      "learning_rate": 5.672430392803423e-06,
+      "loss": 0.6223,
+      "step": 16663
+    },
+    {
+      "epoch": 0.9171666024547306,
+      "grad_norm": 0.7401498556137085,
+      "learning_rate": 5.6720008614948296e-06,
+      "loss": 0.7722,
+      "step": 16664
+    },
+    {
+      "epoch": 0.9172216412570863,
+      "grad_norm": 0.7386918067932129,
+      "learning_rate": 5.6715713251356575e-06,
+      "loss": 0.7338,
+      "step": 16665
+    },
+    {
+      "epoch": 0.9172766800594419,
+      "grad_norm": 0.6631183624267578,
+      "learning_rate": 5.671141783729135e-06,
+      "loss": 0.7338,
+      "step": 16666
+    },
+    {
+      "epoch": 0.9173317188617975,
+      "grad_norm": 0.7793235778808594,
+      "learning_rate": 5.670712237278491e-06,
+      "loss": 0.7597,
+      "step": 16667
+    },
+    {
+      "epoch": 0.9173867576641532,
+      "grad_norm": 0.6626643538475037,
+      "learning_rate": 5.670282685786953e-06,
+      "loss": 0.6953,
+      "step": 16668
+    },
+    {
+      "epoch": 0.9174417964665089,
+      "grad_norm": 0.6923056840896606,
+      "learning_rate": 5.66985312925775e-06,
+      "loss": 0.6598,
+      "step": 16669
+    },
+    {
+      "epoch": 0.9174968352688645,
+      "grad_norm": 0.7240145802497864,
+      "learning_rate": 5.669423567694109e-06,
+      "loss": 0.7927,
+      "step": 16670
+    },
+    {
+      "epoch": 0.9175518740712202,
+      "grad_norm": 0.7715887427330017,
+      "learning_rate": 5.66899400109926e-06,
+      "loss": 0.7028,
+      "step": 16671
+    },
+    {
+      "epoch": 0.9176069128735759,
+      "grad_norm": 0.636465311050415,
+      "learning_rate": 5.668564429476433e-06,
+      "loss": 0.6956,
+      "step": 16672
+    },
+    {
+      "epoch": 0.9176619516759316,
+      "grad_norm": 0.7405944466590881,
+      "learning_rate": 5.668134852828853e-06,
+      "loss": 0.8548,
+      "step": 16673
+    },
+    {
+      "epoch": 0.9177169904782871,
+      "grad_norm": 0.7019765973091125,
+      "learning_rate": 5.6677052711597505e-06,
+      "loss": 0.7633,
+      "step": 16674
+    },
+    {
+      "epoch": 0.9177720292806428,
+      "grad_norm": 0.6769872903823853,
+      "learning_rate": 5.667275684472353e-06,
+      "loss": 0.7114,
+      "step": 16675
+    },
+    {
+      "epoch": 0.9178270680829985,
+      "grad_norm": 0.6730767488479614,
+      "learning_rate": 5.666846092769892e-06,
+      "loss": 0.6733,
+      "step": 16676
+    },
+    {
+      "epoch": 0.9178821068853542,
+      "grad_norm": 0.7447411417961121,
+      "learning_rate": 5.6664164960555935e-06,
+      "loss": 0.742,
+      "step": 16677
+    },
+    {
+      "epoch": 0.9179371456877098,
+      "grad_norm": 0.6733657717704773,
+      "learning_rate": 5.665986894332687e-06,
+      "loss": 0.7784,
+      "step": 16678
+    },
+    {
+      "epoch": 0.9179921844900655,
+      "grad_norm": 0.6672250032424927,
+      "learning_rate": 5.665557287604402e-06,
+      "loss": 0.7409,
+      "step": 16679
+    },
+    {
+      "epoch": 0.9180472232924212,
+      "grad_norm": 0.8029616475105286,
+      "learning_rate": 5.665127675873966e-06,
+      "loss": 0.8032,
+      "step": 16680
+    },
+    {
+      "epoch": 0.9181022620947769,
+      "grad_norm": 0.739986777305603,
+      "learning_rate": 5.664698059144609e-06,
+      "loss": 0.7831,
+      "step": 16681
+    },
+    {
+      "epoch": 0.9181573008971324,
+      "grad_norm": 0.6360197067260742,
+      "learning_rate": 5.6642684374195565e-06,
+      "loss": 0.739,
+      "step": 16682
+    },
+    {
+      "epoch": 0.9182123396994881,
+      "grad_norm": 0.6870883703231812,
+      "learning_rate": 5.663838810702043e-06,
+      "loss": 0.8035,
+      "step": 16683
+    },
+    {
+      "epoch": 0.9182673785018438,
+      "grad_norm": 0.6486376523971558,
+      "learning_rate": 5.663409178995294e-06,
+      "loss": 0.7534,
+      "step": 16684
+    },
+    {
+      "epoch": 0.9183224173041995,
+      "grad_norm": 0.6808338761329651,
+      "learning_rate": 5.662979542302539e-06,
+      "loss": 0.7155,
+      "step": 16685
+    },
+    {
+      "epoch": 0.9183774561065551,
+      "grad_norm": 0.7058351039886475,
+      "learning_rate": 5.6625499006270066e-06,
+      "loss": 0.7006,
+      "step": 16686
+    },
+    {
+      "epoch": 0.9184324949089108,
+      "grad_norm": 0.8551416397094727,
+      "learning_rate": 5.662120253971928e-06,
+      "loss": 0.6619,
+      "step": 16687
+    },
+    {
+      "epoch": 0.9184875337112665,
+      "grad_norm": 0.7475470304489136,
+      "learning_rate": 5.66169060234053e-06,
+      "loss": 0.8396,
+      "step": 16688
+    },
+    {
+      "epoch": 0.9185425725136221,
+      "grad_norm": 0.8063628673553467,
+      "learning_rate": 5.661260945736043e-06,
+      "loss": 0.7052,
+      "step": 16689
+    },
+    {
+      "epoch": 0.9185976113159777,
+      "grad_norm": 0.725739598274231,
+      "learning_rate": 5.660831284161695e-06,
+      "loss": 0.7793,
+      "step": 16690
+    },
+    {
+      "epoch": 0.9186526501183334,
+      "grad_norm": 0.6293089389801025,
+      "learning_rate": 5.660401617620718e-06,
+      "loss": 0.7364,
+      "step": 16691
+    },
+    {
+      "epoch": 0.9187076889206891,
+      "grad_norm": 0.7716445326805115,
+      "learning_rate": 5.659971946116337e-06,
+      "loss": 0.7512,
+      "step": 16692
+    },
+    {
+      "epoch": 0.9187627277230448,
+      "grad_norm": 0.671196460723877,
+      "learning_rate": 5.659542269651784e-06,
+      "loss": 0.724,
+      "step": 16693
+    },
+    {
+      "epoch": 0.9188177665254004,
+      "grad_norm": 0.679114043712616,
+      "learning_rate": 5.659112588230288e-06,
+      "loss": 0.8198,
+      "step": 16694
+    },
+    {
+      "epoch": 0.9188728053277561,
+      "grad_norm": 1.0804789066314697,
+      "learning_rate": 5.658682901855078e-06,
+      "loss": 0.7444,
+      "step": 16695
+    },
+    {
+      "epoch": 0.9189278441301117,
+      "grad_norm": 0.6216247081756592,
+      "learning_rate": 5.658253210529384e-06,
+      "loss": 0.7313,
+      "step": 16696
+    },
+    {
+      "epoch": 0.9189828829324674,
+      "grad_norm": 0.9815022349357605,
+      "learning_rate": 5.657823514256436e-06,
+      "loss": 0.7936,
+      "step": 16697
+    },
+    {
+      "epoch": 0.919037921734823,
+      "grad_norm": 0.6335939168930054,
+      "learning_rate": 5.6573938130394604e-06,
+      "loss": 0.6899,
+      "step": 16698
+    },
+    {
+      "epoch": 0.9190929605371787,
+      "grad_norm": 0.7065087556838989,
+      "learning_rate": 5.6569641068816895e-06,
+      "loss": 0.6891,
+      "step": 16699
+    },
+    {
+      "epoch": 0.9191479993395344,
+      "grad_norm": 0.6492599844932556,
+      "learning_rate": 5.656534395786353e-06,
+      "loss": 0.7571,
+      "step": 16700
+    },
+    {
+      "epoch": 0.9192030381418901,
+      "grad_norm": 0.9063841700553894,
+      "learning_rate": 5.656104679756679e-06,
+      "loss": 0.7297,
+      "step": 16701
+    },
+    {
+      "epoch": 0.9192580769442457,
+      "grad_norm": 0.7018927335739136,
+      "learning_rate": 5.655674958795898e-06,
+      "loss": 0.8285,
+      "step": 16702
+    },
+    {
+      "epoch": 0.9193131157466014,
+      "grad_norm": 0.7087565660476685,
+      "learning_rate": 5.65524523290724e-06,
+      "loss": 0.8399,
+      "step": 16703
+    },
+    {
+      "epoch": 0.919368154548957,
+      "grad_norm": 0.7383968830108643,
+      "learning_rate": 5.6548155020939344e-06,
+      "loss": 0.6995,
+      "step": 16704
+    },
+    {
+      "epoch": 0.9194231933513127,
+      "grad_norm": 0.7523428797721863,
+      "learning_rate": 5.6543857663592104e-06,
+      "loss": 0.7332,
+      "step": 16705
+    },
+    {
+      "epoch": 0.9194782321536683,
+      "grad_norm": 0.761272668838501,
+      "learning_rate": 5.653956025706299e-06,
+      "loss": 0.8085,
+      "step": 16706
+    },
+    {
+      "epoch": 0.919533270956024,
+      "grad_norm": 0.8626916408538818,
+      "learning_rate": 5.653526280138427e-06,
+      "loss": 0.8026,
+      "step": 16707
+    },
+    {
+      "epoch": 0.9195883097583797,
+      "grad_norm": 0.7932292222976685,
+      "learning_rate": 5.653096529658828e-06,
+      "loss": 0.7962,
+      "step": 16708
+    },
+    {
+      "epoch": 0.9196433485607353,
+      "grad_norm": 0.6478299498558044,
+      "learning_rate": 5.65266677427073e-06,
+      "loss": 0.7286,
+      "step": 16709
+    },
+    {
+      "epoch": 0.919698387363091,
+      "grad_norm": 0.5808230042457581,
+      "learning_rate": 5.652237013977363e-06,
+      "loss": 0.6787,
+      "step": 16710
+    },
+    {
+      "epoch": 0.9197534261654466,
+      "grad_norm": 0.6502211689949036,
+      "learning_rate": 5.651807248781956e-06,
+      "loss": 0.6574,
+      "step": 16711
+    },
+    {
+      "epoch": 0.9198084649678023,
+      "grad_norm": 0.7240869998931885,
+      "learning_rate": 5.651377478687743e-06,
+      "loss": 0.8532,
+      "step": 16712
+    },
+    {
+      "epoch": 0.9198635037701579,
+      "grad_norm": 0.8461616635322571,
+      "learning_rate": 5.6509477036979495e-06,
+      "loss": 0.8315,
+      "step": 16713
+    },
+    {
+      "epoch": 0.9199185425725136,
+      "grad_norm": 0.6744678616523743,
+      "learning_rate": 5.650517923815808e-06,
+      "loss": 0.7464,
+      "step": 16714
+    },
+    {
+      "epoch": 0.9199735813748693,
+      "grad_norm": 0.6698602437973022,
+      "learning_rate": 5.650088139044547e-06,
+      "loss": 0.7194,
+      "step": 16715
+    },
+    {
+      "epoch": 0.920028620177225,
+      "grad_norm": 0.7550846934318542,
+      "learning_rate": 5.649658349387399e-06,
+      "loss": 0.754,
+      "step": 16716
+    },
+    {
+      "epoch": 0.9200836589795806,
+      "grad_norm": 0.7450975775718689,
+      "learning_rate": 5.649228554847592e-06,
+      "loss": 0.7989,
+      "step": 16717
+    },
+    {
+      "epoch": 0.9201386977819362,
+      "grad_norm": 0.6790910363197327,
+      "learning_rate": 5.648798755428357e-06,
+      "loss": 0.7259,
+      "step": 16718
+    },
+    {
+      "epoch": 0.9201937365842919,
+      "grad_norm": 0.6760334372520447,
+      "learning_rate": 5.648368951132923e-06,
+      "loss": 0.7317,
+      "step": 16719
+    },
+    {
+      "epoch": 0.9202487753866476,
+      "grad_norm": 0.6798209547996521,
+      "learning_rate": 5.6479391419645225e-06,
+      "loss": 0.7226,
+      "step": 16720
+    },
+    {
+      "epoch": 0.9203038141890032,
+      "grad_norm": 0.6167044639587402,
+      "learning_rate": 5.647509327926386e-06,
+      "loss": 0.7325,
+      "step": 16721
+    },
+    {
+      "epoch": 0.9203588529913589,
+      "grad_norm": 0.7588209509849548,
+      "learning_rate": 5.6470795090217414e-06,
+      "loss": 0.8487,
+      "step": 16722
+    },
+    {
+      "epoch": 0.9204138917937146,
+      "grad_norm": 0.6457192301750183,
+      "learning_rate": 5.64664968525382e-06,
+      "loss": 0.7436,
+      "step": 16723
+    },
+    {
+      "epoch": 0.9204689305960703,
+      "grad_norm": 0.7606278657913208,
+      "learning_rate": 5.6462198566258534e-06,
+      "loss": 0.7431,
+      "step": 16724
+    },
+    {
+      "epoch": 0.9205239693984258,
+      "grad_norm": 0.7064768075942993,
+      "learning_rate": 5.645790023141071e-06,
+      "loss": 0.7231,
+      "step": 16725
+    },
+    {
+      "epoch": 0.9205790082007815,
+      "grad_norm": 0.6990096569061279,
+      "learning_rate": 5.645360184802704e-06,
+      "loss": 0.7883,
+      "step": 16726
+    },
+    {
+      "epoch": 0.9206340470031372,
+      "grad_norm": 0.6305971741676331,
+      "learning_rate": 5.6449303416139835e-06,
+      "loss": 0.7124,
+      "step": 16727
+    },
+    {
+      "epoch": 0.9206890858054929,
+      "grad_norm": 0.6472275853157043,
+      "learning_rate": 5.644500493578138e-06,
+      "loss": 0.6776,
+      "step": 16728
+    },
+    {
+      "epoch": 0.9207441246078485,
+      "grad_norm": 0.6112511157989502,
+      "learning_rate": 5.644070640698399e-06,
+      "loss": 0.6532,
+      "step": 16729
+    },
+    {
+      "epoch": 0.9207991634102042,
+      "grad_norm": 0.6955081224441528,
+      "learning_rate": 5.643640782977998e-06,
+      "loss": 0.6718,
+      "step": 16730
+    },
+    {
+      "epoch": 0.9208542022125599,
+      "grad_norm": 0.6596464514732361,
+      "learning_rate": 5.643210920420165e-06,
+      "loss": 0.7317,
+      "step": 16731
+    },
+    {
+      "epoch": 0.9209092410149156,
+      "grad_norm": 0.8585140705108643,
+      "learning_rate": 5.642781053028131e-06,
+      "loss": 0.747,
+      "step": 16732
+    },
+    {
+      "epoch": 0.9209642798172711,
+      "grad_norm": 0.6630815863609314,
+      "learning_rate": 5.642351180805127e-06,
+      "loss": 0.7175,
+      "step": 16733
+    },
+    {
+      "epoch": 0.9210193186196268,
+      "grad_norm": 0.6956483721733093,
+      "learning_rate": 5.641921303754383e-06,
+      "loss": 0.7509,
+      "step": 16734
+    },
+    {
+      "epoch": 0.9210743574219825,
+      "grad_norm": 0.7407554388046265,
+      "learning_rate": 5.641491421879131e-06,
+      "loss": 0.7198,
+      "step": 16735
+    },
+    {
+      "epoch": 0.9211293962243382,
+      "grad_norm": 0.760361909866333,
+      "learning_rate": 5.6410615351826e-06,
+      "loss": 0.8267,
+      "step": 16736
+    },
+    {
+      "epoch": 0.9211844350266938,
+      "grad_norm": 0.6338782906532288,
+      "learning_rate": 5.640631643668023e-06,
+      "loss": 0.7637,
+      "step": 16737
+    },
+    {
+      "epoch": 0.9212394738290495,
+      "grad_norm": 0.6906964182853699,
+      "learning_rate": 5.64020174733863e-06,
+      "loss": 0.7809,
+      "step": 16738
+    },
+    {
+      "epoch": 0.9212945126314052,
+      "grad_norm": 0.7165959477424622,
+      "learning_rate": 5.639771846197652e-06,
+      "loss": 0.7886,
+      "step": 16739
+    },
+    {
+      "epoch": 0.9213495514337608,
+      "grad_norm": 0.8010871410369873,
+      "learning_rate": 5.639341940248319e-06,
+      "loss": 0.9206,
+      "step": 16740
+    },
+    {
+      "epoch": 0.9214045902361164,
+      "grad_norm": 0.715826153755188,
+      "learning_rate": 5.638912029493865e-06,
+      "loss": 0.7848,
+      "step": 16741
+    },
+    {
+      "epoch": 0.9214596290384721,
+      "grad_norm": 0.7436164617538452,
+      "learning_rate": 5.638482113937519e-06,
+      "loss": 0.7267,
+      "step": 16742
+    },
+    {
+      "epoch": 0.9215146678408278,
+      "grad_norm": 0.6157627105712891,
+      "learning_rate": 5.6380521935825105e-06,
+      "loss": 0.6634,
+      "step": 16743
+    },
+    {
+      "epoch": 0.9215697066431835,
+      "grad_norm": 0.8817352652549744,
+      "learning_rate": 5.637622268432074e-06,
+      "loss": 0.8532,
+      "step": 16744
+    },
+    {
+      "epoch": 0.9216247454455391,
+      "grad_norm": 0.8750252723693848,
+      "learning_rate": 5.637192338489439e-06,
+      "loss": 0.732,
+      "step": 16745
+    },
+    {
+      "epoch": 0.9216797842478948,
+      "grad_norm": 0.787200391292572,
+      "learning_rate": 5.636762403757836e-06,
+      "loss": 0.7485,
+      "step": 16746
+    },
+    {
+      "epoch": 0.9217348230502505,
+      "grad_norm": 0.7141642570495605,
+      "learning_rate": 5.636332464240498e-06,
+      "loss": 0.6631,
+      "step": 16747
+    },
+    {
+      "epoch": 0.9217898618526061,
+      "grad_norm": 0.636040449142456,
+      "learning_rate": 5.6359025199406555e-06,
+      "loss": 0.7597,
+      "step": 16748
+    },
+    {
+      "epoch": 0.9218449006549617,
+      "grad_norm": 0.6707825660705566,
+      "learning_rate": 5.63547257086154e-06,
+      "loss": 0.7931,
+      "step": 16749
+    },
+    {
+      "epoch": 0.9218999394573174,
+      "grad_norm": 0.7076492309570312,
+      "learning_rate": 5.635042617006383e-06,
+      "loss": 0.7502,
+      "step": 16750
+    },
+    {
+      "epoch": 0.9219549782596731,
+      "grad_norm": 0.7007712721824646,
+      "learning_rate": 5.6346126583784156e-06,
+      "loss": 0.7147,
+      "step": 16751
+    },
+    {
+      "epoch": 0.9220100170620287,
+      "grad_norm": 0.6958491802215576,
+      "learning_rate": 5.634182694980869e-06,
+      "loss": 0.7522,
+      "step": 16752
+    },
+    {
+      "epoch": 0.9220650558643844,
+      "grad_norm": 0.7442490458488464,
+      "learning_rate": 5.633752726816976e-06,
+      "loss": 0.8033,
+      "step": 16753
+    },
+    {
+      "epoch": 0.92212009466674,
+      "grad_norm": 0.6746878623962402,
+      "learning_rate": 5.633322753889966e-06,
+      "loss": 0.6874,
+      "step": 16754
+    },
+    {
+      "epoch": 0.9221751334690957,
+      "grad_norm": 0.5542443990707397,
+      "learning_rate": 5.63289277620307e-06,
+      "loss": 0.5964,
+      "step": 16755
+    },
+    {
+      "epoch": 0.9222301722714513,
+      "grad_norm": 0.7200949192047119,
+      "learning_rate": 5.632462793759524e-06,
+      "loss": 0.7213,
+      "step": 16756
+    },
+    {
+      "epoch": 0.922285211073807,
+      "grad_norm": 0.6765961050987244,
+      "learning_rate": 5.632032806562557e-06,
+      "loss": 0.7509,
+      "step": 16757
+    },
+    {
+      "epoch": 0.9223402498761627,
+      "grad_norm": 0.6492508053779602,
+      "learning_rate": 5.6316028146154e-06,
+      "loss": 0.7374,
+      "step": 16758
+    },
+    {
+      "epoch": 0.9223952886785184,
+      "grad_norm": 0.6642841100692749,
+      "learning_rate": 5.631172817921284e-06,
+      "loss": 0.701,
+      "step": 16759
+    },
+    {
+      "epoch": 0.922450327480874,
+      "grad_norm": 0.6529877781867981,
+      "learning_rate": 5.630742816483443e-06,
+      "loss": 0.7891,
+      "step": 16760
+    },
+    {
+      "epoch": 0.9225053662832297,
+      "grad_norm": 0.7389785051345825,
+      "learning_rate": 5.630312810305109e-06,
+      "loss": 0.7081,
+      "step": 16761
+    },
+    {
+      "epoch": 0.9225604050855853,
+      "grad_norm": 0.6740532517433167,
+      "learning_rate": 5.629882799389511e-06,
+      "loss": 0.599,
+      "step": 16762
+    },
+    {
+      "epoch": 0.922615443887941,
+      "grad_norm": 0.6529923677444458,
+      "learning_rate": 5.629452783739884e-06,
+      "loss": 0.7476,
+      "step": 16763
+    },
+    {
+      "epoch": 0.9226704826902966,
+      "grad_norm": 0.7116754651069641,
+      "learning_rate": 5.629022763359457e-06,
+      "loss": 0.7355,
+      "step": 16764
+    },
+    {
+      "epoch": 0.9227255214926523,
+      "grad_norm": 0.6048651337623596,
+      "learning_rate": 5.628592738251465e-06,
+      "loss": 0.6945,
+      "step": 16765
+    },
+    {
+      "epoch": 0.922780560295008,
+      "grad_norm": 0.5997273921966553,
+      "learning_rate": 5.628162708419137e-06,
+      "loss": 0.7473,
+      "step": 16766
+    },
+    {
+      "epoch": 0.9228355990973637,
+      "grad_norm": 0.6367392539978027,
+      "learning_rate": 5.627732673865705e-06,
+      "loss": 0.7156,
+      "step": 16767
+    },
+    {
+      "epoch": 0.9228906378997193,
+      "grad_norm": 0.6315346360206604,
+      "learning_rate": 5.627302634594404e-06,
+      "loss": 0.5789,
+      "step": 16768
+    },
+    {
+      "epoch": 0.922945676702075,
+      "grad_norm": 0.6221422553062439,
+      "learning_rate": 5.626872590608465e-06,
+      "loss": 0.742,
+      "step": 16769
+    },
+    {
+      "epoch": 0.9230007155044306,
+      "grad_norm": 0.6551609039306641,
+      "learning_rate": 5.626442541911117e-06,
+      "loss": 0.7986,
+      "step": 16770
+    },
+    {
+      "epoch": 0.9230557543067863,
+      "grad_norm": 0.6395416259765625,
+      "learning_rate": 5.626012488505595e-06,
+      "loss": 0.719,
+      "step": 16771
+    },
+    {
+      "epoch": 0.9231107931091419,
+      "grad_norm": 0.784221351146698,
+      "learning_rate": 5.625582430395132e-06,
+      "loss": 0.8514,
+      "step": 16772
+    },
+    {
+      "epoch": 0.9231658319114976,
+      "grad_norm": 0.6133732199668884,
+      "learning_rate": 5.625152367582959e-06,
+      "loss": 0.6405,
+      "step": 16773
+    },
+    {
+      "epoch": 0.9232208707138533,
+      "grad_norm": 0.6891661286354065,
+      "learning_rate": 5.6247223000723075e-06,
+      "loss": 0.6904,
+      "step": 16774
+    },
+    {
+      "epoch": 0.923275909516209,
+      "grad_norm": 0.664445161819458,
+      "learning_rate": 5.62429222786641e-06,
+      "loss": 0.8226,
+      "step": 16775
+    },
+    {
+      "epoch": 0.9233309483185645,
+      "grad_norm": 0.8657387495040894,
+      "learning_rate": 5.6238621509685e-06,
+      "loss": 0.7789,
+      "step": 16776
+    },
+    {
+      "epoch": 0.9233859871209202,
+      "grad_norm": 0.6581356525421143,
+      "learning_rate": 5.623432069381809e-06,
+      "loss": 0.6324,
+      "step": 16777
+    },
+    {
+      "epoch": 0.9234410259232759,
+      "grad_norm": 0.6915359497070312,
+      "learning_rate": 5.623001983109569e-06,
+      "loss": 0.7759,
+      "step": 16778
+    },
+    {
+      "epoch": 0.9234960647256316,
+      "grad_norm": 0.6104419827461243,
+      "learning_rate": 5.622571892155013e-06,
+      "loss": 0.6507,
+      "step": 16779
+    },
+    {
+      "epoch": 0.9235511035279872,
+      "grad_norm": 0.7881014943122864,
+      "learning_rate": 5.622141796521373e-06,
+      "loss": 0.6996,
+      "step": 16780
+    },
+    {
+      "epoch": 0.9236061423303429,
+      "grad_norm": 0.6843274831771851,
+      "learning_rate": 5.621711696211882e-06,
+      "loss": 0.757,
+      "step": 16781
+    },
+    {
+      "epoch": 0.9236611811326986,
+      "grad_norm": 0.6329718232154846,
+      "learning_rate": 5.6212815912297735e-06,
+      "loss": 0.7618,
+      "step": 16782
+    },
+    {
+      "epoch": 0.9237162199350543,
+      "grad_norm": 0.7012547254562378,
+      "learning_rate": 5.6208514815782765e-06,
+      "loss": 0.8123,
+      "step": 16783
+    },
+    {
+      "epoch": 0.9237712587374098,
+      "grad_norm": 0.5882138013839722,
+      "learning_rate": 5.6204213672606275e-06,
+      "loss": 0.5835,
+      "step": 16784
+    },
+    {
+      "epoch": 0.9238262975397655,
+      "grad_norm": 0.7064856886863708,
+      "learning_rate": 5.619991248280058e-06,
+      "loss": 0.7661,
+      "step": 16785
+    },
+    {
+      "epoch": 0.9238813363421212,
+      "grad_norm": 0.6935905814170837,
+      "learning_rate": 5.619561124639801e-06,
+      "loss": 0.7613,
+      "step": 16786
+    },
+    {
+      "epoch": 0.9239363751444769,
+      "grad_norm": 0.6331823468208313,
+      "learning_rate": 5.619130996343086e-06,
+      "loss": 0.7639,
+      "step": 16787
+    },
+    {
+      "epoch": 0.9239914139468325,
+      "grad_norm": 0.9952171444892883,
+      "learning_rate": 5.61870086339315e-06,
+      "loss": 0.7417,
+      "step": 16788
+    },
+    {
+      "epoch": 0.9240464527491882,
+      "grad_norm": 0.7930924892425537,
+      "learning_rate": 5.618270725793223e-06,
+      "loss": 0.8565,
+      "step": 16789
+    },
+    {
+      "epoch": 0.9241014915515439,
+      "grad_norm": 0.6192849278450012,
+      "learning_rate": 5.617840583546539e-06,
+      "loss": 0.6792,
+      "step": 16790
+    },
+    {
+      "epoch": 0.9241565303538996,
+      "grad_norm": 0.7150148749351501,
+      "learning_rate": 5.61741043665633e-06,
+      "loss": 0.8016,
+      "step": 16791
+    },
+    {
+      "epoch": 0.9242115691562551,
+      "grad_norm": 0.7186499238014221,
+      "learning_rate": 5.616980285125829e-06,
+      "loss": 0.8054,
+      "step": 16792
+    },
+    {
+      "epoch": 0.9242666079586108,
+      "grad_norm": 0.765373706817627,
+      "learning_rate": 5.616550128958271e-06,
+      "loss": 0.7506,
+      "step": 16793
+    },
+    {
+      "epoch": 0.9243216467609665,
+      "grad_norm": 0.7462981939315796,
+      "learning_rate": 5.6161199681568865e-06,
+      "loss": 0.7942,
+      "step": 16794
+    },
+    {
+      "epoch": 0.9243766855633221,
+      "grad_norm": 0.8685716390609741,
+      "learning_rate": 5.615689802724908e-06,
+      "loss": 0.7345,
+      "step": 16795
+    },
+    {
+      "epoch": 0.9244317243656778,
+      "grad_norm": 0.7249016165733337,
+      "learning_rate": 5.615259632665572e-06,
+      "loss": 0.7717,
+      "step": 16796
+    },
+    {
+      "epoch": 0.9244867631680335,
+      "grad_norm": 0.6986450552940369,
+      "learning_rate": 5.614829457982108e-06,
+      "loss": 0.7059,
+      "step": 16797
+    },
+    {
+      "epoch": 0.9245418019703892,
+      "grad_norm": 0.6127453446388245,
+      "learning_rate": 5.614399278677753e-06,
+      "loss": 0.679,
+      "step": 16798
+    },
+    {
+      "epoch": 0.9245968407727447,
+      "grad_norm": 0.8027729392051697,
+      "learning_rate": 5.613969094755734e-06,
+      "loss": 0.8873,
+      "step": 16799
+    },
+    {
+      "epoch": 0.9246518795751004,
+      "grad_norm": 0.7740058898925781,
+      "learning_rate": 5.613538906219288e-06,
+      "loss": 0.7232,
+      "step": 16800
+    },
+    {
+      "epoch": 0.9247069183774561,
+      "grad_norm": 0.6654518246650696,
+      "learning_rate": 5.613108713071649e-06,
+      "loss": 0.6444,
+      "step": 16801
+    },
+    {
+      "epoch": 0.9247619571798118,
+      "grad_norm": 0.8171139359474182,
+      "learning_rate": 5.612678515316049e-06,
+      "loss": 0.7811,
+      "step": 16802
+    },
+    {
+      "epoch": 0.9248169959821674,
+      "grad_norm": 0.6470557451248169,
+      "learning_rate": 5.612248312955719e-06,
+      "loss": 0.724,
+      "step": 16803
+    },
+    {
+      "epoch": 0.9248720347845231,
+      "grad_norm": 0.6510154604911804,
+      "learning_rate": 5.611818105993897e-06,
+      "loss": 0.7413,
+      "step": 16804
+    },
+    {
+      "epoch": 0.9249270735868788,
+      "grad_norm": 0.6152756810188293,
+      "learning_rate": 5.611387894433814e-06,
+      "loss": 0.6736,
+      "step": 16805
+    },
+    {
+      "epoch": 0.9249821123892344,
+      "grad_norm": 0.7700890898704529,
+      "learning_rate": 5.610957678278702e-06,
+      "loss": 0.7899,
+      "step": 16806
+    },
+    {
+      "epoch": 0.92503715119159,
+      "grad_norm": 0.6250771284103394,
+      "learning_rate": 5.610527457531796e-06,
+      "loss": 0.6656,
+      "step": 16807
+    },
+    {
+      "epoch": 0.9250921899939457,
+      "grad_norm": 0.7004884481430054,
+      "learning_rate": 5.6100972321963295e-06,
+      "loss": 0.7679,
+      "step": 16808
+    },
+    {
+      "epoch": 0.9251472287963014,
+      "grad_norm": 0.7237122058868408,
+      "learning_rate": 5.609667002275535e-06,
+      "loss": 0.7014,
+      "step": 16809
+    },
+    {
+      "epoch": 0.9252022675986571,
+      "grad_norm": 0.7693857550621033,
+      "learning_rate": 5.609236767772649e-06,
+      "loss": 0.6886,
+      "step": 16810
+    },
+    {
+      "epoch": 0.9252573064010127,
+      "grad_norm": 0.738211989402771,
+      "learning_rate": 5.608806528690899e-06,
+      "loss": 0.759,
+      "step": 16811
+    },
+    {
+      "epoch": 0.9253123452033684,
+      "grad_norm": 0.6153057813644409,
+      "learning_rate": 5.608376285033524e-06,
+      "loss": 0.6771,
+      "step": 16812
+    },
+    {
+      "epoch": 0.925367384005724,
+      "grad_norm": 0.6428161263465881,
+      "learning_rate": 5.607946036803755e-06,
+      "loss": 0.5967,
+      "step": 16813
+    },
+    {
+      "epoch": 0.9254224228080797,
+      "grad_norm": 0.659237802028656,
+      "learning_rate": 5.607515784004827e-06,
+      "loss": 0.7997,
+      "step": 16814
+    },
+    {
+      "epoch": 0.9254774616104353,
+      "grad_norm": 0.7318007946014404,
+      "learning_rate": 5.60708552663997e-06,
+      "loss": 0.6912,
+      "step": 16815
+    },
+    {
+      "epoch": 0.925532500412791,
+      "grad_norm": 0.65810227394104,
+      "learning_rate": 5.606655264712424e-06,
+      "loss": 0.7702,
+      "step": 16816
+    },
+    {
+      "epoch": 0.9255875392151467,
+      "grad_norm": 0.6517447233200073,
+      "learning_rate": 5.606224998225419e-06,
+      "loss": 0.7757,
+      "step": 16817
+    },
+    {
+      "epoch": 0.9256425780175024,
+      "grad_norm": 0.8019997477531433,
+      "learning_rate": 5.605794727182189e-06,
+      "loss": 0.755,
+      "step": 16818
+    },
+    {
+      "epoch": 0.925697616819858,
+      "grad_norm": 0.5735001564025879,
+      "learning_rate": 5.605364451585966e-06,
+      "loss": 0.6594,
+      "step": 16819
+    },
+    {
+      "epoch": 0.9257526556222136,
+      "grad_norm": 0.7946392297744751,
+      "learning_rate": 5.604934171439989e-06,
+      "loss": 0.8566,
+      "step": 16820
+    },
+    {
+      "epoch": 0.9258076944245693,
+      "grad_norm": 0.6836885809898376,
+      "learning_rate": 5.6045038867474865e-06,
+      "loss": 0.7309,
+      "step": 16821
+    },
+    {
+      "epoch": 0.925862733226925,
+      "grad_norm": 0.7148752808570862,
+      "learning_rate": 5.604073597511697e-06,
+      "loss": 0.7961,
+      "step": 16822
+    },
+    {
+      "epoch": 0.9259177720292806,
+      "grad_norm": 0.5976021885871887,
+      "learning_rate": 5.60364330373585e-06,
+      "loss": 0.702,
+      "step": 16823
+    },
+    {
+      "epoch": 0.9259728108316363,
+      "grad_norm": 0.7915990948677063,
+      "learning_rate": 5.6032130054231815e-06,
+      "loss": 0.7248,
+      "step": 16824
+    },
+    {
+      "epoch": 0.926027849633992,
+      "grad_norm": 0.8617003560066223,
+      "learning_rate": 5.602782702576925e-06,
+      "loss": 0.7586,
+      "step": 16825
+    },
+    {
+      "epoch": 0.9260828884363477,
+      "grad_norm": 0.707922637462616,
+      "learning_rate": 5.602352395200317e-06,
+      "loss": 0.7649,
+      "step": 16826
+    },
+    {
+      "epoch": 0.9261379272387033,
+      "grad_norm": 0.702974259853363,
+      "learning_rate": 5.601922083296588e-06,
+      "loss": 0.7868,
+      "step": 16827
+    },
+    {
+      "epoch": 0.9261929660410589,
+      "grad_norm": 0.8666204810142517,
+      "learning_rate": 5.601491766868974e-06,
+      "loss": 0.6743,
+      "step": 16828
+    },
+    {
+      "epoch": 0.9262480048434146,
+      "grad_norm": 0.7228319644927979,
+      "learning_rate": 5.60106144592071e-06,
+      "loss": 0.9038,
+      "step": 16829
+    },
+    {
+      "epoch": 0.9263030436457703,
+      "grad_norm": 0.6495303511619568,
+      "learning_rate": 5.600631120455028e-06,
+      "loss": 0.675,
+      "step": 16830
+    },
+    {
+      "epoch": 0.9263580824481259,
+      "grad_norm": 0.909911572933197,
+      "learning_rate": 5.600200790475163e-06,
+      "loss": 0.8512,
+      "step": 16831
+    },
+    {
+      "epoch": 0.9264131212504816,
+      "grad_norm": 0.8628181219100952,
+      "learning_rate": 5.59977045598435e-06,
+      "loss": 0.8263,
+      "step": 16832
+    },
+    {
+      "epoch": 0.9264681600528373,
+      "grad_norm": 0.7378986477851868,
+      "learning_rate": 5.599340116985823e-06,
+      "loss": 0.6147,
+      "step": 16833
+    },
+    {
+      "epoch": 0.926523198855193,
+      "grad_norm": 0.7630714774131775,
+      "learning_rate": 5.598909773482818e-06,
+      "loss": 0.7488,
+      "step": 16834
+    },
+    {
+      "epoch": 0.9265782376575485,
+      "grad_norm": 0.6660358905792236,
+      "learning_rate": 5.598479425478564e-06,
+      "loss": 0.7846,
+      "step": 16835
+    },
+    {
+      "epoch": 0.9266332764599042,
+      "grad_norm": 0.633204996585846,
+      "learning_rate": 5.5980490729763e-06,
+      "loss": 0.719,
+      "step": 16836
+    },
+    {
+      "epoch": 0.9266883152622599,
+      "grad_norm": 0.7990162968635559,
+      "learning_rate": 5.5976187159792605e-06,
+      "loss": 0.7774,
+      "step": 16837
+    },
+    {
+      "epoch": 0.9267433540646155,
+      "grad_norm": 0.7380030155181885,
+      "learning_rate": 5.597188354490678e-06,
+      "loss": 0.6434,
+      "step": 16838
+    },
+    {
+      "epoch": 0.9267983928669712,
+      "grad_norm": 0.7048906683921814,
+      "learning_rate": 5.596757988513788e-06,
+      "loss": 0.7559,
+      "step": 16839
+    },
+    {
+      "epoch": 0.9268534316693269,
+      "grad_norm": 0.8689755201339722,
+      "learning_rate": 5.596327618051823e-06,
+      "loss": 0.7003,
+      "step": 16840
+    },
+    {
+      "epoch": 0.9269084704716826,
+      "grad_norm": 0.6776423454284668,
+      "learning_rate": 5.595897243108021e-06,
+      "loss": 0.7535,
+      "step": 16841
+    },
+    {
+      "epoch": 0.9269635092740381,
+      "grad_norm": 0.6138951778411865,
+      "learning_rate": 5.595466863685614e-06,
+      "loss": 0.7471,
+      "step": 16842
+    },
+    {
+      "epoch": 0.9270185480763938,
+      "grad_norm": 0.6530849933624268,
+      "learning_rate": 5.595036479787838e-06,
+      "loss": 0.7595,
+      "step": 16843
+    },
+    {
+      "epoch": 0.9270735868787495,
+      "grad_norm": 0.672170102596283,
+      "learning_rate": 5.594606091417926e-06,
+      "loss": 0.619,
+      "step": 16844
+    },
+    {
+      "epoch": 0.9271286256811052,
+      "grad_norm": 0.7782242894172668,
+      "learning_rate": 5.594175698579114e-06,
+      "loss": 0.7405,
+      "step": 16845
+    },
+    {
+      "epoch": 0.9271836644834608,
+      "grad_norm": 0.7308391332626343,
+      "learning_rate": 5.593745301274639e-06,
+      "loss": 0.766,
+      "step": 16846
+    },
+    {
+      "epoch": 0.9272387032858165,
+      "grad_norm": 0.8123987913131714,
+      "learning_rate": 5.593314899507731e-06,
+      "loss": 0.8137,
+      "step": 16847
+    },
+    {
+      "epoch": 0.9272937420881722,
+      "grad_norm": 0.6876010894775391,
+      "learning_rate": 5.592884493281626e-06,
+      "loss": 0.6828,
+      "step": 16848
+    },
+    {
+      "epoch": 0.9273487808905279,
+      "grad_norm": 0.7311702966690063,
+      "learning_rate": 5.5924540825995624e-06,
+      "loss": 0.8712,
+      "step": 16849
+    },
+    {
+      "epoch": 0.9274038196928834,
+      "grad_norm": 0.6796336770057678,
+      "learning_rate": 5.592023667464771e-06,
+      "loss": 0.7453,
+      "step": 16850
+    },
+    {
+      "epoch": 0.9274588584952391,
+      "grad_norm": 1.1471986770629883,
+      "learning_rate": 5.591593247880489e-06,
+      "loss": 0.7521,
+      "step": 16851
+    },
+    {
+      "epoch": 0.9275138972975948,
+      "grad_norm": 0.6603145599365234,
+      "learning_rate": 5.5911628238499496e-06,
+      "loss": 0.7757,
+      "step": 16852
+    },
+    {
+      "epoch": 0.9275689360999505,
+      "grad_norm": 0.6421443819999695,
+      "learning_rate": 5.590732395376388e-06,
+      "loss": 0.679,
+      "step": 16853
+    },
+    {
+      "epoch": 0.9276239749023061,
+      "grad_norm": 0.6478089690208435,
+      "learning_rate": 5.590301962463041e-06,
+      "loss": 0.6818,
+      "step": 16854
+    },
+    {
+      "epoch": 0.9276790137046618,
+      "grad_norm": 0.742957353591919,
+      "learning_rate": 5.589871525113143e-06,
+      "loss": 0.7611,
+      "step": 16855
+    },
+    {
+      "epoch": 0.9277340525070175,
+      "grad_norm": 0.760917603969574,
+      "learning_rate": 5.589441083329927e-06,
+      "loss": 0.7634,
+      "step": 16856
+    },
+    {
+      "epoch": 0.9277890913093731,
+      "grad_norm": 0.8060788512229919,
+      "learning_rate": 5.589010637116632e-06,
+      "loss": 0.8003,
+      "step": 16857
+    },
+    {
+      "epoch": 0.9278441301117287,
+      "grad_norm": 0.6400671601295471,
+      "learning_rate": 5.588580186476489e-06,
+      "loss": 0.6459,
+      "step": 16858
+    },
+    {
+      "epoch": 0.9278991689140844,
+      "grad_norm": 0.6653194427490234,
+      "learning_rate": 5.5881497314127354e-06,
+      "loss": 0.6536,
+      "step": 16859
+    },
+    {
+      "epoch": 0.9279542077164401,
+      "grad_norm": 0.7482097148895264,
+      "learning_rate": 5.587719271928605e-06,
+      "loss": 0.6866,
+      "step": 16860
+    },
+    {
+      "epoch": 0.9280092465187958,
+      "grad_norm": 0.6429533958435059,
+      "learning_rate": 5.5872888080273345e-06,
+      "loss": 0.7135,
+      "step": 16861
+    },
+    {
+      "epoch": 0.9280642853211514,
+      "grad_norm": 0.6756870746612549,
+      "learning_rate": 5.586858339712159e-06,
+      "loss": 0.6738,
+      "step": 16862
+    },
+    {
+      "epoch": 0.9281193241235071,
+      "grad_norm": 0.6506288051605225,
+      "learning_rate": 5.586427866986313e-06,
+      "loss": 0.6872,
+      "step": 16863
+    },
+    {
+      "epoch": 0.9281743629258628,
+      "grad_norm": 0.6222309470176697,
+      "learning_rate": 5.585997389853031e-06,
+      "loss": 0.6227,
+      "step": 16864
+    },
+    {
+      "epoch": 0.9282294017282184,
+      "grad_norm": 0.7656785845756531,
+      "learning_rate": 5.58556690831555e-06,
+      "loss": 0.7487,
+      "step": 16865
+    },
+    {
+      "epoch": 0.928284440530574,
+      "grad_norm": 0.8085651993751526,
+      "learning_rate": 5.585136422377105e-06,
+      "loss": 0.7495,
+      "step": 16866
+    },
+    {
+      "epoch": 0.9283394793329297,
+      "grad_norm": 0.7220577597618103,
+      "learning_rate": 5.584705932040933e-06,
+      "loss": 0.7616,
+      "step": 16867
+    },
+    {
+      "epoch": 0.9283945181352854,
+      "grad_norm": 0.6376538872718811,
+      "learning_rate": 5.5842754373102646e-06,
+      "loss": 0.6959,
+      "step": 16868
+    },
+    {
+      "epoch": 0.9284495569376411,
+      "grad_norm": 0.8235114216804504,
+      "learning_rate": 5.583844938188341e-06,
+      "loss": 0.8405,
+      "step": 16869
+    },
+    {
+      "epoch": 0.9285045957399967,
+      "grad_norm": 0.6148403286933899,
+      "learning_rate": 5.583414434678395e-06,
+      "loss": 0.6496,
+      "step": 16870
+    },
+    {
+      "epoch": 0.9285596345423524,
+      "grad_norm": 0.7475830912590027,
+      "learning_rate": 5.5829839267836606e-06,
+      "loss": 0.7728,
+      "step": 16871
+    },
+    {
+      "epoch": 0.928614673344708,
+      "grad_norm": 0.6615459322929382,
+      "learning_rate": 5.582553414507376e-06,
+      "loss": 0.7168,
+      "step": 16872
+    },
+    {
+      "epoch": 0.9286697121470637,
+      "grad_norm": 0.7308976054191589,
+      "learning_rate": 5.582122897852776e-06,
+      "loss": 0.8668,
+      "step": 16873
+    },
+    {
+      "epoch": 0.9287247509494193,
+      "grad_norm": 2.0933802127838135,
+      "learning_rate": 5.5816923768230955e-06,
+      "loss": 0.8744,
+      "step": 16874
+    },
+    {
+      "epoch": 0.928779789751775,
+      "grad_norm": 0.8197827935218811,
+      "learning_rate": 5.581261851421571e-06,
+      "loss": 0.965,
+      "step": 16875
+    },
+    {
+      "epoch": 0.9288348285541307,
+      "grad_norm": 0.7161306738853455,
+      "learning_rate": 5.580831321651437e-06,
+      "loss": 0.7199,
+      "step": 16876
+    },
+    {
+      "epoch": 0.9288898673564864,
+      "grad_norm": 0.6676005721092224,
+      "learning_rate": 5.580400787515932e-06,
+      "loss": 0.8178,
+      "step": 16877
+    },
+    {
+      "epoch": 0.928944906158842,
+      "grad_norm": 0.6874394416809082,
+      "learning_rate": 5.579970249018291e-06,
+      "loss": 0.7827,
+      "step": 16878
+    },
+    {
+      "epoch": 0.9289999449611976,
+      "grad_norm": 0.6925203800201416,
+      "learning_rate": 5.579539706161746e-06,
+      "loss": 0.7628,
+      "step": 16879
+    },
+    {
+      "epoch": 0.9290549837635533,
+      "grad_norm": 0.7088871598243713,
+      "learning_rate": 5.5791091589495375e-06,
+      "loss": 0.8104,
+      "step": 16880
+    },
+    {
+      "epoch": 0.9291100225659089,
+      "grad_norm": 0.640397846698761,
+      "learning_rate": 5.5786786073849e-06,
+      "loss": 0.7041,
+      "step": 16881
+    },
+    {
+      "epoch": 0.9291650613682646,
+      "grad_norm": 0.6750220060348511,
+      "learning_rate": 5.578248051471068e-06,
+      "loss": 0.7124,
+      "step": 16882
+    },
+    {
+      "epoch": 0.9292201001706203,
+      "grad_norm": 0.6361757516860962,
+      "learning_rate": 5.577817491211279e-06,
+      "loss": 0.6831,
+      "step": 16883
+    },
+    {
+      "epoch": 0.929275138972976,
+      "grad_norm": 0.710574746131897,
+      "learning_rate": 5.577386926608766e-06,
+      "loss": 0.8622,
+      "step": 16884
+    },
+    {
+      "epoch": 0.9293301777753316,
+      "grad_norm": 0.6523045301437378,
+      "learning_rate": 5.576956357666771e-06,
+      "loss": 0.7014,
+      "step": 16885
+    },
+    {
+      "epoch": 0.9293852165776872,
+      "grad_norm": 0.7073860764503479,
+      "learning_rate": 5.576525784388525e-06,
+      "loss": 0.7735,
+      "step": 16886
+    },
+    {
+      "epoch": 0.9294402553800429,
+      "grad_norm": 0.6221209764480591,
+      "learning_rate": 5.576095206777266e-06,
+      "loss": 0.7262,
+      "step": 16887
+    },
+    {
+      "epoch": 0.9294952941823986,
+      "grad_norm": 0.7045557498931885,
+      "learning_rate": 5.575664624836228e-06,
+      "loss": 0.8519,
+      "step": 16888
+    },
+    {
+      "epoch": 0.9295503329847542,
+      "grad_norm": 0.8938118815422058,
+      "learning_rate": 5.575234038568648e-06,
+      "loss": 0.8044,
+      "step": 16889
+    },
+    {
+      "epoch": 0.9296053717871099,
+      "grad_norm": 0.6333356499671936,
+      "learning_rate": 5.574803447977766e-06,
+      "loss": 0.7061,
+      "step": 16890
+    },
+    {
+      "epoch": 0.9296604105894656,
+      "grad_norm": 0.7838292717933655,
+      "learning_rate": 5.574372853066814e-06,
+      "loss": 0.8472,
+      "step": 16891
+    },
+    {
+      "epoch": 0.9297154493918213,
+      "grad_norm": 0.6740962862968445,
+      "learning_rate": 5.5739422538390285e-06,
+      "loss": 0.8737,
+      "step": 16892
+    },
+    {
+      "epoch": 0.9297704881941768,
+      "grad_norm": 0.6610066890716553,
+      "learning_rate": 5.57351165029765e-06,
+      "loss": 0.7749,
+      "step": 16893
+    },
+    {
+      "epoch": 0.9298255269965325,
+      "grad_norm": 0.6941941380500793,
+      "learning_rate": 5.573081042445908e-06,
+      "loss": 0.7125,
+      "step": 16894
+    },
+    {
+      "epoch": 0.9298805657988882,
+      "grad_norm": 0.6376040577888489,
+      "learning_rate": 5.572650430287043e-06,
+      "loss": 0.7232,
+      "step": 16895
+    },
+    {
+      "epoch": 0.9299356046012439,
+      "grad_norm": 0.6908338665962219,
+      "learning_rate": 5.57221981382429e-06,
+      "loss": 0.7282,
+      "step": 16896
+    },
+    {
+      "epoch": 0.9299906434035995,
+      "grad_norm": 0.7384136915206909,
+      "learning_rate": 5.571789193060887e-06,
+      "loss": 0.7049,
+      "step": 16897
+    },
+    {
+      "epoch": 0.9300456822059552,
+      "grad_norm": 0.6418812870979309,
+      "learning_rate": 5.57135856800007e-06,
+      "loss": 0.6943,
+      "step": 16898
+    },
+    {
+      "epoch": 0.9301007210083109,
+      "grad_norm": 0.6272100806236267,
+      "learning_rate": 5.570927938645075e-06,
+      "loss": 0.6018,
+      "step": 16899
+    },
+    {
+      "epoch": 0.9301557598106666,
+      "grad_norm": 0.6637504696846008,
+      "learning_rate": 5.570497304999136e-06,
+      "loss": 0.7459,
+      "step": 16900
+    },
+    {
+      "epoch": 0.9302107986130221,
+      "grad_norm": 0.6815645098686218,
+      "learning_rate": 5.570066667065495e-06,
+      "loss": 0.8375,
+      "step": 16901
+    },
+    {
+      "epoch": 0.9302658374153778,
+      "grad_norm": 0.7291944026947021,
+      "learning_rate": 5.569636024847384e-06,
+      "loss": 0.7447,
+      "step": 16902
+    },
+    {
+      "epoch": 0.9303208762177335,
+      "grad_norm": 0.7493074536323547,
+      "learning_rate": 5.569205378348041e-06,
+      "loss": 0.7485,
+      "step": 16903
+    },
+    {
+      "epoch": 0.9303759150200892,
+      "grad_norm": 0.6949830651283264,
+      "learning_rate": 5.568774727570702e-06,
+      "loss": 0.6928,
+      "step": 16904
+    },
+    {
+      "epoch": 0.9304309538224448,
+      "grad_norm": 0.7263016104698181,
+      "learning_rate": 5.568344072518606e-06,
+      "loss": 0.7687,
+      "step": 16905
+    },
+    {
+      "epoch": 0.9304859926248005,
+      "grad_norm": 0.6831687688827515,
+      "learning_rate": 5.567913413194989e-06,
+      "loss": 0.8186,
+      "step": 16906
+    },
+    {
+      "epoch": 0.9305410314271562,
+      "grad_norm": 0.6471529603004456,
+      "learning_rate": 5.567482749603086e-06,
+      "loss": 0.809,
+      "step": 16907
+    },
+    {
+      "epoch": 0.9305960702295119,
+      "grad_norm": 0.716443657875061,
+      "learning_rate": 5.5670520817461324e-06,
+      "loss": 0.7556,
+      "step": 16908
+    },
+    {
+      "epoch": 0.9306511090318674,
+      "grad_norm": 0.7684258222579956,
+      "learning_rate": 5.566621409627369e-06,
+      "loss": 0.7163,
+      "step": 16909
+    },
+    {
+      "epoch": 0.9307061478342231,
+      "grad_norm": 0.6364058256149292,
+      "learning_rate": 5.566190733250031e-06,
+      "loss": 0.7008,
+      "step": 16910
+    },
+    {
+      "epoch": 0.9307611866365788,
+      "grad_norm": 0.7371893525123596,
+      "learning_rate": 5.565760052617354e-06,
+      "loss": 0.7377,
+      "step": 16911
+    },
+    {
+      "epoch": 0.9308162254389345,
+      "grad_norm": 0.6785818934440613,
+      "learning_rate": 5.565329367732576e-06,
+      "loss": 0.6958,
+      "step": 16912
+    },
+    {
+      "epoch": 0.9308712642412901,
+      "grad_norm": 0.6949000954627991,
+      "learning_rate": 5.564898678598933e-06,
+      "loss": 0.7061,
+      "step": 16913
+    },
+    {
+      "epoch": 0.9309263030436458,
+      "grad_norm": 0.834689199924469,
+      "learning_rate": 5.5644679852196645e-06,
+      "loss": 0.7584,
+      "step": 16914
+    },
+    {
+      "epoch": 0.9309813418460015,
+      "grad_norm": 0.6549750566482544,
+      "learning_rate": 5.564037287598005e-06,
+      "loss": 0.7846,
+      "step": 16915
+    },
+    {
+      "epoch": 0.9310363806483571,
+      "grad_norm": 0.660965085029602,
+      "learning_rate": 5.563606585737192e-06,
+      "loss": 0.7106,
+      "step": 16916
+    },
+    {
+      "epoch": 0.9310914194507127,
+      "grad_norm": 0.6218705773353577,
+      "learning_rate": 5.563175879640464e-06,
+      "loss": 0.7455,
+      "step": 16917
+    },
+    {
+      "epoch": 0.9311464582530684,
+      "grad_norm": 0.6786651015281677,
+      "learning_rate": 5.562745169311057e-06,
+      "loss": 0.7967,
+      "step": 16918
+    },
+    {
+      "epoch": 0.9312014970554241,
+      "grad_norm": 0.6778925657272339,
+      "learning_rate": 5.562314454752207e-06,
+      "loss": 0.7567,
+      "step": 16919
+    },
+    {
+      "epoch": 0.9312565358577798,
+      "grad_norm": 0.7895017862319946,
+      "learning_rate": 5.561883735967151e-06,
+      "loss": 0.7321,
+      "step": 16920
+    },
+    {
+      "epoch": 0.9313115746601354,
+      "grad_norm": 0.7806753516197205,
+      "learning_rate": 5.561453012959128e-06,
+      "loss": 0.7469,
+      "step": 16921
+    },
+    {
+      "epoch": 0.931366613462491,
+      "grad_norm": 0.8205820918083191,
+      "learning_rate": 5.561022285731376e-06,
+      "loss": 0.7461,
+      "step": 16922
+    },
+    {
+      "epoch": 0.9314216522648467,
+      "grad_norm": 0.6211751103401184,
+      "learning_rate": 5.560591554287129e-06,
+      "loss": 0.668,
+      "step": 16923
+    },
+    {
+      "epoch": 0.9314766910672023,
+      "grad_norm": 0.7944413423538208,
+      "learning_rate": 5.5601608186296255e-06,
+      "loss": 0.7866,
+      "step": 16924
+    },
+    {
+      "epoch": 0.931531729869558,
+      "grad_norm": 0.7663695812225342,
+      "learning_rate": 5.559730078762103e-06,
+      "loss": 0.7402,
+      "step": 16925
+    },
+    {
+      "epoch": 0.9315867686719137,
+      "grad_norm": 0.6584487557411194,
+      "learning_rate": 5.559299334687801e-06,
+      "loss": 0.7251,
+      "step": 16926
+    },
+    {
+      "epoch": 0.9316418074742694,
+      "grad_norm": 0.740992546081543,
+      "learning_rate": 5.558868586409954e-06,
+      "loss": 0.7572,
+      "step": 16927
+    },
+    {
+      "epoch": 0.931696846276625,
+      "grad_norm": 0.7777884006500244,
+      "learning_rate": 5.558437833931801e-06,
+      "loss": 0.7817,
+      "step": 16928
+    },
+    {
+      "epoch": 0.9317518850789807,
+      "grad_norm": 0.6684641242027283,
+      "learning_rate": 5.5580070772565774e-06,
+      "loss": 0.7154,
+      "step": 16929
+    },
+    {
+      "epoch": 0.9318069238813363,
+      "grad_norm": 0.8053527474403381,
+      "learning_rate": 5.557576316387523e-06,
+      "loss": 0.7699,
+      "step": 16930
+    },
+    {
+      "epoch": 0.931861962683692,
+      "grad_norm": 0.6722555756568909,
+      "learning_rate": 5.557145551327875e-06,
+      "loss": 0.6554,
+      "step": 16931
+    },
+    {
+      "epoch": 0.9319170014860476,
+      "grad_norm": 0.6901096701622009,
+      "learning_rate": 5.556714782080867e-06,
+      "loss": 0.7889,
+      "step": 16932
+    },
+    {
+      "epoch": 0.9319720402884033,
+      "grad_norm": 0.6796066761016846,
+      "learning_rate": 5.556284008649743e-06,
+      "loss": 0.6839,
+      "step": 16933
+    },
+    {
+      "epoch": 0.932027079090759,
+      "grad_norm": 0.673685610294342,
+      "learning_rate": 5.555853231037735e-06,
+      "loss": 0.7677,
+      "step": 16934
+    },
+    {
+      "epoch": 0.9320821178931147,
+      "grad_norm": 0.6612395644187927,
+      "learning_rate": 5.555422449248085e-06,
+      "loss": 0.7566,
+      "step": 16935
+    },
+    {
+      "epoch": 0.9321371566954703,
+      "grad_norm": 0.6591265201568604,
+      "learning_rate": 5.554991663284026e-06,
+      "loss": 0.7503,
+      "step": 16936
+    },
+    {
+      "epoch": 0.932192195497826,
+      "grad_norm": 0.6197401285171509,
+      "learning_rate": 5.5545608731488e-06,
+      "loss": 0.7203,
+      "step": 16937
+    },
+    {
+      "epoch": 0.9322472343001816,
+      "grad_norm": 0.6715179681777954,
+      "learning_rate": 5.554130078845643e-06,
+      "loss": 0.7496,
+      "step": 16938
+    },
+    {
+      "epoch": 0.9323022731025373,
+      "grad_norm": 0.686224639415741,
+      "learning_rate": 5.553699280377793e-06,
+      "loss": 0.7466,
+      "step": 16939
+    },
+    {
+      "epoch": 0.9323573119048929,
+      "grad_norm": 0.7548978328704834,
+      "learning_rate": 5.553268477748487e-06,
+      "loss": 0.8279,
+      "step": 16940
+    },
+    {
+      "epoch": 0.9324123507072486,
+      "grad_norm": 0.5969768166542053,
+      "learning_rate": 5.552837670960962e-06,
+      "loss": 0.641,
+      "step": 16941
+    },
+    {
+      "epoch": 0.9324673895096043,
+      "grad_norm": 0.6729679703712463,
+      "learning_rate": 5.552406860018459e-06,
+      "loss": 0.718,
+      "step": 16942
+    },
+    {
+      "epoch": 0.93252242831196,
+      "grad_norm": 0.6350606679916382,
+      "learning_rate": 5.551976044924213e-06,
+      "loss": 0.7033,
+      "step": 16943
+    },
+    {
+      "epoch": 0.9325774671143156,
+      "grad_norm": 0.6564341187477112,
+      "learning_rate": 5.551545225681462e-06,
+      "loss": 0.6638,
+      "step": 16944
+    },
+    {
+      "epoch": 0.9326325059166712,
+      "grad_norm": 0.7151914238929749,
+      "learning_rate": 5.551114402293445e-06,
+      "loss": 0.8052,
+      "step": 16945
+    },
+    {
+      "epoch": 0.9326875447190269,
+      "grad_norm": 0.647455096244812,
+      "learning_rate": 5.5506835747634e-06,
+      "loss": 0.6903,
+      "step": 16946
+    },
+    {
+      "epoch": 0.9327425835213826,
+      "grad_norm": 0.7067738771438599,
+      "learning_rate": 5.5502527430945655e-06,
+      "loss": 0.7173,
+      "step": 16947
+    },
+    {
+      "epoch": 0.9327976223237382,
+      "grad_norm": 0.6301264762878418,
+      "learning_rate": 5.549821907290177e-06,
+      "loss": 0.7138,
+      "step": 16948
+    },
+    {
+      "epoch": 0.9328526611260939,
+      "grad_norm": 0.6322819590568542,
+      "learning_rate": 5.549391067353476e-06,
+      "loss": 0.7379,
+      "step": 16949
+    },
+    {
+      "epoch": 0.9329076999284496,
+      "grad_norm": 0.6628162860870361,
+      "learning_rate": 5.548960223287697e-06,
+      "loss": 0.7422,
+      "step": 16950
+    },
+    {
+      "epoch": 0.9329627387308053,
+      "grad_norm": 0.8131056427955627,
+      "learning_rate": 5.548529375096081e-06,
+      "loss": 0.808,
+      "step": 16951
+    },
+    {
+      "epoch": 0.9330177775331608,
+      "grad_norm": 0.6634977459907532,
+      "learning_rate": 5.548098522781866e-06,
+      "loss": 0.7963,
+      "step": 16952
+    },
+    {
+      "epoch": 0.9330728163355165,
+      "grad_norm": 0.706670343875885,
+      "learning_rate": 5.547667666348288e-06,
+      "loss": 0.7847,
+      "step": 16953
+    },
+    {
+      "epoch": 0.9331278551378722,
+      "grad_norm": 0.797380268573761,
+      "learning_rate": 5.547236805798587e-06,
+      "loss": 0.8007,
+      "step": 16954
+    },
+    {
+      "epoch": 0.9331828939402279,
+      "grad_norm": 0.6721376776695251,
+      "learning_rate": 5.546805941136e-06,
+      "loss": 0.7537,
+      "step": 16955
+    },
+    {
+      "epoch": 0.9332379327425835,
+      "grad_norm": 0.63679039478302,
+      "learning_rate": 5.546375072363765e-06,
+      "loss": 0.7198,
+      "step": 16956
+    },
+    {
+      "epoch": 0.9332929715449392,
+      "grad_norm": 0.6420717239379883,
+      "learning_rate": 5.5459441994851225e-06,
+      "loss": 0.7361,
+      "step": 16957
+    },
+    {
+      "epoch": 0.9333480103472949,
+      "grad_norm": 0.7549602389335632,
+      "learning_rate": 5.54551332250331e-06,
+      "loss": 0.7626,
+      "step": 16958
+    },
+    {
+      "epoch": 0.9334030491496506,
+      "grad_norm": 0.7319654226303101,
+      "learning_rate": 5.5450824414215656e-06,
+      "loss": 0.6942,
+      "step": 16959
+    },
+    {
+      "epoch": 0.9334580879520061,
+      "grad_norm": 0.7357600331306458,
+      "learning_rate": 5.544651556243124e-06,
+      "loss": 0.8573,
+      "step": 16960
+    },
+    {
+      "epoch": 0.9335131267543618,
+      "grad_norm": 0.6782594323158264,
+      "learning_rate": 5.5442206669712304e-06,
+      "loss": 0.7654,
+      "step": 16961
+    },
+    {
+      "epoch": 0.9335681655567175,
+      "grad_norm": 0.6621338129043579,
+      "learning_rate": 5.5437897736091194e-06,
+      "loss": 0.5882,
+      "step": 16962
+    },
+    {
+      "epoch": 0.9336232043590732,
+      "grad_norm": 0.6803955435752869,
+      "learning_rate": 5.543358876160031e-06,
+      "loss": 0.6574,
+      "step": 16963
+    },
+    {
+      "epoch": 0.9336782431614288,
+      "grad_norm": 0.6741437911987305,
+      "learning_rate": 5.542927974627202e-06,
+      "loss": 0.866,
+      "step": 16964
+    },
+    {
+      "epoch": 0.9337332819637845,
+      "grad_norm": 0.6904686093330383,
+      "learning_rate": 5.542497069013871e-06,
+      "loss": 0.7894,
+      "step": 16965
+    },
+    {
+      "epoch": 0.9337883207661402,
+      "grad_norm": 0.6671770215034485,
+      "learning_rate": 5.542066159323278e-06,
+      "loss": 0.7062,
+      "step": 16966
+    },
+    {
+      "epoch": 0.9338433595684957,
+      "grad_norm": 0.7025963664054871,
+      "learning_rate": 5.541635245558661e-06,
+      "loss": 0.7782,
+      "step": 16967
+    },
+    {
+      "epoch": 0.9338983983708514,
+      "grad_norm": 0.6948807239532471,
+      "learning_rate": 5.5412043277232566e-06,
+      "loss": 0.6972,
+      "step": 16968
+    },
+    {
+      "epoch": 0.9339534371732071,
+      "grad_norm": 0.6727368235588074,
+      "learning_rate": 5.540773405820307e-06,
+      "loss": 0.7303,
+      "step": 16969
+    },
+    {
+      "epoch": 0.9340084759755628,
+      "grad_norm": 0.7015120983123779,
+      "learning_rate": 5.540342479853048e-06,
+      "loss": 0.7417,
+      "step": 16970
+    },
+    {
+      "epoch": 0.9340635147779184,
+      "grad_norm": 0.6439093351364136,
+      "learning_rate": 5.539911549824722e-06,
+      "loss": 0.7465,
+      "step": 16971
+    },
+    {
+      "epoch": 0.9341185535802741,
+      "grad_norm": 0.6767433881759644,
+      "learning_rate": 5.539480615738562e-06,
+      "loss": 0.7867,
+      "step": 16972
+    },
+    {
+      "epoch": 0.9341735923826298,
+      "grad_norm": 0.6318436861038208,
+      "learning_rate": 5.539049677597812e-06,
+      "loss": 0.6693,
+      "step": 16973
+    },
+    {
+      "epoch": 0.9342286311849854,
+      "grad_norm": 0.604857861995697,
+      "learning_rate": 5.538618735405709e-06,
+      "loss": 0.6735,
+      "step": 16974
+    },
+    {
+      "epoch": 0.934283669987341,
+      "grad_norm": 0.7811105847358704,
+      "learning_rate": 5.538187789165491e-06,
+      "loss": 0.7997,
+      "step": 16975
+    },
+    {
+      "epoch": 0.9343387087896967,
+      "grad_norm": 0.644146203994751,
+      "learning_rate": 5.537756838880399e-06,
+      "loss": 0.7147,
+      "step": 16976
+    },
+    {
+      "epoch": 0.9343937475920524,
+      "grad_norm": 0.7074788808822632,
+      "learning_rate": 5.5373258845536705e-06,
+      "loss": 0.7487,
+      "step": 16977
+    },
+    {
+      "epoch": 0.9344487863944081,
+      "grad_norm": 0.597263753414154,
+      "learning_rate": 5.536894926188543e-06,
+      "loss": 0.6645,
+      "step": 16978
+    },
+    {
+      "epoch": 0.9345038251967637,
+      "grad_norm": 0.617344081401825,
+      "learning_rate": 5.5364639637882574e-06,
+      "loss": 0.6744,
+      "step": 16979
+    },
+    {
+      "epoch": 0.9345588639991194,
+      "grad_norm": 0.9057487845420837,
+      "learning_rate": 5.536032997356053e-06,
+      "loss": 0.8575,
+      "step": 16980
+    },
+    {
+      "epoch": 0.934613902801475,
+      "grad_norm": 0.6194222569465637,
+      "learning_rate": 5.535602026895167e-06,
+      "loss": 0.6823,
+      "step": 16981
+    },
+    {
+      "epoch": 0.9346689416038307,
+      "grad_norm": 0.7065734267234802,
+      "learning_rate": 5.535171052408841e-06,
+      "loss": 0.7394,
+      "step": 16982
+    },
+    {
+      "epoch": 0.9347239804061863,
+      "grad_norm": 0.6653978228569031,
+      "learning_rate": 5.534740073900311e-06,
+      "loss": 0.7709,
+      "step": 16983
+    },
+    {
+      "epoch": 0.934779019208542,
+      "grad_norm": 0.6902819275856018,
+      "learning_rate": 5.53430909137282e-06,
+      "loss": 0.7993,
+      "step": 16984
+    },
+    {
+      "epoch": 0.9348340580108977,
+      "grad_norm": 0.8288851976394653,
+      "learning_rate": 5.533878104829601e-06,
+      "loss": 0.6989,
+      "step": 16985
+    },
+    {
+      "epoch": 0.9348890968132534,
+      "grad_norm": 0.6423464417457581,
+      "learning_rate": 5.5334471142739e-06,
+      "loss": 0.6609,
+      "step": 16986
+    },
+    {
+      "epoch": 0.934944135615609,
+      "grad_norm": 0.7889821529388428,
+      "learning_rate": 5.533016119708953e-06,
+      "loss": 0.7466,
+      "step": 16987
+    },
+    {
+      "epoch": 0.9349991744179647,
+      "grad_norm": 0.6482957601547241,
+      "learning_rate": 5.5325851211379985e-06,
+      "loss": 0.7382,
+      "step": 16988
+    },
+    {
+      "epoch": 0.9350542132203203,
+      "grad_norm": 0.7158783078193665,
+      "learning_rate": 5.532154118564279e-06,
+      "loss": 0.7066,
+      "step": 16989
+    },
+    {
+      "epoch": 0.935109252022676,
+      "grad_norm": 0.708812952041626,
+      "learning_rate": 5.53172311199103e-06,
+      "loss": 0.7094,
+      "step": 16990
+    },
+    {
+      "epoch": 0.9351642908250316,
+      "grad_norm": 0.7140359282493591,
+      "learning_rate": 5.531292101421493e-06,
+      "loss": 0.7292,
+      "step": 16991
+    },
+    {
+      "epoch": 0.9352193296273873,
+      "grad_norm": 0.6114800572395325,
+      "learning_rate": 5.530861086858906e-06,
+      "loss": 0.6856,
+      "step": 16992
+    },
+    {
+      "epoch": 0.935274368429743,
+      "grad_norm": 0.6589435338973999,
+      "learning_rate": 5.53043006830651e-06,
+      "loss": 0.6854,
+      "step": 16993
+    },
+    {
+      "epoch": 0.9353294072320987,
+      "grad_norm": 0.6713253259658813,
+      "learning_rate": 5.5299990457675425e-06,
+      "loss": 0.7712,
+      "step": 16994
+    },
+    {
+      "epoch": 0.9353844460344543,
+      "grad_norm": 0.6345391273498535,
+      "learning_rate": 5.529568019245244e-06,
+      "loss": 0.6893,
+      "step": 16995
+    },
+    {
+      "epoch": 0.9354394848368099,
+      "grad_norm": 0.8601719737052917,
+      "learning_rate": 5.529136988742855e-06,
+      "loss": 0.7676,
+      "step": 16996
+    },
+    {
+      "epoch": 0.9354945236391656,
+      "grad_norm": 0.7791833281517029,
+      "learning_rate": 5.528705954263613e-06,
+      "loss": 0.7257,
+      "step": 16997
+    },
+    {
+      "epoch": 0.9355495624415213,
+      "grad_norm": 0.7072179317474365,
+      "learning_rate": 5.5282749158107594e-06,
+      "loss": 0.7292,
+      "step": 16998
+    },
+    {
+      "epoch": 0.9356046012438769,
+      "grad_norm": 0.762479841709137,
+      "learning_rate": 5.527843873387533e-06,
+      "loss": 0.8424,
+      "step": 16999
+    },
+    {
+      "epoch": 0.9356596400462326,
+      "grad_norm": 0.6964772939682007,
+      "learning_rate": 5.527412826997174e-06,
+      "loss": 0.6888,
+      "step": 17000
+    },
+    {
+      "epoch": 0.9357146788485883,
+      "grad_norm": 0.7196127772331238,
+      "learning_rate": 5.52698177664292e-06,
+      "loss": 0.6836,
+      "step": 17001
+    },
+    {
+      "epoch": 0.935769717650944,
+      "grad_norm": 0.6779214143753052,
+      "learning_rate": 5.526550722328012e-06,
+      "loss": 0.6476,
+      "step": 17002
+    },
+    {
+      "epoch": 0.9358247564532995,
+      "grad_norm": 0.7174015045166016,
+      "learning_rate": 5.526119664055691e-06,
+      "loss": 0.7115,
+      "step": 17003
+    },
+    {
+      "epoch": 0.9358797952556552,
+      "grad_norm": 0.6855511665344238,
+      "learning_rate": 5.525688601829195e-06,
+      "loss": 0.6881,
+      "step": 17004
+    },
+    {
+      "epoch": 0.9359348340580109,
+      "grad_norm": 0.716136634349823,
+      "learning_rate": 5.525257535651763e-06,
+      "loss": 0.7429,
+      "step": 17005
+    },
+    {
+      "epoch": 0.9359898728603666,
+      "grad_norm": 0.8645130395889282,
+      "learning_rate": 5.5248264655266374e-06,
+      "loss": 0.7413,
+      "step": 17006
+    },
+    {
+      "epoch": 0.9360449116627222,
+      "grad_norm": 0.6923744082450867,
+      "learning_rate": 5.5243953914570555e-06,
+      "loss": 0.6794,
+      "step": 17007
+    },
+    {
+      "epoch": 0.9360999504650779,
+      "grad_norm": 0.6838575601577759,
+      "learning_rate": 5.52396431344626e-06,
+      "loss": 0.8012,
+      "step": 17008
+    },
+    {
+      "epoch": 0.9361549892674336,
+      "grad_norm": 0.8331907987594604,
+      "learning_rate": 5.523533231497487e-06,
+      "loss": 0.8244,
+      "step": 17009
+    },
+    {
+      "epoch": 0.9362100280697891,
+      "grad_norm": 0.6853833794593811,
+      "learning_rate": 5.523102145613979e-06,
+      "loss": 0.7783,
+      "step": 17010
+    },
+    {
+      "epoch": 0.9362650668721448,
+      "grad_norm": 0.6622136831283569,
+      "learning_rate": 5.522671055798975e-06,
+      "loss": 0.7168,
+      "step": 17011
+    },
+    {
+      "epoch": 0.9363201056745005,
+      "grad_norm": 0.7320037484169006,
+      "learning_rate": 5.522239962055716e-06,
+      "loss": 0.7635,
+      "step": 17012
+    },
+    {
+      "epoch": 0.9363751444768562,
+      "grad_norm": 0.7859386801719666,
+      "learning_rate": 5.5218088643874415e-06,
+      "loss": 0.8936,
+      "step": 17013
+    },
+    {
+      "epoch": 0.9364301832792118,
+      "grad_norm": 0.7455195188522339,
+      "learning_rate": 5.521377762797391e-06,
+      "loss": 0.6984,
+      "step": 17014
+    },
+    {
+      "epoch": 0.9364852220815675,
+      "grad_norm": 0.6779624223709106,
+      "learning_rate": 5.520946657288804e-06,
+      "loss": 0.7338,
+      "step": 17015
+    },
+    {
+      "epoch": 0.9365402608839232,
+      "grad_norm": 0.6418783068656921,
+      "learning_rate": 5.520515547864922e-06,
+      "loss": 0.6768,
+      "step": 17016
+    },
+    {
+      "epoch": 0.9365952996862789,
+      "grad_norm": 0.8031061291694641,
+      "learning_rate": 5.520084434528984e-06,
+      "loss": 0.8285,
+      "step": 17017
+    },
+    {
+      "epoch": 0.9366503384886344,
+      "grad_norm": 0.638831377029419,
+      "learning_rate": 5.51965331728423e-06,
+      "loss": 0.6655,
+      "step": 17018
+    },
+    {
+      "epoch": 0.9367053772909901,
+      "grad_norm": 0.6730309724807739,
+      "learning_rate": 5.5192221961339024e-06,
+      "loss": 0.7191,
+      "step": 17019
+    },
+    {
+      "epoch": 0.9367604160933458,
+      "grad_norm": 0.7811747193336487,
+      "learning_rate": 5.51879107108124e-06,
+      "loss": 0.7116,
+      "step": 17020
+    },
+    {
+      "epoch": 0.9368154548957015,
+      "grad_norm": 0.6636793613433838,
+      "learning_rate": 5.51835994212948e-06,
+      "loss": 0.7178,
+      "step": 17021
+    },
+    {
+      "epoch": 0.9368704936980571,
+      "grad_norm": 0.6776058673858643,
+      "learning_rate": 5.517928809281867e-06,
+      "loss": 0.7542,
+      "step": 17022
+    },
+    {
+      "epoch": 0.9369255325004128,
+      "grad_norm": 0.6820049285888672,
+      "learning_rate": 5.5174976725416404e-06,
+      "loss": 0.7052,
+      "step": 17023
+    },
+    {
+      "epoch": 0.9369805713027685,
+      "grad_norm": 0.6562666296958923,
+      "learning_rate": 5.51706653191204e-06,
+      "loss": 0.7362,
+      "step": 17024
+    },
+    {
+      "epoch": 0.9370356101051242,
+      "grad_norm": 0.6575721502304077,
+      "learning_rate": 5.516635387396305e-06,
+      "loss": 0.7171,
+      "step": 17025
+    },
+    {
+      "epoch": 0.9370906489074797,
+      "grad_norm": 0.7353058457374573,
+      "learning_rate": 5.516204238997678e-06,
+      "loss": 0.7371,
+      "step": 17026
+    },
+    {
+      "epoch": 0.9371456877098354,
+      "grad_norm": 0.6880764961242676,
+      "learning_rate": 5.515773086719397e-06,
+      "loss": 0.7069,
+      "step": 17027
+    },
+    {
+      "epoch": 0.9372007265121911,
+      "grad_norm": 0.6665347814559937,
+      "learning_rate": 5.515341930564704e-06,
+      "loss": 0.7313,
+      "step": 17028
+    },
+    {
+      "epoch": 0.9372557653145468,
+      "grad_norm": 0.6450551152229309,
+      "learning_rate": 5.514910770536838e-06,
+      "loss": 0.6579,
+      "step": 17029
+    },
+    {
+      "epoch": 0.9373108041169024,
+      "grad_norm": 0.6448618769645691,
+      "learning_rate": 5.514479606639041e-06,
+      "loss": 0.6725,
+      "step": 17030
+    },
+    {
+      "epoch": 0.9373658429192581,
+      "grad_norm": 0.6275889873504639,
+      "learning_rate": 5.514048438874553e-06,
+      "loss": 0.7157,
+      "step": 17031
+    },
+    {
+      "epoch": 0.9374208817216138,
+      "grad_norm": 0.7542771100997925,
+      "learning_rate": 5.513617267246615e-06,
+      "loss": 0.8409,
+      "step": 17032
+    },
+    {
+      "epoch": 0.9374759205239694,
+      "grad_norm": 0.6698876619338989,
+      "learning_rate": 5.513186091758466e-06,
+      "loss": 0.7883,
+      "step": 17033
+    },
+    {
+      "epoch": 0.937530959326325,
+      "grad_norm": 0.6699510216712952,
+      "learning_rate": 5.512754912413349e-06,
+      "loss": 0.6719,
+      "step": 17034
+    },
+    {
+      "epoch": 0.9375859981286807,
+      "grad_norm": 0.6606268882751465,
+      "learning_rate": 5.512323729214502e-06,
+      "loss": 0.7626,
+      "step": 17035
+    },
+    {
+      "epoch": 0.9376410369310364,
+      "grad_norm": 0.6428219676017761,
+      "learning_rate": 5.511892542165169e-06,
+      "loss": 0.7135,
+      "step": 17036
+    },
+    {
+      "epoch": 0.9376960757333921,
+      "grad_norm": 0.6446773409843445,
+      "learning_rate": 5.511461351268588e-06,
+      "loss": 0.7064,
+      "step": 17037
+    },
+    {
+      "epoch": 0.9377511145357477,
+      "grad_norm": 0.730904757976532,
+      "learning_rate": 5.511030156528e-06,
+      "loss": 0.7543,
+      "step": 17038
+    },
+    {
+      "epoch": 0.9378061533381034,
+      "grad_norm": 0.6265544891357422,
+      "learning_rate": 5.5105989579466454e-06,
+      "loss": 0.648,
+      "step": 17039
+    },
+    {
+      "epoch": 0.937861192140459,
+      "grad_norm": 0.5883828401565552,
+      "learning_rate": 5.510167755527767e-06,
+      "loss": 0.6948,
+      "step": 17040
+    },
+    {
+      "epoch": 0.9379162309428147,
+      "grad_norm": 0.7379502654075623,
+      "learning_rate": 5.5097365492746024e-06,
+      "loss": 0.7034,
+      "step": 17041
+    },
+    {
+      "epoch": 0.9379712697451703,
+      "grad_norm": 0.7929210066795349,
+      "learning_rate": 5.509305339190396e-06,
+      "loss": 0.7117,
+      "step": 17042
+    },
+    {
+      "epoch": 0.938026308547526,
+      "grad_norm": 0.696765661239624,
+      "learning_rate": 5.5088741252783875e-06,
+      "loss": 0.7152,
+      "step": 17043
+    },
+    {
+      "epoch": 0.9380813473498817,
+      "grad_norm": 0.6686639785766602,
+      "learning_rate": 5.508442907541817e-06,
+      "loss": 0.7658,
+      "step": 17044
+    },
+    {
+      "epoch": 0.9381363861522374,
+      "grad_norm": 0.618567168712616,
+      "learning_rate": 5.508011685983923e-06,
+      "loss": 0.7133,
+      "step": 17045
+    },
+    {
+      "epoch": 0.938191424954593,
+      "grad_norm": 0.6887465119361877,
+      "learning_rate": 5.507580460607952e-06,
+      "loss": 0.7333,
+      "step": 17046
+    },
+    {
+      "epoch": 0.9382464637569486,
+      "grad_norm": 0.642488956451416,
+      "learning_rate": 5.5071492314171414e-06,
+      "loss": 0.7204,
+      "step": 17047
+    },
+    {
+      "epoch": 0.9383015025593043,
+      "grad_norm": 0.6381024718284607,
+      "learning_rate": 5.506717998414733e-06,
+      "loss": 0.7191,
+      "step": 17048
+    },
+    {
+      "epoch": 0.93835654136166,
+      "grad_norm": 0.7331018447875977,
+      "learning_rate": 5.506286761603968e-06,
+      "loss": 0.8042,
+      "step": 17049
+    },
+    {
+      "epoch": 0.9384115801640156,
+      "grad_norm": 0.636921763420105,
+      "learning_rate": 5.505855520988087e-06,
+      "loss": 0.6913,
+      "step": 17050
+    },
+    {
+      "epoch": 0.9384666189663713,
+      "grad_norm": 0.6855827569961548,
+      "learning_rate": 5.505424276570331e-06,
+      "loss": 0.8026,
+      "step": 17051
+    },
+    {
+      "epoch": 0.938521657768727,
+      "grad_norm": 0.6682537198066711,
+      "learning_rate": 5.504993028353942e-06,
+      "loss": 0.7942,
+      "step": 17052
+    },
+    {
+      "epoch": 0.9385766965710826,
+      "grad_norm": 0.6845211982727051,
+      "learning_rate": 5.50456177634216e-06,
+      "loss": 0.6827,
+      "step": 17053
+    },
+    {
+      "epoch": 0.9386317353734382,
+      "grad_norm": 0.7216134071350098,
+      "learning_rate": 5.504130520538227e-06,
+      "loss": 0.7557,
+      "step": 17054
+    },
+    {
+      "epoch": 0.9386867741757939,
+      "grad_norm": 0.6481857299804688,
+      "learning_rate": 5.503699260945384e-06,
+      "loss": 0.742,
+      "step": 17055
+    },
+    {
+      "epoch": 0.9387418129781496,
+      "grad_norm": 0.64216548204422,
+      "learning_rate": 5.503267997566873e-06,
+      "loss": 0.7219,
+      "step": 17056
+    },
+    {
+      "epoch": 0.9387968517805052,
+      "grad_norm": 0.7181951403617859,
+      "learning_rate": 5.502836730405931e-06,
+      "loss": 0.8311,
+      "step": 17057
+    },
+    {
+      "epoch": 0.9388518905828609,
+      "grad_norm": 0.841126024723053,
+      "learning_rate": 5.502405459465806e-06,
+      "loss": 0.7644,
+      "step": 17058
+    },
+    {
+      "epoch": 0.9389069293852166,
+      "grad_norm": 0.712687611579895,
+      "learning_rate": 5.5019741847497355e-06,
+      "loss": 0.5891,
+      "step": 17059
+    },
+    {
+      "epoch": 0.9389619681875723,
+      "grad_norm": 0.5507431030273438,
+      "learning_rate": 5.501542906260962e-06,
+      "loss": 0.6119,
+      "step": 17060
+    },
+    {
+      "epoch": 0.9390170069899278,
+      "grad_norm": 0.7180385589599609,
+      "learning_rate": 5.501111624002725e-06,
+      "loss": 0.7465,
+      "step": 17061
+    },
+    {
+      "epoch": 0.9390720457922835,
+      "grad_norm": 0.8892782330513,
+      "learning_rate": 5.500680337978268e-06,
+      "loss": 0.7639,
+      "step": 17062
+    },
+    {
+      "epoch": 0.9391270845946392,
+      "grad_norm": 0.7344280481338501,
+      "learning_rate": 5.500249048190832e-06,
+      "loss": 0.7659,
+      "step": 17063
+    },
+    {
+      "epoch": 0.9391821233969949,
+      "grad_norm": 0.6756631731987,
+      "learning_rate": 5.499817754643657e-06,
+      "loss": 0.6477,
+      "step": 17064
+    },
+    {
+      "epoch": 0.9392371621993505,
+      "grad_norm": 0.7020494341850281,
+      "learning_rate": 5.499386457339986e-06,
+      "loss": 0.6786,
+      "step": 17065
+    },
+    {
+      "epoch": 0.9392922010017062,
+      "grad_norm": 0.7595643997192383,
+      "learning_rate": 5.49895515628306e-06,
+      "loss": 0.7532,
+      "step": 17066
+    },
+    {
+      "epoch": 0.9393472398040619,
+      "grad_norm": 0.6780978441238403,
+      "learning_rate": 5.498523851476121e-06,
+      "loss": 0.6844,
+      "step": 17067
+    },
+    {
+      "epoch": 0.9394022786064176,
+      "grad_norm": 0.6908407807350159,
+      "learning_rate": 5.49809254292241e-06,
+      "loss": 0.6804,
+      "step": 17068
+    },
+    {
+      "epoch": 0.9394573174087731,
+      "grad_norm": 0.8891296982765198,
+      "learning_rate": 5.497661230625166e-06,
+      "loss": 0.7669,
+      "step": 17069
+    },
+    {
+      "epoch": 0.9395123562111288,
+      "grad_norm": 0.6702411770820618,
+      "learning_rate": 5.497229914587637e-06,
+      "loss": 0.7438,
+      "step": 17070
+    },
+    {
+      "epoch": 0.9395673950134845,
+      "grad_norm": 0.653181254863739,
+      "learning_rate": 5.4967985948130595e-06,
+      "loss": 0.7306,
+      "step": 17071
+    },
+    {
+      "epoch": 0.9396224338158402,
+      "grad_norm": 0.6638855934143066,
+      "learning_rate": 5.496367271304678e-06,
+      "loss": 0.7292,
+      "step": 17072
+    },
+    {
+      "epoch": 0.9396774726181958,
+      "grad_norm": 0.7425099015235901,
+      "learning_rate": 5.4959359440657324e-06,
+      "loss": 0.6555,
+      "step": 17073
+    },
+    {
+      "epoch": 0.9397325114205515,
+      "grad_norm": 0.6774802803993225,
+      "learning_rate": 5.495504613099465e-06,
+      "loss": 0.6969,
+      "step": 17074
+    },
+    {
+      "epoch": 0.9397875502229072,
+      "grad_norm": 0.7952350378036499,
+      "learning_rate": 5.495073278409118e-06,
+      "loss": 0.7323,
+      "step": 17075
+    },
+    {
+      "epoch": 0.9398425890252629,
+      "grad_norm": 0.7487592101097107,
+      "learning_rate": 5.494641939997932e-06,
+      "loss": 0.659,
+      "step": 17076
+    },
+    {
+      "epoch": 0.9398976278276184,
+      "grad_norm": 0.6290369033813477,
+      "learning_rate": 5.494210597869148e-06,
+      "loss": 0.5732,
+      "step": 17077
+    },
+    {
+      "epoch": 0.9399526666299741,
+      "grad_norm": 0.6333632469177246,
+      "learning_rate": 5.493779252026013e-06,
+      "loss": 0.7434,
+      "step": 17078
+    },
+    {
+      "epoch": 0.9400077054323298,
+      "grad_norm": 0.6158192157745361,
+      "learning_rate": 5.4933479024717625e-06,
+      "loss": 0.6443,
+      "step": 17079
+    },
+    {
+      "epoch": 0.9400627442346855,
+      "grad_norm": 0.639827311038971,
+      "learning_rate": 5.4929165492096425e-06,
+      "loss": 0.7668,
+      "step": 17080
+    },
+    {
+      "epoch": 0.9401177830370411,
+      "grad_norm": 0.6548694968223572,
+      "learning_rate": 5.492485192242893e-06,
+      "loss": 0.7439,
+      "step": 17081
+    },
+    {
+      "epoch": 0.9401728218393968,
+      "grad_norm": 0.7560970783233643,
+      "learning_rate": 5.492053831574756e-06,
+      "loss": 0.8069,
+      "step": 17082
+    },
+    {
+      "epoch": 0.9402278606417525,
+      "grad_norm": 0.792142391204834,
+      "learning_rate": 5.491622467208476e-06,
+      "loss": 0.7916,
+      "step": 17083
+    },
+    {
+      "epoch": 0.9402828994441081,
+      "grad_norm": 0.8506911993026733,
+      "learning_rate": 5.491191099147293e-06,
+      "loss": 0.6172,
+      "step": 17084
+    },
+    {
+      "epoch": 0.9403379382464637,
+      "grad_norm": 0.682349443435669,
+      "learning_rate": 5.490759727394448e-06,
+      "loss": 0.788,
+      "step": 17085
+    },
+    {
+      "epoch": 0.9403929770488194,
+      "grad_norm": 0.8406410813331604,
+      "learning_rate": 5.490328351953185e-06,
+      "loss": 0.7693,
+      "step": 17086
+    },
+    {
+      "epoch": 0.9404480158511751,
+      "grad_norm": 0.7017484903335571,
+      "learning_rate": 5.489896972826745e-06,
+      "loss": 0.8121,
+      "step": 17087
+    },
+    {
+      "epoch": 0.9405030546535308,
+      "grad_norm": 0.5880059599876404,
+      "learning_rate": 5.48946559001837e-06,
+      "loss": 0.6335,
+      "step": 17088
+    },
+    {
+      "epoch": 0.9405580934558864,
+      "grad_norm": 0.6772021651268005,
+      "learning_rate": 5.489034203531303e-06,
+      "loss": 0.7218,
+      "step": 17089
+    },
+    {
+      "epoch": 0.9406131322582421,
+      "grad_norm": 0.7308957576751709,
+      "learning_rate": 5.4886028133687865e-06,
+      "loss": 0.7566,
+      "step": 17090
+    },
+    {
+      "epoch": 0.9406681710605977,
+      "grad_norm": 0.5989949703216553,
+      "learning_rate": 5.488171419534061e-06,
+      "loss": 0.668,
+      "step": 17091
+    },
+    {
+      "epoch": 0.9407232098629534,
+      "grad_norm": 0.7060475945472717,
+      "learning_rate": 5.487740022030371e-06,
+      "loss": 0.7811,
+      "step": 17092
+    },
+    {
+      "epoch": 0.940778248665309,
+      "grad_norm": 0.7582124471664429,
+      "learning_rate": 5.487308620860956e-06,
+      "loss": 0.8345,
+      "step": 17093
+    },
+    {
+      "epoch": 0.9408332874676647,
+      "grad_norm": 0.67522794008255,
+      "learning_rate": 5.48687721602906e-06,
+      "loss": 0.6474,
+      "step": 17094
+    },
+    {
+      "epoch": 0.9408883262700204,
+      "grad_norm": 0.6196714639663696,
+      "learning_rate": 5.486445807537926e-06,
+      "loss": 0.6508,
+      "step": 17095
+    },
+    {
+      "epoch": 0.940943365072376,
+      "grad_norm": 0.7253809571266174,
+      "learning_rate": 5.486014395390796e-06,
+      "loss": 0.7439,
+      "step": 17096
+    },
+    {
+      "epoch": 0.9409984038747317,
+      "grad_norm": 0.7475833296775818,
+      "learning_rate": 5.485582979590912e-06,
+      "loss": 0.6686,
+      "step": 17097
+    },
+    {
+      "epoch": 0.9410534426770873,
+      "grad_norm": 0.598025918006897,
+      "learning_rate": 5.485151560141515e-06,
+      "loss": 0.6862,
+      "step": 17098
+    },
+    {
+      "epoch": 0.941108481479443,
+      "grad_norm": 0.644935131072998,
+      "learning_rate": 5.484720137045851e-06,
+      "loss": 0.6571,
+      "step": 17099
+    },
+    {
+      "epoch": 0.9411635202817986,
+      "grad_norm": 1.0013047456741333,
+      "learning_rate": 5.484288710307157e-06,
+      "loss": 0.7726,
+      "step": 17100
+    },
+    {
+      "epoch": 0.9412185590841543,
+      "grad_norm": 0.8886626958847046,
+      "learning_rate": 5.483857279928681e-06,
+      "loss": 0.8842,
+      "step": 17101
+    },
+    {
+      "epoch": 0.94127359788651,
+      "grad_norm": 0.6320830583572388,
+      "learning_rate": 5.483425845913662e-06,
+      "loss": 0.7534,
+      "step": 17102
+    },
+    {
+      "epoch": 0.9413286366888657,
+      "grad_norm": 0.684315025806427,
+      "learning_rate": 5.482994408265344e-06,
+      "loss": 0.777,
+      "step": 17103
+    },
+    {
+      "epoch": 0.9413836754912213,
+      "grad_norm": 0.6782847046852112,
+      "learning_rate": 5.48256296698697e-06,
+      "loss": 0.7018,
+      "step": 17104
+    },
+    {
+      "epoch": 0.941438714293577,
+      "grad_norm": 0.8408985137939453,
+      "learning_rate": 5.48213152208178e-06,
+      "loss": 0.8425,
+      "step": 17105
+    },
+    {
+      "epoch": 0.9414937530959326,
+      "grad_norm": 0.7615265846252441,
+      "learning_rate": 5.48170007355302e-06,
+      "loss": 0.7091,
+      "step": 17106
+    },
+    {
+      "epoch": 0.9415487918982883,
+      "grad_norm": 0.6782113313674927,
+      "learning_rate": 5.481268621403932e-06,
+      "loss": 0.7273,
+      "step": 17107
+    },
+    {
+      "epoch": 0.9416038307006439,
+      "grad_norm": 0.6413006782531738,
+      "learning_rate": 5.480837165637757e-06,
+      "loss": 0.6691,
+      "step": 17108
+    },
+    {
+      "epoch": 0.9416588695029996,
+      "grad_norm": 0.689999520778656,
+      "learning_rate": 5.480405706257739e-06,
+      "loss": 0.7615,
+      "step": 17109
+    },
+    {
+      "epoch": 0.9417139083053553,
+      "grad_norm": 0.7439923286437988,
+      "learning_rate": 5.47997424326712e-06,
+      "loss": 0.7569,
+      "step": 17110
+    },
+    {
+      "epoch": 0.941768947107711,
+      "grad_norm": 0.6936306953430176,
+      "learning_rate": 5.479542776669143e-06,
+      "loss": 0.713,
+      "step": 17111
+    },
+    {
+      "epoch": 0.9418239859100666,
+      "grad_norm": 1.0158600807189941,
+      "learning_rate": 5.479111306467051e-06,
+      "loss": 0.7874,
+      "step": 17112
+    },
+    {
+      "epoch": 0.9418790247124222,
+      "grad_norm": 0.7467468976974487,
+      "learning_rate": 5.478679832664087e-06,
+      "loss": 0.5757,
+      "step": 17113
+    },
+    {
+      "epoch": 0.9419340635147779,
+      "grad_norm": 0.6923582553863525,
+      "learning_rate": 5.478248355263492e-06,
+      "loss": 0.6983,
+      "step": 17114
+    },
+    {
+      "epoch": 0.9419891023171336,
+      "grad_norm": 0.6621877551078796,
+      "learning_rate": 5.4778168742685115e-06,
+      "loss": 0.7642,
+      "step": 17115
+    },
+    {
+      "epoch": 0.9420441411194892,
+      "grad_norm": 0.7422150373458862,
+      "learning_rate": 5.477385389682388e-06,
+      "loss": 0.8372,
+      "step": 17116
+    },
+    {
+      "epoch": 0.9420991799218449,
+      "grad_norm": 0.6306797862052917,
+      "learning_rate": 5.476953901508363e-06,
+      "loss": 0.698,
+      "step": 17117
+    },
+    {
+      "epoch": 0.9421542187242006,
+      "grad_norm": 0.6767764091491699,
+      "learning_rate": 5.476522409749679e-06,
+      "loss": 0.7251,
+      "step": 17118
+    },
+    {
+      "epoch": 0.9422092575265563,
+      "grad_norm": 0.86384117603302,
+      "learning_rate": 5.476090914409581e-06,
+      "loss": 0.7531,
+      "step": 17119
+    },
+    {
+      "epoch": 0.9422642963289118,
+      "grad_norm": 0.6221064329147339,
+      "learning_rate": 5.475659415491311e-06,
+      "loss": 0.735,
+      "step": 17120
+    },
+    {
+      "epoch": 0.9423193351312675,
+      "grad_norm": 0.6867205500602722,
+      "learning_rate": 5.475227912998113e-06,
+      "loss": 0.8036,
+      "step": 17121
+    },
+    {
+      "epoch": 0.9423743739336232,
+      "grad_norm": 0.6117410659790039,
+      "learning_rate": 5.474796406933229e-06,
+      "loss": 0.7011,
+      "step": 17122
+    },
+    {
+      "epoch": 0.9424294127359789,
+      "grad_norm": 0.7704444527626038,
+      "learning_rate": 5.474364897299902e-06,
+      "loss": 0.7836,
+      "step": 17123
+    },
+    {
+      "epoch": 0.9424844515383345,
+      "grad_norm": 0.7088763117790222,
+      "learning_rate": 5.473933384101376e-06,
+      "loss": 0.7613,
+      "step": 17124
+    },
+    {
+      "epoch": 0.9425394903406902,
+      "grad_norm": 0.7116881608963013,
+      "learning_rate": 5.473501867340892e-06,
+      "loss": 0.5805,
+      "step": 17125
+    },
+    {
+      "epoch": 0.9425945291430459,
+      "grad_norm": 0.7216660380363464,
+      "learning_rate": 5.473070347021695e-06,
+      "loss": 0.8896,
+      "step": 17126
+    },
+    {
+      "epoch": 0.9426495679454016,
+      "grad_norm": 0.6709206700325012,
+      "learning_rate": 5.4726388231470295e-06,
+      "loss": 0.7507,
+      "step": 17127
+    },
+    {
+      "epoch": 0.9427046067477571,
+      "grad_norm": 0.6911896467208862,
+      "learning_rate": 5.472207295720135e-06,
+      "loss": 0.6555,
+      "step": 17128
+    },
+    {
+      "epoch": 0.9427596455501128,
+      "grad_norm": 0.7614631652832031,
+      "learning_rate": 5.4717757647442595e-06,
+      "loss": 0.7568,
+      "step": 17129
+    },
+    {
+      "epoch": 0.9428146843524685,
+      "grad_norm": 0.6653177738189697,
+      "learning_rate": 5.47134423022264e-06,
+      "loss": 0.7626,
+      "step": 17130
+    },
+    {
+      "epoch": 0.9428697231548242,
+      "grad_norm": 0.7247204780578613,
+      "learning_rate": 5.470912692158527e-06,
+      "loss": 0.7038,
+      "step": 17131
+    },
+    {
+      "epoch": 0.9429247619571798,
+      "grad_norm": 0.7067399621009827,
+      "learning_rate": 5.470481150555158e-06,
+      "loss": 0.8066,
+      "step": 17132
+    },
+    {
+      "epoch": 0.9429798007595355,
+      "grad_norm": 0.6550384163856506,
+      "learning_rate": 5.47004960541578e-06,
+      "loss": 0.6447,
+      "step": 17133
+    },
+    {
+      "epoch": 0.9430348395618912,
+      "grad_norm": 0.6620935797691345,
+      "learning_rate": 5.469618056743635e-06,
+      "loss": 0.7816,
+      "step": 17134
+    },
+    {
+      "epoch": 0.9430898783642468,
+      "grad_norm": 0.6929067373275757,
+      "learning_rate": 5.469186504541966e-06,
+      "loss": 0.743,
+      "step": 17135
+    },
+    {
+      "epoch": 0.9431449171666024,
+      "grad_norm": 0.7585862278938293,
+      "learning_rate": 5.468754948814017e-06,
+      "loss": 0.7276,
+      "step": 17136
+    },
+    {
+      "epoch": 0.9431999559689581,
+      "grad_norm": 0.7555626034736633,
+      "learning_rate": 5.46832338956303e-06,
+      "loss": 0.7551,
+      "step": 17137
+    },
+    {
+      "epoch": 0.9432549947713138,
+      "grad_norm": 0.6756212115287781,
+      "learning_rate": 5.467891826792251e-06,
+      "loss": 0.6925,
+      "step": 17138
+    },
+    {
+      "epoch": 0.9433100335736694,
+      "grad_norm": 0.7939002513885498,
+      "learning_rate": 5.467460260504922e-06,
+      "loss": 0.7021,
+      "step": 17139
+    },
+    {
+      "epoch": 0.9433650723760251,
+      "grad_norm": 0.6600267887115479,
+      "learning_rate": 5.467028690704287e-06,
+      "loss": 0.7502,
+      "step": 17140
+    },
+    {
+      "epoch": 0.9434201111783808,
+      "grad_norm": 0.6323744058609009,
+      "learning_rate": 5.46659711739359e-06,
+      "loss": 0.6778,
+      "step": 17141
+    },
+    {
+      "epoch": 0.9434751499807364,
+      "grad_norm": 0.7724072933197021,
+      "learning_rate": 5.4661655405760716e-06,
+      "loss": 0.7722,
+      "step": 17142
+    },
+    {
+      "epoch": 0.943530188783092,
+      "grad_norm": 0.784764289855957,
+      "learning_rate": 5.465733960254979e-06,
+      "loss": 0.6578,
+      "step": 17143
+    },
+    {
+      "epoch": 0.9435852275854477,
+      "grad_norm": 0.6818285584449768,
+      "learning_rate": 5.465302376433556e-06,
+      "loss": 0.7913,
+      "step": 17144
+    },
+    {
+      "epoch": 0.9436402663878034,
+      "grad_norm": 0.6029468774795532,
+      "learning_rate": 5.464870789115043e-06,
+      "loss": 0.5936,
+      "step": 17145
+    },
+    {
+      "epoch": 0.9436953051901591,
+      "grad_norm": 0.7085309028625488,
+      "learning_rate": 5.464439198302687e-06,
+      "loss": 0.8497,
+      "step": 17146
+    },
+    {
+      "epoch": 0.9437503439925147,
+      "grad_norm": 0.6141021251678467,
+      "learning_rate": 5.46400760399973e-06,
+      "loss": 0.6056,
+      "step": 17147
+    },
+    {
+      "epoch": 0.9438053827948704,
+      "grad_norm": 0.6860122680664062,
+      "learning_rate": 5.4635760062094154e-06,
+      "loss": 0.8339,
+      "step": 17148
+    },
+    {
+      "epoch": 0.943860421597226,
+      "grad_norm": 0.8153133392333984,
+      "learning_rate": 5.463144404934989e-06,
+      "loss": 0.7799,
+      "step": 17149
+    },
+    {
+      "epoch": 0.9439154603995817,
+      "grad_norm": 0.7422170639038086,
+      "learning_rate": 5.462712800179691e-06,
+      "loss": 0.7581,
+      "step": 17150
+    },
+    {
+      "epoch": 0.9439704992019373,
+      "grad_norm": 0.6634333729743958,
+      "learning_rate": 5.462281191946769e-06,
+      "loss": 0.7292,
+      "step": 17151
+    },
+    {
+      "epoch": 0.944025538004293,
+      "grad_norm": 0.6939749717712402,
+      "learning_rate": 5.4618495802394645e-06,
+      "loss": 0.7568,
+      "step": 17152
+    },
+    {
+      "epoch": 0.9440805768066487,
+      "grad_norm": 0.6564925909042358,
+      "learning_rate": 5.461417965061023e-06,
+      "loss": 0.6955,
+      "step": 17153
+    },
+    {
+      "epoch": 0.9441356156090044,
+      "grad_norm": 0.7528273463249207,
+      "learning_rate": 5.460986346414686e-06,
+      "loss": 0.8456,
+      "step": 17154
+    },
+    {
+      "epoch": 0.94419065441136,
+      "grad_norm": 0.6330366730690002,
+      "learning_rate": 5.4605547243037015e-06,
+      "loss": 0.6921,
+      "step": 17155
+    },
+    {
+      "epoch": 0.9442456932137157,
+      "grad_norm": 0.8115606307983398,
+      "learning_rate": 5.460123098731309e-06,
+      "loss": 0.8116,
+      "step": 17156
+    },
+    {
+      "epoch": 0.9443007320160713,
+      "grad_norm": 0.6605154275894165,
+      "learning_rate": 5.459691469700756e-06,
+      "loss": 0.7205,
+      "step": 17157
+    },
+    {
+      "epoch": 0.944355770818427,
+      "grad_norm": 0.6622677445411682,
+      "learning_rate": 5.459259837215284e-06,
+      "loss": 0.7324,
+      "step": 17158
+    },
+    {
+      "epoch": 0.9444108096207826,
+      "grad_norm": 0.6821356415748596,
+      "learning_rate": 5.458828201278139e-06,
+      "loss": 0.7984,
+      "step": 17159
+    },
+    {
+      "epoch": 0.9444658484231383,
+      "grad_norm": 0.6941211223602295,
+      "learning_rate": 5.4583965618925625e-06,
+      "loss": 0.7305,
+      "step": 17160
+    },
+    {
+      "epoch": 0.944520887225494,
+      "grad_norm": 0.7761551141738892,
+      "learning_rate": 5.4579649190618025e-06,
+      "loss": 0.7898,
+      "step": 17161
+    },
+    {
+      "epoch": 0.9445759260278497,
+      "grad_norm": 0.6097804307937622,
+      "learning_rate": 5.457533272789097e-06,
+      "loss": 0.7151,
+      "step": 17162
+    },
+    {
+      "epoch": 0.9446309648302053,
+      "grad_norm": 0.6361019611358643,
+      "learning_rate": 5.457101623077696e-06,
+      "loss": 0.7037,
+      "step": 17163
+    },
+    {
+      "epoch": 0.9446860036325609,
+      "grad_norm": 0.638758659362793,
+      "learning_rate": 5.456669969930842e-06,
+      "loss": 0.6714,
+      "step": 17164
+    },
+    {
+      "epoch": 0.9447410424349166,
+      "grad_norm": 0.6280048489570618,
+      "learning_rate": 5.4562383133517795e-06,
+      "loss": 0.6839,
+      "step": 17165
+    },
+    {
+      "epoch": 0.9447960812372723,
+      "grad_norm": 0.6959577798843384,
+      "learning_rate": 5.4558066533437495e-06,
+      "loss": 0.7493,
+      "step": 17166
+    },
+    {
+      "epoch": 0.9448511200396279,
+      "grad_norm": 0.7218848466873169,
+      "learning_rate": 5.455374989910001e-06,
+      "loss": 0.8123,
+      "step": 17167
+    },
+    {
+      "epoch": 0.9449061588419836,
+      "grad_norm": 0.6597707271575928,
+      "learning_rate": 5.454943323053774e-06,
+      "loss": 0.7575,
+      "step": 17168
+    },
+    {
+      "epoch": 0.9449611976443393,
+      "grad_norm": 0.6173719763755798,
+      "learning_rate": 5.454511652778316e-06,
+      "loss": 0.6715,
+      "step": 17169
+    },
+    {
+      "epoch": 0.945016236446695,
+      "grad_norm": 0.7458024024963379,
+      "learning_rate": 5.45407997908687e-06,
+      "loss": 0.7526,
+      "step": 17170
+    },
+    {
+      "epoch": 0.9450712752490505,
+      "grad_norm": 0.7513238787651062,
+      "learning_rate": 5.4536483019826806e-06,
+      "loss": 0.7001,
+      "step": 17171
+    },
+    {
+      "epoch": 0.9451263140514062,
+      "grad_norm": 0.6999634504318237,
+      "learning_rate": 5.453216621468992e-06,
+      "loss": 0.7665,
+      "step": 17172
+    },
+    {
+      "epoch": 0.9451813528537619,
+      "grad_norm": 0.6689110994338989,
+      "learning_rate": 5.452784937549048e-06,
+      "loss": 0.7758,
+      "step": 17173
+    },
+    {
+      "epoch": 0.9452363916561176,
+      "grad_norm": 0.61881422996521,
+      "learning_rate": 5.4523532502260935e-06,
+      "loss": 0.6624,
+      "step": 17174
+    },
+    {
+      "epoch": 0.9452914304584732,
+      "grad_norm": 0.748910665512085,
+      "learning_rate": 5.451921559503373e-06,
+      "loss": 0.8298,
+      "step": 17175
+    },
+    {
+      "epoch": 0.9453464692608289,
+      "grad_norm": 0.7247006893157959,
+      "learning_rate": 5.451489865384132e-06,
+      "loss": 0.7738,
+      "step": 17176
+    },
+    {
+      "epoch": 0.9454015080631846,
+      "grad_norm": 0.6910108923912048,
+      "learning_rate": 5.451058167871614e-06,
+      "loss": 0.6645,
+      "step": 17177
+    },
+    {
+      "epoch": 0.9454565468655403,
+      "grad_norm": 0.6405494213104248,
+      "learning_rate": 5.450626466969061e-06,
+      "loss": 0.779,
+      "step": 17178
+    },
+    {
+      "epoch": 0.9455115856678958,
+      "grad_norm": 0.6681149005889893,
+      "learning_rate": 5.450194762679722e-06,
+      "loss": 0.6453,
+      "step": 17179
+    },
+    {
+      "epoch": 0.9455666244702515,
+      "grad_norm": 0.7538934350013733,
+      "learning_rate": 5.44976305500684e-06,
+      "loss": 0.8169,
+      "step": 17180
+    },
+    {
+      "epoch": 0.9456216632726072,
+      "grad_norm": 0.7700850367546082,
+      "learning_rate": 5.4493313439536595e-06,
+      "loss": 0.7508,
+      "step": 17181
+    },
+    {
+      "epoch": 0.9456767020749628,
+      "grad_norm": 0.7019801735877991,
+      "learning_rate": 5.4488996295234256e-06,
+      "loss": 0.7687,
+      "step": 17182
+    },
+    {
+      "epoch": 0.9457317408773185,
+      "grad_norm": 0.6191794276237488,
+      "learning_rate": 5.448467911719381e-06,
+      "loss": 0.6303,
+      "step": 17183
+    },
+    {
+      "epoch": 0.9457867796796742,
+      "grad_norm": 0.738051176071167,
+      "learning_rate": 5.448036190544772e-06,
+      "loss": 0.7625,
+      "step": 17184
+    },
+    {
+      "epoch": 0.9458418184820299,
+      "grad_norm": 0.8286527991294861,
+      "learning_rate": 5.447604466002843e-06,
+      "loss": 0.7328,
+      "step": 17185
+    },
+    {
+      "epoch": 0.9458968572843854,
+      "grad_norm": 0.74991375207901,
+      "learning_rate": 5.447172738096837e-06,
+      "loss": 0.7674,
+      "step": 17186
+    },
+    {
+      "epoch": 0.9459518960867411,
+      "grad_norm": 1.20223069190979,
+      "learning_rate": 5.446741006830002e-06,
+      "loss": 0.8024,
+      "step": 17187
+    },
+    {
+      "epoch": 0.9460069348890968,
+      "grad_norm": 0.6366617679595947,
+      "learning_rate": 5.446309272205581e-06,
+      "loss": 0.702,
+      "step": 17188
+    },
+    {
+      "epoch": 0.9460619736914525,
+      "grad_norm": 0.8702724575996399,
+      "learning_rate": 5.445877534226819e-06,
+      "loss": 0.8237,
+      "step": 17189
+    },
+    {
+      "epoch": 0.9461170124938081,
+      "grad_norm": 0.6849148869514465,
+      "learning_rate": 5.445445792896959e-06,
+      "loss": 0.7883,
+      "step": 17190
+    },
+    {
+      "epoch": 0.9461720512961638,
+      "grad_norm": 0.7968994379043579,
+      "learning_rate": 5.445014048219251e-06,
+      "loss": 0.8019,
+      "step": 17191
+    },
+    {
+      "epoch": 0.9462270900985195,
+      "grad_norm": 0.6820282936096191,
+      "learning_rate": 5.444582300196934e-06,
+      "loss": 0.7058,
+      "step": 17192
+    },
+    {
+      "epoch": 0.9462821289008752,
+      "grad_norm": 0.8677741885185242,
+      "learning_rate": 5.444150548833257e-06,
+      "loss": 0.7753,
+      "step": 17193
+    },
+    {
+      "epoch": 0.9463371677032307,
+      "grad_norm": 0.6906048059463501,
+      "learning_rate": 5.443718794131463e-06,
+      "loss": 0.7275,
+      "step": 17194
+    },
+    {
+      "epoch": 0.9463922065055864,
+      "grad_norm": 0.6943230032920837,
+      "learning_rate": 5.4432870360947966e-06,
+      "loss": 0.7339,
+      "step": 17195
+    },
+    {
+      "epoch": 0.9464472453079421,
+      "grad_norm": 0.9276923537254333,
+      "learning_rate": 5.442855274726504e-06,
+      "loss": 0.8316,
+      "step": 17196
+    },
+    {
+      "epoch": 0.9465022841102978,
+      "grad_norm": 0.6322672963142395,
+      "learning_rate": 5.44242351002983e-06,
+      "loss": 0.7104,
+      "step": 17197
+    },
+    {
+      "epoch": 0.9465573229126534,
+      "grad_norm": 0.6659188270568848,
+      "learning_rate": 5.4419917420080185e-06,
+      "loss": 0.7575,
+      "step": 17198
+    },
+    {
+      "epoch": 0.9466123617150091,
+      "grad_norm": 0.6726362109184265,
+      "learning_rate": 5.441559970664315e-06,
+      "loss": 0.8168,
+      "step": 17199
+    },
+    {
+      "epoch": 0.9466674005173648,
+      "grad_norm": 0.6305935978889465,
+      "learning_rate": 5.441128196001966e-06,
+      "loss": 0.7447,
+      "step": 17200
+    },
+    {
+      "epoch": 0.9467224393197204,
+      "grad_norm": 0.6968098282814026,
+      "learning_rate": 5.4406964180242164e-06,
+      "loss": 0.7592,
+      "step": 17201
+    },
+    {
+      "epoch": 0.946777478122076,
+      "grad_norm": 0.8273503184318542,
+      "learning_rate": 5.440264636734308e-06,
+      "loss": 0.79,
+      "step": 17202
+    },
+    {
+      "epoch": 0.9468325169244317,
+      "grad_norm": 0.6939367055892944,
+      "learning_rate": 5.439832852135489e-06,
+      "loss": 0.7585,
+      "step": 17203
+    },
+    {
+      "epoch": 0.9468875557267874,
+      "grad_norm": 0.7276638746261597,
+      "learning_rate": 5.439401064231006e-06,
+      "loss": 0.8132,
+      "step": 17204
+    },
+    {
+      "epoch": 0.9469425945291431,
+      "grad_norm": 0.6231852769851685,
+      "learning_rate": 5.438969273024102e-06,
+      "loss": 0.7087,
+      "step": 17205
+    },
+    {
+      "epoch": 0.9469976333314987,
+      "grad_norm": 0.6458756923675537,
+      "learning_rate": 5.438537478518021e-06,
+      "loss": 0.6894,
+      "step": 17206
+    },
+    {
+      "epoch": 0.9470526721338544,
+      "grad_norm": 0.7376695871353149,
+      "learning_rate": 5.4381056807160104e-06,
+      "loss": 0.7008,
+      "step": 17207
+    },
+    {
+      "epoch": 0.94710771093621,
+      "grad_norm": 0.6636282205581665,
+      "learning_rate": 5.437673879621315e-06,
+      "loss": 0.7216,
+      "step": 17208
+    },
+    {
+      "epoch": 0.9471627497385657,
+      "grad_norm": 0.6254339814186096,
+      "learning_rate": 5.437242075237179e-06,
+      "loss": 0.7228,
+      "step": 17209
+    },
+    {
+      "epoch": 0.9472177885409213,
+      "grad_norm": 0.6461647748947144,
+      "learning_rate": 5.436810267566849e-06,
+      "loss": 0.6799,
+      "step": 17210
+    },
+    {
+      "epoch": 0.947272827343277,
+      "grad_norm": 0.6700971722602844,
+      "learning_rate": 5.4363784566135694e-06,
+      "loss": 0.8166,
+      "step": 17211
+    },
+    {
+      "epoch": 0.9473278661456327,
+      "grad_norm": 0.6764943599700928,
+      "learning_rate": 5.4359466423805875e-06,
+      "loss": 0.7565,
+      "step": 17212
+    },
+    {
+      "epoch": 0.9473829049479884,
+      "grad_norm": 0.6449782848358154,
+      "learning_rate": 5.4355148248711466e-06,
+      "loss": 0.6943,
+      "step": 17213
+    },
+    {
+      "epoch": 0.947437943750344,
+      "grad_norm": 0.7249940037727356,
+      "learning_rate": 5.435083004088492e-06,
+      "loss": 0.677,
+      "step": 17214
+    },
+    {
+      "epoch": 0.9474929825526996,
+      "grad_norm": 0.6520692706108093,
+      "learning_rate": 5.43465118003587e-06,
+      "loss": 0.8089,
+      "step": 17215
+    },
+    {
+      "epoch": 0.9475480213550553,
+      "grad_norm": 0.826518714427948,
+      "learning_rate": 5.4342193527165275e-06,
+      "loss": 0.6715,
+      "step": 17216
+    },
+    {
+      "epoch": 0.947603060157411,
+      "grad_norm": 0.6832894682884216,
+      "learning_rate": 5.433787522133709e-06,
+      "loss": 0.7615,
+      "step": 17217
+    },
+    {
+      "epoch": 0.9476580989597666,
+      "grad_norm": 0.6764441132545471,
+      "learning_rate": 5.433355688290658e-06,
+      "loss": 0.8082,
+      "step": 17218
+    },
+    {
+      "epoch": 0.9477131377621223,
+      "grad_norm": 0.8981256484985352,
+      "learning_rate": 5.4329238511906226e-06,
+      "loss": 0.7663,
+      "step": 17219
+    },
+    {
+      "epoch": 0.947768176564478,
+      "grad_norm": 0.6509552001953125,
+      "learning_rate": 5.432492010836847e-06,
+      "loss": 0.7375,
+      "step": 17220
+    },
+    {
+      "epoch": 0.9478232153668337,
+      "grad_norm": 0.7138388156890869,
+      "learning_rate": 5.4320601672325775e-06,
+      "loss": 0.8096,
+      "step": 17221
+    },
+    {
+      "epoch": 0.9478782541691892,
+      "grad_norm": 1.0833384990692139,
+      "learning_rate": 5.431628320381058e-06,
+      "loss": 0.7985,
+      "step": 17222
+    },
+    {
+      "epoch": 0.9479332929715449,
+      "grad_norm": 0.685681164264679,
+      "learning_rate": 5.431196470285538e-06,
+      "loss": 0.759,
+      "step": 17223
+    },
+    {
+      "epoch": 0.9479883317739006,
+      "grad_norm": 0.6390333771705627,
+      "learning_rate": 5.430764616949259e-06,
+      "loss": 0.7174,
+      "step": 17224
+    },
+    {
+      "epoch": 0.9480433705762562,
+      "grad_norm": 0.6927287578582764,
+      "learning_rate": 5.43033276037547e-06,
+      "loss": 0.7348,
+      "step": 17225
+    },
+    {
+      "epoch": 0.9480984093786119,
+      "grad_norm": 0.6661121845245361,
+      "learning_rate": 5.429900900567413e-06,
+      "loss": 0.7042,
+      "step": 17226
+    },
+    {
+      "epoch": 0.9481534481809676,
+      "grad_norm": 0.6327020525932312,
+      "learning_rate": 5.429469037528337e-06,
+      "loss": 0.6995,
+      "step": 17227
+    },
+    {
+      "epoch": 0.9482084869833233,
+      "grad_norm": 0.858727216720581,
+      "learning_rate": 5.429037171261488e-06,
+      "loss": 0.7893,
+      "step": 17228
+    },
+    {
+      "epoch": 0.9482635257856789,
+      "grad_norm": 0.6609689593315125,
+      "learning_rate": 5.428605301770109e-06,
+      "loss": 0.7258,
+      "step": 17229
+    },
+    {
+      "epoch": 0.9483185645880345,
+      "grad_norm": 0.811647355556488,
+      "learning_rate": 5.4281734290574475e-06,
+      "loss": 0.7341,
+      "step": 17230
+    },
+    {
+      "epoch": 0.9483736033903902,
+      "grad_norm": 0.5738475322723389,
+      "learning_rate": 5.427741553126748e-06,
+      "loss": 0.6227,
+      "step": 17231
+    },
+    {
+      "epoch": 0.9484286421927459,
+      "grad_norm": 0.6369803547859192,
+      "learning_rate": 5.42730967398126e-06,
+      "loss": 0.7188,
+      "step": 17232
+    },
+    {
+      "epoch": 0.9484836809951015,
+      "grad_norm": 0.6879351735115051,
+      "learning_rate": 5.426877791624225e-06,
+      "loss": 0.6031,
+      "step": 17233
+    },
+    {
+      "epoch": 0.9485387197974572,
+      "grad_norm": 0.872032880783081,
+      "learning_rate": 5.4264459060588905e-06,
+      "loss": 0.7003,
+      "step": 17234
+    },
+    {
+      "epoch": 0.9485937585998129,
+      "grad_norm": 0.6781120300292969,
+      "learning_rate": 5.426014017288504e-06,
+      "loss": 0.7876,
+      "step": 17235
+    },
+    {
+      "epoch": 0.9486487974021686,
+      "grad_norm": 1.026384711265564,
+      "learning_rate": 5.42558212531631e-06,
+      "loss": 0.709,
+      "step": 17236
+    },
+    {
+      "epoch": 0.9487038362045241,
+      "grad_norm": 0.6840795874595642,
+      "learning_rate": 5.4251502301455536e-06,
+      "loss": 0.7513,
+      "step": 17237
+    },
+    {
+      "epoch": 0.9487588750068798,
+      "grad_norm": 0.7462189793586731,
+      "learning_rate": 5.4247183317794825e-06,
+      "loss": 0.6966,
+      "step": 17238
+    },
+    {
+      "epoch": 0.9488139138092355,
+      "grad_norm": 0.7512027025222778,
+      "learning_rate": 5.424286430221342e-06,
+      "loss": 0.8355,
+      "step": 17239
+    },
+    {
+      "epoch": 0.9488689526115912,
+      "grad_norm": 0.8257582187652588,
+      "learning_rate": 5.423854525474379e-06,
+      "loss": 0.7876,
+      "step": 17240
+    },
+    {
+      "epoch": 0.9489239914139468,
+      "grad_norm": 0.6845347881317139,
+      "learning_rate": 5.423422617541838e-06,
+      "loss": 0.6217,
+      "step": 17241
+    },
+    {
+      "epoch": 0.9489790302163025,
+      "grad_norm": 0.7380620837211609,
+      "learning_rate": 5.422990706426967e-06,
+      "loss": 0.7257,
+      "step": 17242
+    },
+    {
+      "epoch": 0.9490340690186582,
+      "grad_norm": 0.703009307384491,
+      "learning_rate": 5.42255879213301e-06,
+      "loss": 0.7212,
+      "step": 17243
+    },
+    {
+      "epoch": 0.9490891078210139,
+      "grad_norm": 0.5682654976844788,
+      "learning_rate": 5.422126874663215e-06,
+      "loss": 0.6095,
+      "step": 17244
+    },
+    {
+      "epoch": 0.9491441466233694,
+      "grad_norm": 0.7798917293548584,
+      "learning_rate": 5.421694954020827e-06,
+      "loss": 0.8202,
+      "step": 17245
+    },
+    {
+      "epoch": 0.9491991854257251,
+      "grad_norm": 0.9012196063995361,
+      "learning_rate": 5.421263030209092e-06,
+      "loss": 0.7277,
+      "step": 17246
+    },
+    {
+      "epoch": 0.9492542242280808,
+      "grad_norm": 0.7454085350036621,
+      "learning_rate": 5.4208311032312575e-06,
+      "loss": 0.709,
+      "step": 17247
+    },
+    {
+      "epoch": 0.9493092630304365,
+      "grad_norm": 0.6721041798591614,
+      "learning_rate": 5.420399173090569e-06,
+      "loss": 0.7932,
+      "step": 17248
+    },
+    {
+      "epoch": 0.9493643018327921,
+      "grad_norm": 0.6498880386352539,
+      "learning_rate": 5.419967239790273e-06,
+      "loss": 0.6983,
+      "step": 17249
+    },
+    {
+      "epoch": 0.9494193406351478,
+      "grad_norm": 0.6301897764205933,
+      "learning_rate": 5.419535303333615e-06,
+      "loss": 0.6722,
+      "step": 17250
+    },
+    {
+      "epoch": 0.9494743794375035,
+      "grad_norm": 0.7664110660552979,
+      "learning_rate": 5.419103363723843e-06,
+      "loss": 0.7823,
+      "step": 17251
+    },
+    {
+      "epoch": 0.9495294182398591,
+      "grad_norm": 0.6919211149215698,
+      "learning_rate": 5.4186714209642026e-06,
+      "loss": 0.7807,
+      "step": 17252
+    },
+    {
+      "epoch": 0.9495844570422147,
+      "grad_norm": 0.6624665260314941,
+      "learning_rate": 5.418239475057938e-06,
+      "loss": 0.7101,
+      "step": 17253
+    },
+    {
+      "epoch": 0.9496394958445704,
+      "grad_norm": 0.6163734197616577,
+      "learning_rate": 5.417807526008299e-06,
+      "loss": 0.6766,
+      "step": 17254
+    },
+    {
+      "epoch": 0.9496945346469261,
+      "grad_norm": 0.6248084306716919,
+      "learning_rate": 5.41737557381853e-06,
+      "loss": 0.7704,
+      "step": 17255
+    },
+    {
+      "epoch": 0.9497495734492818,
+      "grad_norm": 0.6190524101257324,
+      "learning_rate": 5.416943618491878e-06,
+      "loss": 0.7616,
+      "step": 17256
+    },
+    {
+      "epoch": 0.9498046122516374,
+      "grad_norm": 0.7721768617630005,
+      "learning_rate": 5.416511660031589e-06,
+      "loss": 0.7005,
+      "step": 17257
+    },
+    {
+      "epoch": 0.9498596510539931,
+      "grad_norm": 0.6752305030822754,
+      "learning_rate": 5.4160796984409115e-06,
+      "loss": 0.7614,
+      "step": 17258
+    },
+    {
+      "epoch": 0.9499146898563487,
+      "grad_norm": 0.7479132413864136,
+      "learning_rate": 5.415647733723087e-06,
+      "loss": 0.73,
+      "step": 17259
+    },
+    {
+      "epoch": 0.9499697286587044,
+      "grad_norm": 0.6660789847373962,
+      "learning_rate": 5.415215765881367e-06,
+      "loss": 0.6687,
+      "step": 17260
+    },
+    {
+      "epoch": 0.95002476746106,
+      "grad_norm": 0.6333804130554199,
+      "learning_rate": 5.414783794918998e-06,
+      "loss": 0.6641,
+      "step": 17261
+    },
+    {
+      "epoch": 0.9500798062634157,
+      "grad_norm": 0.7207759618759155,
+      "learning_rate": 5.4143518208392245e-06,
+      "loss": 0.7361,
+      "step": 17262
+    },
+    {
+      "epoch": 0.9501348450657714,
+      "grad_norm": 0.9466403722763062,
+      "learning_rate": 5.413919843645291e-06,
+      "loss": 0.7898,
+      "step": 17263
+    },
+    {
+      "epoch": 0.9501898838681271,
+      "grad_norm": 0.7350644469261169,
+      "learning_rate": 5.41348786334045e-06,
+      "loss": 0.7646,
+      "step": 17264
+    },
+    {
+      "epoch": 0.9502449226704827,
+      "grad_norm": 0.7812394499778748,
+      "learning_rate": 5.413055879927943e-06,
+      "loss": 0.807,
+      "step": 17265
+    },
+    {
+      "epoch": 0.9502999614728384,
+      "grad_norm": 0.6254860758781433,
+      "learning_rate": 5.41262389341102e-06,
+      "loss": 0.6722,
+      "step": 17266
+    },
+    {
+      "epoch": 0.950355000275194,
+      "grad_norm": 0.6452184319496155,
+      "learning_rate": 5.412191903792925e-06,
+      "loss": 0.7019,
+      "step": 17267
+    },
+    {
+      "epoch": 0.9504100390775496,
+      "grad_norm": 0.6686402559280396,
+      "learning_rate": 5.411759911076907e-06,
+      "loss": 0.7344,
+      "step": 17268
+    },
+    {
+      "epoch": 0.9504650778799053,
+      "grad_norm": 0.6803160309791565,
+      "learning_rate": 5.411327915266211e-06,
+      "loss": 0.7334,
+      "step": 17269
+    },
+    {
+      "epoch": 0.950520116682261,
+      "grad_norm": 0.6542620658874512,
+      "learning_rate": 5.410895916364085e-06,
+      "loss": 0.7176,
+      "step": 17270
+    },
+    {
+      "epoch": 0.9505751554846167,
+      "grad_norm": 0.6867213845252991,
+      "learning_rate": 5.410463914373775e-06,
+      "loss": 0.6591,
+      "step": 17271
+    },
+    {
+      "epoch": 0.9506301942869723,
+      "grad_norm": 0.7018343210220337,
+      "learning_rate": 5.410031909298528e-06,
+      "loss": 0.8,
+      "step": 17272
+    },
+    {
+      "epoch": 0.950685233089328,
+      "grad_norm": 0.6558470129966736,
+      "learning_rate": 5.409599901141591e-06,
+      "loss": 0.728,
+      "step": 17273
+    },
+    {
+      "epoch": 0.9507402718916836,
+      "grad_norm": 0.7012515068054199,
+      "learning_rate": 5.409167889906211e-06,
+      "loss": 0.7611,
+      "step": 17274
+    },
+    {
+      "epoch": 0.9507953106940393,
+      "grad_norm": 0.661029040813446,
+      "learning_rate": 5.408735875595633e-06,
+      "loss": 0.7216,
+      "step": 17275
+    },
+    {
+      "epoch": 0.9508503494963949,
+      "grad_norm": 0.6965559124946594,
+      "learning_rate": 5.408303858213107e-06,
+      "loss": 0.813,
+      "step": 17276
+    },
+    {
+      "epoch": 0.9509053882987506,
+      "grad_norm": 0.8969947099685669,
+      "learning_rate": 5.407871837761879e-06,
+      "loss": 0.8326,
+      "step": 17277
+    },
+    {
+      "epoch": 0.9509604271011063,
+      "grad_norm": 0.6230607032775879,
+      "learning_rate": 5.407439814245196e-06,
+      "loss": 0.7211,
+      "step": 17278
+    },
+    {
+      "epoch": 0.951015465903462,
+      "grad_norm": 0.6495433449745178,
+      "learning_rate": 5.407007787666303e-06,
+      "loss": 0.6629,
+      "step": 17279
+    },
+    {
+      "epoch": 0.9510705047058176,
+      "grad_norm": 0.782513439655304,
+      "learning_rate": 5.406575758028448e-06,
+      "loss": 0.7655,
+      "step": 17280
+    },
+    {
+      "epoch": 0.9511255435081732,
+      "grad_norm": 0.7411374449729919,
+      "learning_rate": 5.40614372533488e-06,
+      "loss": 0.7893,
+      "step": 17281
+    },
+    {
+      "epoch": 0.9511805823105289,
+      "grad_norm": 0.7352038621902466,
+      "learning_rate": 5.405711689588844e-06,
+      "loss": 0.7761,
+      "step": 17282
+    },
+    {
+      "epoch": 0.9512356211128846,
+      "grad_norm": 0.6346562504768372,
+      "learning_rate": 5.405279650793587e-06,
+      "loss": 0.7511,
+      "step": 17283
+    },
+    {
+      "epoch": 0.9512906599152402,
+      "grad_norm": 0.6457065343856812,
+      "learning_rate": 5.404847608952356e-06,
+      "loss": 0.6964,
+      "step": 17284
+    },
+    {
+      "epoch": 0.9513456987175959,
+      "grad_norm": 0.7764767408370972,
+      "learning_rate": 5.4044155640684e-06,
+      "loss": 0.7979,
+      "step": 17285
+    },
+    {
+      "epoch": 0.9514007375199516,
+      "grad_norm": 0.7626819610595703,
+      "learning_rate": 5.403983516144965e-06,
+      "loss": 0.7799,
+      "step": 17286
+    },
+    {
+      "epoch": 0.9514557763223073,
+      "grad_norm": 0.8251420855522156,
+      "learning_rate": 5.403551465185296e-06,
+      "loss": 0.7631,
+      "step": 17287
+    },
+    {
+      "epoch": 0.9515108151246628,
+      "grad_norm": 0.7766129970550537,
+      "learning_rate": 5.403119411192644e-06,
+      "loss": 0.7879,
+      "step": 17288
+    },
+    {
+      "epoch": 0.9515658539270185,
+      "grad_norm": 0.7220279574394226,
+      "learning_rate": 5.402687354170255e-06,
+      "loss": 0.8042,
+      "step": 17289
+    },
+    {
+      "epoch": 0.9516208927293742,
+      "grad_norm": 0.6045637726783752,
+      "learning_rate": 5.4022552941213745e-06,
+      "loss": 0.6854,
+      "step": 17290
+    },
+    {
+      "epoch": 0.9516759315317299,
+      "grad_norm": 0.6809129118919373,
+      "learning_rate": 5.401823231049251e-06,
+      "loss": 0.6861,
+      "step": 17291
+    },
+    {
+      "epoch": 0.9517309703340855,
+      "grad_norm": 0.654763400554657,
+      "learning_rate": 5.401391164957133e-06,
+      "loss": 0.794,
+      "step": 17292
+    },
+    {
+      "epoch": 0.9517860091364412,
+      "grad_norm": 0.6468398571014404,
+      "learning_rate": 5.4009590958482645e-06,
+      "loss": 0.7368,
+      "step": 17293
+    },
+    {
+      "epoch": 0.9518410479387969,
+      "grad_norm": 0.657477617263794,
+      "learning_rate": 5.400527023725896e-06,
+      "loss": 0.6998,
+      "step": 17294
+    },
+    {
+      "epoch": 0.9518960867411526,
+      "grad_norm": 0.6647237539291382,
+      "learning_rate": 5.400094948593272e-06,
+      "loss": 0.7825,
+      "step": 17295
+    },
+    {
+      "epoch": 0.9519511255435081,
+      "grad_norm": 0.6871651411056519,
+      "learning_rate": 5.399662870453643e-06,
+      "loss": 0.7089,
+      "step": 17296
+    },
+    {
+      "epoch": 0.9520061643458638,
+      "grad_norm": 0.7344084978103638,
+      "learning_rate": 5.399230789310254e-06,
+      "loss": 0.7541,
+      "step": 17297
+    },
+    {
+      "epoch": 0.9520612031482195,
+      "grad_norm": 0.7122900485992432,
+      "learning_rate": 5.3987987051663545e-06,
+      "loss": 0.7669,
+      "step": 17298
+    },
+    {
+      "epoch": 0.9521162419505752,
+      "grad_norm": 0.628433108329773,
+      "learning_rate": 5.39836661802519e-06,
+      "loss": 0.7065,
+      "step": 17299
+    },
+    {
+      "epoch": 0.9521712807529308,
+      "grad_norm": 0.6206123232841492,
+      "learning_rate": 5.397934527890007e-06,
+      "loss": 0.7048,
+      "step": 17300
+    },
+    {
+      "epoch": 0.9522263195552865,
+      "grad_norm": 0.7941209673881531,
+      "learning_rate": 5.397502434764057e-06,
+      "loss": 0.7336,
+      "step": 17301
+    },
+    {
+      "epoch": 0.9522813583576422,
+      "grad_norm": 0.7397454380989075,
+      "learning_rate": 5.397070338650584e-06,
+      "loss": 0.7303,
+      "step": 17302
+    },
+    {
+      "epoch": 0.9523363971599978,
+      "grad_norm": 0.6538659930229187,
+      "learning_rate": 5.396638239552837e-06,
+      "loss": 0.7292,
+      "step": 17303
+    },
+    {
+      "epoch": 0.9523914359623534,
+      "grad_norm": 0.7217223048210144,
+      "learning_rate": 5.396206137474064e-06,
+      "loss": 0.7441,
+      "step": 17304
+    },
+    {
+      "epoch": 0.9524464747647091,
+      "grad_norm": 0.765256404876709,
+      "learning_rate": 5.395774032417511e-06,
+      "loss": 0.7527,
+      "step": 17305
+    },
+    {
+      "epoch": 0.9525015135670648,
+      "grad_norm": 0.6991282105445862,
+      "learning_rate": 5.395341924386426e-06,
+      "loss": 0.5727,
+      "step": 17306
+    },
+    {
+      "epoch": 0.9525565523694205,
+      "grad_norm": 0.7076554894447327,
+      "learning_rate": 5.394909813384056e-06,
+      "loss": 0.7784,
+      "step": 17307
+    },
+    {
+      "epoch": 0.9526115911717761,
+      "grad_norm": 0.7152629494667053,
+      "learning_rate": 5.3944776994136505e-06,
+      "loss": 0.722,
+      "step": 17308
+    },
+    {
+      "epoch": 0.9526666299741318,
+      "grad_norm": 0.7322426438331604,
+      "learning_rate": 5.394045582478457e-06,
+      "loss": 0.7983,
+      "step": 17309
+    },
+    {
+      "epoch": 0.9527216687764875,
+      "grad_norm": 0.7486420273780823,
+      "learning_rate": 5.393613462581721e-06,
+      "loss": 0.7514,
+      "step": 17310
+    },
+    {
+      "epoch": 0.952776707578843,
+      "grad_norm": 0.8110840320587158,
+      "learning_rate": 5.393181339726692e-06,
+      "loss": 0.6736,
+      "step": 17311
+    },
+    {
+      "epoch": 0.9528317463811987,
+      "grad_norm": 0.6394487023353577,
+      "learning_rate": 5.392749213916617e-06,
+      "loss": 0.7039,
+      "step": 17312
+    },
+    {
+      "epoch": 0.9528867851835544,
+      "grad_norm": 0.6446092128753662,
+      "learning_rate": 5.3923170851547445e-06,
+      "loss": 0.6985,
+      "step": 17313
+    },
+    {
+      "epoch": 0.9529418239859101,
+      "grad_norm": 0.6536337733268738,
+      "learning_rate": 5.391884953444323e-06,
+      "loss": 0.7503,
+      "step": 17314
+    },
+    {
+      "epoch": 0.9529968627882657,
+      "grad_norm": 0.7714772820472717,
+      "learning_rate": 5.3914528187885975e-06,
+      "loss": 0.7068,
+      "step": 17315
+    },
+    {
+      "epoch": 0.9530519015906214,
+      "grad_norm": 0.6758143901824951,
+      "learning_rate": 5.391020681190818e-06,
+      "loss": 0.689,
+      "step": 17316
+    },
+    {
+      "epoch": 0.953106940392977,
+      "grad_norm": 0.730246365070343,
+      "learning_rate": 5.390588540654232e-06,
+      "loss": 0.7242,
+      "step": 17317
+    },
+    {
+      "epoch": 0.9531619791953327,
+      "grad_norm": 0.7205057144165039,
+      "learning_rate": 5.390156397182087e-06,
+      "loss": 0.764,
+      "step": 17318
+    },
+    {
+      "epoch": 0.9532170179976883,
+      "grad_norm": 0.6799334287643433,
+      "learning_rate": 5.3897242507776295e-06,
+      "loss": 0.6869,
+      "step": 17319
+    },
+    {
+      "epoch": 0.953272056800044,
+      "grad_norm": 0.7646334171295166,
+      "learning_rate": 5.38929210144411e-06,
+      "loss": 0.7594,
+      "step": 17320
+    },
+    {
+      "epoch": 0.9533270956023997,
+      "grad_norm": 0.8450561761856079,
+      "learning_rate": 5.388859949184776e-06,
+      "loss": 0.841,
+      "step": 17321
+    },
+    {
+      "epoch": 0.9533821344047554,
+      "grad_norm": 0.6637916564941406,
+      "learning_rate": 5.388427794002874e-06,
+      "loss": 0.8445,
+      "step": 17322
+    },
+    {
+      "epoch": 0.953437173207111,
+      "grad_norm": 0.6471973657608032,
+      "learning_rate": 5.387995635901652e-06,
+      "loss": 0.7442,
+      "step": 17323
+    },
+    {
+      "epoch": 0.9534922120094667,
+      "grad_norm": 0.7502676844596863,
+      "learning_rate": 5.3875634748843585e-06,
+      "loss": 0.7112,
+      "step": 17324
+    },
+    {
+      "epoch": 0.9535472508118223,
+      "grad_norm": 0.6649467349052429,
+      "learning_rate": 5.3871313109542435e-06,
+      "loss": 0.7605,
+      "step": 17325
+    },
+    {
+      "epoch": 0.953602289614178,
+      "grad_norm": 0.6515121459960938,
+      "learning_rate": 5.3866991441145535e-06,
+      "loss": 0.7372,
+      "step": 17326
+    },
+    {
+      "epoch": 0.9536573284165336,
+      "grad_norm": 0.6831742525100708,
+      "learning_rate": 5.386266974368533e-06,
+      "loss": 0.762,
+      "step": 17327
+    },
+    {
+      "epoch": 0.9537123672188893,
+      "grad_norm": 0.7404910922050476,
+      "learning_rate": 5.3858348017194364e-06,
+      "loss": 0.7875,
+      "step": 17328
+    },
+    {
+      "epoch": 0.953767406021245,
+      "grad_norm": 0.7611355185508728,
+      "learning_rate": 5.385402626170507e-06,
+      "loss": 0.7679,
+      "step": 17329
+    },
+    {
+      "epoch": 0.9538224448236007,
+      "grad_norm": 0.8916192054748535,
+      "learning_rate": 5.384970447724995e-06,
+      "loss": 0.785,
+      "step": 17330
+    },
+    {
+      "epoch": 0.9538774836259563,
+      "grad_norm": 0.694205641746521,
+      "learning_rate": 5.3845382663861475e-06,
+      "loss": 0.7292,
+      "step": 17331
+    },
+    {
+      "epoch": 0.953932522428312,
+      "grad_norm": 0.7292189598083496,
+      "learning_rate": 5.384106082157214e-06,
+      "loss": 0.8187,
+      "step": 17332
+    },
+    {
+      "epoch": 0.9539875612306676,
+      "grad_norm": 0.7211722135543823,
+      "learning_rate": 5.3836738950414424e-06,
+      "loss": 0.6727,
+      "step": 17333
+    },
+    {
+      "epoch": 0.9540426000330233,
+      "grad_norm": 0.7116650342941284,
+      "learning_rate": 5.38324170504208e-06,
+      "loss": 0.774,
+      "step": 17334
+    },
+    {
+      "epoch": 0.9540976388353789,
+      "grad_norm": 0.6604741215705872,
+      "learning_rate": 5.3828095121623755e-06,
+      "loss": 0.6726,
+      "step": 17335
+    },
+    {
+      "epoch": 0.9541526776377346,
+      "grad_norm": 0.7605896592140198,
+      "learning_rate": 5.382377316405578e-06,
+      "loss": 0.7965,
+      "step": 17336
+    },
+    {
+      "epoch": 0.9542077164400903,
+      "grad_norm": 0.811404287815094,
+      "learning_rate": 5.3819451177749345e-06,
+      "loss": 0.6583,
+      "step": 17337
+    },
+    {
+      "epoch": 0.954262755242446,
+      "grad_norm": 0.6798723340034485,
+      "learning_rate": 5.381512916273695e-06,
+      "loss": 0.6482,
+      "step": 17338
+    },
+    {
+      "epoch": 0.9543177940448015,
+      "grad_norm": 0.8835861682891846,
+      "learning_rate": 5.381080711905105e-06,
+      "loss": 0.8003,
+      "step": 17339
+    },
+    {
+      "epoch": 0.9543728328471572,
+      "grad_norm": 0.6525741219520569,
+      "learning_rate": 5.380648504672413e-06,
+      "loss": 0.7198,
+      "step": 17340
+    },
+    {
+      "epoch": 0.9544278716495129,
+      "grad_norm": 0.738264262676239,
+      "learning_rate": 5.380216294578871e-06,
+      "loss": 0.7832,
+      "step": 17341
+    },
+    {
+      "epoch": 0.9544829104518686,
+      "grad_norm": 0.7376875877380371,
+      "learning_rate": 5.379784081627725e-06,
+      "loss": 0.8544,
+      "step": 17342
+    },
+    {
+      "epoch": 0.9545379492542242,
+      "grad_norm": 0.6113360524177551,
+      "learning_rate": 5.379351865822222e-06,
+      "loss": 0.7242,
+      "step": 17343
+    },
+    {
+      "epoch": 0.9545929880565799,
+      "grad_norm": 0.7817113399505615,
+      "learning_rate": 5.378919647165613e-06,
+      "loss": 0.7488,
+      "step": 17344
+    },
+    {
+      "epoch": 0.9546480268589356,
+      "grad_norm": 0.6994492411613464,
+      "learning_rate": 5.3784874256611455e-06,
+      "loss": 0.7752,
+      "step": 17345
+    },
+    {
+      "epoch": 0.9547030656612913,
+      "grad_norm": 0.7286459803581238,
+      "learning_rate": 5.378055201312067e-06,
+      "loss": 0.6519,
+      "step": 17346
+    },
+    {
+      "epoch": 0.9547581044636468,
+      "grad_norm": 0.6674445271492004,
+      "learning_rate": 5.377622974121627e-06,
+      "loss": 0.8182,
+      "step": 17347
+    },
+    {
+      "epoch": 0.9548131432660025,
+      "grad_norm": 0.6189805269241333,
+      "learning_rate": 5.377190744093074e-06,
+      "loss": 0.6936,
+      "step": 17348
+    },
+    {
+      "epoch": 0.9548681820683582,
+      "grad_norm": 0.6603719592094421,
+      "learning_rate": 5.376758511229656e-06,
+      "loss": 0.6622,
+      "step": 17349
+    },
+    {
+      "epoch": 0.9549232208707139,
+      "grad_norm": 0.71391761302948,
+      "learning_rate": 5.3763262755346245e-06,
+      "loss": 0.7519,
+      "step": 17350
+    },
+    {
+      "epoch": 0.9549782596730695,
+      "grad_norm": 0.6202741265296936,
+      "learning_rate": 5.375894037011222e-06,
+      "loss": 0.6252,
+      "step": 17351
+    },
+    {
+      "epoch": 0.9550332984754252,
+      "grad_norm": 0.6846199631690979,
+      "learning_rate": 5.375461795662702e-06,
+      "loss": 0.7232,
+      "step": 17352
+    },
+    {
+      "epoch": 0.9550883372777809,
+      "grad_norm": 0.6886401772499084,
+      "learning_rate": 5.375029551492312e-06,
+      "loss": 0.6587,
+      "step": 17353
+    },
+    {
+      "epoch": 0.9551433760801364,
+      "grad_norm": 0.6783032417297363,
+      "learning_rate": 5.3745973045033e-06,
+      "loss": 0.761,
+      "step": 17354
+    },
+    {
+      "epoch": 0.9551984148824921,
+      "grad_norm": 0.947640597820282,
+      "learning_rate": 5.374165054698913e-06,
+      "loss": 0.8755,
+      "step": 17355
+    },
+    {
+      "epoch": 0.9552534536848478,
+      "grad_norm": 0.8471985459327698,
+      "learning_rate": 5.3737328020824045e-06,
+      "loss": 0.8288,
+      "step": 17356
+    },
+    {
+      "epoch": 0.9553084924872035,
+      "grad_norm": 0.6486181616783142,
+      "learning_rate": 5.373300546657019e-06,
+      "loss": 0.7299,
+      "step": 17357
+    },
+    {
+      "epoch": 0.9553635312895591,
+      "grad_norm": 0.7250335216522217,
+      "learning_rate": 5.372868288426007e-06,
+      "loss": 0.7624,
+      "step": 17358
+    },
+    {
+      "epoch": 0.9554185700919148,
+      "grad_norm": 0.7069470286369324,
+      "learning_rate": 5.372436027392615e-06,
+      "loss": 0.8266,
+      "step": 17359
+    },
+    {
+      "epoch": 0.9554736088942705,
+      "grad_norm": 0.7567838430404663,
+      "learning_rate": 5.372003763560095e-06,
+      "loss": 0.7348,
+      "step": 17360
+    },
+    {
+      "epoch": 0.9555286476966262,
+      "grad_norm": 0.707920491695404,
+      "learning_rate": 5.371571496931694e-06,
+      "loss": 0.7466,
+      "step": 17361
+    },
+    {
+      "epoch": 0.9555836864989817,
+      "grad_norm": 0.7111881375312805,
+      "learning_rate": 5.371139227510662e-06,
+      "loss": 0.7965,
+      "step": 17362
+    },
+    {
+      "epoch": 0.9556387253013374,
+      "grad_norm": 0.5888195633888245,
+      "learning_rate": 5.370706955300245e-06,
+      "loss": 0.6534,
+      "step": 17363
+    },
+    {
+      "epoch": 0.9556937641036931,
+      "grad_norm": 0.6748310327529907,
+      "learning_rate": 5.370274680303697e-06,
+      "loss": 0.6856,
+      "step": 17364
+    },
+    {
+      "epoch": 0.9557488029060488,
+      "grad_norm": 0.6798354387283325,
+      "learning_rate": 5.369842402524261e-06,
+      "loss": 0.7213,
+      "step": 17365
+    },
+    {
+      "epoch": 0.9558038417084044,
+      "grad_norm": 0.7532860636711121,
+      "learning_rate": 5.369410121965189e-06,
+      "loss": 0.705,
+      "step": 17366
+    },
+    {
+      "epoch": 0.9558588805107601,
+      "grad_norm": 0.6346868872642517,
+      "learning_rate": 5.368977838629728e-06,
+      "loss": 0.6898,
+      "step": 17367
+    },
+    {
+      "epoch": 0.9559139193131158,
+      "grad_norm": 0.6765021681785583,
+      "learning_rate": 5.368545552521131e-06,
+      "loss": 0.7351,
+      "step": 17368
+    },
+    {
+      "epoch": 0.9559689581154714,
+      "grad_norm": 0.8616527915000916,
+      "learning_rate": 5.368113263642643e-06,
+      "loss": 0.8277,
+      "step": 17369
+    },
+    {
+      "epoch": 0.956023996917827,
+      "grad_norm": 0.6542890667915344,
+      "learning_rate": 5.367680971997514e-06,
+      "loss": 0.7375,
+      "step": 17370
+    },
+    {
+      "epoch": 0.9560790357201827,
+      "grad_norm": 0.6682239174842834,
+      "learning_rate": 5.367248677588993e-06,
+      "loss": 0.713,
+      "step": 17371
+    },
+    {
+      "epoch": 0.9561340745225384,
+      "grad_norm": 0.8059317469596863,
+      "learning_rate": 5.36681638042033e-06,
+      "loss": 0.7312,
+      "step": 17372
+    },
+    {
+      "epoch": 0.9561891133248941,
+      "grad_norm": 0.6048741936683655,
+      "learning_rate": 5.366384080494773e-06,
+      "loss": 0.6356,
+      "step": 17373
+    },
+    {
+      "epoch": 0.9562441521272497,
+      "grad_norm": 0.625925600528717,
+      "learning_rate": 5.365951777815572e-06,
+      "loss": 0.7489,
+      "step": 17374
+    },
+    {
+      "epoch": 0.9562991909296054,
+      "grad_norm": 0.6174320578575134,
+      "learning_rate": 5.365519472385975e-06,
+      "loss": 0.674,
+      "step": 17375
+    },
+    {
+      "epoch": 0.956354229731961,
+      "grad_norm": 0.6741634607315063,
+      "learning_rate": 5.365087164209231e-06,
+      "loss": 0.7194,
+      "step": 17376
+    },
+    {
+      "epoch": 0.9564092685343167,
+      "grad_norm": 0.7016188502311707,
+      "learning_rate": 5.36465485328859e-06,
+      "loss": 0.8047,
+      "step": 17377
+    },
+    {
+      "epoch": 0.9564643073366723,
+      "grad_norm": 0.6745482087135315,
+      "learning_rate": 5.364222539627301e-06,
+      "loss": 0.7661,
+      "step": 17378
+    },
+    {
+      "epoch": 0.956519346139028,
+      "grad_norm": 1.3635177612304688,
+      "learning_rate": 5.363790223228611e-06,
+      "loss": 0.7666,
+      "step": 17379
+    },
+    {
+      "epoch": 0.9565743849413837,
+      "grad_norm": 0.6775792837142944,
+      "learning_rate": 5.363357904095773e-06,
+      "loss": 0.7406,
+      "step": 17380
+    },
+    {
+      "epoch": 0.9566294237437394,
+      "grad_norm": 0.6810624003410339,
+      "learning_rate": 5.362925582232035e-06,
+      "loss": 0.6818,
+      "step": 17381
+    },
+    {
+      "epoch": 0.956684462546095,
+      "grad_norm": 0.6558675765991211,
+      "learning_rate": 5.362493257640644e-06,
+      "loss": 0.8256,
+      "step": 17382
+    },
+    {
+      "epoch": 0.9567395013484506,
+      "grad_norm": 0.8928041458129883,
+      "learning_rate": 5.362060930324849e-06,
+      "loss": 0.8213,
+      "step": 17383
+    },
+    {
+      "epoch": 0.9567945401508063,
+      "grad_norm": 0.705581784248352,
+      "learning_rate": 5.361628600287904e-06,
+      "loss": 0.7597,
+      "step": 17384
+    },
+    {
+      "epoch": 0.956849578953162,
+      "grad_norm": 0.701968789100647,
+      "learning_rate": 5.3611962675330545e-06,
+      "loss": 0.7499,
+      "step": 17385
+    },
+    {
+      "epoch": 0.9569046177555176,
+      "grad_norm": 0.671457052230835,
+      "learning_rate": 5.360763932063551e-06,
+      "loss": 0.6984,
+      "step": 17386
+    },
+    {
+      "epoch": 0.9569596565578733,
+      "grad_norm": 0.6033732891082764,
+      "learning_rate": 5.36033159388264e-06,
+      "loss": 0.6182,
+      "step": 17387
+    },
+    {
+      "epoch": 0.957014695360229,
+      "grad_norm": 0.6885262131690979,
+      "learning_rate": 5.359899252993576e-06,
+      "loss": 0.7577,
+      "step": 17388
+    },
+    {
+      "epoch": 0.9570697341625847,
+      "grad_norm": 0.6726160645484924,
+      "learning_rate": 5.359466909399604e-06,
+      "loss": 0.745,
+      "step": 17389
+    },
+    {
+      "epoch": 0.9571247729649403,
+      "grad_norm": 0.6888847947120667,
+      "learning_rate": 5.359034563103975e-06,
+      "loss": 0.8358,
+      "step": 17390
+    },
+    {
+      "epoch": 0.9571798117672959,
+      "grad_norm": 0.6707327961921692,
+      "learning_rate": 5.358602214109939e-06,
+      "loss": 0.7217,
+      "step": 17391
+    },
+    {
+      "epoch": 0.9572348505696516,
+      "grad_norm": 0.72385174036026,
+      "learning_rate": 5.358169862420745e-06,
+      "loss": 0.8016,
+      "step": 17392
+    },
+    {
+      "epoch": 0.9572898893720073,
+      "grad_norm": 0.6428029537200928,
+      "learning_rate": 5.357737508039641e-06,
+      "loss": 0.7241,
+      "step": 17393
+    },
+    {
+      "epoch": 0.9573449281743629,
+      "grad_norm": 0.6318633556365967,
+      "learning_rate": 5.357305150969878e-06,
+      "loss": 0.7515,
+      "step": 17394
+    },
+    {
+      "epoch": 0.9573999669767186,
+      "grad_norm": 0.7007773518562317,
+      "learning_rate": 5.3568727912147055e-06,
+      "loss": 0.7852,
+      "step": 17395
+    },
+    {
+      "epoch": 0.9574550057790743,
+      "grad_norm": 0.6996271014213562,
+      "learning_rate": 5.356440428777372e-06,
+      "loss": 0.8811,
+      "step": 17396
+    },
+    {
+      "epoch": 0.9575100445814299,
+      "grad_norm": 0.6028952598571777,
+      "learning_rate": 5.356008063661128e-06,
+      "loss": 0.7801,
+      "step": 17397
+    },
+    {
+      "epoch": 0.9575650833837855,
+      "grad_norm": 0.6873906850814819,
+      "learning_rate": 5.355575695869224e-06,
+      "loss": 0.6855,
+      "step": 17398
+    },
+    {
+      "epoch": 0.9576201221861412,
+      "grad_norm": 0.6360284686088562,
+      "learning_rate": 5.355143325404908e-06,
+      "loss": 0.7041,
+      "step": 17399
+    },
+    {
+      "epoch": 0.9576751609884969,
+      "grad_norm": 0.7114169597625732,
+      "learning_rate": 5.354710952271427e-06,
+      "loss": 0.7394,
+      "step": 17400
+    },
+    {
+      "epoch": 0.9577301997908525,
+      "grad_norm": 0.6143094301223755,
+      "learning_rate": 5.354278576472036e-06,
+      "loss": 0.7438,
+      "step": 17401
+    },
+    {
+      "epoch": 0.9577852385932082,
+      "grad_norm": 0.6581745147705078,
+      "learning_rate": 5.353846198009981e-06,
+      "loss": 0.7388,
+      "step": 17402
+    },
+    {
+      "epoch": 0.9578402773955639,
+      "grad_norm": 0.735107958316803,
+      "learning_rate": 5.353413816888513e-06,
+      "loss": 0.7853,
+      "step": 17403
+    },
+    {
+      "epoch": 0.9578953161979196,
+      "grad_norm": 0.7374551892280579,
+      "learning_rate": 5.352981433110881e-06,
+      "loss": 0.8393,
+      "step": 17404
+    },
+    {
+      "epoch": 0.9579503550002751,
+      "grad_norm": 0.7229042053222656,
+      "learning_rate": 5.352549046680335e-06,
+      "loss": 0.7052,
+      "step": 17405
+    },
+    {
+      "epoch": 0.9580053938026308,
+      "grad_norm": 0.6857532262802124,
+      "learning_rate": 5.352116657600126e-06,
+      "loss": 0.7484,
+      "step": 17406
+    },
+    {
+      "epoch": 0.9580604326049865,
+      "grad_norm": 0.6766275763511658,
+      "learning_rate": 5.351684265873501e-06,
+      "loss": 0.7739,
+      "step": 17407
+    },
+    {
+      "epoch": 0.9581154714073422,
+      "grad_norm": 0.7877463698387146,
+      "learning_rate": 5.351251871503712e-06,
+      "loss": 0.7681,
+      "step": 17408
+    },
+    {
+      "epoch": 0.9581705102096978,
+      "grad_norm": 0.8620951771736145,
+      "learning_rate": 5.350819474494009e-06,
+      "loss": 0.8691,
+      "step": 17409
+    },
+    {
+      "epoch": 0.9582255490120535,
+      "grad_norm": 0.6989312171936035,
+      "learning_rate": 5.350387074847639e-06,
+      "loss": 0.74,
+      "step": 17410
+    },
+    {
+      "epoch": 0.9582805878144092,
+      "grad_norm": 0.6328219771385193,
+      "learning_rate": 5.349954672567854e-06,
+      "loss": 0.6934,
+      "step": 17411
+    },
+    {
+      "epoch": 0.9583356266167649,
+      "grad_norm": 0.6802984476089478,
+      "learning_rate": 5.349522267657903e-06,
+      "loss": 0.6071,
+      "step": 17412
+    },
+    {
+      "epoch": 0.9583906654191204,
+      "grad_norm": 0.579208493232727,
+      "learning_rate": 5.349089860121037e-06,
+      "loss": 0.6382,
+      "step": 17413
+    },
+    {
+      "epoch": 0.9584457042214761,
+      "grad_norm": 0.712049663066864,
+      "learning_rate": 5.348657449960505e-06,
+      "loss": 0.6865,
+      "step": 17414
+    },
+    {
+      "epoch": 0.9585007430238318,
+      "grad_norm": 0.759441614151001,
+      "learning_rate": 5.3482250371795566e-06,
+      "loss": 0.7735,
+      "step": 17415
+    },
+    {
+      "epoch": 0.9585557818261875,
+      "grad_norm": 0.7504081130027771,
+      "learning_rate": 5.347792621781441e-06,
+      "loss": 0.7704,
+      "step": 17416
+    },
+    {
+      "epoch": 0.9586108206285431,
+      "grad_norm": 0.6927568912506104,
+      "learning_rate": 5.34736020376941e-06,
+      "loss": 0.7529,
+      "step": 17417
+    },
+    {
+      "epoch": 0.9586658594308988,
+      "grad_norm": 0.6337310075759888,
+      "learning_rate": 5.346927783146714e-06,
+      "loss": 0.717,
+      "step": 17418
+    },
+    {
+      "epoch": 0.9587208982332545,
+      "grad_norm": 0.6488094925880432,
+      "learning_rate": 5.3464953599166e-06,
+      "loss": 0.7084,
+      "step": 17419
+    },
+    {
+      "epoch": 0.9587759370356101,
+      "grad_norm": 0.6433645486831665,
+      "learning_rate": 5.34606293408232e-06,
+      "loss": 0.6765,
+      "step": 17420
+    },
+    {
+      "epoch": 0.9588309758379657,
+      "grad_norm": 0.6713243722915649,
+      "learning_rate": 5.345630505647125e-06,
+      "loss": 0.7357,
+      "step": 17421
+    },
+    {
+      "epoch": 0.9588860146403214,
+      "grad_norm": 0.6575330495834351,
+      "learning_rate": 5.345198074614262e-06,
+      "loss": 0.7312,
+      "step": 17422
+    },
+    {
+      "epoch": 0.9589410534426771,
+      "grad_norm": 0.642666757106781,
+      "learning_rate": 5.344765640986984e-06,
+      "loss": 0.7355,
+      "step": 17423
+    },
+    {
+      "epoch": 0.9589960922450328,
+      "grad_norm": 0.7026614546775818,
+      "learning_rate": 5.344333204768537e-06,
+      "loss": 0.7811,
+      "step": 17424
+    },
+    {
+      "epoch": 0.9590511310473884,
+      "grad_norm": 0.6988122463226318,
+      "learning_rate": 5.343900765962177e-06,
+      "loss": 0.6903,
+      "step": 17425
+    },
+    {
+      "epoch": 0.9591061698497441,
+      "grad_norm": 0.682156503200531,
+      "learning_rate": 5.34346832457115e-06,
+      "loss": 0.7504,
+      "step": 17426
+    },
+    {
+      "epoch": 0.9591612086520998,
+      "grad_norm": 0.6704639196395874,
+      "learning_rate": 5.343035880598706e-06,
+      "loss": 0.7432,
+      "step": 17427
+    },
+    {
+      "epoch": 0.9592162474544554,
+      "grad_norm": 0.6490553617477417,
+      "learning_rate": 5.342603434048096e-06,
+      "loss": 0.7347,
+      "step": 17428
+    },
+    {
+      "epoch": 0.959271286256811,
+      "grad_norm": 0.657006561756134,
+      "learning_rate": 5.342170984922571e-06,
+      "loss": 0.772,
+      "step": 17429
+    },
+    {
+      "epoch": 0.9593263250591667,
+      "grad_norm": 0.7124611735343933,
+      "learning_rate": 5.34173853322538e-06,
+      "loss": 0.7326,
+      "step": 17430
+    },
+    {
+      "epoch": 0.9593813638615224,
+      "grad_norm": 0.6327362060546875,
+      "learning_rate": 5.341306078959775e-06,
+      "loss": 0.7248,
+      "step": 17431
+    },
+    {
+      "epoch": 0.9594364026638781,
+      "grad_norm": 0.6160357594490051,
+      "learning_rate": 5.3408736221290025e-06,
+      "loss": 0.6655,
+      "step": 17432
+    },
+    {
+      "epoch": 0.9594914414662337,
+      "grad_norm": 0.729807436466217,
+      "learning_rate": 5.340441162736319e-06,
+      "loss": 0.6996,
+      "step": 17433
+    },
+    {
+      "epoch": 0.9595464802685894,
+      "grad_norm": 0.6158677339553833,
+      "learning_rate": 5.340008700784968e-06,
+      "loss": 0.6189,
+      "step": 17434
+    },
+    {
+      "epoch": 0.959601519070945,
+      "grad_norm": 0.6435769200325012,
+      "learning_rate": 5.339576236278203e-06,
+      "loss": 0.7464,
+      "step": 17435
+    },
+    {
+      "epoch": 0.9596565578733007,
+      "grad_norm": 0.7240917682647705,
+      "learning_rate": 5.339143769219272e-06,
+      "loss": 0.7844,
+      "step": 17436
+    },
+    {
+      "epoch": 0.9597115966756563,
+      "grad_norm": 0.6149145364761353,
+      "learning_rate": 5.33871129961143e-06,
+      "loss": 0.6943,
+      "step": 17437
+    },
+    {
+      "epoch": 0.959766635478012,
+      "grad_norm": 0.7776970267295837,
+      "learning_rate": 5.338278827457924e-06,
+      "loss": 0.7434,
+      "step": 17438
+    },
+    {
+      "epoch": 0.9598216742803677,
+      "grad_norm": 0.5908457636833191,
+      "learning_rate": 5.337846352762005e-06,
+      "loss": 0.6682,
+      "step": 17439
+    },
+    {
+      "epoch": 0.9598767130827233,
+      "grad_norm": 0.7880268692970276,
+      "learning_rate": 5.337413875526923e-06,
+      "loss": 0.7958,
+      "step": 17440
+    },
+    {
+      "epoch": 0.959931751885079,
+      "grad_norm": 0.8126769065856934,
+      "learning_rate": 5.336981395755928e-06,
+      "loss": 0.8199,
+      "step": 17441
+    },
+    {
+      "epoch": 0.9599867906874346,
+      "grad_norm": 0.7309434413909912,
+      "learning_rate": 5.336548913452271e-06,
+      "loss": 0.8179,
+      "step": 17442
+    },
+    {
+      "epoch": 0.9600418294897903,
+      "grad_norm": 0.7369129657745361,
+      "learning_rate": 5.3361164286192034e-06,
+      "loss": 0.7206,
+      "step": 17443
+    },
+    {
+      "epoch": 0.9600968682921459,
+      "grad_norm": 0.716317892074585,
+      "learning_rate": 5.3356839412599736e-06,
+      "loss": 0.6901,
+      "step": 17444
+    },
+    {
+      "epoch": 0.9601519070945016,
+      "grad_norm": 0.9500069618225098,
+      "learning_rate": 5.335251451377835e-06,
+      "loss": 0.7165,
+      "step": 17445
+    },
+    {
+      "epoch": 0.9602069458968573,
+      "grad_norm": 0.6800429821014404,
+      "learning_rate": 5.334818958976035e-06,
+      "loss": 0.684,
+      "step": 17446
+    },
+    {
+      "epoch": 0.960261984699213,
+      "grad_norm": 0.7164038419723511,
+      "learning_rate": 5.334386464057825e-06,
+      "loss": 0.7046,
+      "step": 17447
+    },
+    {
+      "epoch": 0.9603170235015686,
+      "grad_norm": 0.6348811984062195,
+      "learning_rate": 5.3339539666264564e-06,
+      "loss": 0.7413,
+      "step": 17448
+    },
+    {
+      "epoch": 0.9603720623039242,
+      "grad_norm": 0.6681154370307922,
+      "learning_rate": 5.333521466685179e-06,
+      "loss": 0.6845,
+      "step": 17449
+    },
+    {
+      "epoch": 0.9604271011062799,
+      "grad_norm": 0.6277575492858887,
+      "learning_rate": 5.333088964237243e-06,
+      "loss": 0.7471,
+      "step": 17450
+    },
+    {
+      "epoch": 0.9604821399086356,
+      "grad_norm": 0.6492090225219727,
+      "learning_rate": 5.332656459285901e-06,
+      "loss": 0.7499,
+      "step": 17451
+    },
+    {
+      "epoch": 0.9605371787109912,
+      "grad_norm": 0.6493937373161316,
+      "learning_rate": 5.3322239518344e-06,
+      "loss": 0.73,
+      "step": 17452
+    },
+    {
+      "epoch": 0.9605922175133469,
+      "grad_norm": 0.7362397313117981,
+      "learning_rate": 5.331791441885994e-06,
+      "loss": 0.7677,
+      "step": 17453
+    },
+    {
+      "epoch": 0.9606472563157026,
+      "grad_norm": 0.6337111592292786,
+      "learning_rate": 5.331358929443932e-06,
+      "loss": 0.7343,
+      "step": 17454
+    },
+    {
+      "epoch": 0.9607022951180583,
+      "grad_norm": 0.6027660965919495,
+      "learning_rate": 5.330926414511466e-06,
+      "loss": 0.6505,
+      "step": 17455
+    },
+    {
+      "epoch": 0.9607573339204138,
+      "grad_norm": 0.6437423229217529,
+      "learning_rate": 5.330493897091844e-06,
+      "loss": 0.6762,
+      "step": 17456
+    },
+    {
+      "epoch": 0.9608123727227695,
+      "grad_norm": 0.7299374938011169,
+      "learning_rate": 5.33006137718832e-06,
+      "loss": 0.7085,
+      "step": 17457
+    },
+    {
+      "epoch": 0.9608674115251252,
+      "grad_norm": 0.6796537637710571,
+      "learning_rate": 5.329628854804142e-06,
+      "loss": 0.763,
+      "step": 17458
+    },
+    {
+      "epoch": 0.9609224503274809,
+      "grad_norm": 0.6185942888259888,
+      "learning_rate": 5.329196329942563e-06,
+      "loss": 0.672,
+      "step": 17459
+    },
+    {
+      "epoch": 0.9609774891298365,
+      "grad_norm": 0.695350706577301,
+      "learning_rate": 5.32876380260683e-06,
+      "loss": 0.7561,
+      "step": 17460
+    },
+    {
+      "epoch": 0.9610325279321922,
+      "grad_norm": 0.6831729412078857,
+      "learning_rate": 5.328331272800198e-06,
+      "loss": 0.7777,
+      "step": 17461
+    },
+    {
+      "epoch": 0.9610875667345479,
+      "grad_norm": 0.6294922828674316,
+      "learning_rate": 5.327898740525916e-06,
+      "loss": 0.6117,
+      "step": 17462
+    },
+    {
+      "epoch": 0.9611426055369036,
+      "grad_norm": 0.8314191102981567,
+      "learning_rate": 5.327466205787235e-06,
+      "loss": 0.7627,
+      "step": 17463
+    },
+    {
+      "epoch": 0.9611976443392591,
+      "grad_norm": 0.7389507293701172,
+      "learning_rate": 5.327033668587404e-06,
+      "loss": 0.727,
+      "step": 17464
+    },
+    {
+      "epoch": 0.9612526831416148,
+      "grad_norm": 0.7043427228927612,
+      "learning_rate": 5.326601128929677e-06,
+      "loss": 0.7905,
+      "step": 17465
+    },
+    {
+      "epoch": 0.9613077219439705,
+      "grad_norm": 0.7443289756774902,
+      "learning_rate": 5.326168586817303e-06,
+      "loss": 0.7561,
+      "step": 17466
+    },
+    {
+      "epoch": 0.9613627607463262,
+      "grad_norm": 0.687877357006073,
+      "learning_rate": 5.3257360422535345e-06,
+      "loss": 0.7341,
+      "step": 17467
+    },
+    {
+      "epoch": 0.9614177995486818,
+      "grad_norm": 0.6753028035163879,
+      "learning_rate": 5.325303495241618e-06,
+      "loss": 0.7707,
+      "step": 17468
+    },
+    {
+      "epoch": 0.9614728383510375,
+      "grad_norm": 0.7854021787643433,
+      "learning_rate": 5.324870945784811e-06,
+      "loss": 0.7471,
+      "step": 17469
+    },
+    {
+      "epoch": 0.9615278771533932,
+      "grad_norm": 0.7157363891601562,
+      "learning_rate": 5.32443839388636e-06,
+      "loss": 0.8361,
+      "step": 17470
+    },
+    {
+      "epoch": 0.9615829159557489,
+      "grad_norm": 0.7041062712669373,
+      "learning_rate": 5.324005839549517e-06,
+      "loss": 0.7341,
+      "step": 17471
+    },
+    {
+      "epoch": 0.9616379547581044,
+      "grad_norm": 1.242568016052246,
+      "learning_rate": 5.32357328277753e-06,
+      "loss": 0.6966,
+      "step": 17472
+    },
+    {
+      "epoch": 0.9616929935604601,
+      "grad_norm": 0.773711085319519,
+      "learning_rate": 5.323140723573655e-06,
+      "loss": 0.8137,
+      "step": 17473
+    },
+    {
+      "epoch": 0.9617480323628158,
+      "grad_norm": 0.8130469918251038,
+      "learning_rate": 5.322708161941141e-06,
+      "loss": 0.8386,
+      "step": 17474
+    },
+    {
+      "epoch": 0.9618030711651715,
+      "grad_norm": 0.6662439107894897,
+      "learning_rate": 5.322275597883239e-06,
+      "loss": 0.758,
+      "step": 17475
+    },
+    {
+      "epoch": 0.9618581099675271,
+      "grad_norm": 0.6854259371757507,
+      "learning_rate": 5.321843031403197e-06,
+      "loss": 0.7073,
+      "step": 17476
+    },
+    {
+      "epoch": 0.9619131487698828,
+      "grad_norm": 0.7293241024017334,
+      "learning_rate": 5.321410462504273e-06,
+      "loss": 0.7156,
+      "step": 17477
+    },
+    {
+      "epoch": 0.9619681875722385,
+      "grad_norm": 0.9480264782905579,
+      "learning_rate": 5.320977891189712e-06,
+      "loss": 0.7912,
+      "step": 17478
+    },
+    {
+      "epoch": 0.9620232263745941,
+      "grad_norm": 0.7539064884185791,
+      "learning_rate": 5.320545317462767e-06,
+      "loss": 0.7198,
+      "step": 17479
+    },
+    {
+      "epoch": 0.9620782651769497,
+      "grad_norm": 0.6641668081283569,
+      "learning_rate": 5.320112741326691e-06,
+      "loss": 0.6635,
+      "step": 17480
+    },
+    {
+      "epoch": 0.9621333039793054,
+      "grad_norm": 0.6983955502510071,
+      "learning_rate": 5.319680162784732e-06,
+      "loss": 0.7335,
+      "step": 17481
+    },
+    {
+      "epoch": 0.9621883427816611,
+      "grad_norm": 0.5957023501396179,
+      "learning_rate": 5.319247581840143e-06,
+      "loss": 0.6946,
+      "step": 17482
+    },
+    {
+      "epoch": 0.9622433815840167,
+      "grad_norm": 0.7082545757293701,
+      "learning_rate": 5.3188149984961735e-06,
+      "loss": 0.665,
+      "step": 17483
+    },
+    {
+      "epoch": 0.9622984203863724,
+      "grad_norm": 0.62541264295578,
+      "learning_rate": 5.318382412756076e-06,
+      "loss": 0.707,
+      "step": 17484
+    },
+    {
+      "epoch": 0.962353459188728,
+      "grad_norm": 0.7225793600082397,
+      "learning_rate": 5.317949824623102e-06,
+      "loss": 0.7426,
+      "step": 17485
+    },
+    {
+      "epoch": 0.9624084979910837,
+      "grad_norm": 0.6111204624176025,
+      "learning_rate": 5.317517234100502e-06,
+      "loss": 0.6987,
+      "step": 17486
+    },
+    {
+      "epoch": 0.9624635367934393,
+      "grad_norm": 0.6709753274917603,
+      "learning_rate": 5.317084641191528e-06,
+      "loss": 0.7365,
+      "step": 17487
+    },
+    {
+      "epoch": 0.962518575595795,
+      "grad_norm": 0.7895059585571289,
+      "learning_rate": 5.3166520458994286e-06,
+      "loss": 0.8158,
+      "step": 17488
+    },
+    {
+      "epoch": 0.9625736143981507,
+      "grad_norm": 1.045240879058838,
+      "learning_rate": 5.31621944822746e-06,
+      "loss": 0.6528,
+      "step": 17489
+    },
+    {
+      "epoch": 0.9626286532005064,
+      "grad_norm": 0.7611956000328064,
+      "learning_rate": 5.3157868481788685e-06,
+      "loss": 0.8025,
+      "step": 17490
+    },
+    {
+      "epoch": 0.962683692002862,
+      "grad_norm": 0.6797771453857422,
+      "learning_rate": 5.315354245756909e-06,
+      "loss": 0.8072,
+      "step": 17491
+    },
+    {
+      "epoch": 0.9627387308052177,
+      "grad_norm": 1.1472023725509644,
+      "learning_rate": 5.314921640964832e-06,
+      "loss": 0.7167,
+      "step": 17492
+    },
+    {
+      "epoch": 0.9627937696075733,
+      "grad_norm": 0.5967952013015747,
+      "learning_rate": 5.3144890338058875e-06,
+      "loss": 0.6456,
+      "step": 17493
+    },
+    {
+      "epoch": 0.962848808409929,
+      "grad_norm": 0.731985867023468,
+      "learning_rate": 5.314056424283327e-06,
+      "loss": 0.7599,
+      "step": 17494
+    },
+    {
+      "epoch": 0.9629038472122846,
+      "grad_norm": 0.6571676135063171,
+      "learning_rate": 5.313623812400403e-06,
+      "loss": 0.7211,
+      "step": 17495
+    },
+    {
+      "epoch": 0.9629588860146403,
+      "grad_norm": 0.6877505779266357,
+      "learning_rate": 5.313191198160366e-06,
+      "loss": 0.7073,
+      "step": 17496
+    },
+    {
+      "epoch": 0.963013924816996,
+      "grad_norm": 0.7390667796134949,
+      "learning_rate": 5.312758581566468e-06,
+      "loss": 0.7676,
+      "step": 17497
+    },
+    {
+      "epoch": 0.9630689636193517,
+      "grad_norm": 0.6747639179229736,
+      "learning_rate": 5.31232596262196e-06,
+      "loss": 0.7069,
+      "step": 17498
+    },
+    {
+      "epoch": 0.9631240024217073,
+      "grad_norm": 0.6707483530044556,
+      "learning_rate": 5.311893341330094e-06,
+      "loss": 0.7882,
+      "step": 17499
+    },
+    {
+      "epoch": 0.963179041224063,
+      "grad_norm": 0.678855836391449,
+      "learning_rate": 5.31146071769412e-06,
+      "loss": 0.8008,
+      "step": 17500
+    },
+    {
+      "epoch": 0.9632340800264186,
+      "grad_norm": 0.7658770084381104,
+      "learning_rate": 5.311028091717292e-06,
+      "loss": 0.7275,
+      "step": 17501
+    },
+    {
+      "epoch": 0.9632891188287743,
+      "grad_norm": 0.6684555411338806,
+      "learning_rate": 5.31059546340286e-06,
+      "loss": 0.6935,
+      "step": 17502
+    },
+    {
+      "epoch": 0.9633441576311299,
+      "grad_norm": 0.6478149890899658,
+      "learning_rate": 5.310162832754076e-06,
+      "loss": 0.743,
+      "step": 17503
+    },
+    {
+      "epoch": 0.9633991964334856,
+      "grad_norm": 0.7288548946380615,
+      "learning_rate": 5.30973019977419e-06,
+      "loss": 0.6683,
+      "step": 17504
+    },
+    {
+      "epoch": 0.9634542352358413,
+      "grad_norm": 0.7290934920310974,
+      "learning_rate": 5.309297564466455e-06,
+      "loss": 0.6866,
+      "step": 17505
+    },
+    {
+      "epoch": 0.963509274038197,
+      "grad_norm": 0.6582328081130981,
+      "learning_rate": 5.308864926834124e-06,
+      "loss": 0.7615,
+      "step": 17506
+    },
+    {
+      "epoch": 0.9635643128405526,
+      "grad_norm": 0.688706636428833,
+      "learning_rate": 5.308432286880446e-06,
+      "loss": 0.8026,
+      "step": 17507
+    },
+    {
+      "epoch": 0.9636193516429082,
+      "grad_norm": 0.6733599305152893,
+      "learning_rate": 5.307999644608673e-06,
+      "loss": 0.7274,
+      "step": 17508
+    },
+    {
+      "epoch": 0.9636743904452639,
+      "grad_norm": 0.6621059775352478,
+      "learning_rate": 5.307567000022055e-06,
+      "loss": 0.6909,
+      "step": 17509
+    },
+    {
+      "epoch": 0.9637294292476196,
+      "grad_norm": 0.8380454778671265,
+      "learning_rate": 5.30713435312385e-06,
+      "loss": 0.7591,
+      "step": 17510
+    },
+    {
+      "epoch": 0.9637844680499752,
+      "grad_norm": 0.7065680623054504,
+      "learning_rate": 5.306701703917303e-06,
+      "loss": 0.7793,
+      "step": 17511
+    },
+    {
+      "epoch": 0.9638395068523309,
+      "grad_norm": 0.7197430729866028,
+      "learning_rate": 5.306269052405667e-06,
+      "loss": 0.7114,
+      "step": 17512
+    },
+    {
+      "epoch": 0.9638945456546866,
+      "grad_norm": 0.7427436113357544,
+      "learning_rate": 5.3058363985921965e-06,
+      "loss": 0.6783,
+      "step": 17513
+    },
+    {
+      "epoch": 0.9639495844570423,
+      "grad_norm": 0.6908861994743347,
+      "learning_rate": 5.305403742480141e-06,
+      "loss": 0.7839,
+      "step": 17514
+    },
+    {
+      "epoch": 0.9640046232593978,
+      "grad_norm": 0.6575813293457031,
+      "learning_rate": 5.304971084072753e-06,
+      "loss": 0.7608,
+      "step": 17515
+    },
+    {
+      "epoch": 0.9640596620617535,
+      "grad_norm": 0.7817708253860474,
+      "learning_rate": 5.304538423373285e-06,
+      "loss": 0.6945,
+      "step": 17516
+    },
+    {
+      "epoch": 0.9641147008641092,
+      "grad_norm": 0.7376648187637329,
+      "learning_rate": 5.304105760384987e-06,
+      "loss": 0.7437,
+      "step": 17517
+    },
+    {
+      "epoch": 0.9641697396664649,
+      "grad_norm": 0.8665770292282104,
+      "learning_rate": 5.3036730951111105e-06,
+      "loss": 0.6309,
+      "step": 17518
+    },
+    {
+      "epoch": 0.9642247784688205,
+      "grad_norm": 0.7191612124443054,
+      "learning_rate": 5.303240427554909e-06,
+      "loss": 0.7783,
+      "step": 17519
+    },
+    {
+      "epoch": 0.9642798172711762,
+      "grad_norm": 0.6969373822212219,
+      "learning_rate": 5.302807757719633e-06,
+      "loss": 0.7916,
+      "step": 17520
+    },
+    {
+      "epoch": 0.9643348560735319,
+      "grad_norm": 0.7042273283004761,
+      "learning_rate": 5.302375085608535e-06,
+      "loss": 0.7401,
+      "step": 17521
+    },
+    {
+      "epoch": 0.9643898948758876,
+      "grad_norm": 0.6587864756584167,
+      "learning_rate": 5.3019424112248676e-06,
+      "loss": 0.6923,
+      "step": 17522
+    },
+    {
+      "epoch": 0.9644449336782431,
+      "grad_norm": 0.6853402256965637,
+      "learning_rate": 5.301509734571881e-06,
+      "loss": 0.7811,
+      "step": 17523
+    },
+    {
+      "epoch": 0.9644999724805988,
+      "grad_norm": 0.8109264969825745,
+      "learning_rate": 5.301077055652828e-06,
+      "loss": 0.7365,
+      "step": 17524
+    },
+    {
+      "epoch": 0.9645550112829545,
+      "grad_norm": 0.6986764073371887,
+      "learning_rate": 5.300644374470961e-06,
+      "loss": 0.7721,
+      "step": 17525
+    },
+    {
+      "epoch": 0.9646100500853101,
+      "grad_norm": 0.6346173286437988,
+      "learning_rate": 5.300211691029532e-06,
+      "loss": 0.6719,
+      "step": 17526
+    },
+    {
+      "epoch": 0.9646650888876658,
+      "grad_norm": 0.7523975372314453,
+      "learning_rate": 5.299779005331792e-06,
+      "loss": 0.718,
+      "step": 17527
+    },
+    {
+      "epoch": 0.9647201276900215,
+      "grad_norm": 0.7034041881561279,
+      "learning_rate": 5.2993463173809925e-06,
+      "loss": 0.7858,
+      "step": 17528
+    },
+    {
+      "epoch": 0.9647751664923772,
+      "grad_norm": 0.7195404767990112,
+      "learning_rate": 5.2989136271803864e-06,
+      "loss": 0.7105,
+      "step": 17529
+    },
+    {
+      "epoch": 0.9648302052947327,
+      "grad_norm": 0.6533499360084534,
+      "learning_rate": 5.298480934733226e-06,
+      "loss": 0.7639,
+      "step": 17530
+    },
+    {
+      "epoch": 0.9648852440970884,
+      "grad_norm": 0.7042137980461121,
+      "learning_rate": 5.298048240042762e-06,
+      "loss": 0.8254,
+      "step": 17531
+    },
+    {
+      "epoch": 0.9649402828994441,
+      "grad_norm": 0.7466374635696411,
+      "learning_rate": 5.297615543112249e-06,
+      "loss": 0.7034,
+      "step": 17532
+    },
+    {
+      "epoch": 0.9649953217017998,
+      "grad_norm": 0.7622665166854858,
+      "learning_rate": 5.297182843944934e-06,
+      "loss": 0.6669,
+      "step": 17533
+    },
+    {
+      "epoch": 0.9650503605041554,
+      "grad_norm": 0.6359787583351135,
+      "learning_rate": 5.296750142544075e-06,
+      "loss": 0.701,
+      "step": 17534
+    },
+    {
+      "epoch": 0.9651053993065111,
+      "grad_norm": 0.6798627376556396,
+      "learning_rate": 5.29631743891292e-06,
+      "loss": 0.7014,
+      "step": 17535
+    },
+    {
+      "epoch": 0.9651604381088668,
+      "grad_norm": 0.7485163807868958,
+      "learning_rate": 5.295884733054723e-06,
+      "loss": 0.758,
+      "step": 17536
+    },
+    {
+      "epoch": 0.9652154769112224,
+      "grad_norm": 0.6358698606491089,
+      "learning_rate": 5.295452024972735e-06,
+      "loss": 0.7043,
+      "step": 17537
+    },
+    {
+      "epoch": 0.965270515713578,
+      "grad_norm": 0.7724503874778748,
+      "learning_rate": 5.29501931467021e-06,
+      "loss": 0.8167,
+      "step": 17538
+    },
+    {
+      "epoch": 0.9653255545159337,
+      "grad_norm": 0.695549726486206,
+      "learning_rate": 5.294586602150398e-06,
+      "loss": 0.7455,
+      "step": 17539
+    },
+    {
+      "epoch": 0.9653805933182894,
+      "grad_norm": 0.6918785572052002,
+      "learning_rate": 5.294153887416552e-06,
+      "loss": 0.7833,
+      "step": 17540
+    },
+    {
+      "epoch": 0.9654356321206451,
+      "grad_norm": 0.6794410943984985,
+      "learning_rate": 5.293721170471924e-06,
+      "loss": 0.6935,
+      "step": 17541
+    },
+    {
+      "epoch": 0.9654906709230007,
+      "grad_norm": 0.8318895101547241,
+      "learning_rate": 5.293288451319767e-06,
+      "loss": 0.9019,
+      "step": 17542
+    },
+    {
+      "epoch": 0.9655457097253564,
+      "grad_norm": 0.6449755430221558,
+      "learning_rate": 5.292855729963332e-06,
+      "loss": 0.7133,
+      "step": 17543
+    },
+    {
+      "epoch": 0.965600748527712,
+      "grad_norm": 0.6092707514762878,
+      "learning_rate": 5.292423006405871e-06,
+      "loss": 0.654,
+      "step": 17544
+    },
+    {
+      "epoch": 0.9656557873300677,
+      "grad_norm": 0.6690982580184937,
+      "learning_rate": 5.291990280650637e-06,
+      "loss": 0.7909,
+      "step": 17545
+    },
+    {
+      "epoch": 0.9657108261324233,
+      "grad_norm": 0.7012231945991516,
+      "learning_rate": 5.291557552700883e-06,
+      "loss": 0.755,
+      "step": 17546
+    },
+    {
+      "epoch": 0.965765864934779,
+      "grad_norm": 0.6244442462921143,
+      "learning_rate": 5.29112482255986e-06,
+      "loss": 0.6907,
+      "step": 17547
+    },
+    {
+      "epoch": 0.9658209037371347,
+      "grad_norm": 0.6664257049560547,
+      "learning_rate": 5.2906920902308215e-06,
+      "loss": 0.8044,
+      "step": 17548
+    },
+    {
+      "epoch": 0.9658759425394904,
+      "grad_norm": 0.7767570614814758,
+      "learning_rate": 5.290259355717018e-06,
+      "loss": 0.8169,
+      "step": 17549
+    },
+    {
+      "epoch": 0.965930981341846,
+      "grad_norm": 0.8205220699310303,
+      "learning_rate": 5.2898266190217025e-06,
+      "loss": 0.8301,
+      "step": 17550
+    },
+    {
+      "epoch": 0.9659860201442017,
+      "grad_norm": 0.5859615206718445,
+      "learning_rate": 5.2893938801481295e-06,
+      "loss": 0.6929,
+      "step": 17551
+    },
+    {
+      "epoch": 0.9660410589465573,
+      "grad_norm": 0.6787375211715698,
+      "learning_rate": 5.288961139099549e-06,
+      "loss": 0.76,
+      "step": 17552
+    },
+    {
+      "epoch": 0.966096097748913,
+      "grad_norm": 0.6915340423583984,
+      "learning_rate": 5.288528395879214e-06,
+      "loss": 0.6894,
+      "step": 17553
+    },
+    {
+      "epoch": 0.9661511365512686,
+      "grad_norm": 0.7098847031593323,
+      "learning_rate": 5.2880956504903765e-06,
+      "loss": 0.6713,
+      "step": 17554
+    },
+    {
+      "epoch": 0.9662061753536243,
+      "grad_norm": 0.6361868977546692,
+      "learning_rate": 5.287662902936289e-06,
+      "loss": 0.6729,
+      "step": 17555
+    },
+    {
+      "epoch": 0.96626121415598,
+      "grad_norm": 0.6745426654815674,
+      "learning_rate": 5.287230153220205e-06,
+      "loss": 0.7441,
+      "step": 17556
+    },
+    {
+      "epoch": 0.9663162529583357,
+      "grad_norm": 0.6941108703613281,
+      "learning_rate": 5.286797401345374e-06,
+      "loss": 0.7468,
+      "step": 17557
+    },
+    {
+      "epoch": 0.9663712917606913,
+      "grad_norm": 0.6621807217597961,
+      "learning_rate": 5.286364647315052e-06,
+      "loss": 0.8005,
+      "step": 17558
+    },
+    {
+      "epoch": 0.9664263305630469,
+      "grad_norm": 0.5958573222160339,
+      "learning_rate": 5.28593189113249e-06,
+      "loss": 0.7685,
+      "step": 17559
+    },
+    {
+      "epoch": 0.9664813693654026,
+      "grad_norm": 0.6775822043418884,
+      "learning_rate": 5.285499132800941e-06,
+      "loss": 0.6828,
+      "step": 17560
+    },
+    {
+      "epoch": 0.9665364081677583,
+      "grad_norm": 0.7009457945823669,
+      "learning_rate": 5.285066372323655e-06,
+      "loss": 0.8319,
+      "step": 17561
+    },
+    {
+      "epoch": 0.9665914469701139,
+      "grad_norm": 0.6704908609390259,
+      "learning_rate": 5.284633609703887e-06,
+      "loss": 0.8294,
+      "step": 17562
+    },
+    {
+      "epoch": 0.9666464857724696,
+      "grad_norm": 0.6544950604438782,
+      "learning_rate": 5.284200844944891e-06,
+      "loss": 0.7558,
+      "step": 17563
+    },
+    {
+      "epoch": 0.9667015245748253,
+      "grad_norm": 0.6781743168830872,
+      "learning_rate": 5.283768078049915e-06,
+      "loss": 0.7093,
+      "step": 17564
+    },
+    {
+      "epoch": 0.966756563377181,
+      "grad_norm": 0.6010674238204956,
+      "learning_rate": 5.283335309022216e-06,
+      "loss": 0.6998,
+      "step": 17565
+    },
+    {
+      "epoch": 0.9668116021795365,
+      "grad_norm": 0.6528247594833374,
+      "learning_rate": 5.282902537865044e-06,
+      "loss": 0.7307,
+      "step": 17566
+    },
+    {
+      "epoch": 0.9668666409818922,
+      "grad_norm": 0.6252716779708862,
+      "learning_rate": 5.282469764581652e-06,
+      "loss": 0.6509,
+      "step": 17567
+    },
+    {
+      "epoch": 0.9669216797842479,
+      "grad_norm": 0.6801924705505371,
+      "learning_rate": 5.282036989175294e-06,
+      "loss": 0.7988,
+      "step": 17568
+    },
+    {
+      "epoch": 0.9669767185866035,
+      "grad_norm": 0.6112244129180908,
+      "learning_rate": 5.281604211649219e-06,
+      "loss": 0.6524,
+      "step": 17569
+    },
+    {
+      "epoch": 0.9670317573889592,
+      "grad_norm": 0.7375578284263611,
+      "learning_rate": 5.281171432006684e-06,
+      "loss": 0.7292,
+      "step": 17570
+    },
+    {
+      "epoch": 0.9670867961913149,
+      "grad_norm": 0.60760897397995,
+      "learning_rate": 5.28073865025094e-06,
+      "loss": 0.6621,
+      "step": 17571
+    },
+    {
+      "epoch": 0.9671418349936706,
+      "grad_norm": 0.6792896389961243,
+      "learning_rate": 5.280305866385239e-06,
+      "loss": 0.7818,
+      "step": 17572
+    },
+    {
+      "epoch": 0.9671968737960261,
+      "grad_norm": 0.7662883400917053,
+      "learning_rate": 5.279873080412833e-06,
+      "loss": 0.8246,
+      "step": 17573
+    },
+    {
+      "epoch": 0.9672519125983818,
+      "grad_norm": 0.7479282021522522,
+      "learning_rate": 5.279440292336977e-06,
+      "loss": 0.7974,
+      "step": 17574
+    },
+    {
+      "epoch": 0.9673069514007375,
+      "grad_norm": 0.8267967104911804,
+      "learning_rate": 5.279007502160922e-06,
+      "loss": 0.6963,
+      "step": 17575
+    },
+    {
+      "epoch": 0.9673619902030932,
+      "grad_norm": 0.7001194357872009,
+      "learning_rate": 5.278574709887923e-06,
+      "loss": 0.6533,
+      "step": 17576
+    },
+    {
+      "epoch": 0.9674170290054488,
+      "grad_norm": 0.6630476117134094,
+      "learning_rate": 5.27814191552123e-06,
+      "loss": 0.7644,
+      "step": 17577
+    },
+    {
+      "epoch": 0.9674720678078045,
+      "grad_norm": 0.6703379154205322,
+      "learning_rate": 5.277709119064096e-06,
+      "loss": 0.7285,
+      "step": 17578
+    },
+    {
+      "epoch": 0.9675271066101602,
+      "grad_norm": 0.7047588229179382,
+      "learning_rate": 5.277276320519775e-06,
+      "loss": 0.7993,
+      "step": 17579
+    },
+    {
+      "epoch": 0.9675821454125159,
+      "grad_norm": 0.7320923209190369,
+      "learning_rate": 5.27684351989152e-06,
+      "loss": 0.8014,
+      "step": 17580
+    },
+    {
+      "epoch": 0.9676371842148714,
+      "grad_norm": 0.6852779388427734,
+      "learning_rate": 5.276410717182582e-06,
+      "loss": 0.7805,
+      "step": 17581
+    },
+    {
+      "epoch": 0.9676922230172271,
+      "grad_norm": 0.7098143696784973,
+      "learning_rate": 5.275977912396216e-06,
+      "loss": 0.7196,
+      "step": 17582
+    },
+    {
+      "epoch": 0.9677472618195828,
+      "grad_norm": 0.6526379585266113,
+      "learning_rate": 5.275545105535674e-06,
+      "loss": 0.6948,
+      "step": 17583
+    },
+    {
+      "epoch": 0.9678023006219385,
+      "grad_norm": 0.6695026159286499,
+      "learning_rate": 5.275112296604209e-06,
+      "loss": 0.7167,
+      "step": 17584
+    },
+    {
+      "epoch": 0.9678573394242941,
+      "grad_norm": 0.825526773929596,
+      "learning_rate": 5.274679485605072e-06,
+      "loss": 0.6848,
+      "step": 17585
+    },
+    {
+      "epoch": 0.9679123782266498,
+      "grad_norm": 0.7333941459655762,
+      "learning_rate": 5.2742466725415185e-06,
+      "loss": 0.7888,
+      "step": 17586
+    },
+    {
+      "epoch": 0.9679674170290055,
+      "grad_norm": 0.6386739611625671,
+      "learning_rate": 5.2738138574168e-06,
+      "loss": 0.6829,
+      "step": 17587
+    },
+    {
+      "epoch": 0.9680224558313612,
+      "grad_norm": 0.7038518190383911,
+      "learning_rate": 5.273381040234172e-06,
+      "loss": 0.8408,
+      "step": 17588
+    },
+    {
+      "epoch": 0.9680774946337167,
+      "grad_norm": 0.6579585075378418,
+      "learning_rate": 5.272948220996883e-06,
+      "loss": 0.7605,
+      "step": 17589
+    },
+    {
+      "epoch": 0.9681325334360724,
+      "grad_norm": 0.614625096321106,
+      "learning_rate": 5.272515399708189e-06,
+      "loss": 0.6965,
+      "step": 17590
+    },
+    {
+      "epoch": 0.9681875722384281,
+      "grad_norm": 0.6300467848777771,
+      "learning_rate": 5.272082576371342e-06,
+      "loss": 0.7148,
+      "step": 17591
+    },
+    {
+      "epoch": 0.9682426110407838,
+      "grad_norm": 0.634800136089325,
+      "learning_rate": 5.271649750989596e-06,
+      "loss": 0.7119,
+      "step": 17592
+    },
+    {
+      "epoch": 0.9682976498431394,
+      "grad_norm": 0.7334730625152588,
+      "learning_rate": 5.271216923566201e-06,
+      "loss": 0.7616,
+      "step": 17593
+    },
+    {
+      "epoch": 0.9683526886454951,
+      "grad_norm": 0.649853527545929,
+      "learning_rate": 5.270784094104414e-06,
+      "loss": 0.718,
+      "step": 17594
+    },
+    {
+      "epoch": 0.9684077274478508,
+      "grad_norm": 0.7468090653419495,
+      "learning_rate": 5.270351262607486e-06,
+      "loss": 0.723,
+      "step": 17595
+    },
+    {
+      "epoch": 0.9684627662502064,
+      "grad_norm": 0.8167023658752441,
+      "learning_rate": 5.269918429078671e-06,
+      "loss": 0.7052,
+      "step": 17596
+    },
+    {
+      "epoch": 0.968517805052562,
+      "grad_norm": 0.6515017747879028,
+      "learning_rate": 5.26948559352122e-06,
+      "loss": 0.7679,
+      "step": 17597
+    },
+    {
+      "epoch": 0.9685728438549177,
+      "grad_norm": 0.6585719585418701,
+      "learning_rate": 5.269052755938388e-06,
+      "loss": 0.7396,
+      "step": 17598
+    },
+    {
+      "epoch": 0.9686278826572734,
+      "grad_norm": 0.6206550598144531,
+      "learning_rate": 5.268619916333427e-06,
+      "loss": 0.7786,
+      "step": 17599
+    },
+    {
+      "epoch": 0.9686829214596291,
+      "grad_norm": 0.6914154291152954,
+      "learning_rate": 5.268187074709591e-06,
+      "loss": 0.6792,
+      "step": 17600
+    },
+    {
+      "epoch": 0.9687379602619847,
+      "grad_norm": 0.619126558303833,
+      "learning_rate": 5.267754231070134e-06,
+      "loss": 0.7567,
+      "step": 17601
+    },
+    {
+      "epoch": 0.9687929990643404,
+      "grad_norm": 0.6083783507347107,
+      "learning_rate": 5.267321385418306e-06,
+      "loss": 0.6357,
+      "step": 17602
+    },
+    {
+      "epoch": 0.968848037866696,
+      "grad_norm": 0.6827127933502197,
+      "learning_rate": 5.266888537757363e-06,
+      "loss": 0.7282,
+      "step": 17603
+    },
+    {
+      "epoch": 0.9689030766690517,
+      "grad_norm": 0.7914609909057617,
+      "learning_rate": 5.266455688090557e-06,
+      "loss": 0.8448,
+      "step": 17604
+    },
+    {
+      "epoch": 0.9689581154714073,
+      "grad_norm": 0.6890830397605896,
+      "learning_rate": 5.26602283642114e-06,
+      "loss": 0.7391,
+      "step": 17605
+    },
+    {
+      "epoch": 0.969013154273763,
+      "grad_norm": 0.7047440409660339,
+      "learning_rate": 5.265589982752368e-06,
+      "loss": 0.6348,
+      "step": 17606
+    },
+    {
+      "epoch": 0.9690681930761187,
+      "grad_norm": 0.7744457125663757,
+      "learning_rate": 5.265157127087493e-06,
+      "loss": 0.7639,
+      "step": 17607
+    },
+    {
+      "epoch": 0.9691232318784744,
+      "grad_norm": 0.6565386056900024,
+      "learning_rate": 5.264724269429767e-06,
+      "loss": 0.7674,
+      "step": 17608
+    },
+    {
+      "epoch": 0.96917827068083,
+      "grad_norm": 0.7823582887649536,
+      "learning_rate": 5.264291409782445e-06,
+      "loss": 0.7053,
+      "step": 17609
+    },
+    {
+      "epoch": 0.9692333094831856,
+      "grad_norm": 0.6871463656425476,
+      "learning_rate": 5.263858548148779e-06,
+      "loss": 0.6983,
+      "step": 17610
+    },
+    {
+      "epoch": 0.9692883482855413,
+      "grad_norm": 0.6161954402923584,
+      "learning_rate": 5.263425684532023e-06,
+      "loss": 0.7442,
+      "step": 17611
+    },
+    {
+      "epoch": 0.9693433870878969,
+      "grad_norm": 0.6763638854026794,
+      "learning_rate": 5.26299281893543e-06,
+      "loss": 0.7545,
+      "step": 17612
+    },
+    {
+      "epoch": 0.9693984258902526,
+      "grad_norm": 0.6834864020347595,
+      "learning_rate": 5.2625599513622534e-06,
+      "loss": 0.6359,
+      "step": 17613
+    },
+    {
+      "epoch": 0.9694534646926083,
+      "grad_norm": 0.8389854431152344,
+      "learning_rate": 5.262127081815748e-06,
+      "loss": 0.7517,
+      "step": 17614
+    },
+    {
+      "epoch": 0.969508503494964,
+      "grad_norm": 0.6721356511116028,
+      "learning_rate": 5.261694210299164e-06,
+      "loss": 0.6466,
+      "step": 17615
+    },
+    {
+      "epoch": 0.9695635422973196,
+      "grad_norm": 0.6551144123077393,
+      "learning_rate": 5.261261336815757e-06,
+      "loss": 0.7571,
+      "step": 17616
+    },
+    {
+      "epoch": 0.9696185810996752,
+      "grad_norm": 0.6491858959197998,
+      "learning_rate": 5.260828461368779e-06,
+      "loss": 0.7274,
+      "step": 17617
+    },
+    {
+      "epoch": 0.9696736199020309,
+      "grad_norm": 0.6955786347389221,
+      "learning_rate": 5.260395583961484e-06,
+      "loss": 0.7589,
+      "step": 17618
+    },
+    {
+      "epoch": 0.9697286587043866,
+      "grad_norm": 0.6789944171905518,
+      "learning_rate": 5.259962704597127e-06,
+      "loss": 0.7439,
+      "step": 17619
+    },
+    {
+      "epoch": 0.9697836975067422,
+      "grad_norm": 0.6398529410362244,
+      "learning_rate": 5.259529823278958e-06,
+      "loss": 0.6743,
+      "step": 17620
+    },
+    {
+      "epoch": 0.9698387363090979,
+      "grad_norm": 0.7643429040908813,
+      "learning_rate": 5.2590969400102335e-06,
+      "loss": 0.7176,
+      "step": 17621
+    },
+    {
+      "epoch": 0.9698937751114536,
+      "grad_norm": 0.6158431768417358,
+      "learning_rate": 5.258664054794205e-06,
+      "loss": 0.6963,
+      "step": 17622
+    },
+    {
+      "epoch": 0.9699488139138093,
+      "grad_norm": 0.644513726234436,
+      "learning_rate": 5.258231167634127e-06,
+      "loss": 0.7126,
+      "step": 17623
+    },
+    {
+      "epoch": 0.9700038527161648,
+      "grad_norm": 0.6312246918678284,
+      "learning_rate": 5.257798278533253e-06,
+      "loss": 0.7946,
+      "step": 17624
+    },
+    {
+      "epoch": 0.9700588915185205,
+      "grad_norm": 0.8247089385986328,
+      "learning_rate": 5.257365387494836e-06,
+      "loss": 0.8014,
+      "step": 17625
+    },
+    {
+      "epoch": 0.9701139303208762,
+      "grad_norm": 0.6787695288658142,
+      "learning_rate": 5.25693249452213e-06,
+      "loss": 0.775,
+      "step": 17626
+    },
+    {
+      "epoch": 0.9701689691232319,
+      "grad_norm": 0.675240695476532,
+      "learning_rate": 5.256499599618388e-06,
+      "loss": 0.7655,
+      "step": 17627
+    },
+    {
+      "epoch": 0.9702240079255875,
+      "grad_norm": 0.604271411895752,
+      "learning_rate": 5.256066702786864e-06,
+      "loss": 0.7278,
+      "step": 17628
+    },
+    {
+      "epoch": 0.9702790467279432,
+      "grad_norm": 0.7096530795097351,
+      "learning_rate": 5.2556338040308095e-06,
+      "loss": 0.8267,
+      "step": 17629
+    },
+    {
+      "epoch": 0.9703340855302989,
+      "grad_norm": 0.7375852465629578,
+      "learning_rate": 5.255200903353481e-06,
+      "loss": 0.7844,
+      "step": 17630
+    },
+    {
+      "epoch": 0.9703891243326546,
+      "grad_norm": 0.7275679707527161,
+      "learning_rate": 5.25476800075813e-06,
+      "loss": 0.7814,
+      "step": 17631
+    },
+    {
+      "epoch": 0.9704441631350101,
+      "grad_norm": 0.655462384223938,
+      "learning_rate": 5.254335096248012e-06,
+      "loss": 0.7401,
+      "step": 17632
+    },
+    {
+      "epoch": 0.9704992019373658,
+      "grad_norm": 0.6601149439811707,
+      "learning_rate": 5.253902189826379e-06,
+      "loss": 0.6787,
+      "step": 17633
+    },
+    {
+      "epoch": 0.9705542407397215,
+      "grad_norm": 0.7337445020675659,
+      "learning_rate": 5.253469281496485e-06,
+      "loss": 0.744,
+      "step": 17634
+    },
+    {
+      "epoch": 0.9706092795420772,
+      "grad_norm": 0.7127711772918701,
+      "learning_rate": 5.253036371261584e-06,
+      "loss": 0.6862,
+      "step": 17635
+    },
+    {
+      "epoch": 0.9706643183444328,
+      "grad_norm": 0.6720691919326782,
+      "learning_rate": 5.25260345912493e-06,
+      "loss": 0.6828,
+      "step": 17636
+    },
+    {
+      "epoch": 0.9707193571467885,
+      "grad_norm": 0.707231879234314,
+      "learning_rate": 5.252170545089775e-06,
+      "loss": 0.7162,
+      "step": 17637
+    },
+    {
+      "epoch": 0.9707743959491442,
+      "grad_norm": 0.561342179775238,
+      "learning_rate": 5.251737629159374e-06,
+      "loss": 0.558,
+      "step": 17638
+    },
+    {
+      "epoch": 0.9708294347514999,
+      "grad_norm": 0.6514197587966919,
+      "learning_rate": 5.251304711336981e-06,
+      "loss": 0.699,
+      "step": 17639
+    },
+    {
+      "epoch": 0.9708844735538554,
+      "grad_norm": 0.5802120566368103,
+      "learning_rate": 5.250871791625849e-06,
+      "loss": 0.5976,
+      "step": 17640
+    },
+    {
+      "epoch": 0.9709395123562111,
+      "grad_norm": 0.659600019454956,
+      "learning_rate": 5.25043887002923e-06,
+      "loss": 0.6831,
+      "step": 17641
+    },
+    {
+      "epoch": 0.9709945511585668,
+      "grad_norm": 0.7798045873641968,
+      "learning_rate": 5.250005946550381e-06,
+      "loss": 0.7274,
+      "step": 17642
+    },
+    {
+      "epoch": 0.9710495899609225,
+      "grad_norm": 0.6244889497756958,
+      "learning_rate": 5.249573021192553e-06,
+      "loss": 0.6973,
+      "step": 17643
+    },
+    {
+      "epoch": 0.9711046287632781,
+      "grad_norm": 0.7149257659912109,
+      "learning_rate": 5.249140093959003e-06,
+      "loss": 0.7197,
+      "step": 17644
+    },
+    {
+      "epoch": 0.9711596675656338,
+      "grad_norm": 0.7044658660888672,
+      "learning_rate": 5.248707164852981e-06,
+      "loss": 0.7254,
+      "step": 17645
+    },
+    {
+      "epoch": 0.9712147063679895,
+      "grad_norm": 0.8093852400779724,
+      "learning_rate": 5.248274233877742e-06,
+      "loss": 0.7622,
+      "step": 17646
+    },
+    {
+      "epoch": 0.9712697451703451,
+      "grad_norm": 0.6021914482116699,
+      "learning_rate": 5.2478413010365425e-06,
+      "loss": 0.6356,
+      "step": 17647
+    },
+    {
+      "epoch": 0.9713247839727007,
+      "grad_norm": 0.7819526791572571,
+      "learning_rate": 5.247408366332633e-06,
+      "loss": 0.6843,
+      "step": 17648
+    },
+    {
+      "epoch": 0.9713798227750564,
+      "grad_norm": 0.7436261773109436,
+      "learning_rate": 5.246975429769269e-06,
+      "loss": 0.7383,
+      "step": 17649
+    },
+    {
+      "epoch": 0.9714348615774121,
+      "grad_norm": 0.6402625441551208,
+      "learning_rate": 5.246542491349703e-06,
+      "loss": 0.7297,
+      "step": 17650
+    },
+    {
+      "epoch": 0.9714899003797678,
+      "grad_norm": 0.6183507442474365,
+      "learning_rate": 5.246109551077191e-06,
+      "loss": 0.6971,
+      "step": 17651
+    },
+    {
+      "epoch": 0.9715449391821234,
+      "grad_norm": 0.693608820438385,
+      "learning_rate": 5.245676608954985e-06,
+      "loss": 0.7626,
+      "step": 17652
+    },
+    {
+      "epoch": 0.9715999779844791,
+      "grad_norm": 0.6693928837776184,
+      "learning_rate": 5.245243664986337e-06,
+      "loss": 0.7379,
+      "step": 17653
+    },
+    {
+      "epoch": 0.9716550167868347,
+      "grad_norm": 0.6615895628929138,
+      "learning_rate": 5.2448107191745055e-06,
+      "loss": 0.6611,
+      "step": 17654
+    },
+    {
+      "epoch": 0.9717100555891903,
+      "grad_norm": 0.6616193056106567,
+      "learning_rate": 5.244377771522741e-06,
+      "loss": 0.8005,
+      "step": 17655
+    },
+    {
+      "epoch": 0.971765094391546,
+      "grad_norm": 0.7025266289710999,
+      "learning_rate": 5.243944822034301e-06,
+      "loss": 0.756,
+      "step": 17656
+    },
+    {
+      "epoch": 0.9718201331939017,
+      "grad_norm": 0.6651642322540283,
+      "learning_rate": 5.243511870712434e-06,
+      "loss": 0.7113,
+      "step": 17657
+    },
+    {
+      "epoch": 0.9718751719962574,
+      "grad_norm": 0.6264922618865967,
+      "learning_rate": 5.243078917560398e-06,
+      "loss": 0.7355,
+      "step": 17658
+    },
+    {
+      "epoch": 0.971930210798613,
+      "grad_norm": 0.6707214117050171,
+      "learning_rate": 5.2426459625814475e-06,
+      "loss": 0.7703,
+      "step": 17659
+    },
+    {
+      "epoch": 0.9719852496009687,
+      "grad_norm": 0.7076506614685059,
+      "learning_rate": 5.242213005778833e-06,
+      "loss": 0.7704,
+      "step": 17660
+    },
+    {
+      "epoch": 0.9720402884033243,
+      "grad_norm": 0.7259695529937744,
+      "learning_rate": 5.241780047155811e-06,
+      "loss": 0.6814,
+      "step": 17661
+    },
+    {
+      "epoch": 0.97209532720568,
+      "grad_norm": 0.6943030953407288,
+      "learning_rate": 5.241347086715636e-06,
+      "loss": 0.696,
+      "step": 17662
+    },
+    {
+      "epoch": 0.9721503660080356,
+      "grad_norm": 1.216528296470642,
+      "learning_rate": 5.240914124461559e-06,
+      "loss": 0.7686,
+      "step": 17663
+    },
+    {
+      "epoch": 0.9722054048103913,
+      "grad_norm": 0.6558219790458679,
+      "learning_rate": 5.240481160396838e-06,
+      "loss": 0.7056,
+      "step": 17664
+    },
+    {
+      "epoch": 0.972260443612747,
+      "grad_norm": 0.7323386669158936,
+      "learning_rate": 5.2400481945247224e-06,
+      "loss": 0.7147,
+      "step": 17665
+    },
+    {
+      "epoch": 0.9723154824151027,
+      "grad_norm": 0.7845921516418457,
+      "learning_rate": 5.23961522684847e-06,
+      "loss": 0.7745,
+      "step": 17666
+    },
+    {
+      "epoch": 0.9723705212174583,
+      "grad_norm": 0.6414793729782104,
+      "learning_rate": 5.239182257371335e-06,
+      "loss": 0.7127,
+      "step": 17667
+    },
+    {
+      "epoch": 0.972425560019814,
+      "grad_norm": 0.7045019268989563,
+      "learning_rate": 5.238749286096568e-06,
+      "loss": 0.7783,
+      "step": 17668
+    },
+    {
+      "epoch": 0.9724805988221696,
+      "grad_norm": 0.6710691452026367,
+      "learning_rate": 5.2383163130274274e-06,
+      "loss": 0.8069,
+      "step": 17669
+    },
+    {
+      "epoch": 0.9725356376245253,
+      "grad_norm": 0.7799990773200989,
+      "learning_rate": 5.2378833381671626e-06,
+      "loss": 0.7133,
+      "step": 17670
+    },
+    {
+      "epoch": 0.9725906764268809,
+      "grad_norm": 0.6748518943786621,
+      "learning_rate": 5.237450361519032e-06,
+      "loss": 0.7588,
+      "step": 17671
+    },
+    {
+      "epoch": 0.9726457152292366,
+      "grad_norm": 0.6174660325050354,
+      "learning_rate": 5.237017383086288e-06,
+      "loss": 0.6363,
+      "step": 17672
+    },
+    {
+      "epoch": 0.9727007540315923,
+      "grad_norm": 0.7074185013771057,
+      "learning_rate": 5.236584402872185e-06,
+      "loss": 0.7307,
+      "step": 17673
+    },
+    {
+      "epoch": 0.972755792833948,
+      "grad_norm": 0.6636162996292114,
+      "learning_rate": 5.236151420879977e-06,
+      "loss": 0.7476,
+      "step": 17674
+    },
+    {
+      "epoch": 0.9728108316363036,
+      "grad_norm": 0.8502483367919922,
+      "learning_rate": 5.2357184371129175e-06,
+      "loss": 0.8123,
+      "step": 17675
+    },
+    {
+      "epoch": 0.9728658704386592,
+      "grad_norm": 0.7215339541435242,
+      "learning_rate": 5.235285451574262e-06,
+      "loss": 0.6557,
+      "step": 17676
+    },
+    {
+      "epoch": 0.9729209092410149,
+      "grad_norm": 0.6733965277671814,
+      "learning_rate": 5.2348524642672635e-06,
+      "loss": 0.7316,
+      "step": 17677
+    },
+    {
+      "epoch": 0.9729759480433706,
+      "grad_norm": 0.6186164617538452,
+      "learning_rate": 5.234419475195176e-06,
+      "loss": 0.741,
+      "step": 17678
+    },
+    {
+      "epoch": 0.9730309868457262,
+      "grad_norm": 0.6414602994918823,
+      "learning_rate": 5.233986484361255e-06,
+      "loss": 0.7266,
+      "step": 17679
+    },
+    {
+      "epoch": 0.9730860256480819,
+      "grad_norm": 0.8314804434776306,
+      "learning_rate": 5.2335534917687545e-06,
+      "loss": 0.7368,
+      "step": 17680
+    },
+    {
+      "epoch": 0.9731410644504376,
+      "grad_norm": 0.6001927256584167,
+      "learning_rate": 5.23312049742093e-06,
+      "loss": 0.6717,
+      "step": 17681
+    },
+    {
+      "epoch": 0.9731961032527933,
+      "grad_norm": 0.7511568069458008,
+      "learning_rate": 5.232687501321031e-06,
+      "loss": 0.8037,
+      "step": 17682
+    },
+    {
+      "epoch": 0.9732511420551488,
+      "grad_norm": 0.6779069900512695,
+      "learning_rate": 5.232254503472317e-06,
+      "loss": 0.7341,
+      "step": 17683
+    },
+    {
+      "epoch": 0.9733061808575045,
+      "grad_norm": 1.6375025510787964,
+      "learning_rate": 5.23182150387804e-06,
+      "loss": 1.0429,
+      "step": 17684
+    },
+    {
+      "epoch": 0.9733612196598602,
+      "grad_norm": 0.7395290732383728,
+      "learning_rate": 5.231388502541454e-06,
+      "loss": 0.7704,
+      "step": 17685
+    },
+    {
+      "epoch": 0.9734162584622159,
+      "grad_norm": 0.6683037281036377,
+      "learning_rate": 5.230955499465815e-06,
+      "loss": 0.755,
+      "step": 17686
+    },
+    {
+      "epoch": 0.9734712972645715,
+      "grad_norm": 1.0014123916625977,
+      "learning_rate": 5.230522494654377e-06,
+      "loss": 0.902,
+      "step": 17687
+    },
+    {
+      "epoch": 0.9735263360669272,
+      "grad_norm": 0.8084937334060669,
+      "learning_rate": 5.230089488110392e-06,
+      "loss": 0.7252,
+      "step": 17688
+    },
+    {
+      "epoch": 0.9735813748692829,
+      "grad_norm": 0.6769001483917236,
+      "learning_rate": 5.229656479837117e-06,
+      "loss": 0.7287,
+      "step": 17689
+    },
+    {
+      "epoch": 0.9736364136716386,
+      "grad_norm": 0.6600124835968018,
+      "learning_rate": 5.229223469837804e-06,
+      "loss": 0.7659,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9736914524739941,
+      "grad_norm": 0.7093310952186584,
+      "learning_rate": 5.22879045811571e-06,
+      "loss": 0.83,
+      "step": 17691
+    },
+    {
+      "epoch": 0.9737464912763498,
+      "grad_norm": 0.6043664216995239,
+      "learning_rate": 5.228357444674088e-06,
+      "loss": 0.6321,
+      "step": 17692
+    },
+    {
+      "epoch": 0.9738015300787055,
+      "grad_norm": 0.7609572410583496,
+      "learning_rate": 5.227924429516192e-06,
+      "loss": 0.7391,
+      "step": 17693
+    },
+    {
+      "epoch": 0.9738565688810612,
+      "grad_norm": 0.6878551244735718,
+      "learning_rate": 5.227491412645277e-06,
+      "loss": 0.7947,
+      "step": 17694
+    },
+    {
+      "epoch": 0.9739116076834168,
+      "grad_norm": 0.6588627099990845,
+      "learning_rate": 5.227058394064598e-06,
+      "loss": 0.6916,
+      "step": 17695
+    },
+    {
+      "epoch": 0.9739666464857725,
+      "grad_norm": 0.6774247288703918,
+      "learning_rate": 5.22662537377741e-06,
+      "loss": 0.6332,
+      "step": 17696
+    },
+    {
+      "epoch": 0.9740216852881282,
+      "grad_norm": 0.9965986013412476,
+      "learning_rate": 5.226192351786965e-06,
+      "loss": 0.7915,
+      "step": 17697
+    },
+    {
+      "epoch": 0.9740767240904837,
+      "grad_norm": 0.6406721472740173,
+      "learning_rate": 5.225759328096519e-06,
+      "loss": 0.7314,
+      "step": 17698
+    },
+    {
+      "epoch": 0.9741317628928394,
+      "grad_norm": 0.8169941306114197,
+      "learning_rate": 5.225326302709326e-06,
+      "loss": 0.8155,
+      "step": 17699
+    },
+    {
+      "epoch": 0.9741868016951951,
+      "grad_norm": 0.6238623857498169,
+      "learning_rate": 5.224893275628642e-06,
+      "loss": 0.7004,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9742418404975508,
+      "grad_norm": 0.7599614858627319,
+      "learning_rate": 5.224460246857719e-06,
+      "loss": 0.7761,
+      "step": 17701
+    },
+    {
+      "epoch": 0.9742968792999064,
+      "grad_norm": 0.5887364745140076,
+      "learning_rate": 5.224027216399812e-06,
+      "loss": 0.6816,
+      "step": 17702
+    },
+    {
+      "epoch": 0.9743519181022621,
+      "grad_norm": 0.8065565228462219,
+      "learning_rate": 5.2235941842581785e-06,
+      "loss": 0.8206,
+      "step": 17703
+    },
+    {
+      "epoch": 0.9744069569046178,
+      "grad_norm": 0.7422066926956177,
+      "learning_rate": 5.2231611504360705e-06,
+      "loss": 0.8321,
+      "step": 17704
+    },
+    {
+      "epoch": 0.9744619957069734,
+      "grad_norm": 0.6249799132347107,
+      "learning_rate": 5.222728114936742e-06,
+      "loss": 0.7404,
+      "step": 17705
+    },
+    {
+      "epoch": 0.974517034509329,
+      "grad_norm": 0.5743978023529053,
+      "learning_rate": 5.222295077763448e-06,
+      "loss": 0.663,
+      "step": 17706
+    },
+    {
+      "epoch": 0.9745720733116847,
+      "grad_norm": 0.6484057307243347,
+      "learning_rate": 5.221862038919446e-06,
+      "loss": 0.7183,
+      "step": 17707
+    },
+    {
+      "epoch": 0.9746271121140404,
+      "grad_norm": 0.7640897631645203,
+      "learning_rate": 5.221428998407988e-06,
+      "loss": 0.7929,
+      "step": 17708
+    },
+    {
+      "epoch": 0.9746821509163961,
+      "grad_norm": 0.9045720100402832,
+      "learning_rate": 5.220995956232327e-06,
+      "loss": 0.8641,
+      "step": 17709
+    },
+    {
+      "epoch": 0.9747371897187517,
+      "grad_norm": 0.6404321789741516,
+      "learning_rate": 5.2205629123957225e-06,
+      "loss": 0.7055,
+      "step": 17710
+    },
+    {
+      "epoch": 0.9747922285211074,
+      "grad_norm": 0.5746439099311829,
+      "learning_rate": 5.220129866901424e-06,
+      "loss": 0.6194,
+      "step": 17711
+    },
+    {
+      "epoch": 0.974847267323463,
+      "grad_norm": 0.7163496017456055,
+      "learning_rate": 5.2196968197526885e-06,
+      "loss": 0.7532,
+      "step": 17712
+    },
+    {
+      "epoch": 0.9749023061258187,
+      "grad_norm": 0.6287344098091125,
+      "learning_rate": 5.219263770952771e-06,
+      "loss": 0.6701,
+      "step": 17713
+    },
+    {
+      "epoch": 0.9749573449281743,
+      "grad_norm": 0.6174989938735962,
+      "learning_rate": 5.218830720504925e-06,
+      "loss": 0.7362,
+      "step": 17714
+    },
+    {
+      "epoch": 0.97501238373053,
+      "grad_norm": 0.6240310668945312,
+      "learning_rate": 5.218397668412407e-06,
+      "loss": 0.6776,
+      "step": 17715
+    },
+    {
+      "epoch": 0.9750674225328857,
+      "grad_norm": 0.699640154838562,
+      "learning_rate": 5.2179646146784704e-06,
+      "loss": 0.7044,
+      "step": 17716
+    },
+    {
+      "epoch": 0.9751224613352414,
+      "grad_norm": 0.9073681235313416,
+      "learning_rate": 5.2175315593063695e-06,
+      "loss": 0.7662,
+      "step": 17717
+    },
+    {
+      "epoch": 0.975177500137597,
+      "grad_norm": 0.7450598478317261,
+      "learning_rate": 5.2170985022993595e-06,
+      "loss": 0.7926,
+      "step": 17718
+    },
+    {
+      "epoch": 0.9752325389399527,
+      "grad_norm": 0.8336604833602905,
+      "learning_rate": 5.2166654436606955e-06,
+      "loss": 0.7843,
+      "step": 17719
+    },
+    {
+      "epoch": 0.9752875777423083,
+      "grad_norm": 0.7027361392974854,
+      "learning_rate": 5.216232383393633e-06,
+      "loss": 0.7175,
+      "step": 17720
+    },
+    {
+      "epoch": 0.975342616544664,
+      "grad_norm": 1.23917818069458,
+      "learning_rate": 5.215799321501424e-06,
+      "loss": 0.8413,
+      "step": 17721
+    },
+    {
+      "epoch": 0.9753976553470196,
+      "grad_norm": 0.6831318736076355,
+      "learning_rate": 5.215366257987326e-06,
+      "loss": 0.735,
+      "step": 17722
+    },
+    {
+      "epoch": 0.9754526941493753,
+      "grad_norm": 0.6283305883407593,
+      "learning_rate": 5.2149331928545935e-06,
+      "loss": 0.6441,
+      "step": 17723
+    },
+    {
+      "epoch": 0.975507732951731,
+      "grad_norm": 0.7063291072845459,
+      "learning_rate": 5.21450012610648e-06,
+      "loss": 0.6778,
+      "step": 17724
+    },
+    {
+      "epoch": 0.9755627717540867,
+      "grad_norm": 0.6691693067550659,
+      "learning_rate": 5.21406705774624e-06,
+      "loss": 0.7197,
+      "step": 17725
+    },
+    {
+      "epoch": 0.9756178105564423,
+      "grad_norm": 0.7046432495117188,
+      "learning_rate": 5.21363398777713e-06,
+      "loss": 0.7688,
+      "step": 17726
+    },
+    {
+      "epoch": 0.975672849358798,
+      "grad_norm": 0.836552619934082,
+      "learning_rate": 5.213200916202404e-06,
+      "loss": 0.6908,
+      "step": 17727
+    },
+    {
+      "epoch": 0.9757278881611536,
+      "grad_norm": 0.6683027744293213,
+      "learning_rate": 5.2127678430253185e-06,
+      "loss": 0.7254,
+      "step": 17728
+    },
+    {
+      "epoch": 0.9757829269635093,
+      "grad_norm": 0.7955693006515503,
+      "learning_rate": 5.212334768249125e-06,
+      "loss": 0.7707,
+      "step": 17729
+    },
+    {
+      "epoch": 0.9758379657658649,
+      "grad_norm": 0.7328525185585022,
+      "learning_rate": 5.21190169187708e-06,
+      "loss": 0.742,
+      "step": 17730
+    },
+    {
+      "epoch": 0.9758930045682206,
+      "grad_norm": 0.7649689316749573,
+      "learning_rate": 5.21146861391244e-06,
+      "loss": 0.8682,
+      "step": 17731
+    },
+    {
+      "epoch": 0.9759480433705763,
+      "grad_norm": 0.697281002998352,
+      "learning_rate": 5.2110355343584585e-06,
+      "loss": 0.7577,
+      "step": 17732
+    },
+    {
+      "epoch": 0.976003082172932,
+      "grad_norm": 0.736832320690155,
+      "learning_rate": 5.21060245321839e-06,
+      "loss": 0.6747,
+      "step": 17733
+    },
+    {
+      "epoch": 0.9760581209752875,
+      "grad_norm": 0.7473930716514587,
+      "learning_rate": 5.210169370495491e-06,
+      "loss": 0.781,
+      "step": 17734
+    },
+    {
+      "epoch": 0.9761131597776432,
+      "grad_norm": 0.6661664247512817,
+      "learning_rate": 5.209736286193013e-06,
+      "loss": 0.7436,
+      "step": 17735
+    },
+    {
+      "epoch": 0.9761681985799989,
+      "grad_norm": 0.6759940385818481,
+      "learning_rate": 5.209303200314215e-06,
+      "loss": 0.831,
+      "step": 17736
+    },
+    {
+      "epoch": 0.9762232373823546,
+      "grad_norm": 0.7469375133514404,
+      "learning_rate": 5.20887011286235e-06,
+      "loss": 0.7741,
+      "step": 17737
+    },
+    {
+      "epoch": 0.9762782761847102,
+      "grad_norm": 0.7340208888053894,
+      "learning_rate": 5.208437023840671e-06,
+      "loss": 0.8867,
+      "step": 17738
+    },
+    {
+      "epoch": 0.9763333149870659,
+      "grad_norm": 0.8390399217605591,
+      "learning_rate": 5.208003933252437e-06,
+      "loss": 0.7465,
+      "step": 17739
+    },
+    {
+      "epoch": 0.9763883537894216,
+      "grad_norm": 0.6884977221488953,
+      "learning_rate": 5.207570841100901e-06,
+      "loss": 0.8155,
+      "step": 17740
+    },
+    {
+      "epoch": 0.9764433925917771,
+      "grad_norm": 0.7563072443008423,
+      "learning_rate": 5.207137747389318e-06,
+      "loss": 0.6696,
+      "step": 17741
+    },
+    {
+      "epoch": 0.9764984313941328,
+      "grad_norm": 0.5912995934486389,
+      "learning_rate": 5.206704652120942e-06,
+      "loss": 0.6378,
+      "step": 17742
+    },
+    {
+      "epoch": 0.9765534701964885,
+      "grad_norm": 0.6421001553535461,
+      "learning_rate": 5.2062715552990304e-06,
+      "loss": 0.7501,
+      "step": 17743
+    },
+    {
+      "epoch": 0.9766085089988442,
+      "grad_norm": 0.7305091023445129,
+      "learning_rate": 5.205838456926837e-06,
+      "loss": 0.7331,
+      "step": 17744
+    },
+    {
+      "epoch": 0.9766635478011998,
+      "grad_norm": 0.7133939862251282,
+      "learning_rate": 5.2054053570076165e-06,
+      "loss": 0.7701,
+      "step": 17745
+    },
+    {
+      "epoch": 0.9767185866035555,
+      "grad_norm": 0.7652276754379272,
+      "learning_rate": 5.204972255544626e-06,
+      "loss": 0.72,
+      "step": 17746
+    },
+    {
+      "epoch": 0.9767736254059112,
+      "grad_norm": 0.6148221492767334,
+      "learning_rate": 5.204539152541117e-06,
+      "loss": 0.6939,
+      "step": 17747
+    },
+    {
+      "epoch": 0.9768286642082669,
+      "grad_norm": 0.6623834371566772,
+      "learning_rate": 5.204106048000347e-06,
+      "loss": 0.7687,
+      "step": 17748
+    },
+    {
+      "epoch": 0.9768837030106224,
+      "grad_norm": 0.68714439868927,
+      "learning_rate": 5.2036729419255705e-06,
+      "loss": 0.8574,
+      "step": 17749
+    },
+    {
+      "epoch": 0.9769387418129781,
+      "grad_norm": 0.7681630849838257,
+      "learning_rate": 5.203239834320042e-06,
+      "loss": 0.8091,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9769937806153338,
+      "grad_norm": 0.7655820250511169,
+      "learning_rate": 5.202806725187018e-06,
+      "loss": 0.7951,
+      "step": 17751
+    },
+    {
+      "epoch": 0.9770488194176895,
+      "grad_norm": 0.6182354092597961,
+      "learning_rate": 5.202373614529754e-06,
+      "loss": 0.6923,
+      "step": 17752
+    },
+    {
+      "epoch": 0.9771038582200451,
+      "grad_norm": 0.6287481188774109,
+      "learning_rate": 5.2019405023515024e-06,
+      "loss": 0.7334,
+      "step": 17753
+    },
+    {
+      "epoch": 0.9771588970224008,
+      "grad_norm": 0.7055238485336304,
+      "learning_rate": 5.20150738865552e-06,
+      "loss": 0.8105,
+      "step": 17754
+    },
+    {
+      "epoch": 0.9772139358247565,
+      "grad_norm": 0.8190330862998962,
+      "learning_rate": 5.201074273445063e-06,
+      "loss": 0.8578,
+      "step": 17755
+    },
+    {
+      "epoch": 0.9772689746271122,
+      "grad_norm": 0.6595628261566162,
+      "learning_rate": 5.200641156723385e-06,
+      "loss": 0.6637,
+      "step": 17756
+    },
+    {
+      "epoch": 0.9773240134294677,
+      "grad_norm": 0.742751955986023,
+      "learning_rate": 5.200208038493743e-06,
+      "loss": 0.7245,
+      "step": 17757
+    },
+    {
+      "epoch": 0.9773790522318234,
+      "grad_norm": 0.6058782935142517,
+      "learning_rate": 5.199774918759389e-06,
+      "loss": 0.6248,
+      "step": 17758
+    },
+    {
+      "epoch": 0.9774340910341791,
+      "grad_norm": 0.9763246178627014,
+      "learning_rate": 5.199341797523582e-06,
+      "loss": 0.7447,
+      "step": 17759
+    },
+    {
+      "epoch": 0.9774891298365348,
+      "grad_norm": 0.6947417259216309,
+      "learning_rate": 5.198908674789575e-06,
+      "loss": 0.7376,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9775441686388904,
+      "grad_norm": 0.7201378345489502,
+      "learning_rate": 5.198475550560623e-06,
+      "loss": 0.7177,
+      "step": 17761
+    },
+    {
+      "epoch": 0.9775992074412461,
+      "grad_norm": 0.6808243989944458,
+      "learning_rate": 5.198042424839981e-06,
+      "loss": 0.7674,
+      "step": 17762
+    },
+    {
+      "epoch": 0.9776542462436018,
+      "grad_norm": 0.7210538983345032,
+      "learning_rate": 5.197609297630907e-06,
+      "loss": 0.744,
+      "step": 17763
+    },
+    {
+      "epoch": 0.9777092850459574,
+      "grad_norm": 0.6032211780548096,
+      "learning_rate": 5.197176168936653e-06,
+      "loss": 0.7356,
+      "step": 17764
+    },
+    {
+      "epoch": 0.977764323848313,
+      "grad_norm": 0.6586793661117554,
+      "learning_rate": 5.1967430387604766e-06,
+      "loss": 0.8117,
+      "step": 17765
+    },
+    {
+      "epoch": 0.9778193626506687,
+      "grad_norm": 0.6356725692749023,
+      "learning_rate": 5.196309907105631e-06,
+      "loss": 0.6829,
+      "step": 17766
+    },
+    {
+      "epoch": 0.9778744014530244,
+      "grad_norm": 0.6270668506622314,
+      "learning_rate": 5.1958767739753745e-06,
+      "loss": 0.6983,
+      "step": 17767
+    },
+    {
+      "epoch": 0.9779294402553801,
+      "grad_norm": 0.7375897169113159,
+      "learning_rate": 5.19544363937296e-06,
+      "loss": 0.7163,
+      "step": 17768
+    },
+    {
+      "epoch": 0.9779844790577357,
+      "grad_norm": 0.8138049244880676,
+      "learning_rate": 5.195010503301643e-06,
+      "loss": 0.7936,
+      "step": 17769
+    },
+    {
+      "epoch": 0.9780395178600914,
+      "grad_norm": 0.7183161377906799,
+      "learning_rate": 5.1945773657646795e-06,
+      "loss": 0.7003,
+      "step": 17770
+    },
+    {
+      "epoch": 0.978094556662447,
+      "grad_norm": 0.6597039699554443,
+      "learning_rate": 5.194144226765325e-06,
+      "loss": 0.6303,
+      "step": 17771
+    },
+    {
+      "epoch": 0.9781495954648027,
+      "grad_norm": 0.6552296876907349,
+      "learning_rate": 5.193711086306834e-06,
+      "loss": 0.7564,
+      "step": 17772
+    },
+    {
+      "epoch": 0.9782046342671583,
+      "grad_norm": 0.7361578941345215,
+      "learning_rate": 5.193277944392463e-06,
+      "loss": 0.8041,
+      "step": 17773
+    },
+    {
+      "epoch": 0.978259673069514,
+      "grad_norm": 0.6351678967475891,
+      "learning_rate": 5.192844801025465e-06,
+      "loss": 0.6201,
+      "step": 17774
+    },
+    {
+      "epoch": 0.9783147118718697,
+      "grad_norm": 0.6355697512626648,
+      "learning_rate": 5.192411656209097e-06,
+      "loss": 0.7349,
+      "step": 17775
+    },
+    {
+      "epoch": 0.9783697506742254,
+      "grad_norm": 0.6302974820137024,
+      "learning_rate": 5.191978509946617e-06,
+      "loss": 0.6817,
+      "step": 17776
+    },
+    {
+      "epoch": 0.978424789476581,
+      "grad_norm": 0.6403247117996216,
+      "learning_rate": 5.191545362241276e-06,
+      "loss": 0.7492,
+      "step": 17777
+    },
+    {
+      "epoch": 0.9784798282789366,
+      "grad_norm": 0.6228166818618774,
+      "learning_rate": 5.191112213096331e-06,
+      "loss": 0.7064,
+      "step": 17778
+    },
+    {
+      "epoch": 0.9785348670812923,
+      "grad_norm": 0.6551523804664612,
+      "learning_rate": 5.190679062515039e-06,
+      "loss": 0.8182,
+      "step": 17779
+    },
+    {
+      "epoch": 0.978589905883648,
+      "grad_norm": 0.6780593991279602,
+      "learning_rate": 5.190245910500654e-06,
+      "loss": 0.6726,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9786449446860036,
+      "grad_norm": 0.7101402282714844,
+      "learning_rate": 5.189812757056431e-06,
+      "loss": 0.7933,
+      "step": 17781
+    },
+    {
+      "epoch": 0.9786999834883593,
+      "grad_norm": 0.9489981532096863,
+      "learning_rate": 5.189379602185628e-06,
+      "loss": 0.7285,
+      "step": 17782
+    },
+    {
+      "epoch": 0.978755022290715,
+      "grad_norm": 0.7283034324645996,
+      "learning_rate": 5.188946445891497e-06,
+      "loss": 0.7448,
+      "step": 17783
+    },
+    {
+      "epoch": 0.9788100610930706,
+      "grad_norm": 0.7265629768371582,
+      "learning_rate": 5.188513288177296e-06,
+      "loss": 0.7444,
+      "step": 17784
+    },
+    {
+      "epoch": 0.9788650998954262,
+      "grad_norm": 0.7622523903846741,
+      "learning_rate": 5.188080129046279e-06,
+      "loss": 0.7651,
+      "step": 17785
+    },
+    {
+      "epoch": 0.9789201386977819,
+      "grad_norm": 0.6390808820724487,
+      "learning_rate": 5.1876469685017e-06,
+      "loss": 0.7189,
+      "step": 17786
+    },
+    {
+      "epoch": 0.9789751775001376,
+      "grad_norm": 0.6932275295257568,
+      "learning_rate": 5.1872138065468195e-06,
+      "loss": 0.7565,
+      "step": 17787
+    },
+    {
+      "epoch": 0.9790302163024932,
+      "grad_norm": 0.5994375348091125,
+      "learning_rate": 5.18678064318489e-06,
+      "loss": 0.6513,
+      "step": 17788
+    },
+    {
+      "epoch": 0.9790852551048489,
+      "grad_norm": 0.687844455242157,
+      "learning_rate": 5.1863474784191655e-06,
+      "loss": 0.7315,
+      "step": 17789
+    },
+    {
+      "epoch": 0.9791402939072046,
+      "grad_norm": 0.6444111466407776,
+      "learning_rate": 5.1859143122529046e-06,
+      "loss": 0.8145,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9791953327095603,
+      "grad_norm": 0.6670806407928467,
+      "learning_rate": 5.18548114468936e-06,
+      "loss": 0.7711,
+      "step": 17791
+    },
+    {
+      "epoch": 0.9792503715119159,
+      "grad_norm": 0.7144771814346313,
+      "learning_rate": 5.18504797573179e-06,
+      "loss": 0.7265,
+      "step": 17792
+    },
+    {
+      "epoch": 0.9793054103142715,
+      "grad_norm": 0.7024721503257751,
+      "learning_rate": 5.18461480538345e-06,
+      "loss": 0.7601,
+      "step": 17793
+    },
+    {
+      "epoch": 0.9793604491166272,
+      "grad_norm": 0.7342267632484436,
+      "learning_rate": 5.184181633647592e-06,
+      "loss": 0.8719,
+      "step": 17794
+    },
+    {
+      "epoch": 0.9794154879189829,
+      "grad_norm": 0.6706385016441345,
+      "learning_rate": 5.183748460527477e-06,
+      "loss": 0.7708,
+      "step": 17795
+    },
+    {
+      "epoch": 0.9794705267213385,
+      "grad_norm": 0.6513368487358093,
+      "learning_rate": 5.183315286026356e-06,
+      "loss": 0.7379,
+      "step": 17796
+    },
+    {
+      "epoch": 0.9795255655236942,
+      "grad_norm": 0.698735773563385,
+      "learning_rate": 5.182882110147487e-06,
+      "loss": 0.8,
+      "step": 17797
+    },
+    {
+      "epoch": 0.9795806043260499,
+      "grad_norm": 0.6671596765518188,
+      "learning_rate": 5.182448932894123e-06,
+      "loss": 0.7867,
+      "step": 17798
+    },
+    {
+      "epoch": 0.9796356431284056,
+      "grad_norm": 0.6650397181510925,
+      "learning_rate": 5.182015754269524e-06,
+      "loss": 0.6933,
+      "step": 17799
+    },
+    {
+      "epoch": 0.9796906819307611,
+      "grad_norm": 0.6519697904586792,
+      "learning_rate": 5.181582574276942e-06,
+      "loss": 0.6558,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9797457207331168,
+      "grad_norm": 0.7513225078582764,
+      "learning_rate": 5.181149392919635e-06,
+      "loss": 0.7471,
+      "step": 17801
+    },
+    {
+      "epoch": 0.9798007595354725,
+      "grad_norm": 0.8155935406684875,
+      "learning_rate": 5.1807162102008564e-06,
+      "loss": 0.7971,
+      "step": 17802
+    },
+    {
+      "epoch": 0.9798557983378282,
+      "grad_norm": 0.6526888012886047,
+      "learning_rate": 5.180283026123864e-06,
+      "loss": 0.7107,
+      "step": 17803
+    },
+    {
+      "epoch": 0.9799108371401838,
+      "grad_norm": 0.699776291847229,
+      "learning_rate": 5.179849840691913e-06,
+      "loss": 0.7673,
+      "step": 17804
+    },
+    {
+      "epoch": 0.9799658759425395,
+      "grad_norm": 0.6678556799888611,
+      "learning_rate": 5.179416653908259e-06,
+      "loss": 0.7455,
+      "step": 17805
+    },
+    {
+      "epoch": 0.9800209147448952,
+      "grad_norm": 0.7483236789703369,
+      "learning_rate": 5.178983465776156e-06,
+      "loss": 0.7547,
+      "step": 17806
+    },
+    {
+      "epoch": 0.9800759535472509,
+      "grad_norm": 0.6862025260925293,
+      "learning_rate": 5.178550276298862e-06,
+      "loss": 0.725,
+      "step": 17807
+    },
+    {
+      "epoch": 0.9801309923496064,
+      "grad_norm": 0.7763676643371582,
+      "learning_rate": 5.178117085479633e-06,
+      "loss": 0.8345,
+      "step": 17808
+    },
+    {
+      "epoch": 0.9801860311519621,
+      "grad_norm": 0.670272707939148,
+      "learning_rate": 5.177683893321722e-06,
+      "loss": 0.7093,
+      "step": 17809
+    },
+    {
+      "epoch": 0.9802410699543178,
+      "grad_norm": 0.651607871055603,
+      "learning_rate": 5.177250699828388e-06,
+      "loss": 0.6746,
+      "step": 17810
+    },
+    {
+      "epoch": 0.9802961087566735,
+      "grad_norm": 0.6883462071418762,
+      "learning_rate": 5.176817505002884e-06,
+      "loss": 0.7514,
+      "step": 17811
+    },
+    {
+      "epoch": 0.9803511475590291,
+      "grad_norm": 0.6640946269035339,
+      "learning_rate": 5.1763843088484665e-06,
+      "loss": 0.7558,
+      "step": 17812
+    },
+    {
+      "epoch": 0.9804061863613848,
+      "grad_norm": 0.7128181457519531,
+      "learning_rate": 5.175951111368393e-06,
+      "loss": 0.7759,
+      "step": 17813
+    },
+    {
+      "epoch": 0.9804612251637405,
+      "grad_norm": 0.6555806398391724,
+      "learning_rate": 5.1755179125659175e-06,
+      "loss": 0.7697,
+      "step": 17814
+    },
+    {
+      "epoch": 0.9805162639660961,
+      "grad_norm": 0.7027010917663574,
+      "learning_rate": 5.175084712444295e-06,
+      "loss": 0.7325,
+      "step": 17815
+    },
+    {
+      "epoch": 0.9805713027684517,
+      "grad_norm": 0.6129525303840637,
+      "learning_rate": 5.1746515110067845e-06,
+      "loss": 0.7128,
+      "step": 17816
+    },
+    {
+      "epoch": 0.9806263415708074,
+      "grad_norm": 0.719196081161499,
+      "learning_rate": 5.1742183082566386e-06,
+      "loss": 0.8094,
+      "step": 17817
+    },
+    {
+      "epoch": 0.9806813803731631,
+      "grad_norm": 0.6782718896865845,
+      "learning_rate": 5.173785104197115e-06,
+      "loss": 0.7794,
+      "step": 17818
+    },
+    {
+      "epoch": 0.9807364191755188,
+      "grad_norm": 0.6790881752967834,
+      "learning_rate": 5.1733518988314705e-06,
+      "loss": 0.7603,
+      "step": 17819
+    },
+    {
+      "epoch": 0.9807914579778744,
+      "grad_norm": 0.691223680973053,
+      "learning_rate": 5.1729186921629595e-06,
+      "loss": 0.8255,
+      "step": 17820
+    },
+    {
+      "epoch": 0.9808464967802301,
+      "grad_norm": 0.6151034235954285,
+      "learning_rate": 5.172485484194836e-06,
+      "loss": 0.7015,
+      "step": 17821
+    },
+    {
+      "epoch": 0.9809015355825857,
+      "grad_norm": 0.6329746246337891,
+      "learning_rate": 5.172052274930359e-06,
+      "loss": 0.6421,
+      "step": 17822
+    },
+    {
+      "epoch": 0.9809565743849414,
+      "grad_norm": 0.6794832348823547,
+      "learning_rate": 5.171619064372781e-06,
+      "loss": 0.786,
+      "step": 17823
+    },
+    {
+      "epoch": 0.981011613187297,
+      "grad_norm": 0.7153908014297485,
+      "learning_rate": 5.171185852525363e-06,
+      "loss": 0.639,
+      "step": 17824
+    },
+    {
+      "epoch": 0.9810666519896527,
+      "grad_norm": 0.6464501619338989,
+      "learning_rate": 5.170752639391357e-06,
+      "loss": 0.7714,
+      "step": 17825
+    },
+    {
+      "epoch": 0.9811216907920084,
+      "grad_norm": 0.8912076950073242,
+      "learning_rate": 5.17031942497402e-06,
+      "loss": 0.8305,
+      "step": 17826
+    },
+    {
+      "epoch": 0.981176729594364,
+      "grad_norm": 0.6318126320838928,
+      "learning_rate": 5.169886209276607e-06,
+      "loss": 0.704,
+      "step": 17827
+    },
+    {
+      "epoch": 0.9812317683967197,
+      "grad_norm": 0.6664106249809265,
+      "learning_rate": 5.1694529923023755e-06,
+      "loss": 0.7478,
+      "step": 17828
+    },
+    {
+      "epoch": 0.9812868071990754,
+      "grad_norm": 0.6642717123031616,
+      "learning_rate": 5.16901977405458e-06,
+      "loss": 0.6419,
+      "step": 17829
+    },
+    {
+      "epoch": 0.981341846001431,
+      "grad_norm": 0.6837498545646667,
+      "learning_rate": 5.168586554536478e-06,
+      "loss": 0.709,
+      "step": 17830
+    },
+    {
+      "epoch": 0.9813968848037866,
+      "grad_norm": 0.7012450695037842,
+      "learning_rate": 5.168153333751323e-06,
+      "loss": 0.7386,
+      "step": 17831
+    },
+    {
+      "epoch": 0.9814519236061423,
+      "grad_norm": 0.7230706214904785,
+      "learning_rate": 5.1677201117023744e-06,
+      "loss": 0.6954,
+      "step": 17832
+    },
+    {
+      "epoch": 0.981506962408498,
+      "grad_norm": 0.9957376718521118,
+      "learning_rate": 5.167286888392886e-06,
+      "loss": 0.9409,
+      "step": 17833
+    },
+    {
+      "epoch": 0.9815620012108537,
+      "grad_norm": 0.6738237142562866,
+      "learning_rate": 5.1668536638261135e-06,
+      "loss": 0.7469,
+      "step": 17834
+    },
+    {
+      "epoch": 0.9816170400132093,
+      "grad_norm": 0.689182460308075,
+      "learning_rate": 5.166420438005313e-06,
+      "loss": 0.8144,
+      "step": 17835
+    },
+    {
+      "epoch": 0.981672078815565,
+      "grad_norm": 0.7391964793205261,
+      "learning_rate": 5.165987210933741e-06,
+      "loss": 0.8118,
+      "step": 17836
+    },
+    {
+      "epoch": 0.9817271176179206,
+      "grad_norm": 0.6194493770599365,
+      "learning_rate": 5.165553982614655e-06,
+      "loss": 0.6802,
+      "step": 17837
+    },
+    {
+      "epoch": 0.9817821564202763,
+      "grad_norm": 0.7108197808265686,
+      "learning_rate": 5.165120753051309e-06,
+      "loss": 0.812,
+      "step": 17838
+    },
+    {
+      "epoch": 0.9818371952226319,
+      "grad_norm": 0.599499523639679,
+      "learning_rate": 5.164687522246959e-06,
+      "loss": 0.6162,
+      "step": 17839
+    },
+    {
+      "epoch": 0.9818922340249876,
+      "grad_norm": 0.636405885219574,
+      "learning_rate": 5.164254290204862e-06,
+      "loss": 0.741,
+      "step": 17840
+    },
+    {
+      "epoch": 0.9819472728273433,
+      "grad_norm": 0.6845248937606812,
+      "learning_rate": 5.163821056928274e-06,
+      "loss": 0.723,
+      "step": 17841
+    },
+    {
+      "epoch": 0.982002311629699,
+      "grad_norm": 0.6328433156013489,
+      "learning_rate": 5.1633878224204515e-06,
+      "loss": 0.6717,
+      "step": 17842
+    },
+    {
+      "epoch": 0.9820573504320546,
+      "grad_norm": 0.6898068189620972,
+      "learning_rate": 5.162954586684649e-06,
+      "loss": 0.7641,
+      "step": 17843
+    },
+    {
+      "epoch": 0.9821123892344102,
+      "grad_norm": 0.563658595085144,
+      "learning_rate": 5.162521349724123e-06,
+      "loss": 0.6147,
+      "step": 17844
+    },
+    {
+      "epoch": 0.9821674280367659,
+      "grad_norm": 0.6722010970115662,
+      "learning_rate": 5.1620881115421314e-06,
+      "loss": 0.7132,
+      "step": 17845
+    },
+    {
+      "epoch": 0.9822224668391216,
+      "grad_norm": 0.6915282607078552,
+      "learning_rate": 5.161654872141928e-06,
+      "loss": 0.6803,
+      "step": 17846
+    },
+    {
+      "epoch": 0.9822775056414772,
+      "grad_norm": 0.6965909004211426,
+      "learning_rate": 5.161221631526769e-06,
+      "loss": 0.7267,
+      "step": 17847
+    },
+    {
+      "epoch": 0.9823325444438329,
+      "grad_norm": 0.704317569732666,
+      "learning_rate": 5.1607883896999126e-06,
+      "loss": 0.732,
+      "step": 17848
+    },
+    {
+      "epoch": 0.9823875832461886,
+      "grad_norm": 0.6609948873519897,
+      "learning_rate": 5.160355146664614e-06,
+      "loss": 0.7353,
+      "step": 17849
+    },
+    {
+      "epoch": 0.9824426220485443,
+      "grad_norm": 0.6089136600494385,
+      "learning_rate": 5.159921902424128e-06,
+      "loss": 0.6752,
+      "step": 17850
+    },
+    {
+      "epoch": 0.9824976608508998,
+      "grad_norm": 0.7590814828872681,
+      "learning_rate": 5.159488656981712e-06,
+      "loss": 0.7075,
+      "step": 17851
+    },
+    {
+      "epoch": 0.9825526996532555,
+      "grad_norm": 0.7506054639816284,
+      "learning_rate": 5.159055410340622e-06,
+      "loss": 0.6706,
+      "step": 17852
+    },
+    {
+      "epoch": 0.9826077384556112,
+      "grad_norm": 0.726391613483429,
+      "learning_rate": 5.158622162504115e-06,
+      "loss": 0.7227,
+      "step": 17853
+    },
+    {
+      "epoch": 0.9826627772579669,
+      "grad_norm": 0.7026383280754089,
+      "learning_rate": 5.158188913475445e-06,
+      "loss": 0.7934,
+      "step": 17854
+    },
+    {
+      "epoch": 0.9827178160603225,
+      "grad_norm": 0.6941219568252563,
+      "learning_rate": 5.1577556632578705e-06,
+      "loss": 0.7664,
+      "step": 17855
+    },
+    {
+      "epoch": 0.9827728548626782,
+      "grad_norm": 0.6923757195472717,
+      "learning_rate": 5.157322411854646e-06,
+      "loss": 0.8274,
+      "step": 17856
+    },
+    {
+      "epoch": 0.9828278936650339,
+      "grad_norm": 0.6614245772361755,
+      "learning_rate": 5.156889159269029e-06,
+      "loss": 0.7289,
+      "step": 17857
+    },
+    {
+      "epoch": 0.9828829324673896,
+      "grad_norm": 0.7060726284980774,
+      "learning_rate": 5.156455905504275e-06,
+      "loss": 0.6819,
+      "step": 17858
+    },
+    {
+      "epoch": 0.9829379712697451,
+      "grad_norm": 0.6427240967750549,
+      "learning_rate": 5.156022650563639e-06,
+      "loss": 0.7237,
+      "step": 17859
+    },
+    {
+      "epoch": 0.9829930100721008,
+      "grad_norm": 0.6128491163253784,
+      "learning_rate": 5.15558939445038e-06,
+      "loss": 0.7319,
+      "step": 17860
+    },
+    {
+      "epoch": 0.9830480488744565,
+      "grad_norm": 0.6768588423728943,
+      "learning_rate": 5.155156137167751e-06,
+      "loss": 0.673,
+      "step": 17861
+    },
+    {
+      "epoch": 0.9831030876768122,
+      "grad_norm": 0.7171456217765808,
+      "learning_rate": 5.1547228787190125e-06,
+      "loss": 0.7823,
+      "step": 17862
+    },
+    {
+      "epoch": 0.9831581264791678,
+      "grad_norm": 0.6802380681037903,
+      "learning_rate": 5.154289619107416e-06,
+      "loss": 0.7811,
+      "step": 17863
+    },
+    {
+      "epoch": 0.9832131652815235,
+      "grad_norm": 0.7872791886329651,
+      "learning_rate": 5.1538563583362215e-06,
+      "loss": 0.8189,
+      "step": 17864
+    },
+    {
+      "epoch": 0.9832682040838792,
+      "grad_norm": 0.7060483694076538,
+      "learning_rate": 5.153423096408683e-06,
+      "loss": 0.7343,
+      "step": 17865
+    },
+    {
+      "epoch": 0.9833232428862348,
+      "grad_norm": 0.710662305355072,
+      "learning_rate": 5.152989833328058e-06,
+      "loss": 0.7252,
+      "step": 17866
+    },
+    {
+      "epoch": 0.9833782816885904,
+      "grad_norm": 0.6512863039970398,
+      "learning_rate": 5.1525565690976035e-06,
+      "loss": 0.73,
+      "step": 17867
+    },
+    {
+      "epoch": 0.9834333204909461,
+      "grad_norm": 0.7715367078781128,
+      "learning_rate": 5.152123303720573e-06,
+      "loss": 0.736,
+      "step": 17868
+    },
+    {
+      "epoch": 0.9834883592933018,
+      "grad_norm": 0.6657119393348694,
+      "learning_rate": 5.151690037200225e-06,
+      "loss": 0.7094,
+      "step": 17869
+    },
+    {
+      "epoch": 0.9835433980956574,
+      "grad_norm": 0.6604039669036865,
+      "learning_rate": 5.151256769539815e-06,
+      "loss": 0.7337,
+      "step": 17870
+    },
+    {
+      "epoch": 0.9835984368980131,
+      "grad_norm": 0.6652948260307312,
+      "learning_rate": 5.150823500742599e-06,
+      "loss": 0.7307,
+      "step": 17871
+    },
+    {
+      "epoch": 0.9836534757003688,
+      "grad_norm": 0.7212419509887695,
+      "learning_rate": 5.150390230811835e-06,
+      "loss": 0.728,
+      "step": 17872
+    },
+    {
+      "epoch": 0.9837085145027245,
+      "grad_norm": 0.7573513984680176,
+      "learning_rate": 5.149956959750778e-06,
+      "loss": 0.7901,
+      "step": 17873
+    },
+    {
+      "epoch": 0.98376355330508,
+      "grad_norm": 0.6553870439529419,
+      "learning_rate": 5.149523687562685e-06,
+      "loss": 0.7,
+      "step": 17874
+    },
+    {
+      "epoch": 0.9838185921074357,
+      "grad_norm": 0.6913946270942688,
+      "learning_rate": 5.14909041425081e-06,
+      "loss": 0.7547,
+      "step": 17875
+    },
+    {
+      "epoch": 0.9838736309097914,
+      "grad_norm": 0.6566945910453796,
+      "learning_rate": 5.148657139818413e-06,
+      "loss": 0.7256,
+      "step": 17876
+    },
+    {
+      "epoch": 0.9839286697121471,
+      "grad_norm": 0.6469670534133911,
+      "learning_rate": 5.14822386426875e-06,
+      "loss": 0.6803,
+      "step": 17877
+    },
+    {
+      "epoch": 0.9839837085145027,
+      "grad_norm": 0.6760450005531311,
+      "learning_rate": 5.147790587605076e-06,
+      "loss": 0.7567,
+      "step": 17878
+    },
+    {
+      "epoch": 0.9840387473168584,
+      "grad_norm": 0.7145360708236694,
+      "learning_rate": 5.1473573098306455e-06,
+      "loss": 0.7758,
+      "step": 17879
+    },
+    {
+      "epoch": 0.984093786119214,
+      "grad_norm": 0.8869039416313171,
+      "learning_rate": 5.1469240309487175e-06,
+      "loss": 0.708,
+      "step": 17880
+    },
+    {
+      "epoch": 0.9841488249215697,
+      "grad_norm": 0.7354567050933838,
+      "learning_rate": 5.146490750962548e-06,
+      "loss": 0.7731,
+      "step": 17881
+    },
+    {
+      "epoch": 0.9842038637239253,
+      "grad_norm": 0.6989348530769348,
+      "learning_rate": 5.146057469875394e-06,
+      "loss": 0.7419,
+      "step": 17882
+    },
+    {
+      "epoch": 0.984258902526281,
+      "grad_norm": 0.6871246695518494,
+      "learning_rate": 5.145624187690509e-06,
+      "loss": 0.7305,
+      "step": 17883
+    },
+    {
+      "epoch": 0.9843139413286367,
+      "grad_norm": 1.8921427726745605,
+      "learning_rate": 5.1451909044111535e-06,
+      "loss": 0.918,
+      "step": 17884
+    },
+    {
+      "epoch": 0.9843689801309924,
+      "grad_norm": 0.7058244347572327,
+      "learning_rate": 5.144757620040582e-06,
+      "loss": 0.7247,
+      "step": 17885
+    },
+    {
+      "epoch": 0.984424018933348,
+      "grad_norm": 0.811503529548645,
+      "learning_rate": 5.14432433458205e-06,
+      "loss": 0.8942,
+      "step": 17886
+    },
+    {
+      "epoch": 0.9844790577357037,
+      "grad_norm": 0.7346917390823364,
+      "learning_rate": 5.143891048038815e-06,
+      "loss": 0.847,
+      "step": 17887
+    },
+    {
+      "epoch": 0.9845340965380593,
+      "grad_norm": 0.6721924543380737,
+      "learning_rate": 5.143457760414135e-06,
+      "loss": 0.708,
+      "step": 17888
+    },
+    {
+      "epoch": 0.984589135340415,
+      "grad_norm": 0.6473612785339355,
+      "learning_rate": 5.143024471711263e-06,
+      "loss": 0.6688,
+      "step": 17889
+    },
+    {
+      "epoch": 0.9846441741427706,
+      "grad_norm": 0.7375944256782532,
+      "learning_rate": 5.14259118193346e-06,
+      "loss": 0.8455,
+      "step": 17890
+    },
+    {
+      "epoch": 0.9846992129451263,
+      "grad_norm": 0.7279950976371765,
+      "learning_rate": 5.142157891083977e-06,
+      "loss": 0.7632,
+      "step": 17891
+    },
+    {
+      "epoch": 0.984754251747482,
+      "grad_norm": 0.7861246466636658,
+      "learning_rate": 5.141724599166074e-06,
+      "loss": 0.7698,
+      "step": 17892
+    },
+    {
+      "epoch": 0.9848092905498377,
+      "grad_norm": 0.6509001851081848,
+      "learning_rate": 5.141291306183007e-06,
+      "loss": 0.7519,
+      "step": 17893
+    },
+    {
+      "epoch": 0.9848643293521933,
+      "grad_norm": 0.6098498702049255,
+      "learning_rate": 5.140858012138032e-06,
+      "loss": 0.6977,
+      "step": 17894
+    },
+    {
+      "epoch": 0.984919368154549,
+      "grad_norm": 0.7857867479324341,
+      "learning_rate": 5.140424717034405e-06,
+      "loss": 0.6591,
+      "step": 17895
+    },
+    {
+      "epoch": 0.9849744069569046,
+      "grad_norm": 0.7155032157897949,
+      "learning_rate": 5.139991420875386e-06,
+      "loss": 0.821,
+      "step": 17896
+    },
+    {
+      "epoch": 0.9850294457592603,
+      "grad_norm": 0.6748823523521423,
+      "learning_rate": 5.139558123664228e-06,
+      "loss": 0.7852,
+      "step": 17897
+    },
+    {
+      "epoch": 0.9850844845616159,
+      "grad_norm": 0.7474187016487122,
+      "learning_rate": 5.139124825404188e-06,
+      "loss": 0.735,
+      "step": 17898
+    },
+    {
+      "epoch": 0.9851395233639716,
+      "grad_norm": 0.6562691926956177,
+      "learning_rate": 5.138691526098522e-06,
+      "loss": 0.775,
+      "step": 17899
+    },
+    {
+      "epoch": 0.9851945621663273,
+      "grad_norm": 0.6765697002410889,
+      "learning_rate": 5.13825822575049e-06,
+      "loss": 0.6383,
+      "step": 17900
+    },
+    {
+      "epoch": 0.985249600968683,
+      "grad_norm": 0.6247798800468445,
+      "learning_rate": 5.137824924363345e-06,
+      "loss": 0.7331,
+      "step": 17901
+    },
+    {
+      "epoch": 0.9853046397710385,
+      "grad_norm": 0.6596871018409729,
+      "learning_rate": 5.1373916219403465e-06,
+      "loss": 0.7698,
+      "step": 17902
+    },
+    {
+      "epoch": 0.9853596785733942,
+      "grad_norm": 0.7124335765838623,
+      "learning_rate": 5.136958318484746e-06,
+      "loss": 0.8027,
+      "step": 17903
+    },
+    {
+      "epoch": 0.9854147173757499,
+      "grad_norm": 0.6025542616844177,
+      "learning_rate": 5.136525013999805e-06,
+      "loss": 0.6862,
+      "step": 17904
+    },
+    {
+      "epoch": 0.9854697561781056,
+      "grad_norm": 0.8168695569038391,
+      "learning_rate": 5.136091708488779e-06,
+      "loss": 0.7608,
+      "step": 17905
+    },
+    {
+      "epoch": 0.9855247949804612,
+      "grad_norm": 0.660522997379303,
+      "learning_rate": 5.135658401954925e-06,
+      "loss": 0.7814,
+      "step": 17906
+    },
+    {
+      "epoch": 0.9855798337828169,
+      "grad_norm": 0.6121677160263062,
+      "learning_rate": 5.1352250944014955e-06,
+      "loss": 0.6691,
+      "step": 17907
+    },
+    {
+      "epoch": 0.9856348725851726,
+      "grad_norm": 0.7214170694351196,
+      "learning_rate": 5.134791785831753e-06,
+      "loss": 0.8601,
+      "step": 17908
+    },
+    {
+      "epoch": 0.9856899113875283,
+      "grad_norm": 0.7069474458694458,
+      "learning_rate": 5.1343584762489515e-06,
+      "loss": 0.6515,
+      "step": 17909
+    },
+    {
+      "epoch": 0.9857449501898838,
+      "grad_norm": 0.6847478747367859,
+      "learning_rate": 5.133925165656347e-06,
+      "loss": 0.8301,
+      "step": 17910
+    },
+    {
+      "epoch": 0.9857999889922395,
+      "grad_norm": 0.6535064578056335,
+      "learning_rate": 5.133491854057196e-06,
+      "loss": 0.7305,
+      "step": 17911
+    },
+    {
+      "epoch": 0.9858550277945952,
+      "grad_norm": 0.6921623349189758,
+      "learning_rate": 5.133058541454757e-06,
+      "loss": 0.7876,
+      "step": 17912
+    },
+    {
+      "epoch": 0.9859100665969508,
+      "grad_norm": 0.6774876117706299,
+      "learning_rate": 5.132625227852286e-06,
+      "loss": 0.7518,
+      "step": 17913
+    },
+    {
+      "epoch": 0.9859651053993065,
+      "grad_norm": 0.7022778987884521,
+      "learning_rate": 5.1321919132530405e-06,
+      "loss": 0.733,
+      "step": 17914
+    },
+    {
+      "epoch": 0.9860201442016622,
+      "grad_norm": 0.6287358403205872,
+      "learning_rate": 5.131758597660273e-06,
+      "loss": 0.7287,
+      "step": 17915
+    },
+    {
+      "epoch": 0.9860751830040179,
+      "grad_norm": 0.6294651031494141,
+      "learning_rate": 5.131325281077245e-06,
+      "loss": 0.6955,
+      "step": 17916
+    },
+    {
+      "epoch": 0.9861302218063734,
+      "grad_norm": 0.6313046813011169,
+      "learning_rate": 5.130891963507211e-06,
+      "loss": 0.6264,
+      "step": 17917
+    },
+    {
+      "epoch": 0.9861852606087291,
+      "grad_norm": 0.6451496481895447,
+      "learning_rate": 5.130458644953428e-06,
+      "loss": 0.6956,
+      "step": 17918
+    },
+    {
+      "epoch": 0.9862402994110848,
+      "grad_norm": 0.8261640071868896,
+      "learning_rate": 5.130025325419151e-06,
+      "loss": 0.7542,
+      "step": 17919
+    },
+    {
+      "epoch": 0.9862953382134405,
+      "grad_norm": 0.7422053813934326,
+      "learning_rate": 5.129592004907641e-06,
+      "loss": 0.8106,
+      "step": 17920
+    },
+    {
+      "epoch": 0.9863503770157961,
+      "grad_norm": 0.6633789539337158,
+      "learning_rate": 5.12915868342215e-06,
+      "loss": 0.7814,
+      "step": 17921
+    },
+    {
+      "epoch": 0.9864054158181518,
+      "grad_norm": 0.716140627861023,
+      "learning_rate": 5.12872536096594e-06,
+      "loss": 0.7172,
+      "step": 17922
+    },
+    {
+      "epoch": 0.9864604546205075,
+      "grad_norm": 0.8944414854049683,
+      "learning_rate": 5.128292037542261e-06,
+      "loss": 0.8734,
+      "step": 17923
+    },
+    {
+      "epoch": 0.9865154934228632,
+      "grad_norm": 0.6309584379196167,
+      "learning_rate": 5.127858713154375e-06,
+      "loss": 0.722,
+      "step": 17924
+    },
+    {
+      "epoch": 0.9865705322252187,
+      "grad_norm": 0.6496456861495972,
+      "learning_rate": 5.127425387805538e-06,
+      "loss": 0.6808,
+      "step": 17925
+    },
+    {
+      "epoch": 0.9866255710275744,
+      "grad_norm": 0.7771233320236206,
+      "learning_rate": 5.126992061499007e-06,
+      "loss": 0.8073,
+      "step": 17926
+    },
+    {
+      "epoch": 0.9866806098299301,
+      "grad_norm": 0.6993723511695862,
+      "learning_rate": 5.126558734238034e-06,
+      "loss": 0.689,
+      "step": 17927
+    },
+    {
+      "epoch": 0.9867356486322858,
+      "grad_norm": 0.7060848474502563,
+      "learning_rate": 5.126125406025881e-06,
+      "loss": 0.7604,
+      "step": 17928
+    },
+    {
+      "epoch": 0.9867906874346414,
+      "grad_norm": 0.7408806681632996,
+      "learning_rate": 5.125692076865804e-06,
+      "loss": 0.787,
+      "step": 17929
+    },
+    {
+      "epoch": 0.9868457262369971,
+      "grad_norm": 0.6351509094238281,
+      "learning_rate": 5.125258746761059e-06,
+      "loss": 0.7295,
+      "step": 17930
+    },
+    {
+      "epoch": 0.9869007650393528,
+      "grad_norm": 0.6362680196762085,
+      "learning_rate": 5.124825415714902e-06,
+      "loss": 0.7115,
+      "step": 17931
+    },
+    {
+      "epoch": 0.9869558038417084,
+      "grad_norm": 0.651877224445343,
+      "learning_rate": 5.12439208373059e-06,
+      "loss": 0.7006,
+      "step": 17932
+    },
+    {
+      "epoch": 0.987010842644064,
+      "grad_norm": 0.6457502245903015,
+      "learning_rate": 5.123958750811382e-06,
+      "loss": 0.6986,
+      "step": 17933
+    },
+    {
+      "epoch": 0.9870658814464197,
+      "grad_norm": 0.6840323805809021,
+      "learning_rate": 5.123525416960533e-06,
+      "loss": 0.8027,
+      "step": 17934
+    },
+    {
+      "epoch": 0.9871209202487754,
+      "grad_norm": 0.6861988306045532,
+      "learning_rate": 5.123092082181298e-06,
+      "loss": 0.7044,
+      "step": 17935
+    },
+    {
+      "epoch": 0.9871759590511311,
+      "grad_norm": 0.6497409343719482,
+      "learning_rate": 5.122658746476937e-06,
+      "loss": 0.6721,
+      "step": 17936
+    },
+    {
+      "epoch": 0.9872309978534867,
+      "grad_norm": 0.7505937814712524,
+      "learning_rate": 5.122225409850707e-06,
+      "loss": 0.7436,
+      "step": 17937
+    },
+    {
+      "epoch": 0.9872860366558424,
+      "grad_norm": 0.7507444024085999,
+      "learning_rate": 5.1217920723058635e-06,
+      "loss": 0.8291,
+      "step": 17938
+    },
+    {
+      "epoch": 0.987341075458198,
+      "grad_norm": 0.793416440486908,
+      "learning_rate": 5.1213587338456615e-06,
+      "loss": 0.7431,
+      "step": 17939
+    },
+    {
+      "epoch": 0.9873961142605537,
+      "grad_norm": 0.6958576440811157,
+      "learning_rate": 5.120925394473362e-06,
+      "loss": 0.7345,
+      "step": 17940
+    },
+    {
+      "epoch": 0.9874511530629093,
+      "grad_norm": 0.6618613600730896,
+      "learning_rate": 5.120492054192218e-06,
+      "loss": 0.7458,
+      "step": 17941
+    },
+    {
+      "epoch": 0.987506191865265,
+      "grad_norm": 0.6849207282066345,
+      "learning_rate": 5.120058713005488e-06,
+      "loss": 0.7965,
+      "step": 17942
+    },
+    {
+      "epoch": 0.9875612306676207,
+      "grad_norm": 0.717399001121521,
+      "learning_rate": 5.119625370916429e-06,
+      "loss": 0.7462,
+      "step": 17943
+    },
+    {
+      "epoch": 0.9876162694699764,
+      "grad_norm": 0.6913535594940186,
+      "learning_rate": 5.119192027928297e-06,
+      "loss": 0.6916,
+      "step": 17944
+    },
+    {
+      "epoch": 0.987671308272332,
+      "grad_norm": 0.6635895371437073,
+      "learning_rate": 5.118758684044351e-06,
+      "loss": 0.7063,
+      "step": 17945
+    },
+    {
+      "epoch": 0.9877263470746876,
+      "grad_norm": 0.6347431540489197,
+      "learning_rate": 5.1183253392678465e-06,
+      "loss": 0.7639,
+      "step": 17946
+    },
+    {
+      "epoch": 0.9877813858770433,
+      "grad_norm": 0.7334713935852051,
+      "learning_rate": 5.11789199360204e-06,
+      "loss": 0.7947,
+      "step": 17947
+    },
+    {
+      "epoch": 0.987836424679399,
+      "grad_norm": 0.6637980937957764,
+      "learning_rate": 5.117458647050188e-06,
+      "loss": 0.7521,
+      "step": 17948
+    },
+    {
+      "epoch": 0.9878914634817546,
+      "grad_norm": 0.656301736831665,
+      "learning_rate": 5.11702529961555e-06,
+      "loss": 0.5714,
+      "step": 17949
+    },
+    {
+      "epoch": 0.9879465022841103,
+      "grad_norm": 0.7977020144462585,
+      "learning_rate": 5.116591951301381e-06,
+      "loss": 0.7905,
+      "step": 17950
+    },
+    {
+      "epoch": 0.988001541086466,
+      "grad_norm": 0.6231863498687744,
+      "learning_rate": 5.1161586021109365e-06,
+      "loss": 0.6689,
+      "step": 17951
+    },
+    {
+      "epoch": 0.9880565798888217,
+      "grad_norm": 0.6863840818405151,
+      "learning_rate": 5.115725252047475e-06,
+      "loss": 0.8173,
+      "step": 17952
+    },
+    {
+      "epoch": 0.9881116186911773,
+      "grad_norm": 0.6148532032966614,
+      "learning_rate": 5.115291901114255e-06,
+      "loss": 0.734,
+      "step": 17953
+    },
+    {
+      "epoch": 0.9881666574935329,
+      "grad_norm": 0.6900758743286133,
+      "learning_rate": 5.114858549314531e-06,
+      "loss": 0.7317,
+      "step": 17954
+    },
+    {
+      "epoch": 0.9882216962958886,
+      "grad_norm": 0.6875699758529663,
+      "learning_rate": 5.1144251966515615e-06,
+      "loss": 0.6788,
+      "step": 17955
+    },
+    {
+      "epoch": 0.9882767350982442,
+      "grad_norm": 0.6588704586029053,
+      "learning_rate": 5.1139918431286014e-06,
+      "loss": 0.7728,
+      "step": 17956
+    },
+    {
+      "epoch": 0.9883317739005999,
+      "grad_norm": 0.6850310564041138,
+      "learning_rate": 5.113558488748911e-06,
+      "loss": 0.6942,
+      "step": 17957
+    },
+    {
+      "epoch": 0.9883868127029556,
+      "grad_norm": 0.7054475545883179,
+      "learning_rate": 5.113125133515744e-06,
+      "loss": 0.7564,
+      "step": 17958
+    },
+    {
+      "epoch": 0.9884418515053113,
+      "grad_norm": 0.9986534118652344,
+      "learning_rate": 5.112691777432361e-06,
+      "loss": 0.9043,
+      "step": 17959
+    },
+    {
+      "epoch": 0.9884968903076669,
+      "grad_norm": 0.6844736933708191,
+      "learning_rate": 5.112258420502013e-06,
+      "loss": 0.7999,
+      "step": 17960
+    },
+    {
+      "epoch": 0.9885519291100225,
+      "grad_norm": 0.5971188545227051,
+      "learning_rate": 5.111825062727964e-06,
+      "loss": 0.625,
+      "step": 17961
+    },
+    {
+      "epoch": 0.9886069679123782,
+      "grad_norm": 0.9815204739570618,
+      "learning_rate": 5.111391704113467e-06,
+      "loss": 0.6988,
+      "step": 17962
+    },
+    {
+      "epoch": 0.9886620067147339,
+      "grad_norm": 0.867161750793457,
+      "learning_rate": 5.11095834466178e-06,
+      "loss": 0.7014,
+      "step": 17963
+    },
+    {
+      "epoch": 0.9887170455170895,
+      "grad_norm": 0.8310801982879639,
+      "learning_rate": 5.1105249843761585e-06,
+      "loss": 0.7411,
+      "step": 17964
+    },
+    {
+      "epoch": 0.9887720843194452,
+      "grad_norm": 0.7921817898750305,
+      "learning_rate": 5.1100916232598614e-06,
+      "loss": 0.718,
+      "step": 17965
+    },
+    {
+      "epoch": 0.9888271231218009,
+      "grad_norm": 0.6440842747688293,
+      "learning_rate": 5.109658261316146e-06,
+      "loss": 0.6906,
+      "step": 17966
+    },
+    {
+      "epoch": 0.9888821619241566,
+      "grad_norm": 0.6158735156059265,
+      "learning_rate": 5.109224898548267e-06,
+      "loss": 0.6686,
+      "step": 17967
+    },
+    {
+      "epoch": 0.9889372007265121,
+      "grad_norm": 0.6632601618766785,
+      "learning_rate": 5.108791534959483e-06,
+      "loss": 0.7527,
+      "step": 17968
+    },
+    {
+      "epoch": 0.9889922395288678,
+      "grad_norm": 0.6995930671691895,
+      "learning_rate": 5.108358170553051e-06,
+      "loss": 0.8132,
+      "step": 17969
+    },
+    {
+      "epoch": 0.9890472783312235,
+      "grad_norm": 0.6986364126205444,
+      "learning_rate": 5.107924805332229e-06,
+      "loss": 0.7744,
+      "step": 17970
+    },
+    {
+      "epoch": 0.9891023171335792,
+      "grad_norm": 0.6733060479164124,
+      "learning_rate": 5.107491439300274e-06,
+      "loss": 0.6433,
+      "step": 17971
+    },
+    {
+      "epoch": 0.9891573559359348,
+      "grad_norm": 0.6814376711845398,
+      "learning_rate": 5.107058072460439e-06,
+      "loss": 0.6767,
+      "step": 17972
+    },
+    {
+      "epoch": 0.9892123947382905,
+      "grad_norm": 0.7519477009773254,
+      "learning_rate": 5.106624704815988e-06,
+      "loss": 0.7345,
+      "step": 17973
+    },
+    {
+      "epoch": 0.9892674335406462,
+      "grad_norm": 0.8123548030853271,
+      "learning_rate": 5.106191336370172e-06,
+      "loss": 0.7438,
+      "step": 17974
+    },
+    {
+      "epoch": 0.9893224723430019,
+      "grad_norm": 0.7073908448219299,
+      "learning_rate": 5.10575796712625e-06,
+      "loss": 0.8435,
+      "step": 17975
+    },
+    {
+      "epoch": 0.9893775111453574,
+      "grad_norm": 0.8263893723487854,
+      "learning_rate": 5.10532459708748e-06,
+      "loss": 0.9237,
+      "step": 17976
+    },
+    {
+      "epoch": 0.9894325499477131,
+      "grad_norm": 0.7631207704544067,
+      "learning_rate": 5.104891226257118e-06,
+      "loss": 0.7481,
+      "step": 17977
+    },
+    {
+      "epoch": 0.9894875887500688,
+      "grad_norm": 0.7013441324234009,
+      "learning_rate": 5.104457854638423e-06,
+      "loss": 0.6627,
+      "step": 17978
+    },
+    {
+      "epoch": 0.9895426275524245,
+      "grad_norm": 0.6746618151664734,
+      "learning_rate": 5.1040244822346496e-06,
+      "loss": 0.7585,
+      "step": 17979
+    },
+    {
+      "epoch": 0.9895976663547801,
+      "grad_norm": 0.707109808921814,
+      "learning_rate": 5.1035911090490555e-06,
+      "loss": 0.7628,
+      "step": 17980
+    },
+    {
+      "epoch": 0.9896527051571358,
+      "grad_norm": 0.7457565665245056,
+      "learning_rate": 5.103157735084899e-06,
+      "loss": 0.7375,
+      "step": 17981
+    },
+    {
+      "epoch": 0.9897077439594915,
+      "grad_norm": 0.8174850344657898,
+      "learning_rate": 5.102724360345437e-06,
+      "loss": 0.732,
+      "step": 17982
+    },
+    {
+      "epoch": 0.9897627827618471,
+      "grad_norm": 0.7620667219161987,
+      "learning_rate": 5.1022909848339265e-06,
+      "loss": 0.6926,
+      "step": 17983
+    },
+    {
+      "epoch": 0.9898178215642027,
+      "grad_norm": 0.6727546453475952,
+      "learning_rate": 5.101857608553622e-06,
+      "loss": 0.6986,
+      "step": 17984
+    },
+    {
+      "epoch": 0.9898728603665584,
+      "grad_norm": 0.6382732391357422,
+      "learning_rate": 5.101424231507786e-06,
+      "loss": 0.7489,
+      "step": 17985
+    },
+    {
+      "epoch": 0.9899278991689141,
+      "grad_norm": 0.678092360496521,
+      "learning_rate": 5.100990853699672e-06,
+      "loss": 0.7242,
+      "step": 17986
+    },
+    {
+      "epoch": 0.9899829379712698,
+      "grad_norm": 0.7694140672683716,
+      "learning_rate": 5.100557475132537e-06,
+      "loss": 0.8451,
+      "step": 17987
+    },
+    {
+      "epoch": 0.9900379767736254,
+      "grad_norm": 0.672201931476593,
+      "learning_rate": 5.100124095809638e-06,
+      "loss": 0.7942,
+      "step": 17988
+    },
+    {
+      "epoch": 0.9900930155759811,
+      "grad_norm": 0.7260831594467163,
+      "learning_rate": 5.099690715734234e-06,
+      "loss": 0.7073,
+      "step": 17989
+    },
+    {
+      "epoch": 0.9901480543783368,
+      "grad_norm": 0.6417271494865417,
+      "learning_rate": 5.099257334909582e-06,
+      "loss": 0.7272,
+      "step": 17990
+    },
+    {
+      "epoch": 0.9902030931806924,
+      "grad_norm": 0.6794359683990479,
+      "learning_rate": 5.098823953338939e-06,
+      "loss": 0.7286,
+      "step": 17991
+    },
+    {
+      "epoch": 0.990258131983048,
+      "grad_norm": 0.7701132297515869,
+      "learning_rate": 5.098390571025559e-06,
+      "loss": 0.8191,
+      "step": 17992
+    },
+    {
+      "epoch": 0.9903131707854037,
+      "grad_norm": 0.7108728289604187,
+      "learning_rate": 5.097957187972703e-06,
+      "loss": 0.6644,
+      "step": 17993
+    },
+    {
+      "epoch": 0.9903682095877594,
+      "grad_norm": 0.6622050404548645,
+      "learning_rate": 5.097523804183627e-06,
+      "loss": 0.713,
+      "step": 17994
+    },
+    {
+      "epoch": 0.9904232483901151,
+      "grad_norm": 0.664191722869873,
+      "learning_rate": 5.097090419661589e-06,
+      "loss": 0.7217,
+      "step": 17995
+    },
+    {
+      "epoch": 0.9904782871924707,
+      "grad_norm": 0.6588128209114075,
+      "learning_rate": 5.096657034409844e-06,
+      "loss": 0.8139,
+      "step": 17996
+    },
+    {
+      "epoch": 0.9905333259948264,
+      "grad_norm": 0.6252101063728333,
+      "learning_rate": 5.096223648431653e-06,
+      "loss": 0.6455,
+      "step": 17997
+    },
+    {
+      "epoch": 0.990588364797182,
+      "grad_norm": 0.6657758951187134,
+      "learning_rate": 5.095790261730269e-06,
+      "loss": 0.7416,
+      "step": 17998
+    },
+    {
+      "epoch": 0.9906434035995376,
+      "grad_norm": 0.8094693422317505,
+      "learning_rate": 5.095356874308952e-06,
+      "loss": 0.8645,
+      "step": 17999
+    },
+    {
+      "epoch": 0.9906984424018933,
+      "grad_norm": 0.7278556823730469,
+      "learning_rate": 5.094923486170957e-06,
+      "loss": 0.7261,
+      "step": 18000
+    },
+    {
+      "epoch": 0.990753481204249,
+      "grad_norm": 0.6950508952140808,
+      "learning_rate": 5.094490097319542e-06,
+      "loss": 0.7876,
+      "step": 18001
+    },
+    {
+      "epoch": 0.9908085200066047,
+      "grad_norm": 0.6594102382659912,
+      "learning_rate": 5.094056707757966e-06,
+      "loss": 0.8644,
+      "step": 18002
+    },
+    {
+      "epoch": 0.9908635588089603,
+      "grad_norm": 0.7983774542808533,
+      "learning_rate": 5.0936233174894855e-06,
+      "loss": 0.7356,
+      "step": 18003
+    },
+    {
+      "epoch": 0.990918597611316,
+      "grad_norm": 0.6786965727806091,
+      "learning_rate": 5.093189926517356e-06,
+      "loss": 0.6377,
+      "step": 18004
+    },
+    {
+      "epoch": 0.9909736364136716,
+      "grad_norm": 0.7368496656417847,
+      "learning_rate": 5.092756534844836e-06,
+      "loss": 0.7468,
+      "step": 18005
+    },
+    {
+      "epoch": 0.9910286752160273,
+      "grad_norm": 0.8692933917045593,
+      "learning_rate": 5.092323142475183e-06,
+      "loss": 0.7972,
+      "step": 18006
+    },
+    {
+      "epoch": 0.9910837140183829,
+      "grad_norm": 0.7743621468544006,
+      "learning_rate": 5.091889749411655e-06,
+      "loss": 0.643,
+      "step": 18007
+    },
+    {
+      "epoch": 0.9911387528207386,
+      "grad_norm": 0.6345173716545105,
+      "learning_rate": 5.0914563556575056e-06,
+      "loss": 0.7626,
+      "step": 18008
+    },
+    {
+      "epoch": 0.9911937916230943,
+      "grad_norm": 0.7068833112716675,
+      "learning_rate": 5.0910229612159974e-06,
+      "loss": 0.7439,
+      "step": 18009
+    },
+    {
+      "epoch": 0.99124883042545,
+      "grad_norm": 0.7917591333389282,
+      "learning_rate": 5.090589566090385e-06,
+      "loss": 0.8218,
+      "step": 18010
+    },
+    {
+      "epoch": 0.9913038692278056,
+      "grad_norm": 0.609967827796936,
+      "learning_rate": 5.0901561702839255e-06,
+      "loss": 0.6452,
+      "step": 18011
+    },
+    {
+      "epoch": 0.9913589080301612,
+      "grad_norm": 0.7628803253173828,
+      "learning_rate": 5.0897227737998745e-06,
+      "loss": 0.7487,
+      "step": 18012
+    },
+    {
+      "epoch": 0.9914139468325169,
+      "grad_norm": 0.6240081787109375,
+      "learning_rate": 5.089289376641492e-06,
+      "loss": 0.6527,
+      "step": 18013
+    },
+    {
+      "epoch": 0.9914689856348726,
+      "grad_norm": 0.789800763130188,
+      "learning_rate": 5.088855978812035e-06,
+      "loss": 0.7043,
+      "step": 18014
+    },
+    {
+      "epoch": 0.9915240244372282,
+      "grad_norm": 0.9605777263641357,
+      "learning_rate": 5.0884225803147604e-06,
+      "loss": 0.8486,
+      "step": 18015
+    },
+    {
+      "epoch": 0.9915790632395839,
+      "grad_norm": 0.9741759300231934,
+      "learning_rate": 5.087989181152924e-06,
+      "loss": 0.737,
+      "step": 18016
+    },
+    {
+      "epoch": 0.9916341020419396,
+      "grad_norm": 0.7788803577423096,
+      "learning_rate": 5.087555781329785e-06,
+      "loss": 0.8205,
+      "step": 18017
+    },
+    {
+      "epoch": 0.9916891408442953,
+      "grad_norm": 0.782712459564209,
+      "learning_rate": 5.0871223808486e-06,
+      "loss": 0.6982,
+      "step": 18018
+    },
+    {
+      "epoch": 0.9917441796466508,
+      "grad_norm": 0.6824339628219604,
+      "learning_rate": 5.086688979712627e-06,
+      "loss": 0.7448,
+      "step": 18019
+    },
+    {
+      "epoch": 0.9917992184490065,
+      "grad_norm": 0.6738336682319641,
+      "learning_rate": 5.0862555779251235e-06,
+      "loss": 0.6756,
+      "step": 18020
+    },
+    {
+      "epoch": 0.9918542572513622,
+      "grad_norm": 0.6139102578163147,
+      "learning_rate": 5.0858221754893455e-06,
+      "loss": 0.669,
+      "step": 18021
+    },
+    {
+      "epoch": 0.9919092960537179,
+      "grad_norm": 0.7581914067268372,
+      "learning_rate": 5.085388772408552e-06,
+      "loss": 0.7653,
+      "step": 18022
+    },
+    {
+      "epoch": 0.9919643348560735,
+      "grad_norm": 0.6985335946083069,
+      "learning_rate": 5.084955368685999e-06,
+      "loss": 0.7317,
+      "step": 18023
+    },
+    {
+      "epoch": 0.9920193736584292,
+      "grad_norm": 0.7850139737129211,
+      "learning_rate": 5.084521964324942e-06,
+      "loss": 0.8323,
+      "step": 18024
+    },
+    {
+      "epoch": 0.9920744124607849,
+      "grad_norm": 0.6971372365951538,
+      "learning_rate": 5.084088559328643e-06,
+      "loss": 0.8073,
+      "step": 18025
+    },
+    {
+      "epoch": 0.9921294512631406,
+      "grad_norm": 0.7196096181869507,
+      "learning_rate": 5.083655153700355e-06,
+      "loss": 0.793,
+      "step": 18026
+    },
+    {
+      "epoch": 0.9921844900654961,
+      "grad_norm": 0.6821334362030029,
+      "learning_rate": 5.083221747443339e-06,
+      "loss": 0.7266,
+      "step": 18027
+    },
+    {
+      "epoch": 0.9922395288678518,
+      "grad_norm": 0.6544287204742432,
+      "learning_rate": 5.082788340560849e-06,
+      "loss": 0.7812,
+      "step": 18028
+    },
+    {
+      "epoch": 0.9922945676702075,
+      "grad_norm": 0.6568353176116943,
+      "learning_rate": 5.082354933056145e-06,
+      "loss": 0.5682,
+      "step": 18029
+    },
+    {
+      "epoch": 0.9923496064725632,
+      "grad_norm": 0.7107832431793213,
+      "learning_rate": 5.081921524932483e-06,
+      "loss": 0.7068,
+      "step": 18030
+    },
+    {
+      "epoch": 0.9924046452749188,
+      "grad_norm": 0.6856156587600708,
+      "learning_rate": 5.081488116193122e-06,
+      "loss": 0.774,
+      "step": 18031
+    },
+    {
+      "epoch": 0.9924596840772745,
+      "grad_norm": 0.7203328013420105,
+      "learning_rate": 5.0810547068413165e-06,
+      "loss": 0.736,
+      "step": 18032
+    },
+    {
+      "epoch": 0.9925147228796302,
+      "grad_norm": 0.6304858326911926,
+      "learning_rate": 5.080621296880327e-06,
+      "loss": 0.6814,
+      "step": 18033
+    },
+    {
+      "epoch": 0.9925697616819859,
+      "grad_norm": 0.645539402961731,
+      "learning_rate": 5.080187886313408e-06,
+      "loss": 0.7984,
+      "step": 18034
+    },
+    {
+      "epoch": 0.9926248004843414,
+      "grad_norm": 0.655221700668335,
+      "learning_rate": 5.079754475143819e-06,
+      "loss": 0.7022,
+      "step": 18035
+    },
+    {
+      "epoch": 0.9926798392866971,
+      "grad_norm": 0.6705572605133057,
+      "learning_rate": 5.079321063374816e-06,
+      "loss": 0.82,
+      "step": 18036
+    },
+    {
+      "epoch": 0.9927348780890528,
+      "grad_norm": 0.6417540907859802,
+      "learning_rate": 5.078887651009657e-06,
+      "loss": 0.6941,
+      "step": 18037
+    },
+    {
+      "epoch": 0.9927899168914085,
+      "grad_norm": 0.6728167533874512,
+      "learning_rate": 5.078454238051601e-06,
+      "loss": 0.655,
+      "step": 18038
+    },
+    {
+      "epoch": 0.9928449556937641,
+      "grad_norm": 0.6670491099357605,
+      "learning_rate": 5.078020824503903e-06,
+      "loss": 0.7286,
+      "step": 18039
+    },
+    {
+      "epoch": 0.9928999944961198,
+      "grad_norm": 0.7217230200767517,
+      "learning_rate": 5.077587410369821e-06,
+      "loss": 0.7635,
+      "step": 18040
+    },
+    {
+      "epoch": 0.9929550332984755,
+      "grad_norm": 0.6922209858894348,
+      "learning_rate": 5.077153995652614e-06,
+      "loss": 0.6762,
+      "step": 18041
+    },
+    {
+      "epoch": 0.993010072100831,
+      "grad_norm": 0.8064020872116089,
+      "learning_rate": 5.0767205803555376e-06,
+      "loss": 0.6568,
+      "step": 18042
+    },
+    {
+      "epoch": 0.9930651109031867,
+      "grad_norm": 0.6940247416496277,
+      "learning_rate": 5.07628716448185e-06,
+      "loss": 0.7986,
+      "step": 18043
+    },
+    {
+      "epoch": 0.9931201497055424,
+      "grad_norm": 0.6729159951210022,
+      "learning_rate": 5.075853748034808e-06,
+      "loss": 0.7476,
+      "step": 18044
+    },
+    {
+      "epoch": 0.9931751885078981,
+      "grad_norm": 0.7298180460929871,
+      "learning_rate": 5.07542033101767e-06,
+      "loss": 0.7146,
+      "step": 18045
+    },
+    {
+      "epoch": 0.9932302273102537,
+      "grad_norm": 0.6649488806724548,
+      "learning_rate": 5.0749869134336935e-06,
+      "loss": 0.7668,
+      "step": 18046
+    },
+    {
+      "epoch": 0.9932852661126094,
+      "grad_norm": 0.6678916811943054,
+      "learning_rate": 5.0745534952861346e-06,
+      "loss": 0.6365,
+      "step": 18047
+    },
+    {
+      "epoch": 0.993340304914965,
+      "grad_norm": 0.8051722049713135,
+      "learning_rate": 5.0741200765782506e-06,
+      "loss": 0.7426,
+      "step": 18048
+    },
+    {
+      "epoch": 0.9933953437173207,
+      "grad_norm": 0.6319746375083923,
+      "learning_rate": 5.073686657313302e-06,
+      "loss": 0.6855,
+      "step": 18049
+    },
+    {
+      "epoch": 0.9934503825196763,
+      "grad_norm": 0.6837119460105896,
+      "learning_rate": 5.0732532374945434e-06,
+      "loss": 0.7657,
+      "step": 18050
+    },
+    {
+      "epoch": 0.993505421322032,
+      "grad_norm": 0.6625218391418457,
+      "learning_rate": 5.072819817125233e-06,
+      "loss": 0.7836,
+      "step": 18051
+    },
+    {
+      "epoch": 0.9935604601243877,
+      "grad_norm": 0.6353194117546082,
+      "learning_rate": 5.072386396208627e-06,
+      "loss": 0.7298,
+      "step": 18052
+    },
+    {
+      "epoch": 0.9936154989267434,
+      "grad_norm": 0.702965259552002,
+      "learning_rate": 5.071952974747986e-06,
+      "loss": 0.8423,
+      "step": 18053
+    },
+    {
+      "epoch": 0.993670537729099,
+      "grad_norm": 0.7337152361869812,
+      "learning_rate": 5.071519552746566e-06,
+      "loss": 0.734,
+      "step": 18054
+    },
+    {
+      "epoch": 0.9937255765314547,
+      "grad_norm": 0.6457077264785767,
+      "learning_rate": 5.071086130207623e-06,
+      "loss": 0.8179,
+      "step": 18055
+    },
+    {
+      "epoch": 0.9937806153338103,
+      "grad_norm": 0.6137452125549316,
+      "learning_rate": 5.070652707134416e-06,
+      "loss": 0.6268,
+      "step": 18056
+    },
+    {
+      "epoch": 0.993835654136166,
+      "grad_norm": 0.6552117466926575,
+      "learning_rate": 5.070219283530202e-06,
+      "loss": 0.7754,
+      "step": 18057
+    },
+    {
+      "epoch": 0.9938906929385216,
+      "grad_norm": 0.7026165127754211,
+      "learning_rate": 5.06978585939824e-06,
+      "loss": 0.7684,
+      "step": 18058
+    },
+    {
+      "epoch": 0.9939457317408773,
+      "grad_norm": 0.5770609974861145,
+      "learning_rate": 5.069352434741786e-06,
+      "loss": 0.6603,
+      "step": 18059
+    },
+    {
+      "epoch": 0.994000770543233,
+      "grad_norm": 0.6504804491996765,
+      "learning_rate": 5.068919009564095e-06,
+      "loss": 0.7602,
+      "step": 18060
+    },
+    {
+      "epoch": 0.9940558093455887,
+      "grad_norm": 0.7996575832366943,
+      "learning_rate": 5.068485583868429e-06,
+      "loss": 0.8378,
+      "step": 18061
+    },
+    {
+      "epoch": 0.9941108481479443,
+      "grad_norm": 0.6521992087364197,
+      "learning_rate": 5.068052157658044e-06,
+      "loss": 0.7303,
+      "step": 18062
+    },
+    {
+      "epoch": 0.9941658869503,
+      "grad_norm": 0.7057589292526245,
+      "learning_rate": 5.067618730936197e-06,
+      "loss": 0.7986,
+      "step": 18063
+    },
+    {
+      "epoch": 0.9942209257526556,
+      "grad_norm": 0.7683775424957275,
+      "learning_rate": 5.067185303706144e-06,
+      "loss": 0.7051,
+      "step": 18064
+    },
+    {
+      "epoch": 0.9942759645550113,
+      "grad_norm": 0.6067343950271606,
+      "learning_rate": 5.0667518759711465e-06,
+      "loss": 0.6399,
+      "step": 18065
+    },
+    {
+      "epoch": 0.9943310033573669,
+      "grad_norm": 0.7575204372406006,
+      "learning_rate": 5.066318447734459e-06,
+      "loss": 0.7124,
+      "step": 18066
+    },
+    {
+      "epoch": 0.9943860421597226,
+      "grad_norm": 0.7211230397224426,
+      "learning_rate": 5.06588501899934e-06,
+      "loss": 0.7719,
+      "step": 18067
+    },
+    {
+      "epoch": 0.9944410809620783,
+      "grad_norm": 0.8543131351470947,
+      "learning_rate": 5.065451589769047e-06,
+      "loss": 0.7033,
+      "step": 18068
+    },
+    {
+      "epoch": 0.994496119764434,
+      "grad_norm": 0.6685993075370789,
+      "learning_rate": 5.065018160046837e-06,
+      "loss": 0.7331,
+      "step": 18069
+    },
+    {
+      "epoch": 0.9945511585667896,
+      "grad_norm": 0.7711308002471924,
+      "learning_rate": 5.064584729835969e-06,
+      "loss": 0.7389,
+      "step": 18070
+    },
+    {
+      "epoch": 0.9946061973691452,
+      "grad_norm": 0.8691509366035461,
+      "learning_rate": 5.064151299139698e-06,
+      "loss": 0.7457,
+      "step": 18071
+    },
+    {
+      "epoch": 0.9946612361715009,
+      "grad_norm": 0.7130528688430786,
+      "learning_rate": 5.063717867961282e-06,
+      "loss": 0.7442,
+      "step": 18072
+    },
+    {
+      "epoch": 0.9947162749738566,
+      "grad_norm": 0.6554980278015137,
+      "learning_rate": 5.063284436303982e-06,
+      "loss": 0.6587,
+      "step": 18073
+    },
+    {
+      "epoch": 0.9947713137762122,
+      "grad_norm": 0.7127482295036316,
+      "learning_rate": 5.062851004171052e-06,
+      "loss": 0.7282,
+      "step": 18074
+    },
+    {
+      "epoch": 0.9948263525785679,
+      "grad_norm": 0.7662289142608643,
+      "learning_rate": 5.062417571565751e-06,
+      "loss": 0.7576,
+      "step": 18075
+    },
+    {
+      "epoch": 0.9948813913809236,
+      "grad_norm": 0.9702945351600647,
+      "learning_rate": 5.061984138491335e-06,
+      "loss": 0.8,
+      "step": 18076
+    },
+    {
+      "epoch": 0.9949364301832793,
+      "grad_norm": 0.6914392709732056,
+      "learning_rate": 5.061550704951064e-06,
+      "loss": 0.6912,
+      "step": 18077
+    },
+    {
+      "epoch": 0.9949914689856348,
+      "grad_norm": 0.5724602341651917,
+      "learning_rate": 5.061117270948194e-06,
+      "loss": 0.636,
+      "step": 18078
+    },
+    {
+      "epoch": 0.9950465077879905,
+      "grad_norm": 0.8456641435623169,
+      "learning_rate": 5.060683836485983e-06,
+      "loss": 0.7298,
+      "step": 18079
+    },
+    {
+      "epoch": 0.9951015465903462,
+      "grad_norm": 0.7085068821907043,
+      "learning_rate": 5.060250401567689e-06,
+      "loss": 0.7567,
+      "step": 18080
+    },
+    {
+      "epoch": 0.9951565853927019,
+      "grad_norm": 0.7004247307777405,
+      "learning_rate": 5.059816966196569e-06,
+      "loss": 0.7666,
+      "step": 18081
+    },
+    {
+      "epoch": 0.9952116241950575,
+      "grad_norm": 0.6310703158378601,
+      "learning_rate": 5.059383530375881e-06,
+      "loss": 0.712,
+      "step": 18082
+    },
+    {
+      "epoch": 0.9952666629974132,
+      "grad_norm": 0.6922436356544495,
+      "learning_rate": 5.058950094108882e-06,
+      "loss": 0.775,
+      "step": 18083
+    },
+    {
+      "epoch": 0.9953217017997689,
+      "grad_norm": 0.7536594867706299,
+      "learning_rate": 5.058516657398829e-06,
+      "loss": 0.7872,
+      "step": 18084
+    },
+    {
+      "epoch": 0.9953767406021244,
+      "grad_norm": 0.6446521282196045,
+      "learning_rate": 5.058083220248981e-06,
+      "loss": 0.6646,
+      "step": 18085
+    },
+    {
+      "epoch": 0.9954317794044801,
+      "grad_norm": 0.6904979348182678,
+      "learning_rate": 5.057649782662595e-06,
+      "loss": 0.7559,
+      "step": 18086
+    },
+    {
+      "epoch": 0.9954868182068358,
+      "grad_norm": 0.7692407965660095,
+      "learning_rate": 5.057216344642929e-06,
+      "loss": 0.7622,
+      "step": 18087
+    },
+    {
+      "epoch": 0.9955418570091915,
+      "grad_norm": 0.660089910030365,
+      "learning_rate": 5.0567829061932405e-06,
+      "loss": 0.7823,
+      "step": 18088
+    },
+    {
+      "epoch": 0.9955968958115471,
+      "grad_norm": 0.6194438934326172,
+      "learning_rate": 5.056349467316785e-06,
+      "loss": 0.6406,
+      "step": 18089
+    },
+    {
+      "epoch": 0.9956519346139028,
+      "grad_norm": 0.7209805250167847,
+      "learning_rate": 5.055916028016824e-06,
+      "loss": 0.7373,
+      "step": 18090
+    },
+    {
+      "epoch": 0.9957069734162585,
+      "grad_norm": 0.8307274580001831,
+      "learning_rate": 5.055482588296613e-06,
+      "loss": 0.7486,
+      "step": 18091
+    },
+    {
+      "epoch": 0.9957620122186142,
+      "grad_norm": 0.6076368093490601,
+      "learning_rate": 5.055049148159409e-06,
+      "loss": 0.7009,
+      "step": 18092
+    },
+    {
+      "epoch": 0.9958170510209697,
+      "grad_norm": 0.6484249830245972,
+      "learning_rate": 5.054615707608471e-06,
+      "loss": 0.7158,
+      "step": 18093
+    },
+    {
+      "epoch": 0.9958720898233254,
+      "grad_norm": 0.6629005670547485,
+      "learning_rate": 5.054182266647055e-06,
+      "loss": 0.7392,
+      "step": 18094
+    },
+    {
+      "epoch": 0.9959271286256811,
+      "grad_norm": 0.8084813952445984,
+      "learning_rate": 5.0537488252784205e-06,
+      "loss": 0.7418,
+      "step": 18095
+    },
+    {
+      "epoch": 0.9959821674280368,
+      "grad_norm": 0.9468874931335449,
+      "learning_rate": 5.053315383505824e-06,
+      "loss": 0.8699,
+      "step": 18096
+    },
+    {
+      "epoch": 0.9960372062303924,
+      "grad_norm": 0.7731055617332458,
+      "learning_rate": 5.052881941332522e-06,
+      "loss": 0.9144,
+      "step": 18097
+    },
+    {
+      "epoch": 0.9960922450327481,
+      "grad_norm": 0.7422712445259094,
+      "learning_rate": 5.052448498761776e-06,
+      "loss": 0.7075,
+      "step": 18098
+    },
+    {
+      "epoch": 0.9961472838351038,
+      "grad_norm": 0.7030091881752014,
+      "learning_rate": 5.05201505579684e-06,
+      "loss": 0.782,
+      "step": 18099
+    },
+    {
+      "epoch": 0.9962023226374594,
+      "grad_norm": 0.7698554396629333,
+      "learning_rate": 5.051581612440972e-06,
+      "loss": 0.7743,
+      "step": 18100
+    },
+    {
+      "epoch": 0.996257361439815,
+      "grad_norm": 0.6205971837043762,
+      "learning_rate": 5.051148168697431e-06,
+      "loss": 0.6862,
+      "step": 18101
+    },
+    {
+      "epoch": 0.9963124002421707,
+      "grad_norm": 0.7322680354118347,
+      "learning_rate": 5.050714724569474e-06,
+      "loss": 0.7658,
+      "step": 18102
+    },
+    {
+      "epoch": 0.9963674390445264,
+      "grad_norm": 1.250335454940796,
+      "learning_rate": 5.050281280060358e-06,
+      "loss": 0.7925,
+      "step": 18103
+    },
+    {
+      "epoch": 0.9964224778468821,
+      "grad_norm": 1.0011868476867676,
+      "learning_rate": 5.049847835173344e-06,
+      "loss": 0.6378,
+      "step": 18104
+    },
+    {
+      "epoch": 0.9964775166492377,
+      "grad_norm": 0.6508066058158875,
+      "learning_rate": 5.049414389911684e-06,
+      "loss": 0.7329,
+      "step": 18105
+    },
+    {
+      "epoch": 0.9965325554515934,
+      "grad_norm": 0.6632688641548157,
+      "learning_rate": 5.0489809442786406e-06,
+      "loss": 0.7599,
+      "step": 18106
+    },
+    {
+      "epoch": 0.996587594253949,
+      "grad_norm": 0.7015318274497986,
+      "learning_rate": 5.048547498277469e-06,
+      "loss": 0.7626,
+      "step": 18107
+    },
+    {
+      "epoch": 0.9966426330563047,
+      "grad_norm": 0.7433881759643555,
+      "learning_rate": 5.048114051911427e-06,
+      "loss": 0.7612,
+      "step": 18108
+    },
+    {
+      "epoch": 0.9966976718586603,
+      "grad_norm": 0.7714030742645264,
+      "learning_rate": 5.047680605183772e-06,
+      "loss": 0.875,
+      "step": 18109
+    },
+    {
+      "epoch": 0.996752710661016,
+      "grad_norm": 0.7467747330665588,
+      "learning_rate": 5.047247158097764e-06,
+      "loss": 0.8541,
+      "step": 18110
+    },
+    {
+      "epoch": 0.9968077494633717,
+      "grad_norm": 0.6674730181694031,
+      "learning_rate": 5.0468137106566586e-06,
+      "loss": 0.7139,
+      "step": 18111
+    },
+    {
+      "epoch": 0.9968627882657274,
+      "grad_norm": 0.6534250974655151,
+      "learning_rate": 5.046380262863713e-06,
+      "loss": 0.782,
+      "step": 18112
+    },
+    {
+      "epoch": 0.996917827068083,
+      "grad_norm": 0.6498619318008423,
+      "learning_rate": 5.0459468147221846e-06,
+      "loss": 0.7743,
+      "step": 18113
+    },
+    {
+      "epoch": 0.9969728658704387,
+      "grad_norm": 0.8043835163116455,
+      "learning_rate": 5.045513366235335e-06,
+      "loss": 0.7288,
+      "step": 18114
+    },
+    {
+      "epoch": 0.9970279046727943,
+      "grad_norm": 0.7055554389953613,
+      "learning_rate": 5.045079917406418e-06,
+      "loss": 0.7245,
+      "step": 18115
+    },
+    {
+      "epoch": 0.99708294347515,
+      "grad_norm": 0.7295365333557129,
+      "learning_rate": 5.044646468238692e-06,
+      "loss": 0.7464,
+      "step": 18116
+    },
+    {
+      "epoch": 0.9971379822775056,
+      "grad_norm": 0.6557313799858093,
+      "learning_rate": 5.0442130187354155e-06,
+      "loss": 0.6831,
+      "step": 18117
+    },
+    {
+      "epoch": 0.9971930210798613,
+      "grad_norm": 0.7435111403465271,
+      "learning_rate": 5.043779568899846e-06,
+      "loss": 0.8412,
+      "step": 18118
+    },
+    {
+      "epoch": 0.997248059882217,
+      "grad_norm": 0.6655508875846863,
+      "learning_rate": 5.043346118735242e-06,
+      "loss": 0.7027,
+      "step": 18119
+    },
+    {
+      "epoch": 0.9973030986845727,
+      "grad_norm": 0.5845352411270142,
+      "learning_rate": 5.042912668244858e-06,
+      "loss": 0.6496,
+      "step": 18120
+    },
+    {
+      "epoch": 0.9973581374869283,
+      "grad_norm": 0.8248149156570435,
+      "learning_rate": 5.042479217431954e-06,
+      "loss": 0.8128,
+      "step": 18121
+    },
+    {
+      "epoch": 0.9974131762892839,
+      "grad_norm": 0.7040210962295532,
+      "learning_rate": 5.042045766299789e-06,
+      "loss": 0.7452,
+      "step": 18122
+    },
+    {
+      "epoch": 0.9974682150916396,
+      "grad_norm": 0.6569299101829529,
+      "learning_rate": 5.0416123148516185e-06,
+      "loss": 0.7664,
+      "step": 18123
+    },
+    {
+      "epoch": 0.9975232538939953,
+      "grad_norm": 0.6373575329780579,
+      "learning_rate": 5.041178863090701e-06,
+      "loss": 0.7861,
+      "step": 18124
+    },
+    {
+      "epoch": 0.9975782926963509,
+      "grad_norm": 0.793760359287262,
+      "learning_rate": 5.040745411020294e-06,
+      "loss": 0.7909,
+      "step": 18125
+    },
+    {
+      "epoch": 0.9976333314987066,
+      "grad_norm": 0.6696873307228088,
+      "learning_rate": 5.0403119586436555e-06,
+      "loss": 0.754,
+      "step": 18126
+    },
+    {
+      "epoch": 0.9976883703010623,
+      "grad_norm": 0.6404829621315002,
+      "learning_rate": 5.039878505964043e-06,
+      "loss": 0.7401,
+      "step": 18127
+    },
+    {
+      "epoch": 0.9977434091034179,
+      "grad_norm": 0.7170735001564026,
+      "learning_rate": 5.039445052984715e-06,
+      "loss": 0.7789,
+      "step": 18128
+    },
+    {
+      "epoch": 0.9977984479057735,
+      "grad_norm": 0.7306370735168457,
+      "learning_rate": 5.039011599708928e-06,
+      "loss": 0.7786,
+      "step": 18129
+    },
+    {
+      "epoch": 0.9978534867081292,
+      "grad_norm": 0.5817254781723022,
+      "learning_rate": 5.0385781461399405e-06,
+      "loss": 0.6853,
+      "step": 18130
+    },
+    {
+      "epoch": 0.9979085255104849,
+      "grad_norm": 0.6285548210144043,
+      "learning_rate": 5.038144692281011e-06,
+      "loss": 0.7597,
+      "step": 18131
+    },
+    {
+      "epoch": 0.9979635643128405,
+      "grad_norm": 0.6597462892532349,
+      "learning_rate": 5.037711238135395e-06,
+      "loss": 0.7448,
+      "step": 18132
+    },
+    {
+      "epoch": 0.9980186031151962,
+      "grad_norm": 0.6609601378440857,
+      "learning_rate": 5.0372777837063505e-06,
+      "loss": 0.7324,
+      "step": 18133
+    },
+    {
+      "epoch": 0.9980736419175519,
+      "grad_norm": 0.6518424153327942,
+      "learning_rate": 5.036844328997138e-06,
+      "loss": 0.8015,
+      "step": 18134
+    },
+    {
+      "epoch": 0.9981286807199076,
+      "grad_norm": 0.7128679156303406,
+      "learning_rate": 5.036410874011013e-06,
+      "loss": 0.7436,
+      "step": 18135
+    },
+    {
+      "epoch": 0.9981837195222631,
+      "grad_norm": 0.7561799883842468,
+      "learning_rate": 5.0359774187512324e-06,
+      "loss": 0.8486,
+      "step": 18136
+    },
+    {
+      "epoch": 0.9982387583246188,
+      "grad_norm": 0.702702522277832,
+      "learning_rate": 5.035543963221056e-06,
+      "loss": 0.7645,
+      "step": 18137
+    },
+    {
+      "epoch": 0.9982937971269745,
+      "grad_norm": 0.636001467704773,
+      "learning_rate": 5.0351105074237405e-06,
+      "loss": 0.6667,
+      "step": 18138
+    },
+    {
+      "epoch": 0.9983488359293302,
+      "grad_norm": 0.6555122137069702,
+      "learning_rate": 5.034677051362545e-06,
+      "loss": 0.6966,
+      "step": 18139
+    },
+    {
+      "epoch": 0.9984038747316858,
+      "grad_norm": 0.7465553879737854,
+      "learning_rate": 5.034243595040724e-06,
+      "loss": 0.8167,
+      "step": 18140
+    },
+    {
+      "epoch": 0.9984589135340415,
+      "grad_norm": 0.7388782501220703,
+      "learning_rate": 5.033810138461539e-06,
+      "loss": 0.7637,
+      "step": 18141
+    },
+    {
+      "epoch": 0.9985139523363972,
+      "grad_norm": 0.7342116832733154,
+      "learning_rate": 5.033376681628245e-06,
+      "loss": 0.6742,
+      "step": 18142
+    },
+    {
+      "epoch": 0.9985689911387529,
+      "grad_norm": 0.6627763509750366,
+      "learning_rate": 5.0329432245441016e-06,
+      "loss": 0.6908,
+      "step": 18143
+    },
+    {
+      "epoch": 0.9986240299411084,
+      "grad_norm": 0.6392931938171387,
+      "learning_rate": 5.032509767212366e-06,
+      "loss": 0.6588,
+      "step": 18144
+    },
+    {
+      "epoch": 0.9986790687434641,
+      "grad_norm": 0.7912692427635193,
+      "learning_rate": 5.032076309636294e-06,
+      "loss": 0.8287,
+      "step": 18145
+    },
+    {
+      "epoch": 0.9987341075458198,
+      "grad_norm": 0.6602104306221008,
+      "learning_rate": 5.031642851819146e-06,
+      "loss": 0.7309,
+      "step": 18146
+    },
+    {
+      "epoch": 0.9987891463481755,
+      "grad_norm": 0.7000023126602173,
+      "learning_rate": 5.0312093937641794e-06,
+      "loss": 0.6837,
+      "step": 18147
+    },
+    {
+      "epoch": 0.9988441851505311,
+      "grad_norm": 0.6883202791213989,
+      "learning_rate": 5.03077593547465e-06,
+      "loss": 0.7644,
+      "step": 18148
+    },
+    {
+      "epoch": 0.9988992239528868,
+      "grad_norm": 0.6720017194747925,
+      "learning_rate": 5.030342476953817e-06,
+      "loss": 0.7135,
+      "step": 18149
+    },
+    {
+      "epoch": 0.9989542627552425,
+      "grad_norm": 0.7414998412132263,
+      "learning_rate": 5.029909018204939e-06,
+      "loss": 0.7674,
+      "step": 18150
+    },
+    {
+      "epoch": 0.9990093015575982,
+      "grad_norm": 0.7004234790802002,
+      "learning_rate": 5.029475559231273e-06,
+      "loss": 0.8119,
+      "step": 18151
+    },
+    {
+      "epoch": 0.9990643403599537,
+      "grad_norm": 0.5972886681556702,
+      "learning_rate": 5.029042100036077e-06,
+      "loss": 0.6716,
+      "step": 18152
+    },
+    {
+      "epoch": 0.9991193791623094,
+      "grad_norm": 0.6771349906921387,
+      "learning_rate": 5.028608640622606e-06,
+      "loss": 0.7743,
+      "step": 18153
+    },
+    {
+      "epoch": 0.9991744179646651,
+      "grad_norm": 0.6927672028541565,
+      "learning_rate": 5.0281751809941225e-06,
+      "loss": 0.7718,
+      "step": 18154
+    },
+    {
+      "epoch": 0.9992294567670208,
+      "grad_norm": 0.8227388262748718,
+      "learning_rate": 5.0277417211538815e-06,
+      "loss": 0.823,
+      "step": 18155
+    },
+    {
+      "epoch": 0.9992844955693764,
+      "grad_norm": 0.6523411870002747,
+      "learning_rate": 5.027308261105141e-06,
+      "loss": 0.6624,
+      "step": 18156
+    },
+    {
+      "epoch": 0.9993395343717321,
+      "grad_norm": 0.7758201956748962,
+      "learning_rate": 5.026874800851157e-06,
+      "loss": 0.8567,
+      "step": 18157
+    },
+    {
+      "epoch": 0.9993945731740878,
+      "grad_norm": 0.6323235630989075,
+      "learning_rate": 5.026441340395192e-06,
+      "loss": 0.7494,
+      "step": 18158
+    },
+    {
+      "epoch": 0.9994496119764434,
+      "grad_norm": 0.6327700018882751,
+      "learning_rate": 5.0260078797405e-06,
+      "loss": 0.67,
+      "step": 18159
+    },
+    {
+      "epoch": 0.999504650778799,
+      "grad_norm": 0.6610170602798462,
+      "learning_rate": 5.025574418890339e-06,
+      "loss": 0.7279,
+      "step": 18160
+    },
+    {
+      "epoch": 0.9995596895811547,
+      "grad_norm": 0.677026629447937,
+      "learning_rate": 5.025140957847967e-06,
+      "loss": 0.7256,
+      "step": 18161
+    },
+    {
+      "epoch": 0.9996147283835104,
+      "grad_norm": 0.8088963031768799,
+      "learning_rate": 5.024707496616643e-06,
+      "loss": 0.7456,
+      "step": 18162
+    },
+    {
+      "epoch": 0.9996697671858661,
+      "grad_norm": 0.6300736665725708,
+      "learning_rate": 5.024274035199626e-06,
+      "loss": 0.758,
+      "step": 18163
+    },
+    {
+      "epoch": 0.9997248059882217,
+      "grad_norm": 0.6184163093566895,
+      "learning_rate": 5.023840573600169e-06,
+      "loss": 0.6011,
+      "step": 18164
+    },
+    {
+      "epoch": 0.9997798447905774,
+      "grad_norm": 0.7885013222694397,
+      "learning_rate": 5.023407111821534e-06,
+      "loss": 0.7626,
+      "step": 18165
+    },
+    {
+      "epoch": 0.999834883592933,
+      "grad_norm": 0.8335184454917908,
+      "learning_rate": 5.022973649866978e-06,
+      "loss": 0.7157,
+      "step": 18166
+    },
+    {
+      "epoch": 0.9998899223952887,
+      "grad_norm": 0.6559926867485046,
+      "learning_rate": 5.022540187739757e-06,
+      "loss": 0.6273,
+      "step": 18167
+    },
+    {
+      "epoch": 0.9999449611976443,
+      "grad_norm": 0.6130450963973999,
+      "learning_rate": 5.022106725443131e-06,
+      "loss": 0.6352,
+      "step": 18168
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7267221212387085,
+      "learning_rate": 5.0216732629803554e-06,
+      "loss": 0.6457,
+      "step": 18169
+    },
+    {
+      "epoch": 1.0000550388023557,
+      "grad_norm": 0.8285889029502869,
+      "learning_rate": 5.021239800354691e-06,
+      "loss": 0.7876,
+      "step": 18170
+    },
+    {
+      "epoch": 1.0001100776047114,
+      "grad_norm": 0.7533310651779175,
+      "learning_rate": 5.020806337569392e-06,
+      "loss": 0.7404,
+      "step": 18171
+    },
+    {
+      "epoch": 1.000165116407067,
+      "grad_norm": 0.7649720311164856,
+      "learning_rate": 5.02037287462772e-06,
+      "loss": 0.7648,
+      "step": 18172
+    },
+    {
+      "epoch": 1.0002201552094225,
+      "grad_norm": 0.687062680721283,
+      "learning_rate": 5.019939411532929e-06,
+      "loss": 0.7159,
+      "step": 18173
+    },
+    {
+      "epoch": 1.0002751940117782,
+      "grad_norm": 0.6438779830932617,
+      "learning_rate": 5.01950594828828e-06,
+      "loss": 0.6121,
+      "step": 18174
+    },
+    {
+      "epoch": 1.000330232814134,
+      "grad_norm": 0.7010402679443359,
+      "learning_rate": 5.0190724848970296e-06,
+      "loss": 0.8288,
+      "step": 18175
+    },
+    {
+      "epoch": 1.0003852716164896,
+      "grad_norm": 0.6656293869018555,
+      "learning_rate": 5.018639021362436e-06,
+      "loss": 0.7333,
+      "step": 18176
+    },
+    {
+      "epoch": 1.0004403104188453,
+      "grad_norm": 0.7373145222663879,
+      "learning_rate": 5.018205557687756e-06,
+      "loss": 0.7492,
+      "step": 18177
+    },
+    {
+      "epoch": 1.000495349221201,
+      "grad_norm": 0.6980772018432617,
+      "learning_rate": 5.017772093876248e-06,
+      "loss": 0.7217,
+      "step": 18178
+    },
+    {
+      "epoch": 1.0005503880235567,
+      "grad_norm": 0.6385942101478577,
+      "learning_rate": 5.01733862993117e-06,
+      "loss": 0.6618,
+      "step": 18179
+    },
+    {
+      "epoch": 1.0006054268259124,
+      "grad_norm": 0.6605677008628845,
+      "learning_rate": 5.016905165855779e-06,
+      "loss": 0.7343,
+      "step": 18180
     }
   ],
   "logging_steps": 1,
@@ -108197,7 +127286,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.56027501400739e+19,
+  "total_flos": 5.3650294282439885e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null