diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9834 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7138122673736808,
+  "eval_steps": 500,
+  "global_step": 14000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005098659052669148,
+      "grad_norm": 316.62525281525757,
+      "learning_rate": 7.640067911714771e-08,
+      "loss": 10.4749,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010197318105338296,
+      "grad_norm": 285.7401162496275,
+      "learning_rate": 1.6129032258064518e-07,
+      "loss": 10.6201,
+      "step": 20
+    },
+    {
+      "epoch": 0.0015295977158007444,
+      "grad_norm": 312.2873919163466,
+      "learning_rate": 2.461799660441426e-07,
+      "loss": 9.8721,
+      "step": 30
+    },
+    {
+      "epoch": 0.002039463621067659,
+      "grad_norm": 243.94179574653327,
+      "learning_rate": 3.310696095076401e-07,
+      "loss": 8.295,
+      "step": 40
+    },
+    {
+      "epoch": 0.002549329526334574,
+      "grad_norm": 140.37820408824535,
+      "learning_rate": 4.1595925297113753e-07,
+      "loss": 6.5498,
+      "step": 50
+    },
+    {
+      "epoch": 0.0030591954316014887,
+      "grad_norm": 74.21515072344681,
+      "learning_rate": 5.00848896434635e-07,
+      "loss": 5.3838,
+      "step": 60
+    },
+    {
+      "epoch": 0.0035690613368684035,
+      "grad_norm": 93.57353432385545,
+      "learning_rate": 5.857385398981324e-07,
+      "loss": 4.775,
+      "step": 70
+    },
+    {
+      "epoch": 0.004078927242135318,
+      "grad_norm": 68.41964079965516,
+      "learning_rate": 6.706281833616299e-07,
+      "loss": 4.4661,
+      "step": 80
+    },
+    {
+      "epoch": 0.0045887931474022335,
+      "grad_norm": 54.81800032711122,
+      "learning_rate": 7.555178268251275e-07,
+      "loss": 4.1719,
+      "step": 90
+    },
+    {
+      "epoch": 0.005098659052669148,
+      "grad_norm": 234.65660116280884,
+      "learning_rate": 8.404074702886249e-07,
+      "loss": 4.1875,
+      "step": 100
+    },
+    {
+      "epoch": 0.005608524957936063,
+      "grad_norm": 51.75897318253527,
+      "learning_rate": 9.252971137521223e-07,
+      "loss": 4.1807,
+      "step": 110
+    },
+    {
+      "epoch": 0.0061183908632029774,
+      "grad_norm": 32.87813340481878,
+      "learning_rate": 1.0101867572156197e-06,
+      "loss": 4.0487,
+      "step": 120
+    },
+    {
+      "epoch": 0.006628256768469893,
+      "grad_norm": 47.12160023497351,
+      "learning_rate": 1.0950764006791174e-06,
+      "loss": 4.057,
+      "step": 130
+    },
+    {
+      "epoch": 0.007138122673736807,
+      "grad_norm": 33.36787554080754,
+      "learning_rate": 1.1799660441426147e-06,
+      "loss": 4.0638,
+      "step": 140
+    },
+    {
+      "epoch": 0.007647988579003722,
+      "grad_norm": 36.22615493740882,
+      "learning_rate": 1.2648556876061122e-06,
+      "loss": 3.9743,
+      "step": 150
+    },
+    {
+      "epoch": 0.008157854484270637,
+      "grad_norm": 25.548896106296255,
+      "learning_rate": 1.3497453310696096e-06,
+      "loss": 4.0058,
+      "step": 160
+    },
+    {
+      "epoch": 0.008667720389537552,
+      "grad_norm": 28.71639199269464,
+      "learning_rate": 1.434634974533107e-06,
+      "loss": 3.9652,
+      "step": 170
+    },
+    {
+      "epoch": 0.009177586294804467,
+      "grad_norm": 92.81634578219439,
+      "learning_rate": 1.5195246179966044e-06,
+      "loss": 4.0078,
+      "step": 180
+    },
+    {
+      "epoch": 0.00968745220007138,
+      "grad_norm": 46.254937054936384,
+      "learning_rate": 1.604414261460102e-06,
+      "loss": 3.9596,
+      "step": 190
+    },
+    {
+      "epoch": 0.010197318105338296,
+      "grad_norm": 936.4754280338494,
+      "learning_rate": 1.6893039049235995e-06,
+      "loss": 3.855,
+      "step": 200
+    },
+    {
+      "epoch": 0.010707184010605211,
+      "grad_norm": 496.26665459390637,
+      "learning_rate": 1.774193548387097e-06,
+      "loss": 3.7976,
+      "step": 210
+    },
+    {
+      "epoch": 0.011217049915872126,
+      "grad_norm": 59.013534542348424,
+      "learning_rate": 1.8590831918505945e-06,
+      "loss": 3.7332,
+      "step": 220
+    },
+    {
+      "epoch": 0.01172691582113904,
+      "grad_norm": 697.3633039430886,
+      "learning_rate": 1.943972835314092e-06,
+      "loss": 3.6656,
+      "step": 230
+    },
+    {
+      "epoch": 0.012236781726405955,
+      "grad_norm": 52.874578205264044,
+      "learning_rate": 2.028862478777589e-06,
+      "loss": 3.5203,
+      "step": 240
+    },
+    {
+      "epoch": 0.01274664763167287,
+      "grad_norm": 76.57920146938653,
+      "learning_rate": 2.113752122241087e-06,
+      "loss": 3.2779,
+      "step": 250
+    },
+    {
+      "epoch": 0.013256513536939785,
+      "grad_norm": 85.19750945653773,
+      "learning_rate": 2.1986417657045842e-06,
+      "loss": 3.1369,
+      "step": 260
+    },
+    {
+      "epoch": 0.013766379442206699,
+      "grad_norm": 65.41060472799023,
+      "learning_rate": 2.2835314091680816e-06,
+      "loss": 2.8289,
+      "step": 270
+    },
+    {
+      "epoch": 0.014276245347473614,
+      "grad_norm": 53.689335922333726,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 2.6487,
+      "step": 280
+    },
+    {
+      "epoch": 0.01478611125274053,
+      "grad_norm": 53.93921545430968,
+      "learning_rate": 2.4533106960950766e-06,
+      "loss": 2.3619,
+      "step": 290
+    },
+    {
+      "epoch": 0.015295977158007444,
+      "grad_norm": 58.944377639899294,
+      "learning_rate": 2.538200339558574e-06,
+      "loss": 2.1433,
+      "step": 300
+    },
+    {
+      "epoch": 0.015805843063274358,
+      "grad_norm": 46.41298005691738,
+      "learning_rate": 2.6230899830220713e-06,
+      "loss": 1.8862,
+      "step": 310
+    },
+    {
+      "epoch": 0.016315708968541273,
+      "grad_norm": 122.74189427146665,
+      "learning_rate": 2.707979626485569e-06,
+      "loss": 1.6422,
+      "step": 320
+    },
+    {
+      "epoch": 0.01682557487380819,
+      "grad_norm": 81.65446781620822,
+      "learning_rate": 2.7928692699490667e-06,
+      "loss": 1.3886,
+      "step": 330
+    },
+    {
+      "epoch": 0.017335440779075104,
+      "grad_norm": 38.92723075682003,
+      "learning_rate": 2.877758913412564e-06,
+      "loss": 1.3017,
+      "step": 340
+    },
+    {
+      "epoch": 0.01784530668434202,
+      "grad_norm": 27.287077460988606,
+      "learning_rate": 2.9626485568760614e-06,
+      "loss": 1.151,
+      "step": 350
+    },
+    {
+      "epoch": 0.018355172589608934,
+      "grad_norm": 32.79014822857911,
+      "learning_rate": 3.0475382003395587e-06,
+      "loss": 1.0401,
+      "step": 360
+    },
+    {
+      "epoch": 0.01886503849487585,
+      "grad_norm": 24.31108717981766,
+      "learning_rate": 3.1324278438030564e-06,
+      "loss": 1.079,
+      "step": 370
+    },
+    {
+      "epoch": 0.01937490440014276,
+      "grad_norm": 21.523545006358592,
+      "learning_rate": 3.2173174872665538e-06,
+      "loss": 0.9804,
+      "step": 380
+    },
+    {
+      "epoch": 0.019884770305409676,
+      "grad_norm": 21.37428335689284,
+      "learning_rate": 3.302207130730051e-06,
+      "loss": 0.887,
+      "step": 390
+    },
+    {
+      "epoch": 0.02039463621067659,
+      "grad_norm": 18.688340602869662,
+      "learning_rate": 3.3870967741935484e-06,
+      "loss": 0.8609,
+      "step": 400
+    },
+    {
+      "epoch": 0.020904502115943507,
+      "grad_norm": 64.36276305999914,
+      "learning_rate": 3.471986417657046e-06,
+      "loss": 0.8907,
+      "step": 410
+    },
+    {
+      "epoch": 0.021414368021210422,
+      "grad_norm": 47.17306058904959,
+      "learning_rate": 3.556876061120544e-06,
+      "loss": 0.8588,
+      "step": 420
+    },
+    {
+      "epoch": 0.021924233926477337,
+      "grad_norm": 29.113641286019497,
+      "learning_rate": 3.641765704584041e-06,
+      "loss": 0.9311,
+      "step": 430
+    },
+    {
+      "epoch": 0.022434099831744252,
+      "grad_norm": 34.98515248575744,
+      "learning_rate": 3.7266553480475385e-06,
+      "loss": 0.8532,
+      "step": 440
+    },
+    {
+      "epoch": 0.022943965737011168,
+      "grad_norm": 17.956084637737842,
+      "learning_rate": 3.811544991511036e-06,
+      "loss": 0.846,
+      "step": 450
+    },
+    {
+      "epoch": 0.02345383164227808,
+      "grad_norm": 27.7234720431503,
+      "learning_rate": 3.896434634974533e-06,
+      "loss": 0.8954,
+      "step": 460
+    },
+    {
+      "epoch": 0.023963697547544995,
+      "grad_norm": 19.560756884372157,
+      "learning_rate": 3.981324278438031e-06,
+      "loss": 0.8641,
+      "step": 470
+    },
+    {
+      "epoch": 0.02447356345281191,
+      "grad_norm": 57.59529673932827,
+      "learning_rate": 4.066213921901529e-06,
+      "loss": 0.8708,
+      "step": 480
+    },
+    {
+      "epoch": 0.024983429358078825,
+      "grad_norm": 17.446075098012066,
+      "learning_rate": 4.1511035653650255e-06,
+      "loss": 0.8501,
+      "step": 490
+    },
+    {
+      "epoch": 0.02549329526334574,
+      "grad_norm": 30.267237186402376,
+      "learning_rate": 4.235993208828523e-06,
+      "loss": 0.904,
+      "step": 500
+    },
+    {
+      "epoch": 0.026003161168612655,
+      "grad_norm": 13.743819054307822,
+      "learning_rate": 4.320882852292021e-06,
+      "loss": 0.8243,
+      "step": 510
+    },
+    {
+      "epoch": 0.02651302707387957,
+      "grad_norm": 80.97279138588813,
+      "learning_rate": 4.405772495755518e-06,
+      "loss": 0.9007,
+      "step": 520
+    },
+    {
+      "epoch": 0.027022892979146486,
+      "grad_norm": 16.040486362513338,
+      "learning_rate": 4.490662139219016e-06,
+      "loss": 0.7898,
+      "step": 530
+    },
+    {
+      "epoch": 0.027532758884413398,
+      "grad_norm": 10.675993103430974,
+      "learning_rate": 4.575551782682513e-06,
+      "loss": 0.8358,
+      "step": 540
+    },
+    {
+      "epoch": 0.028042624789680313,
+      "grad_norm": 22.25156917130235,
+      "learning_rate": 4.66044142614601e-06,
+      "loss": 0.8365,
+      "step": 550
+    },
+    {
+      "epoch": 0.028552490694947228,
+      "grad_norm": 23.67520091302,
+      "learning_rate": 4.745331069609508e-06,
+      "loss": 0.8394,
+      "step": 560
+    },
+    {
+      "epoch": 0.029062356600214143,
+      "grad_norm": 15.025526899274565,
+      "learning_rate": 4.830220713073006e-06,
+      "loss": 0.7973,
+      "step": 570
+    },
+    {
+      "epoch": 0.02957222250548106,
+      "grad_norm": 20.041273748960393,
+      "learning_rate": 4.915110356536503e-06,
+      "loss": 0.8677,
+      "step": 580
+    },
+    {
+      "epoch": 0.030082088410747974,
+      "grad_norm": 44.45852961277365,
+      "learning_rate": 5e-06,
+      "loss": 0.8133,
+      "step": 590
+    },
+    {
+      "epoch": 0.03059195431601489,
+      "grad_norm": 12.016989828995413,
+      "learning_rate": 4.999996591164963e-06,
+      "loss": 0.8235,
+      "step": 600
+    },
+    {
+      "epoch": 0.031101820221281804,
+      "grad_norm": 14.441095000429819,
+      "learning_rate": 4.999986364669145e-06,
+      "loss": 0.7902,
+      "step": 610
+    },
+    {
+      "epoch": 0.031611686126548716,
+      "grad_norm": 25.24752988501315,
+      "learning_rate": 4.999969320540435e-06,
+      "loss": 0.7942,
+      "step": 620
+    },
+    {
+      "epoch": 0.03212155203181563,
+      "grad_norm": 15.168503999969841,
+      "learning_rate": 4.999945458825315e-06,
+      "loss": 0.7884,
+      "step": 630
+    },
+    {
+      "epoch": 0.032631417937082546,
+      "grad_norm": 10.83021089779209,
+      "learning_rate": 4.9999147795888545e-06,
+      "loss": 0.7717,
+      "step": 640
+    },
+    {
+      "epoch": 0.03314128384234946,
+      "grad_norm": 19.445450635811248,
+      "learning_rate": 4.999877282914722e-06,
+      "loss": 0.8135,
+      "step": 650
+    },
+    {
+      "epoch": 0.03365114974761638,
+      "grad_norm": 22.554572824298493,
+      "learning_rate": 4.999832968905171e-06,
+      "loss": 0.8295,
+      "step": 660
+    },
+    {
+      "epoch": 0.03416101565288329,
+      "grad_norm": 42.731702899490074,
+      "learning_rate": 4.999781837681048e-06,
+      "loss": 0.8874,
+      "step": 670
+    },
+    {
+      "epoch": 0.03467088155815021,
+      "grad_norm": 20.307886633961164,
+      "learning_rate": 4.999723889381793e-06,
+      "loss": 0.8232,
+      "step": 680
+    },
+    {
+      "epoch": 0.03518074746341712,
+      "grad_norm": 10.52969344532376,
+      "learning_rate": 4.999659124165434e-06,
+      "loss": 0.8182,
+      "step": 690
+    },
+    {
+      "epoch": 0.03569061336868404,
+      "grad_norm": 14.518720996860537,
+      "learning_rate": 4.999587542208591e-06,
+      "loss": 0.7534,
+      "step": 700
+    },
+    {
+      "epoch": 0.03620047927395095,
+      "grad_norm": 19.483740313351362,
+      "learning_rate": 4.999509143706472e-06,
+      "loss": 0.8737,
+      "step": 710
+    },
+    {
+      "epoch": 0.03671034517921787,
+      "grad_norm": 54.41221044485456,
+      "learning_rate": 4.999423928872876e-06,
+      "loss": 0.9167,
+      "step": 720
+    },
+    {
+      "epoch": 0.03722021108448478,
+      "grad_norm": 23.662531439864207,
+      "learning_rate": 4.999331897940189e-06,
+      "loss": 0.8138,
+      "step": 730
+    },
+    {
+      "epoch": 0.0377300769897517,
+      "grad_norm": 27.072148977063062,
+      "learning_rate": 4.999233051159385e-06,
+      "loss": 0.7962,
+      "step": 740
+    },
+    {
+      "epoch": 0.03823994289501861,
+      "grad_norm": 24.651843711278424,
+      "learning_rate": 4.999127388800028e-06,
+      "loss": 0.8065,
+      "step": 750
+    },
+    {
+      "epoch": 0.03874980880028552,
+      "grad_norm": 21.94483810373911,
+      "learning_rate": 4.999014911150264e-06,
+      "loss": 0.8254,
+      "step": 760
+    },
+    {
+      "epoch": 0.03925967470555244,
+      "grad_norm": 28.802217532211035,
+      "learning_rate": 4.998895618516829e-06,
+      "loss": 0.8433,
+      "step": 770
+    },
+    {
+      "epoch": 0.03976954061081935,
+      "grad_norm": 11.32904166159406,
+      "learning_rate": 4.998769511225041e-06,
+      "loss": 0.7309,
+      "step": 780
+    },
+    {
+      "epoch": 0.04027940651608627,
+      "grad_norm": 99.07364050040725,
+      "learning_rate": 4.998636589618803e-06,
+      "loss": 0.838,
+      "step": 790
+    },
+    {
+      "epoch": 0.04078927242135318,
+      "grad_norm": 34.42370873450888,
+      "learning_rate": 4.998496854060603e-06,
+      "loss": 0.7781,
+      "step": 800
+    },
+    {
+      "epoch": 0.0412991383266201,
+      "grad_norm": 8.488083542574639,
+      "learning_rate": 4.998350304931507e-06,
+      "loss": 0.8022,
+      "step": 810
+    },
+    {
+      "epoch": 0.04180900423188701,
+      "grad_norm": 14.994383374892706,
+      "learning_rate": 4.998196942631166e-06,
+      "loss": 0.7474,
+      "step": 820
+    },
+    {
+      "epoch": 0.04231887013715393,
+      "grad_norm": 26.626123416012348,
+      "learning_rate": 4.998036767577809e-06,
+      "loss": 0.7616,
+      "step": 830
+    },
+    {
+      "epoch": 0.042828736042420844,
+      "grad_norm": 18.654351581860734,
+      "learning_rate": 4.997869780208244e-06,
+      "loss": 0.8281,
+      "step": 840
+    },
+    {
+      "epoch": 0.04333860194768776,
+      "grad_norm": 27.303240160352054,
+      "learning_rate": 4.997695980977858e-06,
+      "loss": 0.7595,
+      "step": 850
+    },
+    {
+      "epoch": 0.043848467852954674,
+      "grad_norm": 9.408631234143192,
+      "learning_rate": 4.997515370360613e-06,
+      "loss": 0.7614,
+      "step": 860
+    },
+    {
+      "epoch": 0.04435833375822159,
+      "grad_norm": 8.49071108290611,
+      "learning_rate": 4.997327948849044e-06,
+      "loss": 0.715,
+      "step": 870
+    },
+    {
+      "epoch": 0.044868199663488505,
+      "grad_norm": 17.76405503635825,
+      "learning_rate": 4.997133716954266e-06,
+      "loss": 0.7866,
+      "step": 880
+    },
+    {
+      "epoch": 0.04537806556875542,
+      "grad_norm": 15.798376541913443,
+      "learning_rate": 4.996932675205961e-06,
+      "loss": 0.7586,
+      "step": 890
+    },
+    {
+      "epoch": 0.045887931474022335,
+      "grad_norm": 13.142262117486272,
+      "learning_rate": 4.996724824152382e-06,
+      "loss": 0.7853,
+      "step": 900
+    },
+    {
+      "epoch": 0.04639779737928925,
+      "grad_norm": 85.17749253553107,
+      "learning_rate": 4.996510164360355e-06,
+      "loss": 0.7684,
+      "step": 910
+    },
+    {
+      "epoch": 0.04690766328455616,
+      "grad_norm": 34.70487061745637,
+      "learning_rate": 4.996288696415271e-06,
+      "loss": 0.7657,
+      "step": 920
+    },
+    {
+      "epoch": 0.047417529189823074,
+      "grad_norm": 11.797913133577842,
+      "learning_rate": 4.9960604209210885e-06,
+      "loss": 0.7401,
+      "step": 930
+    },
+    {
+      "epoch": 0.04792739509508999,
+      "grad_norm": 18.459225779080683,
+      "learning_rate": 4.99582533850033e-06,
+      "loss": 0.7398,
+      "step": 940
+    },
+    {
+      "epoch": 0.048437261000356904,
+      "grad_norm": 13.742483693841917,
+      "learning_rate": 4.995583449794081e-06,
+      "loss": 0.7655,
+      "step": 950
+    },
+    {
+      "epoch": 0.04894712690562382,
+      "grad_norm": 50.003928825953324,
+      "learning_rate": 4.99533475546199e-06,
+      "loss": 0.6877,
+      "step": 960
+    },
+    {
+      "epoch": 0.049456992810890735,
+      "grad_norm": 8.082504134207067,
+      "learning_rate": 4.995079256182261e-06,
+      "loss": 0.7962,
+      "step": 970
+    },
+    {
+      "epoch": 0.04996685871615765,
+      "grad_norm": 19.5746887217413,
+      "learning_rate": 4.99481695265166e-06,
+      "loss": 0.7655,
+      "step": 980
+    },
+    {
+      "epoch": 0.050476724621424565,
+      "grad_norm": 67.6550051771218,
+      "learning_rate": 4.994547845585505e-06,
+      "loss": 0.7461,
+      "step": 990
+    },
+    {
+      "epoch": 0.05098659052669148,
+      "grad_norm": 22.637134522224294,
+      "learning_rate": 4.994271935717671e-06,
+      "loss": 0.803,
+      "step": 1000
+    },
+    {
+      "epoch": 0.051496456431958396,
+      "grad_norm": 18.105250320469835,
+      "learning_rate": 4.99398922380058e-06,
+      "loss": 0.8435,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05200632233722531,
+      "grad_norm": 7.6508740287140125,
+      "learning_rate": 4.99369971060521e-06,
+      "loss": 0.7596,
+      "step": 1020
+    },
+    {
+      "epoch": 0.052516188242492226,
+      "grad_norm": 15.513939265611556,
+      "learning_rate": 4.993403396921082e-06,
+      "loss": 0.7175,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05302605414775914,
+      "grad_norm": 16.59172849330501,
+      "learning_rate": 4.993100283556262e-06,
+      "loss": 0.6755,
+      "step": 1040
+    },
+    {
+      "epoch": 0.053535920053026056,
+      "grad_norm": 14.05251425869156,
+      "learning_rate": 4.992790371337363e-06,
+      "loss": 0.7593,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05404578595829297,
+      "grad_norm": 14.57170517221848,
+      "learning_rate": 4.992473661109535e-06,
+      "loss": 0.7366,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05455565186355989,
+      "grad_norm": 12.125948500236293,
+      "learning_rate": 4.99215015373647e-06,
+      "loss": 0.7425,
+      "step": 1070
+    },
+    {
+      "epoch": 0.055065517768826795,
+      "grad_norm": 8.24534642564121,
+      "learning_rate": 4.991819850100394e-06,
+      "loss": 0.7838,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05557538367409371,
+      "grad_norm": 56.96441026445971,
+      "learning_rate": 4.991482751102066e-06,
+      "loss": 0.6824,
+      "step": 1090
+    },
+    {
+      "epoch": 0.056085249579360626,
+      "grad_norm": 24.02412886675572,
+      "learning_rate": 4.99113885766078e-06,
+      "loss": 0.7026,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05659511548462754,
+      "grad_norm": 52.756101119639034,
+      "learning_rate": 4.990788170714355e-06,
+      "loss": 0.725,
+      "step": 1110
+    },
+    {
+      "epoch": 0.057104981389894456,
+      "grad_norm": 17.38614067478378,
+      "learning_rate": 4.99043069121914e-06,
+      "loss": 0.7609,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05761484729516137,
+      "grad_norm": 8.055654893368551,
+      "learning_rate": 4.990066420150005e-06,
+      "loss": 0.6845,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05812471320042829,
+      "grad_norm": 18.036549926044923,
+      "learning_rate": 4.989695358500342e-06,
+      "loss": 0.7608,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0586345791056952,
+      "grad_norm": 79.44498639307135,
+      "learning_rate": 4.989317507282061e-06,
+      "loss": 0.7873,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05914444501096212,
+      "grad_norm": 10.75610853089343,
+      "learning_rate": 4.988932867525588e-06,
+      "loss": 0.6532,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05965431091622903,
+      "grad_norm": 21.466559907484747,
+      "learning_rate": 4.988541440279862e-06,
+      "loss": 0.7931,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06016417682149595,
+      "grad_norm": 39.97021653705112,
+      "learning_rate": 4.988143226612333e-06,
+      "loss": 0.7459,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06067404272676286,
+      "grad_norm": 26.23183983177856,
+      "learning_rate": 4.987738227608954e-06,
+      "loss": 0.7495,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06118390863202978,
+      "grad_norm": 30.186219658432467,
+      "learning_rate": 4.987326444374189e-06,
+      "loss": 0.7752,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06169377453729669,
+      "grad_norm": 19.81958951098629,
+      "learning_rate": 4.986907878030994e-06,
+      "loss": 0.6576,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06220364044256361,
+      "grad_norm": 10.29760641694682,
+      "learning_rate": 4.986482529720831e-06,
+      "loss": 0.7398,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06271350634783052,
+      "grad_norm": 23.848823577624334,
+      "learning_rate": 4.986050400603653e-06,
+      "loss": 0.7755,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06322337225309743,
+      "grad_norm": 16.981268241082418,
+      "learning_rate": 4.985611491857906e-06,
+      "loss": 0.6892,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06373323815836435,
+      "grad_norm": 12.203438943223416,
+      "learning_rate": 4.9851658046805226e-06,
+      "loss": 0.697,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06424310406363126,
+      "grad_norm": 22.820805465543035,
+      "learning_rate": 4.9847133402869235e-06,
+      "loss": 0.7316,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06475296996889818,
+      "grad_norm": 15.965046033606288,
+      "learning_rate": 4.984254099911009e-06,
+      "loss": 0.6669,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06526283587416509,
+      "grad_norm": 51.03956406155315,
+      "learning_rate": 4.98378808480516e-06,
+      "loss": 0.7309,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06577270177943201,
+      "grad_norm": 11.580015861232175,
+      "learning_rate": 4.98331529624023e-06,
+      "loss": 0.693,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06628256768469892,
+      "grad_norm": 23.393195377173306,
+      "learning_rate": 4.982835735505545e-06,
+      "loss": 0.7701,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06679243358996584,
+      "grad_norm": 21.230670744931253,
+      "learning_rate": 4.982349403908902e-06,
+      "loss": 0.7009,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06730229949523275,
+      "grad_norm": 7.105260679653341,
+      "learning_rate": 4.98185630277656e-06,
+      "loss": 0.727,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06781216540049967,
+      "grad_norm": 54.34592659188367,
+      "learning_rate": 4.981356433453238e-06,
+      "loss": 0.6632,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06832203130576658,
+      "grad_norm": 9.535587626528804,
+      "learning_rate": 4.9808497973021146e-06,
+      "loss": 0.7859,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0688318972110335,
+      "grad_norm": 16.49512859613625,
+      "learning_rate": 4.98033639570482e-06,
+      "loss": 0.7587,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06934176311630041,
+      "grad_norm": 14.125512381840414,
+      "learning_rate": 4.979816230061436e-06,
+      "loss": 0.7243,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06985162902156733,
+      "grad_norm": 17.80662494813945,
+      "learning_rate": 4.97928930179049e-06,
+      "loss": 0.7113,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07036149492683424,
+      "grad_norm": 18.57586486945156,
+      "learning_rate": 4.978755612328951e-06,
+      "loss": 0.7561,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07087136083210116,
+      "grad_norm": 8.283766059803575,
+      "learning_rate": 4.978215163132226e-06,
+      "loss": 0.7463,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07138122673736808,
+      "grad_norm": 14.653696663915117,
+      "learning_rate": 4.977667955674158e-06,
+      "loss": 0.7461,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07189109264263499,
+      "grad_norm": 13.507517215885278,
+      "learning_rate": 4.977113991447017e-06,
+      "loss": 0.7121,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0724009585479019,
+      "grad_norm": 9.288908938652135,
+      "learning_rate": 4.976553271961503e-06,
+      "loss": 0.8075,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07291082445316882,
+      "grad_norm": 21.236905071975933,
+      "learning_rate": 4.975985798746736e-06,
+      "loss": 0.7781,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07342069035843574,
+      "grad_norm": 13.572369167896932,
+      "learning_rate": 4.975411573350252e-06,
+      "loss": 0.7895,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07393055626370265,
+      "grad_norm": 13.55096914139037,
+      "learning_rate": 4.974830597338004e-06,
+      "loss": 0.736,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07444042216896957,
+      "grad_norm": 13.542801484459083,
+      "learning_rate": 4.974242872294354e-06,
+      "loss": 0.7303,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07495028807423648,
+      "grad_norm": 40.44101408661269,
+      "learning_rate": 4.973648399822068e-06,
+      "loss": 0.7475,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0754601539795034,
+      "grad_norm": 25.97183290799639,
+      "learning_rate": 4.9730471815423124e-06,
+      "loss": 0.7864,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07597001988477031,
+      "grad_norm": 26.74042060051638,
+      "learning_rate": 4.972439219094649e-06,
+      "loss": 0.6983,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07647988579003721,
+      "grad_norm": 11.80538671359037,
+      "learning_rate": 4.971824514137035e-06,
+      "loss": 0.6881,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07698975169530413,
+      "grad_norm": 8.2222875767639,
+      "learning_rate": 4.971203068345811e-06,
+      "loss": 0.7066,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07749961760057104,
+      "grad_norm": 8.370648963437903,
+      "learning_rate": 4.970574883415704e-06,
+      "loss": 0.7455,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07800948350583796,
+      "grad_norm": 16.048489718931286,
+      "learning_rate": 4.969939961059814e-06,
+      "loss": 0.7628,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07851934941110487,
+      "grad_norm": 7.234659506066708,
+      "learning_rate": 4.969298303009621e-06,
+      "loss": 0.7262,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07902921531637179,
+      "grad_norm": 31.829241494176976,
+      "learning_rate": 4.968649911014967e-06,
+      "loss": 0.6757,
+      "step": 1550
+    },
+    {
+      "epoch": 0.0795390812216387,
+      "grad_norm": 19.35414711975273,
+      "learning_rate": 4.9679947868440625e-06,
+      "loss": 0.7596,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08004894712690562,
+      "grad_norm": 8.711492612161729,
+      "learning_rate": 4.967332932283476e-06,
+      "loss": 0.6841,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08055881303217254,
+      "grad_norm": 8.342431288830609,
+      "learning_rate": 4.966664349138129e-06,
+      "loss": 0.6874,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08106867893743945,
+      "grad_norm": 10.392016002654769,
+      "learning_rate": 4.9659890392312935e-06,
+      "loss": 0.7102,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08157854484270637,
+      "grad_norm": 7.781469416808468,
+      "learning_rate": 4.965307004404586e-06,
+      "loss": 0.7509,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08208841074797328,
+      "grad_norm": 7.184310384526179,
+      "learning_rate": 4.964618246517962e-06,
+      "loss": 0.7089,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0825982766532402,
+      "grad_norm": 28.801020894955254,
+      "learning_rate": 4.96392276744971e-06,
+      "loss": 0.6766,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08310814255850711,
+      "grad_norm": 11.015804077658716,
+      "learning_rate": 4.9632205690964505e-06,
+      "loss": 0.6858,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08361800846377403,
+      "grad_norm": 8.76079381481738,
+      "learning_rate": 4.962511653373124e-06,
+      "loss": 0.6786,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08412787436904094,
+      "grad_norm": 6.02245638719929,
+      "learning_rate": 4.961796022212994e-06,
+      "loss": 0.7022,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08463774027430786,
+      "grad_norm": 15.612214027251115,
+      "learning_rate": 4.961073677567634e-06,
+      "loss": 0.6816,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08514760617957477,
+      "grad_norm": 16.85101160492237,
+      "learning_rate": 4.960344621406927e-06,
+      "loss": 0.8359,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08565747208484169,
+      "grad_norm": 19.376664337749983,
+      "learning_rate": 4.959608855719059e-06,
+      "loss": 0.6726,
+      "step": 1680
+    },
+    {
+      "epoch": 0.0861673379901086,
+      "grad_norm": 21.493545021821046,
+      "learning_rate": 4.958866382510515e-06,
+      "loss": 0.7395,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08667720389537552,
+      "grad_norm": 23.71442741073946,
+      "learning_rate": 4.958117203806067e-06,
+      "loss": 0.6864,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08718706980064243,
+      "grad_norm": 33.989133046895674,
+      "learning_rate": 4.957361321648777e-06,
+      "loss": 0.6745,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08769693570590935,
+      "grad_norm": 26.31554851843956,
+      "learning_rate": 4.956598738099988e-06,
+      "loss": 0.7354,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08820680161117626,
+      "grad_norm": 29.998869144940723,
+      "learning_rate": 4.955829455239316e-06,
+      "loss": 0.7275,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08871666751644318,
+      "grad_norm": 23.577129784492897,
+      "learning_rate": 4.95505347516465e-06,
+      "loss": 0.7156,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0892265334217101,
+      "grad_norm": 12.752943472931328,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7123,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08973639932697701,
+      "grad_norm": 14.685020491827581,
+      "learning_rate": 4.953481431856189e-06,
+      "loss": 0.7625,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09024626523224392,
+      "grad_norm": 8.228343997618332,
+      "learning_rate": 4.952685372909465e-06,
+      "loss": 0.6541,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09075613113751084,
+      "grad_norm": 9.63598214058154,
+      "learning_rate": 4.951882625322871e-06,
+      "loss": 0.7265,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09126599704277776,
+      "grad_norm": 14.152116737412488,
+      "learning_rate": 4.951073191285555e-06,
+      "loss": 0.727,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09177586294804467,
+      "grad_norm": 17.44279955167637,
+      "learning_rate": 4.9502570730048995e-06,
+      "loss": 0.7907,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09228572885331159,
+      "grad_norm": 13.60771925339826,
+      "learning_rate": 4.949434272706514e-06,
+      "loss": 0.6227,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0927955947585785,
+      "grad_norm": 6.367815102413218,
+      "learning_rate": 4.9486047926342294e-06,
+      "loss": 0.6355,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0933054606638454,
+      "grad_norm": 15.713131366094526,
+      "learning_rate": 4.947768635050098e-06,
+      "loss": 0.6764,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09381532656911232,
+      "grad_norm": 8.563023673169502,
+      "learning_rate": 4.946925802234373e-06,
+      "loss": 0.6923,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09432519247437923,
+      "grad_norm": 7.363234533978469,
+      "learning_rate": 4.946076296485522e-06,
+      "loss": 0.6625,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09483505837964615,
+      "grad_norm": 20.01812955600952,
+      "learning_rate": 4.945220120120203e-06,
+      "loss": 0.6543,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09534492428491306,
+      "grad_norm": 13.172376589466024,
+      "learning_rate": 4.9443572754732675e-06,
+      "loss": 0.6707,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09585479019017998,
+      "grad_norm": 6.822990704870548,
+      "learning_rate": 4.943487764897749e-06,
+      "loss": 0.6938,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0963646560954469,
+      "grad_norm": 15.932520477808335,
+      "learning_rate": 4.942611590764866e-06,
+      "loss": 0.6879,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09687452200071381,
+      "grad_norm": 24.250098360288504,
+      "learning_rate": 4.941728755464003e-06,
+      "loss": 0.6847,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09738438790598072,
+      "grad_norm": 17.44623443605144,
+      "learning_rate": 4.940839261402711e-06,
+      "loss": 0.7219,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09789425381124764,
+      "grad_norm": 12.45136746215644,
+      "learning_rate": 4.939943111006702e-06,
+      "loss": 0.6549,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09840411971651455,
+      "grad_norm": 18.583572409276037,
+      "learning_rate": 4.93904030671984e-06,
+      "loss": 0.6624,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09891398562178147,
+      "grad_norm": 7.2053286836275605,
+      "learning_rate": 4.938130851004131e-06,
+      "loss": 0.7675,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09942385152704838,
+      "grad_norm": 12.915641303278997,
+      "learning_rate": 4.937214746339726e-06,
+      "loss": 0.7149,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0999337174323153,
+      "grad_norm": 11.31073060509851,
+      "learning_rate": 4.936291995224902e-06,
+      "loss": 0.7683,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10044358333758222,
+      "grad_norm": 6.40606866544193,
+      "learning_rate": 4.935362600176064e-06,
+      "loss": 0.7258,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10095344924284913,
+      "grad_norm": 11.827179756342,
+      "learning_rate": 4.934426563727739e-06,
+      "loss": 0.691,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10146331514811605,
+      "grad_norm": 13.898857798518174,
+      "learning_rate": 4.933483888432558e-06,
+      "loss": 0.6964,
+      "step": 1990
+    },
+    {
+      "epoch": 0.10197318105338296,
+      "grad_norm": 14.73925860829306,
+      "learning_rate": 4.932534576861263e-06,
+      "loss": 0.6815,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10248304695864988,
+      "grad_norm": 10.503261101844386,
+      "learning_rate": 4.931578631602691e-06,
+      "loss": 0.7233,
+      "step": 2010
+    },
+    {
+      "epoch": 0.10299291286391679,
+      "grad_norm": 19.799601878435443,
+      "learning_rate": 4.930616055263768e-06,
+      "loss": 0.6772,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1035027787691837,
+      "grad_norm": 16.215950711270004,
+      "learning_rate": 4.9296468504695075e-06,
+      "loss": 0.6837,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10401264467445062,
+      "grad_norm": 6.487253559165473,
+      "learning_rate": 4.928671019862995e-06,
+      "loss": 0.6732,
+      "step": 2040
+    },
+    {
+      "epoch": 0.10452251057971754,
+      "grad_norm": 66.35910128060324,
+      "learning_rate": 4.927688566105388e-06,
+      "loss": 0.6883,
+      "step": 2050
+    },
+    {
+      "epoch": 0.10503237648498445,
+      "grad_norm": 22.70308591251781,
+      "learning_rate": 4.926699491875905e-06,
+      "loss": 0.73,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10554224239025137,
+      "grad_norm": 14.473695567395094,
+      "learning_rate": 4.925703799871818e-06,
+      "loss": 0.6991,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10605210829551828,
+      "grad_norm": 15.581472109798852,
+      "learning_rate": 4.924701492808447e-06,
+      "loss": 0.6915,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1065619742007852,
+      "grad_norm": 7.980568405700353,
+      "learning_rate": 4.923692573419152e-06,
+      "loss": 0.6913,
+      "step": 2090
+    },
+    {
+      "epoch": 0.10707184010605211,
+      "grad_norm": 23.31011562340674,
+      "learning_rate": 4.922677044455324e-06,
+      "loss": 0.7805,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10758170601131903,
+      "grad_norm": 14.87523420325485,
+      "learning_rate": 4.921654908686381e-06,
+      "loss": 0.6911,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10809157191658594,
+      "grad_norm": 14.14543897390739,
+      "learning_rate": 4.920626168899755e-06,
+      "loss": 0.7109,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10860143782185286,
+      "grad_norm": 10.881342080162556,
+      "learning_rate": 4.91959082790089e-06,
+      "loss": 0.7369,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10911130372711977,
+      "grad_norm": 21.17172349808835,
+      "learning_rate": 4.918548888513232e-06,
+      "loss": 0.6307,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10962116963238669,
+      "grad_norm": 9.13096463033165,
+      "learning_rate": 4.91750035357822e-06,
+      "loss": 0.6476,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11013103553765359,
+      "grad_norm": 11.89326391176972,
+      "learning_rate": 4.9164452259552805e-06,
+      "loss": 0.7066,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1106409014429205,
+      "grad_norm": 8.975708683266294,
+      "learning_rate": 4.9153835085218175e-06,
+      "loss": 0.7041,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11115076734818742,
+      "grad_norm": 8.585258277188357,
+      "learning_rate": 4.9143152041732074e-06,
+      "loss": 0.6879,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11166063325345434,
+      "grad_norm": 11.58529748100363,
+      "learning_rate": 4.91324031582279e-06,
+      "loss": 0.7314,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11217049915872125,
+      "grad_norm": 11.908156658951144,
+      "learning_rate": 4.9121588464018555e-06,
+      "loss": 0.663,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11268036506398817,
+      "grad_norm": 10.24067908361276,
+      "learning_rate": 4.911070798859647e-06,
+      "loss": 0.6701,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11319023096925508,
+      "grad_norm": 23.539758621052474,
+      "learning_rate": 4.909976176163345e-06,
+      "loss": 0.6944,
+      "step": 2220
+    },
+    {
+      "epoch": 0.113700096874522,
+      "grad_norm": 19.14611387579224,
+      "learning_rate": 4.908874981298058e-06,
+      "loss": 0.7148,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11420996277978891,
+      "grad_norm": 11.887671840791135,
+      "learning_rate": 4.90776721726682e-06,
+      "loss": 0.6832,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11471982868505583,
+      "grad_norm": 25.33947936645936,
+      "learning_rate": 4.90665288709058e-06,
+      "loss": 0.6461,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11522969459032274,
+      "grad_norm": 19.152392455583414,
+      "learning_rate": 4.905531993808191e-06,
+      "loss": 0.6192,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11573956049558966,
+      "grad_norm": 15.300011798292733,
+      "learning_rate": 4.904404540476405e-06,
+      "loss": 0.6853,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11624942640085657,
+      "grad_norm": 10.628181469407775,
+      "learning_rate": 4.903270530169865e-06,
+      "loss": 0.666,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11675929230612349,
+      "grad_norm": 19.385929263319337,
+      "learning_rate": 4.902129965981094e-06,
+      "loss": 0.7041,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1172691582113904,
+      "grad_norm": 57.98147128023247,
+      "learning_rate": 4.900982851020487e-06,
+      "loss": 0.7214,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11777902411665732,
+      "grad_norm": 14.01968653085331,
+      "learning_rate": 4.899829188416306e-06,
+      "loss": 0.653,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11828889002192423,
+      "grad_norm": 21.742692522690568,
+      "learning_rate": 4.898668981314667e-06,
+      "loss": 0.6654,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11879875592719115,
+      "grad_norm": 11.319460979734869,
+      "learning_rate": 4.897502232879533e-06,
+      "loss": 0.716,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11930862183245806,
+      "grad_norm": 13.991539570566289,
+      "learning_rate": 4.896328946292706e-06,
+      "loss": 0.6435,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11981848773772498,
+      "grad_norm": 21.2320524992128,
+      "learning_rate": 4.895149124753821e-06,
+      "loss": 0.6812,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1203283536429919,
+      "grad_norm": 10.28385295525001,
+      "learning_rate": 4.893962771480329e-06,
+      "loss": 0.6905,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12083821954825881,
+      "grad_norm": 17.31710761218729,
+      "learning_rate": 4.892769889707497e-06,
+      "loss": 0.7062,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12134808545352573,
+      "grad_norm": 13.401798655701699,
+      "learning_rate": 4.891570482688395e-06,
+      "loss": 0.6135,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12185795135879264,
+      "grad_norm": 26.789093785209946,
+      "learning_rate": 4.890364553693886e-06,
+      "loss": 0.6903,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12236781726405956,
+      "grad_norm": 9.750731917531368,
+      "learning_rate": 4.889152106012623e-06,
+      "loss": 0.655,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12287768316932647,
+      "grad_norm": 26.758659449058616,
+      "learning_rate": 4.88793314295103e-06,
+      "loss": 0.6116,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12338754907459339,
+      "grad_norm": 13.440739094685654,
+      "learning_rate": 4.886707667833306e-06,
+      "loss": 0.6829,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1238974149798603,
+      "grad_norm": 49.978223621638385,
+      "learning_rate": 4.885475684001401e-06,
+      "loss": 0.6415,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12440728088512722,
+      "grad_norm": 9.760273316885899,
+      "learning_rate": 4.884237194815023e-06,
+      "loss": 0.6308,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12491714679039413,
+      "grad_norm": 27.651331346216452,
+      "learning_rate": 4.882992203651613e-06,
+      "loss": 0.7362,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12542701269566103,
+      "grad_norm": 13.634588558328543,
+      "learning_rate": 4.881740713906348e-06,
+      "loss": 0.7433,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12593687860092795,
+      "grad_norm": 13.602097173034727,
+      "learning_rate": 4.880482728992126e-06,
+      "loss": 0.7717,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12644674450619486,
+      "grad_norm": 13.612161583437892,
+      "learning_rate": 4.8792182523395555e-06,
+      "loss": 0.644,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12695661041146178,
+      "grad_norm": 12.619409906208562,
+      "learning_rate": 4.877947287396952e-06,
+      "loss": 0.6215,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1274664763167287,
+      "grad_norm": 10.09193885674217,
+      "learning_rate": 4.876669837630324e-06,
+      "loss": 0.691,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1279763422219956,
+      "grad_norm": 14.76981467720561,
+      "learning_rate": 4.875385906523361e-06,
+      "loss": 0.6645,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12848620812726252,
+      "grad_norm": 7.371746141172926,
+      "learning_rate": 4.874095497577434e-06,
+      "loss": 0.6466,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12899607403252944,
+      "grad_norm": 6.835699156756462,
+      "learning_rate": 4.872798614311574e-06,
+      "loss": 0.6535,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12950593993779635,
+      "grad_norm": 21.566628352018,
+      "learning_rate": 4.87149526026247e-06,
+      "loss": 0.6597,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13001580584306327,
+      "grad_norm": 15.703751182885975,
+      "learning_rate": 4.870185438984458e-06,
+      "loss": 0.6666,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13052567174833019,
+      "grad_norm": 9.583980407078485,
+      "learning_rate": 4.868869154049509e-06,
+      "loss": 0.6803,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1310355376535971,
+      "grad_norm": 30.731237091152327,
+      "learning_rate": 4.867546409047221e-06,
+      "loss": 0.7388,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13154540355886402,
+      "grad_norm": 14.357083066524277,
+      "learning_rate": 4.866217207584811e-06,
+      "loss": 0.6337,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13205526946413093,
+      "grad_norm": 8.506886453649118,
+      "learning_rate": 4.864881553287101e-06,
+      "loss": 0.5825,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13256513536939785,
+      "grad_norm": 9.014994788890833,
+      "learning_rate": 4.863539449796511e-06,
+      "loss": 0.6527,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13307500127466476,
+      "grad_norm": 6.859684932885404,
+      "learning_rate": 4.86219090077305e-06,
+      "loss": 0.6557,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13358486717993168,
+      "grad_norm": 10.206698448625279,
+      "learning_rate": 4.8608359098943014e-06,
+      "loss": 0.6416,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1340947330851986,
+      "grad_norm": 16.90354656461581,
+      "learning_rate": 4.859474480855417e-06,
+      "loss": 0.6386,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1346045989904655,
+      "grad_norm": 8.345581201286654,
+      "learning_rate": 4.858106617369108e-06,
+      "loss": 0.6757,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13511446489573242,
+      "grad_norm": 13.073017499813815,
+      "learning_rate": 4.85673232316563e-06,
+      "loss": 0.7058,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13562433080099934,
+      "grad_norm": 12.30994463616197,
+      "learning_rate": 4.855351601992777e-06,
+      "loss": 0.7062,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13613419670626625,
+      "grad_norm": 18.52389808957781,
+      "learning_rate": 4.853964457615871e-06,
+      "loss": 0.659,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13664406261153317,
+      "grad_norm": 8.300091892691642,
+      "learning_rate": 4.852570893817747e-06,
+      "loss": 0.6531,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13715392851680008,
+      "grad_norm": 12.03361121338335,
+      "learning_rate": 4.851170914398749e-06,
+      "loss": 0.645,
+      "step": 2690
+    },
+    {
+      "epoch": 0.137663794422067,
+      "grad_norm": 15.761500517093225,
+      "learning_rate": 4.849764523176716e-06,
+      "loss": 0.7158,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1381736603273339,
+      "grad_norm": 9.61582603163959,
+      "learning_rate": 4.848351723986974e-06,
+      "loss": 0.7035,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13868352623260083,
+      "grad_norm": 12.19449916010554,
+      "learning_rate": 4.84693252068232e-06,
+      "loss": 0.6123,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13919339213786774,
+      "grad_norm": 10.928920087576591,
+      "learning_rate": 4.845506917133021e-06,
+      "loss": 0.6498,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13970325804313466,
+      "grad_norm": 14.276233849917213,
+      "learning_rate": 4.844074917226792e-06,
+      "loss": 0.6456,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14021312394840157,
+      "grad_norm": 11.803572897065552,
+      "learning_rate": 4.842636524868796e-06,
+      "loss": 0.6318,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1407229898536685,
+      "grad_norm": 7.925735834843732,
+      "learning_rate": 4.8411917439816245e-06,
+      "loss": 0.6248,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1412328557589354,
+      "grad_norm": 11.37915018985988,
+      "learning_rate": 4.839740578505297e-06,
+      "loss": 0.6717,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14174272166420232,
+      "grad_norm": 28.129852376592233,
+      "learning_rate": 4.838283032397237e-06,
+      "loss": 0.7101,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14225258756946924,
+      "grad_norm": 17.754960236458654,
+      "learning_rate": 4.8368191096322734e-06,
+      "loss": 0.7285,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14276245347473615,
+      "grad_norm": 11.980565371011531,
+      "learning_rate": 4.835348814202624e-06,
+      "loss": 0.6997,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14327231938000307,
+      "grad_norm": 11.536185693713957,
+      "learning_rate": 4.833872150117883e-06,
+      "loss": 0.6932,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14378218528526998,
+      "grad_norm": 8.134945726109256,
+      "learning_rate": 4.832389121405013e-06,
+      "loss": 0.7011,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1442920511905369,
+      "grad_norm": 13.154971513839428,
+      "learning_rate": 4.830899732108337e-06,
+      "loss": 0.6592,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1448019170958038,
+      "grad_norm": 16.74245961522776,
+      "learning_rate": 4.829403986289519e-06,
+      "loss": 0.6715,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14531178300107073,
+      "grad_norm": 13.034354542099656,
+      "learning_rate": 4.827901888027561e-06,
+      "loss": 0.6125,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14582164890633764,
+      "grad_norm": 27.871553670050698,
+      "learning_rate": 4.826393441418785e-06,
+      "loss": 0.7305,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14633151481160456,
+      "grad_norm": 15.95016686949317,
+      "learning_rate": 4.824878650576829e-06,
+      "loss": 0.676,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14684138071687147,
+      "grad_norm": 7.303492060471926,
+      "learning_rate": 4.823357519632631e-06,
+      "loss": 0.6396,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1473512466221384,
+      "grad_norm": 9.781495067337557,
+      "learning_rate": 4.821830052734418e-06,
+      "loss": 0.6509,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1478611125274053,
+      "grad_norm": 29.408307046395446,
+      "learning_rate": 4.820296254047695e-06,
+      "loss": 0.6778,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14837097843267222,
+      "grad_norm": 96.84824967607089,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.7329,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14888084433793913,
+      "grad_norm": 6.615844488316659,
+      "learning_rate": 4.817209678057073e-06,
+      "loss": 0.6763,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14939071024320605,
+      "grad_norm": 19.12515725566358,
+      "learning_rate": 4.815656909170476e-06,
+      "loss": 0.6168,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14990057614847296,
+      "grad_norm": 23.734925698529885,
+      "learning_rate": 4.814097825329953e-06,
+      "loss": 0.6631,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15041044205373988,
+      "grad_norm": 11.93938418941973,
+      "learning_rate": 4.81253243078723e-06,
+      "loss": 0.6259,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1509203079590068,
+      "grad_norm": 62.39893614222443,
+      "learning_rate": 4.810960729811247e-06,
+      "loss": 0.6687,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1514301738642737,
+      "grad_norm": 17.461388805483967,
+      "learning_rate": 4.8093827266881375e-06,
+      "loss": 0.6018,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15194003976954062,
+      "grad_norm": 11.072920322971628,
+      "learning_rate": 4.807798425721224e-06,
+      "loss": 0.6557,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1524499056748075,
+      "grad_norm": 72.73945006415794,
+      "learning_rate": 4.806207831231004e-06,
+      "loss": 0.6651,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15295977158007443,
+      "grad_norm": 136.05019813495343,
+      "learning_rate": 4.804610947555135e-06,
+      "loss": 0.6657,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15346963748534134,
+      "grad_norm": 49.62155036405917,
+      "learning_rate": 4.803007779048428e-06,
+      "loss": 0.6058,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15397950339060826,
+      "grad_norm": 150.29526061483784,
+      "learning_rate": 4.801398330082834e-06,
+      "loss": 0.7119,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15448936929587517,
+      "grad_norm": 39.16164201556088,
+      "learning_rate": 4.7997826050474284e-06,
+      "loss": 0.6403,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1549992352011421,
+      "grad_norm": 13.033901470600805,
+      "learning_rate": 4.798160608348404e-06,
+      "loss": 0.669,
+      "step": 3040
+    },
+    {
+      "epoch": 0.155509101106409,
+      "grad_norm": 10.436175563590364,
+      "learning_rate": 4.796532344409055e-06,
+      "loss": 0.6787,
+      "step": 3050
+    },
+    {
+      "epoch": 0.15601896701167592,
+      "grad_norm": 8.593652019091126,
+      "learning_rate": 4.794897817669769e-06,
+      "loss": 0.679,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15652883291694283,
+      "grad_norm": 73.66768580455573,
+      "learning_rate": 4.7932570325880114e-06,
+      "loss": 0.637,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15703869882220975,
+      "grad_norm": 9.16547202941871,
+      "learning_rate": 4.791609993638315e-06,
+      "loss": 0.6113,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15754856472747666,
+      "grad_norm": 27.807854408085937,
+      "learning_rate": 4.789956705312266e-06,
+      "loss": 0.7046,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15805843063274358,
+      "grad_norm": 158.66445861192776,
+      "learning_rate": 4.7882971721184955e-06,
+      "loss": 0.6328,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1585682965380105,
+      "grad_norm": 8.77759522628308,
+      "learning_rate": 4.786631398582663e-06,
+      "loss": 0.5866,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1590781624432774,
+      "grad_norm": 20.82303103042944,
+      "learning_rate": 4.784959389247445e-06,
+      "loss": 0.6648,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15958802834854433,
+      "grad_norm": 30.15207335672682,
+      "learning_rate": 4.7832811486725275e-06,
+      "loss": 0.7311,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16009789425381124,
+      "grad_norm": 48.14398580346578,
+      "learning_rate": 4.781596681434584e-06,
+      "loss": 0.6556,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16060776015907816,
+      "grad_norm": 240.90567407324733,
+      "learning_rate": 4.779905992127273e-06,
+      "loss": 0.6415,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16111762606434507,
+      "grad_norm": 13.587128797890388,
+      "learning_rate": 4.778209085361217e-06,
+      "loss": 0.7353,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16162749196961199,
+      "grad_norm": 26.152194406935024,
+      "learning_rate": 4.776505965763999e-06,
+      "loss": 0.6146,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1621373578748789,
+      "grad_norm": 28.08752066614632,
+      "learning_rate": 4.77479663798014e-06,
+      "loss": 0.6987,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16264722378014582,
+      "grad_norm": 18.4700415367194,
+      "learning_rate": 4.773081106671094e-06,
+      "loss": 0.6776,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16315708968541273,
+      "grad_norm": 26.687478554750435,
+      "learning_rate": 4.771359376515231e-06,
+      "loss": 0.6865,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16366695559067965,
+      "grad_norm": 10.56334023559032,
+      "learning_rate": 4.769631452207828e-06,
+      "loss": 0.6037,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16417682149594656,
+      "grad_norm": 30.24248577609317,
+      "learning_rate": 4.76789733846105e-06,
+      "loss": 0.6766,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16468668740121348,
+      "grad_norm": 6.118922498908346,
+      "learning_rate": 4.766157040003944e-06,
+      "loss": 0.576,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1651965533064804,
+      "grad_norm": 43.10186647344212,
+      "learning_rate": 4.7644105615824226e-06,
+      "loss": 0.6617,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1657064192117473,
+      "grad_norm": 22.58741112886033,
+      "learning_rate": 4.7626579079592504e-06,
+      "loss": 0.6257,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16621628511701422,
+      "grad_norm": 12.207542296775456,
+      "learning_rate": 4.760899083914035e-06,
+      "loss": 0.6313,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16672615102228114,
+      "grad_norm": 45.71668407841048,
+      "learning_rate": 4.759134094243206e-06,
+      "loss": 0.6591,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16723601692754805,
+      "grad_norm": 14.848732025059714,
+      "learning_rate": 4.757362943760013e-06,
+      "loss": 0.5773,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16774588283281497,
+      "grad_norm": 17.646815540917046,
+      "learning_rate": 4.755585637294503e-06,
+      "loss": 0.6431,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16825574873808188,
+      "grad_norm": 51.21492133232567,
+      "learning_rate": 4.753802179693512e-06,
+      "loss": 0.6482,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1687656146433488,
+      "grad_norm": 19.112461161721722,
+      "learning_rate": 4.7520125758206495e-06,
+      "loss": 0.6671,
+      "step": 3310
+    },
+    {
+      "epoch": 0.16927548054861571,
+      "grad_norm": 16.191200196997656,
+      "learning_rate": 4.750216830556287e-06,
+      "loss": 0.6296,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16978534645388263,
+      "grad_norm": 27.51474330611919,
+      "learning_rate": 4.748414948797545e-06,
+      "loss": 0.6612,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17029521235914954,
+      "grad_norm": 23.75272142335703,
+      "learning_rate": 4.746606935458277e-06,
+      "loss": 0.641,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17080507826441646,
+      "grad_norm": 11.809633635411776,
+      "learning_rate": 4.744792795469058e-06,
+      "loss": 0.6834,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17131494416968338,
+      "grad_norm": 189.33582561579482,
+      "learning_rate": 4.742972533777172e-06,
+      "loss": 0.6559,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1718248100749503,
+      "grad_norm": 11.941107362922496,
+      "learning_rate": 4.741146155346596e-06,
+      "loss": 0.6373,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1723346759802172,
+      "grad_norm": 7.436609726007049,
+      "learning_rate": 4.739313665157988e-06,
+      "loss": 0.6129,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17284454188548412,
+      "grad_norm": 23.5332384240844,
+      "learning_rate": 4.737475068208673e-06,
+      "loss": 0.6544,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17335440779075104,
+      "grad_norm": 26.16689346814709,
+      "learning_rate": 4.7356303695126315e-06,
+      "loss": 0.6598,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17386427369601795,
+      "grad_norm": 8.349460123034687,
+      "learning_rate": 4.733779574100482e-06,
+      "loss": 0.6285,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17437413960128487,
+      "grad_norm": 5.208194637835858,
+      "learning_rate": 4.7319226870194676e-06,
+      "loss": 0.6597,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17488400550655178,
+      "grad_norm": 10.345782004711861,
+      "learning_rate": 4.730059713333448e-06,
+      "loss": 0.6892,
+      "step": 3430
+    },
+    {
+      "epoch": 0.1753938714118187,
+      "grad_norm": 21.377764003397456,
+      "learning_rate": 4.728190658122878e-06,
+      "loss": 0.58,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1759037373170856,
+      "grad_norm": 30.6794980257094,
+      "learning_rate": 4.726315526484799e-06,
+      "loss": 0.6423,
+      "step": 3450
+    },
+    {
+      "epoch": 0.17641360322235253,
+      "grad_norm": 13.709883604900671,
+      "learning_rate": 4.724434323532822e-06,
+      "loss": 0.6154,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17692346912761944,
+      "grad_norm": 8.479320584002648,
+      "learning_rate": 4.722547054397114e-06,
+      "loss": 0.628,
+      "step": 3470
+    },
+    {
+      "epoch": 0.17743333503288636,
+      "grad_norm": 13.685563027122981,
+      "learning_rate": 4.720653724224389e-06,
+      "loss": 0.6556,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17794320093815327,
+      "grad_norm": 8.260254371701032,
+      "learning_rate": 4.718754338177887e-06,
+      "loss": 0.5759,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1784530668434202,
+      "grad_norm": 13.913564384357112,
+      "learning_rate": 4.716848901437361e-06,
+      "loss": 0.6285,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1789629327486871,
+      "grad_norm": 45.21035989952079,
+      "learning_rate": 4.714937419199067e-06,
+      "loss": 0.6304,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17947279865395402,
+      "grad_norm": 17.61340217915247,
+      "learning_rate": 4.713019896675749e-06,
+      "loss": 0.7353,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17998266455922093,
+      "grad_norm": 51.73719722197582,
+      "learning_rate": 4.711096339096619e-06,
+      "loss": 0.6661,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18049253046448785,
+      "grad_norm": 41.42271286545349,
+      "learning_rate": 4.709166751707351e-06,
+      "loss": 0.6626,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18100239636975476,
+      "grad_norm": 7.6574270650897605,
+      "learning_rate": 4.7072311397700605e-06,
+      "loss": 0.6286,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18151226227502168,
+      "grad_norm": 55.46468850803111,
+      "learning_rate": 4.705289508563293e-06,
+      "loss": 0.6502,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1820221281802886,
+      "grad_norm": 23.031046436781327,
+      "learning_rate": 4.703341863382009e-06,
+      "loss": 0.6367,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1825319940855555,
+      "grad_norm": 8.278399987077327,
+      "learning_rate": 4.701388209537569e-06,
+      "loss": 0.654,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18304185999082243,
+      "grad_norm": 16.796980125582998,
+      "learning_rate": 4.69942855235772e-06,
+      "loss": 0.6853,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18355172589608934,
+      "grad_norm": 14.118030028068347,
+      "learning_rate": 4.697462897186581e-06,
+      "loss": 0.6337,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18406159180135626,
+      "grad_norm": 15.497810143418386,
+      "learning_rate": 4.695491249384628e-06,
+      "loss": 0.6486,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18457145770662317,
+      "grad_norm": 18.32562759382926,
+      "learning_rate": 4.693513614328676e-06,
+      "loss": 0.6651,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1850813236118901,
+      "grad_norm": 10.993090480177178,
+      "learning_rate": 4.691529997411873e-06,
+      "loss": 0.7046,
+      "step": 3630
+    },
+    {
+      "epoch": 0.185591189517157,
+      "grad_norm": 9.354322331456606,
+      "learning_rate": 4.689540404043677e-06,
+      "loss": 0.6626,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1861010554224239,
+      "grad_norm": 28.01431660577904,
+      "learning_rate": 4.687544839649842e-06,
+      "loss": 0.6499,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1866109213276908,
+      "grad_norm": 9.512985903502441,
+      "learning_rate": 4.685543309672411e-06,
+      "loss": 0.5949,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18712078723295772,
+      "grad_norm": 24.4025944922159,
+      "learning_rate": 4.683535819569691e-06,
+      "loss": 0.6727,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18763065313822463,
+      "grad_norm": 50.8641661183589,
+      "learning_rate": 4.681522374816244e-06,
+      "loss": 0.6116,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18814051904349155,
+      "grad_norm": 17.587190798690532,
+      "learning_rate": 4.679502980902871e-06,
+      "loss": 0.6195,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18865038494875846,
+      "grad_norm": 18.670293843348876,
+      "learning_rate": 4.6774776433365965e-06,
+      "loss": 0.6008,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18916025085402538,
+      "grad_norm": 10.999046092407664,
+      "learning_rate": 4.6754463676406545e-06,
+      "loss": 0.6525,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1896701167592923,
+      "grad_norm": 16.578146994304092,
+      "learning_rate": 4.6734091593544705e-06,
+      "loss": 0.6329,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1901799826645592,
+      "grad_norm": 14.938986449153218,
+      "learning_rate": 4.671366024033651e-06,
+      "loss": 0.6031,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19068984856982613,
+      "grad_norm": 53.66910906959681,
+      "learning_rate": 4.669316967249966e-06,
+      "loss": 0.6748,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19119971447509304,
+      "grad_norm": 38.64005506480217,
+      "learning_rate": 4.667261994591331e-06,
+      "loss": 0.643,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19170958038035996,
+      "grad_norm": 20.83213603104761,
+      "learning_rate": 4.665201111661797e-06,
+      "loss": 0.6829,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19221944628562687,
+      "grad_norm": 7.6025512197039085,
+      "learning_rate": 4.663134324081533e-06,
+      "loss": 0.6365,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1927293121908938,
+      "grad_norm": 17.681281980648198,
+      "learning_rate": 4.6610616374868066e-06,
+      "loss": 0.7101,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1932391780961607,
+      "grad_norm": 5.961528341747031,
+      "learning_rate": 4.658983057529978e-06,
+      "loss": 0.5333,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19374904400142762,
+      "grad_norm": 8.500212586402798,
+      "learning_rate": 4.656898589879475e-06,
+      "loss": 0.6249,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19425890990669453,
+      "grad_norm": 7.489518295782326,
+      "learning_rate": 4.654808240219782e-06,
+      "loss": 0.6581,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19476877581196145,
+      "grad_norm": 6.794369508294625,
+      "learning_rate": 4.652712014251426e-06,
+      "loss": 0.6068,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19527864171722836,
+      "grad_norm": 26.68786642749212,
+      "learning_rate": 4.650609917690957e-06,
+      "loss": 0.5868,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19578850762249528,
+      "grad_norm": 24.634611602664084,
+      "learning_rate": 4.648501956270936e-06,
+      "loss": 0.6491,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1962983735277622,
+      "grad_norm": 7.916826238082026,
+      "learning_rate": 4.646388135739915e-06,
+      "loss": 0.6238,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1968082394330291,
+      "grad_norm": 21.827979294670318,
+      "learning_rate": 4.64426846186243e-06,
+      "loss": 0.6028,
+      "step": 3860
+    },
+    {
+      "epoch": 0.19731810533829602,
+      "grad_norm": 10.817798837200906,
+      "learning_rate": 4.642142940418973e-06,
+      "loss": 0.7177,
+      "step": 3870
+    },
+    {
+      "epoch": 0.19782797124356294,
+      "grad_norm": 7.343593825755392,
+      "learning_rate": 4.640011577205987e-06,
+      "loss": 0.6465,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19833783714882985,
+      "grad_norm": 9.578351561035468,
+      "learning_rate": 4.637874378035845e-06,
+      "loss": 0.6709,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19884770305409677,
+      "grad_norm": 11.509018820886228,
+      "learning_rate": 4.635731348736832e-06,
+      "loss": 0.6401,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19935756895936368,
+      "grad_norm": 7.148943238715311,
+      "learning_rate": 4.633582495153137e-06,
+      "loss": 0.6461,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1998674348646306,
+      "grad_norm": 6.102496878351305,
+      "learning_rate": 4.631427823144829e-06,
+      "loss": 0.6007,
+      "step": 3920
+    },
+    {
+      "epoch": 0.20037730076989752,
+      "grad_norm": 12.872084204664713,
+      "learning_rate": 4.6292673385878466e-06,
+      "loss": 0.6813,
+      "step": 3930
+    },
+    {
+      "epoch": 0.20088716667516443,
+      "grad_norm": 28.92020036551762,
+      "learning_rate": 4.6271010473739754e-06,
+      "loss": 0.7501,
+      "step": 3940
+    },
+    {
+      "epoch": 0.20139703258043135,
+      "grad_norm": 26.1419017695497,
+      "learning_rate": 4.624928955410841e-06,
+      "loss": 0.5818,
+      "step": 3950
+    },
+    {
+      "epoch": 0.20190689848569826,
+      "grad_norm": 8.520817831827328,
+      "learning_rate": 4.622751068621886e-06,
+      "loss": 0.619,
+      "step": 3960
+    },
+    {
+      "epoch": 0.20241676439096518,
+      "grad_norm": 10.60690821999077,
+      "learning_rate": 4.620567392946355e-06,
+      "loss": 0.6238,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2029266302962321,
+      "grad_norm": 6.111505810115736,
+      "learning_rate": 4.618377934339279e-06,
+      "loss": 0.6251,
+      "step": 3980
+    },
+    {
+      "epoch": 0.203436496201499,
+      "grad_norm": 11.501436686181162,
+      "learning_rate": 4.616182698771463e-06,
+      "loss": 0.59,
+      "step": 3990
+    },
+    {
+      "epoch": 0.20394636210676592,
+      "grad_norm": 21.165945980312195,
+      "learning_rate": 4.613981692229462e-06,
+      "loss": 0.6075,
+      "step": 4000
+    },
+    {
+      "epoch": 0.20445622801203284,
+      "grad_norm": 15.671878583013148,
+      "learning_rate": 4.611774920715572e-06,
+      "loss": 0.6139,
+      "step": 4010
+    },
+    {
+      "epoch": 0.20496609391729975,
+      "grad_norm": 10.969313336572679,
+      "learning_rate": 4.609562390247808e-06,
+      "loss": 0.6505,
+      "step": 4020
+    },
+    {
+      "epoch": 0.20547595982256667,
+      "grad_norm": 12.862905465982983,
+      "learning_rate": 4.607344106859891e-06,
+      "loss": 0.6395,
+      "step": 4030
+    },
+    {
+      "epoch": 0.20598582572783358,
+      "grad_norm": 11.382723251669816,
+      "learning_rate": 4.605120076601231e-06,
+      "loss": 0.6574,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2064956916331005,
+      "grad_norm": 35.899532032459916,
+      "learning_rate": 4.602890305536911e-06,
+      "loss": 0.665,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2070055575383674,
+      "grad_norm": 133.21898922751015,
+      "learning_rate": 4.6006547997476666e-06,
+      "loss": 0.6725,
+      "step": 4060
+    },
+    {
+      "epoch": 0.20751542344363433,
+      "grad_norm": 52.27751505587636,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.6175,
+      "step": 4070
+    },
+    {
+      "epoch": 0.20802528934890124,
+      "grad_norm": 22.607365372261278,
+      "learning_rate": 4.596166608395535e-06,
+      "loss": 0.6543,
+      "step": 4080
+    },
+    {
+      "epoch": 0.20853515525416816,
+      "grad_norm": 18.893786224066538,
+      "learning_rate": 4.593913935072251e-06,
+      "loss": 0.6588,
+      "step": 4090
+    },
+    {
+      "epoch": 0.20904502115943507,
+      "grad_norm": 36.925520109780834,
+      "learning_rate": 4.591655551503215e-06,
+      "loss": 0.6452,
+      "step": 4100
+    },
+    {
+      "epoch": 0.209554887064702,
+      "grad_norm": 18.989453127592082,
+      "learning_rate": 4.589391463847194e-06,
+      "loss": 0.7321,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2100647529699689,
+      "grad_norm": 21.559296328550772,
+      "learning_rate": 4.58712167827851e-06,
+      "loss": 0.6381,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21057461887523582,
+      "grad_norm": 7.4422471218088235,
+      "learning_rate": 4.584846200987022e-06,
+      "loss": 0.6695,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21108448478050273,
+      "grad_norm": 8.598103890275697,
+      "learning_rate": 4.582565038178109e-06,
+      "loss": 0.6393,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21159435068576965,
+      "grad_norm": 10.859833811937909,
+      "learning_rate": 4.58027819607266e-06,
+      "loss": 0.6611,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21210421659103657,
+      "grad_norm": 9.455817856990365,
+      "learning_rate": 4.577985680907049e-06,
+      "loss": 0.6438,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21261408249630348,
+      "grad_norm": 13.005173788491543,
+      "learning_rate": 4.575687498933119e-06,
+      "loss": 0.6639,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2131239484015704,
+      "grad_norm": 10.115654248919899,
+      "learning_rate": 4.573383656418169e-06,
+      "loss": 0.6944,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2136338143068373,
+      "grad_norm": 37.56438546169335,
+      "learning_rate": 4.571074159644936e-06,
+      "loss": 0.6497,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21414368021210423,
+      "grad_norm": 5.861898760817187,
+      "learning_rate": 4.568759014911573e-06,
+      "loss": 0.6125,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21465354611737114,
+      "grad_norm": 8.783410942135434,
+      "learning_rate": 4.566438228531638e-06,
+      "loss": 0.6775,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21516341202263806,
+      "grad_norm": 6.639944275088382,
+      "learning_rate": 4.564111806834073e-06,
+      "loss": 0.56,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21567327792790497,
+      "grad_norm": 6.214883955719362,
+      "learning_rate": 4.5617797561631885e-06,
+      "loss": 0.689,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2161831438331719,
+      "grad_norm": 10.229291895282838,
+      "learning_rate": 4.559442082878645e-06,
+      "loss": 0.641,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2166930097384388,
+      "grad_norm": 14.962817683964534,
+      "learning_rate": 4.557098793355436e-06,
+      "loss": 0.5747,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21720287564370572,
+      "grad_norm": 8.53213097749228,
+      "learning_rate": 4.554749893983874e-06,
+      "loss": 0.5826,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21771274154897263,
+      "grad_norm": 16.549243515768183,
+      "learning_rate": 4.552395391169564e-06,
+      "loss": 0.6758,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21822260745423955,
+      "grad_norm": 18.938696755433465,
+      "learning_rate": 4.550035291333398e-06,
+      "loss": 0.5707,
+      "step": 4280
+    },
+    {
+      "epoch": 0.21873247335950646,
+      "grad_norm": 35.395198976128206,
+      "learning_rate": 4.547669600911527e-06,
+      "loss": 0.6877,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21924233926477338,
+      "grad_norm": 8.277494608992038,
+      "learning_rate": 4.545298326355351e-06,
+      "loss": 0.629,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21975220517004027,
+      "grad_norm": 79.25983653731949,
+      "learning_rate": 4.542921474131497e-06,
+      "loss": 0.678,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22026207107530718,
+      "grad_norm": 9.856307804445956,
+      "learning_rate": 4.540539050721801e-06,
+      "loss": 0.676,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2207719369805741,
+      "grad_norm": 8.615502828092712,
+      "learning_rate": 4.538151062623296e-06,
+      "loss": 0.5812,
+      "step": 4330
+    },
+    {
+      "epoch": 0.221281802885841,
+      "grad_norm": 22.168060830576543,
+      "learning_rate": 4.535757516348186e-06,
+      "loss": 0.6231,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22179166879110793,
+      "grad_norm": 11.320048528710553,
+      "learning_rate": 4.533358418423837e-06,
+      "loss": 0.6178,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22230153469637484,
+      "grad_norm": 6.466250445752336,
+      "learning_rate": 4.530953775392749e-06,
+      "loss": 0.6187,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22281140060164176,
+      "grad_norm": 24.18925309517145,
+      "learning_rate": 4.52854359381255e-06,
+      "loss": 0.6119,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22332126650690867,
+      "grad_norm": 8.571095234461708,
+      "learning_rate": 4.5261278802559675e-06,
+      "loss": 0.6014,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2238311324121756,
+      "grad_norm": 5.982288593053768,
+      "learning_rate": 4.523706641310817e-06,
+      "loss": 0.6207,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2243409983174425,
+      "grad_norm": 19.10284583967391,
+      "learning_rate": 4.521279883579982e-06,
+      "loss": 0.5721,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22485086422270942,
+      "grad_norm": 19.058976511095665,
+      "learning_rate": 4.518847613681397e-06,
+      "loss": 0.6835,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22536073012797633,
+      "grad_norm": 17.345690564027926,
+      "learning_rate": 4.516409838248026e-06,
+      "loss": 0.6342,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22587059603324325,
+      "grad_norm": 18.002628820726464,
+      "learning_rate": 4.513966563927849e-06,
+      "loss": 0.6207,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22638046193851016,
+      "grad_norm": 10.76507448906313,
+      "learning_rate": 4.511517797383841e-06,
+      "loss": 0.6504,
+      "step": 4440
+    },
+    {
+      "epoch": 0.22689032784377708,
+      "grad_norm": 37.963122971372236,
+      "learning_rate": 4.509063545293954e-06,
+      "loss": 0.6718,
+      "step": 4450
+    },
+    {
+      "epoch": 0.227400193749044,
+      "grad_norm": 31.081526202955846,
+      "learning_rate": 4.506603814351103e-06,
+      "loss": 0.5657,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2279100596543109,
+      "grad_norm": 9.090363098538898,
+      "learning_rate": 4.5041386112631394e-06,
+      "loss": 0.6365,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22841992555957782,
+      "grad_norm": 29.102701668422192,
+      "learning_rate": 4.501667942752841e-06,
+      "loss": 0.565,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22892979146484474,
+      "grad_norm": 9.40166219606075,
+      "learning_rate": 4.499191815557888e-06,
+      "loss": 0.6468,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22943965737011165,
+      "grad_norm": 24.98331649199063,
+      "learning_rate": 4.496710236430848e-06,
+      "loss": 0.6094,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22994952327537857,
+      "grad_norm": 14.24156171953044,
+      "learning_rate": 4.4942232121391565e-06,
+      "loss": 0.618,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23045938918064549,
+      "grad_norm": 25.888527040924036,
+      "learning_rate": 4.4917307494650975e-06,
+      "loss": 0.5663,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2309692550859124,
+      "grad_norm": 11.32099980777747,
+      "learning_rate": 4.489232855205787e-06,
+      "loss": 0.6577,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23147912099117932,
+      "grad_norm": 7.52041696012304,
+      "learning_rate": 4.4867295361731515e-06,
+      "loss": 0.6623,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23198898689644623,
+      "grad_norm": 86.35249632825519,
+      "learning_rate": 4.484220799193913e-06,
+      "loss": 0.6061,
+      "step": 4550
+    },
+    {
+      "epoch": 0.23249885280171315,
+      "grad_norm": 11.304016464447672,
+      "learning_rate": 4.481706651109567e-06,
+      "loss": 0.5761,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23300871870698006,
+      "grad_norm": 18.086580599155557,
+      "learning_rate": 4.479187098776368e-06,
+      "loss": 0.6349,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23351858461224698,
+      "grad_norm": 12.005880118436764,
+      "learning_rate": 4.476662149065306e-06,
+      "loss": 0.6567,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2340284505175139,
+      "grad_norm": 13.992259943255434,
+      "learning_rate": 4.474131808862089e-06,
+      "loss": 0.7011,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2345383164227808,
+      "grad_norm": 22.858022758700816,
+      "learning_rate": 4.471596085067129e-06,
+      "loss": 0.6229,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23504818232804772,
+      "grad_norm": 17.13933779909115,
+      "learning_rate": 4.469054984595517e-06,
+      "loss": 0.5982,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23555804823331464,
+      "grad_norm": 11.220115247254418,
+      "learning_rate": 4.466508514377006e-06,
+      "loss": 0.6181,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23606791413858155,
+      "grad_norm": 8.87668293968533,
+      "learning_rate": 4.463956681355993e-06,
+      "loss": 0.6309,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23657778004384847,
+      "grad_norm": 8.331451854861665,
+      "learning_rate": 4.461399492491502e-06,
+      "loss": 0.6389,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23708764594911538,
+      "grad_norm": 25.657125436012297,
+      "learning_rate": 4.458836954757161e-06,
+      "loss": 0.6381,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2375975118543823,
+      "grad_norm": 18.084465806692844,
+      "learning_rate": 4.456269075141183e-06,
+      "loss": 0.6064,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2381073777596492,
+      "grad_norm": 12.540649074725133,
+      "learning_rate": 4.4536958606463506e-06,
+      "loss": 0.6177,
+      "step": 4670
+    },
+    {
+      "epoch": 0.23861724366491613,
+      "grad_norm": 10.689133786862486,
+      "learning_rate": 4.451117318289996e-06,
+      "loss": 0.646,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23912710957018304,
+      "grad_norm": 7.958824117114735,
+      "learning_rate": 4.448533455103979e-06,
+      "loss": 0.6104,
+      "step": 4690
+    },
+    {
+      "epoch": 0.23963697547544996,
+      "grad_norm": 10.36267483447428,
+      "learning_rate": 4.445944278134671e-06,
+      "loss": 0.6753,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24014684138071687,
+      "grad_norm": 11.175782587324658,
+      "learning_rate": 4.4433497944429325e-06,
+      "loss": 0.5777,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2406567072859838,
+      "grad_norm": 27.8631592366726,
+      "learning_rate": 4.440750011104098e-06,
+      "loss": 0.5881,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2411665731912507,
+      "grad_norm": 9.81409196930052,
+      "learning_rate": 4.438144935207953e-06,
+      "loss": 0.5875,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24167643909651762,
+      "grad_norm": 9.171704497921569,
+      "learning_rate": 4.435534573858717e-06,
+      "loss": 0.6058,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24218630500178454,
+      "grad_norm": 19.37721728099732,
+      "learning_rate": 4.432918934175023e-06,
+      "loss": 0.6167,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24269617090705145,
+      "grad_norm": 14.658589343659886,
+      "learning_rate": 4.430298023289897e-06,
+      "loss": 0.6457,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24320603681231837,
+      "grad_norm": 17.4652006391887,
+      "learning_rate": 4.427671848350744e-06,
+      "loss": 0.6085,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24371590271758528,
+      "grad_norm": 17.988496277713374,
+      "learning_rate": 4.425040416519319e-06,
+      "loss": 0.5356,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2442257686228522,
+      "grad_norm": 40.73874761370533,
+      "learning_rate": 4.422403734971718e-06,
+      "loss": 0.5709,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2447356345281191,
+      "grad_norm": 11.464972225824136,
+      "learning_rate": 4.419761810898349e-06,
+      "loss": 0.5991,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24524550043338603,
+      "grad_norm": 27.356260220415578,
+      "learning_rate": 4.4171146515039206e-06,
+      "loss": 0.6223,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24575536633865294,
+      "grad_norm": 23.849384135826682,
+      "learning_rate": 4.414462264007414e-06,
+      "loss": 0.6609,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24626523224391986,
+      "grad_norm": 28.00533721425015,
+      "learning_rate": 4.4118046556420725e-06,
+      "loss": 0.5876,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24677509814918677,
+      "grad_norm": 12.977686770179183,
+      "learning_rate": 4.409141833655375e-06,
+      "loss": 0.5651,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2472849640544537,
+      "grad_norm": 15.526266224793934,
+      "learning_rate": 4.406473805309016e-06,
+      "loss": 0.657,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2477948299597206,
+      "grad_norm": 12.62411203239521,
+      "learning_rate": 4.403800577878892e-06,
+      "loss": 0.5806,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24830469586498752,
+      "grad_norm": 21.678019580304962,
+      "learning_rate": 4.401122158655076e-06,
+      "loss": 0.6082,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24881456177025443,
+      "grad_norm": 15.217926723899035,
+      "learning_rate": 4.3984385549418e-06,
+      "loss": 0.6143,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24932442767552135,
+      "grad_norm": 9.534920956087968,
+      "learning_rate": 4.395749774057432e-06,
+      "loss": 0.6445,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24983429358078826,
+      "grad_norm": 20.694698271569113,
+      "learning_rate": 4.393055823334461e-06,
+      "loss": 0.6482,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25034415948605515,
+      "grad_norm": 10.440590354492034,
+      "learning_rate": 4.390356710119476e-06,
+      "loss": 0.644,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25085402539132207,
+      "grad_norm": 14.01886170757573,
+      "learning_rate": 4.38765244177314e-06,
+      "loss": 0.61,
+      "step": 4920
+    },
+    {
+      "epoch": 0.251363891296589,
+      "grad_norm": 12.121295206565357,
+      "learning_rate": 4.3849430256701765e-06,
+      "loss": 0.5708,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2518737572018559,
+      "grad_norm": 11.071313569953784,
+      "learning_rate": 4.38222846919935e-06,
+      "loss": 0.6051,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2523836231071228,
+      "grad_norm": 14.68565361360073,
+      "learning_rate": 4.379508779763438e-06,
+      "loss": 0.7019,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2528934890123897,
+      "grad_norm": 17.274945106451703,
+      "learning_rate": 4.376783964779221e-06,
+      "loss": 0.625,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25340335491765664,
+      "grad_norm": 5.602426927797183,
+      "learning_rate": 4.3740540316774535e-06,
+      "loss": 0.5799,
+      "step": 4970
+    },
+    {
+      "epoch": 0.25391322082292356,
+      "grad_norm": 7.201877021121772,
+      "learning_rate": 4.3713189879028485e-06,
+      "loss": 0.5485,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2544230867281905,
+      "grad_norm": 13.951892269527397,
+      "learning_rate": 4.3685788409140564e-06,
+      "loss": 0.645,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2549329526334574,
+      "grad_norm": 27.947954866045787,
+      "learning_rate": 4.365833598183645e-06,
+      "loss": 0.6404,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2554428185387243,
+      "grad_norm": 10.514953272318815,
+      "learning_rate": 4.363083267198079e-06,
+      "loss": 0.5863,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2559526844439912,
+      "grad_norm": 11.245703293221581,
+      "learning_rate": 4.360327855457696e-06,
+      "loss": 0.6097,
+      "step": 5020
+    },
+    {
+      "epoch": 0.25646255034925813,
+      "grad_norm": 8.599634462633494,
+      "learning_rate": 4.357567370476693e-06,
+      "loss": 0.5853,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25697241625452505,
+      "grad_norm": 6.095105421466429,
+      "learning_rate": 4.354801819783099e-06,
+      "loss": 0.5669,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25748228215979196,
+      "grad_norm": 7.616602354562931,
+      "learning_rate": 4.35203121091876e-06,
+      "loss": 0.5345,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2579921480650589,
+      "grad_norm": 23.42239940092673,
+      "learning_rate": 4.349255551439314e-06,
+      "loss": 0.6427,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2585020139703258,
+      "grad_norm": 22.048908516961838,
+      "learning_rate": 4.346474848914174e-06,
+      "loss": 0.6484,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2590118798755927,
+      "grad_norm": 9.251706608352913,
+      "learning_rate": 4.343689110926504e-06,
+      "loss": 0.64,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2595217457808596,
+      "grad_norm": 13.321339625150785,
+      "learning_rate": 4.340898345073202e-06,
+      "loss": 0.7136,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26003161168612654,
+      "grad_norm": 10.441597949681997,
+      "learning_rate": 4.338102558964876e-06,
+      "loss": 0.6608,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26054147759139346,
+      "grad_norm": 12.745953519040231,
+      "learning_rate": 4.335301760225824e-06,
+      "loss": 0.5851,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26105134349666037,
+      "grad_norm": 65.54384924061196,
+      "learning_rate": 4.3324959564940165e-06,
+      "loss": 0.5452,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2615612094019273,
+      "grad_norm": 6.375930327766279,
+      "learning_rate": 4.329685155421069e-06,
+      "loss": 0.5983,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2620710753071942,
+      "grad_norm": 3.6975754694433514,
+      "learning_rate": 4.326869364672229e-06,
+      "loss": 0.6113,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2625809412124611,
+      "grad_norm": 17.984395739677577,
+      "learning_rate": 4.324048591926349e-06,
+      "loss": 0.6007,
+      "step": 5150
+    },
+    {
+      "epoch": 0.26309080711772803,
+      "grad_norm": 7.288254040012573,
+      "learning_rate": 4.321222844875869e-06,
+      "loss": 0.6186,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26360067302299495,
+      "grad_norm": 5.127848398893114,
+      "learning_rate": 4.318392131226791e-06,
+      "loss": 0.6471,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26411053892826186,
+      "grad_norm": 7.5315117397182645,
+      "learning_rate": 4.315556458698665e-06,
+      "loss": 0.6237,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2646204048335288,
+      "grad_norm": 8.723509314324257,
+      "learning_rate": 4.312715835024565e-06,
+      "loss": 0.6453,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2651302707387957,
+      "grad_norm": 10.824429436685376,
+      "learning_rate": 4.309870267951061e-06,
+      "loss": 0.7068,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2656401366440626,
+      "grad_norm": 8.298367387251412,
+      "learning_rate": 4.30701976523821e-06,
+      "loss": 0.6317,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2661500025493295,
+      "grad_norm": 8.313379507545704,
+      "learning_rate": 4.3041643346595285e-06,
+      "loss": 0.5736,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26665986845459644,
+      "grad_norm": 17.233146557176756,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.6084,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26716973435986335,
+      "grad_norm": 16.00967256982166,
+      "learning_rate": 4.298438721065899e-06,
+      "loss": 0.6074,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26767960026513027,
+      "grad_norm": 15.624861264682277,
+      "learning_rate": 4.295568553665089e-06,
+      "loss": 0.6282,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2681894661703972,
+      "grad_norm": 21.12100148713912,
+      "learning_rate": 4.292693489626681e-06,
+      "loss": 0.6039,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2686993320756641,
+      "grad_norm": 7.6235473567087535,
+      "learning_rate": 4.289813536791168e-06,
+      "loss": 0.5629,
+      "step": 5270
+    },
+    {
+      "epoch": 0.269209197980931,
+      "grad_norm": 26.078669039774066,
+      "learning_rate": 4.2869287030123786e-06,
+      "loss": 0.6521,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26971906388619793,
+      "grad_norm": 12.187667815535157,
+      "learning_rate": 4.284038996157451e-06,
+      "loss": 0.6104,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27022892979146484,
+      "grad_norm": 7.139360580690056,
+      "learning_rate": 4.2811444241068115e-06,
+      "loss": 0.5735,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27073879569673176,
+      "grad_norm": 13.039442344005854,
+      "learning_rate": 4.278244994754155e-06,
+      "loss": 0.5772,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2712486616019987,
+      "grad_norm": 63.202208365540564,
+      "learning_rate": 4.275340716006424e-06,
+      "loss": 0.5382,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2717585275072656,
+      "grad_norm": 14.792564476876887,
+      "learning_rate": 4.272431595783783e-06,
+      "loss": 0.5846,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2722683934125325,
+      "grad_norm": 36.8084652376033,
+      "learning_rate": 4.269517642019601e-06,
+      "loss": 0.6567,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2727782593177994,
+      "grad_norm": 12.876578464531136,
+      "learning_rate": 4.2665988626604285e-06,
+      "loss": 0.5491,
+      "step": 5350
+    },
+    {
+      "epoch": 0.27328812522306634,
+      "grad_norm": 30.51025196353506,
+      "learning_rate": 4.2636752656659745e-06,
+      "loss": 0.6141,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27379799112833325,
+      "grad_norm": 20.777520310083748,
+      "learning_rate": 4.260746859009087e-06,
+      "loss": 0.6084,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27430785703360017,
+      "grad_norm": 9.044259321645782,
+      "learning_rate": 4.257813650675732e-06,
+      "loss": 0.623,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2748177229388671,
+      "grad_norm": 9.516832146327737,
+      "learning_rate": 4.254875648664965e-06,
+      "loss": 0.5885,
+      "step": 5390
+    },
+    {
+      "epoch": 0.275327588844134,
+      "grad_norm": 30.337005568568806,
+      "learning_rate": 4.251932860988921e-06,
+      "loss": 0.5376,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2758374547494009,
+      "grad_norm": 15.785533045132347,
+      "learning_rate": 4.24898529567278e-06,
+      "loss": 0.6343,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2763473206546678,
+      "grad_norm": 11.074354948231605,
+      "learning_rate": 4.246032960754753e-06,
+      "loss": 0.6324,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27685718655993474,
+      "grad_norm": 7.969861802125547,
+      "learning_rate": 4.243075864286059e-06,
+      "loss": 0.5494,
+      "step": 5430
+    },
+    {
+      "epoch": 0.27736705246520166,
+      "grad_norm": 27.380085381498265,
+      "learning_rate": 4.2401140143309e-06,
+      "loss": 0.5518,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2778769183704686,
+      "grad_norm": 13.480748579865969,
+      "learning_rate": 4.237147418966444e-06,
+      "loss": 0.5737,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2783867842757355,
+      "grad_norm": 10.01561427263943,
+      "learning_rate": 4.234176086282797e-06,
+      "loss": 0.5716,
+      "step": 5460
+    },
+    {
+      "epoch": 0.2788966501810024,
+      "grad_norm": 5.812847097454107,
+      "learning_rate": 4.231200024382987e-06,
+      "loss": 0.6192,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2794065160862693,
+      "grad_norm": 73.32354897034128,
+      "learning_rate": 4.228219241382936e-06,
+      "loss": 0.5402,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27991638199153623,
+      "grad_norm": 20.44580128899934,
+      "learning_rate": 4.2252337454114426e-06,
+      "loss": 0.6573,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28042624789680315,
+      "grad_norm": 12.430764888197537,
+      "learning_rate": 4.2222435446101555e-06,
+      "loss": 0.6103,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28093611380207006,
+      "grad_norm": 16.727645466395536,
+      "learning_rate": 4.219248647133559e-06,
+      "loss": 0.5864,
+      "step": 5510
+    },
+    {
+      "epoch": 0.281445979707337,
+      "grad_norm": 9.940584675976092,
+      "learning_rate": 4.216249061148939e-06,
+      "loss": 0.6636,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2819558456126039,
+      "grad_norm": 24.070584051307645,
+      "learning_rate": 4.213244794836373e-06,
+      "loss": 0.6432,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2824657115178708,
+      "grad_norm": 32.11346155712176,
+      "learning_rate": 4.210235856388699e-06,
+      "loss": 0.6097,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2829755774231377,
+      "grad_norm": 7.822795209697631,
+      "learning_rate": 4.2072222540114965e-06,
+      "loss": 0.633,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28348544332840464,
+      "grad_norm": 8.645615052891959,
+      "learning_rate": 4.204203995923064e-06,
+      "loss": 0.5731,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28399530923367156,
+      "grad_norm": 13.28668413130762,
+      "learning_rate": 4.201181090354396e-06,
+      "loss": 0.6167,
+      "step": 5570
+    },
+    {
+      "epoch": 0.28450517513893847,
+      "grad_norm": 7.542368445511174,
+      "learning_rate": 4.198153545549164e-06,
+      "loss": 0.5833,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2850150410442054,
+      "grad_norm": 13.289497206111815,
+      "learning_rate": 4.195121369763687e-06,
+      "loss": 0.6001,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2855249069494723,
+      "grad_norm": 13.00491506318039,
+      "learning_rate": 4.192084571266915e-06,
+      "loss": 0.5839,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2860347728547392,
+      "grad_norm": 6.256915488118059,
+      "learning_rate": 4.189043158340403e-06,
+      "loss": 0.592,
+      "step": 5610
+    },
+    {
+      "epoch": 0.28654463876000613,
+      "grad_norm": 14.22336372739089,
+      "learning_rate": 4.185997139278292e-06,
+      "loss": 0.5594,
+      "step": 5620
+    },
+    {
+      "epoch": 0.28705450466527305,
+      "grad_norm": 17.9794598260869,
+      "learning_rate": 4.182946522387283e-06,
+      "loss": 0.5584,
+      "step": 5630
+    },
+    {
+      "epoch": 0.28756437057053996,
+      "grad_norm": 26.826093361889214,
+      "learning_rate": 4.179891315986617e-06,
+      "loss": 0.5976,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2880742364758069,
+      "grad_norm": 26.854750297006724,
+      "learning_rate": 4.1768315284080475e-06,
+      "loss": 0.5782,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2885841023810738,
+      "grad_norm": 20.85565602863907,
+      "learning_rate": 4.173767167995825e-06,
+      "loss": 0.6018,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2890939682863407,
+      "grad_norm": 40.44526401777264,
+      "learning_rate": 4.170698243106668e-06,
+      "loss": 0.6106,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2896038341916076,
+      "grad_norm": 11.5937114201011,
+      "learning_rate": 4.1676247621097445e-06,
+      "loss": 0.5584,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29011370009687454,
+      "grad_norm": 18.80756096170308,
+      "learning_rate": 4.164546733386644e-06,
+      "loss": 0.5946,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29062356600214145,
+      "grad_norm": 16.134109720458465,
+      "learning_rate": 4.161464165331363e-06,
+      "loss": 0.6808,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29113343190740837,
+      "grad_norm": 62.99927751669872,
+      "learning_rate": 4.158377066350273e-06,
+      "loss": 0.5772,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2916432978126753,
+      "grad_norm": 9.81929767488598,
+      "learning_rate": 4.1552854448621025e-06,
+      "loss": 0.6214,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2921531637179422,
+      "grad_norm": 22.729864470532355,
+      "learning_rate": 4.152189309297914e-06,
+      "loss": 0.5995,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2926630296232091,
+      "grad_norm": 12.477684258820998,
+      "learning_rate": 4.14908866810108e-06,
+      "loss": 0.5958,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29317289552847603,
+      "grad_norm": 18.84572272547121,
+      "learning_rate": 4.14598352972726e-06,
+      "loss": 0.5632,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29368276143374294,
+      "grad_norm": 25.846100170430304,
+      "learning_rate": 4.142873902644378e-06,
+      "loss": 0.6216,
+      "step": 5760
+    },
+    {
+      "epoch": 0.29419262733900986,
+      "grad_norm": 10.199726471823542,
+      "learning_rate": 4.139759795332597e-06,
+      "loss": 0.6134,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2947024932442768,
+      "grad_norm": 22.019306648150717,
+      "learning_rate": 4.1366412162843015e-06,
+      "loss": 0.5522,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2952123591495437,
+      "grad_norm": 13.552518209230131,
+      "learning_rate": 4.133518174004068e-06,
+      "loss": 0.5765,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2957222250548106,
+      "grad_norm": 8.60015583805641,
+      "learning_rate": 4.130390677008644e-06,
+      "loss": 0.6265,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2962320909600775,
+      "grad_norm": 12.209720854395725,
+      "learning_rate": 4.127258733826929e-06,
+      "loss": 0.5497,
+      "step": 5810
+    },
+    {
+      "epoch": 0.29674195686534444,
+      "grad_norm": 22.49340812520029,
+      "learning_rate": 4.1241223529999425e-06,
+      "loss": 0.5843,
+      "step": 5820
+    },
+    {
+      "epoch": 0.29725182277061135,
+      "grad_norm": 16.498403886804276,
+      "learning_rate": 4.12098154308081e-06,
+      "loss": 0.5801,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29776168867587827,
+      "grad_norm": 24.812587542691467,
+      "learning_rate": 4.117836312634734e-06,
+      "loss": 0.5884,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2982715545811452,
+      "grad_norm": 14.33912906673135,
+      "learning_rate": 4.114686670238971e-06,
+      "loss": 0.5561,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2987814204864121,
+      "grad_norm": 18.845657492416535,
+      "learning_rate": 4.111532624482811e-06,
+      "loss": 0.6124,
+      "step": 5860
+    },
+    {
+      "epoch": 0.299291286391679,
+      "grad_norm": 6.225197949040313,
+      "learning_rate": 4.108374183967549e-06,
+      "loss": 0.6058,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2998011522969459,
+      "grad_norm": 14.94669704332347,
+      "learning_rate": 4.10521135730647e-06,
+      "loss": 0.6446,
+      "step": 5880
+    },
+    {
+      "epoch": 0.30031101820221284,
+      "grad_norm": 11.67084221046296,
+      "learning_rate": 4.1020441531248165e-06,
+      "loss": 0.5561,
+      "step": 5890
+    },
+    {
+      "epoch": 0.30082088410747976,
+      "grad_norm": 13.68291404953535,
+      "learning_rate": 4.09887258005977e-06,
+      "loss": 0.5402,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3013307500127467,
+      "grad_norm": 11.694111191588059,
+      "learning_rate": 4.095696646760425e-06,
+      "loss": 0.5883,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3018406159180136,
+      "grad_norm": 9.147828359283288,
+      "learning_rate": 4.09251636188777e-06,
+      "loss": 0.6205,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3023504818232805,
+      "grad_norm": 6.078824528176967,
+      "learning_rate": 4.0893317341146545e-06,
+      "loss": 0.6233,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3028603477285474,
+      "grad_norm": 10.338227098716464,
+      "learning_rate": 4.086142772125779e-06,
+      "loss": 0.5932,
+      "step": 5940
+    },
+    {
+      "epoch": 0.30337021363381433,
+      "grad_norm": 34.03454922469082,
+      "learning_rate": 4.082949484617656e-06,
+      "loss": 0.5736,
+      "step": 5950
+    },
+    {
+      "epoch": 0.30388007953908125,
+      "grad_norm": 27.717990588485844,
+      "learning_rate": 4.079751880298601e-06,
+      "loss": 0.6005,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3043899454443481,
+      "grad_norm": 20.62906430272811,
+      "learning_rate": 4.076549967888697e-06,
+      "loss": 0.625,
+      "step": 5970
+    },
+    {
+      "epoch": 0.304899811349615,
+      "grad_norm": 8.843859127048827,
+      "learning_rate": 4.073343756119778e-06,
+      "loss": 0.5647,
+      "step": 5980
+    },
+    {
+      "epoch": 0.30540967725488194,
+      "grad_norm": 15.825921945819855,
+      "learning_rate": 4.070133253735399e-06,
+      "loss": 0.6106,
+      "step": 5990
+    },
+    {
+      "epoch": 0.30591954316014885,
+      "grad_norm": 7.034079033908886,
+      "learning_rate": 4.066918469490822e-06,
+      "loss": 0.5998,
+      "step": 6000
+    },
+    {
+      "epoch": 0.30642940906541577,
+      "grad_norm": 10.725159040580532,
+      "learning_rate": 4.063699412152979e-06,
+      "loss": 0.5741,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3069392749706827,
+      "grad_norm": 25.954938413714736,
+      "learning_rate": 4.060476090500462e-06,
+      "loss": 0.6323,
+      "step": 6020
+    },
+    {
+      "epoch": 0.3074491408759496,
+      "grad_norm": 19.147343982242354,
+      "learning_rate": 4.057248513323484e-06,
+      "loss": 0.6109,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3079590067812165,
+      "grad_norm": 11.067841286474719,
+      "learning_rate": 4.054016689423871e-06,
+      "loss": 0.5934,
+      "step": 6040
+    },
+    {
+      "epoch": 0.30846887268648343,
+      "grad_norm": 15.374555769962404,
+      "learning_rate": 4.050780627615025e-06,
+      "loss": 0.605,
+      "step": 6050
+    },
+    {
+      "epoch": 0.30897873859175035,
+      "grad_norm": 8.74236544222671,
+      "learning_rate": 4.047540336721909e-06,
+      "loss": 0.5638,
+      "step": 6060
+    },
+    {
+      "epoch": 0.30948860449701726,
+      "grad_norm": 9.692794262472,
+      "learning_rate": 4.044295825581013e-06,
+      "loss": 0.5688,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3099984704022842,
+      "grad_norm": 19.37396547964373,
+      "learning_rate": 4.041047103040343e-06,
+      "loss": 0.5662,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3105083363075511,
+      "grad_norm": 4.850686203291434,
+      "learning_rate": 4.0377941779593835e-06,
+      "loss": 0.5754,
+      "step": 6090
+    },
+    {
+      "epoch": 0.311018202212818,
+      "grad_norm": 19.351654987382656,
+      "learning_rate": 4.034537059209085e-06,
+      "loss": 0.5249,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3115280681180849,
+      "grad_norm": 48.11582019696415,
+      "learning_rate": 4.03127575567183e-06,
+      "loss": 0.5772,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31203793402335184,
+      "grad_norm": 9.755571290958523,
+      "learning_rate": 4.028010276241416e-06,
+      "loss": 0.5537,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31254779992861875,
+      "grad_norm": 19.376346542781313,
+      "learning_rate": 4.0247406298230285e-06,
+      "loss": 0.5969,
+      "step": 6130
+    },
+    {
+      "epoch": 0.31305766583388567,
+      "grad_norm": 8.574897664066954,
+      "learning_rate": 4.021466825333215e-06,
+      "loss": 0.5945,
+      "step": 6140
+    },
+    {
+      "epoch": 0.3135675317391526,
+      "grad_norm": 11.648947877849336,
+      "learning_rate": 4.018188871699861e-06,
+      "loss": 0.6026,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3140773976444195,
+      "grad_norm": 20.788831004876297,
+      "learning_rate": 4.014906777862172e-06,
+      "loss": 0.6057,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3145872635496864,
+      "grad_norm": 7.0208577775784695,
+      "learning_rate": 4.01162055277064e-06,
+      "loss": 0.654,
+      "step": 6170
+    },
+    {
+      "epoch": 0.31509712945495333,
+      "grad_norm": 8.214032627488498,
+      "learning_rate": 4.008330205387024e-06,
+      "loss": 0.5925,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31560699536022024,
+      "grad_norm": 31.67405111863016,
+      "learning_rate": 4.005035744684325e-06,
+      "loss": 0.5596,
+      "step": 6190
+    },
+    {
+      "epoch": 0.31611686126548716,
+      "grad_norm": 7.266395367405267,
+      "learning_rate": 4.0017371796467635e-06,
+      "loss": 0.5927,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3166267271707541,
+      "grad_norm": 16.510677156143704,
+      "learning_rate": 3.998434519269749e-06,
+      "loss": 0.5535,
+      "step": 6210
+    },
+    {
+      "epoch": 0.317136593076021,
+      "grad_norm": 6.854338605125168,
+      "learning_rate": 3.9951277725598604e-06,
+      "loss": 0.6465,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3176464589812879,
+      "grad_norm": 18.111246303610876,
+      "learning_rate": 3.991816948534823e-06,
+      "loss": 0.5545,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3181563248865548,
+      "grad_norm": 8.117561150966036,
+      "learning_rate": 3.988502056223477e-06,
+      "loss": 0.571,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31866619079182174,
+      "grad_norm": 13.084742587714015,
+      "learning_rate": 3.98518310466576e-06,
+      "loss": 0.5753,
+      "step": 6250
+    },
+    {
+      "epoch": 0.31917605669708865,
+      "grad_norm": 57.142935902173484,
+      "learning_rate": 3.98186010291268e-06,
+      "loss": 0.5618,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31968592260235557,
+      "grad_norm": 7.11621758234793,
+      "learning_rate": 3.978533060026288e-06,
+      "loss": 0.641,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3201957885076225,
+      "grad_norm": 16.4098953136186,
+      "learning_rate": 3.975201985079655e-06,
+      "loss": 0.5694,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3207056544128894,
+      "grad_norm": 23.527870364341037,
+      "learning_rate": 3.971866887156851e-06,
+      "loss": 0.5663,
+      "step": 6290
+    },
+    {
+      "epoch": 0.3212155203181563,
+      "grad_norm": 13.608447802374062,
+      "learning_rate": 3.968527775352914e-06,
+      "loss": 0.6021,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3217253862234232,
+      "grad_norm": 7.505925346358185,
+      "learning_rate": 3.965184658773828e-06,
+      "loss": 0.6106,
+      "step": 6310
+    },
+    {
+      "epoch": 0.32223525212869014,
+      "grad_norm": 35.6313686416139,
+      "learning_rate": 3.961837546536501e-06,
+      "loss": 0.5912,
+      "step": 6320
+    },
+    {
+      "epoch": 0.32274511803395706,
+      "grad_norm": 5.634896177598433,
+      "learning_rate": 3.958486447768736e-06,
+      "loss": 0.5676,
+      "step": 6330
+    },
+    {
+      "epoch": 0.32325498393922397,
+      "grad_norm": 6.353712880952279,
+      "learning_rate": 3.955131371609206e-06,
+      "loss": 0.6278,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3237648498444909,
+      "grad_norm": 7.2854542328463054,
+      "learning_rate": 3.951772327207432e-06,
+      "loss": 0.6439,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3242747157497578,
+      "grad_norm": 11.258387362973808,
+      "learning_rate": 3.948409323723756e-06,
+      "loss": 0.5684,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3247845816550247,
+      "grad_norm": 9.882782910086245,
+      "learning_rate": 3.945042370329319e-06,
+      "loss": 0.5296,
+      "step": 6370
+    },
+    {
+      "epoch": 0.32529444756029163,
+      "grad_norm": 19.15860492822133,
+      "learning_rate": 3.9416714762060325e-06,
+      "loss": 0.6402,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32580431346555855,
+      "grad_norm": 16.143665735042397,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.5862,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32631417937082546,
+      "grad_norm": 4.106912978049323,
+      "learning_rate": 3.934917902554257e-06,
+      "loss": 0.5537,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3268240452760924,
+      "grad_norm": 17.22897555904163,
+      "learning_rate": 3.931535241443225e-06,
+      "loss": 0.5744,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3273339111813593,
+      "grad_norm": 7.739685203051684,
+      "learning_rate": 3.9281486764382e-06,
+      "loss": 0.5905,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3278437770866262,
+      "grad_norm": 5.280772793174786,
+      "learning_rate": 3.924758216774579e-06,
+      "loss": 0.6035,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3283536429918931,
+      "grad_norm": 9.56456482861663,
+      "learning_rate": 3.921363871698372e-06,
+      "loss": 0.6449,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32886350889716004,
+      "grad_norm": 19.76743516233812,
+      "learning_rate": 3.917965650466192e-06,
+      "loss": 0.5873,
+      "step": 6450
+    },
+    {
+      "epoch": 0.32937337480242695,
+      "grad_norm": 12.805680819672576,
+      "learning_rate": 3.914563562345218e-06,
+      "loss": 0.5929,
+      "step": 6460
+    },
+    {
+      "epoch": 0.32988324070769387,
+      "grad_norm": 7.3410821734516425,
+      "learning_rate": 3.911157616613176e-06,
+      "loss": 0.6349,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3303931066129608,
+      "grad_norm": 9.91815295194112,
+      "learning_rate": 3.9077478225583115e-06,
+      "loss": 0.5611,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3309029725182277,
+      "grad_norm": 10.58322112787284,
+      "learning_rate": 3.904334189479366e-06,
+      "loss": 0.6239,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3314128384234946,
+      "grad_norm": 5.115542327229815,
+      "learning_rate": 3.900916726685547e-06,
+      "loss": 0.563,
+      "step": 6500
+    },
+    {
+      "epoch": 0.33192270432876153,
+      "grad_norm": 11.574317620807081,
+      "learning_rate": 3.897495443496511e-06,
+      "loss": 0.6234,
+      "step": 6510
+    },
+    {
+      "epoch": 0.33243257023402845,
+      "grad_norm": 6.82635077786538,
+      "learning_rate": 3.894070349242328e-06,
+      "loss": 0.6331,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33294243613929536,
+      "grad_norm": 12.572812590923363,
+      "learning_rate": 3.890641453263463e-06,
+      "loss": 0.6089,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3334523020445623,
+      "grad_norm": 8.528905748581712,
+      "learning_rate": 3.887208764910749e-06,
+      "loss": 0.5721,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3339621679498292,
+      "grad_norm": 11.470816287216191,
+      "learning_rate": 3.8837722935453615e-06,
+      "loss": 0.6266,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3344720338550961,
+      "grad_norm": 25.624061552242477,
+      "learning_rate": 3.880332048538789e-06,
+      "loss": 0.571,
+      "step": 6560
+    },
+    {
+      "epoch": 0.334981899760363,
+      "grad_norm": 6.79483649892652,
+      "learning_rate": 3.876888039272818e-06,
+      "loss": 0.6398,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33549176566562994,
+      "grad_norm": 13.785631626853508,
+      "learning_rate": 3.8734402751394925e-06,
+      "loss": 0.5693,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33600163157089685,
+      "grad_norm": 11.493447222682974,
+      "learning_rate": 3.869988765541101e-06,
+      "loss": 0.5496,
+      "step": 6590
+    },
+    {
+      "epoch": 0.33651149747616377,
+      "grad_norm": 10.373154022440099,
+      "learning_rate": 3.866533519890145e-06,
+      "loss": 0.5145,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3370213633814307,
+      "grad_norm": 5.525677922272744,
+      "learning_rate": 3.863074547609314e-06,
+      "loss": 0.5442,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3375312292866976,
+      "grad_norm": 9.785016845624112,
+      "learning_rate": 3.859611858131461e-06,
+      "loss": 0.6107,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3380410951919645,
+      "grad_norm": 9.48430417779256,
+      "learning_rate": 3.8561454608995765e-06,
+      "loss": 0.6144,
+      "step": 6630
+    },
+    {
+      "epoch": 0.33855096109723143,
+      "grad_norm": 40.189131615359535,
+      "learning_rate": 3.852675365366762e-06,
+      "loss": 0.5367,
+      "step": 6640
+    },
+    {
+      "epoch": 0.33906082700249834,
+      "grad_norm": 8.524943943823194,
+      "learning_rate": 3.849201580996201e-06,
+      "loss": 0.5867,
+      "step": 6650
+    },
+    {
+      "epoch": 0.33957069290776526,
+      "grad_norm": 13.618979522501697,
+      "learning_rate": 3.845724117261142e-06,
+      "loss": 0.5926,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3400805588130322,
+      "grad_norm": 8.693144356860516,
+      "learning_rate": 3.8422429836448665e-06,
+      "loss": 0.6188,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3405904247182991,
+      "grad_norm": 11.74695610219484,
+      "learning_rate": 3.83875818964066e-06,
+      "loss": 0.5656,
+      "step": 6680
+    },
+    {
+      "epoch": 0.341100290623566,
+      "grad_norm": 9.198063906387935,
+      "learning_rate": 3.835269744751796e-06,
+      "loss": 0.5905,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3416101565288329,
+      "grad_norm": 5.334788699784293,
+      "learning_rate": 3.831777658491497e-06,
+      "loss": 0.6202,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34212002243409984,
+      "grad_norm": 24.177391127776264,
+      "learning_rate": 3.828281940382923e-06,
+      "loss": 0.6503,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34262988833936675,
+      "grad_norm": 23.2850171618304,
+      "learning_rate": 3.824782599959134e-06,
+      "loss": 0.6448,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34313975424463367,
+      "grad_norm": 4.4728799548336955,
+      "learning_rate": 3.8212796467630685e-06,
+      "loss": 0.5452,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3436496201499006,
+      "grad_norm": 7.442759570311029,
+      "learning_rate": 3.8177730903475195e-06,
+      "loss": 0.5569,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3441594860551675,
+      "grad_norm": 7.801573446892921,
+      "learning_rate": 3.8142629402751047e-06,
+      "loss": 0.589,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3446693519604344,
+      "grad_norm": 13.098329982961182,
+      "learning_rate": 3.8107492061182418e-06,
+      "loss": 0.6401,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3451792178657013,
+      "grad_norm": 14.400990901182896,
+      "learning_rate": 3.8072318974591233e-06,
+      "loss": 0.5683,
+      "step": 6770
+    },
+    {
+      "epoch": 0.34568908377096824,
+      "grad_norm": 12.27633537035338,
+      "learning_rate": 3.803711023889688e-06,
+      "loss": 0.5933,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34619894967623516,
+      "grad_norm": 14.372056846033948,
+      "learning_rate": 3.800186595011599e-06,
+      "loss": 0.5881,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3467088155815021,
+      "grad_norm": 85.50410333963302,
+      "learning_rate": 3.7966586204362137e-06,
+      "loss": 0.5933,
+      "step": 6800
+    },
+    {
+      "epoch": 0.347218681486769,
+      "grad_norm": 22.517826922116864,
+      "learning_rate": 3.793127109784558e-06,
+      "loss": 0.5916,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3477285473920359,
+      "grad_norm": 14.84798611353888,
+      "learning_rate": 3.789592072687302e-06,
+      "loss": 0.6776,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3482384132973028,
+      "grad_norm": 15.270402470251646,
+      "learning_rate": 3.7860535187847326e-06,
+      "loss": 0.5683,
+      "step": 6830
+    },
+    {
+      "epoch": 0.34874827920256973,
+      "grad_norm": 10.286151783016678,
+      "learning_rate": 3.782511457726725e-06,
+      "loss": 0.6363,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34925814510783665,
+      "grad_norm": 11.341726028096593,
+      "learning_rate": 3.7789658991727242e-06,
+      "loss": 0.5448,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34976801101310356,
+      "grad_norm": 5.429208019721334,
+      "learning_rate": 3.775416852791707e-06,
+      "loss": 0.577,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3502778769183705,
+      "grad_norm": 17.74331297520172,
+      "learning_rate": 3.7718643282621648e-06,
+      "loss": 0.5379,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3507877428236374,
+      "grad_norm": 18.701045945994224,
+      "learning_rate": 3.768308335272075e-06,
+      "loss": 0.5417,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3512976087289043,
+      "grad_norm": 11.322040410058435,
+      "learning_rate": 3.7647488835188705e-06,
+      "loss": 0.5552,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3518074746341712,
+      "grad_norm": 9.526508052327046,
+      "learning_rate": 3.76118598270942e-06,
+      "loss": 0.6219,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35231734053943814,
+      "grad_norm": 9.618221118643495,
+      "learning_rate": 3.7576196425599944e-06,
+      "loss": 0.561,
+      "step": 6910
+    },
+    {
+      "epoch": 0.35282720644470505,
+      "grad_norm": 5.285098061934564,
+      "learning_rate": 3.7540498727962483e-06,
+      "loss": 0.5897,
+      "step": 6920
+    },
+    {
+      "epoch": 0.35333707234997197,
+      "grad_norm": 6.319348479139971,
+      "learning_rate": 3.7504766831531848e-06,
+      "loss": 0.5981,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3538469382552389,
+      "grad_norm": 14.26337724889404,
+      "learning_rate": 3.746900083375137e-06,
+      "loss": 0.5847,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3543568041605058,
+      "grad_norm": 8.842973608099559,
+      "learning_rate": 3.7433200832157333e-06,
+      "loss": 0.687,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3548666700657727,
+      "grad_norm": 26.378636644956504,
+      "learning_rate": 3.7397366924378797e-06,
+      "loss": 0.5838,
+      "step": 6960
+    },
+    {
+      "epoch": 0.35537653597103963,
+      "grad_norm": 12.43417363530746,
+      "learning_rate": 3.736149920813726e-06,
+      "loss": 0.5844,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35588640187630655,
+      "grad_norm": 6.8624589370405715,
+      "learning_rate": 3.7325597781246426e-06,
+      "loss": 0.5796,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35639626778157346,
+      "grad_norm": 34.23759579395761,
+      "learning_rate": 3.728966274161192e-06,
+      "loss": 0.5426,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3569061336868404,
+      "grad_norm": 12.036373216153141,
+      "learning_rate": 3.725369418723106e-06,
+      "loss": 0.5407,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3574159995921073,
+      "grad_norm": 7.6541152081261306,
+      "learning_rate": 3.721769221619252e-06,
+      "loss": 0.5307,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3579258654973742,
+      "grad_norm": 5.221962987657324,
+      "learning_rate": 3.718165692667613e-06,
+      "loss": 0.5638,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3584357314026411,
+      "grad_norm": 9.769128260322358,
+      "learning_rate": 3.714558841695258e-06,
+      "loss": 0.6025,
+      "step": 7030
+    },
+    {
+      "epoch": 0.35894559730790804,
+      "grad_norm": 5.4841935802743516,
+      "learning_rate": 3.710948678538314e-06,
+      "loss": 0.5845,
+      "step": 7040
+    },
+    {
+      "epoch": 0.35945546321317495,
+      "grad_norm": 4.441919327602813,
+      "learning_rate": 3.7073352130419436e-06,
+      "loss": 0.6463,
+      "step": 7050
+    },
+    {
+      "epoch": 0.35996532911844187,
+      "grad_norm": 10.50879690961124,
+      "learning_rate": 3.703718455060311e-06,
+      "loss": 0.5851,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3604751950237088,
+      "grad_norm": 47.663384741510505,
+      "learning_rate": 3.7000984144565617e-06,
+      "loss": 0.6206,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3609850609289757,
+      "grad_norm": 10.628757506788443,
+      "learning_rate": 3.6964751011027933e-06,
+      "loss": 0.5544,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3614949268342426,
+      "grad_norm": 7.288945575368875,
+      "learning_rate": 3.6928485248800273e-06,
+      "loss": 0.5495,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36200479273950953,
+      "grad_norm": 7.992239176835611,
+      "learning_rate": 3.689218695678184e-06,
+      "loss": 0.5572,
+      "step": 7100
+    },
+    {
+      "epoch": 0.36251465864477644,
+      "grad_norm": 7.216006748679708,
+      "learning_rate": 3.685585623396055e-06,
+      "loss": 0.5966,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36302452455004336,
+      "grad_norm": 10.354746806457767,
+      "learning_rate": 3.681949317941275e-06,
+      "loss": 0.5422,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3635343904553103,
+      "grad_norm": 8.329338901239497,
+      "learning_rate": 3.6783097892302967e-06,
+      "loss": 0.6023,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3640442563605772,
+      "grad_norm": 10.42649234263997,
+      "learning_rate": 3.6746670471883616e-06,
+      "loss": 0.5793,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3645541222658441,
+      "grad_norm": 4.914161385304957,
+      "learning_rate": 3.671021101749476e-06,
+      "loss": 0.5533,
+      "step": 7150
+    },
+    {
+      "epoch": 0.365063988171111,
+      "grad_norm": 6.108063070133632,
+      "learning_rate": 3.6673719628563808e-06,
+      "loss": 0.563,
+      "step": 7160
+    },
+    {
+      "epoch": 0.36557385407637794,
+      "grad_norm": 7.615397623687595,
+      "learning_rate": 3.6637196404605257e-06,
+      "loss": 0.5659,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36608371998164485,
+      "grad_norm": 14.69964181406416,
+      "learning_rate": 3.660064144522043e-06,
+      "loss": 0.543,
+      "step": 7180
+    },
+    {
+      "epoch": 0.36659358588691177,
+      "grad_norm": 3.3695135716741245,
+      "learning_rate": 3.656405485009719e-06,
+      "loss": 0.5664,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3671034517921787,
+      "grad_norm": 43.664567891879734,
+      "learning_rate": 3.652743671900967e-06,
+      "loss": 0.5566,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3676133176974456,
+      "grad_norm": 21.87781936577931,
+      "learning_rate": 3.6490787151817986e-06,
+      "loss": 0.5362,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3681231836027125,
+      "grad_norm": 6.9291162528960175,
+      "learning_rate": 3.6454106248468024e-06,
+      "loss": 0.5888,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3686330495079794,
+      "grad_norm": 6.711268461016556,
+      "learning_rate": 3.6417394108991096e-06,
+      "loss": 0.5572,
+      "step": 7230
+    },
+    {
+      "epoch": 0.36914291541324634,
+      "grad_norm": 9.234151539211476,
+      "learning_rate": 3.6380650833503705e-06,
+      "loss": 0.6033,
+      "step": 7240
+    },
+    {
+      "epoch": 0.36965278131851326,
+      "grad_norm": 72.08082771250515,
+      "learning_rate": 3.6343876522207253e-06,
+      "loss": 0.6168,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3701626472237802,
+      "grad_norm": 14.232739813895227,
+      "learning_rate": 3.6307071275387807e-06,
+      "loss": 0.5784,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3706725131290471,
+      "grad_norm": 8.312159079086417,
+      "learning_rate": 3.6270235193415754e-06,
+      "loss": 0.5503,
+      "step": 7270
+    },
+    {
+      "epoch": 0.371182379034314,
+      "grad_norm": 5.219185861295808,
+      "learning_rate": 3.6233368376745616e-06,
+      "loss": 0.6168,
+      "step": 7280
+    },
+    {
+      "epoch": 0.37169224493958086,
+      "grad_norm": 4.783574085733863,
+      "learning_rate": 3.6196470925915705e-06,
+      "loss": 0.6012,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3722021108448478,
+      "grad_norm": 6.14441502544535,
+      "learning_rate": 3.6159542941547883e-06,
+      "loss": 0.5437,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3727119767501147,
+      "grad_norm": 4.913727327218928,
+      "learning_rate": 3.6122584524347267e-06,
+      "loss": 0.6013,
+      "step": 7310
+    },
+    {
+      "epoch": 0.3732218426553816,
+      "grad_norm": 4.673399055747235,
+      "learning_rate": 3.608559577510198e-06,
+      "loss": 0.5395,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3737317085606485,
+      "grad_norm": 5.707554591431282,
+      "learning_rate": 3.604857679468285e-06,
+      "loss": 0.5463,
+      "step": 7330
+    },
+    {
+      "epoch": 0.37424157446591544,
+      "grad_norm": 10.008003378212784,
+      "learning_rate": 3.601152768404317e-06,
+      "loss": 0.5848,
+      "step": 7340
+    },
+    {
+      "epoch": 0.37475144037118235,
+      "grad_norm": 15.561381138932529,
+      "learning_rate": 3.597444854421837e-06,
+      "loss": 0.5833,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37526130627644927,
+      "grad_norm": 23.018403169410053,
+      "learning_rate": 3.59373394763258e-06,
+      "loss": 0.5587,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3757711721817162,
+      "grad_norm": 7.150363888762786,
+      "learning_rate": 3.5900200581564403e-06,
+      "loss": 0.6373,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3762810380869831,
+      "grad_norm": 7.142174562438807,
+      "learning_rate": 3.586303196121447e-06,
+      "loss": 0.5655,
+      "step": 7380
+    },
+    {
+      "epoch": 0.37679090399225,
+      "grad_norm": 13.760948907375717,
+      "learning_rate": 3.5825833716637364e-06,
+      "loss": 0.65,
+      "step": 7390
+    },
+    {
+      "epoch": 0.37730076989751693,
+      "grad_norm": 9.225428328804446,
+      "learning_rate": 3.5788605949275237e-06,
+      "loss": 0.5367,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37781063580278385,
+      "grad_norm": 24.730845210760858,
+      "learning_rate": 3.5751348760650722e-06,
+      "loss": 0.6252,
+      "step": 7410
+    },
+    {
+      "epoch": 0.37832050170805076,
+      "grad_norm": 12.795430386888327,
+      "learning_rate": 3.5714062252366723e-06,
+      "loss": 0.6394,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3788303676133177,
+      "grad_norm": 5.9277660577921765,
+      "learning_rate": 3.5676746526106084e-06,
+      "loss": 0.6379,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3793402335185846,
+      "grad_norm": 5.044275508529918,
+      "learning_rate": 3.5639401683631314e-06,
+      "loss": 0.6292,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3798500994238515,
+      "grad_norm": 9.004761341755255,
+      "learning_rate": 3.5602027826784356e-06,
+      "loss": 0.628,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3803599653291184,
+      "grad_norm": 4.017290839777303,
+      "learning_rate": 3.556462505748625e-06,
+      "loss": 0.5683,
+      "step": 7460
+    },
+    {
+      "epoch": 0.38086983123438534,
+      "grad_norm": 6.958902888219458,
+      "learning_rate": 3.5527193477736903e-06,
+      "loss": 0.5694,
+      "step": 7470
+    },
+    {
+      "epoch": 0.38137969713965225,
+      "grad_norm": 5.5519005203253275,
+      "learning_rate": 3.548973318961477e-06,
+      "loss": 0.5316,
+      "step": 7480
+    },
+    {
+      "epoch": 0.38188956304491917,
+      "grad_norm": 6.163222258445389,
+      "learning_rate": 3.5452244295276604e-06,
+      "loss": 0.6386,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3823994289501861,
+      "grad_norm": 31.506161393196194,
+      "learning_rate": 3.541472689695718e-06,
+      "loss": 0.5672,
+      "step": 7500
+    },
+    {
+      "epoch": 0.382909294855453,
+      "grad_norm": 13.297329690802785,
+      "learning_rate": 3.537718109696899e-06,
+      "loss": 0.5918,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3834191607607199,
+      "grad_norm": 6.188203092221159,
+      "learning_rate": 3.5339606997701975e-06,
+      "loss": 0.5354,
+      "step": 7520
+    },
+    {
+      "epoch": 0.38392902666598683,
+      "grad_norm": 5.844684135462748,
+      "learning_rate": 3.530200470162328e-06,
+      "loss": 0.6198,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38443889257125374,
+      "grad_norm": 6.155255221678391,
+      "learning_rate": 3.5264374311276904e-06,
+      "loss": 0.4963,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38494875847652066,
+      "grad_norm": 44.826506547381406,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.5906,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3854586243817876,
+      "grad_norm": 5.697741044215528,
+      "learning_rate": 3.518902965834003e-06,
+      "loss": 0.5611,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3859684902870545,
+      "grad_norm": 19.71448408084545,
+      "learning_rate": 3.5151315601219517e-06,
+      "loss": 0.536,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3864783561923214,
+      "grad_norm": 37.42877624913661,
+      "learning_rate": 3.5113573860770755e-06,
+      "loss": 0.5678,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3869882220975883,
+      "grad_norm": 5.897945109676117,
+      "learning_rate": 3.5075804539918047e-06,
+      "loss": 0.6857,
+      "step": 7590
+    },
+    {
+      "epoch": 0.38749808800285523,
+      "grad_norm": 8.59608065779492,
+      "learning_rate": 3.5038007741660895e-06,
+      "loss": 0.5904,
+      "step": 7600
+    },
+    {
+      "epoch": 0.38800795390812215,
+      "grad_norm": 8.668612428257326,
+      "learning_rate": 3.5000183569073743e-06,
+      "loss": 0.6183,
+      "step": 7610
+    },
+    {
+      "epoch": 0.38851781981338906,
+      "grad_norm": 7.116728646113035,
+      "learning_rate": 3.4962332125305686e-06,
+      "loss": 0.5345,
+      "step": 7620
+    },
+    {
+      "epoch": 0.389027685718656,
+      "grad_norm": 4.3910147461794775,
+      "learning_rate": 3.492445351358018e-06,
+      "loss": 0.5716,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3895375516239229,
+      "grad_norm": 6.723186930639564,
+      "learning_rate": 3.488654783719477e-06,
+      "loss": 0.5315,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3900474175291898,
+      "grad_norm": 7.9122335533789245,
+      "learning_rate": 3.484861519952083e-06,
+      "loss": 0.5747,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3905572834344567,
+      "grad_norm": 10.613463496682332,
+      "learning_rate": 3.4810655704003237e-06,
+      "loss": 0.5648,
+      "step": 7660
+    },
+    {
+      "epoch": 0.39106714933972364,
+      "grad_norm": 18.243032282223737,
+      "learning_rate": 3.4772669454160113e-06,
+      "loss": 0.5482,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39157701524499056,
+      "grad_norm": 10.660339605638224,
+      "learning_rate": 3.473465655358255e-06,
+      "loss": 0.5209,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39208688115025747,
+      "grad_norm": 6.602625160974067,
+      "learning_rate": 3.469661710593431e-06,
+      "loss": 0.5632,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3925967470555244,
+      "grad_norm": 4.708196667408552,
+      "learning_rate": 3.465855121495156e-06,
+      "loss": 0.4498,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3931066129607913,
+      "grad_norm": 10.823694154178865,
+      "learning_rate": 3.4620458984442564e-06,
+      "loss": 0.5415,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3936164788660582,
+      "grad_norm": 7.351637451106162,
+      "learning_rate": 3.4582340518287444e-06,
+      "loss": 0.5646,
+      "step": 7720
+    },
+    {
+      "epoch": 0.39412634477132513,
+      "grad_norm": 7.415628332609013,
+      "learning_rate": 3.4544195920437834e-06,
+      "loss": 0.5688,
+      "step": 7730
+    },
+    {
+      "epoch": 0.39463621067659205,
+      "grad_norm": 11.07313306100438,
+      "learning_rate": 3.4506025294916656e-06,
+      "loss": 0.5557,
+      "step": 7740
+    },
+    {
+      "epoch": 0.39514607658185896,
+      "grad_norm": 22.03291235562454,
+      "learning_rate": 3.44678287458178e-06,
+      "loss": 0.5875,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3956559424871259,
+      "grad_norm": 5.959836162345017,
+      "learning_rate": 3.4429606377305847e-06,
+      "loss": 0.6191,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3961658083923928,
+      "grad_norm": 5.5293587959812545,
+      "learning_rate": 3.439135829361581e-06,
+      "loss": 0.5281,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3966756742976597,
+      "grad_norm": 17.335795805582098,
+      "learning_rate": 3.435308459905281e-06,
+      "loss": 0.6464,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3971855402029266,
+      "grad_norm": 8.319895170860779,
+      "learning_rate": 3.4314785397991814e-06,
+      "loss": 0.5358,
+      "step": 7790
+    },
+    {
+      "epoch": 0.39769540610819354,
+      "grad_norm": 7.611870921179878,
+      "learning_rate": 3.4276460794877343e-06,
+      "loss": 0.5625,
+      "step": 7800
+    },
+    {
+      "epoch": 0.39820527201346045,
+      "grad_norm": 7.800640649074299,
+      "learning_rate": 3.4238110894223205e-06,
+      "loss": 0.6573,
+      "step": 7810
+    },
+    {
+      "epoch": 0.39871513791872737,
+      "grad_norm": 5.387338032817162,
+      "learning_rate": 3.419973580061218e-06,
+      "loss": 0.5583,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3992250038239943,
+      "grad_norm": 11.84450391470328,
+      "learning_rate": 3.4161335618695774e-06,
+      "loss": 0.6167,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3997348697292612,
+      "grad_norm": 7.947927109201323,
+      "learning_rate": 3.4122910453193885e-06,
+      "loss": 0.5733,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4002447356345281,
+      "grad_norm": 18.10683231004345,
+      "learning_rate": 3.4084460408894553e-06,
+      "loss": 0.6145,
+      "step": 7850
+    },
+    {
+      "epoch": 0.40075460153979503,
+      "grad_norm": 15.775686326362997,
+      "learning_rate": 3.4045985590653667e-06,
+      "loss": 0.6136,
+      "step": 7860
+    },
+    {
+      "epoch": 0.40126446744506195,
+      "grad_norm": 20.43302274856715,
+      "learning_rate": 3.4007486103394678e-06,
+      "loss": 0.5913,
+      "step": 7870
+    },
+    {
+      "epoch": 0.40177433335032886,
+      "grad_norm": 6.018831322305339,
+      "learning_rate": 3.3968962052108288e-06,
+      "loss": 0.5405,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4022841992555958,
+      "grad_norm": 13.816860148733037,
+      "learning_rate": 3.3930413541852235e-06,
+      "loss": 0.5905,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4027940651608627,
+      "grad_norm": 20.814324321343378,
+      "learning_rate": 3.389184067775091e-06,
+      "loss": 0.5492,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4033039310661296,
+      "grad_norm": 13.912098590539072,
+      "learning_rate": 3.385324356499513e-06,
+      "loss": 0.5587,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4038137969713965,
+      "grad_norm": 7.463579403533143,
+      "learning_rate": 3.381462230884187e-06,
+      "loss": 0.5488,
+      "step": 7920
+    },
+    {
+      "epoch": 0.40432366287666344,
+      "grad_norm": 4.290865201752349,
+      "learning_rate": 3.377597701461391e-06,
+      "loss": 0.5305,
+      "step": 7930
+    },
+    {
+      "epoch": 0.40483352878193035,
+      "grad_norm": 15.062253635702449,
+      "learning_rate": 3.3737307787699593e-06,
+      "loss": 0.5701,
+      "step": 7940
+    },
+    {
+      "epoch": 0.40534339468719727,
+      "grad_norm": 10.230006901428354,
+      "learning_rate": 3.3698614733552537e-06,
+      "loss": 0.5206,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4058532605924642,
+      "grad_norm": 36.51788649803484,
+      "learning_rate": 3.3659897957691334e-06,
+      "loss": 0.6388,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4063631264977311,
+      "grad_norm": 12.636380989469782,
+      "learning_rate": 3.3621157565699265e-06,
+      "loss": 0.5556,
+      "step": 7970
+    },
+    {
+      "epoch": 0.406872992402998,
+      "grad_norm": 7.421584163218057,
+      "learning_rate": 3.3582393663224012e-06,
+      "loss": 0.5578,
+      "step": 7980
+    },
+    {
+      "epoch": 0.40738285830826493,
+      "grad_norm": 5.88874288509878,
+      "learning_rate": 3.3543606355977377e-06,
+      "loss": 0.5361,
+      "step": 7990
+    },
+    {
+      "epoch": 0.40789272421353184,
+      "grad_norm": 8.416084341931638,
+      "learning_rate": 3.350479574973498e-06,
+      "loss": 0.5752,
+      "step": 8000
+    },
+    {
+      "epoch": 0.40840259011879876,
+      "grad_norm": 11.531771923553494,
+      "learning_rate": 3.3465961950335994e-06,
+      "loss": 0.5378,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4089124560240657,
+      "grad_norm": 23.79130168481462,
+      "learning_rate": 3.342710506368282e-06,
+      "loss": 0.5193,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4094223219293326,
+      "grad_norm": 15.518320494738575,
+      "learning_rate": 3.3388225195740844e-06,
+      "loss": 0.5552,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4099321878345995,
+      "grad_norm": 15.601458428258836,
+      "learning_rate": 3.33493224525381e-06,
+      "loss": 0.5102,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4104420537398664,
+      "grad_norm": 11.821263743959667,
+      "learning_rate": 3.3310396940165025e-06,
+      "loss": 0.5871,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41095191964513333,
+      "grad_norm": 14.165232414945066,
+      "learning_rate": 3.327144876477413e-06,
+      "loss": 0.6122,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41146178555040025,
+      "grad_norm": 10.710392958957403,
+      "learning_rate": 3.3232478032579746e-06,
+      "loss": 0.5539,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41197165145566716,
+      "grad_norm": 14.032550679107109,
+      "learning_rate": 3.319348484985771e-06,
+      "loss": 0.5259,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4124815173609341,
+      "grad_norm": 29.595540369097325,
+      "learning_rate": 3.3154469322945083e-06,
+      "loss": 0.5379,
+      "step": 8090
+    },
+    {
+      "epoch": 0.412991383266201,
+      "grad_norm": 5.9496106362966605,
+      "learning_rate": 3.311543155823985e-06,
+      "loss": 0.6027,
+      "step": 8100
+    },
+    {
+      "epoch": 0.4135012491714679,
+      "grad_norm": 10.511494733407762,
+      "learning_rate": 3.3076371662200666e-06,
+      "loss": 0.5592,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4140111150767348,
+      "grad_norm": 10.936585430312546,
+      "learning_rate": 3.303728974134653e-06,
+      "loss": 0.5493,
+      "step": 8120
+    },
+    {
+      "epoch": 0.41452098098200174,
+      "grad_norm": 5.7440596533209005,
+      "learning_rate": 3.2998185902256475e-06,
+      "loss": 0.5621,
+      "step": 8130
+    },
+    {
+      "epoch": 0.41503084688726866,
+      "grad_norm": 10.573257615848048,
+      "learning_rate": 3.295906025156935e-06,
+      "loss": 0.6092,
+      "step": 8140
+    },
+    {
+      "epoch": 0.41554071279253557,
+      "grad_norm": 32.59649320062619,
+      "learning_rate": 3.291991289598347e-06,
+      "loss": 0.5696,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4160505786978025,
+      "grad_norm": 24.663894982370834,
+      "learning_rate": 3.288074394225632e-06,
+      "loss": 0.6018,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4165604446030694,
+      "grad_norm": 16.106665031926827,
+      "learning_rate": 3.284155349720431e-06,
+      "loss": 0.584,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4170703105083363,
+      "grad_norm": 7.162412242443836,
+      "learning_rate": 3.2802341667702448e-06,
+      "loss": 0.6162,
+      "step": 8180
+    },
+    {
+      "epoch": 0.41758017641360323,
+      "grad_norm": 3.99419268665801,
+      "learning_rate": 3.276310856068406e-06,
+      "loss": 0.624,
+      "step": 8190
+    },
+    {
+      "epoch": 0.41809004231887015,
+      "grad_norm": 18.830509521933557,
+      "learning_rate": 3.27238542831405e-06,
+      "loss": 0.5375,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41859990822413706,
+      "grad_norm": 13.211612862937661,
+      "learning_rate": 3.2684578942120853e-06,
+      "loss": 0.5065,
+      "step": 8210
+    },
+    {
+      "epoch": 0.419109774129404,
+      "grad_norm": 11.187081189820343,
+      "learning_rate": 3.2645282644731648e-06,
+      "loss": 0.5975,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4196196400346709,
+      "grad_norm": 11.56668755275729,
+      "learning_rate": 3.2605965498136554e-06,
+      "loss": 0.5661,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4201295059399378,
+      "grad_norm": 7.676472574372089,
+      "learning_rate": 3.2566627609556117e-06,
+      "loss": 0.5583,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4206393718452047,
+      "grad_norm": 12.256179639373476,
+      "learning_rate": 3.252726908626742e-06,
+      "loss": 0.6085,
+      "step": 8250
+    },
+    {
+      "epoch": 0.42114923775047164,
+      "grad_norm": 8.779295692765299,
+      "learning_rate": 3.248789003560385e-06,
+      "loss": 0.4848,
+      "step": 8260
+    },
+    {
+      "epoch": 0.42165910365573855,
+      "grad_norm": 8.906852808489715,
+      "learning_rate": 3.2448490564954744e-06,
+      "loss": 0.5238,
+      "step": 8270
+    },
+    {
+      "epoch": 0.42216896956100547,
+      "grad_norm": 11.99953771698637,
+      "learning_rate": 3.2409070781765147e-06,
+      "loss": 0.5521,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4226788354662724,
+      "grad_norm": 13.350283128074937,
+      "learning_rate": 3.236963079353548e-06,
+      "loss": 0.5584,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4231887013715393,
+      "grad_norm": 6.821633592421078,
+      "learning_rate": 3.2330170707821283e-06,
+      "loss": 0.5664,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4236985672768062,
+      "grad_norm": 7.056662121640499,
+      "learning_rate": 3.229069063223289e-06,
+      "loss": 0.5773,
+      "step": 8310
+    },
+    {
+      "epoch": 0.42420843318207313,
+      "grad_norm": 9.355395068848395,
+      "learning_rate": 3.225119067443515e-06,
+      "loss": 0.5793,
+      "step": 8320
+    },
+    {
+      "epoch": 0.42471829908734005,
+      "grad_norm": 5.264154463454432,
+      "learning_rate": 3.2211670942147144e-06,
+      "loss": 0.5646,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42522816499260696,
+      "grad_norm": 9.501578808920266,
+      "learning_rate": 3.2172131543141865e-06,
+      "loss": 0.5505,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4257380308978739,
+      "grad_norm": 7.232532803125666,
+      "learning_rate": 3.2132572585245946e-06,
+      "loss": 0.6165,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4262478968031408,
+      "grad_norm": 9.113984571929457,
+      "learning_rate": 3.209299417633936e-06,
+      "loss": 0.5958,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4267577627084077,
+      "grad_norm": 7.578216172807771,
+      "learning_rate": 3.2053396424355105e-06,
+      "loss": 0.6117,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4272676286136746,
+      "grad_norm": 7.320511070141178,
+      "learning_rate": 3.201377943727896e-06,
+      "loss": 0.551,
+      "step": 8380
+    },
+    {
+      "epoch": 0.42777749451894154,
+      "grad_norm": 14.982785634275919,
+      "learning_rate": 3.197414332314914e-06,
+      "loss": 0.4761,
+      "step": 8390
+    },
+    {
+      "epoch": 0.42828736042420845,
+      "grad_norm": 7.019223513877667,
+      "learning_rate": 3.193448819005603e-06,
+      "loss": 0.5969,
+      "step": 8400
+    },
+    {
+      "epoch": 0.42879722632947537,
+      "grad_norm": 19.840525872281795,
+      "learning_rate": 3.189481414614186e-06,
+      "loss": 0.5495,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4293070922347423,
+      "grad_norm": 22.995303030548026,
+      "learning_rate": 3.1855121299600454e-06,
+      "loss": 0.637,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4298169581400092,
+      "grad_norm": 5.869244255618574,
+      "learning_rate": 3.1815409758676917e-06,
+      "loss": 0.6386,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4303268240452761,
+      "grad_norm": 4.604957329242786,
+      "learning_rate": 3.1775679631667306e-06,
+      "loss": 0.5725,
+      "step": 8440
+    },
+    {
+      "epoch": 0.43083668995054303,
+      "grad_norm": 8.65884034038621,
+      "learning_rate": 3.1735931026918393e-06,
+      "loss": 0.6049,
+      "step": 8450
+    },
+    {
+      "epoch": 0.43134655585580994,
+      "grad_norm": 13.3608961962584,
+      "learning_rate": 3.1696164052827318e-06,
+      "loss": 0.5236,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43185642176107686,
+      "grad_norm": 25.089515960282526,
+      "learning_rate": 3.165637881784133e-06,
+      "loss": 0.4994,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4323662876663438,
+      "grad_norm": 8.409728068654365,
+      "learning_rate": 3.161657543045747e-06,
+      "loss": 0.6067,
+      "step": 8480
+    },
+    {
+      "epoch": 0.4328761535716107,
+      "grad_norm": 20.97435073144917,
+      "learning_rate": 3.1576753999222275e-06,
+      "loss": 0.5705,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4333860194768776,
+      "grad_norm": 33.23045329460443,
+      "learning_rate": 3.1536914632731512e-06,
+      "loss": 0.5923,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4338958853821445,
+      "grad_norm": 6.16600806964113,
+      "learning_rate": 3.1497057439629836e-06,
+      "loss": 0.5394,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43440575128741143,
+      "grad_norm": 10.017018232729027,
+      "learning_rate": 3.1457182528610526e-06,
+      "loss": 0.5935,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43491561719267835,
+      "grad_norm": 12.120848803469393,
+      "learning_rate": 3.1417290008415167e-06,
+      "loss": 0.5272,
+      "step": 8530
+    },
+    {
+      "epoch": 0.43542548309794527,
+      "grad_norm": 26.662754232184135,
+      "learning_rate": 3.1377379987833395e-06,
+      "loss": 0.5517,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4359353490032122,
+      "grad_norm": 23.485595791019445,
+      "learning_rate": 3.133745257570253e-06,
+      "loss": 0.6245,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4364452149084791,
+      "grad_norm": 10.087847255985833,
+      "learning_rate": 3.1297507880907357e-06,
+      "loss": 0.4898,
+      "step": 8560
+    },
+    {
+      "epoch": 0.436955080813746,
+      "grad_norm": 7.147241397833923,
+      "learning_rate": 3.1257546012379775e-06,
+      "loss": 0.543,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4374649467190129,
+      "grad_norm": 6.6109218274816195,
+      "learning_rate": 3.121756707909851e-06,
+      "loss": 0.5064,
+      "step": 8580
+    },
+    {
+      "epoch": 0.43797481262427984,
+      "grad_norm": 8.054310764050582,
+      "learning_rate": 3.1177571190088833e-06,
+      "loss": 0.6016,
+      "step": 8590
+    },
+    {
+      "epoch": 0.43848467852954676,
+      "grad_norm": 10.629143445576705,
+      "learning_rate": 3.1137558454422266e-06,
+      "loss": 0.5692,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43899454443481367,
+      "grad_norm": 4.656946338708994,
+      "learning_rate": 3.1097528981216245e-06,
+      "loss": 0.5431,
+      "step": 8610
+    },
+    {
+      "epoch": 0.43950441034008053,
+      "grad_norm": 7.686273185864347,
+      "learning_rate": 3.105748287963388e-06,
+      "loss": 0.5942,
+      "step": 8620
+    },
+    {
+      "epoch": 0.44001427624534745,
+      "grad_norm": 6.4017419151248784,
+      "learning_rate": 3.1017420258883607e-06,
+      "loss": 0.5951,
+      "step": 8630
+    },
+    {
+      "epoch": 0.44052414215061436,
+      "grad_norm": 7.715398459808364,
+      "learning_rate": 3.0977341228218916e-06,
+      "loss": 0.5274,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4410340080558813,
+      "grad_norm": 6.437445703730611,
+      "learning_rate": 3.0937245896938062e-06,
+      "loss": 0.496,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4415438739611482,
+      "grad_norm": 12.040187268475353,
+      "learning_rate": 3.089713437438373e-06,
+      "loss": 0.5444,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4420537398664151,
+      "grad_norm": 15.97318478204675,
+      "learning_rate": 3.085700676994277e-06,
+      "loss": 0.4989,
+      "step": 8670
+    },
+    {
+      "epoch": 0.442563605771682,
+      "grad_norm": 5.170088323073286,
+      "learning_rate": 3.0816863193045898e-06,
+      "loss": 0.5174,
+      "step": 8680
+    },
+    {
+      "epoch": 0.44307347167694894,
+      "grad_norm": 11.31806742780724,
+      "learning_rate": 3.0776703753167382e-06,
+      "loss": 0.5313,
+      "step": 8690
+    },
+    {
+      "epoch": 0.44358333758221585,
+      "grad_norm": 8.897162337165318,
+      "learning_rate": 3.0736528559824736e-06,
+      "loss": 0.5452,
+      "step": 8700
+    },
+    {
+      "epoch": 0.44409320348748277,
+      "grad_norm": 4.49436311321466,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.5471,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4446030693927497,
+      "grad_norm": 42.610576484138235,
+      "learning_rate": 3.0656131351031663e-06,
+      "loss": 0.5498,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4451129352980166,
+      "grad_norm": 12.31209500660843,
+      "learning_rate": 3.06159095548299e-06,
+      "loss": 0.5345,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4456228012032835,
+      "grad_norm": 8.692810512828606,
+      "learning_rate": 3.057567244366072e-06,
+      "loss": 0.5101,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44613266710855043,
+      "grad_norm": 10.979837107264158,
+      "learning_rate": 3.053542012725347e-06,
+      "loss": 0.5878,
+      "step": 8750
+    },
+    {
+      "epoch": 0.44664253301381734,
+      "grad_norm": 7.20795773464883,
+      "learning_rate": 3.049515271537896e-06,
+      "loss": 0.5345,
+      "step": 8760
+    },
+    {
+      "epoch": 0.44715239891908426,
+      "grad_norm": 23.767407508839465,
+      "learning_rate": 3.045487031784916e-06,
+      "loss": 0.5482,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4476622648243512,
+      "grad_norm": 8.421964173524868,
+      "learning_rate": 3.041457304451691e-06,
+      "loss": 0.5528,
+      "step": 8780
+    },
+    {
+      "epoch": 0.4481721307296181,
+      "grad_norm": 9.859187171670458,
+      "learning_rate": 3.0374261005275606e-06,
+      "loss": 0.5486,
+      "step": 8790
+    },
+    {
+      "epoch": 0.448681996634885,
+      "grad_norm": 6.790273628842832,
+      "learning_rate": 3.033393431005893e-06,
+      "loss": 0.5345,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4491918625401519,
+      "grad_norm": 10.253844615931142,
+      "learning_rate": 3.0293593068840514e-06,
+      "loss": 0.5658,
+      "step": 8810
+    },
+    {
+      "epoch": 0.44970172844541884,
+      "grad_norm": 6.617293173795455,
+      "learning_rate": 3.0253237391633684e-06,
+      "loss": 0.5748,
+      "step": 8820
+    },
+    {
+      "epoch": 0.45021159435068575,
+      "grad_norm": 11.268408655058423,
+      "learning_rate": 3.02128673884911e-06,
+      "loss": 0.5413,
+      "step": 8830
+    },
+    {
+      "epoch": 0.45072146025595267,
+      "grad_norm": 18.781346474180374,
+      "learning_rate": 3.017248316950452e-06,
+      "loss": 0.5501,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4512313261612196,
+      "grad_norm": 7.880045771676749,
+      "learning_rate": 3.0132084844804444e-06,
+      "loss": 0.5556,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4517411920664865,
+      "grad_norm": 22.59802767837297,
+      "learning_rate": 3.0091672524559855e-06,
+      "loss": 0.521,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4522510579717534,
+      "grad_norm": 6.6299156399345085,
+      "learning_rate": 3.0051246318977913e-06,
+      "loss": 0.5611,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4527609238770203,
+      "grad_norm": 98.67067204131763,
+      "learning_rate": 3.0010806338303615e-06,
+      "loss": 0.6875,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45327078978228724,
+      "grad_norm": 11.230266616071487,
+      "learning_rate": 2.997035269281955e-06,
+      "loss": 0.6012,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45378065568755416,
+      "grad_norm": 12.834508374143617,
+      "learning_rate": 2.9929885492845555e-06,
+      "loss": 0.5707,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4542905215928211,
+      "grad_norm": 8.358587417826085,
+      "learning_rate": 2.9889404848738434e-06,
+      "loss": 0.5561,
+      "step": 8910
+    },
+    {
+      "epoch": 0.454800387498088,
+      "grad_norm": 8.590838262002269,
+      "learning_rate": 2.9848910870891667e-06,
+      "loss": 0.559,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4553102534033549,
+      "grad_norm": 6.4262883397092425,
+      "learning_rate": 2.980840366973508e-06,
+      "loss": 0.6292,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4558201193086218,
+      "grad_norm": 8.386109243133818,
+      "learning_rate": 2.9767883355734567e-06,
+      "loss": 0.5761,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45632998521388873,
+      "grad_norm": 8.476299499333017,
+      "learning_rate": 2.9727350039391782e-06,
+      "loss": 0.5891,
+      "step": 8950
+    },
+    {
+      "epoch": 0.45683985111915565,
+      "grad_norm": 69.39163092666091,
+      "learning_rate": 2.968680383124384e-06,
+      "loss": 0.5577,
+      "step": 8960
+    },
+    {
+      "epoch": 0.45734971702442256,
+      "grad_norm": 6.599382708009659,
+      "learning_rate": 2.9646244841862996e-06,
+      "loss": 0.5711,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4578595829296895,
+      "grad_norm": 16.290153939994777,
+      "learning_rate": 2.9605673181856386e-06,
+      "loss": 0.5651,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4583694488349564,
+      "grad_norm": 9.201640373968438,
+      "learning_rate": 2.956508896186569e-06,
+      "loss": 0.5515,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4588793147402233,
+      "grad_norm": 19.028256485342254,
+      "learning_rate": 2.9524492292566824e-06,
+      "loss": 0.5783,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4593891806454902,
+      "grad_norm": 4.33654654210295,
+      "learning_rate": 2.948388328466968e-06,
+      "loss": 0.535,
+      "step": 9010
+    },
+    {
+      "epoch": 0.45989904655075714,
+      "grad_norm": 5.283839053332392,
+      "learning_rate": 2.944326204891777e-06,
+      "loss": 0.5128,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46040891245602406,
+      "grad_norm": 8.155730327753941,
+      "learning_rate": 2.940262869608798e-06,
+      "loss": 0.5879,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46091877836129097,
+      "grad_norm": 8.045593849591542,
+      "learning_rate": 2.9361983336990217e-06,
+      "loss": 0.5142,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4614286442665579,
+      "grad_norm": 7.886643075363955,
+      "learning_rate": 2.932132608246715e-06,
+      "loss": 0.5502,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4619385101718248,
+      "grad_norm": 16.63804533138256,
+      "learning_rate": 2.928065704339388e-06,
+      "loss": 0.5364,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4624483760770917,
+      "grad_norm": 12.882651747731957,
+      "learning_rate": 2.9239976330677627e-06,
+      "loss": 0.5808,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46295824198235863,
+      "grad_norm": 8.13150520823608,
+      "learning_rate": 2.919928405525748e-06,
+      "loss": 0.6199,
+      "step": 9080
+    },
+    {
+      "epoch": 0.46346810788762555,
+      "grad_norm": 11.681625440582726,
+      "learning_rate": 2.9158580328104036e-06,
+      "loss": 0.5239,
+      "step": 9090
+    },
+    {
+      "epoch": 0.46397797379289246,
+      "grad_norm": 6.198613990956929,
+      "learning_rate": 2.9117865260219117e-06,
+      "loss": 0.6272,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4644878396981594,
+      "grad_norm": 5.556099110950296,
+      "learning_rate": 2.907713896263551e-06,
+      "loss": 0.5443,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4649977056034263,
+      "grad_norm": 8.501028484471917,
+      "learning_rate": 2.903640154641657e-06,
+      "loss": 0.603,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4655075715086932,
+      "grad_norm": 5.631359258536221,
+      "learning_rate": 2.899565312265602e-06,
+      "loss": 0.5331,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4660174374139601,
+      "grad_norm": 9.654290186742468,
+      "learning_rate": 2.895489380247758e-06,
+      "loss": 0.5103,
+      "step": 9140
+    },
+    {
+      "epoch": 0.46652730331922704,
+      "grad_norm": 9.735034286631029,
+      "learning_rate": 2.891412369703469e-06,
+      "loss": 0.5305,
+      "step": 9150
+    },
+    {
+      "epoch": 0.46703716922449395,
+      "grad_norm": 7.313438134736095,
+      "learning_rate": 2.887334291751019e-06,
+      "loss": 0.557,
+      "step": 9160
+    },
+    {
+      "epoch": 0.46754703512976087,
+      "grad_norm": 11.519400660039398,
+      "learning_rate": 2.883255157511605e-06,
+      "loss": 0.555,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4680569010350278,
+      "grad_norm": 6.659958582649854,
+      "learning_rate": 2.8791749781093036e-06,
+      "loss": 0.5345,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4685667669402947,
+      "grad_norm": 21.81366907771895,
+      "learning_rate": 2.8750937646710416e-06,
+      "loss": 0.5116,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4690766328455616,
+      "grad_norm": 36.00237424353747,
+      "learning_rate": 2.8710115283265655e-06,
+      "loss": 0.6041,
+      "step": 9200
+    },
+    {
+      "epoch": 0.46958649875082853,
+      "grad_norm": 8.5235261068511,
+      "learning_rate": 2.866928280208412e-06,
+      "loss": 0.498,
+      "step": 9210
+    },
+    {
+      "epoch": 0.47009636465609544,
+      "grad_norm": 7.167372979618883,
+      "learning_rate": 2.8628440314518752e-06,
+      "loss": 0.5538,
+      "step": 9220
+    },
+    {
+      "epoch": 0.47060623056136236,
+      "grad_norm": 11.507343763740069,
+      "learning_rate": 2.8587587931949806e-06,
+      "loss": 0.5146,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4711160964666293,
+      "grad_norm": 6.250940117415326,
+      "learning_rate": 2.854672576578451e-06,
+      "loss": 0.5204,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4716259623718962,
+      "grad_norm": 7.530321442704125,
+      "learning_rate": 2.850585392745676e-06,
+      "loss": 0.4885,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4721358282771631,
+      "grad_norm": 9.163440002400211,
+      "learning_rate": 2.8464972528426847e-06,
+      "loss": 0.5226,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47264569418243,
+      "grad_norm": 16.393409728695335,
+      "learning_rate": 2.842408168018112e-06,
+      "loss": 0.6392,
+      "step": 9270
+    },
+    {
+      "epoch": 0.47315556008769694,
+      "grad_norm": 7.851049582591664,
+      "learning_rate": 2.8383181494231714e-06,
+      "loss": 0.5739,
+      "step": 9280
+    },
+    {
+      "epoch": 0.47366542599296385,
+      "grad_norm": 12.269773289481154,
+      "learning_rate": 2.8342272082116214e-06,
+      "loss": 0.5585,
+      "step": 9290
+    },
+    {
+      "epoch": 0.47417529189823077,
+      "grad_norm": 7.346796842344014,
+      "learning_rate": 2.830135355539737e-06,
+      "loss": 0.5394,
+      "step": 9300
+    },
+    {
+      "epoch": 0.4746851578034977,
+      "grad_norm": 7.796805559267158,
+      "learning_rate": 2.826042602566279e-06,
+      "loss": 0.5542,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4751950237087646,
+      "grad_norm": 11.101461591062174,
+      "learning_rate": 2.821948960452463e-06,
+      "loss": 0.5394,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4757048896140315,
+      "grad_norm": 11.397898491808107,
+      "learning_rate": 2.817854440361929e-06,
+      "loss": 0.5437,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4762147555192984,
+      "grad_norm": 8.962711749510742,
+      "learning_rate": 2.8137590534607123e-06,
+      "loss": 0.5814,
+      "step": 9340
+    },
+    {
+      "epoch": 0.47672462142456534,
+      "grad_norm": 7.963488384641039,
+      "learning_rate": 2.8096628109172125e-06,
+      "loss": 0.5331,
+      "step": 9350
+    },
+    {
+      "epoch": 0.47723448732983226,
+      "grad_norm": 8.368271605384814,
+      "learning_rate": 2.8055657239021605e-06,
+      "loss": 0.5423,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4777443532350992,
+      "grad_norm": 20.145593761398835,
+      "learning_rate": 2.8014678035885913e-06,
+      "loss": 0.5616,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4782542191403661,
+      "grad_norm": 10.98955410252151,
+      "learning_rate": 2.7973690611518124e-06,
+      "loss": 0.6045,
+      "step": 9380
+    },
+    {
+      "epoch": 0.478764085045633,
+      "grad_norm": 8.260243475993803,
+      "learning_rate": 2.7932695077693745e-06,
+      "loss": 0.5665,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4792739509508999,
+      "grad_norm": 13.66264667588255,
+      "learning_rate": 2.7891691546210374e-06,
+      "loss": 0.4857,
+      "step": 9400
+    },
+    {
+      "epoch": 0.47978381685616683,
+      "grad_norm": 8.815539885006997,
+      "learning_rate": 2.7850680128887424e-06,
+      "loss": 0.5462,
+      "step": 9410
+    },
+    {
+      "epoch": 0.48029368276143375,
+      "grad_norm": 13.61453326007113,
+      "learning_rate": 2.780966093756584e-06,
+      "loss": 0.5523,
+      "step": 9420
+    },
+    {
+      "epoch": 0.48080354866670066,
+      "grad_norm": 9.12709099410278,
+      "learning_rate": 2.7768634084107736e-06,
+      "loss": 0.5618,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4813134145719676,
+      "grad_norm": 6.502630279665426,
+      "learning_rate": 2.772759968039614e-06,
+      "loss": 0.5186,
+      "step": 9440
+    },
+    {
+      "epoch": 0.4818232804772345,
+      "grad_norm": 13.15622645433065,
+      "learning_rate": 2.7686557838334644e-06,
+      "loss": 0.6274,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4823331463825014,
+      "grad_norm": 9.81593275094567,
+      "learning_rate": 2.764550866984716e-06,
+      "loss": 0.5152,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4828430122877683,
+      "grad_norm": 7.4162131472691,
+      "learning_rate": 2.760445228687755e-06,
+      "loss": 0.5692,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48335287819303524,
+      "grad_norm": 16.883412871544444,
+      "learning_rate": 2.7563388801389386e-06,
+      "loss": 0.5638,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48386274409830216,
+      "grad_norm": 7.817244543088561,
+      "learning_rate": 2.752231832536556e-06,
+      "loss": 0.5365,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48437261000356907,
+      "grad_norm": 5.344543331403643,
+      "learning_rate": 2.7481240970808074e-06,
+      "loss": 0.5585,
+      "step": 9500
+    },
+    {
+      "epoch": 0.484882475908836,
+      "grad_norm": 9.552447608535903,
+      "learning_rate": 2.744015684973766e-06,
+      "loss": 0.599,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4853923418141029,
+      "grad_norm": 7.09922339815139,
+      "learning_rate": 2.739906607419351e-06,
+      "loss": 0.6551,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4859022077193698,
+      "grad_norm": 9.793826991717298,
+      "learning_rate": 2.7357968756232963e-06,
+      "loss": 0.5411,
+      "step": 9530
+    },
+    {
+      "epoch": 0.48641207362463673,
+      "grad_norm": 9.05956350099835,
+      "learning_rate": 2.7316865007931208e-06,
+      "loss": 0.5401,
+      "step": 9540
+    },
+    {
+      "epoch": 0.48692193952990365,
+      "grad_norm": 30.11492757634757,
+      "learning_rate": 2.727575494138096e-06,
+      "loss": 0.5806,
+      "step": 9550
+    },
+    {
+      "epoch": 0.48743180543517056,
+      "grad_norm": 9.075866648406908,
+      "learning_rate": 2.7234638668692166e-06,
+      "loss": 0.514,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4879416713404375,
+      "grad_norm": 6.186508572080813,
+      "learning_rate": 2.7193516301991703e-06,
+      "loss": 0.5615,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4884515372457044,
+      "grad_norm": 10.472335454023918,
+      "learning_rate": 2.7152387953423047e-06,
+      "loss": 0.5532,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4889614031509713,
+      "grad_norm": 4.820108163742004,
+      "learning_rate": 2.711125373514602e-06,
+      "loss": 0.4781,
+      "step": 9590
+    },
+    {
+      "epoch": 0.4894712690562382,
+      "grad_norm": 8.995560394442036,
+      "learning_rate": 2.7070113759336424e-06,
+      "loss": 0.5203,
+      "step": 9600
+    },
+    {
+      "epoch": 0.48998113496150514,
+      "grad_norm": 17.393326675668504,
+      "learning_rate": 2.7028968138185783e-06,
+      "loss": 0.543,
+      "step": 9610
+    },
+    {
+      "epoch": 0.49049100086677205,
+      "grad_norm": 2.8353076560498707,
+      "learning_rate": 2.6987816983900995e-06,
+      "loss": 0.5306,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49100086677203897,
+      "grad_norm": 9.503152576570983,
+      "learning_rate": 2.6946660408704062e-06,
+      "loss": 0.5211,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4915107326773059,
+      "grad_norm": 320.61604317399934,
+      "learning_rate": 2.6905498524831763e-06,
+      "loss": 0.6006,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4920205985825728,
+      "grad_norm": 8.724977975058774,
+      "learning_rate": 2.6864331444535347e-06,
+      "loss": 0.5338,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4925304644878397,
+      "grad_norm": 19.80012541814024,
+      "learning_rate": 2.682315928008026e-06,
+      "loss": 0.4973,
+      "step": 9660
+    },
+    {
+      "epoch": 0.49304033039310663,
+      "grad_norm": 6.900731066386949,
+      "learning_rate": 2.6781982143745776e-06,
+      "loss": 0.5044,
+      "step": 9670
+    },
+    {
+      "epoch": 0.49355019629837354,
+      "grad_norm": 17.42457089480257,
+      "learning_rate": 2.6740800147824764e-06,
+      "loss": 0.5013,
+      "step": 9680
+    },
+    {
+      "epoch": 0.49406006220364046,
+      "grad_norm": 13.80209829216203,
+      "learning_rate": 2.669961340462332e-06,
+      "loss": 0.6106,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4945699281089074,
+      "grad_norm": 9.356816676968172,
+      "learning_rate": 2.6658422026460493e-06,
+      "loss": 0.5256,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4950797940141743,
+      "grad_norm": 9.84379922723854,
+      "learning_rate": 2.6617226125667977e-06,
+      "loss": 0.5762,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4955896599194412,
+      "grad_norm": 15.230940811889656,
+      "learning_rate": 2.65760258145898e-06,
+      "loss": 0.5258,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4960995258247081,
+      "grad_norm": 8.986235423766201,
+      "learning_rate": 2.653482120558201e-06,
+      "loss": 0.593,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49660939172997504,
+      "grad_norm": 35.48183796431665,
+      "learning_rate": 2.6493612411012377e-06,
+      "loss": 0.5196,
+      "step": 9740
+    },
+    {
+      "epoch": 0.49711925763524195,
+      "grad_norm": 7.2752655638575,
+      "learning_rate": 2.645239954326009e-06,
+      "loss": 0.5706,
+      "step": 9750
+    },
+    {
+      "epoch": 0.49762912354050887,
+      "grad_norm": 8.42331057856242,
+      "learning_rate": 2.641118271471543e-06,
+      "loss": 0.4943,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4981389894457758,
+      "grad_norm": 7.965377523949411,
+      "learning_rate": 2.6369962037779513e-06,
+      "loss": 0.5339,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4986488553510427,
+      "grad_norm": 10.01729377702691,
+      "learning_rate": 2.632873762486392e-06,
+      "loss": 0.5791,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4991587212563096,
+      "grad_norm": 4.511761837890751,
+      "learning_rate": 2.6287509588390424e-06,
+      "loss": 0.5034,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4996685871615765,
+      "grad_norm": 9.294212001685729,
+      "learning_rate": 2.6246278040790696e-06,
+      "loss": 0.5604,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5001784530668434,
+      "grad_norm": 18.481234335838135,
+      "learning_rate": 2.620504309450596e-06,
+      "loss": 0.5656,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5006883189721103,
+      "grad_norm": 6.461543556394008,
+      "learning_rate": 2.6163804861986735e-06,
+      "loss": 0.5326,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5011981848773772,
+      "grad_norm": 6.775693055786728,
+      "learning_rate": 2.6122563455692467e-06,
+      "loss": 0.5439,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5017080507826441,
+      "grad_norm": 10.348144692895449,
+      "learning_rate": 2.608131898809129e-06,
+      "loss": 0.5851,
+      "step": 9840
+    },
+    {
+      "epoch": 0.502217916687911,
+      "grad_norm": 19.51021168023746,
+      "learning_rate": 2.6040071571659676e-06,
+      "loss": 0.519,
+      "step": 9850
+    },
+    {
+      "epoch": 0.502727782593178,
+      "grad_norm": 8.087870299255922,
+      "learning_rate": 2.5998821318882127e-06,
+      "loss": 0.5689,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5032376484984449,
+      "grad_norm": 17.52411247215236,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.5387,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5037475144037118,
+      "grad_norm": 5.141607773429428,
+      "learning_rate": 2.5916312754265636e-06,
+      "loss": 0.5233,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5042573803089787,
+      "grad_norm": 5.366874066818395,
+      "learning_rate": 2.587505466743317e-06,
+      "loss": 0.5694,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5047672462142456,
+      "grad_norm": 9.064599177565993,
+      "learning_rate": 2.583379419426709e-06,
+      "loss": 0.5833,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5052771121195125,
+      "grad_norm": 7.81476080754217,
+      "learning_rate": 2.5792531447287515e-06,
+      "loss": 0.5187,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5057869780247795,
+      "grad_norm": 6.046482753128124,
+      "learning_rate": 2.575126653902078e-06,
+      "loss": 0.4487,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5062968439300464,
+      "grad_norm": 5.090405722171164,
+      "learning_rate": 2.570999958199908e-06,
+      "loss": 0.5504,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5068067098353133,
+      "grad_norm": 10.708861623918402,
+      "learning_rate": 2.566873068876022e-06,
+      "loss": 0.5766,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5073165757405802,
+      "grad_norm": 11.907291428749533,
+      "learning_rate": 2.5627459971847264e-06,
+      "loss": 0.5025,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5078264416458471,
+      "grad_norm": 14.482994499858956,
+      "learning_rate": 2.558618754380829e-06,
+      "loss": 0.5443,
+      "step": 9960
+    },
+    {
+      "epoch": 0.508336307551114,
+      "grad_norm": 15.282044708608089,
+      "learning_rate": 2.5544913517196006e-06,
+      "loss": 0.5886,
+      "step": 9970
+    },
+    {
+      "epoch": 0.508846173456381,
+      "grad_norm": 8.06318617079194,
+      "learning_rate": 2.5503638004567487e-06,
+      "loss": 0.5749,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5093560393616479,
+      "grad_norm": 4.210852657585504,
+      "learning_rate": 2.546236111848387e-06,
+      "loss": 0.5345,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5098659052669148,
+      "grad_norm": 10.272690227765827,
+      "learning_rate": 2.5421082971510024e-06,
+      "loss": 0.5987,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5103757711721817,
+      "grad_norm": 18.85294522582985,
+      "learning_rate": 2.537980367621427e-06,
+      "loss": 0.5536,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5108856370774486,
+      "grad_norm": 6.947032878897913,
+      "learning_rate": 2.533852334516805e-06,
+      "loss": 0.5668,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5113955029827155,
+      "grad_norm": 6.178991706671281,
+      "learning_rate": 2.5297242090945638e-06,
+      "loss": 0.547,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5119053688879824,
+      "grad_norm": 9.380242412444959,
+      "learning_rate": 2.5255960026123825e-06,
+      "loss": 0.5135,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5124152347932494,
+      "grad_norm": 17.20190562415558,
+      "learning_rate": 2.5214677263281613e-06,
+      "loss": 0.5542,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5129251006985163,
+      "grad_norm": 4.702811357253125,
+      "learning_rate": 2.5173393914999894e-06,
+      "loss": 0.5898,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5134349666037832,
+      "grad_norm": 5.979778153512825,
+      "learning_rate": 2.5132110093861174e-06,
+      "loss": 0.49,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5139448325090501,
+      "grad_norm": 8.444482069508595,
+      "learning_rate": 2.5090825912449233e-06,
+      "loss": 0.5111,
+      "step": 10080
+    },
+    {
+      "epoch": 0.514454698414317,
+      "grad_norm": 4.728506915605374,
+      "learning_rate": 2.504954148334886e-06,
+      "loss": 0.5028,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5149645643195839,
+      "grad_norm": 26.205972531140272,
+      "learning_rate": 2.500825691914549e-06,
+      "loss": 0.5315,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5154744302248508,
+      "grad_norm": 17.05895922461699,
+      "learning_rate": 2.4966972332424934e-06,
+      "loss": 0.5929,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5159842961301178,
+      "grad_norm": 11.798337674324719,
+      "learning_rate": 2.492568783577308e-06,
+      "loss": 0.5302,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5164941620353847,
+      "grad_norm": 5.8570796040887245,
+      "learning_rate": 2.4884403541775553e-06,
+      "loss": 0.5408,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5170040279406516,
+      "grad_norm": 37.79091695568464,
+      "learning_rate": 2.4843119563017426e-06,
+      "loss": 0.5014,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5175138938459185,
+      "grad_norm": 30.287463725605825,
+      "learning_rate": 2.480183601208293e-06,
+      "loss": 0.5793,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5180237597511854,
+      "grad_norm": 6.331161625657528,
+      "learning_rate": 2.4760553001555105e-06,
+      "loss": 0.5064,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5185336256564523,
+      "grad_norm": 3.476045278909216,
+      "learning_rate": 2.4719270644015533e-06,
+      "loss": 0.5359,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5190434915617193,
+      "grad_norm": 22.474934468593176,
+      "learning_rate": 2.4677989052044022e-06,
+      "loss": 0.5924,
+      "step": 10180
+    },
+    {
+      "epoch": 0.5195533574669862,
+      "grad_norm": 21.07848740636244,
+      "learning_rate": 2.4636708338218267e-06,
+      "loss": 0.5211,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5200632233722531,
+      "grad_norm": 32.20759395836686,
+      "learning_rate": 2.4595428615113596e-06,
+      "loss": 0.512,
+      "step": 10200
+    },
+    {
+      "epoch": 0.52057308927752,
+      "grad_norm": 10.788375544192391,
+      "learning_rate": 2.4554149995302605e-06,
+      "loss": 0.5667,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5210829551827869,
+      "grad_norm": 5.973113159345188,
+      "learning_rate": 2.451287259135491e-06,
+      "loss": 0.5772,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5215928210880538,
+      "grad_norm": 5.622874034318361,
+      "learning_rate": 2.4471596515836797e-06,
+      "loss": 0.4957,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5221026869933207,
+      "grad_norm": 32.96639762729161,
+      "learning_rate": 2.4430321881310928e-06,
+      "loss": 0.5963,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5226125528985877,
+      "grad_norm": 10.127055741762746,
+      "learning_rate": 2.4389048800336056e-06,
+      "loss": 0.4812,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5231224188038546,
+      "grad_norm": 6.503535423951574,
+      "learning_rate": 2.434777738546666e-06,
+      "loss": 0.5536,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5236322847091215,
+      "grad_norm": 8.731984562436182,
+      "learning_rate": 2.4306507749252715e-06,
+      "loss": 0.5391,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5241421506143884,
+      "grad_norm": 14.54106883066779,
+      "learning_rate": 2.426524000423931e-06,
+      "loss": 0.5646,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5246520165196553,
+      "grad_norm": 8.987888618292743,
+      "learning_rate": 2.4223974262966395e-06,
+      "loss": 0.5405,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5251618824249222,
+      "grad_norm": 7.432453186270774,
+      "learning_rate": 2.4182710637968466e-06,
+      "loss": 0.5267,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5256717483301891,
+      "grad_norm": 20.051730795291874,
+      "learning_rate": 2.414144924177422e-06,
+      "loss": 0.4944,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5261816142354561,
+      "grad_norm": 9.238504273020023,
+      "learning_rate": 2.4100190186906304e-06,
+      "loss": 0.5266,
+      "step": 10320
+    },
+    {
+      "epoch": 0.526691480140723,
+      "grad_norm": 6.3325546925437655,
+      "learning_rate": 2.4058933585880958e-06,
+      "loss": 0.4788,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5272013460459899,
+      "grad_norm": 28.46273823581247,
+      "learning_rate": 2.401767955120776e-06,
+      "loss": 0.5674,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5277112119512568,
+      "grad_norm": 11.101555994175811,
+      "learning_rate": 2.3976428195389236e-06,
+      "loss": 0.524,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5282210778565237,
+      "grad_norm": 4.922045076561477,
+      "learning_rate": 2.393517963092066e-06,
+      "loss": 0.5428,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5287309437617906,
+      "grad_norm": 6.089180197710267,
+      "learning_rate": 2.3893933970289677e-06,
+      "loss": 0.4983,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5292408096670576,
+      "grad_norm": 8.7266534647084,
+      "learning_rate": 2.3852691325975996e-06,
+      "loss": 0.5968,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5297506755723245,
+      "grad_norm": 16.173655045642207,
+      "learning_rate": 2.3811451810451132e-06,
+      "loss": 0.5491,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5302605414775914,
+      "grad_norm": 14.188690969216273,
+      "learning_rate": 2.377021553617803e-06,
+      "loss": 0.5229,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5307704073828583,
+      "grad_norm": 15.947197750694937,
+      "learning_rate": 2.372898261561082e-06,
+      "loss": 0.5756,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5312802732881252,
+      "grad_norm": 22.949283972877854,
+      "learning_rate": 2.3687753161194498e-06,
+      "loss": 0.5335,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5317901391933921,
+      "grad_norm": 52.71294742628508,
+      "learning_rate": 2.3646527285364565e-06,
+      "loss": 0.4808,
+      "step": 10430
+    },
+    {
+      "epoch": 0.532300005098659,
+      "grad_norm": 4.074010052195157,
+      "learning_rate": 2.3605305100546807e-06,
+      "loss": 0.5642,
+      "step": 10440
+    },
+    {
+      "epoch": 0.532809871003926,
+      "grad_norm": 9.838693123496997,
+      "learning_rate": 2.356408671915692e-06,
+      "loss": 0.5526,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5333197369091929,
+      "grad_norm": 9.174659419815153,
+      "learning_rate": 2.352287225360024e-06,
+      "loss": 0.5293,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5338296028144598,
+      "grad_norm": 4.384586430304259,
+      "learning_rate": 2.3481661816271413e-06,
+      "loss": 0.5477,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5343394687197267,
+      "grad_norm": 11.497556219158662,
+      "learning_rate": 2.3440455519554096e-06,
+      "loss": 0.5239,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5348493346249936,
+      "grad_norm": 4.9267726313927795,
+      "learning_rate": 2.339925347582069e-06,
+      "loss": 0.4597,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5353592005302605,
+      "grad_norm": 6.460685665908354,
+      "learning_rate": 2.3358055797431945e-06,
+      "loss": 0.5595,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5358690664355275,
+      "grad_norm": 8.16287039128602,
+      "learning_rate": 2.3316862596736752e-06,
+      "loss": 0.5578,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5363789323407944,
+      "grad_norm": 11.339005579992447,
+      "learning_rate": 2.327567398607175e-06,
+      "loss": 0.4851,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5368887982460613,
+      "grad_norm": 10.028313148707044,
+      "learning_rate": 2.3234490077761097e-06,
+      "loss": 0.5108,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5373986641513282,
+      "grad_norm": 4.465502751196889,
+      "learning_rate": 2.3193310984116113e-06,
+      "loss": 0.5397,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5379085300565951,
+      "grad_norm": 5.70092739153591,
+      "learning_rate": 2.3152136817434983e-06,
+      "loss": 0.5576,
+      "step": 10550
+    },
+    {
+      "epoch": 0.538418395961862,
+      "grad_norm": 9.341114511026655,
+      "learning_rate": 2.311096769000247e-06,
+      "loss": 0.5845,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5389282618671289,
+      "grad_norm": 12.418743380860024,
+      "learning_rate": 2.306980371408957e-06,
+      "loss": 0.5421,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5394381277723959,
+      "grad_norm": 30.682568719995146,
+      "learning_rate": 2.302864500195326e-06,
+      "loss": 0.4822,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5399479936776628,
+      "grad_norm": 28.981287434258377,
+      "learning_rate": 2.2987491665836137e-06,
+      "loss": 0.5105,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5404578595829297,
+      "grad_norm": 7.180895523241582,
+      "learning_rate": 2.2946343817966147e-06,
+      "loss": 0.5413,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5409677254881966,
+      "grad_norm": 8.508579601103788,
+      "learning_rate": 2.2905201570556294e-06,
+      "loss": 0.5524,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5414775913934635,
+      "grad_norm": 5.738514236452168,
+      "learning_rate": 2.2864065035804253e-06,
+      "loss": 0.5221,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5419874572987304,
+      "grad_norm": 16.357357206923098,
+      "learning_rate": 2.2822934325892178e-06,
+      "loss": 0.5901,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5424973232039974,
+      "grad_norm": 8.393585681141637,
+      "learning_rate": 2.2781809552986296e-06,
+      "loss": 0.539,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5430071891092643,
+      "grad_norm": 9.054243646581382,
+      "learning_rate": 2.2740690829236672e-06,
+      "loss": 0.5317,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5435170550145312,
+      "grad_norm": 8.715999157087243,
+      "learning_rate": 2.269957826677685e-06,
+      "loss": 0.5474,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5440269209197981,
+      "grad_norm": 6.328287254963458,
+      "learning_rate": 2.2658471977723593e-06,
+      "loss": 0.5335,
+      "step": 10670
+    },
+    {
+      "epoch": 0.544536786825065,
+      "grad_norm": 17.818805969409258,
+      "learning_rate": 2.2617372074176565e-06,
+      "loss": 0.5085,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5450466527303319,
+      "grad_norm": 8.44665589054514,
+      "learning_rate": 2.2576278668217967e-06,
+      "loss": 0.5066,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5455565186355988,
+      "grad_norm": 8.836352465670993,
+      "learning_rate": 2.2535191871912337e-06,
+      "loss": 0.5242,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5460663845408658,
+      "grad_norm": 6.03086653823258,
+      "learning_rate": 2.2494111797306146e-06,
+      "loss": 0.529,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5465762504461327,
+      "grad_norm": 4.725682085559597,
+      "learning_rate": 2.2453038556427557e-06,
+      "loss": 0.5783,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5470861163513996,
+      "grad_norm": 6.910137878700736,
+      "learning_rate": 2.2411972261286107e-06,
+      "loss": 0.505,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5475959822566665,
+      "grad_norm": 13.116572965114555,
+      "learning_rate": 2.2370913023872357e-06,
+      "loss": 0.6158,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5481058481619334,
+      "grad_norm": 5.016888656730006,
+      "learning_rate": 2.2329860956157655e-06,
+      "loss": 0.5148,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5486157140672003,
+      "grad_norm": 7.704134433953211,
+      "learning_rate": 2.2288816170093767e-06,
+      "loss": 0.5255,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5491255799724672,
+      "grad_norm": 6.62013774387146,
+      "learning_rate": 2.224777877761264e-06,
+      "loss": 0.4716,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5496354458777342,
+      "grad_norm": 5.166956877073506,
+      "learning_rate": 2.2206748890626004e-06,
+      "loss": 0.5539,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5501453117830011,
+      "grad_norm": 12.107412717723916,
+      "learning_rate": 2.216572662102518e-06,
+      "loss": 0.5941,
+      "step": 10790
+    },
+    {
+      "epoch": 0.550655177688268,
+      "grad_norm": 12.185456102816504,
+      "learning_rate": 2.212471208068068e-06,
+      "loss": 0.4973,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5511650435935349,
+      "grad_norm": 12.953954914140017,
+      "learning_rate": 2.2083705381441933e-06,
+      "loss": 0.5258,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5516749094988018,
+      "grad_norm": 45.0233604442855,
+      "learning_rate": 2.204270663513702e-06,
+      "loss": 0.5006,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5521847754040687,
+      "grad_norm": 6.859535095869529,
+      "learning_rate": 2.200171595357229e-06,
+      "loss": 0.5132,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5526946413093357,
+      "grad_norm": 8.507813436801658,
+      "learning_rate": 2.196073344853213e-06,
+      "loss": 0.4566,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5532045072146026,
+      "grad_norm": 13.085163419859967,
+      "learning_rate": 2.1919759231778616e-06,
+      "loss": 0.577,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5537143731198695,
+      "grad_norm": 7.089280829159933,
+      "learning_rate": 2.187879341505123e-06,
+      "loss": 0.5239,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5542242390251364,
+      "grad_norm": 8.945899360906559,
+      "learning_rate": 2.1837836110066544e-06,
+      "loss": 0.5751,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5547341049304033,
+      "grad_norm": 13.650888729895103,
+      "learning_rate": 2.17968874285179e-06,
+      "loss": 0.6315,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5552439708356702,
+      "grad_norm": 16.259039053047864,
+      "learning_rate": 2.175594748207516e-06,
+      "loss": 0.5123,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5557538367409371,
+      "grad_norm": 37.239300000407646,
+      "learning_rate": 2.1715016382384314e-06,
+      "loss": 0.5585,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5562637026462041,
+      "grad_norm": 8.293582577335172,
+      "learning_rate": 2.1674094241067275e-06,
+      "loss": 0.4595,
+      "step": 10910
+    },
+    {
+      "epoch": 0.556773568551471,
+      "grad_norm": 7.899576476666288,
+      "learning_rate": 2.1633181169721518e-06,
+      "loss": 0.5378,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5572834344567379,
+      "grad_norm": 4.496485111294446,
+      "learning_rate": 2.159227727991974e-06,
+      "loss": 0.531,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5577933003620048,
+      "grad_norm": 43.58374674974216,
+      "learning_rate": 2.155138268320966e-06,
+      "loss": 0.5692,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5583031662672717,
+      "grad_norm": 17.668955644082533,
+      "learning_rate": 2.151049749111361e-06,
+      "loss": 0.5973,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5588130321725386,
+      "grad_norm": 11.098601977216765,
+      "learning_rate": 2.146962181512829e-06,
+      "loss": 0.5913,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5593228980778056,
+      "grad_norm": 14.526932562809277,
+      "learning_rate": 2.142875576672446e-06,
+      "loss": 0.5381,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5598327639830725,
+      "grad_norm": 6.722695581409926,
+      "learning_rate": 2.13878994573466e-06,
+      "loss": 0.5268,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5603426298883394,
+      "grad_norm": 7.053335246073939,
+      "learning_rate": 2.1347052998412667e-06,
+      "loss": 0.5253,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5608524957936063,
+      "grad_norm": 7.331713955991623,
+      "learning_rate": 2.1306216501313705e-06,
+      "loss": 0.5202,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5613623616988732,
+      "grad_norm": 17.5984337607932,
+      "learning_rate": 2.1265390077413643e-06,
+      "loss": 0.458,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5618722276041401,
+      "grad_norm": 32.27052143652757,
+      "learning_rate": 2.1224573838048894e-06,
+      "loss": 0.5248,
+      "step": 11020
+    },
+    {
+      "epoch": 0.562382093509407,
+      "grad_norm": 7.9197575597868966,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.4883,
+      "step": 11030
+    },
+    {
+      "epoch": 0.562891959414674,
+      "grad_norm": 5.525935578082481,
+      "learning_rate": 2.114297235813196e-06,
+      "loss": 0.5198,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5634018253199409,
+      "grad_norm": 8.462639242102355,
+      "learning_rate": 2.110218734011255e-06,
+      "loss": 0.4974,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5639116912252078,
+      "grad_norm": 7.481252936658193,
+      "learning_rate": 2.106141295169344e-06,
+      "loss": 0.5222,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5644215571304747,
+      "grad_norm": 5.849561031104596,
+      "learning_rate": 2.1020649304069144e-06,
+      "loss": 0.499,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5649314230357416,
+      "grad_norm": 5.449131688881039,
+      "learning_rate": 2.0979896508404917e-06,
+      "loss": 0.5126,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5654412889410085,
+      "grad_norm": 11.72768434163546,
+      "learning_rate": 2.0939154675836407e-06,
+      "loss": 0.4826,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5659511548462755,
+      "grad_norm": 10.72487124116107,
+      "learning_rate": 2.0898423917469344e-06,
+      "loss": 0.5187,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5664610207515424,
+      "grad_norm": 17.57913672580963,
+      "learning_rate": 2.085770434437931e-06,
+      "loss": 0.4703,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5669708866568093,
+      "grad_norm": 10.94435575365214,
+      "learning_rate": 2.0816996067611315e-06,
+      "loss": 0.514,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5674807525620762,
+      "grad_norm": 77.09310463771246,
+      "learning_rate": 2.0776299198179624e-06,
+      "loss": 0.6672,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5679906184673431,
+      "grad_norm": 6.817395796607053,
+      "learning_rate": 2.0735613847067355e-06,
+      "loss": 0.5581,
+      "step": 11140
+    },
+    {
+      "epoch": 0.56850048437261,
+      "grad_norm": 12.77230188101944,
+      "learning_rate": 2.0694940125226224e-06,
+      "loss": 0.5739,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5690103502778769,
+      "grad_norm": 3.4770700526025147,
+      "learning_rate": 2.0654278143576263e-06,
+      "loss": 0.562,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5695202161831439,
+      "grad_norm": 7.673055088082523,
+      "learning_rate": 2.0613628013005437e-06,
+      "loss": 0.569,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5700300820884108,
+      "grad_norm": 27.69235924478429,
+      "learning_rate": 2.0572989844369427e-06,
+      "loss": 0.5193,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5705399479936777,
+      "grad_norm": 8.25982359313178,
+      "learning_rate": 2.053236374849128e-06,
+      "loss": 0.52,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5710498138989446,
+      "grad_norm": 12.446644595207585,
+      "learning_rate": 2.049174983616113e-06,
+      "loss": 0.5377,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5715596798042115,
+      "grad_norm": 7.177576833130309,
+      "learning_rate": 2.045114821813588e-06,
+      "loss": 0.4965,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5720695457094784,
+      "grad_norm": 9.94974275794222,
+      "learning_rate": 2.0410559005138893e-06,
+      "loss": 0.5104,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5725794116147453,
+      "grad_norm": 4.755356595206868,
+      "learning_rate": 2.0369982307859728e-06,
+      "loss": 0.5485,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5730892775200123,
+      "grad_norm": 5.131064490503118,
+      "learning_rate": 2.032941823695378e-06,
+      "loss": 0.49,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5735991434252792,
+      "grad_norm": 30.454278464618394,
+      "learning_rate": 2.0288866903042055e-06,
+      "loss": 0.5414,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5741090093305461,
+      "grad_norm": 9.761350727583425,
+      "learning_rate": 2.024832841671077e-06,
+      "loss": 0.555,
+      "step": 11260
+    },
+    {
+      "epoch": 0.574618875235813,
+      "grad_norm": 9.070832317219523,
+      "learning_rate": 2.0207802888511155e-06,
+      "loss": 0.5431,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5751287411410799,
+      "grad_norm": 12.80477246245281,
+      "learning_rate": 2.0167290428959082e-06,
+      "loss": 0.5229,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5756386070463468,
+      "grad_norm": 20.566934327233376,
+      "learning_rate": 2.0126791148534777e-06,
+      "loss": 0.5416,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5761484729516138,
+      "grad_norm": 9.72319235009137,
+      "learning_rate": 2.0086305157682546e-06,
+      "loss": 0.4794,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5766583388568807,
+      "grad_norm": 17.305246483705588,
+      "learning_rate": 2.004583256681042e-06,
+      "loss": 0.5459,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5771682047621476,
+      "grad_norm": 6.1869155697423395,
+      "learning_rate": 2.0005373486289932e-06,
+      "loss": 0.5315,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5776780706674145,
+      "grad_norm": 7.02173694085358,
+      "learning_rate": 1.9964928026455715e-06,
+      "loss": 0.4243,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5781879365726814,
+      "grad_norm": 8.536789719667972,
+      "learning_rate": 1.9924496297605315e-06,
+      "loss": 0.5404,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5786978024779483,
+      "grad_norm": 6.550119622890504,
+      "learning_rate": 1.988407840999881e-06,
+      "loss": 0.4879,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5792076683832152,
+      "grad_norm": 9.137261097383348,
+      "learning_rate": 1.984367447385851e-06,
+      "loss": 0.5653,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5797175342884822,
+      "grad_norm": 7.224113742426661,
+      "learning_rate": 1.9803284599368704e-06,
+      "loss": 0.5484,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5802274001937491,
+      "grad_norm": 28.75988441418933,
+      "learning_rate": 1.976290889667533e-06,
+      "loss": 0.5581,
+      "step": 11380
+    },
+    {
+      "epoch": 0.580737266099016,
+      "grad_norm": 17.034587606823564,
+      "learning_rate": 1.9722547475885685e-06,
+      "loss": 0.5403,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5812471320042829,
+      "grad_norm": 8.363382694116806,
+      "learning_rate": 1.9682200447068095e-06,
+      "loss": 0.5392,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5817569979095498,
+      "grad_norm": 6.142045262902703,
+      "learning_rate": 1.9641867920251655e-06,
+      "loss": 0.5267,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5822668638148167,
+      "grad_norm": 5.818725842542696,
+      "learning_rate": 1.9601550005425925e-06,
+      "loss": 0.4801,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5827767297200837,
+      "grad_norm": 6.139658242859134,
+      "learning_rate": 1.9561246812540572e-06,
+      "loss": 0.5558,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5832865956253506,
+      "grad_norm": 8.805001339523916,
+      "learning_rate": 1.952095845150518e-06,
+      "loss": 0.4811,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5837964615306175,
+      "grad_norm": 13.305203178035956,
+      "learning_rate": 1.9480685032188816e-06,
+      "loss": 0.5293,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5843063274358844,
+      "grad_norm": 10.067357266300476,
+      "learning_rate": 1.9440426664419855e-06,
+      "loss": 0.5679,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5848161933411513,
+      "grad_norm": 6.7367825841837305,
+      "learning_rate": 1.940018345798561e-06,
+      "loss": 0.486,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5853260592464182,
+      "grad_norm": 4.7734179339308636,
+      "learning_rate": 1.935995552263202e-06,
+      "loss": 0.5227,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5858359251516851,
+      "grad_norm": 6.440122499331953,
+      "learning_rate": 1.931974296806342e-06,
+      "loss": 0.5004,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5863457910569521,
+      "grad_norm": 6.5132289741934315,
+      "learning_rate": 1.9279545903942174e-06,
+      "loss": 0.5213,
+      "step": 11500
+    },
+    {
+      "epoch": 0.586855656962219,
+      "grad_norm": 18.800925704781697,
+      "learning_rate": 1.923936443988841e-06,
+      "loss": 0.5312,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5873655228674859,
+      "grad_norm": 14.57112932904238,
+      "learning_rate": 1.9199198685479732e-06,
+      "loss": 0.5311,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5878753887727528,
+      "grad_norm": 10.180178069678226,
+      "learning_rate": 1.9159048750250855e-06,
+      "loss": 0.5216,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5883852546780197,
+      "grad_norm": 10.195783008921715,
+      "learning_rate": 1.9118914743693407e-06,
+      "loss": 0.597,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5888951205832866,
+      "grad_norm": 10.974937494383452,
+      "learning_rate": 1.907879677525554e-06,
+      "loss": 0.5168,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5894049864885536,
+      "grad_norm": 8.849536966836947,
+      "learning_rate": 1.9038694954341697e-06,
+      "loss": 0.4529,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5899148523938205,
+      "grad_norm": 3.3795489934918295,
+      "learning_rate": 1.8998609390312251e-06,
+      "loss": 0.4619,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5904247182990874,
+      "grad_norm": 6.395874103233876,
+      "learning_rate": 1.8958540192483277e-06,
+      "loss": 0.5475,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5909345842043543,
+      "grad_norm": 26.128124043964572,
+      "learning_rate": 1.8918487470126207e-06,
+      "loss": 0.5457,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5914444501096212,
+      "grad_norm": 7.414428933895366,
+      "learning_rate": 1.8878451332467529e-06,
+      "loss": 0.4924,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5919543160148881,
+      "grad_norm": 12.17596702715116,
+      "learning_rate": 1.8838431888688528e-06,
+      "loss": 0.5839,
+      "step": 11610
+    },
+    {
+      "epoch": 0.592464181920155,
+      "grad_norm": 7.214396505430331,
+      "learning_rate": 1.879842924792493e-06,
+      "loss": 0.5426,
+      "step": 11620
+    },
+    {
+      "epoch": 0.592974047825422,
+      "grad_norm": 3.3425281166048286,
+      "learning_rate": 1.8758443519266667e-06,
+      "loss": 0.498,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5934839137306889,
+      "grad_norm": 7.942972243447627,
+      "learning_rate": 1.8718474811757553e-06,
+      "loss": 0.4897,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5939937796359558,
+      "grad_norm": 9.82831623989867,
+      "learning_rate": 1.8678523234394956e-06,
+      "loss": 0.5543,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5945036455412227,
+      "grad_norm": 3.5276416237463493,
+      "learning_rate": 1.863858889612956e-06,
+      "loss": 0.4956,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5950135114464896,
+      "grad_norm": 5.747155748080999,
+      "learning_rate": 1.8598671905865002e-06,
+      "loss": 0.5495,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5955233773517565,
+      "grad_norm": 10.434001686139608,
+      "learning_rate": 1.8558772372457647e-06,
+      "loss": 0.532,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5960332432570234,
+      "grad_norm": 7.468458684107875,
+      "learning_rate": 1.8518890404716227e-06,
+      "loss": 0.4994,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5965431091622904,
+      "grad_norm": 5.821387671858344,
+      "learning_rate": 1.8479026111401594e-06,
+      "loss": 0.5085,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5970529750675573,
+      "grad_norm": 6.498163167113197,
+      "learning_rate": 1.8439179601226376e-06,
+      "loss": 0.4926,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5975628409728242,
+      "grad_norm": 3.903883912104386,
+      "learning_rate": 1.8399350982854717e-06,
+      "loss": 0.47,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5980727068780911,
+      "grad_norm": 17.69494525041306,
+      "learning_rate": 1.835954036490198e-06,
+      "loss": 0.5112,
+      "step": 11730
+    },
+    {
+      "epoch": 0.598582572783358,
+      "grad_norm": 8.871473917185979,
+      "learning_rate": 1.8319747855934416e-06,
+      "loss": 0.5604,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5990924386886249,
+      "grad_norm": 6.957782916536802,
+      "learning_rate": 1.8279973564468906e-06,
+      "loss": 0.4845,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5996023045938919,
+      "grad_norm": 15.674418544975776,
+      "learning_rate": 1.8240217598972665e-06,
+      "loss": 0.5703,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6001121704991588,
+      "grad_norm": 5.457187043090276,
+      "learning_rate": 1.8200480067862888e-06,
+      "loss": 0.5326,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6006220364044257,
+      "grad_norm": 19.100409980702484,
+      "learning_rate": 1.8160761079506553e-06,
+      "loss": 0.5163,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6011319023096926,
+      "grad_norm": 12.483490528108064,
+      "learning_rate": 1.812106074222002e-06,
+      "loss": 0.5,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6016417682149595,
+      "grad_norm": 10.438949202796984,
+      "learning_rate": 1.8081379164268826e-06,
+      "loss": 0.543,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6021516341202264,
+      "grad_norm": 10.21693346478386,
+      "learning_rate": 1.804171645386732e-06,
+      "loss": 0.5265,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6026615000254933,
+      "grad_norm": 10.090100648916717,
+      "learning_rate": 1.800207271917842e-06,
+      "loss": 0.5225,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6031713659307603,
+      "grad_norm": 9.33838726133408,
+      "learning_rate": 1.7962448068313298e-06,
+      "loss": 0.538,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6036812318360272,
+      "grad_norm": 9.295998599675029,
+      "learning_rate": 1.7922842609331053e-06,
+      "loss": 0.504,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6041910977412941,
+      "grad_norm": 7.436412661247303,
+      "learning_rate": 1.788325645023848e-06,
+      "loss": 0.4834,
+      "step": 11850
+    },
+    {
+      "epoch": 0.604700963646561,
+      "grad_norm": 11.230956866328153,
+      "learning_rate": 1.7843689698989715e-06,
+      "loss": 0.4796,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6052108295518279,
+      "grad_norm": 13.322406594718249,
+      "learning_rate": 1.7804142463486e-06,
+      "loss": 0.4953,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6057206954570948,
+      "grad_norm": 6.739605043613098,
+      "learning_rate": 1.776461485157531e-06,
+      "loss": 0.5836,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6062305613623618,
+      "grad_norm": 3.6900610967993064,
+      "learning_rate": 1.7725106971052147e-06,
+      "loss": 0.5154,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6067404272676287,
+      "grad_norm": 5.6557903709902675,
+      "learning_rate": 1.7685618929657193e-06,
+      "loss": 0.5439,
+      "step": 11900
+    },
+    {
+      "epoch": 0.6072502931728956,
+      "grad_norm": 19.661102790117276,
+      "learning_rate": 1.7646150835077014e-06,
+      "loss": 0.5327,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6077601590781625,
+      "grad_norm": 6.91807500636854,
+      "learning_rate": 1.7606702794943803e-06,
+      "loss": 0.5193,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6082700249834293,
+      "grad_norm": 16.566079544705957,
+      "learning_rate": 1.756727491683503e-06,
+      "loss": 0.5436,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6087798908886962,
+      "grad_norm": 39.024318471176294,
+      "learning_rate": 1.7527867308273211e-06,
+      "loss": 0.5198,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6092897567939631,
+      "grad_norm": 15.767456247322672,
+      "learning_rate": 1.7488480076725584e-06,
+      "loss": 0.5732,
+      "step": 11950
+    },
+    {
+      "epoch": 0.60979962269923,
+      "grad_norm": 6.138523190826617,
+      "learning_rate": 1.7449113329603787e-06,
+      "loss": 0.5496,
+      "step": 11960
+    },
+    {
+      "epoch": 0.610309488604497,
+      "grad_norm": 36.330040916142245,
+      "learning_rate": 1.7409767174263643e-06,
+      "loss": 0.5342,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6108193545097639,
+      "grad_norm": 12.436489842915702,
+      "learning_rate": 1.7370441718004771e-06,
+      "loss": 0.6011,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6113292204150308,
+      "grad_norm": 5.2678899426017605,
+      "learning_rate": 1.733113706807038e-06,
+      "loss": 0.532,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6118390863202977,
+      "grad_norm": 11.889046920515222,
+      "learning_rate": 1.7291853331646917e-06,
+      "loss": 0.5039,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6123489522255646,
+      "grad_norm": 12.823042645212187,
+      "learning_rate": 1.7252590615863809e-06,
+      "loss": 0.5318,
+      "step": 12010
+    },
+    {
+      "epoch": 0.6128588181308315,
+      "grad_norm": 14.421157906088819,
+      "learning_rate": 1.7213349027793153e-06,
+      "loss": 0.5196,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6133686840360985,
+      "grad_norm": 19.97193610453995,
+      "learning_rate": 1.7174128674449422e-06,
+      "loss": 0.4755,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6138785499413654,
+      "grad_norm": 9.066846582449964,
+      "learning_rate": 1.7134929662789204e-06,
+      "loss": 0.5236,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6143884158466323,
+      "grad_norm": 17.112162242451333,
+      "learning_rate": 1.709575209971085e-06,
+      "loss": 0.467,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6148982817518992,
+      "grad_norm": 17.40291640222636,
+      "learning_rate": 1.7056596092054245e-06,
+      "loss": 0.493,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6154081476571661,
+      "grad_norm": 4.683129228842877,
+      "learning_rate": 1.7017461746600506e-06,
+      "loss": 0.5323,
+      "step": 12070
+    },
+    {
+      "epoch": 0.615918013562433,
+      "grad_norm": 3.7109051124059502,
+      "learning_rate": 1.697834917007163e-06,
+      "loss": 0.5268,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6164278794677,
+      "grad_norm": 9.258106753637072,
+      "learning_rate": 1.6939258469130288e-06,
+      "loss": 0.534,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6169377453729669,
+      "grad_norm": 30.938152269344794,
+      "learning_rate": 1.6900189750379469e-06,
+      "loss": 0.5247,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6174476112782338,
+      "grad_norm": 6.162255674210957,
+      "learning_rate": 1.6861143120362239e-06,
+      "loss": 0.5215,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6179574771835007,
+      "grad_norm": 9.513714545588028,
+      "learning_rate": 1.6822118685561403e-06,
+      "loss": 0.5121,
+      "step": 12120
+    },
+    {
+      "epoch": 0.6184673430887676,
+      "grad_norm": 6.834967431212922,
+      "learning_rate": 1.6783116552399258e-06,
+      "loss": 0.5008,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6189772089940345,
+      "grad_norm": 4.829247413785687,
+      "learning_rate": 1.6744136827237283e-06,
+      "loss": 0.5184,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6194870748993014,
+      "grad_norm": 7.175050633523205,
+      "learning_rate": 1.670517961637582e-06,
+      "loss": 0.4854,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6199969408045684,
+      "grad_norm": 10.676152611521568,
+      "learning_rate": 1.666624502605385e-06,
+      "loss": 0.5476,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6205068067098353,
+      "grad_norm": 19.354648974770278,
+      "learning_rate": 1.6627333162448638e-06,
+      "loss": 0.5143,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6210166726151022,
+      "grad_norm": 9.35966299661283,
+      "learning_rate": 1.6588444131675486e-06,
+      "loss": 0.5187,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6215265385203691,
+      "grad_norm": 8.87291800305617,
+      "learning_rate": 1.6549578039787436e-06,
+      "loss": 0.5525,
+      "step": 12190
+    },
+    {
+      "epoch": 0.622036404425636,
+      "grad_norm": 6.982194117119849,
+      "learning_rate": 1.6510734992774953e-06,
+      "loss": 0.4918,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6225462703309029,
+      "grad_norm": 8.091889155656451,
+      "learning_rate": 1.647191509656567e-06,
+      "loss": 0.4945,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6230561362361698,
+      "grad_norm": 5.905044063744679,
+      "learning_rate": 1.6433118457024094e-06,
+      "loss": 0.5587,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6235660021414368,
+      "grad_norm": 25.530868877288977,
+      "learning_rate": 1.6394345179951293e-06,
+      "loss": 0.4604,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6240758680467037,
+      "grad_norm": 5.695133583269696,
+      "learning_rate": 1.6355595371084627e-06,
+      "loss": 0.4881,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6245857339519706,
+      "grad_norm": 15.30654531084609,
+      "learning_rate": 1.6316869136097467e-06,
+      "loss": 0.5455,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6250955998572375,
+      "grad_norm": 3.749941919122588,
+      "learning_rate": 1.6278166580598897e-06,
+      "loss": 0.4457,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6256054657625044,
+      "grad_norm": 12.003227562589707,
+      "learning_rate": 1.6239487810133404e-06,
+      "loss": 0.5192,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6261153316677713,
+      "grad_norm": 4.426958208572336,
+      "learning_rate": 1.6200832930180643e-06,
+      "loss": 0.4482,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6266251975730383,
+      "grad_norm": 8.717956824567121,
+      "learning_rate": 1.6162202046155085e-06,
+      "loss": 0.5408,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6271350634783052,
+      "grad_norm": 5.105859555737084,
+      "learning_rate": 1.6123595263405783e-06,
+      "loss": 0.5138,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6276449293835721,
+      "grad_norm": 6.392138296699517,
+      "learning_rate": 1.6085012687216078e-06,
+      "loss": 0.4492,
+      "step": 12310
+    },
+    {
+      "epoch": 0.628154795288839,
+      "grad_norm": 18.93794396910775,
+      "learning_rate": 1.6046454422803253e-06,
+      "loss": 0.5681,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6286646611941059,
+      "grad_norm": 6.3885973165769085,
+      "learning_rate": 1.6007920575318334e-06,
+      "loss": 0.5657,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6291745270993728,
+      "grad_norm": 15.232212507178879,
+      "learning_rate": 1.5969411249845737e-06,
+      "loss": 0.4983,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6296843930046397,
+      "grad_norm": 6.38450536379092,
+      "learning_rate": 1.5930926551403025e-06,
+      "loss": 0.4535,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6301942589099067,
+      "grad_norm": 6.810440003124526,
+      "learning_rate": 1.5892466584940574e-06,
+      "loss": 0.5246,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6307041248151736,
+      "grad_norm": 92.33699964113386,
+      "learning_rate": 1.5854031455341332e-06,
+      "loss": 0.4888,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6312139907204405,
+      "grad_norm": 5.952166627327409,
+      "learning_rate": 1.5815621267420526e-06,
+      "loss": 0.4983,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6317238566257074,
+      "grad_norm": 13.330565091112442,
+      "learning_rate": 1.5777236125925333e-06,
+      "loss": 0.4552,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6322337225309743,
+      "grad_norm": 13.431149694108665,
+      "learning_rate": 1.5738876135534659e-06,
+      "loss": 0.5294,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6327435884362412,
+      "grad_norm": 5.8613937499911435,
+      "learning_rate": 1.5700541400858793e-06,
+      "loss": 0.5536,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6332534543415081,
+      "grad_norm": 13.497522029092826,
+      "learning_rate": 1.5662232026439172e-06,
+      "loss": 0.5,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6337633202467751,
+      "grad_norm": 6.317092643414451,
+      "learning_rate": 1.5623948116748074e-06,
+      "loss": 0.4692,
+      "step": 12430
+    },
+    {
+      "epoch": 0.634273186152042,
+      "grad_norm": 10.31916177700157,
+      "learning_rate": 1.5585689776188321e-06,
+      "loss": 0.5556,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6347830520573089,
+      "grad_norm": 7.051415888978958,
+      "learning_rate": 1.5547457109093004e-06,
+      "loss": 0.433,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6352929179625758,
+      "grad_norm": 13.194214008195562,
+      "learning_rate": 1.5509250219725207e-06,
+      "loss": 0.4866,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6358027838678427,
+      "grad_norm": 5.496980733703801,
+      "learning_rate": 1.5471069212277729e-06,
+      "loss": 0.4726,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6363126497731096,
+      "grad_norm": 27.394391903919008,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.5034,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6368225156783766,
+      "grad_norm": 8.800161090855273,
+      "learning_rate": 1.539478525956164e-06,
+      "loss": 0.5441,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6373323815836435,
+      "grad_norm": 12.660866908159553,
+      "learning_rate": 1.5356682522324578e-06,
+      "loss": 0.4604,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6378422474889104,
+      "grad_norm": 14.972027838813567,
+      "learning_rate": 1.5318606083070305e-06,
+      "loss": 0.5573,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6383521133941773,
+      "grad_norm": 8.968890917955061,
+      "learning_rate": 1.5280556045635881e-06,
+      "loss": 0.51,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6388619792994442,
+      "grad_norm": 9.16886223869151,
+      "learning_rate": 1.5242532513786334e-06,
+      "loss": 0.4978,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6393718452047111,
+      "grad_norm": 7.865488171989786,
+      "learning_rate": 1.5204535591214428e-06,
+      "loss": 0.5459,
+      "step": 12540
+    },
+    {
+      "epoch": 0.639881711109978,
+      "grad_norm": 7.322671706067244,
+      "learning_rate": 1.516656538154035e-06,
+      "loss": 0.5672,
+      "step": 12550
+    },
+    {
+      "epoch": 0.640391577015245,
+      "grad_norm": 6.981912098501564,
+      "learning_rate": 1.5128621988311448e-06,
+      "loss": 0.4719,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6409014429205119,
+      "grad_norm": 13.018524518385442,
+      "learning_rate": 1.5090705515001949e-06,
+      "loss": 0.566,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6414113088257788,
+      "grad_norm": 28.22848869993069,
+      "learning_rate": 1.5052816065012635e-06,
+      "loss": 0.4726,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6419211747310457,
+      "grad_norm": 6.130618010338353,
+      "learning_rate": 1.501495374167063e-06,
+      "loss": 0.5574,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6424310406363126,
+      "grad_norm": 3.801489463317786,
+      "learning_rate": 1.497711864822905e-06,
+      "loss": 0.4853,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6429409065415795,
+      "grad_norm": 8.339613236306585,
+      "learning_rate": 1.4939310887866775e-06,
+      "loss": 0.5063,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6434507724468465,
+      "grad_norm": 6.973148186606896,
+      "learning_rate": 1.4901530563688154e-06,
+      "loss": 0.4856,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6439606383521134,
+      "grad_norm": 12.994572347779247,
+      "learning_rate": 1.4863777778722682e-06,
+      "loss": 0.5193,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6444705042573803,
+      "grad_norm": 6.558333340395644,
+      "learning_rate": 1.482605263592478e-06,
+      "loss": 0.4978,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6449803701626472,
+      "grad_norm": 5.141682909872482,
+      "learning_rate": 1.4788355238173473e-06,
+      "loss": 0.5004,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6454902360679141,
+      "grad_norm": 6.522893352262928,
+      "learning_rate": 1.4750685688272143e-06,
+      "loss": 0.5,
+      "step": 12660
+    },
+    {
+      "epoch": 0.646000101973181,
+      "grad_norm": 6.603165179621692,
+      "learning_rate": 1.4713044088948197e-06,
+      "loss": 0.4654,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6465099678784479,
+      "grad_norm": 5.3467446412016315,
+      "learning_rate": 1.4675430542852848e-06,
+      "loss": 0.5501,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6470198337837149,
+      "grad_norm": 19.14659553255697,
+      "learning_rate": 1.4637845152560804e-06,
+      "loss": 0.526,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6475296996889818,
+      "grad_norm": 8.229901563411332,
+      "learning_rate": 1.4600288020569959e-06,
+      "loss": 0.5258,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6480395655942487,
+      "grad_norm": 9.759256471709117,
+      "learning_rate": 1.4562759249301185e-06,
+      "loss": 0.607,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6485494314995156,
+      "grad_norm": 6.935966414378126,
+      "learning_rate": 1.4525258941097985e-06,
+      "loss": 0.5217,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6490592974047825,
+      "grad_norm": 15.54602309816824,
+      "learning_rate": 1.4487787198226244e-06,
+      "loss": 0.5363,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6495691633100494,
+      "grad_norm": 4.928647102455407,
+      "learning_rate": 1.4450344122873985e-06,
+      "loss": 0.4742,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6500790292153164,
+      "grad_norm": 12.495812231264386,
+      "learning_rate": 1.4412929817150982e-06,
+      "loss": 0.5405,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6505888951205833,
+      "grad_norm": 8.425342337716248,
+      "learning_rate": 1.437554438308863e-06,
+      "loss": 0.4444,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6510987610258502,
+      "grad_norm": 9.84721696599807,
+      "learning_rate": 1.4338187922639506e-06,
+      "loss": 0.5144,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6516086269311171,
+      "grad_norm": 5.167726505845119,
+      "learning_rate": 1.430086053767726e-06,
+      "loss": 0.5276,
+      "step": 12780
+    },
+    {
+      "epoch": 0.652118492836384,
+      "grad_norm": 12.668625938673603,
+      "learning_rate": 1.4263562329996194e-06,
+      "loss": 0.505,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6526283587416509,
+      "grad_norm": 15.763800394384761,
+      "learning_rate": 1.422629340131106e-06,
+      "loss": 0.5763,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6531382246469178,
+      "grad_norm": 8.02198435905354,
+      "learning_rate": 1.4189053853256757e-06,
+      "loss": 0.5103,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6536480905521848,
+      "grad_norm": 11.950637204603689,
+      "learning_rate": 1.4151843787388062e-06,
+      "loss": 0.5635,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6541579564574517,
+      "grad_norm": 8.293082362914374,
+      "learning_rate": 1.4114663305179382e-06,
+      "loss": 0.5072,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6546678223627186,
+      "grad_norm": 4.3737022288493765,
+      "learning_rate": 1.4077512508024382e-06,
+      "loss": 0.4543,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6551776882679855,
+      "grad_norm": 18.10698828602165,
+      "learning_rate": 1.4040391497235845e-06,
+      "loss": 0.5085,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6556875541732524,
+      "grad_norm": 7.07731221377441,
+      "learning_rate": 1.4003300374045283e-06,
+      "loss": 0.4776,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6561974200785193,
+      "grad_norm": 4.675481069069686,
+      "learning_rate": 1.396623923960271e-06,
+      "loss": 0.466,
+      "step": 12870
+    },
+    {
+      "epoch": 0.6567072859837862,
+      "grad_norm": 6.033379297093717,
+      "learning_rate": 1.3929208194976362e-06,
+      "loss": 0.5344,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6572171518890532,
+      "grad_norm": 8.628525763507344,
+      "learning_rate": 1.3892207341152416e-06,
+      "loss": 0.5409,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6577270177943201,
+      "grad_norm": 5.409542842182235,
+      "learning_rate": 1.385523677903472e-06,
+      "loss": 0.4883,
+      "step": 12900
+    },
+    {
+      "epoch": 0.658236883699587,
+      "grad_norm": 11.783825873969915,
+      "learning_rate": 1.38182966094445e-06,
+      "loss": 0.5546,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6587467496048539,
+      "grad_norm": 14.457282758726171,
+      "learning_rate": 1.3781386933120133e-06,
+      "loss": 0.4842,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6592566155101208,
+      "grad_norm": 7.7607092167340355,
+      "learning_rate": 1.3744507850716804e-06,
+      "loss": 0.5051,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6597664814153877,
+      "grad_norm": 7.137912593306867,
+      "learning_rate": 1.3707659462806284e-06,
+      "loss": 0.4834,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6602763473206547,
+      "grad_norm": 6.143705561272351,
+      "learning_rate": 1.367084186987663e-06,
+      "loss": 0.5108,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6607862132259216,
+      "grad_norm": 8.731079626973258,
+      "learning_rate": 1.3634055172331926e-06,
+      "loss": 0.4852,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6612960791311885,
+      "grad_norm": 5.371437695099816,
+      "learning_rate": 1.3597299470491986e-06,
+      "loss": 0.5675,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6618059450364554,
+      "grad_norm": 7.673890908642838,
+      "learning_rate": 1.356057486459214e-06,
+      "loss": 0.5427,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6623158109417223,
+      "grad_norm": 5.580626395226384,
+      "learning_rate": 1.352388145478285e-06,
+      "loss": 0.5475,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6628256768469892,
+      "grad_norm": 14.001010484519423,
+      "learning_rate": 1.3487219341129566e-06,
+      "loss": 0.4702,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6633355427522561,
+      "grad_norm": 11.362626293229546,
+      "learning_rate": 1.3450588623612353e-06,
+      "loss": 0.5142,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6638454086575231,
+      "grad_norm": 4.961569081167888,
+      "learning_rate": 1.3413989402125682e-06,
+      "loss": 0.5187,
+      "step": 13020
+    },
+    {
+      "epoch": 0.66435527456279,
+      "grad_norm": 4.915951893745529,
+      "learning_rate": 1.3377421776478111e-06,
+      "loss": 0.5272,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6648651404680569,
+      "grad_norm": 5.045332996976456,
+      "learning_rate": 1.3340885846392032e-06,
+      "loss": 0.5923,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6653750063733238,
+      "grad_norm": 5.509953458864324,
+      "learning_rate": 1.3304381711503444e-06,
+      "loss": 0.49,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6658848722785907,
+      "grad_norm": 7.9456633485728165,
+      "learning_rate": 1.3267909471361574e-06,
+      "loss": 0.5562,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6663947381838576,
+      "grad_norm": 24.233820150621078,
+      "learning_rate": 1.3231469225428726e-06,
+      "loss": 0.5177,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6669046040891246,
+      "grad_norm": 7.9754418427625255,
+      "learning_rate": 1.3195061073079901e-06,
+      "loss": 0.5494,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6674144699943915,
+      "grad_norm": 7.739534240779945,
+      "learning_rate": 1.3158685113602636e-06,
+      "loss": 0.4843,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6679243358996584,
+      "grad_norm": 14.361556260016403,
+      "learning_rate": 1.3122341446196636e-06,
+      "loss": 0.5515,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6684342018049253,
+      "grad_norm": 4.830484020015252,
+      "learning_rate": 1.3086030169973552e-06,
+      "loss": 0.5049,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6689440677101922,
+      "grad_norm": 11.617407710197574,
+      "learning_rate": 1.3049751383956707e-06,
+      "loss": 0.4704,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6694539336154591,
+      "grad_norm": 20.61436121006635,
+      "learning_rate": 1.301350518708081e-06,
+      "loss": 0.5146,
+      "step": 13130
+    },
+    {
+      "epoch": 0.669963799520726,
+      "grad_norm": 22.54013582179851,
+      "learning_rate": 1.2977291678191733e-06,
+      "loss": 0.4731,
+      "step": 13140
+    },
+    {
+      "epoch": 0.670473665425993,
+      "grad_norm": 5.086290366289214,
+      "learning_rate": 1.2941110956046142e-06,
+      "loss": 0.5069,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6709835313312599,
+      "grad_norm": 7.472775727129846,
+      "learning_rate": 1.290496311931135e-06,
+      "loss": 0.5016,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6714933972365268,
+      "grad_norm": 9.605057298095543,
+      "learning_rate": 1.2868848266564964e-06,
+      "loss": 0.4805,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6720032631417937,
+      "grad_norm": 24.030989984043877,
+      "learning_rate": 1.2832766496294647e-06,
+      "loss": 0.4861,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6725131290470606,
+      "grad_norm": 7.688398644867407,
+      "learning_rate": 1.2796717906897831e-06,
+      "loss": 0.5021,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6730229949523275,
+      "grad_norm": 8.08782204632997,
+      "learning_rate": 1.2760702596681478e-06,
+      "loss": 0.5308,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6735328608575945,
+      "grad_norm": 6.724965448279945,
+      "learning_rate": 1.2724720663861792e-06,
+      "loss": 0.4748,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6740427267628614,
+      "grad_norm": 15.869715866274502,
+      "learning_rate": 1.2688772206563938e-06,
+      "loss": 0.5366,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6745525926681283,
+      "grad_norm": 17.33308634891779,
+      "learning_rate": 1.2652857322821821e-06,
+      "loss": 0.4798,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6750624585733952,
+      "grad_norm": 6.413858029938686,
+      "learning_rate": 1.2616976110577766e-06,
+      "loss": 0.5286,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6755723244786621,
+      "grad_norm": 27.500826205776104,
+      "learning_rate": 1.2581128667682286e-06,
+      "loss": 0.525,
+      "step": 13250
+    },
+    {
+      "epoch": 0.676082190383929,
+      "grad_norm": 4.587601999321133,
+      "learning_rate": 1.2545315091893784e-06,
+      "loss": 0.4952,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6765920562891959,
+      "grad_norm": 8.682141735935351,
+      "learning_rate": 1.2509535480878325e-06,
+      "loss": 0.4982,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6771019221944629,
+      "grad_norm": 5.875702610092279,
+      "learning_rate": 1.2473789932209331e-06,
+      "loss": 0.4626,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6776117880997298,
+      "grad_norm": 11.57688664010269,
+      "learning_rate": 1.2438078543367377e-06,
+      "loss": 0.5797,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6781216540049967,
+      "grad_norm": 7.625967931326347,
+      "learning_rate": 1.2402401411739806e-06,
+      "loss": 0.5266,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6786315199102636,
+      "grad_norm": 9.280506265403062,
+      "learning_rate": 1.2366758634620615e-06,
+      "loss": 0.5461,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6791413858155305,
+      "grad_norm": 22.898801072596648,
+      "learning_rate": 1.2331150309210075e-06,
+      "loss": 0.5171,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6796512517207974,
+      "grad_norm": 2.926333613028221,
+      "learning_rate": 1.2295576532614506e-06,
+      "loss": 0.4778,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6801611176260643,
+      "grad_norm": 9.094467060554189,
+      "learning_rate": 1.226003740184602e-06,
+      "loss": 0.4894,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6806709835313313,
+      "grad_norm": 8.219844480956706,
+      "learning_rate": 1.2224533013822237e-06,
+      "loss": 0.5202,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6811808494365982,
+      "grad_norm": 10.280263133918698,
+      "learning_rate": 1.2189063465366064e-06,
+      "loss": 0.6116,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6816907153418651,
+      "grad_norm": 12.133501298347303,
+      "learning_rate": 1.2153628853205336e-06,
+      "loss": 0.5026,
+      "step": 13370
+    },
+    {
+      "epoch": 0.682200581247132,
+      "grad_norm": 7.117093334590749,
+      "learning_rate": 1.2118229273972684e-06,
+      "loss": 0.5053,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6827104471523989,
+      "grad_norm": 9.511696988963454,
+      "learning_rate": 1.2082864824205138e-06,
+      "loss": 0.4878,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6832203130576658,
+      "grad_norm": 13.917897942567095,
+      "learning_rate": 1.2047535600343984e-06,
+      "loss": 0.4947,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6837301789629328,
+      "grad_norm": 9.625477395299297,
+      "learning_rate": 1.2012241698734408e-06,
+      "loss": 0.5615,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6842400448681997,
+      "grad_norm": 9.318591854910585,
+      "learning_rate": 1.1976983215625285e-06,
+      "loss": 0.5031,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6847499107734666,
+      "grad_norm": 10.712263091770266,
+      "learning_rate": 1.1941760247168893e-06,
+      "loss": 0.5237,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6852597766787335,
+      "grad_norm": 4.196412737538298,
+      "learning_rate": 1.1906572889420655e-06,
+      "loss": 0.4728,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6857696425840004,
+      "grad_norm": 7.784431942907173,
+      "learning_rate": 1.1871421238338917e-06,
+      "loss": 0.5148,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6862795084892673,
+      "grad_norm": 6.63260049619592,
+      "learning_rate": 1.1836305389784588e-06,
+      "loss": 0.5399,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6867893743945342,
+      "grad_norm": 4.469197479198041,
+      "learning_rate": 1.1801225439521003e-06,
+      "loss": 0.4436,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6872992402998012,
+      "grad_norm": 14.254868714509106,
+      "learning_rate": 1.176618148321356e-06,
+      "loss": 0.4758,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6878091062050681,
+      "grad_norm": 3.7814670801049695,
+      "learning_rate": 1.1731173616429514e-06,
+      "loss": 0.4495,
+      "step": 13490
+    },
+    {
+      "epoch": 0.688318972110335,
+      "grad_norm": 5.161779138546176,
+      "learning_rate": 1.16962019346377e-06,
+      "loss": 0.4625,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6888288380156019,
+      "grad_norm": 10.993559836072276,
+      "learning_rate": 1.1661266533208274e-06,
+      "loss": 0.5132,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6893387039208688,
+      "grad_norm": 10.908435267498524,
+      "learning_rate": 1.1626367507412443e-06,
+      "loss": 0.4824,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6898485698261357,
+      "grad_norm": 6.856450877530777,
+      "learning_rate": 1.1591504952422243e-06,
+      "loss": 0.5084,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6903584357314027,
+      "grad_norm": 4.982024981770672,
+      "learning_rate": 1.1556678963310222e-06,
+      "loss": 0.5541,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6908683016366696,
+      "grad_norm": 4.906060311571326,
+      "learning_rate": 1.152188963504922e-06,
+      "loss": 0.4915,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6913781675419365,
+      "grad_norm": 29.493906037211964,
+      "learning_rate": 1.148713706251211e-06,
+      "loss": 0.4951,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6918880334472034,
+      "grad_norm": 8.97699120582468,
+      "learning_rate": 1.1452421340471514e-06,
+      "loss": 0.5329,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6923978993524703,
+      "grad_norm": 5.699186471774075,
+      "learning_rate": 1.1417742563599568e-06,
+      "loss": 0.5537,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6929077652577372,
+      "grad_norm": 6.4448339310820435,
+      "learning_rate": 1.1383100826467653e-06,
+      "loss": 0.5556,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6934176311630041,
+      "grad_norm": 4.7807112185144565,
+      "learning_rate": 1.1348496223546162e-06,
+      "loss": 0.498,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6939274970682711,
+      "grad_norm": 17.810465336560668,
+      "learning_rate": 1.131392884920417e-06,
+      "loss": 0.4422,
+      "step": 13610
+    },
+    {
+      "epoch": 0.694437362973538,
+      "grad_norm": 119.33725015633846,
+      "learning_rate": 1.1279398797709293e-06,
+      "loss": 0.5334,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6949472288788049,
+      "grad_norm": 8.000152099817969,
+      "learning_rate": 1.1244906163227295e-06,
+      "loss": 0.5405,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6954570947840718,
+      "grad_norm": 5.822401607695464,
+      "learning_rate": 1.1210451039821965e-06,
+      "loss": 0.5167,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6959669606893387,
+      "grad_norm": 11.163768611048248,
+      "learning_rate": 1.1176033521454758e-06,
+      "loss": 0.4986,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6964768265946056,
+      "grad_norm": 6.217197235375367,
+      "learning_rate": 1.114165370198458e-06,
+      "loss": 0.5233,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6969866924998726,
+      "grad_norm": 10.532638730674618,
+      "learning_rate": 1.1107311675167558e-06,
+      "loss": 0.5622,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6974965584051395,
+      "grad_norm": 9.982847064316687,
+      "learning_rate": 1.1073007534656712e-06,
+      "loss": 0.5113,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6980064243104064,
+      "grad_norm": 11.286075006858832,
+      "learning_rate": 1.1038741374001793e-06,
+      "loss": 0.508,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6985162902156733,
+      "grad_norm": 13.847965895609494,
+      "learning_rate": 1.1004513286648922e-06,
+      "loss": 0.4808,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6990261561209402,
+      "grad_norm": 3.750806517115084,
+      "learning_rate": 1.0970323365940443e-06,
+      "loss": 0.4999,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6995360220262071,
+      "grad_norm": 4.49734953510548,
+      "learning_rate": 1.093617170511459e-06,
+      "loss": 0.5244,
+      "step": 13720
+    },
+    {
+      "epoch": 0.700045887931474,
+      "grad_norm": 3.5635278395876604,
+      "learning_rate": 1.0902058397305268e-06,
+      "loss": 0.5034,
+      "step": 13730
+    },
+    {
+      "epoch": 0.700555753836741,
+      "grad_norm": 7.630693193745469,
+      "learning_rate": 1.0867983535541785e-06,
+      "loss": 0.5485,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7010656197420079,
+      "grad_norm": 20.88551938761256,
+      "learning_rate": 1.0833947212748597e-06,
+      "loss": 0.5409,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7015754856472748,
+      "grad_norm": 20.23048041237047,
+      "learning_rate": 1.07999495217451e-06,
+      "loss": 0.5645,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7020853515525417,
+      "grad_norm": 4.762845462769943,
+      "learning_rate": 1.0765990555245275e-06,
+      "loss": 0.5196,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7025952174578086,
+      "grad_norm": 5.618109909368287,
+      "learning_rate": 1.0732070405857562e-06,
+      "loss": 0.5406,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7031050833630755,
+      "grad_norm": 10.640295482886632,
+      "learning_rate": 1.0698189166084501e-06,
+      "loss": 0.4254,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7036149492683424,
+      "grad_norm": 4.911022067208778,
+      "learning_rate": 1.0664346928322547e-06,
+      "loss": 0.5113,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7041248151736094,
+      "grad_norm": 9.081453850201648,
+      "learning_rate": 1.063054378486178e-06,
+      "loss": 0.5947,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7046346810788763,
+      "grad_norm": 6.77043436264162,
+      "learning_rate": 1.059677982788567e-06,
+      "loss": 0.4636,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7051445469841432,
+      "grad_norm": 7.5449704839316425,
+      "learning_rate": 1.056305514947082e-06,
+      "loss": 0.4751,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7056544128894101,
+      "grad_norm": 11.690143819464407,
+      "learning_rate": 1.0529369841586743e-06,
+      "loss": 0.4504,
+      "step": 13840
+    },
+    {
+      "epoch": 0.706164278794677,
+      "grad_norm": 4.31175777202485,
+      "learning_rate": 1.0495723996095533e-06,
+      "loss": 0.4495,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7066741446999439,
+      "grad_norm": 13.202730632004345,
+      "learning_rate": 1.046211770475173e-06,
+      "loss": 0.4702,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7071840106052109,
+      "grad_norm": 7.249743343482479,
+      "learning_rate": 1.0428551059201964e-06,
+      "loss": 0.4651,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7076938765104778,
+      "grad_norm": 7.318651462556601,
+      "learning_rate": 1.039502415098476e-06,
+      "loss": 0.5248,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7082037424157447,
+      "grad_norm": 8.280309990686213,
+      "learning_rate": 1.0361537071530277e-06,
+      "loss": 0.5498,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7087136083210116,
+      "grad_norm": 7.704516862067239,
+      "learning_rate": 1.0328089912160055e-06,
+      "loss": 0.5427,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7092234742262785,
+      "grad_norm": 5.1790978507620276,
+      "learning_rate": 1.0294682764086794e-06,
+      "loss": 0.5272,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7097333401315454,
+      "grad_norm": 6.6383138164611974,
+      "learning_rate": 1.0261315718414028e-06,
+      "loss": 0.5172,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7102432060368123,
+      "grad_norm": 10.878766001476842,
+      "learning_rate": 1.0227988866135995e-06,
+      "loss": 0.4949,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7107530719420793,
+      "grad_norm": 10.28873182386919,
+      "learning_rate": 1.0194702298137251e-06,
+      "loss": 0.4777,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7112629378473462,
+      "grad_norm": 6.691725960477623,
+      "learning_rate": 1.016145610519256e-06,
+      "loss": 0.5215,
+      "step": 13950
+    },
+    {
+      "epoch": 0.7117728037526131,
+      "grad_norm": 8.8432699489868,
+      "learning_rate": 1.0128250377966545e-06,
+      "loss": 0.5249,
+      "step": 13960
+    },
+    {
+      "epoch": 0.71228266965788,
+      "grad_norm": 6.205838612221995,
+      "learning_rate": 1.009508520701347e-06,
+      "loss": 0.4928,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7127925355631469,
+      "grad_norm": 7.022108699926237,
+      "learning_rate": 1.006196068277704e-06,
+      "loss": 0.4703,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7133024014684138,
+      "grad_norm": 4.790738966564985,
+      "learning_rate": 1.002887689559005e-06,
+      "loss": 0.4899,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7138122673736808,
+      "grad_norm": 7.20639824109062,
+      "learning_rate": 9.99583393567428e-07,
+      "loss": 0.4402,
+      "step": 14000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 19613,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0829919385288704e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}