diff --git "a/checkpoint-15581/trainer_state.json" "b/checkpoint-15581/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-15581/trainer_state.json"
@@ -0,0 +1,10940 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 15581,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000641843374170618,
+      "grad_norm": 2.778578758239746,
+      "learning_rate": 3.6e-07,
+      "loss": 2.0918,
+      "step": 10
+    },
+    {
+      "epoch": 0.001283686748341236,
+      "grad_norm": 3.490725517272949,
+      "learning_rate": 7.6e-07,
+      "loss": 2.0914,
+      "step": 20
+    },
+    {
+      "epoch": 0.0019255301225118541,
+      "grad_norm": 2.9524786472320557,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 2.0855,
+      "step": 30
+    },
+    {
+      "epoch": 0.002567373496682472,
+      "grad_norm": 2.4857563972473145,
+      "learning_rate": 1.56e-06,
+      "loss": 2.0858,
+      "step": 40
+    },
+    {
+      "epoch": 0.00320921687085309,
+      "grad_norm": 4.684011459350586,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 2.0894,
+      "step": 50
+    },
+    {
+      "epoch": 0.0038510602450237083,
+      "grad_norm": 2.8892273902893066,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 2.0796,
+      "step": 60
+    },
+    {
+      "epoch": 0.004492903619194326,
+      "grad_norm": 2.8607919216156006,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 2.0779,
+      "step": 70
+    },
+    {
+      "epoch": 0.005134746993364944,
+      "grad_norm": 3.9530575275421143,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 2.0703,
+      "step": 80
+    },
+    {
+      "epoch": 0.005776590367535562,
+      "grad_norm": 2.9849867820739746,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 2.0624,
+      "step": 90
+    },
+    {
+      "epoch": 0.00641843374170618,
+      "grad_norm": 2.684105634689331,
+      "learning_rate": 3.96e-06,
+      "loss": 2.0596,
+      "step": 100
+    },
+    {
+      "epoch": 0.007060277115876798,
+      "grad_norm": 3.2845048904418945,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 2.0403,
+      "step": 110
+    },
+    {
+      "epoch": 0.0077021204900474165,
+      "grad_norm": 3.709146022796631,
+      "learning_rate": 4.76e-06,
+      "loss": 1.9946,
+      "step": 120
+    },
+    {
+      "epoch": 0.008343963864218035,
+      "grad_norm": 5.354912757873535,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 1.9471,
+      "step": 130
+    },
+    {
+      "epoch": 0.008985807238388653,
+      "grad_norm": 5.101510524749756,
+      "learning_rate": 5.560000000000001e-06,
+      "loss": 1.8538,
+      "step": 140
+    },
+    {
+      "epoch": 0.00962765061255927,
+      "grad_norm": Infinity,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 1.7965,
+      "step": 150
+    },
+    {
+      "epoch": 0.010269493986729888,
+      "grad_norm": 9.85523509979248,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 1.7491,
+      "step": 160
+    },
+    {
+      "epoch": 0.010911337360900506,
+      "grad_norm": 13.500116348266602,
+      "learning_rate": 6.760000000000001e-06,
+      "loss": 1.6258,
+      "step": 170
+    },
+    {
+      "epoch": 0.011553180735071123,
+      "grad_norm": 12.905332565307617,
+      "learning_rate": 7.16e-06,
+      "loss": 1.6156,
+      "step": 180
+    },
+    {
+      "epoch": 0.012195024109241743,
+      "grad_norm": 5.522826194763184,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 1.6384,
+      "step": 190
+    },
+    {
+      "epoch": 0.01283686748341236,
+      "grad_norm": 9.346458435058594,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 1.5583,
+      "step": 200
+    },
+    {
+      "epoch": 0.013478710857582978,
+      "grad_norm": 10.575640678405762,
+      "learning_rate": 8.36e-06,
+      "loss": 1.4437,
+      "step": 210
+    },
+    {
+      "epoch": 0.014120554231753596,
+      "grad_norm": 20.125003814697266,
+      "learning_rate": 8.76e-06,
+      "loss": 1.4157,
+      "step": 220
+    },
+    {
+      "epoch": 0.014762397605924214,
+      "grad_norm": 12.119187355041504,
+      "learning_rate": 9.16e-06,
+      "loss": 1.3255,
+      "step": 230
+    },
+    {
+      "epoch": 0.015404240980094833,
+      "grad_norm": 15.212540626525879,
+      "learning_rate": 9.56e-06,
+      "loss": 1.2985,
+      "step": 240
+    },
+    {
+      "epoch": 0.01604608435426545,
+      "grad_norm": 13.789387702941895,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 1.1824,
+      "step": 250
+    },
+    {
+      "epoch": 0.01668792772843607,
+      "grad_norm": 17.67810821533203,
+      "learning_rate": 1.036e-05,
+      "loss": 1.1609,
+      "step": 260
+    },
+    {
+      "epoch": 0.017329771102606688,
+      "grad_norm": 13.683328628540039,
+      "learning_rate": 1.0760000000000002e-05,
+      "loss": 1.0169,
+      "step": 270
+    },
+    {
+      "epoch": 0.017971614476777305,
+      "grad_norm": 13.746779441833496,
+      "learning_rate": 1.1160000000000002e-05,
+      "loss": 0.9953,
+      "step": 280
+    },
+    {
+      "epoch": 0.018613457850947923,
+      "grad_norm": 17.402944564819336,
+      "learning_rate": 1.156e-05,
+      "loss": 0.8542,
+      "step": 290
+    },
+    {
+      "epoch": 0.01925530122511854,
+      "grad_norm": 18.10095977783203,
+      "learning_rate": 1.196e-05,
+      "loss": 0.807,
+      "step": 300
+    },
+    {
+      "epoch": 0.01989714459928916,
+      "grad_norm": 17.38810920715332,
+      "learning_rate": 1.236e-05,
+      "loss": 0.7492,
+      "step": 310
+    },
+    {
+      "epoch": 0.020538987973459776,
+      "grad_norm": 16.033836364746094,
+      "learning_rate": 1.2760000000000001e-05,
+      "loss": 0.7984,
+      "step": 320
+    },
+    {
+      "epoch": 0.021180831347630394,
+      "grad_norm": 18.93714141845703,
+      "learning_rate": 1.3160000000000001e-05,
+      "loss": 0.632,
+      "step": 330
+    },
+    {
+      "epoch": 0.02182267472180101,
+      "grad_norm": 17.007225036621094,
+      "learning_rate": 1.3560000000000002e-05,
+      "loss": 0.6511,
+      "step": 340
+    },
+    {
+      "epoch": 0.02246451809597163,
+      "grad_norm": 15.338671684265137,
+      "learning_rate": 1.396e-05,
+      "loss": 0.6407,
+      "step": 350
+    },
+    {
+      "epoch": 0.023106361470142247,
+      "grad_norm": 16.871875762939453,
+      "learning_rate": 1.4360000000000001e-05,
+      "loss": 0.5841,
+      "step": 360
+    },
+    {
+      "epoch": 0.023748204844312868,
+      "grad_norm": 12.31517505645752,
+      "learning_rate": 1.4760000000000001e-05,
+      "loss": 0.5787,
+      "step": 370
+    },
+    {
+      "epoch": 0.024390048218483486,
+      "grad_norm": 12.55438232421875,
+      "learning_rate": 1.516e-05,
+      "loss": 0.4979,
+      "step": 380
+    },
+    {
+      "epoch": 0.025031891592654103,
+      "grad_norm": 11.683804512023926,
+      "learning_rate": 1.556e-05,
+      "loss": 0.409,
+      "step": 390
+    },
+    {
+      "epoch": 0.02567373496682472,
+      "grad_norm": 13.466601371765137,
+      "learning_rate": 1.5960000000000003e-05,
+      "loss": 0.4344,
+      "step": 400
+    },
+    {
+      "epoch": 0.02631557834099534,
+      "grad_norm": 13.933389663696289,
+      "learning_rate": 1.636e-05,
+      "loss": 0.4355,
+      "step": 410
+    },
+    {
+      "epoch": 0.026957421715165956,
+      "grad_norm": 14.962430953979492,
+      "learning_rate": 1.6760000000000002e-05,
+      "loss": 0.4358,
+      "step": 420
+    },
+    {
+      "epoch": 0.027599265089336574,
+      "grad_norm": 11.194101333618164,
+      "learning_rate": 1.7160000000000002e-05,
+      "loss": 0.3418,
+      "step": 430
+    },
+    {
+      "epoch": 0.028241108463507192,
+      "grad_norm": 10.003135681152344,
+      "learning_rate": 1.756e-05,
+      "loss": 0.407,
+      "step": 440
+    },
+    {
+      "epoch": 0.02888295183767781,
+      "grad_norm": 13.722594261169434,
+      "learning_rate": 1.796e-05,
+      "loss": 0.3784,
+      "step": 450
+    },
+    {
+      "epoch": 0.029524795211848427,
+      "grad_norm": 11.653929710388184,
+      "learning_rate": 1.8360000000000004e-05,
+      "loss": 0.3438,
+      "step": 460
+    },
+    {
+      "epoch": 0.03016663858601905,
+      "grad_norm": 10.154895782470703,
+      "learning_rate": 1.876e-05,
+      "loss": 0.3102,
+      "step": 470
+    },
+    {
+      "epoch": 0.030808481960189666,
+      "grad_norm": 13.267715454101562,
+      "learning_rate": 1.916e-05,
+      "loss": 0.3173,
+      "step": 480
+    },
+    {
+      "epoch": 0.031450325334360284,
+      "grad_norm": 9.964168548583984,
+      "learning_rate": 1.9560000000000002e-05,
+      "loss": 0.3096,
+      "step": 490
+    },
+    {
+      "epoch": 0.0320921687085309,
+      "grad_norm": 13.796014785766602,
+      "learning_rate": 1.9960000000000002e-05,
+      "loss": 0.345,
+      "step": 500
+    },
+    {
+      "epoch": 0.03273401208270152,
+      "grad_norm": 10.034880638122559,
+      "learning_rate": 1.998806445195942e-05,
+      "loss": 0.3417,
+      "step": 510
+    },
+    {
+      "epoch": 0.03337585545687214,
+      "grad_norm": 9.3333740234375,
+      "learning_rate": 1.9974802731914332e-05,
+      "loss": 0.2927,
+      "step": 520
+    },
+    {
+      "epoch": 0.034017698831042754,
+      "grad_norm": 9.258805274963379,
+      "learning_rate": 1.996154101186924e-05,
+      "loss": 0.2624,
+      "step": 530
+    },
+    {
+      "epoch": 0.034659542205213376,
+      "grad_norm": 6.755455017089844,
+      "learning_rate": 1.994827929182415e-05,
+      "loss": 0.2602,
+      "step": 540
+    },
+    {
+      "epoch": 0.03530138557938399,
+      "grad_norm": 14.036195755004883,
+      "learning_rate": 1.9935017571779062e-05,
+      "loss": 0.3068,
+      "step": 550
+    },
+    {
+      "epoch": 0.03594322895355461,
+      "grad_norm": 7.500053405761719,
+      "learning_rate": 1.992175585173397e-05,
+      "loss": 0.2405,
+      "step": 560
+    },
+    {
+      "epoch": 0.036585072327725225,
+      "grad_norm": 10.683878898620605,
+      "learning_rate": 1.990849413168888e-05,
+      "loss": 0.2651,
+      "step": 570
+    },
+    {
+      "epoch": 0.037226915701895846,
+      "grad_norm": 9.001018524169922,
+      "learning_rate": 1.9895232411643792e-05,
+      "loss": 0.3062,
+      "step": 580
+    },
+    {
+      "epoch": 0.03786875907606646,
+      "grad_norm": 7.354406833648682,
+      "learning_rate": 1.9881970691598703e-05,
+      "loss": 0.2741,
+      "step": 590
+    },
+    {
+      "epoch": 0.03851060245023708,
+      "grad_norm": 7.920845031738281,
+      "learning_rate": 1.986870897155361e-05,
+      "loss": 0.2483,
+      "step": 600
+    },
+    {
+      "epoch": 0.039152445824407696,
+      "grad_norm": 11.448545455932617,
+      "learning_rate": 1.9855447251508522e-05,
+      "loss": 0.1963,
+      "step": 610
+    },
+    {
+      "epoch": 0.03979428919857832,
+      "grad_norm": 8.437763214111328,
+      "learning_rate": 1.984218553146343e-05,
+      "loss": 0.2262,
+      "step": 620
+    },
+    {
+      "epoch": 0.04043613257274894,
+      "grad_norm": 9.328960418701172,
+      "learning_rate": 1.9828923811418345e-05,
+      "loss": 0.1793,
+      "step": 630
+    },
+    {
+      "epoch": 0.04107797594691955,
+      "grad_norm": 7.899652481079102,
+      "learning_rate": 1.9815662091373252e-05,
+      "loss": 0.2089,
+      "step": 640
+    },
+    {
+      "epoch": 0.041719819321090174,
+      "grad_norm": 8.917160987854004,
+      "learning_rate": 1.9802400371328164e-05,
+      "loss": 0.2036,
+      "step": 650
+    },
+    {
+      "epoch": 0.04236166269526079,
+      "grad_norm": 8.51976490020752,
+      "learning_rate": 1.978913865128307e-05,
+      "loss": 0.2127,
+      "step": 660
+    },
+    {
+      "epoch": 0.04300350606943141,
+      "grad_norm": 8.080279350280762,
+      "learning_rate": 1.9775876931237982e-05,
+      "loss": 0.2343,
+      "step": 670
+    },
+    {
+      "epoch": 0.04364534944360202,
+      "grad_norm": 8.942863464355469,
+      "learning_rate": 1.9762615211192894e-05,
+      "loss": 0.1784,
+      "step": 680
+    },
+    {
+      "epoch": 0.044287192817772644,
+      "grad_norm": 8.20798397064209,
+      "learning_rate": 1.97493534911478e-05,
+      "loss": 0.149,
+      "step": 690
+    },
+    {
+      "epoch": 0.04492903619194326,
+      "grad_norm": 7.868164539337158,
+      "learning_rate": 1.9736091771102716e-05,
+      "loss": 0.1537,
+      "step": 700
+    },
+    {
+      "epoch": 0.04557087956611388,
+      "grad_norm": 8.475685119628906,
+      "learning_rate": 1.9722830051057624e-05,
+      "loss": 0.1234,
+      "step": 710
+    },
+    {
+      "epoch": 0.046212722940284494,
+      "grad_norm": 13.600449562072754,
+      "learning_rate": 1.9709568331012535e-05,
+      "loss": 0.1871,
+      "step": 720
+    },
+    {
+      "epoch": 0.046854566314455115,
+      "grad_norm": 7.583765029907227,
+      "learning_rate": 1.9696306610967443e-05,
+      "loss": 0.1772,
+      "step": 730
+    },
+    {
+      "epoch": 0.047496409688625736,
+      "grad_norm": 7.121690273284912,
+      "learning_rate": 1.9683044890922354e-05,
+      "loss": 0.196,
+      "step": 740
+    },
+    {
+      "epoch": 0.04813825306279635,
+      "grad_norm": 7.272078514099121,
+      "learning_rate": 1.9669783170877265e-05,
+      "loss": 0.1647,
+      "step": 750
+    },
+    {
+      "epoch": 0.04878009643696697,
+      "grad_norm": 8.071809768676758,
+      "learning_rate": 1.9656521450832176e-05,
+      "loss": 0.1741,
+      "step": 760
+    },
+    {
+      "epoch": 0.049421939811137586,
+      "grad_norm": 8.809086799621582,
+      "learning_rate": 1.9643259730787084e-05,
+      "loss": 0.1501,
+      "step": 770
+    },
+    {
+      "epoch": 0.05006378318530821,
+      "grad_norm": 7.277254581451416,
+      "learning_rate": 1.9629998010741995e-05,
+      "loss": 0.1812,
+      "step": 780
+    },
+    {
+      "epoch": 0.05070562655947882,
+      "grad_norm": 6.261189937591553,
+      "learning_rate": 1.9616736290696906e-05,
+      "loss": 0.1599,
+      "step": 790
+    },
+    {
+      "epoch": 0.05134746993364944,
+      "grad_norm": 5.358989715576172,
+      "learning_rate": 1.9603474570651814e-05,
+      "loss": 0.1563,
+      "step": 800
+    },
+    {
+      "epoch": 0.051989313307820056,
+      "grad_norm": 5.942981719970703,
+      "learning_rate": 1.9590212850606725e-05,
+      "loss": 0.1131,
+      "step": 810
+    },
+    {
+      "epoch": 0.05263115668199068,
+      "grad_norm": 11.299034118652344,
+      "learning_rate": 1.9576951130561633e-05,
+      "loss": 0.1872,
+      "step": 820
+    },
+    {
+      "epoch": 0.05327300005616129,
+      "grad_norm": 3.3547017574310303,
+      "learning_rate": 1.9563689410516548e-05,
+      "loss": 0.1696,
+      "step": 830
+    },
+    {
+      "epoch": 0.05391484343033191,
+      "grad_norm": 7.165664196014404,
+      "learning_rate": 1.9550427690471455e-05,
+      "loss": 0.1668,
+      "step": 840
+    },
+    {
+      "epoch": 0.054556686804502534,
+      "grad_norm": 9.479928970336914,
+      "learning_rate": 1.9537165970426367e-05,
+      "loss": 0.1488,
+      "step": 850
+    },
+    {
+      "epoch": 0.05519853017867315,
+      "grad_norm": 5.311110019683838,
+      "learning_rate": 1.9523904250381274e-05,
+      "loss": 0.1503,
+      "step": 860
+    },
+    {
+      "epoch": 0.05584037355284377,
+      "grad_norm": 9.265485763549805,
+      "learning_rate": 1.9510642530336186e-05,
+      "loss": 0.1382,
+      "step": 870
+    },
+    {
+      "epoch": 0.056482216927014384,
+      "grad_norm": 11.654985427856445,
+      "learning_rate": 1.9497380810291097e-05,
+      "loss": 0.1764,
+      "step": 880
+    },
+    {
+      "epoch": 0.057124060301185005,
+      "grad_norm": 5.929287433624268,
+      "learning_rate": 1.9484119090246008e-05,
+      "loss": 0.1459,
+      "step": 890
+    },
+    {
+      "epoch": 0.05776590367535562,
+      "grad_norm": 12.011093139648438,
+      "learning_rate": 1.9470857370200916e-05,
+      "loss": 0.1202,
+      "step": 900
+    },
+    {
+      "epoch": 0.05840774704952624,
+      "grad_norm": 5.606518745422363,
+      "learning_rate": 1.9457595650155827e-05,
+      "loss": 0.1153,
+      "step": 910
+    },
+    {
+      "epoch": 0.059049590423696854,
+      "grad_norm": 5.869558811187744,
+      "learning_rate": 1.9444333930110738e-05,
+      "loss": 0.1428,
+      "step": 920
+    },
+    {
+      "epoch": 0.059691433797867476,
+      "grad_norm": 9.52774715423584,
+      "learning_rate": 1.9431072210065646e-05,
+      "loss": 0.1347,
+      "step": 930
+    },
+    {
+      "epoch": 0.0603332771720381,
+      "grad_norm": 8.32546329498291,
+      "learning_rate": 1.9417810490020557e-05,
+      "loss": 0.153,
+      "step": 940
+    },
+    {
+      "epoch": 0.06097512054620871,
+      "grad_norm": 9.355123519897461,
+      "learning_rate": 1.9404548769975465e-05,
+      "loss": 0.123,
+      "step": 950
+    },
+    {
+      "epoch": 0.06161696392037933,
+      "grad_norm": 6.419095039367676,
+      "learning_rate": 1.939128704993038e-05,
+      "loss": 0.1075,
+      "step": 960
+    },
+    {
+      "epoch": 0.062258807294549946,
+      "grad_norm": 8.51015567779541,
+      "learning_rate": 1.9378025329885287e-05,
+      "loss": 0.1322,
+      "step": 970
+    },
+    {
+      "epoch": 0.06290065066872057,
+      "grad_norm": 6.859710216522217,
+      "learning_rate": 1.93647636098402e-05,
+      "loss": 0.1363,
+      "step": 980
+    },
+    {
+      "epoch": 0.06354249404289118,
+      "grad_norm": 6.191632270812988,
+      "learning_rate": 1.935150188979511e-05,
+      "loss": 0.1287,
+      "step": 990
+    },
+    {
+      "epoch": 0.0641843374170618,
+      "grad_norm": 2.526578903198242,
+      "learning_rate": 1.9338240169750017e-05,
+      "loss": 0.1466,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06482618079123242,
+      "grad_norm": 7.311110019683838,
+      "learning_rate": 1.932497844970493e-05,
+      "loss": 0.1341,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06546802416540304,
+      "grad_norm": 1.3056310415267944,
+      "learning_rate": 1.931171672965984e-05,
+      "loss": 0.1211,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06610986753957365,
+      "grad_norm": 9.120423316955566,
+      "learning_rate": 1.929845500961475e-05,
+      "loss": 0.1268,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06675171091374428,
+      "grad_norm": 2.8570845127105713,
+      "learning_rate": 1.928519328956966e-05,
+      "loss": 0.0942,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0673935542879149,
+      "grad_norm": 9.288402557373047,
+      "learning_rate": 1.927193156952457e-05,
+      "loss": 0.147,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06803539766208551,
+      "grad_norm": 6.629462242126465,
+      "learning_rate": 1.9258669849479478e-05,
+      "loss": 0.1461,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06867724103625612,
+      "grad_norm": 3.9713406562805176,
+      "learning_rate": 1.924540812943439e-05,
+      "loss": 0.117,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06931908441042675,
+      "grad_norm": 8.020031929016113,
+      "learning_rate": 1.92321464093893e-05,
+      "loss": 0.1106,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06996092778459737,
+      "grad_norm": 5.471112251281738,
+      "learning_rate": 1.921888468934421e-05,
+      "loss": 0.114,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07060277115876798,
+      "grad_norm": 2.482114315032959,
+      "learning_rate": 1.920562296929912e-05,
+      "loss": 0.0784,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0712446145329386,
+      "grad_norm": 6.366470813751221,
+      "learning_rate": 1.919236124925403e-05,
+      "loss": 0.1158,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07188645790710922,
+      "grad_norm": 8.483400344848633,
+      "learning_rate": 1.917909952920894e-05,
+      "loss": 0.1098,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07252830128127984,
+      "grad_norm": 7.386050701141357,
+      "learning_rate": 1.916583780916385e-05,
+      "loss": 0.1484,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07317014465545045,
+      "grad_norm": 4.264742374420166,
+      "learning_rate": 1.915257608911876e-05,
+      "loss": 0.1496,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07381198802962108,
+      "grad_norm": 6.322717666625977,
+      "learning_rate": 1.9139314369073668e-05,
+      "loss": 0.1267,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07445383140379169,
+      "grad_norm": 4.288897514343262,
+      "learning_rate": 1.9126052649028582e-05,
+      "loss": 0.1293,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0750956747779623,
+      "grad_norm": 3.701270818710327,
+      "learning_rate": 1.911279092898349e-05,
+      "loss": 0.1303,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07573751815213292,
+      "grad_norm": 5.492668628692627,
+      "learning_rate": 1.90995292089384e-05,
+      "loss": 0.0987,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07637936152630355,
+      "grad_norm": 4.863729953765869,
+      "learning_rate": 1.908626748889331e-05,
+      "loss": 0.1186,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07702120490047416,
+      "grad_norm": 5.729610443115234,
+      "learning_rate": 1.907300576884822e-05,
+      "loss": 0.1224,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07766304827464478,
+      "grad_norm": 3.333712100982666,
+      "learning_rate": 1.905974404880313e-05,
+      "loss": 0.1084,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07830489164881539,
+      "grad_norm": 9.325702667236328,
+      "learning_rate": 1.9046482328758043e-05,
+      "loss": 0.0838,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07894673502298602,
+      "grad_norm": 4.56976842880249,
+      "learning_rate": 1.9033220608712954e-05,
+      "loss": 0.0924,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07958857839715663,
+      "grad_norm": 5.612900733947754,
+      "learning_rate": 1.9019958888667862e-05,
+      "loss": 0.1233,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08023042177132725,
+      "grad_norm": 5.123150825500488,
+      "learning_rate": 1.9006697168622773e-05,
+      "loss": 0.0937,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08087226514549788,
+      "grad_norm": 5.567653179168701,
+      "learning_rate": 1.899343544857768e-05,
+      "loss": 0.0755,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08151410851966849,
+      "grad_norm": 1.4452855587005615,
+      "learning_rate": 1.8980173728532592e-05,
+      "loss": 0.0767,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0821559518938391,
+      "grad_norm": 3.7622110843658447,
+      "learning_rate": 1.8966912008487503e-05,
+      "loss": 0.087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08279779526800972,
+      "grad_norm": 1.222861886024475,
+      "learning_rate": 1.8953650288442414e-05,
+      "loss": 0.0841,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08343963864218035,
+      "grad_norm": 5.492730140686035,
+      "learning_rate": 1.8940388568397322e-05,
+      "loss": 0.0691,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08408148201635096,
+      "grad_norm": 4.395420074462891,
+      "learning_rate": 1.8927126848352233e-05,
+      "loss": 0.0668,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08472332539052158,
+      "grad_norm": 6.103671550750732,
+      "learning_rate": 1.8913865128307144e-05,
+      "loss": 0.0723,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08536516876469219,
+      "grad_norm": 5.228280067443848,
+      "learning_rate": 1.8900603408262052e-05,
+      "loss": 0.1032,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08600701213886282,
+      "grad_norm": 1.1769144535064697,
+      "learning_rate": 1.8887341688216963e-05,
+      "loss": 0.0815,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08664885551303343,
+      "grad_norm": 6.19387149810791,
+      "learning_rate": 1.8874079968171874e-05,
+      "loss": 0.0913,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08729069888720405,
+      "grad_norm": 2.781297206878662,
+      "learning_rate": 1.8860818248126786e-05,
+      "loss": 0.0752,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08793254226137467,
+      "grad_norm": 3.160879611968994,
+      "learning_rate": 1.8847556528081693e-05,
+      "loss": 0.0811,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08857438563554529,
+      "grad_norm": 4.549344539642334,
+      "learning_rate": 1.8834294808036605e-05,
+      "loss": 0.1094,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0892162290097159,
+      "grad_norm": 6.160617828369141,
+      "learning_rate": 1.8821033087991512e-05,
+      "loss": 0.0858,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08985807238388652,
+      "grad_norm": 7.712894916534424,
+      "learning_rate": 1.8807771367946424e-05,
+      "loss": 0.0855,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09049991575805715,
+      "grad_norm": 6.89589262008667,
+      "learning_rate": 1.8794509647901335e-05,
+      "loss": 0.0782,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09114175913222776,
+      "grad_norm": 4.766812324523926,
+      "learning_rate": 1.8781247927856246e-05,
+      "loss": 0.0917,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09178360250639837,
+      "grad_norm": 7.581903457641602,
+      "learning_rate": 1.8767986207811154e-05,
+      "loss": 0.1413,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09242544588056899,
+      "grad_norm": 7.758519649505615,
+      "learning_rate": 1.8754724487766065e-05,
+      "loss": 0.0994,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09306728925473962,
+      "grad_norm": 1.8544800281524658,
+      "learning_rate": 1.8741462767720976e-05,
+      "loss": 0.0913,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09370913262891023,
+      "grad_norm": 6.1761674880981445,
+      "learning_rate": 1.8728201047675884e-05,
+      "loss": 0.1086,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09435097600308084,
+      "grad_norm": 6.579728603363037,
+      "learning_rate": 1.8714939327630795e-05,
+      "loss": 0.1467,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09499281937725147,
+      "grad_norm": 6.048295974731445,
+      "learning_rate": 1.8701677607585706e-05,
+      "loss": 0.0818,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09563466275142209,
+      "grad_norm": 5.720274448394775,
+      "learning_rate": 1.8688415887540617e-05,
+      "loss": 0.0729,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0962765061255927,
+      "grad_norm": 8.690102577209473,
+      "learning_rate": 1.8675154167495525e-05,
+      "loss": 0.0647,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09691834949976331,
+      "grad_norm": 3.565424680709839,
+      "learning_rate": 1.8661892447450436e-05,
+      "loss": 0.0977,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09756019287393394,
+      "grad_norm": 8.023870468139648,
+      "learning_rate": 1.8648630727405347e-05,
+      "loss": 0.1096,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09820203624810456,
+      "grad_norm": 5.473697185516357,
+      "learning_rate": 1.8635369007360255e-05,
+      "loss": 0.0933,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09884387962227517,
+      "grad_norm": 1.821287751197815,
+      "learning_rate": 1.8622107287315166e-05,
+      "loss": 0.0839,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09948572299644579,
+      "grad_norm": 6.229073524475098,
+      "learning_rate": 1.8608845567270078e-05,
+      "loss": 0.1008,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10012756637061641,
+      "grad_norm": 5.2744550704956055,
+      "learning_rate": 1.859558384722499e-05,
+      "loss": 0.0841,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10076940974478703,
+      "grad_norm": 0.6284772753715515,
+      "learning_rate": 1.8582322127179897e-05,
+      "loss": 0.0721,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10141125311895764,
+      "grad_norm": 4.158771991729736,
+      "learning_rate": 1.8569060407134808e-05,
+      "loss": 0.0918,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10205309649312827,
+      "grad_norm": 1.499607801437378,
+      "learning_rate": 1.8555798687089715e-05,
+      "loss": 0.0836,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10269493986729888,
+      "grad_norm": 5.547268867492676,
+      "learning_rate": 1.8542536967044627e-05,
+      "loss": 0.0638,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1033367832414695,
+      "grad_norm": 1.7448383569717407,
+      "learning_rate": 1.8529275246999538e-05,
+      "loss": 0.0832,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10397862661564011,
+      "grad_norm": 7.011538028717041,
+      "learning_rate": 1.851601352695445e-05,
+      "loss": 0.0867,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10462046998981074,
+      "grad_norm": 5.350290298461914,
+      "learning_rate": 1.8502751806909357e-05,
+      "loss": 0.0698,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10526231336398136,
+      "grad_norm": 2.6952505111694336,
+      "learning_rate": 1.8489490086864268e-05,
+      "loss": 0.0789,
+      "step": 1640
+    },
+    {
+      "epoch": 0.10590415673815197,
+      "grad_norm": 2.261810779571533,
+      "learning_rate": 1.847622836681918e-05,
+      "loss": 0.0778,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10654600011232258,
+      "grad_norm": 4.009234428405762,
+      "learning_rate": 1.8462966646774087e-05,
+      "loss": 0.0621,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10718784348649321,
+      "grad_norm": 7.318385601043701,
+      "learning_rate": 1.8449704926728998e-05,
+      "loss": 0.0589,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10782968686066383,
+      "grad_norm": 1.8443090915679932,
+      "learning_rate": 1.843644320668391e-05,
+      "loss": 0.0832,
+      "step": 1680
+    },
+    {
+      "epoch": 0.10847153023483444,
+      "grad_norm": 6.533380508422852,
+      "learning_rate": 1.842318148663882e-05,
+      "loss": 0.0944,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10911337360900507,
+      "grad_norm": 7.015242099761963,
+      "learning_rate": 1.8409919766593728e-05,
+      "loss": 0.0743,
+      "step": 1700
+    },
+    {
+      "epoch": 0.10975521698317568,
+      "grad_norm": 5.167893409729004,
+      "learning_rate": 1.839665804654864e-05,
+      "loss": 0.0946,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1103970603573463,
+      "grad_norm": 6.0605058670043945,
+      "learning_rate": 1.8383396326503547e-05,
+      "loss": 0.0796,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11103890373151691,
+      "grad_norm": 6.715466022491455,
+      "learning_rate": 1.837013460645846e-05,
+      "loss": 0.0782,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11168074710568754,
+      "grad_norm": 2.545156955718994,
+      "learning_rate": 1.835687288641337e-05,
+      "loss": 0.0713,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11232259047985815,
+      "grad_norm": 2.0205812454223633,
+      "learning_rate": 1.834361116636828e-05,
+      "loss": 0.0783,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11296443385402877,
+      "grad_norm": 7.414153575897217,
+      "learning_rate": 1.833034944632319e-05,
+      "loss": 0.0533,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1136062772281994,
+      "grad_norm": 1.9587926864624023,
+      "learning_rate": 1.83170877262781e-05,
+      "loss": 0.064,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11424812060237001,
+      "grad_norm": 4.822171211242676,
+      "learning_rate": 1.830382600623301e-05,
+      "loss": 0.0537,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11488996397654062,
+      "grad_norm": 7.510979652404785,
+      "learning_rate": 1.829056428618792e-05,
+      "loss": 0.0759,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11553180735071124,
+      "grad_norm": 2.2131752967834473,
+      "learning_rate": 1.827730256614283e-05,
+      "loss": 0.0611,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11617365072488187,
+      "grad_norm": 6.622641563415527,
+      "learning_rate": 1.826404084609774e-05,
+      "loss": 0.0564,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11681549409905248,
+      "grad_norm": 2.9961423873901367,
+      "learning_rate": 1.8250779126052652e-05,
+      "loss": 0.0731,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1174573374732231,
+      "grad_norm": 0.5653957724571228,
+      "learning_rate": 1.823751740600756e-05,
+      "loss": 0.0794,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11809918084739371,
+      "grad_norm": 4.936248302459717,
+      "learning_rate": 1.822425568596247e-05,
+      "loss": 0.0809,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11874102422156434,
+      "grad_norm": 2.4790220260620117,
+      "learning_rate": 1.8210993965917382e-05,
+      "loss": 0.0462,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11938286759573495,
+      "grad_norm": 5.853545665740967,
+      "learning_rate": 1.819773224587229e-05,
+      "loss": 0.0746,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12002471096990557,
+      "grad_norm": 2.9562971591949463,
+      "learning_rate": 1.81844705258272e-05,
+      "loss": 0.0618,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1206665543440762,
+      "grad_norm": 5.309076309204102,
+      "learning_rate": 1.8171208805782112e-05,
+      "loss": 0.0814,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12130839771824681,
+      "grad_norm": 3.951395273208618,
+      "learning_rate": 1.8157947085737024e-05,
+      "loss": 0.0993,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12195024109241742,
+      "grad_norm": 3.8987228870391846,
+      "learning_rate": 1.814468536569193e-05,
+      "loss": 0.0684,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12259208446658804,
+      "grad_norm": 2.9505648612976074,
+      "learning_rate": 1.8131423645646842e-05,
+      "loss": 0.057,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12323392784075866,
+      "grad_norm": 4.24475622177124,
+      "learning_rate": 1.811816192560175e-05,
+      "loss": 0.0593,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12387577121492928,
+      "grad_norm": 5.541717052459717,
+      "learning_rate": 1.810490020555666e-05,
+      "loss": 0.0644,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12451761458909989,
+      "grad_norm": 4.3561248779296875,
+      "learning_rate": 1.8091638485511573e-05,
+      "loss": 0.0463,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12515945796327052,
+      "grad_norm": 0.40995126962661743,
+      "learning_rate": 1.8078376765466484e-05,
+      "loss": 0.0758,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12580130133744113,
+      "grad_norm": 1.771988868713379,
+      "learning_rate": 1.806511504542139e-05,
+      "loss": 0.0722,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12644314471161175,
+      "grad_norm": 1.8136435747146606,
+      "learning_rate": 1.8051853325376303e-05,
+      "loss": 0.0673,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12708498808578236,
+      "grad_norm": 4.0566277503967285,
+      "learning_rate": 1.8038591605331214e-05,
+      "loss": 0.0566,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12772683145995298,
+      "grad_norm": 4.5531439781188965,
+      "learning_rate": 1.8025329885286122e-05,
+      "loss": 0.0653,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1283686748341236,
+      "grad_norm": 3.038701057434082,
+      "learning_rate": 1.8012068165241033e-05,
+      "loss": 0.048,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12901051820829423,
+      "grad_norm": 10.788609504699707,
+      "learning_rate": 1.7998806445195944e-05,
+      "loss": 0.0777,
+      "step": 2010
+    },
+    {
+      "epoch": 0.12965236158246485,
+      "grad_norm": 7.363852024078369,
+      "learning_rate": 1.7985544725150855e-05,
+      "loss": 0.0692,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13029420495663546,
+      "grad_norm": 4.210390090942383,
+      "learning_rate": 1.7972283005105763e-05,
+      "loss": 0.089,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13093604833080608,
+      "grad_norm": 3.144101858139038,
+      "learning_rate": 1.7959021285060674e-05,
+      "loss": 0.0683,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1315778917049767,
+      "grad_norm": 2.3627800941467285,
+      "learning_rate": 1.7945759565015582e-05,
+      "loss": 0.0427,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1322197350791473,
+      "grad_norm": 3.693427562713623,
+      "learning_rate": 1.7932497844970493e-05,
+      "loss": 0.0698,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13286157845331792,
+      "grad_norm": 6.718416690826416,
+      "learning_rate": 1.7919236124925404e-05,
+      "loss": 0.0688,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13350342182748856,
+      "grad_norm": 3.3696110248565674,
+      "learning_rate": 1.7905974404880315e-05,
+      "loss": 0.0684,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13414526520165918,
+      "grad_norm": 4.205007553100586,
+      "learning_rate": 1.7892712684835227e-05,
+      "loss": 0.0947,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1347871085758298,
+      "grad_norm": 6.637697696685791,
+      "learning_rate": 1.7879450964790134e-05,
+      "loss": 0.0672,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1354289519500004,
+      "grad_norm": 5.024839401245117,
+      "learning_rate": 1.7866189244745046e-05,
+      "loss": 0.0548,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13607079532417102,
+      "grad_norm": 2.4971957206726074,
+      "learning_rate": 1.7852927524699953e-05,
+      "loss": 0.0721,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13671263869834163,
+      "grad_norm": 4.841691017150879,
+      "learning_rate": 1.7839665804654865e-05,
+      "loss": 0.0579,
+      "step": 2130
+    },
+    {
+      "epoch": 0.13735448207251225,
+      "grad_norm": 1.5802949666976929,
+      "learning_rate": 1.7826404084609776e-05,
+      "loss": 0.0478,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13799632544668286,
+      "grad_norm": 3.178330421447754,
+      "learning_rate": 1.7813142364564687e-05,
+      "loss": 0.0382,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1386381688208535,
+      "grad_norm": 1.2376867532730103,
+      "learning_rate": 1.7799880644519595e-05,
+      "loss": 0.0493,
+      "step": 2160
+    },
+    {
+      "epoch": 0.13928001219502412,
+      "grad_norm": 4.373589515686035,
+      "learning_rate": 1.7786618924474506e-05,
+      "loss": 0.0722,
+      "step": 2170
+    },
+    {
+      "epoch": 0.13992185556919473,
+      "grad_norm": 8.756216049194336,
+      "learning_rate": 1.7773357204429417e-05,
+      "loss": 0.0661,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14056369894336534,
+      "grad_norm": 3.211627960205078,
+      "learning_rate": 1.7760095484384325e-05,
+      "loss": 0.0808,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14120554231753596,
+      "grad_norm": 2.4077908992767334,
+      "learning_rate": 1.7746833764339236e-05,
+      "loss": 0.0597,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14184738569170657,
+      "grad_norm": 3.222778081893921,
+      "learning_rate": 1.7733572044294147e-05,
+      "loss": 0.0823,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1424892290658772,
+      "grad_norm": 6.3043742179870605,
+      "learning_rate": 1.772031032424906e-05,
+      "loss": 0.053,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14313107244004783,
+      "grad_norm": 3.8473875522613525,
+      "learning_rate": 1.7707048604203966e-05,
+      "loss": 0.0458,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14377291581421844,
+      "grad_norm": 1.6303006410598755,
+      "learning_rate": 1.7693786884158877e-05,
+      "loss": 0.0436,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14441475918838906,
+      "grad_norm": 6.238514423370361,
+      "learning_rate": 1.7680525164113785e-05,
+      "loss": 0.0694,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14505660256255967,
+      "grad_norm": 1.9263882637023926,
+      "learning_rate": 1.7667263444068696e-05,
+      "loss": 0.052,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1456984459367303,
+      "grad_norm": 5.976217746734619,
+      "learning_rate": 1.7654001724023607e-05,
+      "loss": 0.0585,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1463402893109009,
+      "grad_norm": 2.3731586933135986,
+      "learning_rate": 1.764074000397852e-05,
+      "loss": 0.0548,
+      "step": 2280
+    },
+    {
+      "epoch": 0.14698213268507151,
+      "grad_norm": 2.778264045715332,
+      "learning_rate": 1.7627478283933426e-05,
+      "loss": 0.0442,
+      "step": 2290
+    },
+    {
+      "epoch": 0.14762397605924216,
+      "grad_norm": 2.549161672592163,
+      "learning_rate": 1.7614216563888338e-05,
+      "loss": 0.0512,
+      "step": 2300
+    },
+    {
+      "epoch": 0.14826581943341277,
+      "grad_norm": 2.6218628883361816,
+      "learning_rate": 1.760095484384325e-05,
+      "loss": 0.0504,
+      "step": 2310
+    },
+    {
+      "epoch": 0.14890766280758339,
+      "grad_norm": 2.735814094543457,
+      "learning_rate": 1.7587693123798157e-05,
+      "loss": 0.0486,
+      "step": 2320
+    },
+    {
+      "epoch": 0.149549506181754,
+      "grad_norm": 5.613633155822754,
+      "learning_rate": 1.7574431403753068e-05,
+      "loss": 0.0499,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1501913495559246,
+      "grad_norm": 4.932735443115234,
+      "learning_rate": 1.756116968370798e-05,
+      "loss": 0.094,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15083319293009523,
+      "grad_norm": 1.0233629941940308,
+      "learning_rate": 1.754790796366289e-05,
+      "loss": 0.052,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15147503630426584,
+      "grad_norm": 1.32416832447052,
+      "learning_rate": 1.7534646243617798e-05,
+      "loss": 0.0529,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15211687967843646,
+      "grad_norm": 2.5237696170806885,
+      "learning_rate": 1.752138452357271e-05,
+      "loss": 0.054,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1527587230526071,
+      "grad_norm": 9.520683288574219,
+      "learning_rate": 1.750812280352762e-05,
+      "loss": 0.0693,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1534005664267777,
+      "grad_norm": 5.118333339691162,
+      "learning_rate": 1.7494861083482528e-05,
+      "loss": 0.0591,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15404240980094833,
+      "grad_norm": 1.8272361755371094,
+      "learning_rate": 1.748159936343744e-05,
+      "loss": 0.0525,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15468425317511894,
+      "grad_norm": 5.692248344421387,
+      "learning_rate": 1.746833764339235e-05,
+      "loss": 0.0603,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15532609654928956,
+      "grad_norm": 1.802622675895691,
+      "learning_rate": 1.745507592334726e-05,
+      "loss": 0.0453,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15596793992346017,
+      "grad_norm": 8.569664001464844,
+      "learning_rate": 1.744181420330217e-05,
+      "loss": 0.0684,
+      "step": 2430
+    },
+    {
+      "epoch": 0.15660978329763078,
+      "grad_norm": 1.4416598081588745,
+      "learning_rate": 1.742855248325708e-05,
+      "loss": 0.0574,
+      "step": 2440
+    },
+    {
+      "epoch": 0.15725162667180143,
+      "grad_norm": 1.8214036226272583,
+      "learning_rate": 1.7415290763211988e-05,
+      "loss": 0.0694,
+      "step": 2450
+    },
+    {
+      "epoch": 0.15789347004597204,
+      "grad_norm": 2.4023277759552,
+      "learning_rate": 1.74020290431669e-05,
+      "loss": 0.0577,
+      "step": 2460
+    },
+    {
+      "epoch": 0.15853531342014265,
+      "grad_norm": 9.605439186096191,
+      "learning_rate": 1.738876732312181e-05,
+      "loss": 0.0671,
+      "step": 2470
+    },
+    {
+      "epoch": 0.15917715679431327,
+      "grad_norm": 0.8370853066444397,
+      "learning_rate": 1.7375505603076722e-05,
+      "loss": 0.0528,
+      "step": 2480
+    },
+    {
+      "epoch": 0.15981900016848388,
+      "grad_norm": 3.8083314895629883,
+      "learning_rate": 1.736224388303163e-05,
+      "loss": 0.0632,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1604608435426545,
+      "grad_norm": 4.146091938018799,
+      "learning_rate": 1.734898216298654e-05,
+      "loss": 0.048,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1611026869168251,
+      "grad_norm": 3.8874735832214355,
+      "learning_rate": 1.7335720442941452e-05,
+      "loss": 0.0645,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16174453029099575,
+      "grad_norm": 2.019355535507202,
+      "learning_rate": 1.732245872289636e-05,
+      "loss": 0.0493,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16238637366516637,
+      "grad_norm": 3.533977746963501,
+      "learning_rate": 1.730919700285127e-05,
+      "loss": 0.0529,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16302821703933698,
+      "grad_norm": 3.5773158073425293,
+      "learning_rate": 1.7295935282806182e-05,
+      "loss": 0.0462,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1636700604135076,
+      "grad_norm": 2.2680728435516357,
+      "learning_rate": 1.7282673562761093e-05,
+      "loss": 0.0772,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1643119037876782,
+      "grad_norm": 4.510113716125488,
+      "learning_rate": 1.7269411842716e-05,
+      "loss": 0.049,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16495374716184882,
+      "grad_norm": 6.012620449066162,
+      "learning_rate": 1.7256150122670912e-05,
+      "loss": 0.0701,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16559559053601944,
+      "grad_norm": 4.502586841583252,
+      "learning_rate": 1.724288840262582e-05,
+      "loss": 0.058,
+      "step": 2580
+    },
+    {
+      "epoch": 0.16623743391019005,
+      "grad_norm": 0.7591032981872559,
+      "learning_rate": 1.722962668258073e-05,
+      "loss": 0.0219,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1668792772843607,
+      "grad_norm": 3.2919363975524902,
+      "learning_rate": 1.7216364962535642e-05,
+      "loss": 0.0649,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1675211206585313,
+      "grad_norm": 2.932372570037842,
+      "learning_rate": 1.7203103242490553e-05,
+      "loss": 0.0419,
+      "step": 2610
+    },
+    {
+      "epoch": 0.16816296403270192,
+      "grad_norm": 2.4418764114379883,
+      "learning_rate": 1.7189841522445465e-05,
+      "loss": 0.0495,
+      "step": 2620
+    },
+    {
+      "epoch": 0.16880480740687254,
+      "grad_norm": 3.327172040939331,
+      "learning_rate": 1.7176579802400372e-05,
+      "loss": 0.0531,
+      "step": 2630
+    },
+    {
+      "epoch": 0.16944665078104315,
+      "grad_norm": 3.8956868648529053,
+      "learning_rate": 1.7163318082355284e-05,
+      "loss": 0.05,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17008849415521377,
+      "grad_norm": 2.2222423553466797,
+      "learning_rate": 1.715005636231019e-05,
+      "loss": 0.0582,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17073033752938438,
+      "grad_norm": 2.0240089893341064,
+      "learning_rate": 1.7136794642265102e-05,
+      "loss": 0.0755,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17137218090355502,
+      "grad_norm": 2.35172176361084,
+      "learning_rate": 1.7123532922220014e-05,
+      "loss": 0.0713,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17201402427772564,
+      "grad_norm": 2.8345654010772705,
+      "learning_rate": 1.7110271202174925e-05,
+      "loss": 0.0456,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17265586765189625,
+      "grad_norm": 0.52686607837677,
+      "learning_rate": 1.7097009482129833e-05,
+      "loss": 0.0634,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17329771102606686,
+      "grad_norm": 4.439308166503906,
+      "learning_rate": 1.7083747762084744e-05,
+      "loss": 0.0523,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17393955440023748,
+      "grad_norm": 1.9147131443023682,
+      "learning_rate": 1.7070486042039655e-05,
+      "loss": 0.0418,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1745813977744081,
+      "grad_norm": 0.44640451669692993,
+      "learning_rate": 1.7057224321994563e-05,
+      "loss": 0.0447,
+      "step": 2720
+    },
+    {
+      "epoch": 0.1752232411485787,
+      "grad_norm": 3.7846157550811768,
+      "learning_rate": 1.7043962601949474e-05,
+      "loss": 0.0424,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17586508452274935,
+      "grad_norm": 4.769283294677734,
+      "learning_rate": 1.7030700881904385e-05,
+      "loss": 0.0394,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17650692789691996,
+      "grad_norm": 0.8889997005462646,
+      "learning_rate": 1.7017439161859296e-05,
+      "loss": 0.0422,
+      "step": 2750
+    },
+    {
+      "epoch": 0.17714877127109058,
+      "grad_norm": 2.665426731109619,
+      "learning_rate": 1.7004177441814204e-05,
+      "loss": 0.0495,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1777906146452612,
+      "grad_norm": 2.6268789768218994,
+      "learning_rate": 1.6990915721769115e-05,
+      "loss": 0.0503,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1784324580194318,
+      "grad_norm": 3.4581639766693115,
+      "learning_rate": 1.6977654001724023e-05,
+      "loss": 0.0483,
+      "step": 2780
+    },
+    {
+      "epoch": 0.17907430139360242,
+      "grad_norm": 1.1500879526138306,
+      "learning_rate": 1.6964392281678934e-05,
+      "loss": 0.0438,
+      "step": 2790
+    },
+    {
+      "epoch": 0.17971614476777303,
+      "grad_norm": 6.2053022384643555,
+      "learning_rate": 1.6951130561633845e-05,
+      "loss": 0.0591,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18035798814194368,
+      "grad_norm": 3.427523136138916,
+      "learning_rate": 1.6937868841588757e-05,
+      "loss": 0.0602,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1809998315161143,
+      "grad_norm": 6.429181098937988,
+      "learning_rate": 1.6924607121543664e-05,
+      "loss": 0.0462,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1816416748902849,
+      "grad_norm": 1.9181201457977295,
+      "learning_rate": 1.6911345401498575e-05,
+      "loss": 0.0402,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18228351826445552,
+      "grad_norm": 1.2943888902664185,
+      "learning_rate": 1.6898083681453487e-05,
+      "loss": 0.0529,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18292536163862613,
+      "grad_norm": 3.2357335090637207,
+      "learning_rate": 1.6884821961408394e-05,
+      "loss": 0.0726,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18356720501279675,
+      "grad_norm": 0.8399866223335266,
+      "learning_rate": 1.6871560241363306e-05,
+      "loss": 0.034,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18420904838696736,
+      "grad_norm": 1.1773546934127808,
+      "learning_rate": 1.6858298521318217e-05,
+      "loss": 0.0533,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18485089176113798,
+      "grad_norm": 1.484147071838379,
+      "learning_rate": 1.6845036801273128e-05,
+      "loss": 0.0405,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18549273513530862,
+      "grad_norm": 1.646011233329773,
+      "learning_rate": 1.6831775081228036e-05,
+      "loss": 0.0766,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18613457850947923,
+      "grad_norm": 4.303348064422607,
+      "learning_rate": 1.6818513361182947e-05,
+      "loss": 0.0484,
+      "step": 2900
+    },
+    {
+      "epoch": 0.18677642188364985,
+      "grad_norm": 1.86674165725708,
+      "learning_rate": 1.6805251641137855e-05,
+      "loss": 0.0373,
+      "step": 2910
+    },
+    {
+      "epoch": 0.18741826525782046,
+      "grad_norm": 4.927422523498535,
+      "learning_rate": 1.6791989921092766e-05,
+      "loss": 0.059,
+      "step": 2920
+    },
+    {
+      "epoch": 0.18806010863199107,
+      "grad_norm": 1.2013154029846191,
+      "learning_rate": 1.6778728201047677e-05,
+      "loss": 0.0484,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1887019520061617,
+      "grad_norm": 1.8602125644683838,
+      "learning_rate": 1.6765466481002588e-05,
+      "loss": 0.0651,
+      "step": 2940
+    },
+    {
+      "epoch": 0.1893437953803323,
+      "grad_norm": 2.0056393146514893,
+      "learning_rate": 1.67522047609575e-05,
+      "loss": 0.0347,
+      "step": 2950
+    },
+    {
+      "epoch": 0.18998563875450294,
+      "grad_norm": 4.299374580383301,
+      "learning_rate": 1.6738943040912407e-05,
+      "loss": 0.0415,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19062748212867356,
+      "grad_norm": 1.701195240020752,
+      "learning_rate": 1.672568132086732e-05,
+      "loss": 0.0427,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19126932550284417,
+      "grad_norm": 3.681411027908325,
+      "learning_rate": 1.6712419600822226e-05,
+      "loss": 0.037,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1919111688770148,
+      "grad_norm": 2.3046762943267822,
+      "learning_rate": 1.6699157880777137e-05,
+      "loss": 0.0467,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1925530122511854,
+      "grad_norm": 3.976912260055542,
+      "learning_rate": 1.668589616073205e-05,
+      "loss": 0.0516,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19319485562535602,
+      "grad_norm": 2.6646742820739746,
+      "learning_rate": 1.667263444068696e-05,
+      "loss": 0.0407,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19383669899952663,
+      "grad_norm": 1.8901029825210571,
+      "learning_rate": 1.6659372720641867e-05,
+      "loss": 0.0321,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19447854237369727,
+      "grad_norm": 2.3338727951049805,
+      "learning_rate": 1.664611100059678e-05,
+      "loss": 0.0609,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19512038574786789,
+      "grad_norm": 5.7270121574401855,
+      "learning_rate": 1.663284928055169e-05,
+      "loss": 0.0466,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1957622291220385,
+      "grad_norm": 4.083347320556641,
+      "learning_rate": 1.6619587560506598e-05,
+      "loss": 0.0491,
+      "step": 3050
+    },
+    {
+      "epoch": 0.19640407249620911,
+      "grad_norm": 1.9552839994430542,
+      "learning_rate": 1.660632584046151e-05,
+      "loss": 0.0524,
+      "step": 3060
+    },
+    {
+      "epoch": 0.19704591587037973,
+      "grad_norm": 2.211383104324341,
+      "learning_rate": 1.659306412041642e-05,
+      "loss": 0.038,
+      "step": 3070
+    },
+    {
+      "epoch": 0.19768775924455034,
+      "grad_norm": 0.9306750297546387,
+      "learning_rate": 1.657980240037133e-05,
+      "loss": 0.0488,
+      "step": 3080
+    },
+    {
+      "epoch": 0.19832960261872096,
+      "grad_norm": 2.246899366378784,
+      "learning_rate": 1.656654068032624e-05,
+      "loss": 0.0413,
+      "step": 3090
+    },
+    {
+      "epoch": 0.19897144599289157,
+      "grad_norm": 3.014577627182007,
+      "learning_rate": 1.655327896028115e-05,
+      "loss": 0.0397,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1996132893670622,
+      "grad_norm": 5.5264201164245605,
+      "learning_rate": 1.6540017240236058e-05,
+      "loss": 0.0371,
+      "step": 3110
+    },
+    {
+      "epoch": 0.20025513274123283,
+      "grad_norm": 3.594407320022583,
+      "learning_rate": 1.652675552019097e-05,
+      "loss": 0.0415,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20089697611540344,
+      "grad_norm": 1.3521718978881836,
+      "learning_rate": 1.651349380014588e-05,
+      "loss": 0.0588,
+      "step": 3130
+    },
+    {
+      "epoch": 0.20153881948957406,
+      "grad_norm": 9.015791893005371,
+      "learning_rate": 1.650023208010079e-05,
+      "loss": 0.0596,
+      "step": 3140
+    },
+    {
+      "epoch": 0.20218066286374467,
+      "grad_norm": 2.759474754333496,
+      "learning_rate": 1.64869703600557e-05,
+      "loss": 0.0498,
+      "step": 3150
+    },
+    {
+      "epoch": 0.20282250623791528,
+      "grad_norm": 4.701976776123047,
+      "learning_rate": 1.647370864001061e-05,
+      "loss": 0.0404,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2034643496120859,
+      "grad_norm": 6.130093097686768,
+      "learning_rate": 1.646044691996552e-05,
+      "loss": 0.0672,
+      "step": 3170
+    },
+    {
+      "epoch": 0.20410619298625654,
+      "grad_norm": Infinity,
+      "learning_rate": 1.644718519992043e-05,
+      "loss": 0.0585,
+      "step": 3180
+    },
+    {
+      "epoch": 0.20474803636042715,
+      "grad_norm": 2.636387348175049,
+      "learning_rate": 1.6433923479875344e-05,
+      "loss": 0.0632,
+      "step": 3190
+    },
+    {
+      "epoch": 0.20538987973459777,
+      "grad_norm": 1.0386542081832886,
+      "learning_rate": 1.642066175983025e-05,
+      "loss": 0.0323,
+      "step": 3200
+    },
+    {
+      "epoch": 0.20603172310876838,
+      "grad_norm": 6.843395709991455,
+      "learning_rate": 1.6407400039785163e-05,
+      "loss": 0.0596,
+      "step": 3210
+    },
+    {
+      "epoch": 0.206673566482939,
+      "grad_norm": 4.445530891418457,
+      "learning_rate": 1.639413831974007e-05,
+      "loss": 0.0532,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2073154098571096,
+      "grad_norm": 0.7461674213409424,
+      "learning_rate": 1.6380876599694982e-05,
+      "loss": 0.0511,
+      "step": 3230
+    },
+    {
+      "epoch": 0.20795725323128023,
+      "grad_norm": 7.616549015045166,
+      "learning_rate": 1.6367614879649893e-05,
+      "loss": 0.0434,
+      "step": 3240
+    },
+    {
+      "epoch": 0.20859909660545087,
+      "grad_norm": 1.3540281057357788,
+      "learning_rate": 1.63543531596048e-05,
+      "loss": 0.0312,
+      "step": 3250
+    },
+    {
+      "epoch": 0.20924093997962148,
+      "grad_norm": 0.9516130685806274,
+      "learning_rate": 1.6341091439559712e-05,
+      "loss": 0.0552,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2098827833537921,
+      "grad_norm": 0.5843936800956726,
+      "learning_rate": 1.6327829719514623e-05,
+      "loss": 0.0163,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2105246267279627,
+      "grad_norm": 2.156327486038208,
+      "learning_rate": 1.6314567999469534e-05,
+      "loss": 0.0367,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21116647010213332,
+      "grad_norm": 1.6830333471298218,
+      "learning_rate": 1.6301306279424442e-05,
+      "loss": 0.033,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21180831347630394,
+      "grad_norm": 4.959284782409668,
+      "learning_rate": 1.6288044559379353e-05,
+      "loss": 0.0334,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21245015685047455,
+      "grad_norm": 5.351144313812256,
+      "learning_rate": 1.627478283933426e-05,
+      "loss": 0.0266,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21309200022464517,
+      "grad_norm": 2.5864357948303223,
+      "learning_rate": 1.6261521119289175e-05,
+      "loss": 0.0276,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2137338435988158,
+      "grad_norm": 4.892467021942139,
+      "learning_rate": 1.6248259399244083e-05,
+      "loss": 0.0535,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21437568697298642,
+      "grad_norm": 4.954229354858398,
+      "learning_rate": 1.6234997679198994e-05,
+      "loss": 0.0344,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21501753034715704,
+      "grad_norm": 3.781883955001831,
+      "learning_rate": 1.6221735959153902e-05,
+      "loss": 0.0301,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21565937372132765,
+      "grad_norm": 2.5582525730133057,
+      "learning_rate": 1.6208474239108813e-05,
+      "loss": 0.0316,
+      "step": 3360
+    },
+    {
+      "epoch": 0.21630121709549827,
+      "grad_norm": 4.6287736892700195,
+      "learning_rate": 1.6195212519063725e-05,
+      "loss": 0.0445,
+      "step": 3370
+    },
+    {
+      "epoch": 0.21694306046966888,
+      "grad_norm": 3.5289618968963623,
+      "learning_rate": 1.6181950799018632e-05,
+      "loss": 0.0493,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2175849038438395,
+      "grad_norm": 2.802454710006714,
+      "learning_rate": 1.6168689078973544e-05,
+      "loss": 0.031,
+      "step": 3390
+    },
+    {
+      "epoch": 0.21822674721801014,
+      "grad_norm": 2.012895107269287,
+      "learning_rate": 1.6155427358928455e-05,
+      "loss": 0.04,
+      "step": 3400
+    },
+    {
+      "epoch": 0.21886859059218075,
+      "grad_norm": 1.0591882467269897,
+      "learning_rate": 1.6142165638883366e-05,
+      "loss": 0.0432,
+      "step": 3410
+    },
+    {
+      "epoch": 0.21951043396635136,
+      "grad_norm": 1.5135844945907593,
+      "learning_rate": 1.6128903918838274e-05,
+      "loss": 0.0289,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22015227734052198,
+      "grad_norm": 0.37961867451667786,
+      "learning_rate": 1.6115642198793185e-05,
+      "loss": 0.0399,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2207941207146926,
+      "grad_norm": 0.45994833111763,
+      "learning_rate": 1.6102380478748093e-05,
+      "loss": 0.0348,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2214359640888632,
+      "grad_norm": 2.2288460731506348,
+      "learning_rate": 1.6089118758703004e-05,
+      "loss": 0.0483,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22207780746303382,
+      "grad_norm": 1.327779769897461,
+      "learning_rate": 1.6075857038657915e-05,
+      "loss": 0.0448,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22271965083720446,
+      "grad_norm": 0.829675555229187,
+      "learning_rate": 1.6062595318612826e-05,
+      "loss": 0.0497,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22336149421137508,
+      "grad_norm": 3.0021743774414062,
+      "learning_rate": 1.6049333598567737e-05,
+      "loss": 0.0422,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2240033375855457,
+      "grad_norm": 3.0357425212860107,
+      "learning_rate": 1.6036071878522645e-05,
+      "loss": 0.0509,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2246451809597163,
+      "grad_norm": 1.5336086750030518,
+      "learning_rate": 1.6022810158477556e-05,
+      "loss": 0.0457,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22528702433388692,
+      "grad_norm": 1.462803840637207,
+      "learning_rate": 1.6009548438432464e-05,
+      "loss": 0.0188,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22592886770805753,
+      "grad_norm": 5.490964889526367,
+      "learning_rate": 1.599628671838738e-05,
+      "loss": 0.0472,
+      "step": 3520
+    },
+    {
+      "epoch": 0.22657071108222815,
+      "grad_norm": 5.324558258056641,
+      "learning_rate": 1.5983024998342286e-05,
+      "loss": 0.0233,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2272125544563988,
+      "grad_norm": 3.37449312210083,
+      "learning_rate": 1.5969763278297198e-05,
+      "loss": 0.0694,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2278543978305694,
+      "grad_norm": 2.1609532833099365,
+      "learning_rate": 1.5956501558252105e-05,
+      "loss": 0.0357,
+      "step": 3550
+    },
+    {
+      "epoch": 0.22849624120474002,
+      "grad_norm": 0.04418246075510979,
+      "learning_rate": 1.5943239838207017e-05,
+      "loss": 0.0224,
+      "step": 3560
+    },
+    {
+      "epoch": 0.22913808457891063,
+      "grad_norm": 5.424873352050781,
+      "learning_rate": 1.5929978118161928e-05,
+      "loss": 0.043,
+      "step": 3570
+    },
+    {
+      "epoch": 0.22977992795308125,
+      "grad_norm": 4.027209281921387,
+      "learning_rate": 1.5916716398116835e-05,
+      "loss": 0.0266,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23042177132725186,
+      "grad_norm": 1.5458093881607056,
+      "learning_rate": 1.5903454678071747e-05,
+      "loss": 0.0493,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23106361470142248,
+      "grad_norm": 2.4570200443267822,
+      "learning_rate": 1.5890192958026658e-05,
+      "loss": 0.04,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2317054580755931,
+      "grad_norm": 2.155552625656128,
+      "learning_rate": 1.587693123798157e-05,
+      "loss": 0.0494,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23234730144976373,
+      "grad_norm": 1.6081950664520264,
+      "learning_rate": 1.5863669517936477e-05,
+      "loss": 0.0399,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23298914482393435,
+      "grad_norm": 0.6873506903648376,
+      "learning_rate": 1.5850407797891388e-05,
+      "loss": 0.0386,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23363098819810496,
+      "grad_norm": 4.3657050132751465,
+      "learning_rate": 1.5837146077846296e-05,
+      "loss": 0.0418,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23427283157227557,
+      "grad_norm": 7.361201286315918,
+      "learning_rate": 1.582388435780121e-05,
+      "loss": 0.0516,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2349146749464462,
+      "grad_norm": 5.6615118980407715,
+      "learning_rate": 1.5810622637756118e-05,
+      "loss": 0.0575,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2355565183206168,
+      "grad_norm": 2.371455192565918,
+      "learning_rate": 1.579736091771103e-05,
+      "loss": 0.0639,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23619836169478742,
+      "grad_norm": 1.730076789855957,
+      "learning_rate": 1.5784099197665937e-05,
+      "loss": 0.0376,
+      "step": 3680
+    },
+    {
+      "epoch": 0.23684020506895806,
+      "grad_norm": 3.1411263942718506,
+      "learning_rate": 1.5770837477620848e-05,
+      "loss": 0.044,
+      "step": 3690
+    },
+    {
+      "epoch": 0.23748204844312867,
+      "grad_norm": 0.9767663478851318,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.0359,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2381238918172993,
+      "grad_norm": 0.32298144698143005,
+      "learning_rate": 1.5744314037530667e-05,
+      "loss": 0.0371,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2387657351914699,
+      "grad_norm": 0.5231297016143799,
+      "learning_rate": 1.5731052317485582e-05,
+      "loss": 0.0485,
+      "step": 3720
+    },
+    {
+      "epoch": 0.23940757856564052,
+      "grad_norm": 1.5875322818756104,
+      "learning_rate": 1.571779059744049e-05,
+      "loss": 0.0349,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24004942193981113,
+      "grad_norm": 4.602473735809326,
+      "learning_rate": 1.57045288773954e-05,
+      "loss": 0.048,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24069126531398174,
+      "grad_norm": 3.969989061355591,
+      "learning_rate": 1.569126715735031e-05,
+      "loss": 0.0355,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2413331086881524,
+      "grad_norm": 0.8934308290481567,
+      "learning_rate": 1.567800543730522e-05,
+      "loss": 0.0501,
+      "step": 3760
+    },
+    {
+      "epoch": 0.241974952062323,
+      "grad_norm": 5.02524995803833,
+      "learning_rate": 1.5664743717260127e-05,
+      "loss": 0.0579,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24261679543649362,
+      "grad_norm": 4.93678092956543,
+      "learning_rate": 1.5651481997215042e-05,
+      "loss": 0.0429,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24325863881066423,
+      "grad_norm": 3.8344504833221436,
+      "learning_rate": 1.563822027716995e-05,
+      "loss": 0.0421,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24390048218483484,
+      "grad_norm": 4.208110809326172,
+      "learning_rate": 1.562495855712486e-05,
+      "loss": 0.0423,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24454232555900546,
+      "grad_norm": 5.809897422790527,
+      "learning_rate": 1.5611696837079772e-05,
+      "loss": 0.063,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24518416893317607,
+      "grad_norm": 5.782272815704346,
+      "learning_rate": 1.559843511703468e-05,
+      "loss": 0.0523,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2458260123073467,
+      "grad_norm": 1.3277156352996826,
+      "learning_rate": 1.558517339698959e-05,
+      "loss": 0.0345,
+      "step": 3830
+    },
+    {
+      "epoch": 0.24646785568151733,
+      "grad_norm": 8.235041618347168,
+      "learning_rate": 1.55719116769445e-05,
+      "loss": 0.0589,
+      "step": 3840
+    },
+    {
+      "epoch": 0.24710969905568794,
+      "grad_norm": 2.093726396560669,
+      "learning_rate": 1.5558649956899413e-05,
+      "loss": 0.0435,
+      "step": 3850
+    },
+    {
+      "epoch": 0.24775154242985856,
+      "grad_norm": 6.27501106262207,
+      "learning_rate": 1.554538823685432e-05,
+      "loss": 0.0328,
+      "step": 3860
+    },
+    {
+      "epoch": 0.24839338580402917,
+      "grad_norm": 0.37880775332450867,
+      "learning_rate": 1.5532126516809232e-05,
+      "loss": 0.0237,
+      "step": 3870
+    },
+    {
+      "epoch": 0.24903522917819979,
+      "grad_norm": 2.049382209777832,
+      "learning_rate": 1.551886479676414e-05,
+      "loss": 0.0421,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2496770725523704,
+      "grad_norm": 2.3936824798583984,
+      "learning_rate": 1.550560307671905e-05,
+      "loss": 0.0403,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25031891592654104,
+      "grad_norm": 1.6919853687286377,
+      "learning_rate": 1.5492341356673962e-05,
+      "loss": 0.0303,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25096075930071166,
+      "grad_norm": 7.648709774017334,
+      "learning_rate": 1.5479079636628874e-05,
+      "loss": 0.037,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25160260267488227,
+      "grad_norm": 4.136491298675537,
+      "learning_rate": 1.546581791658378e-05,
+      "loss": 0.0607,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2522444460490529,
+      "grad_norm": 0.24048294126987457,
+      "learning_rate": 1.5452556196538693e-05,
+      "loss": 0.0564,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2528862894232235,
+      "grad_norm": 0.7890238761901855,
+      "learning_rate": 1.5439294476493604e-05,
+      "loss": 0.0416,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2535281327973941,
+      "grad_norm": 5.72561502456665,
+      "learning_rate": 1.542603275644851e-05,
+      "loss": 0.0606,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2541699761715647,
+      "grad_norm": 1.5823456048965454,
+      "learning_rate": 1.5412771036403423e-05,
+      "loss": 0.0284,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25481181954573534,
+      "grad_norm": 2.717170238494873,
+      "learning_rate": 1.539950931635833e-05,
+      "loss": 0.0302,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25545366291990595,
+      "grad_norm": 2.981813669204712,
+      "learning_rate": 1.5386247596313245e-05,
+      "loss": 0.0504,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25609550629407657,
+      "grad_norm": 2.449775457382202,
+      "learning_rate": 1.5372985876268153e-05,
+      "loss": 0.025,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2567373496682472,
+      "grad_norm": 0.42797595262527466,
+      "learning_rate": 1.5359724156223064e-05,
+      "loss": 0.0451,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2573791930424178,
+      "grad_norm": 1.5074821710586548,
+      "learning_rate": 1.5346462436177972e-05,
+      "loss": 0.0376,
+      "step": 4010
+    },
+    {
+      "epoch": 0.25802103641658847,
+      "grad_norm": 4.754857540130615,
+      "learning_rate": 1.5333200716132883e-05,
+      "loss": 0.0357,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2586628797907591,
+      "grad_norm": 4.347653388977051,
+      "learning_rate": 1.5319938996087794e-05,
+      "loss": 0.0509,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2593047231649297,
+      "grad_norm": 4.191555500030518,
+      "learning_rate": 1.5306677276042702e-05,
+      "loss": 0.0343,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2599465665391003,
+      "grad_norm": 3.382261037826538,
+      "learning_rate": 1.5293415555997617e-05,
+      "loss": 0.0535,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2605884099132709,
+      "grad_norm": 3.161862373352051,
+      "learning_rate": 1.5280153835952524e-05,
+      "loss": 0.0481,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26123025328744154,
+      "grad_norm": 1.660260558128357,
+      "learning_rate": 1.5266892115907435e-05,
+      "loss": 0.0297,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26187209666161215,
+      "grad_norm": 0.232618048787117,
+      "learning_rate": 1.5253630395862345e-05,
+      "loss": 0.0371,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26251394003578277,
+      "grad_norm": 3.75004243850708,
+      "learning_rate": 1.5240368675817254e-05,
+      "loss": 0.0469,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2631557834099534,
+      "grad_norm": 3.747248888015747,
+      "learning_rate": 1.5227106955772164e-05,
+      "loss": 0.0612,
+      "step": 4100
+    },
+    {
+      "epoch": 0.263797626784124,
+      "grad_norm": 9.171794891357422,
+      "learning_rate": 1.5213845235727077e-05,
+      "loss": 0.0599,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2644394701582946,
+      "grad_norm": 0.9723399877548218,
+      "learning_rate": 1.5200583515681986e-05,
+      "loss": 0.0275,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2650813135324652,
+      "grad_norm": 1.1156556606292725,
+      "learning_rate": 1.5187321795636896e-05,
+      "loss": 0.0306,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26572315690663584,
+      "grad_norm": 2.5860493183135986,
+      "learning_rate": 1.5174060075591805e-05,
+      "loss": 0.0321,
+      "step": 4140
+    },
+    {
+      "epoch": 0.26636500028080645,
+      "grad_norm": 5.233616828918457,
+      "learning_rate": 1.5160798355546715e-05,
+      "loss": 0.0463,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2670068436549771,
+      "grad_norm": 4.714346408843994,
+      "learning_rate": 1.5147536635501626e-05,
+      "loss": 0.029,
+      "step": 4160
+    },
+    {
+      "epoch": 0.26764868702914774,
+      "grad_norm": 5.325324058532715,
+      "learning_rate": 1.5134274915456535e-05,
+      "loss": 0.0323,
+      "step": 4170
+    },
+    {
+      "epoch": 0.26829053040331835,
+      "grad_norm": 2.1289708614349365,
+      "learning_rate": 1.5121013195411447e-05,
+      "loss": 0.0542,
+      "step": 4180
+    },
+    {
+      "epoch": 0.26893237377748896,
+      "grad_norm": 1.2925282716751099,
+      "learning_rate": 1.5107751475366356e-05,
+      "loss": 0.034,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2695742171516596,
+      "grad_norm": 0.7632297873497009,
+      "learning_rate": 1.5094489755321267e-05,
+      "loss": 0.0482,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2702160605258302,
+      "grad_norm": 1.8542927503585815,
+      "learning_rate": 1.5081228035276177e-05,
+      "loss": 0.0441,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2708579039000008,
+      "grad_norm": 1.7105222940444946,
+      "learning_rate": 1.5067966315231086e-05,
+      "loss": 0.0316,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2714997472741714,
+      "grad_norm": 0.39926835894584656,
+      "learning_rate": 1.5054704595185996e-05,
+      "loss": 0.0362,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27214159064834204,
+      "grad_norm": 1.7719646692276,
+      "learning_rate": 1.5041442875140908e-05,
+      "loss": 0.0351,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27278343402251265,
+      "grad_norm": 1.2764171361923218,
+      "learning_rate": 1.5028181155095818e-05,
+      "loss": 0.0344,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27342527739668326,
+      "grad_norm": 0.1933438628911972,
+      "learning_rate": 1.5014919435050727e-05,
+      "loss": 0.0301,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2740671207708539,
+      "grad_norm": 0.7615514397621155,
+      "learning_rate": 1.5001657715005637e-05,
+      "loss": 0.0457,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2747089641450245,
+      "grad_norm": 1.1360198259353638,
+      "learning_rate": 1.4988395994960548e-05,
+      "loss": 0.0301,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2753508075191951,
+      "grad_norm": 2.9905829429626465,
+      "learning_rate": 1.4975134274915458e-05,
+      "loss": 0.0341,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2759926508933657,
+      "grad_norm": 1.4750293493270874,
+      "learning_rate": 1.4961872554870367e-05,
+      "loss": 0.0333,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2766344942675364,
+      "grad_norm": 0.11904127895832062,
+      "learning_rate": 1.4948610834825278e-05,
+      "loss": 0.0293,
+      "step": 4310
+    },
+    {
+      "epoch": 0.277276337641707,
+      "grad_norm": 2.860715389251709,
+      "learning_rate": 1.493534911478019e-05,
+      "loss": 0.0262,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2779181810158776,
+      "grad_norm": 3.2877347469329834,
+      "learning_rate": 1.4922087394735099e-05,
+      "loss": 0.0259,
+      "step": 4330
+    },
+    {
+      "epoch": 0.27856002439004823,
+      "grad_norm": 0.1747274398803711,
+      "learning_rate": 1.4908825674690008e-05,
+      "loss": 0.0203,
+      "step": 4340
+    },
+    {
+      "epoch": 0.27920186776421885,
+      "grad_norm": 1.7592897415161133,
+      "learning_rate": 1.4895563954644918e-05,
+      "loss": 0.0332,
+      "step": 4350
+    },
+    {
+      "epoch": 0.27984371113838946,
+      "grad_norm": 3.454744338989258,
+      "learning_rate": 1.4882302234599827e-05,
+      "loss": 0.0387,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2804855545125601,
+      "grad_norm": 2.7865166664123535,
+      "learning_rate": 1.486904051455474e-05,
+      "loss": 0.026,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2811273978867307,
+      "grad_norm": 1.9861980676651,
+      "learning_rate": 1.485577879450965e-05,
+      "loss": 0.0349,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2817692412609013,
+      "grad_norm": 1.646071434020996,
+      "learning_rate": 1.4842517074464559e-05,
+      "loss": 0.0297,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2824110846350719,
+      "grad_norm": 0.5253218412399292,
+      "learning_rate": 1.482925535441947e-05,
+      "loss": 0.0311,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28305292800924253,
+      "grad_norm": 0.11210356652736664,
+      "learning_rate": 1.481599363437438e-05,
+      "loss": 0.0344,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28369477138341315,
+      "grad_norm": 6.261229038238525,
+      "learning_rate": 1.480273191432929e-05,
+      "loss": 0.0278,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28433661475758376,
+      "grad_norm": 0.5738747715950012,
+      "learning_rate": 1.4789470194284199e-05,
+      "loss": 0.038,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2849784581317544,
+      "grad_norm": 1.1812382936477661,
+      "learning_rate": 1.4776208474239112e-05,
+      "loss": 0.0408,
+      "step": 4440
+    },
+    {
+      "epoch": 0.285620301505925,
+      "grad_norm": 6.855234146118164,
+      "learning_rate": 1.4762946754194021e-05,
+      "loss": 0.0403,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28626214488009566,
+      "grad_norm": 0.6492354273796082,
+      "learning_rate": 1.474968503414893e-05,
+      "loss": 0.0487,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2869039882542663,
+      "grad_norm": 7.180405139923096,
+      "learning_rate": 1.473642331410384e-05,
+      "loss": 0.0401,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2875458316284369,
+      "grad_norm": 2.133906602859497,
+      "learning_rate": 1.472316159405875e-05,
+      "loss": 0.0323,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2881876750026075,
+      "grad_norm": 4.206572532653809,
+      "learning_rate": 1.470989987401366e-05,
+      "loss": 0.036,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2888295183767781,
+      "grad_norm": 4.9672417640686035,
+      "learning_rate": 1.469663815396857e-05,
+      "loss": 0.0449,
+      "step": 4500
+    },
+    {
+      "epoch": 0.28947136175094873,
+      "grad_norm": 3.2657997608184814,
+      "learning_rate": 1.4683376433923481e-05,
+      "loss": 0.0287,
+      "step": 4510
+    },
+    {
+      "epoch": 0.29011320512511934,
+      "grad_norm": 0.8293973207473755,
+      "learning_rate": 1.4670114713878392e-05,
+      "loss": 0.0585,
+      "step": 4520
+    },
+    {
+      "epoch": 0.29075504849928996,
+      "grad_norm": 3.678178071975708,
+      "learning_rate": 1.4656852993833302e-05,
+      "loss": 0.0351,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2913968918734606,
+      "grad_norm": 0.9538152813911438,
+      "learning_rate": 1.4643591273788211e-05,
+      "loss": 0.0496,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2920387352476312,
+      "grad_norm": 1.1784247159957886,
+      "learning_rate": 1.4630329553743121e-05,
+      "loss": 0.0353,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2926805786218018,
+      "grad_norm": 1.781537413597107,
+      "learning_rate": 1.461706783369803e-05,
+      "loss": 0.0378,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2933224219959724,
+      "grad_norm": 2.990455150604248,
+      "learning_rate": 1.4603806113652943e-05,
+      "loss": 0.0408,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29396426537014303,
+      "grad_norm": 5.851528167724609,
+      "learning_rate": 1.4590544393607853e-05,
+      "loss": 0.0328,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29460610874431364,
+      "grad_norm": 0.7276738882064819,
+      "learning_rate": 1.4577282673562762e-05,
+      "loss": 0.0546,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2952479521184843,
+      "grad_norm": 1.7559603452682495,
+      "learning_rate": 1.4564020953517672e-05,
+      "loss": 0.0305,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29588979549265493,
+      "grad_norm": 2.0152103900909424,
+      "learning_rate": 1.4550759233472583e-05,
+      "loss": 0.0302,
+      "step": 4610
+    },
+    {
+      "epoch": 0.29653163886682554,
+      "grad_norm": 0.8355568647384644,
+      "learning_rate": 1.4537497513427492e-05,
+      "loss": 0.0314,
+      "step": 4620
+    },
+    {
+      "epoch": 0.29717348224099616,
+      "grad_norm": 1.1026408672332764,
+      "learning_rate": 1.4524235793382402e-05,
+      "loss": 0.036,
+      "step": 4630
+    },
+    {
+      "epoch": 0.29781532561516677,
+      "grad_norm": 4.210587024688721,
+      "learning_rate": 1.4510974073337315e-05,
+      "loss": 0.0299,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2984571689893374,
+      "grad_norm": 2.904656171798706,
+      "learning_rate": 1.4497712353292224e-05,
+      "loss": 0.039,
+      "step": 4650
+    },
+    {
+      "epoch": 0.299099012363508,
+      "grad_norm": 3.1882083415985107,
+      "learning_rate": 1.4484450633247134e-05,
+      "loss": 0.0524,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2997408557376786,
+      "grad_norm": 1.4921969175338745,
+      "learning_rate": 1.4471188913202043e-05,
+      "loss": 0.0407,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3003826991118492,
+      "grad_norm": 0.5533859729766846,
+      "learning_rate": 1.4457927193156953e-05,
+      "loss": 0.0419,
+      "step": 4680
+    },
+    {
+      "epoch": 0.30102454248601984,
+      "grad_norm": 0.90987628698349,
+      "learning_rate": 1.4444665473111864e-05,
+      "loss": 0.0489,
+      "step": 4690
+    },
+    {
+      "epoch": 0.30166638586019046,
+      "grad_norm": 4.835458755493164,
+      "learning_rate": 1.4431403753066775e-05,
+      "loss": 0.0305,
+      "step": 4700
+    },
+    {
+      "epoch": 0.30230822923436107,
+      "grad_norm": 1.4399257898330688,
+      "learning_rate": 1.4418142033021684e-05,
+      "loss": 0.0261,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3029500726085317,
+      "grad_norm": 0.8075177073478699,
+      "learning_rate": 1.4404880312976594e-05,
+      "loss": 0.039,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3035919159827023,
+      "grad_norm": 0.26249435544013977,
+      "learning_rate": 1.4391618592931505e-05,
+      "loss": 0.0238,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3042337593568729,
+      "grad_norm": 2.859839677810669,
+      "learning_rate": 1.4378356872886415e-05,
+      "loss": 0.0348,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3048756027310436,
+      "grad_norm": 5.632012844085693,
+      "learning_rate": 1.4365095152841324e-05,
+      "loss": 0.0601,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3055174461052142,
+      "grad_norm": 0.5732155442237854,
+      "learning_rate": 1.4351833432796234e-05,
+      "loss": 0.0209,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3061592894793848,
+      "grad_norm": 3.579094171524048,
+      "learning_rate": 1.4338571712751146e-05,
+      "loss": 0.0273,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3068011328535554,
+      "grad_norm": 3.8557450771331787,
+      "learning_rate": 1.4325309992706056e-05,
+      "loss": 0.0309,
+      "step": 4780
+    },
+    {
+      "epoch": 0.30744297622772604,
+      "grad_norm": 1.1692198514938354,
+      "learning_rate": 1.4312048272660965e-05,
+      "loss": 0.0249,
+      "step": 4790
+    },
+    {
+      "epoch": 0.30808481960189665,
+      "grad_norm": 5.022751331329346,
+      "learning_rate": 1.4298786552615875e-05,
+      "loss": 0.0445,
+      "step": 4800
+    },
+    {
+      "epoch": 0.30872666297606727,
+      "grad_norm": 0.4565708041191101,
+      "learning_rate": 1.4285524832570784e-05,
+      "loss": 0.0185,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3093685063502379,
+      "grad_norm": 0.2485363483428955,
+      "learning_rate": 1.4272263112525695e-05,
+      "loss": 0.0334,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3100103497244085,
+      "grad_norm": 0.2797245681285858,
+      "learning_rate": 1.4259001392480607e-05,
+      "loss": 0.0408,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3106521930985791,
+      "grad_norm": 3.8306257724761963,
+      "learning_rate": 1.4245739672435516e-05,
+      "loss": 0.0486,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3112940364727497,
+      "grad_norm": 3.26303768157959,
+      "learning_rate": 1.4232477952390427e-05,
+      "loss": 0.0245,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31193587984692034,
+      "grad_norm": 0.26074540615081787,
+      "learning_rate": 1.4219216232345337e-05,
+      "loss": 0.0282,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31257772322109095,
+      "grad_norm": 4.135842800140381,
+      "learning_rate": 1.4205954512300246e-05,
+      "loss": 0.0367,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31321956659526157,
+      "grad_norm": 2.937664031982422,
+      "learning_rate": 1.4192692792255156e-05,
+      "loss": 0.0286,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31386140996943224,
+      "grad_norm": 0.9610151052474976,
+      "learning_rate": 1.4179431072210065e-05,
+      "loss": 0.0262,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31450325334360285,
+      "grad_norm": 1.348313570022583,
+      "learning_rate": 1.4166169352164978e-05,
+      "loss": 0.0471,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31514509671777347,
+      "grad_norm": 1.0628247261047363,
+      "learning_rate": 1.4152907632119888e-05,
+      "loss": 0.0278,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3157869400919441,
+      "grad_norm": 2.1277549266815186,
+      "learning_rate": 1.4139645912074797e-05,
+      "loss": 0.0422,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3164287834661147,
+      "grad_norm": 4.375251293182373,
+      "learning_rate": 1.4126384192029707e-05,
+      "loss": 0.0448,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3170706268402853,
+      "grad_norm": 2.4289703369140625,
+      "learning_rate": 1.4113122471984618e-05,
+      "loss": 0.0291,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3177124702144559,
+      "grad_norm": 1.17922043800354,
+      "learning_rate": 1.4099860751939527e-05,
+      "loss": 0.024,
+      "step": 4950
+    },
+    {
+      "epoch": 0.31835431358862654,
+      "grad_norm": 0.7155043482780457,
+      "learning_rate": 1.4086599031894437e-05,
+      "loss": 0.0164,
+      "step": 4960
+    },
+    {
+      "epoch": 0.31899615696279715,
+      "grad_norm": 2.958674907684326,
+      "learning_rate": 1.407333731184935e-05,
+      "loss": 0.0408,
+      "step": 4970
+    },
+    {
+      "epoch": 0.31963800033696776,
+      "grad_norm": 1.0707948207855225,
+      "learning_rate": 1.4060075591804259e-05,
+      "loss": 0.0351,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3202798437111384,
+      "grad_norm": 9.325459480285645,
+      "learning_rate": 1.4046813871759168e-05,
+      "loss": 0.0592,
+      "step": 4990
+    },
+    {
+      "epoch": 0.320921687085309,
+      "grad_norm": 0.5737670660018921,
+      "learning_rate": 1.4033552151714078e-05,
+      "loss": 0.0245,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3215635304594796,
+      "grad_norm": 1.7294018268585205,
+      "learning_rate": 1.4020290431668987e-05,
+      "loss": 0.0341,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3222053738336502,
+      "grad_norm": 3.3348608016967773,
+      "learning_rate": 1.4007028711623899e-05,
+      "loss": 0.0312,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32284721720782084,
+      "grad_norm": 1.2371169328689575,
+      "learning_rate": 1.399376699157881e-05,
+      "loss": 0.0314,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3234890605819915,
+      "grad_norm": 0.5545613765716553,
+      "learning_rate": 1.398050527153372e-05,
+      "loss": 0.0248,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3241309039561621,
+      "grad_norm": 0.1487692892551422,
+      "learning_rate": 1.3967243551488629e-05,
+      "loss": 0.0245,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32477274733033273,
+      "grad_norm": 5.844099044799805,
+      "learning_rate": 1.395398183144354e-05,
+      "loss": 0.0333,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32541459070450335,
+      "grad_norm": 1.8279471397399902,
+      "learning_rate": 1.394072011139845e-05,
+      "loss": 0.0327,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32605643407867396,
+      "grad_norm": 4.413207054138184,
+      "learning_rate": 1.3927458391353359e-05,
+      "loss": 0.0415,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3266982774528446,
+      "grad_norm": 4.457259178161621,
+      "learning_rate": 1.3914196671308268e-05,
+      "loss": 0.0412,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3273401208270152,
+      "grad_norm": 0.5590922236442566,
+      "learning_rate": 1.3900934951263181e-05,
+      "loss": 0.0333,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3279819642011858,
+      "grad_norm": 3.6390750408172607,
+      "learning_rate": 1.388767323121809e-05,
+      "loss": 0.0509,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3286238075753564,
+      "grad_norm": 1.944118618965149,
+      "learning_rate": 1.3874411511173e-05,
+      "loss": 0.0313,
+      "step": 5120
+    },
+    {
+      "epoch": 0.32926565094952703,
+      "grad_norm": 1.424323558807373,
+      "learning_rate": 1.386114979112791e-05,
+      "loss": 0.026,
+      "step": 5130
+    },
+    {
+      "epoch": 0.32990749432369765,
+      "grad_norm": 2.957878351211548,
+      "learning_rate": 1.384788807108282e-05,
+      "loss": 0.0238,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33054933769786826,
+      "grad_norm": 1.0304313898086548,
+      "learning_rate": 1.383462635103773e-05,
+      "loss": 0.0261,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3311911810720389,
+      "grad_norm": 2.29116153717041,
+      "learning_rate": 1.3821364630992641e-05,
+      "loss": 0.0362,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3318330244462095,
+      "grad_norm": 1.6272333860397339,
+      "learning_rate": 1.3808102910947551e-05,
+      "loss": 0.0172,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3324748678203801,
+      "grad_norm": 0.1753595918416977,
+      "learning_rate": 1.3794841190902462e-05,
+      "loss": 0.033,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3331167111945508,
+      "grad_norm": 0.24735307693481445,
+      "learning_rate": 1.3781579470857372e-05,
+      "loss": 0.0464,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3337585545687214,
+      "grad_norm": 1.140425205230713,
+      "learning_rate": 1.3768317750812281e-05,
+      "loss": 0.02,
+      "step": 5200
+    },
+    {
+      "epoch": 0.334400397942892,
+      "grad_norm": 4.770532131195068,
+      "learning_rate": 1.375505603076719e-05,
+      "loss": 0.0331,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3350422413170626,
+      "grad_norm": 1.9777387380599976,
+      "learning_rate": 1.37417943107221e-05,
+      "loss": 0.0236,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33568408469123323,
+      "grad_norm": 5.254188537597656,
+      "learning_rate": 1.3728532590677013e-05,
+      "loss": 0.0427,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33632592806540385,
+      "grad_norm": 1.9346436262130737,
+      "learning_rate": 1.3715270870631922e-05,
+      "loss": 0.0456,
+      "step": 5240
+    },
+    {
+      "epoch": 0.33696777143957446,
+      "grad_norm": 5.422431945800781,
+      "learning_rate": 1.3702009150586832e-05,
+      "loss": 0.043,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3376096148137451,
+      "grad_norm": 4.341948986053467,
+      "learning_rate": 1.3688747430541743e-05,
+      "loss": 0.0433,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3382514581879157,
+      "grad_norm": 0.7508522868156433,
+      "learning_rate": 1.3675485710496652e-05,
+      "loss": 0.0302,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3388933015620863,
+      "grad_norm": 2.358661651611328,
+      "learning_rate": 1.3662223990451562e-05,
+      "loss": 0.0213,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3395351449362569,
+      "grad_norm": 0.9036428332328796,
+      "learning_rate": 1.3648962270406473e-05,
+      "loss": 0.0288,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34017698831042753,
+      "grad_norm": 2.609917640686035,
+      "learning_rate": 1.3635700550361384e-05,
+      "loss": 0.0263,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34081883168459814,
+      "grad_norm": 0.23868480324745178,
+      "learning_rate": 1.3622438830316294e-05,
+      "loss": 0.0103,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34146067505876876,
+      "grad_norm": 5.040385723114014,
+      "learning_rate": 1.3609177110271203e-05,
+      "loss": 0.0375,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34210251843293943,
+      "grad_norm": 1.8747702836990356,
+      "learning_rate": 1.3595915390226113e-05,
+      "loss": 0.0234,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34274436180711004,
+      "grad_norm": 3.535433053970337,
+      "learning_rate": 1.3582653670181022e-05,
+      "loss": 0.0337,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34338620518128066,
+      "grad_norm": 1.0234447717666626,
+      "learning_rate": 1.3569391950135933e-05,
+      "loss": 0.0278,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34402804855545127,
+      "grad_norm": 4.323075771331787,
+      "learning_rate": 1.3556130230090845e-05,
+      "loss": 0.0292,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3446698919296219,
+      "grad_norm": 6.044609069824219,
+      "learning_rate": 1.3542868510045754e-05,
+      "loss": 0.0336,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3453117353037925,
+      "grad_norm": 0.4520716369152069,
+      "learning_rate": 1.3529606790000665e-05,
+      "loss": 0.0386,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3459535786779631,
+      "grad_norm": 1.0738428831100464,
+      "learning_rate": 1.3516345069955575e-05,
+      "loss": 0.0349,
+      "step": 5390
+    },
+    {
+      "epoch": 0.34659542205213373,
+      "grad_norm": 4.519723415374756,
+      "learning_rate": 1.3503083349910484e-05,
+      "loss": 0.0434,
+      "step": 5400
+    },
+    {
+      "epoch": 0.34723726542630434,
+      "grad_norm": 1.12299382686615,
+      "learning_rate": 1.3489821629865394e-05,
+      "loss": 0.0409,
+      "step": 5410
+    },
+    {
+      "epoch": 0.34787910880047496,
+      "grad_norm": 1.1750973463058472,
+      "learning_rate": 1.3476559909820303e-05,
+      "loss": 0.0175,
+      "step": 5420
+    },
+    {
+      "epoch": 0.34852095217464557,
+      "grad_norm": 0.6646749973297119,
+      "learning_rate": 1.3463298189775216e-05,
+      "loss": 0.0244,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3491627955488162,
+      "grad_norm": 0.40298768877983093,
+      "learning_rate": 1.3450036469730125e-05,
+      "loss": 0.0322,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3498046389229868,
+      "grad_norm": 5.02630090713501,
+      "learning_rate": 1.3436774749685035e-05,
+      "loss": 0.0469,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3504464822971574,
+      "grad_norm": 0.8722269535064697,
+      "learning_rate": 1.3423513029639944e-05,
+      "loss": 0.0376,
+      "step": 5460
+    },
+    {
+      "epoch": 0.351088325671328,
+      "grad_norm": 0.7096695303916931,
+      "learning_rate": 1.3410251309594856e-05,
+      "loss": 0.0206,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3517301690454987,
+      "grad_norm": 0.901951014995575,
+      "learning_rate": 1.3396989589549765e-05,
+      "loss": 0.0215,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3523720124196693,
+      "grad_norm": 0.25191473960876465,
+      "learning_rate": 1.3383727869504676e-05,
+      "loss": 0.0231,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3530138557938399,
+      "grad_norm": 4.8230180740356445,
+      "learning_rate": 1.3370466149459587e-05,
+      "loss": 0.0414,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35365569916801054,
+      "grad_norm": 3.088327169418335,
+      "learning_rate": 1.3357204429414497e-05,
+      "loss": 0.0278,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35429754254218115,
+      "grad_norm": 1.8572773933410645,
+      "learning_rate": 1.3343942709369406e-05,
+      "loss": 0.0201,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35493938591635177,
+      "grad_norm": 4.562084674835205,
+      "learning_rate": 1.3330680989324316e-05,
+      "loss": 0.0196,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3555812292905224,
+      "grad_norm": 0.2433650940656662,
+      "learning_rate": 1.3317419269279225e-05,
+      "loss": 0.0395,
+      "step": 5540
+    },
+    {
+      "epoch": 0.356223072664693,
+      "grad_norm": 3.763524055480957,
+      "learning_rate": 1.3304157549234137e-05,
+      "loss": 0.0419,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3568649160388636,
+      "grad_norm": 1.4221765995025635,
+      "learning_rate": 1.3290895829189048e-05,
+      "loss": 0.0238,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3575067594130342,
+      "grad_norm": 0.5821696519851685,
+      "learning_rate": 1.3277634109143957e-05,
+      "loss": 0.0279,
+      "step": 5570
+    },
+    {
+      "epoch": 0.35814860278720484,
+      "grad_norm": 2.7642388343811035,
+      "learning_rate": 1.3264372389098867e-05,
+      "loss": 0.0429,
+      "step": 5580
+    },
+    {
+      "epoch": 0.35879044616137545,
+      "grad_norm": 5.3998565673828125,
+      "learning_rate": 1.3251110669053778e-05,
+      "loss": 0.031,
+      "step": 5590
+    },
+    {
+      "epoch": 0.35943228953554607,
+      "grad_norm": 0.39380407333374023,
+      "learning_rate": 1.3237848949008687e-05,
+      "loss": 0.0312,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3600741329097167,
+      "grad_norm": 0.33455953001976013,
+      "learning_rate": 1.3224587228963597e-05,
+      "loss": 0.0281,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36071597628388735,
+      "grad_norm": 1.693926453590393,
+      "learning_rate": 1.321132550891851e-05,
+      "loss": 0.0218,
+      "step": 5620
+    },
+    {
+      "epoch": 0.36135781965805797,
+      "grad_norm": 0.2952551245689392,
+      "learning_rate": 1.3198063788873419e-05,
+      "loss": 0.0291,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3619996630322286,
+      "grad_norm": 1.343604326248169,
+      "learning_rate": 1.3184802068828329e-05,
+      "loss": 0.0385,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3626415064063992,
+      "grad_norm": 0.8492736220359802,
+      "learning_rate": 1.3171540348783238e-05,
+      "loss": 0.0367,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3632833497805698,
+      "grad_norm": 1.1233781576156616,
+      "learning_rate": 1.3158278628738148e-05,
+      "loss": 0.0181,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3639251931547404,
+      "grad_norm": 0.07301205396652222,
+      "learning_rate": 1.3145016908693059e-05,
+      "loss": 0.0211,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36456703652891104,
+      "grad_norm": 2.1094229221343994,
+      "learning_rate": 1.3131755188647968e-05,
+      "loss": 0.0218,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36520887990308165,
+      "grad_norm": 0.391240656375885,
+      "learning_rate": 1.311849346860288e-05,
+      "loss": 0.0343,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36585072327725227,
+      "grad_norm": 0.3236871063709259,
+      "learning_rate": 1.3105231748557789e-05,
+      "loss": 0.0285,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3664925666514229,
+      "grad_norm": 0.9702367186546326,
+      "learning_rate": 1.30919700285127e-05,
+      "loss": 0.0343,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3671344100255935,
+      "grad_norm": 0.8269650936126709,
+      "learning_rate": 1.307870830846761e-05,
+      "loss": 0.0289,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3677762533997641,
+      "grad_norm": 1.4295783042907715,
+      "learning_rate": 1.3065446588422519e-05,
+      "loss": 0.0367,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3684180967739347,
+      "grad_norm": 1.9564656019210815,
+      "learning_rate": 1.3052184868377428e-05,
+      "loss": 0.0265,
+      "step": 5740
+    },
+    {
+      "epoch": 0.36905994014810534,
+      "grad_norm": 0.7807095050811768,
+      "learning_rate": 1.3038923148332341e-05,
+      "loss": 0.0476,
+      "step": 5750
+    },
+    {
+      "epoch": 0.36970178352227595,
+      "grad_norm": 1.0274103879928589,
+      "learning_rate": 1.302566142828725e-05,
+      "loss": 0.028,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3703436268964466,
+      "grad_norm": 3.565110921859741,
+      "learning_rate": 1.301239970824216e-05,
+      "loss": 0.0322,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37098547027061723,
+      "grad_norm": 1.3256715536117554,
+      "learning_rate": 1.299913798819707e-05,
+      "loss": 0.0332,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37162731364478785,
+      "grad_norm": 2.2184195518493652,
+      "learning_rate": 1.2985876268151981e-05,
+      "loss": 0.0281,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37226915701895846,
+      "grad_norm": 1.855116367340088,
+      "learning_rate": 1.297261454810689e-05,
+      "loss": 0.0434,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3729110003931291,
+      "grad_norm": 0.612144947052002,
+      "learning_rate": 1.29593528280618e-05,
+      "loss": 0.0361,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3735528437672997,
+      "grad_norm": 2.8559155464172363,
+      "learning_rate": 1.2946091108016711e-05,
+      "loss": 0.0387,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3741946871414703,
+      "grad_norm": 0.6401782631874084,
+      "learning_rate": 1.2932829387971622e-05,
+      "loss": 0.0418,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3748365305156409,
+      "grad_norm": 2.0480880737304688,
+      "learning_rate": 1.2919567667926532e-05,
+      "loss": 0.0366,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37547837388981153,
+      "grad_norm": 0.23915986716747284,
+      "learning_rate": 1.2906305947881441e-05,
+      "loss": 0.0304,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37612021726398215,
+      "grad_norm": 1.4414054155349731,
+      "learning_rate": 1.289304422783635e-05,
+      "loss": 0.0254,
+      "step": 5860
+    },
+    {
+      "epoch": 0.37676206063815276,
+      "grad_norm": 1.5166568756103516,
+      "learning_rate": 1.287978250779126e-05,
+      "loss": 0.05,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3774039040123234,
+      "grad_norm": 1.248548984527588,
+      "learning_rate": 1.2866520787746171e-05,
+      "loss": 0.0259,
+      "step": 5880
+    },
+    {
+      "epoch": 0.378045747386494,
+      "grad_norm": 1.0585907697677612,
+      "learning_rate": 1.2853259067701082e-05,
+      "loss": 0.0313,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3786875907606646,
+      "grad_norm": 3.518057346343994,
+      "learning_rate": 1.2839997347655992e-05,
+      "loss": 0.0359,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3793294341348352,
+      "grad_norm": 5.588145732879639,
+      "learning_rate": 1.2826735627610901e-05,
+      "loss": 0.0443,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3799712775090059,
+      "grad_norm": 3.4836244583129883,
+      "learning_rate": 1.2813473907565813e-05,
+      "loss": 0.0302,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3806131208831765,
+      "grad_norm": 0.4826532006263733,
+      "learning_rate": 1.2800212187520722e-05,
+      "loss": 0.0215,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3812549642573471,
+      "grad_norm": 1.2696845531463623,
+      "learning_rate": 1.2786950467475632e-05,
+      "loss": 0.0239,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38189680763151773,
+      "grad_norm": 1.2803621292114258,
+      "learning_rate": 1.2773688747430544e-05,
+      "loss": 0.0291,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38253865100568835,
+      "grad_norm": 2.4219958782196045,
+      "learning_rate": 1.2760427027385454e-05,
+      "loss": 0.0381,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38318049437985896,
+      "grad_norm": 3.6280503273010254,
+      "learning_rate": 1.2747165307340363e-05,
+      "loss": 0.0316,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3838223377540296,
+      "grad_norm": 1.5109280347824097,
+      "learning_rate": 1.2733903587295273e-05,
+      "loss": 0.0226,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3844641811282002,
+      "grad_norm": 1.0172524452209473,
+      "learning_rate": 1.2720641867250182e-05,
+      "loss": 0.0219,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3851060245023708,
+      "grad_norm": 1.4693063497543335,
+      "learning_rate": 1.2707380147205094e-05,
+      "loss": 0.023,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3857478678765414,
+      "grad_norm": 1.2674403190612793,
+      "learning_rate": 1.2694118427160003e-05,
+      "loss": 0.0265,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38638971125071203,
+      "grad_norm": 1.6904219388961792,
+      "learning_rate": 1.2680856707114914e-05,
+      "loss": 0.0252,
+      "step": 6020
+    },
+    {
+      "epoch": 0.38703155462488265,
+      "grad_norm": 5.279049873352051,
+      "learning_rate": 1.2667594987069824e-05,
+      "loss": 0.0386,
+      "step": 6030
+    },
+    {
+      "epoch": 0.38767339799905326,
+      "grad_norm": 0.1736469864845276,
+      "learning_rate": 1.2654333267024735e-05,
+      "loss": 0.0279,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3883152413732239,
+      "grad_norm": 0.937554657459259,
+      "learning_rate": 1.2641071546979644e-05,
+      "loss": 0.0373,
+      "step": 6050
+    },
+    {
+      "epoch": 0.38895708474739454,
+      "grad_norm": 5.91543436050415,
+      "learning_rate": 1.2627809826934554e-05,
+      "loss": 0.042,
+      "step": 6060
+    },
+    {
+      "epoch": 0.38959892812156516,
+      "grad_norm": 2.749023675918579,
+      "learning_rate": 1.2614548106889463e-05,
+      "loss": 0.0364,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39024077149573577,
+      "grad_norm": 2.4227447509765625,
+      "learning_rate": 1.2601286386844376e-05,
+      "loss": 0.0218,
+      "step": 6080
+    },
+    {
+      "epoch": 0.3908826148699064,
+      "grad_norm": 0.1743806004524231,
+      "learning_rate": 1.2588024666799286e-05,
+      "loss": 0.0462,
+      "step": 6090
+    },
+    {
+      "epoch": 0.391524458244077,
+      "grad_norm": 4.252072811126709,
+      "learning_rate": 1.2574762946754195e-05,
+      "loss": 0.0476,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3921663016182476,
+      "grad_norm": 1.6621201038360596,
+      "learning_rate": 1.2561501226709105e-05,
+      "loss": 0.0275,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39280814499241823,
+      "grad_norm": 1.6763057708740234,
+      "learning_rate": 1.2548239506664016e-05,
+      "loss": 0.0312,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39344998836658884,
+      "grad_norm": 1.4106769561767578,
+      "learning_rate": 1.2534977786618925e-05,
+      "loss": 0.0168,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39409183174075946,
+      "grad_norm": 1.142337679862976,
+      "learning_rate": 1.2521716066573835e-05,
+      "loss": 0.0296,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39473367511493007,
+      "grad_norm": 1.2936738729476929,
+      "learning_rate": 1.2508454346528746e-05,
+      "loss": 0.0378,
+      "step": 6150
+    },
+    {
+      "epoch": 0.3953755184891007,
+      "grad_norm": 1.678970456123352,
+      "learning_rate": 1.2495192626483657e-05,
+      "loss": 0.0393,
+      "step": 6160
+    },
+    {
+      "epoch": 0.3960173618632713,
+      "grad_norm": 0.9166494011878967,
+      "learning_rate": 1.2481930906438567e-05,
+      "loss": 0.0265,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3966592052374419,
+      "grad_norm": 0.7071855068206787,
+      "learning_rate": 1.2468669186393476e-05,
+      "loss": 0.0177,
+      "step": 6180
+    },
+    {
+      "epoch": 0.39730104861161253,
+      "grad_norm": 1.9155601263046265,
+      "learning_rate": 1.2455407466348385e-05,
+      "loss": 0.0231,
+      "step": 6190
+    },
+    {
+      "epoch": 0.39794289198578314,
+      "grad_norm": 3.443380832672119,
+      "learning_rate": 1.2442145746303295e-05,
+      "loss": 0.0311,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3985847353599538,
+      "grad_norm": 3.7133545875549316,
+      "learning_rate": 1.2428884026258208e-05,
+      "loss": 0.0298,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3992265787341244,
+      "grad_norm": 0.8525023460388184,
+      "learning_rate": 1.2415622306213117e-05,
+      "loss": 0.0372,
+      "step": 6220
+    },
+    {
+      "epoch": 0.39986842210829504,
+      "grad_norm": 2.423804759979248,
+      "learning_rate": 1.2402360586168027e-05,
+      "loss": 0.0219,
+      "step": 6230
+    },
+    {
+      "epoch": 0.40051026548246565,
+      "grad_norm": 0.8703656792640686,
+      "learning_rate": 1.2389098866122938e-05,
+      "loss": 0.0161,
+      "step": 6240
+    },
+    {
+      "epoch": 0.40115210885663627,
+      "grad_norm": 4.453926086425781,
+      "learning_rate": 1.2375837146077847e-05,
+      "loss": 0.0337,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4017939522308069,
+      "grad_norm": 1.66751229763031,
+      "learning_rate": 1.2362575426032757e-05,
+      "loss": 0.0248,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4024357956049775,
+      "grad_norm": 3.351344108581543,
+      "learning_rate": 1.2349313705987666e-05,
+      "loss": 0.0435,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4030776389791481,
+      "grad_norm": 6.344804286956787,
+      "learning_rate": 1.233605198594258e-05,
+      "loss": 0.0325,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4037194823533187,
+      "grad_norm": 2.798166513442993,
+      "learning_rate": 1.2322790265897489e-05,
+      "loss": 0.0413,
+      "step": 6290
+    },
+    {
+      "epoch": 0.40436132572748934,
+      "grad_norm": 1.518452763557434,
+      "learning_rate": 1.2309528545852398e-05,
+      "loss": 0.0467,
+      "step": 6300
+    },
+    {
+      "epoch": 0.40500316910165995,
+      "grad_norm": 0.12372729927301407,
+      "learning_rate": 1.2296266825807308e-05,
+      "loss": 0.0333,
+      "step": 6310
+    },
+    {
+      "epoch": 0.40564501247583057,
+      "grad_norm": 3.486853837966919,
+      "learning_rate": 1.2283005105762217e-05,
+      "loss": 0.0283,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4062868558500012,
+      "grad_norm": 1.3288218975067139,
+      "learning_rate": 1.2269743385717128e-05,
+      "loss": 0.0201,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4069286992241718,
+      "grad_norm": 1.4961128234863281,
+      "learning_rate": 1.2256481665672038e-05,
+      "loss": 0.0276,
+      "step": 6340
+    },
+    {
+      "epoch": 0.40757054259834247,
+      "grad_norm": 0.3987443447113037,
+      "learning_rate": 1.2243219945626949e-05,
+      "loss": 0.0363,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4082123859725131,
+      "grad_norm": 0.6574511528015137,
+      "learning_rate": 1.222995822558186e-05,
+      "loss": 0.0355,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4088542293466837,
+      "grad_norm": 5.984050750732422,
+      "learning_rate": 1.221669650553677e-05,
+      "loss": 0.028,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4094960727208543,
+      "grad_norm": 0.4951924681663513,
+      "learning_rate": 1.2203434785491679e-05,
+      "loss": 0.0343,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4101379160950249,
+      "grad_norm": 0.14430026710033417,
+      "learning_rate": 1.2190173065446589e-05,
+      "loss": 0.0275,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41077975946919554,
+      "grad_norm": 1.4995652437210083,
+      "learning_rate": 1.2176911345401498e-05,
+      "loss": 0.0282,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41142160284336615,
+      "grad_norm": 0.6720727682113647,
+      "learning_rate": 1.2163649625356411e-05,
+      "loss": 0.019,
+      "step": 6410
+    },
+    {
+      "epoch": 0.41206344621753677,
+      "grad_norm": 3.248645544052124,
+      "learning_rate": 1.215038790531132e-05,
+      "loss": 0.0251,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4127052895917074,
+      "grad_norm": 0.5490819215774536,
+      "learning_rate": 1.213712618526623e-05,
+      "loss": 0.0496,
+      "step": 6430
+    },
+    {
+      "epoch": 0.413347132965878,
+      "grad_norm": 1.7990186214447021,
+      "learning_rate": 1.212386446522114e-05,
+      "loss": 0.0496,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4139889763400486,
+      "grad_norm": 0.8325052261352539,
+      "learning_rate": 1.211060274517605e-05,
+      "loss": 0.0334,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4146308197142192,
+      "grad_norm": 2.4705615043640137,
+      "learning_rate": 1.209734102513096e-05,
+      "loss": 0.0259,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41527266308838984,
+      "grad_norm": 3.4971210956573486,
+      "learning_rate": 1.208407930508587e-05,
+      "loss": 0.024,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41591450646256045,
+      "grad_norm": 0.32653504610061646,
+      "learning_rate": 1.2070817585040782e-05,
+      "loss": 0.0478,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41655634983673107,
+      "grad_norm": 0.8252073526382446,
+      "learning_rate": 1.2057555864995692e-05,
+      "loss": 0.0242,
+      "step": 6490
+    },
+    {
+      "epoch": 0.41719819321090174,
+      "grad_norm": 0.4332980513572693,
+      "learning_rate": 1.2044294144950601e-05,
+      "loss": 0.0168,
+      "step": 6500
+    },
+    {
+      "epoch": 0.41784003658507235,
+      "grad_norm": 1.0085843801498413,
+      "learning_rate": 1.203103242490551e-05,
+      "loss": 0.0286,
+      "step": 6510
+    },
+    {
+      "epoch": 0.41848187995924296,
+      "grad_norm": 0.031135300174355507,
+      "learning_rate": 1.201777070486042e-05,
+      "loss": 0.0331,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4191237233334136,
+      "grad_norm": 1.3585585355758667,
+      "learning_rate": 1.2004508984815331e-05,
+      "loss": 0.0313,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4197655667075842,
+      "grad_norm": 0.3975318968296051,
+      "learning_rate": 1.1991247264770243e-05,
+      "loss": 0.0424,
+      "step": 6540
+    },
+    {
+      "epoch": 0.4204074100817548,
+      "grad_norm": 1.7473927736282349,
+      "learning_rate": 1.1977985544725152e-05,
+      "loss": 0.0311,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4210492534559254,
+      "grad_norm": 0.8991397023200989,
+      "learning_rate": 1.1964723824680062e-05,
+      "loss": 0.0262,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42169109683009603,
+      "grad_norm": 0.20126798748970032,
+      "learning_rate": 1.1951462104634973e-05,
+      "loss": 0.0364,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42233294020426665,
+      "grad_norm": 1.0025629997253418,
+      "learning_rate": 1.1938200384589882e-05,
+      "loss": 0.0209,
+      "step": 6580
+    },
+    {
+      "epoch": 0.42297478357843726,
+      "grad_norm": 1.4650754928588867,
+      "learning_rate": 1.1924938664544792e-05,
+      "loss": 0.0331,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4236166269526079,
+      "grad_norm": 0.48315685987472534,
+      "learning_rate": 1.1911676944499701e-05,
+      "loss": 0.0347,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4242584703267785,
+      "grad_norm": 6.905628681182861,
+      "learning_rate": 1.1898415224454614e-05,
+      "loss": 0.03,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4249003137009491,
+      "grad_norm": 3.8432912826538086,
+      "learning_rate": 1.1885153504409524e-05,
+      "loss": 0.0229,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4255421570751197,
+      "grad_norm": 0.7360489964485168,
+      "learning_rate": 1.1871891784364433e-05,
+      "loss": 0.0254,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42618400044929033,
+      "grad_norm": 0.9641789793968201,
+      "learning_rate": 1.1858630064319342e-05,
+      "loss": 0.0311,
+      "step": 6640
+    },
+    {
+      "epoch": 0.426825843823461,
+      "grad_norm": 0.14324475824832916,
+      "learning_rate": 1.1845368344274254e-05,
+      "loss": 0.0276,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4274676871976316,
+      "grad_norm": 0.7097848653793335,
+      "learning_rate": 1.1832106624229163e-05,
+      "loss": 0.0229,
+      "step": 6660
+    },
+    {
+      "epoch": 0.42810953057180223,
+      "grad_norm": 5.492007255554199,
+      "learning_rate": 1.1818844904184074e-05,
+      "loss": 0.0476,
+      "step": 6670
+    },
+    {
+      "epoch": 0.42875137394597285,
+      "grad_norm": 3.966278076171875,
+      "learning_rate": 1.1805583184138984e-05,
+      "loss": 0.0393,
+      "step": 6680
+    },
+    {
+      "epoch": 0.42939321732014346,
+      "grad_norm": 0.7945524454116821,
+      "learning_rate": 1.1792321464093895e-05,
+      "loss": 0.0422,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4300350606943141,
+      "grad_norm": 0.11064202338457108,
+      "learning_rate": 1.1779059744048804e-05,
+      "loss": 0.0339,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4306769040684847,
+      "grad_norm": 0.8005304336547852,
+      "learning_rate": 1.1765798024003714e-05,
+      "loss": 0.0253,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4313187474426553,
+      "grad_norm": 1.859842300415039,
+      "learning_rate": 1.1752536303958623e-05,
+      "loss": 0.0226,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4319605908168259,
+      "grad_norm": 1.9041091203689575,
+      "learning_rate": 1.1739274583913533e-05,
+      "loss": 0.041,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43260243419099653,
+      "grad_norm": 0.13686195015907288,
+      "learning_rate": 1.1726012863868446e-05,
+      "loss": 0.0236,
+      "step": 6740
+    },
+    {
+      "epoch": 0.43324427756516715,
+      "grad_norm": 0.42238569259643555,
+      "learning_rate": 1.1712751143823355e-05,
+      "loss": 0.0488,
+      "step": 6750
+    },
+    {
+      "epoch": 0.43388612093933776,
+      "grad_norm": 0.4500075578689575,
+      "learning_rate": 1.1699489423778265e-05,
+      "loss": 0.0202,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4345279643135084,
+      "grad_norm": 0.783301055431366,
+      "learning_rate": 1.1686227703733176e-05,
+      "loss": 0.0174,
+      "step": 6770
+    },
+    {
+      "epoch": 0.435169807687679,
+      "grad_norm": 0.05518011748790741,
+      "learning_rate": 1.1672965983688085e-05,
+      "loss": 0.0476,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43581165106184966,
+      "grad_norm": 0.25636357069015503,
+      "learning_rate": 1.1659704263642995e-05,
+      "loss": 0.0358,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4364534944360203,
+      "grad_norm": 0.5332306623458862,
+      "learning_rate": 1.1646442543597904e-05,
+      "loss": 0.0298,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4370953378101909,
+      "grad_norm": 6.301130771636963,
+      "learning_rate": 1.1633180823552817e-05,
+      "loss": 0.0344,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4377371811843615,
+      "grad_norm": 1.3313417434692383,
+      "learning_rate": 1.1619919103507727e-05,
+      "loss": 0.026,
+      "step": 6820
+    },
+    {
+      "epoch": 0.4383790245585321,
+      "grad_norm": 2.344592332839966,
+      "learning_rate": 1.1606657383462636e-05,
+      "loss": 0.0378,
+      "step": 6830
+    },
+    {
+      "epoch": 0.43902086793270273,
+      "grad_norm": 5.8374857902526855,
+      "learning_rate": 1.1593395663417546e-05,
+      "loss": 0.0327,
+      "step": 6840
+    },
+    {
+      "epoch": 0.43966271130687334,
+      "grad_norm": 1.945421814918518,
+      "learning_rate": 1.1580133943372455e-05,
+      "loss": 0.0367,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44030455468104396,
+      "grad_norm": 1.068772315979004,
+      "learning_rate": 1.1566872223327366e-05,
+      "loss": 0.0215,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4409463980552146,
+      "grad_norm": 0.36997637152671814,
+      "learning_rate": 1.1553610503282277e-05,
+      "loss": 0.0323,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4415882414293852,
+      "grad_norm": 0.7297510504722595,
+      "learning_rate": 1.1540348783237187e-05,
+      "loss": 0.0177,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4422300848035558,
+      "grad_norm": 0.4879355728626251,
+      "learning_rate": 1.1527087063192098e-05,
+      "loss": 0.0347,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4428719281777264,
+      "grad_norm": 2.4448482990264893,
+      "learning_rate": 1.1513825343147008e-05,
+      "loss": 0.0291,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44351377155189703,
+      "grad_norm": 0.1498849242925644,
+      "learning_rate": 1.1500563623101917e-05,
+      "loss": 0.0154,
+      "step": 6910
+    },
+    {
+      "epoch": 0.44415561492606764,
+      "grad_norm": 0.27776220440864563,
+      "learning_rate": 1.1487301903056827e-05,
+      "loss": 0.0306,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44479745830023826,
+      "grad_norm": 0.4810338020324707,
+      "learning_rate": 1.1474040183011736e-05,
+      "loss": 0.0216,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4454393016744089,
+      "grad_norm": 0.4320123493671417,
+      "learning_rate": 1.1460778462966649e-05,
+      "loss": 0.0418,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44608114504857954,
+      "grad_norm": 1.5858159065246582,
+      "learning_rate": 1.1447516742921558e-05,
+      "loss": 0.0277,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44672298842275016,
+      "grad_norm": 1.1560159921646118,
+      "learning_rate": 1.1434255022876468e-05,
+      "loss": 0.0232,
+      "step": 6960
+    },
+    {
+      "epoch": 0.44736483179692077,
+      "grad_norm": 0.48158320784568787,
+      "learning_rate": 1.1420993302831377e-05,
+      "loss": 0.0308,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4480066751710914,
+      "grad_norm": 0.22836768627166748,
+      "learning_rate": 1.1407731582786288e-05,
+      "loss": 0.0287,
+      "step": 6980
+    },
+    {
+      "epoch": 0.448648518545262,
+      "grad_norm": 0.2631066143512726,
+      "learning_rate": 1.1394469862741198e-05,
+      "loss": 0.0289,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4492903619194326,
+      "grad_norm": 3.866243362426758,
+      "learning_rate": 1.1381208142696109e-05,
+      "loss": 0.0442,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4499322052936032,
+      "grad_norm": 2.3988699913024902,
+      "learning_rate": 1.1367946422651019e-05,
+      "loss": 0.0123,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45057404866777384,
+      "grad_norm": 3.825685739517212,
+      "learning_rate": 1.135468470260593e-05,
+      "loss": 0.0354,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45121589204194446,
+      "grad_norm": 5.321695804595947,
+      "learning_rate": 1.134142298256084e-05,
+      "loss": 0.0204,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45185773541611507,
+      "grad_norm": 3.1179020404815674,
+      "learning_rate": 1.1328161262515749e-05,
+      "loss": 0.0323,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4524995787902857,
+      "grad_norm": 4.537142276763916,
+      "learning_rate": 1.1314899542470658e-05,
+      "loss": 0.027,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4531414221644563,
+      "grad_norm": 0.16674059629440308,
+      "learning_rate": 1.1301637822425568e-05,
+      "loss": 0.0271,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4537832655386269,
+      "grad_norm": 0.692891001701355,
+      "learning_rate": 1.128837610238048e-05,
+      "loss": 0.0207,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4544251089127976,
+      "grad_norm": 1.5163105726242065,
+      "learning_rate": 1.127511438233539e-05,
+      "loss": 0.0339,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4550669522869682,
+      "grad_norm": 0.9091341495513916,
+      "learning_rate": 1.12618526622903e-05,
+      "loss": 0.0243,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4557087956611388,
+      "grad_norm": 2.6304235458374023,
+      "learning_rate": 1.124859094224521e-05,
+      "loss": 0.024,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4563506390353094,
+      "grad_norm": 5.431898593902588,
+      "learning_rate": 1.123532922220012e-05,
+      "loss": 0.0479,
+      "step": 7110
+    },
+    {
+      "epoch": 0.45699248240948004,
+      "grad_norm": 0.8807118535041809,
+      "learning_rate": 1.122206750215503e-05,
+      "loss": 0.0181,
+      "step": 7120
+    },
+    {
+      "epoch": 0.45763432578365065,
+      "grad_norm": 1.353542447090149,
+      "learning_rate": 1.120880578210994e-05,
+      "loss": 0.0156,
+      "step": 7130
+    },
+    {
+      "epoch": 0.45827616915782127,
+      "grad_norm": 0.13024044036865234,
+      "learning_rate": 1.1195544062064852e-05,
+      "loss": 0.0222,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4589180125319919,
+      "grad_norm": 0.5227543711662292,
+      "learning_rate": 1.1182282342019761e-05,
+      "loss": 0.0295,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4595598559061625,
+      "grad_norm": 4.043930530548096,
+      "learning_rate": 1.1169020621974671e-05,
+      "loss": 0.0254,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4602016992803331,
+      "grad_norm": 1.1538896560668945,
+      "learning_rate": 1.115575890192958e-05,
+      "loss": 0.0218,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4608435426545037,
+      "grad_norm": 1.619653344154358,
+      "learning_rate": 1.114249718188449e-05,
+      "loss": 0.0327,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46148538602867434,
+      "grad_norm": 2.511518955230713,
+      "learning_rate": 1.1129235461839401e-05,
+      "loss": 0.0326,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46212722940284495,
+      "grad_norm": 0.5349030494689941,
+      "learning_rate": 1.1115973741794312e-05,
+      "loss": 0.0213,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46276907277701557,
+      "grad_norm": 1.1988285779953003,
+      "learning_rate": 1.1102712021749222e-05,
+      "loss": 0.0281,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4634109161511862,
+      "grad_norm": 0.8555197715759277,
+      "learning_rate": 1.1089450301704133e-05,
+      "loss": 0.0349,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46405275952535685,
+      "grad_norm": 1.372574806213379,
+      "learning_rate": 1.1076188581659042e-05,
+      "loss": 0.0395,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46469460289952746,
+      "grad_norm": 0.7914135456085205,
+      "learning_rate": 1.1062926861613952e-05,
+      "loss": 0.0234,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4653364462736981,
+      "grad_norm": 0.6850894093513489,
+      "learning_rate": 1.1049665141568861e-05,
+      "loss": 0.0229,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4659782896478687,
+      "grad_norm": 0.6751859188079834,
+      "learning_rate": 1.1036403421523774e-05,
+      "loss": 0.0317,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4666201330220393,
+      "grad_norm": 0.9388216137886047,
+      "learning_rate": 1.1023141701478684e-05,
+      "loss": 0.0339,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4672619763962099,
+      "grad_norm": 0.6718018651008606,
+      "learning_rate": 1.1009879981433593e-05,
+      "loss": 0.0233,
+      "step": 7280
+    },
+    {
+      "epoch": 0.46790381977038054,
+      "grad_norm": 0.6536454558372498,
+      "learning_rate": 1.0996618261388503e-05,
+      "loss": 0.028,
+      "step": 7290
+    },
+    {
+      "epoch": 0.46854566314455115,
+      "grad_norm": 0.9818311333656311,
+      "learning_rate": 1.0983356541343412e-05,
+      "loss": 0.0244,
+      "step": 7300
+    },
+    {
+      "epoch": 0.46918750651872176,
+      "grad_norm": 4.402950286865234,
+      "learning_rate": 1.0970094821298323e-05,
+      "loss": 0.0205,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4698293498928924,
+      "grad_norm": 1.2141329050064087,
+      "learning_rate": 1.0956833101253233e-05,
+      "loss": 0.0237,
+      "step": 7320
+    },
+    {
+      "epoch": 0.470471193267063,
+      "grad_norm": 2.2326622009277344,
+      "learning_rate": 1.0943571381208144e-05,
+      "loss": 0.0254,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4711130366412336,
+      "grad_norm": 6.558351039886475,
+      "learning_rate": 1.0930309661163055e-05,
+      "loss": 0.0466,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4717548800154042,
+      "grad_norm": 0.7890551686286926,
+      "learning_rate": 1.0917047941117965e-05,
+      "loss": 0.036,
+      "step": 7350
+    },
+    {
+      "epoch": 0.47239672338957484,
+      "grad_norm": 5.157413005828857,
+      "learning_rate": 1.0903786221072874e-05,
+      "loss": 0.012,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47303856676374545,
+      "grad_norm": 0.047313038259744644,
+      "learning_rate": 1.0890524501027784e-05,
+      "loss": 0.0325,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4736804101379161,
+      "grad_norm": 1.082221269607544,
+      "learning_rate": 1.0877262780982693e-05,
+      "loss": 0.0236,
+      "step": 7380
+    },
+    {
+      "epoch": 0.47432225351208673,
+      "grad_norm": 1.6152721643447876,
+      "learning_rate": 1.0864001060937604e-05,
+      "loss": 0.0393,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47496409688625735,
+      "grad_norm": 3.0274405479431152,
+      "learning_rate": 1.0850739340892515e-05,
+      "loss": 0.0323,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47560594026042796,
+      "grad_norm": 0.7774971127510071,
+      "learning_rate": 1.0837477620847425e-05,
+      "loss": 0.0381,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4762477836345986,
+      "grad_norm": 0.2556317448616028,
+      "learning_rate": 1.0824215900802334e-05,
+      "loss": 0.0545,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4768896270087692,
+      "grad_norm": 1.066426396369934,
+      "learning_rate": 1.0810954180757245e-05,
+      "loss": 0.025,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4775314703829398,
+      "grad_norm": 3.27673602104187,
+      "learning_rate": 1.0797692460712155e-05,
+      "loss": 0.0275,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4781733137571104,
+      "grad_norm": 2.6178624629974365,
+      "learning_rate": 1.0784430740667064e-05,
+      "loss": 0.02,
+      "step": 7450
+    },
+    {
+      "epoch": 0.47881515713128103,
+      "grad_norm": 0.1021733283996582,
+      "learning_rate": 1.0771169020621977e-05,
+      "loss": 0.0222,
+      "step": 7460
+    },
+    {
+      "epoch": 0.47945700050545165,
+      "grad_norm": 2.5011208057403564,
+      "learning_rate": 1.0757907300576887e-05,
+      "loss": 0.0293,
+      "step": 7470
+    },
+    {
+      "epoch": 0.48009884387962226,
+      "grad_norm": 0.8056000471115112,
+      "learning_rate": 1.0744645580531796e-05,
+      "loss": 0.013,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4807406872537929,
+      "grad_norm": 1.466461420059204,
+      "learning_rate": 1.0731383860486706e-05,
+      "loss": 0.0137,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4813825306279635,
+      "grad_norm": 1.2518103122711182,
+      "learning_rate": 1.0718122140441615e-05,
+      "loss": 0.0241,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4820243740021341,
+      "grad_norm": 0.7515360713005066,
+      "learning_rate": 1.0704860420396526e-05,
+      "loss": 0.0181,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4826662173763048,
+      "grad_norm": 1.9555808305740356,
+      "learning_rate": 1.0691598700351436e-05,
+      "loss": 0.0119,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4833080607504754,
+      "grad_norm": 2.6431872844696045,
+      "learning_rate": 1.0678336980306347e-05,
+      "loss": 0.0209,
+      "step": 7530
+    },
+    {
+      "epoch": 0.483949904124646,
+      "grad_norm": 4.683318614959717,
+      "learning_rate": 1.0665075260261257e-05,
+      "loss": 0.0312,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4845917474988166,
+      "grad_norm": 0.357537180185318,
+      "learning_rate": 1.0651813540216168e-05,
+      "loss": 0.0335,
+      "step": 7550
+    },
+    {
+      "epoch": 0.48523359087298723,
+      "grad_norm": 0.3757706880569458,
+      "learning_rate": 1.0638551820171077e-05,
+      "loss": 0.0294,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48587543424715784,
+      "grad_norm": 2.468076467514038,
+      "learning_rate": 1.0625290100125987e-05,
+      "loss": 0.0131,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48651727762132846,
+      "grad_norm": 0.47705045342445374,
+      "learning_rate": 1.0612028380080896e-05,
+      "loss": 0.0195,
+      "step": 7580
+    },
+    {
+      "epoch": 0.4871591209954991,
+      "grad_norm": 0.9728513360023499,
+      "learning_rate": 1.0598766660035809e-05,
+      "loss": 0.0446,
+      "step": 7590
+    },
+    {
+      "epoch": 0.4878009643696697,
+      "grad_norm": 1.766430139541626,
+      "learning_rate": 1.0585504939990718e-05,
+      "loss": 0.0255,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4884428077438403,
+      "grad_norm": 0.12747463583946228,
+      "learning_rate": 1.0572243219945628e-05,
+      "loss": 0.0222,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4890846511180109,
+      "grad_norm": 0.43220123648643494,
+      "learning_rate": 1.0558981499900537e-05,
+      "loss": 0.0291,
+      "step": 7620
+    },
+    {
+      "epoch": 0.48972649449218153,
+      "grad_norm": 0.8481399416923523,
+      "learning_rate": 1.0545719779855449e-05,
+      "loss": 0.0131,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49036833786635214,
+      "grad_norm": 0.293948769569397,
+      "learning_rate": 1.0532458059810358e-05,
+      "loss": 0.042,
+      "step": 7640
+    },
+    {
+      "epoch": 0.49101018124052276,
+      "grad_norm": 0.6220739483833313,
+      "learning_rate": 1.0519196339765268e-05,
+      "loss": 0.0227,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4916520246146934,
+      "grad_norm": 0.4679778814315796,
+      "learning_rate": 1.0505934619720179e-05,
+      "loss": 0.0212,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49229386798886404,
+      "grad_norm": 1.4122891426086426,
+      "learning_rate": 1.049267289967509e-05,
+      "loss": 0.0365,
+      "step": 7670
+    },
+    {
+      "epoch": 0.49293571136303466,
+      "grad_norm": 0.41941073536872864,
+      "learning_rate": 1.047941117963e-05,
+      "loss": 0.0417,
+      "step": 7680
+    },
+    {
+      "epoch": 0.49357755473720527,
+      "grad_norm": 0.9413459300994873,
+      "learning_rate": 1.0466149459584909e-05,
+      "loss": 0.0349,
+      "step": 7690
+    },
+    {
+      "epoch": 0.4942193981113759,
+      "grad_norm": 0.16081929206848145,
+      "learning_rate": 1.0452887739539818e-05,
+      "loss": 0.0355,
+      "step": 7700
+    },
+    {
+      "epoch": 0.4948612414855465,
+      "grad_norm": 0.12057320773601532,
+      "learning_rate": 1.0439626019494728e-05,
+      "loss": 0.0237,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4955030848597171,
+      "grad_norm": 0.39029595255851746,
+      "learning_rate": 1.042636429944964e-05,
+      "loss": 0.0242,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4961449282338877,
+      "grad_norm": 1.100891351699829,
+      "learning_rate": 1.041310257940455e-05,
+      "loss": 0.0207,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49678677160805834,
+      "grad_norm": 0.3278030455112457,
+      "learning_rate": 1.039984085935946e-05,
+      "loss": 0.0255,
+      "step": 7740
+    },
+    {
+      "epoch": 0.49742861498222896,
+      "grad_norm": 2.3268234729766846,
+      "learning_rate": 1.038657913931437e-05,
+      "loss": 0.0242,
+      "step": 7750
+    },
+    {
+      "epoch": 0.49807045835639957,
+      "grad_norm": 4.574901580810547,
+      "learning_rate": 1.037331741926928e-05,
+      "loss": 0.0303,
+      "step": 7760
+    },
+    {
+      "epoch": 0.4987123017305702,
+      "grad_norm": 0.822098433971405,
+      "learning_rate": 1.036005569922419e-05,
+      "loss": 0.0191,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4993541451047408,
+      "grad_norm": 0.3044750988483429,
+      "learning_rate": 1.03467939791791e-05,
+      "loss": 0.0235,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4999959884789114,
+      "grad_norm": 0.9926797747612,
+      "learning_rate": 1.0333532259134012e-05,
+      "loss": 0.0132,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5006378318530821,
+      "grad_norm": 1.0424890518188477,
+      "learning_rate": 1.0320270539088922e-05,
+      "loss": 0.0331,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5012796752272527,
+      "grad_norm": 0.3462810516357422,
+      "learning_rate": 1.0307008819043831e-05,
+      "loss": 0.0198,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5019215186014233,
+      "grad_norm": 0.24118022620677948,
+      "learning_rate": 1.029374709899874e-05,
+      "loss": 0.0217,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5025633619755939,
+      "grad_norm": 0.46797555685043335,
+      "learning_rate": 1.028048537895365e-05,
+      "loss": 0.0366,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5032052053497645,
+      "grad_norm": 0.6193893551826477,
+      "learning_rate": 1.0267223658908561e-05,
+      "loss": 0.0103,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5038470487239352,
+      "grad_norm": 0.8424329161643982,
+      "learning_rate": 1.025396193886347e-05,
+      "loss": 0.0321,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5044888920981058,
+      "grad_norm": 0.15049193799495697,
+      "learning_rate": 1.0240700218818382e-05,
+      "loss": 0.0183,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5051307354722764,
+      "grad_norm": 5.053986072540283,
+      "learning_rate": 1.0227438498773293e-05,
+      "loss": 0.0242,
+      "step": 7870
+    },
+    {
+      "epoch": 0.505772578846447,
+      "grad_norm": 2.5194997787475586,
+      "learning_rate": 1.0214176778728202e-05,
+      "loss": 0.0182,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5064144222206176,
+      "grad_norm": 0.05712272971868515,
+      "learning_rate": 1.0200915058683112e-05,
+      "loss": 0.0207,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5070562655947882,
+      "grad_norm": 0.5997760891914368,
+      "learning_rate": 1.0187653338638021e-05,
+      "loss": 0.0385,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5076981089689588,
+      "grad_norm": 0.05409073084592819,
+      "learning_rate": 1.0174391618592931e-05,
+      "loss": 0.0293,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5083399523431295,
+      "grad_norm": 0.35536444187164307,
+      "learning_rate": 1.0161129898547844e-05,
+      "loss": 0.0201,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5089817957173001,
+      "grad_norm": 0.0880308747291565,
+      "learning_rate": 1.0147868178502753e-05,
+      "loss": 0.0137,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5096236390914707,
+      "grad_norm": 0.22606980800628662,
+      "learning_rate": 1.0134606458457663e-05,
+      "loss": 0.0244,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5102654824656413,
+      "grad_norm": 0.05226253718137741,
+      "learning_rate": 1.0121344738412572e-05,
+      "loss": 0.011,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5109073258398119,
+      "grad_norm": 0.9334781765937805,
+      "learning_rate": 1.0108083018367483e-05,
+      "loss": 0.0314,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5115491692139825,
+      "grad_norm": 0.446389764547348,
+      "learning_rate": 1.0094821298322393e-05,
+      "loss": 0.0238,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5121910125881531,
+      "grad_norm": 0.25008928775787354,
+      "learning_rate": 1.0081559578277302e-05,
+      "loss": 0.0191,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5128328559623238,
+      "grad_norm": 1.671425461769104,
+      "learning_rate": 1.0068297858232215e-05,
+      "loss": 0.0342,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5134746993364944,
+      "grad_norm": 0.8550741076469421,
+      "learning_rate": 1.0055036138187125e-05,
+      "loss": 0.0215,
+      "step": 8000
+    },
+    {
+      "epoch": 0.514116542710665,
+      "grad_norm": 0.5535662174224854,
+      "learning_rate": 1.0041774418142034e-05,
+      "loss": 0.0402,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5147583860848356,
+      "grad_norm": 0.6770220398902893,
+      "learning_rate": 1.0028512698096944e-05,
+      "loss": 0.0194,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5154002294590063,
+      "grad_norm": 1.0631548166275024,
+      "learning_rate": 1.0015250978051853e-05,
+      "loss": 0.0329,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5160420728331769,
+      "grad_norm": 0.944418728351593,
+      "learning_rate": 1.0001989258006763e-05,
+      "loss": 0.0195,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5166839162073475,
+      "grad_norm": 0.7196710109710693,
+      "learning_rate": 9.988727537961674e-06,
+      "loss": 0.0229,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5173257595815182,
+      "grad_norm": 0.5939407348632812,
+      "learning_rate": 9.975465817916585e-06,
+      "loss": 0.0214,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5179676029556888,
+      "grad_norm": 8.581120491027832,
+      "learning_rate": 9.962204097871494e-06,
+      "loss": 0.0299,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5186094463298594,
+      "grad_norm": 1.4366955757141113,
+      "learning_rate": 9.948942377826406e-06,
+      "loss": 0.0222,
+      "step": 8080
+    },
+    {
+      "epoch": 0.51925128970403,
+      "grad_norm": 0.4379299581050873,
+      "learning_rate": 9.935680657781315e-06,
+      "loss": 0.0278,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5198931330782006,
+      "grad_norm": 0.5662280917167664,
+      "learning_rate": 9.922418937736226e-06,
+      "loss": 0.0226,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5205349764523712,
+      "grad_norm": 0.5420593619346619,
+      "learning_rate": 9.909157217691136e-06,
+      "loss": 0.0208,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5211768198265418,
+      "grad_norm": 5.630989074707031,
+      "learning_rate": 9.895895497646045e-06,
+      "loss": 0.0375,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5218186632007125,
+      "grad_norm": 4.6729230880737305,
+      "learning_rate": 9.882633777600955e-06,
+      "loss": 0.0321,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5224605065748831,
+      "grad_norm": 1.7400848865509033,
+      "learning_rate": 9.869372057555866e-06,
+      "loss": 0.0141,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5231023499490537,
+      "grad_norm": 1.0144647359848022,
+      "learning_rate": 9.856110337510775e-06,
+      "loss": 0.0378,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5237441933232243,
+      "grad_norm": 0.250291645526886,
+      "learning_rate": 9.842848617465685e-06,
+      "loss": 0.0197,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5243860366973949,
+      "grad_norm": 0.33458101749420166,
+      "learning_rate": 9.829586897420596e-06,
+      "loss": 0.0201,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5250278800715655,
+      "grad_norm": 1.0443096160888672,
+      "learning_rate": 9.816325177375507e-06,
+      "loss": 0.0222,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5256697234457361,
+      "grad_norm": 0.8161584734916687,
+      "learning_rate": 9.803063457330417e-06,
+      "loss": 0.016,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5263115668199068,
+      "grad_norm": 0.007811207789927721,
+      "learning_rate": 9.789801737285328e-06,
+      "loss": 0.0231,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5269534101940774,
+      "grad_norm": 1.5142780542373657,
+      "learning_rate": 9.776540017240237e-06,
+      "loss": 0.0305,
+      "step": 8210
+    },
+    {
+      "epoch": 0.527595253568248,
+      "grad_norm": 0.3009796738624573,
+      "learning_rate": 9.763278297195147e-06,
+      "loss": 0.018,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5282370969424186,
+      "grad_norm": 0.023198135197162628,
+      "learning_rate": 9.750016577150058e-06,
+      "loss": 0.0345,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5288789403165892,
+      "grad_norm": 1.20603609085083,
+      "learning_rate": 9.736754857104967e-06,
+      "loss": 0.0293,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5295207836907598,
+      "grad_norm": 1.3458225727081299,
+      "learning_rate": 9.723493137059877e-06,
+      "loss": 0.0284,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5301626270649304,
+      "grad_norm": 1.0462092161178589,
+      "learning_rate": 9.710231417014786e-06,
+      "loss": 0.0246,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5308044704391011,
+      "grad_norm": 0.3493569493293762,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.0296,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5314463138132717,
+      "grad_norm": 0.17700675129890442,
+      "learning_rate": 9.683707976924607e-06,
+      "loss": 0.0327,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5320881571874423,
+      "grad_norm": 0.3968372344970703,
+      "learning_rate": 9.670446256879518e-06,
+      "loss": 0.0393,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5327300005616129,
+      "grad_norm": 0.5520299077033997,
+      "learning_rate": 9.65718453683443e-06,
+      "loss": 0.0219,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5333718439357835,
+      "grad_norm": 3.568875312805176,
+      "learning_rate": 9.643922816789339e-06,
+      "loss": 0.0194,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5340136873099542,
+      "grad_norm": 2.2202658653259277,
+      "learning_rate": 9.630661096744248e-06,
+      "loss": 0.0308,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5346555306841249,
+      "grad_norm": 0.727830708026886,
+      "learning_rate": 9.61739937669916e-06,
+      "loss": 0.0281,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5352973740582955,
+      "grad_norm": 1.5965157747268677,
+      "learning_rate": 9.604137656654069e-06,
+      "loss": 0.0304,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5359392174324661,
+      "grad_norm": 0.17011073231697083,
+      "learning_rate": 9.590875936608978e-06,
+      "loss": 0.0461,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5365810608066367,
+      "grad_norm": 0.8461958169937134,
+      "learning_rate": 9.577614216563888e-06,
+      "loss": 0.0254,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5372229041808073,
+      "grad_norm": 4.257885456085205,
+      "learning_rate": 9.564352496518799e-06,
+      "loss": 0.0286,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5378647475549779,
+      "grad_norm": 0.5877963900566101,
+      "learning_rate": 9.551090776473709e-06,
+      "loss": 0.0202,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5385065909291485,
+      "grad_norm": 0.26890161633491516,
+      "learning_rate": 9.53782905642862e-06,
+      "loss": 0.025,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5391484343033192,
+      "grad_norm": 0.25851166248321533,
+      "learning_rate": 9.52456733638353e-06,
+      "loss": 0.0319,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5397902776774898,
+      "grad_norm": 0.757095992565155,
+      "learning_rate": 9.51130561633844e-06,
+      "loss": 0.0299,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5404321210516604,
+      "grad_norm": 0.4859978258609772,
+      "learning_rate": 9.49804389629335e-06,
+      "loss": 0.0251,
+      "step": 8420
+    },
+    {
+      "epoch": 0.541073964425831,
+      "grad_norm": 1.4098316431045532,
+      "learning_rate": 9.484782176248261e-06,
+      "loss": 0.0293,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5417158078000016,
+      "grad_norm": 1.0507205724716187,
+      "learning_rate": 9.47152045620317e-06,
+      "loss": 0.0382,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5423576511741722,
+      "grad_norm": 1.2821528911590576,
+      "learning_rate": 9.45825873615808e-06,
+      "loss": 0.0254,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5429994945483428,
+      "grad_norm": 3.5782406330108643,
+      "learning_rate": 9.444997016112991e-06,
+      "loss": 0.0311,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5436413379225135,
+      "grad_norm": 1.255115032196045,
+      "learning_rate": 9.4317352960679e-06,
+      "loss": 0.0283,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5442831812966841,
+      "grad_norm": 2.0047855377197266,
+      "learning_rate": 9.41847357602281e-06,
+      "loss": 0.0286,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5449250246708547,
+      "grad_norm": 3.766130208969116,
+      "learning_rate": 9.405211855977721e-06,
+      "loss": 0.0249,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5455668680450253,
+      "grad_norm": 3.5647730827331543,
+      "learning_rate": 9.39195013593263e-06,
+      "loss": 0.0328,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5462087114191959,
+      "grad_norm": 3.9039595127105713,
+      "learning_rate": 9.378688415887542e-06,
+      "loss": 0.0239,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5468505547933665,
+      "grad_norm": 1.170214056968689,
+      "learning_rate": 9.365426695842451e-06,
+      "loss": 0.0272,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5474923981675371,
+      "grad_norm": 0.36973121762275696,
+      "learning_rate": 9.352164975797363e-06,
+      "loss": 0.0262,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5481342415417078,
+      "grad_norm": 0.8599039912223816,
+      "learning_rate": 9.338903255752272e-06,
+      "loss": 0.0246,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5487760849158784,
+      "grad_norm": 0.2630428671836853,
+      "learning_rate": 9.325641535707182e-06,
+      "loss": 0.0254,
+      "step": 8550
+    },
+    {
+      "epoch": 0.549417928290049,
+      "grad_norm": 0.9511770009994507,
+      "learning_rate": 9.312379815662093e-06,
+      "loss": 0.0255,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5500597716642196,
+      "grad_norm": 0.2588956654071808,
+      "learning_rate": 9.299118095617002e-06,
+      "loss": 0.0307,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5507016150383902,
+      "grad_norm": 0.17903345823287964,
+      "learning_rate": 9.285856375571912e-06,
+      "loss": 0.0152,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5513434584125608,
+      "grad_norm": 2.2372515201568604,
+      "learning_rate": 9.272594655526823e-06,
+      "loss": 0.0291,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5519853017867314,
+      "grad_norm": 0.02844194695353508,
+      "learning_rate": 9.259332935481732e-06,
+      "loss": 0.017,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5526271451609021,
+      "grad_norm": 0.294089138507843,
+      "learning_rate": 9.246071215436644e-06,
+      "loss": 0.0166,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5532689885350728,
+      "grad_norm": 0.9256640672683716,
+      "learning_rate": 9.232809495391553e-06,
+      "loss": 0.0302,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5539108319092434,
+      "grad_norm": 0.5491749048233032,
+      "learning_rate": 9.219547775346464e-06,
+      "loss": 0.012,
+      "step": 8630
+    },
+    {
+      "epoch": 0.554552675283414,
+      "grad_norm": 3.515266180038452,
+      "learning_rate": 9.206286055301374e-06,
+      "loss": 0.0239,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5551945186575846,
+      "grad_norm": 3.337519407272339,
+      "learning_rate": 9.193024335256283e-06,
+      "loss": 0.0384,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5558363620317552,
+      "grad_norm": 0.6898011565208435,
+      "learning_rate": 9.179762615211194e-06,
+      "loss": 0.0196,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5564782054059259,
+      "grad_norm": 1.0568803548812866,
+      "learning_rate": 9.166500895166104e-06,
+      "loss": 0.0287,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5571200487800965,
+      "grad_norm": 0.8259192109107971,
+      "learning_rate": 9.153239175121013e-06,
+      "loss": 0.0126,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5577618921542671,
+      "grad_norm": 0.4048657715320587,
+      "learning_rate": 9.139977455075924e-06,
+      "loss": 0.0292,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5584037355284377,
+      "grad_norm": 0.6915555596351624,
+      "learning_rate": 9.126715735030834e-06,
+      "loss": 0.0184,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5590455789026083,
+      "grad_norm": 0.19845692813396454,
+      "learning_rate": 9.113454014985743e-06,
+      "loss": 0.0232,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5596874222767789,
+      "grad_norm": 0.4381246864795685,
+      "learning_rate": 9.100192294940655e-06,
+      "loss": 0.0441,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5603292656509495,
+      "grad_norm": 1.0122863054275513,
+      "learning_rate": 9.086930574895566e-06,
+      "loss": 0.0272,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5609711090251202,
+      "grad_norm": 0.18714962899684906,
+      "learning_rate": 9.073668854850475e-06,
+      "loss": 0.017,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5616129523992908,
+      "grad_norm": 0.34267953038215637,
+      "learning_rate": 9.060407134805385e-06,
+      "loss": 0.0257,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5622547957734614,
+      "grad_norm": 0.4609805643558502,
+      "learning_rate": 9.047145414760296e-06,
+      "loss": 0.0256,
+      "step": 8760
+    },
+    {
+      "epoch": 0.562896639147632,
+      "grad_norm": 0.06432998180389404,
+      "learning_rate": 9.033883694715205e-06,
+      "loss": 0.0141,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5635384825218026,
+      "grad_norm": 0.09321536868810654,
+      "learning_rate": 9.020621974670115e-06,
+      "loss": 0.0212,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5641803258959732,
+      "grad_norm": 7.035615921020508,
+      "learning_rate": 9.007360254625026e-06,
+      "loss": 0.0337,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5648221692701438,
+      "grad_norm": 0.7401273250579834,
+      "learning_rate": 8.994098534579935e-06,
+      "loss": 0.0252,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5654640126443145,
+      "grad_norm": 4.558418273925781,
+      "learning_rate": 8.980836814534845e-06,
+      "loss": 0.0336,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5661058560184851,
+      "grad_norm": 0.634662389755249,
+      "learning_rate": 8.967575094489756e-06,
+      "loss": 0.0163,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5667476993926557,
+      "grad_norm": 0.8215422630310059,
+      "learning_rate": 8.954313374444666e-06,
+      "loss": 0.0277,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5673895427668263,
+      "grad_norm": 0.9724500179290771,
+      "learning_rate": 8.941051654399577e-06,
+      "loss": 0.02,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5680313861409969,
+      "grad_norm": 5.943985939025879,
+      "learning_rate": 8.927789934354486e-06,
+      "loss": 0.0225,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5686732295151675,
+      "grad_norm": 2.096219301223755,
+      "learning_rate": 8.914528214309397e-06,
+      "loss": 0.0417,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5693150728893381,
+      "grad_norm": 4.453114986419678,
+      "learning_rate": 8.901266494264307e-06,
+      "loss": 0.0248,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5699569162635088,
+      "grad_norm": 4.170821189880371,
+      "learning_rate": 8.888004774219216e-06,
+      "loss": 0.0231,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5705987596376794,
+      "grad_norm": 1.1110122203826904,
+      "learning_rate": 8.874743054174128e-06,
+      "loss": 0.0282,
+      "step": 8890
+    },
+    {
+      "epoch": 0.57124060301185,
+      "grad_norm": 2.6466012001037598,
+      "learning_rate": 8.861481334129037e-06,
+      "loss": 0.0221,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5718824463860207,
+      "grad_norm": 0.25523126125335693,
+      "learning_rate": 8.848219614083947e-06,
+      "loss": 0.0127,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5725242897601913,
+      "grad_norm": 3.594900608062744,
+      "learning_rate": 8.834957894038858e-06,
+      "loss": 0.0209,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5731661331343619,
+      "grad_norm": 0.8159608840942383,
+      "learning_rate": 8.821696173993767e-06,
+      "loss": 0.018,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5738079765085325,
+      "grad_norm": 0.9161319732666016,
+      "learning_rate": 8.808434453948678e-06,
+      "loss": 0.0368,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5744498198827032,
+      "grad_norm": 0.16555462777614594,
+      "learning_rate": 8.795172733903588e-06,
+      "loss": 0.0199,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5750916632568738,
+      "grad_norm": 3.5098259449005127,
+      "learning_rate": 8.781911013858499e-06,
+      "loss": 0.0296,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5757335066310444,
+      "grad_norm": 0.0742286741733551,
+      "learning_rate": 8.768649293813408e-06,
+      "loss": 0.038,
+      "step": 8970
+    },
+    {
+      "epoch": 0.576375350005215,
+      "grad_norm": 1.3032357692718506,
+      "learning_rate": 8.755387573768318e-06,
+      "loss": 0.0373,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5770171933793856,
+      "grad_norm": 0.9205700755119324,
+      "learning_rate": 8.742125853723229e-06,
+      "loss": 0.0328,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5776590367535562,
+      "grad_norm": 1.2080702781677246,
+      "learning_rate": 8.728864133678139e-06,
+      "loss": 0.0313,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5783008801277268,
+      "grad_norm": 0.5622536540031433,
+      "learning_rate": 8.715602413633048e-06,
+      "loss": 0.0206,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5789427235018975,
+      "grad_norm": 0.46471625566482544,
+      "learning_rate": 8.70234069358796e-06,
+      "loss": 0.0324,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5795845668760681,
+      "grad_norm": 0.054508015513420105,
+      "learning_rate": 8.689078973542869e-06,
+      "loss": 0.0242,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5802264102502387,
+      "grad_norm": 8.564746856689453,
+      "learning_rate": 8.67581725349778e-06,
+      "loss": 0.0328,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5808682536244093,
+      "grad_norm": 0.13193635642528534,
+      "learning_rate": 8.66255553345269e-06,
+      "loss": 0.0239,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5815100969985799,
+      "grad_norm": 0.4824690520763397,
+      "learning_rate": 8.6492938134076e-06,
+      "loss": 0.0225,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5821519403727505,
+      "grad_norm": 0.7100673913955688,
+      "learning_rate": 8.63603209336251e-06,
+      "loss": 0.0334,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5827937837469211,
+      "grad_norm": 0.1487302929162979,
+      "learning_rate": 8.62277037331742e-06,
+      "loss": 0.0325,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5834356271210918,
+      "grad_norm": 0.15965911746025085,
+      "learning_rate": 8.60950865327233e-06,
+      "loss": 0.0248,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5840774704952624,
+      "grad_norm": 3.24701189994812,
+      "learning_rate": 8.59624693322724e-06,
+      "loss": 0.034,
+      "step": 9100
+    },
+    {
+      "epoch": 0.584719313869433,
+      "grad_norm": 0.09657014906406403,
+      "learning_rate": 8.58298521318215e-06,
+      "loss": 0.032,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5853611572436036,
+      "grad_norm": 2.3377416133880615,
+      "learning_rate": 8.56972349313706e-06,
+      "loss": 0.0441,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5860030006177742,
+      "grad_norm": 1.443271279335022,
+      "learning_rate": 8.55646177309197e-06,
+      "loss": 0.0247,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5866448439919448,
+      "grad_norm": 3.8856918811798096,
+      "learning_rate": 8.54320005304688e-06,
+      "loss": 0.0189,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5872866873661154,
+      "grad_norm": 0.44424521923065186,
+      "learning_rate": 8.529938333001791e-06,
+      "loss": 0.0216,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5879285307402861,
+      "grad_norm": 0.9118013978004456,
+      "learning_rate": 8.516676612956702e-06,
+      "loss": 0.0252,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5885703741144567,
+      "grad_norm": 1.9455515146255493,
+      "learning_rate": 8.503414892911612e-06,
+      "loss": 0.0174,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5892122174886273,
+      "grad_norm": 0.43138858675956726,
+      "learning_rate": 8.490153172866521e-06,
+      "loss": 0.0227,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5898540608627979,
+      "grad_norm": 1.8877631425857544,
+      "learning_rate": 8.476891452821432e-06,
+      "loss": 0.0376,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5904959042369686,
+      "grad_norm": 0.6855882406234741,
+      "learning_rate": 8.463629732776342e-06,
+      "loss": 0.0261,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5911377476111392,
+      "grad_norm": 1.141603946685791,
+      "learning_rate": 8.450368012731251e-06,
+      "loss": 0.0382,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5917795909853099,
+      "grad_norm": 0.5965757369995117,
+      "learning_rate": 8.437106292686162e-06,
+      "loss": 0.0264,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5924214343594805,
+      "grad_norm": 6.3784637451171875,
+      "learning_rate": 8.423844572641072e-06,
+      "loss": 0.0331,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5930632777336511,
+      "grad_norm": 1.8399165868759155,
+      "learning_rate": 8.410582852595981e-06,
+      "loss": 0.0178,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5937051211078217,
+      "grad_norm": 0.6760837435722351,
+      "learning_rate": 8.397321132550892e-06,
+      "loss": 0.0219,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5943469644819923,
+      "grad_norm": 0.23297323286533356,
+      "learning_rate": 8.384059412505802e-06,
+      "loss": 0.0161,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5949888078561629,
+      "grad_norm": 0.9712881445884705,
+      "learning_rate": 8.370797692460713e-06,
+      "loss": 0.0139,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5956306512303335,
+      "grad_norm": 0.10685652494430542,
+      "learning_rate": 8.357535972415624e-06,
+      "loss": 0.0258,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5962724946045042,
+      "grad_norm": 1.6632583141326904,
+      "learning_rate": 8.344274252370534e-06,
+      "loss": 0.0251,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5969143379786748,
+      "grad_norm": 0.5271076560020447,
+      "learning_rate": 8.331012532325443e-06,
+      "loss": 0.0291,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5975561813528454,
+      "grad_norm": 1.0587668418884277,
+      "learning_rate": 8.317750812280353e-06,
+      "loss": 0.0257,
+      "step": 9310
+    },
+    {
+      "epoch": 0.598198024727016,
+      "grad_norm": 1.011917233467102,
+      "learning_rate": 8.304489092235264e-06,
+      "loss": 0.0309,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5988398681011866,
+      "grad_norm": 0.80730801820755,
+      "learning_rate": 8.291227372190173e-06,
+      "loss": 0.0256,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5994817114753572,
+      "grad_norm": 0.4712038040161133,
+      "learning_rate": 8.277965652145083e-06,
+      "loss": 0.021,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6001235548495278,
+      "grad_norm": 0.05072961002588272,
+      "learning_rate": 8.264703932099994e-06,
+      "loss": 0.0118,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6007653982236985,
+      "grad_norm": 5.919791221618652,
+      "learning_rate": 8.251442212054904e-06,
+      "loss": 0.0259,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6014072415978691,
+      "grad_norm": 0.14993645250797272,
+      "learning_rate": 8.238180492009815e-06,
+      "loss": 0.0088,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6020490849720397,
+      "grad_norm": 4.77583646774292,
+      "learning_rate": 8.224918771964724e-06,
+      "loss": 0.0176,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6026909283462103,
+      "grad_norm": 0.11115249246358871,
+      "learning_rate": 8.211657051919635e-06,
+      "loss": 0.0282,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6033327717203809,
+      "grad_norm": 1.8065743446350098,
+      "learning_rate": 8.198395331874545e-06,
+      "loss": 0.0161,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6039746150945515,
+      "grad_norm": 0.1328521966934204,
+      "learning_rate": 8.185133611829454e-06,
+      "loss": 0.0223,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6046164584687221,
+      "grad_norm": 1.3424381017684937,
+      "learning_rate": 8.171871891784365e-06,
+      "loss": 0.0189,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6052583018428928,
+      "grad_norm": 1.714998722076416,
+      "learning_rate": 8.158610171739275e-06,
+      "loss": 0.0251,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6059001452170634,
+      "grad_norm": 2.506918430328369,
+      "learning_rate": 8.145348451694184e-06,
+      "loss": 0.0187,
+      "step": 9440
+    },
+    {
+      "epoch": 0.606541988591234,
+      "grad_norm": 3.1794984340667725,
+      "learning_rate": 8.132086731649096e-06,
+      "loss": 0.0297,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6071838319654046,
+      "grad_norm": 0.3733360469341278,
+      "learning_rate": 8.118825011604005e-06,
+      "loss": 0.0181,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6078256753395752,
+      "grad_norm": 2.3479220867156982,
+      "learning_rate": 8.105563291558916e-06,
+      "loss": 0.023,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6084675187137458,
+      "grad_norm": 1.5137450695037842,
+      "learning_rate": 8.092301571513826e-06,
+      "loss": 0.0137,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6091093620879166,
+      "grad_norm": 0.19078929722309113,
+      "learning_rate": 8.079039851468737e-06,
+      "loss": 0.021,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6097512054620872,
+      "grad_norm": 2.02891206741333,
+      "learning_rate": 8.065778131423646e-06,
+      "loss": 0.0186,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6103930488362578,
+      "grad_norm": 2.559931993484497,
+      "learning_rate": 8.052516411378558e-06,
+      "loss": 0.0236,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6110348922104284,
+      "grad_norm": 0.35723981261253357,
+      "learning_rate": 8.039254691333467e-06,
+      "loss": 0.013,
+      "step": 9520
+    },
+    {
+      "epoch": 0.611676735584599,
+      "grad_norm": 1.501561164855957,
+      "learning_rate": 8.025992971288377e-06,
+      "loss": 0.0262,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6123185789587696,
+      "grad_norm": 1.1635639667510986,
+      "learning_rate": 8.012731251243286e-06,
+      "loss": 0.0238,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6129604223329402,
+      "grad_norm": 0.27196788787841797,
+      "learning_rate": 7.999469531198197e-06,
+      "loss": 0.0198,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6136022657071108,
+      "grad_norm": 2.936884880065918,
+      "learning_rate": 7.986207811153107e-06,
+      "loss": 0.0324,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6142441090812815,
+      "grad_norm": 0.5064958930015564,
+      "learning_rate": 7.972946091108018e-06,
+      "loss": 0.0149,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6148859524554521,
+      "grad_norm": 0.12043922394514084,
+      "learning_rate": 7.959684371062927e-06,
+      "loss": 0.0205,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6155277958296227,
+      "grad_norm": 0.04327112436294556,
+      "learning_rate": 7.946422651017838e-06,
+      "loss": 0.0392,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6161696392037933,
+      "grad_norm": 0.17598643898963928,
+      "learning_rate": 7.933160930972748e-06,
+      "loss": 0.018,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6168114825779639,
+      "grad_norm": 3.665308713912964,
+      "learning_rate": 7.919899210927659e-06,
+      "loss": 0.0239,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6174533259521345,
+      "grad_norm": 4.262577056884766,
+      "learning_rate": 7.906637490882569e-06,
+      "loss": 0.027,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6180951693263051,
+      "grad_norm": 2.2056829929351807,
+      "learning_rate": 7.893375770837478e-06,
+      "loss": 0.0276,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6187370127004758,
+      "grad_norm": 4.004680633544922,
+      "learning_rate": 7.880114050792388e-06,
+      "loss": 0.0194,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6193788560746464,
+      "grad_norm": 0.2112032175064087,
+      "learning_rate": 7.866852330747299e-06,
+      "loss": 0.0177,
+      "step": 9650
+    },
+    {
+      "epoch": 0.620020699448817,
+      "grad_norm": 1.0510079860687256,
+      "learning_rate": 7.853590610702208e-06,
+      "loss": 0.0263,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6206625428229876,
+      "grad_norm": 0.08418594300746918,
+      "learning_rate": 7.840328890657118e-06,
+      "loss": 0.0236,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6213043861971582,
+      "grad_norm": 2.5858445167541504,
+      "learning_rate": 7.827067170612029e-06,
+      "loss": 0.0228,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6219462295713288,
+      "grad_norm": 0.09412383288145065,
+      "learning_rate": 7.813805450566938e-06,
+      "loss": 0.0235,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6225880729454994,
+      "grad_norm": 1.0433266162872314,
+      "learning_rate": 7.80054373052185e-06,
+      "loss": 0.0243,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6232299163196701,
+      "grad_norm": 1.1485860347747803,
+      "learning_rate": 7.78728201047676e-06,
+      "loss": 0.0281,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6238717596938407,
+      "grad_norm": 2.220686674118042,
+      "learning_rate": 7.77402029043167e-06,
+      "loss": 0.0242,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6245136030680113,
+      "grad_norm": 1.1384696960449219,
+      "learning_rate": 7.76075857038658e-06,
+      "loss": 0.0226,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6251554464421819,
+      "grad_norm": 3.107987880706787,
+      "learning_rate": 7.74749685034149e-06,
+      "loss": 0.034,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6257972898163525,
+      "grad_norm": 0.5255154967308044,
+      "learning_rate": 7.7342351302964e-06,
+      "loss": 0.0157,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6264391331905231,
+      "grad_norm": 0.11366035789251328,
+      "learning_rate": 7.72097341025131e-06,
+      "loss": 0.0122,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6270809765646937,
+      "grad_norm": 2.00895094871521,
+      "learning_rate": 7.70771169020622e-06,
+      "loss": 0.0222,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6277228199388645,
+      "grad_norm": 3.02500319480896,
+      "learning_rate": 7.69444997016113e-06,
+      "loss": 0.016,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6283646633130351,
+      "grad_norm": 4.960433483123779,
+      "learning_rate": 7.68118825011604e-06,
+      "loss": 0.0247,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6290065066872057,
+      "grad_norm": 0.007853507064282894,
+      "learning_rate": 7.667926530070951e-06,
+      "loss": 0.0364,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6296483500613763,
+      "grad_norm": 2.501887559890747,
+      "learning_rate": 7.65466481002586e-06,
+      "loss": 0.0265,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6302901934355469,
+      "grad_norm": 0.5652506351470947,
+      "learning_rate": 7.641403089980772e-06,
+      "loss": 0.0146,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6309320368097175,
+      "grad_norm": 4.806582927703857,
+      "learning_rate": 7.628141369935681e-06,
+      "loss": 0.0161,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6315738801838882,
+      "grad_norm": 1.12467622756958,
+      "learning_rate": 7.6148796498905915e-06,
+      "loss": 0.0278,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6322157235580588,
+      "grad_norm": 1.5796605348587036,
+      "learning_rate": 7.601617929845502e-06,
+      "loss": 0.0246,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6328575669322294,
+      "grad_norm": 1.3487434387207031,
+      "learning_rate": 7.588356209800411e-06,
+      "loss": 0.0126,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6334994103064,
+      "grad_norm": 0.05349859222769737,
+      "learning_rate": 7.575094489755322e-06,
+      "loss": 0.0215,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6341412536805706,
+      "grad_norm": 1.3048077821731567,
+      "learning_rate": 7.561832769710232e-06,
+      "loss": 0.0223,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6347830970547412,
+      "grad_norm": 0.46381476521492004,
+      "learning_rate": 7.548571049665142e-06,
+      "loss": 0.0167,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6354249404289118,
+      "grad_norm": 0.19564686715602875,
+      "learning_rate": 7.535309329620052e-06,
+      "loss": 0.0129,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6360667838030825,
+      "grad_norm": 1.4157600402832031,
+      "learning_rate": 7.522047609574963e-06,
+      "loss": 0.0121,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6367086271772531,
+      "grad_norm": 0.7896568179130554,
+      "learning_rate": 7.5087858895298724e-06,
+      "loss": 0.021,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6373504705514237,
+      "grad_norm": 0.48979702591896057,
+      "learning_rate": 7.495524169484783e-06,
+      "loss": 0.0423,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6379923139255943,
+      "grad_norm": 0.4577849805355072,
+      "learning_rate": 7.482262449439693e-06,
+      "loss": 0.0214,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6386341572997649,
+      "grad_norm": 0.2398267239332199,
+      "learning_rate": 7.469000729394603e-06,
+      "loss": 0.0307,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6392760006739355,
+      "grad_norm": 0.11448551714420319,
+      "learning_rate": 7.455739009349513e-06,
+      "loss": 0.0222,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6399178440481061,
+      "grad_norm": 3.8593854904174805,
+      "learning_rate": 7.442477289304424e-06,
+      "loss": 0.0319,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6405596874222768,
+      "grad_norm": 0.3872242271900177,
+      "learning_rate": 7.4292155692593335e-06,
+      "loss": 0.0297,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6412015307964474,
+      "grad_norm": 0.513504683971405,
+      "learning_rate": 7.415953849214244e-06,
+      "loss": 0.0307,
+      "step": 9990
+    },
+    {
+      "epoch": 0.641843374170618,
+      "grad_norm": 5.903943061828613,
+      "learning_rate": 7.402692129169153e-06,
+      "loss": 0.0392,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6424852175447886,
+      "grad_norm": 0.5244473218917847,
+      "learning_rate": 7.3894304091240645e-06,
+      "loss": 0.034,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6431270609189592,
+      "grad_norm": 0.04969165101647377,
+      "learning_rate": 7.376168689078974e-06,
+      "loss": 0.012,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6437689042931298,
+      "grad_norm": 2.9083058834075928,
+      "learning_rate": 7.362906969033884e-06,
+      "loss": 0.0293,
+      "step": 10030
+    },
+    {
+      "epoch": 0.6444107476673004,
+      "grad_norm": 0.10136771202087402,
+      "learning_rate": 7.349645248988795e-06,
+      "loss": 0.0228,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6450525910414711,
+      "grad_norm": 0.04527024179697037,
+      "learning_rate": 7.336383528943705e-06,
+      "loss": 0.024,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6456944344156417,
+      "grad_norm": 0.4695492684841156,
+      "learning_rate": 7.3231218088986144e-06,
+      "loss": 0.0185,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6463362777898123,
+      "grad_norm": 2.390000343322754,
+      "learning_rate": 7.309860088853526e-06,
+      "loss": 0.0189,
+      "step": 10070
+    },
+    {
+      "epoch": 0.646978121163983,
+      "grad_norm": 0.30994078516960144,
+      "learning_rate": 7.296598368808435e-06,
+      "loss": 0.0212,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6476199645381536,
+      "grad_norm": 1.3190053701400757,
+      "learning_rate": 7.283336648763345e-06,
+      "loss": 0.0312,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6482618079123242,
+      "grad_norm": 5.3458662033081055,
+      "learning_rate": 7.270074928718255e-06,
+      "loss": 0.0191,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6489036512864949,
+      "grad_norm": 1.1833816766738892,
+      "learning_rate": 7.256813208673166e-06,
+      "loss": 0.0311,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6495454946606655,
+      "grad_norm": 0.6074949502944946,
+      "learning_rate": 7.2435514886280755e-06,
+      "loss": 0.0258,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6501873380348361,
+      "grad_norm": 1.357013463973999,
+      "learning_rate": 7.230289768582985e-06,
+      "loss": 0.0253,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6508291814090067,
+      "grad_norm": 0.33180567622184753,
+      "learning_rate": 7.217028048537896e-06,
+      "loss": 0.0288,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6514710247831773,
+      "grad_norm": 0.6609655022621155,
+      "learning_rate": 7.2037663284928065e-06,
+      "loss": 0.0206,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6521128681573479,
+      "grad_norm": 0.3761395812034607,
+      "learning_rate": 7.190504608447716e-06,
+      "loss": 0.0189,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6527547115315185,
+      "grad_norm": 0.29698699712753296,
+      "learning_rate": 7.177242888402627e-06,
+      "loss": 0.0219,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6533965549056892,
+      "grad_norm": 0.464578777551651,
+      "learning_rate": 7.163981168357537e-06,
+      "loss": 0.0272,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6540383982798598,
+      "grad_norm": 2.6864356994628906,
+      "learning_rate": 7.150719448312446e-06,
+      "loss": 0.0167,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6546802416540304,
+      "grad_norm": 2.0054922103881836,
+      "learning_rate": 7.137457728267357e-06,
+      "loss": 0.0113,
+      "step": 10200
+    },
+    {
+      "epoch": 0.655322085028201,
+      "grad_norm": 1.1829546689987183,
+      "learning_rate": 7.124196008222267e-06,
+      "loss": 0.0173,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6559639284023716,
+      "grad_norm": 0.9671972990036011,
+      "learning_rate": 7.110934288177177e-06,
+      "loss": 0.0224,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6566057717765422,
+      "grad_norm": 1.9605666399002075,
+      "learning_rate": 7.097672568132087e-06,
+      "loss": 0.0228,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6572476151507128,
+      "grad_norm": 1.4973517656326294,
+      "learning_rate": 7.084410848086998e-06,
+      "loss": 0.0363,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6578894585248835,
+      "grad_norm": 0.1850045919418335,
+      "learning_rate": 7.071149128041907e-06,
+      "loss": 0.0302,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6585313018990541,
+      "grad_norm": 0.5483193397521973,
+      "learning_rate": 7.0578874079968176e-06,
+      "loss": 0.0183,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6591731452732247,
+      "grad_norm": 1.6930410861968994,
+      "learning_rate": 7.044625687951728e-06,
+      "loss": 0.0204,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6598149886473953,
+      "grad_norm": 0.10521617531776428,
+      "learning_rate": 7.031363967906638e-06,
+      "loss": 0.025,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6604568320215659,
+      "grad_norm": 0.20267446339130402,
+      "learning_rate": 7.018102247861548e-06,
+      "loss": 0.0288,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6610986753957365,
+      "grad_norm": 0.15048721432685852,
+      "learning_rate": 7.004840527816459e-06,
+      "loss": 0.0309,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6617405187699071,
+      "grad_norm": 0.7346864938735962,
+      "learning_rate": 6.991578807771368e-06,
+      "loss": 0.0363,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6623823621440778,
+      "grad_norm": 0.04179545119404793,
+      "learning_rate": 6.978317087726279e-06,
+      "loss": 0.0145,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6630242055182484,
+      "grad_norm": 0.3221127390861511,
+      "learning_rate": 6.965055367681188e-06,
+      "loss": 0.0381,
+      "step": 10330
+    },
+    {
+      "epoch": 0.663666048892419,
+      "grad_norm": 0.3161191940307617,
+      "learning_rate": 6.951793647636099e-06,
+      "loss": 0.0201,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6643078922665896,
+      "grad_norm": 0.6822152137756348,
+      "learning_rate": 6.938531927591009e-06,
+      "loss": 0.0115,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6649497356407602,
+      "grad_norm": 0.29630982875823975,
+      "learning_rate": 6.925270207545919e-06,
+      "loss": 0.02,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6655915790149309,
+      "grad_norm": 0.5653153657913208,
+      "learning_rate": 6.9120084875008294e-06,
+      "loss": 0.0197,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6662334223891015,
+      "grad_norm": 0.08264129608869553,
+      "learning_rate": 6.89874676745574e-06,
+      "loss": 0.0173,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6668752657632722,
+      "grad_norm": 1.7407037019729614,
+      "learning_rate": 6.885485047410649e-06,
+      "loss": 0.0172,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6675171091374428,
+      "grad_norm": 0.17370440065860748,
+      "learning_rate": 6.87222332736556e-06,
+      "loss": 0.0233,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6681589525116134,
+      "grad_norm": 0.6718391180038452,
+      "learning_rate": 6.85896160732047e-06,
+      "loss": 0.029,
+      "step": 10410
+    },
+    {
+      "epoch": 0.668800795885784,
+      "grad_norm": 0.21962031722068787,
+      "learning_rate": 6.84569988727538e-06,
+      "loss": 0.0157,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6694426392599546,
+      "grad_norm": 1.031297206878662,
+      "learning_rate": 6.8324381672302905e-06,
+      "loss": 0.0296,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6700844826341252,
+      "grad_norm": 3.4837470054626465,
+      "learning_rate": 6.819176447185201e-06,
+      "loss": 0.0358,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6707263260082958,
+      "grad_norm": 0.05151769146323204,
+      "learning_rate": 6.80591472714011e-06,
+      "loss": 0.0316,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6713681693824665,
+      "grad_norm": 0.693645715713501,
+      "learning_rate": 6.792653007095021e-06,
+      "loss": 0.034,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6720100127566371,
+      "grad_norm": 1.7422330379486084,
+      "learning_rate": 6.779391287049931e-06,
+      "loss": 0.0133,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6726518561308077,
+      "grad_norm": 1.208640456199646,
+      "learning_rate": 6.766129567004841e-06,
+      "loss": 0.0339,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6732936995049783,
+      "grad_norm": 3.527818202972412,
+      "learning_rate": 6.752867846959751e-06,
+      "loss": 0.0204,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6739355428791489,
+      "grad_norm": 0.4659969210624695,
+      "learning_rate": 6.739606126914662e-06,
+      "loss": 0.0231,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6745773862533195,
+      "grad_norm": 0.316992849111557,
+      "learning_rate": 6.7263444068695715e-06,
+      "loss": 0.0151,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6752192296274901,
+      "grad_norm": 0.5036879181861877,
+      "learning_rate": 6.713082686824482e-06,
+      "loss": 0.0274,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6758610730016608,
+      "grad_norm": 2.00020170211792,
+      "learning_rate": 6.699820966779392e-06,
+      "loss": 0.0209,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6765029163758314,
+      "grad_norm": 0.012486232444643974,
+      "learning_rate": 6.6865592467343024e-06,
+      "loss": 0.0187,
+      "step": 10540
+    },
+    {
+      "epoch": 0.677144759750002,
+      "grad_norm": 0.03590549901127815,
+      "learning_rate": 6.673297526689212e-06,
+      "loss": 0.0298,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6777866031241726,
+      "grad_norm": 0.11736287921667099,
+      "learning_rate": 6.660035806644121e-06,
+      "loss": 0.0214,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6784284464983432,
+      "grad_norm": 0.8670855164527893,
+      "learning_rate": 6.6467740865990326e-06,
+      "loss": 0.015,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6790702898725138,
+      "grad_norm": 5.439849376678467,
+      "learning_rate": 6.633512366553943e-06,
+      "loss": 0.0259,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6797121332466844,
+      "grad_norm": 0.3086478114128113,
+      "learning_rate": 6.620250646508852e-06,
+      "loss": 0.018,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6803539766208551,
+      "grad_norm": 0.7371939420700073,
+      "learning_rate": 6.6069889264637635e-06,
+      "loss": 0.0192,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6809958199950257,
+      "grad_norm": 0.051234494894742966,
+      "learning_rate": 6.593727206418673e-06,
+      "loss": 0.0196,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6816376633691963,
+      "grad_norm": 0.4612034559249878,
+      "learning_rate": 6.5804654863735825e-06,
+      "loss": 0.0163,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6822795067433669,
+      "grad_norm": 0.1792960911989212,
+      "learning_rate": 6.567203766328494e-06,
+      "loss": 0.0277,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6829213501175375,
+      "grad_norm": 0.02334243804216385,
+      "learning_rate": 6.553942046283404e-06,
+      "loss": 0.0167,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6835631934917081,
+      "grad_norm": 0.13844627141952515,
+      "learning_rate": 6.5406803262383135e-06,
+      "loss": 0.02,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6842050368658789,
+      "grad_norm": 1.1921093463897705,
+      "learning_rate": 6.527418606193225e-06,
+      "loss": 0.0408,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6848468802400495,
+      "grad_norm": 6.113221645355225,
+      "learning_rate": 6.514156886148134e-06,
+      "loss": 0.0202,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6854887236142201,
+      "grad_norm": 2.2112908363342285,
+      "learning_rate": 6.500895166103044e-06,
+      "loss": 0.0449,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6861305669883907,
+      "grad_norm": 0.7895561456680298,
+      "learning_rate": 6.487633446057954e-06,
+      "loss": 0.0246,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6867724103625613,
+      "grad_norm": 0.5855046510696411,
+      "learning_rate": 6.474371726012865e-06,
+      "loss": 0.0248,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6874142537367319,
+      "grad_norm": 6.148792266845703,
+      "learning_rate": 6.4611100059677746e-06,
+      "loss": 0.024,
+      "step": 10710
+    },
+    {
+      "epoch": 0.6880560971109025,
+      "grad_norm": 0.5399889945983887,
+      "learning_rate": 6.447848285922684e-06,
+      "loss": 0.0314,
+      "step": 10720
+    },
+    {
+      "epoch": 0.6886979404850732,
+      "grad_norm": 1.1208041906356812,
+      "learning_rate": 6.434586565877595e-06,
+      "loss": 0.0102,
+      "step": 10730
+    },
+    {
+      "epoch": 0.6893397838592438,
+      "grad_norm": 0.8087825775146484,
+      "learning_rate": 6.421324845832505e-06,
+      "loss": 0.0272,
+      "step": 10740
+    },
+    {
+      "epoch": 0.6899816272334144,
+      "grad_norm": 0.9773532152175903,
+      "learning_rate": 6.408063125787415e-06,
+      "loss": 0.0196,
+      "step": 10750
+    },
+    {
+      "epoch": 0.690623470607585,
+      "grad_norm": 0.06517565995454788,
+      "learning_rate": 6.394801405742325e-06,
+      "loss": 0.0133,
+      "step": 10760
+    },
+    {
+      "epoch": 0.6912653139817556,
+      "grad_norm": 0.08576376736164093,
+      "learning_rate": 6.381539685697236e-06,
+      "loss": 0.0205,
+      "step": 10770
+    },
+    {
+      "epoch": 0.6919071573559262,
+      "grad_norm": 0.09191477298736572,
+      "learning_rate": 6.368277965652145e-06,
+      "loss": 0.0237,
+      "step": 10780
+    },
+    {
+      "epoch": 0.6925490007300968,
+      "grad_norm": 1.1734598875045776,
+      "learning_rate": 6.3550162456070555e-06,
+      "loss": 0.0203,
+      "step": 10790
+    },
+    {
+      "epoch": 0.6931908441042675,
+      "grad_norm": 1.6206161975860596,
+      "learning_rate": 6.341754525561966e-06,
+      "loss": 0.0267,
+      "step": 10800
+    },
+    {
+      "epoch": 0.6938326874784381,
+      "grad_norm": 0.42549261450767517,
+      "learning_rate": 6.328492805516876e-06,
+      "loss": 0.0261,
+      "step": 10810
+    },
+    {
+      "epoch": 0.6944745308526087,
+      "grad_norm": 0.44076070189476013,
+      "learning_rate": 6.315231085471786e-06,
+      "loss": 0.0248,
+      "step": 10820
+    },
+    {
+      "epoch": 0.6951163742267793,
+      "grad_norm": 0.2725442349910736,
+      "learning_rate": 6.301969365426697e-06,
+      "loss": 0.0228,
+      "step": 10830
+    },
+    {
+      "epoch": 0.6957582176009499,
+      "grad_norm": 3.0624146461486816,
+      "learning_rate": 6.288707645381606e-06,
+      "loss": 0.0415,
+      "step": 10840
+    },
+    {
+      "epoch": 0.6964000609751205,
+      "grad_norm": 2.1957530975341797,
+      "learning_rate": 6.275445925336517e-06,
+      "loss": 0.0277,
+      "step": 10850
+    },
+    {
+      "epoch": 0.6970419043492911,
+      "grad_norm": 0.498536080121994,
+      "learning_rate": 6.262184205291427e-06,
+      "loss": 0.0342,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6976837477234618,
+      "grad_norm": 0.23503071069717407,
+      "learning_rate": 6.248922485246337e-06,
+      "loss": 0.0289,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6983255910976324,
+      "grad_norm": 0.30300644040107727,
+      "learning_rate": 6.235660765201247e-06,
+      "loss": 0.016,
+      "step": 10880
+    },
+    {
+      "epoch": 0.698967434471803,
+      "grad_norm": 0.22650690376758575,
+      "learning_rate": 6.222399045156158e-06,
+      "loss": 0.0103,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6996092778459736,
+      "grad_norm": 0.1359551101922989,
+      "learning_rate": 6.209137325111067e-06,
+      "loss": 0.0233,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7002511212201442,
+      "grad_norm": 0.02224060520529747,
+      "learning_rate": 6.195875605065978e-06,
+      "loss": 0.0158,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7008929645943148,
+      "grad_norm": 1.749082088470459,
+      "learning_rate": 6.182613885020887e-06,
+      "loss": 0.0183,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7015348079684854,
+      "grad_norm": 0.18207766115665436,
+      "learning_rate": 6.169352164975798e-06,
+      "loss": 0.0156,
+      "step": 10930
+    },
+    {
+      "epoch": 0.702176651342656,
+      "grad_norm": 6.35283899307251,
+      "learning_rate": 6.156090444930708e-06,
+      "loss": 0.0488,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7028184947168268,
+      "grad_norm": 0.07517176121473312,
+      "learning_rate": 6.142828724885618e-06,
+      "loss": 0.0235,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7034603380909974,
+      "grad_norm": 3.0476794242858887,
+      "learning_rate": 6.1295670048405285e-06,
+      "loss": 0.0201,
+      "step": 10960
+    },
+    {
+      "epoch": 0.704102181465168,
+      "grad_norm": 0.7774531841278076,
+      "learning_rate": 6.116305284795439e-06,
+      "loss": 0.0192,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7047440248393386,
+      "grad_norm": 0.27221471071243286,
+      "learning_rate": 6.103043564750348e-06,
+      "loss": 0.0283,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7053858682135092,
+      "grad_norm": 1.1724214553833008,
+      "learning_rate": 6.0897818447052594e-06,
+      "loss": 0.0131,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7060277115876799,
+      "grad_norm": 0.020140307024121284,
+      "learning_rate": 6.076520124660169e-06,
+      "loss": 0.0216,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7066695549618505,
+      "grad_norm": 1.0828317403793335,
+      "learning_rate": 6.063258404615079e-06,
+      "loss": 0.0184,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7073113983360211,
+      "grad_norm": 2.071780204772949,
+      "learning_rate": 6.049996684569989e-06,
+      "loss": 0.0354,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7079532417101917,
+      "grad_norm": 0.17378775775432587,
+      "learning_rate": 6.0367349645249e-06,
+      "loss": 0.028,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7085950850843623,
+      "grad_norm": 0.9045881628990173,
+      "learning_rate": 6.023473244479809e-06,
+      "loss": 0.0242,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7092369284585329,
+      "grad_norm": 0.824508011341095,
+      "learning_rate": 6.010211524434719e-06,
+      "loss": 0.0131,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7098787718327035,
+      "grad_norm": 0.4981328248977661,
+      "learning_rate": 5.99694980438963e-06,
+      "loss": 0.0233,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7105206152068742,
+      "grad_norm": 2.5603647232055664,
+      "learning_rate": 5.98368808434454e-06,
+      "loss": 0.0295,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7111624585810448,
+      "grad_norm": 2.7192113399505615,
+      "learning_rate": 5.97042636429945e-06,
+      "loss": 0.0177,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7118043019552154,
+      "grad_norm": 1.1225374937057495,
+      "learning_rate": 5.957164644254361e-06,
+      "loss": 0.0203,
+      "step": 11090
+    },
+    {
+      "epoch": 0.712446145329386,
+      "grad_norm": 1.057196021080017,
+      "learning_rate": 5.9439029242092705e-06,
+      "loss": 0.0283,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7130879887035566,
+      "grad_norm": 0.12977996468544006,
+      "learning_rate": 5.93064120416418e-06,
+      "loss": 0.0115,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7137298320777272,
+      "grad_norm": 1.014158010482788,
+      "learning_rate": 5.917379484119091e-06,
+      "loss": 0.0252,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7143716754518978,
+      "grad_norm": 2.2357592582702637,
+      "learning_rate": 5.9041177640740015e-06,
+      "loss": 0.0299,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7150135188260685,
+      "grad_norm": 0.671495258808136,
+      "learning_rate": 5.890856044028911e-06,
+      "loss": 0.0199,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7156553622002391,
+      "grad_norm": 5.646879196166992,
+      "learning_rate": 5.87759432398382e-06,
+      "loss": 0.0361,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7162972055744097,
+      "grad_norm": 0.16122344136238098,
+      "learning_rate": 5.864332603938732e-06,
+      "loss": 0.0269,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7169390489485803,
+      "grad_norm": 1.2174190282821655,
+      "learning_rate": 5.851070883893641e-06,
+      "loss": 0.02,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7175808923227509,
+      "grad_norm": 0.2357230931520462,
+      "learning_rate": 5.837809163848551e-06,
+      "loss": 0.0167,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7182227356969215,
+      "grad_norm": 2.912205934524536,
+      "learning_rate": 5.8245474438034626e-06,
+      "loss": 0.0169,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7188645790710921,
+      "grad_norm": 4.412971496582031,
+      "learning_rate": 5.811285723758372e-06,
+      "loss": 0.0242,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7195064224452627,
+      "grad_norm": 5.053261756896973,
+      "learning_rate": 5.7980240037132815e-06,
+      "loss": 0.0276,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7201482658194334,
+      "grad_norm": 0.7474462985992432,
+      "learning_rate": 5.784762283668193e-06,
+      "loss": 0.0176,
+      "step": 11220
+    },
+    {
+      "epoch": 0.720790109193604,
+      "grad_norm": 0.7813795208930969,
+      "learning_rate": 5.771500563623102e-06,
+      "loss": 0.019,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7214319525677747,
+      "grad_norm": 5.405880928039551,
+      "learning_rate": 5.7582388435780125e-06,
+      "loss": 0.0382,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7220737959419453,
+      "grad_norm": 0.7542135119438171,
+      "learning_rate": 5.744977123532922e-06,
+      "loss": 0.0354,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7227156393161159,
+      "grad_norm": 0.039900798350572586,
+      "learning_rate": 5.731715403487833e-06,
+      "loss": 0.0155,
+      "step": 11260
+    },
+    {
+      "epoch": 0.7233574826902865,
+      "grad_norm": 0.44073763489723206,
+      "learning_rate": 5.718453683442743e-06,
+      "loss": 0.0196,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7239993260644572,
+      "grad_norm": 0.04682408273220062,
+      "learning_rate": 5.705191963397653e-06,
+      "loss": 0.0328,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7246411694386278,
+      "grad_norm": 0.0709882602095604,
+      "learning_rate": 5.691930243352563e-06,
+      "loss": 0.0161,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7252830128127984,
+      "grad_norm": 2.7055585384368896,
+      "learning_rate": 5.678668523307474e-06,
+      "loss": 0.013,
+      "step": 11300
+    },
+    {
+      "epoch": 0.725924856186969,
+      "grad_norm": 0.7219206690788269,
+      "learning_rate": 5.665406803262383e-06,
+      "loss": 0.0145,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7265666995611396,
+      "grad_norm": 0.4283300042152405,
+      "learning_rate": 5.652145083217294e-06,
+      "loss": 0.0217,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7272085429353102,
+      "grad_norm": 1.872422218322754,
+      "learning_rate": 5.638883363172204e-06,
+      "loss": 0.0171,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7278503863094808,
+      "grad_norm": 0.4110535681247711,
+      "learning_rate": 5.625621643127114e-06,
+      "loss": 0.0284,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7284922296836515,
+      "grad_norm": 1.1330406665802002,
+      "learning_rate": 5.612359923082024e-06,
+      "loss": 0.0264,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7291340730578221,
+      "grad_norm": 0.03569392114877701,
+      "learning_rate": 5.599098203036935e-06,
+      "loss": 0.03,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7297759164319927,
+      "grad_norm": 0.01630323752760887,
+      "learning_rate": 5.585836482991844e-06,
+      "loss": 0.0202,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7304177598061633,
+      "grad_norm": 0.09886395186185837,
+      "learning_rate": 5.5725747629467545e-06,
+      "loss": 0.0302,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7310596031803339,
+      "grad_norm": 0.378889262676239,
+      "learning_rate": 5.559313042901665e-06,
+      "loss": 0.03,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7317014465545045,
+      "grad_norm": 0.40000680088996887,
+      "learning_rate": 5.546051322856575e-06,
+      "loss": 0.0276,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7323432899286751,
+      "grad_norm": 1.454674482345581,
+      "learning_rate": 5.532789602811485e-06,
+      "loss": 0.0221,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7329851333028458,
+      "grad_norm": 0.9703409671783447,
+      "learning_rate": 5.519527882766396e-06,
+      "loss": 0.0375,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7336269766770164,
+      "grad_norm": 5.4596452713012695,
+      "learning_rate": 5.506266162721305e-06,
+      "loss": 0.0341,
+      "step": 11430
+    },
+    {
+      "epoch": 0.734268820051187,
+      "grad_norm": 0.9400603175163269,
+      "learning_rate": 5.493004442676216e-06,
+      "loss": 0.0263,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7349106634253576,
+      "grad_norm": 0.7164040207862854,
+      "learning_rate": 5.479742722631126e-06,
+      "loss": 0.0145,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7355525067995282,
+      "grad_norm": 0.09559702128171921,
+      "learning_rate": 5.466481002586036e-06,
+      "loss": 0.0212,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7361943501736988,
+      "grad_norm": 0.4513958692550659,
+      "learning_rate": 5.453219282540946e-06,
+      "loss": 0.0267,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7368361935478694,
+      "grad_norm": 0.009279868565499783,
+      "learning_rate": 5.439957562495855e-06,
+      "loss": 0.0144,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7374780369220401,
+      "grad_norm": 1.7123267650604248,
+      "learning_rate": 5.426695842450766e-06,
+      "loss": 0.0311,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7381198802962107,
+      "grad_norm": 1.537968397140503,
+      "learning_rate": 5.413434122405677e-06,
+      "loss": 0.0332,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7387617236703813,
+      "grad_norm": 2.3726558685302734,
+      "learning_rate": 5.400172402360586e-06,
+      "loss": 0.0282,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7394035670445519,
+      "grad_norm": 1.8804244995117188,
+      "learning_rate": 5.386910682315497e-06,
+      "loss": 0.0279,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7400454104187225,
+      "grad_norm": 1.1779565811157227,
+      "learning_rate": 5.373648962270407e-06,
+      "loss": 0.0281,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7406872537928932,
+      "grad_norm": 1.5714831352233887,
+      "learning_rate": 5.360387242225316e-06,
+      "loss": 0.0244,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7413290971670639,
+      "grad_norm": 0.5432150363922119,
+      "learning_rate": 5.3471255221802275e-06,
+      "loss": 0.0152,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7419709405412345,
+      "grad_norm": 0.4140097200870514,
+      "learning_rate": 5.333863802135138e-06,
+      "loss": 0.0193,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7426127839154051,
+      "grad_norm": 0.08977576345205307,
+      "learning_rate": 5.320602082090047e-06,
+      "loss": 0.0494,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7432546272895757,
+      "grad_norm": 0.5331017374992371,
+      "learning_rate": 5.3073403620449585e-06,
+      "loss": 0.0246,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7438964706637463,
+      "grad_norm": 0.1366458535194397,
+      "learning_rate": 5.294078641999868e-06,
+      "loss": 0.03,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7445383140379169,
+      "grad_norm": 7.025169849395752,
+      "learning_rate": 5.2808169219547774e-06,
+      "loss": 0.0223,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7451801574120875,
+      "grad_norm": 0.025911634787917137,
+      "learning_rate": 5.267555201909688e-06,
+      "loss": 0.0221,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7458220007862582,
+      "grad_norm": 1.230663776397705,
+      "learning_rate": 5.254293481864599e-06,
+      "loss": 0.0248,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7464638441604288,
+      "grad_norm": 1.829215168952942,
+      "learning_rate": 5.241031761819508e-06,
+      "loss": 0.0292,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7471056875345994,
+      "grad_norm": 1.1922662258148193,
+      "learning_rate": 5.227770041774418e-06,
+      "loss": 0.0125,
+      "step": 11640
+    },
+    {
+      "epoch": 0.74774753090877,
+      "grad_norm": 0.18102401494979858,
+      "learning_rate": 5.214508321729329e-06,
+      "loss": 0.0125,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7483893742829406,
+      "grad_norm": 0.04952188581228256,
+      "learning_rate": 5.2012466016842385e-06,
+      "loss": 0.0144,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7490312176571112,
+      "grad_norm": 0.40653935074806213,
+      "learning_rate": 5.187984881639149e-06,
+      "loss": 0.0149,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7496730610312818,
+      "grad_norm": 2.0297279357910156,
+      "learning_rate": 5.17472316159406e-06,
+      "loss": 0.0232,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7503149044054525,
+      "grad_norm": 1.4835153818130493,
+      "learning_rate": 5.1614614415489695e-06,
+      "loss": 0.0114,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7509567477796231,
+      "grad_norm": 0.1499241292476654,
+      "learning_rate": 5.148199721503879e-06,
+      "loss": 0.0109,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7515985911537937,
+      "grad_norm": 4.053781509399414,
+      "learning_rate": 5.13493800145879e-06,
+      "loss": 0.0152,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7522404345279643,
+      "grad_norm": 0.0353599414229393,
+      "learning_rate": 5.1216762814137e-06,
+      "loss": 0.0158,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7528822779021349,
+      "grad_norm": 0.36769697070121765,
+      "learning_rate": 5.10841456136861e-06,
+      "loss": 0.0146,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7535241212763055,
+      "grad_norm": 1.03465735912323,
+      "learning_rate": 5.0951528413235194e-06,
+      "loss": 0.0285,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7541659646504761,
+      "grad_norm": 0.2146172970533371,
+      "learning_rate": 5.081891121278431e-06,
+      "loss": 0.021,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7548078080246468,
+      "grad_norm": 0.05770499259233475,
+      "learning_rate": 5.06862940123334e-06,
+      "loss": 0.0268,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7554496513988174,
+      "grad_norm": 1.8668345212936401,
+      "learning_rate": 5.05536768118825e-06,
+      "loss": 0.0312,
+      "step": 11770
+    },
+    {
+      "epoch": 0.756091494772988,
+      "grad_norm": 1.0731773376464844,
+      "learning_rate": 5.042105961143161e-06,
+      "loss": 0.0298,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7567333381471586,
+      "grad_norm": 0.1279563009738922,
+      "learning_rate": 5.028844241098071e-06,
+      "loss": 0.0183,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7573751815213292,
+      "grad_norm": 0.774218738079071,
+      "learning_rate": 5.0155825210529805e-06,
+      "loss": 0.0139,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7580170248954998,
+      "grad_norm": 0.9003334045410156,
+      "learning_rate": 5.002320801007892e-06,
+      "loss": 0.0223,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7586588682696704,
+      "grad_norm": 0.5011104941368103,
+      "learning_rate": 4.989059080962801e-06,
+      "loss": 0.0193,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7593007116438412,
+      "grad_norm": 0.020797917619347572,
+      "learning_rate": 4.9757973609177115e-06,
+      "loss": 0.0285,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7599425550180118,
+      "grad_norm": 0.2721422612667084,
+      "learning_rate": 4.962535640872622e-06,
+      "loss": 0.0282,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7605843983921824,
+      "grad_norm": 4.101003170013428,
+      "learning_rate": 4.949273920827532e-06,
+      "loss": 0.0191,
+      "step": 11850
+    },
+    {
+      "epoch": 0.761226241766353,
+      "grad_norm": 0.16916042566299438,
+      "learning_rate": 4.936012200782442e-06,
+      "loss": 0.016,
+      "step": 11860
+    },
+    {
+      "epoch": 0.7618680851405236,
+      "grad_norm": 0.5883210301399231,
+      "learning_rate": 4.922750480737352e-06,
+      "loss": 0.0136,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7625099285146942,
+      "grad_norm": 0.5709671378135681,
+      "learning_rate": 4.909488760692262e-06,
+      "loss": 0.0191,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7631517718888648,
+      "grad_norm": 1.265303611755371,
+      "learning_rate": 4.896227040647173e-06,
+      "loss": 0.0156,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7637936152630355,
+      "grad_norm": 0.45923763513565063,
+      "learning_rate": 4.882965320602083e-06,
+      "loss": 0.0299,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7644354586372061,
+      "grad_norm": 2.511776924133301,
+      "learning_rate": 4.8697036005569924e-06,
+      "loss": 0.0202,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7650773020113767,
+      "grad_norm": 4.2928595542907715,
+      "learning_rate": 4.856441880511903e-06,
+      "loss": 0.0342,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7657191453855473,
+      "grad_norm": 0.0511680543422699,
+      "learning_rate": 4.843180160466813e-06,
+      "loss": 0.014,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7663609887597179,
+      "grad_norm": 0.743169903755188,
+      "learning_rate": 4.829918440421723e-06,
+      "loss": 0.0244,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7670028321338885,
+      "grad_norm": 0.6161398887634277,
+      "learning_rate": 4.816656720376634e-06,
+      "loss": 0.0128,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7676446755080591,
+      "grad_norm": 0.8354362845420837,
+      "learning_rate": 4.803395000331543e-06,
+      "loss": 0.0229,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7682865188822298,
+      "grad_norm": 0.6134613156318665,
+      "learning_rate": 4.7901332802864535e-06,
+      "loss": 0.0188,
+      "step": 11970
+    },
+    {
+      "epoch": 0.7689283622564004,
+      "grad_norm": 0.854161262512207,
+      "learning_rate": 4.776871560241364e-06,
+      "loss": 0.019,
+      "step": 11980
+    },
+    {
+      "epoch": 0.769570205630571,
+      "grad_norm": 1.0251282453536987,
+      "learning_rate": 4.763609840196274e-06,
+      "loss": 0.02,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7702120490047416,
+      "grad_norm": 0.2006016969680786,
+      "learning_rate": 4.7503481201511845e-06,
+      "loss": 0.0191,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7708538923789122,
+      "grad_norm": 2.0554537773132324,
+      "learning_rate": 4.737086400106094e-06,
+      "loss": 0.0169,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7714957357530828,
+      "grad_norm": 0.5309330821037292,
+      "learning_rate": 4.723824680061004e-06,
+      "loss": 0.014,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7721375791272534,
+      "grad_norm": 0.23457559943199158,
+      "learning_rate": 4.710562960015914e-06,
+      "loss": 0.0185,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7727794225014241,
+      "grad_norm": 1.5569202899932861,
+      "learning_rate": 4.697301239970825e-06,
+      "loss": 0.0201,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7734212658755947,
+      "grad_norm": 0.03183150291442871,
+      "learning_rate": 4.684039519925735e-06,
+      "loss": 0.0124,
+      "step": 12050
+    },
+    {
+      "epoch": 0.7740631092497653,
+      "grad_norm": 1.1277698278427124,
+      "learning_rate": 4.670777799880645e-06,
+      "loss": 0.0149,
+      "step": 12060
+    },
+    {
+      "epoch": 0.7747049526239359,
+      "grad_norm": 0.15539094805717468,
+      "learning_rate": 4.657516079835555e-06,
+      "loss": 0.0352,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7753467959981065,
+      "grad_norm": 0.05271551385521889,
+      "learning_rate": 4.644254359790465e-06,
+      "loss": 0.0087,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7759886393722771,
+      "grad_norm": 0.9591747522354126,
+      "learning_rate": 4.630992639745375e-06,
+      "loss": 0.0201,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7766304827464477,
+      "grad_norm": 0.04693359509110451,
+      "learning_rate": 4.617730919700285e-06,
+      "loss": 0.027,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7772723261206184,
+      "grad_norm": 0.5332335233688354,
+      "learning_rate": 4.6044691996551955e-06,
+      "loss": 0.0159,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7779141694947891,
+      "grad_norm": 0.5745136737823486,
+      "learning_rate": 4.591207479610106e-06,
+      "loss": 0.0435,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7785560128689597,
+      "grad_norm": 0.2693396210670471,
+      "learning_rate": 4.577945759565016e-06,
+      "loss": 0.0255,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7791978562431303,
+      "grad_norm": 0.05429879203438759,
+      "learning_rate": 4.564684039519926e-06,
+      "loss": 0.0226,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7798396996173009,
+      "grad_norm": 5.823631286621094,
+      "learning_rate": 4.551422319474836e-06,
+      "loss": 0.0164,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7804815429914715,
+      "grad_norm": 0.052716001868247986,
+      "learning_rate": 4.538160599429746e-06,
+      "loss": 0.0231,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7811233863656422,
+      "grad_norm": 0.6026047468185425,
+      "learning_rate": 4.524898879384657e-06,
+      "loss": 0.0085,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7817652297398128,
+      "grad_norm": 0.8200371861457825,
+      "learning_rate": 4.511637159339567e-06,
+      "loss": 0.0269,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7824070731139834,
+      "grad_norm": 1.132227897644043,
+      "learning_rate": 4.4983754392944765e-06,
+      "loss": 0.0489,
+      "step": 12190
+    },
+    {
+      "epoch": 0.783048916488154,
+      "grad_norm": 0.8593686819076538,
+      "learning_rate": 4.485113719249387e-06,
+      "loss": 0.0203,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7836907598623246,
+      "grad_norm": 1.9194788932800293,
+      "learning_rate": 4.471851999204297e-06,
+      "loss": 0.02,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7843326032364952,
+      "grad_norm": 5.477539539337158,
+      "learning_rate": 4.4585902791592074e-06,
+      "loss": 0.0317,
+      "step": 12220
+    },
+    {
+      "epoch": 0.7849744466106658,
+      "grad_norm": 0.8278726935386658,
+      "learning_rate": 4.445328559114118e-06,
+      "loss": 0.0174,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7856162899848365,
+      "grad_norm": 2.025526762008667,
+      "learning_rate": 4.432066839069027e-06,
+      "loss": 0.018,
+      "step": 12240
+    },
+    {
+      "epoch": 0.7862581333590071,
+      "grad_norm": 0.141770601272583,
+      "learning_rate": 4.4188051190239376e-06,
+      "loss": 0.0251,
+      "step": 12250
+    },
+    {
+      "epoch": 0.7868999767331777,
+      "grad_norm": 0.11860339343547821,
+      "learning_rate": 4.405543398978848e-06,
+      "loss": 0.0143,
+      "step": 12260
+    },
+    {
+      "epoch": 0.7875418201073483,
+      "grad_norm": 1.2235369682312012,
+      "learning_rate": 4.392281678933758e-06,
+      "loss": 0.0373,
+      "step": 12270
+    },
+    {
+      "epoch": 0.7881836634815189,
+      "grad_norm": 0.4883560240268707,
+      "learning_rate": 4.3790199588886685e-06,
+      "loss": 0.0307,
+      "step": 12280
+    },
+    {
+      "epoch": 0.7888255068556895,
+      "grad_norm": 3.3470137119293213,
+      "learning_rate": 4.365758238843578e-06,
+      "loss": 0.0108,
+      "step": 12290
+    },
+    {
+      "epoch": 0.7894673502298601,
+      "grad_norm": 0.5170362591743469,
+      "learning_rate": 4.352496518798488e-06,
+      "loss": 0.02,
+      "step": 12300
+    },
+    {
+      "epoch": 0.7901091936040308,
+      "grad_norm": 0.05397709831595421,
+      "learning_rate": 4.339234798753399e-06,
+      "loss": 0.0178,
+      "step": 12310
+    },
+    {
+      "epoch": 0.7907510369782014,
+      "grad_norm": 2.172746419906616,
+      "learning_rate": 4.325973078708309e-06,
+      "loss": 0.0161,
+      "step": 12320
+    },
+    {
+      "epoch": 0.791392880352372,
+      "grad_norm": 0.5354034900665283,
+      "learning_rate": 4.312711358663219e-06,
+      "loss": 0.0215,
+      "step": 12330
+    },
+    {
+      "epoch": 0.7920347237265426,
+      "grad_norm": 0.03487497940659523,
+      "learning_rate": 4.299449638618129e-06,
+      "loss": 0.018,
+      "step": 12340
+    },
+    {
+      "epoch": 0.7926765671007132,
+      "grad_norm": 0.5163410305976868,
+      "learning_rate": 4.286187918573039e-06,
+      "loss": 0.0216,
+      "step": 12350
+    },
+    {
+      "epoch": 0.7933184104748838,
+      "grad_norm": 2.64249324798584,
+      "learning_rate": 4.2729261985279494e-06,
+      "loss": 0.025,
+      "step": 12360
+    },
+    {
+      "epoch": 0.7939602538490544,
+      "grad_norm": 0.49127480387687683,
+      "learning_rate": 4.25966447848286e-06,
+      "loss": 0.0157,
+      "step": 12370
+    },
+    {
+      "epoch": 0.7946020972232251,
+      "grad_norm": 0.04757243022322655,
+      "learning_rate": 4.24640275843777e-06,
+      "loss": 0.0132,
+      "step": 12380
+    },
+    {
+      "epoch": 0.7952439405973957,
+      "grad_norm": 0.28864434361457825,
+      "learning_rate": 4.2331410383926796e-06,
+      "loss": 0.0368,
+      "step": 12390
+    },
+    {
+      "epoch": 0.7958857839715663,
+      "grad_norm": 2.3907408714294434,
+      "learning_rate": 4.21987931834759e-06,
+      "loss": 0.0155,
+      "step": 12400
+    },
+    {
+      "epoch": 0.796527627345737,
+      "grad_norm": 0.3429820239543915,
+      "learning_rate": 4.2066175983025e-06,
+      "loss": 0.0155,
+      "step": 12410
+    },
+    {
+      "epoch": 0.7971694707199076,
+      "grad_norm": 0.8417637944221497,
+      "learning_rate": 4.1933558782574105e-06,
+      "loss": 0.0234,
+      "step": 12420
+    },
+    {
+      "epoch": 0.7978113140940782,
+      "grad_norm": 2.053241014480591,
+      "learning_rate": 4.180094158212321e-06,
+      "loss": 0.0259,
+      "step": 12430
+    },
+    {
+      "epoch": 0.7984531574682489,
+      "grad_norm": 3.192744731903076,
+      "learning_rate": 4.16683243816723e-06,
+      "loss": 0.0431,
+      "step": 12440
+    },
+    {
+      "epoch": 0.7990950008424195,
+      "grad_norm": 0.042512197047472,
+      "learning_rate": 4.153570718122141e-06,
+      "loss": 0.0114,
+      "step": 12450
+    },
+    {
+      "epoch": 0.7997368442165901,
+      "grad_norm": 0.07811237871646881,
+      "learning_rate": 4.140308998077051e-06,
+      "loss": 0.0147,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8003786875907607,
+      "grad_norm": 0.7863467335700989,
+      "learning_rate": 4.127047278031961e-06,
+      "loss": 0.0103,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8010205309649313,
+      "grad_norm": 0.0853208601474762,
+      "learning_rate": 4.113785557986872e-06,
+      "loss": 0.0225,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8016623743391019,
+      "grad_norm": 0.1661083996295929,
+      "learning_rate": 4.100523837941782e-06,
+      "loss": 0.0276,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8023042177132725,
+      "grad_norm": 0.7978856563568115,
+      "learning_rate": 4.0872621178966915e-06,
+      "loss": 0.03,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8029460610874432,
+      "grad_norm": 5.746982097625732,
+      "learning_rate": 4.074000397851602e-06,
+      "loss": 0.0242,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8035879044616138,
+      "grad_norm": 0.921909511089325,
+      "learning_rate": 4.060738677806511e-06,
+      "loss": 0.0114,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8042297478357844,
+      "grad_norm": 0.4040222465991974,
+      "learning_rate": 4.0474769577614224e-06,
+      "loss": 0.0117,
+      "step": 12530
+    },
+    {
+      "epoch": 0.804871591209955,
+      "grad_norm": 0.5128666758537292,
+      "learning_rate": 4.034215237716333e-06,
+      "loss": 0.0222,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8055134345841256,
+      "grad_norm": 0.8506461977958679,
+      "learning_rate": 4.020953517671242e-06,
+      "loss": 0.0242,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8061552779582962,
+      "grad_norm": 0.9422674179077148,
+      "learning_rate": 4.0076917976261526e-06,
+      "loss": 0.0264,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8067971213324668,
+      "grad_norm": 0.8520615696907043,
+      "learning_rate": 3.994430077581062e-06,
+      "loss": 0.0115,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8074389647066375,
+      "grad_norm": 0.06659626215696335,
+      "learning_rate": 3.981168357535972e-06,
+      "loss": 0.0207,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.9921087026596069,
+      "learning_rate": 3.9679066374908835e-06,
+      "loss": 0.0154,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8087226514549787,
+      "grad_norm": 0.06613270938396454,
+      "learning_rate": 3.954644917445793e-06,
+      "loss": 0.0121,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8093644948291493,
+      "grad_norm": 0.22420759499073029,
+      "learning_rate": 3.941383197400703e-06,
+      "loss": 0.0278,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8100063382033199,
+      "grad_norm": 1.3557567596435547,
+      "learning_rate": 3.928121477355613e-06,
+      "loss": 0.0187,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8106481815774905,
+      "grad_norm": 1.4381176233291626,
+      "learning_rate": 3.914859757310523e-06,
+      "loss": 0.022,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8112900249516611,
+      "grad_norm": 0.49772679805755615,
+      "learning_rate": 3.9015980372654335e-06,
+      "loss": 0.0105,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8119318683258318,
+      "grad_norm": 1.049343466758728,
+      "learning_rate": 3.888336317220344e-06,
+      "loss": 0.0229,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8125737117000024,
+      "grad_norm": 3.795696496963501,
+      "learning_rate": 3.875074597175254e-06,
+      "loss": 0.0209,
+      "step": 12660
+    },
+    {
+      "epoch": 0.813215555074173,
+      "grad_norm": 0.19811248779296875,
+      "learning_rate": 3.861812877130164e-06,
+      "loss": 0.0096,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8138573984483436,
+      "grad_norm": 0.3538595736026764,
+      "learning_rate": 3.848551157085074e-06,
+      "loss": 0.0218,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8144992418225142,
+      "grad_norm": 0.6877108812332153,
+      "learning_rate": 3.835289437039984e-06,
+      "loss": 0.0391,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8151410851966849,
+      "grad_norm": 0.044599391520023346,
+      "learning_rate": 3.8220277169948946e-06,
+      "loss": 0.0245,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8157829285708555,
+      "grad_norm": 0.8634452819824219,
+      "learning_rate": 3.808765996949805e-06,
+      "loss": 0.0241,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8164247719450262,
+      "grad_norm": 0.04034089669585228,
+      "learning_rate": 3.7955042769047152e-06,
+      "loss": 0.0278,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8170666153191968,
+      "grad_norm": 0.028350872918963432,
+      "learning_rate": 3.782242556859625e-06,
+      "loss": 0.0108,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8177084586933674,
+      "grad_norm": 0.8466787934303284,
+      "learning_rate": 3.7689808368145354e-06,
+      "loss": 0.0154,
+      "step": 12740
+    },
+    {
+      "epoch": 0.818350302067538,
+      "grad_norm": 0.27883994579315186,
+      "learning_rate": 3.755719116769445e-06,
+      "loss": 0.016,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8189921454417086,
+      "grad_norm": 2.9189789295196533,
+      "learning_rate": 3.7424573967243557e-06,
+      "loss": 0.0158,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8196339888158792,
+      "grad_norm": 0.6996471881866455,
+      "learning_rate": 3.729195676679266e-06,
+      "loss": 0.0141,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8202758321900498,
+      "grad_norm": 0.2102804332971573,
+      "learning_rate": 3.7159339566341755e-06,
+      "loss": 0.0135,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8209176755642205,
+      "grad_norm": 0.013778652995824814,
+      "learning_rate": 3.7026722365890862e-06,
+      "loss": 0.014,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8215595189383911,
+      "grad_norm": 0.8039547204971313,
+      "learning_rate": 3.6894105165439957e-06,
+      "loss": 0.0134,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8222013623125617,
+      "grad_norm": 3.9042227268218994,
+      "learning_rate": 3.676148796498906e-06,
+      "loss": 0.0202,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8228432056867323,
+      "grad_norm": 0.5314901471138,
+      "learning_rate": 3.6628870764538168e-06,
+      "loss": 0.0164,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8234850490609029,
+      "grad_norm": 0.8493812680244446,
+      "learning_rate": 3.6496253564087263e-06,
+      "loss": 0.022,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8241268924350735,
+      "grad_norm": 1.8056285381317139,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.0339,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8247687358092441,
+      "grad_norm": 0.9738245606422424,
+      "learning_rate": 3.6231019163185465e-06,
+      "loss": 0.0314,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8254105791834148,
+      "grad_norm": 6.427189350128174,
+      "learning_rate": 3.609840196273457e-06,
+      "loss": 0.0307,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8260524225575854,
+      "grad_norm": 0.38021722435951233,
+      "learning_rate": 3.596578476228367e-06,
+      "loss": 0.0084,
+      "step": 12870
+    },
+    {
+      "epoch": 0.826694265931756,
+      "grad_norm": 0.674687922000885,
+      "learning_rate": 3.583316756183277e-06,
+      "loss": 0.021,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8273361093059266,
+      "grad_norm": 2.1990878582000732,
+      "learning_rate": 3.5700550361381874e-06,
+      "loss": 0.0144,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8279779526800972,
+      "grad_norm": 0.09193415939807892,
+      "learning_rate": 3.5567933160930973e-06,
+      "loss": 0.0214,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8286197960542678,
+      "grad_norm": 0.4719434082508087,
+      "learning_rate": 3.5435315960480076e-06,
+      "loss": 0.023,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8292616394284384,
+      "grad_norm": 0.2065659463405609,
+      "learning_rate": 3.530269876002918e-06,
+      "loss": 0.0231,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8299034828026091,
+      "grad_norm": 0.38893452286720276,
+      "learning_rate": 3.517008155957828e-06,
+      "loss": 0.0246,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8305453261767797,
+      "grad_norm": 1.6100575923919678,
+      "learning_rate": 3.503746435912738e-06,
+      "loss": 0.0173,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8311871695509503,
+      "grad_norm": 0.8549080491065979,
+      "learning_rate": 3.4904847158676485e-06,
+      "loss": 0.0355,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8318290129251209,
+      "grad_norm": 1.1523066759109497,
+      "learning_rate": 3.4772229958225584e-06,
+      "loss": 0.026,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8324708562992915,
+      "grad_norm": 0.8753595352172852,
+      "learning_rate": 3.4639612757774687e-06,
+      "loss": 0.0283,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8331126996734621,
+      "grad_norm": 0.05473419278860092,
+      "learning_rate": 3.4506995557323786e-06,
+      "loss": 0.0117,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8337545430476329,
+      "grad_norm": 0.1066131442785263,
+      "learning_rate": 3.437437835687289e-06,
+      "loss": 0.0189,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8343963864218035,
+      "grad_norm": 1.1916576623916626,
+      "learning_rate": 3.4241761156421992e-06,
+      "loss": 0.0299,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8350382297959741,
+      "grad_norm": 0.3121644854545593,
+      "learning_rate": 3.410914395597109e-06,
+      "loss": 0.0301,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8356800731701447,
+      "grad_norm": 0.024094169959425926,
+      "learning_rate": 3.3976526755520195e-06,
+      "loss": 0.0063,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8363219165443153,
+      "grad_norm": 0.6124148964881897,
+      "learning_rate": 3.3843909555069294e-06,
+      "loss": 0.0212,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8369637599184859,
+      "grad_norm": 4.273738861083984,
+      "learning_rate": 3.3711292354618397e-06,
+      "loss": 0.0227,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8376056032926565,
+      "grad_norm": 0.7729155421257019,
+      "learning_rate": 3.35786751541675e-06,
+      "loss": 0.0144,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8382474466668272,
+      "grad_norm": 0.6459729075431824,
+      "learning_rate": 3.34460579537166e-06,
+      "loss": 0.0243,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8388892900409978,
+      "grad_norm": 0.844441831111908,
+      "learning_rate": 3.3313440753265703e-06,
+      "loss": 0.0239,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8395311334151684,
+      "grad_norm": 0.9636744260787964,
+      "learning_rate": 3.31808235528148e-06,
+      "loss": 0.0183,
+      "step": 13080
+    },
+    {
+      "epoch": 0.840172976789339,
+      "grad_norm": 0.8931861519813538,
+      "learning_rate": 3.3048206352363905e-06,
+      "loss": 0.0296,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8408148201635096,
+      "grad_norm": 0.22203336656093597,
+      "learning_rate": 3.291558915191301e-06,
+      "loss": 0.0167,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8414566635376802,
+      "grad_norm": 1.675729751586914,
+      "learning_rate": 3.2782971951462107e-06,
+      "loss": 0.0187,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8420985069118508,
+      "grad_norm": 0.29589271545410156,
+      "learning_rate": 3.265035475101121e-06,
+      "loss": 0.0273,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8427403502860215,
+      "grad_norm": 2.055467367172241,
+      "learning_rate": 3.251773755056031e-06,
+      "loss": 0.0188,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8433821936601921,
+      "grad_norm": 3.291701078414917,
+      "learning_rate": 3.2385120350109413e-06,
+      "loss": 0.0133,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8440240370343627,
+      "grad_norm": 0.6344001889228821,
+      "learning_rate": 3.2252503149658516e-06,
+      "loss": 0.0246,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8446658804085333,
+      "grad_norm": 0.5008775591850281,
+      "learning_rate": 3.2119885949207615e-06,
+      "loss": 0.0297,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8453077237827039,
+      "grad_norm": 0.3949437439441681,
+      "learning_rate": 3.198726874875672e-06,
+      "loss": 0.0275,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8459495671568745,
+      "grad_norm": 2.543548345565796,
+      "learning_rate": 3.185465154830582e-06,
+      "loss": 0.0219,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8465914105310451,
+      "grad_norm": 0.8906174302101135,
+      "learning_rate": 3.172203434785492e-06,
+      "loss": 0.0162,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8472332539052158,
+      "grad_norm": 7.785815238952637,
+      "learning_rate": 3.1589417147404024e-06,
+      "loss": 0.0298,
+      "step": 13200
+    },
+    {
+      "epoch": 0.8478750972793864,
+      "grad_norm": 5.277609825134277,
+      "learning_rate": 3.145679994695312e-06,
+      "loss": 0.0329,
+      "step": 13210
+    },
+    {
+      "epoch": 0.848516940653557,
+      "grad_norm": 2.5439512729644775,
+      "learning_rate": 3.1324182746502226e-06,
+      "loss": 0.0244,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8491587840277276,
+      "grad_norm": 1.0475541353225708,
+      "learning_rate": 3.119156554605133e-06,
+      "loss": 0.0211,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8498006274018982,
+      "grad_norm": 0.20279256999492645,
+      "learning_rate": 3.1058948345600424e-06,
+      "loss": 0.0187,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8504424707760688,
+      "grad_norm": 1.6604892015457153,
+      "learning_rate": 3.092633114514953e-06,
+      "loss": 0.0213,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8510843141502394,
+      "grad_norm": 0.35537654161453247,
+      "learning_rate": 3.0793713944698626e-06,
+      "loss": 0.0362,
+      "step": 13260
+    },
+    {
+      "epoch": 0.85172615752441,
+      "grad_norm": 1.9353057146072388,
+      "learning_rate": 3.066109674424773e-06,
+      "loss": 0.0195,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8523680008985807,
+      "grad_norm": 0.24915455281734467,
+      "learning_rate": 3.0528479543796837e-06,
+      "loss": 0.0239,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8530098442727514,
+      "grad_norm": 0.3387570381164551,
+      "learning_rate": 3.039586234334593e-06,
+      "loss": 0.017,
+      "step": 13290
+    },
+    {
+      "epoch": 0.853651687646922,
+      "grad_norm": 0.6907055377960205,
+      "learning_rate": 3.0263245142895035e-06,
+      "loss": 0.0212,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8542935310210926,
+      "grad_norm": 0.5599414110183716,
+      "learning_rate": 3.0130627942444134e-06,
+      "loss": 0.0354,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8549353743952632,
+      "grad_norm": 0.10817236453294754,
+      "learning_rate": 2.9998010741993237e-06,
+      "loss": 0.0124,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8555772177694339,
+      "grad_norm": 0.4376680254936218,
+      "learning_rate": 2.986539354154234e-06,
+      "loss": 0.0218,
+      "step": 13330
+    },
+    {
+      "epoch": 0.8562190611436045,
+      "grad_norm": 0.790965735912323,
+      "learning_rate": 2.973277634109144e-06,
+      "loss": 0.0219,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8568609045177751,
+      "grad_norm": 0.57417893409729,
+      "learning_rate": 2.9600159140640543e-06,
+      "loss": 0.0179,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8575027478919457,
+      "grad_norm": 0.20554496347904205,
+      "learning_rate": 2.946754194018964e-06,
+      "loss": 0.0215,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8581445912661163,
+      "grad_norm": 0.1840430051088333,
+      "learning_rate": 2.9334924739738745e-06,
+      "loss": 0.0194,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8587864346402869,
+      "grad_norm": 0.9449292421340942,
+      "learning_rate": 2.920230753928785e-06,
+      "loss": 0.0177,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8594282780144575,
+      "grad_norm": 0.7500403523445129,
+      "learning_rate": 2.9069690338836947e-06,
+      "loss": 0.044,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8600701213886282,
+      "grad_norm": 4.72554349899292,
+      "learning_rate": 2.893707313838605e-06,
+      "loss": 0.0221,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8607119647627988,
+      "grad_norm": 3.8151938915252686,
+      "learning_rate": 2.8804455937935154e-06,
+      "loss": 0.0208,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8613538081369694,
+      "grad_norm": 2.1088836193084717,
+      "learning_rate": 2.8671838737484253e-06,
+      "loss": 0.0342,
+      "step": 13420
+    },
+    {
+      "epoch": 0.86199565151114,
+      "grad_norm": 2.671661138534546,
+      "learning_rate": 2.8539221537033356e-06,
+      "loss": 0.0301,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8626374948853106,
+      "grad_norm": 0.2565511167049408,
+      "learning_rate": 2.8406604336582455e-06,
+      "loss": 0.019,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8632793382594812,
+      "grad_norm": 0.8613461256027222,
+      "learning_rate": 2.827398713613156e-06,
+      "loss": 0.0124,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8639211816336518,
+      "grad_norm": 0.024694399908185005,
+      "learning_rate": 2.814136993568066e-06,
+      "loss": 0.0274,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8645630250078224,
+      "grad_norm": 0.3943841755390167,
+      "learning_rate": 2.800875273522976e-06,
+      "loss": 0.0173,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8652048683819931,
+      "grad_norm": 0.1581508219242096,
+      "learning_rate": 2.7876135534778864e-06,
+      "loss": 0.0091,
+      "step": 13480
+    },
+    {
+      "epoch": 0.8658467117561637,
+      "grad_norm": 0.8672017455101013,
+      "learning_rate": 2.7743518334327963e-06,
+      "loss": 0.0178,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8664885551303343,
+      "grad_norm": 0.06939632445573807,
+      "learning_rate": 2.7610901133877066e-06,
+      "loss": 0.0228,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8671303985045049,
+      "grad_norm": 0.6396413445472717,
+      "learning_rate": 2.747828393342617e-06,
+      "loss": 0.0237,
+      "step": 13510
+    },
+    {
+      "epoch": 0.8677722418786755,
+      "grad_norm": 0.2894430458545685,
+      "learning_rate": 2.734566673297527e-06,
+      "loss": 0.012,
+      "step": 13520
+    },
+    {
+      "epoch": 0.8684140852528461,
+      "grad_norm": 0.6599921584129333,
+      "learning_rate": 2.721304953252437e-06,
+      "loss": 0.0208,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8690559286270167,
+      "grad_norm": 0.015496225096285343,
+      "learning_rate": 2.708043233207347e-06,
+      "loss": 0.0197,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8696977720011874,
+      "grad_norm": 0.09208979457616806,
+      "learning_rate": 2.6947815131622574e-06,
+      "loss": 0.021,
+      "step": 13550
+    },
+    {
+      "epoch": 0.870339615375358,
+      "grad_norm": 0.05079632252454758,
+      "learning_rate": 2.6815197931171677e-06,
+      "loss": 0.0154,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8709814587495286,
+      "grad_norm": 0.5705692172050476,
+      "learning_rate": 2.6682580730720776e-06,
+      "loss": 0.0187,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8716233021236993,
+      "grad_norm": 1.234823226928711,
+      "learning_rate": 2.654996353026988e-06,
+      "loss": 0.0292,
+      "step": 13580
+    },
+    {
+      "epoch": 0.8722651454978699,
+      "grad_norm": 0.5673630237579346,
+      "learning_rate": 2.641734632981898e-06,
+      "loss": 0.0275,
+      "step": 13590
+    },
+    {
+      "epoch": 0.8729069888720405,
+      "grad_norm": 0.007527688052505255,
+      "learning_rate": 2.628472912936808e-06,
+      "loss": 0.0218,
+      "step": 13600
+    },
+    {
+      "epoch": 0.8735488322462112,
+      "grad_norm": 1.007513165473938,
+      "learning_rate": 2.6152111928917185e-06,
+      "loss": 0.0129,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8741906756203818,
+      "grad_norm": 1.1092039346694946,
+      "learning_rate": 2.6019494728466284e-06,
+      "loss": 0.0167,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8748325189945524,
+      "grad_norm": 1.0379366874694824,
+      "learning_rate": 2.5886877528015387e-06,
+      "loss": 0.0184,
+      "step": 13630
+    },
+    {
+      "epoch": 0.875474362368723,
+      "grad_norm": 0.5166342854499817,
+      "learning_rate": 2.575426032756449e-06,
+      "loss": 0.0265,
+      "step": 13640
+    },
+    {
+      "epoch": 0.8761162057428936,
+      "grad_norm": 1.019518256187439,
+      "learning_rate": 2.562164312711359e-06,
+      "loss": 0.0188,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8767580491170642,
+      "grad_norm": 0.8891957402229309,
+      "learning_rate": 2.5489025926662693e-06,
+      "loss": 0.0195,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8773998924912348,
+      "grad_norm": 0.3497367799282074,
+      "learning_rate": 2.535640872621179e-06,
+      "loss": 0.0279,
+      "step": 13670
+    },
+    {
+      "epoch": 0.8780417358654055,
+      "grad_norm": 0.20192396640777588,
+      "learning_rate": 2.5223791525760895e-06,
+      "loss": 0.0227,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8786835792395761,
+      "grad_norm": 0.2110779732465744,
+      "learning_rate": 2.509117432531e-06,
+      "loss": 0.0164,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8793254226137467,
+      "grad_norm": 0.18230389058589935,
+      "learning_rate": 2.4958557124859093e-06,
+      "loss": 0.02,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8799672659879173,
+      "grad_norm": 0.035150956362485886,
+      "learning_rate": 2.48259399244082e-06,
+      "loss": 0.022,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8806091093620879,
+      "grad_norm": 2.311920166015625,
+      "learning_rate": 2.46933227239573e-06,
+      "loss": 0.0207,
+      "step": 13720
+    },
+    {
+      "epoch": 0.8812509527362585,
+      "grad_norm": 0.016587447375059128,
+      "learning_rate": 2.45607055235064e-06,
+      "loss": 0.0206,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8818927961104291,
+      "grad_norm": 0.6804785132408142,
+      "learning_rate": 2.44280883230555e-06,
+      "loss": 0.0222,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8825346394845998,
+      "grad_norm": 0.03751286119222641,
+      "learning_rate": 2.42954711226046e-06,
+      "loss": 0.0077,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8831764828587704,
+      "grad_norm": 0.02884281426668167,
+      "learning_rate": 2.4162853922153704e-06,
+      "loss": 0.0218,
+      "step": 13760
+    },
+    {
+      "epoch": 0.883818326232941,
+      "grad_norm": 0.7173253893852234,
+      "learning_rate": 2.4030236721702807e-06,
+      "loss": 0.0301,
+      "step": 13770
+    },
+    {
+      "epoch": 0.8844601696071116,
+      "grad_norm": 2.145676851272583,
+      "learning_rate": 2.3897619521251906e-06,
+      "loss": 0.0274,
+      "step": 13780
+    },
+    {
+      "epoch": 0.8851020129812822,
+      "grad_norm": 0.00921548716723919,
+      "learning_rate": 2.376500232080101e-06,
+      "loss": 0.0207,
+      "step": 13790
+    },
+    {
+      "epoch": 0.8857438563554528,
+      "grad_norm": 0.2222190648317337,
+      "learning_rate": 2.3632385120350113e-06,
+      "loss": 0.0294,
+      "step": 13800
+    },
+    {
+      "epoch": 0.8863856997296234,
+      "grad_norm": 0.701604962348938,
+      "learning_rate": 2.349976791989921e-06,
+      "loss": 0.0127,
+      "step": 13810
+    },
+    {
+      "epoch": 0.8870275431037941,
+      "grad_norm": 1.019660472869873,
+      "learning_rate": 2.3367150719448315e-06,
+      "loss": 0.0241,
+      "step": 13820
+    },
+    {
+      "epoch": 0.8876693864779647,
+      "grad_norm": 4.041540145874023,
+      "learning_rate": 2.3234533518997414e-06,
+      "loss": 0.0181,
+      "step": 13830
+    },
+    {
+      "epoch": 0.8883112298521353,
+      "grad_norm": 0.7163257598876953,
+      "learning_rate": 2.3101916318546517e-06,
+      "loss": 0.0111,
+      "step": 13840
+    },
+    {
+      "epoch": 0.8889530732263059,
+      "grad_norm": 0.4597417712211609,
+      "learning_rate": 2.296929911809562e-06,
+      "loss": 0.0328,
+      "step": 13850
+    },
+    {
+      "epoch": 0.8895949166004765,
+      "grad_norm": 0.07502664625644684,
+      "learning_rate": 2.283668191764472e-06,
+      "loss": 0.0309,
+      "step": 13860
+    },
+    {
+      "epoch": 0.8902367599746472,
+      "grad_norm": 0.32292598485946655,
+      "learning_rate": 2.2704064717193823e-06,
+      "loss": 0.014,
+      "step": 13870
+    },
+    {
+      "epoch": 0.8908786033488179,
+      "grad_norm": 0.6518355011940002,
+      "learning_rate": 2.257144751674292e-06,
+      "loss": 0.0152,
+      "step": 13880
+    },
+    {
+      "epoch": 0.8915204467229885,
+      "grad_norm": 0.5202192068099976,
+      "learning_rate": 2.2438830316292025e-06,
+      "loss": 0.0183,
+      "step": 13890
+    },
+    {
+      "epoch": 0.8921622900971591,
+      "grad_norm": 0.04122181981801987,
+      "learning_rate": 2.230621311584113e-06,
+      "loss": 0.0303,
+      "step": 13900
+    },
+    {
+      "epoch": 0.8928041334713297,
+      "grad_norm": 0.09100360423326492,
+      "learning_rate": 2.2173595915390228e-06,
+      "loss": 0.015,
+      "step": 13910
+    },
+    {
+      "epoch": 0.8934459768455003,
+      "grad_norm": 0.41373616456985474,
+      "learning_rate": 2.204097871493933e-06,
+      "loss": 0.022,
+      "step": 13920
+    },
+    {
+      "epoch": 0.8940878202196709,
+      "grad_norm": 0.007583377417176962,
+      "learning_rate": 2.190836151448843e-06,
+      "loss": 0.013,
+      "step": 13930
+    },
+    {
+      "epoch": 0.8947296635938415,
+      "grad_norm": 0.731808602809906,
+      "learning_rate": 2.1775744314037533e-06,
+      "loss": 0.0154,
+      "step": 13940
+    },
+    {
+      "epoch": 0.8953715069680122,
+      "grad_norm": 2.2925314903259277,
+      "learning_rate": 2.1643127113586636e-06,
+      "loss": 0.022,
+      "step": 13950
+    },
+    {
+      "epoch": 0.8960133503421828,
+      "grad_norm": 0.034649644047021866,
+      "learning_rate": 2.1510509913135735e-06,
+      "loss": 0.0147,
+      "step": 13960
+    },
+    {
+      "epoch": 0.8966551937163534,
+      "grad_norm": 0.47291600704193115,
+      "learning_rate": 2.1377892712684834e-06,
+      "loss": 0.01,
+      "step": 13970
+    },
+    {
+      "epoch": 0.897297037090524,
+      "grad_norm": 7.475317001342773,
+      "learning_rate": 2.1245275512233938e-06,
+      "loss": 0.0187,
+      "step": 13980
+    },
+    {
+      "epoch": 0.8979388804646946,
+      "grad_norm": 0.4022504389286041,
+      "learning_rate": 2.111265831178304e-06,
+      "loss": 0.0194,
+      "step": 13990
+    },
+    {
+      "epoch": 0.8985807238388652,
+      "grad_norm": 0.17653189599514008,
+      "learning_rate": 2.098004111133214e-06,
+      "loss": 0.029,
+      "step": 14000
+    },
+    {
+      "epoch": 0.8992225672130358,
+      "grad_norm": 2.2175135612487793,
+      "learning_rate": 2.0847423910881243e-06,
+      "loss": 0.0499,
+      "step": 14010
+    },
+    {
+      "epoch": 0.8998644105872065,
+      "grad_norm": 0.11678501218557358,
+      "learning_rate": 2.0714806710430342e-06,
+      "loss": 0.015,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9005062539613771,
+      "grad_norm": 2.840829849243164,
+      "learning_rate": 2.0582189509979445e-06,
+      "loss": 0.032,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9011480973355477,
+      "grad_norm": 3.5912017822265625,
+      "learning_rate": 2.044957230952855e-06,
+      "loss": 0.0145,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9017899407097183,
+      "grad_norm": 0.4161267876625061,
+      "learning_rate": 2.0316955109077648e-06,
+      "loss": 0.0111,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9024317840838889,
+      "grad_norm": 2.903442621231079,
+      "learning_rate": 2.018433790862675e-06,
+      "loss": 0.0216,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9030736274580595,
+      "grad_norm": 2.279383420944214,
+      "learning_rate": 2.005172070817585e-06,
+      "loss": 0.0265,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9037154708322301,
+      "grad_norm": 6.031994342803955,
+      "learning_rate": 1.9919103507724953e-06,
+      "loss": 0.0236,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9043573142064008,
+      "grad_norm": 1.7201488018035889,
+      "learning_rate": 1.9786486307274056e-06,
+      "loss": 0.0318,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9049991575805714,
+      "grad_norm": 0.8770641684532166,
+      "learning_rate": 1.9653869106823155e-06,
+      "loss": 0.0129,
+      "step": 14100
+    },
+    {
+      "epoch": 0.905641000954742,
+      "grad_norm": 0.23001974821090698,
+      "learning_rate": 1.952125190637226e-06,
+      "loss": 0.0192,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9062828443289126,
+      "grad_norm": 2.5015745162963867,
+      "learning_rate": 1.938863470592136e-06,
+      "loss": 0.0173,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9069246877030832,
+      "grad_norm": 0.13634377717971802,
+      "learning_rate": 1.925601750547046e-06,
+      "loss": 0.0292,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9075665310772538,
+      "grad_norm": 1.0855636596679688,
+      "learning_rate": 1.9123400305019564e-06,
+      "loss": 0.0187,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9082083744514244,
+      "grad_norm": 0.20524372160434723,
+      "learning_rate": 1.8990783104568663e-06,
+      "loss": 0.0153,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9088502178255952,
+      "grad_norm": 0.9702785611152649,
+      "learning_rate": 1.8858165904117764e-06,
+      "loss": 0.0243,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9094920611997658,
+      "grad_norm": 0.0294529739767313,
+      "learning_rate": 1.8725548703666868e-06,
+      "loss": 0.0296,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9101339045739364,
+      "grad_norm": 1.1509523391723633,
+      "learning_rate": 1.8592931503215969e-06,
+      "loss": 0.0264,
+      "step": 14180
+    },
+    {
+      "epoch": 0.910775747948107,
+      "grad_norm": 0.44765394926071167,
+      "learning_rate": 1.846031430276507e-06,
+      "loss": 0.0185,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9114175913222776,
+      "grad_norm": 0.05895282328128815,
+      "learning_rate": 1.832769710231417e-06,
+      "loss": 0.0239,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9120594346964482,
+      "grad_norm": 0.01649278774857521,
+      "learning_rate": 1.8195079901863272e-06,
+      "loss": 0.0216,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9127012780706188,
+      "grad_norm": 0.04506438225507736,
+      "learning_rate": 1.8062462701412375e-06,
+      "loss": 0.0172,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9133431214447895,
+      "grad_norm": 0.364442378282547,
+      "learning_rate": 1.7929845500961477e-06,
+      "loss": 0.0346,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9139849648189601,
+      "grad_norm": 0.5869020223617554,
+      "learning_rate": 1.7797228300510578e-06,
+      "loss": 0.0207,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9146268081931307,
+      "grad_norm": 0.6192695498466492,
+      "learning_rate": 1.7664611100059679e-06,
+      "loss": 0.012,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9152686515673013,
+      "grad_norm": 0.13403886556625366,
+      "learning_rate": 1.7531993899608782e-06,
+      "loss": 0.0269,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9159104949414719,
+      "grad_norm": 7.840417385101318,
+      "learning_rate": 1.7399376699157883e-06,
+      "loss": 0.0236,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9165523383156425,
+      "grad_norm": 0.1548127382993698,
+      "learning_rate": 1.7266759498706984e-06,
+      "loss": 0.0379,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9171941816898131,
+      "grad_norm": 0.05730681121349335,
+      "learning_rate": 1.7134142298256085e-06,
+      "loss": 0.0151,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9178360250639838,
+      "grad_norm": 2.952178955078125,
+      "learning_rate": 1.7001525097805185e-06,
+      "loss": 0.0124,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9184778684381544,
+      "grad_norm": 0.4452185332775116,
+      "learning_rate": 1.686890789735429e-06,
+      "loss": 0.0208,
+      "step": 14310
+    },
+    {
+      "epoch": 0.919119711812325,
+      "grad_norm": 0.05769308656454086,
+      "learning_rate": 1.673629069690339e-06,
+      "loss": 0.0225,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9197615551864956,
+      "grad_norm": 0.30611440539360046,
+      "learning_rate": 1.660367349645249e-06,
+      "loss": 0.0153,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9204033985606662,
+      "grad_norm": 4.835531234741211,
+      "learning_rate": 1.6471056296001591e-06,
+      "loss": 0.0267,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9210452419348368,
+      "grad_norm": 6.112522602081299,
+      "learning_rate": 1.6338439095550697e-06,
+      "loss": 0.0235,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9216870853090074,
+      "grad_norm": 0.17406339943408966,
+      "learning_rate": 1.6205821895099796e-06,
+      "loss": 0.0303,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9223289286831781,
+      "grad_norm": 0.049876339733600616,
+      "learning_rate": 1.6073204694648897e-06,
+      "loss": 0.0132,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9229707720573487,
+      "grad_norm": 0.455746591091156,
+      "learning_rate": 1.5940587494197998e-06,
+      "loss": 0.0104,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9236126154315193,
+      "grad_norm": 0.1712111383676529,
+      "learning_rate": 1.58079702937471e-06,
+      "loss": 0.0189,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9242544588056899,
+      "grad_norm": 1.3781015872955322,
+      "learning_rate": 1.5675353093296202e-06,
+      "loss": 0.0215,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9248963021798605,
+      "grad_norm": 0.18812592327594757,
+      "learning_rate": 1.5542735892845303e-06,
+      "loss": 0.0259,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9255381455540311,
+      "grad_norm": 0.16782377660274506,
+      "learning_rate": 1.5410118692394404e-06,
+      "loss": 0.0327,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9261799889282017,
+      "grad_norm": 0.07186288386583328,
+      "learning_rate": 1.5277501491943506e-06,
+      "loss": 0.0182,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9268218323023724,
+      "grad_norm": 0.7693184018135071,
+      "learning_rate": 1.5144884291492609e-06,
+      "loss": 0.0235,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9274636756765431,
+      "grad_norm": 0.5924553275108337,
+      "learning_rate": 1.501226709104171e-06,
+      "loss": 0.0281,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9281055190507137,
+      "grad_norm": 0.031777169555425644,
+      "learning_rate": 1.4879649890590811e-06,
+      "loss": 0.0155,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9287473624248843,
+      "grad_norm": 0.011459583416581154,
+      "learning_rate": 1.4747032690139912e-06,
+      "loss": 0.0345,
+      "step": 14470
+    },
+    {
+      "epoch": 0.9293892057990549,
+      "grad_norm": 0.35816437005996704,
+      "learning_rate": 1.4614415489689013e-06,
+      "loss": 0.0273,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9300310491732255,
+      "grad_norm": 0.29088306427001953,
+      "learning_rate": 1.4481798289238117e-06,
+      "loss": 0.0264,
+      "step": 14490
+    },
+    {
+      "epoch": 0.9306728925473962,
+      "grad_norm": 0.4196803569793701,
+      "learning_rate": 1.4349181088787218e-06,
+      "loss": 0.0053,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9313147359215668,
+      "grad_norm": 0.1188807487487793,
+      "learning_rate": 1.4216563888336319e-06,
+      "loss": 0.0238,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9319565792957374,
+      "grad_norm": 1.8770768642425537,
+      "learning_rate": 1.408394668788542e-06,
+      "loss": 0.0292,
+      "step": 14520
+    },
+    {
+      "epoch": 0.932598422669908,
+      "grad_norm": 0.8623146414756775,
+      "learning_rate": 1.395132948743452e-06,
+      "loss": 0.0105,
+      "step": 14530
+    },
+    {
+      "epoch": 0.9332402660440786,
+      "grad_norm": 0.8951078057289124,
+      "learning_rate": 1.3818712286983624e-06,
+      "loss": 0.0246,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9338821094182492,
+      "grad_norm": 0.2718936800956726,
+      "learning_rate": 1.3686095086532726e-06,
+      "loss": 0.0123,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9345239527924198,
+      "grad_norm": 0.025857457891106606,
+      "learning_rate": 1.3553477886081825e-06,
+      "loss": 0.0129,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9351657961665905,
+      "grad_norm": 0.4448324143886566,
+      "learning_rate": 1.3420860685630926e-06,
+      "loss": 0.0206,
+      "step": 14570
+    },
+    {
+      "epoch": 0.9358076395407611,
+      "grad_norm": 0.21557696163654327,
+      "learning_rate": 1.3288243485180031e-06,
+      "loss": 0.0345,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9364494829149317,
+      "grad_norm": 0.2089679092168808,
+      "learning_rate": 1.315562628472913e-06,
+      "loss": 0.0187,
+      "step": 14590
+    },
+    {
+      "epoch": 0.9370913262891023,
+      "grad_norm": 2.0699353218078613,
+      "learning_rate": 1.3023009084278231e-06,
+      "loss": 0.0217,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9377331696632729,
+      "grad_norm": 2.9068894386291504,
+      "learning_rate": 1.2890391883827332e-06,
+      "loss": 0.0247,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9383750130374435,
+      "grad_norm": 1.3320043087005615,
+      "learning_rate": 1.2757774683376434e-06,
+      "loss": 0.0106,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9390168564116141,
+      "grad_norm": 2.405679702758789,
+      "learning_rate": 1.2625157482925537e-06,
+      "loss": 0.0189,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9396586997857848,
+      "grad_norm": 0.5062136650085449,
+      "learning_rate": 1.2492540282474638e-06,
+      "loss": 0.0228,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9403005431599554,
+      "grad_norm": 1.138634204864502,
+      "learning_rate": 1.235992308202374e-06,
+      "loss": 0.0163,
+      "step": 14650
+    },
+    {
+      "epoch": 0.940942386534126,
+      "grad_norm": 0.49596962332725525,
+      "learning_rate": 1.2227305881572842e-06,
+      "loss": 0.0198,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9415842299082966,
+      "grad_norm": 0.5795552730560303,
+      "learning_rate": 1.2094688681121943e-06,
+      "loss": 0.0145,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9422260732824672,
+      "grad_norm": 1.0133116245269775,
+      "learning_rate": 1.1962071480671042e-06,
+      "loss": 0.0144,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9428679166566378,
+      "grad_norm": 0.3072698712348938,
+      "learning_rate": 1.1829454280220146e-06,
+      "loss": 0.0223,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9435097600308084,
+      "grad_norm": 3.633915901184082,
+      "learning_rate": 1.1696837079769247e-06,
+      "loss": 0.0249,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9441516034049791,
+      "grad_norm": 1.0597184896469116,
+      "learning_rate": 1.1564219879318348e-06,
+      "loss": 0.014,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9447934467791497,
+      "grad_norm": 1.0187700986862183,
+      "learning_rate": 1.143160267886745e-06,
+      "loss": 0.0198,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9454352901533203,
+      "grad_norm": 0.5023682117462158,
+      "learning_rate": 1.1298985478416552e-06,
+      "loss": 0.0311,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9460771335274909,
+      "grad_norm": 2.0689022541046143,
+      "learning_rate": 1.1166368277965654e-06,
+      "loss": 0.0197,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9467189769016616,
+      "grad_norm": 7.048704147338867,
+      "learning_rate": 1.1033751077514755e-06,
+      "loss": 0.0235,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9473608202758322,
+      "grad_norm": 0.6049743890762329,
+      "learning_rate": 1.0901133877063856e-06,
+      "loss": 0.0213,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9480026636500029,
+      "grad_norm": 1.2432732582092285,
+      "learning_rate": 1.0768516676612957e-06,
+      "loss": 0.0244,
+      "step": 14770
+    },
+    {
+      "epoch": 0.9486445070241735,
+      "grad_norm": 1.8908295631408691,
+      "learning_rate": 1.063589947616206e-06,
+      "loss": 0.0304,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9492863503983441,
+      "grad_norm": 0.16690313816070557,
+      "learning_rate": 1.050328227571116e-06,
+      "loss": 0.0116,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9499281937725147,
+      "grad_norm": 1.6157116889953613,
+      "learning_rate": 1.0370665075260262e-06,
+      "loss": 0.0251,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9505700371466853,
+      "grad_norm": 0.6109140515327454,
+      "learning_rate": 1.0238047874809364e-06,
+      "loss": 0.0253,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9512118805208559,
+      "grad_norm": 1.494762659072876,
+      "learning_rate": 1.0105430674358465e-06,
+      "loss": 0.0138,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9518537238950265,
+      "grad_norm": 0.47034692764282227,
+      "learning_rate": 9.972813473907566e-07,
+      "loss": 0.0162,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9524955672691972,
+      "grad_norm": 0.22934949398040771,
+      "learning_rate": 9.840196273456667e-07,
+      "loss": 0.024,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9531374106433678,
+      "grad_norm": 0.8337307572364807,
+      "learning_rate": 9.70757907300577e-07,
+      "loss": 0.0251,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9537792540175384,
+      "grad_norm": 2.82429575920105,
+      "learning_rate": 9.574961872554871e-07,
+      "loss": 0.0198,
+      "step": 14860
+    },
+    {
+      "epoch": 0.954421097391709,
+      "grad_norm": 0.22207744419574738,
+      "learning_rate": 9.442344672103972e-07,
+      "loss": 0.0143,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9550629407658796,
+      "grad_norm": 0.013904299587011337,
+      "learning_rate": 9.309727471653074e-07,
+      "loss": 0.0214,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9557047841400502,
+      "grad_norm": 7.341753959655762,
+      "learning_rate": 9.177110271202176e-07,
+      "loss": 0.0161,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9563466275142208,
+      "grad_norm": 0.29478883743286133,
+      "learning_rate": 9.044493070751277e-07,
+      "loss": 0.0157,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9569884708883915,
+      "grad_norm": 0.01080330554395914,
+      "learning_rate": 8.911875870300379e-07,
+      "loss": 0.0217,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9576303142625621,
+      "grad_norm": 0.024894973263144493,
+      "learning_rate": 8.77925866984948e-07,
+      "loss": 0.0208,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9582721576367327,
+      "grad_norm": 5.743207931518555,
+      "learning_rate": 8.646641469398581e-07,
+      "loss": 0.0209,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9589140010109033,
+      "grad_norm": 0.9241563677787781,
+      "learning_rate": 8.514024268947684e-07,
+      "loss": 0.0315,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9595558443850739,
+      "grad_norm": 0.06609442830085754,
+      "learning_rate": 8.381407068496785e-07,
+      "loss": 0.0153,
+      "step": 14950
+    },
+    {
+      "epoch": 0.9601976877592445,
+      "grad_norm": 0.5711029767990112,
+      "learning_rate": 8.248789868045887e-07,
+      "loss": 0.0254,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9608395311334151,
+      "grad_norm": 3.806605100631714,
+      "learning_rate": 8.116172667594988e-07,
+      "loss": 0.0291,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9614813745075858,
+      "grad_norm": 0.03836154565215111,
+      "learning_rate": 7.98355546714409e-07,
+      "loss": 0.0177,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9621232178817564,
+      "grad_norm": 1.2240588665008545,
+      "learning_rate": 7.85093826669319e-07,
+      "loss": 0.0102,
+      "step": 14990
+    },
+    {
+      "epoch": 0.962765061255927,
+      "grad_norm": 1.0644761323928833,
+      "learning_rate": 7.718321066242291e-07,
+      "loss": 0.0426,
+      "step": 15000
+    },
+    {
+      "epoch": 0.9634069046300976,
+      "grad_norm": 0.48844975233078003,
+      "learning_rate": 7.585703865791394e-07,
+      "loss": 0.0362,
+      "step": 15010
+    },
+    {
+      "epoch": 0.9640487480042682,
+      "grad_norm": 1.1872748136520386,
+      "learning_rate": 7.453086665340495e-07,
+      "loss": 0.0327,
+      "step": 15020
+    },
+    {
+      "epoch": 0.9646905913784388,
+      "grad_norm": 1.2432113885879517,
+      "learning_rate": 7.320469464889597e-07,
+      "loss": 0.0236,
+      "step": 15030
+    },
+    {
+      "epoch": 0.9653324347526095,
+      "grad_norm": 0.12339625507593155,
+      "learning_rate": 7.187852264438698e-07,
+      "loss": 0.0213,
+      "step": 15040
+    },
+    {
+      "epoch": 0.9659742781267802,
+      "grad_norm": 0.8140659332275391,
+      "learning_rate": 7.0552350639878e-07,
+      "loss": 0.0208,
+      "step": 15050
+    },
+    {
+      "epoch": 0.9666161215009508,
+      "grad_norm": 0.10417089611291885,
+      "learning_rate": 6.922617863536901e-07,
+      "loss": 0.0185,
+      "step": 15060
+    },
+    {
+      "epoch": 0.9672579648751214,
+      "grad_norm": 1.1666573286056519,
+      "learning_rate": 6.790000663086003e-07,
+      "loss": 0.016,
+      "step": 15070
+    },
+    {
+      "epoch": 0.967899808249292,
+      "grad_norm": 2.4468142986297607,
+      "learning_rate": 6.657383462635105e-07,
+      "loss": 0.0177,
+      "step": 15080
+    },
+    {
+      "epoch": 0.9685416516234626,
+      "grad_norm": 0.02574695274233818,
+      "learning_rate": 6.524766262184205e-07,
+      "loss": 0.0127,
+      "step": 15090
+    },
+    {
+      "epoch": 0.9691834949976332,
+      "grad_norm": 0.8955485224723816,
+      "learning_rate": 6.392149061733308e-07,
+      "loss": 0.0222,
+      "step": 15100
+    },
+    {
+      "epoch": 0.9698253383718038,
+      "grad_norm": 0.622428297996521,
+      "learning_rate": 6.259531861282408e-07,
+      "loss": 0.0133,
+      "step": 15110
+    },
+    {
+      "epoch": 0.9704671817459745,
+      "grad_norm": 0.1865653395652771,
+      "learning_rate": 6.12691466083151e-07,
+      "loss": 0.0222,
+      "step": 15120
+    },
+    {
+      "epoch": 0.9711090251201451,
+      "grad_norm": 0.1449609100818634,
+      "learning_rate": 5.994297460380612e-07,
+      "loss": 0.0369,
+      "step": 15130
+    },
+    {
+      "epoch": 0.9717508684943157,
+      "grad_norm": 1.6518489122390747,
+      "learning_rate": 5.861680259929714e-07,
+      "loss": 0.0306,
+      "step": 15140
+    },
+    {
+      "epoch": 0.9723927118684863,
+      "grad_norm": 0.9899628758430481,
+      "learning_rate": 5.729063059478815e-07,
+      "loss": 0.0112,
+      "step": 15150
+    },
+    {
+      "epoch": 0.9730345552426569,
+      "grad_norm": 0.517170786857605,
+      "learning_rate": 5.596445859027916e-07,
+      "loss": 0.0186,
+      "step": 15160
+    },
+    {
+      "epoch": 0.9736763986168275,
+      "grad_norm": 0.008162170648574829,
+      "learning_rate": 5.463828658577018e-07,
+      "loss": 0.0054,
+      "step": 15170
+    },
+    {
+      "epoch": 0.9743182419909981,
+      "grad_norm": 0.7698886394500732,
+      "learning_rate": 5.331211458126119e-07,
+      "loss": 0.0282,
+      "step": 15180
+    },
+    {
+      "epoch": 0.9749600853651688,
+      "grad_norm": 4.941947937011719,
+      "learning_rate": 5.19859425767522e-07,
+      "loss": 0.0326,
+      "step": 15190
+    },
+    {
+      "epoch": 0.9756019287393394,
+      "grad_norm": 0.18007877469062805,
+      "learning_rate": 5.065977057224323e-07,
+      "loss": 0.0252,
+      "step": 15200
+    },
+    {
+      "epoch": 0.97624377211351,
+      "grad_norm": 0.21914184093475342,
+      "learning_rate": 4.933359856773424e-07,
+      "loss": 0.0333,
+      "step": 15210
+    },
+    {
+      "epoch": 0.9768856154876806,
+      "grad_norm": 0.45805874466896057,
+      "learning_rate": 4.800742656322525e-07,
+      "loss": 0.0149,
+      "step": 15220
+    },
+    {
+      "epoch": 0.9775274588618512,
+      "grad_norm": 6.29951810836792,
+      "learning_rate": 4.668125455871627e-07,
+      "loss": 0.0216,
+      "step": 15230
+    },
+    {
+      "epoch": 0.9781693022360218,
+      "grad_norm": 0.764335036277771,
+      "learning_rate": 4.535508255420729e-07,
+      "loss": 0.017,
+      "step": 15240
+    },
+    {
+      "epoch": 0.9788111456101924,
+      "grad_norm": 10.674691200256348,
+      "learning_rate": 4.40289105496983e-07,
+      "loss": 0.0288,
+      "step": 15250
+    },
+    {
+      "epoch": 0.9794529889843631,
+      "grad_norm": 0.6469502449035645,
+      "learning_rate": 4.2702738545189316e-07,
+      "loss": 0.0089,
+      "step": 15260
+    },
+    {
+      "epoch": 0.9800948323585337,
+      "grad_norm": 1.4636619091033936,
+      "learning_rate": 4.1376566540680327e-07,
+      "loss": 0.014,
+      "step": 15270
+    },
+    {
+      "epoch": 0.9807366757327043,
+      "grad_norm": 1.0619542598724365,
+      "learning_rate": 4.0050394536171344e-07,
+      "loss": 0.0175,
+      "step": 15280
+    },
+    {
+      "epoch": 0.9813785191068749,
+      "grad_norm": 0.06426899135112762,
+      "learning_rate": 3.872422253166236e-07,
+      "loss": 0.0124,
+      "step": 15290
+    },
+    {
+      "epoch": 0.9820203624810455,
+      "grad_norm": 8.639874458312988,
+      "learning_rate": 3.7398050527153377e-07,
+      "loss": 0.0284,
+      "step": 15300
+    },
+    {
+      "epoch": 0.9826622058552161,
+      "grad_norm": 0.9990230798721313,
+      "learning_rate": 3.607187852264439e-07,
+      "loss": 0.0224,
+      "step": 15310
+    },
+    {
+      "epoch": 0.9833040492293867,
+      "grad_norm": 3.5814836025238037,
+      "learning_rate": 3.4745706518135405e-07,
+      "loss": 0.0232,
+      "step": 15320
+    },
+    {
+      "epoch": 0.9839458926035575,
+      "grad_norm": 0.4112697243690491,
+      "learning_rate": 3.341953451362642e-07,
+      "loss": 0.0243,
+      "step": 15330
+    },
+    {
+      "epoch": 0.9845877359777281,
+      "grad_norm": 1.136162281036377,
+      "learning_rate": 3.209336250911744e-07,
+      "loss": 0.0056,
+      "step": 15340
+    },
+    {
+      "epoch": 0.9852295793518987,
+      "grad_norm": 0.06789114326238632,
+      "learning_rate": 3.076719050460845e-07,
+      "loss": 0.007,
+      "step": 15350
+    },
+    {
+      "epoch": 0.9858714227260693,
+      "grad_norm": 0.027609622105956078,
+      "learning_rate": 2.944101850009946e-07,
+      "loss": 0.0214,
+      "step": 15360
+    },
+    {
+      "epoch": 0.9865132661002399,
+      "grad_norm": 0.472529798746109,
+      "learning_rate": 2.811484649559048e-07,
+      "loss": 0.0295,
+      "step": 15370
+    },
+    {
+      "epoch": 0.9871551094744105,
+      "grad_norm": 1.0125441551208496,
+      "learning_rate": 2.6788674491081494e-07,
+      "loss": 0.0255,
+      "step": 15380
+    },
+    {
+      "epoch": 0.9877969528485812,
+      "grad_norm": 0.11004765331745148,
+      "learning_rate": 2.546250248657251e-07,
+      "loss": 0.021,
+      "step": 15390
+    },
+    {
+      "epoch": 0.9884387962227518,
+      "grad_norm": 2.7606465816497803,
+      "learning_rate": 2.413633048206352e-07,
+      "loss": 0.0242,
+      "step": 15400
+    },
+    {
+      "epoch": 0.9890806395969224,
+      "grad_norm": 0.45687106251716614,
+      "learning_rate": 2.281015847755454e-07,
+      "loss": 0.027,
+      "step": 15410
+    },
+    {
+      "epoch": 0.989722482971093,
+      "grad_norm": 1.2439097166061401,
+      "learning_rate": 2.1483986473045556e-07,
+      "loss": 0.0398,
+      "step": 15420
+    },
+    {
+      "epoch": 0.9903643263452636,
+      "grad_norm": 0.4050452709197998,
+      "learning_rate": 2.0157814468536572e-07,
+      "loss": 0.0135,
+      "step": 15430
+    },
+    {
+      "epoch": 0.9910061697194342,
+      "grad_norm": 4.158045768737793,
+      "learning_rate": 1.8831642464027584e-07,
+      "loss": 0.0178,
+      "step": 15440
+    },
+    {
+      "epoch": 0.9916480130936048,
+      "grad_norm": 0.40229663252830505,
+      "learning_rate": 1.75054704595186e-07,
+      "loss": 0.0175,
+      "step": 15450
+    },
+    {
+      "epoch": 0.9922898564677755,
+      "grad_norm": 0.27971166372299194,
+      "learning_rate": 1.6179298455009617e-07,
+      "loss": 0.0279,
+      "step": 15460
+    },
+    {
+      "epoch": 0.9929316998419461,
+      "grad_norm": 0.02092203125357628,
+      "learning_rate": 1.485312645050063e-07,
+      "loss": 0.016,
+      "step": 15470
+    },
+    {
+      "epoch": 0.9935735432161167,
+      "grad_norm": 0.5914480090141296,
+      "learning_rate": 1.3526954445991648e-07,
+      "loss": 0.0353,
+      "step": 15480
+    },
+    {
+      "epoch": 0.9942153865902873,
+      "grad_norm": 0.27921462059020996,
+      "learning_rate": 1.2200782441482662e-07,
+      "loss": 0.0139,
+      "step": 15490
+    },
+    {
+      "epoch": 0.9948572299644579,
+      "grad_norm": 0.5014231204986572,
+      "learning_rate": 1.0874610436973677e-07,
+      "loss": 0.0165,
+      "step": 15500
+    },
+    {
+      "epoch": 0.9954990733386285,
+      "grad_norm": 0.6527778506278992,
+      "learning_rate": 9.548438432464691e-08,
+      "loss": 0.0321,
+      "step": 15510
+    },
+    {
+      "epoch": 0.9961409167127991,
+      "grad_norm": 0.018357345834374428,
+      "learning_rate": 8.222266427955706e-08,
+      "loss": 0.0248,
+      "step": 15520
+    },
+    {
+      "epoch": 0.9967827600869698,
+      "grad_norm": 0.30090826749801636,
+      "learning_rate": 6.896094423446722e-08,
+      "loss": 0.0076,
+      "step": 15530
+    },
+    {
+      "epoch": 0.9974246034611404,
+      "grad_norm": 1.3387514352798462,
+      "learning_rate": 5.569922418937737e-08,
+      "loss": 0.0229,
+      "step": 15540
+    },
+    {
+      "epoch": 0.998066446835311,
+      "grad_norm": 2.0242581367492676,
+      "learning_rate": 4.2437504144287516e-08,
+      "loss": 0.0123,
+      "step": 15550
+    },
+    {
+      "epoch": 0.9987082902094816,
+      "grad_norm": 1.3312128782272339,
+      "learning_rate": 2.917578409919767e-08,
+      "loss": 0.0135,
+      "step": 15560
+    },
+    {
+      "epoch": 0.9993501335836522,
+      "grad_norm": 0.10261988639831543,
+      "learning_rate": 1.591406405410782e-08,
+      "loss": 0.0214,
+      "step": 15570
+    },
+    {
+      "epoch": 0.9999919769578228,
+      "grad_norm": 0.041646864265203476,
+      "learning_rate": 2.6523440090179698e-09,
+      "loss": 0.0197,
+      "step": 15580
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 15581,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}